diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,336034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.1425307055127107,
+  "eval_steps": 500,
+  "global_step": 48000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 2.3802723031514804e-05,
+      "grad_norm": 60.16500032446618,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 1
+    },
+    {
+      "epoch": 4.760544606302961e-05,
+      "grad_norm": 13.652008208308827,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 2
+    },
+    {
+      "epoch": 7.140816909454442e-05,
+      "grad_norm": 77.28033331477609,
+      "learning_rate": 2e-05,
+      "loss": 0.8119,
+      "step": 3
+    },
+    {
+      "epoch": 9.521089212605922e-05,
+      "grad_norm": 20.283881476326496,
+      "learning_rate": 2e-05,
+      "loss": 0.5364,
+      "step": 4
+    },
+    {
+      "epoch": 0.00011901361515757403,
+      "grad_norm": 17.26163061394796,
+      "learning_rate": 2e-05,
+      "loss": 0.3477,
+      "step": 5
+    },
+    {
+      "epoch": 0.00014281633818908883,
+      "grad_norm": 42.537768395505196,
+      "learning_rate": 2e-05,
+      "loss": 1.5683,
+      "step": 6
+    },
+    {
+      "epoch": 0.00016661906122060363,
+      "grad_norm": 24.84552603889919,
+      "learning_rate": 2e-05,
+      "loss": 1.534,
+      "step": 7
+    },
+    {
+      "epoch": 0.00019042178425211843,
+      "grad_norm": 15.518419151580815,
+      "learning_rate": 2e-05,
+      "loss": 1.0744,
+      "step": 8
+    },
+    {
+      "epoch": 0.00021422450728363326,
+      "grad_norm": 36.65202843144203,
+      "learning_rate": 2e-05,
+      "loss": 1.1073,
+      "step": 9
+    },
+    {
+      "epoch": 0.00023802723031514806,
+      "grad_norm": 4.922120560793037,
+      "learning_rate": 2e-05,
+      "loss": 0.1793,
+      "step": 10
+    },
+    {
+      "epoch": 0.00026182995334666283,
+      "grad_norm": 9.771279328765855,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 11
+    },
+    {
+      "epoch": 0.00028563267637817766,
+      "grad_norm": 18.233798881139194,
+      "learning_rate": 2e-05,
+      "loss": 2.0026,
+      "step": 12
+    },
+    {
+      "epoch": 0.0003094353994096925,
+      "grad_norm": 19.61003039381246,
+      "learning_rate": 2e-05,
+      "loss": 0.6025,
+      "step": 13
+    },
+    {
+      "epoch": 0.00033323812244120726,
+      "grad_norm": 30.12173919658445,
+      "learning_rate": 2e-05,
+      "loss": 0.7601,
+      "step": 14
+    },
+    {
+      "epoch": 0.0003570408454727221,
+      "grad_norm": 14.874682249879902,
+      "learning_rate": 2e-05,
+      "loss": 0.6313,
+      "step": 15
+    },
+    {
+      "epoch": 0.00038084356850423686,
+      "grad_norm": 17.961860108275516,
+      "learning_rate": 2e-05,
+      "loss": 1.4473,
+      "step": 16
+    },
+    {
+      "epoch": 0.0004046462915357517,
+      "grad_norm": 11.723530640861886,
+      "learning_rate": 2e-05,
+      "loss": 0.276,
+      "step": 17
+    },
+    {
+      "epoch": 0.0004284490145672665,
+      "grad_norm": 16.588950763343185,
+      "learning_rate": 2e-05,
+      "loss": 0.8837,
+      "step": 18
+    },
+    {
+      "epoch": 0.0004522517375987813,
+      "grad_norm": 13.376322790910544,
+      "learning_rate": 2e-05,
+      "loss": 0.5498,
+      "step": 19
+    },
+    {
+      "epoch": 0.0004760544606302961,
+      "grad_norm": 31.161246677847238,
+      "learning_rate": 2e-05,
+      "loss": 0.884,
+      "step": 20
+    },
+    {
+      "epoch": 0.000499857183661811,
+      "grad_norm": 12.930308178403635,
+      "learning_rate": 2e-05,
+      "loss": 0.3179,
+      "step": 21
+    },
+    {
+      "epoch": 0.0005236599066933257,
+      "grad_norm": 4.244741426191347,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 22
+    },
+    {
+      "epoch": 0.0005474626297248405,
+      "grad_norm": 8.199061623171003,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 23
+    },
+    {
+      "epoch": 0.0005712653527563553,
+      "grad_norm": 7.38485218313211,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 24
+    },
+    {
+      "epoch": 0.0005950680757878701,
+      "grad_norm": 15.423799544524497,
+      "learning_rate": 2e-05,
+      "loss": 1.2118,
+      "step": 25
+    },
+    {
+      "epoch": 0.000618870798819385,
+      "grad_norm": 8.955045077298927,
+      "learning_rate": 2e-05,
+      "loss": 0.3651,
+      "step": 26
+    },
+    {
+      "epoch": 0.0006426735218508997,
+      "grad_norm": 18.719645597120923,
+      "learning_rate": 2e-05,
+      "loss": 1.0808,
+      "step": 27
+    },
+    {
+      "epoch": 0.0006664762448824145,
+      "grad_norm": 8.996843098267405,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 28
+    },
+    {
+      "epoch": 0.0006902789679139294,
+      "grad_norm": 11.849976612937507,
+      "learning_rate": 2e-05,
+      "loss": 0.3928,
+      "step": 29
+    },
+    {
+      "epoch": 0.0007140816909454442,
+      "grad_norm": 94.64852004651448,
+      "learning_rate": 2e-05,
+      "loss": 0.4838,
+      "step": 30
+    },
+    {
+      "epoch": 0.000737884413976959,
+      "grad_norm": 14.52790767716268,
+      "learning_rate": 2e-05,
+      "loss": 0.4263,
+      "step": 31
+    },
+    {
+      "epoch": 0.0007616871370084737,
+      "grad_norm": 17.621278052537694,
+      "learning_rate": 2e-05,
+      "loss": 0.8929,
+      "step": 32
+    },
+    {
+      "epoch": 0.0007854898600399886,
+      "grad_norm": 4.015541777701236,
+      "learning_rate": 2e-05,
+      "loss": 0.1965,
+      "step": 33
+    },
+    {
+      "epoch": 0.0008092925830715034,
+      "grad_norm": 2.564089313467316,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 34
+    },
+    {
+      "epoch": 0.0008330953061030182,
+      "grad_norm": 11.729650110287336,
+      "learning_rate": 2e-05,
+      "loss": 1.168,
+      "step": 35
+    },
+    {
+      "epoch": 0.000856898029134533,
+      "grad_norm": 9.728661392015049,
+      "learning_rate": 2e-05,
+      "loss": 0.3376,
+      "step": 36
+    },
+    {
+      "epoch": 0.0008807007521660478,
+      "grad_norm": 5.18620264263325,
+      "learning_rate": 2e-05,
+      "loss": 0.1834,
+      "step": 37
+    },
+    {
+      "epoch": 0.0009045034751975626,
+      "grad_norm": 19.96876213881063,
+      "learning_rate": 2e-05,
+      "loss": 1.5124,
+      "step": 38
+    },
+    {
+      "epoch": 0.0009283061982290774,
+      "grad_norm": 5.514678857528181,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 39
+    },
+    {
+      "epoch": 0.0009521089212605922,
+      "grad_norm": 15.356024486307243,
+      "learning_rate": 2e-05,
+      "loss": 0.4243,
+      "step": 40
+    },
+    {
+      "epoch": 0.0009759116442921071,
+      "grad_norm": 15.747436474167419,
+      "learning_rate": 2e-05,
+      "loss": 0.8178,
+      "step": 41
+    },
+    {
+      "epoch": 0.000999714367323622,
+      "grad_norm": 10.177739597286056,
+      "learning_rate": 2e-05,
+      "loss": 0.5066,
+      "step": 42
+    },
+    {
+      "epoch": 0.0010235170903551366,
+      "grad_norm": 5.809525799799312,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 43
+    },
+    {
+      "epoch": 0.0010473198133866513,
+      "grad_norm": 18.413546860401418,
+      "learning_rate": 2e-05,
+      "loss": 0.5615,
+      "step": 44
+    },
+    {
+      "epoch": 0.0010711225364181663,
+      "grad_norm": 32.60602284260672,
+      "learning_rate": 2e-05,
+      "loss": 0.9214,
+      "step": 45
+    },
+    {
+      "epoch": 0.001094925259449681,
+      "grad_norm": 30.08748974440602,
+      "learning_rate": 2e-05,
+      "loss": 0.9897,
+      "step": 46
+    },
+    {
+      "epoch": 0.001118727982481196,
+      "grad_norm": 67.61835404888267,
+      "learning_rate": 2e-05,
+      "loss": 0.8209,
+      "step": 47
+    },
+    {
+      "epoch": 0.0011425307055127106,
+      "grad_norm": 12.327709299928324,
+      "learning_rate": 2e-05,
+      "loss": 0.7357,
+      "step": 48
+    },
+    {
+      "epoch": 0.0011663334285442254,
+      "grad_norm": 12.733943092963978,
+      "learning_rate": 2e-05,
+      "loss": 0.7761,
+      "step": 49
+    },
+    {
+      "epoch": 0.0011901361515757403,
+      "grad_norm": 11.513049657368066,
+      "learning_rate": 2e-05,
+      "loss": 0.9264,
+      "step": 50
+    },
+    {
+      "epoch": 0.001213938874607255,
+      "grad_norm": 7.328595332843666,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 51
+    },
+    {
+      "epoch": 0.00123774159763877,
+      "grad_norm": 8.912770452536785,
+      "learning_rate": 2e-05,
+      "loss": 0.3152,
+      "step": 52
+    },
+    {
+      "epoch": 0.0012615443206702847,
+      "grad_norm": 8.40762819233605,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 53
+    },
+    {
+      "epoch": 0.0012853470437017994,
+      "grad_norm": 11.684829181116365,
+      "learning_rate": 2e-05,
+      "loss": 0.669,
+      "step": 54
+    },
+    {
+      "epoch": 0.0013091497667333143,
+      "grad_norm": 4.6324174082563605,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 55
+    },
+    {
+      "epoch": 0.001332952489764829,
+      "grad_norm": 3.4365505887944274,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 56
+    },
+    {
+      "epoch": 0.001356755212796344,
+      "grad_norm": 11.830641752943142,
+      "learning_rate": 2e-05,
+      "loss": 0.682,
+      "step": 57
+    },
+    {
+      "epoch": 0.0013805579358278587,
+      "grad_norm": 15.260118534298167,
+      "learning_rate": 2e-05,
+      "loss": 0.6829,
+      "step": 58
+    },
+    {
+      "epoch": 0.0014043606588593734,
+      "grad_norm": 11.034082918371379,
+      "learning_rate": 2e-05,
+      "loss": 0.766,
+      "step": 59
+    },
+    {
+      "epoch": 0.0014281633818908884,
+      "grad_norm": 6.090770436317622,
+      "learning_rate": 2e-05,
+      "loss": 0.279,
+      "step": 60
+    },
+    {
+      "epoch": 0.001451966104922403,
+      "grad_norm": 92.99108435487098,
+      "learning_rate": 2e-05,
+      "loss": 0.9664,
+      "step": 61
+    },
+    {
+      "epoch": 0.001475768827953918,
+      "grad_norm": 15.52450177924201,
+      "learning_rate": 2e-05,
+      "loss": 1.0294,
+      "step": 62
+    },
+    {
+      "epoch": 0.0014995715509854327,
+      "grad_norm": 6.080813692155232,
+      "learning_rate": 2e-05,
+      "loss": 0.2265,
+      "step": 63
+    },
+    {
+      "epoch": 0.0015233742740169475,
+      "grad_norm": 2.8063027060171493,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 64
+    },
+    {
+      "epoch": 0.0015471769970484624,
+      "grad_norm": 14.142518043468375,
+      "learning_rate": 2e-05,
+      "loss": 1.4059,
+      "step": 65
+    },
+    {
+      "epoch": 0.001570979720079977,
+      "grad_norm": 10.148023013127478,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 66
+    },
+    {
+      "epoch": 0.001594782443111492,
+      "grad_norm": 3.9309660321996347,
+      "learning_rate": 2e-05,
+      "loss": 0.2157,
+      "step": 67
+    },
+    {
+      "epoch": 0.0016185851661430068,
+      "grad_norm": 7.356857255464654,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 68
+    },
+    {
+      "epoch": 0.0016423878891745215,
+      "grad_norm": 6.181228934705726,
+      "learning_rate": 2e-05,
+      "loss": 0.3135,
+      "step": 69
+    },
+    {
+      "epoch": 0.0016661906122060364,
+      "grad_norm": 3.5718504616226494,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 70
+    },
+    {
+      "epoch": 0.0016899933352375511,
+      "grad_norm": 9.981672566786424,
+      "learning_rate": 2e-05,
+      "loss": 0.6376,
+      "step": 71
+    },
+    {
+      "epoch": 0.001713796058269066,
+      "grad_norm": 4.638496208895382,
+      "learning_rate": 2e-05,
+      "loss": 0.2268,
+      "step": 72
+    },
+    {
+      "epoch": 0.0017375987813005808,
+      "grad_norm": 7.062187648712185,
+      "learning_rate": 2e-05,
+      "loss": 0.3047,
+      "step": 73
+    },
+    {
+      "epoch": 0.0017614015043320955,
+      "grad_norm": 10.058949283144981,
+      "learning_rate": 2e-05,
+      "loss": 0.7671,
+      "step": 74
+    },
+    {
+      "epoch": 0.0017852042273636104,
+      "grad_norm": 4.8973643048726005,
+      "learning_rate": 2e-05,
+      "loss": 0.3209,
+      "step": 75
+    },
+    {
+      "epoch": 0.0018090069503951252,
+      "grad_norm": 8.467117111769626,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 76
+    },
+    {
+      "epoch": 0.00183280967342664,
+      "grad_norm": 9.768226996813137,
+      "learning_rate": 2e-05,
+      "loss": 0.7537,
+      "step": 77
+    },
+    {
+      "epoch": 0.0018566123964581548,
+      "grad_norm": 20.200862651304973,
+      "learning_rate": 2e-05,
+      "loss": 1.4978,
+      "step": 78
+    },
+    {
+      "epoch": 0.0018804151194896695,
+      "grad_norm": 6.856869018676403,
+      "learning_rate": 2e-05,
+      "loss": 0.3476,
+      "step": 79
+    },
+    {
+      "epoch": 0.0019042178425211845,
+      "grad_norm": 8.570992377646698,
+      "learning_rate": 2e-05,
+      "loss": 0.7819,
+      "step": 80
+    },
+    {
+      "epoch": 0.0019280205655526992,
+      "grad_norm": 11.034534080763954,
+      "learning_rate": 2e-05,
+      "loss": 1.1143,
+      "step": 81
+    },
+    {
+      "epoch": 0.0019518232885842141,
+      "grad_norm": 3.6610365700964205,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 82
+    },
+    {
+      "epoch": 0.0019756260116157286,
+      "grad_norm": 4.707703830289899,
+      "learning_rate": 2e-05,
+      "loss": 0.24,
+      "step": 83
+    },
+    {
+      "epoch": 0.001999428734647244,
+      "grad_norm": 8.104524579339516,
+      "learning_rate": 2e-05,
+      "loss": 0.4059,
+      "step": 84
+    },
+    {
+      "epoch": 0.0020232314576787585,
+      "grad_norm": 7.591354571610886,
+      "learning_rate": 2e-05,
+      "loss": 0.3374,
+      "step": 85
+    },
+    {
+      "epoch": 0.0020470341807102732,
+      "grad_norm": 6.794963268124376,
+      "learning_rate": 2e-05,
+      "loss": 0.3532,
+      "step": 86
+    },
+    {
+      "epoch": 0.002070836903741788,
+      "grad_norm": 7.0087100715194675,
+      "learning_rate": 2e-05,
+      "loss": 0.3799,
+      "step": 87
+    },
+    {
+      "epoch": 0.0020946396267733027,
+      "grad_norm": 6.745806264943257,
+      "learning_rate": 2e-05,
+      "loss": 0.4105,
+      "step": 88
+    },
+    {
+      "epoch": 0.002118442349804818,
+      "grad_norm": 10.82889290441752,
+      "learning_rate": 2e-05,
+      "loss": 0.6613,
+      "step": 89
+    },
+    {
+      "epoch": 0.0021422450728363325,
+      "grad_norm": 3.5462092257584024,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 90
+    },
+    {
+      "epoch": 0.0021660477958678473,
+      "grad_norm": 9.413707136427908,
+      "learning_rate": 2e-05,
+      "loss": 0.2825,
+      "step": 91
+    },
+    {
+      "epoch": 0.002189850518899362,
+      "grad_norm": 7.003809331352576,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 92
+    },
+    {
+      "epoch": 0.0022136532419308767,
+      "grad_norm": 3.9911202710948404,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 93
+    },
+    {
+      "epoch": 0.002237455964962392,
+      "grad_norm": 8.650206629019943,
+      "learning_rate": 2e-05,
+      "loss": 0.6296,
+      "step": 94
+    },
+    {
+      "epoch": 0.0022612586879939066,
+      "grad_norm": 7.683343020005095,
+      "learning_rate": 2e-05,
+      "loss": 0.2803,
+      "step": 95
+    },
+    {
+      "epoch": 0.0022850614110254213,
+      "grad_norm": 7.314281108780528,
+      "learning_rate": 2e-05,
+      "loss": 0.7889,
+      "step": 96
+    },
+    {
+      "epoch": 0.002308864134056936,
+      "grad_norm": 10.731952971535012,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 97
+    },
+    {
+      "epoch": 0.0023326668570884507,
+      "grad_norm": 6.932183859947297,
+      "learning_rate": 2e-05,
+      "loss": 0.3268,
+      "step": 98
+    },
+    {
+      "epoch": 0.002356469580119966,
+      "grad_norm": 10.319433633104243,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 99
+    },
+    {
+      "epoch": 0.0023802723031514806,
+      "grad_norm": 3.891591250817752,
+      "learning_rate": 2e-05,
+      "loss": 0.3027,
+      "step": 100
+    },
+    {
+      "epoch": 0.0024040750261829953,
+      "grad_norm": 15.7975362998252,
+      "learning_rate": 2e-05,
+      "loss": 0.682,
+      "step": 101
+    },
+    {
+      "epoch": 0.00242787774921451,
+      "grad_norm": 10.027139765093784,
+      "learning_rate": 2e-05,
+      "loss": 0.7866,
+      "step": 102
+    },
+    {
+      "epoch": 0.0024516804722460248,
+      "grad_norm": 5.00311942201846,
+      "learning_rate": 2e-05,
+      "loss": 0.2119,
+      "step": 103
+    },
+    {
+      "epoch": 0.00247548319527754,
+      "grad_norm": 14.54958885009365,
+      "learning_rate": 2e-05,
+      "loss": 1.5939,
+      "step": 104
+    },
+    {
+      "epoch": 0.0024992859183090546,
+      "grad_norm": 5.946510868133136,
+      "learning_rate": 2e-05,
+      "loss": 0.3578,
+      "step": 105
+    },
+    {
+      "epoch": 0.0025230886413405693,
+      "grad_norm": 11.559251173804777,
+      "learning_rate": 2e-05,
+      "loss": 1.6801,
+      "step": 106
+    },
+    {
+      "epoch": 0.002546891364372084,
+      "grad_norm": 11.503925226575927,
+      "learning_rate": 2e-05,
+      "loss": 0.4902,
+      "step": 107
+    },
+    {
+      "epoch": 0.002570694087403599,
+      "grad_norm": 4.022527204898744,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 108
+    },
+    {
+      "epoch": 0.002594496810435114,
+      "grad_norm": 4.588453619690811,
+      "learning_rate": 2e-05,
+      "loss": 0.1828,
+      "step": 109
+    },
+    {
+      "epoch": 0.0026182995334666287,
+      "grad_norm": 11.708032112889942,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 110
+    },
+    {
+      "epoch": 0.0026421022564981434,
+      "grad_norm": 5.180418486807895,
+      "learning_rate": 2e-05,
+      "loss": 0.2047,
+      "step": 111
+    },
+    {
+      "epoch": 0.002665904979529658,
+      "grad_norm": 3.3438517143751203,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 112
+    },
+    {
+      "epoch": 0.002689707702561173,
+      "grad_norm": 2.6150211307958466,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 113
+    },
+    {
+      "epoch": 0.002713510425592688,
+      "grad_norm": 2.388177449858711,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 114
+    },
+    {
+      "epoch": 0.0027373131486242027,
+      "grad_norm": 9.41732622311283,
+      "learning_rate": 2e-05,
+      "loss": 0.2306,
+      "step": 115
+    },
+    {
+      "epoch": 0.0027611158716557174,
+      "grad_norm": 11.467793353049162,
+      "learning_rate": 2e-05,
+      "loss": 0.8661,
+      "step": 116
+    },
+    {
+      "epoch": 0.002784918594687232,
+      "grad_norm": 8.457118023077411,
+      "learning_rate": 2e-05,
+      "loss": 1.1077,
+      "step": 117
+    },
+    {
+      "epoch": 0.002808721317718747,
+      "grad_norm": 5.2001699930211815,
+      "learning_rate": 2e-05,
+      "loss": 0.2902,
+      "step": 118
+    },
+    {
+      "epoch": 0.002832524040750262,
+      "grad_norm": 7.022379231757565,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 119
+    },
+    {
+      "epoch": 0.0028563267637817767,
+      "grad_norm": 14.684866840237714,
+      "learning_rate": 2e-05,
+      "loss": 0.5168,
+      "step": 120
+    },
+    {
+      "epoch": 0.0028801294868132914,
+      "grad_norm": 3.16540852760948,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 121
+    },
+    {
+      "epoch": 0.002903932209844806,
+      "grad_norm": 3.1346409809609415,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 122
+    },
+    {
+      "epoch": 0.002927734932876321,
+      "grad_norm": 5.8690723674204595,
+      "learning_rate": 2e-05,
+      "loss": 0.3084,
+      "step": 123
+    },
+    {
+      "epoch": 0.002951537655907836,
+      "grad_norm": 13.281241965903744,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 124
+    },
+    {
+      "epoch": 0.0029753403789393507,
+      "grad_norm": 2.757342452176235,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 125
+    },
+    {
+      "epoch": 0.0029991431019708655,
+      "grad_norm": 8.762814296956549,
+      "learning_rate": 2e-05,
+      "loss": 1.1281,
+      "step": 126
+    },
+    {
+      "epoch": 0.00302294582500238,
+      "grad_norm": 14.922012460586874,
+      "learning_rate": 2e-05,
+      "loss": 0.441,
+      "step": 127
+    },
+    {
+      "epoch": 0.003046748548033895,
+      "grad_norm": 3.2098898659268045,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 128
+    },
+    {
+      "epoch": 0.00307055127106541,
+      "grad_norm": 19.470651940198252,
+      "learning_rate": 2e-05,
+      "loss": 0.5489,
+      "step": 129
+    },
+    {
+      "epoch": 0.0030943539940969248,
+      "grad_norm": 2.8887276742397803,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 130
+    },
+    {
+      "epoch": 0.0031181567171284395,
+      "grad_norm": 4.801674735566091,
+      "learning_rate": 2e-05,
+      "loss": 0.2683,
+      "step": 131
+    },
+    {
+      "epoch": 0.003141959440159954,
+      "grad_norm": 11.927381085842336,
+      "learning_rate": 2e-05,
+      "loss": 0.8386,
+      "step": 132
+    },
+    {
+      "epoch": 0.003165762163191469,
+      "grad_norm": 14.423055781401455,
+      "learning_rate": 2e-05,
+      "loss": 1.8329,
+      "step": 133
+    },
+    {
+      "epoch": 0.003189564886222984,
+      "grad_norm": 8.759459603835404,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 134
+    },
+    {
+      "epoch": 0.003213367609254499,
+      "grad_norm": 8.833148999405712,
+      "learning_rate": 2e-05,
+      "loss": 0.371,
+      "step": 135
+    },
+    {
+      "epoch": 0.0032371703322860135,
+      "grad_norm": 2.519404681190766,
+      "learning_rate": 2e-05,
+      "loss": 0.1522,
+      "step": 136
+    },
+    {
+      "epoch": 0.0032609730553175282,
+      "grad_norm": 5.254208413650274,
+      "learning_rate": 2e-05,
+      "loss": 0.5516,
+      "step": 137
+    },
+    {
+      "epoch": 0.003284775778349043,
+      "grad_norm": 11.754669911354283,
+      "learning_rate": 2e-05,
+      "loss": 0.6791,
+      "step": 138
+    },
+    {
+      "epoch": 0.003308578501380558,
+      "grad_norm": 4.2160536547347744,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 139
+    },
+    {
+      "epoch": 0.003332381224412073,
+      "grad_norm": 10.177405461188538,
+      "learning_rate": 2e-05,
+      "loss": 1.9135,
+      "step": 140
+    },
+    {
+      "epoch": 0.0033561839474435876,
+      "grad_norm": 4.570505283797129,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 141
+    },
+    {
+      "epoch": 0.0033799866704751023,
+      "grad_norm": 9.119502858946824,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 142
+    },
+    {
+      "epoch": 0.003403789393506617,
+      "grad_norm": 3.6018576498822465,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 143
+    },
+    {
+      "epoch": 0.003427592116538132,
+      "grad_norm": 7.74138478376361,
+      "learning_rate": 2e-05,
+      "loss": 0.3272,
+      "step": 144
+    },
+    {
+      "epoch": 0.003451394839569647,
+      "grad_norm": 9.331258540349204,
+      "learning_rate": 2e-05,
+      "loss": 0.9502,
+      "step": 145
+    },
+    {
+      "epoch": 0.0034751975626011616,
+      "grad_norm": 7.664797276449726,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 146
+    },
+    {
+      "epoch": 0.0034990002856326763,
+      "grad_norm": 3.747220491707778,
+      "learning_rate": 2e-05,
+      "loss": 0.2858,
+      "step": 147
+    },
+    {
+      "epoch": 0.003522803008664191,
+      "grad_norm": 10.34155147811758,
+      "learning_rate": 2e-05,
+      "loss": 0.3394,
+      "step": 148
+    },
+    {
+      "epoch": 0.003546605731695706,
+      "grad_norm": 9.599004007846093,
+      "learning_rate": 2e-05,
+      "loss": 0.5745,
+      "step": 149
+    },
+    {
+      "epoch": 0.003570408454727221,
+      "grad_norm": 13.541902143569825,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 150
+    },
+    {
+      "epoch": 0.0035942111777587356,
+      "grad_norm": 9.921230834371029,
+      "learning_rate": 2e-05,
+      "loss": 0.3397,
+      "step": 151
+    },
+    {
+      "epoch": 0.0036180139007902503,
+      "grad_norm": 7.3139871725967,
+      "learning_rate": 2e-05,
+      "loss": 0.8476,
+      "step": 152
+    },
+    {
+      "epoch": 0.003641816623821765,
+      "grad_norm": 4.3300527212549165,
+      "learning_rate": 2e-05,
+      "loss": 0.3756,
+      "step": 153
+    },
+    {
+      "epoch": 0.00366561934685328,
+      "grad_norm": 8.250330486687547,
+      "learning_rate": 2e-05,
+      "loss": 0.305,
+      "step": 154
+    },
+    {
+      "epoch": 0.003689422069884795,
+      "grad_norm": 11.89787198555851,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 155
+    },
+    {
+      "epoch": 0.0037132247929163096,
+      "grad_norm": 7.4973611111781775,
+      "learning_rate": 2e-05,
+      "loss": 0.8376,
+      "step": 156
+    },
+    {
+      "epoch": 0.0037370275159478244,
+      "grad_norm": 11.546341230548803,
+      "learning_rate": 2e-05,
+      "loss": 0.5708,
+      "step": 157
+    },
+    {
+      "epoch": 0.003760830238979339,
+      "grad_norm": 7.494851413408379,
+      "learning_rate": 2e-05,
+      "loss": 0.7297,
+      "step": 158
+    },
+    {
+      "epoch": 0.0037846329620108542,
+      "grad_norm": 10.685710129595707,
+      "learning_rate": 2e-05,
+      "loss": 0.948,
+      "step": 159
+    },
+    {
+      "epoch": 0.003808435685042369,
+      "grad_norm": 10.993294758396164,
+      "learning_rate": 2e-05,
+      "loss": 1.0031,
+      "step": 160
+    },
+    {
+      "epoch": 0.0038322384080738837,
+      "grad_norm": 6.47682355563804,
+      "learning_rate": 2e-05,
+      "loss": 0.3146,
+      "step": 161
+    },
+    {
+      "epoch": 0.0038560411311053984,
+      "grad_norm": 5.2590422865493345,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 162
+    },
+    {
+      "epoch": 0.003879843854136913,
+      "grad_norm": 5.339198596977536,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 163
+    },
+    {
+      "epoch": 0.0039036465771684283,
+      "grad_norm": 5.947873290070159,
+      "learning_rate": 2e-05,
+      "loss": 0.2194,
+      "step": 164
+    },
+    {
+      "epoch": 0.0039274493001999426,
+      "grad_norm": 7.323022506998552,
+      "learning_rate": 2e-05,
+      "loss": 0.6916,
+      "step": 165
+    },
+    {
+      "epoch": 0.003951252023231457,
+      "grad_norm": 13.904571780232281,
+      "learning_rate": 2e-05,
+      "loss": 1.4633,
+      "step": 166
+    },
+    {
+      "epoch": 0.003975054746262973,
+      "grad_norm": 3.8419432835765552,
+      "learning_rate": 2e-05,
+      "loss": 0.2027,
+      "step": 167
+    },
+    {
+      "epoch": 0.003998857469294488,
+      "grad_norm": 4.6299090929745566,
+      "learning_rate": 2e-05,
+      "loss": 0.2918,
+      "step": 168
+    },
+    {
+      "epoch": 0.004022660192326002,
+      "grad_norm": 8.5556694954986,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 169
+    },
+    {
+      "epoch": 0.004046462915357517,
+      "grad_norm": 4.86378832718045,
+      "learning_rate": 2e-05,
+      "loss": 0.1943,
+      "step": 170
+    },
+    {
+      "epoch": 0.004070265638389032,
+      "grad_norm": 10.632647224360309,
+      "learning_rate": 2e-05,
+      "loss": 0.9971,
+      "step": 171
+    },
+    {
+      "epoch": 0.0040940683614205465,
+      "grad_norm": 10.193166153111244,
+      "learning_rate": 2e-05,
+      "loss": 0.8385,
+      "step": 172
+    },
+    {
+      "epoch": 0.004117871084452061,
+      "grad_norm": 6.773982430957191,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 173
+    },
+    {
+      "epoch": 0.004141673807483576,
+      "grad_norm": 4.540354342279047,
+      "learning_rate": 2e-05,
+      "loss": 0.3418,
+      "step": 174
+    },
+    {
+      "epoch": 0.004165476530515091,
+      "grad_norm": 9.453528261009843,
+      "learning_rate": 2e-05,
+      "loss": 0.3908,
+      "step": 175
+    },
+    {
+      "epoch": 0.004189279253546605,
+      "grad_norm": 6.315586940182848,
+      "learning_rate": 2e-05,
+      "loss": 0.1958,
+      "step": 176
+    },
+    {
+      "epoch": 0.004213081976578121,
+      "grad_norm": 6.550402546700018,
+      "learning_rate": 2e-05,
+      "loss": 0.8054,
+      "step": 177
+    },
+    {
+      "epoch": 0.004236884699609636,
+      "grad_norm": 8.576889126509522,
+      "learning_rate": 2e-05,
+      "loss": 0.9722,
+      "step": 178
+    },
+    {
+      "epoch": 0.00426068742264115,
+      "grad_norm": 7.132804368870821,
+      "learning_rate": 2e-05,
+      "loss": 0.6629,
+      "step": 179
+    },
+    {
+      "epoch": 0.004284490145672665,
+      "grad_norm": 5.090260061605224,
+      "learning_rate": 2e-05,
+      "loss": 0.2619,
+      "step": 180
+    },
+    {
+      "epoch": 0.00430829286870418,
+      "grad_norm": 4.611451045751536,
+      "learning_rate": 2e-05,
+      "loss": 0.2416,
+      "step": 181
+    },
+    {
+      "epoch": 0.0043320955917356945,
+      "grad_norm": 14.074099537239036,
+      "learning_rate": 2e-05,
+      "loss": 1.8254,
+      "step": 182
+    },
+    {
+      "epoch": 0.004355898314767209,
+      "grad_norm": 4.438837497114084,
+      "learning_rate": 2e-05,
+      "loss": 0.2939,
+      "step": 183
+    },
+    {
+      "epoch": 0.004379701037798724,
+      "grad_norm": 13.508502421791171,
+      "learning_rate": 2e-05,
+      "loss": 0.6057,
+      "step": 184
+    },
+    {
+      "epoch": 0.004403503760830239,
+      "grad_norm": 5.661129234821868,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 185
+    },
+    {
+      "epoch": 0.004427306483861753,
+      "grad_norm": 17.99263233910478,
+      "learning_rate": 2e-05,
+      "loss": 0.7255,
+      "step": 186
+    },
+    {
+      "epoch": 0.004451109206893269,
+      "grad_norm": 8.299297502411198,
+      "learning_rate": 2e-05,
+      "loss": 0.7336,
+      "step": 187
+    },
+    {
+      "epoch": 0.004474911929924784,
+      "grad_norm": 9.050778460325041,
+      "learning_rate": 2e-05,
+      "loss": 0.3655,
+      "step": 188
+    },
+    {
+      "epoch": 0.004498714652956298,
+      "grad_norm": 3.7962504559522503,
+      "learning_rate": 2e-05,
+      "loss": 0.2888,
+      "step": 189
+    },
+    {
+      "epoch": 0.004522517375987813,
+      "grad_norm": 3.796206878691512,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 190
+    },
+    {
+      "epoch": 0.004546320099019328,
+      "grad_norm": 8.984674670043901,
+      "learning_rate": 2e-05,
+      "loss": 0.3579,
+      "step": 191
+    },
+    {
+      "epoch": 0.004570122822050843,
+      "grad_norm": 4.6246607714000145,
+      "learning_rate": 2e-05,
+      "loss": 0.201,
+      "step": 192
+    },
+    {
+      "epoch": 0.004593925545082357,
+      "grad_norm": 6.736466292740362,
+      "learning_rate": 2e-05,
+      "loss": 0.2761,
+      "step": 193
+    },
+    {
+      "epoch": 0.004617728268113872,
+      "grad_norm": 3.1413557521827844,
+      "learning_rate": 2e-05,
+      "loss": 0.2387,
+      "step": 194
+    },
+    {
+      "epoch": 0.004641530991145387,
+      "grad_norm": 10.538799294989657,
+      "learning_rate": 2e-05,
+      "loss": 0.9887,
+      "step": 195
+    },
+    {
+      "epoch": 0.0046653337141769015,
+      "grad_norm": 7.7064773311793315,
+      "learning_rate": 2e-05,
+      "loss": 0.411,
+      "step": 196
+    },
+    {
+      "epoch": 0.004689136437208417,
+      "grad_norm": 4.195001129313723,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 197
+    },
+    {
+      "epoch": 0.004712939160239932,
+      "grad_norm": 4.274305774043567,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 198
+    },
+    {
+      "epoch": 0.0047367418832714465,
+      "grad_norm": 6.620958095272695,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 199
+    },
+    {
+      "epoch": 0.004760544606302961,
+      "grad_norm": 13.924872586971281,
+      "learning_rate": 2e-05,
+      "loss": 1.0897,
+      "step": 200
+    },
+    {
+      "epoch": 0.004784347329334476,
+      "grad_norm": 2.4168209336359663,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 201
+    },
+    {
+      "epoch": 0.004808150052365991,
+      "grad_norm": 6.249503072747882,
+      "learning_rate": 2e-05,
+      "loss": 0.3816,
+      "step": 202
+    },
+    {
+      "epoch": 0.004831952775397505,
+      "grad_norm": 10.265765132380942,
+      "learning_rate": 2e-05,
+      "loss": 0.883,
+      "step": 203
+    },
+    {
+      "epoch": 0.00485575549842902,
+      "grad_norm": 2.5975250706526776,
+      "learning_rate": 2e-05,
+      "loss": 0.1675,
+      "step": 204
+    },
+    {
+      "epoch": 0.004879558221460535,
+      "grad_norm": 2.1725607965438773,
+      "learning_rate": 2e-05,
+      "loss": 0.1507,
+      "step": 205
+    },
+    {
+      "epoch": 0.0049033609444920495,
+      "grad_norm": 5.030515875420312,
+      "learning_rate": 2e-05,
+      "loss": 0.1605,
+      "step": 206
+    },
+    {
+      "epoch": 0.004927163667523565,
+      "grad_norm": 4.16737505435962,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 207
+    },
+    {
+      "epoch": 0.00495096639055508,
+      "grad_norm": 4.926727494210655,
+      "learning_rate": 2e-05,
+      "loss": 0.2662,
+      "step": 208
+    },
+    {
+      "epoch": 0.0049747691135865945,
+      "grad_norm": 6.03057818709044,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 209
+    },
+    {
+      "epoch": 0.004998571836618109,
+      "grad_norm": 5.01426984954578,
+      "learning_rate": 2e-05,
+      "loss": 0.672,
+      "step": 210
+    },
+    {
+      "epoch": 0.005022374559649624,
+      "grad_norm": 8.541659723302256,
+      "learning_rate": 2e-05,
+      "loss": 0.3529,
+      "step": 211
+    },
+    {
+      "epoch": 0.005046177282681139,
+      "grad_norm": 7.12004198179627,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 212
+    },
+    {
+      "epoch": 0.005069980005712653,
+      "grad_norm": 3.03816041424151,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 213
+    },
+    {
+      "epoch": 0.005093782728744168,
+      "grad_norm": 2.9989427367152452,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 214
+    },
+    {
+      "epoch": 0.005117585451775683,
+      "grad_norm": 3.7924010059260134,
+      "learning_rate": 2e-05,
+      "loss": 0.2157,
+      "step": 215
+    },
+    {
+      "epoch": 0.005141388174807198,
+      "grad_norm": 8.504167656710486,
+      "learning_rate": 2e-05,
+      "loss": 0.9202,
+      "step": 216
+    },
+    {
+      "epoch": 0.005165190897838713,
+      "grad_norm": 5.916047995753012,
+      "learning_rate": 2e-05,
+      "loss": 0.467,
+      "step": 217
+    },
+    {
+      "epoch": 0.005188993620870228,
+      "grad_norm": 16.08951564480272,
+      "learning_rate": 2e-05,
+      "loss": 1.1827,
+      "step": 218
+    },
+    {
+      "epoch": 0.005212796343901743,
+      "grad_norm": 6.058782415333275,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 219
+    },
+    {
+      "epoch": 0.005236599066933257,
+      "grad_norm": 9.601321511097174,
+      "learning_rate": 2e-05,
+      "loss": 0.2629,
+      "step": 220
+    },
+    {
+      "epoch": 0.005260401789964772,
+      "grad_norm": 3.3384221746109226,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 221
+    },
+    {
+      "epoch": 0.005284204512996287,
+      "grad_norm": 4.076950985889785,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 222
+    },
+    {
+      "epoch": 0.0053080072360278015,
+      "grad_norm": 9.184784079488752,
+      "learning_rate": 2e-05,
+      "loss": 0.9258,
+      "step": 223
+    },
+    {
+      "epoch": 0.005331809959059316,
+      "grad_norm": 10.479287746921925,
+      "learning_rate": 2e-05,
+      "loss": 0.8328,
+      "step": 224
+    },
+    {
+      "epoch": 0.005355612682090831,
+      "grad_norm": 7.67848744419972,
+      "learning_rate": 2e-05,
+      "loss": 0.2401,
+      "step": 225
+    },
+    {
+      "epoch": 0.005379415405122346,
+      "grad_norm": 13.857721747970515,
+      "learning_rate": 2e-05,
+      "loss": 1.2108,
+      "step": 226
+    },
+    {
+      "epoch": 0.005403218128153861,
+      "grad_norm": 3.3049768445451178,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 227
+    },
+    {
+      "epoch": 0.005427020851185376,
+      "grad_norm": 5.210710769782937,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 228
+    },
+    {
+      "epoch": 0.005450823574216891,
+      "grad_norm": 11.716300984396641,
+      "learning_rate": 2e-05,
+      "loss": 1.4372,
+      "step": 229
+    },
+    {
+      "epoch": 0.005474626297248405,
+      "grad_norm": 8.167303827278793,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 230
+    },
+    {
+      "epoch": 0.00549842902027992,
+      "grad_norm": 10.392160222267831,
+      "learning_rate": 2e-05,
+      "loss": 0.5664,
+      "step": 231
+    },
+    {
+      "epoch": 0.005522231743311435,
+      "grad_norm": 6.297066798049351,
+      "learning_rate": 2e-05,
+      "loss": 0.4559,
+      "step": 232
+    },
+    {
+      "epoch": 0.0055460344663429495,
+      "grad_norm": 7.007488879343935,
+      "learning_rate": 2e-05,
+      "loss": 0.5199,
+      "step": 233
+    },
+    {
+      "epoch": 0.005569837189374464,
+      "grad_norm": 9.890432892035832,
+      "learning_rate": 2e-05,
+      "loss": 0.4169,
+      "step": 234
+    },
+    {
+      "epoch": 0.005593639912405979,
+      "grad_norm": 7.465400172397628,
+      "learning_rate": 2e-05,
+      "loss": 0.4212,
+      "step": 235
+    },
+    {
+      "epoch": 0.005617442635437494,
+      "grad_norm": 9.586102613001588,
+      "learning_rate": 2e-05,
+      "loss": 0.3336,
+      "step": 236
+    },
+    {
+      "epoch": 0.005641245358469009,
+      "grad_norm": 342.9306059496556,
+      "learning_rate": 2e-05,
+      "loss": 5.5095,
+      "step": 237
+    },
+    {
+      "epoch": 0.005665048081500524,
+      "grad_norm": 98.29055362242472,
+      "learning_rate": 2e-05,
+      "loss": 3.1127,
+      "step": 238
+    },
+    {
+      "epoch": 0.005688850804532039,
+      "grad_norm": 14.566787420707216,
+      "learning_rate": 2e-05,
+      "loss": 0.8438,
+      "step": 239
+    },
+    {
+      "epoch": 0.0057126535275635534,
+      "grad_norm": 8.05111025262108,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 240
+    },
+    {
+      "epoch": 0.005736456250595068,
+      "grad_norm": 7.274214104011209,
+      "learning_rate": 2e-05,
+      "loss": 0.302,
+      "step": 241
+    },
+    {
+      "epoch": 0.005760258973626583,
+      "grad_norm": 8.488512415183846,
+      "learning_rate": 2e-05,
+      "loss": 0.6961,
+      "step": 242
+    },
+    {
+      "epoch": 0.005784061696658098,
+      "grad_norm": 6.695934672069646,
+      "learning_rate": 2e-05,
+      "loss": 0.7345,
+      "step": 243
+    },
+    {
+      "epoch": 0.005807864419689612,
+      "grad_norm": 11.208961171095018,
+      "learning_rate": 2e-05,
+      "loss": 0.8595,
+      "step": 244
+    },
+    {
+      "epoch": 0.005831667142721127,
+      "grad_norm": 11.946677256218974,
+      "learning_rate": 2e-05,
+      "loss": 1.0238,
+      "step": 245
+    },
+    {
+      "epoch": 0.005855469865752642,
+      "grad_norm": 8.235780581101752,
+      "learning_rate": 2e-05,
+      "loss": 0.3881,
+      "step": 246
+    },
+    {
+      "epoch": 0.005879272588784157,
+      "grad_norm": 4.208948954340002,
+      "learning_rate": 2e-05,
+      "loss": 0.2622,
+      "step": 247
+    },
+    {
+      "epoch": 0.005903075311815672,
+      "grad_norm": 4.774894452410603,
+      "learning_rate": 2e-05,
+      "loss": 0.1788,
+      "step": 248
+    },
+    {
+      "epoch": 0.005926878034847187,
+      "grad_norm": 52.16593511196805,
+      "learning_rate": 2e-05,
+      "loss": 0.9382,
+      "step": 249
+    },
+    {
+      "epoch": 0.0059506807578787015,
+      "grad_norm": 4.8572100861324286,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 250
+    },
+    {
+      "epoch": 0.005974483480910216,
+      "grad_norm": 10.834858039813883,
+      "learning_rate": 2e-05,
+      "loss": 1.3558,
+      "step": 251
+    },
+    {
+      "epoch": 0.005998286203941731,
+      "grad_norm": 1170.4627998453318,
+      "learning_rate": 2e-05,
+      "loss": 1.0358,
+      "step": 252
+    },
+    {
+      "epoch": 0.006022088926973246,
+      "grad_norm": 10.231788365239842,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 253
+    },
+    {
+      "epoch": 0.00604589165000476,
+      "grad_norm": 9.469484224992026,
+      "learning_rate": 2e-05,
+      "loss": 0.9913,
+      "step": 254
+    },
+    {
+      "epoch": 0.006069694373036275,
+      "grad_norm": 2.8996108567226884,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 255
+    },
+    {
+      "epoch": 0.00609349709606779,
+      "grad_norm": 11.590784855312924,
+      "learning_rate": 2e-05,
+      "loss": 0.4912,
+      "step": 256
+    },
+    {
+      "epoch": 0.0061172998190993045,
+      "grad_norm": 4.98305863129011,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 257
+    },
+    {
+      "epoch": 0.00614110254213082,
+      "grad_norm": 11.551860474046203,
+      "learning_rate": 2e-05,
+      "loss": 1.0509,
+      "step": 258
+    },
+    {
+      "epoch": 0.006164905265162335,
+      "grad_norm": 3.1765420460809253,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 259
+    },
+    {
+      "epoch": 0.0061887079881938496,
+      "grad_norm": 7.303154516582469,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 260
+    },
+    {
+      "epoch": 0.006212510711225364,
+      "grad_norm": 5.356446782842391,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 261
+    },
+    {
+      "epoch": 0.006236313434256879,
+      "grad_norm": 4.172859068012919,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 262
+    },
+    {
+      "epoch": 0.006260116157288394,
+      "grad_norm": 7.810671841172967,
+      "learning_rate": 2e-05,
+      "loss": 0.337,
+      "step": 263
+    },
+    {
+      "epoch": 0.006283918880319908,
+      "grad_norm": 10.341043566394715,
+      "learning_rate": 2e-05,
+      "loss": 1.5802,
+      "step": 264
+    },
+    {
+      "epoch": 0.006307721603351423,
+      "grad_norm": 16.415068402852558,
+      "learning_rate": 2e-05,
+      "loss": 2.4215,
+      "step": 265
+    },
+    {
+      "epoch": 0.006331524326382938,
+      "grad_norm": 9.685249192054375,
+      "learning_rate": 2e-05,
+      "loss": 0.4179,
+      "step": 266
+    },
+    {
+      "epoch": 0.006355327049414453,
+      "grad_norm": 73.56057527986158,
+      "learning_rate": 2e-05,
+      "loss": 1.0267,
+      "step": 267
+    },
+    {
+      "epoch": 0.006379129772445968,
+      "grad_norm": 13.582005426650237,
+      "learning_rate": 2e-05,
+      "loss": 0.7239,
+      "step": 268
+    },
+    {
+      "epoch": 0.006402932495477483,
+      "grad_norm": 5.84625680668533,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 269
+    },
+    {
+      "epoch": 0.006426735218508998,
+      "grad_norm": 3.77653533698311,
+      "learning_rate": 2e-05,
+      "loss": 0.2249,
+      "step": 270
+    },
+    {
+      "epoch": 0.006450537941540512,
+      "grad_norm": 10.062994801405063,
+      "learning_rate": 2e-05,
+      "loss": 0.2893,
+      "step": 271
+    },
+    {
+      "epoch": 0.006474340664572027,
+      "grad_norm": 8.363618996567773,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 272
+    },
+    {
+      "epoch": 0.006498143387603542,
+      "grad_norm": 9.05717568381465,
+      "learning_rate": 2e-05,
+      "loss": 0.2757,
+      "step": 273
+    },
+    {
+      "epoch": 0.0065219461106350565,
+      "grad_norm": 15.948356309728332,
+      "learning_rate": 2e-05,
+      "loss": 0.4001,
+      "step": 274
+    },
+    {
+      "epoch": 0.006545748833666571,
+      "grad_norm": 5.488990734647741,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 275
+    },
+    {
+      "epoch": 0.006569551556698086,
+      "grad_norm": 3.8750892370676517,
+      "learning_rate": 2e-05,
+      "loss": 0.211,
+      "step": 276
+    },
+    {
+      "epoch": 0.006593354279729601,
+      "grad_norm": 4.616769991213262,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 277
+    },
+    {
+      "epoch": 0.006617157002761116,
+      "grad_norm": 11.190871783718897,
+      "learning_rate": 2e-05,
+      "loss": 0.6435,
+      "step": 278
+    },
+    {
+      "epoch": 0.006640959725792631,
+      "grad_norm": 7.864271826652847,
+      "learning_rate": 2e-05,
+      "loss": 0.275,
+      "step": 279
+    },
+    {
+      "epoch": 0.006664762448824146,
+      "grad_norm": 3.017762128455189,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 280
+    },
+    {
+      "epoch": 0.00668856517185566,
+      "grad_norm": 11.989788309386995,
+      "learning_rate": 2e-05,
+      "loss": 0.8278,
+      "step": 281
+    },
+    {
+      "epoch": 0.006712367894887175,
+      "grad_norm": 6.985016711479004,
+      "learning_rate": 2e-05,
+      "loss": 0.7082,
+      "step": 282
+    },
+    {
+      "epoch": 0.00673617061791869,
+      "grad_norm": 12.620258081639681,
+      "learning_rate": 2e-05,
+      "loss": 1.449,
+      "step": 283
+    },
+    {
+      "epoch": 0.0067599733409502046,
+      "grad_norm": 2.9096341802685743,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 284
+    },
+    {
+      "epoch": 0.006783776063981719,
+      "grad_norm": 4.344372345881651,
+      "learning_rate": 2e-05,
+      "loss": 0.2488,
+      "step": 285
+    },
+    {
+      "epoch": 0.006807578787013234,
+      "grad_norm": 11.506328686253823,
+      "learning_rate": 2e-05,
+      "loss": 1.0215,
+      "step": 286
+    },
+    {
+      "epoch": 0.006831381510044749,
+      "grad_norm": 2.8018805463453815,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 287
+    },
+    {
+      "epoch": 0.006855184233076264,
+      "grad_norm": 7.0309459537321235,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 288
+    },
+    {
+      "epoch": 0.006878986956107779,
+      "grad_norm": 18.764232683137077,
+      "learning_rate": 2e-05,
+      "loss": 0.9613,
+      "step": 289
+    },
+    {
+      "epoch": 0.006902789679139294,
+      "grad_norm": 12.927511637096707,
+      "learning_rate": 2e-05,
+      "loss": 0.419,
+      "step": 290
+    },
+    {
+      "epoch": 0.0069265924021708085,
+      "grad_norm": 4.269360738416405,
+      "learning_rate": 2e-05,
+      "loss": 0.3558,
+      "step": 291
+    },
+    {
+      "epoch": 0.006950395125202323,
+      "grad_norm": 8.339082532823696,
+      "learning_rate": 2e-05,
+      "loss": 0.3467,
+      "step": 292
+    },
+    {
+      "epoch": 0.006974197848233838,
+      "grad_norm": 24.782995117782377,
+      "learning_rate": 2e-05,
+      "loss": 0.5407,
+      "step": 293
+    },
+    {
+      "epoch": 0.006998000571265353,
+      "grad_norm": 709.6144021073233,
+      "learning_rate": 2e-05,
+      "loss": 7.8355,
+      "step": 294
+    },
+    {
+      "epoch": 0.007021803294296867,
+      "grad_norm": 452.8057144926307,
+      "learning_rate": 2e-05,
+      "loss": 3.7161,
+      "step": 295
+    },
+    {
+      "epoch": 0.007045606017328382,
+      "grad_norm": 514.0429621675087,
+      "learning_rate": 2e-05,
+      "loss": 5.4515,
+      "step": 296
+    },
+    {
+      "epoch": 0.007069408740359897,
+      "grad_norm": 221.07817648308142,
+      "learning_rate": 2e-05,
+      "loss": 6.8501,
+      "step": 297
+    },
+    {
+      "epoch": 0.007093211463391412,
+      "grad_norm": 468.75411031447965,
+      "learning_rate": 2e-05,
+      "loss": 3.1921,
+      "step": 298
+    },
+    {
+      "epoch": 0.007117014186422927,
+      "grad_norm": 4.23349522307372,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 299
+    },
+    {
+      "epoch": 0.007140816909454442,
+      "grad_norm": 9.014164146635324,
+      "learning_rate": 2e-05,
+      "loss": 0.3546,
+      "step": 300
+    },
+    {
+      "epoch": 0.0071646196324859565,
+      "grad_norm": 4.074655756233512,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 301
+    },
+    {
+      "epoch": 0.007188422355517471,
+      "grad_norm": 3.2687150588948555,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 302
+    },
+    {
+      "epoch": 0.007212225078548986,
+      "grad_norm": 3.963448549015189,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 303
+    },
+    {
+      "epoch": 0.007236027801580501,
+      "grad_norm": 16.51324235027786,
+      "learning_rate": 2e-05,
+      "loss": 1.5732,
+      "step": 304
+    },
+    {
+      "epoch": 0.007259830524612015,
+      "grad_norm": 4.661840746187736,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 305
+    },
+    {
+      "epoch": 0.00728363324764353,
+      "grad_norm": 3.4885375076644403,
+      "learning_rate": 2e-05,
+      "loss": 0.1924,
+      "step": 306
+    },
+    {
+      "epoch": 0.007307435970675045,
+      "grad_norm": 15.8339521972824,
+      "learning_rate": 2e-05,
+      "loss": 1.9079,
+      "step": 307
+    },
+    {
+      "epoch": 0.00733123869370656,
+      "grad_norm": 8.721496739970055,
+      "learning_rate": 2e-05,
+      "loss": 0.8668,
+      "step": 308
+    },
+    {
+      "epoch": 0.007355041416738075,
+      "grad_norm": 14.374150800293055,
+      "learning_rate": 2e-05,
+      "loss": 1.3146,
+      "step": 309
+    },
+    {
+      "epoch": 0.00737884413976959,
+      "grad_norm": 9.975171644411693,
+      "learning_rate": 2e-05,
+      "loss": 1.0844,
+      "step": 310
+    },
+    {
+      "epoch": 0.007402646862801105,
+      "grad_norm": 4.168346952485528,
+      "learning_rate": 2e-05,
+      "loss": 0.2305,
+      "step": 311
+    },
+    {
+      "epoch": 0.007426449585832619,
+      "grad_norm": 8.264486732272463,
+      "learning_rate": 2e-05,
+      "loss": 1.0608,
+      "step": 312
+    },
+    {
+      "epoch": 0.007450252308864134,
+      "grad_norm": 8.488084309171217,
+      "learning_rate": 2e-05,
+      "loss": 0.3595,
+      "step": 313
+    },
+    {
+      "epoch": 0.007474055031895649,
+      "grad_norm": 12.18305055867038,
+      "learning_rate": 2e-05,
+      "loss": 0.4491,
+      "step": 314
+    },
+    {
+      "epoch": 0.0074978577549271634,
+      "grad_norm": 10.407349824618134,
+      "learning_rate": 2e-05,
+      "loss": 0.9951,
+      "step": 315
+    },
+    {
+      "epoch": 0.007521660477958678,
+      "grad_norm": 5.767103074541441,
+      "learning_rate": 2e-05,
+      "loss": 0.4204,
+      "step": 316
+    },
+    {
+      "epoch": 0.007545463200990193,
+      "grad_norm": 6.105845982958167,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 317
+    },
+    {
+      "epoch": 0.0075692659240217085,
+      "grad_norm": 6.90108466689358,
+      "learning_rate": 2e-05,
+      "loss": 0.6029,
+      "step": 318
+    },
+    {
+      "epoch": 0.007593068647053223,
+      "grad_norm": 11.0291172188518,
+      "learning_rate": 2e-05,
+      "loss": 0.376,
+      "step": 319
+    },
+    {
+      "epoch": 0.007616871370084738,
+      "grad_norm": 2.4268058892860975,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 320
+    },
+    {
+      "epoch": 0.007640674093116253,
+      "grad_norm": 11.441949155658847,
+      "learning_rate": 2e-05,
+      "loss": 0.9931,
+      "step": 321
+    },
+    {
+      "epoch": 0.007664476816147767,
+      "grad_norm": 17.387541060650126,
+      "learning_rate": 2e-05,
+      "loss": 1.2264,
+      "step": 322
+    },
+    {
+      "epoch": 0.007688279539179282,
+      "grad_norm": 3.3194217731373965,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 323
+    },
+    {
+      "epoch": 0.007712082262210797,
+      "grad_norm": 6.2158782213995805,
+      "learning_rate": 2e-05,
+      "loss": 0.4168,
+      "step": 324
+    },
+    {
+      "epoch": 0.0077358849852423115,
+      "grad_norm": 11.131948035742571,
+      "learning_rate": 2e-05,
+      "loss": 0.5027,
+      "step": 325
+    },
+    {
+      "epoch": 0.007759687708273826,
+      "grad_norm": 9.754709449464913,
+      "learning_rate": 2e-05,
+      "loss": 0.5384,
+      "step": 326
+    },
+    {
+      "epoch": 0.007783490431305341,
+      "grad_norm": 7.645283525424699,
+      "learning_rate": 2e-05,
+      "loss": 0.2389,
+      "step": 327
+    },
+    {
+      "epoch": 0.0078072931543368565,
+      "grad_norm": 9.640677474530484,
+      "learning_rate": 2e-05,
+      "loss": 1.5772,
+      "step": 328
+    },
+    {
+      "epoch": 0.00783109587736837,
+      "grad_norm": 17.114295069538827,
+      "learning_rate": 2e-05,
+      "loss": 1.7545,
+      "step": 329
+    },
+    {
+      "epoch": 0.007854898600399885,
+      "grad_norm": 9.796992923565712,
+      "learning_rate": 2e-05,
+      "loss": 0.6743,
+      "step": 330
+    },
+    {
+      "epoch": 0.0078787013234314,
+      "grad_norm": 4.66145848880958,
+      "learning_rate": 2e-05,
+      "loss": 0.2478,
+      "step": 331
+    },
+    {
+      "epoch": 0.007902504046462915,
+      "grad_norm": 5.576258802296461,
+      "learning_rate": 2e-05,
+      "loss": 0.3442,
+      "step": 332
+    },
+    {
+      "epoch": 0.007926306769494431,
+      "grad_norm": 5.148505967386017,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 333
+    },
+    {
+      "epoch": 0.007950109492525946,
+      "grad_norm": 3.8621877625564043,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 334
+    },
+    {
+      "epoch": 0.00797391221555746,
+      "grad_norm": 7.002689917740424,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 335
+    },
+    {
+      "epoch": 0.007997714938588975,
+      "grad_norm": 13.365574518247836,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 336
+    },
+    {
+      "epoch": 0.00802151766162049,
+      "grad_norm": 7.497697856035734,
+      "learning_rate": 2e-05,
+      "loss": 0.8468,
+      "step": 337
+    },
+    {
+      "epoch": 0.008045320384652005,
+      "grad_norm": 10.118084952835165,
+      "learning_rate": 2e-05,
+      "loss": 0.9178,
+      "step": 338
+    },
+    {
+      "epoch": 0.00806912310768352,
+      "grad_norm": 9.006973440080131,
+      "learning_rate": 2e-05,
+      "loss": 0.7884,
+      "step": 339
+    },
+    {
+      "epoch": 0.008092925830715034,
+      "grad_norm": 4.84206196560659,
+      "learning_rate": 2e-05,
+      "loss": 0.3565,
+      "step": 340
+    },
+    {
+      "epoch": 0.008116728553746549,
+      "grad_norm": 4.022580230711682,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 341
+    },
+    {
+      "epoch": 0.008140531276778063,
+      "grad_norm": 4.981086941950912,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 342
+    },
+    {
+      "epoch": 0.008164333999809578,
+      "grad_norm": 6.939876591645685,
+      "learning_rate": 2e-05,
+      "loss": 0.6901,
+      "step": 343
+    },
+    {
+      "epoch": 0.008188136722841093,
+      "grad_norm": 51.853618394163284,
+      "learning_rate": 2e-05,
+      "loss": 1.3191,
+      "step": 344
+    },
+    {
+      "epoch": 0.008211939445872608,
+      "grad_norm": 6.207581750554001,
+      "learning_rate": 2e-05,
+      "loss": 0.2048,
+      "step": 345
+    },
+    {
+      "epoch": 0.008235742168904122,
+      "grad_norm": 4.902427685717893,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 346
+    },
+    {
+      "epoch": 0.008259544891935637,
+      "grad_norm": 4.6299613635145835,
+      "learning_rate": 2e-05,
+      "loss": 0.2263,
+      "step": 347
+    },
+    {
+      "epoch": 0.008283347614967152,
+      "grad_norm": 6.406151389495322,
+      "learning_rate": 2e-05,
+      "loss": 0.2639,
+      "step": 348
+    },
+    {
+      "epoch": 0.008307150337998667,
+      "grad_norm": 7.231289477017987,
+      "learning_rate": 2e-05,
+      "loss": 0.4007,
+      "step": 349
+    },
+    {
+      "epoch": 0.008330953061030181,
+      "grad_norm": 45.148825141648096,
+      "learning_rate": 2e-05,
+      "loss": 0.5668,
+      "step": 350
+    },
+    {
+      "epoch": 0.008354755784061696,
+      "grad_norm": 6.738476177486717,
+      "learning_rate": 2e-05,
+      "loss": 0.2764,
+      "step": 351
+    },
+    {
+      "epoch": 0.00837855850709321,
+      "grad_norm": 8.137635693355513,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 352
+    },
+    {
+      "epoch": 0.008402361230124727,
+      "grad_norm": 8.808926356698557,
+      "learning_rate": 2e-05,
+      "loss": 0.5452,
+      "step": 353
+    },
+    {
+      "epoch": 0.008426163953156242,
+      "grad_norm": 11.635026581223567,
+      "learning_rate": 2e-05,
+      "loss": 0.8617,
+      "step": 354
+    },
+    {
+      "epoch": 0.008449966676187757,
+      "grad_norm": 11.11243958231491,
+      "learning_rate": 2e-05,
+      "loss": 0.3394,
+      "step": 355
+    },
+    {
+      "epoch": 0.008473769399219271,
+      "grad_norm": 9.950308094203013,
+      "learning_rate": 2e-05,
+      "loss": 0.3002,
+      "step": 356
+    },
+    {
+      "epoch": 0.008497572122250786,
+      "grad_norm": 2.622841319632534,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 357
+    },
+    {
+      "epoch": 0.0085213748452823,
+      "grad_norm": 8.311284929556349,
+      "learning_rate": 2e-05,
+      "loss": 1.1483,
+      "step": 358
+    },
+    {
+      "epoch": 0.008545177568313815,
+      "grad_norm": 13.483173754639424,
+      "learning_rate": 2e-05,
+      "loss": 0.9927,
+      "step": 359
+    },
+    {
+      "epoch": 0.00856898029134533,
+      "grad_norm": 3.729336736466892,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 360
+    },
+    {
+      "epoch": 0.008592783014376845,
+      "grad_norm": 9.63742127854641,
+      "learning_rate": 2e-05,
+      "loss": 0.5604,
+      "step": 361
+    },
+    {
+      "epoch": 0.00861658573740836,
+      "grad_norm": 6.611898470271389,
+      "learning_rate": 2e-05,
+      "loss": 0.8191,
+      "step": 362
+    },
+    {
+      "epoch": 0.008640388460439874,
+      "grad_norm": 8.960511697453653,
+      "learning_rate": 2e-05,
+      "loss": 0.7483,
+      "step": 363
+    },
+    {
+      "epoch": 0.008664191183471389,
+      "grad_norm": 3.1002459510448244,
+      "learning_rate": 2e-05,
+      "loss": 0.2482,
+      "step": 364
+    },
+    {
+      "epoch": 0.008687993906502904,
+      "grad_norm": 8.336470707368484,
+      "learning_rate": 2e-05,
+      "loss": 0.2863,
+      "step": 365
+    },
+    {
+      "epoch": 0.008711796629534418,
+      "grad_norm": 8.556434131441705,
+      "learning_rate": 2e-05,
+      "loss": 1.1096,
+      "step": 366
+    },
+    {
+      "epoch": 0.008735599352565933,
+      "grad_norm": 9.554244776040209,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 367
+    },
+    {
+      "epoch": 0.008759402075597448,
+      "grad_norm": 9.228840209244154,
+      "learning_rate": 2e-05,
+      "loss": 1.0148,
+      "step": 368
+    },
+    {
+      "epoch": 0.008783204798628963,
+      "grad_norm": 8.528827902767004,
+      "learning_rate": 2e-05,
+      "loss": 1.1119,
+      "step": 369
+    },
+    {
+      "epoch": 0.008807007521660477,
+      "grad_norm": 10.070780641051378,
+      "learning_rate": 2e-05,
+      "loss": 0.9059,
+      "step": 370
+    },
+    {
+      "epoch": 0.008830810244691992,
+      "grad_norm": 11.043776065229501,
+      "learning_rate": 2e-05,
+      "loss": 1.0566,
+      "step": 371
+    },
+    {
+      "epoch": 0.008854612967723507,
+      "grad_norm": 7.726929682046753,
+      "learning_rate": 2e-05,
+      "loss": 0.6171,
+      "step": 372
+    },
+    {
+      "epoch": 0.008878415690755023,
+      "grad_norm": 2.7101511103599307,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 373
+    },
+    {
+      "epoch": 0.008902218413786538,
+      "grad_norm": 7.759647102635494,
+      "learning_rate": 2e-05,
+      "loss": 1.3435,
+      "step": 374
+    },
+    {
+      "epoch": 0.008926021136818053,
+      "grad_norm": 3.428897255428793,
+      "learning_rate": 2e-05,
+      "loss": 0.2331,
+      "step": 375
+    },
+    {
+      "epoch": 0.008949823859849567,
+      "grad_norm": 3.438027920575452,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 376
+    },
+    {
+      "epoch": 0.008973626582881082,
+      "grad_norm": 11.595884541778288,
+      "learning_rate": 2e-05,
+      "loss": 0.4278,
+      "step": 377
+    },
+    {
+      "epoch": 0.008997429305912597,
+      "grad_norm": 9.611003875803128,
+      "learning_rate": 2e-05,
+      "loss": 0.8037,
+      "step": 378
+    },
+    {
+      "epoch": 0.009021232028944112,
+      "grad_norm": 8.475494391625016,
+      "learning_rate": 2e-05,
+      "loss": 1.1745,
+      "step": 379
+    },
+    {
+      "epoch": 0.009045034751975626,
+      "grad_norm": 5.068810533927868,
+      "learning_rate": 2e-05,
+      "loss": 0.2526,
+      "step": 380
+    },
+    {
+      "epoch": 0.009068837475007141,
+      "grad_norm": 10.96785235752194,
+      "learning_rate": 2e-05,
+      "loss": 1.5685,
+      "step": 381
+    },
+    {
+      "epoch": 0.009092640198038656,
+      "grad_norm": 7.818081051286203,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 382
+    },
+    {
+      "epoch": 0.00911644292107017,
+      "grad_norm": 8.559958289512709,
+      "learning_rate": 2e-05,
+      "loss": 0.7142,
+      "step": 383
+    },
+    {
+      "epoch": 0.009140245644101685,
+      "grad_norm": 2.821818613366687,
+      "learning_rate": 2e-05,
+      "loss": 0.2329,
+      "step": 384
+    },
+    {
+      "epoch": 0.0091640483671332,
+      "grad_norm": 9.528375963404226,
+      "learning_rate": 2e-05,
+      "loss": 0.8696,
+      "step": 385
+    },
+    {
+      "epoch": 0.009187851090164715,
+      "grad_norm": 9.38680687435836,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 386
+    },
+    {
+      "epoch": 0.00921165381319623,
+      "grad_norm": 7.0574280934555835,
+      "learning_rate": 2e-05,
+      "loss": 0.8412,
+      "step": 387
+    },
+    {
+      "epoch": 0.009235456536227744,
+      "grad_norm": 4.1968146978492795,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 388
+    },
+    {
+      "epoch": 0.009259259259259259,
+      "grad_norm": 8.89238127270584,
+      "learning_rate": 2e-05,
+      "loss": 1.1991,
+      "step": 389
+    },
+    {
+      "epoch": 0.009283061982290773,
+      "grad_norm": 9.93033576208142,
+      "learning_rate": 2e-05,
+      "loss": 0.4885,
+      "step": 390
+    },
+    {
+      "epoch": 0.009306864705322288,
+      "grad_norm": 3.794959195365863,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 391
+    },
+    {
+      "epoch": 0.009330667428353803,
+      "grad_norm": 2.4455837884325025,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 392
+    },
+    {
+      "epoch": 0.009354470151385318,
+      "grad_norm": 7.974341131849344,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 393
+    },
+    {
+      "epoch": 0.009378272874416834,
+      "grad_norm": 12.954431308582741,
+      "learning_rate": 2e-05,
+      "loss": 0.7834,
+      "step": 394
+    },
+    {
+      "epoch": 0.009402075597448349,
+      "grad_norm": 5.588758328770761,
+      "learning_rate": 2e-05,
+      "loss": 0.4512,
+      "step": 395
+    },
+    {
+      "epoch": 0.009425878320479864,
+      "grad_norm": 8.774912068781536,
+      "learning_rate": 2e-05,
+      "loss": 0.3757,
+      "step": 396
+    },
+    {
+      "epoch": 0.009449681043511378,
+      "grad_norm": 2.9887731098788564,
+      "learning_rate": 2e-05,
+      "loss": 0.2742,
+      "step": 397
+    },
+    {
+      "epoch": 0.009473483766542893,
+      "grad_norm": 4.046155143321405,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 398
+    },
+    {
+      "epoch": 0.009497286489574408,
+      "grad_norm": 9.584567857431278,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 399
+    },
+    {
+      "epoch": 0.009521089212605922,
+      "grad_norm": 4.345731755603215,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 400
+    },
+    {
+      "epoch": 0.009544891935637437,
+      "grad_norm": 6.5814301312626045,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 401
+    },
+    {
+      "epoch": 0.009568694658668952,
+      "grad_norm": 9.348211605834788,
+      "learning_rate": 2e-05,
+      "loss": 0.6956,
+      "step": 402
+    },
+    {
+      "epoch": 0.009592497381700467,
+      "grad_norm": 8.58865657724457,
+      "learning_rate": 2e-05,
+      "loss": 0.3545,
+      "step": 403
+    },
+    {
+      "epoch": 0.009616300104731981,
+      "grad_norm": 4.330200010822778,
+      "learning_rate": 2e-05,
+      "loss": 0.1913,
+      "step": 404
+    },
+    {
+      "epoch": 0.009640102827763496,
+      "grad_norm": 6.607224362550519,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 405
+    },
+    {
+      "epoch": 0.00966390555079501,
+      "grad_norm": 7.108260350008049,
+      "learning_rate": 2e-05,
+      "loss": 1.1078,
+      "step": 406
+    },
+    {
+      "epoch": 0.009687708273826525,
+      "grad_norm": 6.638414990873814,
+      "learning_rate": 2e-05,
+      "loss": 0.6971,
+      "step": 407
+    },
+    {
+      "epoch": 0.00971151099685804,
+      "grad_norm": 10.56819778914329,
+      "learning_rate": 2e-05,
+      "loss": 0.8982,
+      "step": 408
+    },
+    {
+      "epoch": 0.009735313719889555,
+      "grad_norm": 3.5855470631763904,
+      "learning_rate": 2e-05,
+      "loss": 0.2311,
+      "step": 409
+    },
+    {
+      "epoch": 0.00975911644292107,
+      "grad_norm": 8.643865506593242,
+      "learning_rate": 2e-05,
+      "loss": 0.6096,
+      "step": 410
+    },
+    {
+      "epoch": 0.009782919165952584,
+      "grad_norm": 10.773251837922958,
+      "learning_rate": 2e-05,
+      "loss": 1.042,
+      "step": 411
+    },
+    {
+      "epoch": 0.009806721888984099,
+      "grad_norm": 10.336223154112162,
+      "learning_rate": 2e-05,
+      "loss": 1.6308,
+      "step": 412
+    },
+    {
+      "epoch": 0.009830524612015614,
+      "grad_norm": 6.830826101051822,
+      "learning_rate": 2e-05,
+      "loss": 0.6813,
+      "step": 413
+    },
+    {
+      "epoch": 0.00985432733504713,
+      "grad_norm": 7.123187291383528,
+      "learning_rate": 2e-05,
+      "loss": 0.5855,
+      "step": 414
+    },
+    {
+      "epoch": 0.009878130058078645,
+      "grad_norm": 10.4511760083669,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 415
+    },
+    {
+      "epoch": 0.00990193278111016,
+      "grad_norm": 8.542768667573093,
+      "learning_rate": 2e-05,
+      "loss": 0.6916,
+      "step": 416
+    },
+    {
+      "epoch": 0.009925735504141674,
+      "grad_norm": 3.329233541564618,
+      "learning_rate": 2e-05,
+      "loss": 0.1891,
+      "step": 417
+    },
+    {
+      "epoch": 0.009949538227173189,
+      "grad_norm": 2.5004743439715797,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 418
+    },
+    {
+      "epoch": 0.009973340950204704,
+      "grad_norm": 9.52950400466305,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 419
+    },
+    {
+      "epoch": 0.009997143673236219,
+      "grad_norm": 2.701915051529953,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 420
+    },
+    {
+      "epoch": 0.010020946396267733,
+      "grad_norm": 7.056814373145077,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 421
+    },
+    {
+      "epoch": 0.010044749119299248,
+      "grad_norm": 8.767081740518442,
+      "learning_rate": 2e-05,
+      "loss": 0.3962,
+      "step": 422
+    },
+    {
+      "epoch": 0.010068551842330763,
+      "grad_norm": 8.30629579618822,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 423
+    },
+    {
+      "epoch": 0.010092354565362277,
+      "grad_norm": 4.96216562661547,
+      "learning_rate": 2e-05,
+      "loss": 0.2651,
+      "step": 424
+    },
+    {
+      "epoch": 0.010116157288393792,
+      "grad_norm": 8.529647145591861,
+      "learning_rate": 2e-05,
+      "loss": 0.3446,
+      "step": 425
+    },
+    {
+      "epoch": 0.010139960011425307,
+      "grad_norm": 7.439328880476745,
+      "learning_rate": 2e-05,
+      "loss": 0.8816,
+      "step": 426
+    },
+    {
+      "epoch": 0.010163762734456822,
+      "grad_norm": 8.676823893148208,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 427
+    },
+    {
+      "epoch": 0.010187565457488336,
+      "grad_norm": 8.797447597687054,
+      "learning_rate": 2e-05,
+      "loss": 1.022,
+      "step": 428
+    },
+    {
+      "epoch": 0.010211368180519851,
+      "grad_norm": 3.637758665835904,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 429
+    },
+    {
+      "epoch": 0.010235170903551366,
+      "grad_norm": 7.79067891914131,
+      "learning_rate": 2e-05,
+      "loss": 1.1113,
+      "step": 430
+    },
+    {
+      "epoch": 0.01025897362658288,
+      "grad_norm": 3.4881201110198323,
+      "learning_rate": 2e-05,
+      "loss": 0.1953,
+      "step": 431
+    },
+    {
+      "epoch": 0.010282776349614395,
+      "grad_norm": 4.208948049110526,
+      "learning_rate": 2e-05,
+      "loss": 0.2955,
+      "step": 432
+    },
+    {
+      "epoch": 0.01030657907264591,
+      "grad_norm": 14.178826150730004,
+      "learning_rate": 2e-05,
+      "loss": 2.1197,
+      "step": 433
+    },
+    {
+      "epoch": 0.010330381795677426,
+      "grad_norm": 7.85882787066156,
+      "learning_rate": 2e-05,
+      "loss": 0.6832,
+      "step": 434
+    },
+    {
+      "epoch": 0.010354184518708941,
+      "grad_norm": 14.976567506046758,
+      "learning_rate": 2e-05,
+      "loss": 1.7924,
+      "step": 435
+    },
+    {
+      "epoch": 0.010377987241740456,
+      "grad_norm": 3.933954300129926,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 436
+    },
+    {
+      "epoch": 0.01040178996477197,
+      "grad_norm": 6.880978750904038,
+      "learning_rate": 2e-05,
+      "loss": 0.7961,
+      "step": 437
+    },
+    {
+      "epoch": 0.010425592687803485,
+      "grad_norm": 2.903817250989802,
+      "learning_rate": 2e-05,
+      "loss": 0.1861,
+      "step": 438
+    },
+    {
+      "epoch": 0.010449395410835,
+      "grad_norm": 13.517990460564874,
+      "learning_rate": 2e-05,
+      "loss": 0.6955,
+      "step": 439
+    },
+    {
+      "epoch": 0.010473198133866515,
+      "grad_norm": 5.928383902577188,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 440
+    },
+    {
+      "epoch": 0.01049700085689803,
+      "grad_norm": 6.332823645408876,
+      "learning_rate": 2e-05,
+      "loss": 0.6874,
+      "step": 441
+    },
+    {
+      "epoch": 0.010520803579929544,
+      "grad_norm": 6.00340506380582,
+      "learning_rate": 2e-05,
+      "loss": 0.7123,
+      "step": 442
+    },
+    {
+      "epoch": 0.010544606302961059,
+      "grad_norm": 11.859654953385403,
+      "learning_rate": 2e-05,
+      "loss": 1.6779,
+      "step": 443
+    },
+    {
+      "epoch": 0.010568409025992574,
+      "grad_norm": 8.420167221253738,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 444
+    },
+    {
+      "epoch": 0.010592211749024088,
+      "grad_norm": 3.930025818707532,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 445
+    },
+    {
+      "epoch": 0.010616014472055603,
+      "grad_norm": 15.812610627737026,
+      "learning_rate": 2e-05,
+      "loss": 2.1901,
+      "step": 446
+    },
+    {
+      "epoch": 0.010639817195087118,
+      "grad_norm": 9.649428978109276,
+      "learning_rate": 2e-05,
+      "loss": 0.4879,
+      "step": 447
+    },
+    {
+      "epoch": 0.010663619918118632,
+      "grad_norm": 7.920134611094551,
+      "learning_rate": 2e-05,
+      "loss": 0.3871,
+      "step": 448
+    },
+    {
+      "epoch": 0.010687422641150147,
+      "grad_norm": 6.985724838168024,
+      "learning_rate": 2e-05,
+      "loss": 0.5648,
+      "step": 449
+    },
+    {
+      "epoch": 0.010711225364181662,
+      "grad_norm": 3.957753708010334,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 450
+    },
+    {
+      "epoch": 0.010735028087213177,
+      "grad_norm": 7.877748884729752,
+      "learning_rate": 2e-05,
+      "loss": 0.35,
+      "step": 451
+    },
+    {
+      "epoch": 0.010758830810244691,
+      "grad_norm": 6.827686657448881,
+      "learning_rate": 2e-05,
+      "loss": 0.3854,
+      "step": 452
+    },
+    {
+      "epoch": 0.010782633533276206,
+      "grad_norm": 5.1705744991414,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 453
+    },
+    {
+      "epoch": 0.010806436256307722,
+      "grad_norm": 6.502624653242156,
+      "learning_rate": 2e-05,
+      "loss": 0.8742,
+      "step": 454
+    },
+    {
+      "epoch": 0.010830238979339237,
+      "grad_norm": 10.521687749457156,
+      "learning_rate": 2e-05,
+      "loss": 1.374,
+      "step": 455
+    },
+    {
+      "epoch": 0.010854041702370752,
+      "grad_norm": 2.9832718265873694,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 456
+    },
+    {
+      "epoch": 0.010877844425402267,
+      "grad_norm": 9.889531385754612,
+      "learning_rate": 2e-05,
+      "loss": 0.4296,
+      "step": 457
+    },
+    {
+      "epoch": 0.010901647148433781,
+      "grad_norm": 7.5911952047634825,
+      "learning_rate": 2e-05,
+      "loss": 0.6439,
+      "step": 458
+    },
+    {
+      "epoch": 0.010925449871465296,
+      "grad_norm": 7.053971174328579,
+      "learning_rate": 2e-05,
+      "loss": 0.6821,
+      "step": 459
+    },
+    {
+      "epoch": 0.01094925259449681,
+      "grad_norm": 2.9955880885307296,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 460
+    },
+    {
+      "epoch": 0.010973055317528325,
+      "grad_norm": 3.178885297740543,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 461
+    },
+    {
+      "epoch": 0.01099685804055984,
+      "grad_norm": 5.37364659308789,
+      "learning_rate": 2e-05,
+      "loss": 0.2475,
+      "step": 462
+    },
+    {
+      "epoch": 0.011020660763591355,
+      "grad_norm": 2.773104850974767,
+      "learning_rate": 2e-05,
+      "loss": 0.2763,
+      "step": 463
+    },
+    {
+      "epoch": 0.01104446348662287,
+      "grad_norm": 12.550733906351914,
+      "learning_rate": 2e-05,
+      "loss": 0.3867,
+      "step": 464
+    },
+    {
+      "epoch": 0.011068266209654384,
+      "grad_norm": 6.934989141603399,
+      "learning_rate": 2e-05,
+      "loss": 0.7972,
+      "step": 465
+    },
+    {
+      "epoch": 0.011092068932685899,
+      "grad_norm": 7.040659161926121,
+      "learning_rate": 2e-05,
+      "loss": 0.7563,
+      "step": 466
+    },
+    {
+      "epoch": 0.011115871655717414,
+      "grad_norm": 10.960962172607495,
+      "learning_rate": 2e-05,
+      "loss": 0.7946,
+      "step": 467
+    },
+    {
+      "epoch": 0.011139674378748929,
+      "grad_norm": 4.099795498173935,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 468
+    },
+    {
+      "epoch": 0.011163477101780443,
+      "grad_norm": 7.232044682714852,
+      "learning_rate": 2e-05,
+      "loss": 0.8832,
+      "step": 469
+    },
+    {
+      "epoch": 0.011187279824811958,
+      "grad_norm": 7.866747109645607,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 470
+    },
+    {
+      "epoch": 0.011211082547843473,
+      "grad_norm": 8.153685928649018,
+      "learning_rate": 2e-05,
+      "loss": 0.523,
+      "step": 471
+    },
+    {
+      "epoch": 0.011234885270874987,
+      "grad_norm": 2.226334676770215,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 472
+    },
+    {
+      "epoch": 0.011258687993906502,
+      "grad_norm": 7.679882527853428,
+      "learning_rate": 2e-05,
+      "loss": 0.9674,
+      "step": 473
+    },
+    {
+      "epoch": 0.011282490716938019,
+      "grad_norm": 8.592652061093208,
+      "learning_rate": 2e-05,
+      "loss": 0.7708,
+      "step": 474
+    },
+    {
+      "epoch": 0.011306293439969533,
+      "grad_norm": 2.7055805227599405,
+      "learning_rate": 2e-05,
+      "loss": 0.2016,
+      "step": 475
+    },
+    {
+      "epoch": 0.011330096163001048,
+      "grad_norm": 4.400096599874525,
+      "learning_rate": 2e-05,
+      "loss": 0.3694,
+      "step": 476
+    },
+    {
+      "epoch": 0.011353898886032563,
+      "grad_norm": 4.911370139560448,
+      "learning_rate": 2e-05,
+      "loss": 0.2672,
+      "step": 477
+    },
+    {
+      "epoch": 0.011377701609064077,
+      "grad_norm": 8.42576504400286,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 478
+    },
+    {
+      "epoch": 0.011401504332095592,
+      "grad_norm": 5.210692254440058,
+      "learning_rate": 2e-05,
+      "loss": 0.2772,
+      "step": 479
+    },
+    {
+      "epoch": 0.011425307055127107,
+      "grad_norm": 7.951534281730314,
+      "learning_rate": 2e-05,
+      "loss": 0.9505,
+      "step": 480
+    },
+    {
+      "epoch": 0.011449109778158622,
+      "grad_norm": 6.0077259590768035,
+      "learning_rate": 2e-05,
+      "loss": 0.6591,
+      "step": 481
+    },
+    {
+      "epoch": 0.011472912501190136,
+      "grad_norm": 5.328515415941751,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 482
+    },
+    {
+      "epoch": 0.011496715224221651,
+      "grad_norm": 2.0898515215246207,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 483
+    },
+    {
+      "epoch": 0.011520517947253166,
+      "grad_norm": 9.517291882588939,
+      "learning_rate": 2e-05,
+      "loss": 0.4712,
+      "step": 484
+    },
+    {
+      "epoch": 0.01154432067028468,
+      "grad_norm": 4.510684450430812,
+      "learning_rate": 2e-05,
+      "loss": 0.2972,
+      "step": 485
+    },
+    {
+      "epoch": 0.011568123393316195,
+      "grad_norm": 13.040556788775516,
+      "learning_rate": 2e-05,
+      "loss": 0.9037,
+      "step": 486
+    },
+    {
+      "epoch": 0.01159192611634771,
+      "grad_norm": 7.906999851985984,
+      "learning_rate": 2e-05,
+      "loss": 1.1621,
+      "step": 487
+    },
+    {
+      "epoch": 0.011615728839379225,
+      "grad_norm": 1.6925876490747556,
+      "learning_rate": 2e-05,
+      "loss": 0.1775,
+      "step": 488
+    },
+    {
+      "epoch": 0.01163953156241074,
+      "grad_norm": 5.196796377544965,
+      "learning_rate": 2e-05,
+      "loss": 0.3588,
+      "step": 489
+    },
+    {
+      "epoch": 0.011663334285442254,
+      "grad_norm": 8.599738279429364,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 490
+    },
+    {
+      "epoch": 0.011687137008473769,
+      "grad_norm": 2.96486988505864,
+      "learning_rate": 2e-05,
+      "loss": 0.243,
+      "step": 491
+    },
+    {
+      "epoch": 0.011710939731505284,
+      "grad_norm": 2.4686337570786865,
+      "learning_rate": 2e-05,
+      "loss": 0.1803,
+      "step": 492
+    },
+    {
+      "epoch": 0.011734742454536798,
+      "grad_norm": 5.943780288444335,
+      "learning_rate": 2e-05,
+      "loss": 0.5273,
+      "step": 493
+    },
+    {
+      "epoch": 0.011758545177568315,
+      "grad_norm": 6.099942409626483,
+      "learning_rate": 2e-05,
+      "loss": 0.7056,
+      "step": 494
+    },
+    {
+      "epoch": 0.01178234790059983,
+      "grad_norm": 8.018201330175275,
+      "learning_rate": 2e-05,
+      "loss": 0.4168,
+      "step": 495
+    },
+    {
+      "epoch": 0.011806150623631344,
+      "grad_norm": 7.86217092735597,
+      "learning_rate": 2e-05,
+      "loss": 0.1993,
+      "step": 496
+    },
+    {
+      "epoch": 0.011829953346662859,
+      "grad_norm": 15.091321431273949,
+      "learning_rate": 2e-05,
+      "loss": 0.6389,
+      "step": 497
+    },
+    {
+      "epoch": 0.011853756069694374,
+      "grad_norm": 9.848480455056205,
+      "learning_rate": 2e-05,
+      "loss": 0.8894,
+      "step": 498
+    },
+    {
+      "epoch": 0.011877558792725888,
+      "grad_norm": 8.539587202316731,
+      "learning_rate": 2e-05,
+      "loss": 0.5575,
+      "step": 499
+    },
+    {
+      "epoch": 0.011901361515757403,
+      "grad_norm": 7.68121749004079,
+      "learning_rate": 2e-05,
+      "loss": 0.6338,
+      "step": 500
+    },
+    {
+      "epoch": 0.011925164238788918,
+      "grad_norm": 5.437614858290986,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 501
+    },
+    {
+      "epoch": 0.011948966961820432,
+      "grad_norm": 7.29384646961988,
+      "learning_rate": 2e-05,
+      "loss": 0.9244,
+      "step": 502
+    },
+    {
+      "epoch": 0.011972769684851947,
+      "grad_norm": 4.102586465917825,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 503
+    },
+    {
+      "epoch": 0.011996572407883462,
+      "grad_norm": 7.263536312140448,
+      "learning_rate": 2e-05,
+      "loss": 0.8365,
+      "step": 504
+    },
+    {
+      "epoch": 0.012020375130914977,
+      "grad_norm": 7.743202012635554,
+      "learning_rate": 2e-05,
+      "loss": 0.7823,
+      "step": 505
+    },
+    {
+      "epoch": 0.012044177853946491,
+      "grad_norm": 7.602292210070401,
+      "learning_rate": 2e-05,
+      "loss": 0.2517,
+      "step": 506
+    },
+    {
+      "epoch": 0.012067980576978006,
+      "grad_norm": 11.199056029167789,
+      "learning_rate": 2e-05,
+      "loss": 1.3805,
+      "step": 507
+    },
+    {
+      "epoch": 0.01209178330000952,
+      "grad_norm": 10.463557296328192,
+      "learning_rate": 2e-05,
+      "loss": 0.4141,
+      "step": 508
+    },
+    {
+      "epoch": 0.012115586023041035,
+      "grad_norm": 2.8393077117581687,
+      "learning_rate": 2e-05,
+      "loss": 0.1965,
+      "step": 509
+    },
+    {
+      "epoch": 0.01213938874607255,
+      "grad_norm": 2.179512754810561,
+      "learning_rate": 2e-05,
+      "loss": 0.1418,
+      "step": 510
+    },
+    {
+      "epoch": 0.012163191469104065,
+      "grad_norm": 4.423388844647368,
+      "learning_rate": 2e-05,
+      "loss": 0.3349,
+      "step": 511
+    },
+    {
+      "epoch": 0.01218699419213558,
+      "grad_norm": 3.7679692438421273,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 512
+    },
+    {
+      "epoch": 0.012210796915167094,
+      "grad_norm": 11.3476232334309,
+      "learning_rate": 2e-05,
+      "loss": 0.7477,
+      "step": 513
+    },
+    {
+      "epoch": 0.012234599638198609,
+      "grad_norm": 11.908284762779251,
+      "learning_rate": 2e-05,
+      "loss": 0.5829,
+      "step": 514
+    },
+    {
+      "epoch": 0.012258402361230126,
+      "grad_norm": 11.488464413059848,
+      "learning_rate": 2e-05,
+      "loss": 0.823,
+      "step": 515
+    },
+    {
+      "epoch": 0.01228220508426164,
+      "grad_norm": 9.273179079740839,
+      "learning_rate": 2e-05,
+      "loss": 1.1041,
+      "step": 516
+    },
+    {
+      "epoch": 0.012306007807293155,
+      "grad_norm": 5.401325446958625,
+      "learning_rate": 2e-05,
+      "loss": 0.2676,
+      "step": 517
+    },
+    {
+      "epoch": 0.01232981053032467,
+      "grad_norm": 9.562556632994404,
+      "learning_rate": 2e-05,
+      "loss": 0.7441,
+      "step": 518
+    },
+    {
+      "epoch": 0.012353613253356184,
+      "grad_norm": 12.0508698771407,
+      "learning_rate": 2e-05,
+      "loss": 0.5625,
+      "step": 519
+    },
+    {
+      "epoch": 0.012377415976387699,
+      "grad_norm": 5.550354055300028,
+      "learning_rate": 2e-05,
+      "loss": 0.3338,
+      "step": 520
+    },
+    {
+      "epoch": 0.012401218699419214,
+      "grad_norm": 4.054681213519116,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 521
+    },
+    {
+      "epoch": 0.012425021422450729,
+      "grad_norm": 7.355067265230637,
+      "learning_rate": 2e-05,
+      "loss": 0.508,
+      "step": 522
+    },
+    {
+      "epoch": 0.012448824145482243,
+      "grad_norm": 4.358687271445444,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 523
+    },
+    {
+      "epoch": 0.012472626868513758,
+      "grad_norm": 5.484387502814857,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 524
+    },
+    {
+      "epoch": 0.012496429591545273,
+      "grad_norm": 3.068009649648289,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 525
+    },
+    {
+      "epoch": 0.012520232314576787,
+      "grad_norm": 7.742956345464622,
+      "learning_rate": 2e-05,
+      "loss": 1.1726,
+      "step": 526
+    },
+    {
+      "epoch": 0.012544035037608302,
+      "grad_norm": 6.633333002187697,
+      "learning_rate": 2e-05,
+      "loss": 0.6899,
+      "step": 527
+    },
+    {
+      "epoch": 0.012567837760639817,
+      "grad_norm": 2.7119848131493196,
+      "learning_rate": 2e-05,
+      "loss": 0.1871,
+      "step": 528
+    },
+    {
+      "epoch": 0.012591640483671332,
+      "grad_norm": 5.991962104435221,
+      "learning_rate": 2e-05,
+      "loss": 0.3034,
+      "step": 529
+    },
+    {
+      "epoch": 0.012615443206702846,
+      "grad_norm": 8.733802030038035,
+      "learning_rate": 2e-05,
+      "loss": 1.1839,
+      "step": 530
+    },
+    {
+      "epoch": 0.012639245929734361,
+      "grad_norm": 9.504725741128647,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 531
+    },
+    {
+      "epoch": 0.012663048652765876,
+      "grad_norm": 3.361393980650947,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 532
+    },
+    {
+      "epoch": 0.01268685137579739,
+      "grad_norm": 3.3356558660094695,
+      "learning_rate": 2e-05,
+      "loss": 0.2333,
+      "step": 533
+    },
+    {
+      "epoch": 0.012710654098828905,
+      "grad_norm": 10.672901831456574,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 534
+    },
+    {
+      "epoch": 0.012734456821860422,
+      "grad_norm": 8.036787150840134,
+      "learning_rate": 2e-05,
+      "loss": 0.8704,
+      "step": 535
+    },
+    {
+      "epoch": 0.012758259544891936,
+      "grad_norm": 5.813426937895348,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 536
+    },
+    {
+      "epoch": 0.012782062267923451,
+      "grad_norm": 8.67759603793602,
+      "learning_rate": 2e-05,
+      "loss": 1.0904,
+      "step": 537
+    },
+    {
+      "epoch": 0.012805864990954966,
+      "grad_norm": 6.430443741941927,
+      "learning_rate": 2e-05,
+      "loss": 0.797,
+      "step": 538
+    },
+    {
+      "epoch": 0.01282966771398648,
+      "grad_norm": 4.412970472075495,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 539
+    },
+    {
+      "epoch": 0.012853470437017995,
+      "grad_norm": 8.495411051899964,
+      "learning_rate": 2e-05,
+      "loss": 0.459,
+      "step": 540
+    },
+    {
+      "epoch": 0.01287727316004951,
+      "grad_norm": 9.668403052535734,
+      "learning_rate": 2e-05,
+      "loss": 1.0072,
+      "step": 541
+    },
+    {
+      "epoch": 0.012901075883081025,
+      "grad_norm": 7.710213482272582,
+      "learning_rate": 2e-05,
+      "loss": 1.2784,
+      "step": 542
+    },
+    {
+      "epoch": 0.01292487860611254,
+      "grad_norm": 2.7692935779192327,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 543
+    },
+    {
+      "epoch": 0.012948681329144054,
+      "grad_norm": 2.9792154108647817,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 544
+    },
+    {
+      "epoch": 0.012972484052175569,
+      "grad_norm": 5.06669292311311,
+      "learning_rate": 2e-05,
+      "loss": 0.6154,
+      "step": 545
+    },
+    {
+      "epoch": 0.012996286775207084,
+      "grad_norm": 3.7696093902631462,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 546
+    },
+    {
+      "epoch": 0.013020089498238598,
+      "grad_norm": 6.378314651733786,
+      "learning_rate": 2e-05,
+      "loss": 0.2781,
+      "step": 547
+    },
+    {
+      "epoch": 0.013043892221270113,
+      "grad_norm": 4.311066291190803,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 548
+    },
+    {
+      "epoch": 0.013067694944301628,
+      "grad_norm": 4.17092980545769,
+      "learning_rate": 2e-05,
+      "loss": 0.2933,
+      "step": 549
+    },
+    {
+      "epoch": 0.013091497667333142,
+      "grad_norm": 10.727568472767771,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 550
+    },
+    {
+      "epoch": 0.013115300390364657,
+      "grad_norm": 8.369930761508723,
+      "learning_rate": 2e-05,
+      "loss": 1.0685,
+      "step": 551
+    },
+    {
+      "epoch": 0.013139103113396172,
+      "grad_norm": 3.8048770045595735,
+      "learning_rate": 2e-05,
+      "loss": 0.2158,
+      "step": 552
+    },
+    {
+      "epoch": 0.013162905836427687,
+      "grad_norm": 7.975055908050074,
+      "learning_rate": 2e-05,
+      "loss": 0.4262,
+      "step": 553
+    },
+    {
+      "epoch": 0.013186708559459201,
+      "grad_norm": 4.106110163708218,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 554
+    },
+    {
+      "epoch": 0.013210511282490718,
+      "grad_norm": 14.788105438557137,
+      "learning_rate": 2e-05,
+      "loss": 1.6476,
+      "step": 555
+    },
+    {
+      "epoch": 0.013234314005522232,
+      "grad_norm": 6.6991496091695435,
+      "learning_rate": 2e-05,
+      "loss": 0.7698,
+      "step": 556
+    },
+    {
+      "epoch": 0.013258116728553747,
+      "grad_norm": 5.12251150671194,
+      "learning_rate": 2e-05,
+      "loss": 0.552,
+      "step": 557
+    },
+    {
+      "epoch": 0.013281919451585262,
+      "grad_norm": 4.416788566858104,
+      "learning_rate": 2e-05,
+      "loss": 0.2937,
+      "step": 558
+    },
+    {
+      "epoch": 0.013305722174616777,
+      "grad_norm": 3.0539638714336785,
+      "learning_rate": 2e-05,
+      "loss": 0.216,
+      "step": 559
+    },
+    {
+      "epoch": 0.013329524897648291,
+      "grad_norm": 10.265670268413924,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 560
+    },
+    {
+      "epoch": 0.013353327620679806,
+      "grad_norm": 65.39134743626677,
+      "learning_rate": 2e-05,
+      "loss": 0.8022,
+      "step": 561
+    },
+    {
+      "epoch": 0.01337713034371132,
+      "grad_norm": 8.135119013396455,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 562
+    },
+    {
+      "epoch": 0.013400933066742836,
+      "grad_norm": 6.641881253454329,
+      "learning_rate": 2e-05,
+      "loss": 1.1487,
+      "step": 563
+    },
+    {
+      "epoch": 0.01342473578977435,
+      "grad_norm": 4.577751895320257,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 564
+    },
+    {
+      "epoch": 0.013448538512805865,
+      "grad_norm": 8.390103205583534,
+      "learning_rate": 2e-05,
+      "loss": 1.0702,
+      "step": 565
+    },
+    {
+      "epoch": 0.01347234123583738,
+      "grad_norm": 7.374456227426911,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 566
+    },
+    {
+      "epoch": 0.013496143958868894,
+      "grad_norm": 3.6315499628732737,
+      "learning_rate": 2e-05,
+      "loss": 0.2572,
+      "step": 567
+    },
+    {
+      "epoch": 0.013519946681900409,
+      "grad_norm": 6.851338701327417,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 568
+    },
+    {
+      "epoch": 0.013543749404931924,
+      "grad_norm": 10.136764258741083,
+      "learning_rate": 2e-05,
+      "loss": 0.5243,
+      "step": 569
+    },
+    {
+      "epoch": 0.013567552127963439,
+      "grad_norm": 2.7082432974025226,
+      "learning_rate": 2e-05,
+      "loss": 0.2035,
+      "step": 570
+    },
+    {
+      "epoch": 0.013591354850994953,
+      "grad_norm": 9.734258555293986,
+      "learning_rate": 2e-05,
+      "loss": 1.1465,
+      "step": 571
+    },
+    {
+      "epoch": 0.013615157574026468,
+      "grad_norm": 8.46811964356974,
+      "learning_rate": 2e-05,
+      "loss": 0.7198,
+      "step": 572
+    },
+    {
+      "epoch": 0.013638960297057983,
+      "grad_norm": 3.7056785198286546,
+      "learning_rate": 2e-05,
+      "loss": 0.1581,
+      "step": 573
+    },
+    {
+      "epoch": 0.013662763020089497,
+      "grad_norm": 26.385787575843555,
+      "learning_rate": 2e-05,
+      "loss": 1.3199,
+      "step": 574
+    },
+    {
+      "epoch": 0.013686565743121014,
+      "grad_norm": 3.583666552199114,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 575
+    },
+    {
+      "epoch": 0.013710368466152529,
+      "grad_norm": 6.821520412446888,
+      "learning_rate": 2e-05,
+      "loss": 0.4402,
+      "step": 576
+    },
+    {
+      "epoch": 0.013734171189184043,
+      "grad_norm": 4.604080192365657,
+      "learning_rate": 2e-05,
+      "loss": 0.4019,
+      "step": 577
+    },
+    {
+      "epoch": 0.013757973912215558,
+      "grad_norm": 8.263324256745706,
+      "learning_rate": 2e-05,
+      "loss": 0.992,
+      "step": 578
+    },
+    {
+      "epoch": 0.013781776635247073,
+      "grad_norm": 9.079888192679425,
+      "learning_rate": 2e-05,
+      "loss": 0.7038,
+      "step": 579
+    },
+    {
+      "epoch": 0.013805579358278587,
+      "grad_norm": 8.641520351471765,
+      "learning_rate": 2e-05,
+      "loss": 0.9466,
+      "step": 580
+    },
+    {
+      "epoch": 0.013829382081310102,
+      "grad_norm": 12.258665270709049,
+      "learning_rate": 2e-05,
+      "loss": 0.3565,
+      "step": 581
+    },
+    {
+      "epoch": 0.013853184804341617,
+      "grad_norm": 3.3870466599620324,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 582
+    },
+    {
+      "epoch": 0.013876987527373132,
+      "grad_norm": 6.794165673138632,
+      "learning_rate": 2e-05,
+      "loss": 1.1073,
+      "step": 583
+    },
+    {
+      "epoch": 0.013900790250404646,
+      "grad_norm": 8.467416888658887,
+      "learning_rate": 2e-05,
+      "loss": 0.3736,
+      "step": 584
+    },
+    {
+      "epoch": 0.013924592973436161,
+      "grad_norm": 6.625307748920535,
+      "learning_rate": 2e-05,
+      "loss": 0.3455,
+      "step": 585
+    },
+    {
+      "epoch": 0.013948395696467676,
+      "grad_norm": 8.443789896999167,
+      "learning_rate": 2e-05,
+      "loss": 0.7505,
+      "step": 586
+    },
+    {
+      "epoch": 0.01397219841949919,
+      "grad_norm": 3.079292611512108,
+      "learning_rate": 2e-05,
+      "loss": 0.283,
+      "step": 587
+    },
+    {
+      "epoch": 0.013996001142530705,
+      "grad_norm": 9.64018412413552,
+      "learning_rate": 2e-05,
+      "loss": 1.0835,
+      "step": 588
+    },
+    {
+      "epoch": 0.01401980386556222,
+      "grad_norm": 4.909316801953898,
+      "learning_rate": 2e-05,
+      "loss": 0.3534,
+      "step": 589
+    },
+    {
+      "epoch": 0.014043606588593735,
+      "grad_norm": 9.193053645531482,
+      "learning_rate": 2e-05,
+      "loss": 0.512,
+      "step": 590
+    },
+    {
+      "epoch": 0.01406740931162525,
+      "grad_norm": 4.708855387172484,
+      "learning_rate": 2e-05,
+      "loss": 0.2054,
+      "step": 591
+    },
+    {
+      "epoch": 0.014091212034656764,
+      "grad_norm": 2.431484038139297,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 592
+    },
+    {
+      "epoch": 0.014115014757688279,
+      "grad_norm": 6.232880149456469,
+      "learning_rate": 2e-05,
+      "loss": 0.9272,
+      "step": 593
+    },
+    {
+      "epoch": 0.014138817480719794,
+      "grad_norm": 3.9867866758067163,
+      "learning_rate": 2e-05,
+      "loss": 0.1958,
+      "step": 594
+    },
+    {
+      "epoch": 0.01416262020375131,
+      "grad_norm": 4.063764550673414,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 595
+    },
+    {
+      "epoch": 0.014186422926782825,
+      "grad_norm": 8.078313582528382,
+      "learning_rate": 2e-05,
+      "loss": 1.0205,
+      "step": 596
+    },
+    {
+      "epoch": 0.01421022564981434,
+      "grad_norm": 3.354894658926732,
+      "learning_rate": 2e-05,
+      "loss": 0.282,
+      "step": 597
+    },
+    {
+      "epoch": 0.014234028372845854,
+      "grad_norm": 7.662254080362851,
+      "learning_rate": 2e-05,
+      "loss": 0.9892,
+      "step": 598
+    },
+    {
+      "epoch": 0.014257831095877369,
+      "grad_norm": 10.431954370300277,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 599
+    },
+    {
+      "epoch": 0.014281633818908884,
+      "grad_norm": 9.134904096195678,
+      "learning_rate": 2e-05,
+      "loss": 0.9901,
+      "step": 600
+    },
+    {
+      "epoch": 0.014305436541940398,
+      "grad_norm": 8.460547435738091,
+      "learning_rate": 2e-05,
+      "loss": 0.3389,
+      "step": 601
+    },
+    {
+      "epoch": 0.014329239264971913,
+      "grad_norm": 3.1405865929391474,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 602
+    },
+    {
+      "epoch": 0.014353041988003428,
+      "grad_norm": 8.515662822767466,
+      "learning_rate": 2e-05,
+      "loss": 0.9515,
+      "step": 603
+    },
+    {
+      "epoch": 0.014376844711034942,
+      "grad_norm": 4.0683696251784935,
+      "learning_rate": 2e-05,
+      "loss": 0.1962,
+      "step": 604
+    },
+    {
+      "epoch": 0.014400647434066457,
+      "grad_norm": 8.092930484807919,
+      "learning_rate": 2e-05,
+      "loss": 0.9795,
+      "step": 605
+    },
+    {
+      "epoch": 0.014424450157097972,
+      "grad_norm": 2.6395860549471117,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 606
+    },
+    {
+      "epoch": 0.014448252880129487,
+      "grad_norm": 3.036120789527167,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 607
+    },
+    {
+      "epoch": 0.014472055603161001,
+      "grad_norm": 8.076280390695077,
+      "learning_rate": 2e-05,
+      "loss": 0.5639,
+      "step": 608
+    },
+    {
+      "epoch": 0.014495858326192516,
+      "grad_norm": 2.9348887339185987,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 609
+    },
+    {
+      "epoch": 0.01451966104922403,
+      "grad_norm": 8.699902524282713,
+      "learning_rate": 2e-05,
+      "loss": 0.9504,
+      "step": 610
+    },
+    {
+      "epoch": 0.014543463772255545,
+      "grad_norm": 7.313228564900054,
+      "learning_rate": 2e-05,
+      "loss": 0.3114,
+      "step": 611
+    },
+    {
+      "epoch": 0.01456726649528706,
+      "grad_norm": 9.001546151640351,
+      "learning_rate": 2e-05,
+      "loss": 0.3143,
+      "step": 612
+    },
+    {
+      "epoch": 0.014591069218318575,
+      "grad_norm": 8.590654138451306,
+      "learning_rate": 2e-05,
+      "loss": 0.3168,
+      "step": 613
+    },
+    {
+      "epoch": 0.01461487194135009,
+      "grad_norm": 1.268316803675946,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 614
+    },
+    {
+      "epoch": 0.014638674664381606,
+      "grad_norm": 7.812942988213114,
+      "learning_rate": 2e-05,
+      "loss": 1.0327,
+      "step": 615
+    },
+    {
+      "epoch": 0.01466247738741312,
+      "grad_norm": 7.904588254214626,
+      "learning_rate": 2e-05,
+      "loss": 0.9508,
+      "step": 616
+    },
+    {
+      "epoch": 0.014686280110444636,
+      "grad_norm": 2.909763542726422,
+      "learning_rate": 2e-05,
+      "loss": 0.3224,
+      "step": 617
+    },
+    {
+      "epoch": 0.01471008283347615,
+      "grad_norm": 8.113895993787565,
+      "learning_rate": 2e-05,
+      "loss": 0.4431,
+      "step": 618
+    },
+    {
+      "epoch": 0.014733885556507665,
+      "grad_norm": 5.888760361163379,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 619
+    },
+    {
+      "epoch": 0.01475768827953918,
+      "grad_norm": 7.0489549540385985,
+      "learning_rate": 2e-05,
+      "loss": 0.9017,
+      "step": 620
+    },
+    {
+      "epoch": 0.014781491002570694,
+      "grad_norm": 7.299305461826953,
+      "learning_rate": 2e-05,
+      "loss": 0.9468,
+      "step": 621
+    },
+    {
+      "epoch": 0.01480529372560221,
+      "grad_norm": 2.714703077423036,
+      "learning_rate": 2e-05,
+      "loss": 0.196,
+      "step": 622
+    },
+    {
+      "epoch": 0.014829096448633724,
+      "grad_norm": 7.934174052935155,
+      "learning_rate": 2e-05,
+      "loss": 1.025,
+      "step": 623
+    },
+    {
+      "epoch": 0.014852899171665239,
+      "grad_norm": 9.535951962008497,
+      "learning_rate": 2e-05,
+      "loss": 0.8646,
+      "step": 624
+    },
+    {
+      "epoch": 0.014876701894696753,
+      "grad_norm": 3.874683566875416,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 625
+    },
+    {
+      "epoch": 0.014900504617728268,
+      "grad_norm": 10.87516042865186,
+      "learning_rate": 2e-05,
+      "loss": 1.5455,
+      "step": 626
+    },
+    {
+      "epoch": 0.014924307340759783,
+      "grad_norm": 10.126058097225643,
+      "learning_rate": 2e-05,
+      "loss": 1.0391,
+      "step": 627
+    },
+    {
+      "epoch": 0.014948110063791297,
+      "grad_norm": 6.576820605202964,
+      "learning_rate": 2e-05,
+      "loss": 0.2958,
+      "step": 628
+    },
+    {
+      "epoch": 0.014971912786822812,
+      "grad_norm": 10.516243667030357,
+      "learning_rate": 2e-05,
+      "loss": 1.262,
+      "step": 629
+    },
+    {
+      "epoch": 0.014995715509854327,
+      "grad_norm": 6.937993910864166,
+      "learning_rate": 2e-05,
+      "loss": 0.7914,
+      "step": 630
+    },
+    {
+      "epoch": 0.015019518232885842,
+      "grad_norm": 11.21644414255064,
+      "learning_rate": 2e-05,
+      "loss": 2.1563,
+      "step": 631
+    },
+    {
+      "epoch": 0.015043320955917356,
+      "grad_norm": 2.7652163778173677,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 632
+    },
+    {
+      "epoch": 0.015067123678948871,
+      "grad_norm": 8.714208089397737,
+      "learning_rate": 2e-05,
+      "loss": 0.7446,
+      "step": 633
+    },
+    {
+      "epoch": 0.015090926401980386,
+      "grad_norm": 5.589068282686739,
+      "learning_rate": 2e-05,
+      "loss": 0.7801,
+      "step": 634
+    },
+    {
+      "epoch": 0.0151147291250119,
+      "grad_norm": 5.658838621623116,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 635
+    },
+    {
+      "epoch": 0.015138531848043417,
+      "grad_norm": 4.08689814680749,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 636
+    },
+    {
+      "epoch": 0.015162334571074932,
+      "grad_norm": 3.6461557534802016,
+      "learning_rate": 2e-05,
+      "loss": 0.2886,
+      "step": 637
+    },
+    {
+      "epoch": 0.015186137294106446,
+      "grad_norm": 4.017391434843807,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 638
+    },
+    {
+      "epoch": 0.015209940017137961,
+      "grad_norm": 3.3533202740438433,
+      "learning_rate": 2e-05,
+      "loss": 0.2017,
+      "step": 639
+    },
+    {
+      "epoch": 0.015233742740169476,
+      "grad_norm": 10.089756866449882,
+      "learning_rate": 2e-05,
+      "loss": 1.3514,
+      "step": 640
+    },
+    {
+      "epoch": 0.01525754546320099,
+      "grad_norm": 10.324013871013102,
+      "learning_rate": 2e-05,
+      "loss": 0.4165,
+      "step": 641
+    },
+    {
+      "epoch": 0.015281348186232505,
+      "grad_norm": 7.550964904670765,
+      "learning_rate": 2e-05,
+      "loss": 0.2886,
+      "step": 642
+    },
+    {
+      "epoch": 0.01530515090926402,
+      "grad_norm": 2.6863207046841175,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 643
+    },
+    {
+      "epoch": 0.015328953632295535,
+      "grad_norm": 13.387518079265178,
+      "learning_rate": 2e-05,
+      "loss": 1.494,
+      "step": 644
+    },
+    {
+      "epoch": 0.01535275635532705,
+      "grad_norm": 16.36266875345884,
+      "learning_rate": 2e-05,
+      "loss": 0.6731,
+      "step": 645
+    },
+    {
+      "epoch": 0.015376559078358564,
+      "grad_norm": 9.787349737314855,
+      "learning_rate": 2e-05,
+      "loss": 1.5951,
+      "step": 646
+    },
+    {
+      "epoch": 0.015400361801390079,
+      "grad_norm": 12.589214206713962,
+      "learning_rate": 2e-05,
+      "loss": 1.2729,
+      "step": 647
+    },
+    {
+      "epoch": 0.015424164524421594,
+      "grad_norm": 8.661737784973372,
+      "learning_rate": 2e-05,
+      "loss": 0.5392,
+      "step": 648
+    },
+    {
+      "epoch": 0.015447967247453108,
+      "grad_norm": 3.5915074319797338,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 649
+    },
+    {
+      "epoch": 0.015471769970484623,
+      "grad_norm": 8.54991571907265,
+      "learning_rate": 2e-05,
+      "loss": 0.9735,
+      "step": 650
+    },
+    {
+      "epoch": 0.015495572693516138,
+      "grad_norm": 8.666063880138548,
+      "learning_rate": 2e-05,
+      "loss": 0.7867,
+      "step": 651
+    },
+    {
+      "epoch": 0.015519375416547652,
+      "grad_norm": 3.7459811948114425,
+      "learning_rate": 2e-05,
+      "loss": 0.2721,
+      "step": 652
+    },
+    {
+      "epoch": 0.015543178139579167,
+      "grad_norm": 13.049354397201263,
+      "learning_rate": 2e-05,
+      "loss": 1.4098,
+      "step": 653
+    },
+    {
+      "epoch": 0.015566980862610682,
+      "grad_norm": 4.159876615290511,
+      "learning_rate": 2e-05,
+      "loss": 0.2521,
+      "step": 654
+    },
+    {
+      "epoch": 0.015590783585642197,
+      "grad_norm": 6.650088995137545,
+      "learning_rate": 2e-05,
+      "loss": 0.9155,
+      "step": 655
+    },
+    {
+      "epoch": 0.015614586308673713,
+      "grad_norm": 7.447080763085217,
+      "learning_rate": 2e-05,
+      "loss": 0.8549,
+      "step": 656
+    },
+    {
+      "epoch": 0.015638389031705228,
+      "grad_norm": 6.532469921820161,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 657
+    },
+    {
+      "epoch": 0.01566219175473674,
+      "grad_norm": 3.899612249527908,
+      "learning_rate": 2e-05,
+      "loss": 0.1767,
+      "step": 658
+    },
+    {
+      "epoch": 0.015685994477768257,
+      "grad_norm": 3.572754925439438,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 659
+    },
+    {
+      "epoch": 0.01570979720079977,
+      "grad_norm": 8.787729256084905,
+      "learning_rate": 2e-05,
+      "loss": 0.3495,
+      "step": 660
+    },
+    {
+      "epoch": 0.015733599923831287,
+      "grad_norm": 11.645723015496026,
+      "learning_rate": 2e-05,
+      "loss": 0.3816,
+      "step": 661
+    },
+    {
+      "epoch": 0.0157574026468628,
+      "grad_norm": 3.067459677790437,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 662
+    },
+    {
+      "epoch": 0.015781205369894316,
+      "grad_norm": 4.5771399072865115,
+      "learning_rate": 2e-05,
+      "loss": 0.463,
+      "step": 663
+    },
+    {
+      "epoch": 0.01580500809292583,
+      "grad_norm": 3.8440530807027917,
+      "learning_rate": 2e-05,
+      "loss": 0.2654,
+      "step": 664
+    },
+    {
+      "epoch": 0.015828810815957346,
+      "grad_norm": 9.046816405969071,
+      "learning_rate": 2e-05,
+      "loss": 1.126,
+      "step": 665
+    },
+    {
+      "epoch": 0.015852613538988862,
+      "grad_norm": 4.003096554080491,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 666
+    },
+    {
+      "epoch": 0.015876416262020375,
+      "grad_norm": 1.4873631980678008,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 667
+    },
+    {
+      "epoch": 0.01590021898505189,
+      "grad_norm": 5.4832034581307365,
+      "learning_rate": 2e-05,
+      "loss": 0.3404,
+      "step": 668
+    },
+    {
+      "epoch": 0.015924021708083404,
+      "grad_norm": 8.73449714575008,
+      "learning_rate": 2e-05,
+      "loss": 0.2881,
+      "step": 669
+    },
+    {
+      "epoch": 0.01594782443111492,
+      "grad_norm": 3.7714691478835145,
+      "learning_rate": 2e-05,
+      "loss": 0.3413,
+      "step": 670
+    },
+    {
+      "epoch": 0.015971627154146434,
+      "grad_norm": 7.625013960971287,
+      "learning_rate": 2e-05,
+      "loss": 0.7033,
+      "step": 671
+    },
+    {
+      "epoch": 0.01599542987717795,
+      "grad_norm": 8.22743777098972,
+      "learning_rate": 2e-05,
+      "loss": 0.3429,
+      "step": 672
+    },
+    {
+      "epoch": 0.016019232600209463,
+      "grad_norm": 12.454729188609715,
+      "learning_rate": 2e-05,
+      "loss": 1.2851,
+      "step": 673
+    },
+    {
+      "epoch": 0.01604303532324098,
+      "grad_norm": 4.124780874030716,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 674
+    },
+    {
+      "epoch": 0.016066838046272493,
+      "grad_norm": 6.416718430738503,
+      "learning_rate": 2e-05,
+      "loss": 0.3115,
+      "step": 675
+    },
+    {
+      "epoch": 0.01609064076930401,
+      "grad_norm": 5.237385128118476,
+      "learning_rate": 2e-05,
+      "loss": 0.4118,
+      "step": 676
+    },
+    {
+      "epoch": 0.016114443492335522,
+      "grad_norm": 8.944432096512152,
+      "learning_rate": 2e-05,
+      "loss": 0.341,
+      "step": 677
+    },
+    {
+      "epoch": 0.01613824621536704,
+      "grad_norm": 4.193903465075456,
+      "learning_rate": 2e-05,
+      "loss": 0.1826,
+      "step": 678
+    },
+    {
+      "epoch": 0.01616204893839855,
+      "grad_norm": 5.384138263009054,
+      "learning_rate": 2e-05,
+      "loss": 0.7135,
+      "step": 679
+    },
+    {
+      "epoch": 0.016185851661430068,
+      "grad_norm": 6.991351294700134,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 680
+    },
+    {
+      "epoch": 0.01620965438446158,
+      "grad_norm": 9.764148101884098,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 681
+    },
+    {
+      "epoch": 0.016233457107493098,
+      "grad_norm": 7.483579394881374,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 682
+    },
+    {
+      "epoch": 0.01625725983052461,
+      "grad_norm": 7.497635173427665,
+      "learning_rate": 2e-05,
+      "loss": 1.2904,
+      "step": 683
+    },
+    {
+      "epoch": 0.016281062553556127,
+      "grad_norm": 5.8552355258062185,
+      "learning_rate": 2e-05,
+      "loss": 0.8066,
+      "step": 684
+    },
+    {
+      "epoch": 0.01630486527658764,
+      "grad_norm": 9.799784493586237,
+      "learning_rate": 2e-05,
+      "loss": 0.3848,
+      "step": 685
+    },
+    {
+      "epoch": 0.016328667999619156,
+      "grad_norm": 9.674549196666426,
+      "learning_rate": 2e-05,
+      "loss": 1.0039,
+      "step": 686
+    },
+    {
+      "epoch": 0.016352470722650673,
+      "grad_norm": 18.333998853422887,
+      "learning_rate": 2e-05,
+      "loss": 1.0626,
+      "step": 687
+    },
+    {
+      "epoch": 0.016376273445682186,
+      "grad_norm": 8.084414352511414,
+      "learning_rate": 2e-05,
+      "loss": 1.1766,
+      "step": 688
+    },
+    {
+      "epoch": 0.016400076168713702,
+      "grad_norm": 8.699938429401312,
+      "learning_rate": 2e-05,
+      "loss": 0.5453,
+      "step": 689
+    },
+    {
+      "epoch": 0.016423878891745215,
+      "grad_norm": 3.290401331808441,
+      "learning_rate": 2e-05,
+      "loss": 0.285,
+      "step": 690
+    },
+    {
+      "epoch": 0.01644768161477673,
+      "grad_norm": 10.608541705362052,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 691
+    },
+    {
+      "epoch": 0.016471484337808245,
+      "grad_norm": 10.544700295428507,
+      "learning_rate": 2e-05,
+      "loss": 1.51,
+      "step": 692
+    },
+    {
+      "epoch": 0.01649528706083976,
+      "grad_norm": 2.0687498404966367,
+      "learning_rate": 2e-05,
+      "loss": 0.1806,
+      "step": 693
+    },
+    {
+      "epoch": 0.016519089783871274,
+      "grad_norm": 5.617800431628024,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 694
+    },
+    {
+      "epoch": 0.01654289250690279,
+      "grad_norm": 8.70164205670232,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 695
+    },
+    {
+      "epoch": 0.016566695229934304,
+      "grad_norm": 3.178786731328533,
+      "learning_rate": 2e-05,
+      "loss": 0.2942,
+      "step": 696
+    },
+    {
+      "epoch": 0.01659049795296582,
+      "grad_norm": 6.582398468876956,
+      "learning_rate": 2e-05,
+      "loss": 0.3762,
+      "step": 697
+    },
+    {
+      "epoch": 0.016614300675997333,
+      "grad_norm": 3.8345205817443695,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 698
+    },
+    {
+      "epoch": 0.01663810339902885,
+      "grad_norm": 11.370472017948982,
+      "learning_rate": 2e-05,
+      "loss": 1.2461,
+      "step": 699
+    },
+    {
+      "epoch": 0.016661906122060362,
+      "grad_norm": 7.311174586651351,
+      "learning_rate": 2e-05,
+      "loss": 0.6672,
+      "step": 700
+    },
+    {
+      "epoch": 0.01668570884509188,
+      "grad_norm": 2.688866681582633,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 701
+    },
+    {
+      "epoch": 0.016709511568123392,
+      "grad_norm": 6.92349695496584,
+      "learning_rate": 2e-05,
+      "loss": 0.572,
+      "step": 702
+    },
+    {
+      "epoch": 0.01673331429115491,
+      "grad_norm": 4.018664217035689,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 703
+    },
+    {
+      "epoch": 0.01675711701418642,
+      "grad_norm": 12.34246793725986,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 704
+    },
+    {
+      "epoch": 0.016780919737217938,
+      "grad_norm": 2.906754583060542,
+      "learning_rate": 2e-05,
+      "loss": 0.1996,
+      "step": 705
+    },
+    {
+      "epoch": 0.016804722460249454,
+      "grad_norm": 6.9013127769491085,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 706
+    },
+    {
+      "epoch": 0.016828525183280967,
+      "grad_norm": 3.2891240101211148,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 707
+    },
+    {
+      "epoch": 0.016852327906312484,
+      "grad_norm": 6.139104686649292,
+      "learning_rate": 2e-05,
+      "loss": 0.7948,
+      "step": 708
+    },
+    {
+      "epoch": 0.016876130629343997,
+      "grad_norm": 17.16629570003928,
+      "learning_rate": 2e-05,
+      "loss": 1.2092,
+      "step": 709
+    },
+    {
+      "epoch": 0.016899933352375513,
+      "grad_norm": 9.437415398537182,
+      "learning_rate": 2e-05,
+      "loss": 0.7638,
+      "step": 710
+    },
+    {
+      "epoch": 0.016923736075407026,
+      "grad_norm": 9.159815106276504,
+      "learning_rate": 2e-05,
+      "loss": 0.5682,
+      "step": 711
+    },
+    {
+      "epoch": 0.016947538798438543,
+      "grad_norm": 2.5100108805811234,
+      "learning_rate": 2e-05,
+      "loss": 0.2692,
+      "step": 712
+    },
+    {
+      "epoch": 0.016971341521470056,
+      "grad_norm": 12.227411712908062,
+      "learning_rate": 2e-05,
+      "loss": 0.735,
+      "step": 713
+    },
+    {
+      "epoch": 0.016995144244501572,
+      "grad_norm": 10.899611000544592,
+      "learning_rate": 2e-05,
+      "loss": 0.3273,
+      "step": 714
+    },
+    {
+      "epoch": 0.017018946967533085,
+      "grad_norm": 2.6695198460556235,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 715
+    },
+    {
+      "epoch": 0.0170427496905646,
+      "grad_norm": 7.269572564851064,
+      "learning_rate": 2e-05,
+      "loss": 0.2885,
+      "step": 716
+    },
+    {
+      "epoch": 0.017066552413596114,
+      "grad_norm": 9.548344591067718,
+      "learning_rate": 2e-05,
+      "loss": 0.5185,
+      "step": 717
+    },
+    {
+      "epoch": 0.01709035513662763,
+      "grad_norm": 6.964474221012495,
+      "learning_rate": 2e-05,
+      "loss": 0.8404,
+      "step": 718
+    },
+    {
+      "epoch": 0.017114157859659144,
+      "grad_norm": 8.823704541786341,
+      "learning_rate": 2e-05,
+      "loss": 0.6906,
+      "step": 719
+    },
+    {
+      "epoch": 0.01713796058269066,
+      "grad_norm": 10.568863730546,
+      "learning_rate": 2e-05,
+      "loss": 0.9843,
+      "step": 720
+    },
+    {
+      "epoch": 0.017161763305722173,
+      "grad_norm": 6.871187742004737,
+      "learning_rate": 2e-05,
+      "loss": 1.0058,
+      "step": 721
+    },
+    {
+      "epoch": 0.01718556602875369,
+      "grad_norm": 8.016409756860169,
+      "learning_rate": 2e-05,
+      "loss": 0.4544,
+      "step": 722
+    },
+    {
+      "epoch": 0.017209368751785203,
+      "grad_norm": 5.893334154489839,
+      "learning_rate": 2e-05,
+      "loss": 0.5556,
+      "step": 723
+    },
+    {
+      "epoch": 0.01723317147481672,
+      "grad_norm": 3.1754150430182144,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 724
+    },
+    {
+      "epoch": 0.017256974197848232,
+      "grad_norm": 3.177983570616777,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 725
+    },
+    {
+      "epoch": 0.01728077692087975,
+      "grad_norm": 5.100389150183543,
+      "learning_rate": 2e-05,
+      "loss": 0.1809,
+      "step": 726
+    },
+    {
+      "epoch": 0.017304579643911265,
+      "grad_norm": 12.70552156529217,
+      "learning_rate": 2e-05,
+      "loss": 1.3693,
+      "step": 727
+    },
+    {
+      "epoch": 0.017328382366942778,
+      "grad_norm": 5.52647846959017,
+      "learning_rate": 2e-05,
+      "loss": 0.4177,
+      "step": 728
+    },
+    {
+      "epoch": 0.017352185089974295,
+      "grad_norm": 8.118381005654854,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 729
+    },
+    {
+      "epoch": 0.017375987813005807,
+      "grad_norm": 6.59686484106401,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 730
+    },
+    {
+      "epoch": 0.017399790536037324,
+      "grad_norm": 4.968465349915094,
+      "learning_rate": 2e-05,
+      "loss": 0.431,
+      "step": 731
+    },
+    {
+      "epoch": 0.017423593259068837,
+      "grad_norm": 6.484830110109145,
+      "learning_rate": 2e-05,
+      "loss": 0.3063,
+      "step": 732
+    },
+    {
+      "epoch": 0.017447395982100353,
+      "grad_norm": 6.24705337900673,
+      "learning_rate": 2e-05,
+      "loss": 0.6375,
+      "step": 733
+    },
+    {
+      "epoch": 0.017471198705131866,
+      "grad_norm": 3.172999954383029,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 734
+    },
+    {
+      "epoch": 0.017495001428163383,
+      "grad_norm": 7.830240068176706,
+      "learning_rate": 2e-05,
+      "loss": 0.7585,
+      "step": 735
+    },
+    {
+      "epoch": 0.017518804151194896,
+      "grad_norm": 6.657802102628063,
+      "learning_rate": 2e-05,
+      "loss": 0.487,
+      "step": 736
+    },
+    {
+      "epoch": 0.017542606874226412,
+      "grad_norm": 5.7988807625819945,
+      "learning_rate": 2e-05,
+      "loss": 0.3469,
+      "step": 737
+    },
+    {
+      "epoch": 0.017566409597257925,
+      "grad_norm": 5.233723596422366,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 738
+    },
+    {
+      "epoch": 0.01759021232028944,
+      "grad_norm": 6.996881177328479,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 739
+    },
+    {
+      "epoch": 0.017614015043320955,
+      "grad_norm": 9.914668675331997,
+      "learning_rate": 2e-05,
+      "loss": 0.9764,
+      "step": 740
+    },
+    {
+      "epoch": 0.01763781776635247,
+      "grad_norm": 8.38792420260447,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 741
+    },
+    {
+      "epoch": 0.017661620489383984,
+      "grad_norm": 12.436453836620743,
+      "learning_rate": 2e-05,
+      "loss": 0.4995,
+      "step": 742
+    },
+    {
+      "epoch": 0.0176854232124155,
+      "grad_norm": 18.395613535106634,
+      "learning_rate": 2e-05,
+      "loss": 0.8051,
+      "step": 743
+    },
+    {
+      "epoch": 0.017709225935447014,
+      "grad_norm": 10.747668165781306,
+      "learning_rate": 2e-05,
+      "loss": 0.8405,
+      "step": 744
+    },
+    {
+      "epoch": 0.01773302865847853,
+      "grad_norm": 10.225553298431617,
+      "learning_rate": 2e-05,
+      "loss": 0.839,
+      "step": 745
+    },
+    {
+      "epoch": 0.017756831381510046,
+      "grad_norm": 10.689208001622573,
+      "learning_rate": 2e-05,
+      "loss": 1.0882,
+      "step": 746
+    },
+    {
+      "epoch": 0.01778063410454156,
+      "grad_norm": 5.901453093063511,
+      "learning_rate": 2e-05,
+      "loss": 0.6057,
+      "step": 747
+    },
+    {
+      "epoch": 0.017804436827573076,
+      "grad_norm": 13.372725566126386,
+      "learning_rate": 2e-05,
+      "loss": 0.6784,
+      "step": 748
+    },
+    {
+      "epoch": 0.01782823955060459,
+      "grad_norm": 3.9499548678163316,
+      "learning_rate": 2e-05,
+      "loss": 0.326,
+      "step": 749
+    },
+    {
+      "epoch": 0.017852042273636105,
+      "grad_norm": 6.297022485695715,
+      "learning_rate": 2e-05,
+      "loss": 0.6809,
+      "step": 750
+    },
+    {
+      "epoch": 0.01787584499666762,
+      "grad_norm": 11.246506999965728,
+      "learning_rate": 2e-05,
+      "loss": 1.5063,
+      "step": 751
+    },
+    {
+      "epoch": 0.017899647719699135,
+      "grad_norm": 3.570317649863944,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 752
+    },
+    {
+      "epoch": 0.017923450442730648,
+      "grad_norm": 3.2785259539570593,
+      "learning_rate": 2e-05,
+      "loss": 0.2183,
+      "step": 753
+    },
+    {
+      "epoch": 0.017947253165762164,
+      "grad_norm": 8.551367822473244,
+      "learning_rate": 2e-05,
+      "loss": 1.0034,
+      "step": 754
+    },
+    {
+      "epoch": 0.017971055888793677,
+      "grad_norm": 4.5485187399742,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 755
+    },
+    {
+      "epoch": 0.017994858611825194,
+      "grad_norm": 9.801912498094222,
+      "learning_rate": 2e-05,
+      "loss": 1.2528,
+      "step": 756
+    },
+    {
+      "epoch": 0.018018661334856707,
+      "grad_norm": 2.8795921627795917,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 757
+    },
+    {
+      "epoch": 0.018042464057888223,
+      "grad_norm": 9.90289298367989,
+      "learning_rate": 2e-05,
+      "loss": 0.4543,
+      "step": 758
+    },
+    {
+      "epoch": 0.018066266780919736,
+      "grad_norm": 6.092384303904405,
+      "learning_rate": 2e-05,
+      "loss": 0.255,
+      "step": 759
+    },
+    {
+      "epoch": 0.018090069503951253,
+      "grad_norm": 8.897372733336733,
+      "learning_rate": 2e-05,
+      "loss": 0.8114,
+      "step": 760
+    },
+    {
+      "epoch": 0.018113872226982766,
+      "grad_norm": 8.706659651567415,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 761
+    },
+    {
+      "epoch": 0.018137674950014282,
+      "grad_norm": 5.709125641478857,
+      "learning_rate": 2e-05,
+      "loss": 0.3893,
+      "step": 762
+    },
+    {
+      "epoch": 0.018161477673045795,
+      "grad_norm": 2.075832473585743,
+      "learning_rate": 2e-05,
+      "loss": 0.174,
+      "step": 763
+    },
+    {
+      "epoch": 0.01818528039607731,
+      "grad_norm": 4.06013444240496,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 764
+    },
+    {
+      "epoch": 0.018209083119108824,
+      "grad_norm": 7.235700676472994,
+      "learning_rate": 2e-05,
+      "loss": 0.3121,
+      "step": 765
+    },
+    {
+      "epoch": 0.01823288584214034,
+      "grad_norm": 2.7800570039044095,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 766
+    },
+    {
+      "epoch": 0.018256688565171857,
+      "grad_norm": 8.077534865454012,
+      "learning_rate": 2e-05,
+      "loss": 0.4046,
+      "step": 767
+    },
+    {
+      "epoch": 0.01828049128820337,
+      "grad_norm": 3.607981077696072,
+      "learning_rate": 2e-05,
+      "loss": 0.1999,
+      "step": 768
+    },
+    {
+      "epoch": 0.018304294011234887,
+      "grad_norm": 4.593857979312275,
+      "learning_rate": 2e-05,
+      "loss": 0.2301,
+      "step": 769
+    },
+    {
+      "epoch": 0.0183280967342664,
+      "grad_norm": 8.957905136639583,
+      "learning_rate": 2e-05,
+      "loss": 0.3368,
+      "step": 770
+    },
+    {
+      "epoch": 0.018351899457297916,
+      "grad_norm": 7.1152345335961416,
+      "learning_rate": 2e-05,
+      "loss": 0.2268,
+      "step": 771
+    },
+    {
+      "epoch": 0.01837570218032943,
+      "grad_norm": 1.6590982509683312,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 772
+    },
+    {
+      "epoch": 0.018399504903360946,
+      "grad_norm": 3.5539867329900514,
+      "learning_rate": 2e-05,
+      "loss": 0.3057,
+      "step": 773
+    },
+    {
+      "epoch": 0.01842330762639246,
+      "grad_norm": 5.020051783614432,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 774
+    },
+    {
+      "epoch": 0.018447110349423975,
+      "grad_norm": 4.815984173738828,
+      "learning_rate": 2e-05,
+      "loss": 0.3668,
+      "step": 775
+    },
+    {
+      "epoch": 0.018470913072455488,
+      "grad_norm": 4.491808744193764,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 776
+    },
+    {
+      "epoch": 0.018494715795487005,
+      "grad_norm": 6.159721278602747,
+      "learning_rate": 2e-05,
+      "loss": 0.802,
+      "step": 777
+    },
+    {
+      "epoch": 0.018518518518518517,
+      "grad_norm": 6.463525500028246,
+      "learning_rate": 2e-05,
+      "loss": 0.5695,
+      "step": 778
+    },
+    {
+      "epoch": 0.018542321241550034,
+      "grad_norm": 8.014987832641419,
+      "learning_rate": 2e-05,
+      "loss": 0.8027,
+      "step": 779
+    },
+    {
+      "epoch": 0.018566123964581547,
+      "grad_norm": 4.264305585771582,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 780
+    },
+    {
+      "epoch": 0.018589926687613063,
+      "grad_norm": 11.388575950728676,
+      "learning_rate": 2e-05,
+      "loss": 1.2716,
+      "step": 781
+    },
+    {
+      "epoch": 0.018613729410644576,
+      "grad_norm": 7.451447413070841,
+      "learning_rate": 2e-05,
+      "loss": 0.9007,
+      "step": 782
+    },
+    {
+      "epoch": 0.018637532133676093,
+      "grad_norm": 3.6098541919132994,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 783
+    },
+    {
+      "epoch": 0.018661334856707606,
+      "grad_norm": 4.166754728186654,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 784
+    },
+    {
+      "epoch": 0.018685137579739122,
+      "grad_norm": 7.357861049032737,
+      "learning_rate": 2e-05,
+      "loss": 1.1009,
+      "step": 785
+    },
+    {
+      "epoch": 0.018708940302770635,
+      "grad_norm": 6.392856907135597,
+      "learning_rate": 2e-05,
+      "loss": 0.2814,
+      "step": 786
+    },
+    {
+      "epoch": 0.01873274302580215,
+      "grad_norm": 2.904000388739125,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 787
+    },
+    {
+      "epoch": 0.018756545748833668,
+      "grad_norm": 8.176065901086,
+      "learning_rate": 2e-05,
+      "loss": 0.9694,
+      "step": 788
+    },
+    {
+      "epoch": 0.01878034847186518,
+      "grad_norm": 11.018053991889994,
+      "learning_rate": 2e-05,
+      "loss": 0.4746,
+      "step": 789
+    },
+    {
+      "epoch": 0.018804151194896698,
+      "grad_norm": 7.452868558370451,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 790
+    },
+    {
+      "epoch": 0.01882795391792821,
+      "grad_norm": 6.728027186764491,
+      "learning_rate": 2e-05,
+      "loss": 0.2408,
+      "step": 791
+    },
+    {
+      "epoch": 0.018851756640959727,
+      "grad_norm": 5.172264242130849,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 792
+    },
+    {
+      "epoch": 0.01887555936399124,
+      "grad_norm": 4.075121236935239,
+      "learning_rate": 2e-05,
+      "loss": 0.2825,
+      "step": 793
+    },
+    {
+      "epoch": 0.018899362087022756,
+      "grad_norm": 9.36425643835262,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 794
+    },
+    {
+      "epoch": 0.01892316481005427,
+      "grad_norm": 7.538213088720499,
+      "learning_rate": 2e-05,
+      "loss": 1.3089,
+      "step": 795
+    },
+    {
+      "epoch": 0.018946967533085786,
+      "grad_norm": 10.437463701153762,
+      "learning_rate": 2e-05,
+      "loss": 1.4604,
+      "step": 796
+    },
+    {
+      "epoch": 0.0189707702561173,
+      "grad_norm": 6.041269654072644,
+      "learning_rate": 2e-05,
+      "loss": 0.9132,
+      "step": 797
+    },
+    {
+      "epoch": 0.018994572979148815,
+      "grad_norm": 8.348114315969992,
+      "learning_rate": 2e-05,
+      "loss": 0.8436,
+      "step": 798
+    },
+    {
+      "epoch": 0.01901837570218033,
+      "grad_norm": 8.538010038722772,
+      "learning_rate": 2e-05,
+      "loss": 0.4884,
+      "step": 799
+    },
+    {
+      "epoch": 0.019042178425211845,
+      "grad_norm": 6.893931755791519,
+      "learning_rate": 2e-05,
+      "loss": 0.7036,
+      "step": 800
+    },
+    {
+      "epoch": 0.019065981148243358,
+      "grad_norm": 6.54026475891498,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 801
+    },
+    {
+      "epoch": 0.019089783871274874,
+      "grad_norm": 3.566164558790727,
+      "learning_rate": 2e-05,
+      "loss": 0.2704,
+      "step": 802
+    },
+    {
+      "epoch": 0.019113586594306387,
+      "grad_norm": 2.836582183525883,
+      "learning_rate": 2e-05,
+      "loss": 0.2029,
+      "step": 803
+    },
+    {
+      "epoch": 0.019137389317337904,
+      "grad_norm": 5.065569223068341,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 804
+    },
+    {
+      "epoch": 0.019161192040369417,
+      "grad_norm": 8.37426052871985,
+      "learning_rate": 2e-05,
+      "loss": 0.5389,
+      "step": 805
+    },
+    {
+      "epoch": 0.019184994763400933,
+      "grad_norm": 5.4175090945734246,
+      "learning_rate": 2e-05,
+      "loss": 0.2943,
+      "step": 806
+    },
+    {
+      "epoch": 0.01920879748643245,
+      "grad_norm": 3.349861305481609,
+      "learning_rate": 2e-05,
+      "loss": 0.3047,
+      "step": 807
+    },
+    {
+      "epoch": 0.019232600209463963,
+      "grad_norm": 6.770770147735443,
+      "learning_rate": 2e-05,
+      "loss": 0.585,
+      "step": 808
+    },
+    {
+      "epoch": 0.01925640293249548,
+      "grad_norm": 6.275013209174291,
+      "learning_rate": 2e-05,
+      "loss": 1.0692,
+      "step": 809
+    },
+    {
+      "epoch": 0.019280205655526992,
+      "grad_norm": 6.064473447736023,
+      "learning_rate": 2e-05,
+      "loss": 0.6322,
+      "step": 810
+    },
+    {
+      "epoch": 0.01930400837855851,
+      "grad_norm": 5.245130954940311,
+      "learning_rate": 2e-05,
+      "loss": 0.5622,
+      "step": 811
+    },
+    {
+      "epoch": 0.01932781110159002,
+      "grad_norm": 6.532093594692872,
+      "learning_rate": 2e-05,
+      "loss": 0.4804,
+      "step": 812
+    },
+    {
+      "epoch": 0.019351613824621538,
+      "grad_norm": 7.996474503470674,
+      "learning_rate": 2e-05,
+      "loss": 0.3518,
+      "step": 813
+    },
+    {
+      "epoch": 0.01937541654765305,
+      "grad_norm": 8.204765104338438,
+      "learning_rate": 2e-05,
+      "loss": 0.7514,
+      "step": 814
+    },
+    {
+      "epoch": 0.019399219270684567,
+      "grad_norm": 7.77965408583579,
+      "learning_rate": 2e-05,
+      "loss": 0.4104,
+      "step": 815
+    },
+    {
+      "epoch": 0.01942302199371608,
+      "grad_norm": 5.563073720943101,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 816
+    },
+    {
+      "epoch": 0.019446824716747597,
+      "grad_norm": 4.053974569317071,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 817
+    },
+    {
+      "epoch": 0.01947062743977911,
+      "grad_norm": 4.456756212072241,
+      "learning_rate": 2e-05,
+      "loss": 0.2031,
+      "step": 818
+    },
+    {
+      "epoch": 0.019494430162810626,
+      "grad_norm": 6.686672550481364,
+      "learning_rate": 2e-05,
+      "loss": 0.8548,
+      "step": 819
+    },
+    {
+      "epoch": 0.01951823288584214,
+      "grad_norm": 5.413804345540516,
+      "learning_rate": 2e-05,
+      "loss": 0.5328,
+      "step": 820
+    },
+    {
+      "epoch": 0.019542035608873656,
+      "grad_norm": 6.991317553388221,
+      "learning_rate": 2e-05,
+      "loss": 0.9046,
+      "step": 821
+    },
+    {
+      "epoch": 0.01956583833190517,
+      "grad_norm": 7.233603233344829,
+      "learning_rate": 2e-05,
+      "loss": 0.2967,
+      "step": 822
+    },
+    {
+      "epoch": 0.019589641054936685,
+      "grad_norm": 4.00708677933644,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 823
+    },
+    {
+      "epoch": 0.019613443777968198,
+      "grad_norm": 4.88873093209973,
+      "learning_rate": 2e-05,
+      "loss": 0.3905,
+      "step": 824
+    },
+    {
+      "epoch": 0.019637246500999714,
+      "grad_norm": 3.5797751789603445,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 825
+    },
+    {
+      "epoch": 0.019661049224031227,
+      "grad_norm": 9.420756820648442,
+      "learning_rate": 2e-05,
+      "loss": 0.391,
+      "step": 826
+    },
+    {
+      "epoch": 0.019684851947062744,
+      "grad_norm": 11.687822104692069,
+      "learning_rate": 2e-05,
+      "loss": 0.9761,
+      "step": 827
+    },
+    {
+      "epoch": 0.01970865467009426,
+      "grad_norm": 8.017925478066854,
+      "learning_rate": 2e-05,
+      "loss": 0.6892,
+      "step": 828
+    },
+    {
+      "epoch": 0.019732457393125773,
+      "grad_norm": 9.424103203363746,
+      "learning_rate": 2e-05,
+      "loss": 1.1894,
+      "step": 829
+    },
+    {
+      "epoch": 0.01975626011615729,
+      "grad_norm": 4.472973477801294,
+      "learning_rate": 2e-05,
+      "loss": 0.2137,
+      "step": 830
+    },
+    {
+      "epoch": 0.019780062839188803,
+      "grad_norm": 4.754528742559234,
+      "learning_rate": 2e-05,
+      "loss": 0.2939,
+      "step": 831
+    },
+    {
+      "epoch": 0.01980386556222032,
+      "grad_norm": 3.5635633266974116,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 832
+    },
+    {
+      "epoch": 0.019827668285251832,
+      "grad_norm": 5.776879286806432,
+      "learning_rate": 2e-05,
+      "loss": 0.3692,
+      "step": 833
+    },
+    {
+      "epoch": 0.01985147100828335,
+      "grad_norm": 8.54486146337683,
+      "learning_rate": 2e-05,
+      "loss": 0.7565,
+      "step": 834
+    },
+    {
+      "epoch": 0.01987527373131486,
+      "grad_norm": 3.9682263445514154,
+      "learning_rate": 2e-05,
+      "loss": 0.3548,
+      "step": 835
+    },
+    {
+      "epoch": 0.019899076454346378,
+      "grad_norm": 4.307373100908035,
+      "learning_rate": 2e-05,
+      "loss": 0.5071,
+      "step": 836
+    },
+    {
+      "epoch": 0.01992287917737789,
+      "grad_norm": 6.299745237965031,
+      "learning_rate": 2e-05,
+      "loss": 0.6223,
+      "step": 837
+    },
+    {
+      "epoch": 0.019946681900409408,
+      "grad_norm": 3.070464894061228,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 838
+    },
+    {
+      "epoch": 0.01997048462344092,
+      "grad_norm": 3.165124019583081,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 839
+    },
+    {
+      "epoch": 0.019994287346472437,
+      "grad_norm": 7.629102985077089,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 840
+    },
+    {
+      "epoch": 0.02001809006950395,
+      "grad_norm": 4.809939576246539,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 841
+    },
+    {
+      "epoch": 0.020041892792535466,
+      "grad_norm": 11.92117186766356,
+      "learning_rate": 2e-05,
+      "loss": 1.3879,
+      "step": 842
+    },
+    {
+      "epoch": 0.02006569551556698,
+      "grad_norm": 4.745811891575969,
+      "learning_rate": 2e-05,
+      "loss": 0.2933,
+      "step": 843
+    },
+    {
+      "epoch": 0.020089498238598496,
+      "grad_norm": 6.158264225031314,
+      "learning_rate": 2e-05,
+      "loss": 0.2752,
+      "step": 844
+    },
+    {
+      "epoch": 0.02011330096163001,
+      "grad_norm": 6.04142389683763,
+      "learning_rate": 2e-05,
+      "loss": 0.8394,
+      "step": 845
+    },
+    {
+      "epoch": 0.020137103684661525,
+      "grad_norm": 3.0431773011423298,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 846
+    },
+    {
+      "epoch": 0.020160906407693042,
+      "grad_norm": 2.5134069143278386,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 847
+    },
+    {
+      "epoch": 0.020184709130724555,
+      "grad_norm": 8.256168922905962,
+      "learning_rate": 2e-05,
+      "loss": 1.048,
+      "step": 848
+    },
+    {
+      "epoch": 0.02020851185375607,
+      "grad_norm": 9.174579396433074,
+      "learning_rate": 2e-05,
+      "loss": 0.4926,
+      "step": 849
+    },
+    {
+      "epoch": 0.020232314576787584,
+      "grad_norm": 10.205717029207308,
+      "learning_rate": 2e-05,
+      "loss": 1.4577,
+      "step": 850
+    },
+    {
+      "epoch": 0.0202561172998191,
+      "grad_norm": 3.5554118258110488,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 851
+    },
+    {
+      "epoch": 0.020279920022850614,
+      "grad_norm": 4.5726315976511325,
+      "learning_rate": 2e-05,
+      "loss": 0.2909,
+      "step": 852
+    },
+    {
+      "epoch": 0.02030372274588213,
+      "grad_norm": 6.207202249005705,
+      "learning_rate": 2e-05,
+      "loss": 0.6916,
+      "step": 853
+    },
+    {
+      "epoch": 0.020327525468913643,
+      "grad_norm": 6.702888509384683,
+      "learning_rate": 2e-05,
+      "loss": 0.6409,
+      "step": 854
+    },
+    {
+      "epoch": 0.02035132819194516,
+      "grad_norm": 10.571645537905008,
+      "learning_rate": 2e-05,
+      "loss": 0.9562,
+      "step": 855
+    },
+    {
+      "epoch": 0.020375130914976673,
+      "grad_norm": 6.1492912812180265,
+      "learning_rate": 2e-05,
+      "loss": 0.3255,
+      "step": 856
+    },
+    {
+      "epoch": 0.02039893363800819,
+      "grad_norm": 5.307821460935167,
+      "learning_rate": 2e-05,
+      "loss": 0.3054,
+      "step": 857
+    },
+    {
+      "epoch": 0.020422736361039702,
+      "grad_norm": 3.291049387934895,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 858
+    },
+    {
+      "epoch": 0.02044653908407122,
+      "grad_norm": 6.521319057870931,
+      "learning_rate": 2e-05,
+      "loss": 0.311,
+      "step": 859
+    },
+    {
+      "epoch": 0.02047034180710273,
+      "grad_norm": 4.009149008018147,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 860
+    },
+    {
+      "epoch": 0.020494144530134248,
+      "grad_norm": 5.482127085548258,
+      "learning_rate": 2e-05,
+      "loss": 0.478,
+      "step": 861
+    },
+    {
+      "epoch": 0.02051794725316576,
+      "grad_norm": 7.714411891262548,
+      "learning_rate": 2e-05,
+      "loss": 0.4295,
+      "step": 862
+    },
+    {
+      "epoch": 0.020541749976197277,
+      "grad_norm": 3.4873265387131474,
+      "learning_rate": 2e-05,
+      "loss": 0.253,
+      "step": 863
+    },
+    {
+      "epoch": 0.02056555269922879,
+      "grad_norm": 10.336366810109686,
+      "learning_rate": 2e-05,
+      "loss": 0.8086,
+      "step": 864
+    },
+    {
+      "epoch": 0.020589355422260307,
+      "grad_norm": 19.777432587920657,
+      "learning_rate": 2e-05,
+      "loss": 1.3327,
+      "step": 865
+    },
+    {
+      "epoch": 0.02061315814529182,
+      "grad_norm": 12.616110425073922,
+      "learning_rate": 2e-05,
+      "loss": 1.0819,
+      "step": 866
+    },
+    {
+      "epoch": 0.020636960868323336,
+      "grad_norm": 6.655895030013199,
+      "learning_rate": 2e-05,
+      "loss": 0.358,
+      "step": 867
+    },
+    {
+      "epoch": 0.020660763591354853,
+      "grad_norm": 9.697176722584125,
+      "learning_rate": 2e-05,
+      "loss": 0.7677,
+      "step": 868
+    },
+    {
+      "epoch": 0.020684566314386366,
+      "grad_norm": 10.578002001933065,
+      "learning_rate": 2e-05,
+      "loss": 1.4596,
+      "step": 869
+    },
+    {
+      "epoch": 0.020708369037417882,
+      "grad_norm": 9.66151661080292,
+      "learning_rate": 2e-05,
+      "loss": 0.9944,
+      "step": 870
+    },
+    {
+      "epoch": 0.020732171760449395,
+      "grad_norm": 5.618598572736618,
+      "learning_rate": 2e-05,
+      "loss": 0.4029,
+      "step": 871
+    },
+    {
+      "epoch": 0.02075597448348091,
+      "grad_norm": 9.062711443353237,
+      "learning_rate": 2e-05,
+      "loss": 1.4928,
+      "step": 872
+    },
+    {
+      "epoch": 0.020779777206512424,
+      "grad_norm": 2.735527368636789,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 873
+    },
+    {
+      "epoch": 0.02080357992954394,
+      "grad_norm": 10.171587838385015,
+      "learning_rate": 2e-05,
+      "loss": 1.2452,
+      "step": 874
+    },
+    {
+      "epoch": 0.020827382652575454,
+      "grad_norm": 5.6240863889309365,
+      "learning_rate": 2e-05,
+      "loss": 0.2933,
+      "step": 875
+    },
+    {
+      "epoch": 0.02085118537560697,
+      "grad_norm": 5.904160679042924,
+      "learning_rate": 2e-05,
+      "loss": 0.7154,
+      "step": 876
+    },
+    {
+      "epoch": 0.020874988098638483,
+      "grad_norm": 3.509677220746993,
+      "learning_rate": 2e-05,
+      "loss": 0.1704,
+      "step": 877
+    },
+    {
+      "epoch": 0.02089879082167,
+      "grad_norm": 7.376233452513279,
+      "learning_rate": 2e-05,
+      "loss": 0.442,
+      "step": 878
+    },
+    {
+      "epoch": 0.020922593544701513,
+      "grad_norm": 8.440788233093368,
+      "learning_rate": 2e-05,
+      "loss": 0.9676,
+      "step": 879
+    },
+    {
+      "epoch": 0.02094639626773303,
+      "grad_norm": 6.4625672099480695,
+      "learning_rate": 2e-05,
+      "loss": 0.6198,
+      "step": 880
+    },
+    {
+      "epoch": 0.020970198990764542,
+      "grad_norm": 3.135997613255641,
+      "learning_rate": 2e-05,
+      "loss": 0.1784,
+      "step": 881
+    },
+    {
+      "epoch": 0.02099400171379606,
+      "grad_norm": 4.706743285530092,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 882
+    },
+    {
+      "epoch": 0.02101780443682757,
+      "grad_norm": 3.781017018591294,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 883
+    },
+    {
+      "epoch": 0.021041607159859088,
+      "grad_norm": 4.5073373094656315,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 884
+    },
+    {
+      "epoch": 0.0210654098828906,
+      "grad_norm": 4.013935487804735,
+      "learning_rate": 2e-05,
+      "loss": 0.3382,
+      "step": 885
+    },
+    {
+      "epoch": 0.021089212605922118,
+      "grad_norm": 12.148204603577543,
+      "learning_rate": 2e-05,
+      "loss": 1.1542,
+      "step": 886
+    },
+    {
+      "epoch": 0.02111301532895363,
+      "grad_norm": 3.7931074569797407,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 887
+    },
+    {
+      "epoch": 0.021136818051985147,
+      "grad_norm": 5.712435103455694,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 888
+    },
+    {
+      "epoch": 0.021160620775016663,
+      "grad_norm": 2.3618548527028422,
+      "learning_rate": 2e-05,
+      "loss": 0.2196,
+      "step": 889
+    },
+    {
+      "epoch": 0.021184423498048176,
+      "grad_norm": 7.724075699373179,
+      "learning_rate": 2e-05,
+      "loss": 0.8154,
+      "step": 890
+    },
+    {
+      "epoch": 0.021208226221079693,
+      "grad_norm": 9.886753544856846,
+      "learning_rate": 2e-05,
+      "loss": 1.3864,
+      "step": 891
+    },
+    {
+      "epoch": 0.021232028944111206,
+      "grad_norm": 6.008019698665216,
+      "learning_rate": 2e-05,
+      "loss": 0.7832,
+      "step": 892
+    },
+    {
+      "epoch": 0.021255831667142722,
+      "grad_norm": 3.866804280922993,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 893
+    },
+    {
+      "epoch": 0.021279634390174235,
+      "grad_norm": 6.33498011362733,
+      "learning_rate": 2e-05,
+      "loss": 0.2471,
+      "step": 894
+    },
+    {
+      "epoch": 0.021303437113205752,
+      "grad_norm": 7.365512757919612,
+      "learning_rate": 2e-05,
+      "loss": 0.8704,
+      "step": 895
+    },
+    {
+      "epoch": 0.021327239836237265,
+      "grad_norm": 2.5263396980695974,
+      "learning_rate": 2e-05,
+      "loss": 0.161,
+      "step": 896
+    },
+    {
+      "epoch": 0.02135104255926878,
+      "grad_norm": 5.17023687093144,
+      "learning_rate": 2e-05,
+      "loss": 0.5452,
+      "step": 897
+    },
+    {
+      "epoch": 0.021374845282300294,
+      "grad_norm": 4.028753667490679,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 898
+    },
+    {
+      "epoch": 0.02139864800533181,
+      "grad_norm": 9.757642475025499,
+      "learning_rate": 2e-05,
+      "loss": 1.5525,
+      "step": 899
+    },
+    {
+      "epoch": 0.021422450728363324,
+      "grad_norm": 7.716284940637684,
+      "learning_rate": 2e-05,
+      "loss": 0.5834,
+      "step": 900
+    },
+    {
+      "epoch": 0.02144625345139484,
+      "grad_norm": 8.476816016406481,
+      "learning_rate": 2e-05,
+      "loss": 0.4632,
+      "step": 901
+    },
+    {
+      "epoch": 0.021470056174426353,
+      "grad_norm": 2.545619960494959,
+      "learning_rate": 2e-05,
+      "loss": 0.2072,
+      "step": 902
+    },
+    {
+      "epoch": 0.02149385889745787,
+      "grad_norm": 9.76638672626342,
+      "learning_rate": 2e-05,
+      "loss": 1.1796,
+      "step": 903
+    },
+    {
+      "epoch": 0.021517661620489383,
+      "grad_norm": 8.406455689162982,
+      "learning_rate": 2e-05,
+      "loss": 1.2891,
+      "step": 904
+    },
+    {
+      "epoch": 0.0215414643435209,
+      "grad_norm": 6.836458073620838,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 905
+    },
+    {
+      "epoch": 0.021565267066552412,
+      "grad_norm": 7.223905156740253,
+      "learning_rate": 2e-05,
+      "loss": 0.8778,
+      "step": 906
+    },
+    {
+      "epoch": 0.02158906978958393,
+      "grad_norm": 5.149739594336871,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 907
+    },
+    {
+      "epoch": 0.021612872512615445,
+      "grad_norm": 8.906118319622326,
+      "learning_rate": 2e-05,
+      "loss": 0.3045,
+      "step": 908
+    },
+    {
+      "epoch": 0.021636675235646958,
+      "grad_norm": 13.088463259407746,
+      "learning_rate": 2e-05,
+      "loss": 0.9724,
+      "step": 909
+    },
+    {
+      "epoch": 0.021660477958678474,
+      "grad_norm": 9.50899747184352,
+      "learning_rate": 2e-05,
+      "loss": 0.8514,
+      "step": 910
+    },
+    {
+      "epoch": 0.021684280681709987,
+      "grad_norm": 3.9260011188667145,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 911
+    },
+    {
+      "epoch": 0.021708083404741504,
+      "grad_norm": 2.265997404554876,
+      "learning_rate": 2e-05,
+      "loss": 0.1941,
+      "step": 912
+    },
+    {
+      "epoch": 0.021731886127773017,
+      "grad_norm": 5.170948909801362,
+      "learning_rate": 2e-05,
+      "loss": 0.5127,
+      "step": 913
+    },
+    {
+      "epoch": 0.021755688850804533,
+      "grad_norm": 3.742713139811487,
+      "learning_rate": 2e-05,
+      "loss": 0.1888,
+      "step": 914
+    },
+    {
+      "epoch": 0.021779491573836046,
+      "grad_norm": 7.173260445795334,
+      "learning_rate": 2e-05,
+      "loss": 1.1651,
+      "step": 915
+    },
+    {
+      "epoch": 0.021803294296867563,
+      "grad_norm": 8.677061014424059,
+      "learning_rate": 2e-05,
+      "loss": 0.3496,
+      "step": 916
+    },
+    {
+      "epoch": 0.021827097019899076,
+      "grad_norm": 16.076973316457988,
+      "learning_rate": 2e-05,
+      "loss": 1.3192,
+      "step": 917
+    },
+    {
+      "epoch": 0.021850899742930592,
+      "grad_norm": 6.485534946538841,
+      "learning_rate": 2e-05,
+      "loss": 0.3045,
+      "step": 918
+    },
+    {
+      "epoch": 0.021874702465962105,
+      "grad_norm": 6.035181141857809,
+      "learning_rate": 2e-05,
+      "loss": 0.4446,
+      "step": 919
+    },
+    {
+      "epoch": 0.02189850518899362,
+      "grad_norm": 6.381477821308645,
+      "learning_rate": 2e-05,
+      "loss": 0.2573,
+      "step": 920
+    },
+    {
+      "epoch": 0.021922307912025134,
+      "grad_norm": 6.1181003269143135,
+      "learning_rate": 2e-05,
+      "loss": 0.6939,
+      "step": 921
+    },
+    {
+      "epoch": 0.02194611063505665,
+      "grad_norm": 3.4568518826802044,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 922
+    },
+    {
+      "epoch": 0.021969913358088164,
+      "grad_norm": 9.662356083072865,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 923
+    },
+    {
+      "epoch": 0.02199371608111968,
+      "grad_norm": 3.2662953057790776,
+      "learning_rate": 2e-05,
+      "loss": 0.2563,
+      "step": 924
+    },
+    {
+      "epoch": 0.022017518804151193,
+      "grad_norm": 3.1502295112920136,
+      "learning_rate": 2e-05,
+      "loss": 0.124,
+      "step": 925
+    },
+    {
+      "epoch": 0.02204132152718271,
+      "grad_norm": 12.764164644336036,
+      "learning_rate": 2e-05,
+      "loss": 0.6153,
+      "step": 926
+    },
+    {
+      "epoch": 0.022065124250214223,
+      "grad_norm": 2.9472225192485304,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 927
+    },
+    {
+      "epoch": 0.02208892697324574,
+      "grad_norm": 6.121404943267903,
+      "learning_rate": 2e-05,
+      "loss": 0.2626,
+      "step": 928
+    },
+    {
+      "epoch": 0.022112729696277256,
+      "grad_norm": 5.946632205496467,
+      "learning_rate": 2e-05,
+      "loss": 0.3397,
+      "step": 929
+    },
+    {
+      "epoch": 0.02213653241930877,
+      "grad_norm": 9.604580089513352,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 930
+    },
+    {
+      "epoch": 0.022160335142340285,
+      "grad_norm": 3.940159209843156,
+      "learning_rate": 2e-05,
+      "loss": 0.3014,
+      "step": 931
+    },
+    {
+      "epoch": 0.022184137865371798,
+      "grad_norm": 8.224971235703844,
+      "learning_rate": 2e-05,
+      "loss": 1.3474,
+      "step": 932
+    },
+    {
+      "epoch": 0.022207940588403315,
+      "grad_norm": 6.096807880459876,
+      "learning_rate": 2e-05,
+      "loss": 0.4737,
+      "step": 933
+    },
+    {
+      "epoch": 0.022231743311434828,
+      "grad_norm": 7.946401155586971,
+      "learning_rate": 2e-05,
+      "loss": 0.3009,
+      "step": 934
+    },
+    {
+      "epoch": 0.022255546034466344,
+      "grad_norm": 4.351818697876006,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 935
+    },
+    {
+      "epoch": 0.022279348757497857,
+      "grad_norm": 6.578301769331393,
+      "learning_rate": 2e-05,
+      "loss": 0.6859,
+      "step": 936
+    },
+    {
+      "epoch": 0.022303151480529373,
+      "grad_norm": 9.926426461572724,
+      "learning_rate": 2e-05,
+      "loss": 1.1893,
+      "step": 937
+    },
+    {
+      "epoch": 0.022326954203560886,
+      "grad_norm": 2.052445465986876,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 938
+    },
+    {
+      "epoch": 0.022350756926592403,
+      "grad_norm": 13.002702310081142,
+      "learning_rate": 2e-05,
+      "loss": 0.8722,
+      "step": 939
+    },
+    {
+      "epoch": 0.022374559649623916,
+      "grad_norm": 4.507713419778045,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 940
+    },
+    {
+      "epoch": 0.022398362372655432,
+      "grad_norm": 8.39408083755895,
+      "learning_rate": 2e-05,
+      "loss": 0.9438,
+      "step": 941
+    },
+    {
+      "epoch": 0.022422165095686945,
+      "grad_norm": 3.1061208161720795,
+      "learning_rate": 2e-05,
+      "loss": 0.2178,
+      "step": 942
+    },
+    {
+      "epoch": 0.022445967818718462,
+      "grad_norm": 4.812342163617467,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 943
+    },
+    {
+      "epoch": 0.022469770541749975,
+      "grad_norm": 7.206762830463444,
+      "learning_rate": 2e-05,
+      "loss": 0.7478,
+      "step": 944
+    },
+    {
+      "epoch": 0.02249357326478149,
+      "grad_norm": 4.5186790188730175,
+      "learning_rate": 2e-05,
+      "loss": 0.1767,
+      "step": 945
+    },
+    {
+      "epoch": 0.022517375987813004,
+      "grad_norm": 9.687778196094179,
+      "learning_rate": 2e-05,
+      "loss": 0.3644,
+      "step": 946
+    },
+    {
+      "epoch": 0.02254117871084452,
+      "grad_norm": 2.1113400096484733,
+      "learning_rate": 2e-05,
+      "loss": 0.1959,
+      "step": 947
+    },
+    {
+      "epoch": 0.022564981433876037,
+      "grad_norm": 11.660126750383284,
+      "learning_rate": 2e-05,
+      "loss": 1.8597,
+      "step": 948
+    },
+    {
+      "epoch": 0.02258878415690755,
+      "grad_norm": 8.474921573118534,
+      "learning_rate": 2e-05,
+      "loss": 0.3744,
+      "step": 949
+    },
+    {
+      "epoch": 0.022612586879939067,
+      "grad_norm": 8.77175214638995,
+      "learning_rate": 2e-05,
+      "loss": 0.6804,
+      "step": 950
+    },
+    {
+      "epoch": 0.02263638960297058,
+      "grad_norm": 6.749059479820697,
+      "learning_rate": 2e-05,
+      "loss": 0.9563,
+      "step": 951
+    },
+    {
+      "epoch": 0.022660192326002096,
+      "grad_norm": 3.117338646977159,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 952
+    },
+    {
+      "epoch": 0.02268399504903361,
+      "grad_norm": 4.788768560481948,
+      "learning_rate": 2e-05,
+      "loss": 0.1675,
+      "step": 953
+    },
+    {
+      "epoch": 0.022707797772065125,
+      "grad_norm": 7.208540251602565,
+      "learning_rate": 2e-05,
+      "loss": 0.4269,
+      "step": 954
+    },
+    {
+      "epoch": 0.02273160049509664,
+      "grad_norm": 7.490050421759173,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 955
+    },
+    {
+      "epoch": 0.022755403218128155,
+      "grad_norm": 3.4166389548569946,
+      "learning_rate": 2e-05,
+      "loss": 0.2801,
+      "step": 956
+    },
+    {
+      "epoch": 0.022779205941159668,
+      "grad_norm": 4.1516469510936895,
+      "learning_rate": 2e-05,
+      "loss": 0.1996,
+      "step": 957
+    },
+    {
+      "epoch": 0.022803008664191184,
+      "grad_norm": 8.469052301109322,
+      "learning_rate": 2e-05,
+      "loss": 0.474,
+      "step": 958
+    },
+    {
+      "epoch": 0.022826811387222697,
+      "grad_norm": 9.266871287464653,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 959
+    },
+    {
+      "epoch": 0.022850614110254214,
+      "grad_norm": 5.222272917449646,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 960
+    },
+    {
+      "epoch": 0.022874416833285727,
+      "grad_norm": 4.792360675837933,
+      "learning_rate": 2e-05,
+      "loss": 0.3057,
+      "step": 961
+    },
+    {
+      "epoch": 0.022898219556317243,
+      "grad_norm": 7.780461522388737,
+      "learning_rate": 2e-05,
+      "loss": 0.3601,
+      "step": 962
+    },
+    {
+      "epoch": 0.022922022279348756,
+      "grad_norm": 4.345595387619765,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 963
+    },
+    {
+      "epoch": 0.022945825002380273,
+      "grad_norm": 2.3482562956797657,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 964
+    },
+    {
+      "epoch": 0.022969627725411786,
+      "grad_norm": 4.375451520719754,
+      "learning_rate": 2e-05,
+      "loss": 0.2714,
+      "step": 965
+    },
+    {
+      "epoch": 0.022993430448443302,
+      "grad_norm": 11.620079339328232,
+      "learning_rate": 2e-05,
+      "loss": 1.0963,
+      "step": 966
+    },
+    {
+      "epoch": 0.023017233171474815,
+      "grad_norm": 5.3714790854473415,
+      "learning_rate": 2e-05,
+      "loss": 0.3036,
+      "step": 967
+    },
+    {
+      "epoch": 0.02304103589450633,
+      "grad_norm": 15.574801330577513,
+      "learning_rate": 2e-05,
+      "loss": 1.1517,
+      "step": 968
+    },
+    {
+      "epoch": 0.023064838617537848,
+      "grad_norm": 11.176290383508723,
+      "learning_rate": 2e-05,
+      "loss": 1.3681,
+      "step": 969
+    },
+    {
+      "epoch": 0.02308864134056936,
+      "grad_norm": 6.715076344600935,
+      "learning_rate": 2e-05,
+      "loss": 0.5379,
+      "step": 970
+    },
+    {
+      "epoch": 0.023112444063600877,
+      "grad_norm": 9.351183569637682,
+      "learning_rate": 2e-05,
+      "loss": 0.8985,
+      "step": 971
+    },
+    {
+      "epoch": 0.02313624678663239,
+      "grad_norm": 4.272390637817902,
+      "learning_rate": 2e-05,
+      "loss": 0.2858,
+      "step": 972
+    },
+    {
+      "epoch": 0.023160049509663907,
+      "grad_norm": 12.83893468791205,
+      "learning_rate": 2e-05,
+      "loss": 0.4223,
+      "step": 973
+    },
+    {
+      "epoch": 0.02318385223269542,
+      "grad_norm": 6.403963470109854,
+      "learning_rate": 2e-05,
+      "loss": 0.6715,
+      "step": 974
+    },
+    {
+      "epoch": 0.023207654955726936,
+      "grad_norm": 2.884333531706562,
+      "learning_rate": 2e-05,
+      "loss": 0.2084,
+      "step": 975
+    },
+    {
+      "epoch": 0.02323145767875845,
+      "grad_norm": 3.3134325600936263,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 976
+    },
+    {
+      "epoch": 0.023255260401789966,
+      "grad_norm": 9.236093888212707,
+      "learning_rate": 2e-05,
+      "loss": 0.6956,
+      "step": 977
+    },
+    {
+      "epoch": 0.02327906312482148,
+      "grad_norm": 4.165481363372096,
+      "learning_rate": 2e-05,
+      "loss": 0.1923,
+      "step": 978
+    },
+    {
+      "epoch": 0.023302865847852995,
+      "grad_norm": 7.056087344465907,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 979
+    },
+    {
+      "epoch": 0.023326668570884508,
+      "grad_norm": 6.437276207713748,
+      "learning_rate": 2e-05,
+      "loss": 0.3096,
+      "step": 980
+    },
+    {
+      "epoch": 0.023350471293916025,
+      "grad_norm": 6.365009668636194,
+      "learning_rate": 2e-05,
+      "loss": 0.2645,
+      "step": 981
+    },
+    {
+      "epoch": 0.023374274016947538,
+      "grad_norm": 8.471580715221945,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 982
+    },
+    {
+      "epoch": 0.023398076739979054,
+      "grad_norm": 12.792087522915441,
+      "learning_rate": 2e-05,
+      "loss": 1.9902,
+      "step": 983
+    },
+    {
+      "epoch": 0.023421879463010567,
+      "grad_norm": 9.023188582543344,
+      "learning_rate": 2e-05,
+      "loss": 1.4349,
+      "step": 984
+    },
+    {
+      "epoch": 0.023445682186042083,
+      "grad_norm": 6.040701922117586,
+      "learning_rate": 2e-05,
+      "loss": 0.7748,
+      "step": 985
+    },
+    {
+      "epoch": 0.023469484909073596,
+      "grad_norm": 6.368998236038695,
+      "learning_rate": 2e-05,
+      "loss": 0.9337,
+      "step": 986
+    },
+    {
+      "epoch": 0.023493287632105113,
+      "grad_norm": 2.877658793229062,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 987
+    },
+    {
+      "epoch": 0.02351709035513663,
+      "grad_norm": 7.998013880575342,
+      "learning_rate": 2e-05,
+      "loss": 0.7946,
+      "step": 988
+    },
+    {
+      "epoch": 0.023540893078168142,
+      "grad_norm": 8.444155496008715,
+      "learning_rate": 2e-05,
+      "loss": 0.6951,
+      "step": 989
+    },
+    {
+      "epoch": 0.02356469580119966,
+      "grad_norm": 13.923834339513856,
+      "learning_rate": 2e-05,
+      "loss": 0.7139,
+      "step": 990
+    },
+    {
+      "epoch": 0.023588498524231172,
+      "grad_norm": 7.4869739440042276,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 991
+    },
+    {
+      "epoch": 0.023612301247262688,
+      "grad_norm": 2.2081985609291173,
+      "learning_rate": 2e-05,
+      "loss": 0.1473,
+      "step": 992
+    },
+    {
+      "epoch": 0.0236361039702942,
+      "grad_norm": 8.911150722285438,
+      "learning_rate": 2e-05,
+      "loss": 1.3909,
+      "step": 993
+    },
+    {
+      "epoch": 0.023659906693325718,
+      "grad_norm": 2.9476491451771234,
+      "learning_rate": 2e-05,
+      "loss": 0.2058,
+      "step": 994
+    },
+    {
+      "epoch": 0.02368370941635723,
+      "grad_norm": 5.858808580786447,
+      "learning_rate": 2e-05,
+      "loss": 0.3049,
+      "step": 995
+    },
+    {
+      "epoch": 0.023707512139388747,
+      "grad_norm": 6.980650421676692,
+      "learning_rate": 2e-05,
+      "loss": 0.8619,
+      "step": 996
+    },
+    {
+      "epoch": 0.02373131486242026,
+      "grad_norm": 2.3069900882885372,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 997
+    },
+    {
+      "epoch": 0.023755117585451777,
+      "grad_norm": 8.311410672539422,
+      "learning_rate": 2e-05,
+      "loss": 0.3906,
+      "step": 998
+    },
+    {
+      "epoch": 0.02377892030848329,
+      "grad_norm": 14.354184555624173,
+      "learning_rate": 2e-05,
+      "loss": 1.0881,
+      "step": 999
+    },
+    {
+      "epoch": 0.023802723031514806,
+      "grad_norm": 4.131706015472593,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02382652575454632,
+      "grad_norm": 11.240997188209757,
+      "learning_rate": 2e-05,
+      "loss": 1.0136,
+      "step": 1001
+    },
+    {
+      "epoch": 0.023850328477577835,
+      "grad_norm": 15.502240515781741,
+      "learning_rate": 2e-05,
+      "loss": 1.1251,
+      "step": 1002
+    },
+    {
+      "epoch": 0.02387413120060935,
+      "grad_norm": 3.9917898780453394,
+      "learning_rate": 2e-05,
+      "loss": 0.21,
+      "step": 1003
+    },
+    {
+      "epoch": 0.023897933923640865,
+      "grad_norm": 8.737279703468499,
+      "learning_rate": 2e-05,
+      "loss": 0.8313,
+      "step": 1004
+    },
+    {
+      "epoch": 0.023921736646672378,
+      "grad_norm": 6.963351186446959,
+      "learning_rate": 2e-05,
+      "loss": 0.4492,
+      "step": 1005
+    },
+    {
+      "epoch": 0.023945539369703894,
+      "grad_norm": 2.9108289809367895,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 1006
+    },
+    {
+      "epoch": 0.023969342092735407,
+      "grad_norm": 2.7638039173833713,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 1007
+    },
+    {
+      "epoch": 0.023993144815766924,
+      "grad_norm": 2.9738022076036414,
+      "learning_rate": 2e-05,
+      "loss": 0.2284,
+      "step": 1008
+    },
+    {
+      "epoch": 0.02401694753879844,
+      "grad_norm": 11.991472357665577,
+      "learning_rate": 2e-05,
+      "loss": 1.1598,
+      "step": 1009
+    },
+    {
+      "epoch": 0.024040750261829953,
+      "grad_norm": 3.5411044774932927,
+      "learning_rate": 2e-05,
+      "loss": 0.1573,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02406455298486147,
+      "grad_norm": 8.17113921473503,
+      "learning_rate": 2e-05,
+      "loss": 0.5697,
+      "step": 1011
+    },
+    {
+      "epoch": 0.024088355707892983,
+      "grad_norm": 9.409891898222995,
+      "learning_rate": 2e-05,
+      "loss": 1.0899,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0241121584309245,
+      "grad_norm": 8.826386319176311,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 1013
+    },
+    {
+      "epoch": 0.024135961153956012,
+      "grad_norm": 8.204863258371871,
+      "learning_rate": 2e-05,
+      "loss": 0.4031,
+      "step": 1014
+    },
+    {
+      "epoch": 0.02415976387698753,
+      "grad_norm": 3.493569317487707,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 1015
+    },
+    {
+      "epoch": 0.02418356660001904,
+      "grad_norm": 4.734420576251277,
+      "learning_rate": 2e-05,
+      "loss": 0.1885,
+      "step": 1016
+    },
+    {
+      "epoch": 0.024207369323050558,
+      "grad_norm": 8.68667750237134,
+      "learning_rate": 2e-05,
+      "loss": 0.7348,
+      "step": 1017
+    },
+    {
+      "epoch": 0.02423117204608207,
+      "grad_norm": 11.145306038995914,
+      "learning_rate": 2e-05,
+      "loss": 0.5994,
+      "step": 1018
+    },
+    {
+      "epoch": 0.024254974769113587,
+      "grad_norm": 7.045413972289054,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0242787774921451,
+      "grad_norm": 8.349149782185401,
+      "learning_rate": 2e-05,
+      "loss": 0.8011,
+      "step": 1020
+    },
+    {
+      "epoch": 0.024302580215176617,
+      "grad_norm": 6.242956064157823,
+      "learning_rate": 2e-05,
+      "loss": 0.4948,
+      "step": 1021
+    },
+    {
+      "epoch": 0.02432638293820813,
+      "grad_norm": 6.282590158949662,
+      "learning_rate": 2e-05,
+      "loss": 0.2979,
+      "step": 1022
+    },
+    {
+      "epoch": 0.024350185661239646,
+      "grad_norm": 12.549329564354876,
+      "learning_rate": 2e-05,
+      "loss": 0.982,
+      "step": 1023
+    },
+    {
+      "epoch": 0.02437398838427116,
+      "grad_norm": 10.656019095740437,
+      "learning_rate": 2e-05,
+      "loss": 1.1055,
+      "step": 1024
+    },
+    {
+      "epoch": 0.024397791107302676,
+      "grad_norm": 5.677274116607746,
+      "learning_rate": 2e-05,
+      "loss": 0.9288,
+      "step": 1025
+    },
+    {
+      "epoch": 0.02442159383033419,
+      "grad_norm": 4.33266259653078,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 1026
+    },
+    {
+      "epoch": 0.024445396553365705,
+      "grad_norm": 10.31412571640231,
+      "learning_rate": 2e-05,
+      "loss": 1.2848,
+      "step": 1027
+    },
+    {
+      "epoch": 0.024469199276397218,
+      "grad_norm": 6.424702965444464,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 1028
+    },
+    {
+      "epoch": 0.024493001999428735,
+      "grad_norm": 14.534361860267676,
+      "learning_rate": 2e-05,
+      "loss": 0.8751,
+      "step": 1029
+    },
+    {
+      "epoch": 0.02451680472246025,
+      "grad_norm": 6.974959357585443,
+      "learning_rate": 2e-05,
+      "loss": 0.3418,
+      "step": 1030
+    },
+    {
+      "epoch": 0.024540607445491764,
+      "grad_norm": 12.976762650984009,
+      "learning_rate": 2e-05,
+      "loss": 0.3847,
+      "step": 1031
+    },
+    {
+      "epoch": 0.02456441016852328,
+      "grad_norm": 7.985740457495593,
+      "learning_rate": 2e-05,
+      "loss": 1.4525,
+      "step": 1032
+    },
+    {
+      "epoch": 0.024588212891554793,
+      "grad_norm": 3.1542721199279833,
+      "learning_rate": 2e-05,
+      "loss": 0.1997,
+      "step": 1033
+    },
+    {
+      "epoch": 0.02461201561458631,
+      "grad_norm": 3.121046186579733,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 1034
+    },
+    {
+      "epoch": 0.024635818337617823,
+      "grad_norm": 11.056229511598545,
+      "learning_rate": 2e-05,
+      "loss": 1.021,
+      "step": 1035
+    },
+    {
+      "epoch": 0.02465962106064934,
+      "grad_norm": 5.8090246526653715,
+      "learning_rate": 2e-05,
+      "loss": 0.7646,
+      "step": 1036
+    },
+    {
+      "epoch": 0.024683423783680852,
+      "grad_norm": 6.127035856660801,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 1037
+    },
+    {
+      "epoch": 0.02470722650671237,
+      "grad_norm": 2.2062231012797007,
+      "learning_rate": 2e-05,
+      "loss": 0.1974,
+      "step": 1038
+    },
+    {
+      "epoch": 0.024731029229743882,
+      "grad_norm": 5.357395572812083,
+      "learning_rate": 2e-05,
+      "loss": 0.3127,
+      "step": 1039
+    },
+    {
+      "epoch": 0.024754831952775398,
+      "grad_norm": 3.06232078081393,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02477863467580691,
+      "grad_norm": 10.159382572951003,
+      "learning_rate": 2e-05,
+      "loss": 1.5157,
+      "step": 1041
+    },
+    {
+      "epoch": 0.024802437398838428,
+      "grad_norm": 7.286385517697458,
+      "learning_rate": 2e-05,
+      "loss": 0.9615,
+      "step": 1042
+    },
+    {
+      "epoch": 0.02482624012186994,
+      "grad_norm": 8.486883278120157,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 1043
+    },
+    {
+      "epoch": 0.024850042844901457,
+      "grad_norm": 4.1766412892920775,
+      "learning_rate": 2e-05,
+      "loss": 0.3962,
+      "step": 1044
+    },
+    {
+      "epoch": 0.02487384556793297,
+      "grad_norm": 6.923311936184498,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 1045
+    },
+    {
+      "epoch": 0.024897648290964487,
+      "grad_norm": 3.597796580190192,
+      "learning_rate": 2e-05,
+      "loss": 0.2888,
+      "step": 1046
+    },
+    {
+      "epoch": 0.024921451013996,
+      "grad_norm": 8.826742281797207,
+      "learning_rate": 2e-05,
+      "loss": 0.8072,
+      "step": 1047
+    },
+    {
+      "epoch": 0.024945253737027516,
+      "grad_norm": 2.4860957019822467,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 1048
+    },
+    {
+      "epoch": 0.024969056460059032,
+      "grad_norm": 2.635073481787966,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 1049
+    },
+    {
+      "epoch": 0.024992859183090545,
+      "grad_norm": 2.5799504858204725,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 1050
+    },
+    {
+      "epoch": 0.025016661906122062,
+      "grad_norm": 20.218222931599257,
+      "learning_rate": 2e-05,
+      "loss": 0.8694,
+      "step": 1051
+    },
+    {
+      "epoch": 0.025040464629153575,
+      "grad_norm": 3.930079459224063,
+      "learning_rate": 2e-05,
+      "loss": 0.1789,
+      "step": 1052
+    },
+    {
+      "epoch": 0.02506426735218509,
+      "grad_norm": 5.470526219991793,
+      "learning_rate": 2e-05,
+      "loss": 0.7598,
+      "step": 1053
+    },
+    {
+      "epoch": 0.025088070075216604,
+      "grad_norm": 7.746951193258814,
+      "learning_rate": 2e-05,
+      "loss": 0.7519,
+      "step": 1054
+    },
+    {
+      "epoch": 0.02511187279824812,
+      "grad_norm": 3.1553612739002017,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 1055
+    },
+    {
+      "epoch": 0.025135675521279634,
+      "grad_norm": 3.7174375291037367,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 1056
+    },
+    {
+      "epoch": 0.02515947824431115,
+      "grad_norm": 4.895614318532625,
+      "learning_rate": 2e-05,
+      "loss": 0.4792,
+      "step": 1057
+    },
+    {
+      "epoch": 0.025183280967342663,
+      "grad_norm": 13.260279552591633,
+      "learning_rate": 2e-05,
+      "loss": 0.9087,
+      "step": 1058
+    },
+    {
+      "epoch": 0.02520708369037418,
+      "grad_norm": 8.864281549173494,
+      "learning_rate": 2e-05,
+      "loss": 0.5058,
+      "step": 1059
+    },
+    {
+      "epoch": 0.025230886413405693,
+      "grad_norm": 9.344801683795241,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02525468913643721,
+      "grad_norm": 8.978443987776037,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 1061
+    },
+    {
+      "epoch": 0.025278491859468722,
+      "grad_norm": 3.6640586541489113,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 1062
+    },
+    {
+      "epoch": 0.02530229458250024,
+      "grad_norm": 5.491755156762704,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 1063
+    },
+    {
+      "epoch": 0.02532609730553175,
+      "grad_norm": 8.904227562916816,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 1064
+    },
+    {
+      "epoch": 0.025349900028563268,
+      "grad_norm": 9.550048867345955,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 1065
+    },
+    {
+      "epoch": 0.02537370275159478,
+      "grad_norm": 2.488230335621183,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 1066
+    },
+    {
+      "epoch": 0.025397505474626297,
+      "grad_norm": 6.385244402301714,
+      "learning_rate": 2e-05,
+      "loss": 1.1231,
+      "step": 1067
+    },
+    {
+      "epoch": 0.02542130819765781,
+      "grad_norm": 11.175464352513428,
+      "learning_rate": 2e-05,
+      "loss": 0.7542,
+      "step": 1068
+    },
+    {
+      "epoch": 0.025445110920689327,
+      "grad_norm": 9.426376777409303,
+      "learning_rate": 2e-05,
+      "loss": 0.3899,
+      "step": 1069
+    },
+    {
+      "epoch": 0.025468913643720843,
+      "grad_norm": 10.138743812797719,
+      "learning_rate": 2e-05,
+      "loss": 1.2557,
+      "step": 1070
+    },
+    {
+      "epoch": 0.025492716366752356,
+      "grad_norm": 3.1487837490151906,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 1071
+    },
+    {
+      "epoch": 0.025516519089783873,
+      "grad_norm": 1.735144229154565,
+      "learning_rate": 2e-05,
+      "loss": 0.1891,
+      "step": 1072
+    },
+    {
+      "epoch": 0.025540321812815386,
+      "grad_norm": 4.151238794881497,
+      "learning_rate": 2e-05,
+      "loss": 0.4355,
+      "step": 1073
+    },
+    {
+      "epoch": 0.025564124535846902,
+      "grad_norm": 4.341267885187158,
+      "learning_rate": 2e-05,
+      "loss": 0.3429,
+      "step": 1074
+    },
+    {
+      "epoch": 0.025587927258878415,
+      "grad_norm": 3.994582648192849,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 1075
+    },
+    {
+      "epoch": 0.02561172998190993,
+      "grad_norm": 2.0667085652409876,
+      "learning_rate": 2e-05,
+      "loss": 0.2358,
+      "step": 1076
+    },
+    {
+      "epoch": 0.025635532704941445,
+      "grad_norm": 5.206203748979754,
+      "learning_rate": 2e-05,
+      "loss": 0.391,
+      "step": 1077
+    },
+    {
+      "epoch": 0.02565933542797296,
+      "grad_norm": 7.560250724545947,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 1078
+    },
+    {
+      "epoch": 0.025683138151004474,
+      "grad_norm": 3.344530699037345,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 1079
+    },
+    {
+      "epoch": 0.02570694087403599,
+      "grad_norm": 7.35032740131177,
+      "learning_rate": 2e-05,
+      "loss": 0.4072,
+      "step": 1080
+    },
+    {
+      "epoch": 0.025730743597067503,
+      "grad_norm": 8.520458713285757,
+      "learning_rate": 2e-05,
+      "loss": 0.6577,
+      "step": 1081
+    },
+    {
+      "epoch": 0.02575454632009902,
+      "grad_norm": 6.395254380161653,
+      "learning_rate": 2e-05,
+      "loss": 0.3158,
+      "step": 1082
+    },
+    {
+      "epoch": 0.025778349043130533,
+      "grad_norm": 9.440567322140424,
+      "learning_rate": 2e-05,
+      "loss": 0.7814,
+      "step": 1083
+    },
+    {
+      "epoch": 0.02580215176616205,
+      "grad_norm": 2.310051365573354,
+      "learning_rate": 2e-05,
+      "loss": 0.1675,
+      "step": 1084
+    },
+    {
+      "epoch": 0.025825954489193562,
+      "grad_norm": 8.004863657553962,
+      "learning_rate": 2e-05,
+      "loss": 0.7035,
+      "step": 1085
+    },
+    {
+      "epoch": 0.02584975721222508,
+      "grad_norm": 2.6532181793874217,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 1086
+    },
+    {
+      "epoch": 0.025873559935256592,
+      "grad_norm": 7.007620802989866,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 1087
+    },
+    {
+      "epoch": 0.025897362658288108,
+      "grad_norm": 9.327326619050233,
+      "learning_rate": 2e-05,
+      "loss": 0.9662,
+      "step": 1088
+    },
+    {
+      "epoch": 0.025921165381319625,
+      "grad_norm": 6.023062080876862,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 1089
+    },
+    {
+      "epoch": 0.025944968104351138,
+      "grad_norm": 9.896708188730692,
+      "learning_rate": 2e-05,
+      "loss": 0.887,
+      "step": 1090
+    },
+    {
+      "epoch": 0.025968770827382654,
+      "grad_norm": 3.6408872345565286,
+      "learning_rate": 2e-05,
+      "loss": 0.176,
+      "step": 1091
+    },
+    {
+      "epoch": 0.025992573550414167,
+      "grad_norm": 6.0076037232202,
+      "learning_rate": 2e-05,
+      "loss": 0.8903,
+      "step": 1092
+    },
+    {
+      "epoch": 0.026016376273445684,
+      "grad_norm": 14.720281032568346,
+      "learning_rate": 2e-05,
+      "loss": 1.357,
+      "step": 1093
+    },
+    {
+      "epoch": 0.026040178996477197,
+      "grad_norm": 5.069602784880054,
+      "learning_rate": 2e-05,
+      "loss": 0.2243,
+      "step": 1094
+    },
+    {
+      "epoch": 0.026063981719508713,
+      "grad_norm": 8.977345966488611,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 1095
+    },
+    {
+      "epoch": 0.026087784442540226,
+      "grad_norm": 3.693854841426392,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 1096
+    },
+    {
+      "epoch": 0.026111587165571742,
+      "grad_norm": 10.038813770295514,
+      "learning_rate": 2e-05,
+      "loss": 1.0767,
+      "step": 1097
+    },
+    {
+      "epoch": 0.026135389888603255,
+      "grad_norm": 8.33676720231025,
+      "learning_rate": 2e-05,
+      "loss": 0.4582,
+      "step": 1098
+    },
+    {
+      "epoch": 0.026159192611634772,
+      "grad_norm": 4.284138593029379,
+      "learning_rate": 2e-05,
+      "loss": 0.2971,
+      "step": 1099
+    },
+    {
+      "epoch": 0.026182995334666285,
+      "grad_norm": 10.402748598616437,
+      "learning_rate": 2e-05,
+      "loss": 0.4024,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0262067980576978,
+      "grad_norm": 5.7067548241515516,
+      "learning_rate": 2e-05,
+      "loss": 0.806,
+      "step": 1101
+    },
+    {
+      "epoch": 0.026230600780729314,
+      "grad_norm": 2.7901763591092106,
+      "learning_rate": 2e-05,
+      "loss": 0.2871,
+      "step": 1102
+    },
+    {
+      "epoch": 0.02625440350376083,
+      "grad_norm": 4.983895914381214,
+      "learning_rate": 2e-05,
+      "loss": 0.5804,
+      "step": 1103
+    },
+    {
+      "epoch": 0.026278206226792344,
+      "grad_norm": 4.3630491145578665,
+      "learning_rate": 2e-05,
+      "loss": 0.2741,
+      "step": 1104
+    },
+    {
+      "epoch": 0.02630200894982386,
+      "grad_norm": 2.7956305389190064,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 1105
+    },
+    {
+      "epoch": 0.026325811672855373,
+      "grad_norm": 2.451845280453167,
+      "learning_rate": 2e-05,
+      "loss": 0.2817,
+      "step": 1106
+    },
+    {
+      "epoch": 0.02634961439588689,
+      "grad_norm": 9.4674267661511,
+      "learning_rate": 2e-05,
+      "loss": 0.3925,
+      "step": 1107
+    },
+    {
+      "epoch": 0.026373417118918403,
+      "grad_norm": 4.130189670919089,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 1108
+    },
+    {
+      "epoch": 0.02639721984194992,
+      "grad_norm": 5.849227504093649,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 1109
+    },
+    {
+      "epoch": 0.026421022564981436,
+      "grad_norm": 4.171662924994044,
+      "learning_rate": 2e-05,
+      "loss": 0.1818,
+      "step": 1110
+    },
+    {
+      "epoch": 0.02644482528801295,
+      "grad_norm": 4.537863676865107,
+      "learning_rate": 2e-05,
+      "loss": 0.2438,
+      "step": 1111
+    },
+    {
+      "epoch": 0.026468628011044465,
+      "grad_norm": 11.020037441374663,
+      "learning_rate": 2e-05,
+      "loss": 0.8617,
+      "step": 1112
+    },
+    {
+      "epoch": 0.026492430734075978,
+      "grad_norm": 6.593492019314559,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 1113
+    },
+    {
+      "epoch": 0.026516233457107494,
+      "grad_norm": 3.336212087248474,
+      "learning_rate": 2e-05,
+      "loss": 0.2878,
+      "step": 1114
+    },
+    {
+      "epoch": 0.026540036180139007,
+      "grad_norm": 7.952124643256949,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 1115
+    },
+    {
+      "epoch": 0.026563838903170524,
+      "grad_norm": 8.004814358269556,
+      "learning_rate": 2e-05,
+      "loss": 0.9134,
+      "step": 1116
+    },
+    {
+      "epoch": 0.026587641626202037,
+      "grad_norm": 7.704537755109708,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 1117
+    },
+    {
+      "epoch": 0.026611444349233553,
+      "grad_norm": 2.66585454012258,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 1118
+    },
+    {
+      "epoch": 0.026635247072265066,
+      "grad_norm": 15.42023142794429,
+      "learning_rate": 2e-05,
+      "loss": 1.043,
+      "step": 1119
+    },
+    {
+      "epoch": 0.026659049795296583,
+      "grad_norm": 3.8026129739801195,
+      "learning_rate": 2e-05,
+      "loss": 0.1657,
+      "step": 1120
+    },
+    {
+      "epoch": 0.026682852518328096,
+      "grad_norm": 7.012080882015425,
+      "learning_rate": 2e-05,
+      "loss": 0.7561,
+      "step": 1121
+    },
+    {
+      "epoch": 0.026706655241359612,
+      "grad_norm": 9.13713872139179,
+      "learning_rate": 2e-05,
+      "loss": 1.0582,
+      "step": 1122
+    },
+    {
+      "epoch": 0.026730457964391125,
+      "grad_norm": 6.747820318266067,
+      "learning_rate": 2e-05,
+      "loss": 1.1173,
+      "step": 1123
+    },
+    {
+      "epoch": 0.02675426068742264,
+      "grad_norm": 2.7853665539003827,
+      "learning_rate": 2e-05,
+      "loss": 0.21,
+      "step": 1124
+    },
+    {
+      "epoch": 0.026778063410454155,
+      "grad_norm": 7.467784261157235,
+      "learning_rate": 2e-05,
+      "loss": 0.3848,
+      "step": 1125
+    },
+    {
+      "epoch": 0.02680186613348567,
+      "grad_norm": 8.288405685231183,
+      "learning_rate": 2e-05,
+      "loss": 0.3971,
+      "step": 1126
+    },
+    {
+      "epoch": 0.026825668856517184,
+      "grad_norm": 6.5577945707861725,
+      "learning_rate": 2e-05,
+      "loss": 0.3568,
+      "step": 1127
+    },
+    {
+      "epoch": 0.0268494715795487,
+      "grad_norm": 6.72994396188418,
+      "learning_rate": 2e-05,
+      "loss": 1.159,
+      "step": 1128
+    },
+    {
+      "epoch": 0.026873274302580217,
+      "grad_norm": 8.146618781445708,
+      "learning_rate": 2e-05,
+      "loss": 0.6917,
+      "step": 1129
+    },
+    {
+      "epoch": 0.02689707702561173,
+      "grad_norm": 3.46197239939856,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 1130
+    },
+    {
+      "epoch": 0.026920879748643246,
+      "grad_norm": 9.663950656779207,
+      "learning_rate": 2e-05,
+      "loss": 0.5565,
+      "step": 1131
+    },
+    {
+      "epoch": 0.02694468247167476,
+      "grad_norm": 2.4874053850253417,
+      "learning_rate": 2e-05,
+      "loss": 0.1631,
+      "step": 1132
+    },
+    {
+      "epoch": 0.026968485194706276,
+      "grad_norm": 12.145606923924575,
+      "learning_rate": 2e-05,
+      "loss": 0.9586,
+      "step": 1133
+    },
+    {
+      "epoch": 0.02699228791773779,
+      "grad_norm": 3.3428647480445686,
+      "learning_rate": 2e-05,
+      "loss": 0.2499,
+      "step": 1134
+    },
+    {
+      "epoch": 0.027016090640769305,
+      "grad_norm": 9.192423509782623,
+      "learning_rate": 2e-05,
+      "loss": 0.3765,
+      "step": 1135
+    },
+    {
+      "epoch": 0.027039893363800818,
+      "grad_norm": 8.684992680252071,
+      "learning_rate": 2e-05,
+      "loss": 0.5069,
+      "step": 1136
+    },
+    {
+      "epoch": 0.027063696086832335,
+      "grad_norm": 6.454364950361279,
+      "learning_rate": 2e-05,
+      "loss": 0.8614,
+      "step": 1137
+    },
+    {
+      "epoch": 0.027087498809863848,
+      "grad_norm": 10.468137956247563,
+      "learning_rate": 2e-05,
+      "loss": 1.5052,
+      "step": 1138
+    },
+    {
+      "epoch": 0.027111301532895364,
+      "grad_norm": 8.698807648332512,
+      "learning_rate": 2e-05,
+      "loss": 0.472,
+      "step": 1139
+    },
+    {
+      "epoch": 0.027135104255926877,
+      "grad_norm": 2.8754766209904172,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 1140
+    },
+    {
+      "epoch": 0.027158906978958394,
+      "grad_norm": 3.2788755326508485,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 1141
+    },
+    {
+      "epoch": 0.027182709701989907,
+      "grad_norm": 1.9946448123089773,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 1142
+    },
+    {
+      "epoch": 0.027206512425021423,
+      "grad_norm": 8.709351959610338,
+      "learning_rate": 2e-05,
+      "loss": 0.9035,
+      "step": 1143
+    },
+    {
+      "epoch": 0.027230315148052936,
+      "grad_norm": 4.245608394903351,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 1144
+    },
+    {
+      "epoch": 0.027254117871084452,
+      "grad_norm": 3.7985403003817795,
+      "learning_rate": 2e-05,
+      "loss": 0.2171,
+      "step": 1145
+    },
+    {
+      "epoch": 0.027277920594115965,
+      "grad_norm": 10.21436103196585,
+      "learning_rate": 2e-05,
+      "loss": 1.49,
+      "step": 1146
+    },
+    {
+      "epoch": 0.027301723317147482,
+      "grad_norm": 4.553239983232938,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 1147
+    },
+    {
+      "epoch": 0.027325526040178995,
+      "grad_norm": 3.1313038337495165,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 1148
+    },
+    {
+      "epoch": 0.02734932876321051,
+      "grad_norm": 7.156134153454355,
+      "learning_rate": 2e-05,
+      "loss": 1.0726,
+      "step": 1149
+    },
+    {
+      "epoch": 0.027373131486242028,
+      "grad_norm": 3.1208846009287825,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02739693420927354,
+      "grad_norm": 10.315228892787637,
+      "learning_rate": 2e-05,
+      "loss": 0.3919,
+      "step": 1151
+    },
+    {
+      "epoch": 0.027420736932305057,
+      "grad_norm": 2.961578497127855,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 1152
+    },
+    {
+      "epoch": 0.02744453965533657,
+      "grad_norm": 5.920877675743419,
+      "learning_rate": 2e-05,
+      "loss": 0.4674,
+      "step": 1153
+    },
+    {
+      "epoch": 0.027468342378368087,
+      "grad_norm": 9.680963356947814,
+      "learning_rate": 2e-05,
+      "loss": 0.3944,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0274921451013996,
+      "grad_norm": 6.269943778774134,
+      "learning_rate": 2e-05,
+      "loss": 0.2692,
+      "step": 1155
+    },
+    {
+      "epoch": 0.027515947824431116,
+      "grad_norm": 7.292894163713567,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 1156
+    },
+    {
+      "epoch": 0.02753975054746263,
+      "grad_norm": 5.590358415105337,
+      "learning_rate": 2e-05,
+      "loss": 0.5203,
+      "step": 1157
+    },
+    {
+      "epoch": 0.027563553270494145,
+      "grad_norm": 14.927466541499877,
+      "learning_rate": 2e-05,
+      "loss": 0.6312,
+      "step": 1158
+    },
+    {
+      "epoch": 0.02758735599352566,
+      "grad_norm": 3.199124171415328,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 1159
+    },
+    {
+      "epoch": 0.027611158716557175,
+      "grad_norm": 4.80946754650435,
+      "learning_rate": 2e-05,
+      "loss": 0.2702,
+      "step": 1160
+    },
+    {
+      "epoch": 0.027634961439588688,
+      "grad_norm": 11.741675694793276,
+      "learning_rate": 2e-05,
+      "loss": 1.7657,
+      "step": 1161
+    },
+    {
+      "epoch": 0.027658764162620204,
+      "grad_norm": 4.203575818948262,
+      "learning_rate": 2e-05,
+      "loss": 0.4436,
+      "step": 1162
+    },
+    {
+      "epoch": 0.027682566885651717,
+      "grad_norm": 6.319778707740388,
+      "learning_rate": 2e-05,
+      "loss": 0.3164,
+      "step": 1163
+    },
+    {
+      "epoch": 0.027706369608683234,
+      "grad_norm": 2.313266199775184,
+      "learning_rate": 2e-05,
+      "loss": 0.2355,
+      "step": 1164
+    },
+    {
+      "epoch": 0.027730172331714747,
+      "grad_norm": 11.156054299255354,
+      "learning_rate": 2e-05,
+      "loss": 1.4802,
+      "step": 1165
+    },
+    {
+      "epoch": 0.027753975054746263,
+      "grad_norm": 22.045835637086945,
+      "learning_rate": 2e-05,
+      "loss": 0.6968,
+      "step": 1166
+    },
+    {
+      "epoch": 0.027777777777777776,
+      "grad_norm": 3.9521628811329936,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 1167
+    },
+    {
+      "epoch": 0.027801580500809293,
+      "grad_norm": 2.172989780486741,
+      "learning_rate": 2e-05,
+      "loss": 0.1871,
+      "step": 1168
+    },
+    {
+      "epoch": 0.027825383223840806,
+      "grad_norm": 4.383533375541904,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 1169
+    },
+    {
+      "epoch": 0.027849185946872322,
+      "grad_norm": 9.18572139815676,
+      "learning_rate": 2e-05,
+      "loss": 0.4605,
+      "step": 1170
+    },
+    {
+      "epoch": 0.02787298866990384,
+      "grad_norm": 8.285985747310793,
+      "learning_rate": 2e-05,
+      "loss": 0.9661,
+      "step": 1171
+    },
+    {
+      "epoch": 0.02789679139293535,
+      "grad_norm": 7.149197534644379,
+      "learning_rate": 2e-05,
+      "loss": 0.8982,
+      "step": 1172
+    },
+    {
+      "epoch": 0.027920594115966868,
+      "grad_norm": 6.112369632951675,
+      "learning_rate": 2e-05,
+      "loss": 0.4223,
+      "step": 1173
+    },
+    {
+      "epoch": 0.02794439683899838,
+      "grad_norm": 11.485403007158359,
+      "learning_rate": 2e-05,
+      "loss": 2.0031,
+      "step": 1174
+    },
+    {
+      "epoch": 0.027968199562029897,
+      "grad_norm": 5.544738644644902,
+      "learning_rate": 2e-05,
+      "loss": 0.2543,
+      "step": 1175
+    },
+    {
+      "epoch": 0.02799200228506141,
+      "grad_norm": 3.214706702058267,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 1176
+    },
+    {
+      "epoch": 0.028015805008092927,
+      "grad_norm": 8.50418672812055,
+      "learning_rate": 2e-05,
+      "loss": 0.4063,
+      "step": 1177
+    },
+    {
+      "epoch": 0.02803960773112444,
+      "grad_norm": 8.803683491305211,
+      "learning_rate": 2e-05,
+      "loss": 0.9219,
+      "step": 1178
+    },
+    {
+      "epoch": 0.028063410454155956,
+      "grad_norm": 3.4026169757024634,
+      "learning_rate": 2e-05,
+      "loss": 0.1628,
+      "step": 1179
+    },
+    {
+      "epoch": 0.02808721317718747,
+      "grad_norm": 9.903822904268521,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 1180
+    },
+    {
+      "epoch": 0.028111015900218986,
+      "grad_norm": 4.780670609081801,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 1181
+    },
+    {
+      "epoch": 0.0281348186232505,
+      "grad_norm": 11.637129004197503,
+      "learning_rate": 2e-05,
+      "loss": 0.5405,
+      "step": 1182
+    },
+    {
+      "epoch": 0.028158621346282015,
+      "grad_norm": 11.617905121120854,
+      "learning_rate": 2e-05,
+      "loss": 0.3954,
+      "step": 1183
+    },
+    {
+      "epoch": 0.028182424069313528,
+      "grad_norm": 7.587474857181052,
+      "learning_rate": 2e-05,
+      "loss": 0.8195,
+      "step": 1184
+    },
+    {
+      "epoch": 0.028206226792345045,
+      "grad_norm": 5.116714721733965,
+      "learning_rate": 2e-05,
+      "loss": 0.3645,
+      "step": 1185
+    },
+    {
+      "epoch": 0.028230029515376558,
+      "grad_norm": 7.524526763987946,
+      "learning_rate": 2e-05,
+      "loss": 1.074,
+      "step": 1186
+    },
+    {
+      "epoch": 0.028253832238408074,
+      "grad_norm": 2.6948710835187457,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 1187
+    },
+    {
+      "epoch": 0.028277634961439587,
+      "grad_norm": 3.4309590999288164,
+      "learning_rate": 2e-05,
+      "loss": 0.2316,
+      "step": 1188
+    },
+    {
+      "epoch": 0.028301437684471104,
+      "grad_norm": 6.508073846850974,
+      "learning_rate": 2e-05,
+      "loss": 0.3432,
+      "step": 1189
+    },
+    {
+      "epoch": 0.02832524040750262,
+      "grad_norm": 16.484931102864188,
+      "learning_rate": 2e-05,
+      "loss": 1.0635,
+      "step": 1190
+    },
+    {
+      "epoch": 0.028349043130534133,
+      "grad_norm": 5.859741258438056,
+      "learning_rate": 2e-05,
+      "loss": 1.0223,
+      "step": 1191
+    },
+    {
+      "epoch": 0.02837284585356565,
+      "grad_norm": 10.440338918759359,
+      "learning_rate": 2e-05,
+      "loss": 0.5882,
+      "step": 1192
+    },
+    {
+      "epoch": 0.028396648576597162,
+      "grad_norm": 8.871225526828095,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 1193
+    },
+    {
+      "epoch": 0.02842045129962868,
+      "grad_norm": 9.03275383850097,
+      "learning_rate": 2e-05,
+      "loss": 0.4565,
+      "step": 1194
+    },
+    {
+      "epoch": 0.028444254022660192,
+      "grad_norm": 3.7994025572408914,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 1195
+    },
+    {
+      "epoch": 0.02846805674569171,
+      "grad_norm": 4.73207135806383,
+      "learning_rate": 2e-05,
+      "loss": 0.21,
+      "step": 1196
+    },
+    {
+      "epoch": 0.02849185946872322,
+      "grad_norm": 4.520530717872594,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 1197
+    },
+    {
+      "epoch": 0.028515662191754738,
+      "grad_norm": 7.090680534736859,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 1198
+    },
+    {
+      "epoch": 0.02853946491478625,
+      "grad_norm": 6.821357078205428,
+      "learning_rate": 2e-05,
+      "loss": 0.8193,
+      "step": 1199
+    },
+    {
+      "epoch": 0.028563267637817767,
+      "grad_norm": 5.118451524191978,
+      "learning_rate": 2e-05,
+      "loss": 0.7421,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02858707036084928,
+      "grad_norm": 4.165727253051571,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 1201
+    },
+    {
+      "epoch": 0.028610873083880797,
+      "grad_norm": 5.5791360717405505,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 1202
+    },
+    {
+      "epoch": 0.02863467580691231,
+      "grad_norm": 9.156094456077248,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 1203
+    },
+    {
+      "epoch": 0.028658478529943826,
+      "grad_norm": 4.686781887496338,
+      "learning_rate": 2e-05,
+      "loss": 0.2794,
+      "step": 1204
+    },
+    {
+      "epoch": 0.02868228125297534,
+      "grad_norm": 6.565824335257934,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 1205
+    },
+    {
+      "epoch": 0.028706083976006855,
+      "grad_norm": 10.205129669109596,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 1206
+    },
+    {
+      "epoch": 0.02872988669903837,
+      "grad_norm": 6.460614340491971,
+      "learning_rate": 2e-05,
+      "loss": 0.9467,
+      "step": 1207
+    },
+    {
+      "epoch": 0.028753689422069885,
+      "grad_norm": 6.185054194874075,
+      "learning_rate": 2e-05,
+      "loss": 0.5599,
+      "step": 1208
+    },
+    {
+      "epoch": 0.028777492145101398,
+      "grad_norm": 7.051185890147718,
+      "learning_rate": 2e-05,
+      "loss": 0.3499,
+      "step": 1209
+    },
+    {
+      "epoch": 0.028801294868132914,
+      "grad_norm": 9.80075280638188,
+      "learning_rate": 2e-05,
+      "loss": 1.082,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02882509759116443,
+      "grad_norm": 1.7112204490722995,
+      "learning_rate": 2e-05,
+      "loss": 0.2051,
+      "step": 1211
+    },
+    {
+      "epoch": 0.028848900314195944,
+      "grad_norm": 2.713162597089628,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 1212
+    },
+    {
+      "epoch": 0.02887270303722746,
+      "grad_norm": 4.960837014600847,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 1213
+    },
+    {
+      "epoch": 0.028896505760258973,
+      "grad_norm": 2.5065028000840157,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 1214
+    },
+    {
+      "epoch": 0.02892030848329049,
+      "grad_norm": 5.451470614837919,
+      "learning_rate": 2e-05,
+      "loss": 0.2458,
+      "step": 1215
+    },
+    {
+      "epoch": 0.028944111206322003,
+      "grad_norm": 8.239411380908312,
+      "learning_rate": 2e-05,
+      "loss": 0.7049,
+      "step": 1216
+    },
+    {
+      "epoch": 0.02896791392935352,
+      "grad_norm": 5.515333881627669,
+      "learning_rate": 2e-05,
+      "loss": 0.2471,
+      "step": 1217
+    },
+    {
+      "epoch": 0.028991716652385032,
+      "grad_norm": 8.838055774155395,
+      "learning_rate": 2e-05,
+      "loss": 1.2028,
+      "step": 1218
+    },
+    {
+      "epoch": 0.02901551937541655,
+      "grad_norm": 2.9315357856826094,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 1219
+    },
+    {
+      "epoch": 0.02903932209844806,
+      "grad_norm": 9.10736196378298,
+      "learning_rate": 2e-05,
+      "loss": 0.2832,
+      "step": 1220
+    },
+    {
+      "epoch": 0.029063124821479578,
+      "grad_norm": 4.483637410736259,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 1221
+    },
+    {
+      "epoch": 0.02908692754451109,
+      "grad_norm": 11.2178163684532,
+      "learning_rate": 2e-05,
+      "loss": 1.3962,
+      "step": 1222
+    },
+    {
+      "epoch": 0.029110730267542607,
+      "grad_norm": 6.292605141408625,
+      "learning_rate": 2e-05,
+      "loss": 0.2956,
+      "step": 1223
+    },
+    {
+      "epoch": 0.02913453299057412,
+      "grad_norm": 3.7217811556697393,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 1224
+    },
+    {
+      "epoch": 0.029158335713605637,
+      "grad_norm": 9.05146178039368,
+      "learning_rate": 2e-05,
+      "loss": 0.5503,
+      "step": 1225
+    },
+    {
+      "epoch": 0.02918213843663715,
+      "grad_norm": 2.1203596530582702,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 1226
+    },
+    {
+      "epoch": 0.029205941159668666,
+      "grad_norm": 9.319410447478678,
+      "learning_rate": 2e-05,
+      "loss": 0.3243,
+      "step": 1227
+    },
+    {
+      "epoch": 0.02922974388270018,
+      "grad_norm": 3.008555494000261,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 1228
+    },
+    {
+      "epoch": 0.029253546605731696,
+      "grad_norm": 10.18313115937334,
+      "learning_rate": 2e-05,
+      "loss": 1.6925,
+      "step": 1229
+    },
+    {
+      "epoch": 0.029277349328763212,
+      "grad_norm": 4.333680187249597,
+      "learning_rate": 2e-05,
+      "loss": 0.3384,
+      "step": 1230
+    },
+    {
+      "epoch": 0.029301152051794725,
+      "grad_norm": 6.636394117605125,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 1231
+    },
+    {
+      "epoch": 0.02932495477482624,
+      "grad_norm": 9.884758524813147,
+      "learning_rate": 2e-05,
+      "loss": 0.7187,
+      "step": 1232
+    },
+    {
+      "epoch": 0.029348757497857755,
+      "grad_norm": 3.5255888981970678,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 1233
+    },
+    {
+      "epoch": 0.02937256022088927,
+      "grad_norm": 4.231168195530084,
+      "learning_rate": 2e-05,
+      "loss": 0.2242,
+      "step": 1234
+    },
+    {
+      "epoch": 0.029396362943920784,
+      "grad_norm": 6.614349376105461,
+      "learning_rate": 2e-05,
+      "loss": 1.0078,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0294201656669523,
+      "grad_norm": 6.1384011929384865,
+      "learning_rate": 2e-05,
+      "loss": 0.3253,
+      "step": 1236
+    },
+    {
+      "epoch": 0.029443968389983814,
+      "grad_norm": 11.329706824429161,
+      "learning_rate": 2e-05,
+      "loss": 1.3285,
+      "step": 1237
+    },
+    {
+      "epoch": 0.02946777111301533,
+      "grad_norm": 6.14757778768489,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 1238
+    },
+    {
+      "epoch": 0.029491573836046843,
+      "grad_norm": 8.839462667851862,
+      "learning_rate": 2e-05,
+      "loss": 1.4978,
+      "step": 1239
+    },
+    {
+      "epoch": 0.02951537655907836,
+      "grad_norm": 5.4293271513211705,
+      "learning_rate": 2e-05,
+      "loss": 0.4103,
+      "step": 1240
+    },
+    {
+      "epoch": 0.029539179282109872,
+      "grad_norm": 2.6716434598169863,
+      "learning_rate": 2e-05,
+      "loss": 0.1679,
+      "step": 1241
+    },
+    {
+      "epoch": 0.02956298200514139,
+      "grad_norm": 8.258936590393793,
+      "learning_rate": 2e-05,
+      "loss": 0.3559,
+      "step": 1242
+    },
+    {
+      "epoch": 0.029586784728172902,
+      "grad_norm": 5.01136865846341,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 1243
+    },
+    {
+      "epoch": 0.02961058745120442,
+      "grad_norm": 7.578605613564013,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 1244
+    },
+    {
+      "epoch": 0.02963439017423593,
+      "grad_norm": 3.1992473324781767,
+      "learning_rate": 2e-05,
+      "loss": 0.1924,
+      "step": 1245
+    },
+    {
+      "epoch": 0.029658192897267448,
+      "grad_norm": 4.297617555490742,
+      "learning_rate": 2e-05,
+      "loss": 0.2741,
+      "step": 1246
+    },
+    {
+      "epoch": 0.02968199562029896,
+      "grad_norm": 2.197904238006067,
+      "learning_rate": 2e-05,
+      "loss": 0.1782,
+      "step": 1247
+    },
+    {
+      "epoch": 0.029705798343330477,
+      "grad_norm": 6.300690950475813,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 1248
+    },
+    {
+      "epoch": 0.02972960106636199,
+      "grad_norm": 2.7562122657329233,
+      "learning_rate": 2e-05,
+      "loss": 0.2144,
+      "step": 1249
+    },
+    {
+      "epoch": 0.029753403789393507,
+      "grad_norm": 2.1597810491493963,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 1250
+    },
+    {
+      "epoch": 0.029777206512425023,
+      "grad_norm": 2.7709227202071243,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 1251
+    },
+    {
+      "epoch": 0.029801009235456536,
+      "grad_norm": 1.9592973519144044,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 1252
+    },
+    {
+      "epoch": 0.029824811958488052,
+      "grad_norm": 5.946349000604872,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 1253
+    },
+    {
+      "epoch": 0.029848614681519565,
+      "grad_norm": 8.21835905402584,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 1254
+    },
+    {
+      "epoch": 0.029872417404551082,
+      "grad_norm": 3.8252633184336,
+      "learning_rate": 2e-05,
+      "loss": 0.2833,
+      "step": 1255
+    },
+    {
+      "epoch": 0.029896220127582595,
+      "grad_norm": 8.124910050677988,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 1256
+    },
+    {
+      "epoch": 0.02992002285061411,
+      "grad_norm": 5.1572801388093295,
+      "learning_rate": 2e-05,
+      "loss": 0.3176,
+      "step": 1257
+    },
+    {
+      "epoch": 0.029943825573645624,
+      "grad_norm": 7.191672406080869,
+      "learning_rate": 2e-05,
+      "loss": 0.8491,
+      "step": 1258
+    },
+    {
+      "epoch": 0.02996762829667714,
+      "grad_norm": 4.911632094186031,
+      "learning_rate": 2e-05,
+      "loss": 0.4416,
+      "step": 1259
+    },
+    {
+      "epoch": 0.029991431019708654,
+      "grad_norm": 7.529727244941186,
+      "learning_rate": 2e-05,
+      "loss": 0.8995,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03001523374274017,
+      "grad_norm": 7.774560067235558,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 1261
+    },
+    {
+      "epoch": 0.030039036465771683,
+      "grad_norm": 6.717904731176802,
+      "learning_rate": 2e-05,
+      "loss": 0.3392,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0300628391888032,
+      "grad_norm": 6.615085761705546,
+      "learning_rate": 2e-05,
+      "loss": 0.866,
+      "step": 1263
+    },
+    {
+      "epoch": 0.030086641911834713,
+      "grad_norm": 8.096072294525671,
+      "learning_rate": 2e-05,
+      "loss": 0.3649,
+      "step": 1264
+    },
+    {
+      "epoch": 0.03011044463486623,
+      "grad_norm": 1.730085626008811,
+      "learning_rate": 2e-05,
+      "loss": 0.1863,
+      "step": 1265
+    },
+    {
+      "epoch": 0.030134247357897742,
+      "grad_norm": 5.583392119070744,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 1266
+    },
+    {
+      "epoch": 0.03015805008092926,
+      "grad_norm": 8.651410600142972,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 1267
+    },
+    {
+      "epoch": 0.03018185280396077,
+      "grad_norm": 3.5312742361128113,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 1268
+    },
+    {
+      "epoch": 0.030205655526992288,
+      "grad_norm": 2.716645229234767,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 1269
+    },
+    {
+      "epoch": 0.0302294582500238,
+      "grad_norm": 8.945440117059064,
+      "learning_rate": 2e-05,
+      "loss": 0.5362,
+      "step": 1270
+    },
+    {
+      "epoch": 0.030253260973055317,
+      "grad_norm": 6.025013110998128,
+      "learning_rate": 2e-05,
+      "loss": 0.5702,
+      "step": 1271
+    },
+    {
+      "epoch": 0.030277063696086834,
+      "grad_norm": 6.169125778030104,
+      "learning_rate": 2e-05,
+      "loss": 0.5449,
+      "step": 1272
+    },
+    {
+      "epoch": 0.030300866419118347,
+      "grad_norm": 8.363530888780769,
+      "learning_rate": 2e-05,
+      "loss": 0.9305,
+      "step": 1273
+    },
+    {
+      "epoch": 0.030324669142149863,
+      "grad_norm": 16.34960509306802,
+      "learning_rate": 2e-05,
+      "loss": 0.4869,
+      "step": 1274
+    },
+    {
+      "epoch": 0.030348471865181376,
+      "grad_norm": 7.336929428001511,
+      "learning_rate": 2e-05,
+      "loss": 0.6674,
+      "step": 1275
+    },
+    {
+      "epoch": 0.030372274588212893,
+      "grad_norm": 3.5182706522122134,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 1276
+    },
+    {
+      "epoch": 0.030396077311244406,
+      "grad_norm": 2.9264809524138937,
+      "learning_rate": 2e-05,
+      "loss": 0.211,
+      "step": 1277
+    },
+    {
+      "epoch": 0.030419880034275922,
+      "grad_norm": 11.094735493335588,
+      "learning_rate": 2e-05,
+      "loss": 0.4354,
+      "step": 1278
+    },
+    {
+      "epoch": 0.030443682757307435,
+      "grad_norm": 5.821097739294113,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 1279
+    },
+    {
+      "epoch": 0.03046748548033895,
+      "grad_norm": 5.01048898508687,
+      "learning_rate": 2e-05,
+      "loss": 0.3182,
+      "step": 1280
+    },
+    {
+      "epoch": 0.030491288203370465,
+      "grad_norm": 5.73770091787813,
+      "learning_rate": 2e-05,
+      "loss": 0.7086,
+      "step": 1281
+    },
+    {
+      "epoch": 0.03051509092640198,
+      "grad_norm": 5.188526984001443,
+      "learning_rate": 2e-05,
+      "loss": 0.2211,
+      "step": 1282
+    },
+    {
+      "epoch": 0.030538893649433494,
+      "grad_norm": 6.9054673988536015,
+      "learning_rate": 2e-05,
+      "loss": 1.1125,
+      "step": 1283
+    },
+    {
+      "epoch": 0.03056269637246501,
+      "grad_norm": 3.4214569064475415,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 1284
+    },
+    {
+      "epoch": 0.030586499095496524,
+      "grad_norm": 9.549476875655431,
+      "learning_rate": 2e-05,
+      "loss": 0.7955,
+      "step": 1285
+    },
+    {
+      "epoch": 0.03061030181852804,
+      "grad_norm": 11.42167378038025,
+      "learning_rate": 2e-05,
+      "loss": 0.5324,
+      "step": 1286
+    },
+    {
+      "epoch": 0.030634104541559553,
+      "grad_norm": 6.946341246635333,
+      "learning_rate": 2e-05,
+      "loss": 0.8927,
+      "step": 1287
+    },
+    {
+      "epoch": 0.03065790726459107,
+      "grad_norm": 5.2104059476340385,
+      "learning_rate": 2e-05,
+      "loss": 0.2605,
+      "step": 1288
+    },
+    {
+      "epoch": 0.030681709987622582,
+      "grad_norm": 5.54337955980709,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 1289
+    },
+    {
+      "epoch": 0.0307055127106541,
+      "grad_norm": 5.126623665334584,
+      "learning_rate": 2e-05,
+      "loss": 0.28,
+      "step": 1290
+    },
+    {
+      "epoch": 0.030729315433685615,
+      "grad_norm": 6.716640221708697,
+      "learning_rate": 2e-05,
+      "loss": 0.3052,
+      "step": 1291
+    },
+    {
+      "epoch": 0.03075311815671713,
+      "grad_norm": 5.953535708998535,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 1292
+    },
+    {
+      "epoch": 0.030776920879748645,
+      "grad_norm": 9.89146355175963,
+      "learning_rate": 2e-05,
+      "loss": 0.9595,
+      "step": 1293
+    },
+    {
+      "epoch": 0.030800723602780158,
+      "grad_norm": 9.811187985152227,
+      "learning_rate": 2e-05,
+      "loss": 0.3795,
+      "step": 1294
+    },
+    {
+      "epoch": 0.030824526325811674,
+      "grad_norm": 4.961496506928742,
+      "learning_rate": 2e-05,
+      "loss": 0.6189,
+      "step": 1295
+    },
+    {
+      "epoch": 0.030848329048843187,
+      "grad_norm": 4.58438265124099,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 1296
+    },
+    {
+      "epoch": 0.030872131771874704,
+      "grad_norm": 2.2130094043527944,
+      "learning_rate": 2e-05,
+      "loss": 0.2688,
+      "step": 1297
+    },
+    {
+      "epoch": 0.030895934494906217,
+      "grad_norm": 6.913559143967847,
+      "learning_rate": 2e-05,
+      "loss": 0.7427,
+      "step": 1298
+    },
+    {
+      "epoch": 0.030919737217937733,
+      "grad_norm": 7.1714392018670265,
+      "learning_rate": 2e-05,
+      "loss": 0.6971,
+      "step": 1299
+    },
+    {
+      "epoch": 0.030943539940969246,
+      "grad_norm": 4.145391535742234,
+      "learning_rate": 2e-05,
+      "loss": 0.2508,
+      "step": 1300
+    },
+    {
+      "epoch": 0.030967342664000762,
+      "grad_norm": 3.041239888048388,
+      "learning_rate": 2e-05,
+      "loss": 0.3502,
+      "step": 1301
+    },
+    {
+      "epoch": 0.030991145387032275,
+      "grad_norm": 3.5895949907618827,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 1302
+    },
+    {
+      "epoch": 0.031014948110063792,
+      "grad_norm": 4.103295093045988,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 1303
+    },
+    {
+      "epoch": 0.031038750833095305,
+      "grad_norm": 9.116861262074403,
+      "learning_rate": 2e-05,
+      "loss": 0.3555,
+      "step": 1304
+    },
+    {
+      "epoch": 0.03106255355612682,
+      "grad_norm": 5.164627552960287,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 1305
+    },
+    {
+      "epoch": 0.031086356279158334,
+      "grad_norm": 6.645320046604526,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 1306
+    },
+    {
+      "epoch": 0.03111015900218985,
+      "grad_norm": 10.7322833507533,
+      "learning_rate": 2e-05,
+      "loss": 0.8494,
+      "step": 1307
+    },
+    {
+      "epoch": 0.031133961725221364,
+      "grad_norm": 9.157952762536183,
+      "learning_rate": 2e-05,
+      "loss": 0.6558,
+      "step": 1308
+    },
+    {
+      "epoch": 0.03115776444825288,
+      "grad_norm": 8.727987464344194,
+      "learning_rate": 2e-05,
+      "loss": 1.3172,
+      "step": 1309
+    },
+    {
+      "epoch": 0.031181567171284393,
+      "grad_norm": 5.6715209673834055,
+      "learning_rate": 2e-05,
+      "loss": 0.3365,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03120536989431591,
+      "grad_norm": 15.02034725711573,
+      "learning_rate": 2e-05,
+      "loss": 0.4799,
+      "step": 1311
+    },
+    {
+      "epoch": 0.031229172617347426,
+      "grad_norm": 4.260491865310073,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 1312
+    },
+    {
+      "epoch": 0.03125297534037894,
+      "grad_norm": 6.84307303368819,
+      "learning_rate": 2e-05,
+      "loss": 0.6624,
+      "step": 1313
+    },
+    {
+      "epoch": 0.031276778063410456,
+      "grad_norm": 3.688839555557674,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 1314
+    },
+    {
+      "epoch": 0.03130058078644197,
+      "grad_norm": 4.737406523660888,
+      "learning_rate": 2e-05,
+      "loss": 0.2266,
+      "step": 1315
+    },
+    {
+      "epoch": 0.03132438350947348,
+      "grad_norm": 3.9034189624667266,
+      "learning_rate": 2e-05,
+      "loss": 0.2416,
+      "step": 1316
+    },
+    {
+      "epoch": 0.031348186232505,
+      "grad_norm": 3.7839985430242034,
+      "learning_rate": 2e-05,
+      "loss": 0.2166,
+      "step": 1317
+    },
+    {
+      "epoch": 0.031371988955536514,
+      "grad_norm": 3.382514283846516,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 1318
+    },
+    {
+      "epoch": 0.03139579167856803,
+      "grad_norm": 4.143755648728048,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 1319
+    },
+    {
+      "epoch": 0.03141959440159954,
+      "grad_norm": 6.216415024004827,
+      "learning_rate": 2e-05,
+      "loss": 0.9012,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03144339712463106,
+      "grad_norm": 3.8428319194193463,
+      "learning_rate": 2e-05,
+      "loss": 0.3119,
+      "step": 1321
+    },
+    {
+      "epoch": 0.03146719984766257,
+      "grad_norm": 5.863097707199108,
+      "learning_rate": 2e-05,
+      "loss": 0.7361,
+      "step": 1322
+    },
+    {
+      "epoch": 0.031491002570694086,
+      "grad_norm": 8.396646864730911,
+      "learning_rate": 2e-05,
+      "loss": 1.0662,
+      "step": 1323
+    },
+    {
+      "epoch": 0.0315148052937256,
+      "grad_norm": 5.297262901698231,
+      "learning_rate": 2e-05,
+      "loss": 0.8655,
+      "step": 1324
+    },
+    {
+      "epoch": 0.03153860801675712,
+      "grad_norm": 11.522967305918103,
+      "learning_rate": 2e-05,
+      "loss": 1.9091,
+      "step": 1325
+    },
+    {
+      "epoch": 0.03156241073978863,
+      "grad_norm": 5.914525612691673,
+      "learning_rate": 2e-05,
+      "loss": 0.3633,
+      "step": 1326
+    },
+    {
+      "epoch": 0.031586213462820145,
+      "grad_norm": 4.3492463215109884,
+      "learning_rate": 2e-05,
+      "loss": 0.6067,
+      "step": 1327
+    },
+    {
+      "epoch": 0.03161001618585166,
+      "grad_norm": 5.171940127431177,
+      "learning_rate": 2e-05,
+      "loss": 0.2514,
+      "step": 1328
+    },
+    {
+      "epoch": 0.03163381890888318,
+      "grad_norm": 5.592960020436848,
+      "learning_rate": 2e-05,
+      "loss": 0.718,
+      "step": 1329
+    },
+    {
+      "epoch": 0.03165762163191469,
+      "grad_norm": 5.533049576183051,
+      "learning_rate": 2e-05,
+      "loss": 0.7533,
+      "step": 1330
+    },
+    {
+      "epoch": 0.031681424354946204,
+      "grad_norm": 6.998422106387176,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 1331
+    },
+    {
+      "epoch": 0.031705227077977724,
+      "grad_norm": 7.270513594460998,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 1332
+    },
+    {
+      "epoch": 0.03172902980100924,
+      "grad_norm": 6.511767524119099,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 1333
+    },
+    {
+      "epoch": 0.03175283252404075,
+      "grad_norm": 4.997795173869838,
+      "learning_rate": 2e-05,
+      "loss": 0.2808,
+      "step": 1334
+    },
+    {
+      "epoch": 0.03177663524707226,
+      "grad_norm": 6.8973994402452945,
+      "learning_rate": 2e-05,
+      "loss": 0.7894,
+      "step": 1335
+    },
+    {
+      "epoch": 0.03180043797010378,
+      "grad_norm": 8.774216009394532,
+      "learning_rate": 2e-05,
+      "loss": 0.7187,
+      "step": 1336
+    },
+    {
+      "epoch": 0.031824240693135296,
+      "grad_norm": 4.635004181854665,
+      "learning_rate": 2e-05,
+      "loss": 0.465,
+      "step": 1337
+    },
+    {
+      "epoch": 0.03184804341616681,
+      "grad_norm": 9.596257715329225,
+      "learning_rate": 2e-05,
+      "loss": 0.9103,
+      "step": 1338
+    },
+    {
+      "epoch": 0.03187184613919832,
+      "grad_norm": 12.985665917363946,
+      "learning_rate": 2e-05,
+      "loss": 0.418,
+      "step": 1339
+    },
+    {
+      "epoch": 0.03189564886222984,
+      "grad_norm": 7.296227170060263,
+      "learning_rate": 2e-05,
+      "loss": 0.6207,
+      "step": 1340
+    },
+    {
+      "epoch": 0.031919451585261355,
+      "grad_norm": 2.6769185353561182,
+      "learning_rate": 2e-05,
+      "loss": 0.1775,
+      "step": 1341
+    },
+    {
+      "epoch": 0.03194325430829287,
+      "grad_norm": 7.7987064027709865,
+      "learning_rate": 2e-05,
+      "loss": 0.8854,
+      "step": 1342
+    },
+    {
+      "epoch": 0.03196705703132438,
+      "grad_norm": 10.032349959147824,
+      "learning_rate": 2e-05,
+      "loss": 0.457,
+      "step": 1343
+    },
+    {
+      "epoch": 0.0319908597543559,
+      "grad_norm": 2.7422596585659225,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 1344
+    },
+    {
+      "epoch": 0.032014662477387414,
+      "grad_norm": 6.696800809205331,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 1345
+    },
+    {
+      "epoch": 0.03203846520041893,
+      "grad_norm": 7.143593171810307,
+      "learning_rate": 2e-05,
+      "loss": 1.0855,
+      "step": 1346
+    },
+    {
+      "epoch": 0.03206226792345044,
+      "grad_norm": 3.8483819424147145,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 1347
+    },
+    {
+      "epoch": 0.03208607064648196,
+      "grad_norm": 6.889826709008769,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 1348
+    },
+    {
+      "epoch": 0.03210987336951347,
+      "grad_norm": 7.241009075581849,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 1349
+    },
+    {
+      "epoch": 0.032133676092544985,
+      "grad_norm": 3.907079251452381,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 1350
+    },
+    {
+      "epoch": 0.032157478815576505,
+      "grad_norm": 9.831099918815351,
+      "learning_rate": 2e-05,
+      "loss": 1.1081,
+      "step": 1351
+    },
+    {
+      "epoch": 0.03218128153860802,
+      "grad_norm": 8.760225719533066,
+      "learning_rate": 2e-05,
+      "loss": 0.4304,
+      "step": 1352
+    },
+    {
+      "epoch": 0.03220508426163953,
+      "grad_norm": 11.128347181888122,
+      "learning_rate": 2e-05,
+      "loss": 0.4147,
+      "step": 1353
+    },
+    {
+      "epoch": 0.032228886984671044,
+      "grad_norm": 9.95734767032855,
+      "learning_rate": 2e-05,
+      "loss": 0.5087,
+      "step": 1354
+    },
+    {
+      "epoch": 0.032252689707702564,
+      "grad_norm": 2.5270755704994516,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 1355
+    },
+    {
+      "epoch": 0.03227649243073408,
+      "grad_norm": 8.014597237936172,
+      "learning_rate": 2e-05,
+      "loss": 0.7976,
+      "step": 1356
+    },
+    {
+      "epoch": 0.03230029515376559,
+      "grad_norm": 3.9974986502492817,
+      "learning_rate": 2e-05,
+      "loss": 0.1811,
+      "step": 1357
+    },
+    {
+      "epoch": 0.0323240978767971,
+      "grad_norm": 2.7728159316583296,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 1358
+    },
+    {
+      "epoch": 0.03234790059982862,
+      "grad_norm": 4.438998590251541,
+      "learning_rate": 2e-05,
+      "loss": 0.5478,
+      "step": 1359
+    },
+    {
+      "epoch": 0.032371703322860136,
+      "grad_norm": 3.768846490134993,
+      "learning_rate": 2e-05,
+      "loss": 0.2828,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03239550604589165,
+      "grad_norm": 3.8342922870310043,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 1361
+    },
+    {
+      "epoch": 0.03241930876892316,
+      "grad_norm": 6.596770292394853,
+      "learning_rate": 2e-05,
+      "loss": 0.908,
+      "step": 1362
+    },
+    {
+      "epoch": 0.03244311149195468,
+      "grad_norm": 2.1638523370300375,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 1363
+    },
+    {
+      "epoch": 0.032466914214986195,
+      "grad_norm": 2.7358436517210696,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 1364
+    },
+    {
+      "epoch": 0.03249071693801771,
+      "grad_norm": 2.7901681024821947,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 1365
+    },
+    {
+      "epoch": 0.03251451966104922,
+      "grad_norm": 7.695798647237689,
+      "learning_rate": 2e-05,
+      "loss": 0.6605,
+      "step": 1366
+    },
+    {
+      "epoch": 0.03253832238408074,
+      "grad_norm": 3.5138149839108492,
+      "learning_rate": 2e-05,
+      "loss": 0.3026,
+      "step": 1367
+    },
+    {
+      "epoch": 0.032562125107112254,
+      "grad_norm": 9.363656862324786,
+      "learning_rate": 2e-05,
+      "loss": 1.1417,
+      "step": 1368
+    },
+    {
+      "epoch": 0.03258592783014377,
+      "grad_norm": 6.092485976265792,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 1369
+    },
+    {
+      "epoch": 0.03260973055317528,
+      "grad_norm": 4.152545479203837,
+      "learning_rate": 2e-05,
+      "loss": 0.3417,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0326335332762068,
+      "grad_norm": 5.200689216683012,
+      "learning_rate": 2e-05,
+      "loss": 0.2392,
+      "step": 1371
+    },
+    {
+      "epoch": 0.03265733599923831,
+      "grad_norm": 7.281895927045704,
+      "learning_rate": 2e-05,
+      "loss": 1.0257,
+      "step": 1372
+    },
+    {
+      "epoch": 0.032681138722269826,
+      "grad_norm": 8.485544443159704,
+      "learning_rate": 2e-05,
+      "loss": 1.2017,
+      "step": 1373
+    },
+    {
+      "epoch": 0.032704941445301346,
+      "grad_norm": 12.563467462342729,
+      "learning_rate": 2e-05,
+      "loss": 0.5001,
+      "step": 1374
+    },
+    {
+      "epoch": 0.03272874416833286,
+      "grad_norm": 7.677915424971844,
+      "learning_rate": 2e-05,
+      "loss": 0.8096,
+      "step": 1375
+    },
+    {
+      "epoch": 0.03275254689136437,
+      "grad_norm": 8.465150832761397,
+      "learning_rate": 2e-05,
+      "loss": 1.1383,
+      "step": 1376
+    },
+    {
+      "epoch": 0.032776349614395885,
+      "grad_norm": 6.144879104412188,
+      "learning_rate": 2e-05,
+      "loss": 0.3308,
+      "step": 1377
+    },
+    {
+      "epoch": 0.032800152337427405,
+      "grad_norm": 3.380772858090413,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 1378
+    },
+    {
+      "epoch": 0.03282395506045892,
+      "grad_norm": 2.4287872890456748,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 1379
+    },
+    {
+      "epoch": 0.03284775778349043,
+      "grad_norm": 9.139482339703369,
+      "learning_rate": 2e-05,
+      "loss": 0.9726,
+      "step": 1380
+    },
+    {
+      "epoch": 0.032871560506521944,
+      "grad_norm": 6.09659225657244,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 1381
+    },
+    {
+      "epoch": 0.03289536322955346,
+      "grad_norm": 8.759434522099953,
+      "learning_rate": 2e-05,
+      "loss": 0.9683,
+      "step": 1382
+    },
+    {
+      "epoch": 0.032919165952584976,
+      "grad_norm": 1.7962801196830844,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 1383
+    },
+    {
+      "epoch": 0.03294296867561649,
+      "grad_norm": 7.980594679635942,
+      "learning_rate": 2e-05,
+      "loss": 1.2702,
+      "step": 1384
+    },
+    {
+      "epoch": 0.032966771398648,
+      "grad_norm": 5.832834769700762,
+      "learning_rate": 2e-05,
+      "loss": 0.7267,
+      "step": 1385
+    },
+    {
+      "epoch": 0.03299057412167952,
+      "grad_norm": 4.5505223767988,
+      "learning_rate": 2e-05,
+      "loss": 0.4437,
+      "step": 1386
+    },
+    {
+      "epoch": 0.033014376844711035,
+      "grad_norm": 5.298713325416716,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 1387
+    },
+    {
+      "epoch": 0.03303817956774255,
+      "grad_norm": 6.716266046116549,
+      "learning_rate": 2e-05,
+      "loss": 0.7516,
+      "step": 1388
+    },
+    {
+      "epoch": 0.03306198229077406,
+      "grad_norm": 5.315494380600393,
+      "learning_rate": 2e-05,
+      "loss": 0.429,
+      "step": 1389
+    },
+    {
+      "epoch": 0.03308578501380558,
+      "grad_norm": 10.399883410391157,
+      "learning_rate": 2e-05,
+      "loss": 0.6883,
+      "step": 1390
+    },
+    {
+      "epoch": 0.033109587736837094,
+      "grad_norm": 6.243087913517175,
+      "learning_rate": 2e-05,
+      "loss": 0.3208,
+      "step": 1391
+    },
+    {
+      "epoch": 0.03313339045986861,
+      "grad_norm": 7.289385756845123,
+      "learning_rate": 2e-05,
+      "loss": 1.6506,
+      "step": 1392
+    },
+    {
+      "epoch": 0.03315719318290013,
+      "grad_norm": 4.449363000529249,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 1393
+    },
+    {
+      "epoch": 0.03318099590593164,
+      "grad_norm": 6.971182572631676,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 1394
+    },
+    {
+      "epoch": 0.03320479862896315,
+      "grad_norm": 6.852426564165775,
+      "learning_rate": 2e-05,
+      "loss": 0.7318,
+      "step": 1395
+    },
+    {
+      "epoch": 0.033228601351994666,
+      "grad_norm": 3.9440836867741766,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 1396
+    },
+    {
+      "epoch": 0.033252404075026186,
+      "grad_norm": 9.971333725451297,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 1397
+    },
+    {
+      "epoch": 0.0332762067980577,
+      "grad_norm": 3.274098701151625,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 1398
+    },
+    {
+      "epoch": 0.03330000952108921,
+      "grad_norm": 8.383800392041513,
+      "learning_rate": 2e-05,
+      "loss": 0.5036,
+      "step": 1399
+    },
+    {
+      "epoch": 0.033323812244120725,
+      "grad_norm": 3.4748825590201724,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 1400
+    },
+    {
+      "epoch": 0.033347614967152245,
+      "grad_norm": 7.437314340177959,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 1401
+    },
+    {
+      "epoch": 0.03337141769018376,
+      "grad_norm": 4.740759551500708,
+      "learning_rate": 2e-05,
+      "loss": 0.2168,
+      "step": 1402
+    },
+    {
+      "epoch": 0.03339522041321527,
+      "grad_norm": 7.9606763873730655,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 1403
+    },
+    {
+      "epoch": 0.033419023136246784,
+      "grad_norm": 8.611245924486266,
+      "learning_rate": 2e-05,
+      "loss": 0.875,
+      "step": 1404
+    },
+    {
+      "epoch": 0.033442825859278304,
+      "grad_norm": 2.55931143238468,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 1405
+    },
+    {
+      "epoch": 0.03346662858230982,
+      "grad_norm": 8.175668540109916,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 1406
+    },
+    {
+      "epoch": 0.03349043130534133,
+      "grad_norm": 5.158493288122953,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 1407
+    },
+    {
+      "epoch": 0.03351423402837284,
+      "grad_norm": 2.8703269581544664,
+      "learning_rate": 2e-05,
+      "loss": 0.2408,
+      "step": 1408
+    },
+    {
+      "epoch": 0.03353803675140436,
+      "grad_norm": 5.118672378205785,
+      "learning_rate": 2e-05,
+      "loss": 0.5999,
+      "step": 1409
+    },
+    {
+      "epoch": 0.033561839474435876,
+      "grad_norm": 5.355041029776837,
+      "learning_rate": 2e-05,
+      "loss": 0.8147,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03358564219746739,
+      "grad_norm": 5.594275636545165,
+      "learning_rate": 2e-05,
+      "loss": 0.3973,
+      "step": 1411
+    },
+    {
+      "epoch": 0.03360944492049891,
+      "grad_norm": 6.983441000125302,
+      "learning_rate": 2e-05,
+      "loss": 0.3488,
+      "step": 1412
+    },
+    {
+      "epoch": 0.03363324764353042,
+      "grad_norm": 1.4474940136455527,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 1413
+    },
+    {
+      "epoch": 0.033657050366561934,
+      "grad_norm": 6.367962103234294,
+      "learning_rate": 2e-05,
+      "loss": 0.781,
+      "step": 1414
+    },
+    {
+      "epoch": 0.03368085308959345,
+      "grad_norm": 3.0692265601612556,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 1415
+    },
+    {
+      "epoch": 0.03370465581262497,
+      "grad_norm": 7.574809274272461,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 1416
+    },
+    {
+      "epoch": 0.03372845853565648,
+      "grad_norm": 3.74742065326213,
+      "learning_rate": 2e-05,
+      "loss": 0.3138,
+      "step": 1417
+    },
+    {
+      "epoch": 0.03375226125868799,
+      "grad_norm": 2.259206802209456,
+      "learning_rate": 2e-05,
+      "loss": 0.1986,
+      "step": 1418
+    },
+    {
+      "epoch": 0.033776063981719506,
+      "grad_norm": 11.550725023511703,
+      "learning_rate": 2e-05,
+      "loss": 0.5639,
+      "step": 1419
+    },
+    {
+      "epoch": 0.033799866704751026,
+      "grad_norm": 7.61645787559109,
+      "learning_rate": 2e-05,
+      "loss": 1.5751,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03382366942778254,
+      "grad_norm": 5.357580515956322,
+      "learning_rate": 2e-05,
+      "loss": 0.4505,
+      "step": 1421
+    },
+    {
+      "epoch": 0.03384747215081405,
+      "grad_norm": 3.2031342148907083,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 1422
+    },
+    {
+      "epoch": 0.033871274873845565,
+      "grad_norm": 2.900974651855979,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 1423
+    },
+    {
+      "epoch": 0.033895077596877085,
+      "grad_norm": 3.220182805646375,
+      "learning_rate": 2e-05,
+      "loss": 0.3076,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0339188803199086,
+      "grad_norm": 2.2713176848053647,
+      "learning_rate": 2e-05,
+      "loss": 0.2561,
+      "step": 1425
+    },
+    {
+      "epoch": 0.03394268304294011,
+      "grad_norm": 6.310122687503284,
+      "learning_rate": 2e-05,
+      "loss": 0.4054,
+      "step": 1426
+    },
+    {
+      "epoch": 0.033966485765971624,
+      "grad_norm": 15.181621017862602,
+      "learning_rate": 2e-05,
+      "loss": 1.3193,
+      "step": 1427
+    },
+    {
+      "epoch": 0.033990288489003144,
+      "grad_norm": 3.5636893294768166,
+      "learning_rate": 2e-05,
+      "loss": 0.2335,
+      "step": 1428
+    },
+    {
+      "epoch": 0.03401409121203466,
+      "grad_norm": 7.578493308668005,
+      "learning_rate": 2e-05,
+      "loss": 0.2982,
+      "step": 1429
+    },
+    {
+      "epoch": 0.03403789393506617,
+      "grad_norm": 5.56285898749209,
+      "learning_rate": 2e-05,
+      "loss": 0.3338,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03406169665809768,
+      "grad_norm": 5.742774437768417,
+      "learning_rate": 2e-05,
+      "loss": 0.5051,
+      "step": 1431
+    },
+    {
+      "epoch": 0.0340854993811292,
+      "grad_norm": 6.89848097143806,
+      "learning_rate": 2e-05,
+      "loss": 0.3371,
+      "step": 1432
+    },
+    {
+      "epoch": 0.034109302104160716,
+      "grad_norm": 10.314098410379172,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 1433
+    },
+    {
+      "epoch": 0.03413310482719223,
+      "grad_norm": 3.0577896363239656,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 1434
+    },
+    {
+      "epoch": 0.03415690755022375,
+      "grad_norm": 1.2506330942010997,
+      "learning_rate": 2e-05,
+      "loss": 0.2016,
+      "step": 1435
+    },
+    {
+      "epoch": 0.03418071027325526,
+      "grad_norm": 6.898255218300624,
+      "learning_rate": 2e-05,
+      "loss": 0.3285,
+      "step": 1436
+    },
+    {
+      "epoch": 0.034204512996286775,
+      "grad_norm": 14.742591231337999,
+      "learning_rate": 2e-05,
+      "loss": 0.8858,
+      "step": 1437
+    },
+    {
+      "epoch": 0.03422831571931829,
+      "grad_norm": 4.861284473220486,
+      "learning_rate": 2e-05,
+      "loss": 0.4704,
+      "step": 1438
+    },
+    {
+      "epoch": 0.03425211844234981,
+      "grad_norm": 4.780191873981574,
+      "learning_rate": 2e-05,
+      "loss": 0.379,
+      "step": 1439
+    },
+    {
+      "epoch": 0.03427592116538132,
+      "grad_norm": 8.786432053741994,
+      "learning_rate": 2e-05,
+      "loss": 0.3367,
+      "step": 1440
+    },
+    {
+      "epoch": 0.034299723888412834,
+      "grad_norm": 2.742138097232108,
+      "learning_rate": 2e-05,
+      "loss": 0.2324,
+      "step": 1441
+    },
+    {
+      "epoch": 0.03432352661144435,
+      "grad_norm": 3.8428802301096256,
+      "learning_rate": 2e-05,
+      "loss": 0.246,
+      "step": 1442
+    },
+    {
+      "epoch": 0.034347329334475867,
+      "grad_norm": 9.590055931548292,
+      "learning_rate": 2e-05,
+      "loss": 1.0575,
+      "step": 1443
+    },
+    {
+      "epoch": 0.03437113205750738,
+      "grad_norm": 10.007898721768507,
+      "learning_rate": 2e-05,
+      "loss": 1.2095,
+      "step": 1444
+    },
+    {
+      "epoch": 0.03439493478053889,
+      "grad_norm": 3.709100363102957,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 1445
+    },
+    {
+      "epoch": 0.034418737503570405,
+      "grad_norm": 8.494752446536104,
+      "learning_rate": 2e-05,
+      "loss": 0.4434,
+      "step": 1446
+    },
+    {
+      "epoch": 0.034442540226601925,
+      "grad_norm": 3.98185414986966,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 1447
+    },
+    {
+      "epoch": 0.03446634294963344,
+      "grad_norm": 3.7192390223741247,
+      "learning_rate": 2e-05,
+      "loss": 0.2631,
+      "step": 1448
+    },
+    {
+      "epoch": 0.03449014567266495,
+      "grad_norm": 8.546243217957894,
+      "learning_rate": 2e-05,
+      "loss": 0.3435,
+      "step": 1449
+    },
+    {
+      "epoch": 0.034513948395696464,
+      "grad_norm": 4.734976298287275,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 1450
+    },
+    {
+      "epoch": 0.034537751118727984,
+      "grad_norm": 9.69900667701259,
+      "learning_rate": 2e-05,
+      "loss": 1.5197,
+      "step": 1451
+    },
+    {
+      "epoch": 0.0345615538417595,
+      "grad_norm": 6.20874276941405,
+      "learning_rate": 2e-05,
+      "loss": 0.3574,
+      "step": 1452
+    },
+    {
+      "epoch": 0.03458535656479101,
+      "grad_norm": 3.9791650765550375,
+      "learning_rate": 2e-05,
+      "loss": 0.2951,
+      "step": 1453
+    },
+    {
+      "epoch": 0.03460915928782253,
+      "grad_norm": 3.622052464759086,
+      "learning_rate": 2e-05,
+      "loss": 0.2697,
+      "step": 1454
+    },
+    {
+      "epoch": 0.03463296201085404,
+      "grad_norm": 7.800214514960112,
+      "learning_rate": 2e-05,
+      "loss": 0.9466,
+      "step": 1455
+    },
+    {
+      "epoch": 0.034656764733885556,
+      "grad_norm": 5.249714279686637,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 1456
+    },
+    {
+      "epoch": 0.03468056745691707,
+      "grad_norm": 5.325542971298025,
+      "learning_rate": 2e-05,
+      "loss": 0.6988,
+      "step": 1457
+    },
+    {
+      "epoch": 0.03470437017994859,
+      "grad_norm": 2.9386481603834924,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 1458
+    },
+    {
+      "epoch": 0.0347281729029801,
+      "grad_norm": 2.9673529042738402,
+      "learning_rate": 2e-05,
+      "loss": 0.2222,
+      "step": 1459
+    },
+    {
+      "epoch": 0.034751975626011615,
+      "grad_norm": 9.22616590362787,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03477577834904313,
+      "grad_norm": 5.112286390619808,
+      "learning_rate": 2e-05,
+      "loss": 0.2311,
+      "step": 1461
+    },
+    {
+      "epoch": 0.03479958107207465,
+      "grad_norm": 2.445092084301256,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 1462
+    },
+    {
+      "epoch": 0.03482338379510616,
+      "grad_norm": 8.717179018031333,
+      "learning_rate": 2e-05,
+      "loss": 0.4484,
+      "step": 1463
+    },
+    {
+      "epoch": 0.034847186518137674,
+      "grad_norm": 8.061089309026913,
+      "learning_rate": 2e-05,
+      "loss": 1.0537,
+      "step": 1464
+    },
+    {
+      "epoch": 0.03487098924116919,
+      "grad_norm": 2.3378744412662664,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 1465
+    },
+    {
+      "epoch": 0.03489479196420071,
+      "grad_norm": 3.6814010206949743,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 1466
+    },
+    {
+      "epoch": 0.03491859468723222,
+      "grad_norm": 4.471002053861289,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 1467
+    },
+    {
+      "epoch": 0.03494239741026373,
+      "grad_norm": 3.0823724113809594,
+      "learning_rate": 2e-05,
+      "loss": 0.1738,
+      "step": 1468
+    },
+    {
+      "epoch": 0.034966200133295246,
+      "grad_norm": 3.2172839779174405,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 1469
+    },
+    {
+      "epoch": 0.034990002856326766,
+      "grad_norm": 8.234285539709962,
+      "learning_rate": 2e-05,
+      "loss": 0.3082,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03501380557935828,
+      "grad_norm": 3.602815901458839,
+      "learning_rate": 2e-05,
+      "loss": 0.1625,
+      "step": 1471
+    },
+    {
+      "epoch": 0.03503760830238979,
+      "grad_norm": 2.483077942705756,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 1472
+    },
+    {
+      "epoch": 0.03506141102542131,
+      "grad_norm": 5.986505722721782,
+      "learning_rate": 2e-05,
+      "loss": 0.7304,
+      "step": 1473
+    },
+    {
+      "epoch": 0.035085213748452825,
+      "grad_norm": 6.531844303003439,
+      "learning_rate": 2e-05,
+      "loss": 1.0628,
+      "step": 1474
+    },
+    {
+      "epoch": 0.03510901647148434,
+      "grad_norm": 2.5455963773541463,
+      "learning_rate": 2e-05,
+      "loss": 0.2002,
+      "step": 1475
+    },
+    {
+      "epoch": 0.03513281919451585,
+      "grad_norm": 3.5787638805189808,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 1476
+    },
+    {
+      "epoch": 0.03515662191754737,
+      "grad_norm": 7.932811437794945,
+      "learning_rate": 2e-05,
+      "loss": 0.4179,
+      "step": 1477
+    },
+    {
+      "epoch": 0.03518042464057888,
+      "grad_norm": 9.901644102848298,
+      "learning_rate": 2e-05,
+      "loss": 0.8667,
+      "step": 1478
+    },
+    {
+      "epoch": 0.035204227363610396,
+      "grad_norm": 11.630514822293076,
+      "learning_rate": 2e-05,
+      "loss": 1.3166,
+      "step": 1479
+    },
+    {
+      "epoch": 0.03522803008664191,
+      "grad_norm": 3.8910673414282493,
+      "learning_rate": 2e-05,
+      "loss": 0.2909,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03525183280967343,
+      "grad_norm": 8.355221636916031,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 1481
+    },
+    {
+      "epoch": 0.03527563553270494,
+      "grad_norm": 6.033841335370828,
+      "learning_rate": 2e-05,
+      "loss": 0.3455,
+      "step": 1482
+    },
+    {
+      "epoch": 0.035299438255736455,
+      "grad_norm": 2.30211366941439,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 1483
+    },
+    {
+      "epoch": 0.03532324097876797,
+      "grad_norm": 7.566453334138618,
+      "learning_rate": 2e-05,
+      "loss": 0.5317,
+      "step": 1484
+    },
+    {
+      "epoch": 0.03534704370179949,
+      "grad_norm": 12.938088818962484,
+      "learning_rate": 2e-05,
+      "loss": 1.6136,
+      "step": 1485
+    },
+    {
+      "epoch": 0.035370846424831,
+      "grad_norm": 10.169400905431576,
+      "learning_rate": 2e-05,
+      "loss": 0.9676,
+      "step": 1486
+    },
+    {
+      "epoch": 0.035394649147862514,
+      "grad_norm": 7.3468783591582625,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 1487
+    },
+    {
+      "epoch": 0.03541845187089403,
+      "grad_norm": 4.875405612432816,
+      "learning_rate": 2e-05,
+      "loss": 0.6434,
+      "step": 1488
+    },
+    {
+      "epoch": 0.03544225459392555,
+      "grad_norm": 9.811974788808566,
+      "learning_rate": 2e-05,
+      "loss": 1.2657,
+      "step": 1489
+    },
+    {
+      "epoch": 0.03546605731695706,
+      "grad_norm": 6.9497074331397934,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03548986003998857,
+      "grad_norm": 5.357650107337727,
+      "learning_rate": 2e-05,
+      "loss": 0.7555,
+      "step": 1491
+    },
+    {
+      "epoch": 0.03551366276302009,
+      "grad_norm": 3.1264213212635648,
+      "learning_rate": 2e-05,
+      "loss": 0.206,
+      "step": 1492
+    },
+    {
+      "epoch": 0.035537465486051606,
+      "grad_norm": 2.3226118604907238,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 1493
+    },
+    {
+      "epoch": 0.03556126820908312,
+      "grad_norm": 3.4583836731244313,
+      "learning_rate": 2e-05,
+      "loss": 0.2674,
+      "step": 1494
+    },
+    {
+      "epoch": 0.03558507093211463,
+      "grad_norm": 5.647751432282822,
+      "learning_rate": 2e-05,
+      "loss": 1.0097,
+      "step": 1495
+    },
+    {
+      "epoch": 0.03560887365514615,
+      "grad_norm": 3.4132358033072956,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 1496
+    },
+    {
+      "epoch": 0.035632676378177665,
+      "grad_norm": 10.150589403747428,
+      "learning_rate": 2e-05,
+      "loss": 0.8442,
+      "step": 1497
+    },
+    {
+      "epoch": 0.03565647910120918,
+      "grad_norm": 6.020071583780252,
+      "learning_rate": 2e-05,
+      "loss": 0.3395,
+      "step": 1498
+    },
+    {
+      "epoch": 0.03568028182424069,
+      "grad_norm": 5.5122615490176035,
+      "learning_rate": 2e-05,
+      "loss": 0.2369,
+      "step": 1499
+    },
+    {
+      "epoch": 0.03570408454727221,
+      "grad_norm": 5.2892002365566615,
+      "learning_rate": 2e-05,
+      "loss": 0.8748,
+      "step": 1500
+    },
+    {
+      "epoch": 0.035727887270303724,
+      "grad_norm": 8.026507627006582,
+      "learning_rate": 2e-05,
+      "loss": 0.2993,
+      "step": 1501
+    },
+    {
+      "epoch": 0.03575168999333524,
+      "grad_norm": 7.969253155249404,
+      "learning_rate": 2e-05,
+      "loss": 1.2759,
+      "step": 1502
+    },
+    {
+      "epoch": 0.03577549271636675,
+      "grad_norm": 8.03016572459769,
+      "learning_rate": 2e-05,
+      "loss": 1.0057,
+      "step": 1503
+    },
+    {
+      "epoch": 0.03579929543939827,
+      "grad_norm": 7.469727765539878,
+      "learning_rate": 2e-05,
+      "loss": 0.3052,
+      "step": 1504
+    },
+    {
+      "epoch": 0.03582309816242978,
+      "grad_norm": 4.6702702379298335,
+      "learning_rate": 2e-05,
+      "loss": 0.3298,
+      "step": 1505
+    },
+    {
+      "epoch": 0.035846900885461296,
+      "grad_norm": 3.572138159732307,
+      "learning_rate": 2e-05,
+      "loss": 0.2947,
+      "step": 1506
+    },
+    {
+      "epoch": 0.03587070360849281,
+      "grad_norm": 3.131472168650841,
+      "learning_rate": 2e-05,
+      "loss": 0.1607,
+      "step": 1507
+    },
+    {
+      "epoch": 0.03589450633152433,
+      "grad_norm": 3.870048397958459,
+      "learning_rate": 2e-05,
+      "loss": 0.2325,
+      "step": 1508
+    },
+    {
+      "epoch": 0.03591830905455584,
+      "grad_norm": 6.55052336379481,
+      "learning_rate": 2e-05,
+      "loss": 0.3746,
+      "step": 1509
+    },
+    {
+      "epoch": 0.035942111777587354,
+      "grad_norm": 2.904692331076395,
+      "learning_rate": 2e-05,
+      "loss": 0.2028,
+      "step": 1510
+    },
+    {
+      "epoch": 0.03596591450061887,
+      "grad_norm": 9.5855711010454,
+      "learning_rate": 2e-05,
+      "loss": 0.5368,
+      "step": 1511
+    },
+    {
+      "epoch": 0.03598971722365039,
+      "grad_norm": 8.812001788582931,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 1512
+    },
+    {
+      "epoch": 0.0360135199466819,
+      "grad_norm": 5.108227696645435,
+      "learning_rate": 2e-05,
+      "loss": 0.2609,
+      "step": 1513
+    },
+    {
+      "epoch": 0.03603732266971341,
+      "grad_norm": 3.8343953339233003,
+      "learning_rate": 2e-05,
+      "loss": 0.2689,
+      "step": 1514
+    },
+    {
+      "epoch": 0.03606112539274493,
+      "grad_norm": 3.603021828655885,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 1515
+    },
+    {
+      "epoch": 0.036084928115776446,
+      "grad_norm": 6.969776543371248,
+      "learning_rate": 2e-05,
+      "loss": 0.627,
+      "step": 1516
+    },
+    {
+      "epoch": 0.03610873083880796,
+      "grad_norm": 7.4208988598037235,
+      "learning_rate": 2e-05,
+      "loss": 0.9488,
+      "step": 1517
+    },
+    {
+      "epoch": 0.03613253356183947,
+      "grad_norm": 5.720799576711903,
+      "learning_rate": 2e-05,
+      "loss": 0.8092,
+      "step": 1518
+    },
+    {
+      "epoch": 0.03615633628487099,
+      "grad_norm": 1.8561477480534772,
+      "learning_rate": 2e-05,
+      "loss": 0.1896,
+      "step": 1519
+    },
+    {
+      "epoch": 0.036180139007902505,
+      "grad_norm": 9.78883056889843,
+      "learning_rate": 2e-05,
+      "loss": 1.6195,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03620394173093402,
+      "grad_norm": 5.853037707901123,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 1521
+    },
+    {
+      "epoch": 0.03622774445396553,
+      "grad_norm": 2.242383855628725,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 1522
+    },
+    {
+      "epoch": 0.03625154717699705,
+      "grad_norm": 6.752073666470255,
+      "learning_rate": 2e-05,
+      "loss": 0.362,
+      "step": 1523
+    },
+    {
+      "epoch": 0.036275349900028564,
+      "grad_norm": 7.653133799015589,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 1524
+    },
+    {
+      "epoch": 0.03629915262306008,
+      "grad_norm": 11.530204084667778,
+      "learning_rate": 2e-05,
+      "loss": 1.2745,
+      "step": 1525
+    },
+    {
+      "epoch": 0.03632295534609159,
+      "grad_norm": 8.05230113634637,
+      "learning_rate": 2e-05,
+      "loss": 1.6311,
+      "step": 1526
+    },
+    {
+      "epoch": 0.03634675806912311,
+      "grad_norm": 12.105788247182751,
+      "learning_rate": 2e-05,
+      "loss": 0.9156,
+      "step": 1527
+    },
+    {
+      "epoch": 0.03637056079215462,
+      "grad_norm": 4.576995430713535,
+      "learning_rate": 2e-05,
+      "loss": 0.3881,
+      "step": 1528
+    },
+    {
+      "epoch": 0.036394363515186136,
+      "grad_norm": 8.534963159737412,
+      "learning_rate": 2e-05,
+      "loss": 1.2088,
+      "step": 1529
+    },
+    {
+      "epoch": 0.03641816623821765,
+      "grad_norm": 4.42794018365918,
+      "learning_rate": 2e-05,
+      "loss": 0.2839,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03644196896124917,
+      "grad_norm": 7.439531490866485,
+      "learning_rate": 2e-05,
+      "loss": 0.7721,
+      "step": 1531
+    },
+    {
+      "epoch": 0.03646577168428068,
+      "grad_norm": 9.376507173964233,
+      "learning_rate": 2e-05,
+      "loss": 0.3534,
+      "step": 1532
+    },
+    {
+      "epoch": 0.036489574407312195,
+      "grad_norm": 8.94559128743708,
+      "learning_rate": 2e-05,
+      "loss": 0.5181,
+      "step": 1533
+    },
+    {
+      "epoch": 0.036513377130343715,
+      "grad_norm": 9.193876830739034,
+      "learning_rate": 2e-05,
+      "loss": 0.4127,
+      "step": 1534
+    },
+    {
+      "epoch": 0.03653717985337523,
+      "grad_norm": 2.363054409722164,
+      "learning_rate": 2e-05,
+      "loss": 0.1974,
+      "step": 1535
+    },
+    {
+      "epoch": 0.03656098257640674,
+      "grad_norm": 4.2032326392701505,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 1536
+    },
+    {
+      "epoch": 0.036584785299438254,
+      "grad_norm": 8.698052849578554,
+      "learning_rate": 2e-05,
+      "loss": 0.708,
+      "step": 1537
+    },
+    {
+      "epoch": 0.036608588022469774,
+      "grad_norm": 6.506101166198766,
+      "learning_rate": 2e-05,
+      "loss": 0.6844,
+      "step": 1538
+    },
+    {
+      "epoch": 0.036632390745501286,
+      "grad_norm": 3.826438743496463,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 1539
+    },
+    {
+      "epoch": 0.0366561934685328,
+      "grad_norm": 3.852925418672768,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03667999619156431,
+      "grad_norm": 7.502530257952218,
+      "learning_rate": 2e-05,
+      "loss": 1.3737,
+      "step": 1541
+    },
+    {
+      "epoch": 0.03670379891459583,
+      "grad_norm": 7.255040090407585,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 1542
+    },
+    {
+      "epoch": 0.036727601637627345,
+      "grad_norm": 5.584205667996825,
+      "learning_rate": 2e-05,
+      "loss": 0.2741,
+      "step": 1543
+    },
+    {
+      "epoch": 0.03675140436065886,
+      "grad_norm": 2.9627551276637614,
+      "learning_rate": 2e-05,
+      "loss": 0.2154,
+      "step": 1544
+    },
+    {
+      "epoch": 0.03677520708369037,
+      "grad_norm": 9.588093462394156,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 1545
+    },
+    {
+      "epoch": 0.03679900980672189,
+      "grad_norm": 2.9683122145584844,
+      "learning_rate": 2e-05,
+      "loss": 0.1625,
+      "step": 1546
+    },
+    {
+      "epoch": 0.036822812529753404,
+      "grad_norm": 6.931542166715832,
+      "learning_rate": 2e-05,
+      "loss": 0.8387,
+      "step": 1547
+    },
+    {
+      "epoch": 0.03684661525278492,
+      "grad_norm": 2.1597519997288006,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 1548
+    },
+    {
+      "epoch": 0.03687041797581643,
+      "grad_norm": 7.087591413974149,
+      "learning_rate": 2e-05,
+      "loss": 0.9881,
+      "step": 1549
+    },
+    {
+      "epoch": 0.03689422069884795,
+      "grad_norm": 6.226566300797245,
+      "learning_rate": 2e-05,
+      "loss": 0.7722,
+      "step": 1550
+    },
+    {
+      "epoch": 0.03691802342187946,
+      "grad_norm": 7.26448911144998,
+      "learning_rate": 2e-05,
+      "loss": 0.3428,
+      "step": 1551
+    },
+    {
+      "epoch": 0.036941826144910976,
+      "grad_norm": 2.2795522130159727,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 1552
+    },
+    {
+      "epoch": 0.036965628867942496,
+      "grad_norm": 4.87398245524533,
+      "learning_rate": 2e-05,
+      "loss": 0.6734,
+      "step": 1553
+    },
+    {
+      "epoch": 0.03698943159097401,
+      "grad_norm": 5.997162252023767,
+      "learning_rate": 2e-05,
+      "loss": 0.9069,
+      "step": 1554
+    },
+    {
+      "epoch": 0.03701323431400552,
+      "grad_norm": 4.311264161390445,
+      "learning_rate": 2e-05,
+      "loss": 0.2677,
+      "step": 1555
+    },
+    {
+      "epoch": 0.037037037037037035,
+      "grad_norm": 5.425238028210277,
+      "learning_rate": 2e-05,
+      "loss": 0.7552,
+      "step": 1556
+    },
+    {
+      "epoch": 0.037060839760068555,
+      "grad_norm": 4.352648940754616,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 1557
+    },
+    {
+      "epoch": 0.03708464248310007,
+      "grad_norm": 4.393292747016725,
+      "learning_rate": 2e-05,
+      "loss": 0.2448,
+      "step": 1558
+    },
+    {
+      "epoch": 0.03710844520613158,
+      "grad_norm": 12.932507117282494,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 1559
+    },
+    {
+      "epoch": 0.037132247929163094,
+      "grad_norm": 2.050514350207554,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 1560
+    },
+    {
+      "epoch": 0.037156050652194614,
+      "grad_norm": 10.680782052245737,
+      "learning_rate": 2e-05,
+      "loss": 0.4051,
+      "step": 1561
+    },
+    {
+      "epoch": 0.03717985337522613,
+      "grad_norm": 7.197091186191045,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 1562
+    },
+    {
+      "epoch": 0.03720365609825764,
+      "grad_norm": 3.8235286884435435,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 1563
+    },
+    {
+      "epoch": 0.03722745882128915,
+      "grad_norm": 10.562444726992483,
+      "learning_rate": 2e-05,
+      "loss": 0.4112,
+      "step": 1564
+    },
+    {
+      "epoch": 0.03725126154432067,
+      "grad_norm": 12.32135830495623,
+      "learning_rate": 2e-05,
+      "loss": 1.4042,
+      "step": 1565
+    },
+    {
+      "epoch": 0.037275064267352186,
+      "grad_norm": 3.2751382694086377,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 1566
+    },
+    {
+      "epoch": 0.0372988669903837,
+      "grad_norm": 2.887388302170535,
+      "learning_rate": 2e-05,
+      "loss": 0.161,
+      "step": 1567
+    },
+    {
+      "epoch": 0.03732266971341521,
+      "grad_norm": 11.91165225739744,
+      "learning_rate": 2e-05,
+      "loss": 1.7101,
+      "step": 1568
+    },
+    {
+      "epoch": 0.03734647243644673,
+      "grad_norm": 3.6747001130683574,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 1569
+    },
+    {
+      "epoch": 0.037370275159478245,
+      "grad_norm": 5.57381073490345,
+      "learning_rate": 2e-05,
+      "loss": 0.2255,
+      "step": 1570
+    },
+    {
+      "epoch": 0.03739407788250976,
+      "grad_norm": 4.32461263268277,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 1571
+    },
+    {
+      "epoch": 0.03741788060554127,
+      "grad_norm": 11.534290596101307,
+      "learning_rate": 2e-05,
+      "loss": 1.2685,
+      "step": 1572
+    },
+    {
+      "epoch": 0.03744168332857279,
+      "grad_norm": 2.507350202035047,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 1573
+    },
+    {
+      "epoch": 0.0374654860516043,
+      "grad_norm": 7.429833817602417,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 1574
+    },
+    {
+      "epoch": 0.037489288774635816,
+      "grad_norm": 3.8535856762929326,
+      "learning_rate": 2e-05,
+      "loss": 0.1547,
+      "step": 1575
+    },
+    {
+      "epoch": 0.037513091497667336,
+      "grad_norm": 5.030619453981506,
+      "learning_rate": 2e-05,
+      "loss": 0.5786,
+      "step": 1576
+    },
+    {
+      "epoch": 0.03753689422069885,
+      "grad_norm": 5.664304954621593,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 1577
+    },
+    {
+      "epoch": 0.03756069694373036,
+      "grad_norm": 7.695787120782867,
+      "learning_rate": 2e-05,
+      "loss": 0.4428,
+      "step": 1578
+    },
+    {
+      "epoch": 0.037584499666761875,
+      "grad_norm": 7.738473786555089,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 1579
+    },
+    {
+      "epoch": 0.037608302389793395,
+      "grad_norm": 9.576691346565497,
+      "learning_rate": 2e-05,
+      "loss": 1.0489,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03763210511282491,
+      "grad_norm": 7.557213828516378,
+      "learning_rate": 2e-05,
+      "loss": 0.6162,
+      "step": 1581
+    },
+    {
+      "epoch": 0.03765590783585642,
+      "grad_norm": 5.454806535961331,
+      "learning_rate": 2e-05,
+      "loss": 0.6807,
+      "step": 1582
+    },
+    {
+      "epoch": 0.037679710558887934,
+      "grad_norm": 9.704171889528691,
+      "learning_rate": 2e-05,
+      "loss": 2.3187,
+      "step": 1583
+    },
+    {
+      "epoch": 0.037703513281919454,
+      "grad_norm": 3.4703841160406927,
+      "learning_rate": 2e-05,
+      "loss": 0.1819,
+      "step": 1584
+    },
+    {
+      "epoch": 0.03772731600495097,
+      "grad_norm": 6.9942563673891085,
+      "learning_rate": 2e-05,
+      "loss": 0.5728,
+      "step": 1585
+    },
+    {
+      "epoch": 0.03775111872798248,
+      "grad_norm": 2.0138649029742033,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 1586
+    },
+    {
+      "epoch": 0.03777492145101399,
+      "grad_norm": 3.102420620389574,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 1587
+    },
+    {
+      "epoch": 0.03779872417404551,
+      "grad_norm": 15.011715089225037,
+      "learning_rate": 2e-05,
+      "loss": 0.3867,
+      "step": 1588
+    },
+    {
+      "epoch": 0.037822526897077026,
+      "grad_norm": 2.5513824162897265,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 1589
+    },
+    {
+      "epoch": 0.03784632962010854,
+      "grad_norm": 3.4374296625623075,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 1590
+    },
+    {
+      "epoch": 0.03787013234314005,
+      "grad_norm": 2.3541462655250704,
+      "learning_rate": 2e-05,
+      "loss": 0.1944,
+      "step": 1591
+    },
+    {
+      "epoch": 0.03789393506617157,
+      "grad_norm": 7.554492481493163,
+      "learning_rate": 2e-05,
+      "loss": 0.3083,
+      "step": 1592
+    },
+    {
+      "epoch": 0.037917737789203085,
+      "grad_norm": 3.652389794255674,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 1593
+    },
+    {
+      "epoch": 0.0379415405122346,
+      "grad_norm": 8.733046640575228,
+      "learning_rate": 2e-05,
+      "loss": 1.013,
+      "step": 1594
+    },
+    {
+      "epoch": 0.03796534323526612,
+      "grad_norm": 12.115936386889732,
+      "learning_rate": 2e-05,
+      "loss": 1.2452,
+      "step": 1595
+    },
+    {
+      "epoch": 0.03798914595829763,
+      "grad_norm": 10.09567153450883,
+      "learning_rate": 2e-05,
+      "loss": 1.2477,
+      "step": 1596
+    },
+    {
+      "epoch": 0.038012948681329144,
+      "grad_norm": 12.533368641880713,
+      "learning_rate": 2e-05,
+      "loss": 1.0532,
+      "step": 1597
+    },
+    {
+      "epoch": 0.03803675140436066,
+      "grad_norm": 6.908652148125706,
+      "learning_rate": 2e-05,
+      "loss": 0.8845,
+      "step": 1598
+    },
+    {
+      "epoch": 0.03806055412739218,
+      "grad_norm": 3.1799953231105262,
+      "learning_rate": 2e-05,
+      "loss": 0.3272,
+      "step": 1599
+    },
+    {
+      "epoch": 0.03808435685042369,
+      "grad_norm": 12.210318395601918,
+      "learning_rate": 2e-05,
+      "loss": 1.1615,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0381081595734552,
+      "grad_norm": 2.3087986611342513,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 1601
+    },
+    {
+      "epoch": 0.038131962296486716,
+      "grad_norm": 8.751295374330207,
+      "learning_rate": 2e-05,
+      "loss": 1.0269,
+      "step": 1602
+    },
+    {
+      "epoch": 0.038155765019518235,
+      "grad_norm": 7.278352421760363,
+      "learning_rate": 2e-05,
+      "loss": 0.5436,
+      "step": 1603
+    },
+    {
+      "epoch": 0.03817956774254975,
+      "grad_norm": 5.812961440594782,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 1604
+    },
+    {
+      "epoch": 0.03820337046558126,
+      "grad_norm": 4.238342278176657,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 1605
+    },
+    {
+      "epoch": 0.038227173188612774,
+      "grad_norm": 4.009214544059626,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 1606
+    },
+    {
+      "epoch": 0.038250975911644294,
+      "grad_norm": 6.1725377952326355,
+      "learning_rate": 2e-05,
+      "loss": 0.7344,
+      "step": 1607
+    },
+    {
+      "epoch": 0.03827477863467581,
+      "grad_norm": 6.203232403439919,
+      "learning_rate": 2e-05,
+      "loss": 0.7939,
+      "step": 1608
+    },
+    {
+      "epoch": 0.03829858135770732,
+      "grad_norm": 6.0275840803307785,
+      "learning_rate": 2e-05,
+      "loss": 0.8567,
+      "step": 1609
+    },
+    {
+      "epoch": 0.03832238408073883,
+      "grad_norm": 2.7787600565742405,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03834618680377035,
+      "grad_norm": 4.923733639110688,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 1611
+    },
+    {
+      "epoch": 0.038369989526801866,
+      "grad_norm": 2.506927177284606,
+      "learning_rate": 2e-05,
+      "loss": 0.1925,
+      "step": 1612
+    },
+    {
+      "epoch": 0.03839379224983338,
+      "grad_norm": 2.703208397758811,
+      "learning_rate": 2e-05,
+      "loss": 0.1726,
+      "step": 1613
+    },
+    {
+      "epoch": 0.0384175949728649,
+      "grad_norm": 9.485961655066708,
+      "learning_rate": 2e-05,
+      "loss": 0.3937,
+      "step": 1614
+    },
+    {
+      "epoch": 0.03844139769589641,
+      "grad_norm": 2.4233520313125845,
+      "learning_rate": 2e-05,
+      "loss": 0.2456,
+      "step": 1615
+    },
+    {
+      "epoch": 0.038465200418927925,
+      "grad_norm": 2.4234098463194127,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 1616
+    },
+    {
+      "epoch": 0.03848900314195944,
+      "grad_norm": 4.572590771577121,
+      "learning_rate": 2e-05,
+      "loss": 0.217,
+      "step": 1617
+    },
+    {
+      "epoch": 0.03851280586499096,
+      "grad_norm": 5.17297176587893,
+      "learning_rate": 2e-05,
+      "loss": 0.8181,
+      "step": 1618
+    },
+    {
+      "epoch": 0.03853660858802247,
+      "grad_norm": 4.730954639323053,
+      "learning_rate": 2e-05,
+      "loss": 0.6348,
+      "step": 1619
+    },
+    {
+      "epoch": 0.038560411311053984,
+      "grad_norm": 2.0594696723959944,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 1620
+    },
+    {
+      "epoch": 0.0385842140340855,
+      "grad_norm": 9.359698704416452,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 1621
+    },
+    {
+      "epoch": 0.03860801675711702,
+      "grad_norm": 3.705016603487109,
+      "learning_rate": 2e-05,
+      "loss": 0.3183,
+      "step": 1622
+    },
+    {
+      "epoch": 0.03863181948014853,
+      "grad_norm": 3.2271303338248667,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 1623
+    },
+    {
+      "epoch": 0.03865562220318004,
+      "grad_norm": 3.3342866286989157,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 1624
+    },
+    {
+      "epoch": 0.038679424926211556,
+      "grad_norm": 8.940071866853266,
+      "learning_rate": 2e-05,
+      "loss": 1.0529,
+      "step": 1625
+    },
+    {
+      "epoch": 0.038703227649243076,
+      "grad_norm": 3.0202302377684123,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 1626
+    },
+    {
+      "epoch": 0.03872703037227459,
+      "grad_norm": 4.282290464947116,
+      "learning_rate": 2e-05,
+      "loss": 0.5357,
+      "step": 1627
+    },
+    {
+      "epoch": 0.0387508330953061,
+      "grad_norm": 9.109054625329845,
+      "learning_rate": 2e-05,
+      "loss": 1.1669,
+      "step": 1628
+    },
+    {
+      "epoch": 0.038774635818337615,
+      "grad_norm": 9.342742298819877,
+      "learning_rate": 2e-05,
+      "loss": 0.4013,
+      "step": 1629
+    },
+    {
+      "epoch": 0.038798438541369135,
+      "grad_norm": 7.775529730691509,
+      "learning_rate": 2e-05,
+      "loss": 0.5314,
+      "step": 1630
+    },
+    {
+      "epoch": 0.03882224126440065,
+      "grad_norm": 7.949416259477404,
+      "learning_rate": 2e-05,
+      "loss": 0.9907,
+      "step": 1631
+    },
+    {
+      "epoch": 0.03884604398743216,
+      "grad_norm": 3.501813017380971,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 1632
+    },
+    {
+      "epoch": 0.03886984671046368,
+      "grad_norm": 6.034299161839956,
+      "learning_rate": 2e-05,
+      "loss": 0.6683,
+      "step": 1633
+    },
+    {
+      "epoch": 0.038893649433495193,
+      "grad_norm": 14.42740395355198,
+      "learning_rate": 2e-05,
+      "loss": 1.2918,
+      "step": 1634
+    },
+    {
+      "epoch": 0.038917452156526706,
+      "grad_norm": 9.09627859261468,
+      "learning_rate": 2e-05,
+      "loss": 0.9061,
+      "step": 1635
+    },
+    {
+      "epoch": 0.03894125487955822,
+      "grad_norm": 7.321723930767468,
+      "learning_rate": 2e-05,
+      "loss": 1.0952,
+      "step": 1636
+    },
+    {
+      "epoch": 0.03896505760258974,
+      "grad_norm": 3.226807437625986,
+      "learning_rate": 2e-05,
+      "loss": 0.179,
+      "step": 1637
+    },
+    {
+      "epoch": 0.03898886032562125,
+      "grad_norm": 7.319780417662733,
+      "learning_rate": 2e-05,
+      "loss": 0.4498,
+      "step": 1638
+    },
+    {
+      "epoch": 0.039012663048652765,
+      "grad_norm": 8.412218664946815,
+      "learning_rate": 2e-05,
+      "loss": 0.9459,
+      "step": 1639
+    },
+    {
+      "epoch": 0.03903646577168428,
+      "grad_norm": 3.50133138509313,
+      "learning_rate": 2e-05,
+      "loss": 0.3598,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0390602684947158,
+      "grad_norm": 5.782052266763102,
+      "learning_rate": 2e-05,
+      "loss": 0.8834,
+      "step": 1641
+    },
+    {
+      "epoch": 0.03908407121774731,
+      "grad_norm": 9.46342047485872,
+      "learning_rate": 2e-05,
+      "loss": 0.3854,
+      "step": 1642
+    },
+    {
+      "epoch": 0.039107873940778824,
+      "grad_norm": 6.661703975749041,
+      "learning_rate": 2e-05,
+      "loss": 0.6911,
+      "step": 1643
+    },
+    {
+      "epoch": 0.03913167666381034,
+      "grad_norm": 3.026813849425251,
+      "learning_rate": 2e-05,
+      "loss": 0.2011,
+      "step": 1644
+    },
+    {
+      "epoch": 0.03915547938684186,
+      "grad_norm": 3.483240299465379,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 1645
+    },
+    {
+      "epoch": 0.03917928210987337,
+      "grad_norm": 9.55729886994216,
+      "learning_rate": 2e-05,
+      "loss": 0.5741,
+      "step": 1646
+    },
+    {
+      "epoch": 0.03920308483290488,
+      "grad_norm": 8.782442418641637,
+      "learning_rate": 2e-05,
+      "loss": 1.094,
+      "step": 1647
+    },
+    {
+      "epoch": 0.039226887555936396,
+      "grad_norm": 4.739538132290373,
+      "learning_rate": 2e-05,
+      "loss": 0.4395,
+      "step": 1648
+    },
+    {
+      "epoch": 0.039250690278967916,
+      "grad_norm": 5.39558262753335,
+      "learning_rate": 2e-05,
+      "loss": 0.3222,
+      "step": 1649
+    },
+    {
+      "epoch": 0.03927449300199943,
+      "grad_norm": 6.175474676388335,
+      "learning_rate": 2e-05,
+      "loss": 0.3761,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03929829572503094,
+      "grad_norm": 8.312842313303923,
+      "learning_rate": 2e-05,
+      "loss": 0.8451,
+      "step": 1651
+    },
+    {
+      "epoch": 0.039322098448062455,
+      "grad_norm": 4.467860834825743,
+      "learning_rate": 2e-05,
+      "loss": 0.3304,
+      "step": 1652
+    },
+    {
+      "epoch": 0.039345901171093975,
+      "grad_norm": 5.157232172212781,
+      "learning_rate": 2e-05,
+      "loss": 0.7464,
+      "step": 1653
+    },
+    {
+      "epoch": 0.03936970389412549,
+      "grad_norm": 4.459712950544472,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 1654
+    },
+    {
+      "epoch": 0.039393506617157,
+      "grad_norm": 7.502887856852155,
+      "learning_rate": 2e-05,
+      "loss": 0.3615,
+      "step": 1655
+    },
+    {
+      "epoch": 0.03941730934018852,
+      "grad_norm": 5.541027106061458,
+      "learning_rate": 2e-05,
+      "loss": 0.2654,
+      "step": 1656
+    },
+    {
+      "epoch": 0.039441112063220034,
+      "grad_norm": 5.850239080135157,
+      "learning_rate": 2e-05,
+      "loss": 0.67,
+      "step": 1657
+    },
+    {
+      "epoch": 0.03946491478625155,
+      "grad_norm": 4.5636933919518246,
+      "learning_rate": 2e-05,
+      "loss": 0.4225,
+      "step": 1658
+    },
+    {
+      "epoch": 0.03948871750928306,
+      "grad_norm": 2.441963107295684,
+      "learning_rate": 2e-05,
+      "loss": 0.2432,
+      "step": 1659
+    },
+    {
+      "epoch": 0.03951252023231458,
+      "grad_norm": 8.52760028601765,
+      "learning_rate": 2e-05,
+      "loss": 1.4542,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03953632295534609,
+      "grad_norm": 7.24911354539839,
+      "learning_rate": 2e-05,
+      "loss": 0.4228,
+      "step": 1661
+    },
+    {
+      "epoch": 0.039560125678377606,
+      "grad_norm": 4.532327166894906,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 1662
+    },
+    {
+      "epoch": 0.03958392840140912,
+      "grad_norm": 4.922710767581923,
+      "learning_rate": 2e-05,
+      "loss": 0.6624,
+      "step": 1663
+    },
+    {
+      "epoch": 0.03960773112444064,
+      "grad_norm": 7.7770274040933485,
+      "learning_rate": 2e-05,
+      "loss": 0.837,
+      "step": 1664
+    },
+    {
+      "epoch": 0.03963153384747215,
+      "grad_norm": 5.400560983083997,
+      "learning_rate": 2e-05,
+      "loss": 0.2274,
+      "step": 1665
+    },
+    {
+      "epoch": 0.039655336570503665,
+      "grad_norm": 8.724241238376711,
+      "learning_rate": 2e-05,
+      "loss": 0.7592,
+      "step": 1666
+    },
+    {
+      "epoch": 0.03967913929353518,
+      "grad_norm": 3.9627901812642694,
+      "learning_rate": 2e-05,
+      "loss": 0.2087,
+      "step": 1667
+    },
+    {
+      "epoch": 0.0397029420165667,
+      "grad_norm": 7.148212643269686,
+      "learning_rate": 2e-05,
+      "loss": 0.4216,
+      "step": 1668
+    },
+    {
+      "epoch": 0.03972674473959821,
+      "grad_norm": 1.0440266860567604,
+      "learning_rate": 2e-05,
+      "loss": 0.1983,
+      "step": 1669
+    },
+    {
+      "epoch": 0.03975054746262972,
+      "grad_norm": 6.8374074774194025,
+      "learning_rate": 2e-05,
+      "loss": 0.4793,
+      "step": 1670
+    },
+    {
+      "epoch": 0.039774350185661236,
+      "grad_norm": 9.986265222185944,
+      "learning_rate": 2e-05,
+      "loss": 2.2054,
+      "step": 1671
+    },
+    {
+      "epoch": 0.039798152908692756,
+      "grad_norm": 2.1724026852082434,
+      "learning_rate": 2e-05,
+      "loss": 0.2794,
+      "step": 1672
+    },
+    {
+      "epoch": 0.03982195563172427,
+      "grad_norm": 5.67928820622391,
+      "learning_rate": 2e-05,
+      "loss": 0.9017,
+      "step": 1673
+    },
+    {
+      "epoch": 0.03984575835475578,
+      "grad_norm": 4.41609868942551,
+      "learning_rate": 2e-05,
+      "loss": 0.1875,
+      "step": 1674
+    },
+    {
+      "epoch": 0.0398695610777873,
+      "grad_norm": 27.210863788582962,
+      "learning_rate": 2e-05,
+      "loss": 1.4951,
+      "step": 1675
+    },
+    {
+      "epoch": 0.039893363800818815,
+      "grad_norm": 5.669693178268891,
+      "learning_rate": 2e-05,
+      "loss": 0.6964,
+      "step": 1676
+    },
+    {
+      "epoch": 0.03991716652385033,
+      "grad_norm": 1.6863835368535398,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 1677
+    },
+    {
+      "epoch": 0.03994096924688184,
+      "grad_norm": 8.518020965143041,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 1678
+    },
+    {
+      "epoch": 0.03996477196991336,
+      "grad_norm": 9.301575867062253,
+      "learning_rate": 2e-05,
+      "loss": 1.6215,
+      "step": 1679
+    },
+    {
+      "epoch": 0.039988574692944874,
+      "grad_norm": 10.878715975362548,
+      "learning_rate": 2e-05,
+      "loss": 1.5636,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04001237741597639,
+      "grad_norm": 5.35901443169936,
+      "learning_rate": 2e-05,
+      "loss": 0.8443,
+      "step": 1681
+    },
+    {
+      "epoch": 0.0400361801390079,
+      "grad_norm": 3.701305612547214,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 1682
+    },
+    {
+      "epoch": 0.04005998286203942,
+      "grad_norm": 5.59094100480851,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 1683
+    },
+    {
+      "epoch": 0.04008378558507093,
+      "grad_norm": 8.947061308062851,
+      "learning_rate": 2e-05,
+      "loss": 0.3785,
+      "step": 1684
+    },
+    {
+      "epoch": 0.040107588308102446,
+      "grad_norm": 6.322944847896313,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 1685
+    },
+    {
+      "epoch": 0.04013139103113396,
+      "grad_norm": 6.595854662819733,
+      "learning_rate": 2e-05,
+      "loss": 1.1902,
+      "step": 1686
+    },
+    {
+      "epoch": 0.04015519375416548,
+      "grad_norm": 4.043933848928537,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 1687
+    },
+    {
+      "epoch": 0.04017899647719699,
+      "grad_norm": 7.266478958967772,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 1688
+    },
+    {
+      "epoch": 0.040202799200228505,
+      "grad_norm": 6.924513534281543,
+      "learning_rate": 2e-05,
+      "loss": 1.0298,
+      "step": 1689
+    },
+    {
+      "epoch": 0.04022660192326002,
+      "grad_norm": 4.203633064278548,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04025040464629154,
+      "grad_norm": 10.762930893241899,
+      "learning_rate": 2e-05,
+      "loss": 1.4471,
+      "step": 1691
+    },
+    {
+      "epoch": 0.04027420736932305,
+      "grad_norm": 2.733047012529821,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 1692
+    },
+    {
+      "epoch": 0.040298010092354564,
+      "grad_norm": 15.011672661003647,
+      "learning_rate": 2e-05,
+      "loss": 0.7603,
+      "step": 1693
+    },
+    {
+      "epoch": 0.040321812815386084,
+      "grad_norm": 9.205091095498299,
+      "learning_rate": 2e-05,
+      "loss": 0.6732,
+      "step": 1694
+    },
+    {
+      "epoch": 0.0403456155384176,
+      "grad_norm": 10.175252047276171,
+      "learning_rate": 2e-05,
+      "loss": 0.8298,
+      "step": 1695
+    },
+    {
+      "epoch": 0.04036941826144911,
+      "grad_norm": 3.8347913880875004,
+      "learning_rate": 2e-05,
+      "loss": 0.2309,
+      "step": 1696
+    },
+    {
+      "epoch": 0.04039322098448062,
+      "grad_norm": 28.867321404729033,
+      "learning_rate": 2e-05,
+      "loss": 0.6792,
+      "step": 1697
+    },
+    {
+      "epoch": 0.04041702370751214,
+      "grad_norm": 10.695394949208977,
+      "learning_rate": 2e-05,
+      "loss": 0.5106,
+      "step": 1698
+    },
+    {
+      "epoch": 0.040440826430543655,
+      "grad_norm": 9.116992608079023,
+      "learning_rate": 2e-05,
+      "loss": 1.0885,
+      "step": 1699
+    },
+    {
+      "epoch": 0.04046462915357517,
+      "grad_norm": 3.5564310815654157,
+      "learning_rate": 2e-05,
+      "loss": 0.2221,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04048843187660668,
+      "grad_norm": 3.3464444690722757,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 1701
+    },
+    {
+      "epoch": 0.0405122345996382,
+      "grad_norm": 11.027570782808542,
+      "learning_rate": 2e-05,
+      "loss": 1.1565,
+      "step": 1702
+    },
+    {
+      "epoch": 0.040536037322669714,
+      "grad_norm": 10.783176663710842,
+      "learning_rate": 2e-05,
+      "loss": 1.1849,
+      "step": 1703
+    },
+    {
+      "epoch": 0.04055984004570123,
+      "grad_norm": 4.27564993367701,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 1704
+    },
+    {
+      "epoch": 0.04058364276873274,
+      "grad_norm": 9.553021004700195,
+      "learning_rate": 2e-05,
+      "loss": 1.2798,
+      "step": 1705
+    },
+    {
+      "epoch": 0.04060744549176426,
+      "grad_norm": 4.669121642127749,
+      "learning_rate": 2e-05,
+      "loss": 0.7588,
+      "step": 1706
+    },
+    {
+      "epoch": 0.04063124821479577,
+      "grad_norm": 7.105020945053884,
+      "learning_rate": 2e-05,
+      "loss": 0.889,
+      "step": 1707
+    },
+    {
+      "epoch": 0.040655050937827286,
+      "grad_norm": 8.061019472245704,
+      "learning_rate": 2e-05,
+      "loss": 1.0541,
+      "step": 1708
+    },
+    {
+      "epoch": 0.0406788536608588,
+      "grad_norm": 9.58339156989302,
+      "learning_rate": 2e-05,
+      "loss": 1.4047,
+      "step": 1709
+    },
+    {
+      "epoch": 0.04070265638389032,
+      "grad_norm": 3.4294441739405346,
+      "learning_rate": 2e-05,
+      "loss": 0.17,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04072645910692183,
+      "grad_norm": 2.8579484589309456,
+      "learning_rate": 2e-05,
+      "loss": 0.2633,
+      "step": 1711
+    },
+    {
+      "epoch": 0.040750261829953345,
+      "grad_norm": 5.088770981776327,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 1712
+    },
+    {
+      "epoch": 0.04077406455298486,
+      "grad_norm": 6.301303083175015,
+      "learning_rate": 2e-05,
+      "loss": 0.8201,
+      "step": 1713
+    },
+    {
+      "epoch": 0.04079786727601638,
+      "grad_norm": 9.48604213639334,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 1714
+    },
+    {
+      "epoch": 0.04082166999904789,
+      "grad_norm": 25.60076061546651,
+      "learning_rate": 2e-05,
+      "loss": 0.7371,
+      "step": 1715
+    },
+    {
+      "epoch": 0.040845472722079404,
+      "grad_norm": 3.820115619358662,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 1716
+    },
+    {
+      "epoch": 0.040869275445110924,
+      "grad_norm": 8.275781947798196,
+      "learning_rate": 2e-05,
+      "loss": 1.6757,
+      "step": 1717
+    },
+    {
+      "epoch": 0.04089307816814244,
+      "grad_norm": 4.966815428082522,
+      "learning_rate": 2e-05,
+      "loss": 0.7449,
+      "step": 1718
+    },
+    {
+      "epoch": 0.04091688089117395,
+      "grad_norm": 22.64612618743551,
+      "learning_rate": 2e-05,
+      "loss": 0.9389,
+      "step": 1719
+    },
+    {
+      "epoch": 0.04094068361420546,
+      "grad_norm": 642.5998047354088,
+      "learning_rate": 2e-05,
+      "loss": 2.6773,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04096448633723698,
+      "grad_norm": 4.8582300504935585,
+      "learning_rate": 2e-05,
+      "loss": 0.233,
+      "step": 1721
+    },
+    {
+      "epoch": 0.040988289060268496,
+      "grad_norm": 24.10765847521324,
+      "learning_rate": 2e-05,
+      "loss": 0.8908,
+      "step": 1722
+    },
+    {
+      "epoch": 0.04101209178330001,
+      "grad_norm": 4.437241677707595,
+      "learning_rate": 2e-05,
+      "loss": 0.6778,
+      "step": 1723
+    },
+    {
+      "epoch": 0.04103589450633152,
+      "grad_norm": 10.3995237346797,
+      "learning_rate": 2e-05,
+      "loss": 0.3814,
+      "step": 1724
+    },
+    {
+      "epoch": 0.04105969722936304,
+      "grad_norm": 760.97352769364,
+      "learning_rate": 2e-05,
+      "loss": 4.6387,
+      "step": 1725
+    },
+    {
+      "epoch": 0.041083499952394555,
+      "grad_norm": 7.481965987230018,
+      "learning_rate": 2e-05,
+      "loss": 0.9847,
+      "step": 1726
+    },
+    {
+      "epoch": 0.04110730267542607,
+      "grad_norm": 4.784127703686634,
+      "learning_rate": 2e-05,
+      "loss": 0.2917,
+      "step": 1727
+    },
+    {
+      "epoch": 0.04113110539845758,
+      "grad_norm": 33.34837564728999,
+      "learning_rate": 2e-05,
+      "loss": 3.7383,
+      "step": 1728
+    },
+    {
+      "epoch": 0.0411549081214891,
+      "grad_norm": 62.50491905912209,
+      "learning_rate": 2e-05,
+      "loss": 3.3184,
+      "step": 1729
+    },
+    {
+      "epoch": 0.041178710844520613,
+      "grad_norm": 5.3635402588121455,
+      "learning_rate": 2e-05,
+      "loss": 0.6028,
+      "step": 1730
+    },
+    {
+      "epoch": 0.041202513567552126,
+      "grad_norm": 18.182318913290953,
+      "learning_rate": 2e-05,
+      "loss": 1.5098,
+      "step": 1731
+    },
+    {
+      "epoch": 0.04122631629058364,
+      "grad_norm": 1.4250615116373238,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 1732
+    },
+    {
+      "epoch": 0.04125011901361516,
+      "grad_norm": 3.6519954709893607,
+      "learning_rate": 2e-05,
+      "loss": 0.255,
+      "step": 1733
+    },
+    {
+      "epoch": 0.04127392173664667,
+      "grad_norm": 5.716241807332978,
+      "learning_rate": 2e-05,
+      "loss": 0.495,
+      "step": 1734
+    },
+    {
+      "epoch": 0.041297724459678185,
+      "grad_norm": 34.358349699772724,
+      "learning_rate": 2e-05,
+      "loss": 0.4508,
+      "step": 1735
+    },
+    {
+      "epoch": 0.041321527182709705,
+      "grad_norm": 3.3759813450325518,
+      "learning_rate": 2e-05,
+      "loss": 0.1712,
+      "step": 1736
+    },
+    {
+      "epoch": 0.04134532990574122,
+      "grad_norm": 2.1333425343537056,
+      "learning_rate": 2e-05,
+      "loss": 0.1986,
+      "step": 1737
+    },
+    {
+      "epoch": 0.04136913262877273,
+      "grad_norm": 61.81582580883,
+      "learning_rate": 2e-05,
+      "loss": 0.4473,
+      "step": 1738
+    },
+    {
+      "epoch": 0.041392935351804244,
+      "grad_norm": 131.32016226673153,
+      "learning_rate": 2e-05,
+      "loss": 1.6332,
+      "step": 1739
+    },
+    {
+      "epoch": 0.041416738074835764,
+      "grad_norm": 3.0655354848204435,
+      "learning_rate": 2e-05,
+      "loss": 0.1605,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04144054079786728,
+      "grad_norm": 22.48392196536762,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 1741
+    },
+    {
+      "epoch": 0.04146434352089879,
+      "grad_norm": 3.6380963214704765,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 1742
+    },
+    {
+      "epoch": 0.0414881462439303,
+      "grad_norm": 4.407012348479068,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 1743
+    },
+    {
+      "epoch": 0.04151194896696182,
+      "grad_norm": 6.217615029081966,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 1744
+    },
+    {
+      "epoch": 0.041535751689993336,
+      "grad_norm": 5.808924304018996,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 1745
+    },
+    {
+      "epoch": 0.04155955441302485,
+      "grad_norm": 7.272581247246199,
+      "learning_rate": 2e-05,
+      "loss": 0.8858,
+      "step": 1746
+    },
+    {
+      "epoch": 0.04158335713605636,
+      "grad_norm": 2.5969080346220803,
+      "learning_rate": 2e-05,
+      "loss": 0.1665,
+      "step": 1747
+    },
+    {
+      "epoch": 0.04160715985908788,
+      "grad_norm": 7.872278625706871,
+      "learning_rate": 2e-05,
+      "loss": 1.1673,
+      "step": 1748
+    },
+    {
+      "epoch": 0.041630962582119395,
+      "grad_norm": 4.918617912537381,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 1749
+    },
+    {
+      "epoch": 0.04165476530515091,
+      "grad_norm": 2.1241786046249715,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04167856802818242,
+      "grad_norm": 5.553161991427888,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 1751
+    },
+    {
+      "epoch": 0.04170237075121394,
+      "grad_norm": 15.49509750022506,
+      "learning_rate": 2e-05,
+      "loss": 0.9839,
+      "step": 1752
+    },
+    {
+      "epoch": 0.041726173474245454,
+      "grad_norm": 5.131457088016092,
+      "learning_rate": 2e-05,
+      "loss": 0.2673,
+      "step": 1753
+    },
+    {
+      "epoch": 0.04174997619727697,
+      "grad_norm": 3.063946546138696,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 1754
+    },
+    {
+      "epoch": 0.04177377892030849,
+      "grad_norm": 3.9662331866427736,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 1755
+    },
+    {
+      "epoch": 0.04179758164334,
+      "grad_norm": 5.869688929174795,
+      "learning_rate": 2e-05,
+      "loss": 0.3776,
+      "step": 1756
+    },
+    {
+      "epoch": 0.04182138436637151,
+      "grad_norm": 4.425209960502867,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 1757
+    },
+    {
+      "epoch": 0.041845187089403026,
+      "grad_norm": 160.61860859949437,
+      "learning_rate": 2e-05,
+      "loss": 0.5661,
+      "step": 1758
+    },
+    {
+      "epoch": 0.041868989812434546,
+      "grad_norm": 7.826216098475665,
+      "learning_rate": 2e-05,
+      "loss": 0.2777,
+      "step": 1759
+    },
+    {
+      "epoch": 0.04189279253546606,
+      "grad_norm": 25.729073656176936,
+      "learning_rate": 2e-05,
+      "loss": 0.7955,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04191659525849757,
+      "grad_norm": 4.285511408957504,
+      "learning_rate": 2e-05,
+      "loss": 0.3279,
+      "step": 1761
+    },
+    {
+      "epoch": 0.041940397981529084,
+      "grad_norm": 12.637118375010804,
+      "learning_rate": 2e-05,
+      "loss": 0.6514,
+      "step": 1762
+    },
+    {
+      "epoch": 0.041964200704560604,
+      "grad_norm": 32.77463664922913,
+      "learning_rate": 2e-05,
+      "loss": 1.3786,
+      "step": 1763
+    },
+    {
+      "epoch": 0.04198800342759212,
+      "grad_norm": 3.3724781119833147,
+      "learning_rate": 2e-05,
+      "loss": 0.2871,
+      "step": 1764
+    },
+    {
+      "epoch": 0.04201180615062363,
+      "grad_norm": 32.57247232616135,
+      "learning_rate": 2e-05,
+      "loss": 1.1468,
+      "step": 1765
+    },
+    {
+      "epoch": 0.04203560887365514,
+      "grad_norm": 2.4396491043186876,
+      "learning_rate": 2e-05,
+      "loss": 0.1573,
+      "step": 1766
+    },
+    {
+      "epoch": 0.04205941159668666,
+      "grad_norm": 23.943225882383356,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 1767
+    },
+    {
+      "epoch": 0.042083214319718176,
+      "grad_norm": 6.731284967226584,
+      "learning_rate": 2e-05,
+      "loss": 0.7714,
+      "step": 1768
+    },
+    {
+      "epoch": 0.04210701704274969,
+      "grad_norm": 9.808542055636536,
+      "learning_rate": 2e-05,
+      "loss": 0.4637,
+      "step": 1769
+    },
+    {
+      "epoch": 0.0421308197657812,
+      "grad_norm": 7.833163142298316,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04215462248881272,
+      "grad_norm": 16.65906239022527,
+      "learning_rate": 2e-05,
+      "loss": 0.5663,
+      "step": 1771
+    },
+    {
+      "epoch": 0.042178425211844235,
+      "grad_norm": 6.762392176101987,
+      "learning_rate": 2e-05,
+      "loss": 0.4666,
+      "step": 1772
+    },
+    {
+      "epoch": 0.04220222793487575,
+      "grad_norm": 4.69770769138017,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 1773
+    },
+    {
+      "epoch": 0.04222603065790726,
+      "grad_norm": 4.001024009841318,
+      "learning_rate": 2e-05,
+      "loss": 0.4082,
+      "step": 1774
+    },
+    {
+      "epoch": 0.04224983338093878,
+      "grad_norm": 9.486455896001594,
+      "learning_rate": 2e-05,
+      "loss": 2.2594,
+      "step": 1775
+    },
+    {
+      "epoch": 0.042273636103970294,
+      "grad_norm": 6.08529072852695,
+      "learning_rate": 2e-05,
+      "loss": 0.7068,
+      "step": 1776
+    },
+    {
+      "epoch": 0.04229743882700181,
+      "grad_norm": 7.594987777717986,
+      "learning_rate": 2e-05,
+      "loss": 0.5923,
+      "step": 1777
+    },
+    {
+      "epoch": 0.04232124155003333,
+      "grad_norm": 8.1672516000414,
+      "learning_rate": 2e-05,
+      "loss": 0.4576,
+      "step": 1778
+    },
+    {
+      "epoch": 0.04234504427306484,
+      "grad_norm": 7.1028675992251085,
+      "learning_rate": 2e-05,
+      "loss": 0.6556,
+      "step": 1779
+    },
+    {
+      "epoch": 0.04236884699609635,
+      "grad_norm": 11.038042177871551,
+      "learning_rate": 2e-05,
+      "loss": 0.4058,
+      "step": 1780
+    },
+    {
+      "epoch": 0.042392649719127866,
+      "grad_norm": 3.288480333637832,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 1781
+    },
+    {
+      "epoch": 0.042416452442159386,
+      "grad_norm": 140.7442061712963,
+      "learning_rate": 2e-05,
+      "loss": 0.3873,
+      "step": 1782
+    },
+    {
+      "epoch": 0.0424402551651909,
+      "grad_norm": 7.8468068427519695,
+      "learning_rate": 2e-05,
+      "loss": 0.3934,
+      "step": 1783
+    },
+    {
+      "epoch": 0.04246405788822241,
+      "grad_norm": 44.87285380768141,
+      "learning_rate": 2e-05,
+      "loss": 1.318,
+      "step": 1784
+    },
+    {
+      "epoch": 0.042487860611253925,
+      "grad_norm": 36.65771211627371,
+      "learning_rate": 2e-05,
+      "loss": 1.7927,
+      "step": 1785
+    },
+    {
+      "epoch": 0.042511663334285445,
+      "grad_norm": 7.634492099143382,
+      "learning_rate": 2e-05,
+      "loss": 0.378,
+      "step": 1786
+    },
+    {
+      "epoch": 0.04253546605731696,
+      "grad_norm": 6.196218384888286,
+      "learning_rate": 2e-05,
+      "loss": 0.9666,
+      "step": 1787
+    },
+    {
+      "epoch": 0.04255926878034847,
+      "grad_norm": 7.546676407007476,
+      "learning_rate": 2e-05,
+      "loss": 0.363,
+      "step": 1788
+    },
+    {
+      "epoch": 0.042583071503379984,
+      "grad_norm": 8.561840707270644,
+      "learning_rate": 2e-05,
+      "loss": 0.6346,
+      "step": 1789
+    },
+    {
+      "epoch": 0.042606874226411504,
+      "grad_norm": 20.181868895625588,
+      "learning_rate": 2e-05,
+      "loss": 0.5091,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04263067694944302,
+      "grad_norm": 6.407379908262136,
+      "learning_rate": 2e-05,
+      "loss": 0.8927,
+      "step": 1791
+    },
+    {
+      "epoch": 0.04265447967247453,
+      "grad_norm": 10.257387412572989,
+      "learning_rate": 2e-05,
+      "loss": 0.3553,
+      "step": 1792
+    },
+    {
+      "epoch": 0.04267828239550604,
+      "grad_norm": 10.521340638760334,
+      "learning_rate": 2e-05,
+      "loss": 0.856,
+      "step": 1793
+    },
+    {
+      "epoch": 0.04270208511853756,
+      "grad_norm": 7.295384159039484,
+      "learning_rate": 2e-05,
+      "loss": 1.2373,
+      "step": 1794
+    },
+    {
+      "epoch": 0.042725887841569075,
+      "grad_norm": 7.8999387119920526,
+      "learning_rate": 2e-05,
+      "loss": 0.5773,
+      "step": 1795
+    },
+    {
+      "epoch": 0.04274969056460059,
+      "grad_norm": 5.750643264148015,
+      "learning_rate": 2e-05,
+      "loss": 0.4948,
+      "step": 1796
+    },
+    {
+      "epoch": 0.04277349328763211,
+      "grad_norm": 8.316311179552548,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 1797
+    },
+    {
+      "epoch": 0.04279729601066362,
+      "grad_norm": 3.9574441961116102,
+      "learning_rate": 2e-05,
+      "loss": 0.2344,
+      "step": 1798
+    },
+    {
+      "epoch": 0.042821098733695134,
+      "grad_norm": 8.080853913381056,
+      "learning_rate": 2e-05,
+      "loss": 0.9046,
+      "step": 1799
+    },
+    {
+      "epoch": 0.04284490145672665,
+      "grad_norm": 5.815216736337468,
+      "learning_rate": 2e-05,
+      "loss": 0.6699,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04286870417975817,
+      "grad_norm": 10.656320244420069,
+      "learning_rate": 2e-05,
+      "loss": 0.3665,
+      "step": 1801
+    },
+    {
+      "epoch": 0.04289250690278968,
+      "grad_norm": 6.580836976345682,
+      "learning_rate": 2e-05,
+      "loss": 0.6887,
+      "step": 1802
+    },
+    {
+      "epoch": 0.04291630962582119,
+      "grad_norm": 81.23126037396506,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 1803
+    },
+    {
+      "epoch": 0.042940112348852706,
+      "grad_norm": 9.419120708701225,
+      "learning_rate": 2e-05,
+      "loss": 0.5192,
+      "step": 1804
+    },
+    {
+      "epoch": 0.042963915071884226,
+      "grad_norm": 4.00386807413515,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 1805
+    },
+    {
+      "epoch": 0.04298771779491574,
+      "grad_norm": 6.848016030437998,
+      "learning_rate": 2e-05,
+      "loss": 1.1994,
+      "step": 1806
+    },
+    {
+      "epoch": 0.04301152051794725,
+      "grad_norm": 13.46675561990992,
+      "learning_rate": 2e-05,
+      "loss": 1.5676,
+      "step": 1807
+    },
+    {
+      "epoch": 0.043035323240978765,
+      "grad_norm": 10.654485908017767,
+      "learning_rate": 2e-05,
+      "loss": 2.0689,
+      "step": 1808
+    },
+    {
+      "epoch": 0.043059125964010285,
+      "grad_norm": 5.738796526636342,
+      "learning_rate": 2e-05,
+      "loss": 0.3538,
+      "step": 1809
+    },
+    {
+      "epoch": 0.0430829286870418,
+      "grad_norm": 2.778682703482368,
+      "learning_rate": 2e-05,
+      "loss": 0.2133,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04310673141007331,
+      "grad_norm": 5.943137401620871,
+      "learning_rate": 2e-05,
+      "loss": 1.0596,
+      "step": 1811
+    },
+    {
+      "epoch": 0.043130534133104824,
+      "grad_norm": 3.712970207527707,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 1812
+    },
+    {
+      "epoch": 0.043154336856136344,
+      "grad_norm": 8.32075679913563,
+      "learning_rate": 2e-05,
+      "loss": 0.483,
+      "step": 1813
+    },
+    {
+      "epoch": 0.04317813957916786,
+      "grad_norm": 10.982357882381628,
+      "learning_rate": 2e-05,
+      "loss": 1.1456,
+      "step": 1814
+    },
+    {
+      "epoch": 0.04320194230219937,
+      "grad_norm": 5.066192781519804,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 1815
+    },
+    {
+      "epoch": 0.04322574502523089,
+      "grad_norm": 12.32494853262563,
+      "learning_rate": 2e-05,
+      "loss": 0.4326,
+      "step": 1816
+    },
+    {
+      "epoch": 0.0432495477482624,
+      "grad_norm": 5.196286923410817,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 1817
+    },
+    {
+      "epoch": 0.043273350471293916,
+      "grad_norm": 2.2142142280655097,
+      "learning_rate": 2e-05,
+      "loss": 0.199,
+      "step": 1818
+    },
+    {
+      "epoch": 0.04329715319432543,
+      "grad_norm": 7.266952320588332,
+      "learning_rate": 2e-05,
+      "loss": 0.8523,
+      "step": 1819
+    },
+    {
+      "epoch": 0.04332095591735695,
+      "grad_norm": 2.6558224893484774,
+      "learning_rate": 2e-05,
+      "loss": 0.2004,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04334475864038846,
+      "grad_norm": 16.25716903964995,
+      "learning_rate": 2e-05,
+      "loss": 0.5169,
+      "step": 1821
+    },
+    {
+      "epoch": 0.043368561363419975,
+      "grad_norm": 6.659233001039819,
+      "learning_rate": 2e-05,
+      "loss": 0.9138,
+      "step": 1822
+    },
+    {
+      "epoch": 0.04339236408645149,
+      "grad_norm": 2.5986969054540094,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 1823
+    },
+    {
+      "epoch": 0.04341616680948301,
+      "grad_norm": 4.915641546212147,
+      "learning_rate": 2e-05,
+      "loss": 0.5948,
+      "step": 1824
+    },
+    {
+      "epoch": 0.04343996953251452,
+      "grad_norm": 9.459320413460237,
+      "learning_rate": 2e-05,
+      "loss": 0.987,
+      "step": 1825
+    },
+    {
+      "epoch": 0.04346377225554603,
+      "grad_norm": 3.2106144051347782,
+      "learning_rate": 2e-05,
+      "loss": 0.341,
+      "step": 1826
+    },
+    {
+      "epoch": 0.043487574978577546,
+      "grad_norm": 2.946248228680394,
+      "learning_rate": 2e-05,
+      "loss": 0.235,
+      "step": 1827
+    },
+    {
+      "epoch": 0.043511377701609066,
+      "grad_norm": 8.102296384981525,
+      "learning_rate": 2e-05,
+      "loss": 0.9201,
+      "step": 1828
+    },
+    {
+      "epoch": 0.04353518042464058,
+      "grad_norm": 12.15424872863651,
+      "learning_rate": 2e-05,
+      "loss": 0.6928,
+      "step": 1829
+    },
+    {
+      "epoch": 0.04355898314767209,
+      "grad_norm": 6.48299807354829,
+      "learning_rate": 2e-05,
+      "loss": 0.341,
+      "step": 1830
+    },
+    {
+      "epoch": 0.043582785870703605,
+      "grad_norm": 4.5013456923615305,
+      "learning_rate": 2e-05,
+      "loss": 0.2051,
+      "step": 1831
+    },
+    {
+      "epoch": 0.043606588593735125,
+      "grad_norm": 29.843133425023304,
+      "learning_rate": 2e-05,
+      "loss": 1.1547,
+      "step": 1832
+    },
+    {
+      "epoch": 0.04363039131676664,
+      "grad_norm": 5.426134350335508,
+      "learning_rate": 2e-05,
+      "loss": 0.754,
+      "step": 1833
+    },
+    {
+      "epoch": 0.04365419403979815,
+      "grad_norm": 23.098267207682174,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 1834
+    },
+    {
+      "epoch": 0.04367799676282967,
+      "grad_norm": 3.6652968818404887,
+      "learning_rate": 2e-05,
+      "loss": 0.3077,
+      "step": 1835
+    },
+    {
+      "epoch": 0.043701799485861184,
+      "grad_norm": 9.882168621890921,
+      "learning_rate": 2e-05,
+      "loss": 1.3688,
+      "step": 1836
+    },
+    {
+      "epoch": 0.0437256022088927,
+      "grad_norm": 79.93393783471004,
+      "learning_rate": 2e-05,
+      "loss": 1.0792,
+      "step": 1837
+    },
+    {
+      "epoch": 0.04374940493192421,
+      "grad_norm": 3.1529572075967134,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 1838
+    },
+    {
+      "epoch": 0.04377320765495573,
+      "grad_norm": 17.124196678064777,
+      "learning_rate": 2e-05,
+      "loss": 0.5738,
+      "step": 1839
+    },
+    {
+      "epoch": 0.04379701037798724,
+      "grad_norm": 4.328729560779248,
+      "learning_rate": 2e-05,
+      "loss": 0.2264,
+      "step": 1840
+    },
+    {
+      "epoch": 0.043820813101018756,
+      "grad_norm": 3.041976512390642,
+      "learning_rate": 2e-05,
+      "loss": 0.183,
+      "step": 1841
+    },
+    {
+      "epoch": 0.04384461582405027,
+      "grad_norm": 24.815954093438144,
+      "learning_rate": 2e-05,
+      "loss": 0.3973,
+      "step": 1842
+    },
+    {
+      "epoch": 0.04386841854708179,
+      "grad_norm": 6.6480752979728575,
+      "learning_rate": 2e-05,
+      "loss": 0.3805,
+      "step": 1843
+    },
+    {
+      "epoch": 0.0438922212701133,
+      "grad_norm": 14.36715718445915,
+      "learning_rate": 2e-05,
+      "loss": 0.8449,
+      "step": 1844
+    },
+    {
+      "epoch": 0.043916023993144815,
+      "grad_norm": 7.990095475471983,
+      "learning_rate": 2e-05,
+      "loss": 0.4955,
+      "step": 1845
+    },
+    {
+      "epoch": 0.04393982671617633,
+      "grad_norm": 12.089145289291181,
+      "learning_rate": 2e-05,
+      "loss": 0.9408,
+      "step": 1846
+    },
+    {
+      "epoch": 0.04396362943920785,
+      "grad_norm": 2.890476286174944,
+      "learning_rate": 2e-05,
+      "loss": 0.2102,
+      "step": 1847
+    },
+    {
+      "epoch": 0.04398743216223936,
+      "grad_norm": 6.351143368240524,
+      "learning_rate": 2e-05,
+      "loss": 0.2884,
+      "step": 1848
+    },
+    {
+      "epoch": 0.044011234885270874,
+      "grad_norm": 7.294022083191216,
+      "learning_rate": 2e-05,
+      "loss": 1.2568,
+      "step": 1849
+    },
+    {
+      "epoch": 0.04403503760830239,
+      "grad_norm": 11.291928643836014,
+      "learning_rate": 2e-05,
+      "loss": 1.4941,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04405884033133391,
+      "grad_norm": 7.978618921043193,
+      "learning_rate": 2e-05,
+      "loss": 0.9897,
+      "step": 1851
+    },
+    {
+      "epoch": 0.04408264305436542,
+      "grad_norm": 7.080196892851147,
+      "learning_rate": 2e-05,
+      "loss": 0.2843,
+      "step": 1852
+    },
+    {
+      "epoch": 0.04410644577739693,
+      "grad_norm": 19.685518939783158,
+      "learning_rate": 2e-05,
+      "loss": 0.4251,
+      "step": 1853
+    },
+    {
+      "epoch": 0.044130248500428446,
+      "grad_norm": 8.98266392501956,
+      "learning_rate": 2e-05,
+      "loss": 1.051,
+      "step": 1854
+    },
+    {
+      "epoch": 0.044154051223459966,
+      "grad_norm": 8.59321192727921,
+      "learning_rate": 2e-05,
+      "loss": 1.2826,
+      "step": 1855
+    },
+    {
+      "epoch": 0.04417785394649148,
+      "grad_norm": 12.115549402667977,
+      "learning_rate": 2e-05,
+      "loss": 1.2025,
+      "step": 1856
+    },
+    {
+      "epoch": 0.04420165666952299,
+      "grad_norm": 5.308416424120916,
+      "learning_rate": 2e-05,
+      "loss": 0.4231,
+      "step": 1857
+    },
+    {
+      "epoch": 0.04422545939255451,
+      "grad_norm": 9.425917101624403,
+      "learning_rate": 2e-05,
+      "loss": 2.3268,
+      "step": 1858
+    },
+    {
+      "epoch": 0.044249262115586024,
+      "grad_norm": 6.311705321267611,
+      "learning_rate": 2e-05,
+      "loss": 0.9771,
+      "step": 1859
+    },
+    {
+      "epoch": 0.04427306483861754,
+      "grad_norm": 6.512323069031171,
+      "learning_rate": 2e-05,
+      "loss": 0.5735,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04429686756164905,
+      "grad_norm": 32.88622243014469,
+      "learning_rate": 2e-05,
+      "loss": 0.9699,
+      "step": 1861
+    },
+    {
+      "epoch": 0.04432067028468057,
+      "grad_norm": 10.138804866201841,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 1862
+    },
+    {
+      "epoch": 0.04434447300771208,
+      "grad_norm": 49.86190320391561,
+      "learning_rate": 2e-05,
+      "loss": 0.5673,
+      "step": 1863
+    },
+    {
+      "epoch": 0.044368275730743596,
+      "grad_norm": 35.29736363496885,
+      "learning_rate": 2e-05,
+      "loss": 0.9962,
+      "step": 1864
+    },
+    {
+      "epoch": 0.04439207845377511,
+      "grad_norm": 7.368461910628824,
+      "learning_rate": 2e-05,
+      "loss": 0.4329,
+      "step": 1865
+    },
+    {
+      "epoch": 0.04441588117680663,
+      "grad_norm": 5.176313266127242,
+      "learning_rate": 2e-05,
+      "loss": 0.3448,
+      "step": 1866
+    },
+    {
+      "epoch": 0.04443968389983814,
+      "grad_norm": 10.523928420165882,
+      "learning_rate": 2e-05,
+      "loss": 1.0922,
+      "step": 1867
+    },
+    {
+      "epoch": 0.044463486622869655,
+      "grad_norm": 7.339943907776276,
+      "learning_rate": 2e-05,
+      "loss": 0.9868,
+      "step": 1868
+    },
+    {
+      "epoch": 0.04448728934590117,
+      "grad_norm": 4.623783315225731,
+      "learning_rate": 2e-05,
+      "loss": 0.4659,
+      "step": 1869
+    },
+    {
+      "epoch": 0.04451109206893269,
+      "grad_norm": 12.676972967604451,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0445348947919642,
+      "grad_norm": 6.609766226791667,
+      "learning_rate": 2e-05,
+      "loss": 0.436,
+      "step": 1871
+    },
+    {
+      "epoch": 0.044558697514995714,
+      "grad_norm": 9.83142615520806,
+      "learning_rate": 2e-05,
+      "loss": 0.9461,
+      "step": 1872
+    },
+    {
+      "epoch": 0.04458250023802723,
+      "grad_norm": 13.678763462920411,
+      "learning_rate": 2e-05,
+      "loss": 0.8826,
+      "step": 1873
+    },
+    {
+      "epoch": 0.04460630296105875,
+      "grad_norm": 7.31048415674288,
+      "learning_rate": 2e-05,
+      "loss": 0.6037,
+      "step": 1874
+    },
+    {
+      "epoch": 0.04463010568409026,
+      "grad_norm": 2.3302106776864266,
+      "learning_rate": 2e-05,
+      "loss": 0.3192,
+      "step": 1875
+    },
+    {
+      "epoch": 0.04465390840712177,
+      "grad_norm": 7.266343098670132,
+      "learning_rate": 2e-05,
+      "loss": 0.881,
+      "step": 1876
+    },
+    {
+      "epoch": 0.04467771113015329,
+      "grad_norm": 6.558647990607909,
+      "learning_rate": 2e-05,
+      "loss": 0.2559,
+      "step": 1877
+    },
+    {
+      "epoch": 0.044701513853184806,
+      "grad_norm": 5.332283690038135,
+      "learning_rate": 2e-05,
+      "loss": 0.8521,
+      "step": 1878
+    },
+    {
+      "epoch": 0.04472531657621632,
+      "grad_norm": 3.0505429409797316,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 1879
+    },
+    {
+      "epoch": 0.04474911929924783,
+      "grad_norm": 2.324678794647983,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04477292202227935,
+      "grad_norm": 29.69823381229704,
+      "learning_rate": 2e-05,
+      "loss": 0.9447,
+      "step": 1881
+    },
+    {
+      "epoch": 0.044796724745310865,
+      "grad_norm": 30.25828873025817,
+      "learning_rate": 2e-05,
+      "loss": 0.2748,
+      "step": 1882
+    },
+    {
+      "epoch": 0.04482052746834238,
+      "grad_norm": 4.2285081108712514,
+      "learning_rate": 2e-05,
+      "loss": 0.2631,
+      "step": 1883
+    },
+    {
+      "epoch": 0.04484433019137389,
+      "grad_norm": 13.971257385250054,
+      "learning_rate": 2e-05,
+      "loss": 0.4627,
+      "step": 1884
+    },
+    {
+      "epoch": 0.04486813291440541,
+      "grad_norm": 3.2580294748462895,
+      "learning_rate": 2e-05,
+      "loss": 0.1934,
+      "step": 1885
+    },
+    {
+      "epoch": 0.044891935637436924,
+      "grad_norm": 6.509049672611028,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 1886
+    },
+    {
+      "epoch": 0.04491573836046844,
+      "grad_norm": 1.7781403451010578,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 1887
+    },
+    {
+      "epoch": 0.04493954108349995,
+      "grad_norm": 4.212559356196415,
+      "learning_rate": 2e-05,
+      "loss": 0.2859,
+      "step": 1888
+    },
+    {
+      "epoch": 0.04496334380653147,
+      "grad_norm": 1.8976102097620473,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 1889
+    },
+    {
+      "epoch": 0.04498714652956298,
+      "grad_norm": 4.290609221422359,
+      "learning_rate": 2e-05,
+      "loss": 0.1925,
+      "step": 1890
+    },
+    {
+      "epoch": 0.045010949252594495,
+      "grad_norm": 5.2339243586317865,
+      "learning_rate": 2e-05,
+      "loss": 0.6806,
+      "step": 1891
+    },
+    {
+      "epoch": 0.04503475197562601,
+      "grad_norm": 4.256618995674515,
+      "learning_rate": 2e-05,
+      "loss": 0.186,
+      "step": 1892
+    },
+    {
+      "epoch": 0.04505855469865753,
+      "grad_norm": 3.8210136699161197,
+      "learning_rate": 2e-05,
+      "loss": 0.239,
+      "step": 1893
+    },
+    {
+      "epoch": 0.04508235742168904,
+      "grad_norm": 3.6290958960096775,
+      "learning_rate": 2e-05,
+      "loss": 0.1604,
+      "step": 1894
+    },
+    {
+      "epoch": 0.045106160144720554,
+      "grad_norm": 9.184416547639033,
+      "learning_rate": 2e-05,
+      "loss": 1.9684,
+      "step": 1895
+    },
+    {
+      "epoch": 0.045129962867752074,
+      "grad_norm": 3.612721013667084,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 1896
+    },
+    {
+      "epoch": 0.04515376559078359,
+      "grad_norm": 5.925131675521893,
+      "learning_rate": 2e-05,
+      "loss": 0.4789,
+      "step": 1897
+    },
+    {
+      "epoch": 0.0451775683138151,
+      "grad_norm": 14.60785125150017,
+      "learning_rate": 2e-05,
+      "loss": 2.1285,
+      "step": 1898
+    },
+    {
+      "epoch": 0.04520137103684661,
+      "grad_norm": 6.465844353368751,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 1899
+    },
+    {
+      "epoch": 0.04522517375987813,
+      "grad_norm": 7.366251120733034,
+      "learning_rate": 2e-05,
+      "loss": 0.6003,
+      "step": 1900
+    },
+    {
+      "epoch": 0.045248976482909646,
+      "grad_norm": 18.752851328204294,
+      "learning_rate": 2e-05,
+      "loss": 1.0901,
+      "step": 1901
+    },
+    {
+      "epoch": 0.04527277920594116,
+      "grad_norm": 4.523093357020508,
+      "learning_rate": 2e-05,
+      "loss": 0.1838,
+      "step": 1902
+    },
+    {
+      "epoch": 0.04529658192897267,
+      "grad_norm": 3.486937163473382,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 1903
+    },
+    {
+      "epoch": 0.04532038465200419,
+      "grad_norm": 7.905924534677884,
+      "learning_rate": 2e-05,
+      "loss": 0.844,
+      "step": 1904
+    },
+    {
+      "epoch": 0.045344187375035705,
+      "grad_norm": 4.876687162177432,
+      "learning_rate": 2e-05,
+      "loss": 1.0373,
+      "step": 1905
+    },
+    {
+      "epoch": 0.04536799009806722,
+      "grad_norm": 15.464419332922883,
+      "learning_rate": 2e-05,
+      "loss": 0.7145,
+      "step": 1906
+    },
+    {
+      "epoch": 0.04539179282109873,
+      "grad_norm": 3.2948332260846924,
+      "learning_rate": 2e-05,
+      "loss": 0.2259,
+      "step": 1907
+    },
+    {
+      "epoch": 0.04541559554413025,
+      "grad_norm": 5.744274798601534,
+      "learning_rate": 2e-05,
+      "loss": 0.5456,
+      "step": 1908
+    },
+    {
+      "epoch": 0.045439398267161764,
+      "grad_norm": 1.540452888488954,
+      "learning_rate": 2e-05,
+      "loss": 0.1387,
+      "step": 1909
+    },
+    {
+      "epoch": 0.04546320099019328,
+      "grad_norm": 8.915554910688934,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04548700371322479,
+      "grad_norm": 8.622246833788257,
+      "learning_rate": 2e-05,
+      "loss": 0.8413,
+      "step": 1911
+    },
+    {
+      "epoch": 0.04551080643625631,
+      "grad_norm": 3.64912262736059,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 1912
+    },
+    {
+      "epoch": 0.04553460915928782,
+      "grad_norm": 3.444143316790823,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 1913
+    },
+    {
+      "epoch": 0.045558411882319336,
+      "grad_norm": 2.838575853759494,
+      "learning_rate": 2e-05,
+      "loss": 0.2294,
+      "step": 1914
+    },
+    {
+      "epoch": 0.04558221460535085,
+      "grad_norm": 12.382419535641485,
+      "learning_rate": 2e-05,
+      "loss": 0.4325,
+      "step": 1915
+    },
+    {
+      "epoch": 0.04560601732838237,
+      "grad_norm": 1.6301939900900129,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 1916
+    },
+    {
+      "epoch": 0.04562982005141388,
+      "grad_norm": 7.443799413341765,
+      "learning_rate": 2e-05,
+      "loss": 1.6386,
+      "step": 1917
+    },
+    {
+      "epoch": 0.045653622774445395,
+      "grad_norm": 7.924480552361493,
+      "learning_rate": 2e-05,
+      "loss": 0.996,
+      "step": 1918
+    },
+    {
+      "epoch": 0.045677425497476914,
+      "grad_norm": 8.232837162764186,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 1919
+    },
+    {
+      "epoch": 0.04570122822050843,
+      "grad_norm": 5.153655607866883,
+      "learning_rate": 2e-05,
+      "loss": 0.6801,
+      "step": 1920
+    },
+    {
+      "epoch": 0.04572503094353994,
+      "grad_norm": 6.980166353825178,
+      "learning_rate": 2e-05,
+      "loss": 0.3891,
+      "step": 1921
+    },
+    {
+      "epoch": 0.04574883366657145,
+      "grad_norm": 10.065030966486813,
+      "learning_rate": 2e-05,
+      "loss": 1.9415,
+      "step": 1922
+    },
+    {
+      "epoch": 0.04577263638960297,
+      "grad_norm": 8.445697757280705,
+      "learning_rate": 2e-05,
+      "loss": 0.6545,
+      "step": 1923
+    },
+    {
+      "epoch": 0.045796439112634486,
+      "grad_norm": 2.9685730520355267,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 1924
+    },
+    {
+      "epoch": 0.045820241835666,
+      "grad_norm": 6.815006274888465,
+      "learning_rate": 2e-05,
+      "loss": 0.3249,
+      "step": 1925
+    },
+    {
+      "epoch": 0.04584404455869751,
+      "grad_norm": 5.755640272399018,
+      "learning_rate": 2e-05,
+      "loss": 0.3254,
+      "step": 1926
+    },
+    {
+      "epoch": 0.04586784728172903,
+      "grad_norm": 5.42695882989913,
+      "learning_rate": 2e-05,
+      "loss": 0.8243,
+      "step": 1927
+    },
+    {
+      "epoch": 0.045891650004760545,
+      "grad_norm": 3.887463741709913,
+      "learning_rate": 2e-05,
+      "loss": 0.2525,
+      "step": 1928
+    },
+    {
+      "epoch": 0.04591545272779206,
+      "grad_norm": 9.485071536748931,
+      "learning_rate": 2e-05,
+      "loss": 0.4557,
+      "step": 1929
+    },
+    {
+      "epoch": 0.04593925545082357,
+      "grad_norm": 2.2388020049548976,
+      "learning_rate": 2e-05,
+      "loss": 0.1589,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04596305817385509,
+      "grad_norm": 2.6710909869404795,
+      "learning_rate": 2e-05,
+      "loss": 0.2642,
+      "step": 1931
+    },
+    {
+      "epoch": 0.045986860896886604,
+      "grad_norm": 8.331497516951325,
+      "learning_rate": 2e-05,
+      "loss": 0.4127,
+      "step": 1932
+    },
+    {
+      "epoch": 0.04601066361991812,
+      "grad_norm": 8.075492412148243,
+      "learning_rate": 2e-05,
+      "loss": 1.3227,
+      "step": 1933
+    },
+    {
+      "epoch": 0.04603446634294963,
+      "grad_norm": 5.387806575192997,
+      "learning_rate": 2e-05,
+      "loss": 0.6738,
+      "step": 1934
+    },
+    {
+      "epoch": 0.04605826906598115,
+      "grad_norm": 9.70745926514669,
+      "learning_rate": 2e-05,
+      "loss": 0.4271,
+      "step": 1935
+    },
+    {
+      "epoch": 0.04608207178901266,
+      "grad_norm": 7.072785706508516,
+      "learning_rate": 2e-05,
+      "loss": 1.1978,
+      "step": 1936
+    },
+    {
+      "epoch": 0.046105874512044176,
+      "grad_norm": 3.2146505326345465,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 1937
+    },
+    {
+      "epoch": 0.046129677235075696,
+      "grad_norm": 6.3420967272157025,
+      "learning_rate": 2e-05,
+      "loss": 0.3043,
+      "step": 1938
+    },
+    {
+      "epoch": 0.04615347995810721,
+      "grad_norm": 3.720484720301169,
+      "learning_rate": 2e-05,
+      "loss": 0.7124,
+      "step": 1939
+    },
+    {
+      "epoch": 0.04617728268113872,
+      "grad_norm": 10.362123452599143,
+      "learning_rate": 2e-05,
+      "loss": 1.6814,
+      "step": 1940
+    },
+    {
+      "epoch": 0.046201085404170235,
+      "grad_norm": 9.808575544908251,
+      "learning_rate": 2e-05,
+      "loss": 1.1511,
+      "step": 1941
+    },
+    {
+      "epoch": 0.046224888127201755,
+      "grad_norm": 4.85545429012606,
+      "learning_rate": 2e-05,
+      "loss": 0.2404,
+      "step": 1942
+    },
+    {
+      "epoch": 0.04624869085023327,
+      "grad_norm": 3.475085827940691,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 1943
+    },
+    {
+      "epoch": 0.04627249357326478,
+      "grad_norm": 9.544931489759149,
+      "learning_rate": 2e-05,
+      "loss": 1.3147,
+      "step": 1944
+    },
+    {
+      "epoch": 0.046296296296296294,
+      "grad_norm": 3.6559771091617437,
+      "learning_rate": 2e-05,
+      "loss": 0.2926,
+      "step": 1945
+    },
+    {
+      "epoch": 0.046320099019327814,
+      "grad_norm": 7.008608092310621,
+      "learning_rate": 2e-05,
+      "loss": 0.3057,
+      "step": 1946
+    },
+    {
+      "epoch": 0.04634390174235933,
+      "grad_norm": 6.934961176688675,
+      "learning_rate": 2e-05,
+      "loss": 0.2553,
+      "step": 1947
+    },
+    {
+      "epoch": 0.04636770446539084,
+      "grad_norm": 8.266486480420436,
+      "learning_rate": 2e-05,
+      "loss": 0.9275,
+      "step": 1948
+    },
+    {
+      "epoch": 0.04639150718842235,
+      "grad_norm": 8.656469476874866,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 1949
+    },
+    {
+      "epoch": 0.04641530991145387,
+      "grad_norm": 10.125014641687613,
+      "learning_rate": 2e-05,
+      "loss": 1.5049,
+      "step": 1950
+    },
+    {
+      "epoch": 0.046439112634485386,
+      "grad_norm": 6.243747835726927,
+      "learning_rate": 2e-05,
+      "loss": 0.9813,
+      "step": 1951
+    },
+    {
+      "epoch": 0.0464629153575169,
+      "grad_norm": 3.0967417286422543,
+      "learning_rate": 2e-05,
+      "loss": 0.2076,
+      "step": 1952
+    },
+    {
+      "epoch": 0.04648671808054841,
+      "grad_norm": 8.866086806548616,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 1953
+    },
+    {
+      "epoch": 0.04651052080357993,
+      "grad_norm": 7.306618497367981,
+      "learning_rate": 2e-05,
+      "loss": 0.7555,
+      "step": 1954
+    },
+    {
+      "epoch": 0.046534323526611444,
+      "grad_norm": 3.4136727034799836,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 1955
+    },
+    {
+      "epoch": 0.04655812624964296,
+      "grad_norm": 4.357066387301325,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 1956
+    },
+    {
+      "epoch": 0.04658192897267448,
+      "grad_norm": 5.665689152496913,
+      "learning_rate": 2e-05,
+      "loss": 1.0266,
+      "step": 1957
+    },
+    {
+      "epoch": 0.04660573169570599,
+      "grad_norm": 7.516785561955892,
+      "learning_rate": 2e-05,
+      "loss": 0.3222,
+      "step": 1958
+    },
+    {
+      "epoch": 0.0466295344187375,
+      "grad_norm": 5.798763075428433,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 1959
+    },
+    {
+      "epoch": 0.046653337141769016,
+      "grad_norm": 2.0138869539288757,
+      "learning_rate": 2e-05,
+      "loss": 0.2376,
+      "step": 1960
+    },
+    {
+      "epoch": 0.046677139864800536,
+      "grad_norm": 6.91257371802876,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 1961
+    },
+    {
+      "epoch": 0.04670094258783205,
+      "grad_norm": 2.5939705873906376,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 1962
+    },
+    {
+      "epoch": 0.04672474531086356,
+      "grad_norm": 5.938137786802548,
+      "learning_rate": 2e-05,
+      "loss": 0.6615,
+      "step": 1963
+    },
+    {
+      "epoch": 0.046748548033895075,
+      "grad_norm": 3.7461177303390407,
+      "learning_rate": 2e-05,
+      "loss": 0.2997,
+      "step": 1964
+    },
+    {
+      "epoch": 0.046772350756926595,
+      "grad_norm": 8.689344607161255,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 1965
+    },
+    {
+      "epoch": 0.04679615347995811,
+      "grad_norm": 2.966770656685401,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 1966
+    },
+    {
+      "epoch": 0.04681995620298962,
+      "grad_norm": 4.1571603303651985,
+      "learning_rate": 2e-05,
+      "loss": 0.2544,
+      "step": 1967
+    },
+    {
+      "epoch": 0.046843758926021134,
+      "grad_norm": 12.540643760380025,
+      "learning_rate": 2e-05,
+      "loss": 0.4997,
+      "step": 1968
+    },
+    {
+      "epoch": 0.046867561649052654,
+      "grad_norm": 6.363077198597244,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 1969
+    },
+    {
+      "epoch": 0.04689136437208417,
+      "grad_norm": 8.716368127014878,
+      "learning_rate": 2e-05,
+      "loss": 0.5418,
+      "step": 1970
+    },
+    {
+      "epoch": 0.04691516709511568,
+      "grad_norm": 2.427342956755388,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 1971
+    },
+    {
+      "epoch": 0.04693896981814719,
+      "grad_norm": 7.136776532375391,
+      "learning_rate": 2e-05,
+      "loss": 0.2685,
+      "step": 1972
+    },
+    {
+      "epoch": 0.04696277254117871,
+      "grad_norm": 6.2528277069104155,
+      "learning_rate": 2e-05,
+      "loss": 0.9113,
+      "step": 1973
+    },
+    {
+      "epoch": 0.046986575264210226,
+      "grad_norm": 9.973184312171897,
+      "learning_rate": 2e-05,
+      "loss": 1.1199,
+      "step": 1974
+    },
+    {
+      "epoch": 0.04701037798724174,
+      "grad_norm": 9.25065036000659,
+      "learning_rate": 2e-05,
+      "loss": 0.5792,
+      "step": 1975
+    },
+    {
+      "epoch": 0.04703418071027326,
+      "grad_norm": 9.38162995201773,
+      "learning_rate": 2e-05,
+      "loss": 0.845,
+      "step": 1976
+    },
+    {
+      "epoch": 0.04705798343330477,
+      "grad_norm": 2.751198651347678,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 1977
+    },
+    {
+      "epoch": 0.047081786156336285,
+      "grad_norm": 7.048105920536659,
+      "learning_rate": 2e-05,
+      "loss": 0.9302,
+      "step": 1978
+    },
+    {
+      "epoch": 0.0471055888793678,
+      "grad_norm": 6.4190771312516075,
+      "learning_rate": 2e-05,
+      "loss": 1.0053,
+      "step": 1979
+    },
+    {
+      "epoch": 0.04712939160239932,
+      "grad_norm": 6.472487108972938,
+      "learning_rate": 2e-05,
+      "loss": 0.3592,
+      "step": 1980
+    },
+    {
+      "epoch": 0.04715319432543083,
+      "grad_norm": 18.490430739679145,
+      "learning_rate": 2e-05,
+      "loss": 0.7177,
+      "step": 1981
+    },
+    {
+      "epoch": 0.047176997048462344,
+      "grad_norm": 6.639710416316449,
+      "learning_rate": 2e-05,
+      "loss": 0.922,
+      "step": 1982
+    },
+    {
+      "epoch": 0.047200799771493857,
+      "grad_norm": 4.587224750255897,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 1983
+    },
+    {
+      "epoch": 0.047224602494525376,
+      "grad_norm": 6.883303888903421,
+      "learning_rate": 2e-05,
+      "loss": 0.8666,
+      "step": 1984
+    },
+    {
+      "epoch": 0.04724840521755689,
+      "grad_norm": 2.968214828261328,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 1985
+    },
+    {
+      "epoch": 0.0472722079405884,
+      "grad_norm": 12.848212288478063,
+      "learning_rate": 2e-05,
+      "loss": 0.4326,
+      "step": 1986
+    },
+    {
+      "epoch": 0.047296010663619915,
+      "grad_norm": 2.872050201066074,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 1987
+    },
+    {
+      "epoch": 0.047319813386651435,
+      "grad_norm": 3.809049396339959,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 1988
+    },
+    {
+      "epoch": 0.04734361610968295,
+      "grad_norm": 2.940035527784215,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 1989
+    },
+    {
+      "epoch": 0.04736741883271446,
+      "grad_norm": 5.9435316129498315,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 1990
+    },
+    {
+      "epoch": 0.047391221555745974,
+      "grad_norm": 2.0808520460884794,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 1991
+    },
+    {
+      "epoch": 0.047415024278777494,
+      "grad_norm": 6.776739220969905,
+      "learning_rate": 2e-05,
+      "loss": 0.381,
+      "step": 1992
+    },
+    {
+      "epoch": 0.04743882700180901,
+      "grad_norm": 6.822360726727605,
+      "learning_rate": 2e-05,
+      "loss": 0.3661,
+      "step": 1993
+    },
+    {
+      "epoch": 0.04746262972484052,
+      "grad_norm": 6.878145155581667,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 1994
+    },
+    {
+      "epoch": 0.04748643244787203,
+      "grad_norm": 4.805299838028359,
+      "learning_rate": 2e-05,
+      "loss": 0.4131,
+      "step": 1995
+    },
+    {
+      "epoch": 0.04751023517090355,
+      "grad_norm": 1.76839121422096,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 1996
+    },
+    {
+      "epoch": 0.047534037893935066,
+      "grad_norm": 4.30319815872761,
+      "learning_rate": 2e-05,
+      "loss": 0.3545,
+      "step": 1997
+    },
+    {
+      "epoch": 0.04755784061696658,
+      "grad_norm": 5.781195994777884,
+      "learning_rate": 2e-05,
+      "loss": 0.4565,
+      "step": 1998
+    },
+    {
+      "epoch": 0.0475816433399981,
+      "grad_norm": 1.7823358137338174,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 1999
+    },
+    {
+      "epoch": 0.04760544606302961,
+      "grad_norm": 2.280740574648784,
+      "learning_rate": 2e-05,
+      "loss": 0.1653,
+      "step": 2000
+    },
+    {
+      "epoch": 0.047629248786061125,
+      "grad_norm": 7.225253444413657,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 2001
+    },
+    {
+      "epoch": 0.04765305150909264,
+      "grad_norm": 2.5908012462629024,
+      "learning_rate": 2e-05,
+      "loss": 0.278,
+      "step": 2002
+    },
+    {
+      "epoch": 0.04767685423212416,
+      "grad_norm": 5.655905194219604,
+      "learning_rate": 2e-05,
+      "loss": 0.7265,
+      "step": 2003
+    },
+    {
+      "epoch": 0.04770065695515567,
+      "grad_norm": 4.751809566543113,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 2004
+    },
+    {
+      "epoch": 0.047724459678187184,
+      "grad_norm": 2.7849259714743355,
+      "learning_rate": 2e-05,
+      "loss": 0.1845,
+      "step": 2005
+    },
+    {
+      "epoch": 0.0477482624012187,
+      "grad_norm": 9.786096042607596,
+      "learning_rate": 2e-05,
+      "loss": 0.4893,
+      "step": 2006
+    },
+    {
+      "epoch": 0.04777206512425022,
+      "grad_norm": 7.5468664623060375,
+      "learning_rate": 2e-05,
+      "loss": 0.9664,
+      "step": 2007
+    },
+    {
+      "epoch": 0.04779586784728173,
+      "grad_norm": 2.776627495611324,
+      "learning_rate": 2e-05,
+      "loss": 0.236,
+      "step": 2008
+    },
+    {
+      "epoch": 0.04781967057031324,
+      "grad_norm": 8.645835134012202,
+      "learning_rate": 2e-05,
+      "loss": 1.083,
+      "step": 2009
+    },
+    {
+      "epoch": 0.047843473293344756,
+      "grad_norm": 7.365430497991341,
+      "learning_rate": 2e-05,
+      "loss": 0.3656,
+      "step": 2010
+    },
+    {
+      "epoch": 0.047867276016376276,
+      "grad_norm": 5.057470078943005,
+      "learning_rate": 2e-05,
+      "loss": 0.4939,
+      "step": 2011
+    },
+    {
+      "epoch": 0.04789107873940779,
+      "grad_norm": 6.9047327590888745,
+      "learning_rate": 2e-05,
+      "loss": 0.4552,
+      "step": 2012
+    },
+    {
+      "epoch": 0.0479148814624393,
+      "grad_norm": 3.173077142190149,
+      "learning_rate": 2e-05,
+      "loss": 0.2945,
+      "step": 2013
+    },
+    {
+      "epoch": 0.047938684185470815,
+      "grad_norm": 8.202524927990837,
+      "learning_rate": 2e-05,
+      "loss": 0.8448,
+      "step": 2014
+    },
+    {
+      "epoch": 0.047962486908502334,
+      "grad_norm": 5.213981000315978,
+      "learning_rate": 2e-05,
+      "loss": 0.6301,
+      "step": 2015
+    },
+    {
+      "epoch": 0.04798628963153385,
+      "grad_norm": 8.20818174955678,
+      "learning_rate": 2e-05,
+      "loss": 1.3783,
+      "step": 2016
+    },
+    {
+      "epoch": 0.04801009235456536,
+      "grad_norm": 9.929611726853222,
+      "learning_rate": 2e-05,
+      "loss": 1.3374,
+      "step": 2017
+    },
+    {
+      "epoch": 0.04803389507759688,
+      "grad_norm": 7.851038599185398,
+      "learning_rate": 2e-05,
+      "loss": 0.4746,
+      "step": 2018
+    },
+    {
+      "epoch": 0.04805769780062839,
+      "grad_norm": 3.2129489048733384,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 2019
+    },
+    {
+      "epoch": 0.048081500523659906,
+      "grad_norm": 2.7761442493347226,
+      "learning_rate": 2e-05,
+      "loss": 0.2865,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04810530324669142,
+      "grad_norm": 8.074671567242493,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 2021
+    },
+    {
+      "epoch": 0.04812910596972294,
+      "grad_norm": 9.520497596614673,
+      "learning_rate": 2e-05,
+      "loss": 0.7576,
+      "step": 2022
+    },
+    {
+      "epoch": 0.04815290869275445,
+      "grad_norm": 6.367758424521348,
+      "learning_rate": 2e-05,
+      "loss": 0.4312,
+      "step": 2023
+    },
+    {
+      "epoch": 0.048176711415785965,
+      "grad_norm": 3.8194628217107183,
+      "learning_rate": 2e-05,
+      "loss": 0.1893,
+      "step": 2024
+    },
+    {
+      "epoch": 0.04820051413881748,
+      "grad_norm": 8.146853996658681,
+      "learning_rate": 2e-05,
+      "loss": 1.3826,
+      "step": 2025
+    },
+    {
+      "epoch": 0.048224316861849,
+      "grad_norm": 4.272605999903501,
+      "learning_rate": 2e-05,
+      "loss": 0.2079,
+      "step": 2026
+    },
+    {
+      "epoch": 0.04824811958488051,
+      "grad_norm": 7.945246399167278,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 2027
+    },
+    {
+      "epoch": 0.048271922307912024,
+      "grad_norm": 4.028443456096617,
+      "learning_rate": 2e-05,
+      "loss": 0.3303,
+      "step": 2028
+    },
+    {
+      "epoch": 0.04829572503094354,
+      "grad_norm": 8.630472896436501,
+      "learning_rate": 2e-05,
+      "loss": 1.7849,
+      "step": 2029
+    },
+    {
+      "epoch": 0.04831952775397506,
+      "grad_norm": 6.092873195078469,
+      "learning_rate": 2e-05,
+      "loss": 0.9053,
+      "step": 2030
+    },
+    {
+      "epoch": 0.04834333047700657,
+      "grad_norm": 7.878151611497844,
+      "learning_rate": 2e-05,
+      "loss": 0.6064,
+      "step": 2031
+    },
+    {
+      "epoch": 0.04836713320003808,
+      "grad_norm": 1.0617284669301466,
+      "learning_rate": 2e-05,
+      "loss": 0.1953,
+      "step": 2032
+    },
+    {
+      "epoch": 0.048390935923069596,
+      "grad_norm": 7.3891094998635305,
+      "learning_rate": 2e-05,
+      "loss": 0.3629,
+      "step": 2033
+    },
+    {
+      "epoch": 0.048414738646101116,
+      "grad_norm": 5.1879474729552,
+      "learning_rate": 2e-05,
+      "loss": 0.867,
+      "step": 2034
+    },
+    {
+      "epoch": 0.04843854136913263,
+      "grad_norm": 3.925048612449574,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 2035
+    },
+    {
+      "epoch": 0.04846234409216414,
+      "grad_norm": 2.337101351340266,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 2036
+    },
+    {
+      "epoch": 0.04848614681519566,
+      "grad_norm": 4.948896962796004,
+      "learning_rate": 2e-05,
+      "loss": 0.4775,
+      "step": 2037
+    },
+    {
+      "epoch": 0.048509949538227175,
+      "grad_norm": 3.078536281724137,
+      "learning_rate": 2e-05,
+      "loss": 0.2301,
+      "step": 2038
+    },
+    {
+      "epoch": 0.04853375226125869,
+      "grad_norm": 9.802767411415477,
+      "learning_rate": 2e-05,
+      "loss": 1.2877,
+      "step": 2039
+    },
+    {
+      "epoch": 0.0485575549842902,
+      "grad_norm": 9.176220785602638,
+      "learning_rate": 2e-05,
+      "loss": 0.7712,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04858135770732172,
+      "grad_norm": 6.472001117986856,
+      "learning_rate": 2e-05,
+      "loss": 0.7858,
+      "step": 2041
+    },
+    {
+      "epoch": 0.048605160430353234,
+      "grad_norm": 4.136602587242384,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 2042
+    },
+    {
+      "epoch": 0.04862896315338475,
+      "grad_norm": 3.0612619738705567,
+      "learning_rate": 2e-05,
+      "loss": 0.177,
+      "step": 2043
+    },
+    {
+      "epoch": 0.04865276587641626,
+      "grad_norm": 7.973648442610219,
+      "learning_rate": 2e-05,
+      "loss": 0.3935,
+      "step": 2044
+    },
+    {
+      "epoch": 0.04867656859944778,
+      "grad_norm": 11.61797528627197,
+      "learning_rate": 2e-05,
+      "loss": 0.5326,
+      "step": 2045
+    },
+    {
+      "epoch": 0.04870037132247929,
+      "grad_norm": 4.332534884963099,
+      "learning_rate": 2e-05,
+      "loss": 0.4109,
+      "step": 2046
+    },
+    {
+      "epoch": 0.048724174045510805,
+      "grad_norm": 12.320750982983219,
+      "learning_rate": 2e-05,
+      "loss": 0.449,
+      "step": 2047
+    },
+    {
+      "epoch": 0.04874797676854232,
+      "grad_norm": 4.15130688318919,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 2048
+    },
+    {
+      "epoch": 0.04877177949157384,
+      "grad_norm": 4.357811932655857,
+      "learning_rate": 2e-05,
+      "loss": 0.3258,
+      "step": 2049
+    },
+    {
+      "epoch": 0.04879558221460535,
+      "grad_norm": 2.7247469030696925,
+      "learning_rate": 2e-05,
+      "loss": 0.2738,
+      "step": 2050
+    },
+    {
+      "epoch": 0.048819384937636864,
+      "grad_norm": 82.08477281226287,
+      "learning_rate": 2e-05,
+      "loss": 0.4925,
+      "step": 2051
+    },
+    {
+      "epoch": 0.04884318766066838,
+      "grad_norm": 7.930749461433814,
+      "learning_rate": 2e-05,
+      "loss": 0.4291,
+      "step": 2052
+    },
+    {
+      "epoch": 0.0488669903836999,
+      "grad_norm": 9.949516088933942,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 2053
+    },
+    {
+      "epoch": 0.04889079310673141,
+      "grad_norm": 7.828500339047789,
+      "learning_rate": 2e-05,
+      "loss": 0.4735,
+      "step": 2054
+    },
+    {
+      "epoch": 0.04891459582976292,
+      "grad_norm": 14.048296111496965,
+      "learning_rate": 2e-05,
+      "loss": 1.0598,
+      "step": 2055
+    },
+    {
+      "epoch": 0.048938398552794436,
+      "grad_norm": 5.001486520150071,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 2056
+    },
+    {
+      "epoch": 0.048962201275825956,
+      "grad_norm": 5.528829175887316,
+      "learning_rate": 2e-05,
+      "loss": 0.474,
+      "step": 2057
+    },
+    {
+      "epoch": 0.04898600399885747,
+      "grad_norm": 2.7423348533702985,
+      "learning_rate": 2e-05,
+      "loss": 0.2196,
+      "step": 2058
+    },
+    {
+      "epoch": 0.04900980672188898,
+      "grad_norm": 9.049539681505394,
+      "learning_rate": 2e-05,
+      "loss": 0.2928,
+      "step": 2059
+    },
+    {
+      "epoch": 0.0490336094449205,
+      "grad_norm": 5.280199562045336,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 2060
+    },
+    {
+      "epoch": 0.049057412167952015,
+      "grad_norm": 9.409992342272998,
+      "learning_rate": 2e-05,
+      "loss": 0.3126,
+      "step": 2061
+    },
+    {
+      "epoch": 0.04908121489098353,
+      "grad_norm": 9.398246284883712,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 2062
+    },
+    {
+      "epoch": 0.04910501761401504,
+      "grad_norm": 3.142427565173893,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 2063
+    },
+    {
+      "epoch": 0.04912882033704656,
+      "grad_norm": 8.52595745019591,
+      "learning_rate": 2e-05,
+      "loss": 0.4751,
+      "step": 2064
+    },
+    {
+      "epoch": 0.049152623060078074,
+      "grad_norm": 10.438029340150514,
+      "learning_rate": 2e-05,
+      "loss": 1.3733,
+      "step": 2065
+    },
+    {
+      "epoch": 0.04917642578310959,
+      "grad_norm": 8.937580824412096,
+      "learning_rate": 2e-05,
+      "loss": 1.1629,
+      "step": 2066
+    },
+    {
+      "epoch": 0.0492002285061411,
+      "grad_norm": 6.787018206638613,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 2067
+    },
+    {
+      "epoch": 0.04922403122917262,
+      "grad_norm": 6.405399307881757,
+      "learning_rate": 2e-05,
+      "loss": 0.2577,
+      "step": 2068
+    },
+    {
+      "epoch": 0.04924783395220413,
+      "grad_norm": 33.58641615409838,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 2069
+    },
+    {
+      "epoch": 0.049271636675235646,
+      "grad_norm": 4.817204302810613,
+      "learning_rate": 2e-05,
+      "loss": 0.2695,
+      "step": 2070
+    },
+    {
+      "epoch": 0.04929543939826716,
+      "grad_norm": 9.192689239853163,
+      "learning_rate": 2e-05,
+      "loss": 0.9661,
+      "step": 2071
+    },
+    {
+      "epoch": 0.04931924212129868,
+      "grad_norm": 1.9490449539485513,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 2072
+    },
+    {
+      "epoch": 0.04934304484433019,
+      "grad_norm": 4.889342748473342,
+      "learning_rate": 2e-05,
+      "loss": 0.3998,
+      "step": 2073
+    },
+    {
+      "epoch": 0.049366847567361705,
+      "grad_norm": 7.948106896932901,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 2074
+    },
+    {
+      "epoch": 0.04939065029039322,
+      "grad_norm": 7.206707435176661,
+      "learning_rate": 2e-05,
+      "loss": 1.0171,
+      "step": 2075
+    },
+    {
+      "epoch": 0.04941445301342474,
+      "grad_norm": 6.00856012573662,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 2076
+    },
+    {
+      "epoch": 0.04943825573645625,
+      "grad_norm": 5.281269392930984,
+      "learning_rate": 2e-05,
+      "loss": 0.6573,
+      "step": 2077
+    },
+    {
+      "epoch": 0.049462058459487764,
+      "grad_norm": 5.47804862078233,
+      "learning_rate": 2e-05,
+      "loss": 0.4121,
+      "step": 2078
+    },
+    {
+      "epoch": 0.04948586118251928,
+      "grad_norm": 6.090501114858611,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 2079
+    },
+    {
+      "epoch": 0.049509663905550796,
+      "grad_norm": 6.224916101533382,
+      "learning_rate": 2e-05,
+      "loss": 0.3723,
+      "step": 2080
+    },
+    {
+      "epoch": 0.04953346662858231,
+      "grad_norm": 6.787168443499312,
+      "learning_rate": 2e-05,
+      "loss": 0.349,
+      "step": 2081
+    },
+    {
+      "epoch": 0.04955726935161382,
+      "grad_norm": 4.116658121567522,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 2082
+    },
+    {
+      "epoch": 0.04958107207464534,
+      "grad_norm": 6.856459045634822,
+      "learning_rate": 2e-05,
+      "loss": 0.4042,
+      "step": 2083
+    },
+    {
+      "epoch": 0.049604874797676855,
+      "grad_norm": 7.003797922798057,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 2084
+    },
+    {
+      "epoch": 0.04962867752070837,
+      "grad_norm": 5.692831646784389,
+      "learning_rate": 2e-05,
+      "loss": 0.6806,
+      "step": 2085
+    },
+    {
+      "epoch": 0.04965248024373988,
+      "grad_norm": 9.115807630598301,
+      "learning_rate": 2e-05,
+      "loss": 1.3797,
+      "step": 2086
+    },
+    {
+      "epoch": 0.0496762829667714,
+      "grad_norm": 5.615856848583129,
+      "learning_rate": 2e-05,
+      "loss": 0.2745,
+      "step": 2087
+    },
+    {
+      "epoch": 0.049700085689802914,
+      "grad_norm": 10.248477572203504,
+      "learning_rate": 2e-05,
+      "loss": 1.4197,
+      "step": 2088
+    },
+    {
+      "epoch": 0.04972388841283443,
+      "grad_norm": 5.635243931394386,
+      "learning_rate": 2e-05,
+      "loss": 0.2011,
+      "step": 2089
+    },
+    {
+      "epoch": 0.04974769113586594,
+      "grad_norm": 6.609672714335492,
+      "learning_rate": 2e-05,
+      "loss": 1.1987,
+      "step": 2090
+    },
+    {
+      "epoch": 0.04977149385889746,
+      "grad_norm": 2.5377876771510093,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 2091
+    },
+    {
+      "epoch": 0.04979529658192897,
+      "grad_norm": 3.451882003682898,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 2092
+    },
+    {
+      "epoch": 0.049819099304960486,
+      "grad_norm": 5.274061699259568,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 2093
+    },
+    {
+      "epoch": 0.049842902027992,
+      "grad_norm": 8.240575557692768,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 2094
+    },
+    {
+      "epoch": 0.04986670475102352,
+      "grad_norm": 5.479864133993252,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 2095
+    },
+    {
+      "epoch": 0.04989050747405503,
+      "grad_norm": 5.843308780402969,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 2096
+    },
+    {
+      "epoch": 0.049914310197086545,
+      "grad_norm": 2.2352642839433106,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 2097
+    },
+    {
+      "epoch": 0.049938112920118065,
+      "grad_norm": 4.7740428938865085,
+      "learning_rate": 2e-05,
+      "loss": 0.6567,
+      "step": 2098
+    },
+    {
+      "epoch": 0.04996191564314958,
+      "grad_norm": 8.062756143114203,
+      "learning_rate": 2e-05,
+      "loss": 0.3494,
+      "step": 2099
+    },
+    {
+      "epoch": 0.04998571836618109,
+      "grad_norm": 5.06988547177264,
+      "learning_rate": 2e-05,
+      "loss": 0.7205,
+      "step": 2100
+    },
+    {
+      "epoch": 0.050009521089212604,
+      "grad_norm": 2.177962545019213,
+      "learning_rate": 2e-05,
+      "loss": 0.2279,
+      "step": 2101
+    },
+    {
+      "epoch": 0.050033323812244124,
+      "grad_norm": 6.542610565245354,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 2102
+    },
+    {
+      "epoch": 0.05005712653527564,
+      "grad_norm": 8.547268952320813,
+      "learning_rate": 2e-05,
+      "loss": 1.1173,
+      "step": 2103
+    },
+    {
+      "epoch": 0.05008092925830715,
+      "grad_norm": 4.412853909394814,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 2104
+    },
+    {
+      "epoch": 0.05010473198133866,
+      "grad_norm": 6.453556172354859,
+      "learning_rate": 2e-05,
+      "loss": 0.299,
+      "step": 2105
+    },
+    {
+      "epoch": 0.05012853470437018,
+      "grad_norm": 6.067338308824513,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 2106
+    },
+    {
+      "epoch": 0.050152337427401696,
+      "grad_norm": 5.372769813351141,
+      "learning_rate": 2e-05,
+      "loss": 0.3106,
+      "step": 2107
+    },
+    {
+      "epoch": 0.05017614015043321,
+      "grad_norm": 5.579561210963609,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 2108
+    },
+    {
+      "epoch": 0.05019994287346472,
+      "grad_norm": 8.489138635339854,
+      "learning_rate": 2e-05,
+      "loss": 0.4428,
+      "step": 2109
+    },
+    {
+      "epoch": 0.05022374559649624,
+      "grad_norm": 7.908983773230201,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 2110
+    },
+    {
+      "epoch": 0.050247548319527754,
+      "grad_norm": 6.417954777488671,
+      "learning_rate": 2e-05,
+      "loss": 0.8513,
+      "step": 2111
+    },
+    {
+      "epoch": 0.05027135104255927,
+      "grad_norm": 4.759891445247794,
+      "learning_rate": 2e-05,
+      "loss": 0.278,
+      "step": 2112
+    },
+    {
+      "epoch": 0.05029515376559078,
+      "grad_norm": 7.486699198233538,
+      "learning_rate": 2e-05,
+      "loss": 0.3677,
+      "step": 2113
+    },
+    {
+      "epoch": 0.0503189564886223,
+      "grad_norm": 5.545535012748427,
+      "learning_rate": 2e-05,
+      "loss": 0.908,
+      "step": 2114
+    },
+    {
+      "epoch": 0.05034275921165381,
+      "grad_norm": 4.842727227914949,
+      "learning_rate": 2e-05,
+      "loss": 0.3121,
+      "step": 2115
+    },
+    {
+      "epoch": 0.050366561934685326,
+      "grad_norm": 4.269160310919242,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 2116
+    },
+    {
+      "epoch": 0.050390364657716846,
+      "grad_norm": 6.004353567331738,
+      "learning_rate": 2e-05,
+      "loss": 0.2829,
+      "step": 2117
+    },
+    {
+      "epoch": 0.05041416738074836,
+      "grad_norm": 4.511427054282976,
+      "learning_rate": 2e-05,
+      "loss": 0.2285,
+      "step": 2118
+    },
+    {
+      "epoch": 0.05043797010377987,
+      "grad_norm": 4.682439045419907,
+      "learning_rate": 2e-05,
+      "loss": 0.4255,
+      "step": 2119
+    },
+    {
+      "epoch": 0.050461772826811385,
+      "grad_norm": 8.931783023311343,
+      "learning_rate": 2e-05,
+      "loss": 1.2274,
+      "step": 2120
+    },
+    {
+      "epoch": 0.050485575549842905,
+      "grad_norm": 4.041598796512998,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 2121
+    },
+    {
+      "epoch": 0.05050937827287442,
+      "grad_norm": 2.8698712697687445,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 2122
+    },
+    {
+      "epoch": 0.05053318099590593,
+      "grad_norm": 10.223133318404486,
+      "learning_rate": 2e-05,
+      "loss": 0.9714,
+      "step": 2123
+    },
+    {
+      "epoch": 0.050556983718937444,
+      "grad_norm": 3.0439299106761664,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 2124
+    },
+    {
+      "epoch": 0.050580786441968964,
+      "grad_norm": 7.263033407875099,
+      "learning_rate": 2e-05,
+      "loss": 0.9119,
+      "step": 2125
+    },
+    {
+      "epoch": 0.05060458916500048,
+      "grad_norm": 4.217706275630168,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 2126
+    },
+    {
+      "epoch": 0.05062839188803199,
+      "grad_norm": 1.7976774850068278,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 2127
+    },
+    {
+      "epoch": 0.0506521946110635,
+      "grad_norm": 2.1766620764195994,
+      "learning_rate": 2e-05,
+      "loss": 0.1872,
+      "step": 2128
+    },
+    {
+      "epoch": 0.05067599733409502,
+      "grad_norm": 8.963525309747025,
+      "learning_rate": 2e-05,
+      "loss": 1.2758,
+      "step": 2129
+    },
+    {
+      "epoch": 0.050699800057126536,
+      "grad_norm": 2.624954315437059,
+      "learning_rate": 2e-05,
+      "loss": 0.2975,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05072360278015805,
+      "grad_norm": 5.8545852416638064,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 2131
+    },
+    {
+      "epoch": 0.05074740550318956,
+      "grad_norm": 2.9279594597938696,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 2132
+    },
+    {
+      "epoch": 0.05077120822622108,
+      "grad_norm": 6.6058855047630045,
+      "learning_rate": 2e-05,
+      "loss": 0.4702,
+      "step": 2133
+    },
+    {
+      "epoch": 0.050795010949252595,
+      "grad_norm": 3.7211541891911626,
+      "learning_rate": 2e-05,
+      "loss": 0.2057,
+      "step": 2134
+    },
+    {
+      "epoch": 0.05081881367228411,
+      "grad_norm": 5.122031600056183,
+      "learning_rate": 2e-05,
+      "loss": 0.2911,
+      "step": 2135
+    },
+    {
+      "epoch": 0.05084261639531562,
+      "grad_norm": 15.071198863319545,
+      "learning_rate": 2e-05,
+      "loss": 0.9139,
+      "step": 2136
+    },
+    {
+      "epoch": 0.05086641911834714,
+      "grad_norm": 11.271986419789357,
+      "learning_rate": 2e-05,
+      "loss": 1.1401,
+      "step": 2137
+    },
+    {
+      "epoch": 0.050890221841378654,
+      "grad_norm": 4.088611367435325,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 2138
+    },
+    {
+      "epoch": 0.05091402456441017,
+      "grad_norm": 6.581945739057791,
+      "learning_rate": 2e-05,
+      "loss": 0.8907,
+      "step": 2139
+    },
+    {
+      "epoch": 0.050937827287441687,
+      "grad_norm": 3.3467331169997903,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0509616300104732,
+      "grad_norm": 5.051338937990023,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 2141
+    },
+    {
+      "epoch": 0.05098543273350471,
+      "grad_norm": 1.699928038519774,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 2142
+    },
+    {
+      "epoch": 0.051009235456536225,
+      "grad_norm": 2.5299874899565733,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 2143
+    },
+    {
+      "epoch": 0.051033038179567745,
+      "grad_norm": 2.5355991693810043,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 2144
+    },
+    {
+      "epoch": 0.05105684090259926,
+      "grad_norm": 13.635797248764034,
+      "learning_rate": 2e-05,
+      "loss": 1.0364,
+      "step": 2145
+    },
+    {
+      "epoch": 0.05108064362563077,
+      "grad_norm": 4.035867681612873,
+      "learning_rate": 2e-05,
+      "loss": 0.4159,
+      "step": 2146
+    },
+    {
+      "epoch": 0.051104446348662284,
+      "grad_norm": 9.093179421556414,
+      "learning_rate": 2e-05,
+      "loss": 0.9288,
+      "step": 2147
+    },
+    {
+      "epoch": 0.051128249071693804,
+      "grad_norm": 2.9438462492538386,
+      "learning_rate": 2e-05,
+      "loss": 0.2126,
+      "step": 2148
+    },
+    {
+      "epoch": 0.05115205179472532,
+      "grad_norm": 9.074272466358227,
+      "learning_rate": 2e-05,
+      "loss": 0.3921,
+      "step": 2149
+    },
+    {
+      "epoch": 0.05117585451775683,
+      "grad_norm": 4.720149474435618,
+      "learning_rate": 2e-05,
+      "loss": 0.7162,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05119965724078834,
+      "grad_norm": 4.115394610161105,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 2151
+    },
+    {
+      "epoch": 0.05122345996381986,
+      "grad_norm": 2.5396164426252623,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 2152
+    },
+    {
+      "epoch": 0.051247262686851376,
+      "grad_norm": 4.074151844252262,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 2153
+    },
+    {
+      "epoch": 0.05127106540988289,
+      "grad_norm": 6.426483238446292,
+      "learning_rate": 2e-05,
+      "loss": 0.9217,
+      "step": 2154
+    },
+    {
+      "epoch": 0.0512948681329144,
+      "grad_norm": 9.891437720875942,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 2155
+    },
+    {
+      "epoch": 0.05131867085594592,
+      "grad_norm": 6.567256196917575,
+      "learning_rate": 2e-05,
+      "loss": 0.867,
+      "step": 2156
+    },
+    {
+      "epoch": 0.051342473578977435,
+      "grad_norm": 3.9724717938890777,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 2157
+    },
+    {
+      "epoch": 0.05136627630200895,
+      "grad_norm": 42.10018996032681,
+      "learning_rate": 2e-05,
+      "loss": 0.942,
+      "step": 2158
+    },
+    {
+      "epoch": 0.05139007902504047,
+      "grad_norm": 3.1445462045491417,
+      "learning_rate": 2e-05,
+      "loss": 0.2205,
+      "step": 2159
+    },
+    {
+      "epoch": 0.05141388174807198,
+      "grad_norm": 7.388263352262605,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 2160
+    },
+    {
+      "epoch": 0.051437684471103494,
+      "grad_norm": 5.270949394683509,
+      "learning_rate": 2e-05,
+      "loss": 0.509,
+      "step": 2161
+    },
+    {
+      "epoch": 0.05146148719413501,
+      "grad_norm": 3.777756612123424,
+      "learning_rate": 2e-05,
+      "loss": 0.3747,
+      "step": 2162
+    },
+    {
+      "epoch": 0.05148528991716653,
+      "grad_norm": 10.47471786900156,
+      "learning_rate": 2e-05,
+      "loss": 0.948,
+      "step": 2163
+    },
+    {
+      "epoch": 0.05150909264019804,
+      "grad_norm": 3.48356874109161,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 2164
+    },
+    {
+      "epoch": 0.05153289536322955,
+      "grad_norm": 7.642662691535282,
+      "learning_rate": 2e-05,
+      "loss": 0.7402,
+      "step": 2165
+    },
+    {
+      "epoch": 0.051556698086261066,
+      "grad_norm": 14.23194377718979,
+      "learning_rate": 2e-05,
+      "loss": 0.7176,
+      "step": 2166
+    },
+    {
+      "epoch": 0.051580500809292586,
+      "grad_norm": 3.2638512244240565,
+      "learning_rate": 2e-05,
+      "loss": 0.1507,
+      "step": 2167
+    },
+    {
+      "epoch": 0.0516043035323241,
+      "grad_norm": 4.241302368339444,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 2168
+    },
+    {
+      "epoch": 0.05162810625535561,
+      "grad_norm": 8.206243998492404,
+      "learning_rate": 2e-05,
+      "loss": 1.1083,
+      "step": 2169
+    },
+    {
+      "epoch": 0.051651908978387125,
+      "grad_norm": 10.430810973243897,
+      "learning_rate": 2e-05,
+      "loss": 0.5295,
+      "step": 2170
+    },
+    {
+      "epoch": 0.051675711701418645,
+      "grad_norm": 1.3259369260602059,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 2171
+    },
+    {
+      "epoch": 0.05169951442445016,
+      "grad_norm": 2.6901997835979725,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 2172
+    },
+    {
+      "epoch": 0.05172331714748167,
+      "grad_norm": 125.71050669248262,
+      "learning_rate": 2e-05,
+      "loss": 1.0509,
+      "step": 2173
+    },
+    {
+      "epoch": 0.051747119870513184,
+      "grad_norm": 2.4808792664015993,
+      "learning_rate": 2e-05,
+      "loss": 0.1331,
+      "step": 2174
+    },
+    {
+      "epoch": 0.0517709225935447,
+      "grad_norm": 1.5285306849652256,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 2175
+    },
+    {
+      "epoch": 0.051794725316576216,
+      "grad_norm": 15.48779350818404,
+      "learning_rate": 2e-05,
+      "loss": 0.9377,
+      "step": 2176
+    },
+    {
+      "epoch": 0.05181852803960773,
+      "grad_norm": 10.478340533415443,
+      "learning_rate": 2e-05,
+      "loss": 1.4136,
+      "step": 2177
+    },
+    {
+      "epoch": 0.05184233076263925,
+      "grad_norm": 5.804206888757116,
+      "learning_rate": 2e-05,
+      "loss": 0.3364,
+      "step": 2178
+    },
+    {
+      "epoch": 0.05186613348567076,
+      "grad_norm": 5.492488152431103,
+      "learning_rate": 2e-05,
+      "loss": 0.2748,
+      "step": 2179
+    },
+    {
+      "epoch": 0.051889936208702275,
+      "grad_norm": 2.2188138160998685,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 2180
+    },
+    {
+      "epoch": 0.05191373893173379,
+      "grad_norm": 7.511462135155529,
+      "learning_rate": 2e-05,
+      "loss": 1.1479,
+      "step": 2181
+    },
+    {
+      "epoch": 0.05193754165476531,
+      "grad_norm": 6.96608726253562,
+      "learning_rate": 2e-05,
+      "loss": 0.7541,
+      "step": 2182
+    },
+    {
+      "epoch": 0.05196134437779682,
+      "grad_norm": 4.314588906960862,
+      "learning_rate": 2e-05,
+      "loss": 0.3595,
+      "step": 2183
+    },
+    {
+      "epoch": 0.051985147100828334,
+      "grad_norm": 3.708652016193496,
+      "learning_rate": 2e-05,
+      "loss": 0.2518,
+      "step": 2184
+    },
+    {
+      "epoch": 0.05200894982385985,
+      "grad_norm": 3.58450019319476,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 2185
+    },
+    {
+      "epoch": 0.05203275254689137,
+      "grad_norm": 14.420485586800309,
+      "learning_rate": 2e-05,
+      "loss": 1.2684,
+      "step": 2186
+    },
+    {
+      "epoch": 0.05205655526992288,
+      "grad_norm": 8.72654884805584,
+      "learning_rate": 2e-05,
+      "loss": 0.5711,
+      "step": 2187
+    },
+    {
+      "epoch": 0.05208035799295439,
+      "grad_norm": 3.9186544923379447,
+      "learning_rate": 2e-05,
+      "loss": 0.2924,
+      "step": 2188
+    },
+    {
+      "epoch": 0.052104160715985906,
+      "grad_norm": 8.777094706569677,
+      "learning_rate": 2e-05,
+      "loss": 1.6771,
+      "step": 2189
+    },
+    {
+      "epoch": 0.052127963439017426,
+      "grad_norm": 1.8422441686192041,
+      "learning_rate": 2e-05,
+      "loss": 0.1953,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05215176616204894,
+      "grad_norm": 7.021331689134169,
+      "learning_rate": 2e-05,
+      "loss": 1.3033,
+      "step": 2191
+    },
+    {
+      "epoch": 0.05217556888508045,
+      "grad_norm": 7.170329139938559,
+      "learning_rate": 2e-05,
+      "loss": 0.3088,
+      "step": 2192
+    },
+    {
+      "epoch": 0.052199371608111965,
+      "grad_norm": 6.567438958258353,
+      "learning_rate": 2e-05,
+      "loss": 0.2363,
+      "step": 2193
+    },
+    {
+      "epoch": 0.052223174331143485,
+      "grad_norm": 14.898721470692204,
+      "learning_rate": 2e-05,
+      "loss": 1.4308,
+      "step": 2194
+    },
+    {
+      "epoch": 0.052246977054175,
+      "grad_norm": 7.010073977248735,
+      "learning_rate": 2e-05,
+      "loss": 1.0995,
+      "step": 2195
+    },
+    {
+      "epoch": 0.05227077977720651,
+      "grad_norm": 8.216666368812627,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 2196
+    },
+    {
+      "epoch": 0.052294582500238024,
+      "grad_norm": 10.954575905113966,
+      "learning_rate": 2e-05,
+      "loss": 0.8161,
+      "step": 2197
+    },
+    {
+      "epoch": 0.052318385223269544,
+      "grad_norm": 1.9965660187264291,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 2198
+    },
+    {
+      "epoch": 0.05234218794630106,
+      "grad_norm": 4.109388535596234,
+      "learning_rate": 2e-05,
+      "loss": 0.3245,
+      "step": 2199
+    },
+    {
+      "epoch": 0.05236599066933257,
+      "grad_norm": 8.648274009653724,
+      "learning_rate": 2e-05,
+      "loss": 0.3188,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05238979339236409,
+      "grad_norm": 5.225530461631878,
+      "learning_rate": 2e-05,
+      "loss": 0.2576,
+      "step": 2201
+    },
+    {
+      "epoch": 0.0524135961153956,
+      "grad_norm": 11.24491304262208,
+      "learning_rate": 2e-05,
+      "loss": 0.4741,
+      "step": 2202
+    },
+    {
+      "epoch": 0.052437398838427116,
+      "grad_norm": 2.7486626289544356,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 2203
+    },
+    {
+      "epoch": 0.05246120156145863,
+      "grad_norm": 5.834555208916619,
+      "learning_rate": 2e-05,
+      "loss": 0.3674,
+      "step": 2204
+    },
+    {
+      "epoch": 0.05248500428449015,
+      "grad_norm": 5.2929787146533345,
+      "learning_rate": 2e-05,
+      "loss": 0.6951,
+      "step": 2205
+    },
+    {
+      "epoch": 0.05250880700752166,
+      "grad_norm": 8.528712998005984,
+      "learning_rate": 2e-05,
+      "loss": 1.09,
+      "step": 2206
+    },
+    {
+      "epoch": 0.052532609730553174,
+      "grad_norm": 8.876281999508999,
+      "learning_rate": 2e-05,
+      "loss": 1.0319,
+      "step": 2207
+    },
+    {
+      "epoch": 0.05255641245358469,
+      "grad_norm": 3.2862374422691305,
+      "learning_rate": 2e-05,
+      "loss": 0.2861,
+      "step": 2208
+    },
+    {
+      "epoch": 0.05258021517661621,
+      "grad_norm": 8.712826235761339,
+      "learning_rate": 2e-05,
+      "loss": 0.2885,
+      "step": 2209
+    },
+    {
+      "epoch": 0.05260401789964772,
+      "grad_norm": 2.9787973578278515,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05262782062267923,
+      "grad_norm": 2.0152209747582313,
+      "learning_rate": 2e-05,
+      "loss": 0.1893,
+      "step": 2211
+    },
+    {
+      "epoch": 0.052651623345710746,
+      "grad_norm": 7.544790776765838,
+      "learning_rate": 2e-05,
+      "loss": 0.2887,
+      "step": 2212
+    },
+    {
+      "epoch": 0.052675426068742266,
+      "grad_norm": 7.397693103483878,
+      "learning_rate": 2e-05,
+      "loss": 1.0216,
+      "step": 2213
+    },
+    {
+      "epoch": 0.05269922879177378,
+      "grad_norm": 8.414262425553838,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 2214
+    },
+    {
+      "epoch": 0.05272303151480529,
+      "grad_norm": 3.096711884556172,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 2215
+    },
+    {
+      "epoch": 0.052746834237836805,
+      "grad_norm": 9.826852724596444,
+      "learning_rate": 2e-05,
+      "loss": 0.6095,
+      "step": 2216
+    },
+    {
+      "epoch": 0.052770636960868325,
+      "grad_norm": 2.185260888172663,
+      "learning_rate": 2e-05,
+      "loss": 0.2468,
+      "step": 2217
+    },
+    {
+      "epoch": 0.05279443968389984,
+      "grad_norm": 2.8129605670413005,
+      "learning_rate": 2e-05,
+      "loss": 0.2617,
+      "step": 2218
+    },
+    {
+      "epoch": 0.05281824240693135,
+      "grad_norm": 2.890243828932123,
+      "learning_rate": 2e-05,
+      "loss": 0.1579,
+      "step": 2219
+    },
+    {
+      "epoch": 0.05284204512996287,
+      "grad_norm": 4.020635829554028,
+      "learning_rate": 2e-05,
+      "loss": 0.2891,
+      "step": 2220
+    },
+    {
+      "epoch": 0.052865847852994384,
+      "grad_norm": 4.8758566778456895,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 2221
+    },
+    {
+      "epoch": 0.0528896505760259,
+      "grad_norm": 3.7244529912075532,
+      "learning_rate": 2e-05,
+      "loss": 0.2782,
+      "step": 2222
+    },
+    {
+      "epoch": 0.05291345329905741,
+      "grad_norm": 2.7325144736018725,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 2223
+    },
+    {
+      "epoch": 0.05293725602208893,
+      "grad_norm": 3.531229314810762,
+      "learning_rate": 2e-05,
+      "loss": 0.2441,
+      "step": 2224
+    },
+    {
+      "epoch": 0.05296105874512044,
+      "grad_norm": 6.089706270619202,
+      "learning_rate": 2e-05,
+      "loss": 0.9368,
+      "step": 2225
+    },
+    {
+      "epoch": 0.052984861468151956,
+      "grad_norm": 8.339574510947896,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 2226
+    },
+    {
+      "epoch": 0.05300866419118347,
+      "grad_norm": 10.942545496321555,
+      "learning_rate": 2e-05,
+      "loss": 1.7162,
+      "step": 2227
+    },
+    {
+      "epoch": 0.05303246691421499,
+      "grad_norm": 5.057188525567394,
+      "learning_rate": 2e-05,
+      "loss": 0.6562,
+      "step": 2228
+    },
+    {
+      "epoch": 0.0530562696372465,
+      "grad_norm": 6.015621377693058,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 2229
+    },
+    {
+      "epoch": 0.053080072360278015,
+      "grad_norm": 8.350277537647308,
+      "learning_rate": 2e-05,
+      "loss": 0.3297,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05310387508330953,
+      "grad_norm": 2.1070082572753086,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 2231
+    },
+    {
+      "epoch": 0.05312767780634105,
+      "grad_norm": 10.051612226582467,
+      "learning_rate": 2e-05,
+      "loss": 0.5151,
+      "step": 2232
+    },
+    {
+      "epoch": 0.05315148052937256,
+      "grad_norm": 2.180499518265679,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 2233
+    },
+    {
+      "epoch": 0.053175283252404074,
+      "grad_norm": 5.465825640911398,
+      "learning_rate": 2e-05,
+      "loss": 0.4085,
+      "step": 2234
+    },
+    {
+      "epoch": 0.05319908597543559,
+      "grad_norm": 7.850571658606856,
+      "learning_rate": 2e-05,
+      "loss": 1.166,
+      "step": 2235
+    },
+    {
+      "epoch": 0.053222888698467107,
+      "grad_norm": 2.861725213443775,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 2236
+    },
+    {
+      "epoch": 0.05324669142149862,
+      "grad_norm": 3.5183116422089067,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 2237
+    },
+    {
+      "epoch": 0.05327049414453013,
+      "grad_norm": 8.48730227823551,
+      "learning_rate": 2e-05,
+      "loss": 0.6544,
+      "step": 2238
+    },
+    {
+      "epoch": 0.05329429686756165,
+      "grad_norm": 7.368066996930929,
+      "learning_rate": 2e-05,
+      "loss": 0.3319,
+      "step": 2239
+    },
+    {
+      "epoch": 0.053318099590593165,
+      "grad_norm": 4.952740216060738,
+      "learning_rate": 2e-05,
+      "loss": 0.175,
+      "step": 2240
+    },
+    {
+      "epoch": 0.05334190231362468,
+      "grad_norm": 2.6058112547144887,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 2241
+    },
+    {
+      "epoch": 0.05336570503665619,
+      "grad_norm": 9.572956982501653,
+      "learning_rate": 2e-05,
+      "loss": 2.1483,
+      "step": 2242
+    },
+    {
+      "epoch": 0.05338950775968771,
+      "grad_norm": 3.2339967068287154,
+      "learning_rate": 2e-05,
+      "loss": 0.2586,
+      "step": 2243
+    },
+    {
+      "epoch": 0.053413310482719224,
+      "grad_norm": 10.466644201638907,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 2244
+    },
+    {
+      "epoch": 0.05343711320575074,
+      "grad_norm": 5.224305216598711,
+      "learning_rate": 2e-05,
+      "loss": 0.7845,
+      "step": 2245
+    },
+    {
+      "epoch": 0.05346091592878225,
+      "grad_norm": 13.187228621606703,
+      "learning_rate": 2e-05,
+      "loss": 1.0518,
+      "step": 2246
+    },
+    {
+      "epoch": 0.05348471865181377,
+      "grad_norm": 8.54594317759605,
+      "learning_rate": 2e-05,
+      "loss": 1.0405,
+      "step": 2247
+    },
+    {
+      "epoch": 0.05350852137484528,
+      "grad_norm": 9.680178747265456,
+      "learning_rate": 2e-05,
+      "loss": 0.4437,
+      "step": 2248
+    },
+    {
+      "epoch": 0.053532324097876796,
+      "grad_norm": 3.717895236471582,
+      "learning_rate": 2e-05,
+      "loss": 0.2768,
+      "step": 2249
+    },
+    {
+      "epoch": 0.05355612682090831,
+      "grad_norm": 3.572383519184122,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05357992954393983,
+      "grad_norm": 6.932050982690749,
+      "learning_rate": 2e-05,
+      "loss": 0.8174,
+      "step": 2251
+    },
+    {
+      "epoch": 0.05360373226697134,
+      "grad_norm": 6.081532354235316,
+      "learning_rate": 2e-05,
+      "loss": 0.7837,
+      "step": 2252
+    },
+    {
+      "epoch": 0.053627534990002855,
+      "grad_norm": 2.368916606038742,
+      "learning_rate": 2e-05,
+      "loss": 0.1681,
+      "step": 2253
+    },
+    {
+      "epoch": 0.05365133771303437,
+      "grad_norm": 8.272044362283134,
+      "learning_rate": 2e-05,
+      "loss": 1.0037,
+      "step": 2254
+    },
+    {
+      "epoch": 0.05367514043606589,
+      "grad_norm": 2.3413883428833806,
+      "learning_rate": 2e-05,
+      "loss": 0.2017,
+      "step": 2255
+    },
+    {
+      "epoch": 0.0536989431590974,
+      "grad_norm": 10.008342281621701,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 2256
+    },
+    {
+      "epoch": 0.053722745882128914,
+      "grad_norm": 8.886890481810575,
+      "learning_rate": 2e-05,
+      "loss": 0.4254,
+      "step": 2257
+    },
+    {
+      "epoch": 0.053746548605160434,
+      "grad_norm": 10.764985146615793,
+      "learning_rate": 2e-05,
+      "loss": 1.0226,
+      "step": 2258
+    },
+    {
+      "epoch": 0.05377035132819195,
+      "grad_norm": 6.293599142875985,
+      "learning_rate": 2e-05,
+      "loss": 0.823,
+      "step": 2259
+    },
+    {
+      "epoch": 0.05379415405122346,
+      "grad_norm": 3.3473238654498023,
+      "learning_rate": 2e-05,
+      "loss": 0.2462,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05381795677425497,
+      "grad_norm": 6.899895086058964,
+      "learning_rate": 2e-05,
+      "loss": 1.102,
+      "step": 2261
+    },
+    {
+      "epoch": 0.05384175949728649,
+      "grad_norm": 4.2735124038082235,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 2262
+    },
+    {
+      "epoch": 0.053865562220318006,
+      "grad_norm": 6.077770661658693,
+      "learning_rate": 2e-05,
+      "loss": 0.2538,
+      "step": 2263
+    },
+    {
+      "epoch": 0.05388936494334952,
+      "grad_norm": 8.42435653922165,
+      "learning_rate": 2e-05,
+      "loss": 0.2453,
+      "step": 2264
+    },
+    {
+      "epoch": 0.05391316766638103,
+      "grad_norm": 7.477195743825543,
+      "learning_rate": 2e-05,
+      "loss": 0.3233,
+      "step": 2265
+    },
+    {
+      "epoch": 0.05393697038941255,
+      "grad_norm": 7.909072314435058,
+      "learning_rate": 2e-05,
+      "loss": 0.3606,
+      "step": 2266
+    },
+    {
+      "epoch": 0.053960773112444065,
+      "grad_norm": 8.074619139890935,
+      "learning_rate": 2e-05,
+      "loss": 1.4056,
+      "step": 2267
+    },
+    {
+      "epoch": 0.05398457583547558,
+      "grad_norm": 2.61835767120097,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 2268
+    },
+    {
+      "epoch": 0.05400837855850709,
+      "grad_norm": 3.7670789883650833,
+      "learning_rate": 2e-05,
+      "loss": 0.2004,
+      "step": 2269
+    },
+    {
+      "epoch": 0.05403218128153861,
+      "grad_norm": 13.231461559017395,
+      "learning_rate": 2e-05,
+      "loss": 1.7605,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05405598400457012,
+      "grad_norm": 3.645745124301651,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 2271
+    },
+    {
+      "epoch": 0.054079786727601636,
+      "grad_norm": 6.075505779821767,
+      "learning_rate": 2e-05,
+      "loss": 0.5921,
+      "step": 2272
+    },
+    {
+      "epoch": 0.05410358945063315,
+      "grad_norm": 9.685629432044147,
+      "learning_rate": 2e-05,
+      "loss": 0.3995,
+      "step": 2273
+    },
+    {
+      "epoch": 0.05412739217366467,
+      "grad_norm": 5.918483724536217,
+      "learning_rate": 2e-05,
+      "loss": 0.7761,
+      "step": 2274
+    },
+    {
+      "epoch": 0.05415119489669618,
+      "grad_norm": 10.61245137339365,
+      "learning_rate": 2e-05,
+      "loss": 0.5278,
+      "step": 2275
+    },
+    {
+      "epoch": 0.054174997619727695,
+      "grad_norm": 2.891658458477362,
+      "learning_rate": 2e-05,
+      "loss": 0.2882,
+      "step": 2276
+    },
+    {
+      "epoch": 0.05419880034275921,
+      "grad_norm": 5.845896973885449,
+      "learning_rate": 2e-05,
+      "loss": 0.3859,
+      "step": 2277
+    },
+    {
+      "epoch": 0.05422260306579073,
+      "grad_norm": 2.1942436098554214,
+      "learning_rate": 2e-05,
+      "loss": 0.1845,
+      "step": 2278
+    },
+    {
+      "epoch": 0.05424640578882224,
+      "grad_norm": 5.765998567572599,
+      "learning_rate": 2e-05,
+      "loss": 0.8725,
+      "step": 2279
+    },
+    {
+      "epoch": 0.054270208511853754,
+      "grad_norm": 8.278238925634344,
+      "learning_rate": 2e-05,
+      "loss": 0.9947,
+      "step": 2280
+    },
+    {
+      "epoch": 0.054294011234885274,
+      "grad_norm": 4.261037334327607,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 2281
+    },
+    {
+      "epoch": 0.05431781395791679,
+      "grad_norm": 6.869932981581842,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 2282
+    },
+    {
+      "epoch": 0.0543416166809483,
+      "grad_norm": 4.017222583093106,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 2283
+    },
+    {
+      "epoch": 0.05436541940397981,
+      "grad_norm": 6.861764328739512,
+      "learning_rate": 2e-05,
+      "loss": 0.2284,
+      "step": 2284
+    },
+    {
+      "epoch": 0.05438922212701133,
+      "grad_norm": 3.632012909975843,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 2285
+    },
+    {
+      "epoch": 0.054413024850042846,
+      "grad_norm": 13.793968215534047,
+      "learning_rate": 2e-05,
+      "loss": 0.4814,
+      "step": 2286
+    },
+    {
+      "epoch": 0.05443682757307436,
+      "grad_norm": 8.73653652810315,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 2287
+    },
+    {
+      "epoch": 0.05446063029610587,
+      "grad_norm": 9.619231287130141,
+      "learning_rate": 2e-05,
+      "loss": 0.433,
+      "step": 2288
+    },
+    {
+      "epoch": 0.05448443301913739,
+      "grad_norm": 5.628736091565664,
+      "learning_rate": 2e-05,
+      "loss": 0.8181,
+      "step": 2289
+    },
+    {
+      "epoch": 0.054508235742168905,
+      "grad_norm": 1.0786144075478035,
+      "learning_rate": 2e-05,
+      "loss": 0.1993,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05453203846520042,
+      "grad_norm": 2.779585928985598,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 2291
+    },
+    {
+      "epoch": 0.05455584118823193,
+      "grad_norm": 7.021913321918926,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 2292
+    },
+    {
+      "epoch": 0.05457964391126345,
+      "grad_norm": 2.7724688512364466,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 2293
+    },
+    {
+      "epoch": 0.054603446634294964,
+      "grad_norm": 4.232101859273723,
+      "learning_rate": 2e-05,
+      "loss": 0.3535,
+      "step": 2294
+    },
+    {
+      "epoch": 0.05462724935732648,
+      "grad_norm": 6.467730150802679,
+      "learning_rate": 2e-05,
+      "loss": 0.7865,
+      "step": 2295
+    },
+    {
+      "epoch": 0.05465105208035799,
+      "grad_norm": 8.491805106835695,
+      "learning_rate": 2e-05,
+      "loss": 1.2749,
+      "step": 2296
+    },
+    {
+      "epoch": 0.05467485480338951,
+      "grad_norm": 2.3636010808276056,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 2297
+    },
+    {
+      "epoch": 0.05469865752642102,
+      "grad_norm": 6.159274320101011,
+      "learning_rate": 2e-05,
+      "loss": 0.4728,
+      "step": 2298
+    },
+    {
+      "epoch": 0.054722460249452536,
+      "grad_norm": 7.956979174037071,
+      "learning_rate": 2e-05,
+      "loss": 0.3875,
+      "step": 2299
+    },
+    {
+      "epoch": 0.054746262972484055,
+      "grad_norm": 3.594333977557177,
+      "learning_rate": 2e-05,
+      "loss": 0.2315,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05477006569551557,
+      "grad_norm": 6.452670187265527,
+      "learning_rate": 2e-05,
+      "loss": 1.2452,
+      "step": 2301
+    },
+    {
+      "epoch": 0.05479386841854708,
+      "grad_norm": 5.170189537905692,
+      "learning_rate": 2e-05,
+      "loss": 0.4862,
+      "step": 2302
+    },
+    {
+      "epoch": 0.054817671141578594,
+      "grad_norm": 6.142018925866877,
+      "learning_rate": 2e-05,
+      "loss": 0.3632,
+      "step": 2303
+    },
+    {
+      "epoch": 0.054841473864610114,
+      "grad_norm": 10.109961722496896,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 2304
+    },
+    {
+      "epoch": 0.05486527658764163,
+      "grad_norm": 3.236862758138841,
+      "learning_rate": 2e-05,
+      "loss": 0.2272,
+      "step": 2305
+    },
+    {
+      "epoch": 0.05488907931067314,
+      "grad_norm": 6.739834510504648,
+      "learning_rate": 2e-05,
+      "loss": 0.3197,
+      "step": 2306
+    },
+    {
+      "epoch": 0.05491288203370465,
+      "grad_norm": 3.9176129586059707,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 2307
+    },
+    {
+      "epoch": 0.05493668475673617,
+      "grad_norm": 3.3932646531506645,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 2308
+    },
+    {
+      "epoch": 0.054960487479767686,
+      "grad_norm": 2.0827503886863696,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 2309
+    },
+    {
+      "epoch": 0.0549842902027992,
+      "grad_norm": 4.090510999630105,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05500809292583071,
+      "grad_norm": 3.2053301908009177,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 2311
+    },
+    {
+      "epoch": 0.05503189564886223,
+      "grad_norm": 6.350701446492859,
+      "learning_rate": 2e-05,
+      "loss": 0.3721,
+      "step": 2312
+    },
+    {
+      "epoch": 0.055055698371893745,
+      "grad_norm": 10.815421226734369,
+      "learning_rate": 2e-05,
+      "loss": 1.1795,
+      "step": 2313
+    },
+    {
+      "epoch": 0.05507950109492526,
+      "grad_norm": 8.356427023490589,
+      "learning_rate": 2e-05,
+      "loss": 0.3236,
+      "step": 2314
+    },
+    {
+      "epoch": 0.05510330381795677,
+      "grad_norm": 6.211273625937206,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 2315
+    },
+    {
+      "epoch": 0.05512710654098829,
+      "grad_norm": 3.475693864951567,
+      "learning_rate": 2e-05,
+      "loss": 0.1768,
+      "step": 2316
+    },
+    {
+      "epoch": 0.055150909264019804,
+      "grad_norm": 3.116560854779123,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 2317
+    },
+    {
+      "epoch": 0.05517471198705132,
+      "grad_norm": 2.813072208119713,
+      "learning_rate": 2e-05,
+      "loss": 0.2793,
+      "step": 2318
+    },
+    {
+      "epoch": 0.05519851471008284,
+      "grad_norm": 8.82363753279015,
+      "learning_rate": 2e-05,
+      "loss": 1.507,
+      "step": 2319
+    },
+    {
+      "epoch": 0.05522231743311435,
+      "grad_norm": 6.870921641868455,
+      "learning_rate": 2e-05,
+      "loss": 0.469,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05524612015614586,
+      "grad_norm": 2.099852393222603,
+      "learning_rate": 2e-05,
+      "loss": 0.1913,
+      "step": 2321
+    },
+    {
+      "epoch": 0.055269922879177376,
+      "grad_norm": 3.3762067553966557,
+      "learning_rate": 2e-05,
+      "loss": 0.165,
+      "step": 2322
+    },
+    {
+      "epoch": 0.055293725602208896,
+      "grad_norm": 8.355966909640513,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 2323
+    },
+    {
+      "epoch": 0.05531752832524041,
+      "grad_norm": 10.165556420003963,
+      "learning_rate": 2e-05,
+      "loss": 1.0346,
+      "step": 2324
+    },
+    {
+      "epoch": 0.05534133104827192,
+      "grad_norm": 3.5347930951820157,
+      "learning_rate": 2e-05,
+      "loss": 0.1759,
+      "step": 2325
+    },
+    {
+      "epoch": 0.055365133771303435,
+      "grad_norm": 7.180472164319195,
+      "learning_rate": 2e-05,
+      "loss": 0.3105,
+      "step": 2326
+    },
+    {
+      "epoch": 0.055388936494334955,
+      "grad_norm": 6.594551618642483,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 2327
+    },
+    {
+      "epoch": 0.05541273921736647,
+      "grad_norm": 3.601908674611816,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 2328
+    },
+    {
+      "epoch": 0.05543654194039798,
+      "grad_norm": 4.230756250250823,
+      "learning_rate": 2e-05,
+      "loss": 0.3183,
+      "step": 2329
+    },
+    {
+      "epoch": 0.055460344663429494,
+      "grad_norm": 5.424766367623963,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 2330
+    },
+    {
+      "epoch": 0.055484147386461014,
+      "grad_norm": 7.289317301746611,
+      "learning_rate": 2e-05,
+      "loss": 0.9415,
+      "step": 2331
+    },
+    {
+      "epoch": 0.055507950109492527,
+      "grad_norm": 7.950837834925938,
+      "learning_rate": 2e-05,
+      "loss": 0.2764,
+      "step": 2332
+    },
+    {
+      "epoch": 0.05553175283252404,
+      "grad_norm": 7.255392300929312,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 2333
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 2.9143093984951,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 2334
+    },
+    {
+      "epoch": 0.05557935827858707,
+      "grad_norm": 7.850117622462598,
+      "learning_rate": 2e-05,
+      "loss": 1.1182,
+      "step": 2335
+    },
+    {
+      "epoch": 0.055603161001618585,
+      "grad_norm": 6.096567438582376,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 2336
+    },
+    {
+      "epoch": 0.0556269637246501,
+      "grad_norm": 7.622674438619805,
+      "learning_rate": 2e-05,
+      "loss": 0.7823,
+      "step": 2337
+    },
+    {
+      "epoch": 0.05565076644768161,
+      "grad_norm": 8.0224127715676,
+      "learning_rate": 2e-05,
+      "loss": 0.8157,
+      "step": 2338
+    },
+    {
+      "epoch": 0.05567456917071313,
+      "grad_norm": 5.847406245217139,
+      "learning_rate": 2e-05,
+      "loss": 0.3186,
+      "step": 2339
+    },
+    {
+      "epoch": 0.055698371893744644,
+      "grad_norm": 6.057957372360134,
+      "learning_rate": 2e-05,
+      "loss": 0.67,
+      "step": 2340
+    },
+    {
+      "epoch": 0.05572217461677616,
+      "grad_norm": 9.634051011456089,
+      "learning_rate": 2e-05,
+      "loss": 1.5071,
+      "step": 2341
+    },
+    {
+      "epoch": 0.05574597733980768,
+      "grad_norm": 5.242654376557789,
+      "learning_rate": 2e-05,
+      "loss": 0.3334,
+      "step": 2342
+    },
+    {
+      "epoch": 0.05576978006283919,
+      "grad_norm": 8.55262331266052,
+      "learning_rate": 2e-05,
+      "loss": 1.8555,
+      "step": 2343
+    },
+    {
+      "epoch": 0.0557935827858707,
+      "grad_norm": 3.3915096565480485,
+      "learning_rate": 2e-05,
+      "loss": 0.2988,
+      "step": 2344
+    },
+    {
+      "epoch": 0.055817385508902216,
+      "grad_norm": 3.0918228044690528,
+      "learning_rate": 2e-05,
+      "loss": 0.2223,
+      "step": 2345
+    },
+    {
+      "epoch": 0.055841188231933736,
+      "grad_norm": 2.393948603444888,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 2346
+    },
+    {
+      "epoch": 0.05586499095496525,
+      "grad_norm": 6.657730492488347,
+      "learning_rate": 2e-05,
+      "loss": 0.4406,
+      "step": 2347
+    },
+    {
+      "epoch": 0.05588879367799676,
+      "grad_norm": 5.3500098706013475,
+      "learning_rate": 2e-05,
+      "loss": 0.2973,
+      "step": 2348
+    },
+    {
+      "epoch": 0.055912596401028275,
+      "grad_norm": 5.140522970881745,
+      "learning_rate": 2e-05,
+      "loss": 0.896,
+      "step": 2349
+    },
+    {
+      "epoch": 0.055936399124059795,
+      "grad_norm": 2.823132006526261,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 2350
+    },
+    {
+      "epoch": 0.05596020184709131,
+      "grad_norm": 7.421102244062846,
+      "learning_rate": 2e-05,
+      "loss": 0.2557,
+      "step": 2351
+    },
+    {
+      "epoch": 0.05598400457012282,
+      "grad_norm": 10.198299684262473,
+      "learning_rate": 2e-05,
+      "loss": 1.1485,
+      "step": 2352
+    },
+    {
+      "epoch": 0.056007807293154334,
+      "grad_norm": 6.378330659040908,
+      "learning_rate": 2e-05,
+      "loss": 0.3029,
+      "step": 2353
+    },
+    {
+      "epoch": 0.056031610016185854,
+      "grad_norm": 8.526207731477035,
+      "learning_rate": 2e-05,
+      "loss": 1.4563,
+      "step": 2354
+    },
+    {
+      "epoch": 0.05605541273921737,
+      "grad_norm": 8.165699870628425,
+      "learning_rate": 2e-05,
+      "loss": 0.9684,
+      "step": 2355
+    },
+    {
+      "epoch": 0.05607921546224888,
+      "grad_norm": 6.281715809888741,
+      "learning_rate": 2e-05,
+      "loss": 0.3113,
+      "step": 2356
+    },
+    {
+      "epoch": 0.05610301818528039,
+      "grad_norm": 3.250070329648912,
+      "learning_rate": 2e-05,
+      "loss": 0.2869,
+      "step": 2357
+    },
+    {
+      "epoch": 0.05612682090831191,
+      "grad_norm": 7.553999350778307,
+      "learning_rate": 2e-05,
+      "loss": 0.8656,
+      "step": 2358
+    },
+    {
+      "epoch": 0.056150623631343426,
+      "grad_norm": 5.714127499304419,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 2359
+    },
+    {
+      "epoch": 0.05617442635437494,
+      "grad_norm": 11.575545308988572,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 2360
+    },
+    {
+      "epoch": 0.05619822907740646,
+      "grad_norm": 5.994844682506816,
+      "learning_rate": 2e-05,
+      "loss": 1.2581,
+      "step": 2361
+    },
+    {
+      "epoch": 0.05622203180043797,
+      "grad_norm": 5.8943033431695735,
+      "learning_rate": 2e-05,
+      "loss": 1.2078,
+      "step": 2362
+    },
+    {
+      "epoch": 0.056245834523469485,
+      "grad_norm": 4.406713430583568,
+      "learning_rate": 2e-05,
+      "loss": 0.6224,
+      "step": 2363
+    },
+    {
+      "epoch": 0.056269637246501,
+      "grad_norm": 7.044555203890535,
+      "learning_rate": 2e-05,
+      "loss": 1.1688,
+      "step": 2364
+    },
+    {
+      "epoch": 0.05629343996953252,
+      "grad_norm": 6.390982078665538,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 2365
+    },
+    {
+      "epoch": 0.05631724269256403,
+      "grad_norm": 74.87889879825471,
+      "learning_rate": 2e-05,
+      "loss": 0.6385,
+      "step": 2366
+    },
+    {
+      "epoch": 0.05634104541559554,
+      "grad_norm": 6.167724053437765,
+      "learning_rate": 2e-05,
+      "loss": 0.6022,
+      "step": 2367
+    },
+    {
+      "epoch": 0.056364848138627056,
+      "grad_norm": 9.646662566067981,
+      "learning_rate": 2e-05,
+      "loss": 1.3664,
+      "step": 2368
+    },
+    {
+      "epoch": 0.056388650861658576,
+      "grad_norm": 3.441582461705852,
+      "learning_rate": 2e-05,
+      "loss": 0.2546,
+      "step": 2369
+    },
+    {
+      "epoch": 0.05641245358469009,
+      "grad_norm": 3.2363820643189665,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0564362563077216,
+      "grad_norm": 16.09879765763909,
+      "learning_rate": 2e-05,
+      "loss": 0.9629,
+      "step": 2371
+    },
+    {
+      "epoch": 0.056460059030753115,
+      "grad_norm": 8.143954826913317,
+      "learning_rate": 2e-05,
+      "loss": 1.001,
+      "step": 2372
+    },
+    {
+      "epoch": 0.056483861753784635,
+      "grad_norm": 3.593474601999878,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 2373
+    },
+    {
+      "epoch": 0.05650766447681615,
+      "grad_norm": 3.9104949379760856,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 2374
+    },
+    {
+      "epoch": 0.05653146719984766,
+      "grad_norm": 5.203848876248387,
+      "learning_rate": 2e-05,
+      "loss": 0.2757,
+      "step": 2375
+    },
+    {
+      "epoch": 0.056555269922879174,
+      "grad_norm": 4.257452005835131,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 2376
+    },
+    {
+      "epoch": 0.056579072645910694,
+      "grad_norm": 5.439865372377364,
+      "learning_rate": 2e-05,
+      "loss": 0.7674,
+      "step": 2377
+    },
+    {
+      "epoch": 0.05660287536894221,
+      "grad_norm": 3.104042141044748,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 2378
+    },
+    {
+      "epoch": 0.05662667809197372,
+      "grad_norm": 4.02994474535265,
+      "learning_rate": 2e-05,
+      "loss": 0.3002,
+      "step": 2379
+    },
+    {
+      "epoch": 0.05665048081500524,
+      "grad_norm": 2.4226325418112045,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 2380
+    },
+    {
+      "epoch": 0.05667428353803675,
+      "grad_norm": 3.1882653743761256,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 2381
+    },
+    {
+      "epoch": 0.056698086261068266,
+      "grad_norm": 8.28416963773986,
+      "learning_rate": 2e-05,
+      "loss": 0.4218,
+      "step": 2382
+    },
+    {
+      "epoch": 0.05672188898409978,
+      "grad_norm": 7.194023769664835,
+      "learning_rate": 2e-05,
+      "loss": 1.3022,
+      "step": 2383
+    },
+    {
+      "epoch": 0.0567456917071313,
+      "grad_norm": 3.571583969940701,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 2384
+    },
+    {
+      "epoch": 0.05676949443016281,
+      "grad_norm": 23.450572227739077,
+      "learning_rate": 2e-05,
+      "loss": 1.8479,
+      "step": 2385
+    },
+    {
+      "epoch": 0.056793297153194325,
+      "grad_norm": 6.188684645253961,
+      "learning_rate": 2e-05,
+      "loss": 0.9389,
+      "step": 2386
+    },
+    {
+      "epoch": 0.05681709987622584,
+      "grad_norm": 2.8800580781388017,
+      "learning_rate": 2e-05,
+      "loss": 0.194,
+      "step": 2387
+    },
+    {
+      "epoch": 0.05684090259925736,
+      "grad_norm": 3.0704902368219913,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 2388
+    },
+    {
+      "epoch": 0.05686470532228887,
+      "grad_norm": 3.0629707863367432,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 2389
+    },
+    {
+      "epoch": 0.056888508045320384,
+      "grad_norm": 2.339874593660796,
+      "learning_rate": 2e-05,
+      "loss": 0.1624,
+      "step": 2390
+    },
+    {
+      "epoch": 0.0569123107683519,
+      "grad_norm": 5.634629322684501,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 2391
+    },
+    {
+      "epoch": 0.05693611349138342,
+      "grad_norm": 7.595134042001959,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 2392
+    },
+    {
+      "epoch": 0.05695991621441493,
+      "grad_norm": 7.478284568309938,
+      "learning_rate": 2e-05,
+      "loss": 0.2137,
+      "step": 2393
+    },
+    {
+      "epoch": 0.05698371893744644,
+      "grad_norm": 6.992240420475054,
+      "learning_rate": 2e-05,
+      "loss": 1.0427,
+      "step": 2394
+    },
+    {
+      "epoch": 0.057007521660477956,
+      "grad_norm": 3.3652232029701254,
+      "learning_rate": 2e-05,
+      "loss": 0.1684,
+      "step": 2395
+    },
+    {
+      "epoch": 0.057031324383509475,
+      "grad_norm": 2.4863935853556614,
+      "learning_rate": 2e-05,
+      "loss": 0.2891,
+      "step": 2396
+    },
+    {
+      "epoch": 0.05705512710654099,
+      "grad_norm": 2.9237727738925647,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 2397
+    },
+    {
+      "epoch": 0.0570789298295725,
+      "grad_norm": 2.391972640572644,
+      "learning_rate": 2e-05,
+      "loss": 0.2773,
+      "step": 2398
+    },
+    {
+      "epoch": 0.05710273255260402,
+      "grad_norm": 37.96447393609738,
+      "learning_rate": 2e-05,
+      "loss": 2.2736,
+      "step": 2399
+    },
+    {
+      "epoch": 0.057126535275635534,
+      "grad_norm": 2.2131480418597524,
+      "learning_rate": 2e-05,
+      "loss": 0.1787,
+      "step": 2400
+    },
+    {
+      "epoch": 0.05715033799866705,
+      "grad_norm": 3.417777237017815,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 2401
+    },
+    {
+      "epoch": 0.05717414072169856,
+      "grad_norm": 19.892552328811853,
+      "learning_rate": 2e-05,
+      "loss": 0.5218,
+      "step": 2402
+    },
+    {
+      "epoch": 0.05719794344473008,
+      "grad_norm": 6.139430228751972,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 2403
+    },
+    {
+      "epoch": 0.05722174616776159,
+      "grad_norm": 3.2125857523063277,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 2404
+    },
+    {
+      "epoch": 0.057245548890793106,
+      "grad_norm": 7.071746130272682,
+      "learning_rate": 2e-05,
+      "loss": 0.3398,
+      "step": 2405
+    },
+    {
+      "epoch": 0.05726935161382462,
+      "grad_norm": 9.70294839149497,
+      "learning_rate": 2e-05,
+      "loss": 0.3963,
+      "step": 2406
+    },
+    {
+      "epoch": 0.05729315433685614,
+      "grad_norm": 5.887469682135142,
+      "learning_rate": 2e-05,
+      "loss": 0.6802,
+      "step": 2407
+    },
+    {
+      "epoch": 0.05731695705988765,
+      "grad_norm": 6.29603813197656,
+      "learning_rate": 2e-05,
+      "loss": 1.027,
+      "step": 2408
+    },
+    {
+      "epoch": 0.057340759782919165,
+      "grad_norm": 7.37864205309983,
+      "learning_rate": 2e-05,
+      "loss": 0.9138,
+      "step": 2409
+    },
+    {
+      "epoch": 0.05736456250595068,
+      "grad_norm": 6.459576662217883,
+      "learning_rate": 2e-05,
+      "loss": 0.8255,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0573883652289822,
+      "grad_norm": 3.308579883005965,
+      "learning_rate": 2e-05,
+      "loss": 0.308,
+      "step": 2411
+    },
+    {
+      "epoch": 0.05741216795201371,
+      "grad_norm": 5.636442890292182,
+      "learning_rate": 2e-05,
+      "loss": 0.4151,
+      "step": 2412
+    },
+    {
+      "epoch": 0.057435970675045224,
+      "grad_norm": 2.9039590797562016,
+      "learning_rate": 2e-05,
+      "loss": 0.2531,
+      "step": 2413
+    },
+    {
+      "epoch": 0.05745977339807674,
+      "grad_norm": 17.399259535808163,
+      "learning_rate": 2e-05,
+      "loss": 0.7829,
+      "step": 2414
+    },
+    {
+      "epoch": 0.05748357612110826,
+      "grad_norm": 2.3500927702284184,
+      "learning_rate": 2e-05,
+      "loss": 0.2016,
+      "step": 2415
+    },
+    {
+      "epoch": 0.05750737884413977,
+      "grad_norm": 6.971144150125065,
+      "learning_rate": 2e-05,
+      "loss": 0.9954,
+      "step": 2416
+    },
+    {
+      "epoch": 0.05753118156717128,
+      "grad_norm": 3.4515657728535247,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 2417
+    },
+    {
+      "epoch": 0.057554984290202796,
+      "grad_norm": 5.583764327819624,
+      "learning_rate": 2e-05,
+      "loss": 0.7506,
+      "step": 2418
+    },
+    {
+      "epoch": 0.057578787013234316,
+      "grad_norm": 3.034749816872133,
+      "learning_rate": 2e-05,
+      "loss": 0.3001,
+      "step": 2419
+    },
+    {
+      "epoch": 0.05760258973626583,
+      "grad_norm": 3.375275256138148,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 2420
+    },
+    {
+      "epoch": 0.05762639245929734,
+      "grad_norm": 7.54454901267898,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 2421
+    },
+    {
+      "epoch": 0.05765019518232886,
+      "grad_norm": 5.030404759192223,
+      "learning_rate": 2e-05,
+      "loss": 0.5946,
+      "step": 2422
+    },
+    {
+      "epoch": 0.057673997905360375,
+      "grad_norm": 7.746089986746256,
+      "learning_rate": 2e-05,
+      "loss": 0.6404,
+      "step": 2423
+    },
+    {
+      "epoch": 0.05769780062839189,
+      "grad_norm": 2.1976261661315637,
+      "learning_rate": 2e-05,
+      "loss": 0.2058,
+      "step": 2424
+    },
+    {
+      "epoch": 0.0577216033514234,
+      "grad_norm": 9.450472254825034,
+      "learning_rate": 2e-05,
+      "loss": 0.6481,
+      "step": 2425
+    },
+    {
+      "epoch": 0.05774540607445492,
+      "grad_norm": 2.201131323707734,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 2426
+    },
+    {
+      "epoch": 0.057769208797486434,
+      "grad_norm": 8.864811636780686,
+      "learning_rate": 2e-05,
+      "loss": 0.5454,
+      "step": 2427
+    },
+    {
+      "epoch": 0.057793011520517946,
+      "grad_norm": 9.389451476456685,
+      "learning_rate": 2e-05,
+      "loss": 0.8192,
+      "step": 2428
+    },
+    {
+      "epoch": 0.05781681424354946,
+      "grad_norm": 6.3295040852368265,
+      "learning_rate": 2e-05,
+      "loss": 1.0366,
+      "step": 2429
+    },
+    {
+      "epoch": 0.05784061696658098,
+      "grad_norm": 7.218313634831698,
+      "learning_rate": 2e-05,
+      "loss": 1.0387,
+      "step": 2430
+    },
+    {
+      "epoch": 0.05786441968961249,
+      "grad_norm": 4.384133984991316,
+      "learning_rate": 2e-05,
+      "loss": 0.3277,
+      "step": 2431
+    },
+    {
+      "epoch": 0.057888222412644005,
+      "grad_norm": 4.35076665964926,
+      "learning_rate": 2e-05,
+      "loss": 0.3675,
+      "step": 2432
+    },
+    {
+      "epoch": 0.05791202513567552,
+      "grad_norm": 4.516549255909019,
+      "learning_rate": 2e-05,
+      "loss": 0.2921,
+      "step": 2433
+    },
+    {
+      "epoch": 0.05793582785870704,
+      "grad_norm": 5.9290478259086745,
+      "learning_rate": 2e-05,
+      "loss": 1.1909,
+      "step": 2434
+    },
+    {
+      "epoch": 0.05795963058173855,
+      "grad_norm": 4.33531902635708,
+      "learning_rate": 2e-05,
+      "loss": 0.3261,
+      "step": 2435
+    },
+    {
+      "epoch": 0.057983433304770064,
+      "grad_norm": 6.761657742988242,
+      "learning_rate": 2e-05,
+      "loss": 0.8611,
+      "step": 2436
+    },
+    {
+      "epoch": 0.05800723602780158,
+      "grad_norm": 4.305565659825921,
+      "learning_rate": 2e-05,
+      "loss": 0.293,
+      "step": 2437
+    },
+    {
+      "epoch": 0.0580310387508331,
+      "grad_norm": 7.936137481326512,
+      "learning_rate": 2e-05,
+      "loss": 0.8231,
+      "step": 2438
+    },
+    {
+      "epoch": 0.05805484147386461,
+      "grad_norm": 12.166497355732307,
+      "learning_rate": 2e-05,
+      "loss": 1.9432,
+      "step": 2439
+    },
+    {
+      "epoch": 0.05807864419689612,
+      "grad_norm": 3.304374753632971,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 2440
+    },
+    {
+      "epoch": 0.05810244691992764,
+      "grad_norm": 7.214056365722447,
+      "learning_rate": 2e-05,
+      "loss": 0.6514,
+      "step": 2441
+    },
+    {
+      "epoch": 0.058126249642959156,
+      "grad_norm": 6.0919404854849075,
+      "learning_rate": 2e-05,
+      "loss": 0.2828,
+      "step": 2442
+    },
+    {
+      "epoch": 0.05815005236599067,
+      "grad_norm": 3.251484203508902,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 2443
+    },
+    {
+      "epoch": 0.05817385508902218,
+      "grad_norm": 5.478170668116647,
+      "learning_rate": 2e-05,
+      "loss": 1.0136,
+      "step": 2444
+    },
+    {
+      "epoch": 0.0581976578120537,
+      "grad_norm": 3.9092967158118146,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 2445
+    },
+    {
+      "epoch": 0.058221460535085215,
+      "grad_norm": 2.846669476680181,
+      "learning_rate": 2e-05,
+      "loss": 0.1555,
+      "step": 2446
+    },
+    {
+      "epoch": 0.05824526325811673,
+      "grad_norm": 8.256200171335578,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 2447
+    },
+    {
+      "epoch": 0.05826906598114824,
+      "grad_norm": 9.295473139180352,
+      "learning_rate": 2e-05,
+      "loss": 0.8469,
+      "step": 2448
+    },
+    {
+      "epoch": 0.05829286870417976,
+      "grad_norm": 4.741123678200566,
+      "learning_rate": 2e-05,
+      "loss": 0.7506,
+      "step": 2449
+    },
+    {
+      "epoch": 0.058316671427211274,
+      "grad_norm": 9.34475294152168,
+      "learning_rate": 2e-05,
+      "loss": 0.4454,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05834047415024279,
+      "grad_norm": 3.0898035676125124,
+      "learning_rate": 2e-05,
+      "loss": 0.2907,
+      "step": 2451
+    },
+    {
+      "epoch": 0.0583642768732743,
+      "grad_norm": 5.744861723856981,
+      "learning_rate": 2e-05,
+      "loss": 0.2596,
+      "step": 2452
+    },
+    {
+      "epoch": 0.05838807959630582,
+      "grad_norm": 5.8374474362824,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 2453
+    },
+    {
+      "epoch": 0.05841188231933733,
+      "grad_norm": 5.946255163171158,
+      "learning_rate": 2e-05,
+      "loss": 0.6199,
+      "step": 2454
+    },
+    {
+      "epoch": 0.058435685042368846,
+      "grad_norm": 7.732649285180383,
+      "learning_rate": 2e-05,
+      "loss": 0.6541,
+      "step": 2455
+    },
+    {
+      "epoch": 0.05845948776540036,
+      "grad_norm": 5.795643608548028,
+      "learning_rate": 2e-05,
+      "loss": 0.5348,
+      "step": 2456
+    },
+    {
+      "epoch": 0.05848329048843188,
+      "grad_norm": 7.891939966499033,
+      "learning_rate": 2e-05,
+      "loss": 0.5795,
+      "step": 2457
+    },
+    {
+      "epoch": 0.05850709321146339,
+      "grad_norm": 9.225119276887291,
+      "learning_rate": 2e-05,
+      "loss": 1.098,
+      "step": 2458
+    },
+    {
+      "epoch": 0.058530895934494905,
+      "grad_norm": 7.957922628692141,
+      "learning_rate": 2e-05,
+      "loss": 0.638,
+      "step": 2459
+    },
+    {
+      "epoch": 0.058554698657526424,
+      "grad_norm": 3.072917354396334,
+      "learning_rate": 2e-05,
+      "loss": 0.2279,
+      "step": 2460
+    },
+    {
+      "epoch": 0.05857850138055794,
+      "grad_norm": 2.1474907486007884,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 2461
+    },
+    {
+      "epoch": 0.05860230410358945,
+      "grad_norm": 6.10008960064569,
+      "learning_rate": 2e-05,
+      "loss": 0.7142,
+      "step": 2462
+    },
+    {
+      "epoch": 0.05862610682662096,
+      "grad_norm": 6.649841412303756,
+      "learning_rate": 2e-05,
+      "loss": 0.3627,
+      "step": 2463
+    },
+    {
+      "epoch": 0.05864990954965248,
+      "grad_norm": 7.0635986167167255,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 2464
+    },
+    {
+      "epoch": 0.058673712272683996,
+      "grad_norm": 2.9152750231715268,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 2465
+    },
+    {
+      "epoch": 0.05869751499571551,
+      "grad_norm": 3.523490744768428,
+      "learning_rate": 2e-05,
+      "loss": 0.1824,
+      "step": 2466
+    },
+    {
+      "epoch": 0.05872131771874702,
+      "grad_norm": 5.326429112480473,
+      "learning_rate": 2e-05,
+      "loss": 0.6898,
+      "step": 2467
+    },
+    {
+      "epoch": 0.05874512044177854,
+      "grad_norm": 3.457429840718745,
+      "learning_rate": 2e-05,
+      "loss": 0.1783,
+      "step": 2468
+    },
+    {
+      "epoch": 0.058768923164810055,
+      "grad_norm": 3.513243819025155,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 2469
+    },
+    {
+      "epoch": 0.05879272588784157,
+      "grad_norm": 5.233945400518374,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05881652861087308,
+      "grad_norm": 6.772756089723057,
+      "learning_rate": 2e-05,
+      "loss": 1.0023,
+      "step": 2471
+    },
+    {
+      "epoch": 0.0588403313339046,
+      "grad_norm": 8.968767845018068,
+      "learning_rate": 2e-05,
+      "loss": 0.4508,
+      "step": 2472
+    },
+    {
+      "epoch": 0.058864134056936114,
+      "grad_norm": 3.204353416209323,
+      "learning_rate": 2e-05,
+      "loss": 0.2892,
+      "step": 2473
+    },
+    {
+      "epoch": 0.05888793677996763,
+      "grad_norm": 6.274580838949348,
+      "learning_rate": 2e-05,
+      "loss": 0.7304,
+      "step": 2474
+    },
+    {
+      "epoch": 0.05891173950299914,
+      "grad_norm": 2.879558434135173,
+      "learning_rate": 2e-05,
+      "loss": 0.1987,
+      "step": 2475
+    },
+    {
+      "epoch": 0.05893554222603066,
+      "grad_norm": 9.18215883378984,
+      "learning_rate": 2e-05,
+      "loss": 0.486,
+      "step": 2476
+    },
+    {
+      "epoch": 0.05895934494906217,
+      "grad_norm": 6.249208884675642,
+      "learning_rate": 2e-05,
+      "loss": 0.3365,
+      "step": 2477
+    },
+    {
+      "epoch": 0.058983147672093686,
+      "grad_norm": 10.527277079693667,
+      "learning_rate": 2e-05,
+      "loss": 1.2273,
+      "step": 2478
+    },
+    {
+      "epoch": 0.0590069503951252,
+      "grad_norm": 9.601967051473768,
+      "learning_rate": 2e-05,
+      "loss": 0.8949,
+      "step": 2479
+    },
+    {
+      "epoch": 0.05903075311815672,
+      "grad_norm": 8.721605591066297,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 2480
+    },
+    {
+      "epoch": 0.05905455584118823,
+      "grad_norm": 8.129499522900064,
+      "learning_rate": 2e-05,
+      "loss": 0.2634,
+      "step": 2481
+    },
+    {
+      "epoch": 0.059078358564219745,
+      "grad_norm": 5.574799647197961,
+      "learning_rate": 2e-05,
+      "loss": 0.294,
+      "step": 2482
+    },
+    {
+      "epoch": 0.059102161287251265,
+      "grad_norm": 5.558760363989063,
+      "learning_rate": 2e-05,
+      "loss": 0.3027,
+      "step": 2483
+    },
+    {
+      "epoch": 0.05912596401028278,
+      "grad_norm": 7.4239322009794,
+      "learning_rate": 2e-05,
+      "loss": 0.5845,
+      "step": 2484
+    },
+    {
+      "epoch": 0.05914976673331429,
+      "grad_norm": 2.102624208486325,
+      "learning_rate": 2e-05,
+      "loss": 0.1565,
+      "step": 2485
+    },
+    {
+      "epoch": 0.059173569456345804,
+      "grad_norm": 4.894760553434682,
+      "learning_rate": 2e-05,
+      "loss": 0.2053,
+      "step": 2486
+    },
+    {
+      "epoch": 0.059197372179377324,
+      "grad_norm": 9.776663398285285,
+      "learning_rate": 2e-05,
+      "loss": 0.4353,
+      "step": 2487
+    },
+    {
+      "epoch": 0.05922117490240884,
+      "grad_norm": 5.786471379997944,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 2488
+    },
+    {
+      "epoch": 0.05924497762544035,
+      "grad_norm": 15.483249481204302,
+      "learning_rate": 2e-05,
+      "loss": 0.527,
+      "step": 2489
+    },
+    {
+      "epoch": 0.05926878034847186,
+      "grad_norm": 8.86514628593651,
+      "learning_rate": 2e-05,
+      "loss": 0.4854,
+      "step": 2490
+    },
+    {
+      "epoch": 0.05929258307150338,
+      "grad_norm": 3.393465611730735,
+      "learning_rate": 2e-05,
+      "loss": 0.2965,
+      "step": 2491
+    },
+    {
+      "epoch": 0.059316385794534895,
+      "grad_norm": 4.253932549471064,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 2492
+    },
+    {
+      "epoch": 0.05934018851756641,
+      "grad_norm": 3.3350409435154176,
+      "learning_rate": 2e-05,
+      "loss": 0.1665,
+      "step": 2493
+    },
+    {
+      "epoch": 0.05936399124059792,
+      "grad_norm": 4.225523997573772,
+      "learning_rate": 2e-05,
+      "loss": 0.3303,
+      "step": 2494
+    },
+    {
+      "epoch": 0.05938779396362944,
+      "grad_norm": 3.5959647230933,
+      "learning_rate": 2e-05,
+      "loss": 0.1566,
+      "step": 2495
+    },
+    {
+      "epoch": 0.059411596686660954,
+      "grad_norm": 6.630836327456906,
+      "learning_rate": 2e-05,
+      "loss": 0.5221,
+      "step": 2496
+    },
+    {
+      "epoch": 0.05943539940969247,
+      "grad_norm": 9.134767348430696,
+      "learning_rate": 2e-05,
+      "loss": 0.4389,
+      "step": 2497
+    },
+    {
+      "epoch": 0.05945920213272398,
+      "grad_norm": 9.567320754272673,
+      "learning_rate": 2e-05,
+      "loss": 0.2701,
+      "step": 2498
+    },
+    {
+      "epoch": 0.0594830048557555,
+      "grad_norm": 5.160746114708289,
+      "learning_rate": 2e-05,
+      "loss": 0.4409,
+      "step": 2499
+    },
+    {
+      "epoch": 0.05950680757878701,
+      "grad_norm": 9.310700457595477,
+      "learning_rate": 2e-05,
+      "loss": 0.4406,
+      "step": 2500
+    },
+    {
+      "epoch": 0.059530610301818526,
+      "grad_norm": 2.6856873882956984,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 2501
+    },
+    {
+      "epoch": 0.059554413024850046,
+      "grad_norm": 6.872968482818884,
+      "learning_rate": 2e-05,
+      "loss": 0.3979,
+      "step": 2502
+    },
+    {
+      "epoch": 0.05957821574788156,
+      "grad_norm": 6.379181330077584,
+      "learning_rate": 2e-05,
+      "loss": 0.4441,
+      "step": 2503
+    },
+    {
+      "epoch": 0.05960201847091307,
+      "grad_norm": 8.017961497371074,
+      "learning_rate": 2e-05,
+      "loss": 0.3381,
+      "step": 2504
+    },
+    {
+      "epoch": 0.059625821193944585,
+      "grad_norm": 3.5301510864127854,
+      "learning_rate": 2e-05,
+      "loss": 0.2701,
+      "step": 2505
+    },
+    {
+      "epoch": 0.059649623916976105,
+      "grad_norm": 1.9002668710407067,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 2506
+    },
+    {
+      "epoch": 0.05967342664000762,
+      "grad_norm": 3.572446899243495,
+      "learning_rate": 2e-05,
+      "loss": 0.1644,
+      "step": 2507
+    },
+    {
+      "epoch": 0.05969722936303913,
+      "grad_norm": 7.389639054305788,
+      "learning_rate": 2e-05,
+      "loss": 0.2855,
+      "step": 2508
+    },
+    {
+      "epoch": 0.059721032086070644,
+      "grad_norm": 5.32220782691602,
+      "learning_rate": 2e-05,
+      "loss": 0.7259,
+      "step": 2509
+    },
+    {
+      "epoch": 0.059744834809102164,
+      "grad_norm": 7.511016530550258,
+      "learning_rate": 2e-05,
+      "loss": 0.287,
+      "step": 2510
+    },
+    {
+      "epoch": 0.05976863753213368,
+      "grad_norm": 3.5289321148229944,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 2511
+    },
+    {
+      "epoch": 0.05979244025516519,
+      "grad_norm": 15.089122790766794,
+      "learning_rate": 2e-05,
+      "loss": 0.6948,
+      "step": 2512
+    },
+    {
+      "epoch": 0.0598162429781967,
+      "grad_norm": 6.7937695352750405,
+      "learning_rate": 2e-05,
+      "loss": 0.768,
+      "step": 2513
+    },
+    {
+      "epoch": 0.05984004570122822,
+      "grad_norm": 2.6504221137050172,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 2514
+    },
+    {
+      "epoch": 0.059863848424259736,
+      "grad_norm": 6.962868079090827,
+      "learning_rate": 2e-05,
+      "loss": 0.431,
+      "step": 2515
+    },
+    {
+      "epoch": 0.05988765114729125,
+      "grad_norm": 7.6967557896765175,
+      "learning_rate": 2e-05,
+      "loss": 0.5568,
+      "step": 2516
+    },
+    {
+      "epoch": 0.05991145387032276,
+      "grad_norm": 2.6079393305377407,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 2517
+    },
+    {
+      "epoch": 0.05993525659335428,
+      "grad_norm": 0.9824802442177614,
+      "learning_rate": 2e-05,
+      "loss": 0.1934,
+      "step": 2518
+    },
+    {
+      "epoch": 0.059959059316385795,
+      "grad_norm": 2.934061662717554,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 2519
+    },
+    {
+      "epoch": 0.05998286203941731,
+      "grad_norm": 4.125571807745933,
+      "learning_rate": 2e-05,
+      "loss": 0.3146,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06000666476244883,
+      "grad_norm": 2.53721729465653,
+      "learning_rate": 2e-05,
+      "loss": 0.2788,
+      "step": 2521
+    },
+    {
+      "epoch": 0.06003046748548034,
+      "grad_norm": 7.514301094298597,
+      "learning_rate": 2e-05,
+      "loss": 0.384,
+      "step": 2522
+    },
+    {
+      "epoch": 0.060054270208511853,
+      "grad_norm": 7.1063540220922885,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 2523
+    },
+    {
+      "epoch": 0.060078072931543366,
+      "grad_norm": 7.357902331458016,
+      "learning_rate": 2e-05,
+      "loss": 1.1092,
+      "step": 2524
+    },
+    {
+      "epoch": 0.060101875654574886,
+      "grad_norm": 3.050272718176716,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 2525
+    },
+    {
+      "epoch": 0.0601256783776064,
+      "grad_norm": 6.504190576488973,
+      "learning_rate": 2e-05,
+      "loss": 1.0712,
+      "step": 2526
+    },
+    {
+      "epoch": 0.06014948110063791,
+      "grad_norm": 5.823311266805192,
+      "learning_rate": 2e-05,
+      "loss": 1.0229,
+      "step": 2527
+    },
+    {
+      "epoch": 0.060173283823669425,
+      "grad_norm": 5.612707847430326,
+      "learning_rate": 2e-05,
+      "loss": 0.7948,
+      "step": 2528
+    },
+    {
+      "epoch": 0.060197086546700945,
+      "grad_norm": 3.14239873809933,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 2529
+    },
+    {
+      "epoch": 0.06022088926973246,
+      "grad_norm": 7.453682807351332,
+      "learning_rate": 2e-05,
+      "loss": 0.386,
+      "step": 2530
+    },
+    {
+      "epoch": 0.06024469199276397,
+      "grad_norm": 11.759825608327965,
+      "learning_rate": 2e-05,
+      "loss": 0.4742,
+      "step": 2531
+    },
+    {
+      "epoch": 0.060268494715795484,
+      "grad_norm": 7.268180787418823,
+      "learning_rate": 2e-05,
+      "loss": 1.4917,
+      "step": 2532
+    },
+    {
+      "epoch": 0.060292297438827004,
+      "grad_norm": 3.2263689327117455,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 2533
+    },
+    {
+      "epoch": 0.06031610016185852,
+      "grad_norm": 21.168747935089407,
+      "learning_rate": 2e-05,
+      "loss": 0.9549,
+      "step": 2534
+    },
+    {
+      "epoch": 0.06033990288489003,
+      "grad_norm": 5.138098583530791,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 2535
+    },
+    {
+      "epoch": 0.06036370560792154,
+      "grad_norm": 3.3286774250720197,
+      "learning_rate": 2e-05,
+      "loss": 0.3519,
+      "step": 2536
+    },
+    {
+      "epoch": 0.06038750833095306,
+      "grad_norm": 2.7585971946768524,
+      "learning_rate": 2e-05,
+      "loss": 0.2883,
+      "step": 2537
+    },
+    {
+      "epoch": 0.060411311053984576,
+      "grad_norm": 6.549862687752755,
+      "learning_rate": 2e-05,
+      "loss": 0.995,
+      "step": 2538
+    },
+    {
+      "epoch": 0.06043511377701609,
+      "grad_norm": 7.952326352505554,
+      "learning_rate": 2e-05,
+      "loss": 1.0266,
+      "step": 2539
+    },
+    {
+      "epoch": 0.0604589165000476,
+      "grad_norm": 6.6429542980758205,
+      "learning_rate": 2e-05,
+      "loss": 1.2733,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06048271922307912,
+      "grad_norm": 5.031950170314386,
+      "learning_rate": 2e-05,
+      "loss": 0.898,
+      "step": 2541
+    },
+    {
+      "epoch": 0.060506521946110635,
+      "grad_norm": 3.153826662070658,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 2542
+    },
+    {
+      "epoch": 0.06053032466914215,
+      "grad_norm": 7.1042129419842865,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 2543
+    },
+    {
+      "epoch": 0.06055412739217367,
+      "grad_norm": 5.721463036837453,
+      "learning_rate": 2e-05,
+      "loss": 0.6093,
+      "step": 2544
+    },
+    {
+      "epoch": 0.06057793011520518,
+      "grad_norm": 5.603282253820706,
+      "learning_rate": 2e-05,
+      "loss": 0.616,
+      "step": 2545
+    },
+    {
+      "epoch": 0.060601732838236694,
+      "grad_norm": 8.8169906622162,
+      "learning_rate": 2e-05,
+      "loss": 1.2759,
+      "step": 2546
+    },
+    {
+      "epoch": 0.06062553556126821,
+      "grad_norm": 2.525656076799951,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 2547
+    },
+    {
+      "epoch": 0.06064933828429973,
+      "grad_norm": 9.875268325070143,
+      "learning_rate": 2e-05,
+      "loss": 1.023,
+      "step": 2548
+    },
+    {
+      "epoch": 0.06067314100733124,
+      "grad_norm": 7.622586405353444,
+      "learning_rate": 2e-05,
+      "loss": 0.848,
+      "step": 2549
+    },
+    {
+      "epoch": 0.06069694373036275,
+      "grad_norm": 3.127075325115087,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 2550
+    },
+    {
+      "epoch": 0.060720746453394266,
+      "grad_norm": 2.8294000622407047,
+      "learning_rate": 2e-05,
+      "loss": 0.2827,
+      "step": 2551
+    },
+    {
+      "epoch": 0.060744549176425786,
+      "grad_norm": 6.42499152091342,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 2552
+    },
+    {
+      "epoch": 0.0607683518994573,
+      "grad_norm": 2.587428845610091,
+      "learning_rate": 2e-05,
+      "loss": 0.2184,
+      "step": 2553
+    },
+    {
+      "epoch": 0.06079215462248881,
+      "grad_norm": 4.398431132666361,
+      "learning_rate": 2e-05,
+      "loss": 0.3266,
+      "step": 2554
+    },
+    {
+      "epoch": 0.060815957345520325,
+      "grad_norm": 4.9501567537159765,
+      "learning_rate": 2e-05,
+      "loss": 0.2608,
+      "step": 2555
+    },
+    {
+      "epoch": 0.060839760068551844,
+      "grad_norm": 7.7577666344947644,
+      "learning_rate": 2e-05,
+      "loss": 1.5913,
+      "step": 2556
+    },
+    {
+      "epoch": 0.06086356279158336,
+      "grad_norm": 3.1567670037526385,
+      "learning_rate": 2e-05,
+      "loss": 0.1758,
+      "step": 2557
+    },
+    {
+      "epoch": 0.06088736551461487,
+      "grad_norm": 6.399146678251057,
+      "learning_rate": 2e-05,
+      "loss": 0.2867,
+      "step": 2558
+    },
+    {
+      "epoch": 0.06091116823764638,
+      "grad_norm": 3.222852342759878,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 2559
+    },
+    {
+      "epoch": 0.0609349709606779,
+      "grad_norm": 7.486777878483477,
+      "learning_rate": 2e-05,
+      "loss": 0.494,
+      "step": 2560
+    },
+    {
+      "epoch": 0.060958773683709416,
+      "grad_norm": 8.357308627982777,
+      "learning_rate": 2e-05,
+      "loss": 0.3362,
+      "step": 2561
+    },
+    {
+      "epoch": 0.06098257640674093,
+      "grad_norm": 8.025960092301137,
+      "learning_rate": 2e-05,
+      "loss": 1.0596,
+      "step": 2562
+    },
+    {
+      "epoch": 0.06100637912977245,
+      "grad_norm": 6.762107363769825,
+      "learning_rate": 2e-05,
+      "loss": 1.2185,
+      "step": 2563
+    },
+    {
+      "epoch": 0.06103018185280396,
+      "grad_norm": 7.892324371498659,
+      "learning_rate": 2e-05,
+      "loss": 0.3285,
+      "step": 2564
+    },
+    {
+      "epoch": 0.061053984575835475,
+      "grad_norm": 5.637364002281171,
+      "learning_rate": 2e-05,
+      "loss": 0.3932,
+      "step": 2565
+    },
+    {
+      "epoch": 0.06107778729886699,
+      "grad_norm": 5.363825533379345,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 2566
+    },
+    {
+      "epoch": 0.06110159002189851,
+      "grad_norm": 4.369594213793721,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 2567
+    },
+    {
+      "epoch": 0.06112539274493002,
+      "grad_norm": 1.201829326660626,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 2568
+    },
+    {
+      "epoch": 0.061149195467961534,
+      "grad_norm": 7.148989675177618,
+      "learning_rate": 2e-05,
+      "loss": 0.9852,
+      "step": 2569
+    },
+    {
+      "epoch": 0.06117299819099305,
+      "grad_norm": 2.643694744869802,
+      "learning_rate": 2e-05,
+      "loss": 0.2212,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06119680091402457,
+      "grad_norm": 6.070019162465637,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 2571
+    },
+    {
+      "epoch": 0.06122060363705608,
+      "grad_norm": 2.6644618981829526,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 2572
+    },
+    {
+      "epoch": 0.06124440636008759,
+      "grad_norm": 6.9404124168757635,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 2573
+    },
+    {
+      "epoch": 0.061268209083119106,
+      "grad_norm": 9.944122868968027,
+      "learning_rate": 2e-05,
+      "loss": 0.9948,
+      "step": 2574
+    },
+    {
+      "epoch": 0.061292011806150626,
+      "grad_norm": 5.803067551776539,
+      "learning_rate": 2e-05,
+      "loss": 0.7301,
+      "step": 2575
+    },
+    {
+      "epoch": 0.06131581452918214,
+      "grad_norm": 2.6748866375589757,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 2576
+    },
+    {
+      "epoch": 0.06133961725221365,
+      "grad_norm": 2.6317299923062243,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 2577
+    },
+    {
+      "epoch": 0.061363419975245165,
+      "grad_norm": 4.553134995245056,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 2578
+    },
+    {
+      "epoch": 0.061387222698276685,
+      "grad_norm": 4.51615689313321,
+      "learning_rate": 2e-05,
+      "loss": 0.7936,
+      "step": 2579
+    },
+    {
+      "epoch": 0.0614110254213082,
+      "grad_norm": 9.6974071426181,
+      "learning_rate": 2e-05,
+      "loss": 1.221,
+      "step": 2580
+    },
+    {
+      "epoch": 0.06143482814433971,
+      "grad_norm": 3.616907275494898,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 2581
+    },
+    {
+      "epoch": 0.06145863086737123,
+      "grad_norm": 5.549686703396384,
+      "learning_rate": 2e-05,
+      "loss": 0.9669,
+      "step": 2582
+    },
+    {
+      "epoch": 0.061482433590402744,
+      "grad_norm": 2.6922149923437213,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 2583
+    },
+    {
+      "epoch": 0.06150623631343426,
+      "grad_norm": 7.264020105495618,
+      "learning_rate": 2e-05,
+      "loss": 1.0498,
+      "step": 2584
+    },
+    {
+      "epoch": 0.06153003903646577,
+      "grad_norm": 4.2444009528950355,
+      "learning_rate": 2e-05,
+      "loss": 0.7808,
+      "step": 2585
+    },
+    {
+      "epoch": 0.06155384175949729,
+      "grad_norm": 6.848111973860081,
+      "learning_rate": 2e-05,
+      "loss": 0.6522,
+      "step": 2586
+    },
+    {
+      "epoch": 0.0615776444825288,
+      "grad_norm": 3.008256914733641,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 2587
+    },
+    {
+      "epoch": 0.061601447205560315,
+      "grad_norm": 2.598164152449428,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 2588
+    },
+    {
+      "epoch": 0.06162524992859183,
+      "grad_norm": 5.61080928664,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 2589
+    },
+    {
+      "epoch": 0.06164905265162335,
+      "grad_norm": 7.0295675999371054,
+      "learning_rate": 2e-05,
+      "loss": 1.1002,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06167285537465486,
+      "grad_norm": 55.04505328212287,
+      "learning_rate": 2e-05,
+      "loss": 1.1202,
+      "step": 2591
+    },
+    {
+      "epoch": 0.061696658097686374,
+      "grad_norm": 8.009135956998545,
+      "learning_rate": 2e-05,
+      "loss": 0.6265,
+      "step": 2592
+    },
+    {
+      "epoch": 0.06172046082071789,
+      "grad_norm": 5.4059500590429845,
+      "learning_rate": 2e-05,
+      "loss": 0.9347,
+      "step": 2593
+    },
+    {
+      "epoch": 0.06174426354374941,
+      "grad_norm": 6.946555084694261,
+      "learning_rate": 2e-05,
+      "loss": 0.3692,
+      "step": 2594
+    },
+    {
+      "epoch": 0.06176806626678092,
+      "grad_norm": 4.716971129668744,
+      "learning_rate": 2e-05,
+      "loss": 0.726,
+      "step": 2595
+    },
+    {
+      "epoch": 0.06179186898981243,
+      "grad_norm": 6.005819224174851,
+      "learning_rate": 2e-05,
+      "loss": 0.2746,
+      "step": 2596
+    },
+    {
+      "epoch": 0.061815671712843946,
+      "grad_norm": 6.083340298284522,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 2597
+    },
+    {
+      "epoch": 0.061839474435875466,
+      "grad_norm": 3.252124747710221,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 2598
+    },
+    {
+      "epoch": 0.06186327715890698,
+      "grad_norm": 7.476741945667098,
+      "learning_rate": 2e-05,
+      "loss": 0.5869,
+      "step": 2599
+    },
+    {
+      "epoch": 0.06188707988193849,
+      "grad_norm": 5.853025919118621,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06191088260497001,
+      "grad_norm": 3.3805063657854677,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 2601
+    },
+    {
+      "epoch": 0.061934685328001525,
+      "grad_norm": 7.315176352645741,
+      "learning_rate": 2e-05,
+      "loss": 0.2946,
+      "step": 2602
+    },
+    {
+      "epoch": 0.06195848805103304,
+      "grad_norm": 5.9576548590084,
+      "learning_rate": 2e-05,
+      "loss": 0.415,
+      "step": 2603
+    },
+    {
+      "epoch": 0.06198229077406455,
+      "grad_norm": 5.241287835222309,
+      "learning_rate": 2e-05,
+      "loss": 0.2453,
+      "step": 2604
+    },
+    {
+      "epoch": 0.06200609349709607,
+      "grad_norm": 4.5026867838212565,
+      "learning_rate": 2e-05,
+      "loss": 0.24,
+      "step": 2605
+    },
+    {
+      "epoch": 0.062029896220127584,
+      "grad_norm": 2.4703538196773045,
+      "learning_rate": 2e-05,
+      "loss": 0.2347,
+      "step": 2606
+    },
+    {
+      "epoch": 0.0620536989431591,
+      "grad_norm": 6.416240913941175,
+      "learning_rate": 2e-05,
+      "loss": 0.7102,
+      "step": 2607
+    },
+    {
+      "epoch": 0.06207750166619061,
+      "grad_norm": 4.1037578475702965,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 2608
+    },
+    {
+      "epoch": 0.06210130438922213,
+      "grad_norm": 6.346344321523282,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 2609
+    },
+    {
+      "epoch": 0.06212510711225364,
+      "grad_norm": 5.8121999531531054,
+      "learning_rate": 2e-05,
+      "loss": 0.3234,
+      "step": 2610
+    },
+    {
+      "epoch": 0.062148909835285156,
+      "grad_norm": 2.690297719886912,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 2611
+    },
+    {
+      "epoch": 0.06217271255831667,
+      "grad_norm": 4.4682417240921275,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 2612
+    },
+    {
+      "epoch": 0.06219651528134819,
+      "grad_norm": 5.901991671028361,
+      "learning_rate": 2e-05,
+      "loss": 0.669,
+      "step": 2613
+    },
+    {
+      "epoch": 0.0622203180043797,
+      "grad_norm": 4.498984005323895,
+      "learning_rate": 2e-05,
+      "loss": 0.2496,
+      "step": 2614
+    },
+    {
+      "epoch": 0.062244120727411215,
+      "grad_norm": 4.242192903281145,
+      "learning_rate": 2e-05,
+      "loss": 0.2244,
+      "step": 2615
+    },
+    {
+      "epoch": 0.06226792345044273,
+      "grad_norm": 11.215918825480097,
+      "learning_rate": 2e-05,
+      "loss": 0.5331,
+      "step": 2616
+    },
+    {
+      "epoch": 0.06229172617347425,
+      "grad_norm": 9.22614300699379,
+      "learning_rate": 2e-05,
+      "loss": 1.1395,
+      "step": 2617
+    },
+    {
+      "epoch": 0.06231552889650576,
+      "grad_norm": 5.8011052456502465,
+      "learning_rate": 2e-05,
+      "loss": 0.4518,
+      "step": 2618
+    },
+    {
+      "epoch": 0.062339331619537273,
+      "grad_norm": 5.38155716007437,
+      "learning_rate": 2e-05,
+      "loss": 0.6978,
+      "step": 2619
+    },
+    {
+      "epoch": 0.062363134342568786,
+      "grad_norm": 11.233921147125818,
+      "learning_rate": 2e-05,
+      "loss": 0.4557,
+      "step": 2620
+    },
+    {
+      "epoch": 0.062386937065600306,
+      "grad_norm": 8.084801967296718,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 2621
+    },
+    {
+      "epoch": 0.06241073978863182,
+      "grad_norm": 6.128847022762278,
+      "learning_rate": 2e-05,
+      "loss": 1.0415,
+      "step": 2622
+    },
+    {
+      "epoch": 0.06243454251166333,
+      "grad_norm": 7.445294102394758,
+      "learning_rate": 2e-05,
+      "loss": 0.7081,
+      "step": 2623
+    },
+    {
+      "epoch": 0.06245834523469485,
+      "grad_norm": 6.599500987742294,
+      "learning_rate": 2e-05,
+      "loss": 0.9161,
+      "step": 2624
+    },
+    {
+      "epoch": 0.062482147957726365,
+      "grad_norm": 2.9891937021813977,
+      "learning_rate": 2e-05,
+      "loss": 0.3309,
+      "step": 2625
+    },
+    {
+      "epoch": 0.06250595068075789,
+      "grad_norm": 2.4304049441288136,
+      "learning_rate": 2e-05,
+      "loss": 0.2027,
+      "step": 2626
+    },
+    {
+      "epoch": 0.0625297534037894,
+      "grad_norm": 4.516892473733533,
+      "learning_rate": 2e-05,
+      "loss": 0.5598,
+      "step": 2627
+    },
+    {
+      "epoch": 0.06255355612682091,
+      "grad_norm": 8.542276390748977,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 2628
+    },
+    {
+      "epoch": 0.06257735884985242,
+      "grad_norm": 3.554402360695512,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 2629
+    },
+    {
+      "epoch": 0.06260116157288394,
+      "grad_norm": 14.665767931255873,
+      "learning_rate": 2e-05,
+      "loss": 0.6562,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06262496429591545,
+      "grad_norm": 9.379204174811562,
+      "learning_rate": 2e-05,
+      "loss": 0.7937,
+      "step": 2631
+    },
+    {
+      "epoch": 0.06264876701894696,
+      "grad_norm": 5.472270360018454,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 2632
+    },
+    {
+      "epoch": 0.06267256974197848,
+      "grad_norm": 2.672466695465241,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 2633
+    },
+    {
+      "epoch": 0.06269637246501,
+      "grad_norm": 3.1322990868337053,
+      "learning_rate": 2e-05,
+      "loss": 0.2412,
+      "step": 2634
+    },
+    {
+      "epoch": 0.06272017518804152,
+      "grad_norm": 4.70580678933396,
+      "learning_rate": 2e-05,
+      "loss": 0.9437,
+      "step": 2635
+    },
+    {
+      "epoch": 0.06274397791107303,
+      "grad_norm": 4.309439193452741,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 2636
+    },
+    {
+      "epoch": 0.06276778063410454,
+      "grad_norm": 4.17912165868669,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 2637
+    },
+    {
+      "epoch": 0.06279158335713605,
+      "grad_norm": 2.7141891379853886,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 2638
+    },
+    {
+      "epoch": 0.06281538608016757,
+      "grad_norm": 5.377922165047463,
+      "learning_rate": 2e-05,
+      "loss": 0.641,
+      "step": 2639
+    },
+    {
+      "epoch": 0.06283918880319908,
+      "grad_norm": 9.515812940751799,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 2640
+    },
+    {
+      "epoch": 0.06286299152623061,
+      "grad_norm": 2.7456981803354026,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 2641
+    },
+    {
+      "epoch": 0.06288679424926212,
+      "grad_norm": 5.481968274297726,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 2642
+    },
+    {
+      "epoch": 0.06291059697229363,
+      "grad_norm": 6.59371166531004,
+      "learning_rate": 2e-05,
+      "loss": 0.851,
+      "step": 2643
+    },
+    {
+      "epoch": 0.06293439969532515,
+      "grad_norm": 2.7015236527313684,
+      "learning_rate": 2e-05,
+      "loss": 0.2585,
+      "step": 2644
+    },
+    {
+      "epoch": 0.06295820241835666,
+      "grad_norm": 6.598087291017619,
+      "learning_rate": 2e-05,
+      "loss": 0.4222,
+      "step": 2645
+    },
+    {
+      "epoch": 0.06298200514138817,
+      "grad_norm": 6.194904108594316,
+      "learning_rate": 2e-05,
+      "loss": 1.2858,
+      "step": 2646
+    },
+    {
+      "epoch": 0.06300580786441969,
+      "grad_norm": 4.908820310079301,
+      "learning_rate": 2e-05,
+      "loss": 0.7093,
+      "step": 2647
+    },
+    {
+      "epoch": 0.0630296105874512,
+      "grad_norm": 9.354655843634795,
+      "learning_rate": 2e-05,
+      "loss": 0.9095,
+      "step": 2648
+    },
+    {
+      "epoch": 0.06305341331048273,
+      "grad_norm": 3.370384786769439,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 2649
+    },
+    {
+      "epoch": 0.06307721603351424,
+      "grad_norm": 8.062364274784768,
+      "learning_rate": 2e-05,
+      "loss": 0.8168,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06310101875654575,
+      "grad_norm": 6.905554594388363,
+      "learning_rate": 2e-05,
+      "loss": 0.2922,
+      "step": 2651
+    },
+    {
+      "epoch": 0.06312482147957726,
+      "grad_norm": 4.542289636137598,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 2652
+    },
+    {
+      "epoch": 0.06314862420260878,
+      "grad_norm": 5.412598300544022,
+      "learning_rate": 2e-05,
+      "loss": 0.6651,
+      "step": 2653
+    },
+    {
+      "epoch": 0.06317242692564029,
+      "grad_norm": 3.4270218329999524,
+      "learning_rate": 2e-05,
+      "loss": 0.1621,
+      "step": 2654
+    },
+    {
+      "epoch": 0.0631962296486718,
+      "grad_norm": 8.274654085189297,
+      "learning_rate": 2e-05,
+      "loss": 1.0584,
+      "step": 2655
+    },
+    {
+      "epoch": 0.06322003237170332,
+      "grad_norm": 9.117333003095023,
+      "learning_rate": 2e-05,
+      "loss": 1.8154,
+      "step": 2656
+    },
+    {
+      "epoch": 0.06324383509473484,
+      "grad_norm": 2.482889200106564,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 2657
+    },
+    {
+      "epoch": 0.06326763781776636,
+      "grad_norm": 5.487507257186583,
+      "learning_rate": 2e-05,
+      "loss": 0.5725,
+      "step": 2658
+    },
+    {
+      "epoch": 0.06329144054079787,
+      "grad_norm": 6.504138464602026,
+      "learning_rate": 2e-05,
+      "loss": 0.3874,
+      "step": 2659
+    },
+    {
+      "epoch": 0.06331524326382938,
+      "grad_norm": 2.6365572896375973,
+      "learning_rate": 2e-05,
+      "loss": 0.1933,
+      "step": 2660
+    },
+    {
+      "epoch": 0.0633390459868609,
+      "grad_norm": 1.7162924493009044,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 2661
+    },
+    {
+      "epoch": 0.06336284870989241,
+      "grad_norm": 4.393290313832797,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 2662
+    },
+    {
+      "epoch": 0.06338665143292392,
+      "grad_norm": 7.8132794609744405,
+      "learning_rate": 2e-05,
+      "loss": 0.9566,
+      "step": 2663
+    },
+    {
+      "epoch": 0.06341045415595545,
+      "grad_norm": 10.450558078901162,
+      "learning_rate": 2e-05,
+      "loss": 0.7149,
+      "step": 2664
+    },
+    {
+      "epoch": 0.06343425687898696,
+      "grad_norm": 10.264104453268422,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 2665
+    },
+    {
+      "epoch": 0.06345805960201847,
+      "grad_norm": 5.778863498149967,
+      "learning_rate": 2e-05,
+      "loss": 0.1751,
+      "step": 2666
+    },
+    {
+      "epoch": 0.06348186232504999,
+      "grad_norm": 7.403988571467779,
+      "learning_rate": 2e-05,
+      "loss": 1.0286,
+      "step": 2667
+    },
+    {
+      "epoch": 0.0635056650480815,
+      "grad_norm": 2.2246430145149403,
+      "learning_rate": 2e-05,
+      "loss": 0.1737,
+      "step": 2668
+    },
+    {
+      "epoch": 0.06352946777111301,
+      "grad_norm": 3.579854299272433,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 2669
+    },
+    {
+      "epoch": 0.06355327049414453,
+      "grad_norm": 10.600316666321143,
+      "learning_rate": 2e-05,
+      "loss": 1.6775,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06357707321717604,
+      "grad_norm": 2.3827660861021793,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 2671
+    },
+    {
+      "epoch": 0.06360087594020757,
+      "grad_norm": 4.640441683099359,
+      "learning_rate": 2e-05,
+      "loss": 0.3075,
+      "step": 2672
+    },
+    {
+      "epoch": 0.06362467866323908,
+      "grad_norm": 6.0998416685781605,
+      "learning_rate": 2e-05,
+      "loss": 1.0612,
+      "step": 2673
+    },
+    {
+      "epoch": 0.06364848138627059,
+      "grad_norm": 3.644677743051855,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 2674
+    },
+    {
+      "epoch": 0.0636722841093021,
+      "grad_norm": 7.55678306088575,
+      "learning_rate": 2e-05,
+      "loss": 0.3307,
+      "step": 2675
+    },
+    {
+      "epoch": 0.06369608683233362,
+      "grad_norm": 9.352898647679798,
+      "learning_rate": 2e-05,
+      "loss": 0.4538,
+      "step": 2676
+    },
+    {
+      "epoch": 0.06371988955536513,
+      "grad_norm": 6.50862612182839,
+      "learning_rate": 2e-05,
+      "loss": 1.0628,
+      "step": 2677
+    },
+    {
+      "epoch": 0.06374369227839664,
+      "grad_norm": 1.6689329533613577,
+      "learning_rate": 2e-05,
+      "loss": 0.1423,
+      "step": 2678
+    },
+    {
+      "epoch": 0.06376749500142816,
+      "grad_norm": 9.537707848119714,
+      "learning_rate": 2e-05,
+      "loss": 0.5476,
+      "step": 2679
+    },
+    {
+      "epoch": 0.06379129772445968,
+      "grad_norm": 6.350227626742197,
+      "learning_rate": 2e-05,
+      "loss": 0.3904,
+      "step": 2680
+    },
+    {
+      "epoch": 0.0638151004474912,
+      "grad_norm": 3.3147650650295364,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 2681
+    },
+    {
+      "epoch": 0.06383890317052271,
+      "grad_norm": 2.779652838496434,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 2682
+    },
+    {
+      "epoch": 0.06386270589355422,
+      "grad_norm": 3.9986825673437645,
+      "learning_rate": 2e-05,
+      "loss": 0.2285,
+      "step": 2683
+    },
+    {
+      "epoch": 0.06388650861658574,
+      "grad_norm": 4.472566848422157,
+      "learning_rate": 2e-05,
+      "loss": 0.6236,
+      "step": 2684
+    },
+    {
+      "epoch": 0.06391031133961725,
+      "grad_norm": 9.04169146094152,
+      "learning_rate": 2e-05,
+      "loss": 0.3764,
+      "step": 2685
+    },
+    {
+      "epoch": 0.06393411406264876,
+      "grad_norm": 3.211038561750457,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 2686
+    },
+    {
+      "epoch": 0.06395791678568029,
+      "grad_norm": 4.013423057485346,
+      "learning_rate": 2e-05,
+      "loss": 0.373,
+      "step": 2687
+    },
+    {
+      "epoch": 0.0639817195087118,
+      "grad_norm": 1.7728027316049215,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 2688
+    },
+    {
+      "epoch": 0.06400552223174331,
+      "grad_norm": 9.610589707199178,
+      "learning_rate": 2e-05,
+      "loss": 1.5357,
+      "step": 2689
+    },
+    {
+      "epoch": 0.06402932495477483,
+      "grad_norm": 8.127355032780219,
+      "learning_rate": 2e-05,
+      "loss": 0.3693,
+      "step": 2690
+    },
+    {
+      "epoch": 0.06405312767780634,
+      "grad_norm": 2.9693737399325815,
+      "learning_rate": 2e-05,
+      "loss": 0.202,
+      "step": 2691
+    },
+    {
+      "epoch": 0.06407693040083785,
+      "grad_norm": 4.580911924194436,
+      "learning_rate": 2e-05,
+      "loss": 0.2709,
+      "step": 2692
+    },
+    {
+      "epoch": 0.06410073312386937,
+      "grad_norm": 7.853409115793132,
+      "learning_rate": 2e-05,
+      "loss": 0.8813,
+      "step": 2693
+    },
+    {
+      "epoch": 0.06412453584690088,
+      "grad_norm": 7.732825544902635,
+      "learning_rate": 2e-05,
+      "loss": 0.8744,
+      "step": 2694
+    },
+    {
+      "epoch": 0.0641483385699324,
+      "grad_norm": 2.7061848082016877,
+      "learning_rate": 2e-05,
+      "loss": 0.2151,
+      "step": 2695
+    },
+    {
+      "epoch": 0.06417214129296392,
+      "grad_norm": 7.380522740807149,
+      "learning_rate": 2e-05,
+      "loss": 0.7347,
+      "step": 2696
+    },
+    {
+      "epoch": 0.06419594401599543,
+      "grad_norm": 7.87803940695446,
+      "learning_rate": 2e-05,
+      "loss": 0.3302,
+      "step": 2697
+    },
+    {
+      "epoch": 0.06421974673902694,
+      "grad_norm": 2.7073117238888145,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 2698
+    },
+    {
+      "epoch": 0.06424354946205846,
+      "grad_norm": 2.1509869238898136,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 2699
+    },
+    {
+      "epoch": 0.06426735218508997,
+      "grad_norm": 9.570201206808996,
+      "learning_rate": 2e-05,
+      "loss": 0.837,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06429115490812148,
+      "grad_norm": 3.4298405765388416,
+      "learning_rate": 2e-05,
+      "loss": 0.1875,
+      "step": 2701
+    },
+    {
+      "epoch": 0.06431495763115301,
+      "grad_norm": 2.6770960010933433,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 2702
+    },
+    {
+      "epoch": 0.06433876035418452,
+      "grad_norm": 3.7394211542995808,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 2703
+    },
+    {
+      "epoch": 0.06436256307721604,
+      "grad_norm": 7.95293004543857,
+      "learning_rate": 2e-05,
+      "loss": 0.358,
+      "step": 2704
+    },
+    {
+      "epoch": 0.06438636580024755,
+      "grad_norm": 7.659640324281743,
+      "learning_rate": 2e-05,
+      "loss": 0.7538,
+      "step": 2705
+    },
+    {
+      "epoch": 0.06441016852327906,
+      "grad_norm": 2.982621457766519,
+      "learning_rate": 2e-05,
+      "loss": 0.1345,
+      "step": 2706
+    },
+    {
+      "epoch": 0.06443397124631058,
+      "grad_norm": 6.217510816391318,
+      "learning_rate": 2e-05,
+      "loss": 0.9774,
+      "step": 2707
+    },
+    {
+      "epoch": 0.06445777396934209,
+      "grad_norm": 2.2093906272893067,
+      "learning_rate": 2e-05,
+      "loss": 0.2216,
+      "step": 2708
+    },
+    {
+      "epoch": 0.0644815766923736,
+      "grad_norm": 6.515557110279227,
+      "learning_rate": 2e-05,
+      "loss": 0.4332,
+      "step": 2709
+    },
+    {
+      "epoch": 0.06450537941540513,
+      "grad_norm": 4.930720149361659,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06452918213843664,
+      "grad_norm": 3.240434847266896,
+      "learning_rate": 2e-05,
+      "loss": 0.3334,
+      "step": 2711
+    },
+    {
+      "epoch": 0.06455298486146815,
+      "grad_norm": 11.976912375947421,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 2712
+    },
+    {
+      "epoch": 0.06457678758449967,
+      "grad_norm": 7.282726003505952,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 2713
+    },
+    {
+      "epoch": 0.06460059030753118,
+      "grad_norm": 7.770762454606823,
+      "learning_rate": 2e-05,
+      "loss": 1.6247,
+      "step": 2714
+    },
+    {
+      "epoch": 0.0646243930305627,
+      "grad_norm": 5.843863585571456,
+      "learning_rate": 2e-05,
+      "loss": 0.6224,
+      "step": 2715
+    },
+    {
+      "epoch": 0.0646481957535942,
+      "grad_norm": 3.168015881651497,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 2716
+    },
+    {
+      "epoch": 0.06467199847662572,
+      "grad_norm": 2.887395866633381,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 2717
+    },
+    {
+      "epoch": 0.06469580119965725,
+      "grad_norm": 7.973906461267167,
+      "learning_rate": 2e-05,
+      "loss": 1.2568,
+      "step": 2718
+    },
+    {
+      "epoch": 0.06471960392268876,
+      "grad_norm": 6.864479229254731,
+      "learning_rate": 2e-05,
+      "loss": 0.3645,
+      "step": 2719
+    },
+    {
+      "epoch": 0.06474340664572027,
+      "grad_norm": 2.8889195898809983,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06476720936875179,
+      "grad_norm": 7.009722736219219,
+      "learning_rate": 2e-05,
+      "loss": 0.39,
+      "step": 2721
+    },
+    {
+      "epoch": 0.0647910120917833,
+      "grad_norm": 6.522030437839087,
+      "learning_rate": 2e-05,
+      "loss": 1.2358,
+      "step": 2722
+    },
+    {
+      "epoch": 0.06481481481481481,
+      "grad_norm": 6.243574084427054,
+      "learning_rate": 2e-05,
+      "loss": 0.8642,
+      "step": 2723
+    },
+    {
+      "epoch": 0.06483861753784632,
+      "grad_norm": 7.301895945743411,
+      "learning_rate": 2e-05,
+      "loss": 0.6776,
+      "step": 2724
+    },
+    {
+      "epoch": 0.06486242026087785,
+      "grad_norm": 12.65354301408637,
+      "learning_rate": 2e-05,
+      "loss": 2.0885,
+      "step": 2725
+    },
+    {
+      "epoch": 0.06488622298390936,
+      "grad_norm": 8.862385702497617,
+      "learning_rate": 2e-05,
+      "loss": 1.0134,
+      "step": 2726
+    },
+    {
+      "epoch": 0.06491002570694088,
+      "grad_norm": 2.8593013349327436,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 2727
+    },
+    {
+      "epoch": 0.06493382842997239,
+      "grad_norm": 2.885992957068783,
+      "learning_rate": 2e-05,
+      "loss": 0.2718,
+      "step": 2728
+    },
+    {
+      "epoch": 0.0649576311530039,
+      "grad_norm": 2.618866136658945,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 2729
+    },
+    {
+      "epoch": 0.06498143387603542,
+      "grad_norm": 4.166876575090319,
+      "learning_rate": 2e-05,
+      "loss": 0.2222,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06500523659906693,
+      "grad_norm": 8.020584422675327,
+      "learning_rate": 2e-05,
+      "loss": 0.4303,
+      "step": 2731
+    },
+    {
+      "epoch": 0.06502903932209844,
+      "grad_norm": 11.639415259169743,
+      "learning_rate": 2e-05,
+      "loss": 0.5784,
+      "step": 2732
+    },
+    {
+      "epoch": 0.06505284204512997,
+      "grad_norm": 1.840142084506002,
+      "learning_rate": 2e-05,
+      "loss": 0.2087,
+      "step": 2733
+    },
+    {
+      "epoch": 0.06507664476816148,
+      "grad_norm": 5.851636599558784,
+      "learning_rate": 2e-05,
+      "loss": 0.9222,
+      "step": 2734
+    },
+    {
+      "epoch": 0.065100447491193,
+      "grad_norm": 5.613776827778494,
+      "learning_rate": 2e-05,
+      "loss": 1.0346,
+      "step": 2735
+    },
+    {
+      "epoch": 0.06512425021422451,
+      "grad_norm": 2.9704027364285785,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 2736
+    },
+    {
+      "epoch": 0.06514805293725602,
+      "grad_norm": 3.150547821580531,
+      "learning_rate": 2e-05,
+      "loss": 0.3042,
+      "step": 2737
+    },
+    {
+      "epoch": 0.06517185566028753,
+      "grad_norm": 3.7479390150562524,
+      "learning_rate": 2e-05,
+      "loss": 0.3677,
+      "step": 2738
+    },
+    {
+      "epoch": 0.06519565838331905,
+      "grad_norm": 5.221487526814124,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 2739
+    },
+    {
+      "epoch": 0.06521946110635056,
+      "grad_norm": 6.291944227078775,
+      "learning_rate": 2e-05,
+      "loss": 0.5384,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06524326382938209,
+      "grad_norm": 5.521234157497432,
+      "learning_rate": 2e-05,
+      "loss": 0.4558,
+      "step": 2741
+    },
+    {
+      "epoch": 0.0652670665524136,
+      "grad_norm": 2.3672918200806015,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 2742
+    },
+    {
+      "epoch": 0.06529086927544511,
+      "grad_norm": 4.756699024602986,
+      "learning_rate": 2e-05,
+      "loss": 0.3229,
+      "step": 2743
+    },
+    {
+      "epoch": 0.06531467199847663,
+      "grad_norm": 3.3571465067091584,
+      "learning_rate": 2e-05,
+      "loss": 0.2327,
+      "step": 2744
+    },
+    {
+      "epoch": 0.06533847472150814,
+      "grad_norm": 5.177503092818778,
+      "learning_rate": 2e-05,
+      "loss": 0.8945,
+      "step": 2745
+    },
+    {
+      "epoch": 0.06536227744453965,
+      "grad_norm": 3.802291930621944,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 2746
+    },
+    {
+      "epoch": 0.06538608016757116,
+      "grad_norm": 14.329364175536966,
+      "learning_rate": 2e-05,
+      "loss": 0.553,
+      "step": 2747
+    },
+    {
+      "epoch": 0.06540988289060269,
+      "grad_norm": 3.5252811912389213,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 2748
+    },
+    {
+      "epoch": 0.0654336856136342,
+      "grad_norm": 3.1892952023484704,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 2749
+    },
+    {
+      "epoch": 0.06545748833666572,
+      "grad_norm": 4.780525604609568,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06548129105969723,
+      "grad_norm": 3.365751831130399,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 2751
+    },
+    {
+      "epoch": 0.06550509378272874,
+      "grad_norm": 3.612599093766394,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 2752
+    },
+    {
+      "epoch": 0.06552889650576026,
+      "grad_norm": 6.122860256964392,
+      "learning_rate": 2e-05,
+      "loss": 0.6987,
+      "step": 2753
+    },
+    {
+      "epoch": 0.06555269922879177,
+      "grad_norm": 5.888782114363467,
+      "learning_rate": 2e-05,
+      "loss": 0.87,
+      "step": 2754
+    },
+    {
+      "epoch": 0.06557650195182328,
+      "grad_norm": 4.639713275129618,
+      "learning_rate": 2e-05,
+      "loss": 0.3186,
+      "step": 2755
+    },
+    {
+      "epoch": 0.06560030467485481,
+      "grad_norm": 2.6812712301104655,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 2756
+    },
+    {
+      "epoch": 0.06562410739788632,
+      "grad_norm": 7.076782171559918,
+      "learning_rate": 2e-05,
+      "loss": 0.4533,
+      "step": 2757
+    },
+    {
+      "epoch": 0.06564791012091784,
+      "grad_norm": 4.105986928089539,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 2758
+    },
+    {
+      "epoch": 0.06567171284394935,
+      "grad_norm": 9.013616651997232,
+      "learning_rate": 2e-05,
+      "loss": 1.6338,
+      "step": 2759
+    },
+    {
+      "epoch": 0.06569551556698086,
+      "grad_norm": 2.9102306523396853,
+      "learning_rate": 2e-05,
+      "loss": 0.2069,
+      "step": 2760
+    },
+    {
+      "epoch": 0.06571931829001237,
+      "grad_norm": 3.078069802192271,
+      "learning_rate": 2e-05,
+      "loss": 0.2616,
+      "step": 2761
+    },
+    {
+      "epoch": 0.06574312101304389,
+      "grad_norm": 4.390722150035201,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 2762
+    },
+    {
+      "epoch": 0.06576692373607541,
+      "grad_norm": 7.960589522995565,
+      "learning_rate": 2e-05,
+      "loss": 0.3269,
+      "step": 2763
+    },
+    {
+      "epoch": 0.06579072645910693,
+      "grad_norm": 4.947927606731157,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 2764
+    },
+    {
+      "epoch": 0.06581452918213844,
+      "grad_norm": 7.380533472702038,
+      "learning_rate": 2e-05,
+      "loss": 1.4248,
+      "step": 2765
+    },
+    {
+      "epoch": 0.06583833190516995,
+      "grad_norm": 2.2815952421495114,
+      "learning_rate": 2e-05,
+      "loss": 0.1999,
+      "step": 2766
+    },
+    {
+      "epoch": 0.06586213462820147,
+      "grad_norm": 3.539131112810029,
+      "learning_rate": 2e-05,
+      "loss": 0.4819,
+      "step": 2767
+    },
+    {
+      "epoch": 0.06588593735123298,
+      "grad_norm": 2.2441585938123745,
+      "learning_rate": 2e-05,
+      "loss": 0.1981,
+      "step": 2768
+    },
+    {
+      "epoch": 0.06590974007426449,
+      "grad_norm": 6.100968539152162,
+      "learning_rate": 2e-05,
+      "loss": 1.0889,
+      "step": 2769
+    },
+    {
+      "epoch": 0.065933542797296,
+      "grad_norm": 2.5322209261560182,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 2770
+    },
+    {
+      "epoch": 0.06595734552032753,
+      "grad_norm": 5.926630040895769,
+      "learning_rate": 2e-05,
+      "loss": 0.7054,
+      "step": 2771
+    },
+    {
+      "epoch": 0.06598114824335904,
+      "grad_norm": 3.3433468928860215,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 2772
+    },
+    {
+      "epoch": 0.06600495096639056,
+      "grad_norm": 4.8207161398334515,
+      "learning_rate": 2e-05,
+      "loss": 0.7041,
+      "step": 2773
+    },
+    {
+      "epoch": 0.06602875368942207,
+      "grad_norm": 10.220398200240751,
+      "learning_rate": 2e-05,
+      "loss": 0.5911,
+      "step": 2774
+    },
+    {
+      "epoch": 0.06605255641245358,
+      "grad_norm": 4.760771149063289,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 2775
+    },
+    {
+      "epoch": 0.0660763591354851,
+      "grad_norm": 6.27380120322418,
+      "learning_rate": 2e-05,
+      "loss": 0.2938,
+      "step": 2776
+    },
+    {
+      "epoch": 0.06610016185851661,
+      "grad_norm": 2.734145665771031,
+      "learning_rate": 2e-05,
+      "loss": 0.1773,
+      "step": 2777
+    },
+    {
+      "epoch": 0.06612396458154812,
+      "grad_norm": 8.641559735218497,
+      "learning_rate": 2e-05,
+      "loss": 1.1331,
+      "step": 2778
+    },
+    {
+      "epoch": 0.06614776730457965,
+      "grad_norm": 3.333147158683592,
+      "learning_rate": 2e-05,
+      "loss": 0.2642,
+      "step": 2779
+    },
+    {
+      "epoch": 0.06617157002761116,
+      "grad_norm": 3.889254201240531,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06619537275064268,
+      "grad_norm": 7.24078247821459,
+      "learning_rate": 2e-05,
+      "loss": 1.5236,
+      "step": 2781
+    },
+    {
+      "epoch": 0.06621917547367419,
+      "grad_norm": 6.424590984139892,
+      "learning_rate": 2e-05,
+      "loss": 1.2454,
+      "step": 2782
+    },
+    {
+      "epoch": 0.0662429781967057,
+      "grad_norm": 2.2521613753038103,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 2783
+    },
+    {
+      "epoch": 0.06626678091973721,
+      "grad_norm": 8.274658305136636,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 2784
+    },
+    {
+      "epoch": 0.06629058364276873,
+      "grad_norm": 7.4619932261054185,
+      "learning_rate": 2e-05,
+      "loss": 0.4692,
+      "step": 2785
+    },
+    {
+      "epoch": 0.06631438636580025,
+      "grad_norm": 9.256643824113844,
+      "learning_rate": 2e-05,
+      "loss": 0.5191,
+      "step": 2786
+    },
+    {
+      "epoch": 0.06633818908883177,
+      "grad_norm": 11.091213301421917,
+      "learning_rate": 2e-05,
+      "loss": 0.3688,
+      "step": 2787
+    },
+    {
+      "epoch": 0.06636199181186328,
+      "grad_norm": 2.204459916215135,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 2788
+    },
+    {
+      "epoch": 0.0663857945348948,
+      "grad_norm": 8.551308666278938,
+      "learning_rate": 2e-05,
+      "loss": 1.8128,
+      "step": 2789
+    },
+    {
+      "epoch": 0.0664095972579263,
+      "grad_norm": 6.9724712253578565,
+      "learning_rate": 2e-05,
+      "loss": 0.3888,
+      "step": 2790
+    },
+    {
+      "epoch": 0.06643339998095782,
+      "grad_norm": 3.52076188660014,
+      "learning_rate": 2e-05,
+      "loss": 0.1428,
+      "step": 2791
+    },
+    {
+      "epoch": 0.06645720270398933,
+      "grad_norm": 4.018383090925838,
+      "learning_rate": 2e-05,
+      "loss": 0.2974,
+      "step": 2792
+    },
+    {
+      "epoch": 0.06648100542702085,
+      "grad_norm": 8.694920256680325,
+      "learning_rate": 2e-05,
+      "loss": 0.8644,
+      "step": 2793
+    },
+    {
+      "epoch": 0.06650480815005237,
+      "grad_norm": 5.2164383680925726,
+      "learning_rate": 2e-05,
+      "loss": 0.366,
+      "step": 2794
+    },
+    {
+      "epoch": 0.06652861087308388,
+      "grad_norm": 3.61367230331091,
+      "learning_rate": 2e-05,
+      "loss": 0.2973,
+      "step": 2795
+    },
+    {
+      "epoch": 0.0665524135961154,
+      "grad_norm": 3.5805992619258964,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 2796
+    },
+    {
+      "epoch": 0.06657621631914691,
+      "grad_norm": 6.927371417687817,
+      "learning_rate": 2e-05,
+      "loss": 1.0802,
+      "step": 2797
+    },
+    {
+      "epoch": 0.06660001904217842,
+      "grad_norm": 7.080559516536848,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 2798
+    },
+    {
+      "epoch": 0.06662382176520994,
+      "grad_norm": 10.444700523415927,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 2799
+    },
+    {
+      "epoch": 0.06664762448824145,
+      "grad_norm": 8.008634168566884,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06667142721127296,
+      "grad_norm": 4.050656373388413,
+      "learning_rate": 2e-05,
+      "loss": 0.1649,
+      "step": 2801
+    },
+    {
+      "epoch": 0.06669522993430449,
+      "grad_norm": 8.988209015016519,
+      "learning_rate": 2e-05,
+      "loss": 0.9621,
+      "step": 2802
+    },
+    {
+      "epoch": 0.066719032657336,
+      "grad_norm": 7.312591810764617,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 2803
+    },
+    {
+      "epoch": 0.06674283538036752,
+      "grad_norm": 9.858154345565382,
+      "learning_rate": 2e-05,
+      "loss": 1.195,
+      "step": 2804
+    },
+    {
+      "epoch": 0.06676663810339903,
+      "grad_norm": 5.6756770969109684,
+      "learning_rate": 2e-05,
+      "loss": 0.3455,
+      "step": 2805
+    },
+    {
+      "epoch": 0.06679044082643054,
+      "grad_norm": 5.5128406623660124,
+      "learning_rate": 2e-05,
+      "loss": 0.7146,
+      "step": 2806
+    },
+    {
+      "epoch": 0.06681424354946205,
+      "grad_norm": 6.203326546176729,
+      "learning_rate": 2e-05,
+      "loss": 1.3633,
+      "step": 2807
+    },
+    {
+      "epoch": 0.06683804627249357,
+      "grad_norm": 6.272370722582784,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 2808
+    },
+    {
+      "epoch": 0.0668618489955251,
+      "grad_norm": 7.672578732569932,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 2809
+    },
+    {
+      "epoch": 0.06688565171855661,
+      "grad_norm": 4.271807080888943,
+      "learning_rate": 2e-05,
+      "loss": 0.3481,
+      "step": 2810
+    },
+    {
+      "epoch": 0.06690945444158812,
+      "grad_norm": 2.137848797809003,
+      "learning_rate": 2e-05,
+      "loss": 0.2171,
+      "step": 2811
+    },
+    {
+      "epoch": 0.06693325716461963,
+      "grad_norm": 8.368092095756017,
+      "learning_rate": 2e-05,
+      "loss": 1.0099,
+      "step": 2812
+    },
+    {
+      "epoch": 0.06695705988765115,
+      "grad_norm": 1.921089209071309,
+      "learning_rate": 2e-05,
+      "loss": 0.1499,
+      "step": 2813
+    },
+    {
+      "epoch": 0.06698086261068266,
+      "grad_norm": 8.9978141171459,
+      "learning_rate": 2e-05,
+      "loss": 0.8747,
+      "step": 2814
+    },
+    {
+      "epoch": 0.06700466533371417,
+      "grad_norm": 9.013491860955128,
+      "learning_rate": 2e-05,
+      "loss": 0.4405,
+      "step": 2815
+    },
+    {
+      "epoch": 0.06702846805674569,
+      "grad_norm": 3.886981563128866,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 2816
+    },
+    {
+      "epoch": 0.06705227077977721,
+      "grad_norm": 9.125249062949834,
+      "learning_rate": 2e-05,
+      "loss": 0.457,
+      "step": 2817
+    },
+    {
+      "epoch": 0.06707607350280873,
+      "grad_norm": 4.177701933286063,
+      "learning_rate": 2e-05,
+      "loss": 0.2842,
+      "step": 2818
+    },
+    {
+      "epoch": 0.06709987622584024,
+      "grad_norm": 2.9900738244250005,
+      "learning_rate": 2e-05,
+      "loss": 0.251,
+      "step": 2819
+    },
+    {
+      "epoch": 0.06712367894887175,
+      "grad_norm": 2.71162532473282,
+      "learning_rate": 2e-05,
+      "loss": 0.2083,
+      "step": 2820
+    },
+    {
+      "epoch": 0.06714748167190326,
+      "grad_norm": 5.224422561504363,
+      "learning_rate": 2e-05,
+      "loss": 0.96,
+      "step": 2821
+    },
+    {
+      "epoch": 0.06717128439493478,
+      "grad_norm": 8.221415509452783,
+      "learning_rate": 2e-05,
+      "loss": 0.9178,
+      "step": 2822
+    },
+    {
+      "epoch": 0.06719508711796629,
+      "grad_norm": 6.041100499074039,
+      "learning_rate": 2e-05,
+      "loss": 0.7899,
+      "step": 2823
+    },
+    {
+      "epoch": 0.06721888984099782,
+      "grad_norm": 2.10562937246565,
+      "learning_rate": 2e-05,
+      "loss": 0.1612,
+      "step": 2824
+    },
+    {
+      "epoch": 0.06724269256402933,
+      "grad_norm": 3.1841863605247487,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 2825
+    },
+    {
+      "epoch": 0.06726649528706084,
+      "grad_norm": 8.20923908288878,
+      "learning_rate": 2e-05,
+      "loss": 0.8998,
+      "step": 2826
+    },
+    {
+      "epoch": 0.06729029801009236,
+      "grad_norm": 2.755836719777192,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 2827
+    },
+    {
+      "epoch": 0.06731410073312387,
+      "grad_norm": 2.4156522050401366,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 2828
+    },
+    {
+      "epoch": 0.06733790345615538,
+      "grad_norm": 4.537909162217931,
+      "learning_rate": 2e-05,
+      "loss": 0.3511,
+      "step": 2829
+    },
+    {
+      "epoch": 0.0673617061791869,
+      "grad_norm": 7.275512300203897,
+      "learning_rate": 2e-05,
+      "loss": 0.7608,
+      "step": 2830
+    },
+    {
+      "epoch": 0.06738550890221841,
+      "grad_norm": 2.9347587420977987,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 2831
+    },
+    {
+      "epoch": 0.06740931162524993,
+      "grad_norm": 1.93052651049275,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 2832
+    },
+    {
+      "epoch": 0.06743311434828145,
+      "grad_norm": 5.067232407463607,
+      "learning_rate": 2e-05,
+      "loss": 0.5875,
+      "step": 2833
+    },
+    {
+      "epoch": 0.06745691707131296,
+      "grad_norm": 5.3443679973555405,
+      "learning_rate": 2e-05,
+      "loss": 0.9725,
+      "step": 2834
+    },
+    {
+      "epoch": 0.06748071979434447,
+      "grad_norm": 7.086743212841805,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 2835
+    },
+    {
+      "epoch": 0.06750452251737599,
+      "grad_norm": 5.9816606897778035,
+      "learning_rate": 2e-05,
+      "loss": 0.2546,
+      "step": 2836
+    },
+    {
+      "epoch": 0.0675283252404075,
+      "grad_norm": 7.000410515163164,
+      "learning_rate": 2e-05,
+      "loss": 0.4276,
+      "step": 2837
+    },
+    {
+      "epoch": 0.06755212796343901,
+      "grad_norm": 7.06703952329802,
+      "learning_rate": 2e-05,
+      "loss": 0.4258,
+      "step": 2838
+    },
+    {
+      "epoch": 0.06757593068647053,
+      "grad_norm": 7.807517146280112,
+      "learning_rate": 2e-05,
+      "loss": 1.3307,
+      "step": 2839
+    },
+    {
+      "epoch": 0.06759973340950205,
+      "grad_norm": 9.714893341623764,
+      "learning_rate": 2e-05,
+      "loss": 0.3694,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06762353613253357,
+      "grad_norm": 10.587580791297848,
+      "learning_rate": 2e-05,
+      "loss": 0.9208,
+      "step": 2841
+    },
+    {
+      "epoch": 0.06764733885556508,
+      "grad_norm": 6.211634615731635,
+      "learning_rate": 2e-05,
+      "loss": 0.6339,
+      "step": 2842
+    },
+    {
+      "epoch": 0.06767114157859659,
+      "grad_norm": 6.342478183563558,
+      "learning_rate": 2e-05,
+      "loss": 0.3211,
+      "step": 2843
+    },
+    {
+      "epoch": 0.0676949443016281,
+      "grad_norm": 5.549623098678855,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 2844
+    },
+    {
+      "epoch": 0.06771874702465962,
+      "grad_norm": 10.71796756529859,
+      "learning_rate": 2e-05,
+      "loss": 0.4324,
+      "step": 2845
+    },
+    {
+      "epoch": 0.06774254974769113,
+      "grad_norm": 2.9075312665368993,
+      "learning_rate": 2e-05,
+      "loss": 0.216,
+      "step": 2846
+    },
+    {
+      "epoch": 0.06776635247072266,
+      "grad_norm": 6.7420613228953865,
+      "learning_rate": 2e-05,
+      "loss": 0.3737,
+      "step": 2847
+    },
+    {
+      "epoch": 0.06779015519375417,
+      "grad_norm": 3.3547409166050457,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 2848
+    },
+    {
+      "epoch": 0.06781395791678568,
+      "grad_norm": 3.8659778517569974,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 2849
+    },
+    {
+      "epoch": 0.0678377606398172,
+      "grad_norm": 6.084206597536989,
+      "learning_rate": 2e-05,
+      "loss": 0.8897,
+      "step": 2850
+    },
+    {
+      "epoch": 0.06786156336284871,
+      "grad_norm": 2.8636867387632785,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 2851
+    },
+    {
+      "epoch": 0.06788536608588022,
+      "grad_norm": 1.932466091553941,
+      "learning_rate": 2e-05,
+      "loss": 0.1493,
+      "step": 2852
+    },
+    {
+      "epoch": 0.06790916880891174,
+      "grad_norm": 3.6301953729167273,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 2853
+    },
+    {
+      "epoch": 0.06793297153194325,
+      "grad_norm": 10.49171281629369,
+      "learning_rate": 2e-05,
+      "loss": 0.5268,
+      "step": 2854
+    },
+    {
+      "epoch": 0.06795677425497477,
+      "grad_norm": 5.956715354332126,
+      "learning_rate": 2e-05,
+      "loss": 0.4088,
+      "step": 2855
+    },
+    {
+      "epoch": 0.06798057697800629,
+      "grad_norm": 5.2042697107073375,
+      "learning_rate": 2e-05,
+      "loss": 0.3696,
+      "step": 2856
+    },
+    {
+      "epoch": 0.0680043797010378,
+      "grad_norm": 2.136415206839742,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 2857
+    },
+    {
+      "epoch": 0.06802818242406931,
+      "grad_norm": 2.974435160720859,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 2858
+    },
+    {
+      "epoch": 0.06805198514710083,
+      "grad_norm": 7.763316334593571,
+      "learning_rate": 2e-05,
+      "loss": 0.4974,
+      "step": 2859
+    },
+    {
+      "epoch": 0.06807578787013234,
+      "grad_norm": 3.396241040989403,
+      "learning_rate": 2e-05,
+      "loss": 0.379,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06809959059316385,
+      "grad_norm": 2.9181856843316476,
+      "learning_rate": 2e-05,
+      "loss": 0.2666,
+      "step": 2861
+    },
+    {
+      "epoch": 0.06812339331619537,
+      "grad_norm": 2.9057329435312362,
+      "learning_rate": 2e-05,
+      "loss": 0.1972,
+      "step": 2862
+    },
+    {
+      "epoch": 0.06814719603922689,
+      "grad_norm": 6.94999724259833,
+      "learning_rate": 2e-05,
+      "loss": 0.8924,
+      "step": 2863
+    },
+    {
+      "epoch": 0.0681709987622584,
+      "grad_norm": 9.21607823233594,
+      "learning_rate": 2e-05,
+      "loss": 0.5479,
+      "step": 2864
+    },
+    {
+      "epoch": 0.06819480148528992,
+      "grad_norm": 2.1170701686467854,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 2865
+    },
+    {
+      "epoch": 0.06821860420832143,
+      "grad_norm": 2.863798132157027,
+      "learning_rate": 2e-05,
+      "loss": 0.2091,
+      "step": 2866
+    },
+    {
+      "epoch": 0.06824240693135294,
+      "grad_norm": 8.261379812063337,
+      "learning_rate": 2e-05,
+      "loss": 1.2847,
+      "step": 2867
+    },
+    {
+      "epoch": 0.06826620965438446,
+      "grad_norm": 7.4145978863828885,
+      "learning_rate": 2e-05,
+      "loss": 0.42,
+      "step": 2868
+    },
+    {
+      "epoch": 0.06829001237741597,
+      "grad_norm": 4.111194217773502,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 2869
+    },
+    {
+      "epoch": 0.0683138151004475,
+      "grad_norm": 7.208458510358523,
+      "learning_rate": 2e-05,
+      "loss": 0.8569,
+      "step": 2870
+    },
+    {
+      "epoch": 0.06833761782347901,
+      "grad_norm": 8.072135581124074,
+      "learning_rate": 2e-05,
+      "loss": 1.2329,
+      "step": 2871
+    },
+    {
+      "epoch": 0.06836142054651052,
+      "grad_norm": 5.262491050349569,
+      "learning_rate": 2e-05,
+      "loss": 0.3008,
+      "step": 2872
+    },
+    {
+      "epoch": 0.06838522326954204,
+      "grad_norm": 2.527226175088049,
+      "learning_rate": 2e-05,
+      "loss": 0.1581,
+      "step": 2873
+    },
+    {
+      "epoch": 0.06840902599257355,
+      "grad_norm": 5.870770796703725,
+      "learning_rate": 2e-05,
+      "loss": 0.8313,
+      "step": 2874
+    },
+    {
+      "epoch": 0.06843282871560506,
+      "grad_norm": 2.678787050867996,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 2875
+    },
+    {
+      "epoch": 0.06845663143863658,
+      "grad_norm": 5.476834707415568,
+      "learning_rate": 2e-05,
+      "loss": 0.2816,
+      "step": 2876
+    },
+    {
+      "epoch": 0.06848043416166809,
+      "grad_norm": 1.6242896468638588,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 2877
+    },
+    {
+      "epoch": 0.06850423688469962,
+      "grad_norm": 2.4282495023512602,
+      "learning_rate": 2e-05,
+      "loss": 0.1789,
+      "step": 2878
+    },
+    {
+      "epoch": 0.06852803960773113,
+      "grad_norm": 9.352981806201564,
+      "learning_rate": 2e-05,
+      "loss": 0.4284,
+      "step": 2879
+    },
+    {
+      "epoch": 0.06855184233076264,
+      "grad_norm": 6.960483505310442,
+      "learning_rate": 2e-05,
+      "loss": 0.4343,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06857564505379415,
+      "grad_norm": 4.596542495752711,
+      "learning_rate": 2e-05,
+      "loss": 0.4323,
+      "step": 2881
+    },
+    {
+      "epoch": 0.06859944777682567,
+      "grad_norm": 3.239571322988044,
+      "learning_rate": 2e-05,
+      "loss": 0.2895,
+      "step": 2882
+    },
+    {
+      "epoch": 0.06862325049985718,
+      "grad_norm": 5.193399385372062,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 2883
+    },
+    {
+      "epoch": 0.0686470532228887,
+      "grad_norm": 2.757607408299293,
+      "learning_rate": 2e-05,
+      "loss": 0.2277,
+      "step": 2884
+    },
+    {
+      "epoch": 0.06867085594592022,
+      "grad_norm": 3.5787903410343254,
+      "learning_rate": 2e-05,
+      "loss": 0.2266,
+      "step": 2885
+    },
+    {
+      "epoch": 0.06869465866895173,
+      "grad_norm": 3.160766711383026,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 2886
+    },
+    {
+      "epoch": 0.06871846139198325,
+      "grad_norm": 5.388010193484513,
+      "learning_rate": 2e-05,
+      "loss": 0.7579,
+      "step": 2887
+    },
+    {
+      "epoch": 0.06874226411501476,
+      "grad_norm": 2.557792541225386,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 2888
+    },
+    {
+      "epoch": 0.06876606683804627,
+      "grad_norm": 4.442844953287355,
+      "learning_rate": 2e-05,
+      "loss": 0.6186,
+      "step": 2889
+    },
+    {
+      "epoch": 0.06878986956107778,
+      "grad_norm": 7.8820549443447385,
+      "learning_rate": 2e-05,
+      "loss": 1.2555,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0688136722841093,
+      "grad_norm": 4.45014975301111,
+      "learning_rate": 2e-05,
+      "loss": 0.3124,
+      "step": 2891
+    },
+    {
+      "epoch": 0.06883747500714081,
+      "grad_norm": 7.574951044133139,
+      "learning_rate": 2e-05,
+      "loss": 0.7388,
+      "step": 2892
+    },
+    {
+      "epoch": 0.06886127773017234,
+      "grad_norm": 7.320916621796745,
+      "learning_rate": 2e-05,
+      "loss": 0.7152,
+      "step": 2893
+    },
+    {
+      "epoch": 0.06888508045320385,
+      "grad_norm": 6.291970903097959,
+      "learning_rate": 2e-05,
+      "loss": 0.5234,
+      "step": 2894
+    },
+    {
+      "epoch": 0.06890888317623536,
+      "grad_norm": 2.3109783732597955,
+      "learning_rate": 2e-05,
+      "loss": 0.1621,
+      "step": 2895
+    },
+    {
+      "epoch": 0.06893268589926688,
+      "grad_norm": 8.597681196054463,
+      "learning_rate": 2e-05,
+      "loss": 0.6928,
+      "step": 2896
+    },
+    {
+      "epoch": 0.06895648862229839,
+      "grad_norm": 9.379734736804977,
+      "learning_rate": 2e-05,
+      "loss": 0.503,
+      "step": 2897
+    },
+    {
+      "epoch": 0.0689802913453299,
+      "grad_norm": 3.8728447064489724,
+      "learning_rate": 2e-05,
+      "loss": 0.3322,
+      "step": 2898
+    },
+    {
+      "epoch": 0.06900409406836142,
+      "grad_norm": 3.042705044343069,
+      "learning_rate": 2e-05,
+      "loss": 0.1578,
+      "step": 2899
+    },
+    {
+      "epoch": 0.06902789679139293,
+      "grad_norm": 7.696993482176299,
+      "learning_rate": 2e-05,
+      "loss": 1.2047,
+      "step": 2900
+    },
+    {
+      "epoch": 0.06905169951442446,
+      "grad_norm": 2.5362726654006855,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 2901
+    },
+    {
+      "epoch": 0.06907550223745597,
+      "grad_norm": 8.41252342995201,
+      "learning_rate": 2e-05,
+      "loss": 1.0156,
+      "step": 2902
+    },
+    {
+      "epoch": 0.06909930496048748,
+      "grad_norm": 6.994635351462541,
+      "learning_rate": 2e-05,
+      "loss": 0.7576,
+      "step": 2903
+    },
+    {
+      "epoch": 0.069123107683519,
+      "grad_norm": 2.6497347006816687,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 2904
+    },
+    {
+      "epoch": 0.06914691040655051,
+      "grad_norm": 7.234349103220536,
+      "learning_rate": 2e-05,
+      "loss": 0.7868,
+      "step": 2905
+    },
+    {
+      "epoch": 0.06917071312958202,
+      "grad_norm": 6.750469970295699,
+      "learning_rate": 2e-05,
+      "loss": 0.437,
+      "step": 2906
+    },
+    {
+      "epoch": 0.06919451585261353,
+      "grad_norm": 4.733646076650274,
+      "learning_rate": 2e-05,
+      "loss": 0.4244,
+      "step": 2907
+    },
+    {
+      "epoch": 0.06921831857564506,
+      "grad_norm": 9.647113440209047,
+      "learning_rate": 2e-05,
+      "loss": 1.0789,
+      "step": 2908
+    },
+    {
+      "epoch": 0.06924212129867657,
+      "grad_norm": 2.3150290282162334,
+      "learning_rate": 2e-05,
+      "loss": 0.2037,
+      "step": 2909
+    },
+    {
+      "epoch": 0.06926592402170809,
+      "grad_norm": 6.480485648227689,
+      "learning_rate": 2e-05,
+      "loss": 0.9903,
+      "step": 2910
+    },
+    {
+      "epoch": 0.0692897267447396,
+      "grad_norm": 4.1155232860632385,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 2911
+    },
+    {
+      "epoch": 0.06931352946777111,
+      "grad_norm": 6.976729834785182,
+      "learning_rate": 2e-05,
+      "loss": 0.2824,
+      "step": 2912
+    },
+    {
+      "epoch": 0.06933733219080263,
+      "grad_norm": 5.253582954360467,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 2913
+    },
+    {
+      "epoch": 0.06936113491383414,
+      "grad_norm": 3.1289280587883357,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 2914
+    },
+    {
+      "epoch": 0.06938493763686565,
+      "grad_norm": 3.8367732280124143,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 2915
+    },
+    {
+      "epoch": 0.06940874035989718,
+      "grad_norm": 6.488302656909624,
+      "learning_rate": 2e-05,
+      "loss": 0.3485,
+      "step": 2916
+    },
+    {
+      "epoch": 0.06943254308292869,
+      "grad_norm": 13.41746156031246,
+      "learning_rate": 2e-05,
+      "loss": 0.5687,
+      "step": 2917
+    },
+    {
+      "epoch": 0.0694563458059602,
+      "grad_norm": 8.271632726695325,
+      "learning_rate": 2e-05,
+      "loss": 0.3134,
+      "step": 2918
+    },
+    {
+      "epoch": 0.06948014852899172,
+      "grad_norm": 7.469269363197203,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 2919
+    },
+    {
+      "epoch": 0.06950395125202323,
+      "grad_norm": 4.580027750750358,
+      "learning_rate": 2e-05,
+      "loss": 0.2892,
+      "step": 2920
+    },
+    {
+      "epoch": 0.06952775397505474,
+      "grad_norm": 7.201051962007636,
+      "learning_rate": 2e-05,
+      "loss": 1.2213,
+      "step": 2921
+    },
+    {
+      "epoch": 0.06955155669808626,
+      "grad_norm": 6.543597508272208,
+      "learning_rate": 2e-05,
+      "loss": 0.8679,
+      "step": 2922
+    },
+    {
+      "epoch": 0.06957535942111777,
+      "grad_norm": 8.659914990807643,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 2923
+    },
+    {
+      "epoch": 0.0695991621441493,
+      "grad_norm": 5.495076693055479,
+      "learning_rate": 2e-05,
+      "loss": 1.0174,
+      "step": 2924
+    },
+    {
+      "epoch": 0.06962296486718081,
+      "grad_norm": 6.84503318114582,
+      "learning_rate": 2e-05,
+      "loss": 0.8146,
+      "step": 2925
+    },
+    {
+      "epoch": 0.06964676759021232,
+      "grad_norm": 4.041116488756715,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 2926
+    },
+    {
+      "epoch": 0.06967057031324383,
+      "grad_norm": 7.237616519761568,
+      "learning_rate": 2e-05,
+      "loss": 0.8608,
+      "step": 2927
+    },
+    {
+      "epoch": 0.06969437303627535,
+      "grad_norm": 1.4528577523846566,
+      "learning_rate": 2e-05,
+      "loss": 0.1956,
+      "step": 2928
+    },
+    {
+      "epoch": 0.06971817575930686,
+      "grad_norm": 8.037527530202265,
+      "learning_rate": 2e-05,
+      "loss": 0.9044,
+      "step": 2929
+    },
+    {
+      "epoch": 0.06974197848233837,
+      "grad_norm": 6.9154906690914935,
+      "learning_rate": 2e-05,
+      "loss": 0.8222,
+      "step": 2930
+    },
+    {
+      "epoch": 0.0697657812053699,
+      "grad_norm": 9.805171081302273,
+      "learning_rate": 2e-05,
+      "loss": 0.2522,
+      "step": 2931
+    },
+    {
+      "epoch": 0.06978958392840141,
+      "grad_norm": 2.465354289758053,
+      "learning_rate": 2e-05,
+      "loss": 0.2211,
+      "step": 2932
+    },
+    {
+      "epoch": 0.06981338665143293,
+      "grad_norm": 5.789546052092581,
+      "learning_rate": 2e-05,
+      "loss": 0.3284,
+      "step": 2933
+    },
+    {
+      "epoch": 0.06983718937446444,
+      "grad_norm": 4.531097726551988,
+      "learning_rate": 2e-05,
+      "loss": 0.6238,
+      "step": 2934
+    },
+    {
+      "epoch": 0.06986099209749595,
+      "grad_norm": 27.524500822593485,
+      "learning_rate": 2e-05,
+      "loss": 1.0614,
+      "step": 2935
+    },
+    {
+      "epoch": 0.06988479482052747,
+      "grad_norm": 2.528739988984657,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 2936
+    },
+    {
+      "epoch": 0.06990859754355898,
+      "grad_norm": 1.7753577247635042,
+      "learning_rate": 2e-05,
+      "loss": 0.2001,
+      "step": 2937
+    },
+    {
+      "epoch": 0.06993240026659049,
+      "grad_norm": 5.032853458073444,
+      "learning_rate": 2e-05,
+      "loss": 0.3152,
+      "step": 2938
+    },
+    {
+      "epoch": 0.06995620298962202,
+      "grad_norm": 8.886569741919638,
+      "learning_rate": 2e-05,
+      "loss": 1.1165,
+      "step": 2939
+    },
+    {
+      "epoch": 0.06998000571265353,
+      "grad_norm": 9.488743221710099,
+      "learning_rate": 2e-05,
+      "loss": 1.0408,
+      "step": 2940
+    },
+    {
+      "epoch": 0.07000380843568504,
+      "grad_norm": 10.891892711154673,
+      "learning_rate": 2e-05,
+      "loss": 0.5009,
+      "step": 2941
+    },
+    {
+      "epoch": 0.07002761115871656,
+      "grad_norm": 5.779453061330272,
+      "learning_rate": 2e-05,
+      "loss": 0.6479,
+      "step": 2942
+    },
+    {
+      "epoch": 0.07005141388174807,
+      "grad_norm": 10.191256085650538,
+      "learning_rate": 2e-05,
+      "loss": 2.1059,
+      "step": 2943
+    },
+    {
+      "epoch": 0.07007521660477958,
+      "grad_norm": 2.522145356298826,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 2944
+    },
+    {
+      "epoch": 0.0700990193278111,
+      "grad_norm": 2.6778743240456957,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 2945
+    },
+    {
+      "epoch": 0.07012282205084262,
+      "grad_norm": 6.408703763484512,
+      "learning_rate": 2e-05,
+      "loss": 0.7914,
+      "step": 2946
+    },
+    {
+      "epoch": 0.07014662477387414,
+      "grad_norm": 4.4205096610593335,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 2947
+    },
+    {
+      "epoch": 0.07017042749690565,
+      "grad_norm": 1.252295235904229,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 2948
+    },
+    {
+      "epoch": 0.07019423021993716,
+      "grad_norm": 6.139464386434331,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 2949
+    },
+    {
+      "epoch": 0.07021803294296868,
+      "grad_norm": 3.0303567177965225,
+      "learning_rate": 2e-05,
+      "loss": 0.4788,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07024183566600019,
+      "grad_norm": 6.937466429151537,
+      "learning_rate": 2e-05,
+      "loss": 0.9685,
+      "step": 2951
+    },
+    {
+      "epoch": 0.0702656383890317,
+      "grad_norm": 5.939839623775525,
+      "learning_rate": 2e-05,
+      "loss": 0.6747,
+      "step": 2952
+    },
+    {
+      "epoch": 0.07028944111206321,
+      "grad_norm": 3.6361566602196578,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 2953
+    },
+    {
+      "epoch": 0.07031324383509474,
+      "grad_norm": 5.5276259549136295,
+      "learning_rate": 2e-05,
+      "loss": 0.7226,
+      "step": 2954
+    },
+    {
+      "epoch": 0.07033704655812625,
+      "grad_norm": 0.6498290046764811,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 2955
+    },
+    {
+      "epoch": 0.07036084928115777,
+      "grad_norm": 5.880703954373084,
+      "learning_rate": 2e-05,
+      "loss": 0.7482,
+      "step": 2956
+    },
+    {
+      "epoch": 0.07038465200418928,
+      "grad_norm": 2.400939019716899,
+      "learning_rate": 2e-05,
+      "loss": 0.2014,
+      "step": 2957
+    },
+    {
+      "epoch": 0.07040845472722079,
+      "grad_norm": 3.603370619229645,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 2958
+    },
+    {
+      "epoch": 0.0704322574502523,
+      "grad_norm": 4.634551621063091,
+      "learning_rate": 2e-05,
+      "loss": 0.681,
+      "step": 2959
+    },
+    {
+      "epoch": 0.07045606017328382,
+      "grad_norm": 3.3648710509124378,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 2960
+    },
+    {
+      "epoch": 0.07047986289631533,
+      "grad_norm": 5.714905743686993,
+      "learning_rate": 2e-05,
+      "loss": 0.8605,
+      "step": 2961
+    },
+    {
+      "epoch": 0.07050366561934686,
+      "grad_norm": 6.59290965256529,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 2962
+    },
+    {
+      "epoch": 0.07052746834237837,
+      "grad_norm": 1.8396134462987253,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 2963
+    },
+    {
+      "epoch": 0.07055127106540988,
+      "grad_norm": 6.075359319963808,
+      "learning_rate": 2e-05,
+      "loss": 0.592,
+      "step": 2964
+    },
+    {
+      "epoch": 0.0705750737884414,
+      "grad_norm": 10.939742480924838,
+      "learning_rate": 2e-05,
+      "loss": 1.2347,
+      "step": 2965
+    },
+    {
+      "epoch": 0.07059887651147291,
+      "grad_norm": 3.487424671879242,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 2966
+    },
+    {
+      "epoch": 0.07062267923450442,
+      "grad_norm": 4.5290956992984786,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 2967
+    },
+    {
+      "epoch": 0.07064648195753594,
+      "grad_norm": 6.713911489625496,
+      "learning_rate": 2e-05,
+      "loss": 0.4687,
+      "step": 2968
+    },
+    {
+      "epoch": 0.07067028468056746,
+      "grad_norm": 9.18967579070619,
+      "learning_rate": 2e-05,
+      "loss": 1.0876,
+      "step": 2969
+    },
+    {
+      "epoch": 0.07069408740359898,
+      "grad_norm": 4.902925814742973,
+      "learning_rate": 2e-05,
+      "loss": 0.6712,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07071789012663049,
+      "grad_norm": 3.437099103258392,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 2971
+    },
+    {
+      "epoch": 0.070741692849662,
+      "grad_norm": 3.398434126622357,
+      "learning_rate": 2e-05,
+      "loss": 0.1594,
+      "step": 2972
+    },
+    {
+      "epoch": 0.07076549557269352,
+      "grad_norm": 3.126602408797842,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 2973
+    },
+    {
+      "epoch": 0.07078929829572503,
+      "grad_norm": 7.391204087180505,
+      "learning_rate": 2e-05,
+      "loss": 0.6127,
+      "step": 2974
+    },
+    {
+      "epoch": 0.07081310101875654,
+      "grad_norm": 4.190464151917323,
+      "learning_rate": 2e-05,
+      "loss": 0.2921,
+      "step": 2975
+    },
+    {
+      "epoch": 0.07083690374178805,
+      "grad_norm": 5.304260605401548,
+      "learning_rate": 2e-05,
+      "loss": 0.7906,
+      "step": 2976
+    },
+    {
+      "epoch": 0.07086070646481958,
+      "grad_norm": 14.318429973774498,
+      "learning_rate": 2e-05,
+      "loss": 0.6532,
+      "step": 2977
+    },
+    {
+      "epoch": 0.0708845091878511,
+      "grad_norm": 4.428055035015361,
+      "learning_rate": 2e-05,
+      "loss": 0.2233,
+      "step": 2978
+    },
+    {
+      "epoch": 0.07090831191088261,
+      "grad_norm": 7.847370513676451,
+      "learning_rate": 2e-05,
+      "loss": 0.3825,
+      "step": 2979
+    },
+    {
+      "epoch": 0.07093211463391412,
+      "grad_norm": 3.3098383940231795,
+      "learning_rate": 2e-05,
+      "loss": 0.2345,
+      "step": 2980
+    },
+    {
+      "epoch": 0.07095591735694563,
+      "grad_norm": 2.3918828091434525,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 2981
+    },
+    {
+      "epoch": 0.07097972007997715,
+      "grad_norm": 9.673456517755408,
+      "learning_rate": 2e-05,
+      "loss": 1.7189,
+      "step": 2982
+    },
+    {
+      "epoch": 0.07100352280300866,
+      "grad_norm": 2.395329324127035,
+      "learning_rate": 2e-05,
+      "loss": 0.2548,
+      "step": 2983
+    },
+    {
+      "epoch": 0.07102732552604019,
+      "grad_norm": 5.817293244323631,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 2984
+    },
+    {
+      "epoch": 0.0710511282490717,
+      "grad_norm": 6.178525862496912,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 2985
+    },
+    {
+      "epoch": 0.07107493097210321,
+      "grad_norm": 8.30239967873046,
+      "learning_rate": 2e-05,
+      "loss": 0.4944,
+      "step": 2986
+    },
+    {
+      "epoch": 0.07109873369513472,
+      "grad_norm": 6.968456012443415,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 2987
+    },
+    {
+      "epoch": 0.07112253641816624,
+      "grad_norm": 4.711212590168261,
+      "learning_rate": 2e-05,
+      "loss": 0.6324,
+      "step": 2988
+    },
+    {
+      "epoch": 0.07114633914119775,
+      "grad_norm": 6.996319935804769,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 2989
+    },
+    {
+      "epoch": 0.07117014186422926,
+      "grad_norm": 8.23909620176853,
+      "learning_rate": 2e-05,
+      "loss": 0.9373,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07119394458726078,
+      "grad_norm": 6.755615658469716,
+      "learning_rate": 2e-05,
+      "loss": 0.8604,
+      "step": 2991
+    },
+    {
+      "epoch": 0.0712177473102923,
+      "grad_norm": 8.410720494991892,
+      "learning_rate": 2e-05,
+      "loss": 1.8222,
+      "step": 2992
+    },
+    {
+      "epoch": 0.07124155003332382,
+      "grad_norm": 7.563169495913201,
+      "learning_rate": 2e-05,
+      "loss": 0.4512,
+      "step": 2993
+    },
+    {
+      "epoch": 0.07126535275635533,
+      "grad_norm": 2.6808918866977947,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 2994
+    },
+    {
+      "epoch": 0.07128915547938684,
+      "grad_norm": 6.587651973144588,
+      "learning_rate": 2e-05,
+      "loss": 0.7834,
+      "step": 2995
+    },
+    {
+      "epoch": 0.07131295820241836,
+      "grad_norm": 9.040930356923544,
+      "learning_rate": 2e-05,
+      "loss": 0.3653,
+      "step": 2996
+    },
+    {
+      "epoch": 0.07133676092544987,
+      "grad_norm": 3.0848164040416037,
+      "learning_rate": 2e-05,
+      "loss": 0.2205,
+      "step": 2997
+    },
+    {
+      "epoch": 0.07136056364848138,
+      "grad_norm": 2.520154685427244,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 2998
+    },
+    {
+      "epoch": 0.0713843663715129,
+      "grad_norm": 8.668613695574397,
+      "learning_rate": 2e-05,
+      "loss": 0.9699,
+      "step": 2999
+    },
+    {
+      "epoch": 0.07140816909454442,
+      "grad_norm": 2.6180127587012563,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07143197181757593,
+      "grad_norm": 3.6512132046490624,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 3001
+    },
+    {
+      "epoch": 0.07145577454060745,
+      "grad_norm": 3.8181532734471237,
+      "learning_rate": 2e-05,
+      "loss": 0.288,
+      "step": 3002
+    },
+    {
+      "epoch": 0.07147957726363896,
+      "grad_norm": 8.434056294964417,
+      "learning_rate": 2e-05,
+      "loss": 0.7168,
+      "step": 3003
+    },
+    {
+      "epoch": 0.07150337998667047,
+      "grad_norm": 7.624748110328643,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 3004
+    },
+    {
+      "epoch": 0.07152718270970199,
+      "grad_norm": 6.415673917940566,
+      "learning_rate": 2e-05,
+      "loss": 0.3675,
+      "step": 3005
+    },
+    {
+      "epoch": 0.0715509854327335,
+      "grad_norm": 7.839455111673625,
+      "learning_rate": 2e-05,
+      "loss": 0.509,
+      "step": 3006
+    },
+    {
+      "epoch": 0.07157478815576503,
+      "grad_norm": 2.3867439659421077,
+      "learning_rate": 2e-05,
+      "loss": 0.1581,
+      "step": 3007
+    },
+    {
+      "epoch": 0.07159859087879654,
+      "grad_norm": 8.220943149423043,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 3008
+    },
+    {
+      "epoch": 0.07162239360182805,
+      "grad_norm": 7.4858482924813545,
+      "learning_rate": 2e-05,
+      "loss": 0.7798,
+      "step": 3009
+    },
+    {
+      "epoch": 0.07164619632485957,
+      "grad_norm": 2.060852273550684,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07166999904789108,
+      "grad_norm": 4.901646203738716,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 3011
+    },
+    {
+      "epoch": 0.07169380177092259,
+      "grad_norm": 3.070338386908558,
+      "learning_rate": 2e-05,
+      "loss": 0.3136,
+      "step": 3012
+    },
+    {
+      "epoch": 0.0717176044939541,
+      "grad_norm": 6.6794239226044985,
+      "learning_rate": 2e-05,
+      "loss": 0.3299,
+      "step": 3013
+    },
+    {
+      "epoch": 0.07174140721698562,
+      "grad_norm": 3.576825217610677,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 3014
+    },
+    {
+      "epoch": 0.07176520994001714,
+      "grad_norm": 5.069267775821738,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 3015
+    },
+    {
+      "epoch": 0.07178901266304866,
+      "grad_norm": 8.667815635431984,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 3016
+    },
+    {
+      "epoch": 0.07181281538608017,
+      "grad_norm": 8.078154940835125,
+      "learning_rate": 2e-05,
+      "loss": 1.0895,
+      "step": 3017
+    },
+    {
+      "epoch": 0.07183661810911168,
+      "grad_norm": 8.02855330133369,
+      "learning_rate": 2e-05,
+      "loss": 0.7886,
+      "step": 3018
+    },
+    {
+      "epoch": 0.0718604208321432,
+      "grad_norm": 7.6518790430746595,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 3019
+    },
+    {
+      "epoch": 0.07188422355517471,
+      "grad_norm": 4.537731253317677,
+      "learning_rate": 2e-05,
+      "loss": 0.393,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07190802627820622,
+      "grad_norm": 5.132796150356288,
+      "learning_rate": 2e-05,
+      "loss": 0.6334,
+      "step": 3021
+    },
+    {
+      "epoch": 0.07193182900123773,
+      "grad_norm": 9.787264746525642,
+      "learning_rate": 2e-05,
+      "loss": 1.4539,
+      "step": 3022
+    },
+    {
+      "epoch": 0.07195563172426926,
+      "grad_norm": 3.714521092167474,
+      "learning_rate": 2e-05,
+      "loss": 0.2824,
+      "step": 3023
+    },
+    {
+      "epoch": 0.07197943444730077,
+      "grad_norm": 8.036659183825217,
+      "learning_rate": 2e-05,
+      "loss": 0.4782,
+      "step": 3024
+    },
+    {
+      "epoch": 0.07200323717033229,
+      "grad_norm": 7.0871332782081655,
+      "learning_rate": 2e-05,
+      "loss": 0.2936,
+      "step": 3025
+    },
+    {
+      "epoch": 0.0720270398933638,
+      "grad_norm": 4.043042250889275,
+      "learning_rate": 2e-05,
+      "loss": 0.3571,
+      "step": 3026
+    },
+    {
+      "epoch": 0.07205084261639531,
+      "grad_norm": 19.770426998574006,
+      "learning_rate": 2e-05,
+      "loss": 0.8168,
+      "step": 3027
+    },
+    {
+      "epoch": 0.07207464533942683,
+      "grad_norm": 3.7895999967358662,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 3028
+    },
+    {
+      "epoch": 0.07209844806245834,
+      "grad_norm": 3.6016005920504726,
+      "learning_rate": 2e-05,
+      "loss": 0.2759,
+      "step": 3029
+    },
+    {
+      "epoch": 0.07212225078548987,
+      "grad_norm": 6.2460614722567405,
+      "learning_rate": 2e-05,
+      "loss": 0.265,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07214605350852138,
+      "grad_norm": 7.308030290424988,
+      "learning_rate": 2e-05,
+      "loss": 1.6031,
+      "step": 3031
+    },
+    {
+      "epoch": 0.07216985623155289,
+      "grad_norm": 1.9065446888003406,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 3032
+    },
+    {
+      "epoch": 0.0721936589545844,
+      "grad_norm": 5.185319747156863,
+      "learning_rate": 2e-05,
+      "loss": 0.5599,
+      "step": 3033
+    },
+    {
+      "epoch": 0.07221746167761592,
+      "grad_norm": 1.867994946059192,
+      "learning_rate": 2e-05,
+      "loss": 0.1867,
+      "step": 3034
+    },
+    {
+      "epoch": 0.07224126440064743,
+      "grad_norm": 3.4521238925606603,
+      "learning_rate": 2e-05,
+      "loss": 0.1917,
+      "step": 3035
+    },
+    {
+      "epoch": 0.07226506712367894,
+      "grad_norm": 3.48569085261191,
+      "learning_rate": 2e-05,
+      "loss": 0.1846,
+      "step": 3036
+    },
+    {
+      "epoch": 0.07228886984671046,
+      "grad_norm": 3.870932019570253,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 3037
+    },
+    {
+      "epoch": 0.07231267256974198,
+      "grad_norm": 3.0672193861966415,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 3038
+    },
+    {
+      "epoch": 0.0723364752927735,
+      "grad_norm": 28.370084882763944,
+      "learning_rate": 2e-05,
+      "loss": 1.1147,
+      "step": 3039
+    },
+    {
+      "epoch": 0.07236027801580501,
+      "grad_norm": 9.783379407785283,
+      "learning_rate": 2e-05,
+      "loss": 0.3757,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07238408073883652,
+      "grad_norm": 3.557103540439515,
+      "learning_rate": 2e-05,
+      "loss": 0.296,
+      "step": 3041
+    },
+    {
+      "epoch": 0.07240788346186804,
+      "grad_norm": 8.509396626776375,
+      "learning_rate": 2e-05,
+      "loss": 1.1556,
+      "step": 3042
+    },
+    {
+      "epoch": 0.07243168618489955,
+      "grad_norm": 2.292157366690772,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 3043
+    },
+    {
+      "epoch": 0.07245548890793106,
+      "grad_norm": 4.1863519359014125,
+      "learning_rate": 2e-05,
+      "loss": 0.5123,
+      "step": 3044
+    },
+    {
+      "epoch": 0.07247929163096259,
+      "grad_norm": 7.9791099875449945,
+      "learning_rate": 2e-05,
+      "loss": 0.4434,
+      "step": 3045
+    },
+    {
+      "epoch": 0.0725030943539941,
+      "grad_norm": 9.211385816728466,
+      "learning_rate": 2e-05,
+      "loss": 0.7237,
+      "step": 3046
+    },
+    {
+      "epoch": 0.07252689707702561,
+      "grad_norm": 10.357155368831467,
+      "learning_rate": 2e-05,
+      "loss": 0.6938,
+      "step": 3047
+    },
+    {
+      "epoch": 0.07255069980005713,
+      "grad_norm": 7.863493696494571,
+      "learning_rate": 2e-05,
+      "loss": 0.4697,
+      "step": 3048
+    },
+    {
+      "epoch": 0.07257450252308864,
+      "grad_norm": 1.1205281374825626,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 3049
+    },
+    {
+      "epoch": 0.07259830524612015,
+      "grad_norm": 8.214683652290576,
+      "learning_rate": 2e-05,
+      "loss": 0.8766,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07262210796915167,
+      "grad_norm": 9.209973610580953,
+      "learning_rate": 2e-05,
+      "loss": 1.0149,
+      "step": 3051
+    },
+    {
+      "epoch": 0.07264591069218318,
+      "grad_norm": 4.769426792115809,
+      "learning_rate": 2e-05,
+      "loss": 0.4027,
+      "step": 3052
+    },
+    {
+      "epoch": 0.0726697134152147,
+      "grad_norm": 6.266234250740162,
+      "learning_rate": 2e-05,
+      "loss": 0.7249,
+      "step": 3053
+    },
+    {
+      "epoch": 0.07269351613824622,
+      "grad_norm": 2.334644042579667,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 3054
+    },
+    {
+      "epoch": 0.07271731886127773,
+      "grad_norm": 2.662889557760465,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 3055
+    },
+    {
+      "epoch": 0.07274112158430925,
+      "grad_norm": 7.6744104069109,
+      "learning_rate": 2e-05,
+      "loss": 0.8487,
+      "step": 3056
+    },
+    {
+      "epoch": 0.07276492430734076,
+      "grad_norm": 7.87665715627979,
+      "learning_rate": 2e-05,
+      "loss": 0.9254,
+      "step": 3057
+    },
+    {
+      "epoch": 0.07278872703037227,
+      "grad_norm": 7.119391517776393,
+      "learning_rate": 2e-05,
+      "loss": 0.8556,
+      "step": 3058
+    },
+    {
+      "epoch": 0.07281252975340378,
+      "grad_norm": 8.83381776933689,
+      "learning_rate": 2e-05,
+      "loss": 0.5937,
+      "step": 3059
+    },
+    {
+      "epoch": 0.0728363324764353,
+      "grad_norm": 6.173302828507047,
+      "learning_rate": 2e-05,
+      "loss": 0.5602,
+      "step": 3060
+    },
+    {
+      "epoch": 0.07286013519946682,
+      "grad_norm": 8.015462138850616,
+      "learning_rate": 2e-05,
+      "loss": 0.488,
+      "step": 3061
+    },
+    {
+      "epoch": 0.07288393792249834,
+      "grad_norm": 5.980815731446912,
+      "learning_rate": 2e-05,
+      "loss": 0.7494,
+      "step": 3062
+    },
+    {
+      "epoch": 0.07290774064552985,
+      "grad_norm": 4.872614702038267,
+      "learning_rate": 2e-05,
+      "loss": 0.2607,
+      "step": 3063
+    },
+    {
+      "epoch": 0.07293154336856136,
+      "grad_norm": 3.5411066256259884,
+      "learning_rate": 2e-05,
+      "loss": 0.3484,
+      "step": 3064
+    },
+    {
+      "epoch": 0.07295534609159288,
+      "grad_norm": 2.6106844757027416,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 3065
+    },
+    {
+      "epoch": 0.07297914881462439,
+      "grad_norm": 5.787897668615346,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 3066
+    },
+    {
+      "epoch": 0.0730029515376559,
+      "grad_norm": 7.122885277164452,
+      "learning_rate": 2e-05,
+      "loss": 0.6134,
+      "step": 3067
+    },
+    {
+      "epoch": 0.07302675426068743,
+      "grad_norm": 3.3842948950296594,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 3068
+    },
+    {
+      "epoch": 0.07305055698371894,
+      "grad_norm": 3.143963387993847,
+      "learning_rate": 2e-05,
+      "loss": 0.299,
+      "step": 3069
+    },
+    {
+      "epoch": 0.07307435970675046,
+      "grad_norm": 8.707365569006797,
+      "learning_rate": 2e-05,
+      "loss": 0.966,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07309816242978197,
+      "grad_norm": 2.246225256219838,
+      "learning_rate": 2e-05,
+      "loss": 0.2023,
+      "step": 3071
+    },
+    {
+      "epoch": 0.07312196515281348,
+      "grad_norm": 6.445018445151646,
+      "learning_rate": 2e-05,
+      "loss": 1.3773,
+      "step": 3072
+    },
+    {
+      "epoch": 0.073145767875845,
+      "grad_norm": 6.009512130580472,
+      "learning_rate": 2e-05,
+      "loss": 0.9123,
+      "step": 3073
+    },
+    {
+      "epoch": 0.07316957059887651,
+      "grad_norm": 8.401143935355442,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 3074
+    },
+    {
+      "epoch": 0.07319337332190802,
+      "grad_norm": 2.1427285791665467,
+      "learning_rate": 2e-05,
+      "loss": 0.1584,
+      "step": 3075
+    },
+    {
+      "epoch": 0.07321717604493955,
+      "grad_norm": 5.1588218573943285,
+      "learning_rate": 2e-05,
+      "loss": 0.422,
+      "step": 3076
+    },
+    {
+      "epoch": 0.07324097876797106,
+      "grad_norm": 6.0021876232584495,
+      "learning_rate": 2e-05,
+      "loss": 0.3306,
+      "step": 3077
+    },
+    {
+      "epoch": 0.07326478149100257,
+      "grad_norm": 1.8276870882321317,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 3078
+    },
+    {
+      "epoch": 0.07328858421403409,
+      "grad_norm": 9.98062111589466,
+      "learning_rate": 2e-05,
+      "loss": 1.5122,
+      "step": 3079
+    },
+    {
+      "epoch": 0.0733123869370656,
+      "grad_norm": 3.703452766789673,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07333618966009711,
+      "grad_norm": 3.697639896673973,
+      "learning_rate": 2e-05,
+      "loss": 0.1954,
+      "step": 3081
+    },
+    {
+      "epoch": 0.07335999238312862,
+      "grad_norm": 4.818652455612524,
+      "learning_rate": 2e-05,
+      "loss": 0.3551,
+      "step": 3082
+    },
+    {
+      "epoch": 0.07338379510616014,
+      "grad_norm": 8.482203229159582,
+      "learning_rate": 2e-05,
+      "loss": 0.4808,
+      "step": 3083
+    },
+    {
+      "epoch": 0.07340759782919166,
+      "grad_norm": 2.414335038447362,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 3084
+    },
+    {
+      "epoch": 0.07343140055222318,
+      "grad_norm": 5.6368117091436645,
+      "learning_rate": 2e-05,
+      "loss": 0.8309,
+      "step": 3085
+    },
+    {
+      "epoch": 0.07345520327525469,
+      "grad_norm": 3.6343609025663453,
+      "learning_rate": 2e-05,
+      "loss": 0.2064,
+      "step": 3086
+    },
+    {
+      "epoch": 0.0734790059982862,
+      "grad_norm": 6.647090734406235,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 3087
+    },
+    {
+      "epoch": 0.07350280872131772,
+      "grad_norm": 4.146316261417578,
+      "learning_rate": 2e-05,
+      "loss": 0.4413,
+      "step": 3088
+    },
+    {
+      "epoch": 0.07352661144434923,
+      "grad_norm": 2.7052905280587667,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 3089
+    },
+    {
+      "epoch": 0.07355041416738074,
+      "grad_norm": 2.883624698802168,
+      "learning_rate": 2e-05,
+      "loss": 0.2836,
+      "step": 3090
+    },
+    {
+      "epoch": 0.07357421689041227,
+      "grad_norm": 3.478789163998198,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 3091
+    },
+    {
+      "epoch": 0.07359801961344378,
+      "grad_norm": 3.2554484885650785,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 3092
+    },
+    {
+      "epoch": 0.0736218223364753,
+      "grad_norm": 8.25266242423349,
+      "learning_rate": 2e-05,
+      "loss": 0.5079,
+      "step": 3093
+    },
+    {
+      "epoch": 0.07364562505950681,
+      "grad_norm": 1.9717180932041667,
+      "learning_rate": 2e-05,
+      "loss": 0.1522,
+      "step": 3094
+    },
+    {
+      "epoch": 0.07366942778253832,
+      "grad_norm": 13.014685934994215,
+      "learning_rate": 2e-05,
+      "loss": 1.1938,
+      "step": 3095
+    },
+    {
+      "epoch": 0.07369323050556983,
+      "grad_norm": 7.502890747946628,
+      "learning_rate": 2e-05,
+      "loss": 1.6548,
+      "step": 3096
+    },
+    {
+      "epoch": 0.07371703322860135,
+      "grad_norm": 7.042739200119292,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 3097
+    },
+    {
+      "epoch": 0.07374083595163286,
+      "grad_norm": 3.6032222306140342,
+      "learning_rate": 2e-05,
+      "loss": 0.2866,
+      "step": 3098
+    },
+    {
+      "epoch": 0.07376463867466439,
+      "grad_norm": 5.336544552146485,
+      "learning_rate": 2e-05,
+      "loss": 0.2702,
+      "step": 3099
+    },
+    {
+      "epoch": 0.0737884413976959,
+      "grad_norm": 2.7922009554034513,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07381224412072741,
+      "grad_norm": 7.327299170738407,
+      "learning_rate": 2e-05,
+      "loss": 0.8015,
+      "step": 3101
+    },
+    {
+      "epoch": 0.07383604684375893,
+      "grad_norm": 6.899784035763868,
+      "learning_rate": 2e-05,
+      "loss": 0.6937,
+      "step": 3102
+    },
+    {
+      "epoch": 0.07385984956679044,
+      "grad_norm": 5.546562009556755,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 3103
+    },
+    {
+      "epoch": 0.07388365228982195,
+      "grad_norm": 2.1403414434075367,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 3104
+    },
+    {
+      "epoch": 0.07390745501285347,
+      "grad_norm": 6.440359048974235,
+      "learning_rate": 2e-05,
+      "loss": 1.0885,
+      "step": 3105
+    },
+    {
+      "epoch": 0.07393125773588499,
+      "grad_norm": 6.768050750234474,
+      "learning_rate": 2e-05,
+      "loss": 0.7216,
+      "step": 3106
+    },
+    {
+      "epoch": 0.0739550604589165,
+      "grad_norm": 2.087937853063055,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 3107
+    },
+    {
+      "epoch": 0.07397886318194802,
+      "grad_norm": 2.4432381245321886,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 3108
+    },
+    {
+      "epoch": 0.07400266590497953,
+      "grad_norm": 4.8668499161845045,
+      "learning_rate": 2e-05,
+      "loss": 0.5348,
+      "step": 3109
+    },
+    {
+      "epoch": 0.07402646862801104,
+      "grad_norm": 4.638065416161358,
+      "learning_rate": 2e-05,
+      "loss": 0.7791,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07405027135104256,
+      "grad_norm": 5.260438649681682,
+      "learning_rate": 2e-05,
+      "loss": 0.4908,
+      "step": 3111
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 2.289009102069517,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 3112
+    },
+    {
+      "epoch": 0.07409787679710558,
+      "grad_norm": 3.1381751974625898,
+      "learning_rate": 2e-05,
+      "loss": 0.2305,
+      "step": 3113
+    },
+    {
+      "epoch": 0.07412167952013711,
+      "grad_norm": 8.792762912220889,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 3114
+    },
+    {
+      "epoch": 0.07414548224316862,
+      "grad_norm": 7.108866292421104,
+      "learning_rate": 2e-05,
+      "loss": 0.3753,
+      "step": 3115
+    },
+    {
+      "epoch": 0.07416928496620014,
+      "grad_norm": 2.3878354230750105,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 3116
+    },
+    {
+      "epoch": 0.07419308768923165,
+      "grad_norm": 3.126756640888974,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 3117
+    },
+    {
+      "epoch": 0.07421689041226316,
+      "grad_norm": 6.002913089713759,
+      "learning_rate": 2e-05,
+      "loss": 0.7293,
+      "step": 3118
+    },
+    {
+      "epoch": 0.07424069313529467,
+      "grad_norm": 2.0359402231806847,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 3119
+    },
+    {
+      "epoch": 0.07426449585832619,
+      "grad_norm": 3.4633345770375032,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 3120
+    },
+    {
+      "epoch": 0.0742882985813577,
+      "grad_norm": 2.5048406308106714,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 3121
+    },
+    {
+      "epoch": 0.07431210130438923,
+      "grad_norm": 3.3657739105069933,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 3122
+    },
+    {
+      "epoch": 0.07433590402742074,
+      "grad_norm": 7.874154451589655,
+      "learning_rate": 2e-05,
+      "loss": 1.0067,
+      "step": 3123
+    },
+    {
+      "epoch": 0.07435970675045225,
+      "grad_norm": 9.14620901056601,
+      "learning_rate": 2e-05,
+      "loss": 0.4834,
+      "step": 3124
+    },
+    {
+      "epoch": 0.07438350947348377,
+      "grad_norm": 2.8332133597094242,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 3125
+    },
+    {
+      "epoch": 0.07440731219651528,
+      "grad_norm": 11.259830179515545,
+      "learning_rate": 2e-05,
+      "loss": 2.6184,
+      "step": 3126
+    },
+    {
+      "epoch": 0.07443111491954679,
+      "grad_norm": 7.454290736120652,
+      "learning_rate": 2e-05,
+      "loss": 0.3032,
+      "step": 3127
+    },
+    {
+      "epoch": 0.0744549176425783,
+      "grad_norm": 5.123004702717932,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 3128
+    },
+    {
+      "epoch": 0.07447872036560983,
+      "grad_norm": 5.980445519416589,
+      "learning_rate": 2e-05,
+      "loss": 0.9176,
+      "step": 3129
+    },
+    {
+      "epoch": 0.07450252308864135,
+      "grad_norm": 8.515910788711025,
+      "learning_rate": 2e-05,
+      "loss": 0.9667,
+      "step": 3130
+    },
+    {
+      "epoch": 0.07452632581167286,
+      "grad_norm": 6.587849905382184,
+      "learning_rate": 2e-05,
+      "loss": 0.2649,
+      "step": 3131
+    },
+    {
+      "epoch": 0.07455012853470437,
+      "grad_norm": 4.956352512272401,
+      "learning_rate": 2e-05,
+      "loss": 0.9012,
+      "step": 3132
+    },
+    {
+      "epoch": 0.07457393125773588,
+      "grad_norm": 3.8805541541291353,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 3133
+    },
+    {
+      "epoch": 0.0745977339807674,
+      "grad_norm": 7.348287594290354,
+      "learning_rate": 2e-05,
+      "loss": 1.3447,
+      "step": 3134
+    },
+    {
+      "epoch": 0.07462153670379891,
+      "grad_norm": 2.6894424411509172,
+      "learning_rate": 2e-05,
+      "loss": 0.2054,
+      "step": 3135
+    },
+    {
+      "epoch": 0.07464533942683042,
+      "grad_norm": 2.0874071568133767,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 3136
+    },
+    {
+      "epoch": 0.07466914214986195,
+      "grad_norm": 3.149697862875305,
+      "learning_rate": 2e-05,
+      "loss": 0.2936,
+      "step": 3137
+    },
+    {
+      "epoch": 0.07469294487289346,
+      "grad_norm": 8.539613230186058,
+      "learning_rate": 2e-05,
+      "loss": 0.448,
+      "step": 3138
+    },
+    {
+      "epoch": 0.07471674759592498,
+      "grad_norm": 8.493888245038455,
+      "learning_rate": 2e-05,
+      "loss": 1.8475,
+      "step": 3139
+    },
+    {
+      "epoch": 0.07474055031895649,
+      "grad_norm": 7.3515993542962415,
+      "learning_rate": 2e-05,
+      "loss": 0.419,
+      "step": 3140
+    },
+    {
+      "epoch": 0.074764353041988,
+      "grad_norm": 2.8723121426125084,
+      "learning_rate": 2e-05,
+      "loss": 0.2182,
+      "step": 3141
+    },
+    {
+      "epoch": 0.07478815576501952,
+      "grad_norm": 8.015776306247872,
+      "learning_rate": 2e-05,
+      "loss": 0.6725,
+      "step": 3142
+    },
+    {
+      "epoch": 0.07481195848805103,
+      "grad_norm": 2.1984470241712764,
+      "learning_rate": 2e-05,
+      "loss": 0.1555,
+      "step": 3143
+    },
+    {
+      "epoch": 0.07483576121108254,
+      "grad_norm": 3.576883847859961,
+      "learning_rate": 2e-05,
+      "loss": 0.5967,
+      "step": 3144
+    },
+    {
+      "epoch": 0.07485956393411407,
+      "grad_norm": 5.809783078966722,
+      "learning_rate": 2e-05,
+      "loss": 1.2954,
+      "step": 3145
+    },
+    {
+      "epoch": 0.07488336665714558,
+      "grad_norm": 35.79185500784045,
+      "learning_rate": 2e-05,
+      "loss": 1.0849,
+      "step": 3146
+    },
+    {
+      "epoch": 0.0749071693801771,
+      "grad_norm": 7.861716931155669,
+      "learning_rate": 2e-05,
+      "loss": 1.3056,
+      "step": 3147
+    },
+    {
+      "epoch": 0.0749309721032086,
+      "grad_norm": 4.9619434207130375,
+      "learning_rate": 2e-05,
+      "loss": 0.3087,
+      "step": 3148
+    },
+    {
+      "epoch": 0.07495477482624012,
+      "grad_norm": 2.171980457500472,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 3149
+    },
+    {
+      "epoch": 0.07497857754927163,
+      "grad_norm": 2.8907930735133314,
+      "learning_rate": 2e-05,
+      "loss": 0.2737,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07500238027230315,
+      "grad_norm": 6.471359372493306,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 3151
+    },
+    {
+      "epoch": 0.07502618299533467,
+      "grad_norm": 6.722250274712194,
+      "learning_rate": 2e-05,
+      "loss": 0.7763,
+      "step": 3152
+    },
+    {
+      "epoch": 0.07504998571836619,
+      "grad_norm": 3.6009217409101963,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 3153
+    },
+    {
+      "epoch": 0.0750737884413977,
+      "grad_norm": 3.7216724490010606,
+      "learning_rate": 2e-05,
+      "loss": 0.3105,
+      "step": 3154
+    },
+    {
+      "epoch": 0.07509759116442921,
+      "grad_norm": 9.675220875819225,
+      "learning_rate": 2e-05,
+      "loss": 1.2021,
+      "step": 3155
+    },
+    {
+      "epoch": 0.07512139388746072,
+      "grad_norm": 8.511588939867236,
+      "learning_rate": 2e-05,
+      "loss": 1.0141,
+      "step": 3156
+    },
+    {
+      "epoch": 0.07514519661049224,
+      "grad_norm": 6.017830130950235,
+      "learning_rate": 2e-05,
+      "loss": 0.9032,
+      "step": 3157
+    },
+    {
+      "epoch": 0.07516899933352375,
+      "grad_norm": 1.5390272403951817,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 3158
+    },
+    {
+      "epoch": 0.07519280205655526,
+      "grad_norm": 1.557425111137272,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 3159
+    },
+    {
+      "epoch": 0.07521660477958679,
+      "grad_norm": 3.108480501221995,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 3160
+    },
+    {
+      "epoch": 0.0752404075026183,
+      "grad_norm": 2.265411181274933,
+      "learning_rate": 2e-05,
+      "loss": 0.1815,
+      "step": 3161
+    },
+    {
+      "epoch": 0.07526421022564982,
+      "grad_norm": 7.253860090068471,
+      "learning_rate": 2e-05,
+      "loss": 0.2814,
+      "step": 3162
+    },
+    {
+      "epoch": 0.07528801294868133,
+      "grad_norm": 1.4295448406888622,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 3163
+    },
+    {
+      "epoch": 0.07531181567171284,
+      "grad_norm": 6.64857382005207,
+      "learning_rate": 2e-05,
+      "loss": 1.1404,
+      "step": 3164
+    },
+    {
+      "epoch": 0.07533561839474436,
+      "grad_norm": 8.192027462147685,
+      "learning_rate": 2e-05,
+      "loss": 1.0276,
+      "step": 3165
+    },
+    {
+      "epoch": 0.07535942111777587,
+      "grad_norm": 3.3391814658902694,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 3166
+    },
+    {
+      "epoch": 0.0753832238408074,
+      "grad_norm": 2.050156642415763,
+      "learning_rate": 2e-05,
+      "loss": 0.1547,
+      "step": 3167
+    },
+    {
+      "epoch": 0.07540702656383891,
+      "grad_norm": 7.715657909426436,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 3168
+    },
+    {
+      "epoch": 0.07543082928687042,
+      "grad_norm": 2.9063390200675543,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 3169
+    },
+    {
+      "epoch": 0.07545463200990193,
+      "grad_norm": 11.439750037957124,
+      "learning_rate": 2e-05,
+      "loss": 0.6431,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07547843473293345,
+      "grad_norm": 5.449392858695256,
+      "learning_rate": 2e-05,
+      "loss": 1.2337,
+      "step": 3171
+    },
+    {
+      "epoch": 0.07550223745596496,
+      "grad_norm": 14.525522823356509,
+      "learning_rate": 2e-05,
+      "loss": 0.4449,
+      "step": 3172
+    },
+    {
+      "epoch": 0.07552604017899647,
+      "grad_norm": 7.001832983501225,
+      "learning_rate": 2e-05,
+      "loss": 0.9586,
+      "step": 3173
+    },
+    {
+      "epoch": 0.07554984290202799,
+      "grad_norm": 12.068936256136917,
+      "learning_rate": 2e-05,
+      "loss": 1.2925,
+      "step": 3174
+    },
+    {
+      "epoch": 0.07557364562505951,
+      "grad_norm": 5.669537658869993,
+      "learning_rate": 2e-05,
+      "loss": 0.67,
+      "step": 3175
+    },
+    {
+      "epoch": 0.07559744834809103,
+      "grad_norm": 6.804847333048303,
+      "learning_rate": 2e-05,
+      "loss": 0.3985,
+      "step": 3176
+    },
+    {
+      "epoch": 0.07562125107112254,
+      "grad_norm": 2.445570602076634,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 3177
+    },
+    {
+      "epoch": 0.07564505379415405,
+      "grad_norm": 2.378697117359055,
+      "learning_rate": 2e-05,
+      "loss": 0.1543,
+      "step": 3178
+    },
+    {
+      "epoch": 0.07566885651718556,
+      "grad_norm": 3.001897050432784,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 3179
+    },
+    {
+      "epoch": 0.07569265924021708,
+      "grad_norm": 16.11664932070678,
+      "learning_rate": 2e-05,
+      "loss": 0.422,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07571646196324859,
+      "grad_norm": 4.641794320504068,
+      "learning_rate": 2e-05,
+      "loss": 0.706,
+      "step": 3181
+    },
+    {
+      "epoch": 0.0757402646862801,
+      "grad_norm": 3.984832209664199,
+      "learning_rate": 2e-05,
+      "loss": 0.2881,
+      "step": 3182
+    },
+    {
+      "epoch": 0.07576406740931163,
+      "grad_norm": 4.909223997361426,
+      "learning_rate": 2e-05,
+      "loss": 0.9502,
+      "step": 3183
+    },
+    {
+      "epoch": 0.07578787013234314,
+      "grad_norm": 4.630381869776765,
+      "learning_rate": 2e-05,
+      "loss": 0.4498,
+      "step": 3184
+    },
+    {
+      "epoch": 0.07581167285537466,
+      "grad_norm": 3.0483758747915672,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 3185
+    },
+    {
+      "epoch": 0.07583547557840617,
+      "grad_norm": 5.093751871484865,
+      "learning_rate": 2e-05,
+      "loss": 0.7299,
+      "step": 3186
+    },
+    {
+      "epoch": 0.07585927830143768,
+      "grad_norm": 3.3419317919168847,
+      "learning_rate": 2e-05,
+      "loss": 0.2592,
+      "step": 3187
+    },
+    {
+      "epoch": 0.0758830810244692,
+      "grad_norm": 7.5886538323944315,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 3188
+    },
+    {
+      "epoch": 0.07590688374750071,
+      "grad_norm": 3.7125373948879408,
+      "learning_rate": 2e-05,
+      "loss": 0.1541,
+      "step": 3189
+    },
+    {
+      "epoch": 0.07593068647053224,
+      "grad_norm": 4.395251133416662,
+      "learning_rate": 2e-05,
+      "loss": 0.3694,
+      "step": 3190
+    },
+    {
+      "epoch": 0.07595448919356375,
+      "grad_norm": 6.69615661908931,
+      "learning_rate": 2e-05,
+      "loss": 0.9679,
+      "step": 3191
+    },
+    {
+      "epoch": 0.07597829191659526,
+      "grad_norm": 8.034324208712823,
+      "learning_rate": 2e-05,
+      "loss": 0.3163,
+      "step": 3192
+    },
+    {
+      "epoch": 0.07600209463962677,
+      "grad_norm": 9.177124637612422,
+      "learning_rate": 2e-05,
+      "loss": 0.8338,
+      "step": 3193
+    },
+    {
+      "epoch": 0.07602589736265829,
+      "grad_norm": 9.977022318320653,
+      "learning_rate": 2e-05,
+      "loss": 1.8172,
+      "step": 3194
+    },
+    {
+      "epoch": 0.0760497000856898,
+      "grad_norm": 7.339492342059103,
+      "learning_rate": 2e-05,
+      "loss": 0.402,
+      "step": 3195
+    },
+    {
+      "epoch": 0.07607350280872131,
+      "grad_norm": 3.046760007833163,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 3196
+    },
+    {
+      "epoch": 0.07609730553175283,
+      "grad_norm": 4.544589016534341,
+      "learning_rate": 2e-05,
+      "loss": 0.4328,
+      "step": 3197
+    },
+    {
+      "epoch": 0.07612110825478435,
+      "grad_norm": 5.475382655484444,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 3198
+    },
+    {
+      "epoch": 0.07614491097781587,
+      "grad_norm": 3.9864462582891522,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 3199
+    },
+    {
+      "epoch": 0.07616871370084738,
+      "grad_norm": 7.534842211082511,
+      "learning_rate": 2e-05,
+      "loss": 0.3123,
+      "step": 3200
+    },
+    {
+      "epoch": 0.07619251642387889,
+      "grad_norm": 5.655652517939138,
+      "learning_rate": 2e-05,
+      "loss": 0.2553,
+      "step": 3201
+    },
+    {
+      "epoch": 0.0762163191469104,
+      "grad_norm": 2.2233207604489937,
+      "learning_rate": 2e-05,
+      "loss": 0.1594,
+      "step": 3202
+    },
+    {
+      "epoch": 0.07624012186994192,
+      "grad_norm": 3.741270065104862,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 3203
+    },
+    {
+      "epoch": 0.07626392459297343,
+      "grad_norm": 6.978726407937986,
+      "learning_rate": 2e-05,
+      "loss": 0.2877,
+      "step": 3204
+    },
+    {
+      "epoch": 0.07628772731600494,
+      "grad_norm": 6.635155507083342,
+      "learning_rate": 2e-05,
+      "loss": 0.7935,
+      "step": 3205
+    },
+    {
+      "epoch": 0.07631153003903647,
+      "grad_norm": 7.002108658795269,
+      "learning_rate": 2e-05,
+      "loss": 0.9559,
+      "step": 3206
+    },
+    {
+      "epoch": 0.07633533276206798,
+      "grad_norm": 7.337721722417961,
+      "learning_rate": 2e-05,
+      "loss": 0.9262,
+      "step": 3207
+    },
+    {
+      "epoch": 0.0763591354850995,
+      "grad_norm": 11.693377293572892,
+      "learning_rate": 2e-05,
+      "loss": 0.6014,
+      "step": 3208
+    },
+    {
+      "epoch": 0.07638293820813101,
+      "grad_norm": 12.413575662461723,
+      "learning_rate": 2e-05,
+      "loss": 0.473,
+      "step": 3209
+    },
+    {
+      "epoch": 0.07640674093116252,
+      "grad_norm": 6.251424482215584,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 3210
+    },
+    {
+      "epoch": 0.07643054365419404,
+      "grad_norm": 4.8567317716103755,
+      "learning_rate": 2e-05,
+      "loss": 0.2226,
+      "step": 3211
+    },
+    {
+      "epoch": 0.07645434637722555,
+      "grad_norm": 5.5403341108874695,
+      "learning_rate": 2e-05,
+      "loss": 0.9047,
+      "step": 3212
+    },
+    {
+      "epoch": 0.07647814910025708,
+      "grad_norm": 9.152441645993209,
+      "learning_rate": 2e-05,
+      "loss": 1.1339,
+      "step": 3213
+    },
+    {
+      "epoch": 0.07650195182328859,
+      "grad_norm": 8.396262600042158,
+      "learning_rate": 2e-05,
+      "loss": 0.691,
+      "step": 3214
+    },
+    {
+      "epoch": 0.0765257545463201,
+      "grad_norm": 4.107351548845595,
+      "learning_rate": 2e-05,
+      "loss": 0.7553,
+      "step": 3215
+    },
+    {
+      "epoch": 0.07654955726935161,
+      "grad_norm": 7.629974220900615,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 3216
+    },
+    {
+      "epoch": 0.07657335999238313,
+      "grad_norm": 3.2149953530283817,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 3217
+    },
+    {
+      "epoch": 0.07659716271541464,
+      "grad_norm": 6.303809482904614,
+      "learning_rate": 2e-05,
+      "loss": 0.3754,
+      "step": 3218
+    },
+    {
+      "epoch": 0.07662096543844615,
+      "grad_norm": 6.925993380330019,
+      "learning_rate": 2e-05,
+      "loss": 0.8396,
+      "step": 3219
+    },
+    {
+      "epoch": 0.07664476816147767,
+      "grad_norm": 4.763882514388828,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 3220
+    },
+    {
+      "epoch": 0.0766685708845092,
+      "grad_norm": 3.1348686174282943,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 3221
+    },
+    {
+      "epoch": 0.0766923736075407,
+      "grad_norm": 3.419988155967535,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 3222
+    },
+    {
+      "epoch": 0.07671617633057222,
+      "grad_norm": 6.534631704368822,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 3223
+    },
+    {
+      "epoch": 0.07673997905360373,
+      "grad_norm": 7.69970886264325,
+      "learning_rate": 2e-05,
+      "loss": 1.1387,
+      "step": 3224
+    },
+    {
+      "epoch": 0.07676378177663525,
+      "grad_norm": 3.177256167276254,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 3225
+    },
+    {
+      "epoch": 0.07678758449966676,
+      "grad_norm": 5.1843693108227376,
+      "learning_rate": 2e-05,
+      "loss": 0.7856,
+      "step": 3226
+    },
+    {
+      "epoch": 0.07681138722269827,
+      "grad_norm": 3.0455662225348994,
+      "learning_rate": 2e-05,
+      "loss": 0.2129,
+      "step": 3227
+    },
+    {
+      "epoch": 0.0768351899457298,
+      "grad_norm": 7.436534449704131,
+      "learning_rate": 2e-05,
+      "loss": 0.5814,
+      "step": 3228
+    },
+    {
+      "epoch": 0.07685899266876131,
+      "grad_norm": 6.203424828686114,
+      "learning_rate": 2e-05,
+      "loss": 0.253,
+      "step": 3229
+    },
+    {
+      "epoch": 0.07688279539179282,
+      "grad_norm": 5.537786378095716,
+      "learning_rate": 2e-05,
+      "loss": 0.6567,
+      "step": 3230
+    },
+    {
+      "epoch": 0.07690659811482434,
+      "grad_norm": 5.3631208246278606,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 3231
+    },
+    {
+      "epoch": 0.07693040083785585,
+      "grad_norm": 2.7935640979392677,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 3232
+    },
+    {
+      "epoch": 0.07695420356088736,
+      "grad_norm": 2.627343033483067,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 3233
+    },
+    {
+      "epoch": 0.07697800628391888,
+      "grad_norm": 3.0520549614994965,
+      "learning_rate": 2e-05,
+      "loss": 0.2437,
+      "step": 3234
+    },
+    {
+      "epoch": 0.07700180900695039,
+      "grad_norm": 1.7636117892887646,
+      "learning_rate": 2e-05,
+      "loss": 0.1894,
+      "step": 3235
+    },
+    {
+      "epoch": 0.07702561172998192,
+      "grad_norm": 3.650549731645895,
+      "learning_rate": 2e-05,
+      "loss": 0.3274,
+      "step": 3236
+    },
+    {
+      "epoch": 0.07704941445301343,
+      "grad_norm": 3.9413858720241475,
+      "learning_rate": 2e-05,
+      "loss": 0.2664,
+      "step": 3237
+    },
+    {
+      "epoch": 0.07707321717604494,
+      "grad_norm": 8.041550492937256,
+      "learning_rate": 2e-05,
+      "loss": 0.4365,
+      "step": 3238
+    },
+    {
+      "epoch": 0.07709701989907645,
+      "grad_norm": 8.425937470524094,
+      "learning_rate": 2e-05,
+      "loss": 1.0779,
+      "step": 3239
+    },
+    {
+      "epoch": 0.07712082262210797,
+      "grad_norm": 2.7030580703530775,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07714462534513948,
+      "grad_norm": 5.9064034428368934,
+      "learning_rate": 2e-05,
+      "loss": 0.929,
+      "step": 3241
+    },
+    {
+      "epoch": 0.077168428068171,
+      "grad_norm": 8.033646767044885,
+      "learning_rate": 2e-05,
+      "loss": 0.3768,
+      "step": 3242
+    },
+    {
+      "epoch": 0.0771922307912025,
+      "grad_norm": 11.04031356160318,
+      "learning_rate": 2e-05,
+      "loss": 0.9044,
+      "step": 3243
+    },
+    {
+      "epoch": 0.07721603351423403,
+      "grad_norm": 14.349376311822233,
+      "learning_rate": 2e-05,
+      "loss": 0.712,
+      "step": 3244
+    },
+    {
+      "epoch": 0.07723983623726555,
+      "grad_norm": 7.119580772478245,
+      "learning_rate": 2e-05,
+      "loss": 1.3185,
+      "step": 3245
+    },
+    {
+      "epoch": 0.07726363896029706,
+      "grad_norm": 8.025785718546445,
+      "learning_rate": 2e-05,
+      "loss": 0.3141,
+      "step": 3246
+    },
+    {
+      "epoch": 0.07728744168332857,
+      "grad_norm": 6.06941709190835,
+      "learning_rate": 2e-05,
+      "loss": 0.6315,
+      "step": 3247
+    },
+    {
+      "epoch": 0.07731124440636009,
+      "grad_norm": 3.54057187767356,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 3248
+    },
+    {
+      "epoch": 0.0773350471293916,
+      "grad_norm": 10.57251825244595,
+      "learning_rate": 2e-05,
+      "loss": 1.7469,
+      "step": 3249
+    },
+    {
+      "epoch": 0.07735884985242311,
+      "grad_norm": 6.85685854783932,
+      "learning_rate": 2e-05,
+      "loss": 1.1086,
+      "step": 3250
+    },
+    {
+      "epoch": 0.07738265257545464,
+      "grad_norm": 12.712823180335466,
+      "learning_rate": 2e-05,
+      "loss": 0.9319,
+      "step": 3251
+    },
+    {
+      "epoch": 0.07740645529848615,
+      "grad_norm": 3.32358497504416,
+      "learning_rate": 2e-05,
+      "loss": 0.1687,
+      "step": 3252
+    },
+    {
+      "epoch": 0.07743025802151766,
+      "grad_norm": 5.83026495465771,
+      "learning_rate": 2e-05,
+      "loss": 0.2631,
+      "step": 3253
+    },
+    {
+      "epoch": 0.07745406074454918,
+      "grad_norm": 3.3951960638965213,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 3254
+    },
+    {
+      "epoch": 0.07747786346758069,
+      "grad_norm": 1.546979175693043,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 3255
+    },
+    {
+      "epoch": 0.0775016661906122,
+      "grad_norm": 6.098829235860274,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 3256
+    },
+    {
+      "epoch": 0.07752546891364372,
+      "grad_norm": 6.276602788748124,
+      "learning_rate": 2e-05,
+      "loss": 0.4415,
+      "step": 3257
+    },
+    {
+      "epoch": 0.07754927163667523,
+      "grad_norm": 7.913511472244098,
+      "learning_rate": 2e-05,
+      "loss": 1.0519,
+      "step": 3258
+    },
+    {
+      "epoch": 0.07757307435970676,
+      "grad_norm": 8.2341919680903,
+      "learning_rate": 2e-05,
+      "loss": 0.8865,
+      "step": 3259
+    },
+    {
+      "epoch": 0.07759687708273827,
+      "grad_norm": 7.847811452921009,
+      "learning_rate": 2e-05,
+      "loss": 0.8726,
+      "step": 3260
+    },
+    {
+      "epoch": 0.07762067980576978,
+      "grad_norm": 5.738020720129733,
+      "learning_rate": 2e-05,
+      "loss": 0.3328,
+      "step": 3261
+    },
+    {
+      "epoch": 0.0776444825288013,
+      "grad_norm": 4.21313616959506,
+      "learning_rate": 2e-05,
+      "loss": 0.2127,
+      "step": 3262
+    },
+    {
+      "epoch": 0.07766828525183281,
+      "grad_norm": 3.9670073680111373,
+      "learning_rate": 2e-05,
+      "loss": 0.4978,
+      "step": 3263
+    },
+    {
+      "epoch": 0.07769208797486432,
+      "grad_norm": 2.8024497009382663,
+      "learning_rate": 2e-05,
+      "loss": 0.2656,
+      "step": 3264
+    },
+    {
+      "epoch": 0.07771589069789583,
+      "grad_norm": 5.705669269582298,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 3265
+    },
+    {
+      "epoch": 0.07773969342092736,
+      "grad_norm": 7.7251280793057875,
+      "learning_rate": 2e-05,
+      "loss": 1.5885,
+      "step": 3266
+    },
+    {
+      "epoch": 0.07776349614395887,
+      "grad_norm": 5.268978249490019,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 3267
+    },
+    {
+      "epoch": 0.07778729886699039,
+      "grad_norm": 15.397065389233862,
+      "learning_rate": 2e-05,
+      "loss": 1.4866,
+      "step": 3268
+    },
+    {
+      "epoch": 0.0778111015900219,
+      "grad_norm": 3.2807871959920942,
+      "learning_rate": 2e-05,
+      "loss": 0.1728,
+      "step": 3269
+    },
+    {
+      "epoch": 0.07783490431305341,
+      "grad_norm": 0.9486956455784316,
+      "learning_rate": 2e-05,
+      "loss": 0.1827,
+      "step": 3270
+    },
+    {
+      "epoch": 0.07785870703608493,
+      "grad_norm": 6.768153388068052,
+      "learning_rate": 2e-05,
+      "loss": 0.4192,
+      "step": 3271
+    },
+    {
+      "epoch": 0.07788250975911644,
+      "grad_norm": 4.745509599636779,
+      "learning_rate": 2e-05,
+      "loss": 0.751,
+      "step": 3272
+    },
+    {
+      "epoch": 0.07790631248214795,
+      "grad_norm": 4.353093846629834,
+      "learning_rate": 2e-05,
+      "loss": 0.8632,
+      "step": 3273
+    },
+    {
+      "epoch": 0.07793011520517948,
+      "grad_norm": 6.069250697716206,
+      "learning_rate": 2e-05,
+      "loss": 0.2667,
+      "step": 3274
+    },
+    {
+      "epoch": 0.07795391792821099,
+      "grad_norm": 7.5204135394786045,
+      "learning_rate": 2e-05,
+      "loss": 0.9448,
+      "step": 3275
+    },
+    {
+      "epoch": 0.0779777206512425,
+      "grad_norm": 4.7127474603755655,
+      "learning_rate": 2e-05,
+      "loss": 0.6909,
+      "step": 3276
+    },
+    {
+      "epoch": 0.07800152337427402,
+      "grad_norm": 2.355507567937556,
+      "learning_rate": 2e-05,
+      "loss": 0.2716,
+      "step": 3277
+    },
+    {
+      "epoch": 0.07802532609730553,
+      "grad_norm": 6.071932475192961,
+      "learning_rate": 2e-05,
+      "loss": 0.3307,
+      "step": 3278
+    },
+    {
+      "epoch": 0.07804912882033704,
+      "grad_norm": 6.047978005375356,
+      "learning_rate": 2e-05,
+      "loss": 0.7834,
+      "step": 3279
+    },
+    {
+      "epoch": 0.07807293154336856,
+      "grad_norm": 5.8995800108577345,
+      "learning_rate": 2e-05,
+      "loss": 0.3161,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07809673426640007,
+      "grad_norm": 5.535132948996889,
+      "learning_rate": 2e-05,
+      "loss": 0.289,
+      "step": 3281
+    },
+    {
+      "epoch": 0.0781205369894316,
+      "grad_norm": 10.245315411787203,
+      "learning_rate": 2e-05,
+      "loss": 2.0421,
+      "step": 3282
+    },
+    {
+      "epoch": 0.07814433971246311,
+      "grad_norm": 15.510842771996867,
+      "learning_rate": 2e-05,
+      "loss": 0.7513,
+      "step": 3283
+    },
+    {
+      "epoch": 0.07816814243549462,
+      "grad_norm": 3.7890306076938467,
+      "learning_rate": 2e-05,
+      "loss": 0.3949,
+      "step": 3284
+    },
+    {
+      "epoch": 0.07819194515852614,
+      "grad_norm": 6.533145835065078,
+      "learning_rate": 2e-05,
+      "loss": 1.2579,
+      "step": 3285
+    },
+    {
+      "epoch": 0.07821574788155765,
+      "grad_norm": 2.9225567070882126,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 3286
+    },
+    {
+      "epoch": 0.07823955060458916,
+      "grad_norm": 2.37355558552041,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 3287
+    },
+    {
+      "epoch": 0.07826335332762067,
+      "grad_norm": 2.426426346370771,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 3288
+    },
+    {
+      "epoch": 0.0782871560506522,
+      "grad_norm": 7.838704965816424,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 3289
+    },
+    {
+      "epoch": 0.07831095877368371,
+      "grad_norm": 5.700713692321684,
+      "learning_rate": 2e-05,
+      "loss": 0.7301,
+      "step": 3290
+    },
+    {
+      "epoch": 0.07833476149671523,
+      "grad_norm": 9.968361552925291,
+      "learning_rate": 2e-05,
+      "loss": 0.9275,
+      "step": 3291
+    },
+    {
+      "epoch": 0.07835856421974674,
+      "grad_norm": 1.9053843737651062,
+      "learning_rate": 2e-05,
+      "loss": 0.2063,
+      "step": 3292
+    },
+    {
+      "epoch": 0.07838236694277825,
+      "grad_norm": 2.3534353852308243,
+      "learning_rate": 2e-05,
+      "loss": 0.1603,
+      "step": 3293
+    },
+    {
+      "epoch": 0.07840616966580977,
+      "grad_norm": 4.799775281354782,
+      "learning_rate": 2e-05,
+      "loss": 0.7077,
+      "step": 3294
+    },
+    {
+      "epoch": 0.07842997238884128,
+      "grad_norm": 2.451682497619321,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 3295
+    },
+    {
+      "epoch": 0.07845377511187279,
+      "grad_norm": 6.271428605540201,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 3296
+    },
+    {
+      "epoch": 0.07847757783490432,
+      "grad_norm": 3.1147450132462278,
+      "learning_rate": 2e-05,
+      "loss": 0.2616,
+      "step": 3297
+    },
+    {
+      "epoch": 0.07850138055793583,
+      "grad_norm": 12.495102518488672,
+      "learning_rate": 2e-05,
+      "loss": 0.534,
+      "step": 3298
+    },
+    {
+      "epoch": 0.07852518328096735,
+      "grad_norm": 6.114337795277213,
+      "learning_rate": 2e-05,
+      "loss": 0.2356,
+      "step": 3299
+    },
+    {
+      "epoch": 0.07854898600399886,
+      "grad_norm": 5.73628428688558,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 3300
+    },
+    {
+      "epoch": 0.07857278872703037,
+      "grad_norm": 3.1632425381671143,
+      "learning_rate": 2e-05,
+      "loss": 0.2094,
+      "step": 3301
+    },
+    {
+      "epoch": 0.07859659145006188,
+      "grad_norm": 8.684159551005203,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 3302
+    },
+    {
+      "epoch": 0.0786203941730934,
+      "grad_norm": 2.8255746082940156,
+      "learning_rate": 2e-05,
+      "loss": 0.2114,
+      "step": 3303
+    },
+    {
+      "epoch": 0.07864419689612491,
+      "grad_norm": 6.291276810503535,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 3304
+    },
+    {
+      "epoch": 0.07866799961915644,
+      "grad_norm": 6.547524175807351,
+      "learning_rate": 2e-05,
+      "loss": 0.3071,
+      "step": 3305
+    },
+    {
+      "epoch": 0.07869180234218795,
+      "grad_norm": 9.525189756709919,
+      "learning_rate": 2e-05,
+      "loss": 1.2253,
+      "step": 3306
+    },
+    {
+      "epoch": 0.07871560506521946,
+      "grad_norm": 4.765280050271959,
+      "learning_rate": 2e-05,
+      "loss": 0.8197,
+      "step": 3307
+    },
+    {
+      "epoch": 0.07873940778825098,
+      "grad_norm": 3.4072003644440856,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 3308
+    },
+    {
+      "epoch": 0.07876321051128249,
+      "grad_norm": 3.77848204848549,
+      "learning_rate": 2e-05,
+      "loss": 0.2611,
+      "step": 3309
+    },
+    {
+      "epoch": 0.078787013234314,
+      "grad_norm": 4.910828243972971,
+      "learning_rate": 2e-05,
+      "loss": 0.607,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07881081595734551,
+      "grad_norm": 4.643000088473556,
+      "learning_rate": 2e-05,
+      "loss": 0.4593,
+      "step": 3311
+    },
+    {
+      "epoch": 0.07883461868037704,
+      "grad_norm": 4.004570124087365,
+      "learning_rate": 2e-05,
+      "loss": 0.755,
+      "step": 3312
+    },
+    {
+      "epoch": 0.07885842140340855,
+      "grad_norm": 4.04343404478518,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 3313
+    },
+    {
+      "epoch": 0.07888222412644007,
+      "grad_norm": 6.812534162954691,
+      "learning_rate": 2e-05,
+      "loss": 0.7672,
+      "step": 3314
+    },
+    {
+      "epoch": 0.07890602684947158,
+      "grad_norm": 5.6420853796404105,
+      "learning_rate": 2e-05,
+      "loss": 0.5958,
+      "step": 3315
+    },
+    {
+      "epoch": 0.0789298295725031,
+      "grad_norm": 6.925995201047842,
+      "learning_rate": 2e-05,
+      "loss": 0.7691,
+      "step": 3316
+    },
+    {
+      "epoch": 0.0789536322955346,
+      "grad_norm": 2.113457566853216,
+      "learning_rate": 2e-05,
+      "loss": 0.1629,
+      "step": 3317
+    },
+    {
+      "epoch": 0.07897743501856612,
+      "grad_norm": 2.6168323574298182,
+      "learning_rate": 2e-05,
+      "loss": 0.2132,
+      "step": 3318
+    },
+    {
+      "epoch": 0.07900123774159763,
+      "grad_norm": 6.879163750916024,
+      "learning_rate": 2e-05,
+      "loss": 0.3026,
+      "step": 3319
+    },
+    {
+      "epoch": 0.07902504046462916,
+      "grad_norm": 9.019306766270741,
+      "learning_rate": 2e-05,
+      "loss": 1.2204,
+      "step": 3320
+    },
+    {
+      "epoch": 0.07904884318766067,
+      "grad_norm": 2.200015871419725,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 3321
+    },
+    {
+      "epoch": 0.07907264591069219,
+      "grad_norm": 6.0313927612801175,
+      "learning_rate": 2e-05,
+      "loss": 0.6918,
+      "step": 3322
+    },
+    {
+      "epoch": 0.0790964486337237,
+      "grad_norm": 3.7877995363632304,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 3323
+    },
+    {
+      "epoch": 0.07912025135675521,
+      "grad_norm": 8.03082834953901,
+      "learning_rate": 2e-05,
+      "loss": 0.9277,
+      "step": 3324
+    },
+    {
+      "epoch": 0.07914405407978672,
+      "grad_norm": 6.531365104907055,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 3325
+    },
+    {
+      "epoch": 0.07916785680281824,
+      "grad_norm": 4.763894809589106,
+      "learning_rate": 2e-05,
+      "loss": 0.4758,
+      "step": 3326
+    },
+    {
+      "epoch": 0.07919165952584976,
+      "grad_norm": 8.417373205026122,
+      "learning_rate": 2e-05,
+      "loss": 0.3958,
+      "step": 3327
+    },
+    {
+      "epoch": 0.07921546224888128,
+      "grad_norm": 2.409424110885213,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 3328
+    },
+    {
+      "epoch": 0.07923926497191279,
+      "grad_norm": 2.375066475881937,
+      "learning_rate": 2e-05,
+      "loss": 0.2514,
+      "step": 3329
+    },
+    {
+      "epoch": 0.0792630676949443,
+      "grad_norm": 5.438043901448767,
+      "learning_rate": 2e-05,
+      "loss": 0.6624,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07928687041797582,
+      "grad_norm": 8.64372484516493,
+      "learning_rate": 2e-05,
+      "loss": 0.4774,
+      "step": 3331
+    },
+    {
+      "epoch": 0.07931067314100733,
+      "grad_norm": 5.443734552569742,
+      "learning_rate": 2e-05,
+      "loss": 0.6832,
+      "step": 3332
+    },
+    {
+      "epoch": 0.07933447586403884,
+      "grad_norm": 7.1512976370011385,
+      "learning_rate": 2e-05,
+      "loss": 1.1879,
+      "step": 3333
+    },
+    {
+      "epoch": 0.07935827858707035,
+      "grad_norm": 3.029742040009381,
+      "learning_rate": 2e-05,
+      "loss": 0.2088,
+      "step": 3334
+    },
+    {
+      "epoch": 0.07938208131010188,
+      "grad_norm": 4.702269453983167,
+      "learning_rate": 2e-05,
+      "loss": 0.6391,
+      "step": 3335
+    },
+    {
+      "epoch": 0.0794058840331334,
+      "grad_norm": 5.998596676974773,
+      "learning_rate": 2e-05,
+      "loss": 0.874,
+      "step": 3336
+    },
+    {
+      "epoch": 0.07942968675616491,
+      "grad_norm": 3.7594355958536103,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 3337
+    },
+    {
+      "epoch": 0.07945348947919642,
+      "grad_norm": 3.2315640395298852,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 3338
+    },
+    {
+      "epoch": 0.07947729220222793,
+      "grad_norm": 6.027190778271744,
+      "learning_rate": 2e-05,
+      "loss": 0.3268,
+      "step": 3339
+    },
+    {
+      "epoch": 0.07950109492525945,
+      "grad_norm": 6.989192499418722,
+      "learning_rate": 2e-05,
+      "loss": 1.4109,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07952489764829096,
+      "grad_norm": 6.692711934751686,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 3341
+    },
+    {
+      "epoch": 0.07954870037132247,
+      "grad_norm": 6.089822951442845,
+      "learning_rate": 2e-05,
+      "loss": 0.5724,
+      "step": 3342
+    },
+    {
+      "epoch": 0.079572503094354,
+      "grad_norm": 2.582466521254786,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 3343
+    },
+    {
+      "epoch": 0.07959630581738551,
+      "grad_norm": 5.2299018802615755,
+      "learning_rate": 2e-05,
+      "loss": 0.7684,
+      "step": 3344
+    },
+    {
+      "epoch": 0.07962010854041703,
+      "grad_norm": 7.652686418031736,
+      "learning_rate": 2e-05,
+      "loss": 0.9897,
+      "step": 3345
+    },
+    {
+      "epoch": 0.07964391126344854,
+      "grad_norm": 10.063573907838803,
+      "learning_rate": 2e-05,
+      "loss": 0.3462,
+      "step": 3346
+    },
+    {
+      "epoch": 0.07966771398648005,
+      "grad_norm": 3.018940291665601,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 3347
+    },
+    {
+      "epoch": 0.07969151670951156,
+      "grad_norm": 10.560690368466203,
+      "learning_rate": 2e-05,
+      "loss": 0.6292,
+      "step": 3348
+    },
+    {
+      "epoch": 0.07971531943254308,
+      "grad_norm": 7.07265152390263,
+      "learning_rate": 2e-05,
+      "loss": 0.6286,
+      "step": 3349
+    },
+    {
+      "epoch": 0.0797391221555746,
+      "grad_norm": 5.881596192434305,
+      "learning_rate": 2e-05,
+      "loss": 1.003,
+      "step": 3350
+    },
+    {
+      "epoch": 0.07976292487860612,
+      "grad_norm": 2.578800361557132,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 3351
+    },
+    {
+      "epoch": 0.07978672760163763,
+      "grad_norm": 3.0687798419146692,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 3352
+    },
+    {
+      "epoch": 0.07981053032466914,
+      "grad_norm": 7.658129610937155,
+      "learning_rate": 2e-05,
+      "loss": 0.6226,
+      "step": 3353
+    },
+    {
+      "epoch": 0.07983433304770066,
+      "grad_norm": 8.821061515205926,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 3354
+    },
+    {
+      "epoch": 0.07985813577073217,
+      "grad_norm": 6.841508704665747,
+      "learning_rate": 2e-05,
+      "loss": 0.292,
+      "step": 3355
+    },
+    {
+      "epoch": 0.07988193849376368,
+      "grad_norm": 2.6190083598236185,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 3356
+    },
+    {
+      "epoch": 0.0799057412167952,
+      "grad_norm": 8.191046820213499,
+      "learning_rate": 2e-05,
+      "loss": 0.4357,
+      "step": 3357
+    },
+    {
+      "epoch": 0.07992954393982672,
+      "grad_norm": 6.53548019541113,
+      "learning_rate": 2e-05,
+      "loss": 0.5699,
+      "step": 3358
+    },
+    {
+      "epoch": 0.07995334666285824,
+      "grad_norm": 6.890839812095456,
+      "learning_rate": 2e-05,
+      "loss": 0.3232,
+      "step": 3359
+    },
+    {
+      "epoch": 0.07997714938588975,
+      "grad_norm": 5.510282372787868,
+      "learning_rate": 2e-05,
+      "loss": 0.3754,
+      "step": 3360
+    },
+    {
+      "epoch": 0.08000095210892126,
+      "grad_norm": 7.986181387746643,
+      "learning_rate": 2e-05,
+      "loss": 0.4305,
+      "step": 3361
+    },
+    {
+      "epoch": 0.08002475483195277,
+      "grad_norm": 2.0069188318022264,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 3362
+    },
+    {
+      "epoch": 0.08004855755498429,
+      "grad_norm": 2.9138972917408457,
+      "learning_rate": 2e-05,
+      "loss": 0.2535,
+      "step": 3363
+    },
+    {
+      "epoch": 0.0800723602780158,
+      "grad_norm": 6.565283095929466,
+      "learning_rate": 2e-05,
+      "loss": 0.2718,
+      "step": 3364
+    },
+    {
+      "epoch": 0.08009616300104731,
+      "grad_norm": 2.4099925648985456,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 3365
+    },
+    {
+      "epoch": 0.08011996572407884,
+      "grad_norm": 5.433841438138322,
+      "learning_rate": 2e-05,
+      "loss": 0.8406,
+      "step": 3366
+    },
+    {
+      "epoch": 0.08014376844711035,
+      "grad_norm": 5.8167965218994135,
+      "learning_rate": 2e-05,
+      "loss": 0.9349,
+      "step": 3367
+    },
+    {
+      "epoch": 0.08016757117014187,
+      "grad_norm": 8.806443418704959,
+      "learning_rate": 2e-05,
+      "loss": 1.9123,
+      "step": 3368
+    },
+    {
+      "epoch": 0.08019137389317338,
+      "grad_norm": 2.468419330335678,
+      "learning_rate": 2e-05,
+      "loss": 0.2069,
+      "step": 3369
+    },
+    {
+      "epoch": 0.08021517661620489,
+      "grad_norm": 2.945355682333883,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 3370
+    },
+    {
+      "epoch": 0.0802389793392364,
+      "grad_norm": 7.169724812419689,
+      "learning_rate": 2e-05,
+      "loss": 0.5119,
+      "step": 3371
+    },
+    {
+      "epoch": 0.08026278206226792,
+      "grad_norm": 4.642587070703176,
+      "learning_rate": 2e-05,
+      "loss": 0.5805,
+      "step": 3372
+    },
+    {
+      "epoch": 0.08028658478529944,
+      "grad_norm": 5.19868201080272,
+      "learning_rate": 2e-05,
+      "loss": 0.926,
+      "step": 3373
+    },
+    {
+      "epoch": 0.08031038750833096,
+      "grad_norm": 6.4430135007867255,
+      "learning_rate": 2e-05,
+      "loss": 0.5974,
+      "step": 3374
+    },
+    {
+      "epoch": 0.08033419023136247,
+      "grad_norm": 3.404365929351331,
+      "learning_rate": 2e-05,
+      "loss": 0.3321,
+      "step": 3375
+    },
+    {
+      "epoch": 0.08035799295439398,
+      "grad_norm": 4.272659901414358,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 3376
+    },
+    {
+      "epoch": 0.0803817956774255,
+      "grad_norm": 7.517952289986932,
+      "learning_rate": 2e-05,
+      "loss": 0.3928,
+      "step": 3377
+    },
+    {
+      "epoch": 0.08040559840045701,
+      "grad_norm": 15.392298428207848,
+      "learning_rate": 2e-05,
+      "loss": 1.0685,
+      "step": 3378
+    },
+    {
+      "epoch": 0.08042940112348852,
+      "grad_norm": 8.438146658216452,
+      "learning_rate": 2e-05,
+      "loss": 1.2069,
+      "step": 3379
+    },
+    {
+      "epoch": 0.08045320384652004,
+      "grad_norm": 5.739353160063187,
+      "learning_rate": 2e-05,
+      "loss": 1.2018,
+      "step": 3380
+    },
+    {
+      "epoch": 0.08047700656955156,
+      "grad_norm": 5.421238188855623,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 3381
+    },
+    {
+      "epoch": 0.08050080929258308,
+      "grad_norm": 8.6695517382901,
+      "learning_rate": 2e-05,
+      "loss": 0.9675,
+      "step": 3382
+    },
+    {
+      "epoch": 0.08052461201561459,
+      "grad_norm": 10.92805736403499,
+      "learning_rate": 2e-05,
+      "loss": 1.1532,
+      "step": 3383
+    },
+    {
+      "epoch": 0.0805484147386461,
+      "grad_norm": 5.230395167528913,
+      "learning_rate": 2e-05,
+      "loss": 0.7484,
+      "step": 3384
+    },
+    {
+      "epoch": 0.08057221746167761,
+      "grad_norm": 9.385180483598754,
+      "learning_rate": 2e-05,
+      "loss": 1.3844,
+      "step": 3385
+    },
+    {
+      "epoch": 0.08059602018470913,
+      "grad_norm": 9.124589969022809,
+      "learning_rate": 2e-05,
+      "loss": 0.9682,
+      "step": 3386
+    },
+    {
+      "epoch": 0.08061982290774064,
+      "grad_norm": 4.572166599871773,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 3387
+    },
+    {
+      "epoch": 0.08064362563077217,
+      "grad_norm": 5.56064941616364,
+      "learning_rate": 2e-05,
+      "loss": 0.3559,
+      "step": 3388
+    },
+    {
+      "epoch": 0.08066742835380368,
+      "grad_norm": 2.4145052106308054,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 3389
+    },
+    {
+      "epoch": 0.0806912310768352,
+      "grad_norm": 6.51519036507726,
+      "learning_rate": 2e-05,
+      "loss": 0.7387,
+      "step": 3390
+    },
+    {
+      "epoch": 0.0807150337998667,
+      "grad_norm": 12.996508576607374,
+      "learning_rate": 2e-05,
+      "loss": 0.9949,
+      "step": 3391
+    },
+    {
+      "epoch": 0.08073883652289822,
+      "grad_norm": 5.862668037088078,
+      "learning_rate": 2e-05,
+      "loss": 0.8246,
+      "step": 3392
+    },
+    {
+      "epoch": 0.08076263924592973,
+      "grad_norm": 5.721551160522021,
+      "learning_rate": 2e-05,
+      "loss": 0.3744,
+      "step": 3393
+    },
+    {
+      "epoch": 0.08078644196896125,
+      "grad_norm": 1.9030030668177031,
+      "learning_rate": 2e-05,
+      "loss": 0.2495,
+      "step": 3394
+    },
+    {
+      "epoch": 0.08081024469199276,
+      "grad_norm": 8.037336526745012,
+      "learning_rate": 2e-05,
+      "loss": 0.7519,
+      "step": 3395
+    },
+    {
+      "epoch": 0.08083404741502428,
+      "grad_norm": 2.8072601686048198,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 3396
+    },
+    {
+      "epoch": 0.0808578501380558,
+      "grad_norm": 12.102982971816974,
+      "learning_rate": 2e-05,
+      "loss": 0.9867,
+      "step": 3397
+    },
+    {
+      "epoch": 0.08088165286108731,
+      "grad_norm": 4.04691281282623,
+      "learning_rate": 2e-05,
+      "loss": 0.2762,
+      "step": 3398
+    },
+    {
+      "epoch": 0.08090545558411882,
+      "grad_norm": 6.683504410563889,
+      "learning_rate": 2e-05,
+      "loss": 0.2681,
+      "step": 3399
+    },
+    {
+      "epoch": 0.08092925830715034,
+      "grad_norm": 1.7140405394943434,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08095306103018185,
+      "grad_norm": 16.07721761362571,
+      "learning_rate": 2e-05,
+      "loss": 1.1346,
+      "step": 3401
+    },
+    {
+      "epoch": 0.08097686375321336,
+      "grad_norm": 5.761668227697725,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 3402
+    },
+    {
+      "epoch": 0.08100066647624488,
+      "grad_norm": 8.808832935266903,
+      "learning_rate": 2e-05,
+      "loss": 1.331,
+      "step": 3403
+    },
+    {
+      "epoch": 0.0810244691992764,
+      "grad_norm": 6.228288814315912,
+      "learning_rate": 2e-05,
+      "loss": 0.3904,
+      "step": 3404
+    },
+    {
+      "epoch": 0.08104827192230792,
+      "grad_norm": 4.437348554450726,
+      "learning_rate": 2e-05,
+      "loss": 0.8941,
+      "step": 3405
+    },
+    {
+      "epoch": 0.08107207464533943,
+      "grad_norm": 4.884243036679516,
+      "learning_rate": 2e-05,
+      "loss": 0.2431,
+      "step": 3406
+    },
+    {
+      "epoch": 0.08109587736837094,
+      "grad_norm": 4.260855733207124,
+      "learning_rate": 2e-05,
+      "loss": 0.2917,
+      "step": 3407
+    },
+    {
+      "epoch": 0.08111968009140245,
+      "grad_norm": 3.47009766825757,
+      "learning_rate": 2e-05,
+      "loss": 0.2376,
+      "step": 3408
+    },
+    {
+      "epoch": 0.08114348281443397,
+      "grad_norm": 3.179594304909714,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 3409
+    },
+    {
+      "epoch": 0.08116728553746548,
+      "grad_norm": 1.885447284659232,
+      "learning_rate": 2e-05,
+      "loss": 0.1885,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08119108826049701,
+      "grad_norm": 6.153282144142426,
+      "learning_rate": 2e-05,
+      "loss": 0.3529,
+      "step": 3411
+    },
+    {
+      "epoch": 0.08121489098352852,
+      "grad_norm": 3.0336664309395047,
+      "learning_rate": 2e-05,
+      "loss": 0.3402,
+      "step": 3412
+    },
+    {
+      "epoch": 0.08123869370656003,
+      "grad_norm": 2.1689990992976584,
+      "learning_rate": 2e-05,
+      "loss": 0.2071,
+      "step": 3413
+    },
+    {
+      "epoch": 0.08126249642959155,
+      "grad_norm": 7.3204151616684525,
+      "learning_rate": 2e-05,
+      "loss": 1.2342,
+      "step": 3414
+    },
+    {
+      "epoch": 0.08128629915262306,
+      "grad_norm": 4.678498654352358,
+      "learning_rate": 2e-05,
+      "loss": 0.7438,
+      "step": 3415
+    },
+    {
+      "epoch": 0.08131010187565457,
+      "grad_norm": 7.768391940804872,
+      "learning_rate": 2e-05,
+      "loss": 1.0105,
+      "step": 3416
+    },
+    {
+      "epoch": 0.08133390459868609,
+      "grad_norm": 5.106556924864265,
+      "learning_rate": 2e-05,
+      "loss": 0.5223,
+      "step": 3417
+    },
+    {
+      "epoch": 0.0813577073217176,
+      "grad_norm": 4.921691480104808,
+      "learning_rate": 2e-05,
+      "loss": 0.5197,
+      "step": 3418
+    },
+    {
+      "epoch": 0.08138151004474913,
+      "grad_norm": 3.7094639546584,
+      "learning_rate": 2e-05,
+      "loss": 0.2479,
+      "step": 3419
+    },
+    {
+      "epoch": 0.08140531276778064,
+      "grad_norm": 9.763277488269427,
+      "learning_rate": 2e-05,
+      "loss": 0.3443,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08142911549081215,
+      "grad_norm": 5.3277451896513215,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 3421
+    },
+    {
+      "epoch": 0.08145291821384366,
+      "grad_norm": 5.547731757193517,
+      "learning_rate": 2e-05,
+      "loss": 0.3193,
+      "step": 3422
+    },
+    {
+      "epoch": 0.08147672093687518,
+      "grad_norm": 2.0205216693682964,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 3423
+    },
+    {
+      "epoch": 0.08150052365990669,
+      "grad_norm": 3.831002855269496,
+      "learning_rate": 2e-05,
+      "loss": 0.3606,
+      "step": 3424
+    },
+    {
+      "epoch": 0.0815243263829382,
+      "grad_norm": 3.3424466754410873,
+      "learning_rate": 2e-05,
+      "loss": 0.1904,
+      "step": 3425
+    },
+    {
+      "epoch": 0.08154812910596972,
+      "grad_norm": 7.762370902836087,
+      "learning_rate": 2e-05,
+      "loss": 0.4621,
+      "step": 3426
+    },
+    {
+      "epoch": 0.08157193182900124,
+      "grad_norm": 7.364367193894512,
+      "learning_rate": 2e-05,
+      "loss": 0.9606,
+      "step": 3427
+    },
+    {
+      "epoch": 0.08159573455203276,
+      "grad_norm": 6.1621012926502505,
+      "learning_rate": 2e-05,
+      "loss": 0.9065,
+      "step": 3428
+    },
+    {
+      "epoch": 0.08161953727506427,
+      "grad_norm": 3.5163797629892217,
+      "learning_rate": 2e-05,
+      "loss": 0.1738,
+      "step": 3429
+    },
+    {
+      "epoch": 0.08164333999809578,
+      "grad_norm": 4.032433780041825,
+      "learning_rate": 2e-05,
+      "loss": 0.2854,
+      "step": 3430
+    },
+    {
+      "epoch": 0.0816671427211273,
+      "grad_norm": 7.393740126417192,
+      "learning_rate": 2e-05,
+      "loss": 1.2344,
+      "step": 3431
+    },
+    {
+      "epoch": 0.08169094544415881,
+      "grad_norm": 6.239485368642886,
+      "learning_rate": 2e-05,
+      "loss": 0.4091,
+      "step": 3432
+    },
+    {
+      "epoch": 0.08171474816719032,
+      "grad_norm": 3.1296364568024737,
+      "learning_rate": 2e-05,
+      "loss": 0.2484,
+      "step": 3433
+    },
+    {
+      "epoch": 0.08173855089022185,
+      "grad_norm": 4.986505456053209,
+      "learning_rate": 2e-05,
+      "loss": 0.3389,
+      "step": 3434
+    },
+    {
+      "epoch": 0.08176235361325336,
+      "grad_norm": 4.762730362167197,
+      "learning_rate": 2e-05,
+      "loss": 0.344,
+      "step": 3435
+    },
+    {
+      "epoch": 0.08178615633628487,
+      "grad_norm": 2.4346903129442667,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 3436
+    },
+    {
+      "epoch": 0.08180995905931639,
+      "grad_norm": 9.47359126118461,
+      "learning_rate": 2e-05,
+      "loss": 1.3871,
+      "step": 3437
+    },
+    {
+      "epoch": 0.0818337617823479,
+      "grad_norm": 4.793718357399366,
+      "learning_rate": 2e-05,
+      "loss": 0.3096,
+      "step": 3438
+    },
+    {
+      "epoch": 0.08185756450537941,
+      "grad_norm": 9.13051386095302,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 3439
+    },
+    {
+      "epoch": 0.08188136722841093,
+      "grad_norm": 2.489150513524738,
+      "learning_rate": 2e-05,
+      "loss": 0.2519,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08190516995144244,
+      "grad_norm": 6.177098065970457,
+      "learning_rate": 2e-05,
+      "loss": 0.371,
+      "step": 3441
+    },
+    {
+      "epoch": 0.08192897267447397,
+      "grad_norm": 6.276683677878656,
+      "learning_rate": 2e-05,
+      "loss": 0.3243,
+      "step": 3442
+    },
+    {
+      "epoch": 0.08195277539750548,
+      "grad_norm": 7.842131821647195,
+      "learning_rate": 2e-05,
+      "loss": 1.2825,
+      "step": 3443
+    },
+    {
+      "epoch": 0.08197657812053699,
+      "grad_norm": 8.81974762004923,
+      "learning_rate": 2e-05,
+      "loss": 1.1715,
+      "step": 3444
+    },
+    {
+      "epoch": 0.0820003808435685,
+      "grad_norm": 3.507797843267766,
+      "learning_rate": 2e-05,
+      "loss": 0.2488,
+      "step": 3445
+    },
+    {
+      "epoch": 0.08202418356660002,
+      "grad_norm": 6.624357256162843,
+      "learning_rate": 2e-05,
+      "loss": 1.0338,
+      "step": 3446
+    },
+    {
+      "epoch": 0.08204798628963153,
+      "grad_norm": 10.617574273829717,
+      "learning_rate": 2e-05,
+      "loss": 1.0145,
+      "step": 3447
+    },
+    {
+      "epoch": 0.08207178901266304,
+      "grad_norm": 2.789057729981666,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 3448
+    },
+    {
+      "epoch": 0.08209559173569457,
+      "grad_norm": 4.270432807821059,
+      "learning_rate": 2e-05,
+      "loss": 0.2727,
+      "step": 3449
+    },
+    {
+      "epoch": 0.08211939445872608,
+      "grad_norm": 13.424615118031738,
+      "learning_rate": 2e-05,
+      "loss": 0.5645,
+      "step": 3450
+    },
+    {
+      "epoch": 0.0821431971817576,
+      "grad_norm": 5.659272583458696,
+      "learning_rate": 2e-05,
+      "loss": 0.9061,
+      "step": 3451
+    },
+    {
+      "epoch": 0.08216699990478911,
+      "grad_norm": 2.207216453164425,
+      "learning_rate": 2e-05,
+      "loss": 0.1959,
+      "step": 3452
+    },
+    {
+      "epoch": 0.08219080262782062,
+      "grad_norm": 7.230857914645199,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 3453
+    },
+    {
+      "epoch": 0.08221460535085214,
+      "grad_norm": 1.7620923068615928,
+      "learning_rate": 2e-05,
+      "loss": 0.1784,
+      "step": 3454
+    },
+    {
+      "epoch": 0.08223840807388365,
+      "grad_norm": 2.43674949899827,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 3455
+    },
+    {
+      "epoch": 0.08226221079691516,
+      "grad_norm": 35.20704840097625,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 3456
+    },
+    {
+      "epoch": 0.08228601351994669,
+      "grad_norm": 2.7857615556854265,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 3457
+    },
+    {
+      "epoch": 0.0823098162429782,
+      "grad_norm": 8.336620421208988,
+      "learning_rate": 2e-05,
+      "loss": 1.5507,
+      "step": 3458
+    },
+    {
+      "epoch": 0.08233361896600971,
+      "grad_norm": 7.120312460541482,
+      "learning_rate": 2e-05,
+      "loss": 0.2876,
+      "step": 3459
+    },
+    {
+      "epoch": 0.08235742168904123,
+      "grad_norm": 8.461157725653983,
+      "learning_rate": 2e-05,
+      "loss": 0.9058,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08238122441207274,
+      "grad_norm": 9.401705647236565,
+      "learning_rate": 2e-05,
+      "loss": 0.8777,
+      "step": 3461
+    },
+    {
+      "epoch": 0.08240502713510425,
+      "grad_norm": 11.21550594345508,
+      "learning_rate": 2e-05,
+      "loss": 1.4047,
+      "step": 3462
+    },
+    {
+      "epoch": 0.08242882985813577,
+      "grad_norm": 8.359013914837345,
+      "learning_rate": 2e-05,
+      "loss": 0.5242,
+      "step": 3463
+    },
+    {
+      "epoch": 0.08245263258116728,
+      "grad_norm": 3.9891794418728366,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 3464
+    },
+    {
+      "epoch": 0.0824764353041988,
+      "grad_norm": 3.9501385197267873,
+      "learning_rate": 2e-05,
+      "loss": 0.3278,
+      "step": 3465
+    },
+    {
+      "epoch": 0.08250023802723032,
+      "grad_norm": 2.9151703211442133,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 3466
+    },
+    {
+      "epoch": 0.08252404075026183,
+      "grad_norm": 6.270479922962324,
+      "learning_rate": 2e-05,
+      "loss": 0.9294,
+      "step": 3467
+    },
+    {
+      "epoch": 0.08254784347329334,
+      "grad_norm": 10.334828008425323,
+      "learning_rate": 2e-05,
+      "loss": 1.3109,
+      "step": 3468
+    },
+    {
+      "epoch": 0.08257164619632486,
+      "grad_norm": 3.8850504131880954,
+      "learning_rate": 2e-05,
+      "loss": 0.3384,
+      "step": 3469
+    },
+    {
+      "epoch": 0.08259544891935637,
+      "grad_norm": 2.1696260568273553,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08261925164238788,
+      "grad_norm": 8.49834732373925,
+      "learning_rate": 2e-05,
+      "loss": 1.0628,
+      "step": 3471
+    },
+    {
+      "epoch": 0.08264305436541941,
+      "grad_norm": 5.5720573758270415,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 3472
+    },
+    {
+      "epoch": 0.08266685708845092,
+      "grad_norm": 4.703068116385984,
+      "learning_rate": 2e-05,
+      "loss": 0.4887,
+      "step": 3473
+    },
+    {
+      "epoch": 0.08269065981148244,
+      "grad_norm": 2.0633734406711093,
+      "learning_rate": 2e-05,
+      "loss": 0.1861,
+      "step": 3474
+    },
+    {
+      "epoch": 0.08271446253451395,
+      "grad_norm": 5.1168088174080495,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 3475
+    },
+    {
+      "epoch": 0.08273826525754546,
+      "grad_norm": 5.35373374038817,
+      "learning_rate": 2e-05,
+      "loss": 0.5909,
+      "step": 3476
+    },
+    {
+      "epoch": 0.08276206798057698,
+      "grad_norm": 6.204472317106088,
+      "learning_rate": 2e-05,
+      "loss": 0.2766,
+      "step": 3477
+    },
+    {
+      "epoch": 0.08278587070360849,
+      "grad_norm": 3.345033545405001,
+      "learning_rate": 2e-05,
+      "loss": 0.5952,
+      "step": 3478
+    },
+    {
+      "epoch": 0.08280967342664,
+      "grad_norm": 2.395696219594056,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 3479
+    },
+    {
+      "epoch": 0.08283347614967153,
+      "grad_norm": 3.152780889780851,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08285727887270304,
+      "grad_norm": 4.680514459709189,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 3481
+    },
+    {
+      "epoch": 0.08288108159573455,
+      "grad_norm": 6.197136726226889,
+      "learning_rate": 2e-05,
+      "loss": 0.2751,
+      "step": 3482
+    },
+    {
+      "epoch": 0.08290488431876607,
+      "grad_norm": 6.168521693434393,
+      "learning_rate": 2e-05,
+      "loss": 1.123,
+      "step": 3483
+    },
+    {
+      "epoch": 0.08292868704179758,
+      "grad_norm": 2.142683849496759,
+      "learning_rate": 2e-05,
+      "loss": 0.1543,
+      "step": 3484
+    },
+    {
+      "epoch": 0.0829524897648291,
+      "grad_norm": 2.4958726347590607,
+      "learning_rate": 2e-05,
+      "loss": 0.2151,
+      "step": 3485
+    },
+    {
+      "epoch": 0.0829762924878606,
+      "grad_norm": 5.780556504251436,
+      "learning_rate": 2e-05,
+      "loss": 0.2589,
+      "step": 3486
+    },
+    {
+      "epoch": 0.08300009521089212,
+      "grad_norm": 5.903944847361858,
+      "learning_rate": 2e-05,
+      "loss": 0.7959,
+      "step": 3487
+    },
+    {
+      "epoch": 0.08302389793392365,
+      "grad_norm": 8.477839431738126,
+      "learning_rate": 2e-05,
+      "loss": 0.4896,
+      "step": 3488
+    },
+    {
+      "epoch": 0.08304770065695516,
+      "grad_norm": 6.73921573904242,
+      "learning_rate": 2e-05,
+      "loss": 0.9767,
+      "step": 3489
+    },
+    {
+      "epoch": 0.08307150337998667,
+      "grad_norm": 1.7621084779142506,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08309530610301818,
+      "grad_norm": 11.168257695170112,
+      "learning_rate": 2e-05,
+      "loss": 1.0303,
+      "step": 3491
+    },
+    {
+      "epoch": 0.0831191088260497,
+      "grad_norm": 8.249044770724392,
+      "learning_rate": 2e-05,
+      "loss": 1.1238,
+      "step": 3492
+    },
+    {
+      "epoch": 0.08314291154908121,
+      "grad_norm": 2.9843792973413024,
+      "learning_rate": 2e-05,
+      "loss": 0.2222,
+      "step": 3493
+    },
+    {
+      "epoch": 0.08316671427211272,
+      "grad_norm": 14.629930920864618,
+      "learning_rate": 2e-05,
+      "loss": 0.7518,
+      "step": 3494
+    },
+    {
+      "epoch": 0.08319051699514425,
+      "grad_norm": 4.13033556399775,
+      "learning_rate": 2e-05,
+      "loss": 0.4658,
+      "step": 3495
+    },
+    {
+      "epoch": 0.08321431971817576,
+      "grad_norm": 6.598293474674274,
+      "learning_rate": 2e-05,
+      "loss": 0.3558,
+      "step": 3496
+    },
+    {
+      "epoch": 0.08323812244120728,
+      "grad_norm": 2.7316861758357383,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 3497
+    },
+    {
+      "epoch": 0.08326192516423879,
+      "grad_norm": 3.780430006236144,
+      "learning_rate": 2e-05,
+      "loss": 0.4945,
+      "step": 3498
+    },
+    {
+      "epoch": 0.0832857278872703,
+      "grad_norm": 16.313473031909773,
+      "learning_rate": 2e-05,
+      "loss": 1.142,
+      "step": 3499
+    },
+    {
+      "epoch": 0.08330953061030182,
+      "grad_norm": 4.316225769330245,
+      "learning_rate": 2e-05,
+      "loss": 0.3797,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 6.310691118384085,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 3501
+    },
+    {
+      "epoch": 0.08335713605636484,
+      "grad_norm": 6.570184057144011,
+      "learning_rate": 2e-05,
+      "loss": 0.3599,
+      "step": 3502
+    },
+    {
+      "epoch": 0.08338093877939637,
+      "grad_norm": 8.21424290106511,
+      "learning_rate": 2e-05,
+      "loss": 0.4523,
+      "step": 3503
+    },
+    {
+      "epoch": 0.08340474150242788,
+      "grad_norm": 5.432742603398238,
+      "learning_rate": 2e-05,
+      "loss": 0.7982,
+      "step": 3504
+    },
+    {
+      "epoch": 0.0834285442254594,
+      "grad_norm": 8.255193483629212,
+      "learning_rate": 2e-05,
+      "loss": 1.0684,
+      "step": 3505
+    },
+    {
+      "epoch": 0.08345234694849091,
+      "grad_norm": 5.119927081565769,
+      "learning_rate": 2e-05,
+      "loss": 0.266,
+      "step": 3506
+    },
+    {
+      "epoch": 0.08347614967152242,
+      "grad_norm": 17.08480779953454,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 3507
+    },
+    {
+      "epoch": 0.08349995239455393,
+      "grad_norm": 5.549691658029839,
+      "learning_rate": 2e-05,
+      "loss": 0.4195,
+      "step": 3508
+    },
+    {
+      "epoch": 0.08352375511758545,
+      "grad_norm": 1.8877410603217628,
+      "learning_rate": 2e-05,
+      "loss": 0.1543,
+      "step": 3509
+    },
+    {
+      "epoch": 0.08354755784061697,
+      "grad_norm": 5.271629461266853,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 3510
+    },
+    {
+      "epoch": 0.08357136056364849,
+      "grad_norm": 5.790329141234427,
+      "learning_rate": 2e-05,
+      "loss": 0.53,
+      "step": 3511
+    },
+    {
+      "epoch": 0.08359516328668,
+      "grad_norm": 2.2837733602836847,
+      "learning_rate": 2e-05,
+      "loss": 0.236,
+      "step": 3512
+    },
+    {
+      "epoch": 0.08361896600971151,
+      "grad_norm": 3.158527407567413,
+      "learning_rate": 2e-05,
+      "loss": 0.3274,
+      "step": 3513
+    },
+    {
+      "epoch": 0.08364276873274303,
+      "grad_norm": 3.4388375612773143,
+      "learning_rate": 2e-05,
+      "loss": 0.2817,
+      "step": 3514
+    },
+    {
+      "epoch": 0.08366657145577454,
+      "grad_norm": 13.17367128664752,
+      "learning_rate": 2e-05,
+      "loss": 1.2792,
+      "step": 3515
+    },
+    {
+      "epoch": 0.08369037417880605,
+      "grad_norm": 3.6168321520559017,
+      "learning_rate": 2e-05,
+      "loss": 0.4394,
+      "step": 3516
+    },
+    {
+      "epoch": 0.08371417690183756,
+      "grad_norm": 3.7918652043002417,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 3517
+    },
+    {
+      "epoch": 0.08373797962486909,
+      "grad_norm": 14.767134814791229,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 3518
+    },
+    {
+      "epoch": 0.0837617823479006,
+      "grad_norm": 8.509532741469634,
+      "learning_rate": 2e-05,
+      "loss": 0.7326,
+      "step": 3519
+    },
+    {
+      "epoch": 0.08378558507093212,
+      "grad_norm": 10.643816774790047,
+      "learning_rate": 2e-05,
+      "loss": 0.8729,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08380938779396363,
+      "grad_norm": 3.387755735864147,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 3521
+    },
+    {
+      "epoch": 0.08383319051699514,
+      "grad_norm": 8.171518495847732,
+      "learning_rate": 2e-05,
+      "loss": 0.4136,
+      "step": 3522
+    },
+    {
+      "epoch": 0.08385699324002666,
+      "grad_norm": 2.021943906066322,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 3523
+    },
+    {
+      "epoch": 0.08388079596305817,
+      "grad_norm": 5.384932963997437,
+      "learning_rate": 2e-05,
+      "loss": 0.3214,
+      "step": 3524
+    },
+    {
+      "epoch": 0.08390459868608968,
+      "grad_norm": 6.915415014180779,
+      "learning_rate": 2e-05,
+      "loss": 0.433,
+      "step": 3525
+    },
+    {
+      "epoch": 0.08392840140912121,
+      "grad_norm": 2.4404780025251323,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 3526
+    },
+    {
+      "epoch": 0.08395220413215272,
+      "grad_norm": 3.1105353919680643,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 3527
+    },
+    {
+      "epoch": 0.08397600685518423,
+      "grad_norm": 8.050902136623629,
+      "learning_rate": 2e-05,
+      "loss": 1.2629,
+      "step": 3528
+    },
+    {
+      "epoch": 0.08399980957821575,
+      "grad_norm": 7.4752049967583565,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 3529
+    },
+    {
+      "epoch": 0.08402361230124726,
+      "grad_norm": 4.84898255980558,
+      "learning_rate": 2e-05,
+      "loss": 0.5546,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08404741502427877,
+      "grad_norm": 4.561943010871363,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 3531
+    },
+    {
+      "epoch": 0.08407121774731029,
+      "grad_norm": 7.292769943574225,
+      "learning_rate": 2e-05,
+      "loss": 0.3041,
+      "step": 3532
+    },
+    {
+      "epoch": 0.08409502047034181,
+      "grad_norm": 3.6961117812734976,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 3533
+    },
+    {
+      "epoch": 0.08411882319337333,
+      "grad_norm": 6.592674927446499,
+      "learning_rate": 2e-05,
+      "loss": 0.4634,
+      "step": 3534
+    },
+    {
+      "epoch": 0.08414262591640484,
+      "grad_norm": 8.074423560404377,
+      "learning_rate": 2e-05,
+      "loss": 1.0074,
+      "step": 3535
+    },
+    {
+      "epoch": 0.08416642863943635,
+      "grad_norm": 3.299149923100008,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 3536
+    },
+    {
+      "epoch": 0.08419023136246787,
+      "grad_norm": 5.019165752126257,
+      "learning_rate": 2e-05,
+      "loss": 0.781,
+      "step": 3537
+    },
+    {
+      "epoch": 0.08421403408549938,
+      "grad_norm": 7.796511352160003,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 3538
+    },
+    {
+      "epoch": 0.08423783680853089,
+      "grad_norm": 7.784168050524831,
+      "learning_rate": 2e-05,
+      "loss": 0.9565,
+      "step": 3539
+    },
+    {
+      "epoch": 0.0842616395315624,
+      "grad_norm": 3.5614858682953754,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08428544225459393,
+      "grad_norm": 5.462069726944584,
+      "learning_rate": 2e-05,
+      "loss": 0.6087,
+      "step": 3541
+    },
+    {
+      "epoch": 0.08430924497762544,
+      "grad_norm": 5.872181264280343,
+      "learning_rate": 2e-05,
+      "loss": 1.0069,
+      "step": 3542
+    },
+    {
+      "epoch": 0.08433304770065696,
+      "grad_norm": 6.339895698352778,
+      "learning_rate": 2e-05,
+      "loss": 0.3129,
+      "step": 3543
+    },
+    {
+      "epoch": 0.08435685042368847,
+      "grad_norm": 2.264648370118872,
+      "learning_rate": 2e-05,
+      "loss": 0.1626,
+      "step": 3544
+    },
+    {
+      "epoch": 0.08438065314671998,
+      "grad_norm": 5.568229990505039,
+      "learning_rate": 2e-05,
+      "loss": 0.5551,
+      "step": 3545
+    },
+    {
+      "epoch": 0.0844044558697515,
+      "grad_norm": 2.904614112917955,
+      "learning_rate": 2e-05,
+      "loss": 0.1527,
+      "step": 3546
+    },
+    {
+      "epoch": 0.08442825859278301,
+      "grad_norm": 7.628328440687087,
+      "learning_rate": 2e-05,
+      "loss": 1.235,
+      "step": 3547
+    },
+    {
+      "epoch": 0.08445206131581452,
+      "grad_norm": 2.725803268334022,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 3548
+    },
+    {
+      "epoch": 0.08447586403884605,
+      "grad_norm": 7.362909868352627,
+      "learning_rate": 2e-05,
+      "loss": 0.3854,
+      "step": 3549
+    },
+    {
+      "epoch": 0.08449966676187756,
+      "grad_norm": 5.9568861800016455,
+      "learning_rate": 2e-05,
+      "loss": 0.3305,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08452346948490908,
+      "grad_norm": 7.803798425002514,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 3551
+    },
+    {
+      "epoch": 0.08454727220794059,
+      "grad_norm": 2.589623232089177,
+      "learning_rate": 2e-05,
+      "loss": 0.1906,
+      "step": 3552
+    },
+    {
+      "epoch": 0.0845710749309721,
+      "grad_norm": 3.2525816958198344,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 3553
+    },
+    {
+      "epoch": 0.08459487765400361,
+      "grad_norm": 2.676316538463759,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 3554
+    },
+    {
+      "epoch": 0.08461868037703513,
+      "grad_norm": 2.1588123157458687,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 3555
+    },
+    {
+      "epoch": 0.08464248310006665,
+      "grad_norm": 5.8155411982298215,
+      "learning_rate": 2e-05,
+      "loss": 0.8781,
+      "step": 3556
+    },
+    {
+      "epoch": 0.08466628582309817,
+      "grad_norm": 6.999735186830827,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 3557
+    },
+    {
+      "epoch": 0.08469008854612968,
+      "grad_norm": 3.1849572360964036,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 3558
+    },
+    {
+      "epoch": 0.08471389126916119,
+      "grad_norm": 9.67121924814743,
+      "learning_rate": 2e-05,
+      "loss": 1.2504,
+      "step": 3559
+    },
+    {
+      "epoch": 0.0847376939921927,
+      "grad_norm": 8.212085309136206,
+      "learning_rate": 2e-05,
+      "loss": 0.8925,
+      "step": 3560
+    },
+    {
+      "epoch": 0.08476149671522422,
+      "grad_norm": 3.2133456423758244,
+      "learning_rate": 2e-05,
+      "loss": 0.1492,
+      "step": 3561
+    },
+    {
+      "epoch": 0.08478529943825573,
+      "grad_norm": 3.6508401847888754,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 3562
+    },
+    {
+      "epoch": 0.08480910216128724,
+      "grad_norm": 2.0333963387630485,
+      "learning_rate": 2e-05,
+      "loss": 0.2083,
+      "step": 3563
+    },
+    {
+      "epoch": 0.08483290488431877,
+      "grad_norm": 7.627490098616866,
+      "learning_rate": 2e-05,
+      "loss": 0.9324,
+      "step": 3564
+    },
+    {
+      "epoch": 0.08485670760735028,
+      "grad_norm": 2.498035331390312,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 3565
+    },
+    {
+      "epoch": 0.0848805103303818,
+      "grad_norm": 3.246577641587253,
+      "learning_rate": 2e-05,
+      "loss": 0.2862,
+      "step": 3566
+    },
+    {
+      "epoch": 0.08490431305341331,
+      "grad_norm": 9.973489354186672,
+      "learning_rate": 2e-05,
+      "loss": 1.3469,
+      "step": 3567
+    },
+    {
+      "epoch": 0.08492811577644482,
+      "grad_norm": 8.22366978650805,
+      "learning_rate": 2e-05,
+      "loss": 0.3382,
+      "step": 3568
+    },
+    {
+      "epoch": 0.08495191849947634,
+      "grad_norm": 10.158077887533171,
+      "learning_rate": 2e-05,
+      "loss": 1.5829,
+      "step": 3569
+    },
+    {
+      "epoch": 0.08497572122250785,
+      "grad_norm": 3.427285216129741,
+      "learning_rate": 2e-05,
+      "loss": 0.2098,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08499952394553938,
+      "grad_norm": 5.003738986622653,
+      "learning_rate": 2e-05,
+      "loss": 0.339,
+      "step": 3571
+    },
+    {
+      "epoch": 0.08502332666857089,
+      "grad_norm": 2.995020988026045,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 3572
+    },
+    {
+      "epoch": 0.0850471293916024,
+      "grad_norm": 3.1221604873438977,
+      "learning_rate": 2e-05,
+      "loss": 0.2242,
+      "step": 3573
+    },
+    {
+      "epoch": 0.08507093211463392,
+      "grad_norm": 5.921501021300117,
+      "learning_rate": 2e-05,
+      "loss": 0.3305,
+      "step": 3574
+    },
+    {
+      "epoch": 0.08509473483766543,
+      "grad_norm": 4.836240383919242,
+      "learning_rate": 2e-05,
+      "loss": 0.2798,
+      "step": 3575
+    },
+    {
+      "epoch": 0.08511853756069694,
+      "grad_norm": 5.314716655861037,
+      "learning_rate": 2e-05,
+      "loss": 1.0397,
+      "step": 3576
+    },
+    {
+      "epoch": 0.08514234028372845,
+      "grad_norm": 7.873049987962907,
+      "learning_rate": 2e-05,
+      "loss": 1.0314,
+      "step": 3577
+    },
+    {
+      "epoch": 0.08516614300675997,
+      "grad_norm": 9.058878165979149,
+      "learning_rate": 2e-05,
+      "loss": 0.5606,
+      "step": 3578
+    },
+    {
+      "epoch": 0.0851899457297915,
+      "grad_norm": 8.424303585389879,
+      "learning_rate": 2e-05,
+      "loss": 1.0843,
+      "step": 3579
+    },
+    {
+      "epoch": 0.08521374845282301,
+      "grad_norm": 6.094687012963568,
+      "learning_rate": 2e-05,
+      "loss": 0.8419,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08523755117585452,
+      "grad_norm": 2.0599073721265997,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 3581
+    },
+    {
+      "epoch": 0.08526135389888603,
+      "grad_norm": 3.8955926760684663,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 3582
+    },
+    {
+      "epoch": 0.08528515662191755,
+      "grad_norm": 4.968419717198446,
+      "learning_rate": 2e-05,
+      "loss": 0.671,
+      "step": 3583
+    },
+    {
+      "epoch": 0.08530895934494906,
+      "grad_norm": 3.356185361637485,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 3584
+    },
+    {
+      "epoch": 0.08533276206798057,
+      "grad_norm": 3.0652233806717204,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 3585
+    },
+    {
+      "epoch": 0.08535656479101209,
+      "grad_norm": 9.165928905139904,
+      "learning_rate": 2e-05,
+      "loss": 1.3813,
+      "step": 3586
+    },
+    {
+      "epoch": 0.08538036751404361,
+      "grad_norm": 1.5629811447521218,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 3587
+    },
+    {
+      "epoch": 0.08540417023707512,
+      "grad_norm": 11.573683026003751,
+      "learning_rate": 2e-05,
+      "loss": 1.6816,
+      "step": 3588
+    },
+    {
+      "epoch": 0.08542797296010664,
+      "grad_norm": 10.36674007998849,
+      "learning_rate": 2e-05,
+      "loss": 0.8201,
+      "step": 3589
+    },
+    {
+      "epoch": 0.08545177568313815,
+      "grad_norm": 4.326107423943692,
+      "learning_rate": 2e-05,
+      "loss": 0.1783,
+      "step": 3590
+    },
+    {
+      "epoch": 0.08547557840616966,
+      "grad_norm": 9.155665451354228,
+      "learning_rate": 2e-05,
+      "loss": 0.8671,
+      "step": 3591
+    },
+    {
+      "epoch": 0.08549938112920118,
+      "grad_norm": 4.943149518983503,
+      "learning_rate": 2e-05,
+      "loss": 0.5431,
+      "step": 3592
+    },
+    {
+      "epoch": 0.08552318385223269,
+      "grad_norm": 5.920741376981581,
+      "learning_rate": 2e-05,
+      "loss": 0.8099,
+      "step": 3593
+    },
+    {
+      "epoch": 0.08554698657526422,
+      "grad_norm": 7.445927508799216,
+      "learning_rate": 2e-05,
+      "loss": 0.4051,
+      "step": 3594
+    },
+    {
+      "epoch": 0.08557078929829573,
+      "grad_norm": 2.6508859981976145,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 3595
+    },
+    {
+      "epoch": 0.08559459202132724,
+      "grad_norm": 4.203449569299814,
+      "learning_rate": 2e-05,
+      "loss": 0.3239,
+      "step": 3596
+    },
+    {
+      "epoch": 0.08561839474435876,
+      "grad_norm": 3.052779811417629,
+      "learning_rate": 2e-05,
+      "loss": 0.1623,
+      "step": 3597
+    },
+    {
+      "epoch": 0.08564219746739027,
+      "grad_norm": 3.0618170281916925,
+      "learning_rate": 2e-05,
+      "loss": 0.1596,
+      "step": 3598
+    },
+    {
+      "epoch": 0.08566600019042178,
+      "grad_norm": 5.502523196031523,
+      "learning_rate": 2e-05,
+      "loss": 0.9688,
+      "step": 3599
+    },
+    {
+      "epoch": 0.0856898029134533,
+      "grad_norm": 2.718333994211349,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08571360563648481,
+      "grad_norm": 1.7691324445324137,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 3601
+    },
+    {
+      "epoch": 0.08573740835951633,
+      "grad_norm": 7.946911386418184,
+      "learning_rate": 2e-05,
+      "loss": 1.1619,
+      "step": 3602
+    },
+    {
+      "epoch": 0.08576121108254785,
+      "grad_norm": 6.0899251243892145,
+      "learning_rate": 2e-05,
+      "loss": 0.9416,
+      "step": 3603
+    },
+    {
+      "epoch": 0.08578501380557936,
+      "grad_norm": 9.338558283794733,
+      "learning_rate": 2e-05,
+      "loss": 1.2002,
+      "step": 3604
+    },
+    {
+      "epoch": 0.08580881652861087,
+      "grad_norm": 3.0783553786117825,
+      "learning_rate": 2e-05,
+      "loss": 0.2209,
+      "step": 3605
+    },
+    {
+      "epoch": 0.08583261925164239,
+      "grad_norm": 4.912373265878499,
+      "learning_rate": 2e-05,
+      "loss": 0.7036,
+      "step": 3606
+    },
+    {
+      "epoch": 0.0858564219746739,
+      "grad_norm": 5.422155120225751,
+      "learning_rate": 2e-05,
+      "loss": 0.5912,
+      "step": 3607
+    },
+    {
+      "epoch": 0.08588022469770541,
+      "grad_norm": 7.180375734692217,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 3608
+    },
+    {
+      "epoch": 0.08590402742073694,
+      "grad_norm": 1.9677249584680547,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 3609
+    },
+    {
+      "epoch": 0.08592783014376845,
+      "grad_norm": 5.1454970266214906,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 3610
+    },
+    {
+      "epoch": 0.08595163286679997,
+      "grad_norm": 2.4341338204137237,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 3611
+    },
+    {
+      "epoch": 0.08597543558983148,
+      "grad_norm": 4.035392626977511,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 3612
+    },
+    {
+      "epoch": 0.08599923831286299,
+      "grad_norm": 5.405299491640588,
+      "learning_rate": 2e-05,
+      "loss": 0.5679,
+      "step": 3613
+    },
+    {
+      "epoch": 0.0860230410358945,
+      "grad_norm": 9.146674117851974,
+      "learning_rate": 2e-05,
+      "loss": 1.2758,
+      "step": 3614
+    },
+    {
+      "epoch": 0.08604684375892602,
+      "grad_norm": 6.736748238164435,
+      "learning_rate": 2e-05,
+      "loss": 1.0427,
+      "step": 3615
+    },
+    {
+      "epoch": 0.08607064648195753,
+      "grad_norm": 2.1784505830044725,
+      "learning_rate": 2e-05,
+      "loss": 0.1573,
+      "step": 3616
+    },
+    {
+      "epoch": 0.08609444920498906,
+      "grad_norm": 5.920304695643343,
+      "learning_rate": 2e-05,
+      "loss": 0.9652,
+      "step": 3617
+    },
+    {
+      "epoch": 0.08611825192802057,
+      "grad_norm": 2.408605014120972,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 3618
+    },
+    {
+      "epoch": 0.08614205465105208,
+      "grad_norm": 5.606484830633071,
+      "learning_rate": 2e-05,
+      "loss": 0.4732,
+      "step": 3619
+    },
+    {
+      "epoch": 0.0861658573740836,
+      "grad_norm": 7.2037920547467325,
+      "learning_rate": 2e-05,
+      "loss": 0.5496,
+      "step": 3620
+    },
+    {
+      "epoch": 0.08618966009711511,
+      "grad_norm": 8.944658017116987,
+      "learning_rate": 2e-05,
+      "loss": 0.9891,
+      "step": 3621
+    },
+    {
+      "epoch": 0.08621346282014662,
+      "grad_norm": 4.862956833428392,
+      "learning_rate": 2e-05,
+      "loss": 0.67,
+      "step": 3622
+    },
+    {
+      "epoch": 0.08623726554317813,
+      "grad_norm": 9.131985420169054,
+      "learning_rate": 2e-05,
+      "loss": 0.4967,
+      "step": 3623
+    },
+    {
+      "epoch": 0.08626106826620965,
+      "grad_norm": 10.13860755306079,
+      "learning_rate": 2e-05,
+      "loss": 1.7246,
+      "step": 3624
+    },
+    {
+      "epoch": 0.08628487098924117,
+      "grad_norm": 6.71877559742159,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 3625
+    },
+    {
+      "epoch": 0.08630867371227269,
+      "grad_norm": 5.231196824894196,
+      "learning_rate": 2e-05,
+      "loss": 0.5112,
+      "step": 3626
+    },
+    {
+      "epoch": 0.0863324764353042,
+      "grad_norm": 5.1645409343126465,
+      "learning_rate": 2e-05,
+      "loss": 0.2261,
+      "step": 3627
+    },
+    {
+      "epoch": 0.08635627915833571,
+      "grad_norm": 2.6406446660073764,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 3628
+    },
+    {
+      "epoch": 0.08638008188136723,
+      "grad_norm": 5.936520709291377,
+      "learning_rate": 2e-05,
+      "loss": 0.5559,
+      "step": 3629
+    },
+    {
+      "epoch": 0.08640388460439874,
+      "grad_norm": 2.7853704128451184,
+      "learning_rate": 2e-05,
+      "loss": 0.2096,
+      "step": 3630
+    },
+    {
+      "epoch": 0.08642768732743025,
+      "grad_norm": 10.616303729446111,
+      "learning_rate": 2e-05,
+      "loss": 1.6763,
+      "step": 3631
+    },
+    {
+      "epoch": 0.08645149005046178,
+      "grad_norm": 5.546583754393704,
+      "learning_rate": 2e-05,
+      "loss": 0.7888,
+      "step": 3632
+    },
+    {
+      "epoch": 0.08647529277349329,
+      "grad_norm": 2.4435342750867397,
+      "learning_rate": 2e-05,
+      "loss": 0.1959,
+      "step": 3633
+    },
+    {
+      "epoch": 0.0864990954965248,
+      "grad_norm": 3.2226835115001466,
+      "learning_rate": 2e-05,
+      "loss": 0.1327,
+      "step": 3634
+    },
+    {
+      "epoch": 0.08652289821955632,
+      "grad_norm": 2.286555759737503,
+      "learning_rate": 2e-05,
+      "loss": 0.2342,
+      "step": 3635
+    },
+    {
+      "epoch": 0.08654670094258783,
+      "grad_norm": 2.9561819766612567,
+      "learning_rate": 2e-05,
+      "loss": 0.2601,
+      "step": 3636
+    },
+    {
+      "epoch": 0.08657050366561934,
+      "grad_norm": 6.96072056116807,
+      "learning_rate": 2e-05,
+      "loss": 0.3314,
+      "step": 3637
+    },
+    {
+      "epoch": 0.08659430638865086,
+      "grad_norm": 2.3248366865258263,
+      "learning_rate": 2e-05,
+      "loss": 0.2095,
+      "step": 3638
+    },
+    {
+      "epoch": 0.08661810911168237,
+      "grad_norm": 6.576656431588998,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 3639
+    },
+    {
+      "epoch": 0.0866419118347139,
+      "grad_norm": 7.732976872228325,
+      "learning_rate": 2e-05,
+      "loss": 0.4343,
+      "step": 3640
+    },
+    {
+      "epoch": 0.08666571455774541,
+      "grad_norm": 3.9840226742545783,
+      "learning_rate": 2e-05,
+      "loss": 0.3717,
+      "step": 3641
+    },
+    {
+      "epoch": 0.08668951728077692,
+      "grad_norm": 4.481403240578429,
+      "learning_rate": 2e-05,
+      "loss": 0.5716,
+      "step": 3642
+    },
+    {
+      "epoch": 0.08671332000380844,
+      "grad_norm": 8.23358050332172,
+      "learning_rate": 2e-05,
+      "loss": 0.4229,
+      "step": 3643
+    },
+    {
+      "epoch": 0.08673712272683995,
+      "grad_norm": 7.917338311209721,
+      "learning_rate": 2e-05,
+      "loss": 0.2659,
+      "step": 3644
+    },
+    {
+      "epoch": 0.08676092544987146,
+      "grad_norm": 6.674009961392681,
+      "learning_rate": 2e-05,
+      "loss": 1.2994,
+      "step": 3645
+    },
+    {
+      "epoch": 0.08678472817290298,
+      "grad_norm": 9.166673511151894,
+      "learning_rate": 2e-05,
+      "loss": 0.4891,
+      "step": 3646
+    },
+    {
+      "epoch": 0.08680853089593449,
+      "grad_norm": 7.9960955916599445,
+      "learning_rate": 2e-05,
+      "loss": 0.6895,
+      "step": 3647
+    },
+    {
+      "epoch": 0.08683233361896601,
+      "grad_norm": 4.560952230431845,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 3648
+    },
+    {
+      "epoch": 0.08685613634199753,
+      "grad_norm": 9.483645698473213,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 3649
+    },
+    {
+      "epoch": 0.08687993906502904,
+      "grad_norm": 2.8060959837012476,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 3650
+    },
+    {
+      "epoch": 0.08690374178806055,
+      "grad_norm": 9.756498753705316,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 3651
+    },
+    {
+      "epoch": 0.08692754451109207,
+      "grad_norm": 4.160725786357257,
+      "learning_rate": 2e-05,
+      "loss": 0.3702,
+      "step": 3652
+    },
+    {
+      "epoch": 0.08695134723412358,
+      "grad_norm": 4.199092636018394,
+      "learning_rate": 2e-05,
+      "loss": 0.503,
+      "step": 3653
+    },
+    {
+      "epoch": 0.08697514995715509,
+      "grad_norm": 7.813084343944018,
+      "learning_rate": 2e-05,
+      "loss": 0.9727,
+      "step": 3654
+    },
+    {
+      "epoch": 0.08699895268018662,
+      "grad_norm": 10.74695902551881,
+      "learning_rate": 2e-05,
+      "loss": 0.5504,
+      "step": 3655
+    },
+    {
+      "epoch": 0.08702275540321813,
+      "grad_norm": 2.5064202694353597,
+      "learning_rate": 2e-05,
+      "loss": 0.1934,
+      "step": 3656
+    },
+    {
+      "epoch": 0.08704655812624965,
+      "grad_norm": 6.551341834224667,
+      "learning_rate": 2e-05,
+      "loss": 0.7208,
+      "step": 3657
+    },
+    {
+      "epoch": 0.08707036084928116,
+      "grad_norm": 3.4269810711475137,
+      "learning_rate": 2e-05,
+      "loss": 0.3533,
+      "step": 3658
+    },
+    {
+      "epoch": 0.08709416357231267,
+      "grad_norm": 5.730280147415211,
+      "learning_rate": 2e-05,
+      "loss": 0.8383,
+      "step": 3659
+    },
+    {
+      "epoch": 0.08711796629534418,
+      "grad_norm": 3.159681546464256,
+      "learning_rate": 2e-05,
+      "loss": 0.2631,
+      "step": 3660
+    },
+    {
+      "epoch": 0.0871417690183757,
+      "grad_norm": 2.013674405193459,
+      "learning_rate": 2e-05,
+      "loss": 0.199,
+      "step": 3661
+    },
+    {
+      "epoch": 0.08716557174140721,
+      "grad_norm": 2.7440711373790445,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 3662
+    },
+    {
+      "epoch": 0.08718937446443874,
+      "grad_norm": 5.837664284210508,
+      "learning_rate": 2e-05,
+      "loss": 0.8529,
+      "step": 3663
+    },
+    {
+      "epoch": 0.08721317718747025,
+      "grad_norm": 6.2041450637768625,
+      "learning_rate": 2e-05,
+      "loss": 0.3985,
+      "step": 3664
+    },
+    {
+      "epoch": 0.08723697991050176,
+      "grad_norm": 1.852214376395083,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 3665
+    },
+    {
+      "epoch": 0.08726078263353328,
+      "grad_norm": 10.970198213394351,
+      "learning_rate": 2e-05,
+      "loss": 0.5973,
+      "step": 3666
+    },
+    {
+      "epoch": 0.08728458535656479,
+      "grad_norm": 3.1838647425915174,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 3667
+    },
+    {
+      "epoch": 0.0873083880795963,
+      "grad_norm": 5.696254387796412,
+      "learning_rate": 2e-05,
+      "loss": 0.7711,
+      "step": 3668
+    },
+    {
+      "epoch": 0.08733219080262782,
+      "grad_norm": 6.218431549035542,
+      "learning_rate": 2e-05,
+      "loss": 1.1608,
+      "step": 3669
+    },
+    {
+      "epoch": 0.08735599352565934,
+      "grad_norm": 5.81230461160781,
+      "learning_rate": 2e-05,
+      "loss": 0.8225,
+      "step": 3670
+    },
+    {
+      "epoch": 0.08737979624869086,
+      "grad_norm": 8.20831791605477,
+      "learning_rate": 2e-05,
+      "loss": 0.4895,
+      "step": 3671
+    },
+    {
+      "epoch": 0.08740359897172237,
+      "grad_norm": 7.8614290933747295,
+      "learning_rate": 2e-05,
+      "loss": 0.6376,
+      "step": 3672
+    },
+    {
+      "epoch": 0.08742740169475388,
+      "grad_norm": 3.088919740292324,
+      "learning_rate": 2e-05,
+      "loss": 0.2625,
+      "step": 3673
+    },
+    {
+      "epoch": 0.0874512044177854,
+      "grad_norm": 7.454122638350356,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 3674
+    },
+    {
+      "epoch": 0.08747500714081691,
+      "grad_norm": 1.814387633463329,
+      "learning_rate": 2e-05,
+      "loss": 0.18,
+      "step": 3675
+    },
+    {
+      "epoch": 0.08749880986384842,
+      "grad_norm": 9.794269780811717,
+      "learning_rate": 2e-05,
+      "loss": 1.7237,
+      "step": 3676
+    },
+    {
+      "epoch": 0.08752261258687993,
+      "grad_norm": 8.241466633597414,
+      "learning_rate": 2e-05,
+      "loss": 0.3764,
+      "step": 3677
+    },
+    {
+      "epoch": 0.08754641530991146,
+      "grad_norm": 5.813388817405851,
+      "learning_rate": 2e-05,
+      "loss": 0.7742,
+      "step": 3678
+    },
+    {
+      "epoch": 0.08757021803294297,
+      "grad_norm": 6.7816449735023046,
+      "learning_rate": 2e-05,
+      "loss": 0.3689,
+      "step": 3679
+    },
+    {
+      "epoch": 0.08759402075597449,
+      "grad_norm": 5.100465640773198,
+      "learning_rate": 2e-05,
+      "loss": 0.4233,
+      "step": 3680
+    },
+    {
+      "epoch": 0.087617823479006,
+      "grad_norm": 7.385734357298984,
+      "learning_rate": 2e-05,
+      "loss": 1.1154,
+      "step": 3681
+    },
+    {
+      "epoch": 0.08764162620203751,
+      "grad_norm": 4.14300358677611,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 3682
+    },
+    {
+      "epoch": 0.08766542892506902,
+      "grad_norm": 2.1409293247415646,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 3683
+    },
+    {
+      "epoch": 0.08768923164810054,
+      "grad_norm": 2.732761408259395,
+      "learning_rate": 2e-05,
+      "loss": 0.2216,
+      "step": 3684
+    },
+    {
+      "epoch": 0.08771303437113205,
+      "grad_norm": 6.629048094383313,
+      "learning_rate": 2e-05,
+      "loss": 0.6319,
+      "step": 3685
+    },
+    {
+      "epoch": 0.08773683709416358,
+      "grad_norm": 10.368345360512702,
+      "learning_rate": 2e-05,
+      "loss": 1.2815,
+      "step": 3686
+    },
+    {
+      "epoch": 0.08776063981719509,
+      "grad_norm": 10.090826545755116,
+      "learning_rate": 2e-05,
+      "loss": 0.4702,
+      "step": 3687
+    },
+    {
+      "epoch": 0.0877844425402266,
+      "grad_norm": 10.687891572010399,
+      "learning_rate": 2e-05,
+      "loss": 2.5115,
+      "step": 3688
+    },
+    {
+      "epoch": 0.08780824526325812,
+      "grad_norm": 7.269057350633687,
+      "learning_rate": 2e-05,
+      "loss": 0.4685,
+      "step": 3689
+    },
+    {
+      "epoch": 0.08783204798628963,
+      "grad_norm": 6.057803761875524,
+      "learning_rate": 2e-05,
+      "loss": 1.0677,
+      "step": 3690
+    },
+    {
+      "epoch": 0.08785585070932114,
+      "grad_norm": 17.041418006369238,
+      "learning_rate": 2e-05,
+      "loss": 0.8426,
+      "step": 3691
+    },
+    {
+      "epoch": 0.08787965343235266,
+      "grad_norm": 11.507159513003224,
+      "learning_rate": 2e-05,
+      "loss": 0.6308,
+      "step": 3692
+    },
+    {
+      "epoch": 0.08790345615538418,
+      "grad_norm": 5.949755701470239,
+      "learning_rate": 2e-05,
+      "loss": 0.7866,
+      "step": 3693
+    },
+    {
+      "epoch": 0.0879272588784157,
+      "grad_norm": 5.745566061956415,
+      "learning_rate": 2e-05,
+      "loss": 1.1761,
+      "step": 3694
+    },
+    {
+      "epoch": 0.08795106160144721,
+      "grad_norm": 6.904191227824254,
+      "learning_rate": 2e-05,
+      "loss": 0.5711,
+      "step": 3695
+    },
+    {
+      "epoch": 0.08797486432447872,
+      "grad_norm": 2.987256470893289,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 3696
+    },
+    {
+      "epoch": 0.08799866704751023,
+      "grad_norm": 5.418117805636472,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 3697
+    },
+    {
+      "epoch": 0.08802246977054175,
+      "grad_norm": 7.87055482512057,
+      "learning_rate": 2e-05,
+      "loss": 0.9688,
+      "step": 3698
+    },
+    {
+      "epoch": 0.08804627249357326,
+      "grad_norm": 4.965738125897962,
+      "learning_rate": 2e-05,
+      "loss": 0.7,
+      "step": 3699
+    },
+    {
+      "epoch": 0.08807007521660477,
+      "grad_norm": 12.263146028305343,
+      "learning_rate": 2e-05,
+      "loss": 1.8674,
+      "step": 3700
+    },
+    {
+      "epoch": 0.0880938779396363,
+      "grad_norm": 4.923372494616672,
+      "learning_rate": 2e-05,
+      "loss": 0.8243,
+      "step": 3701
+    },
+    {
+      "epoch": 0.08811768066266781,
+      "grad_norm": 6.471959529919043,
+      "learning_rate": 2e-05,
+      "loss": 0.4094,
+      "step": 3702
+    },
+    {
+      "epoch": 0.08814148338569933,
+      "grad_norm": 7.2046802077440075,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 3703
+    },
+    {
+      "epoch": 0.08816528610873084,
+      "grad_norm": 7.985420319830443,
+      "learning_rate": 2e-05,
+      "loss": 0.2947,
+      "step": 3704
+    },
+    {
+      "epoch": 0.08818908883176235,
+      "grad_norm": 1.192796833719069,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 3705
+    },
+    {
+      "epoch": 0.08821289155479387,
+      "grad_norm": 2.113906670265512,
+      "learning_rate": 2e-05,
+      "loss": 0.1811,
+      "step": 3706
+    },
+    {
+      "epoch": 0.08823669427782538,
+      "grad_norm": 7.627577471757347,
+      "learning_rate": 2e-05,
+      "loss": 0.8887,
+      "step": 3707
+    },
+    {
+      "epoch": 0.08826049700085689,
+      "grad_norm": 8.847932653074263,
+      "learning_rate": 2e-05,
+      "loss": 1.3183,
+      "step": 3708
+    },
+    {
+      "epoch": 0.08828429972388842,
+      "grad_norm": 7.393463838055947,
+      "learning_rate": 2e-05,
+      "loss": 1.5638,
+      "step": 3709
+    },
+    {
+      "epoch": 0.08830810244691993,
+      "grad_norm": 5.904938144468274,
+      "learning_rate": 2e-05,
+      "loss": 0.8629,
+      "step": 3710
+    },
+    {
+      "epoch": 0.08833190516995144,
+      "grad_norm": 7.622586324750284,
+      "learning_rate": 2e-05,
+      "loss": 1.2425,
+      "step": 3711
+    },
+    {
+      "epoch": 0.08835570789298296,
+      "grad_norm": 8.425330339458416,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 3712
+    },
+    {
+      "epoch": 0.08837951061601447,
+      "grad_norm": 4.001306418587627,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 3713
+    },
+    {
+      "epoch": 0.08840331333904598,
+      "grad_norm": 4.19805694970717,
+      "learning_rate": 2e-05,
+      "loss": 0.266,
+      "step": 3714
+    },
+    {
+      "epoch": 0.0884271160620775,
+      "grad_norm": 5.009470467146281,
+      "learning_rate": 2e-05,
+      "loss": 0.3573,
+      "step": 3715
+    },
+    {
+      "epoch": 0.08845091878510902,
+      "grad_norm": 3.8526596634097694,
+      "learning_rate": 2e-05,
+      "loss": 0.2175,
+      "step": 3716
+    },
+    {
+      "epoch": 0.08847472150814054,
+      "grad_norm": 8.05683805453686,
+      "learning_rate": 2e-05,
+      "loss": 1.0893,
+      "step": 3717
+    },
+    {
+      "epoch": 0.08849852423117205,
+      "grad_norm": 4.491228625644723,
+      "learning_rate": 2e-05,
+      "loss": 0.3019,
+      "step": 3718
+    },
+    {
+      "epoch": 0.08852232695420356,
+      "grad_norm": 5.073151860152183,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 3719
+    },
+    {
+      "epoch": 0.08854612967723507,
+      "grad_norm": 2.014137001790449,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 3720
+    },
+    {
+      "epoch": 0.08856993240026659,
+      "grad_norm": 5.80488560364872,
+      "learning_rate": 2e-05,
+      "loss": 0.3662,
+      "step": 3721
+    },
+    {
+      "epoch": 0.0885937351232981,
+      "grad_norm": 1.5363991029394781,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 3722
+    },
+    {
+      "epoch": 0.08861753784632961,
+      "grad_norm": 4.428127761744023,
+      "learning_rate": 2e-05,
+      "loss": 0.3678,
+      "step": 3723
+    },
+    {
+      "epoch": 0.08864134056936114,
+      "grad_norm": 8.624712866230565,
+      "learning_rate": 2e-05,
+      "loss": 0.3414,
+      "step": 3724
+    },
+    {
+      "epoch": 0.08866514329239265,
+      "grad_norm": 7.300395709306911,
+      "learning_rate": 2e-05,
+      "loss": 0.4749,
+      "step": 3725
+    },
+    {
+      "epoch": 0.08868894601542417,
+      "grad_norm": 3.7950654142064644,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 3726
+    },
+    {
+      "epoch": 0.08871274873845568,
+      "grad_norm": 5.000992286341135,
+      "learning_rate": 2e-05,
+      "loss": 0.3616,
+      "step": 3727
+    },
+    {
+      "epoch": 0.08873655146148719,
+      "grad_norm": 6.11305047945745,
+      "learning_rate": 2e-05,
+      "loss": 1.074,
+      "step": 3728
+    },
+    {
+      "epoch": 0.0887603541845187,
+      "grad_norm": 4.218308347762871,
+      "learning_rate": 2e-05,
+      "loss": 0.6393,
+      "step": 3729
+    },
+    {
+      "epoch": 0.08878415690755022,
+      "grad_norm": 1.6608671183424928,
+      "learning_rate": 2e-05,
+      "loss": 0.1718,
+      "step": 3730
+    },
+    {
+      "epoch": 0.08880795963058175,
+      "grad_norm": 7.360778370456367,
+      "learning_rate": 2e-05,
+      "loss": 0.6956,
+      "step": 3731
+    },
+    {
+      "epoch": 0.08883176235361326,
+      "grad_norm": 1.5203019412754888,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 3732
+    },
+    {
+      "epoch": 0.08885556507664477,
+      "grad_norm": 3.17050297667392,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 3733
+    },
+    {
+      "epoch": 0.08887936779967628,
+      "grad_norm": 2.7775812121531067,
+      "learning_rate": 2e-05,
+      "loss": 0.2017,
+      "step": 3734
+    },
+    {
+      "epoch": 0.0889031705227078,
+      "grad_norm": 7.45988918622045,
+      "learning_rate": 2e-05,
+      "loss": 0.3965,
+      "step": 3735
+    },
+    {
+      "epoch": 0.08892697324573931,
+      "grad_norm": 3.659392298654049,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 3736
+    },
+    {
+      "epoch": 0.08895077596877082,
+      "grad_norm": 2.8134299231659963,
+      "learning_rate": 2e-05,
+      "loss": 0.3081,
+      "step": 3737
+    },
+    {
+      "epoch": 0.08897457869180234,
+      "grad_norm": 5.833987370149647,
+      "learning_rate": 2e-05,
+      "loss": 0.2626,
+      "step": 3738
+    },
+    {
+      "epoch": 0.08899838141483386,
+      "grad_norm": 1.4838464986631097,
+      "learning_rate": 2e-05,
+      "loss": 0.1725,
+      "step": 3739
+    },
+    {
+      "epoch": 0.08902218413786538,
+      "grad_norm": 7.521224732330546,
+      "learning_rate": 2e-05,
+      "loss": 0.9114,
+      "step": 3740
+    },
+    {
+      "epoch": 0.08904598686089689,
+      "grad_norm": 5.428537469320224,
+      "learning_rate": 2e-05,
+      "loss": 0.8204,
+      "step": 3741
+    },
+    {
+      "epoch": 0.0890697895839284,
+      "grad_norm": 5.5838379928706505,
+      "learning_rate": 2e-05,
+      "loss": 0.3644,
+      "step": 3742
+    },
+    {
+      "epoch": 0.08909359230695992,
+      "grad_norm": 9.101907761046824,
+      "learning_rate": 2e-05,
+      "loss": 1.0412,
+      "step": 3743
+    },
+    {
+      "epoch": 0.08911739502999143,
+      "grad_norm": 3.1774165470493028,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 3744
+    },
+    {
+      "epoch": 0.08914119775302294,
+      "grad_norm": 7.241518140711836,
+      "learning_rate": 2e-05,
+      "loss": 0.4085,
+      "step": 3745
+    },
+    {
+      "epoch": 0.08916500047605445,
+      "grad_norm": 6.546901460973805,
+      "learning_rate": 2e-05,
+      "loss": 0.2269,
+      "step": 3746
+    },
+    {
+      "epoch": 0.08918880319908598,
+      "grad_norm": 2.7591956194061202,
+      "learning_rate": 2e-05,
+      "loss": 0.2018,
+      "step": 3747
+    },
+    {
+      "epoch": 0.0892126059221175,
+      "grad_norm": 4.298177158809503,
+      "learning_rate": 2e-05,
+      "loss": 0.6342,
+      "step": 3748
+    },
+    {
+      "epoch": 0.089236408645149,
+      "grad_norm": 6.322876328878816,
+      "learning_rate": 2e-05,
+      "loss": 0.6735,
+      "step": 3749
+    },
+    {
+      "epoch": 0.08926021136818052,
+      "grad_norm": 2.3704127327961073,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 3750
+    },
+    {
+      "epoch": 0.08928401409121203,
+      "grad_norm": 2.095684961081757,
+      "learning_rate": 2e-05,
+      "loss": 0.1796,
+      "step": 3751
+    },
+    {
+      "epoch": 0.08930781681424355,
+      "grad_norm": 2.284547798740005,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 3752
+    },
+    {
+      "epoch": 0.08933161953727506,
+      "grad_norm": 2.831984827973127,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 3753
+    },
+    {
+      "epoch": 0.08935542226030659,
+      "grad_norm": 4.413479869787769,
+      "learning_rate": 2e-05,
+      "loss": 0.4244,
+      "step": 3754
+    },
+    {
+      "epoch": 0.0893792249833381,
+      "grad_norm": 2.674694890878432,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 3755
+    },
+    {
+      "epoch": 0.08940302770636961,
+      "grad_norm": 3.7603491813169425,
+      "learning_rate": 2e-05,
+      "loss": 0.3505,
+      "step": 3756
+    },
+    {
+      "epoch": 0.08942683042940112,
+      "grad_norm": 3.7661197999186724,
+      "learning_rate": 2e-05,
+      "loss": 0.2796,
+      "step": 3757
+    },
+    {
+      "epoch": 0.08945063315243264,
+      "grad_norm": 5.533065908290363,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 3758
+    },
+    {
+      "epoch": 0.08947443587546415,
+      "grad_norm": 2.6926443930085666,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 3759
+    },
+    {
+      "epoch": 0.08949823859849566,
+      "grad_norm": 4.855093350512245,
+      "learning_rate": 2e-05,
+      "loss": 0.8225,
+      "step": 3760
+    },
+    {
+      "epoch": 0.08952204132152718,
+      "grad_norm": 10.860161932973394,
+      "learning_rate": 2e-05,
+      "loss": 1.1719,
+      "step": 3761
+    },
+    {
+      "epoch": 0.0895458440445587,
+      "grad_norm": 8.374054064357097,
+      "learning_rate": 2e-05,
+      "loss": 0.7958,
+      "step": 3762
+    },
+    {
+      "epoch": 0.08956964676759022,
+      "grad_norm": 2.796961187577446,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 3763
+    },
+    {
+      "epoch": 0.08959344949062173,
+      "grad_norm": 3.9047924478761633,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 3764
+    },
+    {
+      "epoch": 0.08961725221365324,
+      "grad_norm": 7.40529112735416,
+      "learning_rate": 2e-05,
+      "loss": 0.79,
+      "step": 3765
+    },
+    {
+      "epoch": 0.08964105493668476,
+      "grad_norm": 4.574318039096484,
+      "learning_rate": 2e-05,
+      "loss": 0.6166,
+      "step": 3766
+    },
+    {
+      "epoch": 0.08966485765971627,
+      "grad_norm": 6.714892879203145,
+      "learning_rate": 2e-05,
+      "loss": 1.2497,
+      "step": 3767
+    },
+    {
+      "epoch": 0.08968866038274778,
+      "grad_norm": 6.456305764100982,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 3768
+    },
+    {
+      "epoch": 0.0897124631057793,
+      "grad_norm": 5.676628723383561,
+      "learning_rate": 2e-05,
+      "loss": 0.6073,
+      "step": 3769
+    },
+    {
+      "epoch": 0.08973626582881082,
+      "grad_norm": 6.613954884825323,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 3770
+    },
+    {
+      "epoch": 0.08976006855184233,
+      "grad_norm": 5.965626408079156,
+      "learning_rate": 2e-05,
+      "loss": 0.3096,
+      "step": 3771
+    },
+    {
+      "epoch": 0.08978387127487385,
+      "grad_norm": 5.447770280415772,
+      "learning_rate": 2e-05,
+      "loss": 0.2746,
+      "step": 3772
+    },
+    {
+      "epoch": 0.08980767399790536,
+      "grad_norm": 5.174333560350587,
+      "learning_rate": 2e-05,
+      "loss": 0.6997,
+      "step": 3773
+    },
+    {
+      "epoch": 0.08983147672093687,
+      "grad_norm": 8.310439034564068,
+      "learning_rate": 2e-05,
+      "loss": 0.2997,
+      "step": 3774
+    },
+    {
+      "epoch": 0.08985527944396839,
+      "grad_norm": 5.435792095349835,
+      "learning_rate": 2e-05,
+      "loss": 0.8848,
+      "step": 3775
+    },
+    {
+      "epoch": 0.0898790821669999,
+      "grad_norm": 2.14434752216647,
+      "learning_rate": 2e-05,
+      "loss": 0.1551,
+      "step": 3776
+    },
+    {
+      "epoch": 0.08990288489003143,
+      "grad_norm": 7.272712281779905,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 3777
+    },
+    {
+      "epoch": 0.08992668761306294,
+      "grad_norm": 7.556287242926599,
+      "learning_rate": 2e-05,
+      "loss": 1.0138,
+      "step": 3778
+    },
+    {
+      "epoch": 0.08995049033609445,
+      "grad_norm": 10.408426079967256,
+      "learning_rate": 2e-05,
+      "loss": 1.1847,
+      "step": 3779
+    },
+    {
+      "epoch": 0.08997429305912596,
+      "grad_norm": 5.513026925805462,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 3780
+    },
+    {
+      "epoch": 0.08999809578215748,
+      "grad_norm": 2.3652638546576616,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 3781
+    },
+    {
+      "epoch": 0.09002189850518899,
+      "grad_norm": 2.3533974676879326,
+      "learning_rate": 2e-05,
+      "loss": 0.1684,
+      "step": 3782
+    },
+    {
+      "epoch": 0.0900457012282205,
+      "grad_norm": 6.9266015679751805,
+      "learning_rate": 2e-05,
+      "loss": 0.3399,
+      "step": 3783
+    },
+    {
+      "epoch": 0.09006950395125202,
+      "grad_norm": 7.52948336313336,
+      "learning_rate": 2e-05,
+      "loss": 0.2877,
+      "step": 3784
+    },
+    {
+      "epoch": 0.09009330667428354,
+      "grad_norm": 6.5318060170105445,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 3785
+    },
+    {
+      "epoch": 0.09011710939731506,
+      "grad_norm": 16.70816865188352,
+      "learning_rate": 2e-05,
+      "loss": 1.0361,
+      "step": 3786
+    },
+    {
+      "epoch": 0.09014091212034657,
+      "grad_norm": 4.23040968144254,
+      "learning_rate": 2e-05,
+      "loss": 0.2009,
+      "step": 3787
+    },
+    {
+      "epoch": 0.09016471484337808,
+      "grad_norm": 1.2855014328802252,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 3788
+    },
+    {
+      "epoch": 0.0901885175664096,
+      "grad_norm": 8.021280009908336,
+      "learning_rate": 2e-05,
+      "loss": 0.2832,
+      "step": 3789
+    },
+    {
+      "epoch": 0.09021232028944111,
+      "grad_norm": 6.952418915184488,
+      "learning_rate": 2e-05,
+      "loss": 0.457,
+      "step": 3790
+    },
+    {
+      "epoch": 0.09023612301247262,
+      "grad_norm": 2.0651329780272714,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 3791
+    },
+    {
+      "epoch": 0.09025992573550415,
+      "grad_norm": 7.369737118171388,
+      "learning_rate": 2e-05,
+      "loss": 0.5059,
+      "step": 3792
+    },
+    {
+      "epoch": 0.09028372845853566,
+      "grad_norm": 4.788366648105384,
+      "learning_rate": 2e-05,
+      "loss": 0.6695,
+      "step": 3793
+    },
+    {
+      "epoch": 0.09030753118156717,
+      "grad_norm": 4.72745657853281,
+      "learning_rate": 2e-05,
+      "loss": 0.2951,
+      "step": 3794
+    },
+    {
+      "epoch": 0.09033133390459869,
+      "grad_norm": 2.7761793517477953,
+      "learning_rate": 2e-05,
+      "loss": 0.1793,
+      "step": 3795
+    },
+    {
+      "epoch": 0.0903551366276302,
+      "grad_norm": 5.7685986193101,
+      "learning_rate": 2e-05,
+      "loss": 0.7447,
+      "step": 3796
+    },
+    {
+      "epoch": 0.09037893935066171,
+      "grad_norm": 6.853936585580858,
+      "learning_rate": 2e-05,
+      "loss": 0.3663,
+      "step": 3797
+    },
+    {
+      "epoch": 0.09040274207369323,
+      "grad_norm": 3.8776773615124567,
+      "learning_rate": 2e-05,
+      "loss": 0.1748,
+      "step": 3798
+    },
+    {
+      "epoch": 0.09042654479672474,
+      "grad_norm": 2.8609842281855884,
+      "learning_rate": 2e-05,
+      "loss": 0.2785,
+      "step": 3799
+    },
+    {
+      "epoch": 0.09045034751975627,
+      "grad_norm": 6.773642709096078,
+      "learning_rate": 2e-05,
+      "loss": 0.5919,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09047415024278778,
+      "grad_norm": 9.413495583911645,
+      "learning_rate": 2e-05,
+      "loss": 1.2014,
+      "step": 3801
+    },
+    {
+      "epoch": 0.09049795296581929,
+      "grad_norm": 9.31529817583843,
+      "learning_rate": 2e-05,
+      "loss": 0.9261,
+      "step": 3802
+    },
+    {
+      "epoch": 0.0905217556888508,
+      "grad_norm": 3.483406507091207,
+      "learning_rate": 2e-05,
+      "loss": 0.3469,
+      "step": 3803
+    },
+    {
+      "epoch": 0.09054555841188232,
+      "grad_norm": 3.306999631890484,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 3804
+    },
+    {
+      "epoch": 0.09056936113491383,
+      "grad_norm": 8.771646042732336,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 3805
+    },
+    {
+      "epoch": 0.09059316385794534,
+      "grad_norm": 10.113210855257785,
+      "learning_rate": 2e-05,
+      "loss": 0.3464,
+      "step": 3806
+    },
+    {
+      "epoch": 0.09061696658097686,
+      "grad_norm": 7.94461501641959,
+      "learning_rate": 2e-05,
+      "loss": 1.6822,
+      "step": 3807
+    },
+    {
+      "epoch": 0.09064076930400838,
+      "grad_norm": 7.230441732133678,
+      "learning_rate": 2e-05,
+      "loss": 0.4593,
+      "step": 3808
+    },
+    {
+      "epoch": 0.0906645720270399,
+      "grad_norm": 17.6378255296457,
+      "learning_rate": 2e-05,
+      "loss": 0.6268,
+      "step": 3809
+    },
+    {
+      "epoch": 0.09068837475007141,
+      "grad_norm": 3.953009669900038,
+      "learning_rate": 2e-05,
+      "loss": 0.3109,
+      "step": 3810
+    },
+    {
+      "epoch": 0.09071217747310292,
+      "grad_norm": 2.9797415203475563,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 3811
+    },
+    {
+      "epoch": 0.09073598019613444,
+      "grad_norm": 9.677880161016319,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 3812
+    },
+    {
+      "epoch": 0.09075978291916595,
+      "grad_norm": 3.3399085729493514,
+      "learning_rate": 2e-05,
+      "loss": 0.252,
+      "step": 3813
+    },
+    {
+      "epoch": 0.09078358564219746,
+      "grad_norm": 7.932095896569732,
+      "learning_rate": 2e-05,
+      "loss": 1.0639,
+      "step": 3814
+    },
+    {
+      "epoch": 0.09080738836522899,
+      "grad_norm": 10.38197910276366,
+      "learning_rate": 2e-05,
+      "loss": 1.2319,
+      "step": 3815
+    },
+    {
+      "epoch": 0.0908311910882605,
+      "grad_norm": 8.217725940490439,
+      "learning_rate": 2e-05,
+      "loss": 0.3799,
+      "step": 3816
+    },
+    {
+      "epoch": 0.09085499381129201,
+      "grad_norm": 3.3566111806906,
+      "learning_rate": 2e-05,
+      "loss": 0.1768,
+      "step": 3817
+    },
+    {
+      "epoch": 0.09087879653432353,
+      "grad_norm": 3.2651167034160413,
+      "learning_rate": 2e-05,
+      "loss": 0.301,
+      "step": 3818
+    },
+    {
+      "epoch": 0.09090259925735504,
+      "grad_norm": 2.8883269046052553,
+      "learning_rate": 2e-05,
+      "loss": 0.2083,
+      "step": 3819
+    },
+    {
+      "epoch": 0.09092640198038655,
+      "grad_norm": 7.507621215933463,
+      "learning_rate": 2e-05,
+      "loss": 0.5614,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09095020470341807,
+      "grad_norm": 5.936720327523376,
+      "learning_rate": 2e-05,
+      "loss": 1.0003,
+      "step": 3821
+    },
+    {
+      "epoch": 0.09097400742644958,
+      "grad_norm": 11.594566289230556,
+      "learning_rate": 2e-05,
+      "loss": 1.2548,
+      "step": 3822
+    },
+    {
+      "epoch": 0.0909978101494811,
+      "grad_norm": 2.4899948402550516,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 3823
+    },
+    {
+      "epoch": 0.09102161287251262,
+      "grad_norm": 8.196839737626675,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 3824
+    },
+    {
+      "epoch": 0.09104541559554413,
+      "grad_norm": 7.650914530965817,
+      "learning_rate": 2e-05,
+      "loss": 0.4076,
+      "step": 3825
+    },
+    {
+      "epoch": 0.09106921831857565,
+      "grad_norm": 9.994267647402967,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 3826
+    },
+    {
+      "epoch": 0.09109302104160716,
+      "grad_norm": 6.734872968896421,
+      "learning_rate": 2e-05,
+      "loss": 1.0082,
+      "step": 3827
+    },
+    {
+      "epoch": 0.09111682376463867,
+      "grad_norm": 8.621832736974465,
+      "learning_rate": 2e-05,
+      "loss": 0.4745,
+      "step": 3828
+    },
+    {
+      "epoch": 0.09114062648767018,
+      "grad_norm": 2.740605730650327,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 3829
+    },
+    {
+      "epoch": 0.0911644292107017,
+      "grad_norm": 8.467000528626977,
+      "learning_rate": 2e-05,
+      "loss": 1.2905,
+      "step": 3830
+    },
+    {
+      "epoch": 0.09118823193373322,
+      "grad_norm": 4.841768699520008,
+      "learning_rate": 2e-05,
+      "loss": 0.3337,
+      "step": 3831
+    },
+    {
+      "epoch": 0.09121203465676474,
+      "grad_norm": 7.5098778481926916,
+      "learning_rate": 2e-05,
+      "loss": 0.3112,
+      "step": 3832
+    },
+    {
+      "epoch": 0.09123583737979625,
+      "grad_norm": 2.3143267390375857,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 3833
+    },
+    {
+      "epoch": 0.09125964010282776,
+      "grad_norm": 4.343046943778815,
+      "learning_rate": 2e-05,
+      "loss": 0.654,
+      "step": 3834
+    },
+    {
+      "epoch": 0.09128344282585928,
+      "grad_norm": 4.353347578287936,
+      "learning_rate": 2e-05,
+      "loss": 0.3151,
+      "step": 3835
+    },
+    {
+      "epoch": 0.09130724554889079,
+      "grad_norm": 4.607673237398999,
+      "learning_rate": 2e-05,
+      "loss": 0.3076,
+      "step": 3836
+    },
+    {
+      "epoch": 0.0913310482719223,
+      "grad_norm": 6.131537133964606,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 3837
+    },
+    {
+      "epoch": 0.09135485099495383,
+      "grad_norm": 8.597202920192952,
+      "learning_rate": 2e-05,
+      "loss": 0.3243,
+      "step": 3838
+    },
+    {
+      "epoch": 0.09137865371798534,
+      "grad_norm": 4.653724571060142,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 3839
+    },
+    {
+      "epoch": 0.09140245644101685,
+      "grad_norm": 2.8318253393861093,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09142625916404837,
+      "grad_norm": 9.148687608391143,
+      "learning_rate": 2e-05,
+      "loss": 1.1302,
+      "step": 3841
+    },
+    {
+      "epoch": 0.09145006188707988,
+      "grad_norm": 4.4045086340660635,
+      "learning_rate": 2e-05,
+      "loss": 0.8883,
+      "step": 3842
+    },
+    {
+      "epoch": 0.0914738646101114,
+      "grad_norm": 6.212638522069626,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 3843
+    },
+    {
+      "epoch": 0.0914976673331429,
+      "grad_norm": 8.086688716563547,
+      "learning_rate": 2e-05,
+      "loss": 0.6205,
+      "step": 3844
+    },
+    {
+      "epoch": 0.09152147005617442,
+      "grad_norm": 5.994548487203503,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 3845
+    },
+    {
+      "epoch": 0.09154527277920595,
+      "grad_norm": 1.5120787683172918,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 3846
+    },
+    {
+      "epoch": 0.09156907550223746,
+      "grad_norm": 3.6356224267389003,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 3847
+    },
+    {
+      "epoch": 0.09159287822526897,
+      "grad_norm": 5.107491739675894,
+      "learning_rate": 2e-05,
+      "loss": 0.8023,
+      "step": 3848
+    },
+    {
+      "epoch": 0.09161668094830049,
+      "grad_norm": 10.232783195496362,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 3849
+    },
+    {
+      "epoch": 0.091640483671332,
+      "grad_norm": 2.1320218559804562,
+      "learning_rate": 2e-05,
+      "loss": 0.2455,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09166428639436351,
+      "grad_norm": 5.5346697421322695,
+      "learning_rate": 2e-05,
+      "loss": 0.7389,
+      "step": 3851
+    },
+    {
+      "epoch": 0.09168808911739502,
+      "grad_norm": 4.243900504940997,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 3852
+    },
+    {
+      "epoch": 0.09171189184042655,
+      "grad_norm": 6.236736656500947,
+      "learning_rate": 2e-05,
+      "loss": 0.9301,
+      "step": 3853
+    },
+    {
+      "epoch": 0.09173569456345806,
+      "grad_norm": 2.2372400060627244,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 3854
+    },
+    {
+      "epoch": 0.09175949728648958,
+      "grad_norm": 8.673358692863946,
+      "learning_rate": 2e-05,
+      "loss": 0.914,
+      "step": 3855
+    },
+    {
+      "epoch": 0.09178330000952109,
+      "grad_norm": 5.282329607616499,
+      "learning_rate": 2e-05,
+      "loss": 0.3182,
+      "step": 3856
+    },
+    {
+      "epoch": 0.0918071027325526,
+      "grad_norm": 7.652099985478001,
+      "learning_rate": 2e-05,
+      "loss": 0.4732,
+      "step": 3857
+    },
+    {
+      "epoch": 0.09183090545558412,
+      "grad_norm": 6.047397887223467,
+      "learning_rate": 2e-05,
+      "loss": 0.5224,
+      "step": 3858
+    },
+    {
+      "epoch": 0.09185470817861563,
+      "grad_norm": 4.975844423015639,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 3859
+    },
+    {
+      "epoch": 0.09187851090164714,
+      "grad_norm": 3.552212420535377,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09190231362467867,
+      "grad_norm": 5.094024363521332,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 3861
+    },
+    {
+      "epoch": 0.09192611634771018,
+      "grad_norm": 3.7123961934741105,
+      "learning_rate": 2e-05,
+      "loss": 0.188,
+      "step": 3862
+    },
+    {
+      "epoch": 0.0919499190707417,
+      "grad_norm": 5.014998913915703,
+      "learning_rate": 2e-05,
+      "loss": 0.6682,
+      "step": 3863
+    },
+    {
+      "epoch": 0.09197372179377321,
+      "grad_norm": 1.9803122292392812,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 3864
+    },
+    {
+      "epoch": 0.09199752451680472,
+      "grad_norm": 2.7526606510823477,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 3865
+    },
+    {
+      "epoch": 0.09202132723983623,
+      "grad_norm": 8.165305858408987,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 3866
+    },
+    {
+      "epoch": 0.09204512996286775,
+      "grad_norm": 5.441285488554216,
+      "learning_rate": 2e-05,
+      "loss": 0.6859,
+      "step": 3867
+    },
+    {
+      "epoch": 0.09206893268589926,
+      "grad_norm": 10.063388367823993,
+      "learning_rate": 2e-05,
+      "loss": 1.5189,
+      "step": 3868
+    },
+    {
+      "epoch": 0.09209273540893079,
+      "grad_norm": 3.8686621955002214,
+      "learning_rate": 2e-05,
+      "loss": 0.2108,
+      "step": 3869
+    },
+    {
+      "epoch": 0.0921165381319623,
+      "grad_norm": 2.869232979188315,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09214034085499381,
+      "grad_norm": 5.549508579633911,
+      "learning_rate": 2e-05,
+      "loss": 0.8229,
+      "step": 3871
+    },
+    {
+      "epoch": 0.09216414357802533,
+      "grad_norm": 6.18267961307881,
+      "learning_rate": 2e-05,
+      "loss": 0.6697,
+      "step": 3872
+    },
+    {
+      "epoch": 0.09218794630105684,
+      "grad_norm": 7.8248133363568675,
+      "learning_rate": 2e-05,
+      "loss": 1.3031,
+      "step": 3873
+    },
+    {
+      "epoch": 0.09221174902408835,
+      "grad_norm": 2.2035555690211166,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 3874
+    },
+    {
+      "epoch": 0.09223555174711986,
+      "grad_norm": 2.5319223379515874,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 3875
+    },
+    {
+      "epoch": 0.09225935447015139,
+      "grad_norm": 3.0696789962747557,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 3876
+    },
+    {
+      "epoch": 0.0922831571931829,
+      "grad_norm": 6.379176730356193,
+      "learning_rate": 2e-05,
+      "loss": 0.6559,
+      "step": 3877
+    },
+    {
+      "epoch": 0.09230695991621442,
+      "grad_norm": 4.248926641884477,
+      "learning_rate": 2e-05,
+      "loss": 0.1893,
+      "step": 3878
+    },
+    {
+      "epoch": 0.09233076263924593,
+      "grad_norm": 4.9210808244369835,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 3879
+    },
+    {
+      "epoch": 0.09235456536227744,
+      "grad_norm": 4.353804751752521,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 3880
+    },
+    {
+      "epoch": 0.09237836808530896,
+      "grad_norm": 5.076104922820257,
+      "learning_rate": 2e-05,
+      "loss": 0.6019,
+      "step": 3881
+    },
+    {
+      "epoch": 0.09240217080834047,
+      "grad_norm": 3.0875724971171317,
+      "learning_rate": 2e-05,
+      "loss": 0.175,
+      "step": 3882
+    },
+    {
+      "epoch": 0.09242597353137198,
+      "grad_norm": 3.7305587948688657,
+      "learning_rate": 2e-05,
+      "loss": 0.2359,
+      "step": 3883
+    },
+    {
+      "epoch": 0.09244977625440351,
+      "grad_norm": 6.867616119774752,
+      "learning_rate": 2e-05,
+      "loss": 0.2801,
+      "step": 3884
+    },
+    {
+      "epoch": 0.09247357897743502,
+      "grad_norm": 3.280545737570488,
+      "learning_rate": 2e-05,
+      "loss": 0.2721,
+      "step": 3885
+    },
+    {
+      "epoch": 0.09249738170046654,
+      "grad_norm": 6.4266100446865115,
+      "learning_rate": 2e-05,
+      "loss": 1.1143,
+      "step": 3886
+    },
+    {
+      "epoch": 0.09252118442349805,
+      "grad_norm": 2.9074100866920665,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 3887
+    },
+    {
+      "epoch": 0.09254498714652956,
+      "grad_norm": 7.054709023042463,
+      "learning_rate": 2e-05,
+      "loss": 1.0769,
+      "step": 3888
+    },
+    {
+      "epoch": 0.09256878986956107,
+      "grad_norm": 4.983453763098729,
+      "learning_rate": 2e-05,
+      "loss": 0.7632,
+      "step": 3889
+    },
+    {
+      "epoch": 0.09259259259259259,
+      "grad_norm": 2.058983062324066,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09261639531562411,
+      "grad_norm": 7.927474235826844,
+      "learning_rate": 2e-05,
+      "loss": 0.8599,
+      "step": 3891
+    },
+    {
+      "epoch": 0.09264019803865563,
+      "grad_norm": 5.857887015173121,
+      "learning_rate": 2e-05,
+      "loss": 0.8453,
+      "step": 3892
+    },
+    {
+      "epoch": 0.09266400076168714,
+      "grad_norm": 5.745279200687996,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 3893
+    },
+    {
+      "epoch": 0.09268780348471865,
+      "grad_norm": 4.574547031792981,
+      "learning_rate": 2e-05,
+      "loss": 0.2222,
+      "step": 3894
+    },
+    {
+      "epoch": 0.09271160620775017,
+      "grad_norm": 5.086940523829978,
+      "learning_rate": 2e-05,
+      "loss": 0.4073,
+      "step": 3895
+    },
+    {
+      "epoch": 0.09273540893078168,
+      "grad_norm": 7.553077613292723,
+      "learning_rate": 2e-05,
+      "loss": 0.856,
+      "step": 3896
+    },
+    {
+      "epoch": 0.09275921165381319,
+      "grad_norm": 5.972802632232322,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 3897
+    },
+    {
+      "epoch": 0.0927830143768447,
+      "grad_norm": 1.797131661089796,
+      "learning_rate": 2e-05,
+      "loss": 0.2151,
+      "step": 3898
+    },
+    {
+      "epoch": 0.09280681709987623,
+      "grad_norm": 1.7840794493842296,
+      "learning_rate": 2e-05,
+      "loss": 0.2115,
+      "step": 3899
+    },
+    {
+      "epoch": 0.09283061982290775,
+      "grad_norm": 8.12707890087674,
+      "learning_rate": 2e-05,
+      "loss": 0.9716,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09285442254593926,
+      "grad_norm": 10.200699067345733,
+      "learning_rate": 2e-05,
+      "loss": 1.3111,
+      "step": 3901
+    },
+    {
+      "epoch": 0.09287822526897077,
+      "grad_norm": 2.6538572624064667,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 3902
+    },
+    {
+      "epoch": 0.09290202799200228,
+      "grad_norm": 3.087412568970475,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 3903
+    },
+    {
+      "epoch": 0.0929258307150338,
+      "grad_norm": 5.645292500427728,
+      "learning_rate": 2e-05,
+      "loss": 0.9322,
+      "step": 3904
+    },
+    {
+      "epoch": 0.09294963343806531,
+      "grad_norm": 2.7944288776921247,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 3905
+    },
+    {
+      "epoch": 0.09297343616109682,
+      "grad_norm": 6.539835095799918,
+      "learning_rate": 2e-05,
+      "loss": 0.394,
+      "step": 3906
+    },
+    {
+      "epoch": 0.09299723888412835,
+      "grad_norm": 2.75345197143691,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 3907
+    },
+    {
+      "epoch": 0.09302104160715986,
+      "grad_norm": 4.711061242713716,
+      "learning_rate": 2e-05,
+      "loss": 0.7437,
+      "step": 3908
+    },
+    {
+      "epoch": 0.09304484433019138,
+      "grad_norm": 7.49840049607541,
+      "learning_rate": 2e-05,
+      "loss": 0.8641,
+      "step": 3909
+    },
+    {
+      "epoch": 0.09306864705322289,
+      "grad_norm": 5.444340251196053,
+      "learning_rate": 2e-05,
+      "loss": 0.8149,
+      "step": 3910
+    },
+    {
+      "epoch": 0.0930924497762544,
+      "grad_norm": 4.182824134805353,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 3911
+    },
+    {
+      "epoch": 0.09311625249928591,
+      "grad_norm": 9.602731328639624,
+      "learning_rate": 2e-05,
+      "loss": 1.3686,
+      "step": 3912
+    },
+    {
+      "epoch": 0.09314005522231743,
+      "grad_norm": 2.6348682869539073,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 3913
+    },
+    {
+      "epoch": 0.09316385794534895,
+      "grad_norm": 6.511088599804983,
+      "learning_rate": 2e-05,
+      "loss": 1.0309,
+      "step": 3914
+    },
+    {
+      "epoch": 0.09318766066838047,
+      "grad_norm": 7.95927089187081,
+      "learning_rate": 2e-05,
+      "loss": 1.0182,
+      "step": 3915
+    },
+    {
+      "epoch": 0.09321146339141198,
+      "grad_norm": 4.455733449314921,
+      "learning_rate": 2e-05,
+      "loss": 0.3491,
+      "step": 3916
+    },
+    {
+      "epoch": 0.0932352661144435,
+      "grad_norm": 11.242350199551995,
+      "learning_rate": 2e-05,
+      "loss": 1.103,
+      "step": 3917
+    },
+    {
+      "epoch": 0.093259068837475,
+      "grad_norm": 6.396262223324917,
+      "learning_rate": 2e-05,
+      "loss": 0.8817,
+      "step": 3918
+    },
+    {
+      "epoch": 0.09328287156050652,
+      "grad_norm": 8.514795129737953,
+      "learning_rate": 2e-05,
+      "loss": 1.2307,
+      "step": 3919
+    },
+    {
+      "epoch": 0.09330667428353803,
+      "grad_norm": 9.957360739758732,
+      "learning_rate": 2e-05,
+      "loss": 1.688,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09333047700656955,
+      "grad_norm": 7.077865644588151,
+      "learning_rate": 2e-05,
+      "loss": 1.0529,
+      "step": 3921
+    },
+    {
+      "epoch": 0.09335427972960107,
+      "grad_norm": 6.5586102635317,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 3922
+    },
+    {
+      "epoch": 0.09337808245263259,
+      "grad_norm": 7.115839875944895,
+      "learning_rate": 2e-05,
+      "loss": 0.4495,
+      "step": 3923
+    },
+    {
+      "epoch": 0.0934018851756641,
+      "grad_norm": 7.698707429704378,
+      "learning_rate": 2e-05,
+      "loss": 0.4488,
+      "step": 3924
+    },
+    {
+      "epoch": 0.09342568789869561,
+      "grad_norm": 5.5281930098358645,
+      "learning_rate": 2e-05,
+      "loss": 0.6711,
+      "step": 3925
+    },
+    {
+      "epoch": 0.09344949062172712,
+      "grad_norm": 3.1201231210608866,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 3926
+    },
+    {
+      "epoch": 0.09347329334475864,
+      "grad_norm": 6.303024791275767,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 3927
+    },
+    {
+      "epoch": 0.09349709606779015,
+      "grad_norm": 6.388273136316902,
+      "learning_rate": 2e-05,
+      "loss": 0.6775,
+      "step": 3928
+    },
+    {
+      "epoch": 0.09352089879082166,
+      "grad_norm": 5.160106852742896,
+      "learning_rate": 2e-05,
+      "loss": 0.2943,
+      "step": 3929
+    },
+    {
+      "epoch": 0.09354470151385319,
+      "grad_norm": 5.331210645007362,
+      "learning_rate": 2e-05,
+      "loss": 0.2432,
+      "step": 3930
+    },
+    {
+      "epoch": 0.0935685042368847,
+      "grad_norm": 1.1956700956376962,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 3931
+    },
+    {
+      "epoch": 0.09359230695991622,
+      "grad_norm": 3.670785280341058,
+      "learning_rate": 2e-05,
+      "loss": 0.5432,
+      "step": 3932
+    },
+    {
+      "epoch": 0.09361610968294773,
+      "grad_norm": 2.290165707226185,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 3933
+    },
+    {
+      "epoch": 0.09363991240597924,
+      "grad_norm": 8.172355041491643,
+      "learning_rate": 2e-05,
+      "loss": 1.1789,
+      "step": 3934
+    },
+    {
+      "epoch": 0.09366371512901076,
+      "grad_norm": 4.9973749138707015,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 3935
+    },
+    {
+      "epoch": 0.09368751785204227,
+      "grad_norm": 5.274350058107056,
+      "learning_rate": 2e-05,
+      "loss": 0.8628,
+      "step": 3936
+    },
+    {
+      "epoch": 0.0937113205750738,
+      "grad_norm": 10.491438785900055,
+      "learning_rate": 2e-05,
+      "loss": 0.8184,
+      "step": 3937
+    },
+    {
+      "epoch": 0.09373512329810531,
+      "grad_norm": 3.515938026605608,
+      "learning_rate": 2e-05,
+      "loss": 0.5842,
+      "step": 3938
+    },
+    {
+      "epoch": 0.09375892602113682,
+      "grad_norm": 3.605386846333153,
+      "learning_rate": 2e-05,
+      "loss": 0.2119,
+      "step": 3939
+    },
+    {
+      "epoch": 0.09378272874416833,
+      "grad_norm": 4.361978611698519,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09380653146719985,
+      "grad_norm": 4.294841230902327,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 3941
+    },
+    {
+      "epoch": 0.09383033419023136,
+      "grad_norm": 4.262676710191222,
+      "learning_rate": 2e-05,
+      "loss": 0.4811,
+      "step": 3942
+    },
+    {
+      "epoch": 0.09385413691326287,
+      "grad_norm": 10.734369172994379,
+      "learning_rate": 2e-05,
+      "loss": 0.9691,
+      "step": 3943
+    },
+    {
+      "epoch": 0.09387793963629439,
+      "grad_norm": 6.722578140020667,
+      "learning_rate": 2e-05,
+      "loss": 0.7147,
+      "step": 3944
+    },
+    {
+      "epoch": 0.09390174235932591,
+      "grad_norm": 4.023725468044433,
+      "learning_rate": 2e-05,
+      "loss": 0.3945,
+      "step": 3945
+    },
+    {
+      "epoch": 0.09392554508235743,
+      "grad_norm": 2.7010578908821774,
+      "learning_rate": 2e-05,
+      "loss": 0.1804,
+      "step": 3946
+    },
+    {
+      "epoch": 0.09394934780538894,
+      "grad_norm": 4.807872031085035,
+      "learning_rate": 2e-05,
+      "loss": 0.8742,
+      "step": 3947
+    },
+    {
+      "epoch": 0.09397315052842045,
+      "grad_norm": 6.270262363373841,
+      "learning_rate": 2e-05,
+      "loss": 1.0078,
+      "step": 3948
+    },
+    {
+      "epoch": 0.09399695325145196,
+      "grad_norm": 3.1643309888034676,
+      "learning_rate": 2e-05,
+      "loss": 0.1692,
+      "step": 3949
+    },
+    {
+      "epoch": 0.09402075597448348,
+      "grad_norm": 3.8818216958655443,
+      "learning_rate": 2e-05,
+      "loss": 0.4942,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09404455869751499,
+      "grad_norm": 7.991706728186196,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 3951
+    },
+    {
+      "epoch": 0.09406836142054652,
+      "grad_norm": 9.953535823156823,
+      "learning_rate": 2e-05,
+      "loss": 0.5767,
+      "step": 3952
+    },
+    {
+      "epoch": 0.09409216414357803,
+      "grad_norm": 2.557601892619635,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 3953
+    },
+    {
+      "epoch": 0.09411596686660954,
+      "grad_norm": 9.100447631009592,
+      "learning_rate": 2e-05,
+      "loss": 0.5103,
+      "step": 3954
+    },
+    {
+      "epoch": 0.09413976958964106,
+      "grad_norm": 8.004582883283996,
+      "learning_rate": 2e-05,
+      "loss": 0.3676,
+      "step": 3955
+    },
+    {
+      "epoch": 0.09416357231267257,
+      "grad_norm": 12.99769124878921,
+      "learning_rate": 2e-05,
+      "loss": 1.4026,
+      "step": 3956
+    },
+    {
+      "epoch": 0.09418737503570408,
+      "grad_norm": 4.019229305178031,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 3957
+    },
+    {
+      "epoch": 0.0942111777587356,
+      "grad_norm": 3.8223780858954637,
+      "learning_rate": 2e-05,
+      "loss": 0.2803,
+      "step": 3958
+    },
+    {
+      "epoch": 0.09423498048176711,
+      "grad_norm": 7.581093213391499,
+      "learning_rate": 2e-05,
+      "loss": 1.5709,
+      "step": 3959
+    },
+    {
+      "epoch": 0.09425878320479864,
+      "grad_norm": 1.705604860500628,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 3960
+    },
+    {
+      "epoch": 0.09428258592783015,
+      "grad_norm": 8.417754260873641,
+      "learning_rate": 2e-05,
+      "loss": 0.4162,
+      "step": 3961
+    },
+    {
+      "epoch": 0.09430638865086166,
+      "grad_norm": 5.5038589540634995,
+      "learning_rate": 2e-05,
+      "loss": 0.863,
+      "step": 3962
+    },
+    {
+      "epoch": 0.09433019137389317,
+      "grad_norm": 7.533829303480629,
+      "learning_rate": 2e-05,
+      "loss": 0.3261,
+      "step": 3963
+    },
+    {
+      "epoch": 0.09435399409692469,
+      "grad_norm": 7.515697010428146,
+      "learning_rate": 2e-05,
+      "loss": 0.9015,
+      "step": 3964
+    },
+    {
+      "epoch": 0.0943777968199562,
+      "grad_norm": 5.506909208084105,
+      "learning_rate": 2e-05,
+      "loss": 0.5113,
+      "step": 3965
+    },
+    {
+      "epoch": 0.09440159954298771,
+      "grad_norm": 8.045583613901105,
+      "learning_rate": 2e-05,
+      "loss": 1.5537,
+      "step": 3966
+    },
+    {
+      "epoch": 0.09442540226601923,
+      "grad_norm": 5.4159033429080194,
+      "learning_rate": 2e-05,
+      "loss": 0.6639,
+      "step": 3967
+    },
+    {
+      "epoch": 0.09444920498905075,
+      "grad_norm": 10.22162267978247,
+      "learning_rate": 2e-05,
+      "loss": 1.2015,
+      "step": 3968
+    },
+    {
+      "epoch": 0.09447300771208227,
+      "grad_norm": 5.350539452102051,
+      "learning_rate": 2e-05,
+      "loss": 0.2406,
+      "step": 3969
+    },
+    {
+      "epoch": 0.09449681043511378,
+      "grad_norm": 5.345526493267819,
+      "learning_rate": 2e-05,
+      "loss": 0.7244,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09452061315814529,
+      "grad_norm": 3.668534295573549,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 3971
+    },
+    {
+      "epoch": 0.0945444158811768,
+      "grad_norm": 6.504309736963955,
+      "learning_rate": 2e-05,
+      "loss": 1.0384,
+      "step": 3972
+    },
+    {
+      "epoch": 0.09456821860420832,
+      "grad_norm": 5.166304829826611,
+      "learning_rate": 2e-05,
+      "loss": 0.4053,
+      "step": 3973
+    },
+    {
+      "epoch": 0.09459202132723983,
+      "grad_norm": 7.570238348972861,
+      "learning_rate": 2e-05,
+      "loss": 0.9897,
+      "step": 3974
+    },
+    {
+      "epoch": 0.09461582405027136,
+      "grad_norm": 4.824433964441483,
+      "learning_rate": 2e-05,
+      "loss": 0.5162,
+      "step": 3975
+    },
+    {
+      "epoch": 0.09463962677330287,
+      "grad_norm": 6.276164353130986,
+      "learning_rate": 2e-05,
+      "loss": 1.2305,
+      "step": 3976
+    },
+    {
+      "epoch": 0.09466342949633438,
+      "grad_norm": 6.243470720568949,
+      "learning_rate": 2e-05,
+      "loss": 0.6213,
+      "step": 3977
+    },
+    {
+      "epoch": 0.0946872322193659,
+      "grad_norm": 5.267203228640196,
+      "learning_rate": 2e-05,
+      "loss": 0.795,
+      "step": 3978
+    },
+    {
+      "epoch": 0.09471103494239741,
+      "grad_norm": 7.725965873759842,
+      "learning_rate": 2e-05,
+      "loss": 1.0761,
+      "step": 3979
+    },
+    {
+      "epoch": 0.09473483766542892,
+      "grad_norm": 4.230909243880799,
+      "learning_rate": 2e-05,
+      "loss": 0.2238,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09475864038846044,
+      "grad_norm": 4.175210398149674,
+      "learning_rate": 2e-05,
+      "loss": 0.5146,
+      "step": 3981
+    },
+    {
+      "epoch": 0.09478244311149195,
+      "grad_norm": 4.024938165979686,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 3982
+    },
+    {
+      "epoch": 0.09480624583452348,
+      "grad_norm": 10.37124626386728,
+      "learning_rate": 2e-05,
+      "loss": 0.7415,
+      "step": 3983
+    },
+    {
+      "epoch": 0.09483004855755499,
+      "grad_norm": 7.26521928770864,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 3984
+    },
+    {
+      "epoch": 0.0948538512805865,
+      "grad_norm": 28.89326788758513,
+      "learning_rate": 2e-05,
+      "loss": 0.752,
+      "step": 3985
+    },
+    {
+      "epoch": 0.09487765400361801,
+      "grad_norm": 6.320940121114912,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 3986
+    },
+    {
+      "epoch": 0.09490145672664953,
+      "grad_norm": 2.184604670914432,
+      "learning_rate": 2e-05,
+      "loss": 0.1866,
+      "step": 3987
+    },
+    {
+      "epoch": 0.09492525944968104,
+      "grad_norm": 5.253648718789849,
+      "learning_rate": 2e-05,
+      "loss": 0.2639,
+      "step": 3988
+    },
+    {
+      "epoch": 0.09494906217271255,
+      "grad_norm": 8.005747595777219,
+      "learning_rate": 2e-05,
+      "loss": 0.9837,
+      "step": 3989
+    },
+    {
+      "epoch": 0.09497286489574407,
+      "grad_norm": 10.655038195586624,
+      "learning_rate": 2e-05,
+      "loss": 0.3702,
+      "step": 3990
+    },
+    {
+      "epoch": 0.0949966676187756,
+      "grad_norm": 3.062338851519008,
+      "learning_rate": 2e-05,
+      "loss": 0.2176,
+      "step": 3991
+    },
+    {
+      "epoch": 0.0950204703418071,
+      "grad_norm": 3.4653697541326562,
+      "learning_rate": 2e-05,
+      "loss": 0.4742,
+      "step": 3992
+    },
+    {
+      "epoch": 0.09504427306483862,
+      "grad_norm": 6.058669421546616,
+      "learning_rate": 2e-05,
+      "loss": 0.6818,
+      "step": 3993
+    },
+    {
+      "epoch": 0.09506807578787013,
+      "grad_norm": 3.1190951612547755,
+      "learning_rate": 2e-05,
+      "loss": 0.2181,
+      "step": 3994
+    },
+    {
+      "epoch": 0.09509187851090165,
+      "grad_norm": 11.88786828928471,
+      "learning_rate": 2e-05,
+      "loss": 1.255,
+      "step": 3995
+    },
+    {
+      "epoch": 0.09511568123393316,
+      "grad_norm": 8.488458483477725,
+      "learning_rate": 2e-05,
+      "loss": 0.3833,
+      "step": 3996
+    },
+    {
+      "epoch": 0.09513948395696467,
+      "grad_norm": 4.4477154732101845,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 3997
+    },
+    {
+      "epoch": 0.0951632866799962,
+      "grad_norm": 3.5453856542937965,
+      "learning_rate": 2e-05,
+      "loss": 0.2372,
+      "step": 3998
+    },
+    {
+      "epoch": 0.09518708940302771,
+      "grad_norm": 1.9110226254080847,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 3999
+    },
+    {
+      "epoch": 0.09521089212605922,
+      "grad_norm": 6.2145515908722135,
+      "learning_rate": 2e-05,
+      "loss": 0.3634,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09523469484909074,
+      "grad_norm": 2.531359047969861,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 4001
+    },
+    {
+      "epoch": 0.09525849757212225,
+      "grad_norm": 9.427899036407394,
+      "learning_rate": 2e-05,
+      "loss": 0.9293,
+      "step": 4002
+    },
+    {
+      "epoch": 0.09528230029515376,
+      "grad_norm": 2.836384110924994,
+      "learning_rate": 2e-05,
+      "loss": 0.2811,
+      "step": 4003
+    },
+    {
+      "epoch": 0.09530610301818528,
+      "grad_norm": 9.79802561668679,
+      "learning_rate": 2e-05,
+      "loss": 0.5442,
+      "step": 4004
+    },
+    {
+      "epoch": 0.09532990574121679,
+      "grad_norm": 6.0971050284142345,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 4005
+    },
+    {
+      "epoch": 0.09535370846424832,
+      "grad_norm": 2.9401189261414555,
+      "learning_rate": 2e-05,
+      "loss": 0.2161,
+      "step": 4006
+    },
+    {
+      "epoch": 0.09537751118727983,
+      "grad_norm": 6.160348727693837,
+      "learning_rate": 2e-05,
+      "loss": 0.3233,
+      "step": 4007
+    },
+    {
+      "epoch": 0.09540131391031134,
+      "grad_norm": 9.107262870057083,
+      "learning_rate": 2e-05,
+      "loss": 2.0389,
+      "step": 4008
+    },
+    {
+      "epoch": 0.09542511663334285,
+      "grad_norm": 2.96597130058497,
+      "learning_rate": 2e-05,
+      "loss": 0.2243,
+      "step": 4009
+    },
+    {
+      "epoch": 0.09544891935637437,
+      "grad_norm": 2.586736137137381,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09547272207940588,
+      "grad_norm": 5.969667740005526,
+      "learning_rate": 2e-05,
+      "loss": 0.4761,
+      "step": 4011
+    },
+    {
+      "epoch": 0.0954965248024374,
+      "grad_norm": 5.561074560552017,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 4012
+    },
+    {
+      "epoch": 0.09552032752546892,
+      "grad_norm": 9.088554230253635,
+      "learning_rate": 2e-05,
+      "loss": 0.4764,
+      "step": 4013
+    },
+    {
+      "epoch": 0.09554413024850043,
+      "grad_norm": 4.840968478809785,
+      "learning_rate": 2e-05,
+      "loss": 0.9677,
+      "step": 4014
+    },
+    {
+      "epoch": 0.09556793297153195,
+      "grad_norm": 5.408433738649745,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 4015
+    },
+    {
+      "epoch": 0.09559173569456346,
+      "grad_norm": 8.390281682486409,
+      "learning_rate": 2e-05,
+      "loss": 0.9811,
+      "step": 4016
+    },
+    {
+      "epoch": 0.09561553841759497,
+      "grad_norm": 4.864802733279654,
+      "learning_rate": 2e-05,
+      "loss": 0.211,
+      "step": 4017
+    },
+    {
+      "epoch": 0.09563934114062649,
+      "grad_norm": 7.797922020496496,
+      "learning_rate": 2e-05,
+      "loss": 0.7151,
+      "step": 4018
+    },
+    {
+      "epoch": 0.095663143863658,
+      "grad_norm": 8.588130259164402,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 4019
+    },
+    {
+      "epoch": 0.09568694658668951,
+      "grad_norm": 5.906913890412402,
+      "learning_rate": 2e-05,
+      "loss": 0.6731,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09571074930972104,
+      "grad_norm": 2.6555713171298896,
+      "learning_rate": 2e-05,
+      "loss": 0.2226,
+      "step": 4021
+    },
+    {
+      "epoch": 0.09573455203275255,
+      "grad_norm": 4.512121325557028,
+      "learning_rate": 2e-05,
+      "loss": 0.5162,
+      "step": 4022
+    },
+    {
+      "epoch": 0.09575835475578406,
+      "grad_norm": 13.085870259094163,
+      "learning_rate": 2e-05,
+      "loss": 0.471,
+      "step": 4023
+    },
+    {
+      "epoch": 0.09578215747881558,
+      "grad_norm": 6.16918934117636,
+      "learning_rate": 2e-05,
+      "loss": 0.3615,
+      "step": 4024
+    },
+    {
+      "epoch": 0.09580596020184709,
+      "grad_norm": 6.446547403987311,
+      "learning_rate": 2e-05,
+      "loss": 0.4263,
+      "step": 4025
+    },
+    {
+      "epoch": 0.0958297629248786,
+      "grad_norm": 2.412559922532327,
+      "learning_rate": 2e-05,
+      "loss": 0.2612,
+      "step": 4026
+    },
+    {
+      "epoch": 0.09585356564791012,
+      "grad_norm": 5.575120098537736,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 4027
+    },
+    {
+      "epoch": 0.09587736837094163,
+      "grad_norm": 1.906340869353557,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 4028
+    },
+    {
+      "epoch": 0.09590117109397316,
+      "grad_norm": 6.385177935717529,
+      "learning_rate": 2e-05,
+      "loss": 0.3752,
+      "step": 4029
+    },
+    {
+      "epoch": 0.09592497381700467,
+      "grad_norm": 1.0255185772305324,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 4030
+    },
+    {
+      "epoch": 0.09594877654003618,
+      "grad_norm": 7.099878506002933,
+      "learning_rate": 2e-05,
+      "loss": 0.6034,
+      "step": 4031
+    },
+    {
+      "epoch": 0.0959725792630677,
+      "grad_norm": 4.106352961261855,
+      "learning_rate": 2e-05,
+      "loss": 0.4184,
+      "step": 4032
+    },
+    {
+      "epoch": 0.09599638198609921,
+      "grad_norm": 6.978247886961,
+      "learning_rate": 2e-05,
+      "loss": 0.432,
+      "step": 4033
+    },
+    {
+      "epoch": 0.09602018470913072,
+      "grad_norm": 5.1411482892566145,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 4034
+    },
+    {
+      "epoch": 0.09604398743216223,
+      "grad_norm": 8.964940985808331,
+      "learning_rate": 2e-05,
+      "loss": 0.42,
+      "step": 4035
+    },
+    {
+      "epoch": 0.09606779015519376,
+      "grad_norm": 8.660305262708015,
+      "learning_rate": 2e-05,
+      "loss": 0.3661,
+      "step": 4036
+    },
+    {
+      "epoch": 0.09609159287822527,
+      "grad_norm": 4.266755171040705,
+      "learning_rate": 2e-05,
+      "loss": 0.5558,
+      "step": 4037
+    },
+    {
+      "epoch": 0.09611539560125679,
+      "grad_norm": 7.214741249880248,
+      "learning_rate": 2e-05,
+      "loss": 0.5044,
+      "step": 4038
+    },
+    {
+      "epoch": 0.0961391983242883,
+      "grad_norm": 1.474769488064386,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 4039
+    },
+    {
+      "epoch": 0.09616300104731981,
+      "grad_norm": 5.94825112692556,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 4040
+    },
+    {
+      "epoch": 0.09618680377035133,
+      "grad_norm": 3.175305491586624,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 4041
+    },
+    {
+      "epoch": 0.09621060649338284,
+      "grad_norm": 9.951445219585072,
+      "learning_rate": 2e-05,
+      "loss": 1.1739,
+      "step": 4042
+    },
+    {
+      "epoch": 0.09623440921641435,
+      "grad_norm": 2.3225909676300978,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 4043
+    },
+    {
+      "epoch": 0.09625821193944588,
+      "grad_norm": 2.534231321170464,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 4044
+    },
+    {
+      "epoch": 0.09628201466247739,
+      "grad_norm": 4.258386217043371,
+      "learning_rate": 2e-05,
+      "loss": 0.6177,
+      "step": 4045
+    },
+    {
+      "epoch": 0.0963058173855089,
+      "grad_norm": 5.759132520455434,
+      "learning_rate": 2e-05,
+      "loss": 0.6487,
+      "step": 4046
+    },
+    {
+      "epoch": 0.09632962010854042,
+      "grad_norm": 6.4588174335973205,
+      "learning_rate": 2e-05,
+      "loss": 0.4448,
+      "step": 4047
+    },
+    {
+      "epoch": 0.09635342283157193,
+      "grad_norm": 5.819622428594364,
+      "learning_rate": 2e-05,
+      "loss": 0.7343,
+      "step": 4048
+    },
+    {
+      "epoch": 0.09637722555460344,
+      "grad_norm": 2.277478567276151,
+      "learning_rate": 2e-05,
+      "loss": 0.125,
+      "step": 4049
+    },
+    {
+      "epoch": 0.09640102827763496,
+      "grad_norm": 2.3983882323176267,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 4050
+    },
+    {
+      "epoch": 0.09642483100066647,
+      "grad_norm": 8.280547060236518,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 4051
+    },
+    {
+      "epoch": 0.096448633723698,
+      "grad_norm": 12.416539339155394,
+      "learning_rate": 2e-05,
+      "loss": 0.4272,
+      "step": 4052
+    },
+    {
+      "epoch": 0.09647243644672951,
+      "grad_norm": 6.966826449862693,
+      "learning_rate": 2e-05,
+      "loss": 0.3293,
+      "step": 4053
+    },
+    {
+      "epoch": 0.09649623916976102,
+      "grad_norm": 1.7837602529060637,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 4054
+    },
+    {
+      "epoch": 0.09652004189279254,
+      "grad_norm": 5.690293631102936,
+      "learning_rate": 2e-05,
+      "loss": 0.3283,
+      "step": 4055
+    },
+    {
+      "epoch": 0.09654384461582405,
+      "grad_norm": 2.4290374294703705,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 4056
+    },
+    {
+      "epoch": 0.09656764733885556,
+      "grad_norm": 5.613171179451821,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 4057
+    },
+    {
+      "epoch": 0.09659145006188707,
+      "grad_norm": 6.865293180017134,
+      "learning_rate": 2e-05,
+      "loss": 0.6162,
+      "step": 4058
+    },
+    {
+      "epoch": 0.0966152527849186,
+      "grad_norm": 8.677455710637373,
+      "learning_rate": 2e-05,
+      "loss": 0.4671,
+      "step": 4059
+    },
+    {
+      "epoch": 0.09663905550795011,
+      "grad_norm": 2.954292131125714,
+      "learning_rate": 2e-05,
+      "loss": 0.1914,
+      "step": 4060
+    },
+    {
+      "epoch": 0.09666285823098163,
+      "grad_norm": 2.1819192479894753,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 4061
+    },
+    {
+      "epoch": 0.09668666095401314,
+      "grad_norm": 8.144069188075166,
+      "learning_rate": 2e-05,
+      "loss": 0.9778,
+      "step": 4062
+    },
+    {
+      "epoch": 0.09671046367704465,
+      "grad_norm": 9.147834545234403,
+      "learning_rate": 2e-05,
+      "loss": 1.1741,
+      "step": 4063
+    },
+    {
+      "epoch": 0.09673426640007617,
+      "grad_norm": 9.730036239527436,
+      "learning_rate": 2e-05,
+      "loss": 1.5243,
+      "step": 4064
+    },
+    {
+      "epoch": 0.09675806912310768,
+      "grad_norm": 5.235057959027659,
+      "learning_rate": 2e-05,
+      "loss": 0.4727,
+      "step": 4065
+    },
+    {
+      "epoch": 0.09678187184613919,
+      "grad_norm": 11.399854350016216,
+      "learning_rate": 2e-05,
+      "loss": 0.3674,
+      "step": 4066
+    },
+    {
+      "epoch": 0.09680567456917072,
+      "grad_norm": 6.189328866904493,
+      "learning_rate": 2e-05,
+      "loss": 0.2475,
+      "step": 4067
+    },
+    {
+      "epoch": 0.09682947729220223,
+      "grad_norm": 9.174864006323624,
+      "learning_rate": 2e-05,
+      "loss": 1.6119,
+      "step": 4068
+    },
+    {
+      "epoch": 0.09685328001523374,
+      "grad_norm": 6.876444974373207,
+      "learning_rate": 2e-05,
+      "loss": 0.3259,
+      "step": 4069
+    },
+    {
+      "epoch": 0.09687708273826526,
+      "grad_norm": 7.337964332255387,
+      "learning_rate": 2e-05,
+      "loss": 0.4621,
+      "step": 4070
+    },
+    {
+      "epoch": 0.09690088546129677,
+      "grad_norm": 7.838212723976708,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 4071
+    },
+    {
+      "epoch": 0.09692468818432828,
+      "grad_norm": 5.933785439552234,
+      "learning_rate": 2e-05,
+      "loss": 1.1811,
+      "step": 4072
+    },
+    {
+      "epoch": 0.0969484909073598,
+      "grad_norm": 7.510965603367502,
+      "learning_rate": 2e-05,
+      "loss": 0.7392,
+      "step": 4073
+    },
+    {
+      "epoch": 0.09697229363039132,
+      "grad_norm": 8.710971041682841,
+      "learning_rate": 2e-05,
+      "loss": 0.4806,
+      "step": 4074
+    },
+    {
+      "epoch": 0.09699609635342284,
+      "grad_norm": 3.029121757723987,
+      "learning_rate": 2e-05,
+      "loss": 0.1962,
+      "step": 4075
+    },
+    {
+      "epoch": 0.09701989907645435,
+      "grad_norm": 4.440908695173467,
+      "learning_rate": 2e-05,
+      "loss": 0.7434,
+      "step": 4076
+    },
+    {
+      "epoch": 0.09704370179948586,
+      "grad_norm": 5.306603457173339,
+      "learning_rate": 2e-05,
+      "loss": 0.9857,
+      "step": 4077
+    },
+    {
+      "epoch": 0.09706750452251738,
+      "grad_norm": 4.782886450054822,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 4078
+    },
+    {
+      "epoch": 0.09709130724554889,
+      "grad_norm": 5.939622017558023,
+      "learning_rate": 2e-05,
+      "loss": 0.3166,
+      "step": 4079
+    },
+    {
+      "epoch": 0.0971151099685804,
+      "grad_norm": 8.993916767965358,
+      "learning_rate": 2e-05,
+      "loss": 0.4716,
+      "step": 4080
+    },
+    {
+      "epoch": 0.09713891269161191,
+      "grad_norm": 8.797446048317093,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 4081
+    },
+    {
+      "epoch": 0.09716271541464344,
+      "grad_norm": 8.927078383269501,
+      "learning_rate": 2e-05,
+      "loss": 0.3749,
+      "step": 4082
+    },
+    {
+      "epoch": 0.09718651813767495,
+      "grad_norm": 4.028680196228677,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 4083
+    },
+    {
+      "epoch": 0.09721032086070647,
+      "grad_norm": 2.522728381283681,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 4084
+    },
+    {
+      "epoch": 0.09723412358373798,
+      "grad_norm": 5.43518986701467,
+      "learning_rate": 2e-05,
+      "loss": 0.2937,
+      "step": 4085
+    },
+    {
+      "epoch": 0.0972579263067695,
+      "grad_norm": 6.0673062134280284,
+      "learning_rate": 2e-05,
+      "loss": 1.0948,
+      "step": 4086
+    },
+    {
+      "epoch": 0.097281729029801,
+      "grad_norm": 9.815671469873866,
+      "learning_rate": 2e-05,
+      "loss": 1.2054,
+      "step": 4087
+    },
+    {
+      "epoch": 0.09730553175283252,
+      "grad_norm": 8.499765971371737,
+      "learning_rate": 2e-05,
+      "loss": 0.2743,
+      "step": 4088
+    },
+    {
+      "epoch": 0.09732933447586403,
+      "grad_norm": 5.888940461836281,
+      "learning_rate": 2e-05,
+      "loss": 0.6742,
+      "step": 4089
+    },
+    {
+      "epoch": 0.09735313719889556,
+      "grad_norm": 7.602347457335577,
+      "learning_rate": 2e-05,
+      "loss": 0.7857,
+      "step": 4090
+    },
+    {
+      "epoch": 0.09737693992192707,
+      "grad_norm": 4.006628303852954,
+      "learning_rate": 2e-05,
+      "loss": 0.2849,
+      "step": 4091
+    },
+    {
+      "epoch": 0.09740074264495859,
+      "grad_norm": 3.045712499053027,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 4092
+    },
+    {
+      "epoch": 0.0974245453679901,
+      "grad_norm": 5.6086240809093475,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 4093
+    },
+    {
+      "epoch": 0.09744834809102161,
+      "grad_norm": 6.138670808743818,
+      "learning_rate": 2e-05,
+      "loss": 1.0904,
+      "step": 4094
+    },
+    {
+      "epoch": 0.09747215081405312,
+      "grad_norm": 5.56740921661442,
+      "learning_rate": 2e-05,
+      "loss": 0.6987,
+      "step": 4095
+    },
+    {
+      "epoch": 0.09749595353708464,
+      "grad_norm": 4.12441449455339,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 4096
+    },
+    {
+      "epoch": 0.09751975626011616,
+      "grad_norm": 6.342213736962981,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 4097
+    },
+    {
+      "epoch": 0.09754355898314768,
+      "grad_norm": 3.0877479412418882,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 4098
+    },
+    {
+      "epoch": 0.09756736170617919,
+      "grad_norm": 6.579384311440069,
+      "learning_rate": 2e-05,
+      "loss": 0.9949,
+      "step": 4099
+    },
+    {
+      "epoch": 0.0975911644292107,
+      "grad_norm": 6.399926693542633,
+      "learning_rate": 2e-05,
+      "loss": 0.7586,
+      "step": 4100
+    },
+    {
+      "epoch": 0.09761496715224222,
+      "grad_norm": 7.139586222624733,
+      "learning_rate": 2e-05,
+      "loss": 1.1619,
+      "step": 4101
+    },
+    {
+      "epoch": 0.09763876987527373,
+      "grad_norm": 1.836272449533921,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 4102
+    },
+    {
+      "epoch": 0.09766257259830524,
+      "grad_norm": 2.46939620401521,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 4103
+    },
+    {
+      "epoch": 0.09768637532133675,
+      "grad_norm": 4.686177969738309,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 4104
+    },
+    {
+      "epoch": 0.09771017804436828,
+      "grad_norm": 6.8481526006758235,
+      "learning_rate": 2e-05,
+      "loss": 1.4206,
+      "step": 4105
+    },
+    {
+      "epoch": 0.0977339807673998,
+      "grad_norm": 6.6544445273506865,
+      "learning_rate": 2e-05,
+      "loss": 0.4464,
+      "step": 4106
+    },
+    {
+      "epoch": 0.09775778349043131,
+      "grad_norm": 3.135897809308402,
+      "learning_rate": 2e-05,
+      "loss": 0.2242,
+      "step": 4107
+    },
+    {
+      "epoch": 0.09778158621346282,
+      "grad_norm": 3.27135439112185,
+      "learning_rate": 2e-05,
+      "loss": 0.2938,
+      "step": 4108
+    },
+    {
+      "epoch": 0.09780538893649433,
+      "grad_norm": 2.323977406376225,
+      "learning_rate": 2e-05,
+      "loss": 0.1866,
+      "step": 4109
+    },
+    {
+      "epoch": 0.09782919165952585,
+      "grad_norm": 6.689120237767561,
+      "learning_rate": 2e-05,
+      "loss": 1.3718,
+      "step": 4110
+    },
+    {
+      "epoch": 0.09785299438255736,
+      "grad_norm": 5.584133117990477,
+      "learning_rate": 2e-05,
+      "loss": 0.8626,
+      "step": 4111
+    },
+    {
+      "epoch": 0.09787679710558887,
+      "grad_norm": 5.229244891862241,
+      "learning_rate": 2e-05,
+      "loss": 0.3059,
+      "step": 4112
+    },
+    {
+      "epoch": 0.0979005998286204,
+      "grad_norm": 5.9545600168983945,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 4113
+    },
+    {
+      "epoch": 0.09792440255165191,
+      "grad_norm": 5.938333821243166,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 4114
+    },
+    {
+      "epoch": 0.09794820527468343,
+      "grad_norm": 11.650206797036915,
+      "learning_rate": 2e-05,
+      "loss": 1.3224,
+      "step": 4115
+    },
+    {
+      "epoch": 0.09797200799771494,
+      "grad_norm": 2.2153254491142476,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 4116
+    },
+    {
+      "epoch": 0.09799581072074645,
+      "grad_norm": 4.609590751656729,
+      "learning_rate": 2e-05,
+      "loss": 0.2619,
+      "step": 4117
+    },
+    {
+      "epoch": 0.09801961344377796,
+      "grad_norm": 6.055370607457932,
+      "learning_rate": 2e-05,
+      "loss": 1.0392,
+      "step": 4118
+    },
+    {
+      "epoch": 0.09804341616680948,
+      "grad_norm": 7.051551356308424,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 4119
+    },
+    {
+      "epoch": 0.098067218889841,
+      "grad_norm": 6.582992131199558,
+      "learning_rate": 2e-05,
+      "loss": 0.363,
+      "step": 4120
+    },
+    {
+      "epoch": 0.09809102161287252,
+      "grad_norm": 3.5705768544797727,
+      "learning_rate": 2e-05,
+      "loss": 0.2594,
+      "step": 4121
+    },
+    {
+      "epoch": 0.09811482433590403,
+      "grad_norm": 8.840927230954943,
+      "learning_rate": 2e-05,
+      "loss": 1.1392,
+      "step": 4122
+    },
+    {
+      "epoch": 0.09813862705893554,
+      "grad_norm": 6.864256839084813,
+      "learning_rate": 2e-05,
+      "loss": 0.4797,
+      "step": 4123
+    },
+    {
+      "epoch": 0.09816242978196706,
+      "grad_norm": 2.712644594071356,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 4124
+    },
+    {
+      "epoch": 0.09818623250499857,
+      "grad_norm": 3.3785276521672962,
+      "learning_rate": 2e-05,
+      "loss": 0.2943,
+      "step": 4125
+    },
+    {
+      "epoch": 0.09821003522803008,
+      "grad_norm": 6.002983012445383,
+      "learning_rate": 2e-05,
+      "loss": 0.7786,
+      "step": 4126
+    },
+    {
+      "epoch": 0.0982338379510616,
+      "grad_norm": 4.78130633973861,
+      "learning_rate": 2e-05,
+      "loss": 0.443,
+      "step": 4127
+    },
+    {
+      "epoch": 0.09825764067409312,
+      "grad_norm": 7.576062760786574,
+      "learning_rate": 2e-05,
+      "loss": 1.4029,
+      "step": 4128
+    },
+    {
+      "epoch": 0.09828144339712463,
+      "grad_norm": 3.3908919174440078,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 4129
+    },
+    {
+      "epoch": 0.09830524612015615,
+      "grad_norm": 2.1987517787150694,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 4130
+    },
+    {
+      "epoch": 0.09832904884318766,
+      "grad_norm": 2.849653135779921,
+      "learning_rate": 2e-05,
+      "loss": 0.2563,
+      "step": 4131
+    },
+    {
+      "epoch": 0.09835285156621917,
+      "grad_norm": 3.8114284380075993,
+      "learning_rate": 2e-05,
+      "loss": 0.4142,
+      "step": 4132
+    },
+    {
+      "epoch": 0.09837665428925069,
+      "grad_norm": 1.3424071215533377,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 4133
+    },
+    {
+      "epoch": 0.0984004570122822,
+      "grad_norm": 3.464903451384497,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 4134
+    },
+    {
+      "epoch": 0.09842425973531373,
+      "grad_norm": 2.9492038820617728,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 4135
+    },
+    {
+      "epoch": 0.09844806245834524,
+      "grad_norm": 5.892760364819779,
+      "learning_rate": 2e-05,
+      "loss": 1.0177,
+      "step": 4136
+    },
+    {
+      "epoch": 0.09847186518137675,
+      "grad_norm": 3.7119526100626064,
+      "learning_rate": 2e-05,
+      "loss": 0.1882,
+      "step": 4137
+    },
+    {
+      "epoch": 0.09849566790440827,
+      "grad_norm": 13.480055551530807,
+      "learning_rate": 2e-05,
+      "loss": 0.7694,
+      "step": 4138
+    },
+    {
+      "epoch": 0.09851947062743978,
+      "grad_norm": 6.100450065538307,
+      "learning_rate": 2e-05,
+      "loss": 0.7988,
+      "step": 4139
+    },
+    {
+      "epoch": 0.09854327335047129,
+      "grad_norm": 4.214600936300562,
+      "learning_rate": 2e-05,
+      "loss": 0.1607,
+      "step": 4140
+    },
+    {
+      "epoch": 0.0985670760735028,
+      "grad_norm": 9.810060299156488,
+      "learning_rate": 2e-05,
+      "loss": 1.6439,
+      "step": 4141
+    },
+    {
+      "epoch": 0.09859087879653432,
+      "grad_norm": 7.023689369173949,
+      "learning_rate": 2e-05,
+      "loss": 0.4247,
+      "step": 4142
+    },
+    {
+      "epoch": 0.09861468151956584,
+      "grad_norm": 3.568526360789418,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 4143
+    },
+    {
+      "epoch": 0.09863848424259736,
+      "grad_norm": 5.617041686195628,
+      "learning_rate": 2e-05,
+      "loss": 0.6386,
+      "step": 4144
+    },
+    {
+      "epoch": 0.09866228696562887,
+      "grad_norm": 7.244145649433793,
+      "learning_rate": 2e-05,
+      "loss": 0.6613,
+      "step": 4145
+    },
+    {
+      "epoch": 0.09868608968866038,
+      "grad_norm": 6.642643851030386,
+      "learning_rate": 2e-05,
+      "loss": 0.9044,
+      "step": 4146
+    },
+    {
+      "epoch": 0.0987098924116919,
+      "grad_norm": 7.272527490227795,
+      "learning_rate": 2e-05,
+      "loss": 0.6111,
+      "step": 4147
+    },
+    {
+      "epoch": 0.09873369513472341,
+      "grad_norm": 8.989607016097919,
+      "learning_rate": 2e-05,
+      "loss": 0.2956,
+      "step": 4148
+    },
+    {
+      "epoch": 0.09875749785775492,
+      "grad_norm": 8.766516875434014,
+      "learning_rate": 2e-05,
+      "loss": 0.4527,
+      "step": 4149
+    },
+    {
+      "epoch": 0.09878130058078644,
+      "grad_norm": 2.6339177922680355,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 4150
+    },
+    {
+      "epoch": 0.09880510330381796,
+      "grad_norm": 5.388421928339736,
+      "learning_rate": 2e-05,
+      "loss": 0.7579,
+      "step": 4151
+    },
+    {
+      "epoch": 0.09882890602684948,
+      "grad_norm": 3.5445236000382763,
+      "learning_rate": 2e-05,
+      "loss": 0.2242,
+      "step": 4152
+    },
+    {
+      "epoch": 0.09885270874988099,
+      "grad_norm": 14.420031709892926,
+      "learning_rate": 2e-05,
+      "loss": 0.7431,
+      "step": 4153
+    },
+    {
+      "epoch": 0.0988765114729125,
+      "grad_norm": 2.51314678213112,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 4154
+    },
+    {
+      "epoch": 0.09890031419594401,
+      "grad_norm": 8.81701086842039,
+      "learning_rate": 2e-05,
+      "loss": 1.1912,
+      "step": 4155
+    },
+    {
+      "epoch": 0.09892411691897553,
+      "grad_norm": 2.433234447537164,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 4156
+    },
+    {
+      "epoch": 0.09894791964200704,
+      "grad_norm": 1.8112452218386932,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 4157
+    },
+    {
+      "epoch": 0.09897172236503857,
+      "grad_norm": 8.616163699198587,
+      "learning_rate": 2e-05,
+      "loss": 0.792,
+      "step": 4158
+    },
+    {
+      "epoch": 0.09899552508807008,
+      "grad_norm": 3.7768721850998963,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 4159
+    },
+    {
+      "epoch": 0.09901932781110159,
+      "grad_norm": 10.019722656047998,
+      "learning_rate": 2e-05,
+      "loss": 1.1744,
+      "step": 4160
+    },
+    {
+      "epoch": 0.0990431305341331,
+      "grad_norm": 0.9215369454392178,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 4161
+    },
+    {
+      "epoch": 0.09906693325716462,
+      "grad_norm": 5.570771341044712,
+      "learning_rate": 2e-05,
+      "loss": 0.5488,
+      "step": 4162
+    },
+    {
+      "epoch": 0.09909073598019613,
+      "grad_norm": 6.197372157710824,
+      "learning_rate": 2e-05,
+      "loss": 0.3016,
+      "step": 4163
+    },
+    {
+      "epoch": 0.09911453870322764,
+      "grad_norm": 9.035034259706148,
+      "learning_rate": 2e-05,
+      "loss": 1.5522,
+      "step": 4164
+    },
+    {
+      "epoch": 0.09913834142625916,
+      "grad_norm": 3.8225229258990443,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 4165
+    },
+    {
+      "epoch": 0.09916214414929068,
+      "grad_norm": 3.981233387724312,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 4166
+    },
+    {
+      "epoch": 0.0991859468723222,
+      "grad_norm": 3.892424782186979,
+      "learning_rate": 2e-05,
+      "loss": 0.2216,
+      "step": 4167
+    },
+    {
+      "epoch": 0.09920974959535371,
+      "grad_norm": 11.65766904353416,
+      "learning_rate": 2e-05,
+      "loss": 0.9664,
+      "step": 4168
+    },
+    {
+      "epoch": 0.09923355231838522,
+      "grad_norm": 4.658343680809487,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 4169
+    },
+    {
+      "epoch": 0.09925735504141674,
+      "grad_norm": 2.862839702239789,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 4170
+    },
+    {
+      "epoch": 0.09928115776444825,
+      "grad_norm": 5.473060945205511,
+      "learning_rate": 2e-05,
+      "loss": 0.3856,
+      "step": 4171
+    },
+    {
+      "epoch": 0.09930496048747976,
+      "grad_norm": 4.861538763915929,
+      "learning_rate": 2e-05,
+      "loss": 0.7172,
+      "step": 4172
+    },
+    {
+      "epoch": 0.09932876321051129,
+      "grad_norm": 4.724091932600855,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 4173
+    },
+    {
+      "epoch": 0.0993525659335428,
+      "grad_norm": 6.4585652460997585,
+      "learning_rate": 2e-05,
+      "loss": 0.33,
+      "step": 4174
+    },
+    {
+      "epoch": 0.09937636865657432,
+      "grad_norm": 10.58027493115218,
+      "learning_rate": 2e-05,
+      "loss": 0.9169,
+      "step": 4175
+    },
+    {
+      "epoch": 0.09940017137960583,
+      "grad_norm": 2.1464247017161484,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 4176
+    },
+    {
+      "epoch": 0.09942397410263734,
+      "grad_norm": 4.438473015214491,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 4177
+    },
+    {
+      "epoch": 0.09944777682566885,
+      "grad_norm": 10.652207237396144,
+      "learning_rate": 2e-05,
+      "loss": 0.3791,
+      "step": 4178
+    },
+    {
+      "epoch": 0.09947157954870037,
+      "grad_norm": 4.002305683625052,
+      "learning_rate": 2e-05,
+      "loss": 0.2132,
+      "step": 4179
+    },
+    {
+      "epoch": 0.09949538227173188,
+      "grad_norm": 6.438196626034537,
+      "learning_rate": 2e-05,
+      "loss": 1.1303,
+      "step": 4180
+    },
+    {
+      "epoch": 0.09951918499476341,
+      "grad_norm": 2.405560651520862,
+      "learning_rate": 2e-05,
+      "loss": 0.242,
+      "step": 4181
+    },
+    {
+      "epoch": 0.09954298771779492,
+      "grad_norm": 3.369634276799881,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 4182
+    },
+    {
+      "epoch": 0.09956679044082643,
+      "grad_norm": 7.886553084656275,
+      "learning_rate": 2e-05,
+      "loss": 1.2897,
+      "step": 4183
+    },
+    {
+      "epoch": 0.09959059316385795,
+      "grad_norm": 3.3596040530828186,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 4184
+    },
+    {
+      "epoch": 0.09961439588688946,
+      "grad_norm": 1.9716295221165299,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 4185
+    },
+    {
+      "epoch": 0.09963819860992097,
+      "grad_norm": 6.164677401209043,
+      "learning_rate": 2e-05,
+      "loss": 0.2916,
+      "step": 4186
+    },
+    {
+      "epoch": 0.09966200133295249,
+      "grad_norm": 0.8722380706416086,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 4187
+    },
+    {
+      "epoch": 0.099685804055984,
+      "grad_norm": 3.9944560076529276,
+      "learning_rate": 2e-05,
+      "loss": 0.344,
+      "step": 4188
+    },
+    {
+      "epoch": 0.09970960677901552,
+      "grad_norm": 3.3457346761056885,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 4189
+    },
+    {
+      "epoch": 0.09973340950204704,
+      "grad_norm": 9.74401234215267,
+      "learning_rate": 2e-05,
+      "loss": 1.31,
+      "step": 4190
+    },
+    {
+      "epoch": 0.09975721222507855,
+      "grad_norm": 8.685110800763276,
+      "learning_rate": 2e-05,
+      "loss": 1.023,
+      "step": 4191
+    },
+    {
+      "epoch": 0.09978101494811006,
+      "grad_norm": 3.744032234541835,
+      "learning_rate": 2e-05,
+      "loss": 0.3068,
+      "step": 4192
+    },
+    {
+      "epoch": 0.09980481767114158,
+      "grad_norm": 8.17593328869064,
+      "learning_rate": 2e-05,
+      "loss": 0.4455,
+      "step": 4193
+    },
+    {
+      "epoch": 0.09982862039417309,
+      "grad_norm": 3.0839447049195035,
+      "learning_rate": 2e-05,
+      "loss": 0.2355,
+      "step": 4194
+    },
+    {
+      "epoch": 0.0998524231172046,
+      "grad_norm": 7.641652270823548,
+      "learning_rate": 2e-05,
+      "loss": 1.0226,
+      "step": 4195
+    },
+    {
+      "epoch": 0.09987622584023613,
+      "grad_norm": 6.710938822782175,
+      "learning_rate": 2e-05,
+      "loss": 0.8986,
+      "step": 4196
+    },
+    {
+      "epoch": 0.09990002856326764,
+      "grad_norm": 5.074299566917188,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 4197
+    },
+    {
+      "epoch": 0.09992383128629916,
+      "grad_norm": 6.143797679383359,
+      "learning_rate": 2e-05,
+      "loss": 1.0527,
+      "step": 4198
+    },
+    {
+      "epoch": 0.09994763400933067,
+      "grad_norm": 8.003706887502991,
+      "learning_rate": 2e-05,
+      "loss": 0.6219,
+      "step": 4199
+    },
+    {
+      "epoch": 0.09997143673236218,
+      "grad_norm": 7.398139890496236,
+      "learning_rate": 2e-05,
+      "loss": 0.394,
+      "step": 4200
+    },
+    {
+      "epoch": 0.0999952394553937,
+      "grad_norm": 4.402825955477081,
+      "learning_rate": 2e-05,
+      "loss": 0.7308,
+      "step": 4201
+    },
+    {
+      "epoch": 0.10001904217842521,
+      "grad_norm": 2.3627633539149633,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 4202
+    },
+    {
+      "epoch": 0.10004284490145672,
+      "grad_norm": 6.314043577866975,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 4203
+    },
+    {
+      "epoch": 0.10006664762448825,
+      "grad_norm": 7.346297208071099,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 4204
+    },
+    {
+      "epoch": 0.10009045034751976,
+      "grad_norm": 2.8648445126730384,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 4205
+    },
+    {
+      "epoch": 0.10011425307055127,
+      "grad_norm": 4.144200441723592,
+      "learning_rate": 2e-05,
+      "loss": 0.4339,
+      "step": 4206
+    },
+    {
+      "epoch": 0.10013805579358279,
+      "grad_norm": 5.320831427976409,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 4207
+    },
+    {
+      "epoch": 0.1001618585166143,
+      "grad_norm": 4.273004711572983,
+      "learning_rate": 2e-05,
+      "loss": 0.3329,
+      "step": 4208
+    },
+    {
+      "epoch": 0.10018566123964581,
+      "grad_norm": 3.1628241408603706,
+      "learning_rate": 2e-05,
+      "loss": 0.2837,
+      "step": 4209
+    },
+    {
+      "epoch": 0.10020946396267733,
+      "grad_norm": 5.887318145650522,
+      "learning_rate": 2e-05,
+      "loss": 0.4347,
+      "step": 4210
+    },
+    {
+      "epoch": 0.10023326668570884,
+      "grad_norm": 7.544515577158723,
+      "learning_rate": 2e-05,
+      "loss": 0.608,
+      "step": 4211
+    },
+    {
+      "epoch": 0.10025706940874037,
+      "grad_norm": 4.650092849729776,
+      "learning_rate": 2e-05,
+      "loss": 0.9931,
+      "step": 4212
+    },
+    {
+      "epoch": 0.10028087213177188,
+      "grad_norm": 3.624189477664266,
+      "learning_rate": 2e-05,
+      "loss": 0.2948,
+      "step": 4213
+    },
+    {
+      "epoch": 0.10030467485480339,
+      "grad_norm": 8.100908070262829,
+      "learning_rate": 2e-05,
+      "loss": 0.4352,
+      "step": 4214
+    },
+    {
+      "epoch": 0.1003284775778349,
+      "grad_norm": 13.860885769160399,
+      "learning_rate": 2e-05,
+      "loss": 1.0012,
+      "step": 4215
+    },
+    {
+      "epoch": 0.10035228030086642,
+      "grad_norm": 3.9506779727295656,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 4216
+    },
+    {
+      "epoch": 0.10037608302389793,
+      "grad_norm": 4.800422737461092,
+      "learning_rate": 2e-05,
+      "loss": 0.6741,
+      "step": 4217
+    },
+    {
+      "epoch": 0.10039988574692944,
+      "grad_norm": 2.782838345095043,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 4218
+    },
+    {
+      "epoch": 0.10042368846996097,
+      "grad_norm": 6.644859518087991,
+      "learning_rate": 2e-05,
+      "loss": 0.3482,
+      "step": 4219
+    },
+    {
+      "epoch": 0.10044749119299248,
+      "grad_norm": 5.1715137230044785,
+      "learning_rate": 2e-05,
+      "loss": 0.2707,
+      "step": 4220
+    },
+    {
+      "epoch": 0.100471293916024,
+      "grad_norm": 1.6243683150548407,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 4221
+    },
+    {
+      "epoch": 0.10049509663905551,
+      "grad_norm": 3.898227037998811,
+      "learning_rate": 2e-05,
+      "loss": 0.6029,
+      "step": 4222
+    },
+    {
+      "epoch": 0.10051889936208702,
+      "grad_norm": 7.2074596335243575,
+      "learning_rate": 2e-05,
+      "loss": 0.4022,
+      "step": 4223
+    },
+    {
+      "epoch": 0.10054270208511853,
+      "grad_norm": 2.720826846667429,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 4224
+    },
+    {
+      "epoch": 0.10056650480815005,
+      "grad_norm": 5.640128553623995,
+      "learning_rate": 2e-05,
+      "loss": 0.6547,
+      "step": 4225
+    },
+    {
+      "epoch": 0.10059030753118156,
+      "grad_norm": 5.752658930786399,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 4226
+    },
+    {
+      "epoch": 0.10061411025421309,
+      "grad_norm": 4.97745664290128,
+      "learning_rate": 2e-05,
+      "loss": 0.686,
+      "step": 4227
+    },
+    {
+      "epoch": 0.1006379129772446,
+      "grad_norm": 2.8192786009815114,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 4228
+    },
+    {
+      "epoch": 0.10066171570027611,
+      "grad_norm": 7.249390836962041,
+      "learning_rate": 2e-05,
+      "loss": 0.6459,
+      "step": 4229
+    },
+    {
+      "epoch": 0.10068551842330763,
+      "grad_norm": 6.311021098484708,
+      "learning_rate": 2e-05,
+      "loss": 0.6049,
+      "step": 4230
+    },
+    {
+      "epoch": 0.10070932114633914,
+      "grad_norm": 5.213922162131959,
+      "learning_rate": 2e-05,
+      "loss": 0.4923,
+      "step": 4231
+    },
+    {
+      "epoch": 0.10073312386937065,
+      "grad_norm": 3.454908826515109,
+      "learning_rate": 2e-05,
+      "loss": 0.2406,
+      "step": 4232
+    },
+    {
+      "epoch": 0.10075692659240217,
+      "grad_norm": 4.49605795203601,
+      "learning_rate": 2e-05,
+      "loss": 0.4992,
+      "step": 4233
+    },
+    {
+      "epoch": 0.10078072931543369,
+      "grad_norm": 9.527724282522104,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1008045320384652,
+      "grad_norm": 12.55368003785225,
+      "learning_rate": 2e-05,
+      "loss": 0.8258,
+      "step": 4235
+    },
+    {
+      "epoch": 0.10082833476149672,
+      "grad_norm": 9.02713262094397,
+      "learning_rate": 2e-05,
+      "loss": 0.9114,
+      "step": 4236
+    },
+    {
+      "epoch": 0.10085213748452823,
+      "grad_norm": 3.42624918419604,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 4237
+    },
+    {
+      "epoch": 0.10087594020755974,
+      "grad_norm": 3.738742936756702,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 4238
+    },
+    {
+      "epoch": 0.10089974293059126,
+      "grad_norm": 6.717366502884815,
+      "learning_rate": 2e-05,
+      "loss": 0.7423,
+      "step": 4239
+    },
+    {
+      "epoch": 0.10092354565362277,
+      "grad_norm": 6.036533264625754,
+      "learning_rate": 2e-05,
+      "loss": 0.3111,
+      "step": 4240
+    },
+    {
+      "epoch": 0.10094734837665428,
+      "grad_norm": 5.7997001978446665,
+      "learning_rate": 2e-05,
+      "loss": 0.5195,
+      "step": 4241
+    },
+    {
+      "epoch": 0.10097115109968581,
+      "grad_norm": 6.451263725048694,
+      "learning_rate": 2e-05,
+      "loss": 0.4292,
+      "step": 4242
+    },
+    {
+      "epoch": 0.10099495382271732,
+      "grad_norm": 2.797924083575135,
+      "learning_rate": 2e-05,
+      "loss": 0.1613,
+      "step": 4243
+    },
+    {
+      "epoch": 0.10101875654574884,
+      "grad_norm": 7.494120732593407,
+      "learning_rate": 2e-05,
+      "loss": 0.5259,
+      "step": 4244
+    },
+    {
+      "epoch": 0.10104255926878035,
+      "grad_norm": 2.971317151140312,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 4245
+    },
+    {
+      "epoch": 0.10106636199181186,
+      "grad_norm": 6.579983544050389,
+      "learning_rate": 2e-05,
+      "loss": 0.9802,
+      "step": 4246
+    },
+    {
+      "epoch": 0.10109016471484338,
+      "grad_norm": 1.6168801207351438,
+      "learning_rate": 2e-05,
+      "loss": 0.1971,
+      "step": 4247
+    },
+    {
+      "epoch": 0.10111396743787489,
+      "grad_norm": 6.107166792689191,
+      "learning_rate": 2e-05,
+      "loss": 0.7317,
+      "step": 4248
+    },
+    {
+      "epoch": 0.1011377701609064,
+      "grad_norm": 7.1667477324184565,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 4249
+    },
+    {
+      "epoch": 0.10116157288393793,
+      "grad_norm": 5.577806856357938,
+      "learning_rate": 2e-05,
+      "loss": 0.302,
+      "step": 4250
+    },
+    {
+      "epoch": 0.10118537560696944,
+      "grad_norm": 5.783981669220383,
+      "learning_rate": 2e-05,
+      "loss": 0.2614,
+      "step": 4251
+    },
+    {
+      "epoch": 0.10120917833000095,
+      "grad_norm": 6.54375132074833,
+      "learning_rate": 2e-05,
+      "loss": 0.3319,
+      "step": 4252
+    },
+    {
+      "epoch": 0.10123298105303247,
+      "grad_norm": 6.640743135398611,
+      "learning_rate": 2e-05,
+      "loss": 0.3115,
+      "step": 4253
+    },
+    {
+      "epoch": 0.10125678377606398,
+      "grad_norm": 2.0342982784197035,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 4254
+    },
+    {
+      "epoch": 0.10128058649909549,
+      "grad_norm": 6.031194021647222,
+      "learning_rate": 2e-05,
+      "loss": 0.3436,
+      "step": 4255
+    },
+    {
+      "epoch": 0.101304389222127,
+      "grad_norm": 3.2157024550388855,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 4256
+    },
+    {
+      "epoch": 0.10132819194515853,
+      "grad_norm": 2.5094309785430755,
+      "learning_rate": 2e-05,
+      "loss": 0.2598,
+      "step": 4257
+    },
+    {
+      "epoch": 0.10135199466819005,
+      "grad_norm": 7.648102218339299,
+      "learning_rate": 2e-05,
+      "loss": 1.175,
+      "step": 4258
+    },
+    {
+      "epoch": 0.10137579739122156,
+      "grad_norm": 6.7154936277758965,
+      "learning_rate": 2e-05,
+      "loss": 0.8663,
+      "step": 4259
+    },
+    {
+      "epoch": 0.10139960011425307,
+      "grad_norm": 9.277326474802441,
+      "learning_rate": 2e-05,
+      "loss": 1.246,
+      "step": 4260
+    },
+    {
+      "epoch": 0.10142340283728458,
+      "grad_norm": 5.382826700114263,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1014472055603161,
+      "grad_norm": 10.88524653486083,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 4262
+    },
+    {
+      "epoch": 0.10147100828334761,
+      "grad_norm": 4.048137579603649,
+      "learning_rate": 2e-05,
+      "loss": 0.4634,
+      "step": 4263
+    },
+    {
+      "epoch": 0.10149481100637912,
+      "grad_norm": 8.395009236097021,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 4264
+    },
+    {
+      "epoch": 0.10151861372941065,
+      "grad_norm": 3.588889273287386,
+      "learning_rate": 2e-05,
+      "loss": 0.3605,
+      "step": 4265
+    },
+    {
+      "epoch": 0.10154241645244216,
+      "grad_norm": 8.040764223023539,
+      "learning_rate": 2e-05,
+      "loss": 1.0095,
+      "step": 4266
+    },
+    {
+      "epoch": 0.10156621917547368,
+      "grad_norm": 2.800588495950113,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 4267
+    },
+    {
+      "epoch": 0.10159002189850519,
+      "grad_norm": 5.214344840301342,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 4268
+    },
+    {
+      "epoch": 0.1016138246215367,
+      "grad_norm": 2.8935458874403137,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 4269
+    },
+    {
+      "epoch": 0.10163762734456822,
+      "grad_norm": 4.419285584958037,
+      "learning_rate": 2e-05,
+      "loss": 0.5278,
+      "step": 4270
+    },
+    {
+      "epoch": 0.10166143006759973,
+      "grad_norm": 3.1955457220279277,
+      "learning_rate": 2e-05,
+      "loss": 0.2369,
+      "step": 4271
+    },
+    {
+      "epoch": 0.10168523279063124,
+      "grad_norm": 8.489120987153896,
+      "learning_rate": 2e-05,
+      "loss": 0.9712,
+      "step": 4272
+    },
+    {
+      "epoch": 0.10170903551366277,
+      "grad_norm": 6.163639492633388,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 4273
+    },
+    {
+      "epoch": 0.10173283823669428,
+      "grad_norm": 6.975006505770845,
+      "learning_rate": 2e-05,
+      "loss": 1.5364,
+      "step": 4274
+    },
+    {
+      "epoch": 0.1017566409597258,
+      "grad_norm": 7.839232971603705,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 4275
+    },
+    {
+      "epoch": 0.10178044368275731,
+      "grad_norm": 2.7972882206444853,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 4276
+    },
+    {
+      "epoch": 0.10180424640578882,
+      "grad_norm": 3.75206322281138,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 4277
+    },
+    {
+      "epoch": 0.10182804912882033,
+      "grad_norm": 1.9433013077627985,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 4278
+    },
+    {
+      "epoch": 0.10185185185185185,
+      "grad_norm": 6.019172527173623,
+      "learning_rate": 2e-05,
+      "loss": 0.4442,
+      "step": 4279
+    },
+    {
+      "epoch": 0.10187565457488337,
+      "grad_norm": 3.7188775171222295,
+      "learning_rate": 2e-05,
+      "loss": 0.3,
+      "step": 4280
+    },
+    {
+      "epoch": 0.10189945729791489,
+      "grad_norm": 14.010744180589736,
+      "learning_rate": 2e-05,
+      "loss": 1.5035,
+      "step": 4281
+    },
+    {
+      "epoch": 0.1019232600209464,
+      "grad_norm": 7.6001109285932875,
+      "learning_rate": 2e-05,
+      "loss": 0.4156,
+      "step": 4282
+    },
+    {
+      "epoch": 0.10194706274397791,
+      "grad_norm": 1.3852578081291793,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 4283
+    },
+    {
+      "epoch": 0.10197086546700942,
+      "grad_norm": 3.313303809356776,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 4284
+    },
+    {
+      "epoch": 0.10199466819004094,
+      "grad_norm": 18.64016553620019,
+      "learning_rate": 2e-05,
+      "loss": 1.0683,
+      "step": 4285
+    },
+    {
+      "epoch": 0.10201847091307245,
+      "grad_norm": 2.974011831848878,
+      "learning_rate": 2e-05,
+      "loss": 0.1979,
+      "step": 4286
+    },
+    {
+      "epoch": 0.10204227363610396,
+      "grad_norm": 6.948799284855236,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 4287
+    },
+    {
+      "epoch": 0.10206607635913549,
+      "grad_norm": 1.6356018475143461,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 4288
+    },
+    {
+      "epoch": 0.102089879082167,
+      "grad_norm": 5.8561929208381605,
+      "learning_rate": 2e-05,
+      "loss": 0.6791,
+      "step": 4289
+    },
+    {
+      "epoch": 0.10211368180519852,
+      "grad_norm": 6.356507605578382,
+      "learning_rate": 2e-05,
+      "loss": 0.3479,
+      "step": 4290
+    },
+    {
+      "epoch": 0.10213748452823003,
+      "grad_norm": 8.504733930501022,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 4291
+    },
+    {
+      "epoch": 0.10216128725126154,
+      "grad_norm": 4.604782609492999,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 4292
+    },
+    {
+      "epoch": 0.10218508997429306,
+      "grad_norm": 3.8444782912740623,
+      "learning_rate": 2e-05,
+      "loss": 0.3463,
+      "step": 4293
+    },
+    {
+      "epoch": 0.10220889269732457,
+      "grad_norm": 4.05801586209013,
+      "learning_rate": 2e-05,
+      "loss": 0.2731,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1022326954203561,
+      "grad_norm": 2.2290945323894884,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 4295
+    },
+    {
+      "epoch": 0.10225649814338761,
+      "grad_norm": 6.702278433214842,
+      "learning_rate": 2e-05,
+      "loss": 1.0952,
+      "step": 4296
+    },
+    {
+      "epoch": 0.10228030086641912,
+      "grad_norm": 6.636489712427373,
+      "learning_rate": 2e-05,
+      "loss": 1.2805,
+      "step": 4297
+    },
+    {
+      "epoch": 0.10230410358945063,
+      "grad_norm": 6.633538593119425,
+      "learning_rate": 2e-05,
+      "loss": 0.2715,
+      "step": 4298
+    },
+    {
+      "epoch": 0.10232790631248215,
+      "grad_norm": 2.1241344029551694,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 4299
+    },
+    {
+      "epoch": 0.10235170903551366,
+      "grad_norm": 4.523768679655466,
+      "learning_rate": 2e-05,
+      "loss": 0.3543,
+      "step": 4300
+    },
+    {
+      "epoch": 0.10237551175854517,
+      "grad_norm": 6.493990851560746,
+      "learning_rate": 2e-05,
+      "loss": 1.142,
+      "step": 4301
+    },
+    {
+      "epoch": 0.10239931448157669,
+      "grad_norm": 6.260827758070583,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 4302
+    },
+    {
+      "epoch": 0.10242311720460821,
+      "grad_norm": 2.344872090369988,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 4303
+    },
+    {
+      "epoch": 0.10244691992763973,
+      "grad_norm": 2.863083016360404,
+      "learning_rate": 2e-05,
+      "loss": 0.239,
+      "step": 4304
+    },
+    {
+      "epoch": 0.10247072265067124,
+      "grad_norm": 4.305798433603229,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 4305
+    },
+    {
+      "epoch": 0.10249452537370275,
+      "grad_norm": 8.70886390542185,
+      "learning_rate": 2e-05,
+      "loss": 1.5855,
+      "step": 4306
+    },
+    {
+      "epoch": 0.10251832809673427,
+      "grad_norm": 6.5383634427505255,
+      "learning_rate": 2e-05,
+      "loss": 1.0038,
+      "step": 4307
+    },
+    {
+      "epoch": 0.10254213081976578,
+      "grad_norm": 2.9347306818971637,
+      "learning_rate": 2e-05,
+      "loss": 0.245,
+      "step": 4308
+    },
+    {
+      "epoch": 0.10256593354279729,
+      "grad_norm": 5.0483592050591595,
+      "learning_rate": 2e-05,
+      "loss": 0.794,
+      "step": 4309
+    },
+    {
+      "epoch": 0.1025897362658288,
+      "grad_norm": 2.9506913006847255,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 4310
+    },
+    {
+      "epoch": 0.10261353898886033,
+      "grad_norm": 1.5594692756117337,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 4311
+    },
+    {
+      "epoch": 0.10263734171189184,
+      "grad_norm": 3.3908589282960473,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 4312
+    },
+    {
+      "epoch": 0.10266114443492336,
+      "grad_norm": 4.029232093939132,
+      "learning_rate": 2e-05,
+      "loss": 0.2883,
+      "step": 4313
+    },
+    {
+      "epoch": 0.10268494715795487,
+      "grad_norm": 6.4694772421530855,
+      "learning_rate": 2e-05,
+      "loss": 0.3653,
+      "step": 4314
+    },
+    {
+      "epoch": 0.10270874988098638,
+      "grad_norm": 8.287278366568856,
+      "learning_rate": 2e-05,
+      "loss": 1.8922,
+      "step": 4315
+    },
+    {
+      "epoch": 0.1027325526040179,
+      "grad_norm": 3.915535491133157,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 4316
+    },
+    {
+      "epoch": 0.10275635532704941,
+      "grad_norm": 1.8985789051528945,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 4317
+    },
+    {
+      "epoch": 0.10278015805008094,
+      "grad_norm": 9.10245147588317,
+      "learning_rate": 2e-05,
+      "loss": 1.6697,
+      "step": 4318
+    },
+    {
+      "epoch": 0.10280396077311245,
+      "grad_norm": 4.828129956666978,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 4319
+    },
+    {
+      "epoch": 0.10282776349614396,
+      "grad_norm": 9.64021580128078,
+      "learning_rate": 2e-05,
+      "loss": 0.3703,
+      "step": 4320
+    },
+    {
+      "epoch": 0.10285156621917547,
+      "grad_norm": 8.23440100617799,
+      "learning_rate": 2e-05,
+      "loss": 0.804,
+      "step": 4321
+    },
+    {
+      "epoch": 0.10287536894220699,
+      "grad_norm": 4.827405182870334,
+      "learning_rate": 2e-05,
+      "loss": 0.4798,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1028991716652385,
+      "grad_norm": 5.9831385668913795,
+      "learning_rate": 2e-05,
+      "loss": 0.7585,
+      "step": 4323
+    },
+    {
+      "epoch": 0.10292297438827001,
+      "grad_norm": 2.4629475779224816,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 4324
+    },
+    {
+      "epoch": 0.10294677711130153,
+      "grad_norm": 11.239719899758603,
+      "learning_rate": 2e-05,
+      "loss": 1.2663,
+      "step": 4325
+    },
+    {
+      "epoch": 0.10297057983433305,
+      "grad_norm": 3.970738207178274,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 4326
+    },
+    {
+      "epoch": 0.10299438255736457,
+      "grad_norm": 2.7308136998472117,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 4327
+    },
+    {
+      "epoch": 0.10301818528039608,
+      "grad_norm": 2.6127005400636287,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 4328
+    },
+    {
+      "epoch": 0.10304198800342759,
+      "grad_norm": 2.7176546323344852,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 4329
+    },
+    {
+      "epoch": 0.1030657907264591,
+      "grad_norm": 8.95030299057148,
+      "learning_rate": 2e-05,
+      "loss": 1.1483,
+      "step": 4330
+    },
+    {
+      "epoch": 0.10308959344949062,
+      "grad_norm": 6.871901718516441,
+      "learning_rate": 2e-05,
+      "loss": 0.6433,
+      "step": 4331
+    },
+    {
+      "epoch": 0.10311339617252213,
+      "grad_norm": 8.95996202424848,
+      "learning_rate": 2e-05,
+      "loss": 0.699,
+      "step": 4332
+    },
+    {
+      "epoch": 0.10313719889555364,
+      "grad_norm": 3.8273164903250723,
+      "learning_rate": 2e-05,
+      "loss": 0.2546,
+      "step": 4333
+    },
+    {
+      "epoch": 0.10316100161858517,
+      "grad_norm": 3.631489117502587,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 4334
+    },
+    {
+      "epoch": 0.10318480434161668,
+      "grad_norm": 12.886319479841276,
+      "learning_rate": 2e-05,
+      "loss": 1.6715,
+      "step": 4335
+    },
+    {
+      "epoch": 0.1032086070646482,
+      "grad_norm": 4.233153977327902,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 4336
+    },
+    {
+      "epoch": 0.10323240978767971,
+      "grad_norm": 6.65836022061343,
+      "learning_rate": 2e-05,
+      "loss": 0.9417,
+      "step": 4337
+    },
+    {
+      "epoch": 0.10325621251071122,
+      "grad_norm": 1.179799585347096,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 4338
+    },
+    {
+      "epoch": 0.10328001523374274,
+      "grad_norm": 1.7697331859546632,
+      "learning_rate": 2e-05,
+      "loss": 0.1434,
+      "step": 4339
+    },
+    {
+      "epoch": 0.10330381795677425,
+      "grad_norm": 6.949745700462017,
+      "learning_rate": 2e-05,
+      "loss": 1.6621,
+      "step": 4340
+    },
+    {
+      "epoch": 0.10332762067980578,
+      "grad_norm": 7.7461874877731605,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 4341
+    },
+    {
+      "epoch": 0.10335142340283729,
+      "grad_norm": 2.068446860131958,
+      "learning_rate": 2e-05,
+      "loss": 0.2453,
+      "step": 4342
+    },
+    {
+      "epoch": 0.1033752261258688,
+      "grad_norm": 8.548476596125695,
+      "learning_rate": 2e-05,
+      "loss": 0.9239,
+      "step": 4343
+    },
+    {
+      "epoch": 0.10339902884890032,
+      "grad_norm": 4.720336519857433,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 4344
+    },
+    {
+      "epoch": 0.10342283157193183,
+      "grad_norm": 6.648196538358363,
+      "learning_rate": 2e-05,
+      "loss": 0.2563,
+      "step": 4345
+    },
+    {
+      "epoch": 0.10344663429496334,
+      "grad_norm": 5.286282156076484,
+      "learning_rate": 2e-05,
+      "loss": 0.4883,
+      "step": 4346
+    },
+    {
+      "epoch": 0.10347043701799485,
+      "grad_norm": 2.7801698410894766,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 4347
+    },
+    {
+      "epoch": 0.10349423974102637,
+      "grad_norm": 2.025518904729673,
+      "learning_rate": 2e-05,
+      "loss": 0.1412,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1035180424640579,
+      "grad_norm": 6.494235023207803,
+      "learning_rate": 2e-05,
+      "loss": 0.3456,
+      "step": 4349
+    },
+    {
+      "epoch": 0.1035418451870894,
+      "grad_norm": 10.33040371751668,
+      "learning_rate": 2e-05,
+      "loss": 1.3466,
+      "step": 4350
+    },
+    {
+      "epoch": 0.10356564791012092,
+      "grad_norm": 7.5593402288492735,
+      "learning_rate": 2e-05,
+      "loss": 1.3808,
+      "step": 4351
+    },
+    {
+      "epoch": 0.10358945063315243,
+      "grad_norm": 8.95414371529535,
+      "learning_rate": 2e-05,
+      "loss": 0.2902,
+      "step": 4352
+    },
+    {
+      "epoch": 0.10361325335618395,
+      "grad_norm": 3.3244629900369396,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 4353
+    },
+    {
+      "epoch": 0.10363705607921546,
+      "grad_norm": 4.208796849404146,
+      "learning_rate": 2e-05,
+      "loss": 0.6905,
+      "step": 4354
+    },
+    {
+      "epoch": 0.10366085880224697,
+      "grad_norm": 3.521029434066928,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 4355
+    },
+    {
+      "epoch": 0.1036846615252785,
+      "grad_norm": 6.72761998268911,
+      "learning_rate": 2e-05,
+      "loss": 1.2301,
+      "step": 4356
+    },
+    {
+      "epoch": 0.10370846424831001,
+      "grad_norm": 4.13114341696799,
+      "learning_rate": 2e-05,
+      "loss": 0.2583,
+      "step": 4357
+    },
+    {
+      "epoch": 0.10373226697134152,
+      "grad_norm": 5.17812998822847,
+      "learning_rate": 2e-05,
+      "loss": 0.8244,
+      "step": 4358
+    },
+    {
+      "epoch": 0.10375606969437304,
+      "grad_norm": 2.0630735279742964,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 4359
+    },
+    {
+      "epoch": 0.10377987241740455,
+      "grad_norm": 4.336160410226874,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 4360
+    },
+    {
+      "epoch": 0.10380367514043606,
+      "grad_norm": 1.896442221487095,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 4361
+    },
+    {
+      "epoch": 0.10382747786346758,
+      "grad_norm": 7.130948093324529,
+      "learning_rate": 2e-05,
+      "loss": 0.9405,
+      "step": 4362
+    },
+    {
+      "epoch": 0.10385128058649909,
+      "grad_norm": 3.8123360059186484,
+      "learning_rate": 2e-05,
+      "loss": 0.457,
+      "step": 4363
+    },
+    {
+      "epoch": 0.10387508330953062,
+      "grad_norm": 7.62126797249358,
+      "learning_rate": 2e-05,
+      "loss": 0.3256,
+      "step": 4364
+    },
+    {
+      "epoch": 0.10389888603256213,
+      "grad_norm": 3.6078181001906047,
+      "learning_rate": 2e-05,
+      "loss": 0.3002,
+      "step": 4365
+    },
+    {
+      "epoch": 0.10392268875559364,
+      "grad_norm": 4.2461094127183365,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 4366
+    },
+    {
+      "epoch": 0.10394649147862516,
+      "grad_norm": 1.7643606074591758,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 4367
+    },
+    {
+      "epoch": 0.10397029420165667,
+      "grad_norm": 4.028619948807407,
+      "learning_rate": 2e-05,
+      "loss": 0.2108,
+      "step": 4368
+    },
+    {
+      "epoch": 0.10399409692468818,
+      "grad_norm": 6.551078406483159,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 4369
+    },
+    {
+      "epoch": 0.1040178996477197,
+      "grad_norm": 12.246707328094793,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 4370
+    },
+    {
+      "epoch": 0.10404170237075121,
+      "grad_norm": 8.677117321041464,
+      "learning_rate": 2e-05,
+      "loss": 0.4647,
+      "step": 4371
+    },
+    {
+      "epoch": 0.10406550509378273,
+      "grad_norm": 10.238340608325023,
+      "learning_rate": 2e-05,
+      "loss": 1.5179,
+      "step": 4372
+    },
+    {
+      "epoch": 0.10408930781681425,
+      "grad_norm": 10.399467826515545,
+      "learning_rate": 2e-05,
+      "loss": 1.5667,
+      "step": 4373
+    },
+    {
+      "epoch": 0.10411311053984576,
+      "grad_norm": 1.7632960459386269,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 4374
+    },
+    {
+      "epoch": 0.10413691326287727,
+      "grad_norm": 1.9850315347769596,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 4375
+    },
+    {
+      "epoch": 0.10416071598590879,
+      "grad_norm": 5.778562229225969,
+      "learning_rate": 2e-05,
+      "loss": 0.7731,
+      "step": 4376
+    },
+    {
+      "epoch": 0.1041845187089403,
+      "grad_norm": 1.800067960563204,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 4377
+    },
+    {
+      "epoch": 0.10420832143197181,
+      "grad_norm": 6.818219569872353,
+      "learning_rate": 2e-05,
+      "loss": 0.8785,
+      "step": 4378
+    },
+    {
+      "epoch": 0.10423212415500334,
+      "grad_norm": 3.753265929421643,
+      "learning_rate": 2e-05,
+      "loss": 0.2716,
+      "step": 4379
+    },
+    {
+      "epoch": 0.10425592687803485,
+      "grad_norm": 17.191841562452232,
+      "learning_rate": 2e-05,
+      "loss": 0.9746,
+      "step": 4380
+    },
+    {
+      "epoch": 0.10427972960106636,
+      "grad_norm": 2.721067311602486,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 4381
+    },
+    {
+      "epoch": 0.10430353232409788,
+      "grad_norm": 4.252374953939628,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 4382
+    },
+    {
+      "epoch": 0.10432733504712939,
+      "grad_norm": 17.042525513490517,
+      "learning_rate": 2e-05,
+      "loss": 0.7811,
+      "step": 4383
+    },
+    {
+      "epoch": 0.1043511377701609,
+      "grad_norm": 4.164304541313525,
+      "learning_rate": 2e-05,
+      "loss": 0.2585,
+      "step": 4384
+    },
+    {
+      "epoch": 0.10437494049319242,
+      "grad_norm": 2.429046220818659,
+      "learning_rate": 2e-05,
+      "loss": 0.1344,
+      "step": 4385
+    },
+    {
+      "epoch": 0.10439874321622393,
+      "grad_norm": 6.350482449243129,
+      "learning_rate": 2e-05,
+      "loss": 0.4545,
+      "step": 4386
+    },
+    {
+      "epoch": 0.10442254593925546,
+      "grad_norm": 9.591829142344015,
+      "learning_rate": 2e-05,
+      "loss": 0.9062,
+      "step": 4387
+    },
+    {
+      "epoch": 0.10444634866228697,
+      "grad_norm": 7.6095138067795975,
+      "learning_rate": 2e-05,
+      "loss": 1.195,
+      "step": 4388
+    },
+    {
+      "epoch": 0.10447015138531848,
+      "grad_norm": 5.3328281457814395,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 4389
+    },
+    {
+      "epoch": 0.10449395410835,
+      "grad_norm": 4.834662354856699,
+      "learning_rate": 2e-05,
+      "loss": 0.2887,
+      "step": 4390
+    },
+    {
+      "epoch": 0.10451775683138151,
+      "grad_norm": 10.83029003705759,
+      "learning_rate": 2e-05,
+      "loss": 0.5491,
+      "step": 4391
+    },
+    {
+      "epoch": 0.10454155955441302,
+      "grad_norm": 4.350044799602026,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 4392
+    },
+    {
+      "epoch": 0.10456536227744453,
+      "grad_norm": 7.710955463172829,
+      "learning_rate": 2e-05,
+      "loss": 0.3279,
+      "step": 4393
+    },
+    {
+      "epoch": 0.10458916500047605,
+      "grad_norm": 4.339662937246398,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 4394
+    },
+    {
+      "epoch": 0.10461296772350757,
+      "grad_norm": 3.615519656131051,
+      "learning_rate": 2e-05,
+      "loss": 0.3678,
+      "step": 4395
+    },
+    {
+      "epoch": 0.10463677044653909,
+      "grad_norm": 10.968298822997225,
+      "learning_rate": 2e-05,
+      "loss": 1.12,
+      "step": 4396
+    },
+    {
+      "epoch": 0.1046605731695706,
+      "grad_norm": 15.077431404673586,
+      "learning_rate": 2e-05,
+      "loss": 1.1897,
+      "step": 4397
+    },
+    {
+      "epoch": 0.10468437589260211,
+      "grad_norm": 9.678147084351478,
+      "learning_rate": 2e-05,
+      "loss": 1.1495,
+      "step": 4398
+    },
+    {
+      "epoch": 0.10470817861563363,
+      "grad_norm": 9.917807285219741,
+      "learning_rate": 2e-05,
+      "loss": 1.2076,
+      "step": 4399
+    },
+    {
+      "epoch": 0.10473198133866514,
+      "grad_norm": 4.017719489446733,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 4400
+    },
+    {
+      "epoch": 0.10475578406169665,
+      "grad_norm": 6.187537179948064,
+      "learning_rate": 2e-05,
+      "loss": 0.8761,
+      "step": 4401
+    },
+    {
+      "epoch": 0.10477958678472818,
+      "grad_norm": 1.8309681741966553,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 4402
+    },
+    {
+      "epoch": 0.10480338950775969,
+      "grad_norm": 7.109916757258641,
+      "learning_rate": 2e-05,
+      "loss": 0.3852,
+      "step": 4403
+    },
+    {
+      "epoch": 0.1048271922307912,
+      "grad_norm": 9.761047033706422,
+      "learning_rate": 2e-05,
+      "loss": 0.4904,
+      "step": 4404
+    },
+    {
+      "epoch": 0.10485099495382272,
+      "grad_norm": 3.302352202142689,
+      "learning_rate": 2e-05,
+      "loss": 0.1632,
+      "step": 4405
+    },
+    {
+      "epoch": 0.10487479767685423,
+      "grad_norm": 5.416569576234348,
+      "learning_rate": 2e-05,
+      "loss": 0.1753,
+      "step": 4406
+    },
+    {
+      "epoch": 0.10489860039988574,
+      "grad_norm": 5.2015649369248855,
+      "learning_rate": 2e-05,
+      "loss": 0.5273,
+      "step": 4407
+    },
+    {
+      "epoch": 0.10492240312291726,
+      "grad_norm": 6.489609856226365,
+      "learning_rate": 2e-05,
+      "loss": 0.9752,
+      "step": 4408
+    },
+    {
+      "epoch": 0.10494620584594877,
+      "grad_norm": 4.845978726535792,
+      "learning_rate": 2e-05,
+      "loss": 0.5973,
+      "step": 4409
+    },
+    {
+      "epoch": 0.1049700085689803,
+      "grad_norm": 6.095234100123287,
+      "learning_rate": 2e-05,
+      "loss": 0.8,
+      "step": 4410
+    },
+    {
+      "epoch": 0.10499381129201181,
+      "grad_norm": 8.641651047348782,
+      "learning_rate": 2e-05,
+      "loss": 0.897,
+      "step": 4411
+    },
+    {
+      "epoch": 0.10501761401504332,
+      "grad_norm": 4.638021464494957,
+      "learning_rate": 2e-05,
+      "loss": 0.5658,
+      "step": 4412
+    },
+    {
+      "epoch": 0.10504141673807484,
+      "grad_norm": 6.490522804858352,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 4413
+    },
+    {
+      "epoch": 0.10506521946110635,
+      "grad_norm": 3.4095617505390203,
+      "learning_rate": 2e-05,
+      "loss": 0.1904,
+      "step": 4414
+    },
+    {
+      "epoch": 0.10508902218413786,
+      "grad_norm": 6.491768680504446,
+      "learning_rate": 2e-05,
+      "loss": 0.428,
+      "step": 4415
+    },
+    {
+      "epoch": 0.10511282490716937,
+      "grad_norm": 7.82652375363713,
+      "learning_rate": 2e-05,
+      "loss": 1.0853,
+      "step": 4416
+    },
+    {
+      "epoch": 0.1051366276302009,
+      "grad_norm": 6.334793105920638,
+      "learning_rate": 2e-05,
+      "loss": 0.9308,
+      "step": 4417
+    },
+    {
+      "epoch": 0.10516043035323241,
+      "grad_norm": 2.266322439812622,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 4418
+    },
+    {
+      "epoch": 0.10518423307626393,
+      "grad_norm": 7.384038088078611,
+      "learning_rate": 2e-05,
+      "loss": 1.0474,
+      "step": 4419
+    },
+    {
+      "epoch": 0.10520803579929544,
+      "grad_norm": 5.357269524033694,
+      "learning_rate": 2e-05,
+      "loss": 0.8423,
+      "step": 4420
+    },
+    {
+      "epoch": 0.10523183852232695,
+      "grad_norm": 8.23785699479774,
+      "learning_rate": 2e-05,
+      "loss": 1.2954,
+      "step": 4421
+    },
+    {
+      "epoch": 0.10525564124535847,
+      "grad_norm": 4.392412023013029,
+      "learning_rate": 2e-05,
+      "loss": 0.6716,
+      "step": 4422
+    },
+    {
+      "epoch": 0.10527944396838998,
+      "grad_norm": 5.210447074725651,
+      "learning_rate": 2e-05,
+      "loss": 0.7777,
+      "step": 4423
+    },
+    {
+      "epoch": 0.10530324669142149,
+      "grad_norm": 5.865479243437861,
+      "learning_rate": 2e-05,
+      "loss": 0.2975,
+      "step": 4424
+    },
+    {
+      "epoch": 0.10532704941445302,
+      "grad_norm": 4.936059038700549,
+      "learning_rate": 2e-05,
+      "loss": 0.3697,
+      "step": 4425
+    },
+    {
+      "epoch": 0.10535085213748453,
+      "grad_norm": 2.591024597827286,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 4426
+    },
+    {
+      "epoch": 0.10537465486051605,
+      "grad_norm": 3.065967371806254,
+      "learning_rate": 2e-05,
+      "loss": 0.2621,
+      "step": 4427
+    },
+    {
+      "epoch": 0.10539845758354756,
+      "grad_norm": 4.95122219976639,
+      "learning_rate": 2e-05,
+      "loss": 0.3249,
+      "step": 4428
+    },
+    {
+      "epoch": 0.10542226030657907,
+      "grad_norm": 2.2658296376249067,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 4429
+    },
+    {
+      "epoch": 0.10544606302961058,
+      "grad_norm": 9.078737703839074,
+      "learning_rate": 2e-05,
+      "loss": 1.3028,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1054698657526421,
+      "grad_norm": 5.73825848319768,
+      "learning_rate": 2e-05,
+      "loss": 0.9303,
+      "step": 4431
+    },
+    {
+      "epoch": 0.10549366847567361,
+      "grad_norm": 7.077858178232656,
+      "learning_rate": 2e-05,
+      "loss": 0.8954,
+      "step": 4432
+    },
+    {
+      "epoch": 0.10551747119870514,
+      "grad_norm": 6.748535220360184,
+      "learning_rate": 2e-05,
+      "loss": 0.8904,
+      "step": 4433
+    },
+    {
+      "epoch": 0.10554127392173665,
+      "grad_norm": 1.124500336946559,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 4434
+    },
+    {
+      "epoch": 0.10556507664476816,
+      "grad_norm": 6.075681241648738,
+      "learning_rate": 2e-05,
+      "loss": 1.1609,
+      "step": 4435
+    },
+    {
+      "epoch": 0.10558887936779968,
+      "grad_norm": 8.059149256586029,
+      "learning_rate": 2e-05,
+      "loss": 1.0036,
+      "step": 4436
+    },
+    {
+      "epoch": 0.10561268209083119,
+      "grad_norm": 9.332966104174346,
+      "learning_rate": 2e-05,
+      "loss": 1.584,
+      "step": 4437
+    },
+    {
+      "epoch": 0.1056364848138627,
+      "grad_norm": 7.171415763086174,
+      "learning_rate": 2e-05,
+      "loss": 1.5335,
+      "step": 4438
+    },
+    {
+      "epoch": 0.10566028753689422,
+      "grad_norm": 4.59223367615488,
+      "learning_rate": 2e-05,
+      "loss": 0.2071,
+      "step": 4439
+    },
+    {
+      "epoch": 0.10568409025992574,
+      "grad_norm": 4.455319290891469,
+      "learning_rate": 2e-05,
+      "loss": 0.3333,
+      "step": 4440
+    },
+    {
+      "epoch": 0.10570789298295726,
+      "grad_norm": 2.4919705528392813,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 4441
+    },
+    {
+      "epoch": 0.10573169570598877,
+      "grad_norm": 10.392538579962237,
+      "learning_rate": 2e-05,
+      "loss": 0.4737,
+      "step": 4442
+    },
+    {
+      "epoch": 0.10575549842902028,
+      "grad_norm": 6.316218531577318,
+      "learning_rate": 2e-05,
+      "loss": 0.3678,
+      "step": 4443
+    },
+    {
+      "epoch": 0.1057793011520518,
+      "grad_norm": 8.811403924525361,
+      "learning_rate": 2e-05,
+      "loss": 1.5213,
+      "step": 4444
+    },
+    {
+      "epoch": 0.10580310387508331,
+      "grad_norm": 1.8090663653757717,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 4445
+    },
+    {
+      "epoch": 0.10582690659811482,
+      "grad_norm": 10.711978552795976,
+      "learning_rate": 2e-05,
+      "loss": 0.9804,
+      "step": 4446
+    },
+    {
+      "epoch": 0.10585070932114633,
+      "grad_norm": 3.1501808124127657,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 4447
+    },
+    {
+      "epoch": 0.10587451204417786,
+      "grad_norm": 2.104403880359417,
+      "learning_rate": 2e-05,
+      "loss": 0.2194,
+      "step": 4448
+    },
+    {
+      "epoch": 0.10589831476720937,
+      "grad_norm": 4.3579483437861155,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 4449
+    },
+    {
+      "epoch": 0.10592211749024089,
+      "grad_norm": 6.954543231446513,
+      "learning_rate": 2e-05,
+      "loss": 0.341,
+      "step": 4450
+    },
+    {
+      "epoch": 0.1059459202132724,
+      "grad_norm": 6.152068330027578,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 4451
+    },
+    {
+      "epoch": 0.10596972293630391,
+      "grad_norm": 5.024147719488779,
+      "learning_rate": 2e-05,
+      "loss": 0.6802,
+      "step": 4452
+    },
+    {
+      "epoch": 0.10599352565933542,
+      "grad_norm": 7.041963785913129,
+      "learning_rate": 2e-05,
+      "loss": 0.2823,
+      "step": 4453
+    },
+    {
+      "epoch": 0.10601732838236694,
+      "grad_norm": 3.7407823826524047,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 4454
+    },
+    {
+      "epoch": 0.10604113110539845,
+      "grad_norm": 7.53282112252292,
+      "learning_rate": 2e-05,
+      "loss": 0.3768,
+      "step": 4455
+    },
+    {
+      "epoch": 0.10606493382842998,
+      "grad_norm": 2.065520673069499,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 4456
+    },
+    {
+      "epoch": 0.10608873655146149,
+      "grad_norm": 3.4427854319173226,
+      "learning_rate": 2e-05,
+      "loss": 0.2085,
+      "step": 4457
+    },
+    {
+      "epoch": 0.106112539274493,
+      "grad_norm": 5.182599924880799,
+      "learning_rate": 2e-05,
+      "loss": 0.7074,
+      "step": 4458
+    },
+    {
+      "epoch": 0.10613634199752452,
+      "grad_norm": 4.668022160318704,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 4459
+    },
+    {
+      "epoch": 0.10616014472055603,
+      "grad_norm": 6.41233022274388,
+      "learning_rate": 2e-05,
+      "loss": 0.2955,
+      "step": 4460
+    },
+    {
+      "epoch": 0.10618394744358754,
+      "grad_norm": 3.4937779886036173,
+      "learning_rate": 2e-05,
+      "loss": 0.2236,
+      "step": 4461
+    },
+    {
+      "epoch": 0.10620775016661906,
+      "grad_norm": 4.843006212352732,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 4462
+    },
+    {
+      "epoch": 0.10623155288965058,
+      "grad_norm": 7.30828026276391,
+      "learning_rate": 2e-05,
+      "loss": 0.3052,
+      "step": 4463
+    },
+    {
+      "epoch": 0.1062553556126821,
+      "grad_norm": 4.078820940838465,
+      "learning_rate": 2e-05,
+      "loss": 0.2876,
+      "step": 4464
+    },
+    {
+      "epoch": 0.10627915833571361,
+      "grad_norm": 6.688893973865118,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 4465
+    },
+    {
+      "epoch": 0.10630296105874512,
+      "grad_norm": 6.857492476942638,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 4466
+    },
+    {
+      "epoch": 0.10632676378177663,
+      "grad_norm": 4.264902779357255,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 4467
+    },
+    {
+      "epoch": 0.10635056650480815,
+      "grad_norm": 6.3851016499034445,
+      "learning_rate": 2e-05,
+      "loss": 0.5205,
+      "step": 4468
+    },
+    {
+      "epoch": 0.10637436922783966,
+      "grad_norm": 8.74872839583226,
+      "learning_rate": 2e-05,
+      "loss": 0.9314,
+      "step": 4469
+    },
+    {
+      "epoch": 0.10639817195087117,
+      "grad_norm": 8.032471554174903,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 4470
+    },
+    {
+      "epoch": 0.1064219746739027,
+      "grad_norm": 2.451666741181791,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 4471
+    },
+    {
+      "epoch": 0.10644577739693421,
+      "grad_norm": 3.600326250335843,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 4472
+    },
+    {
+      "epoch": 0.10646958011996573,
+      "grad_norm": 3.050120182283356,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 4473
+    },
+    {
+      "epoch": 0.10649338284299724,
+      "grad_norm": 3.6684844888043755,
+      "learning_rate": 2e-05,
+      "loss": 0.3881,
+      "step": 4474
+    },
+    {
+      "epoch": 0.10651718556602875,
+      "grad_norm": 5.534474347084669,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 4475
+    },
+    {
+      "epoch": 0.10654098828906026,
+      "grad_norm": 10.097568191599104,
+      "learning_rate": 2e-05,
+      "loss": 1.0286,
+      "step": 4476
+    },
+    {
+      "epoch": 0.10656479101209178,
+      "grad_norm": 8.21065477790736,
+      "learning_rate": 2e-05,
+      "loss": 0.2645,
+      "step": 4477
+    },
+    {
+      "epoch": 0.1065885937351233,
+      "grad_norm": 3.976919551027487,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 4478
+    },
+    {
+      "epoch": 0.10661239645815482,
+      "grad_norm": 2.6184202434931376,
+      "learning_rate": 2e-05,
+      "loss": 0.2843,
+      "step": 4479
+    },
+    {
+      "epoch": 0.10663619918118633,
+      "grad_norm": 6.347497930263891,
+      "learning_rate": 2e-05,
+      "loss": 0.4095,
+      "step": 4480
+    },
+    {
+      "epoch": 0.10666000190421784,
+      "grad_norm": 6.556432104732666,
+      "learning_rate": 2e-05,
+      "loss": 1.0706,
+      "step": 4481
+    },
+    {
+      "epoch": 0.10668380462724936,
+      "grad_norm": 2.116350511937,
+      "learning_rate": 2e-05,
+      "loss": 0.2,
+      "step": 4482
+    },
+    {
+      "epoch": 0.10670760735028087,
+      "grad_norm": 6.682947716621133,
+      "learning_rate": 2e-05,
+      "loss": 0.9712,
+      "step": 4483
+    },
+    {
+      "epoch": 0.10673141007331238,
+      "grad_norm": 4.628422510047727,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 4484
+    },
+    {
+      "epoch": 0.1067552127963439,
+      "grad_norm": 9.038490262051186,
+      "learning_rate": 2e-05,
+      "loss": 1.1907,
+      "step": 4485
+    },
+    {
+      "epoch": 0.10677901551937542,
+      "grad_norm": 10.423296044439823,
+      "learning_rate": 2e-05,
+      "loss": 0.58,
+      "step": 4486
+    },
+    {
+      "epoch": 0.10680281824240694,
+      "grad_norm": 10.237830948012832,
+      "learning_rate": 2e-05,
+      "loss": 0.9128,
+      "step": 4487
+    },
+    {
+      "epoch": 0.10682662096543845,
+      "grad_norm": 7.080703008644012,
+      "learning_rate": 2e-05,
+      "loss": 1.7038,
+      "step": 4488
+    },
+    {
+      "epoch": 0.10685042368846996,
+      "grad_norm": 8.3675297823893,
+      "learning_rate": 2e-05,
+      "loss": 0.4501,
+      "step": 4489
+    },
+    {
+      "epoch": 0.10687422641150147,
+      "grad_norm": 2.150241209045118,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 4490
+    },
+    {
+      "epoch": 0.10689802913453299,
+      "grad_norm": 9.833384373502252,
+      "learning_rate": 2e-05,
+      "loss": 0.8162,
+      "step": 4491
+    },
+    {
+      "epoch": 0.1069218318575645,
+      "grad_norm": 3.9851228521553255,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 4492
+    },
+    {
+      "epoch": 0.10694563458059601,
+      "grad_norm": 4.285442110430365,
+      "learning_rate": 2e-05,
+      "loss": 0.8012,
+      "step": 4493
+    },
+    {
+      "epoch": 0.10696943730362754,
+      "grad_norm": 7.938298337208244,
+      "learning_rate": 2e-05,
+      "loss": 0.3821,
+      "step": 4494
+    },
+    {
+      "epoch": 0.10699324002665905,
+      "grad_norm": 2.196941566152502,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 4495
+    },
+    {
+      "epoch": 0.10701704274969057,
+      "grad_norm": 6.091547205704759,
+      "learning_rate": 2e-05,
+      "loss": 0.2987,
+      "step": 4496
+    },
+    {
+      "epoch": 0.10704084547272208,
+      "grad_norm": 4.771217215497714,
+      "learning_rate": 2e-05,
+      "loss": 0.2313,
+      "step": 4497
+    },
+    {
+      "epoch": 0.10706464819575359,
+      "grad_norm": 3.6544793612952,
+      "learning_rate": 2e-05,
+      "loss": 0.1483,
+      "step": 4498
+    },
+    {
+      "epoch": 0.1070884509187851,
+      "grad_norm": 2.964909167142587,
+      "learning_rate": 2e-05,
+      "loss": 0.2319,
+      "step": 4499
+    },
+    {
+      "epoch": 0.10711225364181662,
+      "grad_norm": 1.5238426778938694,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 4500
+    },
+    {
+      "epoch": 0.10713605636484815,
+      "grad_norm": 2.4916718960105193,
+      "learning_rate": 2e-05,
+      "loss": 0.2162,
+      "step": 4501
+    },
+    {
+      "epoch": 0.10715985908787966,
+      "grad_norm": 2.2292950496968027,
+      "learning_rate": 2e-05,
+      "loss": 0.1867,
+      "step": 4502
+    },
+    {
+      "epoch": 0.10718366181091117,
+      "grad_norm": 2.554247179684472,
+      "learning_rate": 2e-05,
+      "loss": 0.218,
+      "step": 4503
+    },
+    {
+      "epoch": 0.10720746453394268,
+      "grad_norm": 3.3757935904267127,
+      "learning_rate": 2e-05,
+      "loss": 0.4854,
+      "step": 4504
+    },
+    {
+      "epoch": 0.1072312672569742,
+      "grad_norm": 10.148087698843476,
+      "learning_rate": 2e-05,
+      "loss": 0.8517,
+      "step": 4505
+    },
+    {
+      "epoch": 0.10725506998000571,
+      "grad_norm": 3.4436057408178007,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 4506
+    },
+    {
+      "epoch": 0.10727887270303722,
+      "grad_norm": 6.477297373352823,
+      "learning_rate": 2e-05,
+      "loss": 0.8029,
+      "step": 4507
+    },
+    {
+      "epoch": 0.10730267542606874,
+      "grad_norm": 8.264231598805248,
+      "learning_rate": 2e-05,
+      "loss": 0.8641,
+      "step": 4508
+    },
+    {
+      "epoch": 0.10732647814910026,
+      "grad_norm": 5.75309170409673,
+      "learning_rate": 2e-05,
+      "loss": 1.2965,
+      "step": 4509
+    },
+    {
+      "epoch": 0.10735028087213178,
+      "grad_norm": 6.044861860615539,
+      "learning_rate": 2e-05,
+      "loss": 0.8065,
+      "step": 4510
+    },
+    {
+      "epoch": 0.10737408359516329,
+      "grad_norm": 6.114538218840502,
+      "learning_rate": 2e-05,
+      "loss": 1.1519,
+      "step": 4511
+    },
+    {
+      "epoch": 0.1073978863181948,
+      "grad_norm": 2.344229381324084,
+      "learning_rate": 2e-05,
+      "loss": 0.2203,
+      "step": 4512
+    },
+    {
+      "epoch": 0.10742168904122631,
+      "grad_norm": 7.265415733541721,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 4513
+    },
+    {
+      "epoch": 0.10744549176425783,
+      "grad_norm": 5.8927225510404675,
+      "learning_rate": 2e-05,
+      "loss": 0.2597,
+      "step": 4514
+    },
+    {
+      "epoch": 0.10746929448728934,
+      "grad_norm": 3.057700833517385,
+      "learning_rate": 2e-05,
+      "loss": 0.3011,
+      "step": 4515
+    },
+    {
+      "epoch": 0.10749309721032087,
+      "grad_norm": 9.255942778948997,
+      "learning_rate": 2e-05,
+      "loss": 1.1007,
+      "step": 4516
+    },
+    {
+      "epoch": 0.10751689993335238,
+      "grad_norm": 6.973598743353489,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 4517
+    },
+    {
+      "epoch": 0.1075407026563839,
+      "grad_norm": 4.742552927183653,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 4518
+    },
+    {
+      "epoch": 0.1075645053794154,
+      "grad_norm": 4.426094315741327,
+      "learning_rate": 2e-05,
+      "loss": 0.9218,
+      "step": 4519
+    },
+    {
+      "epoch": 0.10758830810244692,
+      "grad_norm": 6.12997282529821,
+      "learning_rate": 2e-05,
+      "loss": 0.3602,
+      "step": 4520
+    },
+    {
+      "epoch": 0.10761211082547843,
+      "grad_norm": 6.012123962152854,
+      "learning_rate": 2e-05,
+      "loss": 1.8768,
+      "step": 4521
+    },
+    {
+      "epoch": 0.10763591354850995,
+      "grad_norm": 8.901365450834689,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 4522
+    },
+    {
+      "epoch": 0.10765971627154146,
+      "grad_norm": 2.4455963440020807,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 4523
+    },
+    {
+      "epoch": 0.10768351899457299,
+      "grad_norm": 2.2598377015552003,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 4524
+    },
+    {
+      "epoch": 0.1077073217176045,
+      "grad_norm": 4.794284781149661,
+      "learning_rate": 2e-05,
+      "loss": 0.6861,
+      "step": 4525
+    },
+    {
+      "epoch": 0.10773112444063601,
+      "grad_norm": 8.411441698296354,
+      "learning_rate": 2e-05,
+      "loss": 0.9925,
+      "step": 4526
+    },
+    {
+      "epoch": 0.10775492716366752,
+      "grad_norm": 4.8374771410856585,
+      "learning_rate": 2e-05,
+      "loss": 0.5242,
+      "step": 4527
+    },
+    {
+      "epoch": 0.10777872988669904,
+      "grad_norm": 5.668146260543707,
+      "learning_rate": 2e-05,
+      "loss": 1.1917,
+      "step": 4528
+    },
+    {
+      "epoch": 0.10780253260973055,
+      "grad_norm": 3.701209642532235,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 4529
+    },
+    {
+      "epoch": 0.10782633533276206,
+      "grad_norm": 2.4888714456146457,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 4530
+    },
+    {
+      "epoch": 0.10785013805579358,
+      "grad_norm": 7.762394262035465,
+      "learning_rate": 2e-05,
+      "loss": 0.7457,
+      "step": 4531
+    },
+    {
+      "epoch": 0.1078739407788251,
+      "grad_norm": 1.6161754226418859,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 4532
+    },
+    {
+      "epoch": 0.10789774350185662,
+      "grad_norm": 6.68080339733095,
+      "learning_rate": 2e-05,
+      "loss": 0.5191,
+      "step": 4533
+    },
+    {
+      "epoch": 0.10792154622488813,
+      "grad_norm": 4.988984059701429,
+      "learning_rate": 2e-05,
+      "loss": 0.3431,
+      "step": 4534
+    },
+    {
+      "epoch": 0.10794534894791964,
+      "grad_norm": 5.536083213593785,
+      "learning_rate": 2e-05,
+      "loss": 0.8402,
+      "step": 4535
+    },
+    {
+      "epoch": 0.10796915167095116,
+      "grad_norm": 9.19098228202636,
+      "learning_rate": 2e-05,
+      "loss": 1.1007,
+      "step": 4536
+    },
+    {
+      "epoch": 0.10799295439398267,
+      "grad_norm": 4.11708185551866,
+      "learning_rate": 2e-05,
+      "loss": 0.362,
+      "step": 4537
+    },
+    {
+      "epoch": 0.10801675711701418,
+      "grad_norm": 6.156625651334844,
+      "learning_rate": 2e-05,
+      "loss": 1.0548,
+      "step": 4538
+    },
+    {
+      "epoch": 0.10804055984004571,
+      "grad_norm": 6.628925997398165,
+      "learning_rate": 2e-05,
+      "loss": 0.4096,
+      "step": 4539
+    },
+    {
+      "epoch": 0.10806436256307722,
+      "grad_norm": 5.109010586390749,
+      "learning_rate": 2e-05,
+      "loss": 0.6314,
+      "step": 4540
+    },
+    {
+      "epoch": 0.10808816528610873,
+      "grad_norm": 5.535578338288574,
+      "learning_rate": 2e-05,
+      "loss": 0.61,
+      "step": 4541
+    },
+    {
+      "epoch": 0.10811196800914025,
+      "grad_norm": 10.719122188538757,
+      "learning_rate": 2e-05,
+      "loss": 0.2675,
+      "step": 4542
+    },
+    {
+      "epoch": 0.10813577073217176,
+      "grad_norm": 3.327204240095933,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 4543
+    },
+    {
+      "epoch": 0.10815957345520327,
+      "grad_norm": 6.184400582630355,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 4544
+    },
+    {
+      "epoch": 0.10818337617823479,
+      "grad_norm": 2.4754331896483004,
+      "learning_rate": 2e-05,
+      "loss": 0.1657,
+      "step": 4545
+    },
+    {
+      "epoch": 0.1082071789012663,
+      "grad_norm": 4.488857224434044,
+      "learning_rate": 2e-05,
+      "loss": 0.5158,
+      "step": 4546
+    },
+    {
+      "epoch": 0.10823098162429783,
+      "grad_norm": 5.073472345772578,
+      "learning_rate": 2e-05,
+      "loss": 0.7594,
+      "step": 4547
+    },
+    {
+      "epoch": 0.10825478434732934,
+      "grad_norm": 4.13048883308419,
+      "learning_rate": 2e-05,
+      "loss": 0.2342,
+      "step": 4548
+    },
+    {
+      "epoch": 0.10827858707036085,
+      "grad_norm": 5.924218771609792,
+      "learning_rate": 2e-05,
+      "loss": 0.2719,
+      "step": 4549
+    },
+    {
+      "epoch": 0.10830238979339236,
+      "grad_norm": 7.640371001461447,
+      "learning_rate": 2e-05,
+      "loss": 0.795,
+      "step": 4550
+    },
+    {
+      "epoch": 0.10832619251642388,
+      "grad_norm": 6.980740555928133,
+      "learning_rate": 2e-05,
+      "loss": 1.0136,
+      "step": 4551
+    },
+    {
+      "epoch": 0.10834999523945539,
+      "grad_norm": 5.909080830436893,
+      "learning_rate": 2e-05,
+      "loss": 1.1768,
+      "step": 4552
+    },
+    {
+      "epoch": 0.1083737979624869,
+      "grad_norm": 8.603518484526992,
+      "learning_rate": 2e-05,
+      "loss": 1.272,
+      "step": 4553
+    },
+    {
+      "epoch": 0.10839760068551842,
+      "grad_norm": 2.611400542113255,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 4554
+    },
+    {
+      "epoch": 0.10842140340854994,
+      "grad_norm": 1.4259625146134172,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 4555
+    },
+    {
+      "epoch": 0.10844520613158146,
+      "grad_norm": 4.279818598486146,
+      "learning_rate": 2e-05,
+      "loss": 0.6587,
+      "step": 4556
+    },
+    {
+      "epoch": 0.10846900885461297,
+      "grad_norm": 7.991015280199547,
+      "learning_rate": 2e-05,
+      "loss": 1.2737,
+      "step": 4557
+    },
+    {
+      "epoch": 0.10849281157764448,
+      "grad_norm": 4.540192220313591,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 4558
+    },
+    {
+      "epoch": 0.108516614300676,
+      "grad_norm": 9.565831040459425,
+      "learning_rate": 2e-05,
+      "loss": 0.3849,
+      "step": 4559
+    },
+    {
+      "epoch": 0.10854041702370751,
+      "grad_norm": 6.772290862253984,
+      "learning_rate": 2e-05,
+      "loss": 1.6939,
+      "step": 4560
+    },
+    {
+      "epoch": 0.10856421974673902,
+      "grad_norm": 4.185762831828644,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 4561
+    },
+    {
+      "epoch": 0.10858802246977055,
+      "grad_norm": 12.545137502805538,
+      "learning_rate": 2e-05,
+      "loss": 2.0007,
+      "step": 4562
+    },
+    {
+      "epoch": 0.10861182519280206,
+      "grad_norm": 6.936755593323794,
+      "learning_rate": 2e-05,
+      "loss": 1.0421,
+      "step": 4563
+    },
+    {
+      "epoch": 0.10863562791583357,
+      "grad_norm": 6.670039446170544,
+      "learning_rate": 2e-05,
+      "loss": 1.1598,
+      "step": 4564
+    },
+    {
+      "epoch": 0.10865943063886509,
+      "grad_norm": 2.6274656304853363,
+      "learning_rate": 2e-05,
+      "loss": 0.1943,
+      "step": 4565
+    },
+    {
+      "epoch": 0.1086832333618966,
+      "grad_norm": 3.488969920993666,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 4566
+    },
+    {
+      "epoch": 0.10870703608492811,
+      "grad_norm": 4.6454630618007435,
+      "learning_rate": 2e-05,
+      "loss": 0.6456,
+      "step": 4567
+    },
+    {
+      "epoch": 0.10873083880795963,
+      "grad_norm": 6.614898249111022,
+      "learning_rate": 2e-05,
+      "loss": 0.4184,
+      "step": 4568
+    },
+    {
+      "epoch": 0.10875464153099114,
+      "grad_norm": 6.751682809615974,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 4569
+    },
+    {
+      "epoch": 0.10877844425402267,
+      "grad_norm": 2.6119563464805142,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 4570
+    },
+    {
+      "epoch": 0.10880224697705418,
+      "grad_norm": 3.8512008258512176,
+      "learning_rate": 2e-05,
+      "loss": 0.722,
+      "step": 4571
+    },
+    {
+      "epoch": 0.10882604970008569,
+      "grad_norm": 2.5143883622119922,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 4572
+    },
+    {
+      "epoch": 0.1088498524231172,
+      "grad_norm": 3.834286703701386,
+      "learning_rate": 2e-05,
+      "loss": 0.583,
+      "step": 4573
+    },
+    {
+      "epoch": 0.10887365514614872,
+      "grad_norm": 4.832518726923298,
+      "learning_rate": 2e-05,
+      "loss": 0.2475,
+      "step": 4574
+    },
+    {
+      "epoch": 0.10889745786918023,
+      "grad_norm": 7.380869244983162,
+      "learning_rate": 2e-05,
+      "loss": 1.4768,
+      "step": 4575
+    },
+    {
+      "epoch": 0.10892126059221174,
+      "grad_norm": 3.186181748214506,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 4576
+    },
+    {
+      "epoch": 0.10894506331524327,
+      "grad_norm": 7.118791675175698,
+      "learning_rate": 2e-05,
+      "loss": 1.0144,
+      "step": 4577
+    },
+    {
+      "epoch": 0.10896886603827478,
+      "grad_norm": 5.968786604770417,
+      "learning_rate": 2e-05,
+      "loss": 0.7178,
+      "step": 4578
+    },
+    {
+      "epoch": 0.1089926687613063,
+      "grad_norm": 3.832127274932373,
+      "learning_rate": 2e-05,
+      "loss": 0.4157,
+      "step": 4579
+    },
+    {
+      "epoch": 0.10901647148433781,
+      "grad_norm": 1.5414772751534167,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 4580
+    },
+    {
+      "epoch": 0.10904027420736932,
+      "grad_norm": 34.43332366718043,
+      "learning_rate": 2e-05,
+      "loss": 0.4595,
+      "step": 4581
+    },
+    {
+      "epoch": 0.10906407693040084,
+      "grad_norm": 6.8398666562840935,
+      "learning_rate": 2e-05,
+      "loss": 0.9072,
+      "step": 4582
+    },
+    {
+      "epoch": 0.10908787965343235,
+      "grad_norm": 2.4546216707095136,
+      "learning_rate": 2e-05,
+      "loss": 0.1397,
+      "step": 4583
+    },
+    {
+      "epoch": 0.10911168237646386,
+      "grad_norm": 2.123146206094123,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 4584
+    },
+    {
+      "epoch": 0.10913548509949539,
+      "grad_norm": 7.59126495257331,
+      "learning_rate": 2e-05,
+      "loss": 0.3949,
+      "step": 4585
+    },
+    {
+      "epoch": 0.1091592878225269,
+      "grad_norm": 2.1168568650901847,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 4586
+    },
+    {
+      "epoch": 0.10918309054555841,
+      "grad_norm": 5.701533700878501,
+      "learning_rate": 2e-05,
+      "loss": 0.2813,
+      "step": 4587
+    },
+    {
+      "epoch": 0.10920689326858993,
+      "grad_norm": 6.651887942279691,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 4588
+    },
+    {
+      "epoch": 0.10923069599162144,
+      "grad_norm": 6.090902514482352,
+      "learning_rate": 2e-05,
+      "loss": 0.3716,
+      "step": 4589
+    },
+    {
+      "epoch": 0.10925449871465295,
+      "grad_norm": 5.012875881521851,
+      "learning_rate": 2e-05,
+      "loss": 0.8148,
+      "step": 4590
+    },
+    {
+      "epoch": 0.10927830143768447,
+      "grad_norm": 5.530806263391207,
+      "learning_rate": 2e-05,
+      "loss": 0.2974,
+      "step": 4591
+    },
+    {
+      "epoch": 0.10930210416071598,
+      "grad_norm": 2.2608189634118863,
+      "learning_rate": 2e-05,
+      "loss": 0.2027,
+      "step": 4592
+    },
+    {
+      "epoch": 0.1093259068837475,
+      "grad_norm": 4.435111078576218,
+      "learning_rate": 2e-05,
+      "loss": 0.6079,
+      "step": 4593
+    },
+    {
+      "epoch": 0.10934970960677902,
+      "grad_norm": 5.55785997833065,
+      "learning_rate": 2e-05,
+      "loss": 0.8919,
+      "step": 4594
+    },
+    {
+      "epoch": 0.10937351232981053,
+      "grad_norm": 5.163957918422169,
+      "learning_rate": 2e-05,
+      "loss": 0.7725,
+      "step": 4595
+    },
+    {
+      "epoch": 0.10939731505284205,
+      "grad_norm": 5.52555120130406,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 4596
+    },
+    {
+      "epoch": 0.10942111777587356,
+      "grad_norm": 4.681799275050135,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 4597
+    },
+    {
+      "epoch": 0.10944492049890507,
+      "grad_norm": 7.862134235225315,
+      "learning_rate": 2e-05,
+      "loss": 0.4476,
+      "step": 4598
+    },
+    {
+      "epoch": 0.10946872322193658,
+      "grad_norm": 6.43641754485657,
+      "learning_rate": 2e-05,
+      "loss": 1.1902,
+      "step": 4599
+    },
+    {
+      "epoch": 0.10949252594496811,
+      "grad_norm": 2.7156702946056903,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 4600
+    },
+    {
+      "epoch": 0.10951632866799962,
+      "grad_norm": 5.050164814376367,
+      "learning_rate": 2e-05,
+      "loss": 0.7756,
+      "step": 4601
+    },
+    {
+      "epoch": 0.10954013139103114,
+      "grad_norm": 7.869712638379544,
+      "learning_rate": 2e-05,
+      "loss": 1.0149,
+      "step": 4602
+    },
+    {
+      "epoch": 0.10956393411406265,
+      "grad_norm": 11.717684374272961,
+      "learning_rate": 2e-05,
+      "loss": 0.9932,
+      "step": 4603
+    },
+    {
+      "epoch": 0.10958773683709416,
+      "grad_norm": 4.8616924016124825,
+      "learning_rate": 2e-05,
+      "loss": 0.7089,
+      "step": 4604
+    },
+    {
+      "epoch": 0.10961153956012568,
+      "grad_norm": 9.849004734038582,
+      "learning_rate": 2e-05,
+      "loss": 1.4217,
+      "step": 4605
+    },
+    {
+      "epoch": 0.10963534228315719,
+      "grad_norm": 5.822391917181978,
+      "learning_rate": 2e-05,
+      "loss": 0.7268,
+      "step": 4606
+    },
+    {
+      "epoch": 0.1096591450061887,
+      "grad_norm": 2.855333416607515,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 4607
+    },
+    {
+      "epoch": 0.10968294772922023,
+      "grad_norm": 2.7062127362821102,
+      "learning_rate": 2e-05,
+      "loss": 0.1248,
+      "step": 4608
+    },
+    {
+      "epoch": 0.10970675045225174,
+      "grad_norm": 10.360743114158959,
+      "learning_rate": 2e-05,
+      "loss": 0.3329,
+      "step": 4609
+    },
+    {
+      "epoch": 0.10973055317528325,
+      "grad_norm": 10.448883212010257,
+      "learning_rate": 2e-05,
+      "loss": 0.5901,
+      "step": 4610
+    },
+    {
+      "epoch": 0.10975435589831477,
+      "grad_norm": 5.648364316076067,
+      "learning_rate": 2e-05,
+      "loss": 0.3637,
+      "step": 4611
+    },
+    {
+      "epoch": 0.10977815862134628,
+      "grad_norm": 15.568063725418261,
+      "learning_rate": 2e-05,
+      "loss": 0.6401,
+      "step": 4612
+    },
+    {
+      "epoch": 0.1098019613443778,
+      "grad_norm": 5.777686047917426,
+      "learning_rate": 2e-05,
+      "loss": 0.7298,
+      "step": 4613
+    },
+    {
+      "epoch": 0.1098257640674093,
+      "grad_norm": 4.831985939954256,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 4614
+    },
+    {
+      "epoch": 0.10984956679044082,
+      "grad_norm": 11.387041151143151,
+      "learning_rate": 2e-05,
+      "loss": 0.3783,
+      "step": 4615
+    },
+    {
+      "epoch": 0.10987336951347235,
+      "grad_norm": 2.057473414046138,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 4616
+    },
+    {
+      "epoch": 0.10989717223650386,
+      "grad_norm": 7.6171787093230146,
+      "learning_rate": 2e-05,
+      "loss": 0.3439,
+      "step": 4617
+    },
+    {
+      "epoch": 0.10992097495953537,
+      "grad_norm": 2.3239852707623716,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 4618
+    },
+    {
+      "epoch": 0.10994477768256689,
+      "grad_norm": 10.713973635674323,
+      "learning_rate": 2e-05,
+      "loss": 0.3579,
+      "step": 4619
+    },
+    {
+      "epoch": 0.1099685804055984,
+      "grad_norm": 4.08902866853398,
+      "learning_rate": 2e-05,
+      "loss": 0.4663,
+      "step": 4620
+    },
+    {
+      "epoch": 0.10999238312862991,
+      "grad_norm": 1.9679420135189583,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 4621
+    },
+    {
+      "epoch": 0.11001618585166142,
+      "grad_norm": 4.972960156449517,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 4622
+    },
+    {
+      "epoch": 0.11003998857469295,
+      "grad_norm": 5.653715913135903,
+      "learning_rate": 2e-05,
+      "loss": 0.3803,
+      "step": 4623
+    },
+    {
+      "epoch": 0.11006379129772446,
+      "grad_norm": 5.790379277611699,
+      "learning_rate": 2e-05,
+      "loss": 1.0611,
+      "step": 4624
+    },
+    {
+      "epoch": 0.11008759402075598,
+      "grad_norm": 5.931347881841952,
+      "learning_rate": 2e-05,
+      "loss": 0.2324,
+      "step": 4625
+    },
+    {
+      "epoch": 0.11011139674378749,
+      "grad_norm": 3.2293127278732494,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 4626
+    },
+    {
+      "epoch": 0.110135199466819,
+      "grad_norm": 9.135240926898842,
+      "learning_rate": 2e-05,
+      "loss": 0.3959,
+      "step": 4627
+    },
+    {
+      "epoch": 0.11015900218985052,
+      "grad_norm": 2.5833029063375155,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 4628
+    },
+    {
+      "epoch": 0.11018280491288203,
+      "grad_norm": 7.1586793463315725,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 4629
+    },
+    {
+      "epoch": 0.11020660763591354,
+      "grad_norm": 5.274429914513859,
+      "learning_rate": 2e-05,
+      "loss": 0.8937,
+      "step": 4630
+    },
+    {
+      "epoch": 0.11023041035894507,
+      "grad_norm": 2.8248716004457983,
+      "learning_rate": 2e-05,
+      "loss": 0.2129,
+      "step": 4631
+    },
+    {
+      "epoch": 0.11025421308197658,
+      "grad_norm": 2.451266503430977,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 4632
+    },
+    {
+      "epoch": 0.1102780158050081,
+      "grad_norm": 6.50855805661206,
+      "learning_rate": 2e-05,
+      "loss": 0.3091,
+      "step": 4633
+    },
+    {
+      "epoch": 0.11030181852803961,
+      "grad_norm": 4.122033082043829,
+      "learning_rate": 2e-05,
+      "loss": 0.2591,
+      "step": 4634
+    },
+    {
+      "epoch": 0.11032562125107112,
+      "grad_norm": 5.663715044505105,
+      "learning_rate": 2e-05,
+      "loss": 1.098,
+      "step": 4635
+    },
+    {
+      "epoch": 0.11034942397410263,
+      "grad_norm": 9.566311777705387,
+      "learning_rate": 2e-05,
+      "loss": 1.8073,
+      "step": 4636
+    },
+    {
+      "epoch": 0.11037322669713415,
+      "grad_norm": 5.2572893932019085,
+      "learning_rate": 2e-05,
+      "loss": 0.7137,
+      "step": 4637
+    },
+    {
+      "epoch": 0.11039702942016567,
+      "grad_norm": 11.213094022842865,
+      "learning_rate": 2e-05,
+      "loss": 1.8356,
+      "step": 4638
+    },
+    {
+      "epoch": 0.11042083214319719,
+      "grad_norm": 5.098593017571939,
+      "learning_rate": 2e-05,
+      "loss": 0.9719,
+      "step": 4639
+    },
+    {
+      "epoch": 0.1104446348662287,
+      "grad_norm": 5.9731920085549035,
+      "learning_rate": 2e-05,
+      "loss": 0.982,
+      "step": 4640
+    },
+    {
+      "epoch": 0.11046843758926021,
+      "grad_norm": 4.111026274545169,
+      "learning_rate": 2e-05,
+      "loss": 0.4908,
+      "step": 4641
+    },
+    {
+      "epoch": 0.11049224031229173,
+      "grad_norm": 4.4845582204741055,
+      "learning_rate": 2e-05,
+      "loss": 0.2531,
+      "step": 4642
+    },
+    {
+      "epoch": 0.11051604303532324,
+      "grad_norm": 1.8020380376016678,
+      "learning_rate": 2e-05,
+      "loss": 0.1775,
+      "step": 4643
+    },
+    {
+      "epoch": 0.11053984575835475,
+      "grad_norm": 7.164897216267568,
+      "learning_rate": 2e-05,
+      "loss": 1.5454,
+      "step": 4644
+    },
+    {
+      "epoch": 0.11056364848138626,
+      "grad_norm": 7.216732868663433,
+      "learning_rate": 2e-05,
+      "loss": 0.7974,
+      "step": 4645
+    },
+    {
+      "epoch": 0.11058745120441779,
+      "grad_norm": 1.6682610604061097,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 4646
+    },
+    {
+      "epoch": 0.1106112539274493,
+      "grad_norm": 3.138697751452981,
+      "learning_rate": 2e-05,
+      "loss": 0.2385,
+      "step": 4647
+    },
+    {
+      "epoch": 0.11063505665048082,
+      "grad_norm": 5.056191111796193,
+      "learning_rate": 2e-05,
+      "loss": 0.471,
+      "step": 4648
+    },
+    {
+      "epoch": 0.11065885937351233,
+      "grad_norm": 4.330644924994065,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 4649
+    },
+    {
+      "epoch": 0.11068266209654384,
+      "grad_norm": 4.8647121081723155,
+      "learning_rate": 2e-05,
+      "loss": 0.8708,
+      "step": 4650
+    },
+    {
+      "epoch": 0.11070646481957536,
+      "grad_norm": 3.3854158893700004,
+      "learning_rate": 2e-05,
+      "loss": 0.2317,
+      "step": 4651
+    },
+    {
+      "epoch": 0.11073026754260687,
+      "grad_norm": 7.831318345736432,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 4652
+    },
+    {
+      "epoch": 0.11075407026563838,
+      "grad_norm": 9.33500142476609,
+      "learning_rate": 2e-05,
+      "loss": 0.8157,
+      "step": 4653
+    },
+    {
+      "epoch": 0.11077787298866991,
+      "grad_norm": 8.334325117918667,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 4654
+    },
+    {
+      "epoch": 0.11080167571170142,
+      "grad_norm": 10.3368709330486,
+      "learning_rate": 2e-05,
+      "loss": 0.7256,
+      "step": 4655
+    },
+    {
+      "epoch": 0.11082547843473294,
+      "grad_norm": 6.27664870411815,
+      "learning_rate": 2e-05,
+      "loss": 0.4705,
+      "step": 4656
+    },
+    {
+      "epoch": 0.11084928115776445,
+      "grad_norm": 4.603026899467357,
+      "learning_rate": 2e-05,
+      "loss": 0.4101,
+      "step": 4657
+    },
+    {
+      "epoch": 0.11087308388079596,
+      "grad_norm": 4.767999262754345,
+      "learning_rate": 2e-05,
+      "loss": 0.7802,
+      "step": 4658
+    },
+    {
+      "epoch": 0.11089688660382747,
+      "grad_norm": 7.560535956246845,
+      "learning_rate": 2e-05,
+      "loss": 0.7996,
+      "step": 4659
+    },
+    {
+      "epoch": 0.11092068932685899,
+      "grad_norm": 4.62838323262061,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 4660
+    },
+    {
+      "epoch": 0.11094449204989051,
+      "grad_norm": 2.0378329929625343,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 4661
+    },
+    {
+      "epoch": 0.11096829477292203,
+      "grad_norm": 1.7409082895129686,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 4662
+    },
+    {
+      "epoch": 0.11099209749595354,
+      "grad_norm": 5.280019503999173,
+      "learning_rate": 2e-05,
+      "loss": 0.7629,
+      "step": 4663
+    },
+    {
+      "epoch": 0.11101590021898505,
+      "grad_norm": 7.072143107084512,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 4664
+    },
+    {
+      "epoch": 0.11103970294201657,
+      "grad_norm": 7.0901419390379745,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 4665
+    },
+    {
+      "epoch": 0.11106350566504808,
+      "grad_norm": 2.8527859920332723,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 4666
+    },
+    {
+      "epoch": 0.11108730838807959,
+      "grad_norm": 5.205590680918547,
+      "learning_rate": 2e-05,
+      "loss": 0.3265,
+      "step": 4667
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 3.1348397576040368,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 4668
+    },
+    {
+      "epoch": 0.11113491383414263,
+      "grad_norm": 3.1585038226197018,
+      "learning_rate": 2e-05,
+      "loss": 0.2001,
+      "step": 4669
+    },
+    {
+      "epoch": 0.11115871655717414,
+      "grad_norm": 7.453281669875553,
+      "learning_rate": 2e-05,
+      "loss": 1.092,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11118251928020566,
+      "grad_norm": 8.996957833354529,
+      "learning_rate": 2e-05,
+      "loss": 0.3738,
+      "step": 4671
+    },
+    {
+      "epoch": 0.11120632200323717,
+      "grad_norm": 3.1320576619726497,
+      "learning_rate": 2e-05,
+      "loss": 0.2363,
+      "step": 4672
+    },
+    {
+      "epoch": 0.11123012472626868,
+      "grad_norm": 2.245972350130854,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 4673
+    },
+    {
+      "epoch": 0.1112539274493002,
+      "grad_norm": 5.172600245850242,
+      "learning_rate": 2e-05,
+      "loss": 0.5977,
+      "step": 4674
+    },
+    {
+      "epoch": 0.11127773017233171,
+      "grad_norm": 6.3595287075080575,
+      "learning_rate": 2e-05,
+      "loss": 1.5647,
+      "step": 4675
+    },
+    {
+      "epoch": 0.11130153289536322,
+      "grad_norm": 6.520323595568324,
+      "learning_rate": 2e-05,
+      "loss": 1.195,
+      "step": 4676
+    },
+    {
+      "epoch": 0.11132533561839475,
+      "grad_norm": 3.1235171670003576,
+      "learning_rate": 2e-05,
+      "loss": 0.2251,
+      "step": 4677
+    },
+    {
+      "epoch": 0.11134913834142626,
+      "grad_norm": 5.0100790408772085,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 4678
+    },
+    {
+      "epoch": 0.11137294106445778,
+      "grad_norm": 7.771007144597852,
+      "learning_rate": 2e-05,
+      "loss": 0.3686,
+      "step": 4679
+    },
+    {
+      "epoch": 0.11139674378748929,
+      "grad_norm": 6.100554746248129,
+      "learning_rate": 2e-05,
+      "loss": 0.9331,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1114205465105208,
+      "grad_norm": 10.521221119323103,
+      "learning_rate": 2e-05,
+      "loss": 1.1304,
+      "step": 4681
+    },
+    {
+      "epoch": 0.11144434923355231,
+      "grad_norm": 8.280802476856895,
+      "learning_rate": 2e-05,
+      "loss": 1.5633,
+      "step": 4682
+    },
+    {
+      "epoch": 0.11146815195658383,
+      "grad_norm": 6.814963329181687,
+      "learning_rate": 2e-05,
+      "loss": 0.3799,
+      "step": 4683
+    },
+    {
+      "epoch": 0.11149195467961535,
+      "grad_norm": 7.38058611074462,
+      "learning_rate": 2e-05,
+      "loss": 0.2862,
+      "step": 4684
+    },
+    {
+      "epoch": 0.11151575740264687,
+      "grad_norm": 51.93806292129021,
+      "learning_rate": 2e-05,
+      "loss": 0.8922,
+      "step": 4685
+    },
+    {
+      "epoch": 0.11153956012567838,
+      "grad_norm": 176.24136910356864,
+      "learning_rate": 2e-05,
+      "loss": 1.4595,
+      "step": 4686
+    },
+    {
+      "epoch": 0.1115633628487099,
+      "grad_norm": 209.75655255853533,
+      "learning_rate": 2e-05,
+      "loss": 0.8865,
+      "step": 4687
+    },
+    {
+      "epoch": 0.1115871655717414,
+      "grad_norm": 32.194766004311745,
+      "learning_rate": 2e-05,
+      "loss": 0.6754,
+      "step": 4688
+    },
+    {
+      "epoch": 0.11161096829477292,
+      "grad_norm": 4.788895324198648,
+      "learning_rate": 2e-05,
+      "loss": 0.3784,
+      "step": 4689
+    },
+    {
+      "epoch": 0.11163477101780443,
+      "grad_norm": 6.06970600545383,
+      "learning_rate": 2e-05,
+      "loss": 0.3263,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11165857374083595,
+      "grad_norm": 7.832865199722468,
+      "learning_rate": 2e-05,
+      "loss": 0.7807,
+      "step": 4691
+    },
+    {
+      "epoch": 0.11168237646386747,
+      "grad_norm": 30.99637867400311,
+      "learning_rate": 2e-05,
+      "loss": 0.5182,
+      "step": 4692
+    },
+    {
+      "epoch": 0.11170617918689899,
+      "grad_norm": 226.98004341814269,
+      "learning_rate": 2e-05,
+      "loss": 1.5162,
+      "step": 4693
+    },
+    {
+      "epoch": 0.1117299819099305,
+      "grad_norm": 28.112726933872253,
+      "learning_rate": 2e-05,
+      "loss": 0.8588,
+      "step": 4694
+    },
+    {
+      "epoch": 0.11175378463296201,
+      "grad_norm": 30.781339989146552,
+      "learning_rate": 2e-05,
+      "loss": 1.8588,
+      "step": 4695
+    },
+    {
+      "epoch": 0.11177758735599352,
+      "grad_norm": 11.06271989606179,
+      "learning_rate": 2e-05,
+      "loss": 0.5393,
+      "step": 4696
+    },
+    {
+      "epoch": 0.11180139007902504,
+      "grad_norm": 14.393148414142361,
+      "learning_rate": 2e-05,
+      "loss": 1.2227,
+      "step": 4697
+    },
+    {
+      "epoch": 0.11182519280205655,
+      "grad_norm": 16.560834297473317,
+      "learning_rate": 2e-05,
+      "loss": 0.4025,
+      "step": 4698
+    },
+    {
+      "epoch": 0.11184899552508808,
+      "grad_norm": 75.89836342087284,
+      "learning_rate": 2e-05,
+      "loss": 2.2395,
+      "step": 4699
+    },
+    {
+      "epoch": 0.11187279824811959,
+      "grad_norm": 17.65739920673717,
+      "learning_rate": 2e-05,
+      "loss": 2.1739,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1118966009711511,
+      "grad_norm": 14.725703336928607,
+      "learning_rate": 2e-05,
+      "loss": 1.9405,
+      "step": 4701
+    },
+    {
+      "epoch": 0.11192040369418262,
+      "grad_norm": 126.14502072896944,
+      "learning_rate": 2e-05,
+      "loss": 1.2667,
+      "step": 4702
+    },
+    {
+      "epoch": 0.11194420641721413,
+      "grad_norm": 10.328835316464227,
+      "learning_rate": 2e-05,
+      "loss": 0.6871,
+      "step": 4703
+    },
+    {
+      "epoch": 0.11196800914024564,
+      "grad_norm": 10.642475321197491,
+      "learning_rate": 2e-05,
+      "loss": 1.1279,
+      "step": 4704
+    },
+    {
+      "epoch": 0.11199181186327715,
+      "grad_norm": 14.1019137430177,
+      "learning_rate": 2e-05,
+      "loss": 0.5728,
+      "step": 4705
+    },
+    {
+      "epoch": 0.11201561458630867,
+      "grad_norm": 46.84616719932462,
+      "learning_rate": 2e-05,
+      "loss": 1.0482,
+      "step": 4706
+    },
+    {
+      "epoch": 0.1120394173093402,
+      "grad_norm": 14.288127852159064,
+      "learning_rate": 2e-05,
+      "loss": 0.2538,
+      "step": 4707
+    },
+    {
+      "epoch": 0.11206322003237171,
+      "grad_norm": 10.42223083184922,
+      "learning_rate": 2e-05,
+      "loss": 1.1075,
+      "step": 4708
+    },
+    {
+      "epoch": 0.11208702275540322,
+      "grad_norm": 21.168367792679163,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 4709
+    },
+    {
+      "epoch": 0.11211082547843473,
+      "grad_norm": 1.3500940717467829,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11213462820146625,
+      "grad_norm": 17.98892711446482,
+      "learning_rate": 2e-05,
+      "loss": 1.7839,
+      "step": 4711
+    },
+    {
+      "epoch": 0.11215843092449776,
+      "grad_norm": 3.8991126273845316,
+      "learning_rate": 2e-05,
+      "loss": 0.2289,
+      "step": 4712
+    },
+    {
+      "epoch": 0.11218223364752927,
+      "grad_norm": 34.146279484740134,
+      "learning_rate": 2e-05,
+      "loss": 0.316,
+      "step": 4713
+    },
+    {
+      "epoch": 0.11220603637056079,
+      "grad_norm": 17.18873762375087,
+      "learning_rate": 2e-05,
+      "loss": 0.5353,
+      "step": 4714
+    },
+    {
+      "epoch": 0.11222983909359231,
+      "grad_norm": 182.95782139612226,
+      "learning_rate": 2e-05,
+      "loss": 0.8447,
+      "step": 4715
+    },
+    {
+      "epoch": 0.11225364181662383,
+      "grad_norm": 9.747293352972688,
+      "learning_rate": 2e-05,
+      "loss": 0.5425,
+      "step": 4716
+    },
+    {
+      "epoch": 0.11227744453965534,
+      "grad_norm": 272.4452957460308,
+      "learning_rate": 2e-05,
+      "loss": 1.6448,
+      "step": 4717
+    },
+    {
+      "epoch": 0.11230124726268685,
+      "grad_norm": 299.8354420174475,
+      "learning_rate": 2e-05,
+      "loss": 0.6235,
+      "step": 4718
+    },
+    {
+      "epoch": 0.11232504998571836,
+      "grad_norm": 1143.6570916014787,
+      "learning_rate": 2e-05,
+      "loss": 1.4852,
+      "step": 4719
+    },
+    {
+      "epoch": 0.11234885270874988,
+      "grad_norm": 146.66675314743605,
+      "learning_rate": 2e-05,
+      "loss": 0.9895,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11237265543178139,
+      "grad_norm": 169.66049041488785,
+      "learning_rate": 2e-05,
+      "loss": 3.7403,
+      "step": 4721
+    },
+    {
+      "epoch": 0.11239645815481292,
+      "grad_norm": 391.8466221798284,
+      "learning_rate": 2e-05,
+      "loss": 0.6908,
+      "step": 4722
+    },
+    {
+      "epoch": 0.11242026087784443,
+      "grad_norm": 37.431572706535995,
+      "learning_rate": 2e-05,
+      "loss": 0.3769,
+      "step": 4723
+    },
+    {
+      "epoch": 0.11244406360087594,
+      "grad_norm": 6.827045792601079,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 4724
+    },
+    {
+      "epoch": 0.11246786632390746,
+      "grad_norm": 254.57910222197452,
+      "learning_rate": 2e-05,
+      "loss": 3.1909,
+      "step": 4725
+    },
+    {
+      "epoch": 0.11249166904693897,
+      "grad_norm": 31.59917835110705,
+      "learning_rate": 2e-05,
+      "loss": 0.3233,
+      "step": 4726
+    },
+    {
+      "epoch": 0.11251547176997048,
+      "grad_norm": 16.41769989564341,
+      "learning_rate": 2e-05,
+      "loss": 1.0072,
+      "step": 4727
+    },
+    {
+      "epoch": 0.112539274493002,
+      "grad_norm": 8.553613206491642,
+      "learning_rate": 2e-05,
+      "loss": 1.1338,
+      "step": 4728
+    },
+    {
+      "epoch": 0.11256307721603351,
+      "grad_norm": 3.873185330262571,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 4729
+    },
+    {
+      "epoch": 0.11258687993906503,
+      "grad_norm": 159.528732493748,
+      "learning_rate": 2e-05,
+      "loss": 1.3257,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11261068266209655,
+      "grad_norm": 30.71814254775661,
+      "learning_rate": 2e-05,
+      "loss": 1.0486,
+      "step": 4731
+    },
+    {
+      "epoch": 0.11263448538512806,
+      "grad_norm": 8.792768272957991,
+      "learning_rate": 2e-05,
+      "loss": 1.0253,
+      "step": 4732
+    },
+    {
+      "epoch": 0.11265828810815957,
+      "grad_norm": 26.20838876078616,
+      "learning_rate": 2e-05,
+      "loss": 0.536,
+      "step": 4733
+    },
+    {
+      "epoch": 0.11268209083119109,
+      "grad_norm": 10.041015635878995,
+      "learning_rate": 2e-05,
+      "loss": 0.5039,
+      "step": 4734
+    },
+    {
+      "epoch": 0.1127058935542226,
+      "grad_norm": 8.680389795498309,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 4735
+    },
+    {
+      "epoch": 0.11272969627725411,
+      "grad_norm": 1.7780945481107664,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 4736
+    },
+    {
+      "epoch": 0.11275349900028563,
+      "grad_norm": 23.02118604562067,
+      "learning_rate": 2e-05,
+      "loss": 0.8811,
+      "step": 4737
+    },
+    {
+      "epoch": 0.11277730172331715,
+      "grad_norm": 183.33031046416812,
+      "learning_rate": 2e-05,
+      "loss": 1.3812,
+      "step": 4738
+    },
+    {
+      "epoch": 0.11280110444634867,
+      "grad_norm": 12.636766690937176,
+      "learning_rate": 2e-05,
+      "loss": 0.3718,
+      "step": 4739
+    },
+    {
+      "epoch": 0.11282490716938018,
+      "grad_norm": 18.940287177256412,
+      "learning_rate": 2e-05,
+      "loss": 0.3856,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11284870989241169,
+      "grad_norm": 5.985202979592066,
+      "learning_rate": 2e-05,
+      "loss": 0.7886,
+      "step": 4741
+    },
+    {
+      "epoch": 0.1128725126154432,
+      "grad_norm": 8.345697781062958,
+      "learning_rate": 2e-05,
+      "loss": 0.5479,
+      "step": 4742
+    },
+    {
+      "epoch": 0.11289631533847472,
+      "grad_norm": 64.85517227048155,
+      "learning_rate": 2e-05,
+      "loss": 0.4292,
+      "step": 4743
+    },
+    {
+      "epoch": 0.11292011806150623,
+      "grad_norm": 7.930228949135101,
+      "learning_rate": 2e-05,
+      "loss": 0.5688,
+      "step": 4744
+    },
+    {
+      "epoch": 0.11294392078453776,
+      "grad_norm": 11.15852135382125,
+      "learning_rate": 2e-05,
+      "loss": 0.1193,
+      "step": 4745
+    },
+    {
+      "epoch": 0.11296772350756927,
+      "grad_norm": 14.61536956417175,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 4746
+    },
+    {
+      "epoch": 0.11299152623060078,
+      "grad_norm": 6.891514858310605,
+      "learning_rate": 2e-05,
+      "loss": 0.2594,
+      "step": 4747
+    },
+    {
+      "epoch": 0.1130153289536323,
+      "grad_norm": 11.06647567525723,
+      "learning_rate": 2e-05,
+      "loss": 0.4864,
+      "step": 4748
+    },
+    {
+      "epoch": 0.11303913167666381,
+      "grad_norm": 4.87638285764593,
+      "learning_rate": 2e-05,
+      "loss": 0.3699,
+      "step": 4749
+    },
+    {
+      "epoch": 0.11306293439969532,
+      "grad_norm": 5.8593327409521745,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11308673712272684,
+      "grad_norm": 64.52968465024853,
+      "learning_rate": 2e-05,
+      "loss": 0.9848,
+      "step": 4751
+    },
+    {
+      "epoch": 0.11311053984575835,
+      "grad_norm": 15.822712395252937,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 4752
+    },
+    {
+      "epoch": 0.11313434256878988,
+      "grad_norm": 2.376362601271818,
+      "learning_rate": 2e-05,
+      "loss": 0.144,
+      "step": 4753
+    },
+    {
+      "epoch": 0.11315814529182139,
+      "grad_norm": 12.683628964345449,
+      "learning_rate": 2e-05,
+      "loss": 0.7924,
+      "step": 4754
+    },
+    {
+      "epoch": 0.1131819480148529,
+      "grad_norm": 5.674512681769386,
+      "learning_rate": 2e-05,
+      "loss": 0.17,
+      "step": 4755
+    },
+    {
+      "epoch": 0.11320575073788441,
+      "grad_norm": 4.075985093355071,
+      "learning_rate": 2e-05,
+      "loss": 0.5612,
+      "step": 4756
+    },
+    {
+      "epoch": 0.11322955346091593,
+      "grad_norm": 2.2045702279501684,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 4757
+    },
+    {
+      "epoch": 0.11325335618394744,
+      "grad_norm": 4.317624144161001,
+      "learning_rate": 2e-05,
+      "loss": 0.5873,
+      "step": 4758
+    },
+    {
+      "epoch": 0.11327715890697895,
+      "grad_norm": 4.690602409723907,
+      "learning_rate": 2e-05,
+      "loss": 0.7064,
+      "step": 4759
+    },
+    {
+      "epoch": 0.11330096163001048,
+      "grad_norm": 6.717349305903209,
+      "learning_rate": 2e-05,
+      "loss": 0.2949,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11332476435304199,
+      "grad_norm": 1.8716163038611933,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 4761
+    },
+    {
+      "epoch": 0.1133485670760735,
+      "grad_norm": 4.001448673137405,
+      "learning_rate": 2e-05,
+      "loss": 0.4754,
+      "step": 4762
+    },
+    {
+      "epoch": 0.11337236979910502,
+      "grad_norm": 72.0315285379693,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 4763
+    },
+    {
+      "epoch": 0.11339617252213653,
+      "grad_norm": 6.1284269046485775,
+      "learning_rate": 2e-05,
+      "loss": 0.2536,
+      "step": 4764
+    },
+    {
+      "epoch": 0.11341997524516804,
+      "grad_norm": 4.118041929931271,
+      "learning_rate": 2e-05,
+      "loss": 0.5971,
+      "step": 4765
+    },
+    {
+      "epoch": 0.11344377796819956,
+      "grad_norm": 7.027422764966228,
+      "learning_rate": 2e-05,
+      "loss": 1.9505,
+      "step": 4766
+    },
+    {
+      "epoch": 0.11346758069123107,
+      "grad_norm": 3.226185089030688,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 4767
+    },
+    {
+      "epoch": 0.1134913834142626,
+      "grad_norm": 7.874742610949751,
+      "learning_rate": 2e-05,
+      "loss": 1.3627,
+      "step": 4768
+    },
+    {
+      "epoch": 0.11351518613729411,
+      "grad_norm": 9.35848379865519,
+      "learning_rate": 2e-05,
+      "loss": 0.5639,
+      "step": 4769
+    },
+    {
+      "epoch": 0.11353898886032562,
+      "grad_norm": 3.155513474141396,
+      "learning_rate": 2e-05,
+      "loss": 0.173,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11356279158335714,
+      "grad_norm": 7.73683589753136,
+      "learning_rate": 2e-05,
+      "loss": 1.9498,
+      "step": 4771
+    },
+    {
+      "epoch": 0.11358659430638865,
+      "grad_norm": 6.703264263525258,
+      "learning_rate": 2e-05,
+      "loss": 0.3171,
+      "step": 4772
+    },
+    {
+      "epoch": 0.11361039702942016,
+      "grad_norm": 3.188559093390959,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 4773
+    },
+    {
+      "epoch": 0.11363419975245168,
+      "grad_norm": 8.840691107207885,
+      "learning_rate": 2e-05,
+      "loss": 1.0565,
+      "step": 4774
+    },
+    {
+      "epoch": 0.11365800247548319,
+      "grad_norm": 2.852261187291853,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 4775
+    },
+    {
+      "epoch": 0.11368180519851472,
+      "grad_norm": 7.405802963064173,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 4776
+    },
+    {
+      "epoch": 0.11370560792154623,
+      "grad_norm": 6.108456289550815,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 4777
+    },
+    {
+      "epoch": 0.11372941064457774,
+      "grad_norm": 2.147330074172041,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 4778
+    },
+    {
+      "epoch": 0.11375321336760925,
+      "grad_norm": 1.6269089811767565,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 4779
+    },
+    {
+      "epoch": 0.11377701609064077,
+      "grad_norm": 3.1218053016091147,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 4780
+    },
+    {
+      "epoch": 0.11380081881367228,
+      "grad_norm": 2.7341679116695627,
+      "learning_rate": 2e-05,
+      "loss": 0.2107,
+      "step": 4781
+    },
+    {
+      "epoch": 0.1138246215367038,
+      "grad_norm": 6.360661277472284,
+      "learning_rate": 2e-05,
+      "loss": 0.8105,
+      "step": 4782
+    },
+    {
+      "epoch": 0.11384842425973532,
+      "grad_norm": 7.1167326703843505,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 4783
+    },
+    {
+      "epoch": 0.11387222698276683,
+      "grad_norm": 6.929290983548197,
+      "learning_rate": 2e-05,
+      "loss": 0.9271,
+      "step": 4784
+    },
+    {
+      "epoch": 0.11389602970579835,
+      "grad_norm": 1.5248326289716834,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 4785
+    },
+    {
+      "epoch": 0.11391983242882986,
+      "grad_norm": 2.0435960749847237,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 4786
+    },
+    {
+      "epoch": 0.11394363515186137,
+      "grad_norm": 7.069883924966212,
+      "learning_rate": 2e-05,
+      "loss": 0.4115,
+      "step": 4787
+    },
+    {
+      "epoch": 0.11396743787489289,
+      "grad_norm": 3.6078914491803764,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 4788
+    },
+    {
+      "epoch": 0.1139912405979244,
+      "grad_norm": 2.3041208394036454,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 4789
+    },
+    {
+      "epoch": 0.11401504332095591,
+      "grad_norm": 19.621767613638166,
+      "learning_rate": 2e-05,
+      "loss": 2.0256,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11403884604398744,
+      "grad_norm": 7.0052627425333895,
+      "learning_rate": 2e-05,
+      "loss": 0.3754,
+      "step": 4791
+    },
+    {
+      "epoch": 0.11406264876701895,
+      "grad_norm": 4.938361308799949,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 4792
+    },
+    {
+      "epoch": 0.11408645149005046,
+      "grad_norm": 6.60673339610511,
+      "learning_rate": 2e-05,
+      "loss": 0.2761,
+      "step": 4793
+    },
+    {
+      "epoch": 0.11411025421308198,
+      "grad_norm": 5.572592149913874,
+      "learning_rate": 2e-05,
+      "loss": 0.365,
+      "step": 4794
+    },
+    {
+      "epoch": 0.11413405693611349,
+      "grad_norm": 6.468267575350015,
+      "learning_rate": 2e-05,
+      "loss": 0.8869,
+      "step": 4795
+    },
+    {
+      "epoch": 0.114157859659145,
+      "grad_norm": 2.363779762782282,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 4796
+    },
+    {
+      "epoch": 0.11418166238217652,
+      "grad_norm": 9.578416389063088,
+      "learning_rate": 2e-05,
+      "loss": 0.5427,
+      "step": 4797
+    },
+    {
+      "epoch": 0.11420546510520804,
+      "grad_norm": 6.321843735754471,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 4798
+    },
+    {
+      "epoch": 0.11422926782823956,
+      "grad_norm": 2.856498222870879,
+      "learning_rate": 2e-05,
+      "loss": 0.2202,
+      "step": 4799
+    },
+    {
+      "epoch": 0.11425307055127107,
+      "grad_norm": 3.4982422953188115,
+      "learning_rate": 2e-05,
+      "loss": 0.4313,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11427687327430258,
+      "grad_norm": 10.007432905203475,
+      "learning_rate": 2e-05,
+      "loss": 0.9732,
+      "step": 4801
+    },
+    {
+      "epoch": 0.1143006759973341,
+      "grad_norm": 8.133128536163284,
+      "learning_rate": 2e-05,
+      "loss": 0.6671,
+      "step": 4802
+    },
+    {
+      "epoch": 0.11432447872036561,
+      "grad_norm": 3.7373676376130662,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 4803
+    },
+    {
+      "epoch": 0.11434828144339712,
+      "grad_norm": 5.421139175519154,
+      "learning_rate": 2e-05,
+      "loss": 0.2891,
+      "step": 4804
+    },
+    {
+      "epoch": 0.11437208416642863,
+      "grad_norm": 2.216245757298081,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 4805
+    },
+    {
+      "epoch": 0.11439588688946016,
+      "grad_norm": 1.3907035986174525,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 4806
+    },
+    {
+      "epoch": 0.11441968961249167,
+      "grad_norm": 1.8213465275023613,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 4807
+    },
+    {
+      "epoch": 0.11444349233552319,
+      "grad_norm": 3.416005594153654,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 4808
+    },
+    {
+      "epoch": 0.1144672950585547,
+      "grad_norm": 6.466901781876952,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 4809
+    },
+    {
+      "epoch": 0.11449109778158621,
+      "grad_norm": 5.024471189816357,
+      "learning_rate": 2e-05,
+      "loss": 0.2236,
+      "step": 4810
+    },
+    {
+      "epoch": 0.11451490050461773,
+      "grad_norm": 5.931421215574607,
+      "learning_rate": 2e-05,
+      "loss": 0.8333,
+      "step": 4811
+    },
+    {
+      "epoch": 0.11453870322764924,
+      "grad_norm": 13.540146087408733,
+      "learning_rate": 2e-05,
+      "loss": 0.5501,
+      "step": 4812
+    },
+    {
+      "epoch": 0.11456250595068075,
+      "grad_norm": 7.698983568057343,
+      "learning_rate": 2e-05,
+      "loss": 1.1144,
+      "step": 4813
+    },
+    {
+      "epoch": 0.11458630867371228,
+      "grad_norm": 5.026803763526195,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 4814
+    },
+    {
+      "epoch": 0.11461011139674379,
+      "grad_norm": 2.8835061389006786,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 4815
+    },
+    {
+      "epoch": 0.1146339141197753,
+      "grad_norm": 7.525901956421434,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 4816
+    },
+    {
+      "epoch": 0.11465771684280682,
+      "grad_norm": 1.1642989658715837,
+      "learning_rate": 2e-05,
+      "loss": 0.1302,
+      "step": 4817
+    },
+    {
+      "epoch": 0.11468151956583833,
+      "grad_norm": 2.1156662321433677,
+      "learning_rate": 2e-05,
+      "loss": 0.2145,
+      "step": 4818
+    },
+    {
+      "epoch": 0.11470532228886984,
+      "grad_norm": 7.963416301017296,
+      "learning_rate": 2e-05,
+      "loss": 0.4076,
+      "step": 4819
+    },
+    {
+      "epoch": 0.11472912501190136,
+      "grad_norm": 3.3790358085123406,
+      "learning_rate": 2e-05,
+      "loss": 0.3011,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11475292773493288,
+      "grad_norm": 6.810739474994784,
+      "learning_rate": 2e-05,
+      "loss": 0.8757,
+      "step": 4821
+    },
+    {
+      "epoch": 0.1147767304579644,
+      "grad_norm": 9.295675611403553,
+      "learning_rate": 2e-05,
+      "loss": 1.0363,
+      "step": 4822
+    },
+    {
+      "epoch": 0.11480053318099591,
+      "grad_norm": 5.039001184263474,
+      "learning_rate": 2e-05,
+      "loss": 0.3932,
+      "step": 4823
+    },
+    {
+      "epoch": 0.11482433590402742,
+      "grad_norm": 0.7396771602659673,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 4824
+    },
+    {
+      "epoch": 0.11484813862705893,
+      "grad_norm": 8.916607519225307,
+      "learning_rate": 2e-05,
+      "loss": 0.8323,
+      "step": 4825
+    },
+    {
+      "epoch": 0.11487194135009045,
+      "grad_norm": 7.008113085726575,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 4826
+    },
+    {
+      "epoch": 0.11489574407312196,
+      "grad_norm": 5.2088098050092055,
+      "learning_rate": 2e-05,
+      "loss": 0.3187,
+      "step": 4827
+    },
+    {
+      "epoch": 0.11491954679615347,
+      "grad_norm": 13.269479906772768,
+      "learning_rate": 2e-05,
+      "loss": 0.9736,
+      "step": 4828
+    },
+    {
+      "epoch": 0.114943349519185,
+      "grad_norm": 9.118443869149779,
+      "learning_rate": 2e-05,
+      "loss": 0.9003,
+      "step": 4829
+    },
+    {
+      "epoch": 0.11496715224221651,
+      "grad_norm": 8.362730558158646,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11499095496524803,
+      "grad_norm": 10.629389918032299,
+      "learning_rate": 2e-05,
+      "loss": 1.2374,
+      "step": 4831
+    },
+    {
+      "epoch": 0.11501475768827954,
+      "grad_norm": 2.8057991991521356,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 4832
+    },
+    {
+      "epoch": 0.11503856041131105,
+      "grad_norm": 3.6375927052894044,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 4833
+    },
+    {
+      "epoch": 0.11506236313434257,
+      "grad_norm": 7.756218129012421,
+      "learning_rate": 2e-05,
+      "loss": 0.9058,
+      "step": 4834
+    },
+    {
+      "epoch": 0.11508616585737408,
+      "grad_norm": 2.4993323329701655,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 4835
+    },
+    {
+      "epoch": 0.11510996858040559,
+      "grad_norm": 3.65758323366326,
+      "learning_rate": 2e-05,
+      "loss": 0.2141,
+      "step": 4836
+    },
+    {
+      "epoch": 0.11513377130343712,
+      "grad_norm": 5.430309554305257,
+      "learning_rate": 2e-05,
+      "loss": 0.4737,
+      "step": 4837
+    },
+    {
+      "epoch": 0.11515757402646863,
+      "grad_norm": 3.1752209709970436,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 4838
+    },
+    {
+      "epoch": 0.11518137674950014,
+      "grad_norm": 8.704423821042184,
+      "learning_rate": 2e-05,
+      "loss": 2.1053,
+      "step": 4839
+    },
+    {
+      "epoch": 0.11520517947253166,
+      "grad_norm": 6.942779601054383,
+      "learning_rate": 2e-05,
+      "loss": 0.3176,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11522898219556317,
+      "grad_norm": 7.402850085028529,
+      "learning_rate": 2e-05,
+      "loss": 0.9445,
+      "step": 4841
+    },
+    {
+      "epoch": 0.11525278491859468,
+      "grad_norm": 5.160098946262418,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 4842
+    },
+    {
+      "epoch": 0.1152765876416262,
+      "grad_norm": 9.5456596239957,
+      "learning_rate": 2e-05,
+      "loss": 0.3548,
+      "step": 4843
+    },
+    {
+      "epoch": 0.11530039036465772,
+      "grad_norm": 3.2595791260692386,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 4844
+    },
+    {
+      "epoch": 0.11532419308768924,
+      "grad_norm": 4.3028606593578145,
+      "learning_rate": 2e-05,
+      "loss": 0.6707,
+      "step": 4845
+    },
+    {
+      "epoch": 0.11534799581072075,
+      "grad_norm": 4.936373996170996,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 4846
+    },
+    {
+      "epoch": 0.11537179853375226,
+      "grad_norm": 4.987118290182394,
+      "learning_rate": 2e-05,
+      "loss": 0.4159,
+      "step": 4847
+    },
+    {
+      "epoch": 0.11539560125678378,
+      "grad_norm": 5.351532852120362,
+      "learning_rate": 2e-05,
+      "loss": 1.1742,
+      "step": 4848
+    },
+    {
+      "epoch": 0.11541940397981529,
+      "grad_norm": 6.248566168367903,
+      "learning_rate": 2e-05,
+      "loss": 0.8216,
+      "step": 4849
+    },
+    {
+      "epoch": 0.1154432067028468,
+      "grad_norm": 7.447265100069008,
+      "learning_rate": 2e-05,
+      "loss": 0.8981,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11546700942587831,
+      "grad_norm": 8.051780324564497,
+      "learning_rate": 2e-05,
+      "loss": 0.2993,
+      "step": 4851
+    },
+    {
+      "epoch": 0.11549081214890984,
+      "grad_norm": 5.799334669624975,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 4852
+    },
+    {
+      "epoch": 0.11551461487194135,
+      "grad_norm": 3.2701097727959683,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 4853
+    },
+    {
+      "epoch": 0.11553841759497287,
+      "grad_norm": 2.977311001277156,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 4854
+    },
+    {
+      "epoch": 0.11556222031800438,
+      "grad_norm": 1.5754763204352142,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 4855
+    },
+    {
+      "epoch": 0.11558602304103589,
+      "grad_norm": 7.094878496004383,
+      "learning_rate": 2e-05,
+      "loss": 0.9206,
+      "step": 4856
+    },
+    {
+      "epoch": 0.1156098257640674,
+      "grad_norm": 5.826133117188179,
+      "learning_rate": 2e-05,
+      "loss": 0.406,
+      "step": 4857
+    },
+    {
+      "epoch": 0.11563362848709892,
+      "grad_norm": 5.465662645630504,
+      "learning_rate": 2e-05,
+      "loss": 0.2493,
+      "step": 4858
+    },
+    {
+      "epoch": 0.11565743121013045,
+      "grad_norm": 7.883790971404638,
+      "learning_rate": 2e-05,
+      "loss": 1.1746,
+      "step": 4859
+    },
+    {
+      "epoch": 0.11568123393316196,
+      "grad_norm": 2.72575745967894,
+      "learning_rate": 2e-05,
+      "loss": 0.1944,
+      "step": 4860
+    },
+    {
+      "epoch": 0.11570503665619347,
+      "grad_norm": 6.327742559116763,
+      "learning_rate": 2e-05,
+      "loss": 0.5394,
+      "step": 4861
+    },
+    {
+      "epoch": 0.11572883937922498,
+      "grad_norm": 5.656738868672595,
+      "learning_rate": 2e-05,
+      "loss": 0.2759,
+      "step": 4862
+    },
+    {
+      "epoch": 0.1157526421022565,
+      "grad_norm": 6.444142829417775,
+      "learning_rate": 2e-05,
+      "loss": 0.7893,
+      "step": 4863
+    },
+    {
+      "epoch": 0.11577644482528801,
+      "grad_norm": 2.6664256343133386,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 4864
+    },
+    {
+      "epoch": 0.11580024754831952,
+      "grad_norm": 2.6210093067148383,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 4865
+    },
+    {
+      "epoch": 0.11582405027135104,
+      "grad_norm": 7.351914174075198,
+      "learning_rate": 2e-05,
+      "loss": 1.2036,
+      "step": 4866
+    },
+    {
+      "epoch": 0.11584785299438256,
+      "grad_norm": 3.537236488302681,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 4867
+    },
+    {
+      "epoch": 0.11587165571741408,
+      "grad_norm": 9.598024659822991,
+      "learning_rate": 2e-05,
+      "loss": 0.4592,
+      "step": 4868
+    },
+    {
+      "epoch": 0.11589545844044559,
+      "grad_norm": 4.5584940791741255,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 4869
+    },
+    {
+      "epoch": 0.1159192611634771,
+      "grad_norm": 6.392386629751258,
+      "learning_rate": 2e-05,
+      "loss": 0.8271,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11594306388650862,
+      "grad_norm": 1.73764263014522,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 4871
+    },
+    {
+      "epoch": 0.11596686660954013,
+      "grad_norm": 7.623281473802597,
+      "learning_rate": 2e-05,
+      "loss": 0.4881,
+      "step": 4872
+    },
+    {
+      "epoch": 0.11599066933257164,
+      "grad_norm": 6.33487365146986,
+      "learning_rate": 2e-05,
+      "loss": 0.3918,
+      "step": 4873
+    },
+    {
+      "epoch": 0.11601447205560315,
+      "grad_norm": 5.774652021585197,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 4874
+    },
+    {
+      "epoch": 0.11603827477863468,
+      "grad_norm": 3.5472451413876973,
+      "learning_rate": 2e-05,
+      "loss": 0.2855,
+      "step": 4875
+    },
+    {
+      "epoch": 0.1160620775016662,
+      "grad_norm": 7.3632763303025355,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 4876
+    },
+    {
+      "epoch": 0.11608588022469771,
+      "grad_norm": 2.939117124965315,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 4877
+    },
+    {
+      "epoch": 0.11610968294772922,
+      "grad_norm": 8.319218768956118,
+      "learning_rate": 2e-05,
+      "loss": 0.7118,
+      "step": 4878
+    },
+    {
+      "epoch": 0.11613348567076073,
+      "grad_norm": 2.8590970867819467,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 4879
+    },
+    {
+      "epoch": 0.11615728839379225,
+      "grad_norm": 5.09219922781691,
+      "learning_rate": 2e-05,
+      "loss": 0.2063,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11618109111682376,
+      "grad_norm": 9.705837588648555,
+      "learning_rate": 2e-05,
+      "loss": 1.254,
+      "step": 4881
+    },
+    {
+      "epoch": 0.11620489383985529,
+      "grad_norm": 7.194549811739654,
+      "learning_rate": 2e-05,
+      "loss": 0.352,
+      "step": 4882
+    },
+    {
+      "epoch": 0.1162286965628868,
+      "grad_norm": 8.802888610599867,
+      "learning_rate": 2e-05,
+      "loss": 0.8333,
+      "step": 4883
+    },
+    {
+      "epoch": 0.11625249928591831,
+      "grad_norm": 8.666833611372205,
+      "learning_rate": 2e-05,
+      "loss": 0.2938,
+      "step": 4884
+    },
+    {
+      "epoch": 0.11627630200894983,
+      "grad_norm": 5.928457483157066,
+      "learning_rate": 2e-05,
+      "loss": 1.0604,
+      "step": 4885
+    },
+    {
+      "epoch": 0.11630010473198134,
+      "grad_norm": 4.650207066539362,
+      "learning_rate": 2e-05,
+      "loss": 0.8907,
+      "step": 4886
+    },
+    {
+      "epoch": 0.11632390745501285,
+      "grad_norm": 3.5716553818815195,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 4887
+    },
+    {
+      "epoch": 0.11634771017804436,
+      "grad_norm": 6.844184004590073,
+      "learning_rate": 2e-05,
+      "loss": 1.2606,
+      "step": 4888
+    },
+    {
+      "epoch": 0.11637151290107588,
+      "grad_norm": 6.42672185761277,
+      "learning_rate": 2e-05,
+      "loss": 0.7296,
+      "step": 4889
+    },
+    {
+      "epoch": 0.1163953156241074,
+      "grad_norm": 5.396681202748465,
+      "learning_rate": 2e-05,
+      "loss": 0.7073,
+      "step": 4890
+    },
+    {
+      "epoch": 0.11641911834713892,
+      "grad_norm": 6.080908635713778,
+      "learning_rate": 2e-05,
+      "loss": 0.2933,
+      "step": 4891
+    },
+    {
+      "epoch": 0.11644292107017043,
+      "grad_norm": 2.1687774144826975,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 4892
+    },
+    {
+      "epoch": 0.11646672379320194,
+      "grad_norm": 6.260807895902922,
+      "learning_rate": 2e-05,
+      "loss": 0.4179,
+      "step": 4893
+    },
+    {
+      "epoch": 0.11649052651623346,
+      "grad_norm": 1.314714092884477,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 4894
+    },
+    {
+      "epoch": 0.11651432923926497,
+      "grad_norm": 5.4113335884283496,
+      "learning_rate": 2e-05,
+      "loss": 0.5192,
+      "step": 4895
+    },
+    {
+      "epoch": 0.11653813196229648,
+      "grad_norm": 3.523083841711034,
+      "learning_rate": 2e-05,
+      "loss": 0.2327,
+      "step": 4896
+    },
+    {
+      "epoch": 0.116561934685328,
+      "grad_norm": 6.793629860835849,
+      "learning_rate": 2e-05,
+      "loss": 1.2144,
+      "step": 4897
+    },
+    {
+      "epoch": 0.11658573740835952,
+      "grad_norm": 6.795964342567533,
+      "learning_rate": 2e-05,
+      "loss": 0.408,
+      "step": 4898
+    },
+    {
+      "epoch": 0.11660954013139103,
+      "grad_norm": 47.408034990836164,
+      "learning_rate": 2e-05,
+      "loss": 0.2978,
+      "step": 4899
+    },
+    {
+      "epoch": 0.11663334285442255,
+      "grad_norm": 6.0532414241916435,
+      "learning_rate": 2e-05,
+      "loss": 1.1082,
+      "step": 4900
+    },
+    {
+      "epoch": 0.11665714557745406,
+      "grad_norm": 14.813230403950177,
+      "learning_rate": 2e-05,
+      "loss": 0.7315,
+      "step": 4901
+    },
+    {
+      "epoch": 0.11668094830048557,
+      "grad_norm": 6.334693016141302,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 4902
+    },
+    {
+      "epoch": 0.11670475102351709,
+      "grad_norm": 4.791451019931469,
+      "learning_rate": 2e-05,
+      "loss": 0.2874,
+      "step": 4903
+    },
+    {
+      "epoch": 0.1167285537465486,
+      "grad_norm": 7.292549843843692,
+      "learning_rate": 2e-05,
+      "loss": 0.6662,
+      "step": 4904
+    },
+    {
+      "epoch": 0.11675235646958013,
+      "grad_norm": 31.27526932643581,
+      "learning_rate": 2e-05,
+      "loss": 1.4832,
+      "step": 4905
+    },
+    {
+      "epoch": 0.11677615919261164,
+      "grad_norm": 8.481030728255615,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 4906
+    },
+    {
+      "epoch": 0.11679996191564315,
+      "grad_norm": 10.848179434219682,
+      "learning_rate": 2e-05,
+      "loss": 0.4769,
+      "step": 4907
+    },
+    {
+      "epoch": 0.11682376463867467,
+      "grad_norm": 2.110276330592406,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 4908
+    },
+    {
+      "epoch": 0.11684756736170618,
+      "grad_norm": 2.4043609431086757,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 4909
+    },
+    {
+      "epoch": 0.11687137008473769,
+      "grad_norm": 0.9302854014397872,
+      "learning_rate": 2e-05,
+      "loss": 0.1273,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1168951728077692,
+      "grad_norm": 4.722182793811118,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 4911
+    },
+    {
+      "epoch": 0.11691897553080072,
+      "grad_norm": 11.370779004838376,
+      "learning_rate": 2e-05,
+      "loss": 0.2421,
+      "step": 4912
+    },
+    {
+      "epoch": 0.11694277825383224,
+      "grad_norm": 4.09199180944961,
+      "learning_rate": 2e-05,
+      "loss": 0.2529,
+      "step": 4913
+    },
+    {
+      "epoch": 0.11696658097686376,
+      "grad_norm": 8.940985177704261,
+      "learning_rate": 2e-05,
+      "loss": 0.4776,
+      "step": 4914
+    },
+    {
+      "epoch": 0.11699038369989527,
+      "grad_norm": 25.983236369421004,
+      "learning_rate": 2e-05,
+      "loss": 1.7132,
+      "step": 4915
+    },
+    {
+      "epoch": 0.11701418642292678,
+      "grad_norm": 4.8657945745347435,
+      "learning_rate": 2e-05,
+      "loss": 0.2934,
+      "step": 4916
+    },
+    {
+      "epoch": 0.1170379891459583,
+      "grad_norm": 4.685454893377123,
+      "learning_rate": 2e-05,
+      "loss": 0.3163,
+      "step": 4917
+    },
+    {
+      "epoch": 0.11706179186898981,
+      "grad_norm": 8.893237854967962,
+      "learning_rate": 2e-05,
+      "loss": 1.3086,
+      "step": 4918
+    },
+    {
+      "epoch": 0.11708559459202132,
+      "grad_norm": 19.1918437933905,
+      "learning_rate": 2e-05,
+      "loss": 0.7574,
+      "step": 4919
+    },
+    {
+      "epoch": 0.11710939731505285,
+      "grad_norm": 6.422596029801759,
+      "learning_rate": 2e-05,
+      "loss": 0.2251,
+      "step": 4920
+    },
+    {
+      "epoch": 0.11713320003808436,
+      "grad_norm": 3.687856272224447,
+      "learning_rate": 2e-05,
+      "loss": 0.3421,
+      "step": 4921
+    },
+    {
+      "epoch": 0.11715700276111587,
+      "grad_norm": 3.3047266175285457,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 4922
+    },
+    {
+      "epoch": 0.11718080548414739,
+      "grad_norm": 3.4697925500455553,
+      "learning_rate": 2e-05,
+      "loss": 0.1277,
+      "step": 4923
+    },
+    {
+      "epoch": 0.1172046082071789,
+      "grad_norm": 6.339759838434039,
+      "learning_rate": 2e-05,
+      "loss": 0.6941,
+      "step": 4924
+    },
+    {
+      "epoch": 0.11722841093021041,
+      "grad_norm": 2.890067760555565,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 4925
+    },
+    {
+      "epoch": 0.11725221365324193,
+      "grad_norm": 5.810240957677293,
+      "learning_rate": 2e-05,
+      "loss": 0.8681,
+      "step": 4926
+    },
+    {
+      "epoch": 0.11727601637627344,
+      "grad_norm": 7.305391715353436,
+      "learning_rate": 2e-05,
+      "loss": 1.5496,
+      "step": 4927
+    },
+    {
+      "epoch": 0.11729981909930497,
+      "grad_norm": 4.190583951193446,
+      "learning_rate": 2e-05,
+      "loss": 0.3625,
+      "step": 4928
+    },
+    {
+      "epoch": 0.11732362182233648,
+      "grad_norm": 3.289821545500375,
+      "learning_rate": 2e-05,
+      "loss": 0.2914,
+      "step": 4929
+    },
+    {
+      "epoch": 0.11734742454536799,
+      "grad_norm": 3.7503549257737645,
+      "learning_rate": 2e-05,
+      "loss": 0.3208,
+      "step": 4930
+    },
+    {
+      "epoch": 0.1173712272683995,
+      "grad_norm": 7.155237488902503,
+      "learning_rate": 2e-05,
+      "loss": 1.093,
+      "step": 4931
+    },
+    {
+      "epoch": 0.11739502999143102,
+      "grad_norm": 4.734212424570628,
+      "learning_rate": 2e-05,
+      "loss": 0.242,
+      "step": 4932
+    },
+    {
+      "epoch": 0.11741883271446253,
+      "grad_norm": 4.172531784235488,
+      "learning_rate": 2e-05,
+      "loss": 0.1782,
+      "step": 4933
+    },
+    {
+      "epoch": 0.11744263543749404,
+      "grad_norm": 8.205962333870925,
+      "learning_rate": 2e-05,
+      "loss": 0.7285,
+      "step": 4934
+    },
+    {
+      "epoch": 0.11746643816052556,
+      "grad_norm": 4.128895257320899,
+      "learning_rate": 2e-05,
+      "loss": 0.8227,
+      "step": 4935
+    },
+    {
+      "epoch": 0.11749024088355708,
+      "grad_norm": 8.952061640010504,
+      "learning_rate": 2e-05,
+      "loss": 1.1741,
+      "step": 4936
+    },
+    {
+      "epoch": 0.1175140436065886,
+      "grad_norm": 1.9955548643556005,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 4937
+    },
+    {
+      "epoch": 0.11753784632962011,
+      "grad_norm": 2.5331957811213526,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 4938
+    },
+    {
+      "epoch": 0.11756164905265162,
+      "grad_norm": 7.134302608146519,
+      "learning_rate": 2e-05,
+      "loss": 0.386,
+      "step": 4939
+    },
+    {
+      "epoch": 0.11758545177568314,
+      "grad_norm": 4.783443438953998,
+      "learning_rate": 2e-05,
+      "loss": 0.4804,
+      "step": 4940
+    },
+    {
+      "epoch": 0.11760925449871465,
+      "grad_norm": 5.633583610193467,
+      "learning_rate": 2e-05,
+      "loss": 0.4629,
+      "step": 4941
+    },
+    {
+      "epoch": 0.11763305722174616,
+      "grad_norm": 7.989395760226992,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 4942
+    },
+    {
+      "epoch": 0.11765685994477769,
+      "grad_norm": 4.812374532618999,
+      "learning_rate": 2e-05,
+      "loss": 0.7927,
+      "step": 4943
+    },
+    {
+      "epoch": 0.1176806626678092,
+      "grad_norm": 2.083586240625732,
+      "learning_rate": 2e-05,
+      "loss": 0.1621,
+      "step": 4944
+    },
+    {
+      "epoch": 0.11770446539084072,
+      "grad_norm": 5.157085319684075,
+      "learning_rate": 2e-05,
+      "loss": 0.2431,
+      "step": 4945
+    },
+    {
+      "epoch": 0.11772826811387223,
+      "grad_norm": 13.428533965709539,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 4946
+    },
+    {
+      "epoch": 0.11775207083690374,
+      "grad_norm": 8.348580782481012,
+      "learning_rate": 2e-05,
+      "loss": 0.4304,
+      "step": 4947
+    },
+    {
+      "epoch": 0.11777587355993525,
+      "grad_norm": 3.330500380893224,
+      "learning_rate": 2e-05,
+      "loss": 0.3668,
+      "step": 4948
+    },
+    {
+      "epoch": 0.11779967628296677,
+      "grad_norm": 6.911170763847315,
+      "learning_rate": 2e-05,
+      "loss": 1.366,
+      "step": 4949
+    },
+    {
+      "epoch": 0.11782347900599828,
+      "grad_norm": 22.96573689646586,
+      "learning_rate": 2e-05,
+      "loss": 0.9487,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1178472817290298,
+      "grad_norm": 3.6866909768952247,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 4951
+    },
+    {
+      "epoch": 0.11787108445206132,
+      "grad_norm": 4.692885851807019,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 4952
+    },
+    {
+      "epoch": 0.11789488717509283,
+      "grad_norm": 1.3370893975928626,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 4953
+    },
+    {
+      "epoch": 0.11791868989812435,
+      "grad_norm": 7.0593487793211285,
+      "learning_rate": 2e-05,
+      "loss": 1.089,
+      "step": 4954
+    },
+    {
+      "epoch": 0.11794249262115586,
+      "grad_norm": 7.354014465681345,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 4955
+    },
+    {
+      "epoch": 0.11796629534418737,
+      "grad_norm": 1.8176789773610476,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 4956
+    },
+    {
+      "epoch": 0.11799009806721888,
+      "grad_norm": 4.002880849114806,
+      "learning_rate": 2e-05,
+      "loss": 0.5745,
+      "step": 4957
+    },
+    {
+      "epoch": 0.1180139007902504,
+      "grad_norm": 5.6481409476628075,
+      "learning_rate": 2e-05,
+      "loss": 0.6315,
+      "step": 4958
+    },
+    {
+      "epoch": 0.11803770351328192,
+      "grad_norm": 5.344289595835718,
+      "learning_rate": 2e-05,
+      "loss": 0.5803,
+      "step": 4959
+    },
+    {
+      "epoch": 0.11806150623631344,
+      "grad_norm": 3.5618503913279334,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 4960
+    },
+    {
+      "epoch": 0.11808530895934495,
+      "grad_norm": 1.7817229322520047,
+      "learning_rate": 2e-05,
+      "loss": 0.173,
+      "step": 4961
+    },
+    {
+      "epoch": 0.11810911168237646,
+      "grad_norm": 2.2304670172532597,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 4962
+    },
+    {
+      "epoch": 0.11813291440540798,
+      "grad_norm": 9.25424820620953,
+      "learning_rate": 2e-05,
+      "loss": 0.5525,
+      "step": 4963
+    },
+    {
+      "epoch": 0.11815671712843949,
+      "grad_norm": 4.6401915197150565,
+      "learning_rate": 2e-05,
+      "loss": 0.6213,
+      "step": 4964
+    },
+    {
+      "epoch": 0.118180519851471,
+      "grad_norm": 3.3003173174538283,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 4965
+    },
+    {
+      "epoch": 0.11820432257450253,
+      "grad_norm": 1.0544325392156157,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 4966
+    },
+    {
+      "epoch": 0.11822812529753404,
+      "grad_norm": 3.626443599006333,
+      "learning_rate": 2e-05,
+      "loss": 0.2699,
+      "step": 4967
+    },
+    {
+      "epoch": 0.11825192802056556,
+      "grad_norm": 4.307985601249075,
+      "learning_rate": 2e-05,
+      "loss": 0.7699,
+      "step": 4968
+    },
+    {
+      "epoch": 0.11827573074359707,
+      "grad_norm": 2.3793075202144003,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 4969
+    },
+    {
+      "epoch": 0.11829953346662858,
+      "grad_norm": 10.460510570021755,
+      "learning_rate": 2e-05,
+      "loss": 0.5389,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1183233361896601,
+      "grad_norm": 3.2674899590098363,
+      "learning_rate": 2e-05,
+      "loss": 0.2865,
+      "step": 4971
+    },
+    {
+      "epoch": 0.11834713891269161,
+      "grad_norm": 4.422156332290222,
+      "learning_rate": 2e-05,
+      "loss": 0.2738,
+      "step": 4972
+    },
+    {
+      "epoch": 0.11837094163572312,
+      "grad_norm": 6.617004891197485,
+      "learning_rate": 2e-05,
+      "loss": 0.9254,
+      "step": 4973
+    },
+    {
+      "epoch": 0.11839474435875465,
+      "grad_norm": 3.8503931263247124,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 4974
+    },
+    {
+      "epoch": 0.11841854708178616,
+      "grad_norm": 2.448118252554656,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 4975
+    },
+    {
+      "epoch": 0.11844234980481767,
+      "grad_norm": 2.698518144537628,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 4976
+    },
+    {
+      "epoch": 0.11846615252784919,
+      "grad_norm": 4.750082872903494,
+      "learning_rate": 2e-05,
+      "loss": 0.2904,
+      "step": 4977
+    },
+    {
+      "epoch": 0.1184899552508807,
+      "grad_norm": 1.7734290942349593,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 4978
+    },
+    {
+      "epoch": 0.11851375797391221,
+      "grad_norm": 1.9549612583026632,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 4979
+    },
+    {
+      "epoch": 0.11853756069694373,
+      "grad_norm": 11.21736680583199,
+      "learning_rate": 2e-05,
+      "loss": 0.5015,
+      "step": 4980
+    },
+    {
+      "epoch": 0.11856136341997525,
+      "grad_norm": 7.711152913407559,
+      "learning_rate": 2e-05,
+      "loss": 0.7736,
+      "step": 4981
+    },
+    {
+      "epoch": 0.11858516614300676,
+      "grad_norm": 6.151749724646667,
+      "learning_rate": 2e-05,
+      "loss": 0.8262,
+      "step": 4982
+    },
+    {
+      "epoch": 0.11860896886603828,
+      "grad_norm": 11.883514302743695,
+      "learning_rate": 2e-05,
+      "loss": 0.5628,
+      "step": 4983
+    },
+    {
+      "epoch": 0.11863277158906979,
+      "grad_norm": 2.147413040643141,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 4984
+    },
+    {
+      "epoch": 0.1186565743121013,
+      "grad_norm": 5.325768172274958,
+      "learning_rate": 2e-05,
+      "loss": 0.8349,
+      "step": 4985
+    },
+    {
+      "epoch": 0.11868037703513282,
+      "grad_norm": 0.9304342460671495,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 4986
+    },
+    {
+      "epoch": 0.11870417975816433,
+      "grad_norm": 3.1438137051460453,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 4987
+    },
+    {
+      "epoch": 0.11872798248119584,
+      "grad_norm": 7.7316736496795775,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 4988
+    },
+    {
+      "epoch": 0.11875178520422737,
+      "grad_norm": 9.0734495502962,
+      "learning_rate": 2e-05,
+      "loss": 1.5216,
+      "step": 4989
+    },
+    {
+      "epoch": 0.11877558792725888,
+      "grad_norm": 7.227698562959801,
+      "learning_rate": 2e-05,
+      "loss": 0.3347,
+      "step": 4990
+    },
+    {
+      "epoch": 0.1187993906502904,
+      "grad_norm": 5.61987547621291,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 4991
+    },
+    {
+      "epoch": 0.11882319337332191,
+      "grad_norm": 4.139466374673181,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 4992
+    },
+    {
+      "epoch": 0.11884699609635342,
+      "grad_norm": 3.0032407662697143,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 4993
+    },
+    {
+      "epoch": 0.11887079881938493,
+      "grad_norm": 4.2571249568873775,
+      "learning_rate": 2e-05,
+      "loss": 0.3961,
+      "step": 4994
+    },
+    {
+      "epoch": 0.11889460154241645,
+      "grad_norm": 3.2159630844591343,
+      "learning_rate": 2e-05,
+      "loss": 0.265,
+      "step": 4995
+    },
+    {
+      "epoch": 0.11891840426544796,
+      "grad_norm": 1.591751228553937,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 4996
+    },
+    {
+      "epoch": 0.11894220698847949,
+      "grad_norm": 2.619469633359261,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 4997
+    },
+    {
+      "epoch": 0.118966009711511,
+      "grad_norm": 8.399910263210897,
+      "learning_rate": 2e-05,
+      "loss": 1.6984,
+      "step": 4998
+    },
+    {
+      "epoch": 0.11898981243454251,
+      "grad_norm": 7.285264264466389,
+      "learning_rate": 2e-05,
+      "loss": 1.6943,
+      "step": 4999
+    },
+    {
+      "epoch": 0.11901361515757403,
+      "grad_norm": 4.635128415249327,
+      "learning_rate": 2e-05,
+      "loss": 0.3051,
+      "step": 5000
+    },
+    {
+      "epoch": 0.11903741788060554,
+      "grad_norm": 3.0964009106923522,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 5001
+    },
+    {
+      "epoch": 0.11906122060363705,
+      "grad_norm": 2.9962720545174717,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 5002
+    },
+    {
+      "epoch": 0.11908502332666857,
+      "grad_norm": 6.460352641849011,
+      "learning_rate": 2e-05,
+      "loss": 0.2993,
+      "step": 5003
+    },
+    {
+      "epoch": 0.11910882604970009,
+      "grad_norm": 3.9703976635700617,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 5004
+    },
+    {
+      "epoch": 0.1191326287727316,
+      "grad_norm": 6.884143298966365,
+      "learning_rate": 2e-05,
+      "loss": 0.8106,
+      "step": 5005
+    },
+    {
+      "epoch": 0.11915643149576312,
+      "grad_norm": 4.21922222064274,
+      "learning_rate": 2e-05,
+      "loss": 0.3415,
+      "step": 5006
+    },
+    {
+      "epoch": 0.11918023421879463,
+      "grad_norm": 7.611576885335416,
+      "learning_rate": 2e-05,
+      "loss": 0.4755,
+      "step": 5007
+    },
+    {
+      "epoch": 0.11920403694182614,
+      "grad_norm": 6.644638757140073,
+      "learning_rate": 2e-05,
+      "loss": 1.222,
+      "step": 5008
+    },
+    {
+      "epoch": 0.11922783966485766,
+      "grad_norm": 29.687879954415934,
+      "learning_rate": 2e-05,
+      "loss": 1.9372,
+      "step": 5009
+    },
+    {
+      "epoch": 0.11925164238788917,
+      "grad_norm": 4.049588628835882,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 5010
+    },
+    {
+      "epoch": 0.11927544511092068,
+      "grad_norm": 3.8852719659275614,
+      "learning_rate": 2e-05,
+      "loss": 0.2552,
+      "step": 5011
+    },
+    {
+      "epoch": 0.11929924783395221,
+      "grad_norm": 7.230000201398585,
+      "learning_rate": 2e-05,
+      "loss": 0.7066,
+      "step": 5012
+    },
+    {
+      "epoch": 0.11932305055698372,
+      "grad_norm": 2.4671518750593298,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 5013
+    },
+    {
+      "epoch": 0.11934685328001524,
+      "grad_norm": 7.034903730280832,
+      "learning_rate": 2e-05,
+      "loss": 0.3386,
+      "step": 5014
+    },
+    {
+      "epoch": 0.11937065600304675,
+      "grad_norm": 5.07954734722116,
+      "learning_rate": 2e-05,
+      "loss": 0.622,
+      "step": 5015
+    },
+    {
+      "epoch": 0.11939445872607826,
+      "grad_norm": 7.868789798691688,
+      "learning_rate": 2e-05,
+      "loss": 1.0372,
+      "step": 5016
+    },
+    {
+      "epoch": 0.11941826144910977,
+      "grad_norm": 2.854489355712285,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 5017
+    },
+    {
+      "epoch": 0.11944206417214129,
+      "grad_norm": 6.500927523847056,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 5018
+    },
+    {
+      "epoch": 0.1194658668951728,
+      "grad_norm": 2.9988369618193933,
+      "learning_rate": 2e-05,
+      "loss": 0.1989,
+      "step": 5019
+    },
+    {
+      "epoch": 0.11948966961820433,
+      "grad_norm": 4.054893833177234,
+      "learning_rate": 2e-05,
+      "loss": 0.5817,
+      "step": 5020
+    },
+    {
+      "epoch": 0.11951347234123584,
+      "grad_norm": 4.691321959091021,
+      "learning_rate": 2e-05,
+      "loss": 0.6213,
+      "step": 5021
+    },
+    {
+      "epoch": 0.11953727506426735,
+      "grad_norm": 2.4822747199717123,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 5022
+    },
+    {
+      "epoch": 0.11956107778729887,
+      "grad_norm": 10.009821135679495,
+      "learning_rate": 2e-05,
+      "loss": 0.6685,
+      "step": 5023
+    },
+    {
+      "epoch": 0.11958488051033038,
+      "grad_norm": 2.456625543449981,
+      "learning_rate": 2e-05,
+      "loss": 0.2302,
+      "step": 5024
+    },
+    {
+      "epoch": 0.11960868323336189,
+      "grad_norm": 5.946975990115157,
+      "learning_rate": 2e-05,
+      "loss": 1.1132,
+      "step": 5025
+    },
+    {
+      "epoch": 0.1196324859563934,
+      "grad_norm": 8.94338345276429,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 5026
+    },
+    {
+      "epoch": 0.11965628867942493,
+      "grad_norm": 7.182909207688572,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 5027
+    },
+    {
+      "epoch": 0.11968009140245645,
+      "grad_norm": 6.040309364477843,
+      "learning_rate": 2e-05,
+      "loss": 0.218,
+      "step": 5028
+    },
+    {
+      "epoch": 0.11970389412548796,
+      "grad_norm": 5.602284585391714,
+      "learning_rate": 2e-05,
+      "loss": 0.8904,
+      "step": 5029
+    },
+    {
+      "epoch": 0.11972769684851947,
+      "grad_norm": 1.6227183733668342,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 5030
+    },
+    {
+      "epoch": 0.11975149957155098,
+      "grad_norm": 8.496739399292396,
+      "learning_rate": 2e-05,
+      "loss": 1.0321,
+      "step": 5031
+    },
+    {
+      "epoch": 0.1197753022945825,
+      "grad_norm": 3.5323879304673373,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 5032
+    },
+    {
+      "epoch": 0.11979910501761401,
+      "grad_norm": 9.217899412900236,
+      "learning_rate": 2e-05,
+      "loss": 0.9492,
+      "step": 5033
+    },
+    {
+      "epoch": 0.11982290774064552,
+      "grad_norm": 2.6189620861872984,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 5034
+    },
+    {
+      "epoch": 0.11984671046367705,
+      "grad_norm": 8.906241189746048,
+      "learning_rate": 2e-05,
+      "loss": 0.5074,
+      "step": 5035
+    },
+    {
+      "epoch": 0.11987051318670856,
+      "grad_norm": 3.652282246628854,
+      "learning_rate": 2e-05,
+      "loss": 0.1858,
+      "step": 5036
+    },
+    {
+      "epoch": 0.11989431590974008,
+      "grad_norm": 4.250774699459481,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 5037
+    },
+    {
+      "epoch": 0.11991811863277159,
+      "grad_norm": 4.7373995689915676,
+      "learning_rate": 2e-05,
+      "loss": 0.5995,
+      "step": 5038
+    },
+    {
+      "epoch": 0.1199419213558031,
+      "grad_norm": 7.768676516068686,
+      "learning_rate": 2e-05,
+      "loss": 0.9475,
+      "step": 5039
+    },
+    {
+      "epoch": 0.11996572407883462,
+      "grad_norm": 5.864046966857189,
+      "learning_rate": 2e-05,
+      "loss": 0.4476,
+      "step": 5040
+    },
+    {
+      "epoch": 0.11998952680186613,
+      "grad_norm": 5.54495512302654,
+      "learning_rate": 2e-05,
+      "loss": 0.7659,
+      "step": 5041
+    },
+    {
+      "epoch": 0.12001332952489766,
+      "grad_norm": 3.4308271421098295,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 5042
+    },
+    {
+      "epoch": 0.12003713224792917,
+      "grad_norm": 5.553777088679011,
+      "learning_rate": 2e-05,
+      "loss": 0.6091,
+      "step": 5043
+    },
+    {
+      "epoch": 0.12006093497096068,
+      "grad_norm": 3.863374847429927,
+      "learning_rate": 2e-05,
+      "loss": 0.3059,
+      "step": 5044
+    },
+    {
+      "epoch": 0.1200847376939922,
+      "grad_norm": 1.2415054945556638,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 5045
+    },
+    {
+      "epoch": 0.12010854041702371,
+      "grad_norm": 5.584984532245718,
+      "learning_rate": 2e-05,
+      "loss": 0.3334,
+      "step": 5046
+    },
+    {
+      "epoch": 0.12013234314005522,
+      "grad_norm": 2.7334057776254785,
+      "learning_rate": 2e-05,
+      "loss": 0.199,
+      "step": 5047
+    },
+    {
+      "epoch": 0.12015614586308673,
+      "grad_norm": 9.329185448637197,
+      "learning_rate": 2e-05,
+      "loss": 0.5087,
+      "step": 5048
+    },
+    {
+      "epoch": 0.12017994858611825,
+      "grad_norm": 3.590151011007613,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 5049
+    },
+    {
+      "epoch": 0.12020375130914977,
+      "grad_norm": 7.176767487428391,
+      "learning_rate": 2e-05,
+      "loss": 1.3272,
+      "step": 5050
+    },
+    {
+      "epoch": 0.12022755403218129,
+      "grad_norm": 6.996140851311262,
+      "learning_rate": 2e-05,
+      "loss": 0.2907,
+      "step": 5051
+    },
+    {
+      "epoch": 0.1202513567552128,
+      "grad_norm": 4.969556946387803,
+      "learning_rate": 2e-05,
+      "loss": 0.1934,
+      "step": 5052
+    },
+    {
+      "epoch": 0.12027515947824431,
+      "grad_norm": 7.200332802263738,
+      "learning_rate": 2e-05,
+      "loss": 0.6851,
+      "step": 5053
+    },
+    {
+      "epoch": 0.12029896220127582,
+      "grad_norm": 5.962332071996925,
+      "learning_rate": 2e-05,
+      "loss": 1.1241,
+      "step": 5054
+    },
+    {
+      "epoch": 0.12032276492430734,
+      "grad_norm": 2.862106999169578,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 5055
+    },
+    {
+      "epoch": 0.12034656764733885,
+      "grad_norm": 8.486460513821587,
+      "learning_rate": 2e-05,
+      "loss": 0.8757,
+      "step": 5056
+    },
+    {
+      "epoch": 0.12037037037037036,
+      "grad_norm": 2.5604630688832812,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 5057
+    },
+    {
+      "epoch": 0.12039417309340189,
+      "grad_norm": 5.232037422238777,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 5058
+    },
+    {
+      "epoch": 0.1204179758164334,
+      "grad_norm": 5.519790234832103,
+      "learning_rate": 2e-05,
+      "loss": 0.2518,
+      "step": 5059
+    },
+    {
+      "epoch": 0.12044177853946492,
+      "grad_norm": 7.750636938238305,
+      "learning_rate": 2e-05,
+      "loss": 1.2746,
+      "step": 5060
+    },
+    {
+      "epoch": 0.12046558126249643,
+      "grad_norm": 6.744705910271761,
+      "learning_rate": 2e-05,
+      "loss": 0.4688,
+      "step": 5061
+    },
+    {
+      "epoch": 0.12048938398552794,
+      "grad_norm": 4.504379657185391,
+      "learning_rate": 2e-05,
+      "loss": 0.6912,
+      "step": 5062
+    },
+    {
+      "epoch": 0.12051318670855946,
+      "grad_norm": 3.564841051121886,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 5063
+    },
+    {
+      "epoch": 0.12053698943159097,
+      "grad_norm": 2.0046881177750846,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 5064
+    },
+    {
+      "epoch": 0.1205607921546225,
+      "grad_norm": 1.650219217327689,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 5065
+    },
+    {
+      "epoch": 0.12058459487765401,
+      "grad_norm": 5.2763966301355,
+      "learning_rate": 2e-05,
+      "loss": 0.2696,
+      "step": 5066
+    },
+    {
+      "epoch": 0.12060839760068552,
+      "grad_norm": 6.4362353614703585,
+      "learning_rate": 2e-05,
+      "loss": 0.2321,
+      "step": 5067
+    },
+    {
+      "epoch": 0.12063220032371703,
+      "grad_norm": 2.346187835225107,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 5068
+    },
+    {
+      "epoch": 0.12065600304674855,
+      "grad_norm": 6.33440650807327,
+      "learning_rate": 2e-05,
+      "loss": 0.2855,
+      "step": 5069
+    },
+    {
+      "epoch": 0.12067980576978006,
+      "grad_norm": 6.325999150697398,
+      "learning_rate": 2e-05,
+      "loss": 1.0602,
+      "step": 5070
+    },
+    {
+      "epoch": 0.12070360849281157,
+      "grad_norm": 5.308536374858947,
+      "learning_rate": 2e-05,
+      "loss": 0.6821,
+      "step": 5071
+    },
+    {
+      "epoch": 0.12072741121584309,
+      "grad_norm": 8.579517568969663,
+      "learning_rate": 2e-05,
+      "loss": 1.8527,
+      "step": 5072
+    },
+    {
+      "epoch": 0.12075121393887461,
+      "grad_norm": 6.380817327843093,
+      "learning_rate": 2e-05,
+      "loss": 1.2804,
+      "step": 5073
+    },
+    {
+      "epoch": 0.12077501666190613,
+      "grad_norm": 5.4940327625273575,
+      "learning_rate": 2e-05,
+      "loss": 0.8265,
+      "step": 5074
+    },
+    {
+      "epoch": 0.12079881938493764,
+      "grad_norm": 5.8472217683289704,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 5075
+    },
+    {
+      "epoch": 0.12082262210796915,
+      "grad_norm": 5.92073706179796,
+      "learning_rate": 2e-05,
+      "loss": 1.2919,
+      "step": 5076
+    },
+    {
+      "epoch": 0.12084642483100067,
+      "grad_norm": 2.4040299323436463,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 5077
+    },
+    {
+      "epoch": 0.12087022755403218,
+      "grad_norm": 7.0721750109965935,
+      "learning_rate": 2e-05,
+      "loss": 0.643,
+      "step": 5078
+    },
+    {
+      "epoch": 0.12089403027706369,
+      "grad_norm": 9.885866906980024,
+      "learning_rate": 2e-05,
+      "loss": 0.7522,
+      "step": 5079
+    },
+    {
+      "epoch": 0.1209178330000952,
+      "grad_norm": 2.901290724169067,
+      "learning_rate": 2e-05,
+      "loss": 0.201,
+      "step": 5080
+    },
+    {
+      "epoch": 0.12094163572312673,
+      "grad_norm": 9.715876698140056,
+      "learning_rate": 2e-05,
+      "loss": 0.4932,
+      "step": 5081
+    },
+    {
+      "epoch": 0.12096543844615824,
+      "grad_norm": 6.598504260557666,
+      "learning_rate": 2e-05,
+      "loss": 1.0775,
+      "step": 5082
+    },
+    {
+      "epoch": 0.12098924116918976,
+      "grad_norm": 5.443827349058521,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 5083
+    },
+    {
+      "epoch": 0.12101304389222127,
+      "grad_norm": 9.12400523889355,
+      "learning_rate": 2e-05,
+      "loss": 0.7224,
+      "step": 5084
+    },
+    {
+      "epoch": 0.12103684661525278,
+      "grad_norm": 26.586659986514455,
+      "learning_rate": 2e-05,
+      "loss": 0.3562,
+      "step": 5085
+    },
+    {
+      "epoch": 0.1210606493382843,
+      "grad_norm": 1.6683610998103577,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 5086
+    },
+    {
+      "epoch": 0.12108445206131581,
+      "grad_norm": 1.065272586001698,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 5087
+    },
+    {
+      "epoch": 0.12110825478434734,
+      "grad_norm": 4.445168119600721,
+      "learning_rate": 2e-05,
+      "loss": 0.2353,
+      "step": 5088
+    },
+    {
+      "epoch": 0.12113205750737885,
+      "grad_norm": 7.121955274659212,
+      "learning_rate": 2e-05,
+      "loss": 0.8019,
+      "step": 5089
+    },
+    {
+      "epoch": 0.12115586023041036,
+      "grad_norm": 9.564135125470521,
+      "learning_rate": 2e-05,
+      "loss": 0.5681,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12117966295344187,
+      "grad_norm": 5.062973974193765,
+      "learning_rate": 2e-05,
+      "loss": 0.3063,
+      "step": 5091
+    },
+    {
+      "epoch": 0.12120346567647339,
+      "grad_norm": 3.6308707916592815,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 5092
+    },
+    {
+      "epoch": 0.1212272683995049,
+      "grad_norm": 4.35440912876298,
+      "learning_rate": 2e-05,
+      "loss": 0.8325,
+      "step": 5093
+    },
+    {
+      "epoch": 0.12125107112253641,
+      "grad_norm": 6.131883526187094,
+      "learning_rate": 2e-05,
+      "loss": 0.1771,
+      "step": 5094
+    },
+    {
+      "epoch": 0.12127487384556793,
+      "grad_norm": 7.780990108909742,
+      "learning_rate": 2e-05,
+      "loss": 0.8638,
+      "step": 5095
+    },
+    {
+      "epoch": 0.12129867656859945,
+      "grad_norm": 2.997740133657004,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 5096
+    },
+    {
+      "epoch": 0.12132247929163097,
+      "grad_norm": 4.109989012950907,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 5097
+    },
+    {
+      "epoch": 0.12134628201466248,
+      "grad_norm": 15.187561263766861,
+      "learning_rate": 2e-05,
+      "loss": 1.3504,
+      "step": 5098
+    },
+    {
+      "epoch": 0.12137008473769399,
+      "grad_norm": 8.528274315838564,
+      "learning_rate": 2e-05,
+      "loss": 0.8861,
+      "step": 5099
+    },
+    {
+      "epoch": 0.1213938874607255,
+      "grad_norm": 3.6194284736739406,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12141769018375702,
+      "grad_norm": 8.463357175067268,
+      "learning_rate": 2e-05,
+      "loss": 0.3756,
+      "step": 5101
+    },
+    {
+      "epoch": 0.12144149290678853,
+      "grad_norm": 3.132404068736148,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 5102
+    },
+    {
+      "epoch": 0.12146529562982006,
+      "grad_norm": 7.767023742244202,
+      "learning_rate": 2e-05,
+      "loss": 0.3905,
+      "step": 5103
+    },
+    {
+      "epoch": 0.12148909835285157,
+      "grad_norm": 9.743616719446436,
+      "learning_rate": 2e-05,
+      "loss": 1.1237,
+      "step": 5104
+    },
+    {
+      "epoch": 0.12151290107588308,
+      "grad_norm": 5.483866348522601,
+      "learning_rate": 2e-05,
+      "loss": 0.7414,
+      "step": 5105
+    },
+    {
+      "epoch": 0.1215367037989146,
+      "grad_norm": 4.845519655873116,
+      "learning_rate": 2e-05,
+      "loss": 0.5602,
+      "step": 5106
+    },
+    {
+      "epoch": 0.12156050652194611,
+      "grad_norm": 1.977853542234534,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 5107
+    },
+    {
+      "epoch": 0.12158430924497762,
+      "grad_norm": 7.536323043769548,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 5108
+    },
+    {
+      "epoch": 0.12160811196800914,
+      "grad_norm": 7.8710615957658705,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 5109
+    },
+    {
+      "epoch": 0.12163191469104065,
+      "grad_norm": 6.960032981791569,
+      "learning_rate": 2e-05,
+      "loss": 0.973,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12165571741407218,
+      "grad_norm": 10.211705139532903,
+      "learning_rate": 2e-05,
+      "loss": 1.4154,
+      "step": 5111
+    },
+    {
+      "epoch": 0.12167952013710369,
+      "grad_norm": 7.053541678476359,
+      "learning_rate": 2e-05,
+      "loss": 1.2288,
+      "step": 5112
+    },
+    {
+      "epoch": 0.1217033228601352,
+      "grad_norm": 7.793050461619286,
+      "learning_rate": 2e-05,
+      "loss": 0.6624,
+      "step": 5113
+    },
+    {
+      "epoch": 0.12172712558316671,
+      "grad_norm": 11.903325955100193,
+      "learning_rate": 2e-05,
+      "loss": 0.7264,
+      "step": 5114
+    },
+    {
+      "epoch": 0.12175092830619823,
+      "grad_norm": 8.214657349359513,
+      "learning_rate": 2e-05,
+      "loss": 0.2997,
+      "step": 5115
+    },
+    {
+      "epoch": 0.12177473102922974,
+      "grad_norm": 5.838543106912282,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 5116
+    },
+    {
+      "epoch": 0.12179853375226125,
+      "grad_norm": 19.026798943371382,
+      "learning_rate": 2e-05,
+      "loss": 1.2515,
+      "step": 5117
+    },
+    {
+      "epoch": 0.12182233647529277,
+      "grad_norm": 3.6064087918397463,
+      "learning_rate": 2e-05,
+      "loss": 0.1917,
+      "step": 5118
+    },
+    {
+      "epoch": 0.1218461391983243,
+      "grad_norm": 2.702913971201377,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 5119
+    },
+    {
+      "epoch": 0.1218699419213558,
+      "grad_norm": 2.008618333999127,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12189374464438732,
+      "grad_norm": 8.998056726900149,
+      "learning_rate": 2e-05,
+      "loss": 1.0418,
+      "step": 5121
+    },
+    {
+      "epoch": 0.12191754736741883,
+      "grad_norm": 9.12447032507478,
+      "learning_rate": 2e-05,
+      "loss": 1.0186,
+      "step": 5122
+    },
+    {
+      "epoch": 0.12194135009045035,
+      "grad_norm": 6.814627126128984,
+      "learning_rate": 2e-05,
+      "loss": 0.5999,
+      "step": 5123
+    },
+    {
+      "epoch": 0.12196515281348186,
+      "grad_norm": 7.681145756860988,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 5124
+    },
+    {
+      "epoch": 0.12198895553651337,
+      "grad_norm": 25.25866029745053,
+      "learning_rate": 2e-05,
+      "loss": 0.4939,
+      "step": 5125
+    },
+    {
+      "epoch": 0.1220127582595449,
+      "grad_norm": 2.8146282465651122,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 5126
+    },
+    {
+      "epoch": 0.12203656098257641,
+      "grad_norm": 7.909682347015563,
+      "learning_rate": 2e-05,
+      "loss": 1.5142,
+      "step": 5127
+    },
+    {
+      "epoch": 0.12206036370560792,
+      "grad_norm": 1.753322632655926,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 5128
+    },
+    {
+      "epoch": 0.12208416642863944,
+      "grad_norm": 6.90996572774457,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 5129
+    },
+    {
+      "epoch": 0.12210796915167095,
+      "grad_norm": 7.209259657649389,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12213177187470246,
+      "grad_norm": 2.0065394842614688,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 5131
+    },
+    {
+      "epoch": 0.12215557459773398,
+      "grad_norm": 9.361343347799313,
+      "learning_rate": 2e-05,
+      "loss": 0.9412,
+      "step": 5132
+    },
+    {
+      "epoch": 0.12217937732076549,
+      "grad_norm": 3.7336558545790353,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 5133
+    },
+    {
+      "epoch": 0.12220318004379702,
+      "grad_norm": 5.720509435363221,
+      "learning_rate": 2e-05,
+      "loss": 0.9854,
+      "step": 5134
+    },
+    {
+      "epoch": 0.12222698276682853,
+      "grad_norm": 2.7408437126334806,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 5135
+    },
+    {
+      "epoch": 0.12225078548986004,
+      "grad_norm": 7.611180216708235,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 5136
+    },
+    {
+      "epoch": 0.12227458821289156,
+      "grad_norm": 4.96536911266867,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 5137
+    },
+    {
+      "epoch": 0.12229839093592307,
+      "grad_norm": 3.311515455421506,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 5138
+    },
+    {
+      "epoch": 0.12232219365895458,
+      "grad_norm": 3.929520971218628,
+      "learning_rate": 2e-05,
+      "loss": 0.2743,
+      "step": 5139
+    },
+    {
+      "epoch": 0.1223459963819861,
+      "grad_norm": 7.912748616952659,
+      "learning_rate": 2e-05,
+      "loss": 1.0811,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12236979910501762,
+      "grad_norm": 7.1578602241102605,
+      "learning_rate": 2e-05,
+      "loss": 0.3245,
+      "step": 5141
+    },
+    {
+      "epoch": 0.12239360182804913,
+      "grad_norm": 6.859687706277017,
+      "learning_rate": 2e-05,
+      "loss": 0.381,
+      "step": 5142
+    },
+    {
+      "epoch": 0.12241740455108065,
+      "grad_norm": 7.174800335794734,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 5143
+    },
+    {
+      "epoch": 0.12244120727411216,
+      "grad_norm": 6.037405797031674,
+      "learning_rate": 2e-05,
+      "loss": 0.2829,
+      "step": 5144
+    },
+    {
+      "epoch": 0.12246500999714367,
+      "grad_norm": 8.228014056325103,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 5145
+    },
+    {
+      "epoch": 0.12248881272017519,
+      "grad_norm": 2.9210286315080887,
+      "learning_rate": 2e-05,
+      "loss": 0.224,
+      "step": 5146
+    },
+    {
+      "epoch": 0.1225126154432067,
+      "grad_norm": 2.891676541377833,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 5147
+    },
+    {
+      "epoch": 0.12253641816623821,
+      "grad_norm": 2.456811736280042,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 5148
+    },
+    {
+      "epoch": 0.12256022088926974,
+      "grad_norm": 6.18271981358013,
+      "learning_rate": 2e-05,
+      "loss": 1.2342,
+      "step": 5149
+    },
+    {
+      "epoch": 0.12258402361230125,
+      "grad_norm": 7.38076806491381,
+      "learning_rate": 2e-05,
+      "loss": 1.4041,
+      "step": 5150
+    },
+    {
+      "epoch": 0.12260782633533276,
+      "grad_norm": 7.964599327158505,
+      "learning_rate": 2e-05,
+      "loss": 0.692,
+      "step": 5151
+    },
+    {
+      "epoch": 0.12263162905836428,
+      "grad_norm": 3.376838242732902,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 5152
+    },
+    {
+      "epoch": 0.12265543178139579,
+      "grad_norm": 7.258568123102163,
+      "learning_rate": 2e-05,
+      "loss": 0.7262,
+      "step": 5153
+    },
+    {
+      "epoch": 0.1226792345044273,
+      "grad_norm": 4.902080600367208,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 5154
+    },
+    {
+      "epoch": 0.12270303722745882,
+      "grad_norm": 7.504214347603442,
+      "learning_rate": 2e-05,
+      "loss": 0.6367,
+      "step": 5155
+    },
+    {
+      "epoch": 0.12272683995049033,
+      "grad_norm": 2.4940877162371917,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 5156
+    },
+    {
+      "epoch": 0.12275064267352186,
+      "grad_norm": 2.2358256085906167,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 5157
+    },
+    {
+      "epoch": 0.12277444539655337,
+      "grad_norm": 7.598986356971546,
+      "learning_rate": 2e-05,
+      "loss": 1.144,
+      "step": 5158
+    },
+    {
+      "epoch": 0.12279824811958488,
+      "grad_norm": 4.646771297668541,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 5159
+    },
+    {
+      "epoch": 0.1228220508426164,
+      "grad_norm": 7.43562960744397,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12284585356564791,
+      "grad_norm": 5.986290043814575,
+      "learning_rate": 2e-05,
+      "loss": 0.9246,
+      "step": 5161
+    },
+    {
+      "epoch": 0.12286965628867942,
+      "grad_norm": 7.511228774279981,
+      "learning_rate": 2e-05,
+      "loss": 1.0066,
+      "step": 5162
+    },
+    {
+      "epoch": 0.12289345901171093,
+      "grad_norm": 7.576160462282314,
+      "learning_rate": 2e-05,
+      "loss": 0.5498,
+      "step": 5163
+    },
+    {
+      "epoch": 0.12291726173474246,
+      "grad_norm": 12.933409122636617,
+      "learning_rate": 2e-05,
+      "loss": 0.6924,
+      "step": 5164
+    },
+    {
+      "epoch": 0.12294106445777397,
+      "grad_norm": 4.448022829697306,
+      "learning_rate": 2e-05,
+      "loss": 0.7571,
+      "step": 5165
+    },
+    {
+      "epoch": 0.12296486718080549,
+      "grad_norm": 4.7397133242007214,
+      "learning_rate": 2e-05,
+      "loss": 1.0697,
+      "step": 5166
+    },
+    {
+      "epoch": 0.122988669903837,
+      "grad_norm": 1.9796747895265303,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 5167
+    },
+    {
+      "epoch": 0.12301247262686851,
+      "grad_norm": 6.152162901963146,
+      "learning_rate": 2e-05,
+      "loss": 0.3772,
+      "step": 5168
+    },
+    {
+      "epoch": 0.12303627534990003,
+      "grad_norm": 2.4228234561147426,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 5169
+    },
+    {
+      "epoch": 0.12306007807293154,
+      "grad_norm": 8.592271636610871,
+      "learning_rate": 2e-05,
+      "loss": 1.0862,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12308388079596305,
+      "grad_norm": 6.30412863001888,
+      "learning_rate": 2e-05,
+      "loss": 1.3042,
+      "step": 5171
+    },
+    {
+      "epoch": 0.12310768351899458,
+      "grad_norm": 3.9304208095198208,
+      "learning_rate": 2e-05,
+      "loss": 0.3188,
+      "step": 5172
+    },
+    {
+      "epoch": 0.12313148624202609,
+      "grad_norm": 3.1171211313828664,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 5173
+    },
+    {
+      "epoch": 0.1231552889650576,
+      "grad_norm": 6.548768355034782,
+      "learning_rate": 2e-05,
+      "loss": 1.2775,
+      "step": 5174
+    },
+    {
+      "epoch": 0.12317909168808912,
+      "grad_norm": 6.328724170983738,
+      "learning_rate": 2e-05,
+      "loss": 0.8824,
+      "step": 5175
+    },
+    {
+      "epoch": 0.12320289441112063,
+      "grad_norm": 1.3447282221179264,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 5176
+    },
+    {
+      "epoch": 0.12322669713415214,
+      "grad_norm": 6.2474744389569254,
+      "learning_rate": 2e-05,
+      "loss": 0.3365,
+      "step": 5177
+    },
+    {
+      "epoch": 0.12325049985718366,
+      "grad_norm": 2.3231491311629284,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 5178
+    },
+    {
+      "epoch": 0.12327430258021517,
+      "grad_norm": 6.602209528535699,
+      "learning_rate": 2e-05,
+      "loss": 0.2541,
+      "step": 5179
+    },
+    {
+      "epoch": 0.1232981053032467,
+      "grad_norm": 4.423338223392859,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 5180
+    },
+    {
+      "epoch": 0.12332190802627821,
+      "grad_norm": 1.668234338567132,
+      "learning_rate": 2e-05,
+      "loss": 0.1387,
+      "step": 5181
+    },
+    {
+      "epoch": 0.12334571074930972,
+      "grad_norm": 2.805642664672162,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 5182
+    },
+    {
+      "epoch": 0.12336951347234124,
+      "grad_norm": 5.616860633396894,
+      "learning_rate": 2e-05,
+      "loss": 0.7184,
+      "step": 5183
+    },
+    {
+      "epoch": 0.12339331619537275,
+      "grad_norm": 5.887176773110988,
+      "learning_rate": 2e-05,
+      "loss": 1.2043,
+      "step": 5184
+    },
+    {
+      "epoch": 0.12341711891840426,
+      "grad_norm": 5.982926732827432,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 5185
+    },
+    {
+      "epoch": 0.12344092164143577,
+      "grad_norm": 4.899498940585507,
+      "learning_rate": 2e-05,
+      "loss": 0.8789,
+      "step": 5186
+    },
+    {
+      "epoch": 0.1234647243644673,
+      "grad_norm": 1.654640204613121,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 5187
+    },
+    {
+      "epoch": 0.12348852708749881,
+      "grad_norm": 6.085395519211326,
+      "learning_rate": 2e-05,
+      "loss": 1.1505,
+      "step": 5188
+    },
+    {
+      "epoch": 0.12351232981053033,
+      "grad_norm": 9.487168740416926,
+      "learning_rate": 2e-05,
+      "loss": 1.9012,
+      "step": 5189
+    },
+    {
+      "epoch": 0.12353613253356184,
+      "grad_norm": 5.151976771139367,
+      "learning_rate": 2e-05,
+      "loss": 0.3114,
+      "step": 5190
+    },
+    {
+      "epoch": 0.12355993525659335,
+      "grad_norm": 2.0313394374618765,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 5191
+    },
+    {
+      "epoch": 0.12358373797962487,
+      "grad_norm": 6.581882868361508,
+      "learning_rate": 2e-05,
+      "loss": 0.3424,
+      "step": 5192
+    },
+    {
+      "epoch": 0.12360754070265638,
+      "grad_norm": 6.8102133937251885,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 5193
+    },
+    {
+      "epoch": 0.12363134342568789,
+      "grad_norm": 5.5491425687025275,
+      "learning_rate": 2e-05,
+      "loss": 0.2799,
+      "step": 5194
+    },
+    {
+      "epoch": 0.12365514614871942,
+      "grad_norm": 4.091839532534743,
+      "learning_rate": 2e-05,
+      "loss": 0.3767,
+      "step": 5195
+    },
+    {
+      "epoch": 0.12367894887175093,
+      "grad_norm": 5.75538379095756,
+      "learning_rate": 2e-05,
+      "loss": 0.7394,
+      "step": 5196
+    },
+    {
+      "epoch": 0.12370275159478245,
+      "grad_norm": 5.391548008121377,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 5197
+    },
+    {
+      "epoch": 0.12372655431781396,
+      "grad_norm": 8.75396264618813,
+      "learning_rate": 2e-05,
+      "loss": 0.7302,
+      "step": 5198
+    },
+    {
+      "epoch": 0.12375035704084547,
+      "grad_norm": 7.8536599176101705,
+      "learning_rate": 2e-05,
+      "loss": 0.9311,
+      "step": 5199
+    },
+    {
+      "epoch": 0.12377415976387698,
+      "grad_norm": 2.1149674392745004,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1237979624869085,
+      "grad_norm": 8.870399522514937,
+      "learning_rate": 2e-05,
+      "loss": 1.1997,
+      "step": 5201
+    },
+    {
+      "epoch": 0.12382176520994002,
+      "grad_norm": 2.178490914191959,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 5202
+    },
+    {
+      "epoch": 0.12384556793297154,
+      "grad_norm": 1.2845415745913051,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 5203
+    },
+    {
+      "epoch": 0.12386937065600305,
+      "grad_norm": 4.831348699878771,
+      "learning_rate": 2e-05,
+      "loss": 0.8095,
+      "step": 5204
+    },
+    {
+      "epoch": 0.12389317337903456,
+      "grad_norm": 7.669218136239507,
+      "learning_rate": 2e-05,
+      "loss": 0.448,
+      "step": 5205
+    },
+    {
+      "epoch": 0.12391697610206608,
+      "grad_norm": 8.838690653682539,
+      "learning_rate": 2e-05,
+      "loss": 1.792,
+      "step": 5206
+    },
+    {
+      "epoch": 0.12394077882509759,
+      "grad_norm": 8.0412312651784,
+      "learning_rate": 2e-05,
+      "loss": 0.7437,
+      "step": 5207
+    },
+    {
+      "epoch": 0.1239645815481291,
+      "grad_norm": 6.862873797953978,
+      "learning_rate": 2e-05,
+      "loss": 0.3716,
+      "step": 5208
+    },
+    {
+      "epoch": 0.12398838427116061,
+      "grad_norm": 5.1751685391087126,
+      "learning_rate": 2e-05,
+      "loss": 0.358,
+      "step": 5209
+    },
+    {
+      "epoch": 0.12401218699419214,
+      "grad_norm": 5.941052671853903,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12403598971722365,
+      "grad_norm": 6.124190478757026,
+      "learning_rate": 2e-05,
+      "loss": 1.0284,
+      "step": 5211
+    },
+    {
+      "epoch": 0.12405979244025517,
+      "grad_norm": 5.126106677503554,
+      "learning_rate": 2e-05,
+      "loss": 0.5652,
+      "step": 5212
+    },
+    {
+      "epoch": 0.12408359516328668,
+      "grad_norm": 5.12480532650414,
+      "learning_rate": 2e-05,
+      "loss": 0.3222,
+      "step": 5213
+    },
+    {
+      "epoch": 0.1241073978863182,
+      "grad_norm": 7.381207743644249,
+      "learning_rate": 2e-05,
+      "loss": 1.2547,
+      "step": 5214
+    },
+    {
+      "epoch": 0.1241312006093497,
+      "grad_norm": 3.429439638884868,
+      "learning_rate": 2e-05,
+      "loss": 0.2391,
+      "step": 5215
+    },
+    {
+      "epoch": 0.12415500333238122,
+      "grad_norm": 8.747802155488735,
+      "learning_rate": 2e-05,
+      "loss": 1.2519,
+      "step": 5216
+    },
+    {
+      "epoch": 0.12417880605541273,
+      "grad_norm": 6.558600187701124,
+      "learning_rate": 2e-05,
+      "loss": 1.1452,
+      "step": 5217
+    },
+    {
+      "epoch": 0.12420260877844426,
+      "grad_norm": 6.496907758319055,
+      "learning_rate": 2e-05,
+      "loss": 0.2992,
+      "step": 5218
+    },
+    {
+      "epoch": 0.12422641150147577,
+      "grad_norm": 2.6677091163564346,
+      "learning_rate": 2e-05,
+      "loss": 0.1953,
+      "step": 5219
+    },
+    {
+      "epoch": 0.12425021422450729,
+      "grad_norm": 6.368738650228823,
+      "learning_rate": 2e-05,
+      "loss": 1.0911,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1242740169475388,
+      "grad_norm": 8.085625369558421,
+      "learning_rate": 2e-05,
+      "loss": 0.6724,
+      "step": 5221
+    },
+    {
+      "epoch": 0.12429781967057031,
+      "grad_norm": 7.79064794407653,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 5222
+    },
+    {
+      "epoch": 0.12432162239360182,
+      "grad_norm": 3.5842963074751273,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 5223
+    },
+    {
+      "epoch": 0.12434542511663334,
+      "grad_norm": 2.604644464211887,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 5224
+    },
+    {
+      "epoch": 0.12436922783966486,
+      "grad_norm": 15.511082895945655,
+      "learning_rate": 2e-05,
+      "loss": 1.4613,
+      "step": 5225
+    },
+    {
+      "epoch": 0.12439303056269638,
+      "grad_norm": 2.615212917441948,
+      "learning_rate": 2e-05,
+      "loss": 0.2401,
+      "step": 5226
+    },
+    {
+      "epoch": 0.12441683328572789,
+      "grad_norm": 4.793103135814819,
+      "learning_rate": 2e-05,
+      "loss": 0.7349,
+      "step": 5227
+    },
+    {
+      "epoch": 0.1244406360087594,
+      "grad_norm": 3.222677976840878,
+      "learning_rate": 2e-05,
+      "loss": 0.2035,
+      "step": 5228
+    },
+    {
+      "epoch": 0.12446443873179092,
+      "grad_norm": 8.846664692636951,
+      "learning_rate": 2e-05,
+      "loss": 0.9582,
+      "step": 5229
+    },
+    {
+      "epoch": 0.12448824145482243,
+      "grad_norm": 2.8229538314479443,
+      "learning_rate": 2e-05,
+      "loss": 0.1726,
+      "step": 5230
+    },
+    {
+      "epoch": 0.12451204417785394,
+      "grad_norm": 4.041900471175845,
+      "learning_rate": 2e-05,
+      "loss": 0.894,
+      "step": 5231
+    },
+    {
+      "epoch": 0.12453584690088546,
+      "grad_norm": 7.028816002798233,
+      "learning_rate": 2e-05,
+      "loss": 0.3452,
+      "step": 5232
+    },
+    {
+      "epoch": 0.12455964962391698,
+      "grad_norm": 2.613339635615375,
+      "learning_rate": 2e-05,
+      "loss": 0.1293,
+      "step": 5233
+    },
+    {
+      "epoch": 0.1245834523469485,
+      "grad_norm": 4.106934852418466,
+      "learning_rate": 2e-05,
+      "loss": 0.6435,
+      "step": 5234
+    },
+    {
+      "epoch": 0.12460725506998001,
+      "grad_norm": 9.816647362473688,
+      "learning_rate": 2e-05,
+      "loss": 1.2253,
+      "step": 5235
+    },
+    {
+      "epoch": 0.12463105779301152,
+      "grad_norm": 4.401934318856212,
+      "learning_rate": 2e-05,
+      "loss": 0.3321,
+      "step": 5236
+    },
+    {
+      "epoch": 0.12465486051604303,
+      "grad_norm": 4.230397655979488,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 5237
+    },
+    {
+      "epoch": 0.12467866323907455,
+      "grad_norm": 5.96654836352007,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 5238
+    },
+    {
+      "epoch": 0.12470246596210606,
+      "grad_norm": 2.1458627603582157,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 5239
+    },
+    {
+      "epoch": 0.12472626868513757,
+      "grad_norm": 5.671626708060972,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 5240
+    },
+    {
+      "epoch": 0.1247500714081691,
+      "grad_norm": 3.2335243059378826,
+      "learning_rate": 2e-05,
+      "loss": 0.1602,
+      "step": 5241
+    },
+    {
+      "epoch": 0.12477387413120061,
+      "grad_norm": 5.8851286699644305,
+      "learning_rate": 2e-05,
+      "loss": 0.2689,
+      "step": 5242
+    },
+    {
+      "epoch": 0.12479767685423213,
+      "grad_norm": 6.829663439568497,
+      "learning_rate": 2e-05,
+      "loss": 0.609,
+      "step": 5243
+    },
+    {
+      "epoch": 0.12482147957726364,
+      "grad_norm": 5.167513505453573,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 5244
+    },
+    {
+      "epoch": 0.12484528230029515,
+      "grad_norm": 4.524976238726525,
+      "learning_rate": 2e-05,
+      "loss": 0.4363,
+      "step": 5245
+    },
+    {
+      "epoch": 0.12486908502332666,
+      "grad_norm": 4.002638402354587,
+      "learning_rate": 2e-05,
+      "loss": 0.4075,
+      "step": 5246
+    },
+    {
+      "epoch": 0.12489288774635818,
+      "grad_norm": 3.5498832266822706,
+      "learning_rate": 2e-05,
+      "loss": 0.234,
+      "step": 5247
+    },
+    {
+      "epoch": 0.1249166904693897,
+      "grad_norm": 3.1028164813230985,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 5248
+    },
+    {
+      "epoch": 0.12494049319242122,
+      "grad_norm": 8.434470656424217,
+      "learning_rate": 2e-05,
+      "loss": 0.367,
+      "step": 5249
+    },
+    {
+      "epoch": 0.12496429591545273,
+      "grad_norm": 2.5241555966466893,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12498809863848424,
+      "grad_norm": 3.6346769385617175,
+      "learning_rate": 2e-05,
+      "loss": 0.2385,
+      "step": 5251
+    },
+    {
+      "epoch": 0.12501190136151577,
+      "grad_norm": 7.868339442594795,
+      "learning_rate": 2e-05,
+      "loss": 0.8543,
+      "step": 5252
+    },
+    {
+      "epoch": 0.12503570408454728,
+      "grad_norm": 3.942946220759021,
+      "learning_rate": 2e-05,
+      "loss": 0.2924,
+      "step": 5253
+    },
+    {
+      "epoch": 0.1250595068075788,
+      "grad_norm": 5.169407507074952,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 5254
+    },
+    {
+      "epoch": 0.1250833095306103,
+      "grad_norm": 9.91624636108852,
+      "learning_rate": 2e-05,
+      "loss": 0.8014,
+      "step": 5255
+    },
+    {
+      "epoch": 0.12510711225364182,
+      "grad_norm": 4.943275588534574,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 5256
+    },
+    {
+      "epoch": 0.12513091497667334,
+      "grad_norm": 6.216389407056161,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 5257
+    },
+    {
+      "epoch": 0.12515471769970485,
+      "grad_norm": 8.645939107067003,
+      "learning_rate": 2e-05,
+      "loss": 1.2614,
+      "step": 5258
+    },
+    {
+      "epoch": 0.12517852042273636,
+      "grad_norm": 7.541514880114062,
+      "learning_rate": 2e-05,
+      "loss": 2.0688,
+      "step": 5259
+    },
+    {
+      "epoch": 0.12520232314576787,
+      "grad_norm": 2.889132973222135,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1252261258687994,
+      "grad_norm": 1.979516790328747,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 5261
+    },
+    {
+      "epoch": 0.1252499285918309,
+      "grad_norm": 6.615731995476451,
+      "learning_rate": 2e-05,
+      "loss": 1.2937,
+      "step": 5262
+    },
+    {
+      "epoch": 0.1252737313148624,
+      "grad_norm": 2.1487193189254765,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 5263
+    },
+    {
+      "epoch": 0.12529753403789393,
+      "grad_norm": 9.496978653715182,
+      "learning_rate": 2e-05,
+      "loss": 0.7135,
+      "step": 5264
+    },
+    {
+      "epoch": 0.12532133676092544,
+      "grad_norm": 5.641356884443858,
+      "learning_rate": 2e-05,
+      "loss": 1.2454,
+      "step": 5265
+    },
+    {
+      "epoch": 0.12534513948395695,
+      "grad_norm": 4.54545719177127,
+      "learning_rate": 2e-05,
+      "loss": 0.3638,
+      "step": 5266
+    },
+    {
+      "epoch": 0.1253689422069885,
+      "grad_norm": 4.837595005763997,
+      "learning_rate": 2e-05,
+      "loss": 0.698,
+      "step": 5267
+    },
+    {
+      "epoch": 0.12539274493002,
+      "grad_norm": 2.938176363950165,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 5268
+    },
+    {
+      "epoch": 0.12541654765305152,
+      "grad_norm": 7.377448842330741,
+      "learning_rate": 2e-05,
+      "loss": 0.545,
+      "step": 5269
+    },
+    {
+      "epoch": 0.12544035037608303,
+      "grad_norm": 8.973825789422099,
+      "learning_rate": 2e-05,
+      "loss": 0.4665,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12546415309911454,
+      "grad_norm": 3.6819640543262047,
+      "learning_rate": 2e-05,
+      "loss": 0.5431,
+      "step": 5271
+    },
+    {
+      "epoch": 0.12548795582214606,
+      "grad_norm": 6.7296612299753065,
+      "learning_rate": 2e-05,
+      "loss": 0.2889,
+      "step": 5272
+    },
+    {
+      "epoch": 0.12551175854517757,
+      "grad_norm": 8.76240919393969,
+      "learning_rate": 2e-05,
+      "loss": 0.5205,
+      "step": 5273
+    },
+    {
+      "epoch": 0.12553556126820908,
+      "grad_norm": 3.2417386334781475,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 5274
+    },
+    {
+      "epoch": 0.1255593639912406,
+      "grad_norm": 3.4929433315662104,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 5275
+    },
+    {
+      "epoch": 0.1255831667142721,
+      "grad_norm": 3.35682868846361,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 5276
+    },
+    {
+      "epoch": 0.12560696943730362,
+      "grad_norm": 5.991668060662196,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 5277
+    },
+    {
+      "epoch": 0.12563077216033514,
+      "grad_norm": 6.401963117844133,
+      "learning_rate": 2e-05,
+      "loss": 0.4764,
+      "step": 5278
+    },
+    {
+      "epoch": 0.12565457488336665,
+      "grad_norm": 3.4025040576531898,
+      "learning_rate": 2e-05,
+      "loss": 0.5687,
+      "step": 5279
+    },
+    {
+      "epoch": 0.12567837760639816,
+      "grad_norm": 9.224941511750949,
+      "learning_rate": 2e-05,
+      "loss": 2.0714,
+      "step": 5280
+    },
+    {
+      "epoch": 0.12570218032942967,
+      "grad_norm": 3.6143418448703835,
+      "learning_rate": 2e-05,
+      "loss": 0.2425,
+      "step": 5281
+    },
+    {
+      "epoch": 0.12572598305246122,
+      "grad_norm": 3.6408268461068394,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 5282
+    },
+    {
+      "epoch": 0.12574978577549273,
+      "grad_norm": 6.407384152572705,
+      "learning_rate": 2e-05,
+      "loss": 0.2899,
+      "step": 5283
+    },
+    {
+      "epoch": 0.12577358849852424,
+      "grad_norm": 5.42690486714669,
+      "learning_rate": 2e-05,
+      "loss": 0.8159,
+      "step": 5284
+    },
+    {
+      "epoch": 0.12579739122155575,
+      "grad_norm": 9.257757061094178,
+      "learning_rate": 2e-05,
+      "loss": 1.0101,
+      "step": 5285
+    },
+    {
+      "epoch": 0.12582119394458727,
+      "grad_norm": 1.6429935766079407,
+      "learning_rate": 2e-05,
+      "loss": 0.1411,
+      "step": 5286
+    },
+    {
+      "epoch": 0.12584499666761878,
+      "grad_norm": 3.4007116905209775,
+      "learning_rate": 2e-05,
+      "loss": 0.3021,
+      "step": 5287
+    },
+    {
+      "epoch": 0.1258687993906503,
+      "grad_norm": 3.709201165742549,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 5288
+    },
+    {
+      "epoch": 0.1258926021136818,
+      "grad_norm": 5.7529047003313565,
+      "learning_rate": 2e-05,
+      "loss": 0.3018,
+      "step": 5289
+    },
+    {
+      "epoch": 0.12591640483671332,
+      "grad_norm": 3.1521118806211055,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12594020755974483,
+      "grad_norm": 8.899299922529453,
+      "learning_rate": 2e-05,
+      "loss": 0.4626,
+      "step": 5291
+    },
+    {
+      "epoch": 0.12596401028277635,
+      "grad_norm": 1.7334970524349649,
+      "learning_rate": 2e-05,
+      "loss": 0.133,
+      "step": 5292
+    },
+    {
+      "epoch": 0.12598781300580786,
+      "grad_norm": 1.7871899648664251,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 5293
+    },
+    {
+      "epoch": 0.12601161572883937,
+      "grad_norm": 8.391761188400803,
+      "learning_rate": 2e-05,
+      "loss": 0.4199,
+      "step": 5294
+    },
+    {
+      "epoch": 0.12603541845187088,
+      "grad_norm": 5.548241107312774,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 5295
+    },
+    {
+      "epoch": 0.1260592211749024,
+      "grad_norm": 1.887640179859381,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 5296
+    },
+    {
+      "epoch": 0.1260830238979339,
+      "grad_norm": 6.930473907642873,
+      "learning_rate": 2e-05,
+      "loss": 0.6269,
+      "step": 5297
+    },
+    {
+      "epoch": 0.12610682662096545,
+      "grad_norm": 11.14221356872348,
+      "learning_rate": 2e-05,
+      "loss": 1.1374,
+      "step": 5298
+    },
+    {
+      "epoch": 0.12613062934399696,
+      "grad_norm": 1.7165218454063853,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 5299
+    },
+    {
+      "epoch": 0.12615443206702848,
+      "grad_norm": 2.471515511504019,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12617823479006,
+      "grad_norm": 3.1581891272787415,
+      "learning_rate": 2e-05,
+      "loss": 0.4648,
+      "step": 5301
+    },
+    {
+      "epoch": 0.1262020375130915,
+      "grad_norm": 6.1169025385443945,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 5302
+    },
+    {
+      "epoch": 0.12622584023612302,
+      "grad_norm": 3.0034701664383525,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 5303
+    },
+    {
+      "epoch": 0.12624964295915453,
+      "grad_norm": 1.8654533468413357,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 5304
+    },
+    {
+      "epoch": 0.12627344568218604,
+      "grad_norm": 7.152649334047657,
+      "learning_rate": 2e-05,
+      "loss": 0.2868,
+      "step": 5305
+    },
+    {
+      "epoch": 0.12629724840521755,
+      "grad_norm": 5.431300770415294,
+      "learning_rate": 2e-05,
+      "loss": 0.2455,
+      "step": 5306
+    },
+    {
+      "epoch": 0.12632105112824907,
+      "grad_norm": 2.402010710995579,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 5307
+    },
+    {
+      "epoch": 0.12634485385128058,
+      "grad_norm": 6.242014131552194,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 5308
+    },
+    {
+      "epoch": 0.1263686565743121,
+      "grad_norm": 5.938719270105933,
+      "learning_rate": 2e-05,
+      "loss": 0.291,
+      "step": 5309
+    },
+    {
+      "epoch": 0.1263924592973436,
+      "grad_norm": 4.591445373423992,
+      "learning_rate": 2e-05,
+      "loss": 0.357,
+      "step": 5310
+    },
+    {
+      "epoch": 0.12641626202037512,
+      "grad_norm": 5.301277368900399,
+      "learning_rate": 2e-05,
+      "loss": 0.9136,
+      "step": 5311
+    },
+    {
+      "epoch": 0.12644006474340663,
+      "grad_norm": 3.4159642944261144,
+      "learning_rate": 2e-05,
+      "loss": 0.5014,
+      "step": 5312
+    },
+    {
+      "epoch": 0.12646386746643817,
+      "grad_norm": 4.978760102729938,
+      "learning_rate": 2e-05,
+      "loss": 0.4463,
+      "step": 5313
+    },
+    {
+      "epoch": 0.1264876701894697,
+      "grad_norm": 2.117893408685318,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 5314
+    },
+    {
+      "epoch": 0.1265114729125012,
+      "grad_norm": 3.534394481678272,
+      "learning_rate": 2e-05,
+      "loss": 0.6824,
+      "step": 5315
+    },
+    {
+      "epoch": 0.1265352756355327,
+      "grad_norm": 11.023912272185264,
+      "learning_rate": 2e-05,
+      "loss": 0.3656,
+      "step": 5316
+    },
+    {
+      "epoch": 0.12655907835856423,
+      "grad_norm": 6.4573453331096005,
+      "learning_rate": 2e-05,
+      "loss": 0.8017,
+      "step": 5317
+    },
+    {
+      "epoch": 0.12658288108159574,
+      "grad_norm": 3.219012067047224,
+      "learning_rate": 2e-05,
+      "loss": 0.7614,
+      "step": 5318
+    },
+    {
+      "epoch": 0.12660668380462725,
+      "grad_norm": 3.7993957941851577,
+      "learning_rate": 2e-05,
+      "loss": 0.7943,
+      "step": 5319
+    },
+    {
+      "epoch": 0.12663048652765876,
+      "grad_norm": 3.01079912331898,
+      "learning_rate": 2e-05,
+      "loss": 0.2389,
+      "step": 5320
+    },
+    {
+      "epoch": 0.12665428925069028,
+      "grad_norm": 1.523407283820195,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 5321
+    },
+    {
+      "epoch": 0.1266780919737218,
+      "grad_norm": 2.8191826520739984,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 5322
+    },
+    {
+      "epoch": 0.1267018946967533,
+      "grad_norm": 2.9936490326760894,
+      "learning_rate": 2e-05,
+      "loss": 0.2601,
+      "step": 5323
+    },
+    {
+      "epoch": 0.12672569741978482,
+      "grad_norm": 7.6582684209603835,
+      "learning_rate": 2e-05,
+      "loss": 0.3561,
+      "step": 5324
+    },
+    {
+      "epoch": 0.12674950014281633,
+      "grad_norm": 7.00734882319834,
+      "learning_rate": 2e-05,
+      "loss": 1.5078,
+      "step": 5325
+    },
+    {
+      "epoch": 0.12677330286584784,
+      "grad_norm": 7.120590481874692,
+      "learning_rate": 2e-05,
+      "loss": 0.4809,
+      "step": 5326
+    },
+    {
+      "epoch": 0.12679710558887936,
+      "grad_norm": 5.15503158936039,
+      "learning_rate": 2e-05,
+      "loss": 0.5346,
+      "step": 5327
+    },
+    {
+      "epoch": 0.1268209083119109,
+      "grad_norm": 7.850539357568919,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 5328
+    },
+    {
+      "epoch": 0.1268447110349424,
+      "grad_norm": 5.051114597942671,
+      "learning_rate": 2e-05,
+      "loss": 0.3229,
+      "step": 5329
+    },
+    {
+      "epoch": 0.12686851375797392,
+      "grad_norm": 7.3197633937595725,
+      "learning_rate": 2e-05,
+      "loss": 0.9259,
+      "step": 5330
+    },
+    {
+      "epoch": 0.12689231648100543,
+      "grad_norm": 4.4747221417692655,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 5331
+    },
+    {
+      "epoch": 0.12691611920403695,
+      "grad_norm": 7.905437253921641,
+      "learning_rate": 2e-05,
+      "loss": 0.6618,
+      "step": 5332
+    },
+    {
+      "epoch": 0.12693992192706846,
+      "grad_norm": 7.90717100376134,
+      "learning_rate": 2e-05,
+      "loss": 0.8599,
+      "step": 5333
+    },
+    {
+      "epoch": 0.12696372465009997,
+      "grad_norm": 5.567699995657911,
+      "learning_rate": 2e-05,
+      "loss": 0.8935,
+      "step": 5334
+    },
+    {
+      "epoch": 0.1269875273731315,
+      "grad_norm": 5.63472961799631,
+      "learning_rate": 2e-05,
+      "loss": 0.4288,
+      "step": 5335
+    },
+    {
+      "epoch": 0.127011330096163,
+      "grad_norm": 4.7264849947559595,
+      "learning_rate": 2e-05,
+      "loss": 0.2989,
+      "step": 5336
+    },
+    {
+      "epoch": 0.1270351328191945,
+      "grad_norm": 6.909923423896033,
+      "learning_rate": 2e-05,
+      "loss": 0.9156,
+      "step": 5337
+    },
+    {
+      "epoch": 0.12705893554222603,
+      "grad_norm": 1.2041151213157808,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 5338
+    },
+    {
+      "epoch": 0.12708273826525754,
+      "grad_norm": 9.968838584836682,
+      "learning_rate": 2e-05,
+      "loss": 0.5263,
+      "step": 5339
+    },
+    {
+      "epoch": 0.12710654098828905,
+      "grad_norm": 6.8247376998170335,
+      "learning_rate": 2e-05,
+      "loss": 1.083,
+      "step": 5340
+    },
+    {
+      "epoch": 0.12713034371132056,
+      "grad_norm": 1.4057209552389822,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 5341
+    },
+    {
+      "epoch": 0.12715414643435208,
+      "grad_norm": 27.31772326675147,
+      "learning_rate": 2e-05,
+      "loss": 0.6719,
+      "step": 5342
+    },
+    {
+      "epoch": 0.12717794915738362,
+      "grad_norm": 7.4276898499002195,
+      "learning_rate": 2e-05,
+      "loss": 0.9395,
+      "step": 5343
+    },
+    {
+      "epoch": 0.12720175188041513,
+      "grad_norm": 2.7889849961610165,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 5344
+    },
+    {
+      "epoch": 0.12722555460344664,
+      "grad_norm": 8.228007229165469,
+      "learning_rate": 2e-05,
+      "loss": 1.2363,
+      "step": 5345
+    },
+    {
+      "epoch": 0.12724935732647816,
+      "grad_norm": 6.76252532449662,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 5346
+    },
+    {
+      "epoch": 0.12727316004950967,
+      "grad_norm": 4.346526213951191,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 5347
+    },
+    {
+      "epoch": 0.12729696277254118,
+      "grad_norm": 6.003074147883696,
+      "learning_rate": 2e-05,
+      "loss": 0.287,
+      "step": 5348
+    },
+    {
+      "epoch": 0.1273207654955727,
+      "grad_norm": 7.287134546426803,
+      "learning_rate": 2e-05,
+      "loss": 1.4131,
+      "step": 5349
+    },
+    {
+      "epoch": 0.1273445682186042,
+      "grad_norm": 6.15675142592523,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 5350
+    },
+    {
+      "epoch": 0.12736837094163572,
+      "grad_norm": 7.1416312760500835,
+      "learning_rate": 2e-05,
+      "loss": 0.4347,
+      "step": 5351
+    },
+    {
+      "epoch": 0.12739217366466724,
+      "grad_norm": 3.911361595189692,
+      "learning_rate": 2e-05,
+      "loss": 0.4551,
+      "step": 5352
+    },
+    {
+      "epoch": 0.12741597638769875,
+      "grad_norm": 2.5137048538584015,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 5353
+    },
+    {
+      "epoch": 0.12743977911073026,
+      "grad_norm": 4.58050979561082,
+      "learning_rate": 2e-05,
+      "loss": 0.7825,
+      "step": 5354
+    },
+    {
+      "epoch": 0.12746358183376177,
+      "grad_norm": 6.534831840374162,
+      "learning_rate": 2e-05,
+      "loss": 0.3352,
+      "step": 5355
+    },
+    {
+      "epoch": 0.1274873845567933,
+      "grad_norm": 4.822438103598649,
+      "learning_rate": 2e-05,
+      "loss": 0.2683,
+      "step": 5356
+    },
+    {
+      "epoch": 0.1275111872798248,
+      "grad_norm": 2.5960372568303867,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 5357
+    },
+    {
+      "epoch": 0.1275349900028563,
+      "grad_norm": 2.8580067946886,
+      "learning_rate": 2e-05,
+      "loss": 0.2347,
+      "step": 5358
+    },
+    {
+      "epoch": 0.12755879272588785,
+      "grad_norm": 10.532794579144046,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 5359
+    },
+    {
+      "epoch": 0.12758259544891937,
+      "grad_norm": 6.85649838345079,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 5360
+    },
+    {
+      "epoch": 0.12760639817195088,
+      "grad_norm": 9.665924005460656,
+      "learning_rate": 2e-05,
+      "loss": 0.5101,
+      "step": 5361
+    },
+    {
+      "epoch": 0.1276302008949824,
+      "grad_norm": 5.72420290580316,
+      "learning_rate": 2e-05,
+      "loss": 0.2635,
+      "step": 5362
+    },
+    {
+      "epoch": 0.1276540036180139,
+      "grad_norm": 5.44091825684699,
+      "learning_rate": 2e-05,
+      "loss": 0.8912,
+      "step": 5363
+    },
+    {
+      "epoch": 0.12767780634104542,
+      "grad_norm": 11.318627948532102,
+      "learning_rate": 2e-05,
+      "loss": 0.8388,
+      "step": 5364
+    },
+    {
+      "epoch": 0.12770160906407693,
+      "grad_norm": 2.7463928124783403,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 5365
+    },
+    {
+      "epoch": 0.12772541178710844,
+      "grad_norm": 5.170336177058555,
+      "learning_rate": 2e-05,
+      "loss": 0.2205,
+      "step": 5366
+    },
+    {
+      "epoch": 0.12774921451013996,
+      "grad_norm": 7.027964346495052,
+      "learning_rate": 2e-05,
+      "loss": 1.2593,
+      "step": 5367
+    },
+    {
+      "epoch": 0.12777301723317147,
+      "grad_norm": 5.913140773628373,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 5368
+    },
+    {
+      "epoch": 0.12779681995620298,
+      "grad_norm": 1.978295537714415,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 5369
+    },
+    {
+      "epoch": 0.1278206226792345,
+      "grad_norm": 7.153592278627112,
+      "learning_rate": 2e-05,
+      "loss": 0.893,
+      "step": 5370
+    },
+    {
+      "epoch": 0.127844425402266,
+      "grad_norm": 3.506198354324732,
+      "learning_rate": 2e-05,
+      "loss": 0.4495,
+      "step": 5371
+    },
+    {
+      "epoch": 0.12786822812529752,
+      "grad_norm": 5.892730978632053,
+      "learning_rate": 2e-05,
+      "loss": 0.9659,
+      "step": 5372
+    },
+    {
+      "epoch": 0.12789203084832904,
+      "grad_norm": 7.047919501416969,
+      "learning_rate": 2e-05,
+      "loss": 1.0422,
+      "step": 5373
+    },
+    {
+      "epoch": 0.12791583357136058,
+      "grad_norm": 4.883854251350776,
+      "learning_rate": 2e-05,
+      "loss": 0.8549,
+      "step": 5374
+    },
+    {
+      "epoch": 0.1279396362943921,
+      "grad_norm": 9.481072002880468,
+      "learning_rate": 2e-05,
+      "loss": 1.1331,
+      "step": 5375
+    },
+    {
+      "epoch": 0.1279634390174236,
+      "grad_norm": 1.7013824861829219,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 5376
+    },
+    {
+      "epoch": 0.12798724174045512,
+      "grad_norm": 4.502234377207571,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 5377
+    },
+    {
+      "epoch": 0.12801104446348663,
+      "grad_norm": 7.166308038007169,
+      "learning_rate": 2e-05,
+      "loss": 1.5311,
+      "step": 5378
+    },
+    {
+      "epoch": 0.12803484718651814,
+      "grad_norm": 2.171057488709197,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 5379
+    },
+    {
+      "epoch": 0.12805864990954965,
+      "grad_norm": 6.09996767405627,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 5380
+    },
+    {
+      "epoch": 0.12808245263258117,
+      "grad_norm": 2.1386809042177117,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 5381
+    },
+    {
+      "epoch": 0.12810625535561268,
+      "grad_norm": 5.078352342201599,
+      "learning_rate": 2e-05,
+      "loss": 0.522,
+      "step": 5382
+    },
+    {
+      "epoch": 0.1281300580786442,
+      "grad_norm": 7.556013531468271,
+      "learning_rate": 2e-05,
+      "loss": 1.1363,
+      "step": 5383
+    },
+    {
+      "epoch": 0.1281538608016757,
+      "grad_norm": 6.700250166158824,
+      "learning_rate": 2e-05,
+      "loss": 0.3722,
+      "step": 5384
+    },
+    {
+      "epoch": 0.12817766352470722,
+      "grad_norm": 4.824228537778985,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 5385
+    },
+    {
+      "epoch": 0.12820146624773873,
+      "grad_norm": 9.425438357493679,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 5386
+    },
+    {
+      "epoch": 0.12822526897077025,
+      "grad_norm": 5.659318491998954,
+      "learning_rate": 2e-05,
+      "loss": 0.8642,
+      "step": 5387
+    },
+    {
+      "epoch": 0.12824907169380176,
+      "grad_norm": 6.137502207039746,
+      "learning_rate": 2e-05,
+      "loss": 0.2811,
+      "step": 5388
+    },
+    {
+      "epoch": 0.1282728744168333,
+      "grad_norm": 7.344868875708034,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 5389
+    },
+    {
+      "epoch": 0.1282966771398648,
+      "grad_norm": 3.2189394627330947,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 5390
+    },
+    {
+      "epoch": 0.12832047986289633,
+      "grad_norm": 4.284226827930189,
+      "learning_rate": 2e-05,
+      "loss": 0.3259,
+      "step": 5391
+    },
+    {
+      "epoch": 0.12834428258592784,
+      "grad_norm": 6.99791276318733,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 5392
+    },
+    {
+      "epoch": 0.12836808530895935,
+      "grad_norm": 2.2302493343981906,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 5393
+    },
+    {
+      "epoch": 0.12839188803199086,
+      "grad_norm": 3.4198045450459547,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 5394
+    },
+    {
+      "epoch": 0.12841569075502238,
+      "grad_norm": 3.0625548907453806,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 5395
+    },
+    {
+      "epoch": 0.1284394934780539,
+      "grad_norm": 3.8938062555222297,
+      "learning_rate": 2e-05,
+      "loss": 0.5987,
+      "step": 5396
+    },
+    {
+      "epoch": 0.1284632962010854,
+      "grad_norm": 9.796001583845488,
+      "learning_rate": 2e-05,
+      "loss": 0.9382,
+      "step": 5397
+    },
+    {
+      "epoch": 0.12848709892411692,
+      "grad_norm": 5.216470996174303,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 5398
+    },
+    {
+      "epoch": 0.12851090164714843,
+      "grad_norm": 3.8427895837418773,
+      "learning_rate": 2e-05,
+      "loss": 0.6724,
+      "step": 5399
+    },
+    {
+      "epoch": 0.12853470437017994,
+      "grad_norm": 7.124166488759014,
+      "learning_rate": 2e-05,
+      "loss": 0.509,
+      "step": 5400
+    },
+    {
+      "epoch": 0.12855850709321145,
+      "grad_norm": 3.265932067060808,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 5401
+    },
+    {
+      "epoch": 0.12858230981624297,
+      "grad_norm": 5.028986112891781,
+      "learning_rate": 2e-05,
+      "loss": 0.6154,
+      "step": 5402
+    },
+    {
+      "epoch": 0.12860611253927448,
+      "grad_norm": 4.490142772639719,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 5403
+    },
+    {
+      "epoch": 0.12862991526230602,
+      "grad_norm": 6.620645385746631,
+      "learning_rate": 2e-05,
+      "loss": 0.9558,
+      "step": 5404
+    },
+    {
+      "epoch": 0.12865371798533753,
+      "grad_norm": 7.780334664315145,
+      "learning_rate": 2e-05,
+      "loss": 0.7229,
+      "step": 5405
+    },
+    {
+      "epoch": 0.12867752070836905,
+      "grad_norm": 7.731047121804707,
+      "learning_rate": 2e-05,
+      "loss": 0.3266,
+      "step": 5406
+    },
+    {
+      "epoch": 0.12870132343140056,
+      "grad_norm": 7.149298924077983,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 5407
+    },
+    {
+      "epoch": 0.12872512615443207,
+      "grad_norm": 7.047456590658007,
+      "learning_rate": 2e-05,
+      "loss": 0.4593,
+      "step": 5408
+    },
+    {
+      "epoch": 0.1287489288774636,
+      "grad_norm": 2.8599582616246866,
+      "learning_rate": 2e-05,
+      "loss": 0.2124,
+      "step": 5409
+    },
+    {
+      "epoch": 0.1287727316004951,
+      "grad_norm": 10.449345210741557,
+      "learning_rate": 2e-05,
+      "loss": 2.2128,
+      "step": 5410
+    },
+    {
+      "epoch": 0.1287965343235266,
+      "grad_norm": 4.331580181318673,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 5411
+    },
+    {
+      "epoch": 0.12882033704655813,
+      "grad_norm": 4.052399699476726,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 5412
+    },
+    {
+      "epoch": 0.12884413976958964,
+      "grad_norm": 5.465947805196772,
+      "learning_rate": 2e-05,
+      "loss": 0.6283,
+      "step": 5413
+    },
+    {
+      "epoch": 0.12886794249262115,
+      "grad_norm": 2.4795548094421,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 5414
+    },
+    {
+      "epoch": 0.12889174521565266,
+      "grad_norm": 3.8376270235266663,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 5415
+    },
+    {
+      "epoch": 0.12891554793868418,
+      "grad_norm": 7.5097447088752425,
+      "learning_rate": 2e-05,
+      "loss": 0.4406,
+      "step": 5416
+    },
+    {
+      "epoch": 0.1289393506617157,
+      "grad_norm": 6.968781485943264,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 5417
+    },
+    {
+      "epoch": 0.1289631533847472,
+      "grad_norm": 2.753093605896278,
+      "learning_rate": 2e-05,
+      "loss": 0.2392,
+      "step": 5418
+    },
+    {
+      "epoch": 0.12898695610777872,
+      "grad_norm": 6.587821383318371,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 5419
+    },
+    {
+      "epoch": 0.12901075883081026,
+      "grad_norm": 1.9530717126042618,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 5420
+    },
+    {
+      "epoch": 0.12903456155384177,
+      "grad_norm": 5.41257087543492,
+      "learning_rate": 2e-05,
+      "loss": 0.2646,
+      "step": 5421
+    },
+    {
+      "epoch": 0.12905836427687328,
+      "grad_norm": 4.614025257762996,
+      "learning_rate": 2e-05,
+      "loss": 0.63,
+      "step": 5422
+    },
+    {
+      "epoch": 0.1290821669999048,
+      "grad_norm": 2.578152468122225,
+      "learning_rate": 2e-05,
+      "loss": 0.2132,
+      "step": 5423
+    },
+    {
+      "epoch": 0.1291059697229363,
+      "grad_norm": 4.363577508061899,
+      "learning_rate": 2e-05,
+      "loss": 0.25,
+      "step": 5424
+    },
+    {
+      "epoch": 0.12912977244596782,
+      "grad_norm": 7.080987718434564,
+      "learning_rate": 2e-05,
+      "loss": 1.1694,
+      "step": 5425
+    },
+    {
+      "epoch": 0.12915357516899933,
+      "grad_norm": 5.864729108762773,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 5426
+    },
+    {
+      "epoch": 0.12917737789203085,
+      "grad_norm": 1.3636408170175538,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 5427
+    },
+    {
+      "epoch": 0.12920118061506236,
+      "grad_norm": 5.288800875570107,
+      "learning_rate": 2e-05,
+      "loss": 0.2161,
+      "step": 5428
+    },
+    {
+      "epoch": 0.12922498333809387,
+      "grad_norm": 4.161091943423627,
+      "learning_rate": 2e-05,
+      "loss": 0.2874,
+      "step": 5429
+    },
+    {
+      "epoch": 0.1292487860611254,
+      "grad_norm": 3.0975850772037914,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 5430
+    },
+    {
+      "epoch": 0.1292725887841569,
+      "grad_norm": 4.220936911811931,
+      "learning_rate": 2e-05,
+      "loss": 0.2617,
+      "step": 5431
+    },
+    {
+      "epoch": 0.1292963915071884,
+      "grad_norm": 5.02070534994538,
+      "learning_rate": 2e-05,
+      "loss": 0.4772,
+      "step": 5432
+    },
+    {
+      "epoch": 0.12932019423021993,
+      "grad_norm": 6.847729600916181,
+      "learning_rate": 2e-05,
+      "loss": 1.4812,
+      "step": 5433
+    },
+    {
+      "epoch": 0.12934399695325144,
+      "grad_norm": 1.803306653788371,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 5434
+    },
+    {
+      "epoch": 0.12936779967628298,
+      "grad_norm": 9.51571978853247,
+      "learning_rate": 2e-05,
+      "loss": 1.0035,
+      "step": 5435
+    },
+    {
+      "epoch": 0.1293916023993145,
+      "grad_norm": 8.78589490916932,
+      "learning_rate": 2e-05,
+      "loss": 0.8337,
+      "step": 5436
+    },
+    {
+      "epoch": 0.129415405122346,
+      "grad_norm": 10.212404654817494,
+      "learning_rate": 2e-05,
+      "loss": 0.5561,
+      "step": 5437
+    },
+    {
+      "epoch": 0.12943920784537752,
+      "grad_norm": 4.018296665526583,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 5438
+    },
+    {
+      "epoch": 0.12946301056840903,
+      "grad_norm": 6.16503184303942,
+      "learning_rate": 2e-05,
+      "loss": 0.3292,
+      "step": 5439
+    },
+    {
+      "epoch": 0.12948681329144054,
+      "grad_norm": 5.006080431278278,
+      "learning_rate": 2e-05,
+      "loss": 0.4414,
+      "step": 5440
+    },
+    {
+      "epoch": 0.12951061601447206,
+      "grad_norm": 4.167557465013553,
+      "learning_rate": 2e-05,
+      "loss": 0.5275,
+      "step": 5441
+    },
+    {
+      "epoch": 0.12953441873750357,
+      "grad_norm": 8.962571971404005,
+      "learning_rate": 2e-05,
+      "loss": 0.9455,
+      "step": 5442
+    },
+    {
+      "epoch": 0.12955822146053508,
+      "grad_norm": 7.377110199534639,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 5443
+    },
+    {
+      "epoch": 0.1295820241835666,
+      "grad_norm": 3.6907872307774348,
+      "learning_rate": 2e-05,
+      "loss": 0.6293,
+      "step": 5444
+    },
+    {
+      "epoch": 0.1296058269065981,
+      "grad_norm": 2.893930251020415,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 5445
+    },
+    {
+      "epoch": 0.12962962962962962,
+      "grad_norm": 3.7918664892736555,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 5446
+    },
+    {
+      "epoch": 0.12965343235266114,
+      "grad_norm": 7.923274570016798,
+      "learning_rate": 2e-05,
+      "loss": 0.3223,
+      "step": 5447
+    },
+    {
+      "epoch": 0.12967723507569265,
+      "grad_norm": 5.043774040008407,
+      "learning_rate": 2e-05,
+      "loss": 0.8008,
+      "step": 5448
+    },
+    {
+      "epoch": 0.12970103779872416,
+      "grad_norm": 6.487241262276304,
+      "learning_rate": 2e-05,
+      "loss": 0.5307,
+      "step": 5449
+    },
+    {
+      "epoch": 0.1297248405217557,
+      "grad_norm": 8.349697400881082,
+      "learning_rate": 2e-05,
+      "loss": 0.426,
+      "step": 5450
+    },
+    {
+      "epoch": 0.12974864324478722,
+      "grad_norm": 2.2353731954572367,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 5451
+    },
+    {
+      "epoch": 0.12977244596781873,
+      "grad_norm": 4.83781696668433,
+      "learning_rate": 2e-05,
+      "loss": 0.5323,
+      "step": 5452
+    },
+    {
+      "epoch": 0.12979624869085024,
+      "grad_norm": 2.2898282182527714,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 5453
+    },
+    {
+      "epoch": 0.12982005141388175,
+      "grad_norm": 4.769696641278494,
+      "learning_rate": 2e-05,
+      "loss": 0.8305,
+      "step": 5454
+    },
+    {
+      "epoch": 0.12984385413691327,
+      "grad_norm": 2.4418245753749566,
+      "learning_rate": 2e-05,
+      "loss": 0.157,
+      "step": 5455
+    },
+    {
+      "epoch": 0.12986765685994478,
+      "grad_norm": 5.4173787524864805,
+      "learning_rate": 2e-05,
+      "loss": 0.7489,
+      "step": 5456
+    },
+    {
+      "epoch": 0.1298914595829763,
+      "grad_norm": 9.096281883958497,
+      "learning_rate": 2e-05,
+      "loss": 0.7265,
+      "step": 5457
+    },
+    {
+      "epoch": 0.1299152623060078,
+      "grad_norm": 7.8906573431226095,
+      "learning_rate": 2e-05,
+      "loss": 0.5668,
+      "step": 5458
+    },
+    {
+      "epoch": 0.12993906502903932,
+      "grad_norm": 5.937900136293005,
+      "learning_rate": 2e-05,
+      "loss": 0.7609,
+      "step": 5459
+    },
+    {
+      "epoch": 0.12996286775207083,
+      "grad_norm": 7.4783942102610474,
+      "learning_rate": 2e-05,
+      "loss": 0.6349,
+      "step": 5460
+    },
+    {
+      "epoch": 0.12998667047510234,
+      "grad_norm": 7.660604584640031,
+      "learning_rate": 2e-05,
+      "loss": 1.2708,
+      "step": 5461
+    },
+    {
+      "epoch": 0.13001047319813386,
+      "grad_norm": 3.320514498986285,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 5462
+    },
+    {
+      "epoch": 0.13003427592116537,
+      "grad_norm": 3.723006192839614,
+      "learning_rate": 2e-05,
+      "loss": 0.34,
+      "step": 5463
+    },
+    {
+      "epoch": 0.13005807864419688,
+      "grad_norm": 16.67586686530757,
+      "learning_rate": 2e-05,
+      "loss": 0.8049,
+      "step": 5464
+    },
+    {
+      "epoch": 0.13008188136722842,
+      "grad_norm": 6.7223237390269635,
+      "learning_rate": 2e-05,
+      "loss": 0.5108,
+      "step": 5465
+    },
+    {
+      "epoch": 0.13010568409025994,
+      "grad_norm": 6.0379933848761445,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 5466
+    },
+    {
+      "epoch": 0.13012948681329145,
+      "grad_norm": 6.391981843043015,
+      "learning_rate": 2e-05,
+      "loss": 0.8526,
+      "step": 5467
+    },
+    {
+      "epoch": 0.13015328953632296,
+      "grad_norm": 8.071652067068653,
+      "learning_rate": 2e-05,
+      "loss": 0.5058,
+      "step": 5468
+    },
+    {
+      "epoch": 0.13017709225935448,
+      "grad_norm": 7.464567666922736,
+      "learning_rate": 2e-05,
+      "loss": 1.6908,
+      "step": 5469
+    },
+    {
+      "epoch": 0.130200894982386,
+      "grad_norm": 4.787359603404444,
+      "learning_rate": 2e-05,
+      "loss": 0.8389,
+      "step": 5470
+    },
+    {
+      "epoch": 0.1302246977054175,
+      "grad_norm": 5.828307698042798,
+      "learning_rate": 2e-05,
+      "loss": 0.9084,
+      "step": 5471
+    },
+    {
+      "epoch": 0.13024850042844902,
+      "grad_norm": 1.6706959517958468,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 5472
+    },
+    {
+      "epoch": 0.13027230315148053,
+      "grad_norm": 2.1354358260292656,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 5473
+    },
+    {
+      "epoch": 0.13029610587451204,
+      "grad_norm": 7.618371769468722,
+      "learning_rate": 2e-05,
+      "loss": 0.343,
+      "step": 5474
+    },
+    {
+      "epoch": 0.13031990859754355,
+      "grad_norm": 7.602938036990208,
+      "learning_rate": 2e-05,
+      "loss": 1.3634,
+      "step": 5475
+    },
+    {
+      "epoch": 0.13034371132057507,
+      "grad_norm": 11.03326104791881,
+      "learning_rate": 2e-05,
+      "loss": 0.4165,
+      "step": 5476
+    },
+    {
+      "epoch": 0.13036751404360658,
+      "grad_norm": 1.691487982961677,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 5477
+    },
+    {
+      "epoch": 0.1303913167666381,
+      "grad_norm": 4.0607515536746455,
+      "learning_rate": 2e-05,
+      "loss": 0.2512,
+      "step": 5478
+    },
+    {
+      "epoch": 0.1304151194896696,
+      "grad_norm": 2.6588416187099857,
+      "learning_rate": 2e-05,
+      "loss": 0.2703,
+      "step": 5479
+    },
+    {
+      "epoch": 0.13043892221270112,
+      "grad_norm": 2.174682388355033,
+      "learning_rate": 2e-05,
+      "loss": 0.1971,
+      "step": 5480
+    },
+    {
+      "epoch": 0.13046272493573266,
+      "grad_norm": 13.23736344655271,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 5481
+    },
+    {
+      "epoch": 0.13048652765876417,
+      "grad_norm": 3.2753531382386414,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 5482
+    },
+    {
+      "epoch": 0.1305103303817957,
+      "grad_norm": 7.617819937375373,
+      "learning_rate": 2e-05,
+      "loss": 0.8619,
+      "step": 5483
+    },
+    {
+      "epoch": 0.1305341331048272,
+      "grad_norm": 4.525523356620994,
+      "learning_rate": 2e-05,
+      "loss": 0.5581,
+      "step": 5484
+    },
+    {
+      "epoch": 0.1305579358278587,
+      "grad_norm": 2.1892031693399345,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 5485
+    },
+    {
+      "epoch": 0.13058173855089023,
+      "grad_norm": 7.00725050330179,
+      "learning_rate": 2e-05,
+      "loss": 0.8197,
+      "step": 5486
+    },
+    {
+      "epoch": 0.13060554127392174,
+      "grad_norm": 4.451279909439157,
+      "learning_rate": 2e-05,
+      "loss": 0.6432,
+      "step": 5487
+    },
+    {
+      "epoch": 0.13062934399695325,
+      "grad_norm": 9.963663442263883,
+      "learning_rate": 2e-05,
+      "loss": 1.3195,
+      "step": 5488
+    },
+    {
+      "epoch": 0.13065314671998476,
+      "grad_norm": 4.279259055227695,
+      "learning_rate": 2e-05,
+      "loss": 0.2259,
+      "step": 5489
+    },
+    {
+      "epoch": 0.13067694944301628,
+      "grad_norm": 5.722099457533001,
+      "learning_rate": 2e-05,
+      "loss": 0.3257,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1307007521660478,
+      "grad_norm": 5.04628493380131,
+      "learning_rate": 2e-05,
+      "loss": 0.8123,
+      "step": 5491
+    },
+    {
+      "epoch": 0.1307245548890793,
+      "grad_norm": 6.856659237858877,
+      "learning_rate": 2e-05,
+      "loss": 0.2829,
+      "step": 5492
+    },
+    {
+      "epoch": 0.13074835761211082,
+      "grad_norm": 2.644859802622941,
+      "learning_rate": 2e-05,
+      "loss": 0.1789,
+      "step": 5493
+    },
+    {
+      "epoch": 0.13077216033514233,
+      "grad_norm": 7.9304612348120544,
+      "learning_rate": 2e-05,
+      "loss": 1.2411,
+      "step": 5494
+    },
+    {
+      "epoch": 0.13079596305817384,
+      "grad_norm": 7.693487701763694,
+      "learning_rate": 2e-05,
+      "loss": 0.5522,
+      "step": 5495
+    },
+    {
+      "epoch": 0.13081976578120538,
+      "grad_norm": 5.7314706029767555,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 5496
+    },
+    {
+      "epoch": 0.1308435685042369,
+      "grad_norm": 3.060065272699828,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 5497
+    },
+    {
+      "epoch": 0.1308673712272684,
+      "grad_norm": 2.4579109858793196,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 5498
+    },
+    {
+      "epoch": 0.13089117395029992,
+      "grad_norm": 4.7015600385910306,
+      "learning_rate": 2e-05,
+      "loss": 0.5043,
+      "step": 5499
+    },
+    {
+      "epoch": 0.13091497667333143,
+      "grad_norm": 6.416665354706507,
+      "learning_rate": 2e-05,
+      "loss": 0.8176,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13093877939636295,
+      "grad_norm": 3.773485489619042,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 5501
+    },
+    {
+      "epoch": 0.13096258211939446,
+      "grad_norm": 5.373190287160067,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 5502
+    },
+    {
+      "epoch": 0.13098638484242597,
+      "grad_norm": 4.214101496706285,
+      "learning_rate": 2e-05,
+      "loss": 0.1922,
+      "step": 5503
+    },
+    {
+      "epoch": 0.1310101875654575,
+      "grad_norm": 8.521526935049481,
+      "learning_rate": 2e-05,
+      "loss": 0.3105,
+      "step": 5504
+    },
+    {
+      "epoch": 0.131033990288489,
+      "grad_norm": 7.647791484306461,
+      "learning_rate": 2e-05,
+      "loss": 0.8337,
+      "step": 5505
+    },
+    {
+      "epoch": 0.1310577930115205,
+      "grad_norm": 9.152162527256154,
+      "learning_rate": 2e-05,
+      "loss": 0.835,
+      "step": 5506
+    },
+    {
+      "epoch": 0.13108159573455203,
+      "grad_norm": 9.989233832737746,
+      "learning_rate": 2e-05,
+      "loss": 0.423,
+      "step": 5507
+    },
+    {
+      "epoch": 0.13110539845758354,
+      "grad_norm": 3.488840495903781,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 5508
+    },
+    {
+      "epoch": 0.13112920118061505,
+      "grad_norm": 2.980114128757066,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 5509
+    },
+    {
+      "epoch": 0.13115300390364656,
+      "grad_norm": 5.876623414973603,
+      "learning_rate": 2e-05,
+      "loss": 0.7843,
+      "step": 5510
+    },
+    {
+      "epoch": 0.1311768066266781,
+      "grad_norm": 2.543263345727679,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 5511
+    },
+    {
+      "epoch": 0.13120060934970962,
+      "grad_norm": 6.748683637735871,
+      "learning_rate": 2e-05,
+      "loss": 0.3215,
+      "step": 5512
+    },
+    {
+      "epoch": 0.13122441207274113,
+      "grad_norm": 4.140395344575947,
+      "learning_rate": 2e-05,
+      "loss": 0.8962,
+      "step": 5513
+    },
+    {
+      "epoch": 0.13124821479577264,
+      "grad_norm": 5.458404219970071,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 5514
+    },
+    {
+      "epoch": 0.13127201751880416,
+      "grad_norm": 2.320157120768857,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 5515
+    },
+    {
+      "epoch": 0.13129582024183567,
+      "grad_norm": 6.2588083345725005,
+      "learning_rate": 2e-05,
+      "loss": 1.1237,
+      "step": 5516
+    },
+    {
+      "epoch": 0.13131962296486718,
+      "grad_norm": 7.502845863699795,
+      "learning_rate": 2e-05,
+      "loss": 0.3082,
+      "step": 5517
+    },
+    {
+      "epoch": 0.1313434256878987,
+      "grad_norm": 5.336342726787706,
+      "learning_rate": 2e-05,
+      "loss": 0.762,
+      "step": 5518
+    },
+    {
+      "epoch": 0.1313672284109302,
+      "grad_norm": 6.5669254062598466,
+      "learning_rate": 2e-05,
+      "loss": 0.9912,
+      "step": 5519
+    },
+    {
+      "epoch": 0.13139103113396172,
+      "grad_norm": 5.9645684171585,
+      "learning_rate": 2e-05,
+      "loss": 1.2463,
+      "step": 5520
+    },
+    {
+      "epoch": 0.13141483385699324,
+      "grad_norm": 8.791696185418623,
+      "learning_rate": 2e-05,
+      "loss": 0.3162,
+      "step": 5521
+    },
+    {
+      "epoch": 0.13143863658002475,
+      "grad_norm": 6.3511340467493875,
+      "learning_rate": 2e-05,
+      "loss": 0.8205,
+      "step": 5522
+    },
+    {
+      "epoch": 0.13146243930305626,
+      "grad_norm": 6.581752821680077,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 5523
+    },
+    {
+      "epoch": 0.13148624202608777,
+      "grad_norm": 2.1311294925947637,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 5524
+    },
+    {
+      "epoch": 0.1315100447491193,
+      "grad_norm": 5.8473606391812245,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 5525
+    },
+    {
+      "epoch": 0.13153384747215083,
+      "grad_norm": 6.551323673975393,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 5526
+    },
+    {
+      "epoch": 0.13155765019518234,
+      "grad_norm": 2.3664573877192008,
+      "learning_rate": 2e-05,
+      "loss": 0.2027,
+      "step": 5527
+    },
+    {
+      "epoch": 0.13158145291821385,
+      "grad_norm": 3.9754049483991416,
+      "learning_rate": 2e-05,
+      "loss": 0.332,
+      "step": 5528
+    },
+    {
+      "epoch": 0.13160525564124537,
+      "grad_norm": 4.110954147279235,
+      "learning_rate": 2e-05,
+      "loss": 0.251,
+      "step": 5529
+    },
+    {
+      "epoch": 0.13162905836427688,
+      "grad_norm": 7.659903002774822,
+      "learning_rate": 2e-05,
+      "loss": 1.5021,
+      "step": 5530
+    },
+    {
+      "epoch": 0.1316528610873084,
+      "grad_norm": 5.592985431255331,
+      "learning_rate": 2e-05,
+      "loss": 0.4176,
+      "step": 5531
+    },
+    {
+      "epoch": 0.1316766638103399,
+      "grad_norm": 7.055910776928873,
+      "learning_rate": 2e-05,
+      "loss": 0.9183,
+      "step": 5532
+    },
+    {
+      "epoch": 0.13170046653337142,
+      "grad_norm": 2.8103215573607967,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 5533
+    },
+    {
+      "epoch": 0.13172426925640293,
+      "grad_norm": 3.0315053258956306,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 5534
+    },
+    {
+      "epoch": 0.13174807197943444,
+      "grad_norm": 7.2973921097536305,
+      "learning_rate": 2e-05,
+      "loss": 1.0402,
+      "step": 5535
+    },
+    {
+      "epoch": 0.13177187470246596,
+      "grad_norm": 6.808091498741736,
+      "learning_rate": 2e-05,
+      "loss": 0.3177,
+      "step": 5536
+    },
+    {
+      "epoch": 0.13179567742549747,
+      "grad_norm": 5.811082396420532,
+      "learning_rate": 2e-05,
+      "loss": 0.7915,
+      "step": 5537
+    },
+    {
+      "epoch": 0.13181948014852898,
+      "grad_norm": 1.0528098609006646,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 5538
+    },
+    {
+      "epoch": 0.1318432828715605,
+      "grad_norm": 6.962405740111069,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 5539
+    },
+    {
+      "epoch": 0.131867085594592,
+      "grad_norm": 7.8371498054449935,
+      "learning_rate": 2e-05,
+      "loss": 0.9441,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13189088831762352,
+      "grad_norm": 2.431764703222899,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 5541
+    },
+    {
+      "epoch": 0.13191469104065506,
+      "grad_norm": 2.3959924721803927,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 5542
+    },
+    {
+      "epoch": 0.13193849376368658,
+      "grad_norm": 7.047801774108345,
+      "learning_rate": 2e-05,
+      "loss": 1.2161,
+      "step": 5543
+    },
+    {
+      "epoch": 0.1319622964867181,
+      "grad_norm": 13.743821807779474,
+      "learning_rate": 2e-05,
+      "loss": 0.4663,
+      "step": 5544
+    },
+    {
+      "epoch": 0.1319860992097496,
+      "grad_norm": 3.0846716055990666,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 5545
+    },
+    {
+      "epoch": 0.13200990193278112,
+      "grad_norm": 8.049350268247176,
+      "learning_rate": 2e-05,
+      "loss": 0.2971,
+      "step": 5546
+    },
+    {
+      "epoch": 0.13203370465581263,
+      "grad_norm": 7.7523738485734945,
+      "learning_rate": 2e-05,
+      "loss": 0.8032,
+      "step": 5547
+    },
+    {
+      "epoch": 0.13205750737884414,
+      "grad_norm": 17.439454916727264,
+      "learning_rate": 2e-05,
+      "loss": 0.7377,
+      "step": 5548
+    },
+    {
+      "epoch": 0.13208131010187565,
+      "grad_norm": 4.656173499679599,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 5549
+    },
+    {
+      "epoch": 0.13210511282490717,
+      "grad_norm": 0.9192020988158378,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 5550
+    },
+    {
+      "epoch": 0.13212891554793868,
+      "grad_norm": 22.52907882692252,
+      "learning_rate": 2e-05,
+      "loss": 0.5216,
+      "step": 5551
+    },
+    {
+      "epoch": 0.1321527182709702,
+      "grad_norm": 5.271595617039674,
+      "learning_rate": 2e-05,
+      "loss": 0.2183,
+      "step": 5552
+    },
+    {
+      "epoch": 0.1321765209940017,
+      "grad_norm": 4.42120078903254,
+      "learning_rate": 2e-05,
+      "loss": 0.3814,
+      "step": 5553
+    },
+    {
+      "epoch": 0.13220032371703322,
+      "grad_norm": 4.440803402470033,
+      "learning_rate": 2e-05,
+      "loss": 0.7121,
+      "step": 5554
+    },
+    {
+      "epoch": 0.13222412644006473,
+      "grad_norm": 3.551375685979646,
+      "learning_rate": 2e-05,
+      "loss": 0.6139,
+      "step": 5555
+    },
+    {
+      "epoch": 0.13224792916309625,
+      "grad_norm": 2.0549946548823472,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 5556
+    },
+    {
+      "epoch": 0.13227173188612779,
+      "grad_norm": 31.582081503043224,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 5557
+    },
+    {
+      "epoch": 0.1322955346091593,
+      "grad_norm": 10.05415455108408,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 5558
+    },
+    {
+      "epoch": 0.1323193373321908,
+      "grad_norm": 3.245379654979984,
+      "learning_rate": 2e-05,
+      "loss": 0.1046,
+      "step": 5559
+    },
+    {
+      "epoch": 0.13234314005522232,
+      "grad_norm": 4.438928492164444,
+      "learning_rate": 2e-05,
+      "loss": 0.3993,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13236694277825384,
+      "grad_norm": 4.007213083890448,
+      "learning_rate": 2e-05,
+      "loss": 0.5486,
+      "step": 5561
+    },
+    {
+      "epoch": 0.13239074550128535,
+      "grad_norm": 2.142776200314943,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 5562
+    },
+    {
+      "epoch": 0.13241454822431686,
+      "grad_norm": 7.29769135327356,
+      "learning_rate": 2e-05,
+      "loss": 2.5232,
+      "step": 5563
+    },
+    {
+      "epoch": 0.13243835094734838,
+      "grad_norm": 4.598363986862501,
+      "learning_rate": 2e-05,
+      "loss": 0.2555,
+      "step": 5564
+    },
+    {
+      "epoch": 0.1324621536703799,
+      "grad_norm": 2.84354331518083,
+      "learning_rate": 2e-05,
+      "loss": 0.2397,
+      "step": 5565
+    },
+    {
+      "epoch": 0.1324859563934114,
+      "grad_norm": 6.254562392093335,
+      "learning_rate": 2e-05,
+      "loss": 0.9962,
+      "step": 5566
+    },
+    {
+      "epoch": 0.13250975911644292,
+      "grad_norm": 4.215616755553235,
+      "learning_rate": 2e-05,
+      "loss": 0.5579,
+      "step": 5567
+    },
+    {
+      "epoch": 0.13253356183947443,
+      "grad_norm": 9.74824521944795,
+      "learning_rate": 2e-05,
+      "loss": 0.2051,
+      "step": 5568
+    },
+    {
+      "epoch": 0.13255736456250594,
+      "grad_norm": 5.076471515998615,
+      "learning_rate": 2e-05,
+      "loss": 0.6208,
+      "step": 5569
+    },
+    {
+      "epoch": 0.13258116728553745,
+      "grad_norm": 2.636620956240313,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 5570
+    },
+    {
+      "epoch": 0.13260497000856897,
+      "grad_norm": 5.269777212631838,
+      "learning_rate": 2e-05,
+      "loss": 0.6924,
+      "step": 5571
+    },
+    {
+      "epoch": 0.1326287727316005,
+      "grad_norm": 13.710098098644869,
+      "learning_rate": 2e-05,
+      "loss": 0.7036,
+      "step": 5572
+    },
+    {
+      "epoch": 0.13265257545463202,
+      "grad_norm": 12.74022665549191,
+      "learning_rate": 2e-05,
+      "loss": 0.5677,
+      "step": 5573
+    },
+    {
+      "epoch": 0.13267637817766353,
+      "grad_norm": 4.014741768767597,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 5574
+    },
+    {
+      "epoch": 0.13270018090069505,
+      "grad_norm": 5.764124548605716,
+      "learning_rate": 2e-05,
+      "loss": 0.5478,
+      "step": 5575
+    },
+    {
+      "epoch": 0.13272398362372656,
+      "grad_norm": 7.763622276626942,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 5576
+    },
+    {
+      "epoch": 0.13274778634675807,
+      "grad_norm": 9.86733623838362,
+      "learning_rate": 2e-05,
+      "loss": 1.5184,
+      "step": 5577
+    },
+    {
+      "epoch": 0.1327715890697896,
+      "grad_norm": 10.627674220014253,
+      "learning_rate": 2e-05,
+      "loss": 0.5906,
+      "step": 5578
+    },
+    {
+      "epoch": 0.1327953917928211,
+      "grad_norm": 2.4826653815632964,
+      "learning_rate": 2e-05,
+      "loss": 0.1212,
+      "step": 5579
+    },
+    {
+      "epoch": 0.1328191945158526,
+      "grad_norm": 8.980065300539762,
+      "learning_rate": 2e-05,
+      "loss": 1.2785,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13284299723888413,
+      "grad_norm": 1.561858787189603,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 5581
+    },
+    {
+      "epoch": 0.13286679996191564,
+      "grad_norm": 11.035251763237731,
+      "learning_rate": 2e-05,
+      "loss": 1.2779,
+      "step": 5582
+    },
+    {
+      "epoch": 0.13289060268494715,
+      "grad_norm": 1.1735134072864812,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 5583
+    },
+    {
+      "epoch": 0.13291440540797866,
+      "grad_norm": 6.388942055853214,
+      "learning_rate": 2e-05,
+      "loss": 1.0369,
+      "step": 5584
+    },
+    {
+      "epoch": 0.13293820813101018,
+      "grad_norm": 1.53363965328093,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 5585
+    },
+    {
+      "epoch": 0.1329620108540417,
+      "grad_norm": 3.800708937553252,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 5586
+    },
+    {
+      "epoch": 0.13298581357707323,
+      "grad_norm": 3.75874926374074,
+      "learning_rate": 2e-05,
+      "loss": 0.3061,
+      "step": 5587
+    },
+    {
+      "epoch": 0.13300961630010474,
+      "grad_norm": 6.883164017629324,
+      "learning_rate": 2e-05,
+      "loss": 0.5295,
+      "step": 5588
+    },
+    {
+      "epoch": 0.13303341902313626,
+      "grad_norm": 12.265933115633382,
+      "learning_rate": 2e-05,
+      "loss": 0.8757,
+      "step": 5589
+    },
+    {
+      "epoch": 0.13305722174616777,
+      "grad_norm": 3.4941386783280923,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 5590
+    },
+    {
+      "epoch": 0.13308102446919928,
+      "grad_norm": 6.231234335867161,
+      "learning_rate": 2e-05,
+      "loss": 0.4818,
+      "step": 5591
+    },
+    {
+      "epoch": 0.1331048271922308,
+      "grad_norm": 22.834350145086457,
+      "learning_rate": 2e-05,
+      "loss": 0.3658,
+      "step": 5592
+    },
+    {
+      "epoch": 0.1331286299152623,
+      "grad_norm": 4.38796754176436,
+      "learning_rate": 2e-05,
+      "loss": 0.3232,
+      "step": 5593
+    },
+    {
+      "epoch": 0.13315243263829382,
+      "grad_norm": 2.8318938567907534,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 5594
+    },
+    {
+      "epoch": 0.13317623536132533,
+      "grad_norm": 7.27951191018999,
+      "learning_rate": 2e-05,
+      "loss": 1.0086,
+      "step": 5595
+    },
+    {
+      "epoch": 0.13320003808435685,
+      "grad_norm": 8.07423494082485,
+      "learning_rate": 2e-05,
+      "loss": 0.9546,
+      "step": 5596
+    },
+    {
+      "epoch": 0.13322384080738836,
+      "grad_norm": 4.788613511054308,
+      "learning_rate": 2e-05,
+      "loss": 0.8161,
+      "step": 5597
+    },
+    {
+      "epoch": 0.13324764353041987,
+      "grad_norm": 5.919336207108427,
+      "learning_rate": 2e-05,
+      "loss": 0.2839,
+      "step": 5598
+    },
+    {
+      "epoch": 0.1332714462534514,
+      "grad_norm": 2.654302788181934,
+      "learning_rate": 2e-05,
+      "loss": 0.2626,
+      "step": 5599
+    },
+    {
+      "epoch": 0.1332952489764829,
+      "grad_norm": 5.612114843167641,
+      "learning_rate": 2e-05,
+      "loss": 0.3063,
+      "step": 5600
+    },
+    {
+      "epoch": 0.1333190516995144,
+      "grad_norm": 8.171925752815282,
+      "learning_rate": 2e-05,
+      "loss": 1.632,
+      "step": 5601
+    },
+    {
+      "epoch": 0.13334285442254593,
+      "grad_norm": 2.599193676693188,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 5602
+    },
+    {
+      "epoch": 0.13336665714557747,
+      "grad_norm": 3.726459946886697,
+      "learning_rate": 2e-05,
+      "loss": 0.3197,
+      "step": 5603
+    },
+    {
+      "epoch": 0.13339045986860898,
+      "grad_norm": 5.330204135478402,
+      "learning_rate": 2e-05,
+      "loss": 0.9665,
+      "step": 5604
+    },
+    {
+      "epoch": 0.1334142625916405,
+      "grad_norm": 1.9290321488015703,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 5605
+    },
+    {
+      "epoch": 0.133438065314672,
+      "grad_norm": 8.395845610714503,
+      "learning_rate": 2e-05,
+      "loss": 0.9474,
+      "step": 5606
+    },
+    {
+      "epoch": 0.13346186803770352,
+      "grad_norm": 2.1308014807760434,
+      "learning_rate": 2e-05,
+      "loss": 0.1926,
+      "step": 5607
+    },
+    {
+      "epoch": 0.13348567076073503,
+      "grad_norm": 7.135757616557318,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 5608
+    },
+    {
+      "epoch": 0.13350947348376654,
+      "grad_norm": 2.9208488508383788,
+      "learning_rate": 2e-05,
+      "loss": 0.1463,
+      "step": 5609
+    },
+    {
+      "epoch": 0.13353327620679806,
+      "grad_norm": 7.440317488912332,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13355707892982957,
+      "grad_norm": 7.2805758335214055,
+      "learning_rate": 2e-05,
+      "loss": 0.3532,
+      "step": 5611
+    },
+    {
+      "epoch": 0.13358088165286108,
+      "grad_norm": 2.5676383815128503,
+      "learning_rate": 2e-05,
+      "loss": 0.2814,
+      "step": 5612
+    },
+    {
+      "epoch": 0.1336046843758926,
+      "grad_norm": 6.198134823776082,
+      "learning_rate": 2e-05,
+      "loss": 0.5378,
+      "step": 5613
+    },
+    {
+      "epoch": 0.1336284870989241,
+      "grad_norm": 10.424482090648779,
+      "learning_rate": 2e-05,
+      "loss": 0.9819,
+      "step": 5614
+    },
+    {
+      "epoch": 0.13365228982195562,
+      "grad_norm": 2.0434430130042984,
+      "learning_rate": 2e-05,
+      "loss": 0.2075,
+      "step": 5615
+    },
+    {
+      "epoch": 0.13367609254498714,
+      "grad_norm": 2.216857058028805,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 5616
+    },
+    {
+      "epoch": 0.13369989526801865,
+      "grad_norm": 7.571457061668696,
+      "learning_rate": 2e-05,
+      "loss": 0.3833,
+      "step": 5617
+    },
+    {
+      "epoch": 0.1337236979910502,
+      "grad_norm": 3.174288982255389,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 5618
+    },
+    {
+      "epoch": 0.1337475007140817,
+      "grad_norm": 8.337040888352512,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 5619
+    },
+    {
+      "epoch": 0.13377130343711321,
+      "grad_norm": 8.546224608824598,
+      "learning_rate": 2e-05,
+      "loss": 1.3993,
+      "step": 5620
+    },
+    {
+      "epoch": 0.13379510616014473,
+      "grad_norm": 3.7703014846784972,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 5621
+    },
+    {
+      "epoch": 0.13381890888317624,
+      "grad_norm": 5.902160139832697,
+      "learning_rate": 2e-05,
+      "loss": 1.1371,
+      "step": 5622
+    },
+    {
+      "epoch": 0.13384271160620775,
+      "grad_norm": 6.728689418454657,
+      "learning_rate": 2e-05,
+      "loss": 1.0051,
+      "step": 5623
+    },
+    {
+      "epoch": 0.13386651432923927,
+      "grad_norm": 2.1631150458834965,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 5624
+    },
+    {
+      "epoch": 0.13389031705227078,
+      "grad_norm": 3.9586879038706186,
+      "learning_rate": 2e-05,
+      "loss": 0.4315,
+      "step": 5625
+    },
+    {
+      "epoch": 0.1339141197753023,
+      "grad_norm": 3.9962571250650467,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 5626
+    },
+    {
+      "epoch": 0.1339379224983338,
+      "grad_norm": 2.6626578376782137,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 5627
+    },
+    {
+      "epoch": 0.13396172522136532,
+      "grad_norm": 3.6246954279677794,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 5628
+    },
+    {
+      "epoch": 0.13398552794439683,
+      "grad_norm": 3.01221575798268,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 5629
+    },
+    {
+      "epoch": 0.13400933066742834,
+      "grad_norm": 10.8644328431909,
+      "learning_rate": 2e-05,
+      "loss": 0.3493,
+      "step": 5630
+    },
+    {
+      "epoch": 0.13403313339045986,
+      "grad_norm": 4.620539435069518,
+      "learning_rate": 2e-05,
+      "loss": 0.9022,
+      "step": 5631
+    },
+    {
+      "epoch": 0.13405693611349137,
+      "grad_norm": 2.200703731522039,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 5632
+    },
+    {
+      "epoch": 0.1340807388365229,
+      "grad_norm": 15.310671331822283,
+      "learning_rate": 2e-05,
+      "loss": 0.7616,
+      "step": 5633
+    },
+    {
+      "epoch": 0.13410454155955442,
+      "grad_norm": 4.156492733641971,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 5634
+    },
+    {
+      "epoch": 0.13412834428258594,
+      "grad_norm": 6.3056661929691975,
+      "learning_rate": 2e-05,
+      "loss": 0.3,
+      "step": 5635
+    },
+    {
+      "epoch": 0.13415214700561745,
+      "grad_norm": 3.8642879969848174,
+      "learning_rate": 2e-05,
+      "loss": 0.3159,
+      "step": 5636
+    },
+    {
+      "epoch": 0.13417594972864896,
+      "grad_norm": 5.564983309442964,
+      "learning_rate": 2e-05,
+      "loss": 0.7602,
+      "step": 5637
+    },
+    {
+      "epoch": 0.13419975245168048,
+      "grad_norm": 11.504824226843327,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 5638
+    },
+    {
+      "epoch": 0.134223555174712,
+      "grad_norm": 2.4142970930082748,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 5639
+    },
+    {
+      "epoch": 0.1342473578977435,
+      "grad_norm": 11.061250328977113,
+      "learning_rate": 2e-05,
+      "loss": 1.1393,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13427116062077502,
+      "grad_norm": 8.46638240050809,
+      "learning_rate": 2e-05,
+      "loss": 1.489,
+      "step": 5641
+    },
+    {
+      "epoch": 0.13429496334380653,
+      "grad_norm": 3.615344287332834,
+      "learning_rate": 2e-05,
+      "loss": 0.1651,
+      "step": 5642
+    },
+    {
+      "epoch": 0.13431876606683804,
+      "grad_norm": 7.241496454228647,
+      "learning_rate": 2e-05,
+      "loss": 0.6903,
+      "step": 5643
+    },
+    {
+      "epoch": 0.13434256878986955,
+      "grad_norm": 5.583269206489711,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 5644
+    },
+    {
+      "epoch": 0.13436637151290107,
+      "grad_norm": 5.373566422460356,
+      "learning_rate": 2e-05,
+      "loss": 0.8173,
+      "step": 5645
+    },
+    {
+      "epoch": 0.13439017423593258,
+      "grad_norm": 3.711609814528763,
+      "learning_rate": 2e-05,
+      "loss": 0.3038,
+      "step": 5646
+    },
+    {
+      "epoch": 0.1344139769589641,
+      "grad_norm": 5.179688629718662,
+      "learning_rate": 2e-05,
+      "loss": 0.846,
+      "step": 5647
+    },
+    {
+      "epoch": 0.13443777968199563,
+      "grad_norm": 1.3410228265701165,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 5648
+    },
+    {
+      "epoch": 0.13446158240502715,
+      "grad_norm": 12.844339437284166,
+      "learning_rate": 2e-05,
+      "loss": 2.6019,
+      "step": 5649
+    },
+    {
+      "epoch": 0.13448538512805866,
+      "grad_norm": 5.106725085140729,
+      "learning_rate": 2e-05,
+      "loss": 0.2834,
+      "step": 5650
+    },
+    {
+      "epoch": 0.13450918785109017,
+      "grad_norm": 1.5375998094075574,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 5651
+    },
+    {
+      "epoch": 0.13453299057412169,
+      "grad_norm": 0.9622948501568486,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 5652
+    },
+    {
+      "epoch": 0.1345567932971532,
+      "grad_norm": 3.0599725366674444,
+      "learning_rate": 2e-05,
+      "loss": 0.1506,
+      "step": 5653
+    },
+    {
+      "epoch": 0.1345805960201847,
+      "grad_norm": 8.883504634873008,
+      "learning_rate": 2e-05,
+      "loss": 0.454,
+      "step": 5654
+    },
+    {
+      "epoch": 0.13460439874321622,
+      "grad_norm": 4.03965992948574,
+      "learning_rate": 2e-05,
+      "loss": 0.2667,
+      "step": 5655
+    },
+    {
+      "epoch": 0.13462820146624774,
+      "grad_norm": 2.1376043516084606,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 5656
+    },
+    {
+      "epoch": 0.13465200418927925,
+      "grad_norm": 6.459987195369138,
+      "learning_rate": 2e-05,
+      "loss": 0.213,
+      "step": 5657
+    },
+    {
+      "epoch": 0.13467580691231076,
+      "grad_norm": 7.162703423842383,
+      "learning_rate": 2e-05,
+      "loss": 0.4369,
+      "step": 5658
+    },
+    {
+      "epoch": 0.13469960963534228,
+      "grad_norm": 6.6892664463744085,
+      "learning_rate": 2e-05,
+      "loss": 1.4511,
+      "step": 5659
+    },
+    {
+      "epoch": 0.1347234123583738,
+      "grad_norm": 7.538039489923631,
+      "learning_rate": 2e-05,
+      "loss": 0.5567,
+      "step": 5660
+    },
+    {
+      "epoch": 0.1347472150814053,
+      "grad_norm": 9.019679963342005,
+      "learning_rate": 2e-05,
+      "loss": 0.5075,
+      "step": 5661
+    },
+    {
+      "epoch": 0.13477101780443682,
+      "grad_norm": 3.615971892795576,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 5662
+    },
+    {
+      "epoch": 0.13479482052746833,
+      "grad_norm": 7.288073154179893,
+      "learning_rate": 2e-05,
+      "loss": 1.1696,
+      "step": 5663
+    },
+    {
+      "epoch": 0.13481862325049987,
+      "grad_norm": 4.363357869491243,
+      "learning_rate": 2e-05,
+      "loss": 0.3136,
+      "step": 5664
+    },
+    {
+      "epoch": 0.13484242597353138,
+      "grad_norm": 4.868312651813668,
+      "learning_rate": 2e-05,
+      "loss": 0.2865,
+      "step": 5665
+    },
+    {
+      "epoch": 0.1348662286965629,
+      "grad_norm": 11.008808713492115,
+      "learning_rate": 2e-05,
+      "loss": 1.8278,
+      "step": 5666
+    },
+    {
+      "epoch": 0.1348900314195944,
+      "grad_norm": 7.336803822904419,
+      "learning_rate": 2e-05,
+      "loss": 0.4183,
+      "step": 5667
+    },
+    {
+      "epoch": 0.13491383414262592,
+      "grad_norm": 4.969143408450365,
+      "learning_rate": 2e-05,
+      "loss": 0.7581,
+      "step": 5668
+    },
+    {
+      "epoch": 0.13493763686565743,
+      "grad_norm": 8.420092497963829,
+      "learning_rate": 2e-05,
+      "loss": 0.6727,
+      "step": 5669
+    },
+    {
+      "epoch": 0.13496143958868895,
+      "grad_norm": 7.0058857638739545,
+      "learning_rate": 2e-05,
+      "loss": 0.5296,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13498524231172046,
+      "grad_norm": 1.3474340304837689,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 5671
+    },
+    {
+      "epoch": 0.13500904503475197,
+      "grad_norm": 6.912651189493817,
+      "learning_rate": 2e-05,
+      "loss": 1.549,
+      "step": 5672
+    },
+    {
+      "epoch": 0.1350328477577835,
+      "grad_norm": 3.44346254441335,
+      "learning_rate": 2e-05,
+      "loss": 0.2818,
+      "step": 5673
+    },
+    {
+      "epoch": 0.135056650480815,
+      "grad_norm": 2.903707267574283,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 5674
+    },
+    {
+      "epoch": 0.1350804532038465,
+      "grad_norm": 5.384062341828606,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 5675
+    },
+    {
+      "epoch": 0.13510425592687803,
+      "grad_norm": 3.330575675565638,
+      "learning_rate": 2e-05,
+      "loss": 0.1649,
+      "step": 5676
+    },
+    {
+      "epoch": 0.13512805864990954,
+      "grad_norm": 6.163276382439875,
+      "learning_rate": 2e-05,
+      "loss": 0.7723,
+      "step": 5677
+    },
+    {
+      "epoch": 0.13515186137294105,
+      "grad_norm": 4.134814839891128,
+      "learning_rate": 2e-05,
+      "loss": 0.7327,
+      "step": 5678
+    },
+    {
+      "epoch": 0.1351756640959726,
+      "grad_norm": 5.342811122915468,
+      "learning_rate": 2e-05,
+      "loss": 0.2618,
+      "step": 5679
+    },
+    {
+      "epoch": 0.1351994668190041,
+      "grad_norm": 7.928055037747176,
+      "learning_rate": 2e-05,
+      "loss": 0.4627,
+      "step": 5680
+    },
+    {
+      "epoch": 0.13522326954203562,
+      "grad_norm": 5.1983106367983485,
+      "learning_rate": 2e-05,
+      "loss": 0.234,
+      "step": 5681
+    },
+    {
+      "epoch": 0.13524707226506713,
+      "grad_norm": 5.128112447741888,
+      "learning_rate": 2e-05,
+      "loss": 0.2611,
+      "step": 5682
+    },
+    {
+      "epoch": 0.13527087498809864,
+      "grad_norm": 4.6222745341163956,
+      "learning_rate": 2e-05,
+      "loss": 0.867,
+      "step": 5683
+    },
+    {
+      "epoch": 0.13529467771113016,
+      "grad_norm": 8.556155910480808,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 5684
+    },
+    {
+      "epoch": 0.13531848043416167,
+      "grad_norm": 6.244493944414469,
+      "learning_rate": 2e-05,
+      "loss": 0.8063,
+      "step": 5685
+    },
+    {
+      "epoch": 0.13534228315719318,
+      "grad_norm": 14.750427186770759,
+      "learning_rate": 2e-05,
+      "loss": 0.6601,
+      "step": 5686
+    },
+    {
+      "epoch": 0.1353660858802247,
+      "grad_norm": 2.4237173649457504,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 5687
+    },
+    {
+      "epoch": 0.1353898886032562,
+      "grad_norm": 16.64147378408144,
+      "learning_rate": 2e-05,
+      "loss": 0.9818,
+      "step": 5688
+    },
+    {
+      "epoch": 0.13541369132628772,
+      "grad_norm": 7.467017862369729,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 5689
+    },
+    {
+      "epoch": 0.13543749404931923,
+      "grad_norm": 8.20637748908214,
+      "learning_rate": 2e-05,
+      "loss": 0.7234,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13546129677235075,
+      "grad_norm": 1.895929584451552,
+      "learning_rate": 2e-05,
+      "loss": 0.1272,
+      "step": 5691
+    },
+    {
+      "epoch": 0.13548509949538226,
+      "grad_norm": 9.29023687675074,
+      "learning_rate": 2e-05,
+      "loss": 0.7517,
+      "step": 5692
+    },
+    {
+      "epoch": 0.13550890221841377,
+      "grad_norm": 46.18183399889229,
+      "learning_rate": 2e-05,
+      "loss": 0.3303,
+      "step": 5693
+    },
+    {
+      "epoch": 0.13553270494144531,
+      "grad_norm": 4.559621286179722,
+      "learning_rate": 2e-05,
+      "loss": 0.9966,
+      "step": 5694
+    },
+    {
+      "epoch": 0.13555650766447683,
+      "grad_norm": 10.068225957827554,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 5695
+    },
+    {
+      "epoch": 0.13558031038750834,
+      "grad_norm": 8.9822033473356,
+      "learning_rate": 2e-05,
+      "loss": 0.4055,
+      "step": 5696
+    },
+    {
+      "epoch": 0.13560411311053985,
+      "grad_norm": 3.8327994022455982,
+      "learning_rate": 2e-05,
+      "loss": 0.5323,
+      "step": 5697
+    },
+    {
+      "epoch": 0.13562791583357137,
+      "grad_norm": 2.630537894492204,
+      "learning_rate": 2e-05,
+      "loss": 0.1888,
+      "step": 5698
+    },
+    {
+      "epoch": 0.13565171855660288,
+      "grad_norm": 7.337874507856403,
+      "learning_rate": 2e-05,
+      "loss": 0.4682,
+      "step": 5699
+    },
+    {
+      "epoch": 0.1356755212796344,
+      "grad_norm": 2.6120462928090404,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1356993240026659,
+      "grad_norm": 9.098228939295542,
+      "learning_rate": 2e-05,
+      "loss": 1.0444,
+      "step": 5701
+    },
+    {
+      "epoch": 0.13572312672569742,
+      "grad_norm": 2.0051650353064603,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 5702
+    },
+    {
+      "epoch": 0.13574692944872893,
+      "grad_norm": 10.102033598009646,
+      "learning_rate": 2e-05,
+      "loss": 0.4964,
+      "step": 5703
+    },
+    {
+      "epoch": 0.13577073217176044,
+      "grad_norm": 4.286333324611567,
+      "learning_rate": 2e-05,
+      "loss": 0.2579,
+      "step": 5704
+    },
+    {
+      "epoch": 0.13579453489479196,
+      "grad_norm": 6.091272633372232,
+      "learning_rate": 2e-05,
+      "loss": 1.72,
+      "step": 5705
+    },
+    {
+      "epoch": 0.13581833761782347,
+      "grad_norm": 1.2836436644381675,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 5706
+    },
+    {
+      "epoch": 0.13584214034085498,
+      "grad_norm": 5.56055099087766,
+      "learning_rate": 2e-05,
+      "loss": 0.6738,
+      "step": 5707
+    },
+    {
+      "epoch": 0.1358659430638865,
+      "grad_norm": 8.63229827817011,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 5708
+    },
+    {
+      "epoch": 0.13588974578691804,
+      "grad_norm": 8.355468792535623,
+      "learning_rate": 2e-05,
+      "loss": 0.3535,
+      "step": 5709
+    },
+    {
+      "epoch": 0.13591354850994955,
+      "grad_norm": 11.598013595383994,
+      "learning_rate": 2e-05,
+      "loss": 0.4444,
+      "step": 5710
+    },
+    {
+      "epoch": 0.13593735123298106,
+      "grad_norm": 5.296159526265911,
+      "learning_rate": 2e-05,
+      "loss": 0.9287,
+      "step": 5711
+    },
+    {
+      "epoch": 0.13596115395601258,
+      "grad_norm": 6.826273080403333,
+      "learning_rate": 2e-05,
+      "loss": 1.0755,
+      "step": 5712
+    },
+    {
+      "epoch": 0.1359849566790441,
+      "grad_norm": 6.521791760173186,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 5713
+    },
+    {
+      "epoch": 0.1360087594020756,
+      "grad_norm": 2.267646983995424,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 5714
+    },
+    {
+      "epoch": 0.13603256212510711,
+      "grad_norm": 6.880538148574175,
+      "learning_rate": 2e-05,
+      "loss": 1.0476,
+      "step": 5715
+    },
+    {
+      "epoch": 0.13605636484813863,
+      "grad_norm": 2.121308104545261,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 5716
+    },
+    {
+      "epoch": 0.13608016757117014,
+      "grad_norm": 6.6934483955672635,
+      "learning_rate": 2e-05,
+      "loss": 1.5921,
+      "step": 5717
+    },
+    {
+      "epoch": 0.13610397029420165,
+      "grad_norm": 1.9239819398416742,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 5718
+    },
+    {
+      "epoch": 0.13612777301723317,
+      "grad_norm": 1.9374412030025068,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 5719
+    },
+    {
+      "epoch": 0.13615157574026468,
+      "grad_norm": 5.276661706801381,
+      "learning_rate": 2e-05,
+      "loss": 0.811,
+      "step": 5720
+    },
+    {
+      "epoch": 0.1361753784632962,
+      "grad_norm": 5.784884949907504,
+      "learning_rate": 2e-05,
+      "loss": 0.4945,
+      "step": 5721
+    },
+    {
+      "epoch": 0.1361991811863277,
+      "grad_norm": 1.80424539735417,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 5722
+    },
+    {
+      "epoch": 0.13622298390935922,
+      "grad_norm": 10.080257490590537,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 5723
+    },
+    {
+      "epoch": 0.13624678663239073,
+      "grad_norm": 6.811196240007591,
+      "learning_rate": 2e-05,
+      "loss": 0.4907,
+      "step": 5724
+    },
+    {
+      "epoch": 0.13627058935542227,
+      "grad_norm": 6.6059629310032815,
+      "learning_rate": 2e-05,
+      "loss": 0.4816,
+      "step": 5725
+    },
+    {
+      "epoch": 0.13629439207845379,
+      "grad_norm": 1.9223642494581652,
+      "learning_rate": 2e-05,
+      "loss": 0.1205,
+      "step": 5726
+    },
+    {
+      "epoch": 0.1363181948014853,
+      "grad_norm": 3.994852378558803,
+      "learning_rate": 2e-05,
+      "loss": 0.6368,
+      "step": 5727
+    },
+    {
+      "epoch": 0.1363419975245168,
+      "grad_norm": 6.607720621103164,
+      "learning_rate": 2e-05,
+      "loss": 0.2554,
+      "step": 5728
+    },
+    {
+      "epoch": 0.13636580024754832,
+      "grad_norm": 7.817280157691069,
+      "learning_rate": 2e-05,
+      "loss": 0.3123,
+      "step": 5729
+    },
+    {
+      "epoch": 0.13638960297057984,
+      "grad_norm": 3.321832320377641,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 5730
+    },
+    {
+      "epoch": 0.13641340569361135,
+      "grad_norm": 5.379932049206225,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 5731
+    },
+    {
+      "epoch": 0.13643720841664286,
+      "grad_norm": 8.7499222139998,
+      "learning_rate": 2e-05,
+      "loss": 0.7676,
+      "step": 5732
+    },
+    {
+      "epoch": 0.13646101113967438,
+      "grad_norm": 12.032134088086345,
+      "learning_rate": 2e-05,
+      "loss": 0.6254,
+      "step": 5733
+    },
+    {
+      "epoch": 0.1364848138627059,
+      "grad_norm": 7.124284653614009,
+      "learning_rate": 2e-05,
+      "loss": 0.513,
+      "step": 5734
+    },
+    {
+      "epoch": 0.1365086165857374,
+      "grad_norm": 10.594598057717947,
+      "learning_rate": 2e-05,
+      "loss": 0.7723,
+      "step": 5735
+    },
+    {
+      "epoch": 0.13653241930876892,
+      "grad_norm": 1.754046070383409,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 5736
+    },
+    {
+      "epoch": 0.13655622203180043,
+      "grad_norm": 5.236968700441728,
+      "learning_rate": 2e-05,
+      "loss": 0.6277,
+      "step": 5737
+    },
+    {
+      "epoch": 0.13658002475483194,
+      "grad_norm": 5.58963455480357,
+      "learning_rate": 2e-05,
+      "loss": 0.9764,
+      "step": 5738
+    },
+    {
+      "epoch": 0.13660382747786345,
+      "grad_norm": 5.309152779041187,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 5739
+    },
+    {
+      "epoch": 0.136627630200895,
+      "grad_norm": 4.670643558721637,
+      "learning_rate": 2e-05,
+      "loss": 0.4871,
+      "step": 5740
+    },
+    {
+      "epoch": 0.1366514329239265,
+      "grad_norm": 3.925999415710645,
+      "learning_rate": 2e-05,
+      "loss": 0.3238,
+      "step": 5741
+    },
+    {
+      "epoch": 0.13667523564695802,
+      "grad_norm": 5.293984824858116,
+      "learning_rate": 2e-05,
+      "loss": 0.6437,
+      "step": 5742
+    },
+    {
+      "epoch": 0.13669903836998953,
+      "grad_norm": 4.974241525997733,
+      "learning_rate": 2e-05,
+      "loss": 0.8243,
+      "step": 5743
+    },
+    {
+      "epoch": 0.13672284109302105,
+      "grad_norm": 5.893286754185945,
+      "learning_rate": 2e-05,
+      "loss": 0.8927,
+      "step": 5744
+    },
+    {
+      "epoch": 0.13674664381605256,
+      "grad_norm": 3.384533689043183,
+      "learning_rate": 2e-05,
+      "loss": 0.4297,
+      "step": 5745
+    },
+    {
+      "epoch": 0.13677044653908407,
+      "grad_norm": 1.8490571144877752,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 5746
+    },
+    {
+      "epoch": 0.13679424926211559,
+      "grad_norm": 2.9095964626413706,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 5747
+    },
+    {
+      "epoch": 0.1368180519851471,
+      "grad_norm": 2.235646345817967,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 5748
+    },
+    {
+      "epoch": 0.1368418547081786,
+      "grad_norm": 2.8852346019486337,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 5749
+    },
+    {
+      "epoch": 0.13686565743121012,
+      "grad_norm": 2.6062590251228728,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 5750
+    },
+    {
+      "epoch": 0.13688946015424164,
+      "grad_norm": 7.214481810802201,
+      "learning_rate": 2e-05,
+      "loss": 0.6788,
+      "step": 5751
+    },
+    {
+      "epoch": 0.13691326287727315,
+      "grad_norm": 2.062582925867707,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 5752
+    },
+    {
+      "epoch": 0.13693706560030466,
+      "grad_norm": 2.7851106270542916,
+      "learning_rate": 2e-05,
+      "loss": 0.1537,
+      "step": 5753
+    },
+    {
+      "epoch": 0.13696086832333618,
+      "grad_norm": 8.215443148083205,
+      "learning_rate": 2e-05,
+      "loss": 1.1713,
+      "step": 5754
+    },
+    {
+      "epoch": 0.13698467104636772,
+      "grad_norm": 7.437206840440916,
+      "learning_rate": 2e-05,
+      "loss": 0.8988,
+      "step": 5755
+    },
+    {
+      "epoch": 0.13700847376939923,
+      "grad_norm": 3.4476501137995093,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 5756
+    },
+    {
+      "epoch": 0.13703227649243074,
+      "grad_norm": 6.716151257568919,
+      "learning_rate": 2e-05,
+      "loss": 1.0889,
+      "step": 5757
+    },
+    {
+      "epoch": 0.13705607921546226,
+      "grad_norm": 9.56668339451893,
+      "learning_rate": 2e-05,
+      "loss": 0.9397,
+      "step": 5758
+    },
+    {
+      "epoch": 0.13707988193849377,
+      "grad_norm": 8.712870083010333,
+      "learning_rate": 2e-05,
+      "loss": 1.47,
+      "step": 5759
+    },
+    {
+      "epoch": 0.13710368466152528,
+      "grad_norm": 2.0033287515766927,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 5760
+    },
+    {
+      "epoch": 0.1371274873845568,
+      "grad_norm": 3.308584397013404,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 5761
+    },
+    {
+      "epoch": 0.1371512901075883,
+      "grad_norm": 6.150720036129562,
+      "learning_rate": 2e-05,
+      "loss": 0.8531,
+      "step": 5762
+    },
+    {
+      "epoch": 0.13717509283061982,
+      "grad_norm": 12.113116027126669,
+      "learning_rate": 2e-05,
+      "loss": 0.5737,
+      "step": 5763
+    },
+    {
+      "epoch": 0.13719889555365133,
+      "grad_norm": 6.6476866551978375,
+      "learning_rate": 2e-05,
+      "loss": 0.811,
+      "step": 5764
+    },
+    {
+      "epoch": 0.13722269827668285,
+      "grad_norm": 8.374004480295273,
+      "learning_rate": 2e-05,
+      "loss": 0.3866,
+      "step": 5765
+    },
+    {
+      "epoch": 0.13724650099971436,
+      "grad_norm": 4.776808593313083,
+      "learning_rate": 2e-05,
+      "loss": 0.6024,
+      "step": 5766
+    },
+    {
+      "epoch": 0.13727030372274587,
+      "grad_norm": 8.314235745669295,
+      "learning_rate": 2e-05,
+      "loss": 1.0194,
+      "step": 5767
+    },
+    {
+      "epoch": 0.1372941064457774,
+      "grad_norm": 3.689755611861755,
+      "learning_rate": 2e-05,
+      "loss": 0.4257,
+      "step": 5768
+    },
+    {
+      "epoch": 0.1373179091688089,
+      "grad_norm": 6.239172896643704,
+      "learning_rate": 2e-05,
+      "loss": 1.0511,
+      "step": 5769
+    },
+    {
+      "epoch": 0.13734171189184044,
+      "grad_norm": 5.212666486453696,
+      "learning_rate": 2e-05,
+      "loss": 0.7287,
+      "step": 5770
+    },
+    {
+      "epoch": 0.13736551461487195,
+      "grad_norm": 2.498574735550319,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 5771
+    },
+    {
+      "epoch": 0.13738931733790347,
+      "grad_norm": 1.0097425396456352,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 5772
+    },
+    {
+      "epoch": 0.13741312006093498,
+      "grad_norm": 2.6885928902878082,
+      "learning_rate": 2e-05,
+      "loss": 0.1461,
+      "step": 5773
+    },
+    {
+      "epoch": 0.1374369227839665,
+      "grad_norm": 4.338236638623477,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 5774
+    },
+    {
+      "epoch": 0.137460725506998,
+      "grad_norm": 12.918804769949233,
+      "learning_rate": 2e-05,
+      "loss": 0.9032,
+      "step": 5775
+    },
+    {
+      "epoch": 0.13748452823002952,
+      "grad_norm": 3.9148804003638453,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 5776
+    },
+    {
+      "epoch": 0.13750833095306103,
+      "grad_norm": 2.8415045382250073,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 5777
+    },
+    {
+      "epoch": 0.13753213367609254,
+      "grad_norm": 3.1646730733693573,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 5778
+    },
+    {
+      "epoch": 0.13755593639912406,
+      "grad_norm": 5.867150909309084,
+      "learning_rate": 2e-05,
+      "loss": 0.9866,
+      "step": 5779
+    },
+    {
+      "epoch": 0.13757973912215557,
+      "grad_norm": 4.355249436382998,
+      "learning_rate": 2e-05,
+      "loss": 0.7214,
+      "step": 5780
+    },
+    {
+      "epoch": 0.13760354184518708,
+      "grad_norm": 5.81767097328499,
+      "learning_rate": 2e-05,
+      "loss": 0.6549,
+      "step": 5781
+    },
+    {
+      "epoch": 0.1376273445682186,
+      "grad_norm": 7.370109292611125,
+      "learning_rate": 2e-05,
+      "loss": 1.5118,
+      "step": 5782
+    },
+    {
+      "epoch": 0.1376511472912501,
+      "grad_norm": 15.896981428373929,
+      "learning_rate": 2e-05,
+      "loss": 0.4487,
+      "step": 5783
+    },
+    {
+      "epoch": 0.13767495001428162,
+      "grad_norm": 5.739317947116784,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 5784
+    },
+    {
+      "epoch": 0.13769875273731313,
+      "grad_norm": 1.975581311702142,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 5785
+    },
+    {
+      "epoch": 0.13772255546034468,
+      "grad_norm": 9.862886441340333,
+      "learning_rate": 2e-05,
+      "loss": 1.5895,
+      "step": 5786
+    },
+    {
+      "epoch": 0.1377463581833762,
+      "grad_norm": 4.89285975508718,
+      "learning_rate": 2e-05,
+      "loss": 0.5301,
+      "step": 5787
+    },
+    {
+      "epoch": 0.1377701609064077,
+      "grad_norm": 5.260576586509829,
+      "learning_rate": 2e-05,
+      "loss": 0.5905,
+      "step": 5788
+    },
+    {
+      "epoch": 0.13779396362943921,
+      "grad_norm": 4.368677895469291,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 5789
+    },
+    {
+      "epoch": 0.13781776635247073,
+      "grad_norm": 6.03640823845482,
+      "learning_rate": 2e-05,
+      "loss": 0.8306,
+      "step": 5790
+    },
+    {
+      "epoch": 0.13784156907550224,
+      "grad_norm": 8.553521116203529,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 5791
+    },
+    {
+      "epoch": 0.13786537179853375,
+      "grad_norm": 3.0703811323196457,
+      "learning_rate": 2e-05,
+      "loss": 0.1829,
+      "step": 5792
+    },
+    {
+      "epoch": 0.13788917452156527,
+      "grad_norm": 4.099378217411263,
+      "learning_rate": 2e-05,
+      "loss": 0.7713,
+      "step": 5793
+    },
+    {
+      "epoch": 0.13791297724459678,
+      "grad_norm": 6.6974519797774015,
+      "learning_rate": 2e-05,
+      "loss": 0.8876,
+      "step": 5794
+    },
+    {
+      "epoch": 0.1379367799676283,
+      "grad_norm": 2.234943318788368,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 5795
+    },
+    {
+      "epoch": 0.1379605826906598,
+      "grad_norm": 6.571041326865925,
+      "learning_rate": 2e-05,
+      "loss": 0.2751,
+      "step": 5796
+    },
+    {
+      "epoch": 0.13798438541369132,
+      "grad_norm": 8.783892532571537,
+      "learning_rate": 2e-05,
+      "loss": 1.5082,
+      "step": 5797
+    },
+    {
+      "epoch": 0.13800818813672283,
+      "grad_norm": 8.084539884238858,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 5798
+    },
+    {
+      "epoch": 0.13803199085975434,
+      "grad_norm": 5.2711006857131535,
+      "learning_rate": 2e-05,
+      "loss": 0.6894,
+      "step": 5799
+    },
+    {
+      "epoch": 0.13805579358278586,
+      "grad_norm": 4.589260745490841,
+      "learning_rate": 2e-05,
+      "loss": 0.7873,
+      "step": 5800
+    },
+    {
+      "epoch": 0.1380795963058174,
+      "grad_norm": 6.085794060370995,
+      "learning_rate": 2e-05,
+      "loss": 0.6921,
+      "step": 5801
+    },
+    {
+      "epoch": 0.1381033990288489,
+      "grad_norm": 12.34198747565151,
+      "learning_rate": 2e-05,
+      "loss": 0.4057,
+      "step": 5802
+    },
+    {
+      "epoch": 0.13812720175188042,
+      "grad_norm": 5.9791447683551695,
+      "learning_rate": 2e-05,
+      "loss": 0.6458,
+      "step": 5803
+    },
+    {
+      "epoch": 0.13815100447491194,
+      "grad_norm": 34.04725343492156,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 5804
+    },
+    {
+      "epoch": 0.13817480719794345,
+      "grad_norm": 9.703366071050729,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 5805
+    },
+    {
+      "epoch": 0.13819860992097496,
+      "grad_norm": 3.9300263114696543,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 5806
+    },
+    {
+      "epoch": 0.13822241264400648,
+      "grad_norm": 5.839474723208607,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 5807
+    },
+    {
+      "epoch": 0.138246215367038,
+      "grad_norm": 4.111914295892479,
+      "learning_rate": 2e-05,
+      "loss": 0.443,
+      "step": 5808
+    },
+    {
+      "epoch": 0.1382700180900695,
+      "grad_norm": 7.640518237947095,
+      "learning_rate": 2e-05,
+      "loss": 0.3404,
+      "step": 5809
+    },
+    {
+      "epoch": 0.13829382081310101,
+      "grad_norm": 7.326673866684192,
+      "learning_rate": 2e-05,
+      "loss": 0.7469,
+      "step": 5810
+    },
+    {
+      "epoch": 0.13831762353613253,
+      "grad_norm": 5.903570453321934,
+      "learning_rate": 2e-05,
+      "loss": 0.82,
+      "step": 5811
+    },
+    {
+      "epoch": 0.13834142625916404,
+      "grad_norm": 8.262938395811423,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 5812
+    },
+    {
+      "epoch": 0.13836522898219555,
+      "grad_norm": 5.23313564797633,
+      "learning_rate": 2e-05,
+      "loss": 0.9032,
+      "step": 5813
+    },
+    {
+      "epoch": 0.13838903170522707,
+      "grad_norm": 6.853595089411026,
+      "learning_rate": 2e-05,
+      "loss": 0.8497,
+      "step": 5814
+    },
+    {
+      "epoch": 0.13841283442825858,
+      "grad_norm": 4.826786678316766,
+      "learning_rate": 2e-05,
+      "loss": 0.1265,
+      "step": 5815
+    },
+    {
+      "epoch": 0.13843663715129012,
+      "grad_norm": 5.960484596718809,
+      "learning_rate": 2e-05,
+      "loss": 0.3094,
+      "step": 5816
+    },
+    {
+      "epoch": 0.13846043987432163,
+      "grad_norm": 2.5806321031263617,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 5817
+    },
+    {
+      "epoch": 0.13848424259735315,
+      "grad_norm": 3.071016166220183,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 5818
+    },
+    {
+      "epoch": 0.13850804532038466,
+      "grad_norm": 7.0882813086749765,
+      "learning_rate": 2e-05,
+      "loss": 0.8662,
+      "step": 5819
+    },
+    {
+      "epoch": 0.13853184804341617,
+      "grad_norm": 8.132303133927161,
+      "learning_rate": 2e-05,
+      "loss": 1.2059,
+      "step": 5820
+    },
+    {
+      "epoch": 0.13855565076644769,
+      "grad_norm": 5.303132797828768,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 5821
+    },
+    {
+      "epoch": 0.1385794534894792,
+      "grad_norm": 2.880835829058394,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 5822
+    },
+    {
+      "epoch": 0.1386032562125107,
+      "grad_norm": 3.7446847106465375,
+      "learning_rate": 2e-05,
+      "loss": 0.532,
+      "step": 5823
+    },
+    {
+      "epoch": 0.13862705893554222,
+      "grad_norm": 8.640329212156923,
+      "learning_rate": 2e-05,
+      "loss": 1.1993,
+      "step": 5824
+    },
+    {
+      "epoch": 0.13865086165857374,
+      "grad_norm": 1.6436167042942633,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 5825
+    },
+    {
+      "epoch": 0.13867466438160525,
+      "grad_norm": 19.975202036114656,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 5826
+    },
+    {
+      "epoch": 0.13869846710463676,
+      "grad_norm": 4.320750058432274,
+      "learning_rate": 2e-05,
+      "loss": 1.0391,
+      "step": 5827
+    },
+    {
+      "epoch": 0.13872226982766828,
+      "grad_norm": 4.2998758640526855,
+      "learning_rate": 2e-05,
+      "loss": 0.7319,
+      "step": 5828
+    },
+    {
+      "epoch": 0.1387460725506998,
+      "grad_norm": 17.832131594386677,
+      "learning_rate": 2e-05,
+      "loss": 2.174,
+      "step": 5829
+    },
+    {
+      "epoch": 0.1387698752737313,
+      "grad_norm": 3.122030176080198,
+      "learning_rate": 2e-05,
+      "loss": 0.1731,
+      "step": 5830
+    },
+    {
+      "epoch": 0.13879367799676284,
+      "grad_norm": 6.855537131128323,
+      "learning_rate": 2e-05,
+      "loss": 1.1301,
+      "step": 5831
+    },
+    {
+      "epoch": 0.13881748071979436,
+      "grad_norm": 7.48178596629275,
+      "learning_rate": 2e-05,
+      "loss": 0.2353,
+      "step": 5832
+    },
+    {
+      "epoch": 0.13884128344282587,
+      "grad_norm": 7.864009631141614,
+      "learning_rate": 2e-05,
+      "loss": 0.9233,
+      "step": 5833
+    },
+    {
+      "epoch": 0.13886508616585738,
+      "grad_norm": 13.395675482203755,
+      "learning_rate": 2e-05,
+      "loss": 1.2165,
+      "step": 5834
+    },
+    {
+      "epoch": 0.1388888888888889,
+      "grad_norm": 2.0490945155856437,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 5835
+    },
+    {
+      "epoch": 0.1389126916119204,
+      "grad_norm": 1.1792061004325554,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 5836
+    },
+    {
+      "epoch": 0.13893649433495192,
+      "grad_norm": 9.458988671006512,
+      "learning_rate": 2e-05,
+      "loss": 0.3268,
+      "step": 5837
+    },
+    {
+      "epoch": 0.13896029705798343,
+      "grad_norm": 5.205316123078917,
+      "learning_rate": 2e-05,
+      "loss": 0.2309,
+      "step": 5838
+    },
+    {
+      "epoch": 0.13898409978101495,
+      "grad_norm": 6.544803387312809,
+      "learning_rate": 2e-05,
+      "loss": 0.4117,
+      "step": 5839
+    },
+    {
+      "epoch": 0.13900790250404646,
+      "grad_norm": 1.421039867158823,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 5840
+    },
+    {
+      "epoch": 0.13903170522707797,
+      "grad_norm": 3.460143076794283,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 5841
+    },
+    {
+      "epoch": 0.13905550795010949,
+      "grad_norm": 4.481842528642481,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 5842
+    },
+    {
+      "epoch": 0.139079310673141,
+      "grad_norm": 6.8662695703199566,
+      "learning_rate": 2e-05,
+      "loss": 0.3552,
+      "step": 5843
+    },
+    {
+      "epoch": 0.1391031133961725,
+      "grad_norm": 7.252199779298179,
+      "learning_rate": 2e-05,
+      "loss": 1.2872,
+      "step": 5844
+    },
+    {
+      "epoch": 0.13912691611920402,
+      "grad_norm": 14.098239512417713,
+      "learning_rate": 2e-05,
+      "loss": 0.5601,
+      "step": 5845
+    },
+    {
+      "epoch": 0.13915071884223554,
+      "grad_norm": 2.8201432490683374,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 5846
+    },
+    {
+      "epoch": 0.13917452156526708,
+      "grad_norm": 7.9101849848622745,
+      "learning_rate": 2e-05,
+      "loss": 0.3502,
+      "step": 5847
+    },
+    {
+      "epoch": 0.1391983242882986,
+      "grad_norm": 7.300559555081929,
+      "learning_rate": 2e-05,
+      "loss": 0.3688,
+      "step": 5848
+    },
+    {
+      "epoch": 0.1392221270113301,
+      "grad_norm": 5.855861060212256,
+      "learning_rate": 2e-05,
+      "loss": 0.6771,
+      "step": 5849
+    },
+    {
+      "epoch": 0.13924592973436162,
+      "grad_norm": 1.4935860683614777,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 5850
+    },
+    {
+      "epoch": 0.13926973245739313,
+      "grad_norm": 10.036588415360878,
+      "learning_rate": 2e-05,
+      "loss": 0.2712,
+      "step": 5851
+    },
+    {
+      "epoch": 0.13929353518042464,
+      "grad_norm": 9.69869271284957,
+      "learning_rate": 2e-05,
+      "loss": 1.1692,
+      "step": 5852
+    },
+    {
+      "epoch": 0.13931733790345616,
+      "grad_norm": 15.577486133380656,
+      "learning_rate": 2e-05,
+      "loss": 1.1273,
+      "step": 5853
+    },
+    {
+      "epoch": 0.13934114062648767,
+      "grad_norm": 6.57560772681698,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 5854
+    },
+    {
+      "epoch": 0.13936494334951918,
+      "grad_norm": 4.736076576974065,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 5855
+    },
+    {
+      "epoch": 0.1393887460725507,
+      "grad_norm": 11.562182048355533,
+      "learning_rate": 2e-05,
+      "loss": 0.9573,
+      "step": 5856
+    },
+    {
+      "epoch": 0.1394125487955822,
+      "grad_norm": 9.467886486140445,
+      "learning_rate": 2e-05,
+      "loss": 0.3397,
+      "step": 5857
+    },
+    {
+      "epoch": 0.13943635151861372,
+      "grad_norm": 6.142803818340291,
+      "learning_rate": 2e-05,
+      "loss": 1.2554,
+      "step": 5858
+    },
+    {
+      "epoch": 0.13946015424164523,
+      "grad_norm": 3.861366896207936,
+      "learning_rate": 2e-05,
+      "loss": 0.2353,
+      "step": 5859
+    },
+    {
+      "epoch": 0.13948395696467675,
+      "grad_norm": 5.187553264916121,
+      "learning_rate": 2e-05,
+      "loss": 0.4459,
+      "step": 5860
+    },
+    {
+      "epoch": 0.13950775968770826,
+      "grad_norm": 6.207284215564084,
+      "learning_rate": 2e-05,
+      "loss": 1.2211,
+      "step": 5861
+    },
+    {
+      "epoch": 0.1395315624107398,
+      "grad_norm": 5.32651845874388,
+      "learning_rate": 2e-05,
+      "loss": 0.8973,
+      "step": 5862
+    },
+    {
+      "epoch": 0.13955536513377131,
+      "grad_norm": 4.819011086172669,
+      "learning_rate": 2e-05,
+      "loss": 0.4675,
+      "step": 5863
+    },
+    {
+      "epoch": 0.13957916785680283,
+      "grad_norm": 7.031923035823844,
+      "learning_rate": 2e-05,
+      "loss": 0.8769,
+      "step": 5864
+    },
+    {
+      "epoch": 0.13960297057983434,
+      "grad_norm": 5.07727161414453,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 5865
+    },
+    {
+      "epoch": 0.13962677330286585,
+      "grad_norm": 5.433152627995022,
+      "learning_rate": 2e-05,
+      "loss": 0.4783,
+      "step": 5866
+    },
+    {
+      "epoch": 0.13965057602589737,
+      "grad_norm": 3.14967649514065,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 5867
+    },
+    {
+      "epoch": 0.13967437874892888,
+      "grad_norm": 5.608210545906347,
+      "learning_rate": 2e-05,
+      "loss": 0.5668,
+      "step": 5868
+    },
+    {
+      "epoch": 0.1396981814719604,
+      "grad_norm": 7.809497429905664,
+      "learning_rate": 2e-05,
+      "loss": 1.9266,
+      "step": 5869
+    },
+    {
+      "epoch": 0.1397219841949919,
+      "grad_norm": 2.917323909437179,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 5870
+    },
+    {
+      "epoch": 0.13974578691802342,
+      "grad_norm": 3.921110795843174,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 5871
+    },
+    {
+      "epoch": 0.13976958964105493,
+      "grad_norm": 7.332250341285237,
+      "learning_rate": 2e-05,
+      "loss": 0.4152,
+      "step": 5872
+    },
+    {
+      "epoch": 0.13979339236408644,
+      "grad_norm": 1.9964310859788221,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 5873
+    },
+    {
+      "epoch": 0.13981719508711796,
+      "grad_norm": 3.395394118649524,
+      "learning_rate": 2e-05,
+      "loss": 0.178,
+      "step": 5874
+    },
+    {
+      "epoch": 0.13984099781014947,
+      "grad_norm": 7.503005478964444,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 5875
+    },
+    {
+      "epoch": 0.13986480053318098,
+      "grad_norm": 4.3187997433760525,
+      "learning_rate": 2e-05,
+      "loss": 0.7371,
+      "step": 5876
+    },
+    {
+      "epoch": 0.13988860325621252,
+      "grad_norm": 9.484188482247815,
+      "learning_rate": 2e-05,
+      "loss": 1.3629,
+      "step": 5877
+    },
+    {
+      "epoch": 0.13991240597924404,
+      "grad_norm": 9.872014902125898,
+      "learning_rate": 2e-05,
+      "loss": 1.2911,
+      "step": 5878
+    },
+    {
+      "epoch": 0.13993620870227555,
+      "grad_norm": 7.3216218194432425,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 5879
+    },
+    {
+      "epoch": 0.13996001142530706,
+      "grad_norm": 5.297117705999904,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 5880
+    },
+    {
+      "epoch": 0.13998381414833858,
+      "grad_norm": 5.213959720730124,
+      "learning_rate": 2e-05,
+      "loss": 0.809,
+      "step": 5881
+    },
+    {
+      "epoch": 0.1400076168713701,
+      "grad_norm": 4.584831426523408,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 5882
+    },
+    {
+      "epoch": 0.1400314195944016,
+      "grad_norm": 8.91914250284976,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 5883
+    },
+    {
+      "epoch": 0.14005522231743311,
+      "grad_norm": 6.587792009807234,
+      "learning_rate": 2e-05,
+      "loss": 1.3829,
+      "step": 5884
+    },
+    {
+      "epoch": 0.14007902504046463,
+      "grad_norm": 1.5500830107966663,
+      "learning_rate": 2e-05,
+      "loss": 0.1623,
+      "step": 5885
+    },
+    {
+      "epoch": 0.14010282776349614,
+      "grad_norm": 3.5826482332379457,
+      "learning_rate": 2e-05,
+      "loss": 0.4067,
+      "step": 5886
+    },
+    {
+      "epoch": 0.14012663048652765,
+      "grad_norm": 4.6429121405315925,
+      "learning_rate": 2e-05,
+      "loss": 0.7507,
+      "step": 5887
+    },
+    {
+      "epoch": 0.14015043320955917,
+      "grad_norm": 6.630868313403362,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 5888
+    },
+    {
+      "epoch": 0.14017423593259068,
+      "grad_norm": 4.973936143904459,
+      "learning_rate": 2e-05,
+      "loss": 0.1882,
+      "step": 5889
+    },
+    {
+      "epoch": 0.1401980386556222,
+      "grad_norm": 9.484280075553801,
+      "learning_rate": 2e-05,
+      "loss": 0.6564,
+      "step": 5890
+    },
+    {
+      "epoch": 0.1402218413786537,
+      "grad_norm": 4.034815771985964,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 5891
+    },
+    {
+      "epoch": 0.14024564410168525,
+      "grad_norm": 5.612540785961384,
+      "learning_rate": 2e-05,
+      "loss": 0.6979,
+      "step": 5892
+    },
+    {
+      "epoch": 0.14026944682471676,
+      "grad_norm": 5.432505699753289,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 5893
+    },
+    {
+      "epoch": 0.14029324954774827,
+      "grad_norm": 8.856101446570067,
+      "learning_rate": 2e-05,
+      "loss": 0.8974,
+      "step": 5894
+    },
+    {
+      "epoch": 0.14031705227077979,
+      "grad_norm": 7.489976497350764,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 5895
+    },
+    {
+      "epoch": 0.1403408549938113,
+      "grad_norm": 3.864032684521014,
+      "learning_rate": 2e-05,
+      "loss": 0.2468,
+      "step": 5896
+    },
+    {
+      "epoch": 0.1403646577168428,
+      "grad_norm": 3.233269439161349,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 5897
+    },
+    {
+      "epoch": 0.14038846043987432,
+      "grad_norm": 2.538394247583382,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 5898
+    },
+    {
+      "epoch": 0.14041226316290584,
+      "grad_norm": 2.6460863277007864,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 5899
+    },
+    {
+      "epoch": 0.14043606588593735,
+      "grad_norm": 4.209781683247729,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14045986860896886,
+      "grad_norm": 32.83024261811752,
+      "learning_rate": 2e-05,
+      "loss": 1.5428,
+      "step": 5901
+    },
+    {
+      "epoch": 0.14048367133200038,
+      "grad_norm": 20.156753826711874,
+      "learning_rate": 2e-05,
+      "loss": 1.5693,
+      "step": 5902
+    },
+    {
+      "epoch": 0.1405074740550319,
+      "grad_norm": 6.364414554434794,
+      "learning_rate": 2e-05,
+      "loss": 0.7582,
+      "step": 5903
+    },
+    {
+      "epoch": 0.1405312767780634,
+      "grad_norm": 9.649816291522683,
+      "learning_rate": 2e-05,
+      "loss": 1.0222,
+      "step": 5904
+    },
+    {
+      "epoch": 0.14055507950109491,
+      "grad_norm": 8.80544076912898,
+      "learning_rate": 2e-05,
+      "loss": 0.5149,
+      "step": 5905
+    },
+    {
+      "epoch": 0.14057888222412643,
+      "grad_norm": 45.49173805876656,
+      "learning_rate": 2e-05,
+      "loss": 0.923,
+      "step": 5906
+    },
+    {
+      "epoch": 0.14060268494715797,
+      "grad_norm": 7.3131476926195935,
+      "learning_rate": 2e-05,
+      "loss": 0.5314,
+      "step": 5907
+    },
+    {
+      "epoch": 0.14062648767018948,
+      "grad_norm": 4.739428497693138,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 5908
+    },
+    {
+      "epoch": 0.140650290393221,
+      "grad_norm": 1.1792602692908083,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 5909
+    },
+    {
+      "epoch": 0.1406740931162525,
+      "grad_norm": 5.918162035659991,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 5910
+    },
+    {
+      "epoch": 0.14069789583928402,
+      "grad_norm": 8.462973302160266,
+      "learning_rate": 2e-05,
+      "loss": 0.7818,
+      "step": 5911
+    },
+    {
+      "epoch": 0.14072169856231553,
+      "grad_norm": 5.496753907764335,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 5912
+    },
+    {
+      "epoch": 0.14074550128534705,
+      "grad_norm": 1.9510198268361525,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 5913
+    },
+    {
+      "epoch": 0.14076930400837856,
+      "grad_norm": 5.337610245632298,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 5914
+    },
+    {
+      "epoch": 0.14079310673141007,
+      "grad_norm": 6.4677146917172745,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 5915
+    },
+    {
+      "epoch": 0.14081690945444159,
+      "grad_norm": 1.2049834037093605,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 5916
+    },
+    {
+      "epoch": 0.1408407121774731,
+      "grad_norm": 1.2569426506329953,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 5917
+    },
+    {
+      "epoch": 0.1408645149005046,
+      "grad_norm": 6.798580158129132,
+      "learning_rate": 2e-05,
+      "loss": 0.4322,
+      "step": 5918
+    },
+    {
+      "epoch": 0.14088831762353612,
+      "grad_norm": 9.720525844582326,
+      "learning_rate": 2e-05,
+      "loss": 1.4666,
+      "step": 5919
+    },
+    {
+      "epoch": 0.14091212034656764,
+      "grad_norm": 3.669471662938432,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 5920
+    },
+    {
+      "epoch": 0.14093592306959915,
+      "grad_norm": 10.25183385258545,
+      "learning_rate": 2e-05,
+      "loss": 0.4334,
+      "step": 5921
+    },
+    {
+      "epoch": 0.14095972579263066,
+      "grad_norm": 4.891060282853051,
+      "learning_rate": 2e-05,
+      "loss": 0.5508,
+      "step": 5922
+    },
+    {
+      "epoch": 0.1409835285156622,
+      "grad_norm": 3.7323295579192806,
+      "learning_rate": 2e-05,
+      "loss": 0.2404,
+      "step": 5923
+    },
+    {
+      "epoch": 0.14100733123869372,
+      "grad_norm": 3.520709016028084,
+      "learning_rate": 2e-05,
+      "loss": 0.2799,
+      "step": 5924
+    },
+    {
+      "epoch": 0.14103113396172523,
+      "grad_norm": 7.647847907886687,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 5925
+    },
+    {
+      "epoch": 0.14105493668475674,
+      "grad_norm": 3.664398553457604,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 5926
+    },
+    {
+      "epoch": 0.14107873940778826,
+      "grad_norm": 6.6782170039038355,
+      "learning_rate": 2e-05,
+      "loss": 0.7099,
+      "step": 5927
+    },
+    {
+      "epoch": 0.14110254213081977,
+      "grad_norm": 7.255617255686037,
+      "learning_rate": 2e-05,
+      "loss": 1.1749,
+      "step": 5928
+    },
+    {
+      "epoch": 0.14112634485385128,
+      "grad_norm": 3.8259586995359833,
+      "learning_rate": 2e-05,
+      "loss": 0.2438,
+      "step": 5929
+    },
+    {
+      "epoch": 0.1411501475768828,
+      "grad_norm": 8.519341773095876,
+      "learning_rate": 2e-05,
+      "loss": 0.9116,
+      "step": 5930
+    },
+    {
+      "epoch": 0.1411739502999143,
+      "grad_norm": 9.204032654037539,
+      "learning_rate": 2e-05,
+      "loss": 0.9639,
+      "step": 5931
+    },
+    {
+      "epoch": 0.14119775302294582,
+      "grad_norm": 8.965735981490687,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 5932
+    },
+    {
+      "epoch": 0.14122155574597733,
+      "grad_norm": 10.703675278668957,
+      "learning_rate": 2e-05,
+      "loss": 1.2905,
+      "step": 5933
+    },
+    {
+      "epoch": 0.14124535846900885,
+      "grad_norm": 4.069312730911727,
+      "learning_rate": 2e-05,
+      "loss": 0.3671,
+      "step": 5934
+    },
+    {
+      "epoch": 0.14126916119204036,
+      "grad_norm": 5.198745048673989,
+      "learning_rate": 2e-05,
+      "loss": 0.6326,
+      "step": 5935
+    },
+    {
+      "epoch": 0.14129296391507187,
+      "grad_norm": 6.30071166849407,
+      "learning_rate": 2e-05,
+      "loss": 0.3883,
+      "step": 5936
+    },
+    {
+      "epoch": 0.14131676663810339,
+      "grad_norm": 7.314239509620535,
+      "learning_rate": 2e-05,
+      "loss": 0.3878,
+      "step": 5937
+    },
+    {
+      "epoch": 0.14134056936113493,
+      "grad_norm": 2.3990479116929575,
+      "learning_rate": 2e-05,
+      "loss": 0.1873,
+      "step": 5938
+    },
+    {
+      "epoch": 0.14136437208416644,
+      "grad_norm": 7.097703130950034,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 5939
+    },
+    {
+      "epoch": 0.14138817480719795,
+      "grad_norm": 3.1143182349355705,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14141197753022947,
+      "grad_norm": 5.32711510600681,
+      "learning_rate": 2e-05,
+      "loss": 0.6939,
+      "step": 5941
+    },
+    {
+      "epoch": 0.14143578025326098,
+      "grad_norm": 5.676346514780785,
+      "learning_rate": 2e-05,
+      "loss": 0.6706,
+      "step": 5942
+    },
+    {
+      "epoch": 0.1414595829762925,
+      "grad_norm": 3.9751771230468846,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 5943
+    },
+    {
+      "epoch": 0.141483385699324,
+      "grad_norm": 4.996360307837379,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 5944
+    },
+    {
+      "epoch": 0.14150718842235552,
+      "grad_norm": 8.876686090895305,
+      "learning_rate": 2e-05,
+      "loss": 0.6682,
+      "step": 5945
+    },
+    {
+      "epoch": 0.14153099114538703,
+      "grad_norm": 5.354013786634349,
+      "learning_rate": 2e-05,
+      "loss": 0.653,
+      "step": 5946
+    },
+    {
+      "epoch": 0.14155479386841854,
+      "grad_norm": 2.0560801223215814,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 5947
+    },
+    {
+      "epoch": 0.14157859659145006,
+      "grad_norm": 1.8253938621494659,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 5948
+    },
+    {
+      "epoch": 0.14160239931448157,
+      "grad_norm": 1.5725671359348414,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 5949
+    },
+    {
+      "epoch": 0.14162620203751308,
+      "grad_norm": 4.404530879798171,
+      "learning_rate": 2e-05,
+      "loss": 0.7723,
+      "step": 5950
+    },
+    {
+      "epoch": 0.1416500047605446,
+      "grad_norm": 7.178985862906726,
+      "learning_rate": 2e-05,
+      "loss": 0.3648,
+      "step": 5951
+    },
+    {
+      "epoch": 0.1416738074835761,
+      "grad_norm": 7.475109868991772,
+      "learning_rate": 2e-05,
+      "loss": 0.423,
+      "step": 5952
+    },
+    {
+      "epoch": 0.14169761020660765,
+      "grad_norm": 1.0882407311255315,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 5953
+    },
+    {
+      "epoch": 0.14172141292963916,
+      "grad_norm": 7.599791774740462,
+      "learning_rate": 2e-05,
+      "loss": 0.856,
+      "step": 5954
+    },
+    {
+      "epoch": 0.14174521565267068,
+      "grad_norm": 0.9855016562229616,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 5955
+    },
+    {
+      "epoch": 0.1417690183757022,
+      "grad_norm": 1.845990404952921,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 5956
+    },
+    {
+      "epoch": 0.1417928210987337,
+      "grad_norm": 9.939466026065741,
+      "learning_rate": 2e-05,
+      "loss": 0.5584,
+      "step": 5957
+    },
+    {
+      "epoch": 0.14181662382176521,
+      "grad_norm": 4.715756728869905,
+      "learning_rate": 2e-05,
+      "loss": 0.178,
+      "step": 5958
+    },
+    {
+      "epoch": 0.14184042654479673,
+      "grad_norm": 6.449570500140467,
+      "learning_rate": 2e-05,
+      "loss": 0.4192,
+      "step": 5959
+    },
+    {
+      "epoch": 0.14186422926782824,
+      "grad_norm": 4.314849199382355,
+      "learning_rate": 2e-05,
+      "loss": 0.4595,
+      "step": 5960
+    },
+    {
+      "epoch": 0.14188803199085975,
+      "grad_norm": 4.0181830587845795,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 5961
+    },
+    {
+      "epoch": 0.14191183471389127,
+      "grad_norm": 2.3404197095256256,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 5962
+    },
+    {
+      "epoch": 0.14193563743692278,
+      "grad_norm": 7.6115628900089325,
+      "learning_rate": 2e-05,
+      "loss": 0.4424,
+      "step": 5963
+    },
+    {
+      "epoch": 0.1419594401599543,
+      "grad_norm": 6.204529635558481,
+      "learning_rate": 2e-05,
+      "loss": 0.8603,
+      "step": 5964
+    },
+    {
+      "epoch": 0.1419832428829858,
+      "grad_norm": 4.420908267060261,
+      "learning_rate": 2e-05,
+      "loss": 0.5603,
+      "step": 5965
+    },
+    {
+      "epoch": 0.14200704560601732,
+      "grad_norm": 6.030469283936075,
+      "learning_rate": 2e-05,
+      "loss": 0.6629,
+      "step": 5966
+    },
+    {
+      "epoch": 0.14203084832904883,
+      "grad_norm": 3.7727376038562848,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 5967
+    },
+    {
+      "epoch": 0.14205465105208037,
+      "grad_norm": 5.437122120051045,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 5968
+    },
+    {
+      "epoch": 0.14207845377511188,
+      "grad_norm": 4.393363550816396,
+      "learning_rate": 2e-05,
+      "loss": 0.1528,
+      "step": 5969
+    },
+    {
+      "epoch": 0.1421022564981434,
+      "grad_norm": 3.6367330404111353,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1421260592211749,
+      "grad_norm": 6.544593980336721,
+      "learning_rate": 2e-05,
+      "loss": 0.2759,
+      "step": 5971
+    },
+    {
+      "epoch": 0.14214986194420642,
+      "grad_norm": 4.233678330767257,
+      "learning_rate": 2e-05,
+      "loss": 0.7156,
+      "step": 5972
+    },
+    {
+      "epoch": 0.14217366466723794,
+      "grad_norm": 2.81063989441186,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 5973
+    },
+    {
+      "epoch": 0.14219746739026945,
+      "grad_norm": 3.642003373083241,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 5974
+    },
+    {
+      "epoch": 0.14222127011330096,
+      "grad_norm": 3.2891565826365237,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 5975
+    },
+    {
+      "epoch": 0.14224507283633248,
+      "grad_norm": 9.24062959871678,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 5976
+    },
+    {
+      "epoch": 0.142268875559364,
+      "grad_norm": 9.433566475278727,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 5977
+    },
+    {
+      "epoch": 0.1422926782823955,
+      "grad_norm": 11.044662409976365,
+      "learning_rate": 2e-05,
+      "loss": 1.0217,
+      "step": 5978
+    },
+    {
+      "epoch": 0.14231648100542701,
+      "grad_norm": 7.9892637445256796,
+      "learning_rate": 2e-05,
+      "loss": 0.9753,
+      "step": 5979
+    },
+    {
+      "epoch": 0.14234028372845853,
+      "grad_norm": 5.215875140599528,
+      "learning_rate": 2e-05,
+      "loss": 0.2976,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14236408645149004,
+      "grad_norm": 32.38469926593702,
+      "learning_rate": 2e-05,
+      "loss": 0.8861,
+      "step": 5981
+    },
+    {
+      "epoch": 0.14238788917452155,
+      "grad_norm": 2.3474239127321006,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 5982
+    },
+    {
+      "epoch": 0.14241169189755307,
+      "grad_norm": 5.364708587495443,
+      "learning_rate": 2e-05,
+      "loss": 0.8204,
+      "step": 5983
+    },
+    {
+      "epoch": 0.1424354946205846,
+      "grad_norm": 5.265978715223523,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 5984
+    },
+    {
+      "epoch": 0.14245929734361612,
+      "grad_norm": 15.492811803421743,
+      "learning_rate": 2e-05,
+      "loss": 0.3289,
+      "step": 5985
+    },
+    {
+      "epoch": 0.14248310006664763,
+      "grad_norm": 1.4456606470755218,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 5986
+    },
+    {
+      "epoch": 0.14250690278967915,
+      "grad_norm": 3.9002954427437584,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 5987
+    },
+    {
+      "epoch": 0.14253070551271066,
+      "grad_norm": 4.636712104548355,
+      "learning_rate": 2e-05,
+      "loss": 0.7343,
+      "step": 5988
+    },
+    {
+      "epoch": 0.14255450823574217,
+      "grad_norm": 10.583815922981449,
+      "learning_rate": 2e-05,
+      "loss": 0.8655,
+      "step": 5989
+    },
+    {
+      "epoch": 0.14257831095877369,
+      "grad_norm": 4.507199307543651,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 5990
+    },
+    {
+      "epoch": 0.1426021136818052,
+      "grad_norm": 5.022893510882867,
+      "learning_rate": 2e-05,
+      "loss": 0.3361,
+      "step": 5991
+    },
+    {
+      "epoch": 0.1426259164048367,
+      "grad_norm": 4.832976881236371,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 5992
+    },
+    {
+      "epoch": 0.14264971912786822,
+      "grad_norm": 5.674432581200682,
+      "learning_rate": 2e-05,
+      "loss": 0.8102,
+      "step": 5993
+    },
+    {
+      "epoch": 0.14267352185089974,
+      "grad_norm": 3.548253101387764,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 5994
+    },
+    {
+      "epoch": 0.14269732457393125,
+      "grad_norm": 5.784201381674557,
+      "learning_rate": 2e-05,
+      "loss": 0.7447,
+      "step": 5995
+    },
+    {
+      "epoch": 0.14272112729696276,
+      "grad_norm": 5.4154670813074635,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 5996
+    },
+    {
+      "epoch": 0.14274493001999428,
+      "grad_norm": 1.7408560838998748,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 5997
+    },
+    {
+      "epoch": 0.1427687327430258,
+      "grad_norm": 1.5486655115143555,
+      "learning_rate": 2e-05,
+      "loss": 0.1635,
+      "step": 5998
+    },
+    {
+      "epoch": 0.14279253546605733,
+      "grad_norm": 11.014829413476237,
+      "learning_rate": 2e-05,
+      "loss": 0.708,
+      "step": 5999
+    },
+    {
+      "epoch": 0.14281633818908884,
+      "grad_norm": 6.252626809247937,
+      "learning_rate": 2e-05,
+      "loss": 1.0757,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14284014091212036,
+      "grad_norm": 8.016555502411675,
+      "learning_rate": 2e-05,
+      "loss": 1.0702,
+      "step": 6001
+    },
+    {
+      "epoch": 0.14286394363515187,
+      "grad_norm": 3.387006883655438,
+      "learning_rate": 2e-05,
+      "loss": 0.2182,
+      "step": 6002
+    },
+    {
+      "epoch": 0.14288774635818338,
+      "grad_norm": 2.3116661923620336,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 6003
+    },
+    {
+      "epoch": 0.1429115490812149,
+      "grad_norm": 4.673244412595824,
+      "learning_rate": 2e-05,
+      "loss": 0.2166,
+      "step": 6004
+    },
+    {
+      "epoch": 0.1429353518042464,
+      "grad_norm": 29.43515688206223,
+      "learning_rate": 2e-05,
+      "loss": 0.6063,
+      "step": 6005
+    },
+    {
+      "epoch": 0.14295915452727792,
+      "grad_norm": 7.331689539042746,
+      "learning_rate": 2e-05,
+      "loss": 0.3457,
+      "step": 6006
+    },
+    {
+      "epoch": 0.14298295725030943,
+      "grad_norm": 11.079129739017336,
+      "learning_rate": 2e-05,
+      "loss": 1.0467,
+      "step": 6007
+    },
+    {
+      "epoch": 0.14300675997334095,
+      "grad_norm": 18.093353159328156,
+      "learning_rate": 2e-05,
+      "loss": 0.4399,
+      "step": 6008
+    },
+    {
+      "epoch": 0.14303056269637246,
+      "grad_norm": 4.9190925328303585,
+      "learning_rate": 2e-05,
+      "loss": 0.7813,
+      "step": 6009
+    },
+    {
+      "epoch": 0.14305436541940397,
+      "grad_norm": 2.7512032770062174,
+      "learning_rate": 2e-05,
+      "loss": 0.1658,
+      "step": 6010
+    },
+    {
+      "epoch": 0.14307816814243549,
+      "grad_norm": 5.759000377106228,
+      "learning_rate": 2e-05,
+      "loss": 0.3462,
+      "step": 6011
+    },
+    {
+      "epoch": 0.143101970865467,
+      "grad_norm": 5.944507463001131,
+      "learning_rate": 2e-05,
+      "loss": 0.3866,
+      "step": 6012
+    },
+    {
+      "epoch": 0.1431257735884985,
+      "grad_norm": 7.930152268483025,
+      "learning_rate": 2e-05,
+      "loss": 1.2328,
+      "step": 6013
+    },
+    {
+      "epoch": 0.14314957631153005,
+      "grad_norm": 2.242139990391248,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 6014
+    },
+    {
+      "epoch": 0.14317337903456157,
+      "grad_norm": 4.124080252109726,
+      "learning_rate": 2e-05,
+      "loss": 0.2503,
+      "step": 6015
+    },
+    {
+      "epoch": 0.14319718175759308,
+      "grad_norm": 6.699916203240939,
+      "learning_rate": 2e-05,
+      "loss": 0.2894,
+      "step": 6016
+    },
+    {
+      "epoch": 0.1432209844806246,
+      "grad_norm": 4.308627643542143,
+      "learning_rate": 2e-05,
+      "loss": 0.2576,
+      "step": 6017
+    },
+    {
+      "epoch": 0.1432447872036561,
+      "grad_norm": 7.170582412180755,
+      "learning_rate": 2e-05,
+      "loss": 0.8178,
+      "step": 6018
+    },
+    {
+      "epoch": 0.14326858992668762,
+      "grad_norm": 6.563663437318605,
+      "learning_rate": 2e-05,
+      "loss": 0.6587,
+      "step": 6019
+    },
+    {
+      "epoch": 0.14329239264971913,
+      "grad_norm": 4.382959951737055,
+      "learning_rate": 2e-05,
+      "loss": 0.8412,
+      "step": 6020
+    },
+    {
+      "epoch": 0.14331619537275064,
+      "grad_norm": 2.4087173574997367,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 6021
+    },
+    {
+      "epoch": 0.14333999809578216,
+      "grad_norm": 6.867085846035114,
+      "learning_rate": 2e-05,
+      "loss": 0.7806,
+      "step": 6022
+    },
+    {
+      "epoch": 0.14336380081881367,
+      "grad_norm": 5.507441268251244,
+      "learning_rate": 2e-05,
+      "loss": 0.225,
+      "step": 6023
+    },
+    {
+      "epoch": 0.14338760354184518,
+      "grad_norm": 5.666998904600161,
+      "learning_rate": 2e-05,
+      "loss": 0.366,
+      "step": 6024
+    },
+    {
+      "epoch": 0.1434114062648767,
+      "grad_norm": 5.1519562703975055,
+      "learning_rate": 2e-05,
+      "loss": 0.6878,
+      "step": 6025
+    },
+    {
+      "epoch": 0.1434352089879082,
+      "grad_norm": 6.213777594278343,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 6026
+    },
+    {
+      "epoch": 0.14345901171093972,
+      "grad_norm": 6.168680896019458,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 6027
+    },
+    {
+      "epoch": 0.14348281443397123,
+      "grad_norm": 3.366553318188396,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 6028
+    },
+    {
+      "epoch": 0.14350661715700277,
+      "grad_norm": 0.7179337606427807,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 6029
+    },
+    {
+      "epoch": 0.1435304198800343,
+      "grad_norm": 4.21531658801162,
+      "learning_rate": 2e-05,
+      "loss": 0.3121,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1435542226030658,
+      "grad_norm": 8.582067123774385,
+      "learning_rate": 2e-05,
+      "loss": 0.7062,
+      "step": 6031
+    },
+    {
+      "epoch": 0.1435780253260973,
+      "grad_norm": 4.779314768700341,
+      "learning_rate": 2e-05,
+      "loss": 0.4948,
+      "step": 6032
+    },
+    {
+      "epoch": 0.14360182804912883,
+      "grad_norm": 6.961198457952535,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 6033
+    },
+    {
+      "epoch": 0.14362563077216034,
+      "grad_norm": 2.2299459794221175,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 6034
+    },
+    {
+      "epoch": 0.14364943349519185,
+      "grad_norm": 2.852475155313252,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 6035
+    },
+    {
+      "epoch": 0.14367323621822337,
+      "grad_norm": 5.190987745144132,
+      "learning_rate": 2e-05,
+      "loss": 0.4919,
+      "step": 6036
+    },
+    {
+      "epoch": 0.14369703894125488,
+      "grad_norm": 1.8940558801010858,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 6037
+    },
+    {
+      "epoch": 0.1437208416642864,
+      "grad_norm": 6.092131937357021,
+      "learning_rate": 2e-05,
+      "loss": 0.9716,
+      "step": 6038
+    },
+    {
+      "epoch": 0.1437446443873179,
+      "grad_norm": 4.184216314291221,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 6039
+    },
+    {
+      "epoch": 0.14376844711034942,
+      "grad_norm": 2.966512479958972,
+      "learning_rate": 2e-05,
+      "loss": 0.1516,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14379224983338093,
+      "grad_norm": 2.4401013137629493,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 6041
+    },
+    {
+      "epoch": 0.14381605255641244,
+      "grad_norm": 5.5917923167404355,
+      "learning_rate": 2e-05,
+      "loss": 0.7378,
+      "step": 6042
+    },
+    {
+      "epoch": 0.14383985527944396,
+      "grad_norm": 4.825394565558984,
+      "learning_rate": 2e-05,
+      "loss": 1.1263,
+      "step": 6043
+    },
+    {
+      "epoch": 0.14386365800247547,
+      "grad_norm": 6.628410707857746,
+      "learning_rate": 2e-05,
+      "loss": 0.974,
+      "step": 6044
+    },
+    {
+      "epoch": 0.143887460725507,
+      "grad_norm": 1.5460683614841044,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 6045
+    },
+    {
+      "epoch": 0.14391126344853852,
+      "grad_norm": 6.324648770579446,
+      "learning_rate": 2e-05,
+      "loss": 0.2713,
+      "step": 6046
+    },
+    {
+      "epoch": 0.14393506617157004,
+      "grad_norm": 3.3210430344826194,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 6047
+    },
+    {
+      "epoch": 0.14395886889460155,
+      "grad_norm": 3.7739464755942245,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 6048
+    },
+    {
+      "epoch": 0.14398267161763306,
+      "grad_norm": 5.1955411373352955,
+      "learning_rate": 2e-05,
+      "loss": 0.5142,
+      "step": 6049
+    },
+    {
+      "epoch": 0.14400647434066458,
+      "grad_norm": 11.931553815112935,
+      "learning_rate": 2e-05,
+      "loss": 1.7427,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1440302770636961,
+      "grad_norm": 5.371105169190046,
+      "learning_rate": 2e-05,
+      "loss": 0.1818,
+      "step": 6051
+    },
+    {
+      "epoch": 0.1440540797867276,
+      "grad_norm": 3.9961585842128984,
+      "learning_rate": 2e-05,
+      "loss": 0.3224,
+      "step": 6052
+    },
+    {
+      "epoch": 0.14407788250975911,
+      "grad_norm": 3.8974936642789597,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 6053
+    },
+    {
+      "epoch": 0.14410168523279063,
+      "grad_norm": 4.508775718419119,
+      "learning_rate": 2e-05,
+      "loss": 0.389,
+      "step": 6054
+    },
+    {
+      "epoch": 0.14412548795582214,
+      "grad_norm": 6.354328461881277,
+      "learning_rate": 2e-05,
+      "loss": 0.8257,
+      "step": 6055
+    },
+    {
+      "epoch": 0.14414929067885365,
+      "grad_norm": 5.940069037927338,
+      "learning_rate": 2e-05,
+      "loss": 0.3115,
+      "step": 6056
+    },
+    {
+      "epoch": 0.14417309340188517,
+      "grad_norm": 8.0094549316681,
+      "learning_rate": 2e-05,
+      "loss": 0.3162,
+      "step": 6057
+    },
+    {
+      "epoch": 0.14419689612491668,
+      "grad_norm": 6.498139741396462,
+      "learning_rate": 2e-05,
+      "loss": 1.5578,
+      "step": 6058
+    },
+    {
+      "epoch": 0.1442206988479482,
+      "grad_norm": 8.946915665848945,
+      "learning_rate": 2e-05,
+      "loss": 0.9574,
+      "step": 6059
+    },
+    {
+      "epoch": 0.14424450157097973,
+      "grad_norm": 7.640784358658121,
+      "learning_rate": 2e-05,
+      "loss": 0.8819,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14426830429401125,
+      "grad_norm": 5.891217762104671,
+      "learning_rate": 2e-05,
+      "loss": 0.8275,
+      "step": 6061
+    },
+    {
+      "epoch": 0.14429210701704276,
+      "grad_norm": 2.711976709020013,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 6062
+    },
+    {
+      "epoch": 0.14431590974007427,
+      "grad_norm": 4.011343141516704,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 6063
+    },
+    {
+      "epoch": 0.14433971246310578,
+      "grad_norm": 3.0021937611275513,
+      "learning_rate": 2e-05,
+      "loss": 0.1533,
+      "step": 6064
+    },
+    {
+      "epoch": 0.1443635151861373,
+      "grad_norm": 1.5111879392335377,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 6065
+    },
+    {
+      "epoch": 0.1443873179091688,
+      "grad_norm": 10.360072882691115,
+      "learning_rate": 2e-05,
+      "loss": 1.2348,
+      "step": 6066
+    },
+    {
+      "epoch": 0.14441112063220032,
+      "grad_norm": 7.569243174314645,
+      "learning_rate": 2e-05,
+      "loss": 0.3805,
+      "step": 6067
+    },
+    {
+      "epoch": 0.14443492335523184,
+      "grad_norm": 9.628180722822881,
+      "learning_rate": 2e-05,
+      "loss": 1.7042,
+      "step": 6068
+    },
+    {
+      "epoch": 0.14445872607826335,
+      "grad_norm": 6.486539786299179,
+      "learning_rate": 2e-05,
+      "loss": 1.0072,
+      "step": 6069
+    },
+    {
+      "epoch": 0.14448252880129486,
+      "grad_norm": 10.116127173987675,
+      "learning_rate": 2e-05,
+      "loss": 1.1535,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14450633152432638,
+      "grad_norm": 7.830827818105773,
+      "learning_rate": 2e-05,
+      "loss": 0.8363,
+      "step": 6071
+    },
+    {
+      "epoch": 0.1445301342473579,
+      "grad_norm": 3.106217491828787,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 6072
+    },
+    {
+      "epoch": 0.1445539369703894,
+      "grad_norm": 3.6313108770776648,
+      "learning_rate": 2e-05,
+      "loss": 0.6587,
+      "step": 6073
+    },
+    {
+      "epoch": 0.14457773969342091,
+      "grad_norm": 3.6309148445796344,
+      "learning_rate": 2e-05,
+      "loss": 0.1483,
+      "step": 6074
+    },
+    {
+      "epoch": 0.14460154241645246,
+      "grad_norm": 5.262070206250083,
+      "learning_rate": 2e-05,
+      "loss": 0.3503,
+      "step": 6075
+    },
+    {
+      "epoch": 0.14462534513948397,
+      "grad_norm": 4.3596358453647035,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 6076
+    },
+    {
+      "epoch": 0.14464914786251548,
+      "grad_norm": 6.479172200083938,
+      "learning_rate": 2e-05,
+      "loss": 0.4186,
+      "step": 6077
+    },
+    {
+      "epoch": 0.144672950585547,
+      "grad_norm": 4.484201777134637,
+      "learning_rate": 2e-05,
+      "loss": 0.5998,
+      "step": 6078
+    },
+    {
+      "epoch": 0.1446967533085785,
+      "grad_norm": 3.7308463027990775,
+      "learning_rate": 2e-05,
+      "loss": 0.148,
+      "step": 6079
+    },
+    {
+      "epoch": 0.14472055603161002,
+      "grad_norm": 5.886774771341505,
+      "learning_rate": 2e-05,
+      "loss": 0.2269,
+      "step": 6080
+    },
+    {
+      "epoch": 0.14474435875464153,
+      "grad_norm": 3.459460689122756,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 6081
+    },
+    {
+      "epoch": 0.14476816147767305,
+      "grad_norm": 5.308435881160963,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 6082
+    },
+    {
+      "epoch": 0.14479196420070456,
+      "grad_norm": 2.8505331069717674,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 6083
+    },
+    {
+      "epoch": 0.14481576692373607,
+      "grad_norm": 7.35032239112482,
+      "learning_rate": 2e-05,
+      "loss": 1.0687,
+      "step": 6084
+    },
+    {
+      "epoch": 0.14483956964676759,
+      "grad_norm": 11.727856363342557,
+      "learning_rate": 2e-05,
+      "loss": 0.3849,
+      "step": 6085
+    },
+    {
+      "epoch": 0.1448633723697991,
+      "grad_norm": 13.391881064157232,
+      "learning_rate": 2e-05,
+      "loss": 0.7281,
+      "step": 6086
+    },
+    {
+      "epoch": 0.1448871750928306,
+      "grad_norm": 1.3588709810257955,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 6087
+    },
+    {
+      "epoch": 0.14491097781586212,
+      "grad_norm": 7.378498666218086,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 6088
+    },
+    {
+      "epoch": 0.14493478053889364,
+      "grad_norm": 5.998426797009314,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 6089
+    },
+    {
+      "epoch": 0.14495858326192518,
+      "grad_norm": 8.496945575851964,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1449823859849567,
+      "grad_norm": 1.6310140066832617,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 6091
+    },
+    {
+      "epoch": 0.1450061887079882,
+      "grad_norm": 3.042618650952541,
+      "learning_rate": 2e-05,
+      "loss": 0.1807,
+      "step": 6092
+    },
+    {
+      "epoch": 0.14502999143101972,
+      "grad_norm": 3.9084261571070495,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 6093
+    },
+    {
+      "epoch": 0.14505379415405123,
+      "grad_norm": 3.2696303642186653,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 6094
+    },
+    {
+      "epoch": 0.14507759687708274,
+      "grad_norm": 6.5408086134661145,
+      "learning_rate": 2e-05,
+      "loss": 1.133,
+      "step": 6095
+    },
+    {
+      "epoch": 0.14510139960011426,
+      "grad_norm": 5.082354625206435,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 6096
+    },
+    {
+      "epoch": 0.14512520232314577,
+      "grad_norm": 2.66702519457744,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 6097
+    },
+    {
+      "epoch": 0.14514900504617728,
+      "grad_norm": 3.0414427522714855,
+      "learning_rate": 2e-05,
+      "loss": 0.2433,
+      "step": 6098
+    },
+    {
+      "epoch": 0.1451728077692088,
+      "grad_norm": 1.4511930837217684,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 6099
+    },
+    {
+      "epoch": 0.1451966104922403,
+      "grad_norm": 6.5254312190114465,
+      "learning_rate": 2e-05,
+      "loss": 1.4556,
+      "step": 6100
+    },
+    {
+      "epoch": 0.14522041321527182,
+      "grad_norm": 2.449711115689058,
+      "learning_rate": 2e-05,
+      "loss": 0.1613,
+      "step": 6101
+    },
+    {
+      "epoch": 0.14524421593830333,
+      "grad_norm": 3.4008243174685644,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 6102
+    },
+    {
+      "epoch": 0.14526801866133485,
+      "grad_norm": 8.336587491374736,
+      "learning_rate": 2e-05,
+      "loss": 0.3588,
+      "step": 6103
+    },
+    {
+      "epoch": 0.14529182138436636,
+      "grad_norm": 6.517313827596119,
+      "learning_rate": 2e-05,
+      "loss": 0.395,
+      "step": 6104
+    },
+    {
+      "epoch": 0.14531562410739787,
+      "grad_norm": 4.0497416864313065,
+      "learning_rate": 2e-05,
+      "loss": 0.1374,
+      "step": 6105
+    },
+    {
+      "epoch": 0.1453394268304294,
+      "grad_norm": 4.270451029610299,
+      "learning_rate": 2e-05,
+      "loss": 0.7035,
+      "step": 6106
+    },
+    {
+      "epoch": 0.14536322955346093,
+      "grad_norm": 2.1338631113870727,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 6107
+    },
+    {
+      "epoch": 0.14538703227649244,
+      "grad_norm": 4.387148655960722,
+      "learning_rate": 2e-05,
+      "loss": 0.2119,
+      "step": 6108
+    },
+    {
+      "epoch": 0.14541083499952395,
+      "grad_norm": 4.199106690466582,
+      "learning_rate": 2e-05,
+      "loss": 0.5607,
+      "step": 6109
+    },
+    {
+      "epoch": 0.14543463772255547,
+      "grad_norm": 7.555658126074288,
+      "learning_rate": 2e-05,
+      "loss": 0.2794,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14545844044558698,
+      "grad_norm": 4.639995234882532,
+      "learning_rate": 2e-05,
+      "loss": 0.9191,
+      "step": 6111
+    },
+    {
+      "epoch": 0.1454822431686185,
+      "grad_norm": 5.751830418279817,
+      "learning_rate": 2e-05,
+      "loss": 0.8989,
+      "step": 6112
+    },
+    {
+      "epoch": 0.14550604589165,
+      "grad_norm": 9.664881159339233,
+      "learning_rate": 2e-05,
+      "loss": 0.7982,
+      "step": 6113
+    },
+    {
+      "epoch": 0.14552984861468152,
+      "grad_norm": 0.6876145648828427,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 6114
+    },
+    {
+      "epoch": 0.14555365133771303,
+      "grad_norm": 1.0194868695682269,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 6115
+    },
+    {
+      "epoch": 0.14557745406074454,
+      "grad_norm": 5.254157405121801,
+      "learning_rate": 2e-05,
+      "loss": 0.7199,
+      "step": 6116
+    },
+    {
+      "epoch": 0.14560125678377606,
+      "grad_norm": 7.982008606558851,
+      "learning_rate": 2e-05,
+      "loss": 0.4265,
+      "step": 6117
+    },
+    {
+      "epoch": 0.14562505950680757,
+      "grad_norm": 7.022403171716141,
+      "learning_rate": 2e-05,
+      "loss": 0.3685,
+      "step": 6118
+    },
+    {
+      "epoch": 0.14564886222983908,
+      "grad_norm": 13.838222843665912,
+      "learning_rate": 2e-05,
+      "loss": 1.5558,
+      "step": 6119
+    },
+    {
+      "epoch": 0.1456726649528706,
+      "grad_norm": 4.322683824673668,
+      "learning_rate": 2e-05,
+      "loss": 0.6121,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14569646767590214,
+      "grad_norm": 9.364883359354891,
+      "learning_rate": 2e-05,
+      "loss": 0.4376,
+      "step": 6121
+    },
+    {
+      "epoch": 0.14572027039893365,
+      "grad_norm": 5.632454947225024,
+      "learning_rate": 2e-05,
+      "loss": 0.6111,
+      "step": 6122
+    },
+    {
+      "epoch": 0.14574407312196516,
+      "grad_norm": 4.2202769600778085,
+      "learning_rate": 2e-05,
+      "loss": 0.4055,
+      "step": 6123
+    },
+    {
+      "epoch": 0.14576787584499667,
+      "grad_norm": 3.9737398724811057,
+      "learning_rate": 2e-05,
+      "loss": 0.4881,
+      "step": 6124
+    },
+    {
+      "epoch": 0.1457916785680282,
+      "grad_norm": 10.149975185234897,
+      "learning_rate": 2e-05,
+      "loss": 1.2469,
+      "step": 6125
+    },
+    {
+      "epoch": 0.1458154812910597,
+      "grad_norm": 7.502101892459646,
+      "learning_rate": 2e-05,
+      "loss": 0.5143,
+      "step": 6126
+    },
+    {
+      "epoch": 0.14583928401409121,
+      "grad_norm": 2.7326397031150194,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 6127
+    },
+    {
+      "epoch": 0.14586308673712273,
+      "grad_norm": 7.975338077669337,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 6128
+    },
+    {
+      "epoch": 0.14588688946015424,
+      "grad_norm": 6.405646960753142,
+      "learning_rate": 2e-05,
+      "loss": 1.2759,
+      "step": 6129
+    },
+    {
+      "epoch": 0.14591069218318575,
+      "grad_norm": 1.0255510018950766,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 6130
+    },
+    {
+      "epoch": 0.14593449490621727,
+      "grad_norm": 4.1703643258326775,
+      "learning_rate": 2e-05,
+      "loss": 0.2816,
+      "step": 6131
+    },
+    {
+      "epoch": 0.14595829762924878,
+      "grad_norm": 3.628787396404409,
+      "learning_rate": 2e-05,
+      "loss": 0.5001,
+      "step": 6132
+    },
+    {
+      "epoch": 0.1459821003522803,
+      "grad_norm": 5.029786637862805,
+      "learning_rate": 2e-05,
+      "loss": 0.5787,
+      "step": 6133
+    },
+    {
+      "epoch": 0.1460059030753118,
+      "grad_norm": 10.504361699953249,
+      "learning_rate": 2e-05,
+      "loss": 1.7669,
+      "step": 6134
+    },
+    {
+      "epoch": 0.14602970579834332,
+      "grad_norm": 4.985909263297584,
+      "learning_rate": 2e-05,
+      "loss": 0.3445,
+      "step": 6135
+    },
+    {
+      "epoch": 0.14605350852137486,
+      "grad_norm": 5.356334708819516,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 6136
+    },
+    {
+      "epoch": 0.14607731124440637,
+      "grad_norm": 15.417356097333789,
+      "learning_rate": 2e-05,
+      "loss": 1.0307,
+      "step": 6137
+    },
+    {
+      "epoch": 0.14610111396743788,
+      "grad_norm": 8.069822584787067,
+      "learning_rate": 2e-05,
+      "loss": 1.3965,
+      "step": 6138
+    },
+    {
+      "epoch": 0.1461249166904694,
+      "grad_norm": 4.677998743970654,
+      "learning_rate": 2e-05,
+      "loss": 0.7561,
+      "step": 6139
+    },
+    {
+      "epoch": 0.1461487194135009,
+      "grad_norm": 6.851799675825167,
+      "learning_rate": 2e-05,
+      "loss": 0.7307,
+      "step": 6140
+    },
+    {
+      "epoch": 0.14617252213653242,
+      "grad_norm": 8.447492625835238,
+      "learning_rate": 2e-05,
+      "loss": 0.3463,
+      "step": 6141
+    },
+    {
+      "epoch": 0.14619632485956394,
+      "grad_norm": 7.0565493945162565,
+      "learning_rate": 2e-05,
+      "loss": 1.0949,
+      "step": 6142
+    },
+    {
+      "epoch": 0.14622012758259545,
+      "grad_norm": 6.480445215552085,
+      "learning_rate": 2e-05,
+      "loss": 1.1939,
+      "step": 6143
+    },
+    {
+      "epoch": 0.14624393030562696,
+      "grad_norm": 4.034694072385806,
+      "learning_rate": 2e-05,
+      "loss": 0.8161,
+      "step": 6144
+    },
+    {
+      "epoch": 0.14626773302865848,
+      "grad_norm": 3.511454937580964,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 6145
+    },
+    {
+      "epoch": 0.14629153575169,
+      "grad_norm": 5.855977800090585,
+      "learning_rate": 2e-05,
+      "loss": 0.3231,
+      "step": 6146
+    },
+    {
+      "epoch": 0.1463153384747215,
+      "grad_norm": 4.39574333754101,
+      "learning_rate": 2e-05,
+      "loss": 0.5804,
+      "step": 6147
+    },
+    {
+      "epoch": 0.14633914119775301,
+      "grad_norm": 5.782651825935571,
+      "learning_rate": 2e-05,
+      "loss": 0.8881,
+      "step": 6148
+    },
+    {
+      "epoch": 0.14636294392078453,
+      "grad_norm": 7.500708394799067,
+      "learning_rate": 2e-05,
+      "loss": 1.4473,
+      "step": 6149
+    },
+    {
+      "epoch": 0.14638674664381604,
+      "grad_norm": 3.799943512215976,
+      "learning_rate": 2e-05,
+      "loss": 0.4562,
+      "step": 6150
+    },
+    {
+      "epoch": 0.14641054936684758,
+      "grad_norm": 10.624290005187756,
+      "learning_rate": 2e-05,
+      "loss": 1.9029,
+      "step": 6151
+    },
+    {
+      "epoch": 0.1464343520898791,
+      "grad_norm": 4.222046451585235,
+      "learning_rate": 2e-05,
+      "loss": 0.2243,
+      "step": 6152
+    },
+    {
+      "epoch": 0.1464581548129106,
+      "grad_norm": 3.1492023996900986,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 6153
+    },
+    {
+      "epoch": 0.14648195753594212,
+      "grad_norm": 2.105439790854197,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 6154
+    },
+    {
+      "epoch": 0.14650576025897363,
+      "grad_norm": 3.8838846878007978,
+      "learning_rate": 2e-05,
+      "loss": 0.1725,
+      "step": 6155
+    },
+    {
+      "epoch": 0.14652956298200515,
+      "grad_norm": 9.233600107637542,
+      "learning_rate": 2e-05,
+      "loss": 0.4581,
+      "step": 6156
+    },
+    {
+      "epoch": 0.14655336570503666,
+      "grad_norm": 5.17754908946252,
+      "learning_rate": 2e-05,
+      "loss": 0.5354,
+      "step": 6157
+    },
+    {
+      "epoch": 0.14657716842806817,
+      "grad_norm": 5.061357348747608,
+      "learning_rate": 2e-05,
+      "loss": 0.1937,
+      "step": 6158
+    },
+    {
+      "epoch": 0.14660097115109968,
+      "grad_norm": 6.201233794346515,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 6159
+    },
+    {
+      "epoch": 0.1466247738741312,
+      "grad_norm": 2.9191564813473425,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1466485765971627,
+      "grad_norm": 2.256521592739078,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 6161
+    },
+    {
+      "epoch": 0.14667237932019422,
+      "grad_norm": 9.536603348726157,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 6162
+    },
+    {
+      "epoch": 0.14669618204322574,
+      "grad_norm": 6.740616972853534,
+      "learning_rate": 2e-05,
+      "loss": 0.3602,
+      "step": 6163
+    },
+    {
+      "epoch": 0.14671998476625725,
+      "grad_norm": 6.991522008992981,
+      "learning_rate": 2e-05,
+      "loss": 0.9147,
+      "step": 6164
+    },
+    {
+      "epoch": 0.14674378748928876,
+      "grad_norm": 5.876770055864849,
+      "learning_rate": 2e-05,
+      "loss": 0.7164,
+      "step": 6165
+    },
+    {
+      "epoch": 0.14676759021232028,
+      "grad_norm": 4.696196429768236,
+      "learning_rate": 2e-05,
+      "loss": 0.3673,
+      "step": 6166
+    },
+    {
+      "epoch": 0.14679139293535182,
+      "grad_norm": 2.679426229619155,
+      "learning_rate": 2e-05,
+      "loss": 0.1589,
+      "step": 6167
+    },
+    {
+      "epoch": 0.14681519565838333,
+      "grad_norm": 6.3901219739995225,
+      "learning_rate": 2e-05,
+      "loss": 0.9344,
+      "step": 6168
+    },
+    {
+      "epoch": 0.14683899838141484,
+      "grad_norm": 8.369756238470304,
+      "learning_rate": 2e-05,
+      "loss": 0.3811,
+      "step": 6169
+    },
+    {
+      "epoch": 0.14686280110444636,
+      "grad_norm": 3.2217385046924494,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 6170
+    },
+    {
+      "epoch": 0.14688660382747787,
+      "grad_norm": 4.602769439664888,
+      "learning_rate": 2e-05,
+      "loss": 0.6612,
+      "step": 6171
+    },
+    {
+      "epoch": 0.14691040655050938,
+      "grad_norm": 4.56982575527259,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 6172
+    },
+    {
+      "epoch": 0.1469342092735409,
+      "grad_norm": 2.7481501567147757,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 6173
+    },
+    {
+      "epoch": 0.1469580119965724,
+      "grad_norm": 7.2539257908665205,
+      "learning_rate": 2e-05,
+      "loss": 0.44,
+      "step": 6174
+    },
+    {
+      "epoch": 0.14698181471960392,
+      "grad_norm": 5.039944029614365,
+      "learning_rate": 2e-05,
+      "loss": 0.7034,
+      "step": 6175
+    },
+    {
+      "epoch": 0.14700561744263543,
+      "grad_norm": 6.997709800944312,
+      "learning_rate": 2e-05,
+      "loss": 0.924,
+      "step": 6176
+    },
+    {
+      "epoch": 0.14702942016566695,
+      "grad_norm": 5.520834038600986,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 6177
+    },
+    {
+      "epoch": 0.14705322288869846,
+      "grad_norm": 2.6625595386885443,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 6178
+    },
+    {
+      "epoch": 0.14707702561172997,
+      "grad_norm": 2.880853868552494,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 6179
+    },
+    {
+      "epoch": 0.14710082833476149,
+      "grad_norm": 1.9018105153558487,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 6180
+    },
+    {
+      "epoch": 0.147124631057793,
+      "grad_norm": 8.822279389425711,
+      "learning_rate": 2e-05,
+      "loss": 1.434,
+      "step": 6181
+    },
+    {
+      "epoch": 0.14714843378082454,
+      "grad_norm": 5.7556158480188975,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 6182
+    },
+    {
+      "epoch": 0.14717223650385605,
+      "grad_norm": 5.341932974667055,
+      "learning_rate": 2e-05,
+      "loss": 0.3253,
+      "step": 6183
+    },
+    {
+      "epoch": 0.14719603922688757,
+      "grad_norm": 3.884592747725231,
+      "learning_rate": 2e-05,
+      "loss": 0.2413,
+      "step": 6184
+    },
+    {
+      "epoch": 0.14721984194991908,
+      "grad_norm": 11.66692281373771,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 6185
+    },
+    {
+      "epoch": 0.1472436446729506,
+      "grad_norm": 7.928506952044813,
+      "learning_rate": 2e-05,
+      "loss": 0.4016,
+      "step": 6186
+    },
+    {
+      "epoch": 0.1472674473959821,
+      "grad_norm": 2.1167080445146196,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 6187
+    },
+    {
+      "epoch": 0.14729125011901362,
+      "grad_norm": 6.237118663823149,
+      "learning_rate": 2e-05,
+      "loss": 0.2863,
+      "step": 6188
+    },
+    {
+      "epoch": 0.14731505284204513,
+      "grad_norm": 7.010367994295354,
+      "learning_rate": 2e-05,
+      "loss": 0.5157,
+      "step": 6189
+    },
+    {
+      "epoch": 0.14733885556507664,
+      "grad_norm": 5.3282812268363235,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 6190
+    },
+    {
+      "epoch": 0.14736265828810816,
+      "grad_norm": 6.908938803839362,
+      "learning_rate": 2e-05,
+      "loss": 0.3678,
+      "step": 6191
+    },
+    {
+      "epoch": 0.14738646101113967,
+      "grad_norm": 2.6555316246417315,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 6192
+    },
+    {
+      "epoch": 0.14741026373417118,
+      "grad_norm": 3.995711896982722,
+      "learning_rate": 2e-05,
+      "loss": 0.7871,
+      "step": 6193
+    },
+    {
+      "epoch": 0.1474340664572027,
+      "grad_norm": 5.993133665776208,
+      "learning_rate": 2e-05,
+      "loss": 0.1965,
+      "step": 6194
+    },
+    {
+      "epoch": 0.1474578691802342,
+      "grad_norm": 3.8275908820279505,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 6195
+    },
+    {
+      "epoch": 0.14748167190326572,
+      "grad_norm": 2.979634803500994,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 6196
+    },
+    {
+      "epoch": 0.14750547462629726,
+      "grad_norm": 2.3775664764364537,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 6197
+    },
+    {
+      "epoch": 0.14752927734932877,
+      "grad_norm": 3.0600336053638117,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 6198
+    },
+    {
+      "epoch": 0.1475530800723603,
+      "grad_norm": 3.239974522778443,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 6199
+    },
+    {
+      "epoch": 0.1475768827953918,
+      "grad_norm": 2.6743938465201937,
+      "learning_rate": 2e-05,
+      "loss": 0.2596,
+      "step": 6200
+    },
+    {
+      "epoch": 0.1476006855184233,
+      "grad_norm": 4.402610321056492,
+      "learning_rate": 2e-05,
+      "loss": 0.3773,
+      "step": 6201
+    },
+    {
+      "epoch": 0.14762448824145483,
+      "grad_norm": 6.069314041739251,
+      "learning_rate": 2e-05,
+      "loss": 0.3877,
+      "step": 6202
+    },
+    {
+      "epoch": 0.14764829096448634,
+      "grad_norm": 7.602687128597831,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 6203
+    },
+    {
+      "epoch": 0.14767209368751785,
+      "grad_norm": 7.41243870849958,
+      "learning_rate": 2e-05,
+      "loss": 1.2321,
+      "step": 6204
+    },
+    {
+      "epoch": 0.14769589641054937,
+      "grad_norm": 2.5046009434744403,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 6205
+    },
+    {
+      "epoch": 0.14771969913358088,
+      "grad_norm": 1.678182583215553,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 6206
+    },
+    {
+      "epoch": 0.1477435018566124,
+      "grad_norm": 3.8000775631290016,
+      "learning_rate": 2e-05,
+      "loss": 0.1828,
+      "step": 6207
+    },
+    {
+      "epoch": 0.1477673045796439,
+      "grad_norm": 3.5743606123130256,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 6208
+    },
+    {
+      "epoch": 0.14779110730267542,
+      "grad_norm": 3.3157823389149024,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 6209
+    },
+    {
+      "epoch": 0.14781491002570693,
+      "grad_norm": 6.166869341173732,
+      "learning_rate": 2e-05,
+      "loss": 0.3578,
+      "step": 6210
+    },
+    {
+      "epoch": 0.14783871274873844,
+      "grad_norm": 4.045594547721932,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 6211
+    },
+    {
+      "epoch": 0.14786251547176998,
+      "grad_norm": 5.1901965989918795,
+      "learning_rate": 2e-05,
+      "loss": 0.4336,
+      "step": 6212
+    },
+    {
+      "epoch": 0.1478863181948015,
+      "grad_norm": 3.386215489462334,
+      "learning_rate": 2e-05,
+      "loss": 0.1964,
+      "step": 6213
+    },
+    {
+      "epoch": 0.147910120917833,
+      "grad_norm": 1.2494578655396067,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 6214
+    },
+    {
+      "epoch": 0.14793392364086452,
+      "grad_norm": 3.130560898673312,
+      "learning_rate": 2e-05,
+      "loss": 0.2426,
+      "step": 6215
+    },
+    {
+      "epoch": 0.14795772636389604,
+      "grad_norm": 2.8174438825828245,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 6216
+    },
+    {
+      "epoch": 0.14798152908692755,
+      "grad_norm": 5.372163638965092,
+      "learning_rate": 2e-05,
+      "loss": 0.5909,
+      "step": 6217
+    },
+    {
+      "epoch": 0.14800533180995906,
+      "grad_norm": 5.7399828655902745,
+      "learning_rate": 2e-05,
+      "loss": 0.5017,
+      "step": 6218
+    },
+    {
+      "epoch": 0.14802913453299058,
+      "grad_norm": 7.999714226274758,
+      "learning_rate": 2e-05,
+      "loss": 0.7325,
+      "step": 6219
+    },
+    {
+      "epoch": 0.1480529372560221,
+      "grad_norm": 6.23492159398145,
+      "learning_rate": 2e-05,
+      "loss": 0.5561,
+      "step": 6220
+    },
+    {
+      "epoch": 0.1480767399790536,
+      "grad_norm": 2.1769641297682467,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 6221
+    },
+    {
+      "epoch": 0.14810054270208511,
+      "grad_norm": 7.34130846257066,
+      "learning_rate": 2e-05,
+      "loss": 0.4446,
+      "step": 6222
+    },
+    {
+      "epoch": 0.14812434542511663,
+      "grad_norm": 2.8476516464739534,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 6223
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 3.7555310134615483,
+      "learning_rate": 2e-05,
+      "loss": 0.5788,
+      "step": 6224
+    },
+    {
+      "epoch": 0.14817195087117965,
+      "grad_norm": 7.2780182903363375,
+      "learning_rate": 2e-05,
+      "loss": 1.0359,
+      "step": 6225
+    },
+    {
+      "epoch": 0.14819575359421117,
+      "grad_norm": 5.992626202753531,
+      "learning_rate": 2e-05,
+      "loss": 0.3565,
+      "step": 6226
+    },
+    {
+      "epoch": 0.14821955631724268,
+      "grad_norm": 2.843138239938887,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 6227
+    },
+    {
+      "epoch": 0.14824335904027422,
+      "grad_norm": 1.8172586892830778,
+      "learning_rate": 2e-05,
+      "loss": 0.1504,
+      "step": 6228
+    },
+    {
+      "epoch": 0.14826716176330573,
+      "grad_norm": 5.020826381340244,
+      "learning_rate": 2e-05,
+      "loss": 0.2798,
+      "step": 6229
+    },
+    {
+      "epoch": 0.14829096448633725,
+      "grad_norm": 4.859600416161067,
+      "learning_rate": 2e-05,
+      "loss": 0.954,
+      "step": 6230
+    },
+    {
+      "epoch": 0.14831476720936876,
+      "grad_norm": 3.1172427320275253,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 6231
+    },
+    {
+      "epoch": 0.14833856993240027,
+      "grad_norm": 4.568063849369712,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 6232
+    },
+    {
+      "epoch": 0.14836237265543178,
+      "grad_norm": 4.101134023035856,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 6233
+    },
+    {
+      "epoch": 0.1483861753784633,
+      "grad_norm": 5.7719740510874376,
+      "learning_rate": 2e-05,
+      "loss": 0.7971,
+      "step": 6234
+    },
+    {
+      "epoch": 0.1484099781014948,
+      "grad_norm": 5.145210667939188,
+      "learning_rate": 2e-05,
+      "loss": 0.7863,
+      "step": 6235
+    },
+    {
+      "epoch": 0.14843378082452632,
+      "grad_norm": 5.877613296287649,
+      "learning_rate": 2e-05,
+      "loss": 0.3104,
+      "step": 6236
+    },
+    {
+      "epoch": 0.14845758354755784,
+      "grad_norm": 7.547391330720704,
+      "learning_rate": 2e-05,
+      "loss": 1.4137,
+      "step": 6237
+    },
+    {
+      "epoch": 0.14848138627058935,
+      "grad_norm": 3.936919599500152,
+      "learning_rate": 2e-05,
+      "loss": 0.2633,
+      "step": 6238
+    },
+    {
+      "epoch": 0.14850518899362086,
+      "grad_norm": 7.6861720088485335,
+      "learning_rate": 2e-05,
+      "loss": 1.7479,
+      "step": 6239
+    },
+    {
+      "epoch": 0.14852899171665238,
+      "grad_norm": 10.155584891798089,
+      "learning_rate": 2e-05,
+      "loss": 0.9073,
+      "step": 6240
+    },
+    {
+      "epoch": 0.1485527944396839,
+      "grad_norm": 2.342317312251072,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 6241
+    },
+    {
+      "epoch": 0.1485765971627154,
+      "grad_norm": 7.489595587326252,
+      "learning_rate": 2e-05,
+      "loss": 0.3482,
+      "step": 6242
+    },
+    {
+      "epoch": 0.14860039988574694,
+      "grad_norm": 3.573503492232065,
+      "learning_rate": 2e-05,
+      "loss": 0.4804,
+      "step": 6243
+    },
+    {
+      "epoch": 0.14862420260877846,
+      "grad_norm": 7.143900758739681,
+      "learning_rate": 2e-05,
+      "loss": 0.3628,
+      "step": 6244
+    },
+    {
+      "epoch": 0.14864800533180997,
+      "grad_norm": 7.219315057292773,
+      "learning_rate": 2e-05,
+      "loss": 0.3392,
+      "step": 6245
+    },
+    {
+      "epoch": 0.14867180805484148,
+      "grad_norm": 8.129629718624845,
+      "learning_rate": 2e-05,
+      "loss": 1.1298,
+      "step": 6246
+    },
+    {
+      "epoch": 0.148695610777873,
+      "grad_norm": 7.787213762611517,
+      "learning_rate": 2e-05,
+      "loss": 0.6512,
+      "step": 6247
+    },
+    {
+      "epoch": 0.1487194135009045,
+      "grad_norm": 6.035108218044101,
+      "learning_rate": 2e-05,
+      "loss": 0.7037,
+      "step": 6248
+    },
+    {
+      "epoch": 0.14874321622393602,
+      "grad_norm": 8.532727384776901,
+      "learning_rate": 2e-05,
+      "loss": 1.1935,
+      "step": 6249
+    },
+    {
+      "epoch": 0.14876701894696753,
+      "grad_norm": 7.01100444133837,
+      "learning_rate": 2e-05,
+      "loss": 0.3569,
+      "step": 6250
+    },
+    {
+      "epoch": 0.14879082166999905,
+      "grad_norm": 6.764925550060096,
+      "learning_rate": 2e-05,
+      "loss": 0.2297,
+      "step": 6251
+    },
+    {
+      "epoch": 0.14881462439303056,
+      "grad_norm": 1.2797065369065086,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 6252
+    },
+    {
+      "epoch": 0.14883842711606207,
+      "grad_norm": 5.68766686222203,
+      "learning_rate": 2e-05,
+      "loss": 0.5705,
+      "step": 6253
+    },
+    {
+      "epoch": 0.14886222983909358,
+      "grad_norm": 2.500724222588008,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 6254
+    },
+    {
+      "epoch": 0.1488860325621251,
+      "grad_norm": 5.678776153819123,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 6255
+    },
+    {
+      "epoch": 0.1489098352851566,
+      "grad_norm": 3.2308911413792822,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 6256
+    },
+    {
+      "epoch": 0.14893363800818812,
+      "grad_norm": 5.894819146659045,
+      "learning_rate": 2e-05,
+      "loss": 0.6326,
+      "step": 6257
+    },
+    {
+      "epoch": 0.14895744073121966,
+      "grad_norm": 8.199814466515669,
+      "learning_rate": 2e-05,
+      "loss": 1.2343,
+      "step": 6258
+    },
+    {
+      "epoch": 0.14898124345425118,
+      "grad_norm": 4.2961640597545765,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 6259
+    },
+    {
+      "epoch": 0.1490050461772827,
+      "grad_norm": 2.0666998948255695,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 6260
+    },
+    {
+      "epoch": 0.1490288489003142,
+      "grad_norm": 4.158730030689401,
+      "learning_rate": 2e-05,
+      "loss": 0.301,
+      "step": 6261
+    },
+    {
+      "epoch": 0.14905265162334572,
+      "grad_norm": 7.583262214308366,
+      "learning_rate": 2e-05,
+      "loss": 1.219,
+      "step": 6262
+    },
+    {
+      "epoch": 0.14907645434637723,
+      "grad_norm": 6.330447762236417,
+      "learning_rate": 2e-05,
+      "loss": 0.2762,
+      "step": 6263
+    },
+    {
+      "epoch": 0.14910025706940874,
+      "grad_norm": 2.5653473101876463,
+      "learning_rate": 2e-05,
+      "loss": 0.2201,
+      "step": 6264
+    },
+    {
+      "epoch": 0.14912405979244026,
+      "grad_norm": 5.444307834120276,
+      "learning_rate": 2e-05,
+      "loss": 0.2801,
+      "step": 6265
+    },
+    {
+      "epoch": 0.14914786251547177,
+      "grad_norm": 10.506216749195197,
+      "learning_rate": 2e-05,
+      "loss": 0.8852,
+      "step": 6266
+    },
+    {
+      "epoch": 0.14917166523850328,
+      "grad_norm": 3.681050586754618,
+      "learning_rate": 2e-05,
+      "loss": 0.4473,
+      "step": 6267
+    },
+    {
+      "epoch": 0.1491954679615348,
+      "grad_norm": 5.062440838690365,
+      "learning_rate": 2e-05,
+      "loss": 1.0881,
+      "step": 6268
+    },
+    {
+      "epoch": 0.1492192706845663,
+      "grad_norm": 3.3043184618613233,
+      "learning_rate": 2e-05,
+      "loss": 0.2331,
+      "step": 6269
+    },
+    {
+      "epoch": 0.14924307340759782,
+      "grad_norm": 4.749477937151289,
+      "learning_rate": 2e-05,
+      "loss": 0.2746,
+      "step": 6270
+    },
+    {
+      "epoch": 0.14926687613062933,
+      "grad_norm": 7.313539758768599,
+      "learning_rate": 2e-05,
+      "loss": 0.2569,
+      "step": 6271
+    },
+    {
+      "epoch": 0.14929067885366085,
+      "grad_norm": 1.6170143652380122,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 6272
+    },
+    {
+      "epoch": 0.1493144815766924,
+      "grad_norm": 4.367715341190511,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 6273
+    },
+    {
+      "epoch": 0.1493382842997239,
+      "grad_norm": 4.888780825074628,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 6274
+    },
+    {
+      "epoch": 0.1493620870227554,
+      "grad_norm": 5.139951434614395,
+      "learning_rate": 2e-05,
+      "loss": 0.2085,
+      "step": 6275
+    },
+    {
+      "epoch": 0.14938588974578693,
+      "grad_norm": 3.976229307652182,
+      "learning_rate": 2e-05,
+      "loss": 0.4306,
+      "step": 6276
+    },
+    {
+      "epoch": 0.14940969246881844,
+      "grad_norm": 6.738822545882951,
+      "learning_rate": 2e-05,
+      "loss": 0.4522,
+      "step": 6277
+    },
+    {
+      "epoch": 0.14943349519184995,
+      "grad_norm": 5.317093929696273,
+      "learning_rate": 2e-05,
+      "loss": 0.2435,
+      "step": 6278
+    },
+    {
+      "epoch": 0.14945729791488147,
+      "grad_norm": 2.4616477525062956,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 6279
+    },
+    {
+      "epoch": 0.14948110063791298,
+      "grad_norm": 7.405723450875103,
+      "learning_rate": 2e-05,
+      "loss": 0.3973,
+      "step": 6280
+    },
+    {
+      "epoch": 0.1495049033609445,
+      "grad_norm": 9.358285319746475,
+      "learning_rate": 2e-05,
+      "loss": 1.0991,
+      "step": 6281
+    },
+    {
+      "epoch": 0.149528706083976,
+      "grad_norm": 2.4481029295784866,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 6282
+    },
+    {
+      "epoch": 0.14955250880700752,
+      "grad_norm": 6.077110433307556,
+      "learning_rate": 2e-05,
+      "loss": 1.3089,
+      "step": 6283
+    },
+    {
+      "epoch": 0.14957631153003903,
+      "grad_norm": 6.342549042165264,
+      "learning_rate": 2e-05,
+      "loss": 1.0127,
+      "step": 6284
+    },
+    {
+      "epoch": 0.14960011425307054,
+      "grad_norm": 6.568041091074661,
+      "learning_rate": 2e-05,
+      "loss": 0.7368,
+      "step": 6285
+    },
+    {
+      "epoch": 0.14962391697610206,
+      "grad_norm": 5.193061814868811,
+      "learning_rate": 2e-05,
+      "loss": 0.8782,
+      "step": 6286
+    },
+    {
+      "epoch": 0.14964771969913357,
+      "grad_norm": 8.640714293817798,
+      "learning_rate": 2e-05,
+      "loss": 0.5551,
+      "step": 6287
+    },
+    {
+      "epoch": 0.14967152242216508,
+      "grad_norm": 8.529908494088899,
+      "learning_rate": 2e-05,
+      "loss": 0.3254,
+      "step": 6288
+    },
+    {
+      "epoch": 0.14969532514519662,
+      "grad_norm": 1.9681457704651482,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 6289
+    },
+    {
+      "epoch": 0.14971912786822814,
+      "grad_norm": 11.960134963458145,
+      "learning_rate": 2e-05,
+      "loss": 0.9814,
+      "step": 6290
+    },
+    {
+      "epoch": 0.14974293059125965,
+      "grad_norm": 6.934619035027712,
+      "learning_rate": 2e-05,
+      "loss": 0.4643,
+      "step": 6291
+    },
+    {
+      "epoch": 0.14976673331429116,
+      "grad_norm": 2.5356404817929845,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 6292
+    },
+    {
+      "epoch": 0.14979053603732267,
+      "grad_norm": 5.290837119070645,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 6293
+    },
+    {
+      "epoch": 0.1498143387603542,
+      "grad_norm": 5.02800421044928,
+      "learning_rate": 2e-05,
+      "loss": 0.3056,
+      "step": 6294
+    },
+    {
+      "epoch": 0.1498381414833857,
+      "grad_norm": 2.212644416631184,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 6295
+    },
+    {
+      "epoch": 0.1498619442064172,
+      "grad_norm": 2.2374548315654654,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 6296
+    },
+    {
+      "epoch": 0.14988574692944873,
+      "grad_norm": 2.444740827318729,
+      "learning_rate": 2e-05,
+      "loss": 0.237,
+      "step": 6297
+    },
+    {
+      "epoch": 0.14990954965248024,
+      "grad_norm": 4.7475870822587325,
+      "learning_rate": 2e-05,
+      "loss": 0.3271,
+      "step": 6298
+    },
+    {
+      "epoch": 0.14993335237551175,
+      "grad_norm": 6.120927828110601,
+      "learning_rate": 2e-05,
+      "loss": 0.6991,
+      "step": 6299
+    },
+    {
+      "epoch": 0.14995715509854327,
+      "grad_norm": 1.709646066441277,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 6300
+    },
+    {
+      "epoch": 0.14998095782157478,
+      "grad_norm": 5.146335064584055,
+      "learning_rate": 2e-05,
+      "loss": 0.9914,
+      "step": 6301
+    },
+    {
+      "epoch": 0.1500047605446063,
+      "grad_norm": 6.09664617705841,
+      "learning_rate": 2e-05,
+      "loss": 0.8672,
+      "step": 6302
+    },
+    {
+      "epoch": 0.1500285632676378,
+      "grad_norm": 6.832306553268119,
+      "learning_rate": 2e-05,
+      "loss": 1.0345,
+      "step": 6303
+    },
+    {
+      "epoch": 0.15005236599066935,
+      "grad_norm": 6.2027196150275605,
+      "learning_rate": 2e-05,
+      "loss": 1.1721,
+      "step": 6304
+    },
+    {
+      "epoch": 0.15007616871370086,
+      "grad_norm": 7.191894180185598,
+      "learning_rate": 2e-05,
+      "loss": 0.9391,
+      "step": 6305
+    },
+    {
+      "epoch": 0.15009997143673237,
+      "grad_norm": 4.284736553468097,
+      "learning_rate": 2e-05,
+      "loss": 0.5914,
+      "step": 6306
+    },
+    {
+      "epoch": 0.15012377415976388,
+      "grad_norm": 5.207040142470136,
+      "learning_rate": 2e-05,
+      "loss": 0.8622,
+      "step": 6307
+    },
+    {
+      "epoch": 0.1501475768827954,
+      "grad_norm": 2.0897228403096215,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 6308
+    },
+    {
+      "epoch": 0.1501713796058269,
+      "grad_norm": 6.506628976192823,
+      "learning_rate": 2e-05,
+      "loss": 0.3412,
+      "step": 6309
+    },
+    {
+      "epoch": 0.15019518232885842,
+      "grad_norm": 7.3001189156909705,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 6310
+    },
+    {
+      "epoch": 0.15021898505188994,
+      "grad_norm": 10.535084446393943,
+      "learning_rate": 2e-05,
+      "loss": 0.3707,
+      "step": 6311
+    },
+    {
+      "epoch": 0.15024278777492145,
+      "grad_norm": 2.8429926471868425,
+      "learning_rate": 2e-05,
+      "loss": 0.176,
+      "step": 6312
+    },
+    {
+      "epoch": 0.15026659049795296,
+      "grad_norm": 6.681694071722273,
+      "learning_rate": 2e-05,
+      "loss": 0.3213,
+      "step": 6313
+    },
+    {
+      "epoch": 0.15029039322098448,
+      "grad_norm": 4.224849443367363,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 6314
+    },
+    {
+      "epoch": 0.150314195944016,
+      "grad_norm": 5.3550705511709324,
+      "learning_rate": 2e-05,
+      "loss": 0.8512,
+      "step": 6315
+    },
+    {
+      "epoch": 0.1503379986670475,
+      "grad_norm": 7.915239007723838,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 6316
+    },
+    {
+      "epoch": 0.15036180139007901,
+      "grad_norm": 1.689996832508436,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 6317
+    },
+    {
+      "epoch": 0.15038560411311053,
+      "grad_norm": 3.0417745473474853,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 6318
+    },
+    {
+      "epoch": 0.15040940683614207,
+      "grad_norm": 7.624108101839649,
+      "learning_rate": 2e-05,
+      "loss": 0.3684,
+      "step": 6319
+    },
+    {
+      "epoch": 0.15043320955917358,
+      "grad_norm": 5.7744907444744795,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 6320
+    },
+    {
+      "epoch": 0.1504570122822051,
+      "grad_norm": 8.757597686067832,
+      "learning_rate": 2e-05,
+      "loss": 1.3275,
+      "step": 6321
+    },
+    {
+      "epoch": 0.1504808150052366,
+      "grad_norm": 9.005049411227748,
+      "learning_rate": 2e-05,
+      "loss": 0.2571,
+      "step": 6322
+    },
+    {
+      "epoch": 0.15050461772826812,
+      "grad_norm": 3.218642390355167,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 6323
+    },
+    {
+      "epoch": 0.15052842045129963,
+      "grad_norm": 5.500923316259649,
+      "learning_rate": 2e-05,
+      "loss": 0.7362,
+      "step": 6324
+    },
+    {
+      "epoch": 0.15055222317433115,
+      "grad_norm": 3.48247922680384,
+      "learning_rate": 2e-05,
+      "loss": 0.5574,
+      "step": 6325
+    },
+    {
+      "epoch": 0.15057602589736266,
+      "grad_norm": 3.053903868194372,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 6326
+    },
+    {
+      "epoch": 0.15059982862039417,
+      "grad_norm": 3.325349839890461,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 6327
+    },
+    {
+      "epoch": 0.15062363134342568,
+      "grad_norm": 6.564554630698214,
+      "learning_rate": 2e-05,
+      "loss": 0.965,
+      "step": 6328
+    },
+    {
+      "epoch": 0.1506474340664572,
+      "grad_norm": 6.976212621620935,
+      "learning_rate": 2e-05,
+      "loss": 0.7299,
+      "step": 6329
+    },
+    {
+      "epoch": 0.1506712367894887,
+      "grad_norm": 2.920695558939654,
+      "learning_rate": 2e-05,
+      "loss": 0.216,
+      "step": 6330
+    },
+    {
+      "epoch": 0.15069503951252022,
+      "grad_norm": 5.003888340770782,
+      "learning_rate": 2e-05,
+      "loss": 0.5257,
+      "step": 6331
+    },
+    {
+      "epoch": 0.15071884223555174,
+      "grad_norm": 8.52561960159136,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 6332
+    },
+    {
+      "epoch": 0.15074264495858325,
+      "grad_norm": 2.9477386925205677,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 6333
+    },
+    {
+      "epoch": 0.1507664476816148,
+      "grad_norm": 6.808031465945677,
+      "learning_rate": 2e-05,
+      "loss": 0.393,
+      "step": 6334
+    },
+    {
+      "epoch": 0.1507902504046463,
+      "grad_norm": 2.474576874549095,
+      "learning_rate": 2e-05,
+      "loss": 0.1858,
+      "step": 6335
+    },
+    {
+      "epoch": 0.15081405312767782,
+      "grad_norm": 5.562498176478295,
+      "learning_rate": 2e-05,
+      "loss": 0.8742,
+      "step": 6336
+    },
+    {
+      "epoch": 0.15083785585070933,
+      "grad_norm": 7.068912347768972,
+      "learning_rate": 2e-05,
+      "loss": 1.5705,
+      "step": 6337
+    },
+    {
+      "epoch": 0.15086165857374084,
+      "grad_norm": 4.679235083626676,
+      "learning_rate": 2e-05,
+      "loss": 0.4667,
+      "step": 6338
+    },
+    {
+      "epoch": 0.15088546129677236,
+      "grad_norm": 7.78586163460179,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 6339
+    },
+    {
+      "epoch": 0.15090926401980387,
+      "grad_norm": 3.461586856331455,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 6340
+    },
+    {
+      "epoch": 0.15093306674283538,
+      "grad_norm": 2.1443540261372664,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 6341
+    },
+    {
+      "epoch": 0.1509568694658669,
+      "grad_norm": 7.654741652121135,
+      "learning_rate": 2e-05,
+      "loss": 0.8215,
+      "step": 6342
+    },
+    {
+      "epoch": 0.1509806721888984,
+      "grad_norm": 4.977155789601603,
+      "learning_rate": 2e-05,
+      "loss": 0.7587,
+      "step": 6343
+    },
+    {
+      "epoch": 0.15100447491192992,
+      "grad_norm": 7.194493937045317,
+      "learning_rate": 2e-05,
+      "loss": 0.1907,
+      "step": 6344
+    },
+    {
+      "epoch": 0.15102827763496143,
+      "grad_norm": 4.590344929162424,
+      "learning_rate": 2e-05,
+      "loss": 0.6296,
+      "step": 6345
+    },
+    {
+      "epoch": 0.15105208035799295,
+      "grad_norm": 3.898753139886571,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 6346
+    },
+    {
+      "epoch": 0.15107588308102446,
+      "grad_norm": 2.037366769739658,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 6347
+    },
+    {
+      "epoch": 0.15109968580405597,
+      "grad_norm": 11.221614168878997,
+      "learning_rate": 2e-05,
+      "loss": 1.0618,
+      "step": 6348
+    },
+    {
+      "epoch": 0.15112348852708749,
+      "grad_norm": 5.830273902943531,
+      "learning_rate": 2e-05,
+      "loss": 1.0368,
+      "step": 6349
+    },
+    {
+      "epoch": 0.15114729125011903,
+      "grad_norm": 9.929099238144225,
+      "learning_rate": 2e-05,
+      "loss": 0.2984,
+      "step": 6350
+    },
+    {
+      "epoch": 0.15117109397315054,
+      "grad_norm": 0.5818420197062993,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 6351
+    },
+    {
+      "epoch": 0.15119489669618205,
+      "grad_norm": 6.987280070819168,
+      "learning_rate": 2e-05,
+      "loss": 1.5222,
+      "step": 6352
+    },
+    {
+      "epoch": 0.15121869941921356,
+      "grad_norm": 3.1919956166796357,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 6353
+    },
+    {
+      "epoch": 0.15124250214224508,
+      "grad_norm": 4.884958010222692,
+      "learning_rate": 2e-05,
+      "loss": 0.2639,
+      "step": 6354
+    },
+    {
+      "epoch": 0.1512663048652766,
+      "grad_norm": 12.095416800579553,
+      "learning_rate": 2e-05,
+      "loss": 1.8039,
+      "step": 6355
+    },
+    {
+      "epoch": 0.1512901075883081,
+      "grad_norm": 5.959052343094752,
+      "learning_rate": 2e-05,
+      "loss": 0.9824,
+      "step": 6356
+    },
+    {
+      "epoch": 0.15131391031133962,
+      "grad_norm": 4.154586949413746,
+      "learning_rate": 2e-05,
+      "loss": 0.2713,
+      "step": 6357
+    },
+    {
+      "epoch": 0.15133771303437113,
+      "grad_norm": 4.811711899172389,
+      "learning_rate": 2e-05,
+      "loss": 0.5699,
+      "step": 6358
+    },
+    {
+      "epoch": 0.15136151575740264,
+      "grad_norm": 6.588660478609713,
+      "learning_rate": 2e-05,
+      "loss": 0.6784,
+      "step": 6359
+    },
+    {
+      "epoch": 0.15138531848043416,
+      "grad_norm": 8.375463267051362,
+      "learning_rate": 2e-05,
+      "loss": 0.36,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15140912120346567,
+      "grad_norm": 8.210597484253082,
+      "learning_rate": 2e-05,
+      "loss": 0.8703,
+      "step": 6361
+    },
+    {
+      "epoch": 0.15143292392649718,
+      "grad_norm": 3.5607919251580746,
+      "learning_rate": 2e-05,
+      "loss": 0.2297,
+      "step": 6362
+    },
+    {
+      "epoch": 0.1514567266495287,
+      "grad_norm": 7.680054634334442,
+      "learning_rate": 2e-05,
+      "loss": 0.4725,
+      "step": 6363
+    },
+    {
+      "epoch": 0.1514805293725602,
+      "grad_norm": 4.697676577632059,
+      "learning_rate": 2e-05,
+      "loss": 0.2037,
+      "step": 6364
+    },
+    {
+      "epoch": 0.15150433209559175,
+      "grad_norm": 4.837075891609482,
+      "learning_rate": 2e-05,
+      "loss": 0.6736,
+      "step": 6365
+    },
+    {
+      "epoch": 0.15152813481862326,
+      "grad_norm": 5.54827618272581,
+      "learning_rate": 2e-05,
+      "loss": 0.5043,
+      "step": 6366
+    },
+    {
+      "epoch": 0.15155193754165477,
+      "grad_norm": 3.209505766407974,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 6367
+    },
+    {
+      "epoch": 0.1515757402646863,
+      "grad_norm": 10.578579871567086,
+      "learning_rate": 2e-05,
+      "loss": 1.1042,
+      "step": 6368
+    },
+    {
+      "epoch": 0.1515995429877178,
+      "grad_norm": 4.9887476453558035,
+      "learning_rate": 2e-05,
+      "loss": 0.7596,
+      "step": 6369
+    },
+    {
+      "epoch": 0.1516233457107493,
+      "grad_norm": 14.030732955910873,
+      "learning_rate": 2e-05,
+      "loss": 1.0348,
+      "step": 6370
+    },
+    {
+      "epoch": 0.15164714843378083,
+      "grad_norm": 7.950864243921592,
+      "learning_rate": 2e-05,
+      "loss": 0.3352,
+      "step": 6371
+    },
+    {
+      "epoch": 0.15167095115681234,
+      "grad_norm": 2.0053065363857776,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 6372
+    },
+    {
+      "epoch": 0.15169475387984385,
+      "grad_norm": 5.453089019048944,
+      "learning_rate": 2e-05,
+      "loss": 0.8688,
+      "step": 6373
+    },
+    {
+      "epoch": 0.15171855660287537,
+      "grad_norm": 4.2400021122660885,
+      "learning_rate": 2e-05,
+      "loss": 0.1645,
+      "step": 6374
+    },
+    {
+      "epoch": 0.15174235932590688,
+      "grad_norm": 2.602351932030163,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 6375
+    },
+    {
+      "epoch": 0.1517661620489384,
+      "grad_norm": 5.3827880482392665,
+      "learning_rate": 2e-05,
+      "loss": 0.842,
+      "step": 6376
+    },
+    {
+      "epoch": 0.1517899647719699,
+      "grad_norm": 5.985661015694937,
+      "learning_rate": 2e-05,
+      "loss": 0.7554,
+      "step": 6377
+    },
+    {
+      "epoch": 0.15181376749500142,
+      "grad_norm": 2.7050492520515026,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 6378
+    },
+    {
+      "epoch": 0.15183757021803293,
+      "grad_norm": 2.1926519922229053,
+      "learning_rate": 2e-05,
+      "loss": 0.1623,
+      "step": 6379
+    },
+    {
+      "epoch": 0.15186137294106447,
+      "grad_norm": 5.718178803283316,
+      "learning_rate": 2e-05,
+      "loss": 0.7688,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15188517566409598,
+      "grad_norm": 6.875267314946393,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 6381
+    },
+    {
+      "epoch": 0.1519089783871275,
+      "grad_norm": 18.365338790565346,
+      "learning_rate": 2e-05,
+      "loss": 1.0352,
+      "step": 6382
+    },
+    {
+      "epoch": 0.151932781110159,
+      "grad_norm": 6.582384552826605,
+      "learning_rate": 2e-05,
+      "loss": 0.2971,
+      "step": 6383
+    },
+    {
+      "epoch": 0.15195658383319052,
+      "grad_norm": 2.0511144320278296,
+      "learning_rate": 2e-05,
+      "loss": 0.1569,
+      "step": 6384
+    },
+    {
+      "epoch": 0.15198038655622204,
+      "grad_norm": 4.106387046373503,
+      "learning_rate": 2e-05,
+      "loss": 0.287,
+      "step": 6385
+    },
+    {
+      "epoch": 0.15200418927925355,
+      "grad_norm": 6.737408590809476,
+      "learning_rate": 2e-05,
+      "loss": 0.6948,
+      "step": 6386
+    },
+    {
+      "epoch": 0.15202799200228506,
+      "grad_norm": 10.424481994616347,
+      "learning_rate": 2e-05,
+      "loss": 1.1535,
+      "step": 6387
+    },
+    {
+      "epoch": 0.15205179472531657,
+      "grad_norm": 5.982811187220056,
+      "learning_rate": 2e-05,
+      "loss": 1.0552,
+      "step": 6388
+    },
+    {
+      "epoch": 0.1520755974483481,
+      "grad_norm": 4.682930660685167,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 6389
+    },
+    {
+      "epoch": 0.1520994001713796,
+      "grad_norm": 1.7914905895767348,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 6390
+    },
+    {
+      "epoch": 0.1521232028944111,
+      "grad_norm": 5.908210228859203,
+      "learning_rate": 2e-05,
+      "loss": 0.7147,
+      "step": 6391
+    },
+    {
+      "epoch": 0.15214700561744263,
+      "grad_norm": 9.923391632168212,
+      "learning_rate": 2e-05,
+      "loss": 0.2993,
+      "step": 6392
+    },
+    {
+      "epoch": 0.15217080834047414,
+      "grad_norm": 1.5404556871684656,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 6393
+    },
+    {
+      "epoch": 0.15219461106350565,
+      "grad_norm": 4.417073610366388,
+      "learning_rate": 2e-05,
+      "loss": 0.7592,
+      "step": 6394
+    },
+    {
+      "epoch": 0.1522184137865372,
+      "grad_norm": 8.038205001504066,
+      "learning_rate": 2e-05,
+      "loss": 0.9979,
+      "step": 6395
+    },
+    {
+      "epoch": 0.1522422165095687,
+      "grad_norm": 6.547176832206321,
+      "learning_rate": 2e-05,
+      "loss": 0.7752,
+      "step": 6396
+    },
+    {
+      "epoch": 0.15226601923260022,
+      "grad_norm": 2.1771598407324593,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 6397
+    },
+    {
+      "epoch": 0.15228982195563173,
+      "grad_norm": 3.957442922662358,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 6398
+    },
+    {
+      "epoch": 0.15231362467866325,
+      "grad_norm": 8.758138927479749,
+      "learning_rate": 2e-05,
+      "loss": 0.3395,
+      "step": 6399
+    },
+    {
+      "epoch": 0.15233742740169476,
+      "grad_norm": 3.87871728606953,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 6400
+    },
+    {
+      "epoch": 0.15236123012472627,
+      "grad_norm": 1.5726782508912973,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 6401
+    },
+    {
+      "epoch": 0.15238503284775778,
+      "grad_norm": 9.449835732867067,
+      "learning_rate": 2e-05,
+      "loss": 0.337,
+      "step": 6402
+    },
+    {
+      "epoch": 0.1524088355707893,
+      "grad_norm": 8.140764308600184,
+      "learning_rate": 2e-05,
+      "loss": 0.8183,
+      "step": 6403
+    },
+    {
+      "epoch": 0.1524326382938208,
+      "grad_norm": 6.10893147425468,
+      "learning_rate": 2e-05,
+      "loss": 0.9432,
+      "step": 6404
+    },
+    {
+      "epoch": 0.15245644101685232,
+      "grad_norm": 4.965757799958757,
+      "learning_rate": 2e-05,
+      "loss": 0.6797,
+      "step": 6405
+    },
+    {
+      "epoch": 0.15248024373988384,
+      "grad_norm": 3.4048107650500343,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 6406
+    },
+    {
+      "epoch": 0.15250404646291535,
+      "grad_norm": 4.221104728506743,
+      "learning_rate": 2e-05,
+      "loss": 0.273,
+      "step": 6407
+    },
+    {
+      "epoch": 0.15252784918594686,
+      "grad_norm": 2.1804545202256858,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 6408
+    },
+    {
+      "epoch": 0.15255165190897838,
+      "grad_norm": 4.724772721518244,
+      "learning_rate": 2e-05,
+      "loss": 0.2327,
+      "step": 6409
+    },
+    {
+      "epoch": 0.1525754546320099,
+      "grad_norm": 4.268195937042596,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15259925735504143,
+      "grad_norm": 6.476520883931548,
+      "learning_rate": 2e-05,
+      "loss": 0.7355,
+      "step": 6411
+    },
+    {
+      "epoch": 0.15262306007807294,
+      "grad_norm": 6.89297428831909,
+      "learning_rate": 2e-05,
+      "loss": 0.6459,
+      "step": 6412
+    },
+    {
+      "epoch": 0.15264686280110445,
+      "grad_norm": 6.2790281481128,
+      "learning_rate": 2e-05,
+      "loss": 1.4425,
+      "step": 6413
+    },
+    {
+      "epoch": 0.15267066552413597,
+      "grad_norm": 9.56540167334463,
+      "learning_rate": 2e-05,
+      "loss": 2.2176,
+      "step": 6414
+    },
+    {
+      "epoch": 0.15269446824716748,
+      "grad_norm": 3.989853239601712,
+      "learning_rate": 2e-05,
+      "loss": 0.2001,
+      "step": 6415
+    },
+    {
+      "epoch": 0.152718270970199,
+      "grad_norm": 3.044059963436378,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 6416
+    },
+    {
+      "epoch": 0.1527420736932305,
+      "grad_norm": 4.778823960132845,
+      "learning_rate": 2e-05,
+      "loss": 0.4795,
+      "step": 6417
+    },
+    {
+      "epoch": 0.15276587641626202,
+      "grad_norm": 6.174638981694629,
+      "learning_rate": 2e-05,
+      "loss": 0.9704,
+      "step": 6418
+    },
+    {
+      "epoch": 0.15278967913929353,
+      "grad_norm": 2.9693187338221514,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 6419
+    },
+    {
+      "epoch": 0.15281348186232505,
+      "grad_norm": 6.37382279698733,
+      "learning_rate": 2e-05,
+      "loss": 0.2703,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15283728458535656,
+      "grad_norm": 8.052806954370734,
+      "learning_rate": 2e-05,
+      "loss": 0.3999,
+      "step": 6421
+    },
+    {
+      "epoch": 0.15286108730838807,
+      "grad_norm": 4.744907302630779,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 6422
+    },
+    {
+      "epoch": 0.15288489003141958,
+      "grad_norm": 6.256768914034105,
+      "learning_rate": 2e-05,
+      "loss": 0.3686,
+      "step": 6423
+    },
+    {
+      "epoch": 0.1529086927544511,
+      "grad_norm": 7.51269979403552,
+      "learning_rate": 2e-05,
+      "loss": 1.1873,
+      "step": 6424
+    },
+    {
+      "epoch": 0.1529324954774826,
+      "grad_norm": 4.11271144919317,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 6425
+    },
+    {
+      "epoch": 0.15295629820051415,
+      "grad_norm": 3.22827098178718,
+      "learning_rate": 2e-05,
+      "loss": 0.1863,
+      "step": 6426
+    },
+    {
+      "epoch": 0.15298010092354566,
+      "grad_norm": 10.82904597976927,
+      "learning_rate": 2e-05,
+      "loss": 0.9977,
+      "step": 6427
+    },
+    {
+      "epoch": 0.15300390364657718,
+      "grad_norm": 8.26753542599866,
+      "learning_rate": 2e-05,
+      "loss": 0.386,
+      "step": 6428
+    },
+    {
+      "epoch": 0.1530277063696087,
+      "grad_norm": 8.097395257960592,
+      "learning_rate": 2e-05,
+      "loss": 0.4655,
+      "step": 6429
+    },
+    {
+      "epoch": 0.1530515090926402,
+      "grad_norm": 7.174589169780675,
+      "learning_rate": 2e-05,
+      "loss": 1.845,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15307531181567172,
+      "grad_norm": 6.400794295819629,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 6431
+    },
+    {
+      "epoch": 0.15309911453870323,
+      "grad_norm": 3.705860972143418,
+      "learning_rate": 2e-05,
+      "loss": 0.2499,
+      "step": 6432
+    },
+    {
+      "epoch": 0.15312291726173474,
+      "grad_norm": 5.921136053767657,
+      "learning_rate": 2e-05,
+      "loss": 0.542,
+      "step": 6433
+    },
+    {
+      "epoch": 0.15314671998476626,
+      "grad_norm": 4.486600819045605,
+      "learning_rate": 2e-05,
+      "loss": 0.172,
+      "step": 6434
+    },
+    {
+      "epoch": 0.15317052270779777,
+      "grad_norm": 6.284009661874782,
+      "learning_rate": 2e-05,
+      "loss": 0.9765,
+      "step": 6435
+    },
+    {
+      "epoch": 0.15319432543082928,
+      "grad_norm": 7.281080522252892,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 6436
+    },
+    {
+      "epoch": 0.1532181281538608,
+      "grad_norm": 6.244673174481832,
+      "learning_rate": 2e-05,
+      "loss": 1.4256,
+      "step": 6437
+    },
+    {
+      "epoch": 0.1532419308768923,
+      "grad_norm": 5.56395541662438,
+      "learning_rate": 2e-05,
+      "loss": 0.5642,
+      "step": 6438
+    },
+    {
+      "epoch": 0.15326573359992382,
+      "grad_norm": 6.607722834890134,
+      "learning_rate": 2e-05,
+      "loss": 0.3267,
+      "step": 6439
+    },
+    {
+      "epoch": 0.15328953632295533,
+      "grad_norm": 7.7692416384291665,
+      "learning_rate": 2e-05,
+      "loss": 0.8904,
+      "step": 6440
+    },
+    {
+      "epoch": 0.15331333904598687,
+      "grad_norm": 1.0226457008633776,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 6441
+    },
+    {
+      "epoch": 0.1533371417690184,
+      "grad_norm": 3.6527368979348926,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 6442
+    },
+    {
+      "epoch": 0.1533609444920499,
+      "grad_norm": 1.8874464750627031,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 6443
+    },
+    {
+      "epoch": 0.1533847472150814,
+      "grad_norm": 2.8563484343069057,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 6444
+    },
+    {
+      "epoch": 0.15340854993811293,
+      "grad_norm": 5.08338921440301,
+      "learning_rate": 2e-05,
+      "loss": 0.6853,
+      "step": 6445
+    },
+    {
+      "epoch": 0.15343235266114444,
+      "grad_norm": 9.276773446963668,
+      "learning_rate": 2e-05,
+      "loss": 2.1523,
+      "step": 6446
+    },
+    {
+      "epoch": 0.15345615538417595,
+      "grad_norm": 7.510042044099965,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 6447
+    },
+    {
+      "epoch": 0.15347995810720746,
+      "grad_norm": 7.900473813212469,
+      "learning_rate": 2e-05,
+      "loss": 0.5017,
+      "step": 6448
+    },
+    {
+      "epoch": 0.15350376083023898,
+      "grad_norm": 3.9387302553051757,
+      "learning_rate": 2e-05,
+      "loss": 0.4036,
+      "step": 6449
+    },
+    {
+      "epoch": 0.1535275635532705,
+      "grad_norm": 2.981247707160042,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 6450
+    },
+    {
+      "epoch": 0.153551366276302,
+      "grad_norm": 5.550993364698842,
+      "learning_rate": 2e-05,
+      "loss": 0.9375,
+      "step": 6451
+    },
+    {
+      "epoch": 0.15357516899933352,
+      "grad_norm": 9.684038898928689,
+      "learning_rate": 2e-05,
+      "loss": 0.558,
+      "step": 6452
+    },
+    {
+      "epoch": 0.15359897172236503,
+      "grad_norm": 25.93445020001349,
+      "learning_rate": 2e-05,
+      "loss": 0.6035,
+      "step": 6453
+    },
+    {
+      "epoch": 0.15362277444539654,
+      "grad_norm": 5.046050058313837,
+      "learning_rate": 2e-05,
+      "loss": 0.5555,
+      "step": 6454
+    },
+    {
+      "epoch": 0.15364657716842806,
+      "grad_norm": 4.911519654966419,
+      "learning_rate": 2e-05,
+      "loss": 0.5291,
+      "step": 6455
+    },
+    {
+      "epoch": 0.1536703798914596,
+      "grad_norm": 6.977206106180648,
+      "learning_rate": 2e-05,
+      "loss": 0.5374,
+      "step": 6456
+    },
+    {
+      "epoch": 0.1536941826144911,
+      "grad_norm": 5.514109087580047,
+      "learning_rate": 2e-05,
+      "loss": 0.4779,
+      "step": 6457
+    },
+    {
+      "epoch": 0.15371798533752262,
+      "grad_norm": 6.355997303956701,
+      "learning_rate": 2e-05,
+      "loss": 0.3894,
+      "step": 6458
+    },
+    {
+      "epoch": 0.15374178806055414,
+      "grad_norm": 5.952730579423676,
+      "learning_rate": 2e-05,
+      "loss": 1.0141,
+      "step": 6459
+    },
+    {
+      "epoch": 0.15376559078358565,
+      "grad_norm": 3.1298983293183364,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15378939350661716,
+      "grad_norm": 4.8844322232029596,
+      "learning_rate": 2e-05,
+      "loss": 0.8492,
+      "step": 6461
+    },
+    {
+      "epoch": 0.15381319622964867,
+      "grad_norm": 3.748621219459309,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 6462
+    },
+    {
+      "epoch": 0.1538369989526802,
+      "grad_norm": 2.4674252764880342,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 6463
+    },
+    {
+      "epoch": 0.1538608016757117,
+      "grad_norm": 1.8314176762488013,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 6464
+    },
+    {
+      "epoch": 0.1538846043987432,
+      "grad_norm": 6.9851838316197545,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 6465
+    },
+    {
+      "epoch": 0.15390840712177473,
+      "grad_norm": 9.419051808662914,
+      "learning_rate": 2e-05,
+      "loss": 1.2025,
+      "step": 6466
+    },
+    {
+      "epoch": 0.15393220984480624,
+      "grad_norm": 3.9346506679569706,
+      "learning_rate": 2e-05,
+      "loss": 0.4461,
+      "step": 6467
+    },
+    {
+      "epoch": 0.15395601256783775,
+      "grad_norm": 9.26476798430759,
+      "learning_rate": 2e-05,
+      "loss": 1.798,
+      "step": 6468
+    },
+    {
+      "epoch": 0.15397981529086927,
+      "grad_norm": 4.184454681627088,
+      "learning_rate": 2e-05,
+      "loss": 0.1725,
+      "step": 6469
+    },
+    {
+      "epoch": 0.15400361801390078,
+      "grad_norm": 1.0513554165961971,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 6470
+    },
+    {
+      "epoch": 0.1540274207369323,
+      "grad_norm": 3.860306867956276,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 6471
+    },
+    {
+      "epoch": 0.15405122345996383,
+      "grad_norm": 6.155818872162706,
+      "learning_rate": 2e-05,
+      "loss": 0.9104,
+      "step": 6472
+    },
+    {
+      "epoch": 0.15407502618299534,
+      "grad_norm": 24.532016659585114,
+      "learning_rate": 2e-05,
+      "loss": 0.6496,
+      "step": 6473
+    },
+    {
+      "epoch": 0.15409882890602686,
+      "grad_norm": 2.442732965082403,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 6474
+    },
+    {
+      "epoch": 0.15412263162905837,
+      "grad_norm": 5.136597271203985,
+      "learning_rate": 2e-05,
+      "loss": 0.6904,
+      "step": 6475
+    },
+    {
+      "epoch": 0.15414643435208988,
+      "grad_norm": 8.219902487561235,
+      "learning_rate": 2e-05,
+      "loss": 0.877,
+      "step": 6476
+    },
+    {
+      "epoch": 0.1541702370751214,
+      "grad_norm": 3.5580204622159353,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 6477
+    },
+    {
+      "epoch": 0.1541940397981529,
+      "grad_norm": 2.2622595477831315,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 6478
+    },
+    {
+      "epoch": 0.15421784252118442,
+      "grad_norm": 4.969120826202594,
+      "learning_rate": 2e-05,
+      "loss": 0.5781,
+      "step": 6479
+    },
+    {
+      "epoch": 0.15424164524421594,
+      "grad_norm": 6.033659737932028,
+      "learning_rate": 2e-05,
+      "loss": 1.0401,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15426544796724745,
+      "grad_norm": 12.982268764732597,
+      "learning_rate": 2e-05,
+      "loss": 2.4542,
+      "step": 6481
+    },
+    {
+      "epoch": 0.15428925069027896,
+      "grad_norm": 6.8313433348387935,
+      "learning_rate": 2e-05,
+      "loss": 1.3929,
+      "step": 6482
+    },
+    {
+      "epoch": 0.15431305341331047,
+      "grad_norm": 1.1016681961745587,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 6483
+    },
+    {
+      "epoch": 0.154336856136342,
+      "grad_norm": 0.6490572996661291,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 6484
+    },
+    {
+      "epoch": 0.1543606588593735,
+      "grad_norm": 10.56615995905647,
+      "learning_rate": 2e-05,
+      "loss": 0.4517,
+      "step": 6485
+    },
+    {
+      "epoch": 0.154384461582405,
+      "grad_norm": 14.111624373317305,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 6486
+    },
+    {
+      "epoch": 0.15440826430543655,
+      "grad_norm": 4.087812299332055,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 6487
+    },
+    {
+      "epoch": 0.15443206702846807,
+      "grad_norm": 16.01227364984956,
+      "learning_rate": 2e-05,
+      "loss": 1.5121,
+      "step": 6488
+    },
+    {
+      "epoch": 0.15445586975149958,
+      "grad_norm": 2.416854712237164,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 6489
+    },
+    {
+      "epoch": 0.1544796724745311,
+      "grad_norm": 1.9857522840131607,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 6490
+    },
+    {
+      "epoch": 0.1545034751975626,
+      "grad_norm": 1.768718025148466,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 6491
+    },
+    {
+      "epoch": 0.15452727792059412,
+      "grad_norm": 3.0891361355297753,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 6492
+    },
+    {
+      "epoch": 0.15455108064362563,
+      "grad_norm": 11.432367158074248,
+      "learning_rate": 2e-05,
+      "loss": 1.0057,
+      "step": 6493
+    },
+    {
+      "epoch": 0.15457488336665715,
+      "grad_norm": 4.158218432620868,
+      "learning_rate": 2e-05,
+      "loss": 0.3756,
+      "step": 6494
+    },
+    {
+      "epoch": 0.15459868608968866,
+      "grad_norm": 4.957864177267587,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 6495
+    },
+    {
+      "epoch": 0.15462248881272017,
+      "grad_norm": 10.424495038262638,
+      "learning_rate": 2e-05,
+      "loss": 2.0716,
+      "step": 6496
+    },
+    {
+      "epoch": 0.15464629153575168,
+      "grad_norm": 10.836483541440888,
+      "learning_rate": 2e-05,
+      "loss": 1.792,
+      "step": 6497
+    },
+    {
+      "epoch": 0.1546700942587832,
+      "grad_norm": 8.111684662742723,
+      "learning_rate": 2e-05,
+      "loss": 0.2927,
+      "step": 6498
+    },
+    {
+      "epoch": 0.1546938969818147,
+      "grad_norm": 13.886245440508985,
+      "learning_rate": 2e-05,
+      "loss": 0.5229,
+      "step": 6499
+    },
+    {
+      "epoch": 0.15471769970484622,
+      "grad_norm": 3.0459385073350345,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15474150242787774,
+      "grad_norm": 2.0174935770311513,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 6501
+    },
+    {
+      "epoch": 0.15476530515090928,
+      "grad_norm": 5.863997646702827,
+      "learning_rate": 2e-05,
+      "loss": 0.7871,
+      "step": 6502
+    },
+    {
+      "epoch": 0.1547891078739408,
+      "grad_norm": 3.4567247282411153,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 6503
+    },
+    {
+      "epoch": 0.1548129105969723,
+      "grad_norm": 19.208176716704994,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 6504
+    },
+    {
+      "epoch": 0.15483671332000382,
+      "grad_norm": 6.729451837629397,
+      "learning_rate": 2e-05,
+      "loss": 1.0069,
+      "step": 6505
+    },
+    {
+      "epoch": 0.15486051604303533,
+      "grad_norm": 4.6510828659413725,
+      "learning_rate": 2e-05,
+      "loss": 0.7024,
+      "step": 6506
+    },
+    {
+      "epoch": 0.15488431876606684,
+      "grad_norm": 6.950921472420486,
+      "learning_rate": 2e-05,
+      "loss": 0.4156,
+      "step": 6507
+    },
+    {
+      "epoch": 0.15490812148909835,
+      "grad_norm": 4.400611293527559,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 6508
+    },
+    {
+      "epoch": 0.15493192421212987,
+      "grad_norm": 4.624994289938454,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 6509
+    },
+    {
+      "epoch": 0.15495572693516138,
+      "grad_norm": 15.824582549569032,
+      "learning_rate": 2e-05,
+      "loss": 0.4446,
+      "step": 6510
+    },
+    {
+      "epoch": 0.1549795296581929,
+      "grad_norm": 1.4525562666602478,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 6511
+    },
+    {
+      "epoch": 0.1550033323812244,
+      "grad_norm": 4.779266378742347,
+      "learning_rate": 2e-05,
+      "loss": 0.2783,
+      "step": 6512
+    },
+    {
+      "epoch": 0.15502713510425592,
+      "grad_norm": 10.605448909836994,
+      "learning_rate": 2e-05,
+      "loss": 0.7351,
+      "step": 6513
+    },
+    {
+      "epoch": 0.15505093782728743,
+      "grad_norm": 2.5720331999491504,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 6514
+    },
+    {
+      "epoch": 0.15507474055031895,
+      "grad_norm": 2.2992089772945072,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 6515
+    },
+    {
+      "epoch": 0.15509854327335046,
+      "grad_norm": 4.149809695316531,
+      "learning_rate": 2e-05,
+      "loss": 0.7008,
+      "step": 6516
+    },
+    {
+      "epoch": 0.155122345996382,
+      "grad_norm": 10.980025403125024,
+      "learning_rate": 2e-05,
+      "loss": 0.5857,
+      "step": 6517
+    },
+    {
+      "epoch": 0.1551461487194135,
+      "grad_norm": 4.341788814823889,
+      "learning_rate": 2e-05,
+      "loss": 0.6176,
+      "step": 6518
+    },
+    {
+      "epoch": 0.15516995144244503,
+      "grad_norm": 4.3059481316148025,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 6519
+    },
+    {
+      "epoch": 0.15519375416547654,
+      "grad_norm": 3.9004776963587755,
+      "learning_rate": 2e-05,
+      "loss": 0.3085,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15521755688850805,
+      "grad_norm": 2.68915848018533,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 6521
+    },
+    {
+      "epoch": 0.15524135961153956,
+      "grad_norm": 9.060155334643829,
+      "learning_rate": 2e-05,
+      "loss": 0.6259,
+      "step": 6522
+    },
+    {
+      "epoch": 0.15526516233457108,
+      "grad_norm": 6.457867307987936,
+      "learning_rate": 2e-05,
+      "loss": 0.8779,
+      "step": 6523
+    },
+    {
+      "epoch": 0.1552889650576026,
+      "grad_norm": 3.477084019632811,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 6524
+    },
+    {
+      "epoch": 0.1553127677806341,
+      "grad_norm": 7.081729099930958,
+      "learning_rate": 2e-05,
+      "loss": 0.9838,
+      "step": 6525
+    },
+    {
+      "epoch": 0.15533657050366562,
+      "grad_norm": 3.832304984207106,
+      "learning_rate": 2e-05,
+      "loss": 0.4307,
+      "step": 6526
+    },
+    {
+      "epoch": 0.15536037322669713,
+      "grad_norm": 8.472932411213316,
+      "learning_rate": 2e-05,
+      "loss": 0.4981,
+      "step": 6527
+    },
+    {
+      "epoch": 0.15538417594972864,
+      "grad_norm": 5.987273729642179,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 6528
+    },
+    {
+      "epoch": 0.15540797867276016,
+      "grad_norm": 6.528762269914427,
+      "learning_rate": 2e-05,
+      "loss": 1.4196,
+      "step": 6529
+    },
+    {
+      "epoch": 0.15543178139579167,
+      "grad_norm": 2.2574301760903897,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 6530
+    },
+    {
+      "epoch": 0.15545558411882318,
+      "grad_norm": 3.2939893901691164,
+      "learning_rate": 2e-05,
+      "loss": 0.2586,
+      "step": 6531
+    },
+    {
+      "epoch": 0.15547938684185472,
+      "grad_norm": 5.145571027344723,
+      "learning_rate": 2e-05,
+      "loss": 0.9302,
+      "step": 6532
+    },
+    {
+      "epoch": 0.15550318956488624,
+      "grad_norm": 5.614456746346966,
+      "learning_rate": 2e-05,
+      "loss": 0.2596,
+      "step": 6533
+    },
+    {
+      "epoch": 0.15552699228791775,
+      "grad_norm": 7.270054559759668,
+      "learning_rate": 2e-05,
+      "loss": 0.8352,
+      "step": 6534
+    },
+    {
+      "epoch": 0.15555079501094926,
+      "grad_norm": 7.573239472627789,
+      "learning_rate": 2e-05,
+      "loss": 0.7708,
+      "step": 6535
+    },
+    {
+      "epoch": 0.15557459773398077,
+      "grad_norm": 2.101618132628234,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 6536
+    },
+    {
+      "epoch": 0.1555984004570123,
+      "grad_norm": 4.854714310284943,
+      "learning_rate": 2e-05,
+      "loss": 0.3056,
+      "step": 6537
+    },
+    {
+      "epoch": 0.1556222031800438,
+      "grad_norm": 4.330375715807885,
+      "learning_rate": 2e-05,
+      "loss": 0.7958,
+      "step": 6538
+    },
+    {
+      "epoch": 0.1556460059030753,
+      "grad_norm": 6.082700026621028,
+      "learning_rate": 2e-05,
+      "loss": 0.8912,
+      "step": 6539
+    },
+    {
+      "epoch": 0.15566980862610683,
+      "grad_norm": 4.351262821386691,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15569361134913834,
+      "grad_norm": 8.3212805865397,
+      "learning_rate": 2e-05,
+      "loss": 0.5062,
+      "step": 6541
+    },
+    {
+      "epoch": 0.15571741407216985,
+      "grad_norm": 5.6481000824175,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 6542
+    },
+    {
+      "epoch": 0.15574121679520136,
+      "grad_norm": 3.8413387435748545,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 6543
+    },
+    {
+      "epoch": 0.15576501951823288,
+      "grad_norm": 6.997497400322477,
+      "learning_rate": 2e-05,
+      "loss": 0.6948,
+      "step": 6544
+    },
+    {
+      "epoch": 0.1557888222412644,
+      "grad_norm": 5.053029314533426,
+      "learning_rate": 2e-05,
+      "loss": 0.6773,
+      "step": 6545
+    },
+    {
+      "epoch": 0.1558126249642959,
+      "grad_norm": 4.237738331973776,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 6546
+    },
+    {
+      "epoch": 0.15583642768732742,
+      "grad_norm": 5.6244563311868045,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 6547
+    },
+    {
+      "epoch": 0.15586023041035896,
+      "grad_norm": 4.029207640577062,
+      "learning_rate": 2e-05,
+      "loss": 0.525,
+      "step": 6548
+    },
+    {
+      "epoch": 0.15588403313339047,
+      "grad_norm": 10.452230151943454,
+      "learning_rate": 2e-05,
+      "loss": 1.1089,
+      "step": 6549
+    },
+    {
+      "epoch": 0.15590783585642198,
+      "grad_norm": 6.858224768374269,
+      "learning_rate": 2e-05,
+      "loss": 0.2934,
+      "step": 6550
+    },
+    {
+      "epoch": 0.1559316385794535,
+      "grad_norm": 14.158477786238496,
+      "learning_rate": 2e-05,
+      "loss": 0.9059,
+      "step": 6551
+    },
+    {
+      "epoch": 0.155955441302485,
+      "grad_norm": 2.069044910521819,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 6552
+    },
+    {
+      "epoch": 0.15597924402551652,
+      "grad_norm": 7.521615241502334,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 6553
+    },
+    {
+      "epoch": 0.15600304674854804,
+      "grad_norm": 4.693886634330592,
+      "learning_rate": 2e-05,
+      "loss": 0.7924,
+      "step": 6554
+    },
+    {
+      "epoch": 0.15602684947157955,
+      "grad_norm": 4.00982609542426,
+      "learning_rate": 2e-05,
+      "loss": 0.2937,
+      "step": 6555
+    },
+    {
+      "epoch": 0.15605065219461106,
+      "grad_norm": 5.11358647789339,
+      "learning_rate": 2e-05,
+      "loss": 0.2353,
+      "step": 6556
+    },
+    {
+      "epoch": 0.15607445491764257,
+      "grad_norm": 8.904819333572195,
+      "learning_rate": 2e-05,
+      "loss": 0.8508,
+      "step": 6557
+    },
+    {
+      "epoch": 0.1560982576406741,
+      "grad_norm": 3.4949408071245176,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 6558
+    },
+    {
+      "epoch": 0.1561220603637056,
+      "grad_norm": 6.230164116529361,
+      "learning_rate": 2e-05,
+      "loss": 1.164,
+      "step": 6559
+    },
+    {
+      "epoch": 0.1561458630867371,
+      "grad_norm": 5.674466795795031,
+      "learning_rate": 2e-05,
+      "loss": 0.2338,
+      "step": 6560
+    },
+    {
+      "epoch": 0.15616966580976863,
+      "grad_norm": 12.298636975212775,
+      "learning_rate": 2e-05,
+      "loss": 0.5601,
+      "step": 6561
+    },
+    {
+      "epoch": 0.15619346853280014,
+      "grad_norm": 4.160329197028944,
+      "learning_rate": 2e-05,
+      "loss": 0.2417,
+      "step": 6562
+    },
+    {
+      "epoch": 0.15621727125583168,
+      "grad_norm": 5.883660661262581,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 6563
+    },
+    {
+      "epoch": 0.1562410739788632,
+      "grad_norm": 4.6978683583329515,
+      "learning_rate": 2e-05,
+      "loss": 0.1649,
+      "step": 6564
+    },
+    {
+      "epoch": 0.1562648767018947,
+      "grad_norm": 2.2088137432988177,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 6565
+    },
+    {
+      "epoch": 0.15628867942492622,
+      "grad_norm": 3.749890511156582,
+      "learning_rate": 2e-05,
+      "loss": 0.564,
+      "step": 6566
+    },
+    {
+      "epoch": 0.15631248214795773,
+      "grad_norm": 4.168067168214124,
+      "learning_rate": 2e-05,
+      "loss": 0.7103,
+      "step": 6567
+    },
+    {
+      "epoch": 0.15633628487098924,
+      "grad_norm": 6.560276863501202,
+      "learning_rate": 2e-05,
+      "loss": 0.7601,
+      "step": 6568
+    },
+    {
+      "epoch": 0.15636008759402076,
+      "grad_norm": 11.444439439428937,
+      "learning_rate": 2e-05,
+      "loss": 0.4471,
+      "step": 6569
+    },
+    {
+      "epoch": 0.15638389031705227,
+      "grad_norm": 15.671487583811444,
+      "learning_rate": 2e-05,
+      "loss": 0.9185,
+      "step": 6570
+    },
+    {
+      "epoch": 0.15640769304008378,
+      "grad_norm": 9.688684177715135,
+      "learning_rate": 2e-05,
+      "loss": 0.9813,
+      "step": 6571
+    },
+    {
+      "epoch": 0.1564314957631153,
+      "grad_norm": 3.93632653775888,
+      "learning_rate": 2e-05,
+      "loss": 0.5673,
+      "step": 6572
+    },
+    {
+      "epoch": 0.1564552984861468,
+      "grad_norm": 42.83647133948317,
+      "learning_rate": 2e-05,
+      "loss": 1.0705,
+      "step": 6573
+    },
+    {
+      "epoch": 0.15647910120917832,
+      "grad_norm": 3.0288714725252714,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 6574
+    },
+    {
+      "epoch": 0.15650290393220984,
+      "grad_norm": 8.218741153808613,
+      "learning_rate": 2e-05,
+      "loss": 0.7187,
+      "step": 6575
+    },
+    {
+      "epoch": 0.15652670665524135,
+      "grad_norm": 1.5766289142926424,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 6576
+    },
+    {
+      "epoch": 0.15655050937827286,
+      "grad_norm": 9.882828658175763,
+      "learning_rate": 2e-05,
+      "loss": 0.4152,
+      "step": 6577
+    },
+    {
+      "epoch": 0.1565743121013044,
+      "grad_norm": 4.0440565930895875,
+      "learning_rate": 2e-05,
+      "loss": 0.5514,
+      "step": 6578
+    },
+    {
+      "epoch": 0.15659811482433592,
+      "grad_norm": 3.0736135104909876,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 6579
+    },
+    {
+      "epoch": 0.15662191754736743,
+      "grad_norm": 8.747749264146554,
+      "learning_rate": 2e-05,
+      "loss": 0.4901,
+      "step": 6580
+    },
+    {
+      "epoch": 0.15664572027039894,
+      "grad_norm": 3.104083503419853,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 6581
+    },
+    {
+      "epoch": 0.15666952299343045,
+      "grad_norm": 4.4541311282080835,
+      "learning_rate": 2e-05,
+      "loss": 0.5657,
+      "step": 6582
+    },
+    {
+      "epoch": 0.15669332571646197,
+      "grad_norm": 8.964068568112857,
+      "learning_rate": 2e-05,
+      "loss": 0.4334,
+      "step": 6583
+    },
+    {
+      "epoch": 0.15671712843949348,
+      "grad_norm": 7.210830309436835,
+      "learning_rate": 2e-05,
+      "loss": 0.2713,
+      "step": 6584
+    },
+    {
+      "epoch": 0.156740931162525,
+      "grad_norm": 4.559330424108593,
+      "learning_rate": 2e-05,
+      "loss": 0.6963,
+      "step": 6585
+    },
+    {
+      "epoch": 0.1567647338855565,
+      "grad_norm": 5.146643090108429,
+      "learning_rate": 2e-05,
+      "loss": 0.9541,
+      "step": 6586
+    },
+    {
+      "epoch": 0.15678853660858802,
+      "grad_norm": 6.653217061609706,
+      "learning_rate": 2e-05,
+      "loss": 0.3139,
+      "step": 6587
+    },
+    {
+      "epoch": 0.15681233933161953,
+      "grad_norm": 4.224648657596633,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 6588
+    },
+    {
+      "epoch": 0.15683614205465105,
+      "grad_norm": 10.543981523618534,
+      "learning_rate": 2e-05,
+      "loss": 0.8946,
+      "step": 6589
+    },
+    {
+      "epoch": 0.15685994477768256,
+      "grad_norm": 6.0767672430719415,
+      "learning_rate": 2e-05,
+      "loss": 0.5029,
+      "step": 6590
+    },
+    {
+      "epoch": 0.15688374750071407,
+      "grad_norm": 2.8950725154574983,
+      "learning_rate": 2e-05,
+      "loss": 0.1463,
+      "step": 6591
+    },
+    {
+      "epoch": 0.15690755022374558,
+      "grad_norm": 5.226218868933165,
+      "learning_rate": 2e-05,
+      "loss": 0.7303,
+      "step": 6592
+    },
+    {
+      "epoch": 0.15693135294677713,
+      "grad_norm": 4.330027540178492,
+      "learning_rate": 2e-05,
+      "loss": 0.3082,
+      "step": 6593
+    },
+    {
+      "epoch": 0.15695515566980864,
+      "grad_norm": 9.60980425469151,
+      "learning_rate": 2e-05,
+      "loss": 0.8449,
+      "step": 6594
+    },
+    {
+      "epoch": 0.15697895839284015,
+      "grad_norm": 6.52414530720304,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 6595
+    },
+    {
+      "epoch": 0.15700276111587166,
+      "grad_norm": 6.025844231233478,
+      "learning_rate": 2e-05,
+      "loss": 1.0539,
+      "step": 6596
+    },
+    {
+      "epoch": 0.15702656383890318,
+      "grad_norm": 2.229651851404961,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 6597
+    },
+    {
+      "epoch": 0.1570503665619347,
+      "grad_norm": 2.044970337068743,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 6598
+    },
+    {
+      "epoch": 0.1570741692849662,
+      "grad_norm": 4.074551219703185,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 6599
+    },
+    {
+      "epoch": 0.15709797200799772,
+      "grad_norm": 67.41077641257954,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 6600
+    },
+    {
+      "epoch": 0.15712177473102923,
+      "grad_norm": 1.92295437347665,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 6601
+    },
+    {
+      "epoch": 0.15714557745406074,
+      "grad_norm": 5.809150839275913,
+      "learning_rate": 2e-05,
+      "loss": 1.1836,
+      "step": 6602
+    },
+    {
+      "epoch": 0.15716938017709225,
+      "grad_norm": 4.553906901322872,
+      "learning_rate": 2e-05,
+      "loss": 0.8434,
+      "step": 6603
+    },
+    {
+      "epoch": 0.15719318290012377,
+      "grad_norm": 3.380609618050142,
+      "learning_rate": 2e-05,
+      "loss": 0.2705,
+      "step": 6604
+    },
+    {
+      "epoch": 0.15721698562315528,
+      "grad_norm": 5.468428415061186,
+      "learning_rate": 2e-05,
+      "loss": 0.3707,
+      "step": 6605
+    },
+    {
+      "epoch": 0.1572407883461868,
+      "grad_norm": 7.194553692376134,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 6606
+    },
+    {
+      "epoch": 0.1572645910692183,
+      "grad_norm": 6.06947173832709,
+      "learning_rate": 2e-05,
+      "loss": 0.5355,
+      "step": 6607
+    },
+    {
+      "epoch": 0.15728839379224982,
+      "grad_norm": 3.7730993136938764,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 6608
+    },
+    {
+      "epoch": 0.15731219651528136,
+      "grad_norm": 2.2369105602097314,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 6609
+    },
+    {
+      "epoch": 0.15733599923831287,
+      "grad_norm": 8.959255717933424,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 6610
+    },
+    {
+      "epoch": 0.1573598019613444,
+      "grad_norm": 4.987304795302422,
+      "learning_rate": 2e-05,
+      "loss": 0.5869,
+      "step": 6611
+    },
+    {
+      "epoch": 0.1573836046843759,
+      "grad_norm": 9.025826101653847,
+      "learning_rate": 2e-05,
+      "loss": 0.3875,
+      "step": 6612
+    },
+    {
+      "epoch": 0.1574074074074074,
+      "grad_norm": 6.9969054932186285,
+      "learning_rate": 2e-05,
+      "loss": 0.6681,
+      "step": 6613
+    },
+    {
+      "epoch": 0.15743121013043893,
+      "grad_norm": 6.17695184657204,
+      "learning_rate": 2e-05,
+      "loss": 0.9168,
+      "step": 6614
+    },
+    {
+      "epoch": 0.15745501285347044,
+      "grad_norm": 1.8959557886892835,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 6615
+    },
+    {
+      "epoch": 0.15747881557650195,
+      "grad_norm": 5.553536531584412,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 6616
+    },
+    {
+      "epoch": 0.15750261829953346,
+      "grad_norm": 3.1801059513730356,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 6617
+    },
+    {
+      "epoch": 0.15752642102256498,
+      "grad_norm": 20.682216042584887,
+      "learning_rate": 2e-05,
+      "loss": 1.208,
+      "step": 6618
+    },
+    {
+      "epoch": 0.1575502237455965,
+      "grad_norm": 6.892725674450913,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 6619
+    },
+    {
+      "epoch": 0.157574026468628,
+      "grad_norm": 5.969890741159528,
+      "learning_rate": 2e-05,
+      "loss": 0.8999,
+      "step": 6620
+    },
+    {
+      "epoch": 0.15759782919165952,
+      "grad_norm": 9.112238350588852,
+      "learning_rate": 2e-05,
+      "loss": 0.439,
+      "step": 6621
+    },
+    {
+      "epoch": 0.15762163191469103,
+      "grad_norm": 8.367244681224362,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 6622
+    },
+    {
+      "epoch": 0.15764543463772254,
+      "grad_norm": 3.2081794938130477,
+      "learning_rate": 2e-05,
+      "loss": 0.4859,
+      "step": 6623
+    },
+    {
+      "epoch": 0.15766923736075408,
+      "grad_norm": 1.0081584672169144,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 6624
+    },
+    {
+      "epoch": 0.1576930400837856,
+      "grad_norm": 2.5055953206101025,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 6625
+    },
+    {
+      "epoch": 0.1577168428068171,
+      "grad_norm": 2.171660860677141,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 6626
+    },
+    {
+      "epoch": 0.15774064552984862,
+      "grad_norm": 8.3040742793024,
+      "learning_rate": 2e-05,
+      "loss": 0.304,
+      "step": 6627
+    },
+    {
+      "epoch": 0.15776444825288014,
+      "grad_norm": 4.494178042475107,
+      "learning_rate": 2e-05,
+      "loss": 0.8342,
+      "step": 6628
+    },
+    {
+      "epoch": 0.15778825097591165,
+      "grad_norm": 7.422107406560351,
+      "learning_rate": 2e-05,
+      "loss": 1.5251,
+      "step": 6629
+    },
+    {
+      "epoch": 0.15781205369894316,
+      "grad_norm": 7.629795545262388,
+      "learning_rate": 2e-05,
+      "loss": 0.8684,
+      "step": 6630
+    },
+    {
+      "epoch": 0.15783585642197467,
+      "grad_norm": 9.268008527728377,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 6631
+    },
+    {
+      "epoch": 0.1578596591450062,
+      "grad_norm": 7.572029214517002,
+      "learning_rate": 2e-05,
+      "loss": 0.3136,
+      "step": 6632
+    },
+    {
+      "epoch": 0.1578834618680377,
+      "grad_norm": 1.1006898622221777,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 6633
+    },
+    {
+      "epoch": 0.1579072645910692,
+      "grad_norm": 6.019152382715144,
+      "learning_rate": 2e-05,
+      "loss": 0.4855,
+      "step": 6634
+    },
+    {
+      "epoch": 0.15793106731410073,
+      "grad_norm": 4.716185312482661,
+      "learning_rate": 2e-05,
+      "loss": 0.285,
+      "step": 6635
+    },
+    {
+      "epoch": 0.15795487003713224,
+      "grad_norm": 4.578002049078438,
+      "learning_rate": 2e-05,
+      "loss": 0.2359,
+      "step": 6636
+    },
+    {
+      "epoch": 0.15797867276016375,
+      "grad_norm": 7.057018342295638,
+      "learning_rate": 2e-05,
+      "loss": 0.2643,
+      "step": 6637
+    },
+    {
+      "epoch": 0.15800247548319526,
+      "grad_norm": 0.8354672178350192,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 6638
+    },
+    {
+      "epoch": 0.1580262782062268,
+      "grad_norm": 5.005819311526332,
+      "learning_rate": 2e-05,
+      "loss": 0.8278,
+      "step": 6639
+    },
+    {
+      "epoch": 0.15805008092925832,
+      "grad_norm": 6.192183579254233,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 6640
+    },
+    {
+      "epoch": 0.15807388365228983,
+      "grad_norm": 4.887254148775886,
+      "learning_rate": 2e-05,
+      "loss": 0.4343,
+      "step": 6641
+    },
+    {
+      "epoch": 0.15809768637532134,
+      "grad_norm": 8.32180426788389,
+      "learning_rate": 2e-05,
+      "loss": 0.9737,
+      "step": 6642
+    },
+    {
+      "epoch": 0.15812148909835286,
+      "grad_norm": 5.219075811465048,
+      "learning_rate": 2e-05,
+      "loss": 0.3643,
+      "step": 6643
+    },
+    {
+      "epoch": 0.15814529182138437,
+      "grad_norm": 4.724742730271031,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 6644
+    },
+    {
+      "epoch": 0.15816909454441588,
+      "grad_norm": 11.761160784885416,
+      "learning_rate": 2e-05,
+      "loss": 1.237,
+      "step": 6645
+    },
+    {
+      "epoch": 0.1581928972674474,
+      "grad_norm": 3.221181467496374,
+      "learning_rate": 2e-05,
+      "loss": 0.5084,
+      "step": 6646
+    },
+    {
+      "epoch": 0.1582166999904789,
+      "grad_norm": 2.6516899636431543,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 6647
+    },
+    {
+      "epoch": 0.15824050271351042,
+      "grad_norm": 9.581402391220672,
+      "learning_rate": 2e-05,
+      "loss": 0.746,
+      "step": 6648
+    },
+    {
+      "epoch": 0.15826430543654194,
+      "grad_norm": 4.0653873747690525,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 6649
+    },
+    {
+      "epoch": 0.15828810815957345,
+      "grad_norm": 10.346545667289831,
+      "learning_rate": 2e-05,
+      "loss": 1.5766,
+      "step": 6650
+    },
+    {
+      "epoch": 0.15831191088260496,
+      "grad_norm": 4.589851441145508,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 6651
+    },
+    {
+      "epoch": 0.15833571360563647,
+      "grad_norm": 4.692292785288441,
+      "learning_rate": 2e-05,
+      "loss": 0.2642,
+      "step": 6652
+    },
+    {
+      "epoch": 0.158359516328668,
+      "grad_norm": 9.577346648967769,
+      "learning_rate": 2e-05,
+      "loss": 0.4793,
+      "step": 6653
+    },
+    {
+      "epoch": 0.15838331905169953,
+      "grad_norm": 14.21758174327861,
+      "learning_rate": 2e-05,
+      "loss": 0.8829,
+      "step": 6654
+    },
+    {
+      "epoch": 0.15840712177473104,
+      "grad_norm": 5.59354266545869,
+      "learning_rate": 2e-05,
+      "loss": 0.2991,
+      "step": 6655
+    },
+    {
+      "epoch": 0.15843092449776255,
+      "grad_norm": 2.306134382896716,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 6656
+    },
+    {
+      "epoch": 0.15845472722079407,
+      "grad_norm": 8.274745190476676,
+      "learning_rate": 2e-05,
+      "loss": 0.3669,
+      "step": 6657
+    },
+    {
+      "epoch": 0.15847852994382558,
+      "grad_norm": 5.621123248954976,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 6658
+    },
+    {
+      "epoch": 0.1585023326668571,
+      "grad_norm": 9.600608771311489,
+      "learning_rate": 2e-05,
+      "loss": 1.0098,
+      "step": 6659
+    },
+    {
+      "epoch": 0.1585261353898886,
+      "grad_norm": 8.677751512180377,
+      "learning_rate": 2e-05,
+      "loss": 1.0214,
+      "step": 6660
+    },
+    {
+      "epoch": 0.15854993811292012,
+      "grad_norm": 17.813993943820105,
+      "learning_rate": 2e-05,
+      "loss": 1.2236,
+      "step": 6661
+    },
+    {
+      "epoch": 0.15857374083595163,
+      "grad_norm": 6.395951999206728,
+      "learning_rate": 2e-05,
+      "loss": 0.9733,
+      "step": 6662
+    },
+    {
+      "epoch": 0.15859754355898315,
+      "grad_norm": 3.4430007262815776,
+      "learning_rate": 2e-05,
+      "loss": 0.177,
+      "step": 6663
+    },
+    {
+      "epoch": 0.15862134628201466,
+      "grad_norm": 5.2593891988109736,
+      "learning_rate": 2e-05,
+      "loss": 0.2623,
+      "step": 6664
+    },
+    {
+      "epoch": 0.15864514900504617,
+      "grad_norm": 7.712445260008672,
+      "learning_rate": 2e-05,
+      "loss": 0.6799,
+      "step": 6665
+    },
+    {
+      "epoch": 0.15866895172807768,
+      "grad_norm": 2.2448150388481305,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 6666
+    },
+    {
+      "epoch": 0.1586927544511092,
+      "grad_norm": 4.858861357561691,
+      "learning_rate": 2e-05,
+      "loss": 0.3005,
+      "step": 6667
+    },
+    {
+      "epoch": 0.1587165571741407,
+      "grad_norm": 2.7655090297513296,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 6668
+    },
+    {
+      "epoch": 0.15874035989717222,
+      "grad_norm": 4.141712625404244,
+      "learning_rate": 2e-05,
+      "loss": 0.3213,
+      "step": 6669
+    },
+    {
+      "epoch": 0.15876416262020376,
+      "grad_norm": 12.39859238154156,
+      "learning_rate": 2e-05,
+      "loss": 1.2808,
+      "step": 6670
+    },
+    {
+      "epoch": 0.15878796534323528,
+      "grad_norm": 48.90082803762089,
+      "learning_rate": 2e-05,
+      "loss": 0.6633,
+      "step": 6671
+    },
+    {
+      "epoch": 0.1588117680662668,
+      "grad_norm": 9.338167300779935,
+      "learning_rate": 2e-05,
+      "loss": 0.5346,
+      "step": 6672
+    },
+    {
+      "epoch": 0.1588355707892983,
+      "grad_norm": 4.1614371869625195,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 6673
+    },
+    {
+      "epoch": 0.15885937351232982,
+      "grad_norm": 1.384143911681959,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 6674
+    },
+    {
+      "epoch": 0.15888317623536133,
+      "grad_norm": 205.39053236373806,
+      "learning_rate": 2e-05,
+      "loss": 0.2765,
+      "step": 6675
+    },
+    {
+      "epoch": 0.15890697895839284,
+      "grad_norm": 5.517997621744721,
+      "learning_rate": 2e-05,
+      "loss": 0.9381,
+      "step": 6676
+    },
+    {
+      "epoch": 0.15893078168142435,
+      "grad_norm": 9.968810938295256,
+      "learning_rate": 2e-05,
+      "loss": 1.0273,
+      "step": 6677
+    },
+    {
+      "epoch": 0.15895458440445587,
+      "grad_norm": 3.7973791174087146,
+      "learning_rate": 2e-05,
+      "loss": 0.2644,
+      "step": 6678
+    },
+    {
+      "epoch": 0.15897838712748738,
+      "grad_norm": 6.217455538948318,
+      "learning_rate": 2e-05,
+      "loss": 0.9877,
+      "step": 6679
+    },
+    {
+      "epoch": 0.1590021898505189,
+      "grad_norm": 0.9969649365629765,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1590259925735504,
+      "grad_norm": 5.09694722693809,
+      "learning_rate": 2e-05,
+      "loss": 0.6508,
+      "step": 6681
+    },
+    {
+      "epoch": 0.15904979529658192,
+      "grad_norm": 12.352437840641478,
+      "learning_rate": 2e-05,
+      "loss": 0.9807,
+      "step": 6682
+    },
+    {
+      "epoch": 0.15907359801961343,
+      "grad_norm": 2.286695599636035,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 6683
+    },
+    {
+      "epoch": 0.15909740074264495,
+      "grad_norm": 3.9040516354359225,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 6684
+    },
+    {
+      "epoch": 0.1591212034656765,
+      "grad_norm": 6.838859264907603,
+      "learning_rate": 2e-05,
+      "loss": 1.5214,
+      "step": 6685
+    },
+    {
+      "epoch": 0.159145006188708,
+      "grad_norm": 8.821353503188774,
+      "learning_rate": 2e-05,
+      "loss": 1.1101,
+      "step": 6686
+    },
+    {
+      "epoch": 0.1591688089117395,
+      "grad_norm": 10.007172178850716,
+      "learning_rate": 2e-05,
+      "loss": 0.8829,
+      "step": 6687
+    },
+    {
+      "epoch": 0.15919261163477103,
+      "grad_norm": 8.695836806925529,
+      "learning_rate": 2e-05,
+      "loss": 1.2958,
+      "step": 6688
+    },
+    {
+      "epoch": 0.15921641435780254,
+      "grad_norm": 8.530984266723081,
+      "learning_rate": 2e-05,
+      "loss": 0.4085,
+      "step": 6689
+    },
+    {
+      "epoch": 0.15924021708083405,
+      "grad_norm": 9.755686523145966,
+      "learning_rate": 2e-05,
+      "loss": 0.3394,
+      "step": 6690
+    },
+    {
+      "epoch": 0.15926401980386556,
+      "grad_norm": 1.186017593234886,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 6691
+    },
+    {
+      "epoch": 0.15928782252689708,
+      "grad_norm": 8.79102951751115,
+      "learning_rate": 2e-05,
+      "loss": 1.0842,
+      "step": 6692
+    },
+    {
+      "epoch": 0.1593116252499286,
+      "grad_norm": 8.751235186938427,
+      "learning_rate": 2e-05,
+      "loss": 0.4301,
+      "step": 6693
+    },
+    {
+      "epoch": 0.1593354279729601,
+      "grad_norm": 4.345550143530904,
+      "learning_rate": 2e-05,
+      "loss": 0.771,
+      "step": 6694
+    },
+    {
+      "epoch": 0.15935923069599162,
+      "grad_norm": 8.179218357405231,
+      "learning_rate": 2e-05,
+      "loss": 1.6099,
+      "step": 6695
+    },
+    {
+      "epoch": 0.15938303341902313,
+      "grad_norm": 3.367289200397969,
+      "learning_rate": 2e-05,
+      "loss": 0.2512,
+      "step": 6696
+    },
+    {
+      "epoch": 0.15940683614205464,
+      "grad_norm": 5.177447202403904,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 6697
+    },
+    {
+      "epoch": 0.15943063886508616,
+      "grad_norm": 4.678509827940187,
+      "learning_rate": 2e-05,
+      "loss": 1.066,
+      "step": 6698
+    },
+    {
+      "epoch": 0.15945444158811767,
+      "grad_norm": 2.4742887585161246,
+      "learning_rate": 2e-05,
+      "loss": 0.212,
+      "step": 6699
+    },
+    {
+      "epoch": 0.1594782443111492,
+      "grad_norm": 2.778048721198691,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 6700
+    },
+    {
+      "epoch": 0.15950204703418072,
+      "grad_norm": 6.367597418384385,
+      "learning_rate": 2e-05,
+      "loss": 1.6012,
+      "step": 6701
+    },
+    {
+      "epoch": 0.15952584975721223,
+      "grad_norm": 5.911776189311258,
+      "learning_rate": 2e-05,
+      "loss": 0.2687,
+      "step": 6702
+    },
+    {
+      "epoch": 0.15954965248024375,
+      "grad_norm": 5.235362813006248,
+      "learning_rate": 2e-05,
+      "loss": 0.2776,
+      "step": 6703
+    },
+    {
+      "epoch": 0.15957345520327526,
+      "grad_norm": 5.011499043724648,
+      "learning_rate": 2e-05,
+      "loss": 0.1683,
+      "step": 6704
+    },
+    {
+      "epoch": 0.15959725792630677,
+      "grad_norm": 7.090258273651994,
+      "learning_rate": 2e-05,
+      "loss": 0.8438,
+      "step": 6705
+    },
+    {
+      "epoch": 0.1596210606493383,
+      "grad_norm": 5.797180243574233,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 6706
+    },
+    {
+      "epoch": 0.1596448633723698,
+      "grad_norm": 3.7473102792718356,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 6707
+    },
+    {
+      "epoch": 0.1596686660954013,
+      "grad_norm": 9.962648836580705,
+      "learning_rate": 2e-05,
+      "loss": 0.4007,
+      "step": 6708
+    },
+    {
+      "epoch": 0.15969246881843283,
+      "grad_norm": 3.0761439764127534,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 6709
+    },
+    {
+      "epoch": 0.15971627154146434,
+      "grad_norm": 4.077221507941836,
+      "learning_rate": 2e-05,
+      "loss": 0.1046,
+      "step": 6710
+    },
+    {
+      "epoch": 0.15974007426449585,
+      "grad_norm": 2.9673219042889714,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 6711
+    },
+    {
+      "epoch": 0.15976387698752736,
+      "grad_norm": 5.936035754426252,
+      "learning_rate": 2e-05,
+      "loss": 0.7273,
+      "step": 6712
+    },
+    {
+      "epoch": 0.15978767971055888,
+      "grad_norm": 3.2144345468107733,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 6713
+    },
+    {
+      "epoch": 0.1598114824335904,
+      "grad_norm": 3.3762951884965764,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 6714
+    },
+    {
+      "epoch": 0.15983528515662193,
+      "grad_norm": 5.181170047983322,
+      "learning_rate": 2e-05,
+      "loss": 0.6726,
+      "step": 6715
+    },
+    {
+      "epoch": 0.15985908787965344,
+      "grad_norm": 2.3713905720274373,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 6716
+    },
+    {
+      "epoch": 0.15988289060268496,
+      "grad_norm": 7.48376754964789,
+      "learning_rate": 2e-05,
+      "loss": 0.9584,
+      "step": 6717
+    },
+    {
+      "epoch": 0.15990669332571647,
+      "grad_norm": 4.69645373880449,
+      "learning_rate": 2e-05,
+      "loss": 0.9355,
+      "step": 6718
+    },
+    {
+      "epoch": 0.15993049604874798,
+      "grad_norm": 6.56966539627927,
+      "learning_rate": 2e-05,
+      "loss": 0.2808,
+      "step": 6719
+    },
+    {
+      "epoch": 0.1599542987717795,
+      "grad_norm": 3.7568694356067174,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 6720
+    },
+    {
+      "epoch": 0.159978101494811,
+      "grad_norm": 14.010208619642626,
+      "learning_rate": 2e-05,
+      "loss": 1.1122,
+      "step": 6721
+    },
+    {
+      "epoch": 0.16000190421784252,
+      "grad_norm": 13.809293392407922,
+      "learning_rate": 2e-05,
+      "loss": 0.7862,
+      "step": 6722
+    },
+    {
+      "epoch": 0.16002570694087404,
+      "grad_norm": 15.648082058293005,
+      "learning_rate": 2e-05,
+      "loss": 0.7822,
+      "step": 6723
+    },
+    {
+      "epoch": 0.16004950966390555,
+      "grad_norm": 3.6199492598622283,
+      "learning_rate": 2e-05,
+      "loss": 0.2521,
+      "step": 6724
+    },
+    {
+      "epoch": 0.16007331238693706,
+      "grad_norm": 2.8386276685872875,
+      "learning_rate": 2e-05,
+      "loss": 0.217,
+      "step": 6725
+    },
+    {
+      "epoch": 0.16009711510996857,
+      "grad_norm": 8.50391284377708,
+      "learning_rate": 2e-05,
+      "loss": 1.0091,
+      "step": 6726
+    },
+    {
+      "epoch": 0.1601209178330001,
+      "grad_norm": 5.837564169343749,
+      "learning_rate": 2e-05,
+      "loss": 0.2916,
+      "step": 6727
+    },
+    {
+      "epoch": 0.1601447205560316,
+      "grad_norm": 7.259367770696954,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 6728
+    },
+    {
+      "epoch": 0.1601685232790631,
+      "grad_norm": 6.358205249379292,
+      "learning_rate": 2e-05,
+      "loss": 0.9886,
+      "step": 6729
+    },
+    {
+      "epoch": 0.16019232600209463,
+      "grad_norm": 5.618773286766888,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 6730
+    },
+    {
+      "epoch": 0.16021612872512617,
+      "grad_norm": 3.4870277106883054,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 6731
+    },
+    {
+      "epoch": 0.16023993144815768,
+      "grad_norm": 7.160914424035882,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 6732
+    },
+    {
+      "epoch": 0.1602637341711892,
+      "grad_norm": 4.004144624394978,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 6733
+    },
+    {
+      "epoch": 0.1602875368942207,
+      "grad_norm": 7.897316687915187,
+      "learning_rate": 2e-05,
+      "loss": 0.2929,
+      "step": 6734
+    },
+    {
+      "epoch": 0.16031133961725222,
+      "grad_norm": 6.340512512369419,
+      "learning_rate": 2e-05,
+      "loss": 0.3138,
+      "step": 6735
+    },
+    {
+      "epoch": 0.16033514234028373,
+      "grad_norm": 2.980921002784521,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 6736
+    },
+    {
+      "epoch": 0.16035894506331524,
+      "grad_norm": 3.4178532547972607,
+      "learning_rate": 2e-05,
+      "loss": 0.3792,
+      "step": 6737
+    },
+    {
+      "epoch": 0.16038274778634676,
+      "grad_norm": 8.429056513331403,
+      "learning_rate": 2e-05,
+      "loss": 1.3127,
+      "step": 6738
+    },
+    {
+      "epoch": 0.16040655050937827,
+      "grad_norm": 3.0669469689394107,
+      "learning_rate": 2e-05,
+      "loss": 0.2765,
+      "step": 6739
+    },
+    {
+      "epoch": 0.16043035323240978,
+      "grad_norm": 4.603207655653615,
+      "learning_rate": 2e-05,
+      "loss": 0.6513,
+      "step": 6740
+    },
+    {
+      "epoch": 0.1604541559554413,
+      "grad_norm": 5.04816847524574,
+      "learning_rate": 2e-05,
+      "loss": 0.2768,
+      "step": 6741
+    },
+    {
+      "epoch": 0.1604779586784728,
+      "grad_norm": 8.385271117407475,
+      "learning_rate": 2e-05,
+      "loss": 0.3066,
+      "step": 6742
+    },
+    {
+      "epoch": 0.16050176140150432,
+      "grad_norm": 33.69702012761269,
+      "learning_rate": 2e-05,
+      "loss": 0.7183,
+      "step": 6743
+    },
+    {
+      "epoch": 0.16052556412453584,
+      "grad_norm": 1.8867517750140332,
+      "learning_rate": 2e-05,
+      "loss": 0.1138,
+      "step": 6744
+    },
+    {
+      "epoch": 0.16054936684756735,
+      "grad_norm": 5.709824049644597,
+      "learning_rate": 2e-05,
+      "loss": 0.1662,
+      "step": 6745
+    },
+    {
+      "epoch": 0.1605731695705989,
+      "grad_norm": 5.289029136023813,
+      "learning_rate": 2e-05,
+      "loss": 0.5024,
+      "step": 6746
+    },
+    {
+      "epoch": 0.1605969722936304,
+      "grad_norm": 5.143024406555926,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 6747
+    },
+    {
+      "epoch": 0.16062077501666192,
+      "grad_norm": 7.090298055548788,
+      "learning_rate": 2e-05,
+      "loss": 0.8546,
+      "step": 6748
+    },
+    {
+      "epoch": 0.16064457773969343,
+      "grad_norm": 219.00609535442007,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 6749
+    },
+    {
+      "epoch": 0.16066838046272494,
+      "grad_norm": 2.7703669191101135,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 6750
+    },
+    {
+      "epoch": 0.16069218318575645,
+      "grad_norm": 22.183837169863093,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 6751
+    },
+    {
+      "epoch": 0.16071598590878797,
+      "grad_norm": 2.2160130927132915,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 6752
+    },
+    {
+      "epoch": 0.16073978863181948,
+      "grad_norm": 5.035819137302277,
+      "learning_rate": 2e-05,
+      "loss": 0.688,
+      "step": 6753
+    },
+    {
+      "epoch": 0.160763591354851,
+      "grad_norm": 9.999015362430878,
+      "learning_rate": 2e-05,
+      "loss": 0.6895,
+      "step": 6754
+    },
+    {
+      "epoch": 0.1607873940778825,
+      "grad_norm": 2.279667201932817,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 6755
+    },
+    {
+      "epoch": 0.16081119680091402,
+      "grad_norm": 4.806188530676533,
+      "learning_rate": 2e-05,
+      "loss": 0.6233,
+      "step": 6756
+    },
+    {
+      "epoch": 0.16083499952394553,
+      "grad_norm": 13.53883971497789,
+      "learning_rate": 2e-05,
+      "loss": 0.9078,
+      "step": 6757
+    },
+    {
+      "epoch": 0.16085880224697705,
+      "grad_norm": 21.47435698147785,
+      "learning_rate": 2e-05,
+      "loss": 0.6043,
+      "step": 6758
+    },
+    {
+      "epoch": 0.16088260497000856,
+      "grad_norm": 3.5884735533253016,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 6759
+    },
+    {
+      "epoch": 0.16090640769304007,
+      "grad_norm": 4.68915754224105,
+      "learning_rate": 2e-05,
+      "loss": 0.8206,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1609302104160716,
+      "grad_norm": 2.9389960146491227,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 6761
+    },
+    {
+      "epoch": 0.16095401313910312,
+      "grad_norm": 3.2193358823646196,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 6762
+    },
+    {
+      "epoch": 0.16097781586213464,
+      "grad_norm": 5.718428443174065,
+      "learning_rate": 2e-05,
+      "loss": 0.8439,
+      "step": 6763
+    },
+    {
+      "epoch": 0.16100161858516615,
+      "grad_norm": 11.803133809687118,
+      "learning_rate": 2e-05,
+      "loss": 0.5045,
+      "step": 6764
+    },
+    {
+      "epoch": 0.16102542130819766,
+      "grad_norm": 2.7987284425566386,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 6765
+    },
+    {
+      "epoch": 0.16104922403122918,
+      "grad_norm": 4.943606321733826,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 6766
+    },
+    {
+      "epoch": 0.1610730267542607,
+      "grad_norm": 1.9085254311021327,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 6767
+    },
+    {
+      "epoch": 0.1610968294772922,
+      "grad_norm": 1.6753053288611572,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 6768
+    },
+    {
+      "epoch": 0.16112063220032372,
+      "grad_norm": 9.934862976243654,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 6769
+    },
+    {
+      "epoch": 0.16114443492335523,
+      "grad_norm": 4.653647269729826,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 6770
+    },
+    {
+      "epoch": 0.16116823764638674,
+      "grad_norm": 4.184755814931482,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 6771
+    },
+    {
+      "epoch": 0.16119204036941825,
+      "grad_norm": 4.8479854965595,
+      "learning_rate": 2e-05,
+      "loss": 0.5575,
+      "step": 6772
+    },
+    {
+      "epoch": 0.16121584309244977,
+      "grad_norm": 1.748346005454777,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 6773
+    },
+    {
+      "epoch": 0.16123964581548128,
+      "grad_norm": 13.308325765581957,
+      "learning_rate": 2e-05,
+      "loss": 1.5163,
+      "step": 6774
+    },
+    {
+      "epoch": 0.1612634485385128,
+      "grad_norm": 3.261221260353048,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 6775
+    },
+    {
+      "epoch": 0.16128725126154433,
+      "grad_norm": 12.010111697606712,
+      "learning_rate": 2e-05,
+      "loss": 0.6578,
+      "step": 6776
+    },
+    {
+      "epoch": 0.16131105398457585,
+      "grad_norm": 5.559437876412284,
+      "learning_rate": 2e-05,
+      "loss": 1.053,
+      "step": 6777
+    },
+    {
+      "epoch": 0.16133485670760736,
+      "grad_norm": 7.8946228299746455,
+      "learning_rate": 2e-05,
+      "loss": 0.8352,
+      "step": 6778
+    },
+    {
+      "epoch": 0.16135865943063887,
+      "grad_norm": 3.4395063032726023,
+      "learning_rate": 2e-05,
+      "loss": 0.4631,
+      "step": 6779
+    },
+    {
+      "epoch": 0.1613824621536704,
+      "grad_norm": 1.2957293611376999,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 6780
+    },
+    {
+      "epoch": 0.1614062648767019,
+      "grad_norm": 7.0155915495888275,
+      "learning_rate": 2e-05,
+      "loss": 0.312,
+      "step": 6781
+    },
+    {
+      "epoch": 0.1614300675997334,
+      "grad_norm": 4.427866518817178,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 6782
+    },
+    {
+      "epoch": 0.16145387032276493,
+      "grad_norm": 3.732252759727274,
+      "learning_rate": 2e-05,
+      "loss": 0.2127,
+      "step": 6783
+    },
+    {
+      "epoch": 0.16147767304579644,
+      "grad_norm": 15.907296583371947,
+      "learning_rate": 2e-05,
+      "loss": 1.1705,
+      "step": 6784
+    },
+    {
+      "epoch": 0.16150147576882795,
+      "grad_norm": 2.1744305040456453,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 6785
+    },
+    {
+      "epoch": 0.16152527849185946,
+      "grad_norm": 8.763773213406353,
+      "learning_rate": 2e-05,
+      "loss": 0.8455,
+      "step": 6786
+    },
+    {
+      "epoch": 0.16154908121489098,
+      "grad_norm": 5.173864024591227,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 6787
+    },
+    {
+      "epoch": 0.1615728839379225,
+      "grad_norm": 14.983575830952066,
+      "learning_rate": 2e-05,
+      "loss": 1.1158,
+      "step": 6788
+    },
+    {
+      "epoch": 0.161596686660954,
+      "grad_norm": 6.308817073859772,
+      "learning_rate": 2e-05,
+      "loss": 0.2802,
+      "step": 6789
+    },
+    {
+      "epoch": 0.16162048938398552,
+      "grad_norm": 2.6950994400510435,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 6790
+    },
+    {
+      "epoch": 0.16164429210701703,
+      "grad_norm": 3.97113040005599,
+      "learning_rate": 2e-05,
+      "loss": 0.27,
+      "step": 6791
+    },
+    {
+      "epoch": 0.16166809483004857,
+      "grad_norm": 1.3639362351860718,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 6792
+    },
+    {
+      "epoch": 0.16169189755308008,
+      "grad_norm": 8.228174288375238,
+      "learning_rate": 2e-05,
+      "loss": 0.3255,
+      "step": 6793
+    },
+    {
+      "epoch": 0.1617157002761116,
+      "grad_norm": 3.727054773986279,
+      "learning_rate": 2e-05,
+      "loss": 0.118,
+      "step": 6794
+    },
+    {
+      "epoch": 0.1617395029991431,
+      "grad_norm": 24.40515613597498,
+      "learning_rate": 2e-05,
+      "loss": 0.4226,
+      "step": 6795
+    },
+    {
+      "epoch": 0.16176330572217462,
+      "grad_norm": 1.656033787363332,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 6796
+    },
+    {
+      "epoch": 0.16178710844520613,
+      "grad_norm": 4.503194355478873,
+      "learning_rate": 2e-05,
+      "loss": 0.2837,
+      "step": 6797
+    },
+    {
+      "epoch": 0.16181091116823765,
+      "grad_norm": 1.2388516266271208,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 6798
+    },
+    {
+      "epoch": 0.16183471389126916,
+      "grad_norm": 8.227760310814947,
+      "learning_rate": 2e-05,
+      "loss": 1.1287,
+      "step": 6799
+    },
+    {
+      "epoch": 0.16185851661430067,
+      "grad_norm": 7.476689350522976,
+      "learning_rate": 2e-05,
+      "loss": 0.8573,
+      "step": 6800
+    },
+    {
+      "epoch": 0.1618823193373322,
+      "grad_norm": 3.794827255999415,
+      "learning_rate": 2e-05,
+      "loss": 0.439,
+      "step": 6801
+    },
+    {
+      "epoch": 0.1619061220603637,
+      "grad_norm": 2.817915276286406,
+      "learning_rate": 2e-05,
+      "loss": 0.227,
+      "step": 6802
+    },
+    {
+      "epoch": 0.1619299247833952,
+      "grad_norm": 8.107347824790134,
+      "learning_rate": 2e-05,
+      "loss": 0.297,
+      "step": 6803
+    },
+    {
+      "epoch": 0.16195372750642673,
+      "grad_norm": 7.056609909192699,
+      "learning_rate": 2e-05,
+      "loss": 0.8538,
+      "step": 6804
+    },
+    {
+      "epoch": 0.16197753022945824,
+      "grad_norm": 2.459988285101887,
+      "learning_rate": 2e-05,
+      "loss": 0.1669,
+      "step": 6805
+    },
+    {
+      "epoch": 0.16200133295248975,
+      "grad_norm": 3.3046023502275554,
+      "learning_rate": 2e-05,
+      "loss": 0.157,
+      "step": 6806
+    },
+    {
+      "epoch": 0.1620251356755213,
+      "grad_norm": 9.757978063207958,
+      "learning_rate": 2e-05,
+      "loss": 1.0204,
+      "step": 6807
+    },
+    {
+      "epoch": 0.1620489383985528,
+      "grad_norm": 4.85685455149907,
+      "learning_rate": 2e-05,
+      "loss": 0.6312,
+      "step": 6808
+    },
+    {
+      "epoch": 0.16207274112158432,
+      "grad_norm": 121.77286861919286,
+      "learning_rate": 2e-05,
+      "loss": 0.475,
+      "step": 6809
+    },
+    {
+      "epoch": 0.16209654384461583,
+      "grad_norm": 2.3949719594248307,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16212034656764734,
+      "grad_norm": 6.336850271091669,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 6811
+    },
+    {
+      "epoch": 0.16214414929067886,
+      "grad_norm": 4.283869825358692,
+      "learning_rate": 2e-05,
+      "loss": 0.3878,
+      "step": 6812
+    },
+    {
+      "epoch": 0.16216795201371037,
+      "grad_norm": 7.554790347326891,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 6813
+    },
+    {
+      "epoch": 0.16219175473674188,
+      "grad_norm": 10.816986745715797,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 6814
+    },
+    {
+      "epoch": 0.1622155574597734,
+      "grad_norm": 4.102194934705146,
+      "learning_rate": 2e-05,
+      "loss": 0.5152,
+      "step": 6815
+    },
+    {
+      "epoch": 0.1622393601828049,
+      "grad_norm": 8.34283635075617,
+      "learning_rate": 2e-05,
+      "loss": 0.2619,
+      "step": 6816
+    },
+    {
+      "epoch": 0.16226316290583642,
+      "grad_norm": 6.787929471308212,
+      "learning_rate": 2e-05,
+      "loss": 0.7017,
+      "step": 6817
+    },
+    {
+      "epoch": 0.16228696562886794,
+      "grad_norm": 5.650695493516138,
+      "learning_rate": 2e-05,
+      "loss": 1.1887,
+      "step": 6818
+    },
+    {
+      "epoch": 0.16231076835189945,
+      "grad_norm": 5.416110547957531,
+      "learning_rate": 2e-05,
+      "loss": 0.7619,
+      "step": 6819
+    },
+    {
+      "epoch": 0.16233457107493096,
+      "grad_norm": 10.927479980962922,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 6820
+    },
+    {
+      "epoch": 0.16235837379796247,
+      "grad_norm": 3.172498960379779,
+      "learning_rate": 2e-05,
+      "loss": 0.2305,
+      "step": 6821
+    },
+    {
+      "epoch": 0.16238217652099401,
+      "grad_norm": 3.7559777742391,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 6822
+    },
+    {
+      "epoch": 0.16240597924402553,
+      "grad_norm": 7.262946892250349,
+      "learning_rate": 2e-05,
+      "loss": 0.2552,
+      "step": 6823
+    },
+    {
+      "epoch": 0.16242978196705704,
+      "grad_norm": 2.785646304554813,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 6824
+    },
+    {
+      "epoch": 0.16245358469008855,
+      "grad_norm": 4.004314453160296,
+      "learning_rate": 2e-05,
+      "loss": 0.5673,
+      "step": 6825
+    },
+    {
+      "epoch": 0.16247738741312007,
+      "grad_norm": 6.856228714476234,
+      "learning_rate": 2e-05,
+      "loss": 0.5521,
+      "step": 6826
+    },
+    {
+      "epoch": 0.16250119013615158,
+      "grad_norm": 3.1598939266891617,
+      "learning_rate": 2e-05,
+      "loss": 0.1826,
+      "step": 6827
+    },
+    {
+      "epoch": 0.1625249928591831,
+      "grad_norm": 3.699348738646155,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 6828
+    },
+    {
+      "epoch": 0.1625487955822146,
+      "grad_norm": 14.049316106518788,
+      "learning_rate": 2e-05,
+      "loss": 0.598,
+      "step": 6829
+    },
+    {
+      "epoch": 0.16257259830524612,
+      "grad_norm": 12.693187827379777,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16259640102827763,
+      "grad_norm": 2.9581503463098517,
+      "learning_rate": 2e-05,
+      "loss": 0.2123,
+      "step": 6831
+    },
+    {
+      "epoch": 0.16262020375130914,
+      "grad_norm": 4.73315146844897,
+      "learning_rate": 2e-05,
+      "loss": 0.5762,
+      "step": 6832
+    },
+    {
+      "epoch": 0.16264400647434066,
+      "grad_norm": 6.670366334661905,
+      "learning_rate": 2e-05,
+      "loss": 0.4453,
+      "step": 6833
+    },
+    {
+      "epoch": 0.16266780919737217,
+      "grad_norm": 4.055976351795325,
+      "learning_rate": 2e-05,
+      "loss": 0.5766,
+      "step": 6834
+    },
+    {
+      "epoch": 0.16269161192040368,
+      "grad_norm": 6.825740658311766,
+      "learning_rate": 2e-05,
+      "loss": 0.4011,
+      "step": 6835
+    },
+    {
+      "epoch": 0.1627154146434352,
+      "grad_norm": 3.1055374523992687,
+      "learning_rate": 2e-05,
+      "loss": 0.1309,
+      "step": 6836
+    },
+    {
+      "epoch": 0.16273921736646674,
+      "grad_norm": 3.7305839895836206,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 6837
+    },
+    {
+      "epoch": 0.16276302008949825,
+      "grad_norm": 3.2403057561680133,
+      "learning_rate": 2e-05,
+      "loss": 0.2023,
+      "step": 6838
+    },
+    {
+      "epoch": 0.16278682281252976,
+      "grad_norm": 1.6491090631268823,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 6839
+    },
+    {
+      "epoch": 0.16281062553556128,
+      "grad_norm": 3.882012064332345,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 6840
+    },
+    {
+      "epoch": 0.1628344282585928,
+      "grad_norm": 6.52365261295573,
+      "learning_rate": 2e-05,
+      "loss": 0.9334,
+      "step": 6841
+    },
+    {
+      "epoch": 0.1628582309816243,
+      "grad_norm": 1.7939037456065705,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 6842
+    },
+    {
+      "epoch": 0.16288203370465582,
+      "grad_norm": 4.3407536586885955,
+      "learning_rate": 2e-05,
+      "loss": 0.1805,
+      "step": 6843
+    },
+    {
+      "epoch": 0.16290583642768733,
+      "grad_norm": 4.684700629113748,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 6844
+    },
+    {
+      "epoch": 0.16292963915071884,
+      "grad_norm": 10.577887241064708,
+      "learning_rate": 2e-05,
+      "loss": 0.4015,
+      "step": 6845
+    },
+    {
+      "epoch": 0.16295344187375035,
+      "grad_norm": 6.260694370576632,
+      "learning_rate": 2e-05,
+      "loss": 1.0269,
+      "step": 6846
+    },
+    {
+      "epoch": 0.16297724459678187,
+      "grad_norm": 4.867286256936227,
+      "learning_rate": 2e-05,
+      "loss": 0.8209,
+      "step": 6847
+    },
+    {
+      "epoch": 0.16300104731981338,
+      "grad_norm": 11.436798680044676,
+      "learning_rate": 2e-05,
+      "loss": 0.4939,
+      "step": 6848
+    },
+    {
+      "epoch": 0.1630248500428449,
+      "grad_norm": 1.6074239419461256,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 6849
+    },
+    {
+      "epoch": 0.1630486527658764,
+      "grad_norm": 3.643645696508478,
+      "learning_rate": 2e-05,
+      "loss": 0.1847,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16307245548890792,
+      "grad_norm": 13.908875052037537,
+      "learning_rate": 2e-05,
+      "loss": 0.5326,
+      "step": 6851
+    },
+    {
+      "epoch": 0.16309625821193943,
+      "grad_norm": 3.099726994739542,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 6852
+    },
+    {
+      "epoch": 0.16312006093497097,
+      "grad_norm": 9.102395610065027,
+      "learning_rate": 2e-05,
+      "loss": 0.3775,
+      "step": 6853
+    },
+    {
+      "epoch": 0.16314386365800249,
+      "grad_norm": 13.264963691188019,
+      "learning_rate": 2e-05,
+      "loss": 1.3918,
+      "step": 6854
+    },
+    {
+      "epoch": 0.163167666381034,
+      "grad_norm": 4.7469476090448,
+      "learning_rate": 2e-05,
+      "loss": 0.3139,
+      "step": 6855
+    },
+    {
+      "epoch": 0.1631914691040655,
+      "grad_norm": 2.9129858342765167,
+      "learning_rate": 2e-05,
+      "loss": 0.137,
+      "step": 6856
+    },
+    {
+      "epoch": 0.16321527182709702,
+      "grad_norm": 4.726540603170268,
+      "learning_rate": 2e-05,
+      "loss": 0.5334,
+      "step": 6857
+    },
+    {
+      "epoch": 0.16323907455012854,
+      "grad_norm": 9.806092177152014,
+      "learning_rate": 2e-05,
+      "loss": 1.2179,
+      "step": 6858
+    },
+    {
+      "epoch": 0.16326287727316005,
+      "grad_norm": 19.24293151435851,
+      "learning_rate": 2e-05,
+      "loss": 1.3334,
+      "step": 6859
+    },
+    {
+      "epoch": 0.16328667999619156,
+      "grad_norm": 9.79991427202362,
+      "learning_rate": 2e-05,
+      "loss": 1.2069,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16331048271922308,
+      "grad_norm": 10.33049614514673,
+      "learning_rate": 2e-05,
+      "loss": 1.6897,
+      "step": 6861
+    },
+    {
+      "epoch": 0.1633342854422546,
+      "grad_norm": 5.27526957800746,
+      "learning_rate": 2e-05,
+      "loss": 0.7418,
+      "step": 6862
+    },
+    {
+      "epoch": 0.1633580881652861,
+      "grad_norm": 6.477026672396044,
+      "learning_rate": 2e-05,
+      "loss": 1.1882,
+      "step": 6863
+    },
+    {
+      "epoch": 0.16338189088831762,
+      "grad_norm": 1.1094740327305077,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 6864
+    },
+    {
+      "epoch": 0.16340569361134913,
+      "grad_norm": 6.54126646518194,
+      "learning_rate": 2e-05,
+      "loss": 1.1686,
+      "step": 6865
+    },
+    {
+      "epoch": 0.16342949633438064,
+      "grad_norm": 3.9694856124848323,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 6866
+    },
+    {
+      "epoch": 0.16345329905741215,
+      "grad_norm": 8.792177192709428,
+      "learning_rate": 2e-05,
+      "loss": 0.3728,
+      "step": 6867
+    },
+    {
+      "epoch": 0.1634771017804437,
+      "grad_norm": 5.276389280151382,
+      "learning_rate": 2e-05,
+      "loss": 0.878,
+      "step": 6868
+    },
+    {
+      "epoch": 0.1635009045034752,
+      "grad_norm": 10.040106070098254,
+      "learning_rate": 2e-05,
+      "loss": 0.8015,
+      "step": 6869
+    },
+    {
+      "epoch": 0.16352470722650672,
+      "grad_norm": 2.9747645976660215,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 6870
+    },
+    {
+      "epoch": 0.16354850994953823,
+      "grad_norm": 3.899396466781724,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 6871
+    },
+    {
+      "epoch": 0.16357231267256975,
+      "grad_norm": 18.374961336465905,
+      "learning_rate": 2e-05,
+      "loss": 0.915,
+      "step": 6872
+    },
+    {
+      "epoch": 0.16359611539560126,
+      "grad_norm": 100.43506840266718,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 6873
+    },
+    {
+      "epoch": 0.16361991811863277,
+      "grad_norm": 3.1340035712090937,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 6874
+    },
+    {
+      "epoch": 0.1636437208416643,
+      "grad_norm": 4.875786416281001,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 6875
+    },
+    {
+      "epoch": 0.1636675235646958,
+      "grad_norm": 2.4070455808979956,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 6876
+    },
+    {
+      "epoch": 0.1636913262877273,
+      "grad_norm": 13.753060544790692,
+      "learning_rate": 2e-05,
+      "loss": 0.3348,
+      "step": 6877
+    },
+    {
+      "epoch": 0.16371512901075883,
+      "grad_norm": 2.5451045333245035,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 6878
+    },
+    {
+      "epoch": 0.16373893173379034,
+      "grad_norm": 3.2602369301178338,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 6879
+    },
+    {
+      "epoch": 0.16376273445682185,
+      "grad_norm": 3.9263902879298467,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16378653717985336,
+      "grad_norm": 5.217284510988591,
+      "learning_rate": 2e-05,
+      "loss": 0.6587,
+      "step": 6881
+    },
+    {
+      "epoch": 0.16381033990288488,
+      "grad_norm": 3.662222316179795,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 6882
+    },
+    {
+      "epoch": 0.16383414262591642,
+      "grad_norm": 4.600503658219522,
+      "learning_rate": 2e-05,
+      "loss": 0.1712,
+      "step": 6883
+    },
+    {
+      "epoch": 0.16385794534894793,
+      "grad_norm": 1.395304929134021,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 6884
+    },
+    {
+      "epoch": 0.16388174807197944,
+      "grad_norm": 2.898214578812223,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 6885
+    },
+    {
+      "epoch": 0.16390555079501096,
+      "grad_norm": 4.4470578819911335,
+      "learning_rate": 2e-05,
+      "loss": 0.5711,
+      "step": 6886
+    },
+    {
+      "epoch": 0.16392935351804247,
+      "grad_norm": 10.280868104049903,
+      "learning_rate": 2e-05,
+      "loss": 0.5994,
+      "step": 6887
+    },
+    {
+      "epoch": 0.16395315624107398,
+      "grad_norm": 48.17753583233437,
+      "learning_rate": 2e-05,
+      "loss": 0.389,
+      "step": 6888
+    },
+    {
+      "epoch": 0.1639769589641055,
+      "grad_norm": 5.852364854474705,
+      "learning_rate": 2e-05,
+      "loss": 1.07,
+      "step": 6889
+    },
+    {
+      "epoch": 0.164000761687137,
+      "grad_norm": 4.353338629526814,
+      "learning_rate": 2e-05,
+      "loss": 0.591,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16402456441016852,
+      "grad_norm": 3.156502286260069,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 6891
+    },
+    {
+      "epoch": 0.16404836713320003,
+      "grad_norm": 5.462076521632772,
+      "learning_rate": 2e-05,
+      "loss": 0.8383,
+      "step": 6892
+    },
+    {
+      "epoch": 0.16407216985623155,
+      "grad_norm": 12.744098331827189,
+      "learning_rate": 2e-05,
+      "loss": 0.4936,
+      "step": 6893
+    },
+    {
+      "epoch": 0.16409597257926306,
+      "grad_norm": 598.206538368853,
+      "learning_rate": 2e-05,
+      "loss": 0.4733,
+      "step": 6894
+    },
+    {
+      "epoch": 0.16411977530229457,
+      "grad_norm": 1.8074558130327791,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 6895
+    },
+    {
+      "epoch": 0.1641435780253261,
+      "grad_norm": 10.025215028140439,
+      "learning_rate": 2e-05,
+      "loss": 0.5148,
+      "step": 6896
+    },
+    {
+      "epoch": 0.1641673807483576,
+      "grad_norm": 3.4545290487889417,
+      "learning_rate": 2e-05,
+      "loss": 0.5645,
+      "step": 6897
+    },
+    {
+      "epoch": 0.16419118347138914,
+      "grad_norm": 4.323438048990771,
+      "learning_rate": 2e-05,
+      "loss": 0.2436,
+      "step": 6898
+    },
+    {
+      "epoch": 0.16421498619442065,
+      "grad_norm": 1.5780269460042327,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 6899
+    },
+    {
+      "epoch": 0.16423878891745217,
+      "grad_norm": 7.9276157022143465,
+      "learning_rate": 2e-05,
+      "loss": 1.7061,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16426259164048368,
+      "grad_norm": 5.036009658325434,
+      "learning_rate": 2e-05,
+      "loss": 0.8364,
+      "step": 6901
+    },
+    {
+      "epoch": 0.1642863943635152,
+      "grad_norm": 1.9149653503808952,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 6902
+    },
+    {
+      "epoch": 0.1643101970865467,
+      "grad_norm": 6.996834185441183,
+      "learning_rate": 2e-05,
+      "loss": 1.2459,
+      "step": 6903
+    },
+    {
+      "epoch": 0.16433399980957822,
+      "grad_norm": 9.20936076034121,
+      "learning_rate": 2e-05,
+      "loss": 0.3983,
+      "step": 6904
+    },
+    {
+      "epoch": 0.16435780253260973,
+      "grad_norm": 10.499688376036628,
+      "learning_rate": 2e-05,
+      "loss": 1.1517,
+      "step": 6905
+    },
+    {
+      "epoch": 0.16438160525564124,
+      "grad_norm": 6.734220753694977,
+      "learning_rate": 2e-05,
+      "loss": 1.07,
+      "step": 6906
+    },
+    {
+      "epoch": 0.16440540797867276,
+      "grad_norm": 2.9297156083437903,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 6907
+    },
+    {
+      "epoch": 0.16442921070170427,
+      "grad_norm": 4.913507558590686,
+      "learning_rate": 2e-05,
+      "loss": 0.6972,
+      "step": 6908
+    },
+    {
+      "epoch": 0.16445301342473578,
+      "grad_norm": 4.006956479346471,
+      "learning_rate": 2e-05,
+      "loss": 0.3603,
+      "step": 6909
+    },
+    {
+      "epoch": 0.1644768161477673,
+      "grad_norm": 2.733597408843775,
+      "learning_rate": 2e-05,
+      "loss": 0.1507,
+      "step": 6910
+    },
+    {
+      "epoch": 0.1645006188707988,
+      "grad_norm": 8.304985051485355,
+      "learning_rate": 2e-05,
+      "loss": 0.4465,
+      "step": 6911
+    },
+    {
+      "epoch": 0.16452442159383032,
+      "grad_norm": 3.412078216585933,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 6912
+    },
+    {
+      "epoch": 0.16454822431686184,
+      "grad_norm": 2.7557533023799445,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 6913
+    },
+    {
+      "epoch": 0.16457202703989338,
+      "grad_norm": 1.4196208209886114,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 6914
+    },
+    {
+      "epoch": 0.1645958297629249,
+      "grad_norm": 4.503262469744044,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 6915
+    },
+    {
+      "epoch": 0.1646196324859564,
+      "grad_norm": 7.31498729403946,
+      "learning_rate": 2e-05,
+      "loss": 0.5839,
+      "step": 6916
+    },
+    {
+      "epoch": 0.16464343520898791,
+      "grad_norm": 8.961540784745615,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 6917
+    },
+    {
+      "epoch": 0.16466723793201943,
+      "grad_norm": 3.0068348938408427,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 6918
+    },
+    {
+      "epoch": 0.16469104065505094,
+      "grad_norm": 4.339819148389635,
+      "learning_rate": 2e-05,
+      "loss": 0.6659,
+      "step": 6919
+    },
+    {
+      "epoch": 0.16471484337808245,
+      "grad_norm": 9.874093744439747,
+      "learning_rate": 2e-05,
+      "loss": 0.5141,
+      "step": 6920
+    },
+    {
+      "epoch": 0.16473864610111397,
+      "grad_norm": 16.183379145509083,
+      "learning_rate": 2e-05,
+      "loss": 1.0854,
+      "step": 6921
+    },
+    {
+      "epoch": 0.16476244882414548,
+      "grad_norm": 3.58430601062438,
+      "learning_rate": 2e-05,
+      "loss": 0.3801,
+      "step": 6922
+    },
+    {
+      "epoch": 0.164786251547177,
+      "grad_norm": 8.531863836044115,
+      "learning_rate": 2e-05,
+      "loss": 0.2919,
+      "step": 6923
+    },
+    {
+      "epoch": 0.1648100542702085,
+      "grad_norm": 6.648779165459004,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 6924
+    },
+    {
+      "epoch": 0.16483385699324002,
+      "grad_norm": 8.694170097992282,
+      "learning_rate": 2e-05,
+      "loss": 0.5953,
+      "step": 6925
+    },
+    {
+      "epoch": 0.16485765971627153,
+      "grad_norm": 2.9458177006061,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 6926
+    },
+    {
+      "epoch": 0.16488146243930304,
+      "grad_norm": 2.004817080555612,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 6927
+    },
+    {
+      "epoch": 0.16490526516233456,
+      "grad_norm": 7.548199264940775,
+      "learning_rate": 2e-05,
+      "loss": 0.6516,
+      "step": 6928
+    },
+    {
+      "epoch": 0.1649290678853661,
+      "grad_norm": 6.304968701433249,
+      "learning_rate": 2e-05,
+      "loss": 1.4099,
+      "step": 6929
+    },
+    {
+      "epoch": 0.1649528706083976,
+      "grad_norm": 8.904318703018099,
+      "learning_rate": 2e-05,
+      "loss": 0.3561,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16497667333142912,
+      "grad_norm": 6.540416606210749,
+      "learning_rate": 2e-05,
+      "loss": 0.3695,
+      "step": 6931
+    },
+    {
+      "epoch": 0.16500047605446064,
+      "grad_norm": 9.895975149696385,
+      "learning_rate": 2e-05,
+      "loss": 0.9981,
+      "step": 6932
+    },
+    {
+      "epoch": 0.16502427877749215,
+      "grad_norm": 10.531050440447927,
+      "learning_rate": 2e-05,
+      "loss": 0.7977,
+      "step": 6933
+    },
+    {
+      "epoch": 0.16504808150052366,
+      "grad_norm": 7.094024354922371,
+      "learning_rate": 2e-05,
+      "loss": 0.8581,
+      "step": 6934
+    },
+    {
+      "epoch": 0.16507188422355518,
+      "grad_norm": 6.571231273116495,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 6935
+    },
+    {
+      "epoch": 0.1650956869465867,
+      "grad_norm": 8.693300038349,
+      "learning_rate": 2e-05,
+      "loss": 1.1079,
+      "step": 6936
+    },
+    {
+      "epoch": 0.1651194896696182,
+      "grad_norm": 6.8160302555893955,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 6937
+    },
+    {
+      "epoch": 0.16514329239264972,
+      "grad_norm": 6.406715185774573,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 6938
+    },
+    {
+      "epoch": 0.16516709511568123,
+      "grad_norm": 7.04662425619122,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 6939
+    },
+    {
+      "epoch": 0.16519089783871274,
+      "grad_norm": 4.494934471698903,
+      "learning_rate": 2e-05,
+      "loss": 0.5008,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16521470056174425,
+      "grad_norm": 1.8530786684376124,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 6941
+    },
+    {
+      "epoch": 0.16523850328477577,
+      "grad_norm": 5.522363485026232,
+      "learning_rate": 2e-05,
+      "loss": 0.31,
+      "step": 6942
+    },
+    {
+      "epoch": 0.16526230600780728,
+      "grad_norm": 5.825209275002019,
+      "learning_rate": 2e-05,
+      "loss": 0.247,
+      "step": 6943
+    },
+    {
+      "epoch": 0.16528610873083882,
+      "grad_norm": 4.496819832479326,
+      "learning_rate": 2e-05,
+      "loss": 0.3098,
+      "step": 6944
+    },
+    {
+      "epoch": 0.16530991145387033,
+      "grad_norm": 10.21335367984477,
+      "learning_rate": 2e-05,
+      "loss": 0.9879,
+      "step": 6945
+    },
+    {
+      "epoch": 0.16533371417690185,
+      "grad_norm": 5.11266847559805,
+      "learning_rate": 2e-05,
+      "loss": 0.3104,
+      "step": 6946
+    },
+    {
+      "epoch": 0.16535751689993336,
+      "grad_norm": 2.2648325167695242,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 6947
+    },
+    {
+      "epoch": 0.16538131962296487,
+      "grad_norm": 4.370088303229116,
+      "learning_rate": 2e-05,
+      "loss": 0.6465,
+      "step": 6948
+    },
+    {
+      "epoch": 0.16540512234599639,
+      "grad_norm": 4.421209460556701,
+      "learning_rate": 2e-05,
+      "loss": 0.9195,
+      "step": 6949
+    },
+    {
+      "epoch": 0.1654289250690279,
+      "grad_norm": 3.4802253805729375,
+      "learning_rate": 2e-05,
+      "loss": 0.4031,
+      "step": 6950
+    },
+    {
+      "epoch": 0.1654527277920594,
+      "grad_norm": 1.8477232418118397,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 6951
+    },
+    {
+      "epoch": 0.16547653051509092,
+      "grad_norm": 12.302380756384586,
+      "learning_rate": 2e-05,
+      "loss": 1.6243,
+      "step": 6952
+    },
+    {
+      "epoch": 0.16550033323812244,
+      "grad_norm": 6.024840835911667,
+      "learning_rate": 2e-05,
+      "loss": 1.1128,
+      "step": 6953
+    },
+    {
+      "epoch": 0.16552413596115395,
+      "grad_norm": 7.517514882772353,
+      "learning_rate": 2e-05,
+      "loss": 1.1413,
+      "step": 6954
+    },
+    {
+      "epoch": 0.16554793868418546,
+      "grad_norm": 9.101661822352051,
+      "learning_rate": 2e-05,
+      "loss": 1.0838,
+      "step": 6955
+    },
+    {
+      "epoch": 0.16557174140721698,
+      "grad_norm": 1.5770872228125237,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 6956
+    },
+    {
+      "epoch": 0.1655955441302485,
+      "grad_norm": 1.691946612059975,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 6957
+    },
+    {
+      "epoch": 0.16561934685328,
+      "grad_norm": 10.753681522435482,
+      "learning_rate": 2e-05,
+      "loss": 0.5832,
+      "step": 6958
+    },
+    {
+      "epoch": 0.16564314957631154,
+      "grad_norm": 1.5053088526058906,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 6959
+    },
+    {
+      "epoch": 0.16566695229934306,
+      "grad_norm": 3.159096650148176,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16569075502237457,
+      "grad_norm": 7.227092279582326,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 6961
+    },
+    {
+      "epoch": 0.16571455774540608,
+      "grad_norm": 9.532722632147852,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 6962
+    },
+    {
+      "epoch": 0.1657383604684376,
+      "grad_norm": 6.037431170808814,
+      "learning_rate": 2e-05,
+      "loss": 1.2968,
+      "step": 6963
+    },
+    {
+      "epoch": 0.1657621631914691,
+      "grad_norm": 3.6140988329714556,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 6964
+    },
+    {
+      "epoch": 0.16578596591450062,
+      "grad_norm": 5.550254790450601,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 6965
+    },
+    {
+      "epoch": 0.16580976863753213,
+      "grad_norm": 8.659551781522826,
+      "learning_rate": 2e-05,
+      "loss": 1.5724,
+      "step": 6966
+    },
+    {
+      "epoch": 0.16583357136056365,
+      "grad_norm": 7.115864239552887,
+      "learning_rate": 2e-05,
+      "loss": 0.4422,
+      "step": 6967
+    },
+    {
+      "epoch": 0.16585737408359516,
+      "grad_norm": 5.530389226418741,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 6968
+    },
+    {
+      "epoch": 0.16588117680662667,
+      "grad_norm": 5.260278254594664,
+      "learning_rate": 2e-05,
+      "loss": 0.8632,
+      "step": 6969
+    },
+    {
+      "epoch": 0.1659049795296582,
+      "grad_norm": 5.733621640418696,
+      "learning_rate": 2e-05,
+      "loss": 0.6855,
+      "step": 6970
+    },
+    {
+      "epoch": 0.1659287822526897,
+      "grad_norm": 9.373972889251553,
+      "learning_rate": 2e-05,
+      "loss": 0.8891,
+      "step": 6971
+    },
+    {
+      "epoch": 0.1659525849757212,
+      "grad_norm": 5.688201393958465,
+      "learning_rate": 2e-05,
+      "loss": 0.7396,
+      "step": 6972
+    },
+    {
+      "epoch": 0.16597638769875273,
+      "grad_norm": 2.299793494600313,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 6973
+    },
+    {
+      "epoch": 0.16600019042178424,
+      "grad_norm": 4.604745824866679,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 6974
+    },
+    {
+      "epoch": 0.16602399314481578,
+      "grad_norm": 9.113091155686611,
+      "learning_rate": 2e-05,
+      "loss": 1.0251,
+      "step": 6975
+    },
+    {
+      "epoch": 0.1660477958678473,
+      "grad_norm": 3.4599870535749586,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 6976
+    },
+    {
+      "epoch": 0.1660715985908788,
+      "grad_norm": 1.2873144511731474,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 6977
+    },
+    {
+      "epoch": 0.16609540131391032,
+      "grad_norm": 6.647559167435307,
+      "learning_rate": 2e-05,
+      "loss": 1.4325,
+      "step": 6978
+    },
+    {
+      "epoch": 0.16611920403694183,
+      "grad_norm": 9.113306782102974,
+      "learning_rate": 2e-05,
+      "loss": 0.4507,
+      "step": 6979
+    },
+    {
+      "epoch": 0.16614300675997334,
+      "grad_norm": 6.998990476772732,
+      "learning_rate": 2e-05,
+      "loss": 0.3079,
+      "step": 6980
+    },
+    {
+      "epoch": 0.16616680948300486,
+      "grad_norm": 1.9129034452825842,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 6981
+    },
+    {
+      "epoch": 0.16619061220603637,
+      "grad_norm": 2.079385003000419,
+      "learning_rate": 2e-05,
+      "loss": 0.1657,
+      "step": 6982
+    },
+    {
+      "epoch": 0.16621441492906788,
+      "grad_norm": 7.8166238311896015,
+      "learning_rate": 2e-05,
+      "loss": 0.3396,
+      "step": 6983
+    },
+    {
+      "epoch": 0.1662382176520994,
+      "grad_norm": 1.9602136353840207,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 6984
+    },
+    {
+      "epoch": 0.1662620203751309,
+      "grad_norm": 2.3206559778787277,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 6985
+    },
+    {
+      "epoch": 0.16628582309816242,
+      "grad_norm": 8.48051410626718,
+      "learning_rate": 2e-05,
+      "loss": 1.2344,
+      "step": 6986
+    },
+    {
+      "epoch": 0.16630962582119393,
+      "grad_norm": 4.384686613544552,
+      "learning_rate": 2e-05,
+      "loss": 0.2551,
+      "step": 6987
+    },
+    {
+      "epoch": 0.16633342854422545,
+      "grad_norm": 2.7491160331498223,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 6988
+    },
+    {
+      "epoch": 0.16635723126725696,
+      "grad_norm": 6.194710773231195,
+      "learning_rate": 2e-05,
+      "loss": 1.2494,
+      "step": 6989
+    },
+    {
+      "epoch": 0.1663810339902885,
+      "grad_norm": 3.8354029216020256,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16640483671332001,
+      "grad_norm": 4.871111767064889,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 6991
+    },
+    {
+      "epoch": 0.16642863943635153,
+      "grad_norm": 4.512631843408233,
+      "learning_rate": 2e-05,
+      "loss": 0.6397,
+      "step": 6992
+    },
+    {
+      "epoch": 0.16645244215938304,
+      "grad_norm": 8.978553130819531,
+      "learning_rate": 2e-05,
+      "loss": 0.3628,
+      "step": 6993
+    },
+    {
+      "epoch": 0.16647624488241455,
+      "grad_norm": 9.514605855437473,
+      "learning_rate": 2e-05,
+      "loss": 1.6979,
+      "step": 6994
+    },
+    {
+      "epoch": 0.16650004760544607,
+      "grad_norm": 7.186290253548502,
+      "learning_rate": 2e-05,
+      "loss": 0.291,
+      "step": 6995
+    },
+    {
+      "epoch": 0.16652385032847758,
+      "grad_norm": 1.9391410057317902,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 6996
+    },
+    {
+      "epoch": 0.1665476530515091,
+      "grad_norm": 4.821472659801,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 6997
+    },
+    {
+      "epoch": 0.1665714557745406,
+      "grad_norm": 4.260885564215818,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 6998
+    },
+    {
+      "epoch": 0.16659525849757212,
+      "grad_norm": 3.5687880425356533,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 6999
+    },
+    {
+      "epoch": 0.16661906122060363,
+      "grad_norm": 7.4811920909218,
+      "learning_rate": 2e-05,
+      "loss": 1.0105,
+      "step": 7000
+    },
+    {
+      "epoch": 0.16664286394363514,
+      "grad_norm": 3.582425306164652,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 7001
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 3.1943305310181955,
+      "learning_rate": 2e-05,
+      "loss": 0.1894,
+      "step": 7002
+    },
+    {
+      "epoch": 0.16669046938969817,
+      "grad_norm": 2.161737136679958,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 7003
+    },
+    {
+      "epoch": 0.16671427211272968,
+      "grad_norm": 6.670693164213125,
+      "learning_rate": 2e-05,
+      "loss": 0.429,
+      "step": 7004
+    },
+    {
+      "epoch": 0.16673807483576122,
+      "grad_norm": 2.613062747699946,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 7005
+    },
+    {
+      "epoch": 0.16676187755879274,
+      "grad_norm": 4.759917808249213,
+      "learning_rate": 2e-05,
+      "loss": 0.4928,
+      "step": 7006
+    },
+    {
+      "epoch": 0.16678568028182425,
+      "grad_norm": 4.771637840041966,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 7007
+    },
+    {
+      "epoch": 0.16680948300485576,
+      "grad_norm": 1.6020463817823136,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 7008
+    },
+    {
+      "epoch": 0.16683328572788728,
+      "grad_norm": 4.586094812619131,
+      "learning_rate": 2e-05,
+      "loss": 0.7227,
+      "step": 7009
+    },
+    {
+      "epoch": 0.1668570884509188,
+      "grad_norm": 8.11518265414507,
+      "learning_rate": 2e-05,
+      "loss": 0.7511,
+      "step": 7010
+    },
+    {
+      "epoch": 0.1668808911739503,
+      "grad_norm": 3.765685782845035,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 7011
+    },
+    {
+      "epoch": 0.16690469389698182,
+      "grad_norm": 5.412368616323358,
+      "learning_rate": 2e-05,
+      "loss": 0.505,
+      "step": 7012
+    },
+    {
+      "epoch": 0.16692849662001333,
+      "grad_norm": 4.836418002858445,
+      "learning_rate": 2e-05,
+      "loss": 0.8126,
+      "step": 7013
+    },
+    {
+      "epoch": 0.16695229934304484,
+      "grad_norm": 14.508046895427482,
+      "learning_rate": 2e-05,
+      "loss": 0.4213,
+      "step": 7014
+    },
+    {
+      "epoch": 0.16697610206607635,
+      "grad_norm": 5.82690912992476,
+      "learning_rate": 2e-05,
+      "loss": 0.8537,
+      "step": 7015
+    },
+    {
+      "epoch": 0.16699990478910787,
+      "grad_norm": 3.0401067154168087,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 7016
+    },
+    {
+      "epoch": 0.16702370751213938,
+      "grad_norm": 4.00543924872878,
+      "learning_rate": 2e-05,
+      "loss": 0.2798,
+      "step": 7017
+    },
+    {
+      "epoch": 0.1670475102351709,
+      "grad_norm": 1.2989138071914201,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 7018
+    },
+    {
+      "epoch": 0.1670713129582024,
+      "grad_norm": 3.151477745384327,
+      "learning_rate": 2e-05,
+      "loss": 0.4709,
+      "step": 7019
+    },
+    {
+      "epoch": 0.16709511568123395,
+      "grad_norm": 4.759330784813684,
+      "learning_rate": 2e-05,
+      "loss": 0.5687,
+      "step": 7020
+    },
+    {
+      "epoch": 0.16711891840426546,
+      "grad_norm": 2.1904328348418067,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 7021
+    },
+    {
+      "epoch": 0.16714272112729697,
+      "grad_norm": 2.6229604175927768,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 7022
+    },
+    {
+      "epoch": 0.16716652385032849,
+      "grad_norm": 12.013806702057575,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 7023
+    },
+    {
+      "epoch": 0.16719032657336,
+      "grad_norm": 0.8108053534554999,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 7024
+    },
+    {
+      "epoch": 0.1672141292963915,
+      "grad_norm": 4.185388992400208,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 7025
+    },
+    {
+      "epoch": 0.16723793201942302,
+      "grad_norm": 4.4268239509886245,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 7026
+    },
+    {
+      "epoch": 0.16726173474245454,
+      "grad_norm": 5.3738153334846075,
+      "learning_rate": 2e-05,
+      "loss": 0.7496,
+      "step": 7027
+    },
+    {
+      "epoch": 0.16728553746548605,
+      "grad_norm": 1.8726929890316408,
+      "learning_rate": 2e-05,
+      "loss": 0.1266,
+      "step": 7028
+    },
+    {
+      "epoch": 0.16730934018851756,
+      "grad_norm": 6.200294960864807,
+      "learning_rate": 2e-05,
+      "loss": 0.7565,
+      "step": 7029
+    },
+    {
+      "epoch": 0.16733314291154908,
+      "grad_norm": 10.962688830322376,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1673569456345806,
+      "grad_norm": 2.768445659509598,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 7031
+    },
+    {
+      "epoch": 0.1673807483576121,
+      "grad_norm": 5.925218210487179,
+      "learning_rate": 2e-05,
+      "loss": 0.7315,
+      "step": 7032
+    },
+    {
+      "epoch": 0.16740455108064362,
+      "grad_norm": 1.6103734501805407,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 7033
+    },
+    {
+      "epoch": 0.16742835380367513,
+      "grad_norm": 6.696307799744989,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 7034
+    },
+    {
+      "epoch": 0.16745215652670664,
+      "grad_norm": 3.0635936819886367,
+      "learning_rate": 2e-05,
+      "loss": 0.2822,
+      "step": 7035
+    },
+    {
+      "epoch": 0.16747595924973818,
+      "grad_norm": 1.9029127986453351,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 7036
+    },
+    {
+      "epoch": 0.1674997619727697,
+      "grad_norm": 7.755541957163338,
+      "learning_rate": 2e-05,
+      "loss": 0.4605,
+      "step": 7037
+    },
+    {
+      "epoch": 0.1675235646958012,
+      "grad_norm": 5.381823178982223,
+      "learning_rate": 2e-05,
+      "loss": 0.671,
+      "step": 7038
+    },
+    {
+      "epoch": 0.16754736741883272,
+      "grad_norm": 7.406431283170387,
+      "learning_rate": 2e-05,
+      "loss": 0.5604,
+      "step": 7039
+    },
+    {
+      "epoch": 0.16757117014186423,
+      "grad_norm": 6.1556114551176755,
+      "learning_rate": 2e-05,
+      "loss": 0.7659,
+      "step": 7040
+    },
+    {
+      "epoch": 0.16759497286489575,
+      "grad_norm": 5.353364527682694,
+      "learning_rate": 2e-05,
+      "loss": 0.4542,
+      "step": 7041
+    },
+    {
+      "epoch": 0.16761877558792726,
+      "grad_norm": 4.628511469233716,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 7042
+    },
+    {
+      "epoch": 0.16764257831095877,
+      "grad_norm": 7.504125581025406,
+      "learning_rate": 2e-05,
+      "loss": 0.2591,
+      "step": 7043
+    },
+    {
+      "epoch": 0.16766638103399029,
+      "grad_norm": 7.625278049422768,
+      "learning_rate": 2e-05,
+      "loss": 0.6622,
+      "step": 7044
+    },
+    {
+      "epoch": 0.1676901837570218,
+      "grad_norm": 5.707369138401389,
+      "learning_rate": 2e-05,
+      "loss": 0.9599,
+      "step": 7045
+    },
+    {
+      "epoch": 0.1677139864800533,
+      "grad_norm": 2.0539095410931827,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 7046
+    },
+    {
+      "epoch": 0.16773778920308482,
+      "grad_norm": 5.048058373769643,
+      "learning_rate": 2e-05,
+      "loss": 0.6158,
+      "step": 7047
+    },
+    {
+      "epoch": 0.16776159192611634,
+      "grad_norm": 5.378472338496133,
+      "learning_rate": 2e-05,
+      "loss": 1.1516,
+      "step": 7048
+    },
+    {
+      "epoch": 0.16778539464914785,
+      "grad_norm": 6.514324903802903,
+      "learning_rate": 2e-05,
+      "loss": 1.4252,
+      "step": 7049
+    },
+    {
+      "epoch": 0.16780919737217936,
+      "grad_norm": 5.055219350005336,
+      "learning_rate": 2e-05,
+      "loss": 0.8307,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1678330000952109,
+      "grad_norm": 2.414900752918547,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 7051
+    },
+    {
+      "epoch": 0.16785680281824242,
+      "grad_norm": 8.752610637921174,
+      "learning_rate": 2e-05,
+      "loss": 0.7949,
+      "step": 7052
+    },
+    {
+      "epoch": 0.16788060554127393,
+      "grad_norm": 30.202645802384605,
+      "learning_rate": 2e-05,
+      "loss": 0.8593,
+      "step": 7053
+    },
+    {
+      "epoch": 0.16790440826430544,
+      "grad_norm": 4.7807246277963635,
+      "learning_rate": 2e-05,
+      "loss": 0.5867,
+      "step": 7054
+    },
+    {
+      "epoch": 0.16792821098733696,
+      "grad_norm": 2.064858025703835,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 7055
+    },
+    {
+      "epoch": 0.16795201371036847,
+      "grad_norm": 4.637421616352916,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 7056
+    },
+    {
+      "epoch": 0.16797581643339998,
+      "grad_norm": 6.766710719065444,
+      "learning_rate": 2e-05,
+      "loss": 1.0935,
+      "step": 7057
+    },
+    {
+      "epoch": 0.1679996191564315,
+      "grad_norm": 3.4928690156834707,
+      "learning_rate": 2e-05,
+      "loss": 0.5712,
+      "step": 7058
+    },
+    {
+      "epoch": 0.168023421879463,
+      "grad_norm": 3.3220850960471084,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 7059
+    },
+    {
+      "epoch": 0.16804722460249452,
+      "grad_norm": 9.09966658010135,
+      "learning_rate": 2e-05,
+      "loss": 0.7955,
+      "step": 7060
+    },
+    {
+      "epoch": 0.16807102732552603,
+      "grad_norm": 7.2361939806718505,
+      "learning_rate": 2e-05,
+      "loss": 0.2828,
+      "step": 7061
+    },
+    {
+      "epoch": 0.16809483004855755,
+      "grad_norm": 7.062747046179967,
+      "learning_rate": 2e-05,
+      "loss": 0.8728,
+      "step": 7062
+    },
+    {
+      "epoch": 0.16811863277158906,
+      "grad_norm": 2.3866410284256387,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 7063
+    },
+    {
+      "epoch": 0.16814243549462057,
+      "grad_norm": 5.091151686762353,
+      "learning_rate": 2e-05,
+      "loss": 0.1613,
+      "step": 7064
+    },
+    {
+      "epoch": 0.1681662382176521,
+      "grad_norm": 7.3308205569314,
+      "learning_rate": 2e-05,
+      "loss": 0.3679,
+      "step": 7065
+    },
+    {
+      "epoch": 0.16819004094068363,
+      "grad_norm": 5.851400451495944,
+      "learning_rate": 2e-05,
+      "loss": 1.2012,
+      "step": 7066
+    },
+    {
+      "epoch": 0.16821384366371514,
+      "grad_norm": 10.795527609753112,
+      "learning_rate": 2e-05,
+      "loss": 0.3256,
+      "step": 7067
+    },
+    {
+      "epoch": 0.16823764638674665,
+      "grad_norm": 3.030322989019049,
+      "learning_rate": 2e-05,
+      "loss": 0.2825,
+      "step": 7068
+    },
+    {
+      "epoch": 0.16826144910977817,
+      "grad_norm": 6.710157507375612,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 7069
+    },
+    {
+      "epoch": 0.16828525183280968,
+      "grad_norm": 5.895025239177133,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 7070
+    },
+    {
+      "epoch": 0.1683090545558412,
+      "grad_norm": 7.852744348903419,
+      "learning_rate": 2e-05,
+      "loss": 0.2565,
+      "step": 7071
+    },
+    {
+      "epoch": 0.1683328572788727,
+      "grad_norm": 6.8046872633835775,
+      "learning_rate": 2e-05,
+      "loss": 0.3106,
+      "step": 7072
+    },
+    {
+      "epoch": 0.16835666000190422,
+      "grad_norm": 6.45917334292686,
+      "learning_rate": 2e-05,
+      "loss": 0.2353,
+      "step": 7073
+    },
+    {
+      "epoch": 0.16838046272493573,
+      "grad_norm": 2.161121748892617,
+      "learning_rate": 2e-05,
+      "loss": 0.2004,
+      "step": 7074
+    },
+    {
+      "epoch": 0.16840426544796724,
+      "grad_norm": 4.956069357557766,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 7075
+    },
+    {
+      "epoch": 0.16842806817099876,
+      "grad_norm": 6.339517829788518,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 7076
+    },
+    {
+      "epoch": 0.16845187089403027,
+      "grad_norm": 5.809796502240479,
+      "learning_rate": 2e-05,
+      "loss": 0.5944,
+      "step": 7077
+    },
+    {
+      "epoch": 0.16847567361706178,
+      "grad_norm": 5.637918077932501,
+      "learning_rate": 2e-05,
+      "loss": 0.6458,
+      "step": 7078
+    },
+    {
+      "epoch": 0.1684994763400933,
+      "grad_norm": 2.353841004772552,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 7079
+    },
+    {
+      "epoch": 0.1685232790631248,
+      "grad_norm": 8.464730731724988,
+      "learning_rate": 2e-05,
+      "loss": 0.9388,
+      "step": 7080
+    },
+    {
+      "epoch": 0.16854708178615635,
+      "grad_norm": 3.989575883394919,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 7081
+    },
+    {
+      "epoch": 0.16857088450918786,
+      "grad_norm": 6.516324758655056,
+      "learning_rate": 2e-05,
+      "loss": 0.9815,
+      "step": 7082
+    },
+    {
+      "epoch": 0.16859468723221938,
+      "grad_norm": 2.9254357869931455,
+      "learning_rate": 2e-05,
+      "loss": 0.145,
+      "step": 7083
+    },
+    {
+      "epoch": 0.1686184899552509,
+      "grad_norm": 3.3284393334128546,
+      "learning_rate": 2e-05,
+      "loss": 0.2031,
+      "step": 7084
+    },
+    {
+      "epoch": 0.1686422926782824,
+      "grad_norm": 4.002397676690772,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 7085
+    },
+    {
+      "epoch": 0.16866609540131391,
+      "grad_norm": 0.9263746491702587,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 7086
+    },
+    {
+      "epoch": 0.16868989812434543,
+      "grad_norm": 1.387094820638533,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 7087
+    },
+    {
+      "epoch": 0.16871370084737694,
+      "grad_norm": 4.959559555511714,
+      "learning_rate": 2e-05,
+      "loss": 0.6674,
+      "step": 7088
+    },
+    {
+      "epoch": 0.16873750357040845,
+      "grad_norm": 4.074353415931376,
+      "learning_rate": 2e-05,
+      "loss": 0.752,
+      "step": 7089
+    },
+    {
+      "epoch": 0.16876130629343997,
+      "grad_norm": 7.058078570808565,
+      "learning_rate": 2e-05,
+      "loss": 0.5696,
+      "step": 7090
+    },
+    {
+      "epoch": 0.16878510901647148,
+      "grad_norm": 3.4990312936613455,
+      "learning_rate": 2e-05,
+      "loss": 0.1821,
+      "step": 7091
+    },
+    {
+      "epoch": 0.168808911739503,
+      "grad_norm": 5.728824601209861,
+      "learning_rate": 2e-05,
+      "loss": 0.9726,
+      "step": 7092
+    },
+    {
+      "epoch": 0.1688327144625345,
+      "grad_norm": 10.730834141331002,
+      "learning_rate": 2e-05,
+      "loss": 0.6365,
+      "step": 7093
+    },
+    {
+      "epoch": 0.16885651718556602,
+      "grad_norm": 8.523742655193825,
+      "learning_rate": 2e-05,
+      "loss": 0.9369,
+      "step": 7094
+    },
+    {
+      "epoch": 0.16888031990859753,
+      "grad_norm": 1.9389517549532491,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 7095
+    },
+    {
+      "epoch": 0.16890412263162904,
+      "grad_norm": 7.13016259660473,
+      "learning_rate": 2e-05,
+      "loss": 0.5361,
+      "step": 7096
+    },
+    {
+      "epoch": 0.16892792535466059,
+      "grad_norm": 9.386365705337617,
+      "learning_rate": 2e-05,
+      "loss": 1.1374,
+      "step": 7097
+    },
+    {
+      "epoch": 0.1689517280776921,
+      "grad_norm": 5.7990598200510375,
+      "learning_rate": 2e-05,
+      "loss": 0.7419,
+      "step": 7098
+    },
+    {
+      "epoch": 0.1689755308007236,
+      "grad_norm": 3.6459224228594986,
+      "learning_rate": 2e-05,
+      "loss": 0.584,
+      "step": 7099
+    },
+    {
+      "epoch": 0.16899933352375512,
+      "grad_norm": 7.6209369662991815,
+      "learning_rate": 2e-05,
+      "loss": 0.6565,
+      "step": 7100
+    },
+    {
+      "epoch": 0.16902313624678664,
+      "grad_norm": 4.88890911446824,
+      "learning_rate": 2e-05,
+      "loss": 0.5628,
+      "step": 7101
+    },
+    {
+      "epoch": 0.16904693896981815,
+      "grad_norm": 2.6855149872539488,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 7102
+    },
+    {
+      "epoch": 0.16907074169284966,
+      "grad_norm": 9.235397293791838,
+      "learning_rate": 2e-05,
+      "loss": 0.8541,
+      "step": 7103
+    },
+    {
+      "epoch": 0.16909454441588118,
+      "grad_norm": 4.819381544834709,
+      "learning_rate": 2e-05,
+      "loss": 0.6355,
+      "step": 7104
+    },
+    {
+      "epoch": 0.1691183471389127,
+      "grad_norm": 30.273163176541704,
+      "learning_rate": 2e-05,
+      "loss": 0.8863,
+      "step": 7105
+    },
+    {
+      "epoch": 0.1691421498619442,
+      "grad_norm": 7.415679999817682,
+      "learning_rate": 2e-05,
+      "loss": 0.9015,
+      "step": 7106
+    },
+    {
+      "epoch": 0.16916595258497572,
+      "grad_norm": 5.955673602434953,
+      "learning_rate": 2e-05,
+      "loss": 1.3701,
+      "step": 7107
+    },
+    {
+      "epoch": 0.16918975530800723,
+      "grad_norm": 3.391797296772335,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 7108
+    },
+    {
+      "epoch": 0.16921355803103874,
+      "grad_norm": 5.5977879119565115,
+      "learning_rate": 2e-05,
+      "loss": 0.7855,
+      "step": 7109
+    },
+    {
+      "epoch": 0.16923736075407025,
+      "grad_norm": 5.551112558199651,
+      "learning_rate": 2e-05,
+      "loss": 0.2571,
+      "step": 7110
+    },
+    {
+      "epoch": 0.16926116347710177,
+      "grad_norm": 4.561584318814486,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 7111
+    },
+    {
+      "epoch": 0.1692849662001333,
+      "grad_norm": 3.075728170787214,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 7112
+    },
+    {
+      "epoch": 0.16930876892316482,
+      "grad_norm": 6.320407591921564,
+      "learning_rate": 2e-05,
+      "loss": 0.8799,
+      "step": 7113
+    },
+    {
+      "epoch": 0.16933257164619633,
+      "grad_norm": 6.130280223538212,
+      "learning_rate": 2e-05,
+      "loss": 0.7017,
+      "step": 7114
+    },
+    {
+      "epoch": 0.16935637436922785,
+      "grad_norm": 2.0156940341331713,
+      "learning_rate": 2e-05,
+      "loss": 0.1007,
+      "step": 7115
+    },
+    {
+      "epoch": 0.16938017709225936,
+      "grad_norm": 7.100086537311585,
+      "learning_rate": 2e-05,
+      "loss": 1.2157,
+      "step": 7116
+    },
+    {
+      "epoch": 0.16940397981529087,
+      "grad_norm": 2.8786183868452437,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 7117
+    },
+    {
+      "epoch": 0.16942778253832239,
+      "grad_norm": 4.280939123915036,
+      "learning_rate": 2e-05,
+      "loss": 0.5567,
+      "step": 7118
+    },
+    {
+      "epoch": 0.1694515852613539,
+      "grad_norm": 2.79176051426714,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 7119
+    },
+    {
+      "epoch": 0.1694753879843854,
+      "grad_norm": 2.8074335296002153,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 7120
+    },
+    {
+      "epoch": 0.16949919070741692,
+      "grad_norm": 7.387257346807261,
+      "learning_rate": 2e-05,
+      "loss": 1.3467,
+      "step": 7121
+    },
+    {
+      "epoch": 0.16952299343044844,
+      "grad_norm": 6.484814287045924,
+      "learning_rate": 2e-05,
+      "loss": 0.9202,
+      "step": 7122
+    },
+    {
+      "epoch": 0.16954679615347995,
+      "grad_norm": 9.09463014004676,
+      "learning_rate": 2e-05,
+      "loss": 0.7842,
+      "step": 7123
+    },
+    {
+      "epoch": 0.16957059887651146,
+      "grad_norm": 3.1677850939043934,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 7124
+    },
+    {
+      "epoch": 0.16959440159954298,
+      "grad_norm": 4.6583654113602275,
+      "learning_rate": 2e-05,
+      "loss": 0.617,
+      "step": 7125
+    },
+    {
+      "epoch": 0.1696182043225745,
+      "grad_norm": 4.502967715225665,
+      "learning_rate": 2e-05,
+      "loss": 0.9294,
+      "step": 7126
+    },
+    {
+      "epoch": 0.16964200704560603,
+      "grad_norm": 2.77639049993808,
+      "learning_rate": 2e-05,
+      "loss": 0.2099,
+      "step": 7127
+    },
+    {
+      "epoch": 0.16966580976863754,
+      "grad_norm": 4.393843880521815,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 7128
+    },
+    {
+      "epoch": 0.16968961249166906,
+      "grad_norm": 4.100005597621498,
+      "learning_rate": 2e-05,
+      "loss": 0.2144,
+      "step": 7129
+    },
+    {
+      "epoch": 0.16971341521470057,
+      "grad_norm": 10.399250641404386,
+      "learning_rate": 2e-05,
+      "loss": 0.5802,
+      "step": 7130
+    },
+    {
+      "epoch": 0.16973721793773208,
+      "grad_norm": 5.155545507159411,
+      "learning_rate": 2e-05,
+      "loss": 0.8422,
+      "step": 7131
+    },
+    {
+      "epoch": 0.1697610206607636,
+      "grad_norm": 6.79091418323161,
+      "learning_rate": 2e-05,
+      "loss": 0.2727,
+      "step": 7132
+    },
+    {
+      "epoch": 0.1697848233837951,
+      "grad_norm": 3.7073996249498107,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 7133
+    },
+    {
+      "epoch": 0.16980862610682662,
+      "grad_norm": 3.1902917917604423,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 7134
+    },
+    {
+      "epoch": 0.16983242882985813,
+      "grad_norm": 9.424672105566044,
+      "learning_rate": 2e-05,
+      "loss": 1.1464,
+      "step": 7135
+    },
+    {
+      "epoch": 0.16985623155288965,
+      "grad_norm": 1.4177023065048637,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 7136
+    },
+    {
+      "epoch": 0.16988003427592116,
+      "grad_norm": 4.151847076350571,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 7137
+    },
+    {
+      "epoch": 0.16990383699895267,
+      "grad_norm": 4.604106816117896,
+      "learning_rate": 2e-05,
+      "loss": 0.6062,
+      "step": 7138
+    },
+    {
+      "epoch": 0.16992763972198419,
+      "grad_norm": 5.1135817866792195,
+      "learning_rate": 2e-05,
+      "loss": 0.7261,
+      "step": 7139
+    },
+    {
+      "epoch": 0.1699514424450157,
+      "grad_norm": 7.946638403791122,
+      "learning_rate": 2e-05,
+      "loss": 0.9072,
+      "step": 7140
+    },
+    {
+      "epoch": 0.1699752451680472,
+      "grad_norm": 4.968229626433895,
+      "learning_rate": 2e-05,
+      "loss": 0.3297,
+      "step": 7141
+    },
+    {
+      "epoch": 0.16999904789107875,
+      "grad_norm": 4.89021595511142,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 7142
+    },
+    {
+      "epoch": 0.17002285061411027,
+      "grad_norm": 5.028873090186072,
+      "learning_rate": 2e-05,
+      "loss": 0.1701,
+      "step": 7143
+    },
+    {
+      "epoch": 0.17004665333714178,
+      "grad_norm": 4.866284569644305,
+      "learning_rate": 2e-05,
+      "loss": 0.4981,
+      "step": 7144
+    },
+    {
+      "epoch": 0.1700704560601733,
+      "grad_norm": 1.129384685123748,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 7145
+    },
+    {
+      "epoch": 0.1700942587832048,
+      "grad_norm": 4.321837526107936,
+      "learning_rate": 2e-05,
+      "loss": 0.2698,
+      "step": 7146
+    },
+    {
+      "epoch": 0.17011806150623632,
+      "grad_norm": 3.071570843394183,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 7147
+    },
+    {
+      "epoch": 0.17014186422926783,
+      "grad_norm": 4.744875112064853,
+      "learning_rate": 2e-05,
+      "loss": 0.2289,
+      "step": 7148
+    },
+    {
+      "epoch": 0.17016566695229934,
+      "grad_norm": 4.079524988277126,
+      "learning_rate": 2e-05,
+      "loss": 0.3986,
+      "step": 7149
+    },
+    {
+      "epoch": 0.17018946967533086,
+      "grad_norm": 5.049360624418625,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 7150
+    },
+    {
+      "epoch": 0.17021327239836237,
+      "grad_norm": 3.006711145694638,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 7151
+    },
+    {
+      "epoch": 0.17023707512139388,
+      "grad_norm": 2.529478459171379,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 7152
+    },
+    {
+      "epoch": 0.1702608778444254,
+      "grad_norm": 10.601063844757233,
+      "learning_rate": 2e-05,
+      "loss": 0.5419,
+      "step": 7153
+    },
+    {
+      "epoch": 0.1702846805674569,
+      "grad_norm": 2.994854888038997,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 7154
+    },
+    {
+      "epoch": 0.17030848329048842,
+      "grad_norm": 8.363543349371657,
+      "learning_rate": 2e-05,
+      "loss": 0.8523,
+      "step": 7155
+    },
+    {
+      "epoch": 0.17033228601351993,
+      "grad_norm": 4.718504592824061,
+      "learning_rate": 2e-05,
+      "loss": 0.5556,
+      "step": 7156
+    },
+    {
+      "epoch": 0.17035608873655148,
+      "grad_norm": 4.241079006601236,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 7157
+    },
+    {
+      "epoch": 0.170379891459583,
+      "grad_norm": 3.617260510790185,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 7158
+    },
+    {
+      "epoch": 0.1704036941826145,
+      "grad_norm": 38.55803501056553,
+      "learning_rate": 2e-05,
+      "loss": 0.6739,
+      "step": 7159
+    },
+    {
+      "epoch": 0.17042749690564601,
+      "grad_norm": 4.363318236600872,
+      "learning_rate": 2e-05,
+      "loss": 0.254,
+      "step": 7160
+    },
+    {
+      "epoch": 0.17045129962867753,
+      "grad_norm": 6.237895044654617,
+      "learning_rate": 2e-05,
+      "loss": 0.952,
+      "step": 7161
+    },
+    {
+      "epoch": 0.17047510235170904,
+      "grad_norm": 16.196308765757763,
+      "learning_rate": 2e-05,
+      "loss": 2.0521,
+      "step": 7162
+    },
+    {
+      "epoch": 0.17049890507474055,
+      "grad_norm": 2.112272537516939,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 7163
+    },
+    {
+      "epoch": 0.17052270779777207,
+      "grad_norm": 8.52506072092662,
+      "learning_rate": 2e-05,
+      "loss": 1.712,
+      "step": 7164
+    },
+    {
+      "epoch": 0.17054651052080358,
+      "grad_norm": 4.77359480404339,
+      "learning_rate": 2e-05,
+      "loss": 0.2555,
+      "step": 7165
+    },
+    {
+      "epoch": 0.1705703132438351,
+      "grad_norm": 11.818578989432623,
+      "learning_rate": 2e-05,
+      "loss": 1.3257,
+      "step": 7166
+    },
+    {
+      "epoch": 0.1705941159668666,
+      "grad_norm": 2.092005093265874,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 7167
+    },
+    {
+      "epoch": 0.17061791868989812,
+      "grad_norm": 5.906037271529462,
+      "learning_rate": 2e-05,
+      "loss": 0.547,
+      "step": 7168
+    },
+    {
+      "epoch": 0.17064172141292963,
+      "grad_norm": 5.4494442220795865,
+      "learning_rate": 2e-05,
+      "loss": 1.1205,
+      "step": 7169
+    },
+    {
+      "epoch": 0.17066552413596114,
+      "grad_norm": 3.9305136620189733,
+      "learning_rate": 2e-05,
+      "loss": 0.6193,
+      "step": 7170
+    },
+    {
+      "epoch": 0.17068932685899266,
+      "grad_norm": 5.412711967388369,
+      "learning_rate": 2e-05,
+      "loss": 1.0164,
+      "step": 7171
+    },
+    {
+      "epoch": 0.17071312958202417,
+      "grad_norm": 7.621567421480244,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 7172
+    },
+    {
+      "epoch": 0.1707369323050557,
+      "grad_norm": 2.8622883276490128,
+      "learning_rate": 2e-05,
+      "loss": 0.1626,
+      "step": 7173
+    },
+    {
+      "epoch": 0.17076073502808722,
+      "grad_norm": 21.21734157837524,
+      "learning_rate": 2e-05,
+      "loss": 0.2903,
+      "step": 7174
+    },
+    {
+      "epoch": 0.17078453775111874,
+      "grad_norm": 7.863556022274003,
+      "learning_rate": 2e-05,
+      "loss": 0.8741,
+      "step": 7175
+    },
+    {
+      "epoch": 0.17080834047415025,
+      "grad_norm": 2.44384424279375,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 7176
+    },
+    {
+      "epoch": 0.17083214319718176,
+      "grad_norm": 6.926816713942697,
+      "learning_rate": 2e-05,
+      "loss": 1.0294,
+      "step": 7177
+    },
+    {
+      "epoch": 0.17085594592021328,
+      "grad_norm": 6.806635188973347,
+      "learning_rate": 2e-05,
+      "loss": 0.846,
+      "step": 7178
+    },
+    {
+      "epoch": 0.1708797486432448,
+      "grad_norm": 9.070018541215743,
+      "learning_rate": 2e-05,
+      "loss": 0.5207,
+      "step": 7179
+    },
+    {
+      "epoch": 0.1709035513662763,
+      "grad_norm": 13.289398547420054,
+      "learning_rate": 2e-05,
+      "loss": 2.1948,
+      "step": 7180
+    },
+    {
+      "epoch": 0.17092735408930781,
+      "grad_norm": 5.0574941601347,
+      "learning_rate": 2e-05,
+      "loss": 0.6819,
+      "step": 7181
+    },
+    {
+      "epoch": 0.17095115681233933,
+      "grad_norm": 8.98459454362779,
+      "learning_rate": 2e-05,
+      "loss": 1.367,
+      "step": 7182
+    },
+    {
+      "epoch": 0.17097495953537084,
+      "grad_norm": 4.4560858680908,
+      "learning_rate": 2e-05,
+      "loss": 0.8119,
+      "step": 7183
+    },
+    {
+      "epoch": 0.17099876225840235,
+      "grad_norm": 4.200951792442152,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 7184
+    },
+    {
+      "epoch": 0.17102256498143387,
+      "grad_norm": 8.519016129745523,
+      "learning_rate": 2e-05,
+      "loss": 0.8052,
+      "step": 7185
+    },
+    {
+      "epoch": 0.17104636770446538,
+      "grad_norm": 5.365048059049726,
+      "learning_rate": 2e-05,
+      "loss": 0.5348,
+      "step": 7186
+    },
+    {
+      "epoch": 0.1710701704274969,
+      "grad_norm": 10.032090141011638,
+      "learning_rate": 2e-05,
+      "loss": 0.3642,
+      "step": 7187
+    },
+    {
+      "epoch": 0.17109397315052843,
+      "grad_norm": 3.7588217429209356,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 7188
+    },
+    {
+      "epoch": 0.17111777587355995,
+      "grad_norm": 6.262151924282852,
+      "learning_rate": 2e-05,
+      "loss": 1.5044,
+      "step": 7189
+    },
+    {
+      "epoch": 0.17114157859659146,
+      "grad_norm": 1.7278287174316054,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 7190
+    },
+    {
+      "epoch": 0.17116538131962297,
+      "grad_norm": 4.188530553988361,
+      "learning_rate": 2e-05,
+      "loss": 0.3174,
+      "step": 7191
+    },
+    {
+      "epoch": 0.17118918404265449,
+      "grad_norm": 6.103324425168408,
+      "learning_rate": 2e-05,
+      "loss": 0.8661,
+      "step": 7192
+    },
+    {
+      "epoch": 0.171212986765686,
+      "grad_norm": 2.668404065195157,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 7193
+    },
+    {
+      "epoch": 0.1712367894887175,
+      "grad_norm": 5.649389721107045,
+      "learning_rate": 2e-05,
+      "loss": 0.2857,
+      "step": 7194
+    },
+    {
+      "epoch": 0.17126059221174902,
+      "grad_norm": 1.8353675058840118,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 7195
+    },
+    {
+      "epoch": 0.17128439493478054,
+      "grad_norm": 9.084033256284844,
+      "learning_rate": 2e-05,
+      "loss": 0.493,
+      "step": 7196
+    },
+    {
+      "epoch": 0.17130819765781205,
+      "grad_norm": 2.648482630550943,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 7197
+    },
+    {
+      "epoch": 0.17133200038084356,
+      "grad_norm": 2.3057677327083135,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 7198
+    },
+    {
+      "epoch": 0.17135580310387508,
+      "grad_norm": 2.983892235430335,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 7199
+    },
+    {
+      "epoch": 0.1713796058269066,
+      "grad_norm": 2.1879453156976862,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 7200
+    },
+    {
+      "epoch": 0.1714034085499381,
+      "grad_norm": 4.0818042233132905,
+      "learning_rate": 2e-05,
+      "loss": 0.1996,
+      "step": 7201
+    },
+    {
+      "epoch": 0.17142721127296962,
+      "grad_norm": 4.856895628702071,
+      "learning_rate": 2e-05,
+      "loss": 0.7979,
+      "step": 7202
+    },
+    {
+      "epoch": 0.17145101399600116,
+      "grad_norm": 11.946896686061393,
+      "learning_rate": 2e-05,
+      "loss": 0.9568,
+      "step": 7203
+    },
+    {
+      "epoch": 0.17147481671903267,
+      "grad_norm": 3.339381481572203,
+      "learning_rate": 2e-05,
+      "loss": 0.1331,
+      "step": 7204
+    },
+    {
+      "epoch": 0.17149861944206418,
+      "grad_norm": 2.528108650744825,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 7205
+    },
+    {
+      "epoch": 0.1715224221650957,
+      "grad_norm": 3.87438985129063,
+      "learning_rate": 2e-05,
+      "loss": 0.1602,
+      "step": 7206
+    },
+    {
+      "epoch": 0.1715462248881272,
+      "grad_norm": 4.608765151705503,
+      "learning_rate": 2e-05,
+      "loss": 0.9481,
+      "step": 7207
+    },
+    {
+      "epoch": 0.17157002761115872,
+      "grad_norm": 2.733893322527773,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 7208
+    },
+    {
+      "epoch": 0.17159383033419023,
+      "grad_norm": 1.9037865418571267,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 7209
+    },
+    {
+      "epoch": 0.17161763305722175,
+      "grad_norm": 7.1359493560688225,
+      "learning_rate": 2e-05,
+      "loss": 1.0681,
+      "step": 7210
+    },
+    {
+      "epoch": 0.17164143578025326,
+      "grad_norm": 4.608159333838101,
+      "learning_rate": 2e-05,
+      "loss": 0.7236,
+      "step": 7211
+    },
+    {
+      "epoch": 0.17166523850328477,
+      "grad_norm": 9.363637769842477,
+      "learning_rate": 2e-05,
+      "loss": 1.3662,
+      "step": 7212
+    },
+    {
+      "epoch": 0.17168904122631629,
+      "grad_norm": 13.661171964415665,
+      "learning_rate": 2e-05,
+      "loss": 0.3139,
+      "step": 7213
+    },
+    {
+      "epoch": 0.1717128439493478,
+      "grad_norm": 5.8237209091373305,
+      "learning_rate": 2e-05,
+      "loss": 0.9849,
+      "step": 7214
+    },
+    {
+      "epoch": 0.1717366466723793,
+      "grad_norm": 1.7953793708698995,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 7215
+    },
+    {
+      "epoch": 0.17176044939541082,
+      "grad_norm": 3.412035541556608,
+      "learning_rate": 2e-05,
+      "loss": 0.2622,
+      "step": 7216
+    },
+    {
+      "epoch": 0.17178425211844234,
+      "grad_norm": 9.255923999759615,
+      "learning_rate": 2e-05,
+      "loss": 0.3034,
+      "step": 7217
+    },
+    {
+      "epoch": 0.17180805484147388,
+      "grad_norm": 4.234276644145746,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 7218
+    },
+    {
+      "epoch": 0.1718318575645054,
+      "grad_norm": 14.062686676964306,
+      "learning_rate": 2e-05,
+      "loss": 0.4078,
+      "step": 7219
+    },
+    {
+      "epoch": 0.1718556602875369,
+      "grad_norm": 8.316351742581805,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17187946301056842,
+      "grad_norm": 5.299486481189859,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 7221
+    },
+    {
+      "epoch": 0.17190326573359993,
+      "grad_norm": 4.656147771472867,
+      "learning_rate": 2e-05,
+      "loss": 0.6741,
+      "step": 7222
+    },
+    {
+      "epoch": 0.17192706845663144,
+      "grad_norm": 3.3970568789023097,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 7223
+    },
+    {
+      "epoch": 0.17195087117966296,
+      "grad_norm": 5.300952702457449,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 7224
+    },
+    {
+      "epoch": 0.17197467390269447,
+      "grad_norm": 8.292438300787916,
+      "learning_rate": 2e-05,
+      "loss": 1.2398,
+      "step": 7225
+    },
+    {
+      "epoch": 0.17199847662572598,
+      "grad_norm": 5.540749991556987,
+      "learning_rate": 2e-05,
+      "loss": 1.0641,
+      "step": 7226
+    },
+    {
+      "epoch": 0.1720222793487575,
+      "grad_norm": 0.6321368059712824,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 7227
+    },
+    {
+      "epoch": 0.172046082071789,
+      "grad_norm": 3.196752692767998,
+      "learning_rate": 2e-05,
+      "loss": 0.2417,
+      "step": 7228
+    },
+    {
+      "epoch": 0.17206988479482052,
+      "grad_norm": 2.7362814322514972,
+      "learning_rate": 2e-05,
+      "loss": 0.2154,
+      "step": 7229
+    },
+    {
+      "epoch": 0.17209368751785203,
+      "grad_norm": 5.556854659825578,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17211749024088355,
+      "grad_norm": 5.5775735058805305,
+      "learning_rate": 2e-05,
+      "loss": 0.3058,
+      "step": 7231
+    },
+    {
+      "epoch": 0.17214129296391506,
+      "grad_norm": 1.643220515539108,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 7232
+    },
+    {
+      "epoch": 0.17216509568694657,
+      "grad_norm": 5.9736252302976025,
+      "learning_rate": 2e-05,
+      "loss": 0.9467,
+      "step": 7233
+    },
+    {
+      "epoch": 0.17218889840997811,
+      "grad_norm": 12.320846344696474,
+      "learning_rate": 2e-05,
+      "loss": 0.4355,
+      "step": 7234
+    },
+    {
+      "epoch": 0.17221270113300963,
+      "grad_norm": 16.183169693623533,
+      "learning_rate": 2e-05,
+      "loss": 1.2444,
+      "step": 7235
+    },
+    {
+      "epoch": 0.17223650385604114,
+      "grad_norm": 3.9398823453480603,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 7236
+    },
+    {
+      "epoch": 0.17226030657907265,
+      "grad_norm": 7.764673382913487,
+      "learning_rate": 2e-05,
+      "loss": 0.2356,
+      "step": 7237
+    },
+    {
+      "epoch": 0.17228410930210417,
+      "grad_norm": 5.367886775086443,
+      "learning_rate": 2e-05,
+      "loss": 0.5654,
+      "step": 7238
+    },
+    {
+      "epoch": 0.17230791202513568,
+      "grad_norm": 4.890001957251512,
+      "learning_rate": 2e-05,
+      "loss": 0.7701,
+      "step": 7239
+    },
+    {
+      "epoch": 0.1723317147481672,
+      "grad_norm": 4.80006226391396,
+      "learning_rate": 2e-05,
+      "loss": 0.2873,
+      "step": 7240
+    },
+    {
+      "epoch": 0.1723555174711987,
+      "grad_norm": 1.4005083621719396,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 7241
+    },
+    {
+      "epoch": 0.17237932019423022,
+      "grad_norm": 4.111092919472835,
+      "learning_rate": 2e-05,
+      "loss": 0.5696,
+      "step": 7242
+    },
+    {
+      "epoch": 0.17240312291726173,
+      "grad_norm": 2.6413064838214924,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 7243
+    },
+    {
+      "epoch": 0.17242692564029324,
+      "grad_norm": 13.442491430278924,
+      "learning_rate": 2e-05,
+      "loss": 0.424,
+      "step": 7244
+    },
+    {
+      "epoch": 0.17245072836332476,
+      "grad_norm": 9.12530590786161,
+      "learning_rate": 2e-05,
+      "loss": 0.1563,
+      "step": 7245
+    },
+    {
+      "epoch": 0.17247453108635627,
+      "grad_norm": 3.5893107190199594,
+      "learning_rate": 2e-05,
+      "loss": 0.2187,
+      "step": 7246
+    },
+    {
+      "epoch": 0.17249833380938778,
+      "grad_norm": 2.0125804584318248,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 7247
+    },
+    {
+      "epoch": 0.1725221365324193,
+      "grad_norm": 3.0361401114443733,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 7248
+    },
+    {
+      "epoch": 0.17254593925545084,
+      "grad_norm": 6.3357164672876145,
+      "learning_rate": 2e-05,
+      "loss": 1.3567,
+      "step": 7249
+    },
+    {
+      "epoch": 0.17256974197848235,
+      "grad_norm": 1.3759515621723137,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17259354470151386,
+      "grad_norm": 2.875230437962128,
+      "learning_rate": 2e-05,
+      "loss": 0.2174,
+      "step": 7251
+    },
+    {
+      "epoch": 0.17261734742454538,
+      "grad_norm": 2.6689759947861114,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 7252
+    },
+    {
+      "epoch": 0.1726411501475769,
+      "grad_norm": 8.504028100124684,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 7253
+    },
+    {
+      "epoch": 0.1726649528706084,
+      "grad_norm": 4.0195133006345385,
+      "learning_rate": 2e-05,
+      "loss": 0.5548,
+      "step": 7254
+    },
+    {
+      "epoch": 0.17268875559363991,
+      "grad_norm": 37.108959593328194,
+      "learning_rate": 2e-05,
+      "loss": 0.9208,
+      "step": 7255
+    },
+    {
+      "epoch": 0.17271255831667143,
+      "grad_norm": 3.417619273901354,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 7256
+    },
+    {
+      "epoch": 0.17273636103970294,
+      "grad_norm": 3.310122989385814,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 7257
+    },
+    {
+      "epoch": 0.17276016376273445,
+      "grad_norm": 4.507101664527397,
+      "learning_rate": 2e-05,
+      "loss": 0.7797,
+      "step": 7258
+    },
+    {
+      "epoch": 0.17278396648576597,
+      "grad_norm": 5.375139201343101,
+      "learning_rate": 2e-05,
+      "loss": 0.8543,
+      "step": 7259
+    },
+    {
+      "epoch": 0.17280776920879748,
+      "grad_norm": 3.3990813266394357,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 7260
+    },
+    {
+      "epoch": 0.172831571931829,
+      "grad_norm": 3.8301971629658245,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 7261
+    },
+    {
+      "epoch": 0.1728553746548605,
+      "grad_norm": 4.1652422896504895,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 7262
+    },
+    {
+      "epoch": 0.17287917737789202,
+      "grad_norm": 10.803804391456767,
+      "learning_rate": 2e-05,
+      "loss": 0.8879,
+      "step": 7263
+    },
+    {
+      "epoch": 0.17290298010092356,
+      "grad_norm": 1.2544959179137662,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 7264
+    },
+    {
+      "epoch": 0.17292678282395507,
+      "grad_norm": 8.66805219779877,
+      "learning_rate": 2e-05,
+      "loss": 0.873,
+      "step": 7265
+    },
+    {
+      "epoch": 0.17295058554698658,
+      "grad_norm": 2.2206931834412127,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 7266
+    },
+    {
+      "epoch": 0.1729743882700181,
+      "grad_norm": 7.320632885965264,
+      "learning_rate": 2e-05,
+      "loss": 1.1317,
+      "step": 7267
+    },
+    {
+      "epoch": 0.1729981909930496,
+      "grad_norm": 8.387492953714718,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 7268
+    },
+    {
+      "epoch": 0.17302199371608112,
+      "grad_norm": 0.8146010757977178,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 7269
+    },
+    {
+      "epoch": 0.17304579643911264,
+      "grad_norm": 1.7729324494846257,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17306959916214415,
+      "grad_norm": 5.678453352356136,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 7271
+    },
+    {
+      "epoch": 0.17309340188517566,
+      "grad_norm": 2.422041229000151,
+      "learning_rate": 2e-05,
+      "loss": 0.2042,
+      "step": 7272
+    },
+    {
+      "epoch": 0.17311720460820718,
+      "grad_norm": 4.866219320214256,
+      "learning_rate": 2e-05,
+      "loss": 0.5313,
+      "step": 7273
+    },
+    {
+      "epoch": 0.1731410073312387,
+      "grad_norm": 5.663930253879187,
+      "learning_rate": 2e-05,
+      "loss": 0.2284,
+      "step": 7274
+    },
+    {
+      "epoch": 0.1731648100542702,
+      "grad_norm": 9.5714379370895,
+      "learning_rate": 2e-05,
+      "loss": 1.0451,
+      "step": 7275
+    },
+    {
+      "epoch": 0.17318861277730171,
+      "grad_norm": 5.953500469084791,
+      "learning_rate": 2e-05,
+      "loss": 0.7069,
+      "step": 7276
+    },
+    {
+      "epoch": 0.17321241550033323,
+      "grad_norm": 6.076753142471356,
+      "learning_rate": 2e-05,
+      "loss": 0.118,
+      "step": 7277
+    },
+    {
+      "epoch": 0.17323621822336474,
+      "grad_norm": 4.359776915915057,
+      "learning_rate": 2e-05,
+      "loss": 0.3315,
+      "step": 7278
+    },
+    {
+      "epoch": 0.17326002094639628,
+      "grad_norm": 9.76148425606122,
+      "learning_rate": 2e-05,
+      "loss": 1.015,
+      "step": 7279
+    },
+    {
+      "epoch": 0.1732838236694278,
+      "grad_norm": 7.842941616182232,
+      "learning_rate": 2e-05,
+      "loss": 0.4668,
+      "step": 7280
+    },
+    {
+      "epoch": 0.1733076263924593,
+      "grad_norm": 6.689988347563779,
+      "learning_rate": 2e-05,
+      "loss": 1.2067,
+      "step": 7281
+    },
+    {
+      "epoch": 0.17333142911549082,
+      "grad_norm": 6.807186837822887,
+      "learning_rate": 2e-05,
+      "loss": 1.3172,
+      "step": 7282
+    },
+    {
+      "epoch": 0.17335523183852233,
+      "grad_norm": 2.125137234206982,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 7283
+    },
+    {
+      "epoch": 0.17337903456155385,
+      "grad_norm": 6.354032778461755,
+      "learning_rate": 2e-05,
+      "loss": 0.9334,
+      "step": 7284
+    },
+    {
+      "epoch": 0.17340283728458536,
+      "grad_norm": 4.080278385149727,
+      "learning_rate": 2e-05,
+      "loss": 0.2098,
+      "step": 7285
+    },
+    {
+      "epoch": 0.17342664000761687,
+      "grad_norm": 2.3342993681205857,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 7286
+    },
+    {
+      "epoch": 0.17345044273064839,
+      "grad_norm": 6.566380688432851,
+      "learning_rate": 2e-05,
+      "loss": 1.2856,
+      "step": 7287
+    },
+    {
+      "epoch": 0.1734742454536799,
+      "grad_norm": 9.132032284061046,
+      "learning_rate": 2e-05,
+      "loss": 1.0221,
+      "step": 7288
+    },
+    {
+      "epoch": 0.1734980481767114,
+      "grad_norm": 3.1133847387434623,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 7289
+    },
+    {
+      "epoch": 0.17352185089974292,
+      "grad_norm": 2.945286976040183,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 7290
+    },
+    {
+      "epoch": 0.17354565362277444,
+      "grad_norm": 5.828572409387873,
+      "learning_rate": 2e-05,
+      "loss": 1.3307,
+      "step": 7291
+    },
+    {
+      "epoch": 0.17356945634580595,
+      "grad_norm": 2.127767867317895,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 7292
+    },
+    {
+      "epoch": 0.17359325906883746,
+      "grad_norm": 7.640320099191192,
+      "learning_rate": 2e-05,
+      "loss": 0.4196,
+      "step": 7293
+    },
+    {
+      "epoch": 0.17361706179186898,
+      "grad_norm": 7.836072910547199,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 7294
+    },
+    {
+      "epoch": 0.17364086451490052,
+      "grad_norm": 3.116331804414335,
+      "learning_rate": 2e-05,
+      "loss": 0.1474,
+      "step": 7295
+    },
+    {
+      "epoch": 0.17366466723793203,
+      "grad_norm": 5.2300247476643085,
+      "learning_rate": 2e-05,
+      "loss": 0.2418,
+      "step": 7296
+    },
+    {
+      "epoch": 0.17368846996096354,
+      "grad_norm": 2.916516251784624,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 7297
+    },
+    {
+      "epoch": 0.17371227268399506,
+      "grad_norm": 1.6587428305258414,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 7298
+    },
+    {
+      "epoch": 0.17373607540702657,
+      "grad_norm": 1.5614719189362383,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 7299
+    },
+    {
+      "epoch": 0.17375987813005808,
+      "grad_norm": 2.8853774720437335,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 7300
+    },
+    {
+      "epoch": 0.1737836808530896,
+      "grad_norm": 3.135831287335706,
+      "learning_rate": 2e-05,
+      "loss": 0.1461,
+      "step": 7301
+    },
+    {
+      "epoch": 0.1738074835761211,
+      "grad_norm": 8.629989493458831,
+      "learning_rate": 2e-05,
+      "loss": 0.5008,
+      "step": 7302
+    },
+    {
+      "epoch": 0.17383128629915262,
+      "grad_norm": 7.699172625694012,
+      "learning_rate": 2e-05,
+      "loss": 0.8698,
+      "step": 7303
+    },
+    {
+      "epoch": 0.17385508902218413,
+      "grad_norm": 4.661545901831773,
+      "learning_rate": 2e-05,
+      "loss": 0.5155,
+      "step": 7304
+    },
+    {
+      "epoch": 0.17387889174521565,
+      "grad_norm": 8.832573622618431,
+      "learning_rate": 2e-05,
+      "loss": 0.5487,
+      "step": 7305
+    },
+    {
+      "epoch": 0.17390269446824716,
+      "grad_norm": 11.104984841981508,
+      "learning_rate": 2e-05,
+      "loss": 0.8188,
+      "step": 7306
+    },
+    {
+      "epoch": 0.17392649719127867,
+      "grad_norm": 9.809035491218612,
+      "learning_rate": 2e-05,
+      "loss": 0.5584,
+      "step": 7307
+    },
+    {
+      "epoch": 0.17395029991431019,
+      "grad_norm": 2.2590098416231057,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 7308
+    },
+    {
+      "epoch": 0.1739741026373417,
+      "grad_norm": 2.1603967275288776,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 7309
+    },
+    {
+      "epoch": 0.17399790536037324,
+      "grad_norm": 5.04273596716298,
+      "learning_rate": 2e-05,
+      "loss": 0.4397,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17402170808340475,
+      "grad_norm": 1.8698895134552496,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 7311
+    },
+    {
+      "epoch": 0.17404551080643627,
+      "grad_norm": 13.387537787378088,
+      "learning_rate": 2e-05,
+      "loss": 0.2085,
+      "step": 7312
+    },
+    {
+      "epoch": 0.17406931352946778,
+      "grad_norm": 2.1475770948210617,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 7313
+    },
+    {
+      "epoch": 0.1740931162524993,
+      "grad_norm": 4.617881808842144,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 7314
+    },
+    {
+      "epoch": 0.1741169189755308,
+      "grad_norm": 6.170977110013786,
+      "learning_rate": 2e-05,
+      "loss": 0.8732,
+      "step": 7315
+    },
+    {
+      "epoch": 0.17414072169856232,
+      "grad_norm": 11.77205757225149,
+      "learning_rate": 2e-05,
+      "loss": 0.6009,
+      "step": 7316
+    },
+    {
+      "epoch": 0.17416452442159383,
+      "grad_norm": 8.539798620307009,
+      "learning_rate": 2e-05,
+      "loss": 0.1605,
+      "step": 7317
+    },
+    {
+      "epoch": 0.17418832714462534,
+      "grad_norm": 2.4326985544883133,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 7318
+    },
+    {
+      "epoch": 0.17421212986765686,
+      "grad_norm": 7.716151810977045,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 7319
+    },
+    {
+      "epoch": 0.17423593259068837,
+      "grad_norm": 6.3127335633066854,
+      "learning_rate": 2e-05,
+      "loss": 1.0191,
+      "step": 7320
+    },
+    {
+      "epoch": 0.17425973531371988,
+      "grad_norm": 2.620696321415369,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 7321
+    },
+    {
+      "epoch": 0.1742835380367514,
+      "grad_norm": 6.537004829754783,
+      "learning_rate": 2e-05,
+      "loss": 0.5015,
+      "step": 7322
+    },
+    {
+      "epoch": 0.1743073407597829,
+      "grad_norm": 6.059419443378582,
+      "learning_rate": 2e-05,
+      "loss": 0.1667,
+      "step": 7323
+    },
+    {
+      "epoch": 0.17433114348281442,
+      "grad_norm": 3.9112086595136684,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 7324
+    },
+    {
+      "epoch": 0.17435494620584596,
+      "grad_norm": 2.0139263867588455,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 7325
+    },
+    {
+      "epoch": 0.17437874892887748,
+      "grad_norm": 7.404920183475875,
+      "learning_rate": 2e-05,
+      "loss": 1.0882,
+      "step": 7326
+    },
+    {
+      "epoch": 0.174402551651909,
+      "grad_norm": 5.219100441406521,
+      "learning_rate": 2e-05,
+      "loss": 0.3993,
+      "step": 7327
+    },
+    {
+      "epoch": 0.1744263543749405,
+      "grad_norm": 11.612966347596064,
+      "learning_rate": 2e-05,
+      "loss": 1.9667,
+      "step": 7328
+    },
+    {
+      "epoch": 0.17445015709797201,
+      "grad_norm": 26.364105669356846,
+      "learning_rate": 2e-05,
+      "loss": 0.621,
+      "step": 7329
+    },
+    {
+      "epoch": 0.17447395982100353,
+      "grad_norm": 5.8219869442114485,
+      "learning_rate": 2e-05,
+      "loss": 1.0928,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17449776254403504,
+      "grad_norm": 2.605550362317547,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 7331
+    },
+    {
+      "epoch": 0.17452156526706655,
+      "grad_norm": 9.461202248864659,
+      "learning_rate": 2e-05,
+      "loss": 0.2279,
+      "step": 7332
+    },
+    {
+      "epoch": 0.17454536799009807,
+      "grad_norm": 2.4222353064998066,
+      "learning_rate": 2e-05,
+      "loss": 0.1731,
+      "step": 7333
+    },
+    {
+      "epoch": 0.17456917071312958,
+      "grad_norm": 9.197122733830476,
+      "learning_rate": 2e-05,
+      "loss": 0.7217,
+      "step": 7334
+    },
+    {
+      "epoch": 0.1745929734361611,
+      "grad_norm": 6.061685124249586,
+      "learning_rate": 2e-05,
+      "loss": 0.8328,
+      "step": 7335
+    },
+    {
+      "epoch": 0.1746167761591926,
+      "grad_norm": 5.538060098145679,
+      "learning_rate": 2e-05,
+      "loss": 0.7593,
+      "step": 7336
+    },
+    {
+      "epoch": 0.17464057888222412,
+      "grad_norm": 14.121767720181996,
+      "learning_rate": 2e-05,
+      "loss": 1.7828,
+      "step": 7337
+    },
+    {
+      "epoch": 0.17466438160525563,
+      "grad_norm": 11.798363600325223,
+      "learning_rate": 2e-05,
+      "loss": 0.4458,
+      "step": 7338
+    },
+    {
+      "epoch": 0.17468818432828714,
+      "grad_norm": 9.019851311911417,
+      "learning_rate": 2e-05,
+      "loss": 0.3101,
+      "step": 7339
+    },
+    {
+      "epoch": 0.17471198705131868,
+      "grad_norm": 1.9268526007220843,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 7340
+    },
+    {
+      "epoch": 0.1747357897743502,
+      "grad_norm": 3.6669512004525857,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 7341
+    },
+    {
+      "epoch": 0.1747595924973817,
+      "grad_norm": 8.90234502188956,
+      "learning_rate": 2e-05,
+      "loss": 1.2613,
+      "step": 7342
+    },
+    {
+      "epoch": 0.17478339522041322,
+      "grad_norm": 5.094316491363461,
+      "learning_rate": 2e-05,
+      "loss": 0.9124,
+      "step": 7343
+    },
+    {
+      "epoch": 0.17480719794344474,
+      "grad_norm": 4.840162118175317,
+      "learning_rate": 2e-05,
+      "loss": 0.8756,
+      "step": 7344
+    },
+    {
+      "epoch": 0.17483100066647625,
+      "grad_norm": 9.60279448699385,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 7345
+    },
+    {
+      "epoch": 0.17485480338950776,
+      "grad_norm": 2.0705751213250094,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 7346
+    },
+    {
+      "epoch": 0.17487860611253928,
+      "grad_norm": 5.535061203054105,
+      "learning_rate": 2e-05,
+      "loss": 0.7127,
+      "step": 7347
+    },
+    {
+      "epoch": 0.1749024088355708,
+      "grad_norm": 1.2224997832355022,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 7348
+    },
+    {
+      "epoch": 0.1749262115586023,
+      "grad_norm": 6.784043986614696,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 7349
+    },
+    {
+      "epoch": 0.17495001428163381,
+      "grad_norm": 2.1383694862103986,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 7350
+    },
+    {
+      "epoch": 0.17497381700466533,
+      "grad_norm": 20.559884276433902,
+      "learning_rate": 2e-05,
+      "loss": 0.8592,
+      "step": 7351
+    },
+    {
+      "epoch": 0.17499761972769684,
+      "grad_norm": 5.751168383027525,
+      "learning_rate": 2e-05,
+      "loss": 1.3742,
+      "step": 7352
+    },
+    {
+      "epoch": 0.17502142245072835,
+      "grad_norm": 5.224099107608738,
+      "learning_rate": 2e-05,
+      "loss": 0.4035,
+      "step": 7353
+    },
+    {
+      "epoch": 0.17504522517375987,
+      "grad_norm": 4.222467491387632,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 7354
+    },
+    {
+      "epoch": 0.17506902789679138,
+      "grad_norm": 7.801971740029172,
+      "learning_rate": 2e-05,
+      "loss": 0.9438,
+      "step": 7355
+    },
+    {
+      "epoch": 0.17509283061982292,
+      "grad_norm": 1.6332216989244919,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 7356
+    },
+    {
+      "epoch": 0.17511663334285443,
+      "grad_norm": 3.2971516008394,
+      "learning_rate": 2e-05,
+      "loss": 0.4954,
+      "step": 7357
+    },
+    {
+      "epoch": 0.17514043606588595,
+      "grad_norm": 3.303841492538399,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 7358
+    },
+    {
+      "epoch": 0.17516423878891746,
+      "grad_norm": 11.764680582956503,
+      "learning_rate": 2e-05,
+      "loss": 0.8871,
+      "step": 7359
+    },
+    {
+      "epoch": 0.17518804151194897,
+      "grad_norm": 3.130975305946456,
+      "learning_rate": 2e-05,
+      "loss": 0.2684,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17521184423498048,
+      "grad_norm": 7.097784441425655,
+      "learning_rate": 2e-05,
+      "loss": 0.196,
+      "step": 7361
+    },
+    {
+      "epoch": 0.175235646958012,
+      "grad_norm": 7.100551116881917,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 7362
+    },
+    {
+      "epoch": 0.1752594496810435,
+      "grad_norm": 8.296002028930033,
+      "learning_rate": 2e-05,
+      "loss": 0.3858,
+      "step": 7363
+    },
+    {
+      "epoch": 0.17528325240407502,
+      "grad_norm": 11.519297360729274,
+      "learning_rate": 2e-05,
+      "loss": 1.0605,
+      "step": 7364
+    },
+    {
+      "epoch": 0.17530705512710654,
+      "grad_norm": 6.441151767716289,
+      "learning_rate": 2e-05,
+      "loss": 0.3755,
+      "step": 7365
+    },
+    {
+      "epoch": 0.17533085785013805,
+      "grad_norm": 5.71033653289773,
+      "learning_rate": 2e-05,
+      "loss": 0.3232,
+      "step": 7366
+    },
+    {
+      "epoch": 0.17535466057316956,
+      "grad_norm": 2.2610136855344507,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 7367
+    },
+    {
+      "epoch": 0.17537846329620108,
+      "grad_norm": 4.482947033561537,
+      "learning_rate": 2e-05,
+      "loss": 0.6126,
+      "step": 7368
+    },
+    {
+      "epoch": 0.1754022660192326,
+      "grad_norm": 6.580826665583236,
+      "learning_rate": 2e-05,
+      "loss": 1.1779,
+      "step": 7369
+    },
+    {
+      "epoch": 0.1754260687422641,
+      "grad_norm": 18.739252426749193,
+      "learning_rate": 2e-05,
+      "loss": 0.91,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17544987146529564,
+      "grad_norm": 9.317847736053675,
+      "learning_rate": 2e-05,
+      "loss": 0.4587,
+      "step": 7371
+    },
+    {
+      "epoch": 0.17547367418832716,
+      "grad_norm": 1.4886197974212623,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 7372
+    },
+    {
+      "epoch": 0.17549747691135867,
+      "grad_norm": 6.6993910515224355,
+      "learning_rate": 2e-05,
+      "loss": 0.7136,
+      "step": 7373
+    },
+    {
+      "epoch": 0.17552127963439018,
+      "grad_norm": 7.355059167848569,
+      "learning_rate": 2e-05,
+      "loss": 0.2673,
+      "step": 7374
+    },
+    {
+      "epoch": 0.1755450823574217,
+      "grad_norm": 2.2559503493547584,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 7375
+    },
+    {
+      "epoch": 0.1755688850804532,
+      "grad_norm": 4.261259474739834,
+      "learning_rate": 2e-05,
+      "loss": 0.583,
+      "step": 7376
+    },
+    {
+      "epoch": 0.17559268780348472,
+      "grad_norm": 5.037340412991527,
+      "learning_rate": 2e-05,
+      "loss": 0.9209,
+      "step": 7377
+    },
+    {
+      "epoch": 0.17561649052651623,
+      "grad_norm": 9.520543101941437,
+      "learning_rate": 2e-05,
+      "loss": 0.4585,
+      "step": 7378
+    },
+    {
+      "epoch": 0.17564029324954775,
+      "grad_norm": 15.316242368708696,
+      "learning_rate": 2e-05,
+      "loss": 0.5036,
+      "step": 7379
+    },
+    {
+      "epoch": 0.17566409597257926,
+      "grad_norm": 6.133313855233547,
+      "learning_rate": 2e-05,
+      "loss": 0.8973,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17568789869561077,
+      "grad_norm": 1.6581650200944147,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 7381
+    },
+    {
+      "epoch": 0.17571170141864229,
+      "grad_norm": 8.993169620432303,
+      "learning_rate": 2e-05,
+      "loss": 0.3797,
+      "step": 7382
+    },
+    {
+      "epoch": 0.1757355041416738,
+      "grad_norm": 2.4661505296411077,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 7383
+    },
+    {
+      "epoch": 0.1757593068647053,
+      "grad_norm": 7.221299897727725,
+      "learning_rate": 2e-05,
+      "loss": 0.4213,
+      "step": 7384
+    },
+    {
+      "epoch": 0.17578310958773682,
+      "grad_norm": 15.460006903020316,
+      "learning_rate": 2e-05,
+      "loss": 0.9463,
+      "step": 7385
+    },
+    {
+      "epoch": 0.17580691231076837,
+      "grad_norm": 3.4575081746820095,
+      "learning_rate": 2e-05,
+      "loss": 0.4101,
+      "step": 7386
+    },
+    {
+      "epoch": 0.17583071503379988,
+      "grad_norm": 1.883811936671577,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 7387
+    },
+    {
+      "epoch": 0.1758545177568314,
+      "grad_norm": 5.7809140630212035,
+      "learning_rate": 2e-05,
+      "loss": 0.7172,
+      "step": 7388
+    },
+    {
+      "epoch": 0.1758783204798629,
+      "grad_norm": 8.873323689055715,
+      "learning_rate": 2e-05,
+      "loss": 1.1277,
+      "step": 7389
+    },
+    {
+      "epoch": 0.17590212320289442,
+      "grad_norm": 4.087689981154122,
+      "learning_rate": 2e-05,
+      "loss": 0.9393,
+      "step": 7390
+    },
+    {
+      "epoch": 0.17592592592592593,
+      "grad_norm": 9.517224268844135,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 7391
+    },
+    {
+      "epoch": 0.17594972864895744,
+      "grad_norm": 3.097376615476808,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 7392
+    },
+    {
+      "epoch": 0.17597353137198896,
+      "grad_norm": 2.046965240269777,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 7393
+    },
+    {
+      "epoch": 0.17599733409502047,
+      "grad_norm": 2.587834303955614,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 7394
+    },
+    {
+      "epoch": 0.17602113681805198,
+      "grad_norm": 5.575540390599727,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 7395
+    },
+    {
+      "epoch": 0.1760449395410835,
+      "grad_norm": 2.7779336512376505,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 7396
+    },
+    {
+      "epoch": 0.176068742264115,
+      "grad_norm": 4.06301292017643,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 7397
+    },
+    {
+      "epoch": 0.17609254498714652,
+      "grad_norm": 3.044847585168624,
+      "learning_rate": 2e-05,
+      "loss": 0.132,
+      "step": 7398
+    },
+    {
+      "epoch": 0.17611634771017803,
+      "grad_norm": 3.6632867955711967,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 7399
+    },
+    {
+      "epoch": 0.17614015043320955,
+      "grad_norm": 7.924354256255379,
+      "learning_rate": 2e-05,
+      "loss": 0.3495,
+      "step": 7400
+    },
+    {
+      "epoch": 0.1761639531562411,
+      "grad_norm": 4.828174894109392,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 7401
+    },
+    {
+      "epoch": 0.1761877558792726,
+      "grad_norm": 2.4701121795616428,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 7402
+    },
+    {
+      "epoch": 0.1762115586023041,
+      "grad_norm": 1.9615099901911164,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 7403
+    },
+    {
+      "epoch": 0.17623536132533563,
+      "grad_norm": 2.3062818785060166,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 7404
+    },
+    {
+      "epoch": 0.17625916404836714,
+      "grad_norm": 10.125117559530858,
+      "learning_rate": 2e-05,
+      "loss": 0.4487,
+      "step": 7405
+    },
+    {
+      "epoch": 0.17628296677139865,
+      "grad_norm": 2.294164344391315,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 7406
+    },
+    {
+      "epoch": 0.17630676949443017,
+      "grad_norm": 2.1746413626139782,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 7407
+    },
+    {
+      "epoch": 0.17633057221746168,
+      "grad_norm": 19.808569253331935,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 7408
+    },
+    {
+      "epoch": 0.1763543749404932,
+      "grad_norm": 5.859870328835741,
+      "learning_rate": 2e-05,
+      "loss": 0.9291,
+      "step": 7409
+    },
+    {
+      "epoch": 0.1763781776635247,
+      "grad_norm": 9.516755238015037,
+      "learning_rate": 2e-05,
+      "loss": 0.4986,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17640198038655622,
+      "grad_norm": 4.358785881582254,
+      "learning_rate": 2e-05,
+      "loss": 0.3437,
+      "step": 7411
+    },
+    {
+      "epoch": 0.17642578310958773,
+      "grad_norm": 10.364024680649964,
+      "learning_rate": 2e-05,
+      "loss": 1.293,
+      "step": 7412
+    },
+    {
+      "epoch": 0.17644958583261924,
+      "grad_norm": 6.356058995687072,
+      "learning_rate": 2e-05,
+      "loss": 0.7083,
+      "step": 7413
+    },
+    {
+      "epoch": 0.17647338855565076,
+      "grad_norm": 5.74262886308833,
+      "learning_rate": 2e-05,
+      "loss": 0.6072,
+      "step": 7414
+    },
+    {
+      "epoch": 0.17649719127868227,
+      "grad_norm": 3.4220421053825936,
+      "learning_rate": 2e-05,
+      "loss": 0.394,
+      "step": 7415
+    },
+    {
+      "epoch": 0.17652099400171378,
+      "grad_norm": 5.7706412808770935,
+      "learning_rate": 2e-05,
+      "loss": 0.6519,
+      "step": 7416
+    },
+    {
+      "epoch": 0.17654479672474532,
+      "grad_norm": 14.587539257593177,
+      "learning_rate": 2e-05,
+      "loss": 0.6794,
+      "step": 7417
+    },
+    {
+      "epoch": 0.17656859944777684,
+      "grad_norm": 2.835613215539488,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 7418
+    },
+    {
+      "epoch": 0.17659240217080835,
+      "grad_norm": 1.285668204203725,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 7419
+    },
+    {
+      "epoch": 0.17661620489383986,
+      "grad_norm": 16.564638494490286,
+      "learning_rate": 2e-05,
+      "loss": 0.7946,
+      "step": 7420
+    },
+    {
+      "epoch": 0.17664000761687138,
+      "grad_norm": 1.7520845288423155,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 7421
+    },
+    {
+      "epoch": 0.1766638103399029,
+      "grad_norm": 3.503207470238726,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 7422
+    },
+    {
+      "epoch": 0.1766876130629344,
+      "grad_norm": 13.024375426825154,
+      "learning_rate": 2e-05,
+      "loss": 1.155,
+      "step": 7423
+    },
+    {
+      "epoch": 0.17671141578596591,
+      "grad_norm": 1.0426886544293814,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 7424
+    },
+    {
+      "epoch": 0.17673521850899743,
+      "grad_norm": 3.4739454778049996,
+      "learning_rate": 2e-05,
+      "loss": 0.2772,
+      "step": 7425
+    },
+    {
+      "epoch": 0.17675902123202894,
+      "grad_norm": 4.488036145736124,
+      "learning_rate": 2e-05,
+      "loss": 0.1955,
+      "step": 7426
+    },
+    {
+      "epoch": 0.17678282395506045,
+      "grad_norm": 5.324457752574794,
+      "learning_rate": 2e-05,
+      "loss": 0.6217,
+      "step": 7427
+    },
+    {
+      "epoch": 0.17680662667809197,
+      "grad_norm": 3.56016950185847,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 7428
+    },
+    {
+      "epoch": 0.17683042940112348,
+      "grad_norm": 5.581723278353547,
+      "learning_rate": 2e-05,
+      "loss": 0.4983,
+      "step": 7429
+    },
+    {
+      "epoch": 0.176854232124155,
+      "grad_norm": 2.160727149658101,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 7430
+    },
+    {
+      "epoch": 0.1768780348471865,
+      "grad_norm": 4.276575116636022,
+      "learning_rate": 2e-05,
+      "loss": 0.2911,
+      "step": 7431
+    },
+    {
+      "epoch": 0.17690183757021805,
+      "grad_norm": 3.8451480496880777,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 7432
+    },
+    {
+      "epoch": 0.17692564029324956,
+      "grad_norm": 8.228230725143696,
+      "learning_rate": 2e-05,
+      "loss": 0.3153,
+      "step": 7433
+    },
+    {
+      "epoch": 0.17694944301628107,
+      "grad_norm": 3.862674935735728,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 7434
+    },
+    {
+      "epoch": 0.17697324573931258,
+      "grad_norm": 11.525264191702526,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 7435
+    },
+    {
+      "epoch": 0.1769970484623441,
+      "grad_norm": 2.6132815550825335,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 7436
+    },
+    {
+      "epoch": 0.1770208511853756,
+      "grad_norm": 7.826952323089702,
+      "learning_rate": 2e-05,
+      "loss": 0.3077,
+      "step": 7437
+    },
+    {
+      "epoch": 0.17704465390840712,
+      "grad_norm": 12.689987848790345,
+      "learning_rate": 2e-05,
+      "loss": 0.6834,
+      "step": 7438
+    },
+    {
+      "epoch": 0.17706845663143864,
+      "grad_norm": 16.758378868686684,
+      "learning_rate": 2e-05,
+      "loss": 0.5553,
+      "step": 7439
+    },
+    {
+      "epoch": 0.17709225935447015,
+      "grad_norm": 3.677638619648624,
+      "learning_rate": 2e-05,
+      "loss": 0.3698,
+      "step": 7440
+    },
+    {
+      "epoch": 0.17711606207750166,
+      "grad_norm": 6.806322432269248,
+      "learning_rate": 2e-05,
+      "loss": 0.5615,
+      "step": 7441
+    },
+    {
+      "epoch": 0.17713986480053318,
+      "grad_norm": 9.301270424173063,
+      "learning_rate": 2e-05,
+      "loss": 0.4742,
+      "step": 7442
+    },
+    {
+      "epoch": 0.1771636675235647,
+      "grad_norm": 3.2828835950943414,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 7443
+    },
+    {
+      "epoch": 0.1771874702465962,
+      "grad_norm": 3.531789613391319,
+      "learning_rate": 2e-05,
+      "loss": 0.1673,
+      "step": 7444
+    },
+    {
+      "epoch": 0.17721127296962771,
+      "grad_norm": 11.326798595975177,
+      "learning_rate": 2e-05,
+      "loss": 1.2413,
+      "step": 7445
+    },
+    {
+      "epoch": 0.17723507569265923,
+      "grad_norm": 2.3464257256089462,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 7446
+    },
+    {
+      "epoch": 0.17725887841569077,
+      "grad_norm": 4.50271742917633,
+      "learning_rate": 2e-05,
+      "loss": 0.2805,
+      "step": 7447
+    },
+    {
+      "epoch": 0.17728268113872228,
+      "grad_norm": 3.2835286367518304,
+      "learning_rate": 2e-05,
+      "loss": 0.181,
+      "step": 7448
+    },
+    {
+      "epoch": 0.1773064838617538,
+      "grad_norm": 4.90335657987519,
+      "learning_rate": 2e-05,
+      "loss": 0.7003,
+      "step": 7449
+    },
+    {
+      "epoch": 0.1773302865847853,
+      "grad_norm": 0.6716740218732463,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 7450
+    },
+    {
+      "epoch": 0.17735408930781682,
+      "grad_norm": 2.8649347730324313,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 7451
+    },
+    {
+      "epoch": 0.17737789203084833,
+      "grad_norm": 19.466860430597254,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 7452
+    },
+    {
+      "epoch": 0.17740169475387985,
+      "grad_norm": 2.359680571963659,
+      "learning_rate": 2e-05,
+      "loss": 0.1504,
+      "step": 7453
+    },
+    {
+      "epoch": 0.17742549747691136,
+      "grad_norm": 6.418629759614849,
+      "learning_rate": 2e-05,
+      "loss": 1.0433,
+      "step": 7454
+    },
+    {
+      "epoch": 0.17744930019994287,
+      "grad_norm": 5.906022012408622,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 7455
+    },
+    {
+      "epoch": 0.17747310292297439,
+      "grad_norm": 4.366454277978694,
+      "learning_rate": 2e-05,
+      "loss": 0.2869,
+      "step": 7456
+    },
+    {
+      "epoch": 0.1774969056460059,
+      "grad_norm": 8.412055000546683,
+      "learning_rate": 2e-05,
+      "loss": 0.6649,
+      "step": 7457
+    },
+    {
+      "epoch": 0.1775207083690374,
+      "grad_norm": 5.0263939856370365,
+      "learning_rate": 2e-05,
+      "loss": 0.5162,
+      "step": 7458
+    },
+    {
+      "epoch": 0.17754451109206892,
+      "grad_norm": 1.8529037271322044,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 7459
+    },
+    {
+      "epoch": 0.17756831381510044,
+      "grad_norm": 2.3148360356389217,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 7460
+    },
+    {
+      "epoch": 0.17759211653813195,
+      "grad_norm": 3.2666706375423766,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 7461
+    },
+    {
+      "epoch": 0.1776159192611635,
+      "grad_norm": 4.663821723911158,
+      "learning_rate": 2e-05,
+      "loss": 0.653,
+      "step": 7462
+    },
+    {
+      "epoch": 0.177639721984195,
+      "grad_norm": 6.736132072520694,
+      "learning_rate": 2e-05,
+      "loss": 1.0002,
+      "step": 7463
+    },
+    {
+      "epoch": 0.17766352470722652,
+      "grad_norm": 6.287180856476545,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 7464
+    },
+    {
+      "epoch": 0.17768732743025803,
+      "grad_norm": 4.979841320736399,
+      "learning_rate": 2e-05,
+      "loss": 0.2821,
+      "step": 7465
+    },
+    {
+      "epoch": 0.17771113015328954,
+      "grad_norm": 17.40699931394921,
+      "learning_rate": 2e-05,
+      "loss": 0.4676,
+      "step": 7466
+    },
+    {
+      "epoch": 0.17773493287632106,
+      "grad_norm": 3.6800316446278467,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 7467
+    },
+    {
+      "epoch": 0.17775873559935257,
+      "grad_norm": 4.6612858801961865,
+      "learning_rate": 2e-05,
+      "loss": 0.9295,
+      "step": 7468
+    },
+    {
+      "epoch": 0.17778253832238408,
+      "grad_norm": 2.9668850756772964,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 7469
+    },
+    {
+      "epoch": 0.1778063410454156,
+      "grad_norm": 2.0506112775018117,
+      "learning_rate": 2e-05,
+      "loss": 0.0966,
+      "step": 7470
+    },
+    {
+      "epoch": 0.1778301437684471,
+      "grad_norm": 2.610459412489211,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 7471
+    },
+    {
+      "epoch": 0.17785394649147862,
+      "grad_norm": 5.1497481456061305,
+      "learning_rate": 2e-05,
+      "loss": 0.6366,
+      "step": 7472
+    },
+    {
+      "epoch": 0.17787774921451013,
+      "grad_norm": 6.142003760989738,
+      "learning_rate": 2e-05,
+      "loss": 0.937,
+      "step": 7473
+    },
+    {
+      "epoch": 0.17790155193754165,
+      "grad_norm": 4.802423699672755,
+      "learning_rate": 2e-05,
+      "loss": 0.5146,
+      "step": 7474
+    },
+    {
+      "epoch": 0.17792535466057316,
+      "grad_norm": 11.63298201177241,
+      "learning_rate": 2e-05,
+      "loss": 1.2899,
+      "step": 7475
+    },
+    {
+      "epoch": 0.17794915738360467,
+      "grad_norm": 4.772836216959087,
+      "learning_rate": 2e-05,
+      "loss": 0.8713,
+      "step": 7476
+    },
+    {
+      "epoch": 0.17797296010663619,
+      "grad_norm": 3.6022755731616027,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 7477
+    },
+    {
+      "epoch": 0.17799676282966773,
+      "grad_norm": 4.009476151772133,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 7478
+    },
+    {
+      "epoch": 0.17802056555269924,
+      "grad_norm": 4.665881038149135,
+      "learning_rate": 2e-05,
+      "loss": 1.0698,
+      "step": 7479
+    },
+    {
+      "epoch": 0.17804436827573075,
+      "grad_norm": 8.56630087492668,
+      "learning_rate": 2e-05,
+      "loss": 0.5843,
+      "step": 7480
+    },
+    {
+      "epoch": 0.17806817099876227,
+      "grad_norm": 31.898348988636375,
+      "learning_rate": 2e-05,
+      "loss": 1.3539,
+      "step": 7481
+    },
+    {
+      "epoch": 0.17809197372179378,
+      "grad_norm": 1.682498181583016,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 7482
+    },
+    {
+      "epoch": 0.1781157764448253,
+      "grad_norm": 1.9303147533639262,
+      "learning_rate": 2e-05,
+      "loss": 0.0937,
+      "step": 7483
+    },
+    {
+      "epoch": 0.1781395791678568,
+      "grad_norm": 1.7966034476950434,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 7484
+    },
+    {
+      "epoch": 0.17816338189088832,
+      "grad_norm": 2.982813866208441,
+      "learning_rate": 2e-05,
+      "loss": 0.134,
+      "step": 7485
+    },
+    {
+      "epoch": 0.17818718461391983,
+      "grad_norm": 4.661510111429969,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 7486
+    },
+    {
+      "epoch": 0.17821098733695134,
+      "grad_norm": 10.886246993047584,
+      "learning_rate": 2e-05,
+      "loss": 0.4393,
+      "step": 7487
+    },
+    {
+      "epoch": 0.17823479005998286,
+      "grad_norm": 2.2411485034871665,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 7488
+    },
+    {
+      "epoch": 0.17825859278301437,
+      "grad_norm": 2.6444573143532883,
+      "learning_rate": 2e-05,
+      "loss": 0.1732,
+      "step": 7489
+    },
+    {
+      "epoch": 0.17828239550604588,
+      "grad_norm": 5.678474190685662,
+      "learning_rate": 2e-05,
+      "loss": 0.787,
+      "step": 7490
+    },
+    {
+      "epoch": 0.1783061982290774,
+      "grad_norm": 7.21918973194097,
+      "learning_rate": 2e-05,
+      "loss": 1.13,
+      "step": 7491
+    },
+    {
+      "epoch": 0.1783300009521089,
+      "grad_norm": 2.655508938803968,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 7492
+    },
+    {
+      "epoch": 0.17835380367514045,
+      "grad_norm": 5.770313240080441,
+      "learning_rate": 2e-05,
+      "loss": 0.9839,
+      "step": 7493
+    },
+    {
+      "epoch": 0.17837760639817196,
+      "grad_norm": 5.643397815552452,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 7494
+    },
+    {
+      "epoch": 0.17840140912120347,
+      "grad_norm": 2.8661731830161483,
+      "learning_rate": 2e-05,
+      "loss": 0.1172,
+      "step": 7495
+    },
+    {
+      "epoch": 0.178425211844235,
+      "grad_norm": 1.1419694045736963,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 7496
+    },
+    {
+      "epoch": 0.1784490145672665,
+      "grad_norm": 6.347835030797733,
+      "learning_rate": 2e-05,
+      "loss": 0.5429,
+      "step": 7497
+    },
+    {
+      "epoch": 0.178472817290298,
+      "grad_norm": 5.711270463119168,
+      "learning_rate": 2e-05,
+      "loss": 0.7376,
+      "step": 7498
+    },
+    {
+      "epoch": 0.17849662001332953,
+      "grad_norm": 2.6523427842587273,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 7499
+    },
+    {
+      "epoch": 0.17852042273636104,
+      "grad_norm": 1.4560907490320516,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 7500
+    },
+    {
+      "epoch": 0.17854422545939255,
+      "grad_norm": 4.493185095408024,
+      "learning_rate": 2e-05,
+      "loss": 0.4397,
+      "step": 7501
+    },
+    {
+      "epoch": 0.17856802818242407,
+      "grad_norm": 6.065488946356345,
+      "learning_rate": 2e-05,
+      "loss": 0.9401,
+      "step": 7502
+    },
+    {
+      "epoch": 0.17859183090545558,
+      "grad_norm": 7.061138045686747,
+      "learning_rate": 2e-05,
+      "loss": 1.0014,
+      "step": 7503
+    },
+    {
+      "epoch": 0.1786156336284871,
+      "grad_norm": 4.297415843156935,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 7504
+    },
+    {
+      "epoch": 0.1786394363515186,
+      "grad_norm": 4.3599790259442415,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 7505
+    },
+    {
+      "epoch": 0.17866323907455012,
+      "grad_norm": 3.244528518357138,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 7506
+    },
+    {
+      "epoch": 0.17868704179758163,
+      "grad_norm": 2.587211056648185,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 7507
+    },
+    {
+      "epoch": 0.17871084452061317,
+      "grad_norm": 6.889989103461292,
+      "learning_rate": 2e-05,
+      "loss": 0.6463,
+      "step": 7508
+    },
+    {
+      "epoch": 0.17873464724364468,
+      "grad_norm": 10.930061162596079,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 7509
+    },
+    {
+      "epoch": 0.1787584499666762,
+      "grad_norm": 3.2551236092806417,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 7510
+    },
+    {
+      "epoch": 0.1787822526897077,
+      "grad_norm": 5.656580565837561,
+      "learning_rate": 2e-05,
+      "loss": 0.9083,
+      "step": 7511
+    },
+    {
+      "epoch": 0.17880605541273922,
+      "grad_norm": 34.34370578879444,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 7512
+    },
+    {
+      "epoch": 0.17882985813577074,
+      "grad_norm": 6.901098186295622,
+      "learning_rate": 2e-05,
+      "loss": 0.7798,
+      "step": 7513
+    },
+    {
+      "epoch": 0.17885366085880225,
+      "grad_norm": 6.408837847792235,
+      "learning_rate": 2e-05,
+      "loss": 1.0521,
+      "step": 7514
+    },
+    {
+      "epoch": 0.17887746358183376,
+      "grad_norm": 3.1664283145812617,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 7515
+    },
+    {
+      "epoch": 0.17890126630486528,
+      "grad_norm": 7.203698328760522,
+      "learning_rate": 2e-05,
+      "loss": 0.3368,
+      "step": 7516
+    },
+    {
+      "epoch": 0.1789250690278968,
+      "grad_norm": 6.246890340821796,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 7517
+    },
+    {
+      "epoch": 0.1789488717509283,
+      "grad_norm": 8.472925658936699,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 7518
+    },
+    {
+      "epoch": 0.17897267447395981,
+      "grad_norm": 5.383489305058106,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 7519
+    },
+    {
+      "epoch": 0.17899647719699133,
+      "grad_norm": 2.4358366654728023,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 7520
+    },
+    {
+      "epoch": 0.17902027992002284,
+      "grad_norm": 5.800402716831876,
+      "learning_rate": 2e-05,
+      "loss": 0.5568,
+      "step": 7521
+    },
+    {
+      "epoch": 0.17904408264305435,
+      "grad_norm": 3.856862428423238,
+      "learning_rate": 2e-05,
+      "loss": 0.3211,
+      "step": 7522
+    },
+    {
+      "epoch": 0.1790678853660859,
+      "grad_norm": 3.0406079337617973,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 7523
+    },
+    {
+      "epoch": 0.1790916880891174,
+      "grad_norm": 1.587914917524078,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 7524
+    },
+    {
+      "epoch": 0.17911549081214892,
+      "grad_norm": 6.354667122791061,
+      "learning_rate": 2e-05,
+      "loss": 0.9133,
+      "step": 7525
+    },
+    {
+      "epoch": 0.17913929353518043,
+      "grad_norm": 1.8068614852434384,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 7526
+    },
+    {
+      "epoch": 0.17916309625821195,
+      "grad_norm": 2.335278833731774,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 7527
+    },
+    {
+      "epoch": 0.17918689898124346,
+      "grad_norm": 12.465891388922294,
+      "learning_rate": 2e-05,
+      "loss": 0.9607,
+      "step": 7528
+    },
+    {
+      "epoch": 0.17921070170427497,
+      "grad_norm": 4.60930786654298,
+      "learning_rate": 2e-05,
+      "loss": 0.6724,
+      "step": 7529
+    },
+    {
+      "epoch": 0.17923450442730648,
+      "grad_norm": 5.062456932317272,
+      "learning_rate": 2e-05,
+      "loss": 1.1083,
+      "step": 7530
+    },
+    {
+      "epoch": 0.179258307150338,
+      "grad_norm": 8.061880557834904,
+      "learning_rate": 2e-05,
+      "loss": 0.3564,
+      "step": 7531
+    },
+    {
+      "epoch": 0.1792821098733695,
+      "grad_norm": 4.212572941175243,
+      "learning_rate": 2e-05,
+      "loss": 0.5065,
+      "step": 7532
+    },
+    {
+      "epoch": 0.17930591259640102,
+      "grad_norm": 3.083400980960136,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 7533
+    },
+    {
+      "epoch": 0.17932971531943254,
+      "grad_norm": 2.37587176961505,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 7534
+    },
+    {
+      "epoch": 0.17935351804246405,
+      "grad_norm": 1.9509445901299802,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 7535
+    },
+    {
+      "epoch": 0.17937732076549556,
+      "grad_norm": 9.314803846093087,
+      "learning_rate": 2e-05,
+      "loss": 1.3547,
+      "step": 7536
+    },
+    {
+      "epoch": 0.17940112348852708,
+      "grad_norm": 13.899709028096462,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 7537
+    },
+    {
+      "epoch": 0.1794249262115586,
+      "grad_norm": 2.8711319699511537,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 7538
+    },
+    {
+      "epoch": 0.17944872893459013,
+      "grad_norm": 10.623174479575624,
+      "learning_rate": 2e-05,
+      "loss": 0.2524,
+      "step": 7539
+    },
+    {
+      "epoch": 0.17947253165762164,
+      "grad_norm": 2.866811949481677,
+      "learning_rate": 2e-05,
+      "loss": 0.6048,
+      "step": 7540
+    },
+    {
+      "epoch": 0.17949633438065316,
+      "grad_norm": 13.071887787354152,
+      "learning_rate": 2e-05,
+      "loss": 0.874,
+      "step": 7541
+    },
+    {
+      "epoch": 0.17952013710368467,
+      "grad_norm": 11.357536809123555,
+      "learning_rate": 2e-05,
+      "loss": 1.0978,
+      "step": 7542
+    },
+    {
+      "epoch": 0.17954393982671618,
+      "grad_norm": 4.622213769304759,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 7543
+    },
+    {
+      "epoch": 0.1795677425497477,
+      "grad_norm": 7.115118823928388,
+      "learning_rate": 2e-05,
+      "loss": 1.2654,
+      "step": 7544
+    },
+    {
+      "epoch": 0.1795915452727792,
+      "grad_norm": 2.011746402442547,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 7545
+    },
+    {
+      "epoch": 0.17961534799581072,
+      "grad_norm": 4.034051609136661,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 7546
+    },
+    {
+      "epoch": 0.17963915071884223,
+      "grad_norm": 0.3814017190411563,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 7547
+    },
+    {
+      "epoch": 0.17966295344187375,
+      "grad_norm": 7.6069925845221125,
+      "learning_rate": 2e-05,
+      "loss": 0.4649,
+      "step": 7548
+    },
+    {
+      "epoch": 0.17968675616490526,
+      "grad_norm": 2.7464181660813165,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 7549
+    },
+    {
+      "epoch": 0.17971055888793677,
+      "grad_norm": 5.291823220040867,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 7550
+    },
+    {
+      "epoch": 0.17973436161096829,
+      "grad_norm": 1.689961508446543,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 7551
+    },
+    {
+      "epoch": 0.1797581643339998,
+      "grad_norm": 3.7095596209552895,
+      "learning_rate": 2e-05,
+      "loss": 0.2124,
+      "step": 7552
+    },
+    {
+      "epoch": 0.1797819670570313,
+      "grad_norm": 6.153524591974679,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 7553
+    },
+    {
+      "epoch": 0.17980576978006285,
+      "grad_norm": 2.7046837813508757,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 7554
+    },
+    {
+      "epoch": 0.17982957250309436,
+      "grad_norm": 7.153190427391128,
+      "learning_rate": 2e-05,
+      "loss": 1.0822,
+      "step": 7555
+    },
+    {
+      "epoch": 0.17985337522612588,
+      "grad_norm": 6.45310934558768,
+      "learning_rate": 2e-05,
+      "loss": 0.2438,
+      "step": 7556
+    },
+    {
+      "epoch": 0.1798771779491574,
+      "grad_norm": 9.933848152248448,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 7557
+    },
+    {
+      "epoch": 0.1799009806721889,
+      "grad_norm": 3.464723470891979,
+      "learning_rate": 2e-05,
+      "loss": 0.2305,
+      "step": 7558
+    },
+    {
+      "epoch": 0.17992478339522042,
+      "grad_norm": 5.9000052593723575,
+      "learning_rate": 2e-05,
+      "loss": 0.4105,
+      "step": 7559
+    },
+    {
+      "epoch": 0.17994858611825193,
+      "grad_norm": 3.052468964183582,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 7560
+    },
+    {
+      "epoch": 0.17997238884128344,
+      "grad_norm": 3.621895769707599,
+      "learning_rate": 2e-05,
+      "loss": 0.555,
+      "step": 7561
+    },
+    {
+      "epoch": 0.17999619156431496,
+      "grad_norm": 1.5908750247969077,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 7562
+    },
+    {
+      "epoch": 0.18001999428734647,
+      "grad_norm": 2.76635192923962,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 7563
+    },
+    {
+      "epoch": 0.18004379701037798,
+      "grad_norm": 5.175103379985253,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 7564
+    },
+    {
+      "epoch": 0.1800675997334095,
+      "grad_norm": 8.663074289024028,
+      "learning_rate": 2e-05,
+      "loss": 0.7717,
+      "step": 7565
+    },
+    {
+      "epoch": 0.180091402456441,
+      "grad_norm": 2.903066508908633,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 7566
+    },
+    {
+      "epoch": 0.18011520517947252,
+      "grad_norm": 2.663849673845848,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 7567
+    },
+    {
+      "epoch": 0.18013900790250403,
+      "grad_norm": 1.4346694279468353,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 7568
+    },
+    {
+      "epoch": 0.18016281062553557,
+      "grad_norm": 6.731854469572837,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 7569
+    },
+    {
+      "epoch": 0.1801866133485671,
+      "grad_norm": 12.905697351478695,
+      "learning_rate": 2e-05,
+      "loss": 1.3659,
+      "step": 7570
+    },
+    {
+      "epoch": 0.1802104160715986,
+      "grad_norm": 1.3583441423750167,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 7571
+    },
+    {
+      "epoch": 0.1802342187946301,
+      "grad_norm": 5.849829287834972,
+      "learning_rate": 2e-05,
+      "loss": 0.2862,
+      "step": 7572
+    },
+    {
+      "epoch": 0.18025802151766163,
+      "grad_norm": 3.840845730834769,
+      "learning_rate": 2e-05,
+      "loss": 0.1737,
+      "step": 7573
+    },
+    {
+      "epoch": 0.18028182424069314,
+      "grad_norm": 2.715880894622216,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 7574
+    },
+    {
+      "epoch": 0.18030562696372465,
+      "grad_norm": 1.614029038703274,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 7575
+    },
+    {
+      "epoch": 0.18032942968675617,
+      "grad_norm": 5.9973355081965805,
+      "learning_rate": 2e-05,
+      "loss": 0.7172,
+      "step": 7576
+    },
+    {
+      "epoch": 0.18035323240978768,
+      "grad_norm": 9.0668495010709,
+      "learning_rate": 2e-05,
+      "loss": 1.016,
+      "step": 7577
+    },
+    {
+      "epoch": 0.1803770351328192,
+      "grad_norm": 5.615021745404934,
+      "learning_rate": 2e-05,
+      "loss": 0.2723,
+      "step": 7578
+    },
+    {
+      "epoch": 0.1804008378558507,
+      "grad_norm": 9.730256918464187,
+      "learning_rate": 2e-05,
+      "loss": 0.9051,
+      "step": 7579
+    },
+    {
+      "epoch": 0.18042464057888222,
+      "grad_norm": 7.920183043636886,
+      "learning_rate": 2e-05,
+      "loss": 0.381,
+      "step": 7580
+    },
+    {
+      "epoch": 0.18044844330191373,
+      "grad_norm": 8.726763639844199,
+      "learning_rate": 2e-05,
+      "loss": 1.0493,
+      "step": 7581
+    },
+    {
+      "epoch": 0.18047224602494524,
+      "grad_norm": 2.9889629607615236,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 7582
+    },
+    {
+      "epoch": 0.18049604874797676,
+      "grad_norm": 1.491517557706344,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 7583
+    },
+    {
+      "epoch": 0.1805198514710083,
+      "grad_norm": 6.707454880995858,
+      "learning_rate": 2e-05,
+      "loss": 0.3923,
+      "step": 7584
+    },
+    {
+      "epoch": 0.1805436541940398,
+      "grad_norm": 2.2899843179698482,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 7585
+    },
+    {
+      "epoch": 0.18056745691707132,
+      "grad_norm": 7.505723124290598,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 7586
+    },
+    {
+      "epoch": 0.18059125964010284,
+      "grad_norm": 1.8296565212878488,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 7587
+    },
+    {
+      "epoch": 0.18061506236313435,
+      "grad_norm": 12.326812001421956,
+      "learning_rate": 2e-05,
+      "loss": 0.9449,
+      "step": 7588
+    },
+    {
+      "epoch": 0.18063886508616586,
+      "grad_norm": 1.9584397428657818,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 7589
+    },
+    {
+      "epoch": 0.18066266780919737,
+      "grad_norm": 9.691485212185887,
+      "learning_rate": 2e-05,
+      "loss": 0.3922,
+      "step": 7590
+    },
+    {
+      "epoch": 0.1806864705322289,
+      "grad_norm": 4.324438704536878,
+      "learning_rate": 2e-05,
+      "loss": 0.2874,
+      "step": 7591
+    },
+    {
+      "epoch": 0.1807102732552604,
+      "grad_norm": 3.9673796865842488,
+      "learning_rate": 2e-05,
+      "loss": 0.358,
+      "step": 7592
+    },
+    {
+      "epoch": 0.1807340759782919,
+      "grad_norm": 1.7876170934465583,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 7593
+    },
+    {
+      "epoch": 0.18075787870132343,
+      "grad_norm": 7.0298038722182445,
+      "learning_rate": 2e-05,
+      "loss": 0.3572,
+      "step": 7594
+    },
+    {
+      "epoch": 0.18078168142435494,
+      "grad_norm": 6.664496959489777,
+      "learning_rate": 2e-05,
+      "loss": 0.743,
+      "step": 7595
+    },
+    {
+      "epoch": 0.18080548414738645,
+      "grad_norm": 6.126807182060622,
+      "learning_rate": 2e-05,
+      "loss": 0.1943,
+      "step": 7596
+    },
+    {
+      "epoch": 0.18082928687041797,
+      "grad_norm": 5.033639158065725,
+      "learning_rate": 2e-05,
+      "loss": 0.2978,
+      "step": 7597
+    },
+    {
+      "epoch": 0.18085308959344948,
+      "grad_norm": 6.6954918314495036,
+      "learning_rate": 2e-05,
+      "loss": 1.0978,
+      "step": 7598
+    },
+    {
+      "epoch": 0.180876892316481,
+      "grad_norm": 1.8896906477726054,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 7599
+    },
+    {
+      "epoch": 0.18090069503951253,
+      "grad_norm": 15.010212472961022,
+      "learning_rate": 2e-05,
+      "loss": 0.1462,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18092449776254405,
+      "grad_norm": 4.670149583445492,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 7601
+    },
+    {
+      "epoch": 0.18094830048557556,
+      "grad_norm": 5.818150760107948,
+      "learning_rate": 2e-05,
+      "loss": 0.6146,
+      "step": 7602
+    },
+    {
+      "epoch": 0.18097210320860707,
+      "grad_norm": 5.766448367379064,
+      "learning_rate": 2e-05,
+      "loss": 0.7003,
+      "step": 7603
+    },
+    {
+      "epoch": 0.18099590593163858,
+      "grad_norm": 1.5779479100340381,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 7604
+    },
+    {
+      "epoch": 0.1810197086546701,
+      "grad_norm": 2.0344657102713124,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 7605
+    },
+    {
+      "epoch": 0.1810435113777016,
+      "grad_norm": 2.465632700660533,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 7606
+    },
+    {
+      "epoch": 0.18106731410073312,
+      "grad_norm": 11.334053014345448,
+      "learning_rate": 2e-05,
+      "loss": 0.4834,
+      "step": 7607
+    },
+    {
+      "epoch": 0.18109111682376464,
+      "grad_norm": 1.4868859187875871,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 7608
+    },
+    {
+      "epoch": 0.18111491954679615,
+      "grad_norm": 4.199057570280446,
+      "learning_rate": 2e-05,
+      "loss": 0.7436,
+      "step": 7609
+    },
+    {
+      "epoch": 0.18113872226982766,
+      "grad_norm": 11.782991564986979,
+      "learning_rate": 2e-05,
+      "loss": 1.8724,
+      "step": 7610
+    },
+    {
+      "epoch": 0.18116252499285918,
+      "grad_norm": 5.701729817033718,
+      "learning_rate": 2e-05,
+      "loss": 0.3698,
+      "step": 7611
+    },
+    {
+      "epoch": 0.1811863277158907,
+      "grad_norm": 4.674738928787264,
+      "learning_rate": 2e-05,
+      "loss": 0.6827,
+      "step": 7612
+    },
+    {
+      "epoch": 0.1812101304389222,
+      "grad_norm": 3.56894716522909,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 7613
+    },
+    {
+      "epoch": 0.18123393316195371,
+      "grad_norm": 6.654027415269988,
+      "learning_rate": 2e-05,
+      "loss": 0.9441,
+      "step": 7614
+    },
+    {
+      "epoch": 0.18125773588498525,
+      "grad_norm": 3.046349419496832,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 7615
+    },
+    {
+      "epoch": 0.18128153860801677,
+      "grad_norm": 4.993436891416328,
+      "learning_rate": 2e-05,
+      "loss": 0.2976,
+      "step": 7616
+    },
+    {
+      "epoch": 0.18130534133104828,
+      "grad_norm": 12.196386289484332,
+      "learning_rate": 2e-05,
+      "loss": 1.4618,
+      "step": 7617
+    },
+    {
+      "epoch": 0.1813291440540798,
+      "grad_norm": 2.9640445291181106,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 7618
+    },
+    {
+      "epoch": 0.1813529467771113,
+      "grad_norm": 6.911419314559103,
+      "learning_rate": 2e-05,
+      "loss": 0.3106,
+      "step": 7619
+    },
+    {
+      "epoch": 0.18137674950014282,
+      "grad_norm": 6.221751301387031,
+      "learning_rate": 2e-05,
+      "loss": 0.3367,
+      "step": 7620
+    },
+    {
+      "epoch": 0.18140055222317433,
+      "grad_norm": 4.120780889013574,
+      "learning_rate": 2e-05,
+      "loss": 0.3662,
+      "step": 7621
+    },
+    {
+      "epoch": 0.18142435494620585,
+      "grad_norm": 6.893607287058467,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 7622
+    },
+    {
+      "epoch": 0.18144815766923736,
+      "grad_norm": 3.1347103067539215,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 7623
+    },
+    {
+      "epoch": 0.18147196039226887,
+      "grad_norm": 4.781122176836823,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 7624
+    },
+    {
+      "epoch": 0.18149576311530038,
+      "grad_norm": 11.58493093824369,
+      "learning_rate": 2e-05,
+      "loss": 1.2816,
+      "step": 7625
+    },
+    {
+      "epoch": 0.1815195658383319,
+      "grad_norm": 3.7148748969027108,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 7626
+    },
+    {
+      "epoch": 0.1815433685613634,
+      "grad_norm": 8.730341212817025,
+      "learning_rate": 2e-05,
+      "loss": 0.7207,
+      "step": 7627
+    },
+    {
+      "epoch": 0.18156717128439492,
+      "grad_norm": 5.411232787820713,
+      "learning_rate": 2e-05,
+      "loss": 0.5504,
+      "step": 7628
+    },
+    {
+      "epoch": 0.18159097400742644,
+      "grad_norm": 4.802321242090314,
+      "learning_rate": 2e-05,
+      "loss": 0.6527,
+      "step": 7629
+    },
+    {
+      "epoch": 0.18161477673045798,
+      "grad_norm": 4.933228238991328,
+      "learning_rate": 2e-05,
+      "loss": 0.3274,
+      "step": 7630
+    },
+    {
+      "epoch": 0.1816385794534895,
+      "grad_norm": 4.207732543440836,
+      "learning_rate": 2e-05,
+      "loss": 0.1875,
+      "step": 7631
+    },
+    {
+      "epoch": 0.181662382176521,
+      "grad_norm": 3.9967002270332874,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 7632
+    },
+    {
+      "epoch": 0.18168618489955252,
+      "grad_norm": 10.465833018484556,
+      "learning_rate": 2e-05,
+      "loss": 0.5015,
+      "step": 7633
+    },
+    {
+      "epoch": 0.18170998762258403,
+      "grad_norm": 5.886855268703543,
+      "learning_rate": 2e-05,
+      "loss": 0.9301,
+      "step": 7634
+    },
+    {
+      "epoch": 0.18173379034561554,
+      "grad_norm": 2.0604550221184574,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 7635
+    },
+    {
+      "epoch": 0.18175759306864706,
+      "grad_norm": 4.5023369986924395,
+      "learning_rate": 2e-05,
+      "loss": 0.2945,
+      "step": 7636
+    },
+    {
+      "epoch": 0.18178139579167857,
+      "grad_norm": 1.6868532763819246,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 7637
+    },
+    {
+      "epoch": 0.18180519851471008,
+      "grad_norm": 2.219874590573611,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 7638
+    },
+    {
+      "epoch": 0.1818290012377416,
+      "grad_norm": 4.4808306153973945,
+      "learning_rate": 2e-05,
+      "loss": 0.5747,
+      "step": 7639
+    },
+    {
+      "epoch": 0.1818528039607731,
+      "grad_norm": 2.1463202194716122,
+      "learning_rate": 2e-05,
+      "loss": 0.1498,
+      "step": 7640
+    },
+    {
+      "epoch": 0.18187660668380462,
+      "grad_norm": 6.856666542700157,
+      "learning_rate": 2e-05,
+      "loss": 0.719,
+      "step": 7641
+    },
+    {
+      "epoch": 0.18190040940683613,
+      "grad_norm": 3.8258842673141866,
+      "learning_rate": 2e-05,
+      "loss": 0.2151,
+      "step": 7642
+    },
+    {
+      "epoch": 0.18192421212986765,
+      "grad_norm": 6.396528307811796,
+      "learning_rate": 2e-05,
+      "loss": 0.2872,
+      "step": 7643
+    },
+    {
+      "epoch": 0.18194801485289916,
+      "grad_norm": 2.1004676580792143,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 7644
+    },
+    {
+      "epoch": 0.1819718175759307,
+      "grad_norm": 3.4168801214493807,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 7645
+    },
+    {
+      "epoch": 0.1819956202989622,
+      "grad_norm": 3.53703355999878,
+      "learning_rate": 2e-05,
+      "loss": 0.2003,
+      "step": 7646
+    },
+    {
+      "epoch": 0.18201942302199373,
+      "grad_norm": 3.571555549797237,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 7647
+    },
+    {
+      "epoch": 0.18204322574502524,
+      "grad_norm": 3.7587206545893546,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 7648
+    },
+    {
+      "epoch": 0.18206702846805675,
+      "grad_norm": 4.029298245872579,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 7649
+    },
+    {
+      "epoch": 0.18209083119108826,
+      "grad_norm": 6.722676035534525,
+      "learning_rate": 2e-05,
+      "loss": 0.3509,
+      "step": 7650
+    },
+    {
+      "epoch": 0.18211463391411978,
+      "grad_norm": 3.672539886596665,
+      "learning_rate": 2e-05,
+      "loss": 0.2268,
+      "step": 7651
+    },
+    {
+      "epoch": 0.1821384366371513,
+      "grad_norm": 5.9067034802122,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 7652
+    },
+    {
+      "epoch": 0.1821622393601828,
+      "grad_norm": 4.349900578322507,
+      "learning_rate": 2e-05,
+      "loss": 0.558,
+      "step": 7653
+    },
+    {
+      "epoch": 0.18218604208321432,
+      "grad_norm": 9.498095208460258,
+      "learning_rate": 2e-05,
+      "loss": 0.9028,
+      "step": 7654
+    },
+    {
+      "epoch": 0.18220984480624583,
+      "grad_norm": 17.77289417128967,
+      "learning_rate": 2e-05,
+      "loss": 0.75,
+      "step": 7655
+    },
+    {
+      "epoch": 0.18223364752927734,
+      "grad_norm": 4.933552519518277,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 7656
+    },
+    {
+      "epoch": 0.18225745025230886,
+      "grad_norm": 5.110708709112771,
+      "learning_rate": 2e-05,
+      "loss": 0.983,
+      "step": 7657
+    },
+    {
+      "epoch": 0.18228125297534037,
+      "grad_norm": 3.8840027991418653,
+      "learning_rate": 2e-05,
+      "loss": 0.2717,
+      "step": 7658
+    },
+    {
+      "epoch": 0.18230505569837188,
+      "grad_norm": 5.327677903781476,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 7659
+    },
+    {
+      "epoch": 0.1823288584214034,
+      "grad_norm": 3.980464041785276,
+      "learning_rate": 2e-05,
+      "loss": 0.5843,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18235266114443494,
+      "grad_norm": 9.743868874153403,
+      "learning_rate": 2e-05,
+      "loss": 0.892,
+      "step": 7661
+    },
+    {
+      "epoch": 0.18237646386746645,
+      "grad_norm": 2.3191314470711224,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 7662
+    },
+    {
+      "epoch": 0.18240026659049796,
+      "grad_norm": 7.746238313330466,
+      "learning_rate": 2e-05,
+      "loss": 0.6514,
+      "step": 7663
+    },
+    {
+      "epoch": 0.18242406931352947,
+      "grad_norm": 2.736063140464679,
+      "learning_rate": 2e-05,
+      "loss": 0.1964,
+      "step": 7664
+    },
+    {
+      "epoch": 0.182447872036561,
+      "grad_norm": 3.837533358907921,
+      "learning_rate": 2e-05,
+      "loss": 0.6272,
+      "step": 7665
+    },
+    {
+      "epoch": 0.1824716747595925,
+      "grad_norm": 5.474283096841873,
+      "learning_rate": 2e-05,
+      "loss": 0.6881,
+      "step": 7666
+    },
+    {
+      "epoch": 0.182495477482624,
+      "grad_norm": 3.7429040906618063,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 7667
+    },
+    {
+      "epoch": 0.18251928020565553,
+      "grad_norm": 4.782777442721602,
+      "learning_rate": 2e-05,
+      "loss": 0.742,
+      "step": 7668
+    },
+    {
+      "epoch": 0.18254308292868704,
+      "grad_norm": 6.916840431454583,
+      "learning_rate": 2e-05,
+      "loss": 1.1824,
+      "step": 7669
+    },
+    {
+      "epoch": 0.18256688565171855,
+      "grad_norm": 5.552152402800925,
+      "learning_rate": 2e-05,
+      "loss": 1.534,
+      "step": 7670
+    },
+    {
+      "epoch": 0.18259068837475007,
+      "grad_norm": 8.185163216216802,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 7671
+    },
+    {
+      "epoch": 0.18261449109778158,
+      "grad_norm": 2.2133110285198647,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 7672
+    },
+    {
+      "epoch": 0.1826382938208131,
+      "grad_norm": 8.94493368007437,
+      "learning_rate": 2e-05,
+      "loss": 0.3278,
+      "step": 7673
+    },
+    {
+      "epoch": 0.1826620965438446,
+      "grad_norm": 6.824441388499045,
+      "learning_rate": 2e-05,
+      "loss": 0.2694,
+      "step": 7674
+    },
+    {
+      "epoch": 0.18268589926687612,
+      "grad_norm": 4.630811836820136,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 7675
+    },
+    {
+      "epoch": 0.18270970198990766,
+      "grad_norm": 2.5838503991074058,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 7676
+    },
+    {
+      "epoch": 0.18273350471293917,
+      "grad_norm": 6.360071919354664,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 7677
+    },
+    {
+      "epoch": 0.18275730743597068,
+      "grad_norm": 5.077409432860251,
+      "learning_rate": 2e-05,
+      "loss": 0.1667,
+      "step": 7678
+    },
+    {
+      "epoch": 0.1827811101590022,
+      "grad_norm": 6.400701261269266,
+      "learning_rate": 2e-05,
+      "loss": 1.1049,
+      "step": 7679
+    },
+    {
+      "epoch": 0.1828049128820337,
+      "grad_norm": 5.838633910502178,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 7680
+    },
+    {
+      "epoch": 0.18282871560506522,
+      "grad_norm": 1.598978150027399,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 7681
+    },
+    {
+      "epoch": 0.18285251832809674,
+      "grad_norm": 8.55491067073201,
+      "learning_rate": 2e-05,
+      "loss": 0.3571,
+      "step": 7682
+    },
+    {
+      "epoch": 0.18287632105112825,
+      "grad_norm": 10.90983073896588,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 7683
+    },
+    {
+      "epoch": 0.18290012377415976,
+      "grad_norm": 7.215135030902756,
+      "learning_rate": 2e-05,
+      "loss": 0.6087,
+      "step": 7684
+    },
+    {
+      "epoch": 0.18292392649719127,
+      "grad_norm": 8.271360640777889,
+      "learning_rate": 2e-05,
+      "loss": 0.7757,
+      "step": 7685
+    },
+    {
+      "epoch": 0.1829477292202228,
+      "grad_norm": 8.284447737455855,
+      "learning_rate": 2e-05,
+      "loss": 0.4074,
+      "step": 7686
+    },
+    {
+      "epoch": 0.1829715319432543,
+      "grad_norm": 15.234771206862918,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 7687
+    },
+    {
+      "epoch": 0.1829953346662858,
+      "grad_norm": 2.851391517589848,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 7688
+    },
+    {
+      "epoch": 0.18301913738931733,
+      "grad_norm": 5.0801974777167995,
+      "learning_rate": 2e-05,
+      "loss": 0.7837,
+      "step": 7689
+    },
+    {
+      "epoch": 0.18304294011234884,
+      "grad_norm": 3.9235145137456375,
+      "learning_rate": 2e-05,
+      "loss": 0.3098,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18306674283538038,
+      "grad_norm": 8.576375426789118,
+      "learning_rate": 2e-05,
+      "loss": 1.1085,
+      "step": 7691
+    },
+    {
+      "epoch": 0.1830905455584119,
+      "grad_norm": 2.5688122132983278,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 7692
+    },
+    {
+      "epoch": 0.1831143482814434,
+      "grad_norm": 1.9218072180990995,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 7693
+    },
+    {
+      "epoch": 0.18313815100447492,
+      "grad_norm": 1.6364024293991435,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 7694
+    },
+    {
+      "epoch": 0.18316195372750643,
+      "grad_norm": 2.1407456648013214,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 7695
+    },
+    {
+      "epoch": 0.18318575645053795,
+      "grad_norm": 5.913142929928924,
+      "learning_rate": 2e-05,
+      "loss": 0.2606,
+      "step": 7696
+    },
+    {
+      "epoch": 0.18320955917356946,
+      "grad_norm": 8.663250656226978,
+      "learning_rate": 2e-05,
+      "loss": 0.6507,
+      "step": 7697
+    },
+    {
+      "epoch": 0.18323336189660097,
+      "grad_norm": 4.306562636041668,
+      "learning_rate": 2e-05,
+      "loss": 0.2592,
+      "step": 7698
+    },
+    {
+      "epoch": 0.18325716461963248,
+      "grad_norm": 5.606583379532727,
+      "learning_rate": 2e-05,
+      "loss": 0.3857,
+      "step": 7699
+    },
+    {
+      "epoch": 0.183280967342664,
+      "grad_norm": 8.332687585623754,
+      "learning_rate": 2e-05,
+      "loss": 1.4425,
+      "step": 7700
+    },
+    {
+      "epoch": 0.1833047700656955,
+      "grad_norm": 5.701586952864795,
+      "learning_rate": 2e-05,
+      "loss": 0.8685,
+      "step": 7701
+    },
+    {
+      "epoch": 0.18332857278872702,
+      "grad_norm": 19.068025289949723,
+      "learning_rate": 2e-05,
+      "loss": 1.5236,
+      "step": 7702
+    },
+    {
+      "epoch": 0.18335237551175854,
+      "grad_norm": 3.1745940061416635,
+      "learning_rate": 2e-05,
+      "loss": 0.282,
+      "step": 7703
+    },
+    {
+      "epoch": 0.18337617823479005,
+      "grad_norm": 9.329309385852078,
+      "learning_rate": 2e-05,
+      "loss": 0.6985,
+      "step": 7704
+    },
+    {
+      "epoch": 0.18339998095782156,
+      "grad_norm": 4.235954324605894,
+      "learning_rate": 2e-05,
+      "loss": 0.6918,
+      "step": 7705
+    },
+    {
+      "epoch": 0.1834237836808531,
+      "grad_norm": 5.442616939770885,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 7706
+    },
+    {
+      "epoch": 0.18344758640388462,
+      "grad_norm": 6.974942897937832,
+      "learning_rate": 2e-05,
+      "loss": 0.3808,
+      "step": 7707
+    },
+    {
+      "epoch": 0.18347138912691613,
+      "grad_norm": 5.5260925508062675,
+      "learning_rate": 2e-05,
+      "loss": 0.5755,
+      "step": 7708
+    },
+    {
+      "epoch": 0.18349519184994764,
+      "grad_norm": 12.066177353330543,
+      "learning_rate": 2e-05,
+      "loss": 0.5637,
+      "step": 7709
+    },
+    {
+      "epoch": 0.18351899457297915,
+      "grad_norm": 10.317242750604349,
+      "learning_rate": 2e-05,
+      "loss": 0.4525,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18354279729601067,
+      "grad_norm": 5.128126779768542,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 7711
+    },
+    {
+      "epoch": 0.18356660001904218,
+      "grad_norm": 6.306920424993985,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 7712
+    },
+    {
+      "epoch": 0.1835904027420737,
+      "grad_norm": 6.479178685162702,
+      "learning_rate": 2e-05,
+      "loss": 0.4042,
+      "step": 7713
+    },
+    {
+      "epoch": 0.1836142054651052,
+      "grad_norm": 3.897189532778633,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 7714
+    },
+    {
+      "epoch": 0.18363800818813672,
+      "grad_norm": 1.3720254452033986,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 7715
+    },
+    {
+      "epoch": 0.18366181091116823,
+      "grad_norm": 7.607682243664782,
+      "learning_rate": 2e-05,
+      "loss": 0.7326,
+      "step": 7716
+    },
+    {
+      "epoch": 0.18368561363419975,
+      "grad_norm": 4.014118944308007,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 7717
+    },
+    {
+      "epoch": 0.18370941635723126,
+      "grad_norm": 6.754670348073364,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 7718
+    },
+    {
+      "epoch": 0.18373321908026277,
+      "grad_norm": 0.6571899387622492,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 7719
+    },
+    {
+      "epoch": 0.18375702180329428,
+      "grad_norm": 1.0263050188588443,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 7720
+    },
+    {
+      "epoch": 0.1837808245263258,
+      "grad_norm": 4.952201666321481,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 7721
+    },
+    {
+      "epoch": 0.18380462724935734,
+      "grad_norm": 0.863627265811376,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 7722
+    },
+    {
+      "epoch": 0.18382842997238885,
+      "grad_norm": 5.142597402618179,
+      "learning_rate": 2e-05,
+      "loss": 0.1916,
+      "step": 7723
+    },
+    {
+      "epoch": 0.18385223269542036,
+      "grad_norm": 18.89572264760685,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 7724
+    },
+    {
+      "epoch": 0.18387603541845188,
+      "grad_norm": 4.433841105646019,
+      "learning_rate": 2e-05,
+      "loss": 0.6235,
+      "step": 7725
+    },
+    {
+      "epoch": 0.1838998381414834,
+      "grad_norm": 5.134333276648182,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 7726
+    },
+    {
+      "epoch": 0.1839236408645149,
+      "grad_norm": 2.4766076247106357,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 7727
+    },
+    {
+      "epoch": 0.18394744358754642,
+      "grad_norm": 5.125360017312899,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 7728
+    },
+    {
+      "epoch": 0.18397124631057793,
+      "grad_norm": 11.070190676895056,
+      "learning_rate": 2e-05,
+      "loss": 2.4905,
+      "step": 7729
+    },
+    {
+      "epoch": 0.18399504903360944,
+      "grad_norm": 2.0852301761718532,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18401885175664096,
+      "grad_norm": 9.38473101683413,
+      "learning_rate": 2e-05,
+      "loss": 0.4061,
+      "step": 7731
+    },
+    {
+      "epoch": 0.18404265447967247,
+      "grad_norm": 7.170140907934974,
+      "learning_rate": 2e-05,
+      "loss": 0.868,
+      "step": 7732
+    },
+    {
+      "epoch": 0.18406645720270398,
+      "grad_norm": 5.558744335548106,
+      "learning_rate": 2e-05,
+      "loss": 0.2618,
+      "step": 7733
+    },
+    {
+      "epoch": 0.1840902599257355,
+      "grad_norm": 11.590719873256358,
+      "learning_rate": 2e-05,
+      "loss": 1.0461,
+      "step": 7734
+    },
+    {
+      "epoch": 0.184114062648767,
+      "grad_norm": 4.916679220428235,
+      "learning_rate": 2e-05,
+      "loss": 0.6156,
+      "step": 7735
+    },
+    {
+      "epoch": 0.18413786537179852,
+      "grad_norm": 5.910289301959194,
+      "learning_rate": 2e-05,
+      "loss": 0.8969,
+      "step": 7736
+    },
+    {
+      "epoch": 0.18416166809483006,
+      "grad_norm": 5.887893860678678,
+      "learning_rate": 2e-05,
+      "loss": 0.3905,
+      "step": 7737
+    },
+    {
+      "epoch": 0.18418547081786157,
+      "grad_norm": 5.813528110291901,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 7738
+    },
+    {
+      "epoch": 0.1842092735408931,
+      "grad_norm": 1.9688853617611775,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 7739
+    },
+    {
+      "epoch": 0.1842330762639246,
+      "grad_norm": 2.4302279599129983,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 7740
+    },
+    {
+      "epoch": 0.1842568789869561,
+      "grad_norm": 2.997426681557821,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 7741
+    },
+    {
+      "epoch": 0.18428068170998763,
+      "grad_norm": 3.656871962600397,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 7742
+    },
+    {
+      "epoch": 0.18430448443301914,
+      "grad_norm": 10.09050803788867,
+      "learning_rate": 2e-05,
+      "loss": 0.3779,
+      "step": 7743
+    },
+    {
+      "epoch": 0.18432828715605065,
+      "grad_norm": 1.985582956165571,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 7744
+    },
+    {
+      "epoch": 0.18435208987908216,
+      "grad_norm": 9.567563120323603,
+      "learning_rate": 2e-05,
+      "loss": 0.563,
+      "step": 7745
+    },
+    {
+      "epoch": 0.18437589260211368,
+      "grad_norm": 6.437196192593403,
+      "learning_rate": 2e-05,
+      "loss": 1.0114,
+      "step": 7746
+    },
+    {
+      "epoch": 0.1843996953251452,
+      "grad_norm": 2.8429424447564124,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 7747
+    },
+    {
+      "epoch": 0.1844234980481767,
+      "grad_norm": 5.195705057026693,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 7748
+    },
+    {
+      "epoch": 0.18444730077120822,
+      "grad_norm": 6.776434010180423,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 7749
+    },
+    {
+      "epoch": 0.18447110349423973,
+      "grad_norm": 2.2369062808070894,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18449490621727124,
+      "grad_norm": 2.889032473593571,
+      "learning_rate": 2e-05,
+      "loss": 0.3962,
+      "step": 7751
+    },
+    {
+      "epoch": 0.18451870894030278,
+      "grad_norm": 8.397542249210744,
+      "learning_rate": 2e-05,
+      "loss": 1.0484,
+      "step": 7752
+    },
+    {
+      "epoch": 0.1845425116633343,
+      "grad_norm": 4.661155320118666,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 7753
+    },
+    {
+      "epoch": 0.1845663143863658,
+      "grad_norm": 5.677945106614249,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 7754
+    },
+    {
+      "epoch": 0.18459011710939732,
+      "grad_norm": 6.26792786839461,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 7755
+    },
+    {
+      "epoch": 0.18461391983242884,
+      "grad_norm": 3.1297454179768556,
+      "learning_rate": 2e-05,
+      "loss": 0.2354,
+      "step": 7756
+    },
+    {
+      "epoch": 0.18463772255546035,
+      "grad_norm": 5.908324091130429,
+      "learning_rate": 2e-05,
+      "loss": 0.331,
+      "step": 7757
+    },
+    {
+      "epoch": 0.18466152527849186,
+      "grad_norm": 11.499640450744776,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 7758
+    },
+    {
+      "epoch": 0.18468532800152337,
+      "grad_norm": 9.38110098575402,
+      "learning_rate": 2e-05,
+      "loss": 0.2667,
+      "step": 7759
+    },
+    {
+      "epoch": 0.1847091307245549,
+      "grad_norm": 3.6913132663942325,
+      "learning_rate": 2e-05,
+      "loss": 0.5276,
+      "step": 7760
+    },
+    {
+      "epoch": 0.1847329334475864,
+      "grad_norm": 5.599096325806889,
+      "learning_rate": 2e-05,
+      "loss": 0.7619,
+      "step": 7761
+    },
+    {
+      "epoch": 0.1847567361706179,
+      "grad_norm": 2.334662971614181,
+      "learning_rate": 2e-05,
+      "loss": 0.245,
+      "step": 7762
+    },
+    {
+      "epoch": 0.18478053889364943,
+      "grad_norm": 3.1715292922056717,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 7763
+    },
+    {
+      "epoch": 0.18480434161668094,
+      "grad_norm": 7.2673096972823545,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 7764
+    },
+    {
+      "epoch": 0.18482814433971245,
+      "grad_norm": 8.043636339638205,
+      "learning_rate": 2e-05,
+      "loss": 1.0856,
+      "step": 7765
+    },
+    {
+      "epoch": 0.18485194706274397,
+      "grad_norm": 2.451523428872356,
+      "learning_rate": 2e-05,
+      "loss": 0.1161,
+      "step": 7766
+    },
+    {
+      "epoch": 0.1848757497857755,
+      "grad_norm": 3.7703436772240124,
+      "learning_rate": 2e-05,
+      "loss": 0.2481,
+      "step": 7767
+    },
+    {
+      "epoch": 0.18489955250880702,
+      "grad_norm": 1.2962849283056406,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 7768
+    },
+    {
+      "epoch": 0.18492335523183853,
+      "grad_norm": 4.183852165006461,
+      "learning_rate": 2e-05,
+      "loss": 0.7999,
+      "step": 7769
+    },
+    {
+      "epoch": 0.18494715795487005,
+      "grad_norm": 2.346086707128165,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 7770
+    },
+    {
+      "epoch": 0.18497096067790156,
+      "grad_norm": 3.427211856604487,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 7771
+    },
+    {
+      "epoch": 0.18499476340093307,
+      "grad_norm": 5.4631520540251435,
+      "learning_rate": 2e-05,
+      "loss": 0.6778,
+      "step": 7772
+    },
+    {
+      "epoch": 0.18501856612396458,
+      "grad_norm": 2.3792584985054783,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 7773
+    },
+    {
+      "epoch": 0.1850423688469961,
+      "grad_norm": 4.877081733795375,
+      "learning_rate": 2e-05,
+      "loss": 0.6622,
+      "step": 7774
+    },
+    {
+      "epoch": 0.1850661715700276,
+      "grad_norm": 22.749047865496276,
+      "learning_rate": 2e-05,
+      "loss": 0.6971,
+      "step": 7775
+    },
+    {
+      "epoch": 0.18508997429305912,
+      "grad_norm": 3.0237136131255617,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 7776
+    },
+    {
+      "epoch": 0.18511377701609064,
+      "grad_norm": 7.334092720618438,
+      "learning_rate": 2e-05,
+      "loss": 1.545,
+      "step": 7777
+    },
+    {
+      "epoch": 0.18513757973912215,
+      "grad_norm": 1.504667067898081,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 7778
+    },
+    {
+      "epoch": 0.18516138246215366,
+      "grad_norm": 3.452160437438669,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 7779
+    },
+    {
+      "epoch": 0.18518518518518517,
+      "grad_norm": 4.385631288180579,
+      "learning_rate": 2e-05,
+      "loss": 0.6845,
+      "step": 7780
+    },
+    {
+      "epoch": 0.1852089879082167,
+      "grad_norm": 3.723743372547748,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 7781
+    },
+    {
+      "epoch": 0.18523279063124823,
+      "grad_norm": 4.871275675447934,
+      "learning_rate": 2e-05,
+      "loss": 0.9057,
+      "step": 7782
+    },
+    {
+      "epoch": 0.18525659335427974,
+      "grad_norm": 8.4000936262394,
+      "learning_rate": 2e-05,
+      "loss": 0.2252,
+      "step": 7783
+    },
+    {
+      "epoch": 0.18528039607731125,
+      "grad_norm": 7.9597302002434365,
+      "learning_rate": 2e-05,
+      "loss": 0.753,
+      "step": 7784
+    },
+    {
+      "epoch": 0.18530419880034277,
+      "grad_norm": 3.8723439075574477,
+      "learning_rate": 2e-05,
+      "loss": 0.2684,
+      "step": 7785
+    },
+    {
+      "epoch": 0.18532800152337428,
+      "grad_norm": 8.919948157093284,
+      "learning_rate": 2e-05,
+      "loss": 0.2408,
+      "step": 7786
+    },
+    {
+      "epoch": 0.1853518042464058,
+      "grad_norm": 8.201046857931157,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 7787
+    },
+    {
+      "epoch": 0.1853756069694373,
+      "grad_norm": 4.873969547256841,
+      "learning_rate": 2e-05,
+      "loss": 0.1827,
+      "step": 7788
+    },
+    {
+      "epoch": 0.18539940969246882,
+      "grad_norm": 8.048616197063877,
+      "learning_rate": 2e-05,
+      "loss": 1.3211,
+      "step": 7789
+    },
+    {
+      "epoch": 0.18542321241550033,
+      "grad_norm": 35.216551115588146,
+      "learning_rate": 2e-05,
+      "loss": 0.5882,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18544701513853185,
+      "grad_norm": 15.43445875634598,
+      "learning_rate": 2e-05,
+      "loss": 0.25,
+      "step": 7791
+    },
+    {
+      "epoch": 0.18547081786156336,
+      "grad_norm": 5.204503794320768,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 7792
+    },
+    {
+      "epoch": 0.18549462058459487,
+      "grad_norm": 6.837959957677924,
+      "learning_rate": 2e-05,
+      "loss": 0.7945,
+      "step": 7793
+    },
+    {
+      "epoch": 0.18551842330762638,
+      "grad_norm": 3.6458568322495406,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 7794
+    },
+    {
+      "epoch": 0.1855422260306579,
+      "grad_norm": 7.676872963788263,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 7795
+    },
+    {
+      "epoch": 0.1855660287536894,
+      "grad_norm": 3.344014155206404,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 7796
+    },
+    {
+      "epoch": 0.18558983147672092,
+      "grad_norm": 2.276845032162672,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 7797
+    },
+    {
+      "epoch": 0.18561363419975246,
+      "grad_norm": 17.632670459124647,
+      "learning_rate": 2e-05,
+      "loss": 1.0408,
+      "step": 7798
+    },
+    {
+      "epoch": 0.18563743692278398,
+      "grad_norm": 1.4042259584168122,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 7799
+    },
+    {
+      "epoch": 0.1856612396458155,
+      "grad_norm": 5.005133081739277,
+      "learning_rate": 2e-05,
+      "loss": 0.831,
+      "step": 7800
+    },
+    {
+      "epoch": 0.185685042368847,
+      "grad_norm": 5.622589587287399,
+      "learning_rate": 2e-05,
+      "loss": 0.9686,
+      "step": 7801
+    },
+    {
+      "epoch": 0.18570884509187852,
+      "grad_norm": 3.953277751773898,
+      "learning_rate": 2e-05,
+      "loss": 0.2425,
+      "step": 7802
+    },
+    {
+      "epoch": 0.18573264781491003,
+      "grad_norm": 3.9373009774826686,
+      "learning_rate": 2e-05,
+      "loss": 0.1856,
+      "step": 7803
+    },
+    {
+      "epoch": 0.18575645053794154,
+      "grad_norm": 15.758696828542417,
+      "learning_rate": 2e-05,
+      "loss": 2.1745,
+      "step": 7804
+    },
+    {
+      "epoch": 0.18578025326097306,
+      "grad_norm": 1.2363221357994687,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 7805
+    },
+    {
+      "epoch": 0.18580405598400457,
+      "grad_norm": 7.803933811616833,
+      "learning_rate": 2e-05,
+      "loss": 0.3515,
+      "step": 7806
+    },
+    {
+      "epoch": 0.18582785870703608,
+      "grad_norm": 5.709745501385654,
+      "learning_rate": 2e-05,
+      "loss": 0.9047,
+      "step": 7807
+    },
+    {
+      "epoch": 0.1858516614300676,
+      "grad_norm": 11.665227708379675,
+      "learning_rate": 2e-05,
+      "loss": 0.6082,
+      "step": 7808
+    },
+    {
+      "epoch": 0.1858754641530991,
+      "grad_norm": 2.1107617088265047,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 7809
+    },
+    {
+      "epoch": 0.18589926687613062,
+      "grad_norm": 1.279996161476007,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 7810
+    },
+    {
+      "epoch": 0.18592306959916213,
+      "grad_norm": 6.5291906196827405,
+      "learning_rate": 2e-05,
+      "loss": 1.2881,
+      "step": 7811
+    },
+    {
+      "epoch": 0.18594687232219365,
+      "grad_norm": 4.040838556414923,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 7812
+    },
+    {
+      "epoch": 0.1859706750452252,
+      "grad_norm": 4.602980550329949,
+      "learning_rate": 2e-05,
+      "loss": 0.1819,
+      "step": 7813
+    },
+    {
+      "epoch": 0.1859944777682567,
+      "grad_norm": 4.0318486721950695,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 7814
+    },
+    {
+      "epoch": 0.1860182804912882,
+      "grad_norm": 4.825892937020017,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 7815
+    },
+    {
+      "epoch": 0.18604208321431973,
+      "grad_norm": 5.796851129027722,
+      "learning_rate": 2e-05,
+      "loss": 0.5703,
+      "step": 7816
+    },
+    {
+      "epoch": 0.18606588593735124,
+      "grad_norm": 6.2822176812199215,
+      "learning_rate": 2e-05,
+      "loss": 0.2892,
+      "step": 7817
+    },
+    {
+      "epoch": 0.18608968866038275,
+      "grad_norm": 6.692021583911007,
+      "learning_rate": 2e-05,
+      "loss": 0.1882,
+      "step": 7818
+    },
+    {
+      "epoch": 0.18611349138341426,
+      "grad_norm": 1.1347616219717593,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 7819
+    },
+    {
+      "epoch": 0.18613729410644578,
+      "grad_norm": 8.249805706985624,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 7820
+    },
+    {
+      "epoch": 0.1861610968294773,
+      "grad_norm": 2.4210069869893363,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 7821
+    },
+    {
+      "epoch": 0.1861848995525088,
+      "grad_norm": 8.707372157052585,
+      "learning_rate": 2e-05,
+      "loss": 0.3607,
+      "step": 7822
+    },
+    {
+      "epoch": 0.18620870227554032,
+      "grad_norm": 2.453264837300352,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 7823
+    },
+    {
+      "epoch": 0.18623250499857183,
+      "grad_norm": 6.271642932796718,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 7824
+    },
+    {
+      "epoch": 0.18625630772160334,
+      "grad_norm": 1.721502866434684,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 7825
+    },
+    {
+      "epoch": 0.18628011044463486,
+      "grad_norm": 2.8244347044427323,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 7826
+    },
+    {
+      "epoch": 0.18630391316766637,
+      "grad_norm": 0.973360384148104,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 7827
+    },
+    {
+      "epoch": 0.1863277158906979,
+      "grad_norm": 1.0843111145661009,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 7828
+    },
+    {
+      "epoch": 0.18635151861372942,
+      "grad_norm": 5.198406497826371,
+      "learning_rate": 2e-05,
+      "loss": 0.8531,
+      "step": 7829
+    },
+    {
+      "epoch": 0.18637532133676094,
+      "grad_norm": 4.262864027445727,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 7830
+    },
+    {
+      "epoch": 0.18639912405979245,
+      "grad_norm": 1.6426883695586807,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 7831
+    },
+    {
+      "epoch": 0.18642292678282396,
+      "grad_norm": 5.276337450231596,
+      "learning_rate": 2e-05,
+      "loss": 0.7285,
+      "step": 7832
+    },
+    {
+      "epoch": 0.18644672950585547,
+      "grad_norm": 5.748161865697951,
+      "learning_rate": 2e-05,
+      "loss": 0.9207,
+      "step": 7833
+    },
+    {
+      "epoch": 0.186470532228887,
+      "grad_norm": 4.847570180368297,
+      "learning_rate": 2e-05,
+      "loss": 0.8713,
+      "step": 7834
+    },
+    {
+      "epoch": 0.1864943349519185,
+      "grad_norm": 5.7157454675690325,
+      "learning_rate": 2e-05,
+      "loss": 0.7835,
+      "step": 7835
+    },
+    {
+      "epoch": 0.18651813767495,
+      "grad_norm": 1.660118495534357,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 7836
+    },
+    {
+      "epoch": 0.18654194039798153,
+      "grad_norm": 10.679680764043239,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 7837
+    },
+    {
+      "epoch": 0.18656574312101304,
+      "grad_norm": 6.525077737475017,
+      "learning_rate": 2e-05,
+      "loss": 0.506,
+      "step": 7838
+    },
+    {
+      "epoch": 0.18658954584404455,
+      "grad_norm": 4.405249643388437,
+      "learning_rate": 2e-05,
+      "loss": 0.7501,
+      "step": 7839
+    },
+    {
+      "epoch": 0.18661334856707606,
+      "grad_norm": 7.652149355389675,
+      "learning_rate": 2e-05,
+      "loss": 1.1327,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18663715129010758,
+      "grad_norm": 14.416046445472261,
+      "learning_rate": 2e-05,
+      "loss": 1.1399,
+      "step": 7841
+    },
+    {
+      "epoch": 0.1866609540131391,
+      "grad_norm": 6.220620927838767,
+      "learning_rate": 2e-05,
+      "loss": 1.1266,
+      "step": 7842
+    },
+    {
+      "epoch": 0.18668475673617063,
+      "grad_norm": 3.0828434989112368,
+      "learning_rate": 2e-05,
+      "loss": 0.3593,
+      "step": 7843
+    },
+    {
+      "epoch": 0.18670855945920214,
+      "grad_norm": 6.742492023059215,
+      "learning_rate": 2e-05,
+      "loss": 0.8853,
+      "step": 7844
+    },
+    {
+      "epoch": 0.18673236218223366,
+      "grad_norm": 4.503203475560011,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 7845
+    },
+    {
+      "epoch": 0.18675616490526517,
+      "grad_norm": 4.908137665256274,
+      "learning_rate": 2e-05,
+      "loss": 0.8756,
+      "step": 7846
+    },
+    {
+      "epoch": 0.18677996762829668,
+      "grad_norm": 4.088334924225412,
+      "learning_rate": 2e-05,
+      "loss": 0.3609,
+      "step": 7847
+    },
+    {
+      "epoch": 0.1868037703513282,
+      "grad_norm": 6.844753787901827,
+      "learning_rate": 2e-05,
+      "loss": 1.0745,
+      "step": 7848
+    },
+    {
+      "epoch": 0.1868275730743597,
+      "grad_norm": 3.094840507672431,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 7849
+    },
+    {
+      "epoch": 0.18685137579739122,
+      "grad_norm": 8.465138420673906,
+      "learning_rate": 2e-05,
+      "loss": 1.0071,
+      "step": 7850
+    },
+    {
+      "epoch": 0.18687517852042274,
+      "grad_norm": 3.333935731681205,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 7851
+    },
+    {
+      "epoch": 0.18689898124345425,
+      "grad_norm": 2.4201998867384673,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 7852
+    },
+    {
+      "epoch": 0.18692278396648576,
+      "grad_norm": 7.872790104099792,
+      "learning_rate": 2e-05,
+      "loss": 0.9632,
+      "step": 7853
+    },
+    {
+      "epoch": 0.18694658668951727,
+      "grad_norm": 9.238355891926577,
+      "learning_rate": 2e-05,
+      "loss": 0.4115,
+      "step": 7854
+    },
+    {
+      "epoch": 0.1869703894125488,
+      "grad_norm": 2.566856120348519,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 7855
+    },
+    {
+      "epoch": 0.1869941921355803,
+      "grad_norm": 6.817072580255418,
+      "learning_rate": 2e-05,
+      "loss": 0.9279,
+      "step": 7856
+    },
+    {
+      "epoch": 0.1870179948586118,
+      "grad_norm": 3.6885444450032865,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 7857
+    },
+    {
+      "epoch": 0.18704179758164333,
+      "grad_norm": 6.264147393705672,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 7858
+    },
+    {
+      "epoch": 0.18706560030467487,
+      "grad_norm": 1.65111777043928,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 7859
+    },
+    {
+      "epoch": 0.18708940302770638,
+      "grad_norm": 15.118992359322725,
+      "learning_rate": 2e-05,
+      "loss": 0.506,
+      "step": 7860
+    },
+    {
+      "epoch": 0.1871132057507379,
+      "grad_norm": 1.9148240512038366,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 7861
+    },
+    {
+      "epoch": 0.1871370084737694,
+      "grad_norm": 1.8262066608385128,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 7862
+    },
+    {
+      "epoch": 0.18716081119680092,
+      "grad_norm": 4.358367020392766,
+      "learning_rate": 2e-05,
+      "loss": 0.6942,
+      "step": 7863
+    },
+    {
+      "epoch": 0.18718461391983243,
+      "grad_norm": 2.685269621306914,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 7864
+    },
+    {
+      "epoch": 0.18720841664286395,
+      "grad_norm": 5.199622563513828,
+      "learning_rate": 2e-05,
+      "loss": 0.7177,
+      "step": 7865
+    },
+    {
+      "epoch": 0.18723221936589546,
+      "grad_norm": 4.307600826016461,
+      "learning_rate": 2e-05,
+      "loss": 0.4843,
+      "step": 7866
+    },
+    {
+      "epoch": 0.18725602208892697,
+      "grad_norm": 2.984655517548256,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 7867
+    },
+    {
+      "epoch": 0.18727982481195848,
+      "grad_norm": 3.731926539706104,
+      "learning_rate": 2e-05,
+      "loss": 0.3038,
+      "step": 7868
+    },
+    {
+      "epoch": 0.18730362753499,
+      "grad_norm": 33.84765568786331,
+      "learning_rate": 2e-05,
+      "loss": 0.937,
+      "step": 7869
+    },
+    {
+      "epoch": 0.1873274302580215,
+      "grad_norm": 9.644934483432964,
+      "learning_rate": 2e-05,
+      "loss": 0.4386,
+      "step": 7870
+    },
+    {
+      "epoch": 0.18735123298105302,
+      "grad_norm": 4.4531727851234635,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 7871
+    },
+    {
+      "epoch": 0.18737503570408454,
+      "grad_norm": 3.011866671897737,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 7872
+    },
+    {
+      "epoch": 0.18739883842711605,
+      "grad_norm": 18.94247577189604,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 7873
+    },
+    {
+      "epoch": 0.1874226411501476,
+      "grad_norm": 2.761707862313509,
+      "learning_rate": 2e-05,
+      "loss": 0.2068,
+      "step": 7874
+    },
+    {
+      "epoch": 0.1874464438731791,
+      "grad_norm": 4.264430388738006,
+      "learning_rate": 2e-05,
+      "loss": 0.4831,
+      "step": 7875
+    },
+    {
+      "epoch": 0.18747024659621062,
+      "grad_norm": 16.93925169046529,
+      "learning_rate": 2e-05,
+      "loss": 0.9533,
+      "step": 7876
+    },
+    {
+      "epoch": 0.18749404931924213,
+      "grad_norm": 4.7527279422901465,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 7877
+    },
+    {
+      "epoch": 0.18751785204227364,
+      "grad_norm": 5.844275524778388,
+      "learning_rate": 2e-05,
+      "loss": 0.7654,
+      "step": 7878
+    },
+    {
+      "epoch": 0.18754165476530515,
+      "grad_norm": 3.5742006889557745,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 7879
+    },
+    {
+      "epoch": 0.18756545748833667,
+      "grad_norm": 5.15919106759155,
+      "learning_rate": 2e-05,
+      "loss": 0.7429,
+      "step": 7880
+    },
+    {
+      "epoch": 0.18758926021136818,
+      "grad_norm": 2.8336088422904404,
+      "learning_rate": 2e-05,
+      "loss": 0.3292,
+      "step": 7881
+    },
+    {
+      "epoch": 0.1876130629343997,
+      "grad_norm": 6.667855130341896,
+      "learning_rate": 2e-05,
+      "loss": 0.7629,
+      "step": 7882
+    },
+    {
+      "epoch": 0.1876368656574312,
+      "grad_norm": 8.374846836961737,
+      "learning_rate": 2e-05,
+      "loss": 1.3282,
+      "step": 7883
+    },
+    {
+      "epoch": 0.18766066838046272,
+      "grad_norm": 2.4142173937031126,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 7884
+    },
+    {
+      "epoch": 0.18768447110349423,
+      "grad_norm": 4.276686689173817,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 7885
+    },
+    {
+      "epoch": 0.18770827382652575,
+      "grad_norm": 9.124034996967822,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 7886
+    },
+    {
+      "epoch": 0.18773207654955726,
+      "grad_norm": 4.5927728345304075,
+      "learning_rate": 2e-05,
+      "loss": 0.329,
+      "step": 7887
+    },
+    {
+      "epoch": 0.18775587927258877,
+      "grad_norm": 7.100778521456624,
+      "learning_rate": 2e-05,
+      "loss": 0.279,
+      "step": 7888
+    },
+    {
+      "epoch": 0.1877796819956203,
+      "grad_norm": 9.424199256712491,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 7889
+    },
+    {
+      "epoch": 0.18780348471865183,
+      "grad_norm": 3.983094430588235,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 7890
+    },
+    {
+      "epoch": 0.18782728744168334,
+      "grad_norm": 7.561966147373221,
+      "learning_rate": 2e-05,
+      "loss": 1.3095,
+      "step": 7891
+    },
+    {
+      "epoch": 0.18785109016471485,
+      "grad_norm": 1.970981831846618,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 7892
+    },
+    {
+      "epoch": 0.18787489288774636,
+      "grad_norm": 2.591483470337366,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 7893
+    },
+    {
+      "epoch": 0.18789869561077788,
+      "grad_norm": 6.119412306514257,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 7894
+    },
+    {
+      "epoch": 0.1879224983338094,
+      "grad_norm": 0.9634098509843927,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 7895
+    },
+    {
+      "epoch": 0.1879463010568409,
+      "grad_norm": 4.349943621186806,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 7896
+    },
+    {
+      "epoch": 0.18797010377987242,
+      "grad_norm": 1.8238692704870108,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 7897
+    },
+    {
+      "epoch": 0.18799390650290393,
+      "grad_norm": 4.6277147390222435,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 7898
+    },
+    {
+      "epoch": 0.18801770922593544,
+      "grad_norm": 1.9541314294506724,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 7899
+    },
+    {
+      "epoch": 0.18804151194896696,
+      "grad_norm": 2.5117813803391034,
+      "learning_rate": 2e-05,
+      "loss": 0.1815,
+      "step": 7900
+    },
+    {
+      "epoch": 0.18806531467199847,
+      "grad_norm": 14.613737847188224,
+      "learning_rate": 2e-05,
+      "loss": 1.4507,
+      "step": 7901
+    },
+    {
+      "epoch": 0.18808911739502998,
+      "grad_norm": 9.499772805952219,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 7902
+    },
+    {
+      "epoch": 0.1881129201180615,
+      "grad_norm": 8.929948533752663,
+      "learning_rate": 2e-05,
+      "loss": 1.4915,
+      "step": 7903
+    },
+    {
+      "epoch": 0.18813672284109303,
+      "grad_norm": 5.832335807982901,
+      "learning_rate": 2e-05,
+      "loss": 0.6316,
+      "step": 7904
+    },
+    {
+      "epoch": 0.18816052556412455,
+      "grad_norm": 13.808912747538336,
+      "learning_rate": 2e-05,
+      "loss": 0.9327,
+      "step": 7905
+    },
+    {
+      "epoch": 0.18818432828715606,
+      "grad_norm": 11.093698145089741,
+      "learning_rate": 2e-05,
+      "loss": 0.6394,
+      "step": 7906
+    },
+    {
+      "epoch": 0.18820813101018757,
+      "grad_norm": 11.669730779018547,
+      "learning_rate": 2e-05,
+      "loss": 1.5639,
+      "step": 7907
+    },
+    {
+      "epoch": 0.1882319337332191,
+      "grad_norm": 6.091822379072962,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 7908
+    },
+    {
+      "epoch": 0.1882557364562506,
+      "grad_norm": 1.3206329748178294,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 7909
+    },
+    {
+      "epoch": 0.1882795391792821,
+      "grad_norm": 3.026223590890704,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 7910
+    },
+    {
+      "epoch": 0.18830334190231363,
+      "grad_norm": 9.540052204063507,
+      "learning_rate": 2e-05,
+      "loss": 0.3787,
+      "step": 7911
+    },
+    {
+      "epoch": 0.18832714462534514,
+      "grad_norm": 10.391227947400616,
+      "learning_rate": 2e-05,
+      "loss": 0.9904,
+      "step": 7912
+    },
+    {
+      "epoch": 0.18835094734837665,
+      "grad_norm": 7.175572865371838,
+      "learning_rate": 2e-05,
+      "loss": 0.2545,
+      "step": 7913
+    },
+    {
+      "epoch": 0.18837475007140816,
+      "grad_norm": 5.558397620058704,
+      "learning_rate": 2e-05,
+      "loss": 0.6391,
+      "step": 7914
+    },
+    {
+      "epoch": 0.18839855279443968,
+      "grad_norm": 5.471738451115883,
+      "learning_rate": 2e-05,
+      "loss": 0.4889,
+      "step": 7915
+    },
+    {
+      "epoch": 0.1884223555174712,
+      "grad_norm": 5.240797010890667,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 7916
+    },
+    {
+      "epoch": 0.1884461582405027,
+      "grad_norm": 3.9152039402228227,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 7917
+    },
+    {
+      "epoch": 0.18846996096353422,
+      "grad_norm": 2.474998179887157,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 7918
+    },
+    {
+      "epoch": 0.18849376368656573,
+      "grad_norm": 9.38561010438298,
+      "learning_rate": 2e-05,
+      "loss": 0.3981,
+      "step": 7919
+    },
+    {
+      "epoch": 0.18851756640959727,
+      "grad_norm": 3.816537065496498,
+      "learning_rate": 2e-05,
+      "loss": 0.3379,
+      "step": 7920
+    },
+    {
+      "epoch": 0.18854136913262878,
+      "grad_norm": 7.011621855473579,
+      "learning_rate": 2e-05,
+      "loss": 1.4464,
+      "step": 7921
+    },
+    {
+      "epoch": 0.1885651718556603,
+      "grad_norm": 12.523669249710144,
+      "learning_rate": 2e-05,
+      "loss": 1.2735,
+      "step": 7922
+    },
+    {
+      "epoch": 0.1885889745786918,
+      "grad_norm": 14.418794235895515,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 7923
+    },
+    {
+      "epoch": 0.18861277730172332,
+      "grad_norm": 9.840662157852314,
+      "learning_rate": 2e-05,
+      "loss": 0.6248,
+      "step": 7924
+    },
+    {
+      "epoch": 0.18863658002475484,
+      "grad_norm": 7.467701309797743,
+      "learning_rate": 2e-05,
+      "loss": 0.7819,
+      "step": 7925
+    },
+    {
+      "epoch": 0.18866038274778635,
+      "grad_norm": 9.122873407661348,
+      "learning_rate": 2e-05,
+      "loss": 0.251,
+      "step": 7926
+    },
+    {
+      "epoch": 0.18868418547081786,
+      "grad_norm": 3.8668934496163,
+      "learning_rate": 2e-05,
+      "loss": 0.2063,
+      "step": 7927
+    },
+    {
+      "epoch": 0.18870798819384937,
+      "grad_norm": 24.7573224070373,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 7928
+    },
+    {
+      "epoch": 0.1887317909168809,
+      "grad_norm": 1.9256911797107548,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 7929
+    },
+    {
+      "epoch": 0.1887555936399124,
+      "grad_norm": 19.688030760161517,
+      "learning_rate": 2e-05,
+      "loss": 0.7444,
+      "step": 7930
+    },
+    {
+      "epoch": 0.1887793963629439,
+      "grad_norm": 2.584059833230659,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 7931
+    },
+    {
+      "epoch": 0.18880319908597543,
+      "grad_norm": 2.469942658019229,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 7932
+    },
+    {
+      "epoch": 0.18882700180900694,
+      "grad_norm": 5.050233626328461,
+      "learning_rate": 2e-05,
+      "loss": 0.767,
+      "step": 7933
+    },
+    {
+      "epoch": 0.18885080453203845,
+      "grad_norm": 1.5656687831933787,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 7934
+    },
+    {
+      "epoch": 0.18887460725507,
+      "grad_norm": 8.738667361307957,
+      "learning_rate": 2e-05,
+      "loss": 1.1632,
+      "step": 7935
+    },
+    {
+      "epoch": 0.1888984099781015,
+      "grad_norm": 11.605801450091983,
+      "learning_rate": 2e-05,
+      "loss": 1.2825,
+      "step": 7936
+    },
+    {
+      "epoch": 0.18892221270113302,
+      "grad_norm": 0.9738433284121566,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 7937
+    },
+    {
+      "epoch": 0.18894601542416453,
+      "grad_norm": 5.803674159349993,
+      "learning_rate": 2e-05,
+      "loss": 0.1904,
+      "step": 7938
+    },
+    {
+      "epoch": 0.18896981814719604,
+      "grad_norm": 7.751322109572779,
+      "learning_rate": 2e-05,
+      "loss": 0.9685,
+      "step": 7939
+    },
+    {
+      "epoch": 0.18899362087022756,
+      "grad_norm": 6.13956219817453,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 7940
+    },
+    {
+      "epoch": 0.18901742359325907,
+      "grad_norm": 4.995208079140174,
+      "learning_rate": 2e-05,
+      "loss": 0.8766,
+      "step": 7941
+    },
+    {
+      "epoch": 0.18904122631629058,
+      "grad_norm": 5.501627474109915,
+      "learning_rate": 2e-05,
+      "loss": 0.3091,
+      "step": 7942
+    },
+    {
+      "epoch": 0.1890650290393221,
+      "grad_norm": 6.007496226079622,
+      "learning_rate": 2e-05,
+      "loss": 0.7112,
+      "step": 7943
+    },
+    {
+      "epoch": 0.1890888317623536,
+      "grad_norm": 1.8994048293798347,
+      "learning_rate": 2e-05,
+      "loss": 0.1053,
+      "step": 7944
+    },
+    {
+      "epoch": 0.18911263448538512,
+      "grad_norm": 9.19064114156342,
+      "learning_rate": 2e-05,
+      "loss": 1.4253,
+      "step": 7945
+    },
+    {
+      "epoch": 0.18913643720841664,
+      "grad_norm": 5.703840063700642,
+      "learning_rate": 2e-05,
+      "loss": 0.7896,
+      "step": 7946
+    },
+    {
+      "epoch": 0.18916023993144815,
+      "grad_norm": 4.119015060952482,
+      "learning_rate": 2e-05,
+      "loss": 0.3065,
+      "step": 7947
+    },
+    {
+      "epoch": 0.18918404265447966,
+      "grad_norm": 3.847592540713204,
+      "learning_rate": 2e-05,
+      "loss": 0.2611,
+      "step": 7948
+    },
+    {
+      "epoch": 0.18920784537751117,
+      "grad_norm": 6.604790803759424,
+      "learning_rate": 2e-05,
+      "loss": 0.4089,
+      "step": 7949
+    },
+    {
+      "epoch": 0.18923164810054272,
+      "grad_norm": 15.029537095982853,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 7950
+    },
+    {
+      "epoch": 0.18925545082357423,
+      "grad_norm": 7.113807426850842,
+      "learning_rate": 2e-05,
+      "loss": 1.3397,
+      "step": 7951
+    },
+    {
+      "epoch": 0.18927925354660574,
+      "grad_norm": 4.528726278257381,
+      "learning_rate": 2e-05,
+      "loss": 0.6528,
+      "step": 7952
+    },
+    {
+      "epoch": 0.18930305626963725,
+      "grad_norm": 4.854794205496692,
+      "learning_rate": 2e-05,
+      "loss": 0.2479,
+      "step": 7953
+    },
+    {
+      "epoch": 0.18932685899266877,
+      "grad_norm": 6.837038580746713,
+      "learning_rate": 2e-05,
+      "loss": 1.0797,
+      "step": 7954
+    },
+    {
+      "epoch": 0.18935066171570028,
+      "grad_norm": 8.519482451216414,
+      "learning_rate": 2e-05,
+      "loss": 1.1029,
+      "step": 7955
+    },
+    {
+      "epoch": 0.1893744644387318,
+      "grad_norm": 6.069317007971926,
+      "learning_rate": 2e-05,
+      "loss": 0.7873,
+      "step": 7956
+    },
+    {
+      "epoch": 0.1893982671617633,
+      "grad_norm": 12.02893940856901,
+      "learning_rate": 2e-05,
+      "loss": 0.3469,
+      "step": 7957
+    },
+    {
+      "epoch": 0.18942206988479482,
+      "grad_norm": 4.589112410601893,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 7958
+    },
+    {
+      "epoch": 0.18944587260782633,
+      "grad_norm": 16.211078385887443,
+      "learning_rate": 2e-05,
+      "loss": 0.3971,
+      "step": 7959
+    },
+    {
+      "epoch": 0.18946967533085785,
+      "grad_norm": 7.882566067245421,
+      "learning_rate": 2e-05,
+      "loss": 0.7874,
+      "step": 7960
+    },
+    {
+      "epoch": 0.18949347805388936,
+      "grad_norm": 3.826040867666295,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 7961
+    },
+    {
+      "epoch": 0.18951728077692087,
+      "grad_norm": 4.003044111671301,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 7962
+    },
+    {
+      "epoch": 0.18954108349995238,
+      "grad_norm": 6.026934646192707,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 7963
+    },
+    {
+      "epoch": 0.1895648862229839,
+      "grad_norm": 3.025862168622743,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 7964
+    },
+    {
+      "epoch": 0.18958868894601544,
+      "grad_norm": 7.868225319067987,
+      "learning_rate": 2e-05,
+      "loss": 1.6261,
+      "step": 7965
+    },
+    {
+      "epoch": 0.18961249166904695,
+      "grad_norm": 2.937722953046634,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 7966
+    },
+    {
+      "epoch": 0.18963629439207846,
+      "grad_norm": 21.121095052400786,
+      "learning_rate": 2e-05,
+      "loss": 0.981,
+      "step": 7967
+    },
+    {
+      "epoch": 0.18966009711510998,
+      "grad_norm": 2.955878307488207,
+      "learning_rate": 2e-05,
+      "loss": 0.1838,
+      "step": 7968
+    },
+    {
+      "epoch": 0.1896838998381415,
+      "grad_norm": 3.9529052783842866,
+      "learning_rate": 2e-05,
+      "loss": 0.2002,
+      "step": 7969
+    },
+    {
+      "epoch": 0.189707702561173,
+      "grad_norm": 4.571034829904678,
+      "learning_rate": 2e-05,
+      "loss": 0.1902,
+      "step": 7970
+    },
+    {
+      "epoch": 0.18973150528420452,
+      "grad_norm": 4.0076282225095365,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 7971
+    },
+    {
+      "epoch": 0.18975530800723603,
+      "grad_norm": 2.743947006405304,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 7972
+    },
+    {
+      "epoch": 0.18977911073026754,
+      "grad_norm": 12.189564554585894,
+      "learning_rate": 2e-05,
+      "loss": 0.25,
+      "step": 7973
+    },
+    {
+      "epoch": 0.18980291345329905,
+      "grad_norm": 3.465512567757128,
+      "learning_rate": 2e-05,
+      "loss": 0.1673,
+      "step": 7974
+    },
+    {
+      "epoch": 0.18982671617633057,
+      "grad_norm": 1.9226821615548666,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 7975
+    },
+    {
+      "epoch": 0.18985051889936208,
+      "grad_norm": 4.486750026390434,
+      "learning_rate": 2e-05,
+      "loss": 0.6025,
+      "step": 7976
+    },
+    {
+      "epoch": 0.1898743216223936,
+      "grad_norm": 4.464675508613782,
+      "learning_rate": 2e-05,
+      "loss": 0.9636,
+      "step": 7977
+    },
+    {
+      "epoch": 0.1898981243454251,
+      "grad_norm": 2.4638401379788006,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 7978
+    },
+    {
+      "epoch": 0.18992192706845662,
+      "grad_norm": 11.198698065612684,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 7979
+    },
+    {
+      "epoch": 0.18994572979148813,
+      "grad_norm": 1.679912257277867,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 7980
+    },
+    {
+      "epoch": 0.18996953251451967,
+      "grad_norm": 2.3975648712328965,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 7981
+    },
+    {
+      "epoch": 0.1899933352375512,
+      "grad_norm": 10.502933428528772,
+      "learning_rate": 2e-05,
+      "loss": 0.3515,
+      "step": 7982
+    },
+    {
+      "epoch": 0.1900171379605827,
+      "grad_norm": 12.123916258602746,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 7983
+    },
+    {
+      "epoch": 0.1900409406836142,
+      "grad_norm": 1.621033276213818,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 7984
+    },
+    {
+      "epoch": 0.19006474340664573,
+      "grad_norm": 2.9140439269594207,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 7985
+    },
+    {
+      "epoch": 0.19008854612967724,
+      "grad_norm": 9.986539664019846,
+      "learning_rate": 2e-05,
+      "loss": 0.3899,
+      "step": 7986
+    },
+    {
+      "epoch": 0.19011234885270875,
+      "grad_norm": 2.549352526149301,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 7987
+    },
+    {
+      "epoch": 0.19013615157574026,
+      "grad_norm": 3.4238192931175644,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 7988
+    },
+    {
+      "epoch": 0.19015995429877178,
+      "grad_norm": 2.1824352114521055,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 7989
+    },
+    {
+      "epoch": 0.1901837570218033,
+      "grad_norm": 1.5978395396437477,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 7990
+    },
+    {
+      "epoch": 0.1902075597448348,
+      "grad_norm": 6.073574306799179,
+      "learning_rate": 2e-05,
+      "loss": 0.6708,
+      "step": 7991
+    },
+    {
+      "epoch": 0.19023136246786632,
+      "grad_norm": 5.354499931534632,
+      "learning_rate": 2e-05,
+      "loss": 0.8814,
+      "step": 7992
+    },
+    {
+      "epoch": 0.19025516519089783,
+      "grad_norm": 13.650973609527588,
+      "learning_rate": 2e-05,
+      "loss": 0.824,
+      "step": 7993
+    },
+    {
+      "epoch": 0.19027896791392934,
+      "grad_norm": 9.109765358066822,
+      "learning_rate": 2e-05,
+      "loss": 0.913,
+      "step": 7994
+    },
+    {
+      "epoch": 0.19030277063696086,
+      "grad_norm": 7.706589177951784,
+      "learning_rate": 2e-05,
+      "loss": 0.9327,
+      "step": 7995
+    },
+    {
+      "epoch": 0.1903265733599924,
+      "grad_norm": 4.1372354568501395,
+      "learning_rate": 2e-05,
+      "loss": 0.2861,
+      "step": 7996
+    },
+    {
+      "epoch": 0.1903503760830239,
+      "grad_norm": 4.618209352610745,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 7997
+    },
+    {
+      "epoch": 0.19037417880605542,
+      "grad_norm": 1.0688435909660445,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 7998
+    },
+    {
+      "epoch": 0.19039798152908693,
+      "grad_norm": 4.731783499663978,
+      "learning_rate": 2e-05,
+      "loss": 0.9129,
+      "step": 7999
+    },
+    {
+      "epoch": 0.19042178425211845,
+      "grad_norm": 6.298236121508159,
+      "learning_rate": 2e-05,
+      "loss": 0.8338,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19044558697514996,
+      "grad_norm": 3.586066922005124,
+      "learning_rate": 2e-05,
+      "loss": 0.4599,
+      "step": 8001
+    },
+    {
+      "epoch": 0.19046938969818147,
+      "grad_norm": 5.417657886392859,
+      "learning_rate": 2e-05,
+      "loss": 0.67,
+      "step": 8002
+    },
+    {
+      "epoch": 0.190493192421213,
+      "grad_norm": 5.146157075896047,
+      "learning_rate": 2e-05,
+      "loss": 0.638,
+      "step": 8003
+    },
+    {
+      "epoch": 0.1905169951442445,
+      "grad_norm": 3.170112557691012,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 8004
+    },
+    {
+      "epoch": 0.190540797867276,
+      "grad_norm": 6.846136168561237,
+      "learning_rate": 2e-05,
+      "loss": 1.5598,
+      "step": 8005
+    },
+    {
+      "epoch": 0.19056460059030753,
+      "grad_norm": 6.873685724459038,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 8006
+    },
+    {
+      "epoch": 0.19058840331333904,
+      "grad_norm": 14.60687860544445,
+      "learning_rate": 2e-05,
+      "loss": 1.7494,
+      "step": 8007
+    },
+    {
+      "epoch": 0.19061220603637055,
+      "grad_norm": 7.396912523410416,
+      "learning_rate": 2e-05,
+      "loss": 0.3069,
+      "step": 8008
+    },
+    {
+      "epoch": 0.19063600875940206,
+      "grad_norm": 4.189500416491957,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 8009
+    },
+    {
+      "epoch": 0.19065981148243358,
+      "grad_norm": 7.199780774084227,
+      "learning_rate": 2e-05,
+      "loss": 0.5051,
+      "step": 8010
+    },
+    {
+      "epoch": 0.19068361420546512,
+      "grad_norm": 9.064274737883094,
+      "learning_rate": 2e-05,
+      "loss": 0.2423,
+      "step": 8011
+    },
+    {
+      "epoch": 0.19070741692849663,
+      "grad_norm": 10.565262203469288,
+      "learning_rate": 2e-05,
+      "loss": 0.8182,
+      "step": 8012
+    },
+    {
+      "epoch": 0.19073121965152814,
+      "grad_norm": 11.119440959033927,
+      "learning_rate": 2e-05,
+      "loss": 0.6193,
+      "step": 8013
+    },
+    {
+      "epoch": 0.19075502237455966,
+      "grad_norm": 9.762624225140225,
+      "learning_rate": 2e-05,
+      "loss": 0.4812,
+      "step": 8014
+    },
+    {
+      "epoch": 0.19077882509759117,
+      "grad_norm": 7.161991787262991,
+      "learning_rate": 2e-05,
+      "loss": 0.2747,
+      "step": 8015
+    },
+    {
+      "epoch": 0.19080262782062268,
+      "grad_norm": 6.79549960945083,
+      "learning_rate": 2e-05,
+      "loss": 0.6806,
+      "step": 8016
+    },
+    {
+      "epoch": 0.1908264305436542,
+      "grad_norm": 2.165225889728844,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 8017
+    },
+    {
+      "epoch": 0.1908502332666857,
+      "grad_norm": 3.971709968097541,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 8018
+    },
+    {
+      "epoch": 0.19087403598971722,
+      "grad_norm": 5.106704886191452,
+      "learning_rate": 2e-05,
+      "loss": 0.6786,
+      "step": 8019
+    },
+    {
+      "epoch": 0.19089783871274874,
+      "grad_norm": 3.1583356042931356,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 8020
+    },
+    {
+      "epoch": 0.19092164143578025,
+      "grad_norm": 8.530845522240066,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 8021
+    },
+    {
+      "epoch": 0.19094544415881176,
+      "grad_norm": 1.6913889210251978,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 8022
+    },
+    {
+      "epoch": 0.19096924688184327,
+      "grad_norm": 4.307552708332872,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 8023
+    },
+    {
+      "epoch": 0.1909930496048748,
+      "grad_norm": 23.430646123994396,
+      "learning_rate": 2e-05,
+      "loss": 0.9393,
+      "step": 8024
+    },
+    {
+      "epoch": 0.1910168523279063,
+      "grad_norm": 2.317298242494726,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 8025
+    },
+    {
+      "epoch": 0.19104065505093784,
+      "grad_norm": 6.628842352178679,
+      "learning_rate": 2e-05,
+      "loss": 0.6263,
+      "step": 8026
+    },
+    {
+      "epoch": 0.19106445777396935,
+      "grad_norm": 9.530908866095144,
+      "learning_rate": 2e-05,
+      "loss": 1.2459,
+      "step": 8027
+    },
+    {
+      "epoch": 0.19108826049700087,
+      "grad_norm": 7.6589540208408176,
+      "learning_rate": 2e-05,
+      "loss": 1.7109,
+      "step": 8028
+    },
+    {
+      "epoch": 0.19111206322003238,
+      "grad_norm": 2.998257925047132,
+      "learning_rate": 2e-05,
+      "loss": 0.2042,
+      "step": 8029
+    },
+    {
+      "epoch": 0.1911358659430639,
+      "grad_norm": 2.425202119326164,
+      "learning_rate": 2e-05,
+      "loss": 0.1782,
+      "step": 8030
+    },
+    {
+      "epoch": 0.1911596686660954,
+      "grad_norm": 4.494984162715118,
+      "learning_rate": 2e-05,
+      "loss": 0.6582,
+      "step": 8031
+    },
+    {
+      "epoch": 0.19118347138912692,
+      "grad_norm": 13.466354625201763,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 8032
+    },
+    {
+      "epoch": 0.19120727411215843,
+      "grad_norm": 2.7180092438923573,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 8033
+    },
+    {
+      "epoch": 0.19123107683518994,
+      "grad_norm": 2.618433925747077,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 8034
+    },
+    {
+      "epoch": 0.19125487955822146,
+      "grad_norm": 3.606042186389332,
+      "learning_rate": 2e-05,
+      "loss": 0.7562,
+      "step": 8035
+    },
+    {
+      "epoch": 0.19127868228125297,
+      "grad_norm": 5.547804965761409,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 8036
+    },
+    {
+      "epoch": 0.19130248500428448,
+      "grad_norm": 9.812751261240576,
+      "learning_rate": 2e-05,
+      "loss": 0.8532,
+      "step": 8037
+    },
+    {
+      "epoch": 0.191326287727316,
+      "grad_norm": 3.6088637099752328,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 8038
+    },
+    {
+      "epoch": 0.1913500904503475,
+      "grad_norm": 3.0847383591632034,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 8039
+    },
+    {
+      "epoch": 0.19137389317337902,
+      "grad_norm": 3.8004325263546215,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 8040
+    },
+    {
+      "epoch": 0.19139769589641054,
+      "grad_norm": 8.934698649466558,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 8041
+    },
+    {
+      "epoch": 0.19142149861944208,
+      "grad_norm": 8.677252068905911,
+      "learning_rate": 2e-05,
+      "loss": 1.0156,
+      "step": 8042
+    },
+    {
+      "epoch": 0.1914453013424736,
+      "grad_norm": 4.607435777446774,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 8043
+    },
+    {
+      "epoch": 0.1914691040655051,
+      "grad_norm": 7.41493008648272,
+      "learning_rate": 2e-05,
+      "loss": 0.7106,
+      "step": 8044
+    },
+    {
+      "epoch": 0.19149290678853662,
+      "grad_norm": 5.606869819715633,
+      "learning_rate": 2e-05,
+      "loss": 0.8888,
+      "step": 8045
+    },
+    {
+      "epoch": 0.19151670951156813,
+      "grad_norm": 2.8772191633347184,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 8046
+    },
+    {
+      "epoch": 0.19154051223459964,
+      "grad_norm": 11.25587979536741,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 8047
+    },
+    {
+      "epoch": 0.19156431495763115,
+      "grad_norm": 6.64874833667526,
+      "learning_rate": 2e-05,
+      "loss": 0.7838,
+      "step": 8048
+    },
+    {
+      "epoch": 0.19158811768066267,
+      "grad_norm": 2.407282681217935,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 8049
+    },
+    {
+      "epoch": 0.19161192040369418,
+      "grad_norm": 1.750415104169338,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 8050
+    },
+    {
+      "epoch": 0.1916357231267257,
+      "grad_norm": 2.377870664903526,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 8051
+    },
+    {
+      "epoch": 0.1916595258497572,
+      "grad_norm": 20.43413629622806,
+      "learning_rate": 2e-05,
+      "loss": 0.9766,
+      "step": 8052
+    },
+    {
+      "epoch": 0.19168332857278872,
+      "grad_norm": 21.09768486404267,
+      "learning_rate": 2e-05,
+      "loss": 0.7397,
+      "step": 8053
+    },
+    {
+      "epoch": 0.19170713129582023,
+      "grad_norm": 4.98062780141805,
+      "learning_rate": 2e-05,
+      "loss": 0.6997,
+      "step": 8054
+    },
+    {
+      "epoch": 0.19173093401885175,
+      "grad_norm": 19.891071402274793,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 8055
+    },
+    {
+      "epoch": 0.19175473674188326,
+      "grad_norm": 2.904084670597153,
+      "learning_rate": 2e-05,
+      "loss": 0.2221,
+      "step": 8056
+    },
+    {
+      "epoch": 0.1917785394649148,
+      "grad_norm": 0.9701100689139411,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 8057
+    },
+    {
+      "epoch": 0.1918023421879463,
+      "grad_norm": 14.629314328832164,
+      "learning_rate": 2e-05,
+      "loss": 1.0081,
+      "step": 8058
+    },
+    {
+      "epoch": 0.19182614491097782,
+      "grad_norm": 14.560968052958433,
+      "learning_rate": 2e-05,
+      "loss": 0.3297,
+      "step": 8059
+    },
+    {
+      "epoch": 0.19184994763400934,
+      "grad_norm": 4.582571736258437,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 8060
+    },
+    {
+      "epoch": 0.19187375035704085,
+      "grad_norm": 8.635904216605187,
+      "learning_rate": 2e-05,
+      "loss": 0.4469,
+      "step": 8061
+    },
+    {
+      "epoch": 0.19189755308007236,
+      "grad_norm": 1.6912191523412592,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 8062
+    },
+    {
+      "epoch": 0.19192135580310388,
+      "grad_norm": 4.407910307042707,
+      "learning_rate": 2e-05,
+      "loss": 0.6285,
+      "step": 8063
+    },
+    {
+      "epoch": 0.1919451585261354,
+      "grad_norm": 12.622821656670176,
+      "learning_rate": 2e-05,
+      "loss": 0.2965,
+      "step": 8064
+    },
+    {
+      "epoch": 0.1919689612491669,
+      "grad_norm": 15.487939387381296,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 8065
+    },
+    {
+      "epoch": 0.19199276397219842,
+      "grad_norm": 5.121027262494147,
+      "learning_rate": 2e-05,
+      "loss": 0.6095,
+      "step": 8066
+    },
+    {
+      "epoch": 0.19201656669522993,
+      "grad_norm": 1.612095798961402,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 8067
+    },
+    {
+      "epoch": 0.19204036941826144,
+      "grad_norm": 6.465238030508315,
+      "learning_rate": 2e-05,
+      "loss": 0.7395,
+      "step": 8068
+    },
+    {
+      "epoch": 0.19206417214129295,
+      "grad_norm": 6.996199899152706,
+      "learning_rate": 2e-05,
+      "loss": 1.1876,
+      "step": 8069
+    },
+    {
+      "epoch": 0.19208797486432447,
+      "grad_norm": 5.414683129276133,
+      "learning_rate": 2e-05,
+      "loss": 0.668,
+      "step": 8070
+    },
+    {
+      "epoch": 0.19211177758735598,
+      "grad_norm": 10.03768182919971,
+      "learning_rate": 2e-05,
+      "loss": 0.555,
+      "step": 8071
+    },
+    {
+      "epoch": 0.19213558031038752,
+      "grad_norm": 10.54752139866734,
+      "learning_rate": 2e-05,
+      "loss": 0.2936,
+      "step": 8072
+    },
+    {
+      "epoch": 0.19215938303341903,
+      "grad_norm": 25.60250274952969,
+      "learning_rate": 2e-05,
+      "loss": 1.8618,
+      "step": 8073
+    },
+    {
+      "epoch": 0.19218318575645055,
+      "grad_norm": 5.260277462630005,
+      "learning_rate": 2e-05,
+      "loss": 0.2497,
+      "step": 8074
+    },
+    {
+      "epoch": 0.19220698847948206,
+      "grad_norm": 5.475114091284375,
+      "learning_rate": 2e-05,
+      "loss": 0.2154,
+      "step": 8075
+    },
+    {
+      "epoch": 0.19223079120251357,
+      "grad_norm": 6.830899582362227,
+      "learning_rate": 2e-05,
+      "loss": 0.6582,
+      "step": 8076
+    },
+    {
+      "epoch": 0.1922545939255451,
+      "grad_norm": 3.0299728126952066,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 8077
+    },
+    {
+      "epoch": 0.1922783966485766,
+      "grad_norm": 4.188902098627874,
+      "learning_rate": 2e-05,
+      "loss": 0.2063,
+      "step": 8078
+    },
+    {
+      "epoch": 0.1923021993716081,
+      "grad_norm": 8.954245008093453,
+      "learning_rate": 2e-05,
+      "loss": 0.8077,
+      "step": 8079
+    },
+    {
+      "epoch": 0.19232600209463963,
+      "grad_norm": 4.552708373950664,
+      "learning_rate": 2e-05,
+      "loss": 0.8367,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19234980481767114,
+      "grad_norm": 3.5144716064723323,
+      "learning_rate": 2e-05,
+      "loss": 0.3093,
+      "step": 8081
+    },
+    {
+      "epoch": 0.19237360754070265,
+      "grad_norm": 2.2861420855584518,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 8082
+    },
+    {
+      "epoch": 0.19239741026373416,
+      "grad_norm": 11.527058400871175,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 8083
+    },
+    {
+      "epoch": 0.19242121298676568,
+      "grad_norm": 11.4208125599766,
+      "learning_rate": 2e-05,
+      "loss": 1.1649,
+      "step": 8084
+    },
+    {
+      "epoch": 0.1924450157097972,
+      "grad_norm": 13.29014609274459,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 8085
+    },
+    {
+      "epoch": 0.1924688184328287,
+      "grad_norm": 9.835277191228165,
+      "learning_rate": 2e-05,
+      "loss": 0.8783,
+      "step": 8086
+    },
+    {
+      "epoch": 0.19249262115586024,
+      "grad_norm": 6.34346940998071,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 8087
+    },
+    {
+      "epoch": 0.19251642387889176,
+      "grad_norm": 2.609545675743286,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 8088
+    },
+    {
+      "epoch": 0.19254022660192327,
+      "grad_norm": 11.042722853764127,
+      "learning_rate": 2e-05,
+      "loss": 1.1061,
+      "step": 8089
+    },
+    {
+      "epoch": 0.19256402932495478,
+      "grad_norm": 3.1759734777707966,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 8090
+    },
+    {
+      "epoch": 0.1925878320479863,
+      "grad_norm": 6.3167652726791035,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 8091
+    },
+    {
+      "epoch": 0.1926116347710178,
+      "grad_norm": 5.213676164037381,
+      "learning_rate": 2e-05,
+      "loss": 0.5121,
+      "step": 8092
+    },
+    {
+      "epoch": 0.19263543749404932,
+      "grad_norm": 4.607876895061862,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 8093
+    },
+    {
+      "epoch": 0.19265924021708083,
+      "grad_norm": 2.828820107714285,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 8094
+    },
+    {
+      "epoch": 0.19268304294011235,
+      "grad_norm": 1.820318279160388,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 8095
+    },
+    {
+      "epoch": 0.19270684566314386,
+      "grad_norm": 5.824850607322332,
+      "learning_rate": 2e-05,
+      "loss": 0.9466,
+      "step": 8096
+    },
+    {
+      "epoch": 0.19273064838617537,
+      "grad_norm": 3.6117764059350623,
+      "learning_rate": 2e-05,
+      "loss": 0.1999,
+      "step": 8097
+    },
+    {
+      "epoch": 0.1927544511092069,
+      "grad_norm": 2.848143082802809,
+      "learning_rate": 2e-05,
+      "loss": 0.2309,
+      "step": 8098
+    },
+    {
+      "epoch": 0.1927782538322384,
+      "grad_norm": 13.24430743273579,
+      "learning_rate": 2e-05,
+      "loss": 0.4288,
+      "step": 8099
+    },
+    {
+      "epoch": 0.1928020565552699,
+      "grad_norm": 6.507142811845777,
+      "learning_rate": 2e-05,
+      "loss": 1.202,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19282585927830143,
+      "grad_norm": 3.9947492074400603,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 8101
+    },
+    {
+      "epoch": 0.19284966200133294,
+      "grad_norm": 3.8978730943114415,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 8102
+    },
+    {
+      "epoch": 0.19287346472436448,
+      "grad_norm": 16.555696797113278,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 8103
+    },
+    {
+      "epoch": 0.192897267447396,
+      "grad_norm": 23.23968754883892,
+      "learning_rate": 2e-05,
+      "loss": 1.1379,
+      "step": 8104
+    },
+    {
+      "epoch": 0.1929210701704275,
+      "grad_norm": 4.903806291986812,
+      "learning_rate": 2e-05,
+      "loss": 0.7486,
+      "step": 8105
+    },
+    {
+      "epoch": 0.19294487289345902,
+      "grad_norm": 4.936594540611641,
+      "learning_rate": 2e-05,
+      "loss": 0.1915,
+      "step": 8106
+    },
+    {
+      "epoch": 0.19296867561649053,
+      "grad_norm": 6.336166755762927,
+      "learning_rate": 2e-05,
+      "loss": 0.9321,
+      "step": 8107
+    },
+    {
+      "epoch": 0.19299247833952204,
+      "grad_norm": 17.25170414063332,
+      "learning_rate": 2e-05,
+      "loss": 0.5649,
+      "step": 8108
+    },
+    {
+      "epoch": 0.19301628106255356,
+      "grad_norm": 2.591344966419153,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 8109
+    },
+    {
+      "epoch": 0.19304008378558507,
+      "grad_norm": 1.3440414326754364,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 8110
+    },
+    {
+      "epoch": 0.19306388650861658,
+      "grad_norm": 6.929751336206018,
+      "learning_rate": 2e-05,
+      "loss": 1.6711,
+      "step": 8111
+    },
+    {
+      "epoch": 0.1930876892316481,
+      "grad_norm": 17.574059242471506,
+      "learning_rate": 2e-05,
+      "loss": 0.8762,
+      "step": 8112
+    },
+    {
+      "epoch": 0.1931114919546796,
+      "grad_norm": 3.636967307758046,
+      "learning_rate": 2e-05,
+      "loss": 0.1644,
+      "step": 8113
+    },
+    {
+      "epoch": 0.19313529467771112,
+      "grad_norm": 9.183842440325424,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 8114
+    },
+    {
+      "epoch": 0.19315909740074264,
+      "grad_norm": 3.668890478848789,
+      "learning_rate": 2e-05,
+      "loss": 0.5589,
+      "step": 8115
+    },
+    {
+      "epoch": 0.19318290012377415,
+      "grad_norm": 11.603877084468413,
+      "learning_rate": 2e-05,
+      "loss": 0.2496,
+      "step": 8116
+    },
+    {
+      "epoch": 0.19320670284680566,
+      "grad_norm": 4.639990098212741,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 8117
+    },
+    {
+      "epoch": 0.1932305055698372,
+      "grad_norm": 41.60079708611339,
+      "learning_rate": 2e-05,
+      "loss": 1.004,
+      "step": 8118
+    },
+    {
+      "epoch": 0.19325430829286872,
+      "grad_norm": 3.9495574284223194,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 8119
+    },
+    {
+      "epoch": 0.19327811101590023,
+      "grad_norm": 3.1948641814993795,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19330191373893174,
+      "grad_norm": 6.623851629765438,
+      "learning_rate": 2e-05,
+      "loss": 1.0346,
+      "step": 8121
+    },
+    {
+      "epoch": 0.19332571646196325,
+      "grad_norm": 21.596083161333,
+      "learning_rate": 2e-05,
+      "loss": 1.1843,
+      "step": 8122
+    },
+    {
+      "epoch": 0.19334951918499477,
+      "grad_norm": 4.323767479007234,
+      "learning_rate": 2e-05,
+      "loss": 0.7131,
+      "step": 8123
+    },
+    {
+      "epoch": 0.19337332190802628,
+      "grad_norm": 6.647897498711027,
+      "learning_rate": 2e-05,
+      "loss": 0.2808,
+      "step": 8124
+    },
+    {
+      "epoch": 0.1933971246310578,
+      "grad_norm": 5.654740006146388,
+      "learning_rate": 2e-05,
+      "loss": 0.8004,
+      "step": 8125
+    },
+    {
+      "epoch": 0.1934209273540893,
+      "grad_norm": 1.7086084282869576,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 8126
+    },
+    {
+      "epoch": 0.19344473007712082,
+      "grad_norm": 3.6228407529910838,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 8127
+    },
+    {
+      "epoch": 0.19346853280015233,
+      "grad_norm": 3.602868261694235,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 8128
+    },
+    {
+      "epoch": 0.19349233552318384,
+      "grad_norm": 60.283995516725874,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 8129
+    },
+    {
+      "epoch": 0.19351613824621536,
+      "grad_norm": 17.022684106824173,
+      "learning_rate": 2e-05,
+      "loss": 0.5438,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19353994096924687,
+      "grad_norm": 39.31257526585788,
+      "learning_rate": 2e-05,
+      "loss": 0.2715,
+      "step": 8131
+    },
+    {
+      "epoch": 0.19356374369227838,
+      "grad_norm": 2.585233992339765,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 8132
+    },
+    {
+      "epoch": 0.19358754641530992,
+      "grad_norm": 28.384741256887665,
+      "learning_rate": 2e-05,
+      "loss": 1.0821,
+      "step": 8133
+    },
+    {
+      "epoch": 0.19361134913834144,
+      "grad_norm": 24.56848350936907,
+      "learning_rate": 2e-05,
+      "loss": 0.1156,
+      "step": 8134
+    },
+    {
+      "epoch": 0.19363515186137295,
+      "grad_norm": 5.47120172064612,
+      "learning_rate": 2e-05,
+      "loss": 0.557,
+      "step": 8135
+    },
+    {
+      "epoch": 0.19365895458440446,
+      "grad_norm": 6.4440593731934985,
+      "learning_rate": 2e-05,
+      "loss": 1.0333,
+      "step": 8136
+    },
+    {
+      "epoch": 0.19368275730743598,
+      "grad_norm": 7.496979648477483,
+      "learning_rate": 2e-05,
+      "loss": 0.8099,
+      "step": 8137
+    },
+    {
+      "epoch": 0.1937065600304675,
+      "grad_norm": 5.027003058434383,
+      "learning_rate": 2e-05,
+      "loss": 0.2264,
+      "step": 8138
+    },
+    {
+      "epoch": 0.193730362753499,
+      "grad_norm": 2.085280363465202,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 8139
+    },
+    {
+      "epoch": 0.19375416547653052,
+      "grad_norm": 113.13226652569847,
+      "learning_rate": 2e-05,
+      "loss": 0.237,
+      "step": 8140
+    },
+    {
+      "epoch": 0.19377796819956203,
+      "grad_norm": 14.833410141501625,
+      "learning_rate": 2e-05,
+      "loss": 1.1524,
+      "step": 8141
+    },
+    {
+      "epoch": 0.19380177092259354,
+      "grad_norm": 64.11316199623566,
+      "learning_rate": 2e-05,
+      "loss": 0.235,
+      "step": 8142
+    },
+    {
+      "epoch": 0.19382557364562505,
+      "grad_norm": 4.78474247238643,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 8143
+    },
+    {
+      "epoch": 0.19384937636865657,
+      "grad_norm": 38.039673667701386,
+      "learning_rate": 2e-05,
+      "loss": 0.647,
+      "step": 8144
+    },
+    {
+      "epoch": 0.19387317909168808,
+      "grad_norm": 5.320169858145136,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 8145
+    },
+    {
+      "epoch": 0.1938969818147196,
+      "grad_norm": 2.305510504495167,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 8146
+    },
+    {
+      "epoch": 0.1939207845377511,
+      "grad_norm": 7.231956616185039,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 8147
+    },
+    {
+      "epoch": 0.19394458726078265,
+      "grad_norm": 5.49160961669394,
+      "learning_rate": 2e-05,
+      "loss": 0.6833,
+      "step": 8148
+    },
+    {
+      "epoch": 0.19396838998381416,
+      "grad_norm": 8.419011577843225,
+      "learning_rate": 2e-05,
+      "loss": 1.0499,
+      "step": 8149
+    },
+    {
+      "epoch": 0.19399219270684567,
+      "grad_norm": 14.302288626923355,
+      "learning_rate": 2e-05,
+      "loss": 0.5953,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19401599542987719,
+      "grad_norm": 89.08540346100547,
+      "learning_rate": 2e-05,
+      "loss": 1.0448,
+      "step": 8151
+    },
+    {
+      "epoch": 0.1940397981529087,
+      "grad_norm": 6.163002634477431,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 8152
+    },
+    {
+      "epoch": 0.1940636008759402,
+      "grad_norm": 8.959525334736679,
+      "learning_rate": 2e-05,
+      "loss": 0.4806,
+      "step": 8153
+    },
+    {
+      "epoch": 0.19408740359897173,
+      "grad_norm": 4.741953774895875,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 8154
+    },
+    {
+      "epoch": 0.19411120632200324,
+      "grad_norm": 7.40229880375114,
+      "learning_rate": 2e-05,
+      "loss": 0.909,
+      "step": 8155
+    },
+    {
+      "epoch": 0.19413500904503475,
+      "grad_norm": 6.1931076886699445,
+      "learning_rate": 2e-05,
+      "loss": 0.7996,
+      "step": 8156
+    },
+    {
+      "epoch": 0.19415881176806626,
+      "grad_norm": 1.9008213368396683,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 8157
+    },
+    {
+      "epoch": 0.19418261449109778,
+      "grad_norm": 4.679689061949088,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 8158
+    },
+    {
+      "epoch": 0.1942064172141293,
+      "grad_norm": 6.86797013114434,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 8159
+    },
+    {
+      "epoch": 0.1942302199371608,
+      "grad_norm": 4.0019809564767215,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 8160
+    },
+    {
+      "epoch": 0.19425402266019232,
+      "grad_norm": 3.745364433161317,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 8161
+    },
+    {
+      "epoch": 0.19427782538322383,
+      "grad_norm": 10.484446905259095,
+      "learning_rate": 2e-05,
+      "loss": 0.5464,
+      "step": 8162
+    },
+    {
+      "epoch": 0.19430162810625534,
+      "grad_norm": 92.35875431739638,
+      "learning_rate": 2e-05,
+      "loss": 0.3432,
+      "step": 8163
+    },
+    {
+      "epoch": 0.19432543082928688,
+      "grad_norm": 8.314449956057047,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 8164
+    },
+    {
+      "epoch": 0.1943492335523184,
+      "grad_norm": 7.435689638258095,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 8165
+    },
+    {
+      "epoch": 0.1943730362753499,
+      "grad_norm": 10.040092525846216,
+      "learning_rate": 2e-05,
+      "loss": 0.2678,
+      "step": 8166
+    },
+    {
+      "epoch": 0.19439683899838142,
+      "grad_norm": 13.958561969789983,
+      "learning_rate": 2e-05,
+      "loss": 1.27,
+      "step": 8167
+    },
+    {
+      "epoch": 0.19442064172141293,
+      "grad_norm": 4.772703733597175,
+      "learning_rate": 2e-05,
+      "loss": 0.835,
+      "step": 8168
+    },
+    {
+      "epoch": 0.19444444444444445,
+      "grad_norm": 18.651083145337058,
+      "learning_rate": 2e-05,
+      "loss": 0.3427,
+      "step": 8169
+    },
+    {
+      "epoch": 0.19446824716747596,
+      "grad_norm": 5.641136479677198,
+      "learning_rate": 2e-05,
+      "loss": 1.1004,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19449204989050747,
+      "grad_norm": 11.559445893487231,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 8171
+    },
+    {
+      "epoch": 0.194515852613539,
+      "grad_norm": 6.542449034815342,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 8172
+    },
+    {
+      "epoch": 0.1945396553365705,
+      "grad_norm": 39.44814706155535,
+      "learning_rate": 2e-05,
+      "loss": 0.5933,
+      "step": 8173
+    },
+    {
+      "epoch": 0.194563458059602,
+      "grad_norm": 31.358098237087404,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 8174
+    },
+    {
+      "epoch": 0.19458726078263353,
+      "grad_norm": 13.16362283339123,
+      "learning_rate": 2e-05,
+      "loss": 1.0041,
+      "step": 8175
+    },
+    {
+      "epoch": 0.19461106350566504,
+      "grad_norm": 5.594644153157345,
+      "learning_rate": 2e-05,
+      "loss": 0.8794,
+      "step": 8176
+    },
+    {
+      "epoch": 0.19463486622869655,
+      "grad_norm": 6.690032936661364,
+      "learning_rate": 2e-05,
+      "loss": 0.181,
+      "step": 8177
+    },
+    {
+      "epoch": 0.19465866895172806,
+      "grad_norm": 12.409170013814645,
+      "learning_rate": 2e-05,
+      "loss": 0.3756,
+      "step": 8178
+    },
+    {
+      "epoch": 0.1946824716747596,
+      "grad_norm": 8.287145639737764,
+      "learning_rate": 2e-05,
+      "loss": 0.2487,
+      "step": 8179
+    },
+    {
+      "epoch": 0.19470627439779112,
+      "grad_norm": 5.344448834495308,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19473007712082263,
+      "grad_norm": 5.64115470827068,
+      "learning_rate": 2e-05,
+      "loss": 0.1771,
+      "step": 8181
+    },
+    {
+      "epoch": 0.19475387984385414,
+      "grad_norm": 5.786241189635905,
+      "learning_rate": 2e-05,
+      "loss": 1.3709,
+      "step": 8182
+    },
+    {
+      "epoch": 0.19477768256688566,
+      "grad_norm": 6.623250849204729,
+      "learning_rate": 2e-05,
+      "loss": 1.0746,
+      "step": 8183
+    },
+    {
+      "epoch": 0.19480148528991717,
+      "grad_norm": 14.39641837442935,
+      "learning_rate": 2e-05,
+      "loss": 0.1868,
+      "step": 8184
+    },
+    {
+      "epoch": 0.19482528801294868,
+      "grad_norm": 2.154503293358701,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 8185
+    },
+    {
+      "epoch": 0.1948490907359802,
+      "grad_norm": 8.622447409440232,
+      "learning_rate": 2e-05,
+      "loss": 1.0363,
+      "step": 8186
+    },
+    {
+      "epoch": 0.1948728934590117,
+      "grad_norm": 7.228765258542687,
+      "learning_rate": 2e-05,
+      "loss": 0.9396,
+      "step": 8187
+    },
+    {
+      "epoch": 0.19489669618204322,
+      "grad_norm": 2.5519520106094316,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 8188
+    },
+    {
+      "epoch": 0.19492049890507473,
+      "grad_norm": 8.674773444331938,
+      "learning_rate": 2e-05,
+      "loss": 0.5395,
+      "step": 8189
+    },
+    {
+      "epoch": 0.19494430162810625,
+      "grad_norm": 13.994865325782063,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 8190
+    },
+    {
+      "epoch": 0.19496810435113776,
+      "grad_norm": 8.996529096799653,
+      "learning_rate": 2e-05,
+      "loss": 0.924,
+      "step": 8191
+    },
+    {
+      "epoch": 0.19499190707416927,
+      "grad_norm": 17.605297074995974,
+      "learning_rate": 2e-05,
+      "loss": 1.5662,
+      "step": 8192
+    },
+    {
+      "epoch": 0.1950157097972008,
+      "grad_norm": 19.23465314661634,
+      "learning_rate": 2e-05,
+      "loss": 0.3334,
+      "step": 8193
+    },
+    {
+      "epoch": 0.19503951252023233,
+      "grad_norm": 5.907823409339783,
+      "learning_rate": 2e-05,
+      "loss": 0.548,
+      "step": 8194
+    },
+    {
+      "epoch": 0.19506331524326384,
+      "grad_norm": 5.958202289186783,
+      "learning_rate": 2e-05,
+      "loss": 0.6448,
+      "step": 8195
+    },
+    {
+      "epoch": 0.19508711796629535,
+      "grad_norm": 5.9573872555545835,
+      "learning_rate": 2e-05,
+      "loss": 1.2682,
+      "step": 8196
+    },
+    {
+      "epoch": 0.19511092068932687,
+      "grad_norm": 5.378996428773103,
+      "learning_rate": 2e-05,
+      "loss": 0.5644,
+      "step": 8197
+    },
+    {
+      "epoch": 0.19513472341235838,
+      "grad_norm": 13.368625698344038,
+      "learning_rate": 2e-05,
+      "loss": 1.2401,
+      "step": 8198
+    },
+    {
+      "epoch": 0.1951585261353899,
+      "grad_norm": 1.2386975884822526,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 8199
+    },
+    {
+      "epoch": 0.1951823288584214,
+      "grad_norm": 5.3327043014383495,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 8200
+    },
+    {
+      "epoch": 0.19520613158145292,
+      "grad_norm": 4.81576560520581,
+      "learning_rate": 2e-05,
+      "loss": 0.5995,
+      "step": 8201
+    },
+    {
+      "epoch": 0.19522993430448443,
+      "grad_norm": 5.928083231393209,
+      "learning_rate": 2e-05,
+      "loss": 0.4059,
+      "step": 8202
+    },
+    {
+      "epoch": 0.19525373702751594,
+      "grad_norm": 5.4663059706866965,
+      "learning_rate": 2e-05,
+      "loss": 0.6419,
+      "step": 8203
+    },
+    {
+      "epoch": 0.19527753975054746,
+      "grad_norm": 16.9381811975827,
+      "learning_rate": 2e-05,
+      "loss": 0.8999,
+      "step": 8204
+    },
+    {
+      "epoch": 0.19530134247357897,
+      "grad_norm": 1.4204798979891753,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 8205
+    },
+    {
+      "epoch": 0.19532514519661048,
+      "grad_norm": 7.549518424348386,
+      "learning_rate": 2e-05,
+      "loss": 1.2273,
+      "step": 8206
+    },
+    {
+      "epoch": 0.195348947919642,
+      "grad_norm": 8.131235309637292,
+      "learning_rate": 2e-05,
+      "loss": 0.8612,
+      "step": 8207
+    },
+    {
+      "epoch": 0.1953727506426735,
+      "grad_norm": 1.9511527185299489,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 8208
+    },
+    {
+      "epoch": 0.19539655336570505,
+      "grad_norm": 4.979770577399732,
+      "learning_rate": 2e-05,
+      "loss": 0.6265,
+      "step": 8209
+    },
+    {
+      "epoch": 0.19542035608873656,
+      "grad_norm": 3.619082527065044,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19544415881176808,
+      "grad_norm": 13.759756673339908,
+      "learning_rate": 2e-05,
+      "loss": 0.4843,
+      "step": 8211
+    },
+    {
+      "epoch": 0.1954679615347996,
+      "grad_norm": 1.4488079437838137,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 8212
+    },
+    {
+      "epoch": 0.1954917642578311,
+      "grad_norm": 14.222334588564127,
+      "learning_rate": 2e-05,
+      "loss": 0.9995,
+      "step": 8213
+    },
+    {
+      "epoch": 0.19551556698086262,
+      "grad_norm": 6.428355601200472,
+      "learning_rate": 2e-05,
+      "loss": 1.2917,
+      "step": 8214
+    },
+    {
+      "epoch": 0.19553936970389413,
+      "grad_norm": 19.583600219200964,
+      "learning_rate": 2e-05,
+      "loss": 0.688,
+      "step": 8215
+    },
+    {
+      "epoch": 0.19556317242692564,
+      "grad_norm": 15.33664534625101,
+      "learning_rate": 2e-05,
+      "loss": 0.3069,
+      "step": 8216
+    },
+    {
+      "epoch": 0.19558697514995715,
+      "grad_norm": 5.1826725940689355,
+      "learning_rate": 2e-05,
+      "loss": 0.6522,
+      "step": 8217
+    },
+    {
+      "epoch": 0.19561077787298867,
+      "grad_norm": 1.6709641352959095,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 8218
+    },
+    {
+      "epoch": 0.19563458059602018,
+      "grad_norm": 3.1652779654542393,
+      "learning_rate": 2e-05,
+      "loss": 0.2133,
+      "step": 8219
+    },
+    {
+      "epoch": 0.1956583833190517,
+      "grad_norm": 2.3429844942368114,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 8220
+    },
+    {
+      "epoch": 0.1956821860420832,
+      "grad_norm": 4.42085642427971,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 8221
+    },
+    {
+      "epoch": 0.19570598876511472,
+      "grad_norm": 1.776955481529521,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 8222
+    },
+    {
+      "epoch": 0.19572979148814623,
+      "grad_norm": 2.494606585757494,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 8223
+    },
+    {
+      "epoch": 0.19575359421117774,
+      "grad_norm": 4.004926321105066,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 8224
+    },
+    {
+      "epoch": 0.19577739693420929,
+      "grad_norm": 3.9634827892581037,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 8225
+    },
+    {
+      "epoch": 0.1958011996572408,
+      "grad_norm": 3.802301627922158,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 8226
+    },
+    {
+      "epoch": 0.1958250023802723,
+      "grad_norm": 4.688479661219836,
+      "learning_rate": 2e-05,
+      "loss": 0.1845,
+      "step": 8227
+    },
+    {
+      "epoch": 0.19584880510330382,
+      "grad_norm": 2.6129671105584698,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 8228
+    },
+    {
+      "epoch": 0.19587260782633534,
+      "grad_norm": 6.285188695733909,
+      "learning_rate": 2e-05,
+      "loss": 1.2736,
+      "step": 8229
+    },
+    {
+      "epoch": 0.19589641054936685,
+      "grad_norm": 1.7626038213852138,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19592021327239836,
+      "grad_norm": 11.385402291774652,
+      "learning_rate": 2e-05,
+      "loss": 0.3534,
+      "step": 8231
+    },
+    {
+      "epoch": 0.19594401599542988,
+      "grad_norm": 6.701946271397562,
+      "learning_rate": 2e-05,
+      "loss": 1.0487,
+      "step": 8232
+    },
+    {
+      "epoch": 0.1959678187184614,
+      "grad_norm": 4.2501127113223385,
+      "learning_rate": 2e-05,
+      "loss": 0.655,
+      "step": 8233
+    },
+    {
+      "epoch": 0.1959916214414929,
+      "grad_norm": 3.569369879288568,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 8234
+    },
+    {
+      "epoch": 0.19601542416452442,
+      "grad_norm": 14.87310816152875,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 8235
+    },
+    {
+      "epoch": 0.19603922688755593,
+      "grad_norm": 3.7391399817126723,
+      "learning_rate": 2e-05,
+      "loss": 0.1955,
+      "step": 8236
+    },
+    {
+      "epoch": 0.19606302961058744,
+      "grad_norm": 2.6763677877259324,
+      "learning_rate": 2e-05,
+      "loss": 0.2272,
+      "step": 8237
+    },
+    {
+      "epoch": 0.19608683233361895,
+      "grad_norm": 1.2445295329749397,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 8238
+    },
+    {
+      "epoch": 0.19611063505665047,
+      "grad_norm": 0.8724036044169262,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 8239
+    },
+    {
+      "epoch": 0.196134437779682,
+      "grad_norm": 9.417881252189712,
+      "learning_rate": 2e-05,
+      "loss": 0.7988,
+      "step": 8240
+    },
+    {
+      "epoch": 0.19615824050271352,
+      "grad_norm": 3.65891420308063,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 8241
+    },
+    {
+      "epoch": 0.19618204322574503,
+      "grad_norm": 2.8571287942513663,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 8242
+    },
+    {
+      "epoch": 0.19620584594877655,
+      "grad_norm": 9.830032232669156,
+      "learning_rate": 2e-05,
+      "loss": 1.1955,
+      "step": 8243
+    },
+    {
+      "epoch": 0.19622964867180806,
+      "grad_norm": 12.273686911296235,
+      "learning_rate": 2e-05,
+      "loss": 0.4598,
+      "step": 8244
+    },
+    {
+      "epoch": 0.19625345139483957,
+      "grad_norm": 5.887407180406042,
+      "learning_rate": 2e-05,
+      "loss": 0.4682,
+      "step": 8245
+    },
+    {
+      "epoch": 0.19627725411787109,
+      "grad_norm": 0.34307454850966773,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 8246
+    },
+    {
+      "epoch": 0.1963010568409026,
+      "grad_norm": 3.181230816088385,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 8247
+    },
+    {
+      "epoch": 0.1963248595639341,
+      "grad_norm": 5.500423952776593,
+      "learning_rate": 2e-05,
+      "loss": 0.7217,
+      "step": 8248
+    },
+    {
+      "epoch": 0.19634866228696563,
+      "grad_norm": 5.237589948056954,
+      "learning_rate": 2e-05,
+      "loss": 0.1543,
+      "step": 8249
+    },
+    {
+      "epoch": 0.19637246500999714,
+      "grad_norm": 11.227153936228552,
+      "learning_rate": 2e-05,
+      "loss": 0.4156,
+      "step": 8250
+    },
+    {
+      "epoch": 0.19639626773302865,
+      "grad_norm": 11.5665207407465,
+      "learning_rate": 2e-05,
+      "loss": 0.7641,
+      "step": 8251
+    },
+    {
+      "epoch": 0.19642007045606016,
+      "grad_norm": 2.0665106915736136,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 8252
+    },
+    {
+      "epoch": 0.19644387317909168,
+      "grad_norm": 4.611746431579416,
+      "learning_rate": 2e-05,
+      "loss": 0.2072,
+      "step": 8253
+    },
+    {
+      "epoch": 0.1964676759021232,
+      "grad_norm": 4.904730951895704,
+      "learning_rate": 2e-05,
+      "loss": 0.3167,
+      "step": 8254
+    },
+    {
+      "epoch": 0.19649147862515473,
+      "grad_norm": 7.617335717744342,
+      "learning_rate": 2e-05,
+      "loss": 1.2651,
+      "step": 8255
+    },
+    {
+      "epoch": 0.19651528134818624,
+      "grad_norm": 5.467361230638927,
+      "learning_rate": 2e-05,
+      "loss": 0.5285,
+      "step": 8256
+    },
+    {
+      "epoch": 0.19653908407121776,
+      "grad_norm": 9.053122032329115,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 8257
+    },
+    {
+      "epoch": 0.19656288679424927,
+      "grad_norm": 5.035778385213117,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 8258
+    },
+    {
+      "epoch": 0.19658668951728078,
+      "grad_norm": 0.9829417692534281,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 8259
+    },
+    {
+      "epoch": 0.1966104922403123,
+      "grad_norm": 7.257176424853363,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 8260
+    },
+    {
+      "epoch": 0.1966342949633438,
+      "grad_norm": 5.169849381148398,
+      "learning_rate": 2e-05,
+      "loss": 0.2407,
+      "step": 8261
+    },
+    {
+      "epoch": 0.19665809768637532,
+      "grad_norm": 2.2256498396152598,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 8262
+    },
+    {
+      "epoch": 0.19668190040940683,
+      "grad_norm": 5.105038839897333,
+      "learning_rate": 2e-05,
+      "loss": 0.6503,
+      "step": 8263
+    },
+    {
+      "epoch": 0.19670570313243835,
+      "grad_norm": 4.531675825355472,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 8264
+    },
+    {
+      "epoch": 0.19672950585546986,
+      "grad_norm": 2.8849330025408553,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 8265
+    },
+    {
+      "epoch": 0.19675330857850137,
+      "grad_norm": 2.3092318675165644,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 8266
+    },
+    {
+      "epoch": 0.1967771113015329,
+      "grad_norm": 2.397627487616631,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 8267
+    },
+    {
+      "epoch": 0.1968009140245644,
+      "grad_norm": 7.734277322284766,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 8268
+    },
+    {
+      "epoch": 0.1968247167475959,
+      "grad_norm": 7.12609919410767,
+      "learning_rate": 2e-05,
+      "loss": 0.2589,
+      "step": 8269
+    },
+    {
+      "epoch": 0.19684851947062745,
+      "grad_norm": 4.548616727569129,
+      "learning_rate": 2e-05,
+      "loss": 1.0584,
+      "step": 8270
+    },
+    {
+      "epoch": 0.19687232219365897,
+      "grad_norm": 6.878572567715338,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 8271
+    },
+    {
+      "epoch": 0.19689612491669048,
+      "grad_norm": 3.441867131849701,
+      "learning_rate": 2e-05,
+      "loss": 0.2592,
+      "step": 8272
+    },
+    {
+      "epoch": 0.196919927639722,
+      "grad_norm": 1.7510741708993196,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 8273
+    },
+    {
+      "epoch": 0.1969437303627535,
+      "grad_norm": 6.922716707031501,
+      "learning_rate": 2e-05,
+      "loss": 0.3791,
+      "step": 8274
+    },
+    {
+      "epoch": 0.19696753308578502,
+      "grad_norm": 5.327783775344866,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 8275
+    },
+    {
+      "epoch": 0.19699133580881653,
+      "grad_norm": 4.202663462541398,
+      "learning_rate": 2e-05,
+      "loss": 0.4015,
+      "step": 8276
+    },
+    {
+      "epoch": 0.19701513853184804,
+      "grad_norm": 3.2345700444249443,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 8277
+    },
+    {
+      "epoch": 0.19703894125487956,
+      "grad_norm": 8.351180967963359,
+      "learning_rate": 2e-05,
+      "loss": 0.3936,
+      "step": 8278
+    },
+    {
+      "epoch": 0.19706274397791107,
+      "grad_norm": 9.74488884737714,
+      "learning_rate": 2e-05,
+      "loss": 0.3196,
+      "step": 8279
+    },
+    {
+      "epoch": 0.19708654670094258,
+      "grad_norm": 2.105068652883412,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 8280
+    },
+    {
+      "epoch": 0.1971103494239741,
+      "grad_norm": 5.288976676764121,
+      "learning_rate": 2e-05,
+      "loss": 0.2736,
+      "step": 8281
+    },
+    {
+      "epoch": 0.1971341521470056,
+      "grad_norm": 4.062459929295905,
+      "learning_rate": 2e-05,
+      "loss": 0.2098,
+      "step": 8282
+    },
+    {
+      "epoch": 0.19715795487003712,
+      "grad_norm": 7.10541799674085,
+      "learning_rate": 2e-05,
+      "loss": 0.1387,
+      "step": 8283
+    },
+    {
+      "epoch": 0.19718175759306864,
+      "grad_norm": 5.326789524723656,
+      "learning_rate": 2e-05,
+      "loss": 0.3113,
+      "step": 8284
+    },
+    {
+      "epoch": 0.19720556031610015,
+      "grad_norm": 7.150344677609223,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 8285
+    },
+    {
+      "epoch": 0.1972293630391317,
+      "grad_norm": 4.584673214628423,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 8286
+    },
+    {
+      "epoch": 0.1972531657621632,
+      "grad_norm": 2.9345155812202117,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 8287
+    },
+    {
+      "epoch": 0.19727696848519471,
+      "grad_norm": 17.016864458334567,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 8288
+    },
+    {
+      "epoch": 0.19730077120822623,
+      "grad_norm": 4.000156938646291,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 8289
+    },
+    {
+      "epoch": 0.19732457393125774,
+      "grad_norm": 4.0367979192028125,
+      "learning_rate": 2e-05,
+      "loss": 0.1493,
+      "step": 8290
+    },
+    {
+      "epoch": 0.19734837665428925,
+      "grad_norm": 5.850091570515664,
+      "learning_rate": 2e-05,
+      "loss": 1.0182,
+      "step": 8291
+    },
+    {
+      "epoch": 0.19737217937732077,
+      "grad_norm": 6.1887469005666595,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 8292
+    },
+    {
+      "epoch": 0.19739598210035228,
+      "grad_norm": 4.44690198255282,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 8293
+    },
+    {
+      "epoch": 0.1974197848233838,
+      "grad_norm": 11.393018668994548,
+      "learning_rate": 2e-05,
+      "loss": 1.6948,
+      "step": 8294
+    },
+    {
+      "epoch": 0.1974435875464153,
+      "grad_norm": 1.6117199644826241,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 8295
+    },
+    {
+      "epoch": 0.19746739026944682,
+      "grad_norm": 7.624431648169151,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 8296
+    },
+    {
+      "epoch": 0.19749119299247833,
+      "grad_norm": 2.526499358231623,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 8297
+    },
+    {
+      "epoch": 0.19751499571550984,
+      "grad_norm": 5.974882163951137,
+      "learning_rate": 2e-05,
+      "loss": 0.2814,
+      "step": 8298
+    },
+    {
+      "epoch": 0.19753879843854136,
+      "grad_norm": 5.388437796133777,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 8299
+    },
+    {
+      "epoch": 0.19756260116157287,
+      "grad_norm": 7.472101064305208,
+      "learning_rate": 2e-05,
+      "loss": 0.1813,
+      "step": 8300
+    },
+    {
+      "epoch": 0.1975864038846044,
+      "grad_norm": 33.84393960058466,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 8301
+    },
+    {
+      "epoch": 0.19761020660763592,
+      "grad_norm": 3.1219790205363744,
+      "learning_rate": 2e-05,
+      "loss": 0.1407,
+      "step": 8302
+    },
+    {
+      "epoch": 0.19763400933066744,
+      "grad_norm": 5.685458916173922,
+      "learning_rate": 2e-05,
+      "loss": 0.3987,
+      "step": 8303
+    },
+    {
+      "epoch": 0.19765781205369895,
+      "grad_norm": 7.275810799464361,
+      "learning_rate": 2e-05,
+      "loss": 0.1838,
+      "step": 8304
+    },
+    {
+      "epoch": 0.19768161477673046,
+      "grad_norm": 9.97933963054512,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 8305
+    },
+    {
+      "epoch": 0.19770541749976198,
+      "grad_norm": 2.101236170880599,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 8306
+    },
+    {
+      "epoch": 0.1977292202227935,
+      "grad_norm": 5.935093875264034,
+      "learning_rate": 2e-05,
+      "loss": 0.1973,
+      "step": 8307
+    },
+    {
+      "epoch": 0.197753022945825,
+      "grad_norm": 16.333113188868953,
+      "learning_rate": 2e-05,
+      "loss": 0.4344,
+      "step": 8308
+    },
+    {
+      "epoch": 0.19777682566885652,
+      "grad_norm": 3.8161741966556115,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 8309
+    },
+    {
+      "epoch": 0.19780062839188803,
+      "grad_norm": 12.067881633877482,
+      "learning_rate": 2e-05,
+      "loss": 0.4997,
+      "step": 8310
+    },
+    {
+      "epoch": 0.19782443111491954,
+      "grad_norm": 3.4345983676803553,
+      "learning_rate": 2e-05,
+      "loss": 0.1846,
+      "step": 8311
+    },
+    {
+      "epoch": 0.19784823383795105,
+      "grad_norm": 3.2420113113125355,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 8312
+    },
+    {
+      "epoch": 0.19787203656098257,
+      "grad_norm": 3.988488643839656,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 8313
+    },
+    {
+      "epoch": 0.19789583928401408,
+      "grad_norm": 5.430219822338803,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 8314
+    },
+    {
+      "epoch": 0.1979196420070456,
+      "grad_norm": 5.20079495469595,
+      "learning_rate": 2e-05,
+      "loss": 0.5196,
+      "step": 8315
+    },
+    {
+      "epoch": 0.19794344473007713,
+      "grad_norm": 2.5568072123604715,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 8316
+    },
+    {
+      "epoch": 0.19796724745310865,
+      "grad_norm": 13.203668496391751,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 8317
+    },
+    {
+      "epoch": 0.19799105017614016,
+      "grad_norm": 4.6654781076749865,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 8318
+    },
+    {
+      "epoch": 0.19801485289917167,
+      "grad_norm": 20.39577452424165,
+      "learning_rate": 2e-05,
+      "loss": 0.4281,
+      "step": 8319
+    },
+    {
+      "epoch": 0.19803865562220319,
+      "grad_norm": 3.7037376877846144,
+      "learning_rate": 2e-05,
+      "loss": 0.2265,
+      "step": 8320
+    },
+    {
+      "epoch": 0.1980624583452347,
+      "grad_norm": 2.2530176381177407,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 8321
+    },
+    {
+      "epoch": 0.1980862610682662,
+      "grad_norm": 4.25710624032476,
+      "learning_rate": 2e-05,
+      "loss": 0.7466,
+      "step": 8322
+    },
+    {
+      "epoch": 0.19811006379129772,
+      "grad_norm": 6.232091924691847,
+      "learning_rate": 2e-05,
+      "loss": 0.9418,
+      "step": 8323
+    },
+    {
+      "epoch": 0.19813386651432924,
+      "grad_norm": 3.031630444468995,
+      "learning_rate": 2e-05,
+      "loss": 0.1527,
+      "step": 8324
+    },
+    {
+      "epoch": 0.19815766923736075,
+      "grad_norm": 21.39481626113053,
+      "learning_rate": 2e-05,
+      "loss": 0.7084,
+      "step": 8325
+    },
+    {
+      "epoch": 0.19818147196039226,
+      "grad_norm": 4.998379973191086,
+      "learning_rate": 2e-05,
+      "loss": 0.7877,
+      "step": 8326
+    },
+    {
+      "epoch": 0.19820527468342378,
+      "grad_norm": 3.328295595890113,
+      "learning_rate": 2e-05,
+      "loss": 0.2633,
+      "step": 8327
+    },
+    {
+      "epoch": 0.1982290774064553,
+      "grad_norm": 5.2959354453822245,
+      "learning_rate": 2e-05,
+      "loss": 1.2681,
+      "step": 8328
+    },
+    {
+      "epoch": 0.1982528801294868,
+      "grad_norm": 3.533106759828098,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 8329
+    },
+    {
+      "epoch": 0.19827668285251832,
+      "grad_norm": 5.837073392828597,
+      "learning_rate": 2e-05,
+      "loss": 0.7727,
+      "step": 8330
+    },
+    {
+      "epoch": 0.19830048557554986,
+      "grad_norm": 4.63978389342808,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 8331
+    },
+    {
+      "epoch": 0.19832428829858137,
+      "grad_norm": 7.018011293266333,
+      "learning_rate": 2e-05,
+      "loss": 0.4718,
+      "step": 8332
+    },
+    {
+      "epoch": 0.19834809102161288,
+      "grad_norm": 22.59690151212213,
+      "learning_rate": 2e-05,
+      "loss": 0.4498,
+      "step": 8333
+    },
+    {
+      "epoch": 0.1983718937446444,
+      "grad_norm": 6.220245098918459,
+      "learning_rate": 2e-05,
+      "loss": 0.7933,
+      "step": 8334
+    },
+    {
+      "epoch": 0.1983956964676759,
+      "grad_norm": 4.838457774244191,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 8335
+    },
+    {
+      "epoch": 0.19841949919070742,
+      "grad_norm": 5.011336477648528,
+      "learning_rate": 2e-05,
+      "loss": 0.9286,
+      "step": 8336
+    },
+    {
+      "epoch": 0.19844330191373893,
+      "grad_norm": 4.143098034689945,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 8337
+    },
+    {
+      "epoch": 0.19846710463677045,
+      "grad_norm": 3.3272507546400765,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 8338
+    },
+    {
+      "epoch": 0.19849090735980196,
+      "grad_norm": 4.63198939524583,
+      "learning_rate": 2e-05,
+      "loss": 0.1758,
+      "step": 8339
+    },
+    {
+      "epoch": 0.19851471008283347,
+      "grad_norm": 7.916118165656389,
+      "learning_rate": 2e-05,
+      "loss": 0.5405,
+      "step": 8340
+    },
+    {
+      "epoch": 0.198538512805865,
+      "grad_norm": 3.842987696419106,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 8341
+    },
+    {
+      "epoch": 0.1985623155288965,
+      "grad_norm": 13.689294968830598,
+      "learning_rate": 2e-05,
+      "loss": 0.598,
+      "step": 8342
+    },
+    {
+      "epoch": 0.198586118251928,
+      "grad_norm": 16.271664097212746,
+      "learning_rate": 2e-05,
+      "loss": 0.4448,
+      "step": 8343
+    },
+    {
+      "epoch": 0.19860992097495953,
+      "grad_norm": 7.408324249403971,
+      "learning_rate": 2e-05,
+      "loss": 0.543,
+      "step": 8344
+    },
+    {
+      "epoch": 0.19863372369799104,
+      "grad_norm": 3.4141577036867834,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 8345
+    },
+    {
+      "epoch": 0.19865752642102258,
+      "grad_norm": 4.101052266355589,
+      "learning_rate": 2e-05,
+      "loss": 0.7736,
+      "step": 8346
+    },
+    {
+      "epoch": 0.1986813291440541,
+      "grad_norm": 8.091067179346748,
+      "learning_rate": 2e-05,
+      "loss": 0.8081,
+      "step": 8347
+    },
+    {
+      "epoch": 0.1987051318670856,
+      "grad_norm": 4.678384760772551,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 8348
+    },
+    {
+      "epoch": 0.19872893459011712,
+      "grad_norm": 4.647041393079729,
+      "learning_rate": 2e-05,
+      "loss": 0.6363,
+      "step": 8349
+    },
+    {
+      "epoch": 0.19875273731314863,
+      "grad_norm": 2.11139585314447,
+      "learning_rate": 2e-05,
+      "loss": 0.1265,
+      "step": 8350
+    },
+    {
+      "epoch": 0.19877654003618014,
+      "grad_norm": 3.6864541404732707,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 8351
+    },
+    {
+      "epoch": 0.19880034275921166,
+      "grad_norm": 5.53501334705717,
+      "learning_rate": 2e-05,
+      "loss": 0.2978,
+      "step": 8352
+    },
+    {
+      "epoch": 0.19882414548224317,
+      "grad_norm": 11.236217607254408,
+      "learning_rate": 2e-05,
+      "loss": 0.2352,
+      "step": 8353
+    },
+    {
+      "epoch": 0.19884794820527468,
+      "grad_norm": 11.641836104797532,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 8354
+    },
+    {
+      "epoch": 0.1988717509283062,
+      "grad_norm": 4.457343506405642,
+      "learning_rate": 2e-05,
+      "loss": 0.2274,
+      "step": 8355
+    },
+    {
+      "epoch": 0.1988955536513377,
+      "grad_norm": 7.34570599270079,
+      "learning_rate": 2e-05,
+      "loss": 0.9773,
+      "step": 8356
+    },
+    {
+      "epoch": 0.19891935637436922,
+      "grad_norm": 2.060836854082056,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 8357
+    },
+    {
+      "epoch": 0.19894315909740073,
+      "grad_norm": 1.3913219137799635,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 8358
+    },
+    {
+      "epoch": 0.19896696182043225,
+      "grad_norm": 2.1984362073391197,
+      "learning_rate": 2e-05,
+      "loss": 0.147,
+      "step": 8359
+    },
+    {
+      "epoch": 0.19899076454346376,
+      "grad_norm": 3.510119502848441,
+      "learning_rate": 2e-05,
+      "loss": 0.2368,
+      "step": 8360
+    },
+    {
+      "epoch": 0.19901456726649527,
+      "grad_norm": 10.26091209024547,
+      "learning_rate": 2e-05,
+      "loss": 0.424,
+      "step": 8361
+    },
+    {
+      "epoch": 0.19903836998952681,
+      "grad_norm": 7.146830510656014,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 8362
+    },
+    {
+      "epoch": 0.19906217271255833,
+      "grad_norm": 4.8302161597648094,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 8363
+    },
+    {
+      "epoch": 0.19908597543558984,
+      "grad_norm": 3.3749342604525006,
+      "learning_rate": 2e-05,
+      "loss": 0.1596,
+      "step": 8364
+    },
+    {
+      "epoch": 0.19910977815862135,
+      "grad_norm": 8.956324126626852,
+      "learning_rate": 2e-05,
+      "loss": 0.3906,
+      "step": 8365
+    },
+    {
+      "epoch": 0.19913358088165287,
+      "grad_norm": 2.6594644269744068,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 8366
+    },
+    {
+      "epoch": 0.19915738360468438,
+      "grad_norm": 5.298847231411081,
+      "learning_rate": 2e-05,
+      "loss": 0.497,
+      "step": 8367
+    },
+    {
+      "epoch": 0.1991811863277159,
+      "grad_norm": 2.6474643527921446,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 8368
+    },
+    {
+      "epoch": 0.1992049890507474,
+      "grad_norm": 6.45961125526067,
+      "learning_rate": 2e-05,
+      "loss": 1.2253,
+      "step": 8369
+    },
+    {
+      "epoch": 0.19922879177377892,
+      "grad_norm": 3.4354680900029098,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 8370
+    },
+    {
+      "epoch": 0.19925259449681043,
+      "grad_norm": 6.233774596377994,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 8371
+    },
+    {
+      "epoch": 0.19927639721984194,
+      "grad_norm": 8.795392401148009,
+      "learning_rate": 2e-05,
+      "loss": 0.3061,
+      "step": 8372
+    },
+    {
+      "epoch": 0.19930019994287346,
+      "grad_norm": 5.734045439928501,
+      "learning_rate": 2e-05,
+      "loss": 0.7807,
+      "step": 8373
+    },
+    {
+      "epoch": 0.19932400266590497,
+      "grad_norm": 8.38898934273658,
+      "learning_rate": 2e-05,
+      "loss": 1.2893,
+      "step": 8374
+    },
+    {
+      "epoch": 0.19934780538893648,
+      "grad_norm": 8.822227263042103,
+      "learning_rate": 2e-05,
+      "loss": 0.4082,
+      "step": 8375
+    },
+    {
+      "epoch": 0.199371608111968,
+      "grad_norm": 7.904962024645192,
+      "learning_rate": 2e-05,
+      "loss": 0.9985,
+      "step": 8376
+    },
+    {
+      "epoch": 0.19939541083499954,
+      "grad_norm": 4.544719919544483,
+      "learning_rate": 2e-05,
+      "loss": 0.1923,
+      "step": 8377
+    },
+    {
+      "epoch": 0.19941921355803105,
+      "grad_norm": 18.508831647401305,
+      "learning_rate": 2e-05,
+      "loss": 0.6627,
+      "step": 8378
+    },
+    {
+      "epoch": 0.19944301628106256,
+      "grad_norm": 5.813661198313389,
+      "learning_rate": 2e-05,
+      "loss": 0.2939,
+      "step": 8379
+    },
+    {
+      "epoch": 0.19946681900409408,
+      "grad_norm": 5.892907732769461,
+      "learning_rate": 2e-05,
+      "loss": 0.9859,
+      "step": 8380
+    },
+    {
+      "epoch": 0.1994906217271256,
+      "grad_norm": 4.533276032166949,
+      "learning_rate": 2e-05,
+      "loss": 0.718,
+      "step": 8381
+    },
+    {
+      "epoch": 0.1995144244501571,
+      "grad_norm": 7.176369628215601,
+      "learning_rate": 2e-05,
+      "loss": 1.1441,
+      "step": 8382
+    },
+    {
+      "epoch": 0.19953822717318861,
+      "grad_norm": 7.491697800060027,
+      "learning_rate": 2e-05,
+      "loss": 0.7854,
+      "step": 8383
+    },
+    {
+      "epoch": 0.19956202989622013,
+      "grad_norm": 4.666027737437505,
+      "learning_rate": 2e-05,
+      "loss": 0.377,
+      "step": 8384
+    },
+    {
+      "epoch": 0.19958583261925164,
+      "grad_norm": 4.234343837193354,
+      "learning_rate": 2e-05,
+      "loss": 0.3397,
+      "step": 8385
+    },
+    {
+      "epoch": 0.19960963534228315,
+      "grad_norm": 9.734570965145835,
+      "learning_rate": 2e-05,
+      "loss": 0.2137,
+      "step": 8386
+    },
+    {
+      "epoch": 0.19963343806531467,
+      "grad_norm": 9.052538706707509,
+      "learning_rate": 2e-05,
+      "loss": 0.6345,
+      "step": 8387
+    },
+    {
+      "epoch": 0.19965724078834618,
+      "grad_norm": 3.6511029646333895,
+      "learning_rate": 2e-05,
+      "loss": 0.2504,
+      "step": 8388
+    },
+    {
+      "epoch": 0.1996810435113777,
+      "grad_norm": 1.2448790319871765,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 8389
+    },
+    {
+      "epoch": 0.1997048462344092,
+      "grad_norm": 8.211411922663466,
+      "learning_rate": 2e-05,
+      "loss": 1.8013,
+      "step": 8390
+    },
+    {
+      "epoch": 0.19972864895744072,
+      "grad_norm": 1.4217782607586018,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 8391
+    },
+    {
+      "epoch": 0.19975245168047226,
+      "grad_norm": 8.719066091159654,
+      "learning_rate": 2e-05,
+      "loss": 0.3672,
+      "step": 8392
+    },
+    {
+      "epoch": 0.19977625440350377,
+      "grad_norm": 2.40446452700299,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 8393
+    },
+    {
+      "epoch": 0.19980005712653529,
+      "grad_norm": 15.815346240918597,
+      "learning_rate": 2e-05,
+      "loss": 0.2467,
+      "step": 8394
+    },
+    {
+      "epoch": 0.1998238598495668,
+      "grad_norm": 3.8267728251642645,
+      "learning_rate": 2e-05,
+      "loss": 0.2133,
+      "step": 8395
+    },
+    {
+      "epoch": 0.1998476625725983,
+      "grad_norm": 3.803586225314321,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 8396
+    },
+    {
+      "epoch": 0.19987146529562982,
+      "grad_norm": 4.648355108242539,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 8397
+    },
+    {
+      "epoch": 0.19989526801866134,
+      "grad_norm": 4.385194528533499,
+      "learning_rate": 2e-05,
+      "loss": 0.8087,
+      "step": 8398
+    },
+    {
+      "epoch": 0.19991907074169285,
+      "grad_norm": 3.237541150646476,
+      "learning_rate": 2e-05,
+      "loss": 0.1788,
+      "step": 8399
+    },
+    {
+      "epoch": 0.19994287346472436,
+      "grad_norm": 4.642307585458307,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 8400
+    },
+    {
+      "epoch": 0.19996667618775588,
+      "grad_norm": 1.5537416227716667,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 8401
+    },
+    {
+      "epoch": 0.1999904789107874,
+      "grad_norm": 5.16824982640837,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 8402
+    },
+    {
+      "epoch": 0.2000142816338189,
+      "grad_norm": 3.2503684531508377,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 8403
+    },
+    {
+      "epoch": 0.20003808435685042,
+      "grad_norm": 7.436744553279803,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 8404
+    },
+    {
+      "epoch": 0.20006188707988193,
+      "grad_norm": 11.008285850904155,
+      "learning_rate": 2e-05,
+      "loss": 1.8048,
+      "step": 8405
+    },
+    {
+      "epoch": 0.20008568980291344,
+      "grad_norm": 2.3926467012575308,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 8406
+    },
+    {
+      "epoch": 0.20010949252594498,
+      "grad_norm": 3.1105991382062617,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 8407
+    },
+    {
+      "epoch": 0.2001332952489765,
+      "grad_norm": 2.5469348740208773,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 8408
+    },
+    {
+      "epoch": 0.200157097972008,
+      "grad_norm": 6.0860360095201695,
+      "learning_rate": 2e-05,
+      "loss": 0.9383,
+      "step": 8409
+    },
+    {
+      "epoch": 0.20018090069503952,
+      "grad_norm": 2.77117457700417,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 8410
+    },
+    {
+      "epoch": 0.20020470341807103,
+      "grad_norm": 5.897597757803737,
+      "learning_rate": 2e-05,
+      "loss": 0.7696,
+      "step": 8411
+    },
+    {
+      "epoch": 0.20022850614110255,
+      "grad_norm": 2.419103163674733,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 8412
+    },
+    {
+      "epoch": 0.20025230886413406,
+      "grad_norm": 83.1356631599733,
+      "learning_rate": 2e-05,
+      "loss": 0.3927,
+      "step": 8413
+    },
+    {
+      "epoch": 0.20027611158716557,
+      "grad_norm": 0.9081146257821364,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 8414
+    },
+    {
+      "epoch": 0.20029991431019709,
+      "grad_norm": 40.62070536529564,
+      "learning_rate": 2e-05,
+      "loss": 0.8424,
+      "step": 8415
+    },
+    {
+      "epoch": 0.2003237170332286,
+      "grad_norm": 3.9511025350016014,
+      "learning_rate": 2e-05,
+      "loss": 0.4075,
+      "step": 8416
+    },
+    {
+      "epoch": 0.2003475197562601,
+      "grad_norm": 7.607998790083051,
+      "learning_rate": 2e-05,
+      "loss": 0.7447,
+      "step": 8417
+    },
+    {
+      "epoch": 0.20037132247929162,
+      "grad_norm": 5.055402327579972,
+      "learning_rate": 2e-05,
+      "loss": 0.3481,
+      "step": 8418
+    },
+    {
+      "epoch": 0.20039512520232314,
+      "grad_norm": 2.9167551324587224,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 8419
+    },
+    {
+      "epoch": 0.20041892792535465,
+      "grad_norm": 18.053981396099395,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 8420
+    },
+    {
+      "epoch": 0.20044273064838616,
+      "grad_norm": 32.06219359573824,
+      "learning_rate": 2e-05,
+      "loss": 1.3052,
+      "step": 8421
+    },
+    {
+      "epoch": 0.20046653337141768,
+      "grad_norm": 1.7535498061829715,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 8422
+    },
+    {
+      "epoch": 0.20049033609444922,
+      "grad_norm": 6.775683742703384,
+      "learning_rate": 2e-05,
+      "loss": 0.7897,
+      "step": 8423
+    },
+    {
+      "epoch": 0.20051413881748073,
+      "grad_norm": 5.007841170351438,
+      "learning_rate": 2e-05,
+      "loss": 0.3057,
+      "step": 8424
+    },
+    {
+      "epoch": 0.20053794154051224,
+      "grad_norm": 3.716734572531388,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 8425
+    },
+    {
+      "epoch": 0.20056174426354376,
+      "grad_norm": 11.280138126722905,
+      "learning_rate": 2e-05,
+      "loss": 0.5408,
+      "step": 8426
+    },
+    {
+      "epoch": 0.20058554698657527,
+      "grad_norm": 7.515975374374249,
+      "learning_rate": 2e-05,
+      "loss": 0.2549,
+      "step": 8427
+    },
+    {
+      "epoch": 0.20060934970960678,
+      "grad_norm": 21.75100605121183,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 8428
+    },
+    {
+      "epoch": 0.2006331524326383,
+      "grad_norm": 9.956918515585263,
+      "learning_rate": 2e-05,
+      "loss": 0.5178,
+      "step": 8429
+    },
+    {
+      "epoch": 0.2006569551556698,
+      "grad_norm": 6.718055308647998,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 8430
+    },
+    {
+      "epoch": 0.20068075787870132,
+      "grad_norm": 3.758603804237433,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 8431
+    },
+    {
+      "epoch": 0.20070456060173283,
+      "grad_norm": 3.235410077748677,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 8432
+    },
+    {
+      "epoch": 0.20072836332476435,
+      "grad_norm": 4.011119404111407,
+      "learning_rate": 2e-05,
+      "loss": 0.239,
+      "step": 8433
+    },
+    {
+      "epoch": 0.20075216604779586,
+      "grad_norm": 5.146478551973286,
+      "learning_rate": 2e-05,
+      "loss": 0.8309,
+      "step": 8434
+    },
+    {
+      "epoch": 0.20077596877082737,
+      "grad_norm": 5.243585058324716,
+      "learning_rate": 2e-05,
+      "loss": 0.6286,
+      "step": 8435
+    },
+    {
+      "epoch": 0.2007997714938589,
+      "grad_norm": 5.4119719683000564,
+      "learning_rate": 2e-05,
+      "loss": 0.8719,
+      "step": 8436
+    },
+    {
+      "epoch": 0.2008235742168904,
+      "grad_norm": 16.618469166845482,
+      "learning_rate": 2e-05,
+      "loss": 1.0321,
+      "step": 8437
+    },
+    {
+      "epoch": 0.20084737693992194,
+      "grad_norm": 4.276110701202511,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 8438
+    },
+    {
+      "epoch": 0.20087117966295345,
+      "grad_norm": 6.040436497484373,
+      "learning_rate": 2e-05,
+      "loss": 0.8475,
+      "step": 8439
+    },
+    {
+      "epoch": 0.20089498238598497,
+      "grad_norm": 41.96012705924633,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 8440
+    },
+    {
+      "epoch": 0.20091878510901648,
+      "grad_norm": 7.439590173591966,
+      "learning_rate": 2e-05,
+      "loss": 0.8542,
+      "step": 8441
+    },
+    {
+      "epoch": 0.200942587832048,
+      "grad_norm": 2.7345884212750233,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 8442
+    },
+    {
+      "epoch": 0.2009663905550795,
+      "grad_norm": 6.469401067690647,
+      "learning_rate": 2e-05,
+      "loss": 1.0664,
+      "step": 8443
+    },
+    {
+      "epoch": 0.20099019327811102,
+      "grad_norm": 5.751611848941587,
+      "learning_rate": 2e-05,
+      "loss": 0.5823,
+      "step": 8444
+    },
+    {
+      "epoch": 0.20101399600114253,
+      "grad_norm": 27.017673361917716,
+      "learning_rate": 2e-05,
+      "loss": 0.5995,
+      "step": 8445
+    },
+    {
+      "epoch": 0.20103779872417404,
+      "grad_norm": 10.299239635877948,
+      "learning_rate": 2e-05,
+      "loss": 0.8645,
+      "step": 8446
+    },
+    {
+      "epoch": 0.20106160144720556,
+      "grad_norm": 7.58897094270041,
+      "learning_rate": 2e-05,
+      "loss": 1.0159,
+      "step": 8447
+    },
+    {
+      "epoch": 0.20108540417023707,
+      "grad_norm": 2.7688569660681956,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 8448
+    },
+    {
+      "epoch": 0.20110920689326858,
+      "grad_norm": 6.13593113417737,
+      "learning_rate": 2e-05,
+      "loss": 1.2313,
+      "step": 8449
+    },
+    {
+      "epoch": 0.2011330096163001,
+      "grad_norm": 23.87028173871066,
+      "learning_rate": 2e-05,
+      "loss": 0.3376,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2011568123393316,
+      "grad_norm": 9.756064619437723,
+      "learning_rate": 2e-05,
+      "loss": 1.4471,
+      "step": 8451
+    },
+    {
+      "epoch": 0.20118061506236312,
+      "grad_norm": 12.334997247661569,
+      "learning_rate": 2e-05,
+      "loss": 0.9697,
+      "step": 8452
+    },
+    {
+      "epoch": 0.20120441778539466,
+      "grad_norm": 3.861825357355222,
+      "learning_rate": 2e-05,
+      "loss": 0.2553,
+      "step": 8453
+    },
+    {
+      "epoch": 0.20122822050842618,
+      "grad_norm": 6.43196608776071,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 8454
+    },
+    {
+      "epoch": 0.2012520232314577,
+      "grad_norm": 3.709002914707994,
+      "learning_rate": 2e-05,
+      "loss": 0.144,
+      "step": 8455
+    },
+    {
+      "epoch": 0.2012758259544892,
+      "grad_norm": 14.570015009713336,
+      "learning_rate": 2e-05,
+      "loss": 0.3007,
+      "step": 8456
+    },
+    {
+      "epoch": 0.20129962867752071,
+      "grad_norm": 6.259936296312155,
+      "learning_rate": 2e-05,
+      "loss": 0.8437,
+      "step": 8457
+    },
+    {
+      "epoch": 0.20132343140055223,
+      "grad_norm": 1.43241422610773,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 8458
+    },
+    {
+      "epoch": 0.20134723412358374,
+      "grad_norm": 9.18536337027328,
+      "learning_rate": 2e-05,
+      "loss": 0.3464,
+      "step": 8459
+    },
+    {
+      "epoch": 0.20137103684661525,
+      "grad_norm": 2.448269361150266,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 8460
+    },
+    {
+      "epoch": 0.20139483956964677,
+      "grad_norm": 2.3097745076540055,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 8461
+    },
+    {
+      "epoch": 0.20141864229267828,
+      "grad_norm": 5.150759526098529,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 8462
+    },
+    {
+      "epoch": 0.2014424450157098,
+      "grad_norm": 10.405384035342312,
+      "learning_rate": 2e-05,
+      "loss": 0.469,
+      "step": 8463
+    },
+    {
+      "epoch": 0.2014662477387413,
+      "grad_norm": 5.419620571607893,
+      "learning_rate": 2e-05,
+      "loss": 0.4157,
+      "step": 8464
+    },
+    {
+      "epoch": 0.20149005046177282,
+      "grad_norm": 8.626058940548457,
+      "learning_rate": 2e-05,
+      "loss": 1.1524,
+      "step": 8465
+    },
+    {
+      "epoch": 0.20151385318480433,
+      "grad_norm": 0.37782274843131425,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 8466
+    },
+    {
+      "epoch": 0.20153765590783584,
+      "grad_norm": 3.7745256585580775,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 8467
+    },
+    {
+      "epoch": 0.20156145863086739,
+      "grad_norm": 7.4194389161969605,
+      "learning_rate": 2e-05,
+      "loss": 0.3332,
+      "step": 8468
+    },
+    {
+      "epoch": 0.2015852613538989,
+      "grad_norm": 1.1552765217752976,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 8469
+    },
+    {
+      "epoch": 0.2016090640769304,
+      "grad_norm": 1.2189689075620223,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 8470
+    },
+    {
+      "epoch": 0.20163286679996192,
+      "grad_norm": 7.733823623931605,
+      "learning_rate": 2e-05,
+      "loss": 0.8186,
+      "step": 8471
+    },
+    {
+      "epoch": 0.20165666952299344,
+      "grad_norm": 10.290086332524954,
+      "learning_rate": 2e-05,
+      "loss": 0.1494,
+      "step": 8472
+    },
+    {
+      "epoch": 0.20168047224602495,
+      "grad_norm": 10.674382113819581,
+      "learning_rate": 2e-05,
+      "loss": 0.4681,
+      "step": 8473
+    },
+    {
+      "epoch": 0.20170427496905646,
+      "grad_norm": 7.530407559657052,
+      "learning_rate": 2e-05,
+      "loss": 0.8871,
+      "step": 8474
+    },
+    {
+      "epoch": 0.20172807769208798,
+      "grad_norm": 4.682135508285009,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 8475
+    },
+    {
+      "epoch": 0.2017518804151195,
+      "grad_norm": 7.661015395006889,
+      "learning_rate": 2e-05,
+      "loss": 0.2649,
+      "step": 8476
+    },
+    {
+      "epoch": 0.201775683138151,
+      "grad_norm": 8.263347825189983,
+      "learning_rate": 2e-05,
+      "loss": 1.6029,
+      "step": 8477
+    },
+    {
+      "epoch": 0.20179948586118251,
+      "grad_norm": 22.405625369007108,
+      "learning_rate": 2e-05,
+      "loss": 0.6503,
+      "step": 8478
+    },
+    {
+      "epoch": 0.20182328858421403,
+      "grad_norm": 4.20113988468767,
+      "learning_rate": 2e-05,
+      "loss": 0.7543,
+      "step": 8479
+    },
+    {
+      "epoch": 0.20184709130724554,
+      "grad_norm": 3.9796105354644626,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 8480
+    },
+    {
+      "epoch": 0.20187089403027705,
+      "grad_norm": 2.2246823949379797,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 8481
+    },
+    {
+      "epoch": 0.20189469675330857,
+      "grad_norm": 7.895371005337071,
+      "learning_rate": 2e-05,
+      "loss": 1.4464,
+      "step": 8482
+    },
+    {
+      "epoch": 0.20191849947634008,
+      "grad_norm": 9.73350873890513,
+      "learning_rate": 2e-05,
+      "loss": 0.232,
+      "step": 8483
+    },
+    {
+      "epoch": 0.20194230219937162,
+      "grad_norm": 4.96256878092079,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 8484
+    },
+    {
+      "epoch": 0.20196610492240313,
+      "grad_norm": 9.30375320711463,
+      "learning_rate": 2e-05,
+      "loss": 0.3395,
+      "step": 8485
+    },
+    {
+      "epoch": 0.20198990764543465,
+      "grad_norm": 2.4387616248848754,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 8486
+    },
+    {
+      "epoch": 0.20201371036846616,
+      "grad_norm": 8.822417013644456,
+      "learning_rate": 2e-05,
+      "loss": 1.3246,
+      "step": 8487
+    },
+    {
+      "epoch": 0.20203751309149767,
+      "grad_norm": 18.165926234568147,
+      "learning_rate": 2e-05,
+      "loss": 1.0489,
+      "step": 8488
+    },
+    {
+      "epoch": 0.20206131581452919,
+      "grad_norm": 8.542345695049812,
+      "learning_rate": 2e-05,
+      "loss": 2.0999,
+      "step": 8489
+    },
+    {
+      "epoch": 0.2020851185375607,
+      "grad_norm": 3.2136294986429106,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2021089212605922,
+      "grad_norm": 15.423320338307729,
+      "learning_rate": 2e-05,
+      "loss": 0.7126,
+      "step": 8491
+    },
+    {
+      "epoch": 0.20213272398362372,
+      "grad_norm": 4.092666458136279,
+      "learning_rate": 2e-05,
+      "loss": 0.1128,
+      "step": 8492
+    },
+    {
+      "epoch": 0.20215652670665524,
+      "grad_norm": 0.43409454155086125,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 8493
+    },
+    {
+      "epoch": 0.20218032942968675,
+      "grad_norm": 4.4391640828593575,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 8494
+    },
+    {
+      "epoch": 0.20220413215271826,
+      "grad_norm": 1.5053340429516262,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 8495
+    },
+    {
+      "epoch": 0.20222793487574978,
+      "grad_norm": 4.786852888092633,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 8496
+    },
+    {
+      "epoch": 0.2022517375987813,
+      "grad_norm": 2.1953698550723124,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 8497
+    },
+    {
+      "epoch": 0.2022755403218128,
+      "grad_norm": 2.959585970329295,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 8498
+    },
+    {
+      "epoch": 0.20229934304484434,
+      "grad_norm": 10.021875908640624,
+      "learning_rate": 2e-05,
+      "loss": 1.2677,
+      "step": 8499
+    },
+    {
+      "epoch": 0.20232314576787586,
+      "grad_norm": 10.50915961520215,
+      "learning_rate": 2e-05,
+      "loss": 0.6654,
+      "step": 8500
+    },
+    {
+      "epoch": 0.20234694849090737,
+      "grad_norm": 6.950329641034902,
+      "learning_rate": 2e-05,
+      "loss": 0.479,
+      "step": 8501
+    },
+    {
+      "epoch": 0.20237075121393888,
+      "grad_norm": 12.120673247617214,
+      "learning_rate": 2e-05,
+      "loss": 1.0146,
+      "step": 8502
+    },
+    {
+      "epoch": 0.2023945539369704,
+      "grad_norm": 14.304683753275784,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 8503
+    },
+    {
+      "epoch": 0.2024183566600019,
+      "grad_norm": 0.4138129507482259,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 8504
+    },
+    {
+      "epoch": 0.20244215938303342,
+      "grad_norm": 10.905868882481903,
+      "learning_rate": 2e-05,
+      "loss": 0.5967,
+      "step": 8505
+    },
+    {
+      "epoch": 0.20246596210606493,
+      "grad_norm": 6.3295977306877536,
+      "learning_rate": 2e-05,
+      "loss": 0.8733,
+      "step": 8506
+    },
+    {
+      "epoch": 0.20248976482909645,
+      "grad_norm": 6.403508141098564,
+      "learning_rate": 2e-05,
+      "loss": 0.3497,
+      "step": 8507
+    },
+    {
+      "epoch": 0.20251356755212796,
+      "grad_norm": 9.757642628482476,
+      "learning_rate": 2e-05,
+      "loss": 0.1475,
+      "step": 8508
+    },
+    {
+      "epoch": 0.20253737027515947,
+      "grad_norm": 7.4737323761424665,
+      "learning_rate": 2e-05,
+      "loss": 0.9224,
+      "step": 8509
+    },
+    {
+      "epoch": 0.20256117299819099,
+      "grad_norm": 6.030855615798981,
+      "learning_rate": 2e-05,
+      "loss": 1.0621,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2025849757212225,
+      "grad_norm": 8.381103659717347,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 8511
+    },
+    {
+      "epoch": 0.202608778444254,
+      "grad_norm": 3.609258617646838,
+      "learning_rate": 2e-05,
+      "loss": 0.2761,
+      "step": 8512
+    },
+    {
+      "epoch": 0.20263258116728552,
+      "grad_norm": 4.269626275398919,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 8513
+    },
+    {
+      "epoch": 0.20265638389031707,
+      "grad_norm": 1.7035278849760813,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 8514
+    },
+    {
+      "epoch": 0.20268018661334858,
+      "grad_norm": 12.289530545516495,
+      "learning_rate": 2e-05,
+      "loss": 2.1729,
+      "step": 8515
+    },
+    {
+      "epoch": 0.2027039893363801,
+      "grad_norm": 3.2870923573306747,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 8516
+    },
+    {
+      "epoch": 0.2027277920594116,
+      "grad_norm": 2.340769090870906,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 8517
+    },
+    {
+      "epoch": 0.20275159478244312,
+      "grad_norm": 3.483595834077667,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 8518
+    },
+    {
+      "epoch": 0.20277539750547463,
+      "grad_norm": 2.728902336984641,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 8519
+    },
+    {
+      "epoch": 0.20279920022850614,
+      "grad_norm": 19.858252766057696,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 8520
+    },
+    {
+      "epoch": 0.20282300295153766,
+      "grad_norm": 10.491080362776321,
+      "learning_rate": 2e-05,
+      "loss": 0.4673,
+      "step": 8521
+    },
+    {
+      "epoch": 0.20284680567456917,
+      "grad_norm": 11.15991826940205,
+      "learning_rate": 2e-05,
+      "loss": 1.5256,
+      "step": 8522
+    },
+    {
+      "epoch": 0.20287060839760068,
+      "grad_norm": 3.277138341073863,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 8523
+    },
+    {
+      "epoch": 0.2028944111206322,
+      "grad_norm": 1.7224404956609218,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 8524
+    },
+    {
+      "epoch": 0.2029182138436637,
+      "grad_norm": 5.470671502748618,
+      "learning_rate": 2e-05,
+      "loss": 0.958,
+      "step": 8525
+    },
+    {
+      "epoch": 0.20294201656669522,
+      "grad_norm": 11.6977308610548,
+      "learning_rate": 2e-05,
+      "loss": 0.3367,
+      "step": 8526
+    },
+    {
+      "epoch": 0.20296581928972673,
+      "grad_norm": 4.971722791752444,
+      "learning_rate": 2e-05,
+      "loss": 0.6983,
+      "step": 8527
+    },
+    {
+      "epoch": 0.20298962201275825,
+      "grad_norm": 6.821763307696504,
+      "learning_rate": 2e-05,
+      "loss": 0.2533,
+      "step": 8528
+    },
+    {
+      "epoch": 0.2030134247357898,
+      "grad_norm": 4.2543924122560846,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 8529
+    },
+    {
+      "epoch": 0.2030372274588213,
+      "grad_norm": 5.092166583653885,
+      "learning_rate": 2e-05,
+      "loss": 0.35,
+      "step": 8530
+    },
+    {
+      "epoch": 0.20306103018185281,
+      "grad_norm": 1.6407966105888552,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 8531
+    },
+    {
+      "epoch": 0.20308483290488433,
+      "grad_norm": 10.192069829844625,
+      "learning_rate": 2e-05,
+      "loss": 1.6319,
+      "step": 8532
+    },
+    {
+      "epoch": 0.20310863562791584,
+      "grad_norm": 15.894095349459596,
+      "learning_rate": 2e-05,
+      "loss": 0.7726,
+      "step": 8533
+    },
+    {
+      "epoch": 0.20313243835094735,
+      "grad_norm": 9.087446264734687,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 8534
+    },
+    {
+      "epoch": 0.20315624107397887,
+      "grad_norm": 10.983653249988542,
+      "learning_rate": 2e-05,
+      "loss": 0.4823,
+      "step": 8535
+    },
+    {
+      "epoch": 0.20318004379701038,
+      "grad_norm": 2.24095369244359,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 8536
+    },
+    {
+      "epoch": 0.2032038465200419,
+      "grad_norm": 4.316509854060766,
+      "learning_rate": 2e-05,
+      "loss": 0.3165,
+      "step": 8537
+    },
+    {
+      "epoch": 0.2032276492430734,
+      "grad_norm": 2.7089287754840194,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 8538
+    },
+    {
+      "epoch": 0.20325145196610492,
+      "grad_norm": 4.809262800022662,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 8539
+    },
+    {
+      "epoch": 0.20327525468913643,
+      "grad_norm": 1.4566324928822008,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 8540
+    },
+    {
+      "epoch": 0.20329905741216794,
+      "grad_norm": 24.864011199778172,
+      "learning_rate": 2e-05,
+      "loss": 1.176,
+      "step": 8541
+    },
+    {
+      "epoch": 0.20332286013519946,
+      "grad_norm": 5.32606971085368,
+      "learning_rate": 2e-05,
+      "loss": 0.6585,
+      "step": 8542
+    },
+    {
+      "epoch": 0.20334666285823097,
+      "grad_norm": 10.38331818043811,
+      "learning_rate": 2e-05,
+      "loss": 1.0799,
+      "step": 8543
+    },
+    {
+      "epoch": 0.20337046558126248,
+      "grad_norm": 1.0793834719647677,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 8544
+    },
+    {
+      "epoch": 0.20339426830429402,
+      "grad_norm": 6.374591143380678,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 8545
+    },
+    {
+      "epoch": 0.20341807102732554,
+      "grad_norm": 5.983312425744936,
+      "learning_rate": 2e-05,
+      "loss": 0.5522,
+      "step": 8546
+    },
+    {
+      "epoch": 0.20344187375035705,
+      "grad_norm": 6.481004776739478,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 8547
+    },
+    {
+      "epoch": 0.20346567647338856,
+      "grad_norm": 5.560882370067927,
+      "learning_rate": 2e-05,
+      "loss": 0.2101,
+      "step": 8548
+    },
+    {
+      "epoch": 0.20348947919642008,
+      "grad_norm": 6.0807884547167,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 8549
+    },
+    {
+      "epoch": 0.2035132819194516,
+      "grad_norm": 2.9707333610198563,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2035370846424831,
+      "grad_norm": 1.565957874936794,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 8551
+    },
+    {
+      "epoch": 0.20356088736551461,
+      "grad_norm": 5.219888197940253,
+      "learning_rate": 2e-05,
+      "loss": 0.1941,
+      "step": 8552
+    },
+    {
+      "epoch": 0.20358469008854613,
+      "grad_norm": 13.531652250966793,
+      "learning_rate": 2e-05,
+      "loss": 0.6616,
+      "step": 8553
+    },
+    {
+      "epoch": 0.20360849281157764,
+      "grad_norm": 8.662730403698358,
+      "learning_rate": 2e-05,
+      "loss": 1.0103,
+      "step": 8554
+    },
+    {
+      "epoch": 0.20363229553460915,
+      "grad_norm": 4.814236803516205,
+      "learning_rate": 2e-05,
+      "loss": 0.5701,
+      "step": 8555
+    },
+    {
+      "epoch": 0.20365609825764067,
+      "grad_norm": 11.278649942647272,
+      "learning_rate": 2e-05,
+      "loss": 0.7513,
+      "step": 8556
+    },
+    {
+      "epoch": 0.20367990098067218,
+      "grad_norm": 33.832244022921586,
+      "learning_rate": 2e-05,
+      "loss": 0.4921,
+      "step": 8557
+    },
+    {
+      "epoch": 0.2037037037037037,
+      "grad_norm": 83.85452555889037,
+      "learning_rate": 2e-05,
+      "loss": 0.7055,
+      "step": 8558
+    },
+    {
+      "epoch": 0.2037275064267352,
+      "grad_norm": 11.619137837920983,
+      "learning_rate": 2e-05,
+      "loss": 0.927,
+      "step": 8559
+    },
+    {
+      "epoch": 0.20375130914976675,
+      "grad_norm": 5.121797252498004,
+      "learning_rate": 2e-05,
+      "loss": 0.3053,
+      "step": 8560
+    },
+    {
+      "epoch": 0.20377511187279826,
+      "grad_norm": 55.15489561254798,
+      "learning_rate": 2e-05,
+      "loss": 1.6064,
+      "step": 8561
+    },
+    {
+      "epoch": 0.20379891459582977,
+      "grad_norm": 3.979497755038075,
+      "learning_rate": 2e-05,
+      "loss": 0.3083,
+      "step": 8562
+    },
+    {
+      "epoch": 0.20382271731886129,
+      "grad_norm": 19.625637348365277,
+      "learning_rate": 2e-05,
+      "loss": 0.9796,
+      "step": 8563
+    },
+    {
+      "epoch": 0.2038465200418928,
+      "grad_norm": 3.178069983399387,
+      "learning_rate": 2e-05,
+      "loss": 0.2313,
+      "step": 8564
+    },
+    {
+      "epoch": 0.2038703227649243,
+      "grad_norm": 18.026459622620457,
+      "learning_rate": 2e-05,
+      "loss": 1.2256,
+      "step": 8565
+    },
+    {
+      "epoch": 0.20389412548795582,
+      "grad_norm": 1.3715639885966633,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 8566
+    },
+    {
+      "epoch": 0.20391792821098734,
+      "grad_norm": 1.4243659671938922,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 8567
+    },
+    {
+      "epoch": 0.20394173093401885,
+      "grad_norm": 16.10819143784102,
+      "learning_rate": 2e-05,
+      "loss": 0.6958,
+      "step": 8568
+    },
+    {
+      "epoch": 0.20396553365705036,
+      "grad_norm": 16.76572187377339,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 8569
+    },
+    {
+      "epoch": 0.20398933638008188,
+      "grad_norm": 8.304825916454458,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2040131391031134,
+      "grad_norm": 8.13556489001526,
+      "learning_rate": 2e-05,
+      "loss": 0.6628,
+      "step": 8571
+    },
+    {
+      "epoch": 0.2040369418261449,
+      "grad_norm": 6.906792134602876,
+      "learning_rate": 2e-05,
+      "loss": 0.2191,
+      "step": 8572
+    },
+    {
+      "epoch": 0.20406074454917641,
+      "grad_norm": 7.521002747741644,
+      "learning_rate": 2e-05,
+      "loss": 0.3655,
+      "step": 8573
+    },
+    {
+      "epoch": 0.20408454727220793,
+      "grad_norm": 4.993142451230459,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 8574
+    },
+    {
+      "epoch": 0.20410834999523947,
+      "grad_norm": 2.8986214555472216,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 8575
+    },
+    {
+      "epoch": 0.20413215271827098,
+      "grad_norm": 3.092826946959024,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 8576
+    },
+    {
+      "epoch": 0.2041559554413025,
+      "grad_norm": 6.5933751802704545,
+      "learning_rate": 2e-05,
+      "loss": 1.592,
+      "step": 8577
+    },
+    {
+      "epoch": 0.204179758164334,
+      "grad_norm": 26.529885495563562,
+      "learning_rate": 2e-05,
+      "loss": 0.3425,
+      "step": 8578
+    },
+    {
+      "epoch": 0.20420356088736552,
+      "grad_norm": 9.411763014755568,
+      "learning_rate": 2e-05,
+      "loss": 2.1259,
+      "step": 8579
+    },
+    {
+      "epoch": 0.20422736361039703,
+      "grad_norm": 4.2045227327789965,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 8580
+    },
+    {
+      "epoch": 0.20425116633342855,
+      "grad_norm": 4.419486993001725,
+      "learning_rate": 2e-05,
+      "loss": 0.6581,
+      "step": 8581
+    },
+    {
+      "epoch": 0.20427496905646006,
+      "grad_norm": 3.675195101387582,
+      "learning_rate": 2e-05,
+      "loss": 0.3989,
+      "step": 8582
+    },
+    {
+      "epoch": 0.20429877177949157,
+      "grad_norm": 1.4024434671142763,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 8583
+    },
+    {
+      "epoch": 0.20432257450252309,
+      "grad_norm": 1.362510270539315,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 8584
+    },
+    {
+      "epoch": 0.2043463772255546,
+      "grad_norm": 10.355975060187557,
+      "learning_rate": 2e-05,
+      "loss": 1.1353,
+      "step": 8585
+    },
+    {
+      "epoch": 0.2043701799485861,
+      "grad_norm": 5.894504716003414,
+      "learning_rate": 2e-05,
+      "loss": 1.0159,
+      "step": 8586
+    },
+    {
+      "epoch": 0.20439398267161762,
+      "grad_norm": 24.157510582829797,
+      "learning_rate": 2e-05,
+      "loss": 0.9288,
+      "step": 8587
+    },
+    {
+      "epoch": 0.20441778539464914,
+      "grad_norm": 18.137395041533903,
+      "learning_rate": 2e-05,
+      "loss": 0.6558,
+      "step": 8588
+    },
+    {
+      "epoch": 0.20444158811768065,
+      "grad_norm": 3.8430620965777407,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 8589
+    },
+    {
+      "epoch": 0.2044653908407122,
+      "grad_norm": 4.207380983012105,
+      "learning_rate": 2e-05,
+      "loss": 0.728,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2044891935637437,
+      "grad_norm": 6.562551671996522,
+      "learning_rate": 2e-05,
+      "loss": 0.6222,
+      "step": 8591
+    },
+    {
+      "epoch": 0.20451299628677522,
+      "grad_norm": 7.6364725468006,
+      "learning_rate": 2e-05,
+      "loss": 0.6994,
+      "step": 8592
+    },
+    {
+      "epoch": 0.20453679900980673,
+      "grad_norm": 5.900318995972482,
+      "learning_rate": 2e-05,
+      "loss": 0.5867,
+      "step": 8593
+    },
+    {
+      "epoch": 0.20456060173283824,
+      "grad_norm": 8.320598375165234,
+      "learning_rate": 2e-05,
+      "loss": 1.1499,
+      "step": 8594
+    },
+    {
+      "epoch": 0.20458440445586976,
+      "grad_norm": 25.24776216259217,
+      "learning_rate": 2e-05,
+      "loss": 0.2726,
+      "step": 8595
+    },
+    {
+      "epoch": 0.20460820717890127,
+      "grad_norm": 15.375741376984399,
+      "learning_rate": 2e-05,
+      "loss": 0.173,
+      "step": 8596
+    },
+    {
+      "epoch": 0.20463200990193278,
+      "grad_norm": 4.515610433500372,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 8597
+    },
+    {
+      "epoch": 0.2046558126249643,
+      "grad_norm": 6.396210383066805,
+      "learning_rate": 2e-05,
+      "loss": 1.3531,
+      "step": 8598
+    },
+    {
+      "epoch": 0.2046796153479958,
+      "grad_norm": 4.786826928138963,
+      "learning_rate": 2e-05,
+      "loss": 0.3568,
+      "step": 8599
+    },
+    {
+      "epoch": 0.20470341807102732,
+      "grad_norm": 5.79773443021542,
+      "learning_rate": 2e-05,
+      "loss": 1.3596,
+      "step": 8600
+    },
+    {
+      "epoch": 0.20472722079405883,
+      "grad_norm": 4.422167453950814,
+      "learning_rate": 2e-05,
+      "loss": 0.4591,
+      "step": 8601
+    },
+    {
+      "epoch": 0.20475102351709035,
+      "grad_norm": 4.222467706271645,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 8602
+    },
+    {
+      "epoch": 0.20477482624012186,
+      "grad_norm": 6.56338307984827,
+      "learning_rate": 2e-05,
+      "loss": 0.9347,
+      "step": 8603
+    },
+    {
+      "epoch": 0.20479862896315337,
+      "grad_norm": 8.45076574061663,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 8604
+    },
+    {
+      "epoch": 0.20482243168618489,
+      "grad_norm": 17.631074895787,
+      "learning_rate": 2e-05,
+      "loss": 1.4703,
+      "step": 8605
+    },
+    {
+      "epoch": 0.20484623440921643,
+      "grad_norm": 5.092747261261215,
+      "learning_rate": 2e-05,
+      "loss": 0.6908,
+      "step": 8606
+    },
+    {
+      "epoch": 0.20487003713224794,
+      "grad_norm": 5.527250372091122,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 8607
+    },
+    {
+      "epoch": 0.20489383985527945,
+      "grad_norm": 20.284516956967604,
+      "learning_rate": 2e-05,
+      "loss": 1.6903,
+      "step": 8608
+    },
+    {
+      "epoch": 0.20491764257831097,
+      "grad_norm": 6.854942053101871,
+      "learning_rate": 2e-05,
+      "loss": 0.2988,
+      "step": 8609
+    },
+    {
+      "epoch": 0.20494144530134248,
+      "grad_norm": 4.926054786959917,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 8610
+    },
+    {
+      "epoch": 0.204965248024374,
+      "grad_norm": 5.782776463986279,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 8611
+    },
+    {
+      "epoch": 0.2049890507474055,
+      "grad_norm": 4.4143347520326115,
+      "learning_rate": 2e-05,
+      "loss": 0.4215,
+      "step": 8612
+    },
+    {
+      "epoch": 0.20501285347043702,
+      "grad_norm": 22.42259759739996,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 8613
+    },
+    {
+      "epoch": 0.20503665619346853,
+      "grad_norm": 3.182158561688208,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 8614
+    },
+    {
+      "epoch": 0.20506045891650004,
+      "grad_norm": 7.993863517303273,
+      "learning_rate": 2e-05,
+      "loss": 0.8997,
+      "step": 8615
+    },
+    {
+      "epoch": 0.20508426163953156,
+      "grad_norm": 6.108487275471643,
+      "learning_rate": 2e-05,
+      "loss": 0.844,
+      "step": 8616
+    },
+    {
+      "epoch": 0.20510806436256307,
+      "grad_norm": 1.8094006169455097,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 8617
+    },
+    {
+      "epoch": 0.20513186708559458,
+      "grad_norm": 5.266633957704451,
+      "learning_rate": 2e-05,
+      "loss": 1.0024,
+      "step": 8618
+    },
+    {
+      "epoch": 0.2051556698086261,
+      "grad_norm": 3.849782017291982,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 8619
+    },
+    {
+      "epoch": 0.2051794725316576,
+      "grad_norm": 2.7206160007696383,
+      "learning_rate": 2e-05,
+      "loss": 0.3475,
+      "step": 8620
+    },
+    {
+      "epoch": 0.20520327525468915,
+      "grad_norm": 2.366275365486942,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 8621
+    },
+    {
+      "epoch": 0.20522707797772066,
+      "grad_norm": 4.524964870451096,
+      "learning_rate": 2e-05,
+      "loss": 0.6057,
+      "step": 8622
+    },
+    {
+      "epoch": 0.20525088070075218,
+      "grad_norm": 7.76669465727215,
+      "learning_rate": 2e-05,
+      "loss": 1.1226,
+      "step": 8623
+    },
+    {
+      "epoch": 0.2052746834237837,
+      "grad_norm": 3.3363419182961036,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 8624
+    },
+    {
+      "epoch": 0.2052984861468152,
+      "grad_norm": 7.900304617084714,
+      "learning_rate": 2e-05,
+      "loss": 0.2712,
+      "step": 8625
+    },
+    {
+      "epoch": 0.20532228886984671,
+      "grad_norm": 6.813114666612476,
+      "learning_rate": 2e-05,
+      "loss": 0.5148,
+      "step": 8626
+    },
+    {
+      "epoch": 0.20534609159287823,
+      "grad_norm": 1.3436313819834291,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 8627
+    },
+    {
+      "epoch": 0.20536989431590974,
+      "grad_norm": 1.8448437907221158,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 8628
+    },
+    {
+      "epoch": 0.20539369703894125,
+      "grad_norm": 12.060009282116736,
+      "learning_rate": 2e-05,
+      "loss": 0.3108,
+      "step": 8629
+    },
+    {
+      "epoch": 0.20541749976197277,
+      "grad_norm": 8.86590591619386,
+      "learning_rate": 2e-05,
+      "loss": 2.5304,
+      "step": 8630
+    },
+    {
+      "epoch": 0.20544130248500428,
+      "grad_norm": 5.028361608100928,
+      "learning_rate": 2e-05,
+      "loss": 0.7728,
+      "step": 8631
+    },
+    {
+      "epoch": 0.2054651052080358,
+      "grad_norm": 6.041781457298605,
+      "learning_rate": 2e-05,
+      "loss": 0.3891,
+      "step": 8632
+    },
+    {
+      "epoch": 0.2054889079310673,
+      "grad_norm": 4.508378549902962,
+      "learning_rate": 2e-05,
+      "loss": 0.5136,
+      "step": 8633
+    },
+    {
+      "epoch": 0.20551271065409882,
+      "grad_norm": 7.458647669159358,
+      "learning_rate": 2e-05,
+      "loss": 0.789,
+      "step": 8634
+    },
+    {
+      "epoch": 0.20553651337713033,
+      "grad_norm": 20.74069585148764,
+      "learning_rate": 2e-05,
+      "loss": 1.445,
+      "step": 8635
+    },
+    {
+      "epoch": 0.20556031610016187,
+      "grad_norm": 3.6063704462378876,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 8636
+    },
+    {
+      "epoch": 0.20558411882319338,
+      "grad_norm": 3.3868403543222185,
+      "learning_rate": 2e-05,
+      "loss": 0.4672,
+      "step": 8637
+    },
+    {
+      "epoch": 0.2056079215462249,
+      "grad_norm": 2.1653154461445263,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 8638
+    },
+    {
+      "epoch": 0.2056317242692564,
+      "grad_norm": 3.4212589652515857,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 8639
+    },
+    {
+      "epoch": 0.20565552699228792,
+      "grad_norm": 3.200711739315612,
+      "learning_rate": 2e-05,
+      "loss": 0.1782,
+      "step": 8640
+    },
+    {
+      "epoch": 0.20567932971531944,
+      "grad_norm": 0.7113222073049765,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 8641
+    },
+    {
+      "epoch": 0.20570313243835095,
+      "grad_norm": 9.38220294731782,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 8642
+    },
+    {
+      "epoch": 0.20572693516138246,
+      "grad_norm": 2.330092464900674,
+      "learning_rate": 2e-05,
+      "loss": 0.1635,
+      "step": 8643
+    },
+    {
+      "epoch": 0.20575073788441398,
+      "grad_norm": 3.1116447767177875,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 8644
+    },
+    {
+      "epoch": 0.2057745406074455,
+      "grad_norm": 2.718821012357315,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 8645
+    },
+    {
+      "epoch": 0.205798343330477,
+      "grad_norm": 21.779809315467187,
+      "learning_rate": 2e-05,
+      "loss": 0.44,
+      "step": 8646
+    },
+    {
+      "epoch": 0.20582214605350851,
+      "grad_norm": 4.887387708424819,
+      "learning_rate": 2e-05,
+      "loss": 0.2965,
+      "step": 8647
+    },
+    {
+      "epoch": 0.20584594877654003,
+      "grad_norm": 3.672138900659049,
+      "learning_rate": 2e-05,
+      "loss": 0.31,
+      "step": 8648
+    },
+    {
+      "epoch": 0.20586975149957154,
+      "grad_norm": 13.6967691941106,
+      "learning_rate": 2e-05,
+      "loss": 1.2249,
+      "step": 8649
+    },
+    {
+      "epoch": 0.20589355422260305,
+      "grad_norm": 5.794454574469325,
+      "learning_rate": 2e-05,
+      "loss": 0.6091,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2059173569456346,
+      "grad_norm": 16.79383392174819,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 8651
+    },
+    {
+      "epoch": 0.2059411596686661,
+      "grad_norm": 11.820545622684554,
+      "learning_rate": 2e-05,
+      "loss": 0.4843,
+      "step": 8652
+    },
+    {
+      "epoch": 0.20596496239169762,
+      "grad_norm": 5.703070175977921,
+      "learning_rate": 2e-05,
+      "loss": 0.8021,
+      "step": 8653
+    },
+    {
+      "epoch": 0.20598876511472913,
+      "grad_norm": 2.853504380507949,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 8654
+    },
+    {
+      "epoch": 0.20601256783776065,
+      "grad_norm": 5.135131317087874,
+      "learning_rate": 2e-05,
+      "loss": 0.5999,
+      "step": 8655
+    },
+    {
+      "epoch": 0.20603637056079216,
+      "grad_norm": 10.785980553667628,
+      "learning_rate": 2e-05,
+      "loss": 0.9026,
+      "step": 8656
+    },
+    {
+      "epoch": 0.20606017328382367,
+      "grad_norm": 7.757373013171973,
+      "learning_rate": 2e-05,
+      "loss": 0.6777,
+      "step": 8657
+    },
+    {
+      "epoch": 0.20608397600685519,
+      "grad_norm": 6.169233969973754,
+      "learning_rate": 2e-05,
+      "loss": 0.8488,
+      "step": 8658
+    },
+    {
+      "epoch": 0.2061077787298867,
+      "grad_norm": 1.924836702734557,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 8659
+    },
+    {
+      "epoch": 0.2061315814529182,
+      "grad_norm": 24.160731697461223,
+      "learning_rate": 2e-05,
+      "loss": 1.9253,
+      "step": 8660
+    },
+    {
+      "epoch": 0.20615538417594972,
+      "grad_norm": 2.6621536167892375,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 8661
+    },
+    {
+      "epoch": 0.20617918689898124,
+      "grad_norm": 7.136952829786485,
+      "learning_rate": 2e-05,
+      "loss": 0.3704,
+      "step": 8662
+    },
+    {
+      "epoch": 0.20620298962201275,
+      "grad_norm": 2.5705305775213008,
+      "learning_rate": 2e-05,
+      "loss": 0.225,
+      "step": 8663
+    },
+    {
+      "epoch": 0.20622679234504426,
+      "grad_norm": 6.259575891702202,
+      "learning_rate": 2e-05,
+      "loss": 0.8494,
+      "step": 8664
+    },
+    {
+      "epoch": 0.20625059506807578,
+      "grad_norm": 3.0360318131140125,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 8665
+    },
+    {
+      "epoch": 0.2062743977911073,
+      "grad_norm": 2.0146922721531375,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 8666
+    },
+    {
+      "epoch": 0.20629820051413883,
+      "grad_norm": 2.2814150082133455,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 8667
+    },
+    {
+      "epoch": 0.20632200323717034,
+      "grad_norm": 3.9460851426590566,
+      "learning_rate": 2e-05,
+      "loss": 0.3994,
+      "step": 8668
+    },
+    {
+      "epoch": 0.20634580596020186,
+      "grad_norm": 1.892016495561552,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 8669
+    },
+    {
+      "epoch": 0.20636960868323337,
+      "grad_norm": 3.2309121197183788,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 8670
+    },
+    {
+      "epoch": 0.20639341140626488,
+      "grad_norm": 13.06402133425595,
+      "learning_rate": 2e-05,
+      "loss": 1.1616,
+      "step": 8671
+    },
+    {
+      "epoch": 0.2064172141292964,
+      "grad_norm": 13.013302651661055,
+      "learning_rate": 2e-05,
+      "loss": 0.4799,
+      "step": 8672
+    },
+    {
+      "epoch": 0.2064410168523279,
+      "grad_norm": 2.396019173534902,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 8673
+    },
+    {
+      "epoch": 0.20646481957535942,
+      "grad_norm": 10.250812842881887,
+      "learning_rate": 2e-05,
+      "loss": 0.5793,
+      "step": 8674
+    },
+    {
+      "epoch": 0.20648862229839093,
+      "grad_norm": 0.949558031424507,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 8675
+    },
+    {
+      "epoch": 0.20651242502142245,
+      "grad_norm": 4.8020775003127785,
+      "learning_rate": 2e-05,
+      "loss": 0.6809,
+      "step": 8676
+    },
+    {
+      "epoch": 0.20653622774445396,
+      "grad_norm": 17.51517810716808,
+      "learning_rate": 2e-05,
+      "loss": 1.0199,
+      "step": 8677
+    },
+    {
+      "epoch": 0.20656003046748547,
+      "grad_norm": 3.6667197737286346,
+      "learning_rate": 2e-05,
+      "loss": 0.4103,
+      "step": 8678
+    },
+    {
+      "epoch": 0.20658383319051699,
+      "grad_norm": 1.469127194965033,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 8679
+    },
+    {
+      "epoch": 0.2066076359135485,
+      "grad_norm": 5.718162143376884,
+      "learning_rate": 2e-05,
+      "loss": 0.7725,
+      "step": 8680
+    },
+    {
+      "epoch": 0.20663143863658,
+      "grad_norm": 3.8287072254827956,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 8681
+    },
+    {
+      "epoch": 0.20665524135961155,
+      "grad_norm": 3.236387598871698,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 8682
+    },
+    {
+      "epoch": 0.20667904408264307,
+      "grad_norm": 4.202994822550122,
+      "learning_rate": 2e-05,
+      "loss": 0.3808,
+      "step": 8683
+    },
+    {
+      "epoch": 0.20670284680567458,
+      "grad_norm": 8.62724002273874,
+      "learning_rate": 2e-05,
+      "loss": 0.709,
+      "step": 8684
+    },
+    {
+      "epoch": 0.2067266495287061,
+      "grad_norm": 3.6271359799591827,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 8685
+    },
+    {
+      "epoch": 0.2067504522517376,
+      "grad_norm": 3.697745256803544,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 8686
+    },
+    {
+      "epoch": 0.20677425497476912,
+      "grad_norm": 13.089569395312687,
+      "learning_rate": 2e-05,
+      "loss": 0.4315,
+      "step": 8687
+    },
+    {
+      "epoch": 0.20679805769780063,
+      "grad_norm": 210.2112146828022,
+      "learning_rate": 2e-05,
+      "loss": 0.861,
+      "step": 8688
+    },
+    {
+      "epoch": 0.20682186042083214,
+      "grad_norm": 5.804956071381393,
+      "learning_rate": 2e-05,
+      "loss": 0.265,
+      "step": 8689
+    },
+    {
+      "epoch": 0.20684566314386366,
+      "grad_norm": 3.431424736439423,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 8690
+    },
+    {
+      "epoch": 0.20686946586689517,
+      "grad_norm": 7.507752669600362,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 8691
+    },
+    {
+      "epoch": 0.20689326858992668,
+      "grad_norm": 5.7960140391782335,
+      "learning_rate": 2e-05,
+      "loss": 0.4109,
+      "step": 8692
+    },
+    {
+      "epoch": 0.2069170713129582,
+      "grad_norm": 1.6882647410158829,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 8693
+    },
+    {
+      "epoch": 0.2069408740359897,
+      "grad_norm": 3.2831708139454503,
+      "learning_rate": 2e-05,
+      "loss": 0.2407,
+      "step": 8694
+    },
+    {
+      "epoch": 0.20696467675902122,
+      "grad_norm": 3.4371022024138735,
+      "learning_rate": 2e-05,
+      "loss": 0.2572,
+      "step": 8695
+    },
+    {
+      "epoch": 0.20698847948205273,
+      "grad_norm": 4.36630590378073,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 8696
+    },
+    {
+      "epoch": 0.20701228220508427,
+      "grad_norm": 15.412294045030533,
+      "learning_rate": 2e-05,
+      "loss": 0.7497,
+      "step": 8697
+    },
+    {
+      "epoch": 0.2070360849281158,
+      "grad_norm": 5.919295185751195,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 8698
+    },
+    {
+      "epoch": 0.2070598876511473,
+      "grad_norm": 4.380414744365118,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 8699
+    },
+    {
+      "epoch": 0.2070836903741788,
+      "grad_norm": 7.631909045498375,
+      "learning_rate": 2e-05,
+      "loss": 0.6177,
+      "step": 8700
+    },
+    {
+      "epoch": 0.20710749309721033,
+      "grad_norm": 3.5653431720672746,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 8701
+    },
+    {
+      "epoch": 0.20713129582024184,
+      "grad_norm": 6.755555647010498,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 8702
+    },
+    {
+      "epoch": 0.20715509854327335,
+      "grad_norm": 6.891551898815389,
+      "learning_rate": 2e-05,
+      "loss": 0.7297,
+      "step": 8703
+    },
+    {
+      "epoch": 0.20717890126630487,
+      "grad_norm": 1.7486658845324348,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 8704
+    },
+    {
+      "epoch": 0.20720270398933638,
+      "grad_norm": 7.850421322392905,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 8705
+    },
+    {
+      "epoch": 0.2072265067123679,
+      "grad_norm": 7.701360427294983,
+      "learning_rate": 2e-05,
+      "loss": 0.4016,
+      "step": 8706
+    },
+    {
+      "epoch": 0.2072503094353994,
+      "grad_norm": 5.016063550792986,
+      "learning_rate": 2e-05,
+      "loss": 0.4887,
+      "step": 8707
+    },
+    {
+      "epoch": 0.20727411215843092,
+      "grad_norm": 0.8612266252059176,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 8708
+    },
+    {
+      "epoch": 0.20729791488146243,
+      "grad_norm": 2.363146943137617,
+      "learning_rate": 2e-05,
+      "loss": 0.1993,
+      "step": 8709
+    },
+    {
+      "epoch": 0.20732171760449394,
+      "grad_norm": 6.9581092342606725,
+      "learning_rate": 2e-05,
+      "loss": 0.4603,
+      "step": 8710
+    },
+    {
+      "epoch": 0.20734552032752546,
+      "grad_norm": 9.694368483450278,
+      "learning_rate": 2e-05,
+      "loss": 0.4711,
+      "step": 8711
+    },
+    {
+      "epoch": 0.207369323050557,
+      "grad_norm": 2.084331987544914,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 8712
+    },
+    {
+      "epoch": 0.2073931257735885,
+      "grad_norm": 2.3682203779435356,
+      "learning_rate": 2e-05,
+      "loss": 0.1515,
+      "step": 8713
+    },
+    {
+      "epoch": 0.20741692849662002,
+      "grad_norm": 14.65213255753737,
+      "learning_rate": 2e-05,
+      "loss": 0.3706,
+      "step": 8714
+    },
+    {
+      "epoch": 0.20744073121965154,
+      "grad_norm": 6.6263946735615304,
+      "learning_rate": 2e-05,
+      "loss": 0.745,
+      "step": 8715
+    },
+    {
+      "epoch": 0.20746453394268305,
+      "grad_norm": 4.8344545280028,
+      "learning_rate": 2e-05,
+      "loss": 0.8755,
+      "step": 8716
+    },
+    {
+      "epoch": 0.20748833666571456,
+      "grad_norm": 6.974926509825277,
+      "learning_rate": 2e-05,
+      "loss": 0.332,
+      "step": 8717
+    },
+    {
+      "epoch": 0.20751213938874608,
+      "grad_norm": 1.5183395363881254,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 8718
+    },
+    {
+      "epoch": 0.2075359421117776,
+      "grad_norm": 5.764906613417653,
+      "learning_rate": 2e-05,
+      "loss": 0.2998,
+      "step": 8719
+    },
+    {
+      "epoch": 0.2075597448348091,
+      "grad_norm": 6.6546549982750145,
+      "learning_rate": 2e-05,
+      "loss": 0.8272,
+      "step": 8720
+    },
+    {
+      "epoch": 0.20758354755784061,
+      "grad_norm": 7.193309763026134,
+      "learning_rate": 2e-05,
+      "loss": 1.0231,
+      "step": 8721
+    },
+    {
+      "epoch": 0.20760735028087213,
+      "grad_norm": 4.571318744674314,
+      "learning_rate": 2e-05,
+      "loss": 0.3086,
+      "step": 8722
+    },
+    {
+      "epoch": 0.20763115300390364,
+      "grad_norm": 2.2040899383940724,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 8723
+    },
+    {
+      "epoch": 0.20765495572693515,
+      "grad_norm": 0.3779163264394938,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 8724
+    },
+    {
+      "epoch": 0.20767875844996667,
+      "grad_norm": 3.909538445648716,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 8725
+    },
+    {
+      "epoch": 0.20770256117299818,
+      "grad_norm": 4.7921203603305615,
+      "learning_rate": 2e-05,
+      "loss": 0.782,
+      "step": 8726
+    },
+    {
+      "epoch": 0.2077263638960297,
+      "grad_norm": 7.274140933485529,
+      "learning_rate": 2e-05,
+      "loss": 0.7014,
+      "step": 8727
+    },
+    {
+      "epoch": 0.20775016661906123,
+      "grad_norm": 6.8523159029481135,
+      "learning_rate": 2e-05,
+      "loss": 0.3558,
+      "step": 8728
+    },
+    {
+      "epoch": 0.20777396934209275,
+      "grad_norm": 2.4773432569471225,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 8729
+    },
+    {
+      "epoch": 0.20779777206512426,
+      "grad_norm": 2.5190983852564575,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 8730
+    },
+    {
+      "epoch": 0.20782157478815577,
+      "grad_norm": 7.098546721848233,
+      "learning_rate": 2e-05,
+      "loss": 0.8718,
+      "step": 8731
+    },
+    {
+      "epoch": 0.20784537751118728,
+      "grad_norm": 7.162741215861373,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 8732
+    },
+    {
+      "epoch": 0.2078691802342188,
+      "grad_norm": 7.051181078265853,
+      "learning_rate": 2e-05,
+      "loss": 0.3932,
+      "step": 8733
+    },
+    {
+      "epoch": 0.2078929829572503,
+      "grad_norm": 2.5181757411650323,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 8734
+    },
+    {
+      "epoch": 0.20791678568028182,
+      "grad_norm": 6.04747330704258,
+      "learning_rate": 2e-05,
+      "loss": 0.3717,
+      "step": 8735
+    },
+    {
+      "epoch": 0.20794058840331334,
+      "grad_norm": 5.695673699380661,
+      "learning_rate": 2e-05,
+      "loss": 0.2955,
+      "step": 8736
+    },
+    {
+      "epoch": 0.20796439112634485,
+      "grad_norm": 3.4090508999223608,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 8737
+    },
+    {
+      "epoch": 0.20798819384937636,
+      "grad_norm": 3.084718352109773,
+      "learning_rate": 2e-05,
+      "loss": 0.1193,
+      "step": 8738
+    },
+    {
+      "epoch": 0.20801199657240788,
+      "grad_norm": 1.7373813502976438,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 8739
+    },
+    {
+      "epoch": 0.2080357992954394,
+      "grad_norm": 21.500148905448274,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2080596020184709,
+      "grad_norm": 8.28845554344737,
+      "learning_rate": 2e-05,
+      "loss": 1.0994,
+      "step": 8741
+    },
+    {
+      "epoch": 0.20808340474150241,
+      "grad_norm": 7.242594798074125,
+      "learning_rate": 2e-05,
+      "loss": 0.3105,
+      "step": 8742
+    },
+    {
+      "epoch": 0.20810720746453396,
+      "grad_norm": 2.9543257528929896,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 8743
+    },
+    {
+      "epoch": 0.20813101018756547,
+      "grad_norm": 4.0400677857241,
+      "learning_rate": 2e-05,
+      "loss": 0.2441,
+      "step": 8744
+    },
+    {
+      "epoch": 0.20815481291059698,
+      "grad_norm": 8.845074471612472,
+      "learning_rate": 2e-05,
+      "loss": 1.0739,
+      "step": 8745
+    },
+    {
+      "epoch": 0.2081786156336285,
+      "grad_norm": 6.119675859475063,
+      "learning_rate": 2e-05,
+      "loss": 0.2824,
+      "step": 8746
+    },
+    {
+      "epoch": 0.20820241835666,
+      "grad_norm": 6.747144973652265,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 8747
+    },
+    {
+      "epoch": 0.20822622107969152,
+      "grad_norm": 5.635697569541378,
+      "learning_rate": 2e-05,
+      "loss": 0.6417,
+      "step": 8748
+    },
+    {
+      "epoch": 0.20825002380272303,
+      "grad_norm": 2.6736539531371384,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 8749
+    },
+    {
+      "epoch": 0.20827382652575455,
+      "grad_norm": 4.799707026438669,
+      "learning_rate": 2e-05,
+      "loss": 0.2577,
+      "step": 8750
+    },
+    {
+      "epoch": 0.20829762924878606,
+      "grad_norm": 1.8362627988566351,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 8751
+    },
+    {
+      "epoch": 0.20832143197181757,
+      "grad_norm": 4.113805806715907,
+      "learning_rate": 2e-05,
+      "loss": 0.2786,
+      "step": 8752
+    },
+    {
+      "epoch": 0.20834523469484909,
+      "grad_norm": 9.51797880865214,
+      "learning_rate": 2e-05,
+      "loss": 0.4583,
+      "step": 8753
+    },
+    {
+      "epoch": 0.2083690374178806,
+      "grad_norm": 7.613061900213029,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 8754
+    },
+    {
+      "epoch": 0.2083928401409121,
+      "grad_norm": 4.347353600248249,
+      "learning_rate": 2e-05,
+      "loss": 0.681,
+      "step": 8755
+    },
+    {
+      "epoch": 0.20841664286394362,
+      "grad_norm": 12.190079307016624,
+      "learning_rate": 2e-05,
+      "loss": 0.945,
+      "step": 8756
+    },
+    {
+      "epoch": 0.20844044558697514,
+      "grad_norm": 6.038252113683946,
+      "learning_rate": 2e-05,
+      "loss": 1.1322,
+      "step": 8757
+    },
+    {
+      "epoch": 0.20846424831000668,
+      "grad_norm": 8.550529670034711,
+      "learning_rate": 2e-05,
+      "loss": 0.3562,
+      "step": 8758
+    },
+    {
+      "epoch": 0.2084880510330382,
+      "grad_norm": 9.118056267771626,
+      "learning_rate": 2e-05,
+      "loss": 0.2941,
+      "step": 8759
+    },
+    {
+      "epoch": 0.2085118537560697,
+      "grad_norm": 8.761781547015216,
+      "learning_rate": 2e-05,
+      "loss": 1.6467,
+      "step": 8760
+    },
+    {
+      "epoch": 0.20853565647910122,
+      "grad_norm": 5.077196929733028,
+      "learning_rate": 2e-05,
+      "loss": 0.7065,
+      "step": 8761
+    },
+    {
+      "epoch": 0.20855945920213273,
+      "grad_norm": 3.766870034702855,
+      "learning_rate": 2e-05,
+      "loss": 0.4498,
+      "step": 8762
+    },
+    {
+      "epoch": 0.20858326192516424,
+      "grad_norm": 1.3637745698436494,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 8763
+    },
+    {
+      "epoch": 0.20860706464819576,
+      "grad_norm": 2.4022306166756002,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 8764
+    },
+    {
+      "epoch": 0.20863086737122727,
+      "grad_norm": 6.246165006277148,
+      "learning_rate": 2e-05,
+      "loss": 0.7774,
+      "step": 8765
+    },
+    {
+      "epoch": 0.20865467009425878,
+      "grad_norm": 8.940164940445415,
+      "learning_rate": 2e-05,
+      "loss": 1.5931,
+      "step": 8766
+    },
+    {
+      "epoch": 0.2086784728172903,
+      "grad_norm": 1.825529429425213,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 8767
+    },
+    {
+      "epoch": 0.2087022755403218,
+      "grad_norm": 7.525471902913651,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 8768
+    },
+    {
+      "epoch": 0.20872607826335332,
+      "grad_norm": 6.561994514753461,
+      "learning_rate": 2e-05,
+      "loss": 0.2574,
+      "step": 8769
+    },
+    {
+      "epoch": 0.20874988098638483,
+      "grad_norm": 8.340189364010886,
+      "learning_rate": 2e-05,
+      "loss": 0.473,
+      "step": 8770
+    },
+    {
+      "epoch": 0.20877368370941635,
+      "grad_norm": 3.3775877991670225,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 8771
+    },
+    {
+      "epoch": 0.20879748643244786,
+      "grad_norm": 1.596725354782229,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 8772
+    },
+    {
+      "epoch": 0.2088212891554794,
+      "grad_norm": 7.807129766706653,
+      "learning_rate": 2e-05,
+      "loss": 1.1141,
+      "step": 8773
+    },
+    {
+      "epoch": 0.2088450918785109,
+      "grad_norm": 7.1946191901529355,
+      "learning_rate": 2e-05,
+      "loss": 1.3511,
+      "step": 8774
+    },
+    {
+      "epoch": 0.20886889460154243,
+      "grad_norm": 5.378197661497347,
+      "learning_rate": 2e-05,
+      "loss": 0.4341,
+      "step": 8775
+    },
+    {
+      "epoch": 0.20889269732457394,
+      "grad_norm": 2.93827849779559,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 8776
+    },
+    {
+      "epoch": 0.20891650004760545,
+      "grad_norm": 4.657475913777423,
+      "learning_rate": 2e-05,
+      "loss": 0.2354,
+      "step": 8777
+    },
+    {
+      "epoch": 0.20894030277063697,
+      "grad_norm": 4.397177016355802,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 8778
+    },
+    {
+      "epoch": 0.20896410549366848,
+      "grad_norm": 6.026271971330118,
+      "learning_rate": 2e-05,
+      "loss": 0.2705,
+      "step": 8779
+    },
+    {
+      "epoch": 0.2089879082167,
+      "grad_norm": 6.80410789753539,
+      "learning_rate": 2e-05,
+      "loss": 0.893,
+      "step": 8780
+    },
+    {
+      "epoch": 0.2090117109397315,
+      "grad_norm": 7.747070130410736,
+      "learning_rate": 2e-05,
+      "loss": 1.1604,
+      "step": 8781
+    },
+    {
+      "epoch": 0.20903551366276302,
+      "grad_norm": 3.30154774839961,
+      "learning_rate": 2e-05,
+      "loss": 0.6395,
+      "step": 8782
+    },
+    {
+      "epoch": 0.20905931638579453,
+      "grad_norm": 3.209151507505241,
+      "learning_rate": 2e-05,
+      "loss": 0.2937,
+      "step": 8783
+    },
+    {
+      "epoch": 0.20908311910882604,
+      "grad_norm": 3.486337050378004,
+      "learning_rate": 2e-05,
+      "loss": 0.715,
+      "step": 8784
+    },
+    {
+      "epoch": 0.20910692183185756,
+      "grad_norm": 6.411793214819874,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 8785
+    },
+    {
+      "epoch": 0.20913072455488907,
+      "grad_norm": 6.810359612737128,
+      "learning_rate": 2e-05,
+      "loss": 1.0521,
+      "step": 8786
+    },
+    {
+      "epoch": 0.20915452727792058,
+      "grad_norm": 6.68014968576554,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 8787
+    },
+    {
+      "epoch": 0.2091783300009521,
+      "grad_norm": 7.716784870543493,
+      "learning_rate": 2e-05,
+      "loss": 0.8588,
+      "step": 8788
+    },
+    {
+      "epoch": 0.20920213272398364,
+      "grad_norm": 2.4659035279893105,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 8789
+    },
+    {
+      "epoch": 0.20922593544701515,
+      "grad_norm": 3.5365824564549984,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 8790
+    },
+    {
+      "epoch": 0.20924973817004666,
+      "grad_norm": 8.957633117693522,
+      "learning_rate": 2e-05,
+      "loss": 0.922,
+      "step": 8791
+    },
+    {
+      "epoch": 0.20927354089307817,
+      "grad_norm": 6.0249234151594075,
+      "learning_rate": 2e-05,
+      "loss": 0.2553,
+      "step": 8792
+    },
+    {
+      "epoch": 0.2092973436161097,
+      "grad_norm": 2.984223057979382,
+      "learning_rate": 2e-05,
+      "loss": 0.3769,
+      "step": 8793
+    },
+    {
+      "epoch": 0.2093211463391412,
+      "grad_norm": 4.913011525715766,
+      "learning_rate": 2e-05,
+      "loss": 0.196,
+      "step": 8794
+    },
+    {
+      "epoch": 0.2093449490621727,
+      "grad_norm": 3.658573851086268,
+      "learning_rate": 2e-05,
+      "loss": 0.7223,
+      "step": 8795
+    },
+    {
+      "epoch": 0.20936875178520423,
+      "grad_norm": 7.12374408130211,
+      "learning_rate": 2e-05,
+      "loss": 1.1756,
+      "step": 8796
+    },
+    {
+      "epoch": 0.20939255450823574,
+      "grad_norm": 6.1452818666961315,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 8797
+    },
+    {
+      "epoch": 0.20941635723126725,
+      "grad_norm": 4.348764381311173,
+      "learning_rate": 2e-05,
+      "loss": 0.7022,
+      "step": 8798
+    },
+    {
+      "epoch": 0.20944015995429877,
+      "grad_norm": 5.134821038959485,
+      "learning_rate": 2e-05,
+      "loss": 0.7483,
+      "step": 8799
+    },
+    {
+      "epoch": 0.20946396267733028,
+      "grad_norm": 2.186011788303218,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 8800
+    },
+    {
+      "epoch": 0.2094877654003618,
+      "grad_norm": 16.153602879551247,
+      "learning_rate": 2e-05,
+      "loss": 0.8546,
+      "step": 8801
+    },
+    {
+      "epoch": 0.2095115681233933,
+      "grad_norm": 6.5178653984460375,
+      "learning_rate": 2e-05,
+      "loss": 1.077,
+      "step": 8802
+    },
+    {
+      "epoch": 0.20953537084642482,
+      "grad_norm": 3.8383299417735217,
+      "learning_rate": 2e-05,
+      "loss": 0.2451,
+      "step": 8803
+    },
+    {
+      "epoch": 0.20955917356945636,
+      "grad_norm": 5.564368065614468,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 8804
+    },
+    {
+      "epoch": 0.20958297629248787,
+      "grad_norm": 5.266348215860949,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 8805
+    },
+    {
+      "epoch": 0.20960677901551938,
+      "grad_norm": 15.007637992533892,
+      "learning_rate": 2e-05,
+      "loss": 0.522,
+      "step": 8806
+    },
+    {
+      "epoch": 0.2096305817385509,
+      "grad_norm": 5.772860192556461,
+      "learning_rate": 2e-05,
+      "loss": 0.3961,
+      "step": 8807
+    },
+    {
+      "epoch": 0.2096543844615824,
+      "grad_norm": 2.887083103788737,
+      "learning_rate": 2e-05,
+      "loss": 0.347,
+      "step": 8808
+    },
+    {
+      "epoch": 0.20967818718461392,
+      "grad_norm": 3.9001208800316096,
+      "learning_rate": 2e-05,
+      "loss": 0.4011,
+      "step": 8809
+    },
+    {
+      "epoch": 0.20970198990764544,
+      "grad_norm": 2.460071348238291,
+      "learning_rate": 2e-05,
+      "loss": 0.1106,
+      "step": 8810
+    },
+    {
+      "epoch": 0.20972579263067695,
+      "grad_norm": 4.819776113026093,
+      "learning_rate": 2e-05,
+      "loss": 0.937,
+      "step": 8811
+    },
+    {
+      "epoch": 0.20974959535370846,
+      "grad_norm": 4.167264020937461,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 8812
+    },
+    {
+      "epoch": 0.20977339807673998,
+      "grad_norm": 3.6870541268790618,
+      "learning_rate": 2e-05,
+      "loss": 0.1519,
+      "step": 8813
+    },
+    {
+      "epoch": 0.2097972007997715,
+      "grad_norm": 3.076464103492494,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 8814
+    },
+    {
+      "epoch": 0.209821003522803,
+      "grad_norm": 1.2043884184528308,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 8815
+    },
+    {
+      "epoch": 0.20984480624583451,
+      "grad_norm": 5.535961843444119,
+      "learning_rate": 2e-05,
+      "loss": 0.5959,
+      "step": 8816
+    },
+    {
+      "epoch": 0.20986860896886603,
+      "grad_norm": 1.5380526124900429,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 8817
+    },
+    {
+      "epoch": 0.20989241169189754,
+      "grad_norm": 3.2318751316534633,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 8818
+    },
+    {
+      "epoch": 0.20991621441492908,
+      "grad_norm": 5.64757944426593,
+      "learning_rate": 2e-05,
+      "loss": 0.9049,
+      "step": 8819
+    },
+    {
+      "epoch": 0.2099400171379606,
+      "grad_norm": 2.1113486603467333,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 8820
+    },
+    {
+      "epoch": 0.2099638198609921,
+      "grad_norm": 14.606294625465603,
+      "learning_rate": 2e-05,
+      "loss": 1.1659,
+      "step": 8821
+    },
+    {
+      "epoch": 0.20998762258402362,
+      "grad_norm": 7.223805813843644,
+      "learning_rate": 2e-05,
+      "loss": 1.099,
+      "step": 8822
+    },
+    {
+      "epoch": 0.21001142530705513,
+      "grad_norm": 4.864231857314785,
+      "learning_rate": 2e-05,
+      "loss": 0.7797,
+      "step": 8823
+    },
+    {
+      "epoch": 0.21003522803008665,
+      "grad_norm": 1.3618541766387144,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 8824
+    },
+    {
+      "epoch": 0.21005903075311816,
+      "grad_norm": 7.442011883652388,
+      "learning_rate": 2e-05,
+      "loss": 1.6724,
+      "step": 8825
+    },
+    {
+      "epoch": 0.21008283347614967,
+      "grad_norm": 1.8970917498891295,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 8826
+    },
+    {
+      "epoch": 0.21010663619918118,
+      "grad_norm": 24.132384570355164,
+      "learning_rate": 2e-05,
+      "loss": 0.629,
+      "step": 8827
+    },
+    {
+      "epoch": 0.2101304389222127,
+      "grad_norm": 2.642198088366782,
+      "learning_rate": 2e-05,
+      "loss": 0.1551,
+      "step": 8828
+    },
+    {
+      "epoch": 0.2101542416452442,
+      "grad_norm": 6.021207998251858,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 8829
+    },
+    {
+      "epoch": 0.21017804436827572,
+      "grad_norm": 6.978139586836554,
+      "learning_rate": 2e-05,
+      "loss": 1.6867,
+      "step": 8830
+    },
+    {
+      "epoch": 0.21020184709130724,
+      "grad_norm": 4.223382346020937,
+      "learning_rate": 2e-05,
+      "loss": 0.5798,
+      "step": 8831
+    },
+    {
+      "epoch": 0.21022564981433875,
+      "grad_norm": 4.098632321640452,
+      "learning_rate": 2e-05,
+      "loss": 0.266,
+      "step": 8832
+    },
+    {
+      "epoch": 0.21024945253737026,
+      "grad_norm": 7.723281982992292,
+      "learning_rate": 2e-05,
+      "loss": 1.2456,
+      "step": 8833
+    },
+    {
+      "epoch": 0.2102732552604018,
+      "grad_norm": 3.529494927435862,
+      "learning_rate": 2e-05,
+      "loss": 0.1974,
+      "step": 8834
+    },
+    {
+      "epoch": 0.21029705798343332,
+      "grad_norm": 3.2391318972147753,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 8835
+    },
+    {
+      "epoch": 0.21032086070646483,
+      "grad_norm": 3.136321000796866,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 8836
+    },
+    {
+      "epoch": 0.21034466342949634,
+      "grad_norm": 5.820901663282806,
+      "learning_rate": 2e-05,
+      "loss": 0.8315,
+      "step": 8837
+    },
+    {
+      "epoch": 0.21036846615252786,
+      "grad_norm": 34.08622308599336,
+      "learning_rate": 2e-05,
+      "loss": 0.8167,
+      "step": 8838
+    },
+    {
+      "epoch": 0.21039226887555937,
+      "grad_norm": 3.3809864209125675,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 8839
+    },
+    {
+      "epoch": 0.21041607159859088,
+      "grad_norm": 5.452599493745847,
+      "learning_rate": 2e-05,
+      "loss": 1.0215,
+      "step": 8840
+    },
+    {
+      "epoch": 0.2104398743216224,
+      "grad_norm": 3.2419348840503757,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 8841
+    },
+    {
+      "epoch": 0.2104636770446539,
+      "grad_norm": 1.352987211299835,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 8842
+    },
+    {
+      "epoch": 0.21048747976768542,
+      "grad_norm": 1.973466178469752,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 8843
+    },
+    {
+      "epoch": 0.21051128249071693,
+      "grad_norm": 2.3231545419320168,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 8844
+    },
+    {
+      "epoch": 0.21053508521374845,
+      "grad_norm": 5.137423171747269,
+      "learning_rate": 2e-05,
+      "loss": 0.5533,
+      "step": 8845
+    },
+    {
+      "epoch": 0.21055888793677996,
+      "grad_norm": 3.9155063870803284,
+      "learning_rate": 2e-05,
+      "loss": 0.5867,
+      "step": 8846
+    },
+    {
+      "epoch": 0.21058269065981147,
+      "grad_norm": 6.991178438145708,
+      "learning_rate": 2e-05,
+      "loss": 0.8546,
+      "step": 8847
+    },
+    {
+      "epoch": 0.21060649338284299,
+      "grad_norm": 3.7544955380256386,
+      "learning_rate": 2e-05,
+      "loss": 0.5393,
+      "step": 8848
+    },
+    {
+      "epoch": 0.2106302961058745,
+      "grad_norm": 1.41813819244514,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 8849
+    },
+    {
+      "epoch": 0.21065409882890604,
+      "grad_norm": 3.862004090504433,
+      "learning_rate": 2e-05,
+      "loss": 0.2308,
+      "step": 8850
+    },
+    {
+      "epoch": 0.21067790155193755,
+      "grad_norm": 14.293768471358344,
+      "learning_rate": 2e-05,
+      "loss": 0.7315,
+      "step": 8851
+    },
+    {
+      "epoch": 0.21070170427496906,
+      "grad_norm": 7.501911038515991,
+      "learning_rate": 2e-05,
+      "loss": 1.5145,
+      "step": 8852
+    },
+    {
+      "epoch": 0.21072550699800058,
+      "grad_norm": 7.472471242601133,
+      "learning_rate": 2e-05,
+      "loss": 0.7157,
+      "step": 8853
+    },
+    {
+      "epoch": 0.2107493097210321,
+      "grad_norm": 5.902225051141185,
+      "learning_rate": 2e-05,
+      "loss": 0.9872,
+      "step": 8854
+    },
+    {
+      "epoch": 0.2107731124440636,
+      "grad_norm": 3.416297226054989,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 8855
+    },
+    {
+      "epoch": 0.21079691516709512,
+      "grad_norm": 19.866357746309156,
+      "learning_rate": 2e-05,
+      "loss": 0.3306,
+      "step": 8856
+    },
+    {
+      "epoch": 0.21082071789012663,
+      "grad_norm": 4.3403542473716294,
+      "learning_rate": 2e-05,
+      "loss": 0.9299,
+      "step": 8857
+    },
+    {
+      "epoch": 0.21084452061315814,
+      "grad_norm": 8.14964771918856,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 8858
+    },
+    {
+      "epoch": 0.21086832333618966,
+      "grad_norm": 13.851173941840369,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 8859
+    },
+    {
+      "epoch": 0.21089212605922117,
+      "grad_norm": 2.037209189464389,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 8860
+    },
+    {
+      "epoch": 0.21091592878225268,
+      "grad_norm": 10.731655722393354,
+      "learning_rate": 2e-05,
+      "loss": 0.625,
+      "step": 8861
+    },
+    {
+      "epoch": 0.2109397315052842,
+      "grad_norm": 10.280102710013603,
+      "learning_rate": 2e-05,
+      "loss": 0.302,
+      "step": 8862
+    },
+    {
+      "epoch": 0.2109635342283157,
+      "grad_norm": 3.325013368870263,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 8863
+    },
+    {
+      "epoch": 0.21098733695134722,
+      "grad_norm": 7.780943928175658,
+      "learning_rate": 2e-05,
+      "loss": 0.6158,
+      "step": 8864
+    },
+    {
+      "epoch": 0.21101113967437876,
+      "grad_norm": 5.2836202576042535,
+      "learning_rate": 2e-05,
+      "loss": 0.8546,
+      "step": 8865
+    },
+    {
+      "epoch": 0.21103494239741027,
+      "grad_norm": 4.491341238244604,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 8866
+    },
+    {
+      "epoch": 0.2110587451204418,
+      "grad_norm": 3.1730467806202336,
+      "learning_rate": 2e-05,
+      "loss": 0.1235,
+      "step": 8867
+    },
+    {
+      "epoch": 0.2110825478434733,
+      "grad_norm": 1.9090874757857463,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 8868
+    },
+    {
+      "epoch": 0.2111063505665048,
+      "grad_norm": 3.1023523165616047,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 8869
+    },
+    {
+      "epoch": 0.21113015328953633,
+      "grad_norm": 2.6066633207932,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 8870
+    },
+    {
+      "epoch": 0.21115395601256784,
+      "grad_norm": 7.618085635640803,
+      "learning_rate": 2e-05,
+      "loss": 0.3012,
+      "step": 8871
+    },
+    {
+      "epoch": 0.21117775873559935,
+      "grad_norm": 7.8378846667355155,
+      "learning_rate": 2e-05,
+      "loss": 0.1319,
+      "step": 8872
+    },
+    {
+      "epoch": 0.21120156145863087,
+      "grad_norm": 2.5014694528332075,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 8873
+    },
+    {
+      "epoch": 0.21122536418166238,
+      "grad_norm": 2.552755693354668,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 8874
+    },
+    {
+      "epoch": 0.2112491669046939,
+      "grad_norm": 5.496226390055511,
+      "learning_rate": 2e-05,
+      "loss": 0.7211,
+      "step": 8875
+    },
+    {
+      "epoch": 0.2112729696277254,
+      "grad_norm": 4.976925941708068,
+      "learning_rate": 2e-05,
+      "loss": 0.6251,
+      "step": 8876
+    },
+    {
+      "epoch": 0.21129677235075692,
+      "grad_norm": 4.264561324035488,
+      "learning_rate": 2e-05,
+      "loss": 0.7588,
+      "step": 8877
+    },
+    {
+      "epoch": 0.21132057507378843,
+      "grad_norm": 5.729181526189849,
+      "learning_rate": 2e-05,
+      "loss": 0.2247,
+      "step": 8878
+    },
+    {
+      "epoch": 0.21134437779681994,
+      "grad_norm": 5.618616419556285,
+      "learning_rate": 2e-05,
+      "loss": 0.6159,
+      "step": 8879
+    },
+    {
+      "epoch": 0.21136818051985148,
+      "grad_norm": 2.9759778937367996,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 8880
+    },
+    {
+      "epoch": 0.211391983242883,
+      "grad_norm": 3.8119745350180656,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 8881
+    },
+    {
+      "epoch": 0.2114157859659145,
+      "grad_norm": 1.2976799501685865,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 8882
+    },
+    {
+      "epoch": 0.21143958868894602,
+      "grad_norm": 6.65141870711823,
+      "learning_rate": 2e-05,
+      "loss": 0.9577,
+      "step": 8883
+    },
+    {
+      "epoch": 0.21146339141197754,
+      "grad_norm": 9.25338708580246,
+      "learning_rate": 2e-05,
+      "loss": 1.5962,
+      "step": 8884
+    },
+    {
+      "epoch": 0.21148719413500905,
+      "grad_norm": 6.176419288685049,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 8885
+    },
+    {
+      "epoch": 0.21151099685804056,
+      "grad_norm": 3.8464678474674474,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 8886
+    },
+    {
+      "epoch": 0.21153479958107207,
+      "grad_norm": 4.236453916653599,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 8887
+    },
+    {
+      "epoch": 0.2115586023041036,
+      "grad_norm": 6.902486942859644,
+      "learning_rate": 2e-05,
+      "loss": 1.3642,
+      "step": 8888
+    },
+    {
+      "epoch": 0.2115824050271351,
+      "grad_norm": 4.342216947746846,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 8889
+    },
+    {
+      "epoch": 0.21160620775016661,
+      "grad_norm": 16.290424395672854,
+      "learning_rate": 2e-05,
+      "loss": 0.5041,
+      "step": 8890
+    },
+    {
+      "epoch": 0.21163001047319813,
+      "grad_norm": 2.4324285896822655,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 8891
+    },
+    {
+      "epoch": 0.21165381319622964,
+      "grad_norm": 3.222088908938421,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 8892
+    },
+    {
+      "epoch": 0.21167761591926115,
+      "grad_norm": 12.138781547017285,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 8893
+    },
+    {
+      "epoch": 0.21170141864229267,
+      "grad_norm": 4.598646057725709,
+      "learning_rate": 2e-05,
+      "loss": 0.2441,
+      "step": 8894
+    },
+    {
+      "epoch": 0.2117252213653242,
+      "grad_norm": 2.469617511390471,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 8895
+    },
+    {
+      "epoch": 0.21174902408835572,
+      "grad_norm": 5.141534610310173,
+      "learning_rate": 2e-05,
+      "loss": 0.7882,
+      "step": 8896
+    },
+    {
+      "epoch": 0.21177282681138723,
+      "grad_norm": 3.522482027754064,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 8897
+    },
+    {
+      "epoch": 0.21179662953441875,
+      "grad_norm": 3.3842196160793647,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 8898
+    },
+    {
+      "epoch": 0.21182043225745026,
+      "grad_norm": 6.668241522962881,
+      "learning_rate": 2e-05,
+      "loss": 0.2583,
+      "step": 8899
+    },
+    {
+      "epoch": 0.21184423498048177,
+      "grad_norm": 3.3819746795767482,
+      "learning_rate": 2e-05,
+      "loss": 0.8038,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21186803770351328,
+      "grad_norm": 2.0765973907537556,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 8901
+    },
+    {
+      "epoch": 0.2118918404265448,
+      "grad_norm": 4.023177869541307,
+      "learning_rate": 2e-05,
+      "loss": 0.2302,
+      "step": 8902
+    },
+    {
+      "epoch": 0.2119156431495763,
+      "grad_norm": 1.2931827469534973,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 8903
+    },
+    {
+      "epoch": 0.21193944587260782,
+      "grad_norm": 2.522990339679448,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 8904
+    },
+    {
+      "epoch": 0.21196324859563934,
+      "grad_norm": 5.167861565349234,
+      "learning_rate": 2e-05,
+      "loss": 0.899,
+      "step": 8905
+    },
+    {
+      "epoch": 0.21198705131867085,
+      "grad_norm": 7.325063930105458,
+      "learning_rate": 2e-05,
+      "loss": 0.8023,
+      "step": 8906
+    },
+    {
+      "epoch": 0.21201085404170236,
+      "grad_norm": 2.960379612409495,
+      "learning_rate": 2e-05,
+      "loss": 0.143,
+      "step": 8907
+    },
+    {
+      "epoch": 0.21203465676473388,
+      "grad_norm": 1.2162272365038032,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 8908
+    },
+    {
+      "epoch": 0.2120584594877654,
+      "grad_norm": 3.231977530952188,
+      "learning_rate": 2e-05,
+      "loss": 0.2307,
+      "step": 8909
+    },
+    {
+      "epoch": 0.2120822622107969,
+      "grad_norm": 3.061999920610738,
+      "learning_rate": 2e-05,
+      "loss": 0.169,
+      "step": 8910
+    },
+    {
+      "epoch": 0.21210606493382844,
+      "grad_norm": 3.0892513502518653,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 8911
+    },
+    {
+      "epoch": 0.21212986765685996,
+      "grad_norm": 8.116098014041281,
+      "learning_rate": 2e-05,
+      "loss": 0.5688,
+      "step": 8912
+    },
+    {
+      "epoch": 0.21215367037989147,
+      "grad_norm": 1.8392174150154448,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 8913
+    },
+    {
+      "epoch": 0.21217747310292298,
+      "grad_norm": 5.355148728418359,
+      "learning_rate": 2e-05,
+      "loss": 0.9541,
+      "step": 8914
+    },
+    {
+      "epoch": 0.2122012758259545,
+      "grad_norm": 2.99373773989495,
+      "learning_rate": 2e-05,
+      "loss": 0.1731,
+      "step": 8915
+    },
+    {
+      "epoch": 0.212225078548986,
+      "grad_norm": 4.813786520361415,
+      "learning_rate": 2e-05,
+      "loss": 0.2397,
+      "step": 8916
+    },
+    {
+      "epoch": 0.21224888127201752,
+      "grad_norm": 4.531892318535946,
+      "learning_rate": 2e-05,
+      "loss": 0.4885,
+      "step": 8917
+    },
+    {
+      "epoch": 0.21227268399504903,
+      "grad_norm": 7.641896400697971,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 8918
+    },
+    {
+      "epoch": 0.21229648671808055,
+      "grad_norm": 12.951147888646254,
+      "learning_rate": 2e-05,
+      "loss": 0.8118,
+      "step": 8919
+    },
+    {
+      "epoch": 0.21232028944111206,
+      "grad_norm": 3.4771382800634343,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21234409216414357,
+      "grad_norm": 6.50453426148281,
+      "learning_rate": 2e-05,
+      "loss": 1.4321,
+      "step": 8921
+    },
+    {
+      "epoch": 0.21236789488717508,
+      "grad_norm": 4.350363847343812,
+      "learning_rate": 2e-05,
+      "loss": 0.1673,
+      "step": 8922
+    },
+    {
+      "epoch": 0.2123916976102066,
+      "grad_norm": 5.004365816023517,
+      "learning_rate": 2e-05,
+      "loss": 0.5114,
+      "step": 8923
+    },
+    {
+      "epoch": 0.2124155003332381,
+      "grad_norm": 6.303446042757645,
+      "learning_rate": 2e-05,
+      "loss": 0.415,
+      "step": 8924
+    },
+    {
+      "epoch": 0.21243930305626962,
+      "grad_norm": 5.3762578156703995,
+      "learning_rate": 2e-05,
+      "loss": 0.3431,
+      "step": 8925
+    },
+    {
+      "epoch": 0.21246310577930116,
+      "grad_norm": 1.1444920409998434,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 8926
+    },
+    {
+      "epoch": 0.21248690850233268,
+      "grad_norm": 7.629566792434312,
+      "learning_rate": 2e-05,
+      "loss": 1.1748,
+      "step": 8927
+    },
+    {
+      "epoch": 0.2125107112253642,
+      "grad_norm": 5.1165015436377175,
+      "learning_rate": 2e-05,
+      "loss": 0.2682,
+      "step": 8928
+    },
+    {
+      "epoch": 0.2125345139483957,
+      "grad_norm": 8.36879671494246,
+      "learning_rate": 2e-05,
+      "loss": 1.3752,
+      "step": 8929
+    },
+    {
+      "epoch": 0.21255831667142722,
+      "grad_norm": 6.589114742846347,
+      "learning_rate": 2e-05,
+      "loss": 0.2029,
+      "step": 8930
+    },
+    {
+      "epoch": 0.21258211939445873,
+      "grad_norm": 5.819223457486809,
+      "learning_rate": 2e-05,
+      "loss": 1.319,
+      "step": 8931
+    },
+    {
+      "epoch": 0.21260592211749024,
+      "grad_norm": 16.63356920236866,
+      "learning_rate": 2e-05,
+      "loss": 0.4363,
+      "step": 8932
+    },
+    {
+      "epoch": 0.21262972484052176,
+      "grad_norm": 8.898771049343443,
+      "learning_rate": 2e-05,
+      "loss": 0.9527,
+      "step": 8933
+    },
+    {
+      "epoch": 0.21265352756355327,
+      "grad_norm": 2.8618965935723337,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 8934
+    },
+    {
+      "epoch": 0.21267733028658478,
+      "grad_norm": 7.75761067821763,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 8935
+    },
+    {
+      "epoch": 0.2127011330096163,
+      "grad_norm": 3.065870275514265,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 8936
+    },
+    {
+      "epoch": 0.2127249357326478,
+      "grad_norm": 5.6618980592925,
+      "learning_rate": 2e-05,
+      "loss": 0.968,
+      "step": 8937
+    },
+    {
+      "epoch": 0.21274873845567932,
+      "grad_norm": 1.4685418127885184,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 8938
+    },
+    {
+      "epoch": 0.21277254117871083,
+      "grad_norm": 7.6096247580058245,
+      "learning_rate": 2e-05,
+      "loss": 0.2823,
+      "step": 8939
+    },
+    {
+      "epoch": 0.21279634390174235,
+      "grad_norm": 11.833241282104247,
+      "learning_rate": 2e-05,
+      "loss": 0.2877,
+      "step": 8940
+    },
+    {
+      "epoch": 0.2128201466247739,
+      "grad_norm": 3.633338108911972,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 8941
+    },
+    {
+      "epoch": 0.2128439493478054,
+      "grad_norm": 1.7828327984784718,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 8942
+    },
+    {
+      "epoch": 0.2128677520708369,
+      "grad_norm": 4.280760757609572,
+      "learning_rate": 2e-05,
+      "loss": 0.423,
+      "step": 8943
+    },
+    {
+      "epoch": 0.21289155479386843,
+      "grad_norm": 3.51146092759328,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 8944
+    },
+    {
+      "epoch": 0.21291535751689994,
+      "grad_norm": 9.431874642785747,
+      "learning_rate": 2e-05,
+      "loss": 0.5044,
+      "step": 8945
+    },
+    {
+      "epoch": 0.21293916023993145,
+      "grad_norm": 5.916667163525743,
+      "learning_rate": 2e-05,
+      "loss": 0.2124,
+      "step": 8946
+    },
+    {
+      "epoch": 0.21296296296296297,
+      "grad_norm": 5.536515856882381,
+      "learning_rate": 2e-05,
+      "loss": 1.1509,
+      "step": 8947
+    },
+    {
+      "epoch": 0.21298676568599448,
+      "grad_norm": 1.5966050677748092,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 8948
+    },
+    {
+      "epoch": 0.213010568409026,
+      "grad_norm": 11.097609152814847,
+      "learning_rate": 2e-05,
+      "loss": 0.3634,
+      "step": 8949
+    },
+    {
+      "epoch": 0.2130343711320575,
+      "grad_norm": 2.417802473023446,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21305817385508902,
+      "grad_norm": 1.7027628876229455,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 8951
+    },
+    {
+      "epoch": 0.21308197657812053,
+      "grad_norm": 5.633764159946692,
+      "learning_rate": 2e-05,
+      "loss": 0.2957,
+      "step": 8952
+    },
+    {
+      "epoch": 0.21310577930115204,
+      "grad_norm": 12.357980703181573,
+      "learning_rate": 2e-05,
+      "loss": 0.3124,
+      "step": 8953
+    },
+    {
+      "epoch": 0.21312958202418356,
+      "grad_norm": 4.369264418301259,
+      "learning_rate": 2e-05,
+      "loss": 0.4676,
+      "step": 8954
+    },
+    {
+      "epoch": 0.21315338474721507,
+      "grad_norm": 4.620903216335628,
+      "learning_rate": 2e-05,
+      "loss": 0.6065,
+      "step": 8955
+    },
+    {
+      "epoch": 0.2131771874702466,
+      "grad_norm": 6.204099073208237,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 8956
+    },
+    {
+      "epoch": 0.21320099019327812,
+      "grad_norm": 3.851906126085587,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 8957
+    },
+    {
+      "epoch": 0.21322479291630964,
+      "grad_norm": 4.234204671465297,
+      "learning_rate": 2e-05,
+      "loss": 0.2372,
+      "step": 8958
+    },
+    {
+      "epoch": 0.21324859563934115,
+      "grad_norm": 9.933391140177735,
+      "learning_rate": 2e-05,
+      "loss": 1.4183,
+      "step": 8959
+    },
+    {
+      "epoch": 0.21327239836237266,
+      "grad_norm": 3.6228867348050735,
+      "learning_rate": 2e-05,
+      "loss": 0.4631,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21329620108540417,
+      "grad_norm": 3.098985620102153,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 8961
+    },
+    {
+      "epoch": 0.2133200038084357,
+      "grad_norm": 20.61611782926773,
+      "learning_rate": 2e-05,
+      "loss": 0.5757,
+      "step": 8962
+    },
+    {
+      "epoch": 0.2133438065314672,
+      "grad_norm": 4.926032045540541,
+      "learning_rate": 2e-05,
+      "loss": 0.7883,
+      "step": 8963
+    },
+    {
+      "epoch": 0.2133676092544987,
+      "grad_norm": 8.06568595571825,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 8964
+    },
+    {
+      "epoch": 0.21339141197753023,
+      "grad_norm": 12.95492903833409,
+      "learning_rate": 2e-05,
+      "loss": 0.8772,
+      "step": 8965
+    },
+    {
+      "epoch": 0.21341521470056174,
+      "grad_norm": 3.158514815748658,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 8966
+    },
+    {
+      "epoch": 0.21343901742359325,
+      "grad_norm": 2.095318125328727,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 8967
+    },
+    {
+      "epoch": 0.21346282014662477,
+      "grad_norm": 29.60436587740916,
+      "learning_rate": 2e-05,
+      "loss": 0.6421,
+      "step": 8968
+    },
+    {
+      "epoch": 0.21348662286965628,
+      "grad_norm": 2.5547153032550955,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 8969
+    },
+    {
+      "epoch": 0.2135104255926878,
+      "grad_norm": 15.19794693675342,
+      "learning_rate": 2e-05,
+      "loss": 0.4015,
+      "step": 8970
+    },
+    {
+      "epoch": 0.21353422831571933,
+      "grad_norm": 3.0987367976841735,
+      "learning_rate": 2e-05,
+      "loss": 0.2203,
+      "step": 8971
+    },
+    {
+      "epoch": 0.21355803103875085,
+      "grad_norm": 7.864334529204961,
+      "learning_rate": 2e-05,
+      "loss": 1.2419,
+      "step": 8972
+    },
+    {
+      "epoch": 0.21358183376178236,
+      "grad_norm": 6.257889061955176,
+      "learning_rate": 2e-05,
+      "loss": 1.2238,
+      "step": 8973
+    },
+    {
+      "epoch": 0.21360563648481387,
+      "grad_norm": 13.787996135598405,
+      "learning_rate": 2e-05,
+      "loss": 0.418,
+      "step": 8974
+    },
+    {
+      "epoch": 0.21362943920784538,
+      "grad_norm": 11.58660604193404,
+      "learning_rate": 2e-05,
+      "loss": 0.4239,
+      "step": 8975
+    },
+    {
+      "epoch": 0.2136532419308769,
+      "grad_norm": 6.924649725011439,
+      "learning_rate": 2e-05,
+      "loss": 1.5329,
+      "step": 8976
+    },
+    {
+      "epoch": 0.2136770446539084,
+      "grad_norm": 3.1245938107188316,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 8977
+    },
+    {
+      "epoch": 0.21370084737693992,
+      "grad_norm": 3.5533706278962893,
+      "learning_rate": 2e-05,
+      "loss": 0.6142,
+      "step": 8978
+    },
+    {
+      "epoch": 0.21372465009997144,
+      "grad_norm": 3.5457993833840464,
+      "learning_rate": 2e-05,
+      "loss": 0.4477,
+      "step": 8979
+    },
+    {
+      "epoch": 0.21374845282300295,
+      "grad_norm": 3.987413278021319,
+      "learning_rate": 2e-05,
+      "loss": 0.4075,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21377225554603446,
+      "grad_norm": 3.238873541765119,
+      "learning_rate": 2e-05,
+      "loss": 0.4091,
+      "step": 8981
+    },
+    {
+      "epoch": 0.21379605826906597,
+      "grad_norm": 6.081848929683201,
+      "learning_rate": 2e-05,
+      "loss": 1.1496,
+      "step": 8982
+    },
+    {
+      "epoch": 0.2138198609920975,
+      "grad_norm": 3.2844682802216343,
+      "learning_rate": 2e-05,
+      "loss": 0.4732,
+      "step": 8983
+    },
+    {
+      "epoch": 0.213843663715129,
+      "grad_norm": 7.362791612931339,
+      "learning_rate": 2e-05,
+      "loss": 0.3517,
+      "step": 8984
+    },
+    {
+      "epoch": 0.21386746643816051,
+      "grad_norm": 1.779326891067617,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 8985
+    },
+    {
+      "epoch": 0.21389126916119203,
+      "grad_norm": 4.166248656265898,
+      "learning_rate": 2e-05,
+      "loss": 0.5786,
+      "step": 8986
+    },
+    {
+      "epoch": 0.21391507188422357,
+      "grad_norm": 3.101782450861859,
+      "learning_rate": 2e-05,
+      "loss": 0.2061,
+      "step": 8987
+    },
+    {
+      "epoch": 0.21393887460725508,
+      "grad_norm": 12.398447383475935,
+      "learning_rate": 2e-05,
+      "loss": 0.6275,
+      "step": 8988
+    },
+    {
+      "epoch": 0.2139626773302866,
+      "grad_norm": 12.700690331799096,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 8989
+    },
+    {
+      "epoch": 0.2139864800533181,
+      "grad_norm": 8.106739847955964,
+      "learning_rate": 2e-05,
+      "loss": 0.1868,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21401028277634962,
+      "grad_norm": 4.764236332778487,
+      "learning_rate": 2e-05,
+      "loss": 0.8943,
+      "step": 8991
+    },
+    {
+      "epoch": 0.21403408549938113,
+      "grad_norm": 2.1539900917071444,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 8992
+    },
+    {
+      "epoch": 0.21405788822241265,
+      "grad_norm": 6.023170761877774,
+      "learning_rate": 2e-05,
+      "loss": 0.1679,
+      "step": 8993
+    },
+    {
+      "epoch": 0.21408169094544416,
+      "grad_norm": 109.71472861108374,
+      "learning_rate": 2e-05,
+      "loss": 1.601,
+      "step": 8994
+    },
+    {
+      "epoch": 0.21410549366847567,
+      "grad_norm": 2.4827368131196033,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 8995
+    },
+    {
+      "epoch": 0.21412929639150718,
+      "grad_norm": 3.72166892738006,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 8996
+    },
+    {
+      "epoch": 0.2141530991145387,
+      "grad_norm": 1.5990874956615881,
+      "learning_rate": 2e-05,
+      "loss": 0.1213,
+      "step": 8997
+    },
+    {
+      "epoch": 0.2141769018375702,
+      "grad_norm": 5.324613391800849,
+      "learning_rate": 2e-05,
+      "loss": 0.798,
+      "step": 8998
+    },
+    {
+      "epoch": 0.21420070456060172,
+      "grad_norm": 3.5638639403866232,
+      "learning_rate": 2e-05,
+      "loss": 0.712,
+      "step": 8999
+    },
+    {
+      "epoch": 0.21422450728363324,
+      "grad_norm": 1.7771436434238177,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21424831000666475,
+      "grad_norm": 5.776379521491683,
+      "learning_rate": 2e-05,
+      "loss": 0.8668,
+      "step": 9001
+    },
+    {
+      "epoch": 0.2142721127296963,
+      "grad_norm": 1.8811702803031394,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 9002
+    },
+    {
+      "epoch": 0.2142959154527278,
+      "grad_norm": 3.659882680176265,
+      "learning_rate": 2e-05,
+      "loss": 0.6128,
+      "step": 9003
+    },
+    {
+      "epoch": 0.21431971817575932,
+      "grad_norm": 6.726920983817393,
+      "learning_rate": 2e-05,
+      "loss": 0.9989,
+      "step": 9004
+    },
+    {
+      "epoch": 0.21434352089879083,
+      "grad_norm": 3.3322747489117392,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 9005
+    },
+    {
+      "epoch": 0.21436732362182234,
+      "grad_norm": 8.489358671984917,
+      "learning_rate": 2e-05,
+      "loss": 0.8211,
+      "step": 9006
+    },
+    {
+      "epoch": 0.21439112634485386,
+      "grad_norm": 9.24179220399491,
+      "learning_rate": 2e-05,
+      "loss": 1.2465,
+      "step": 9007
+    },
+    {
+      "epoch": 0.21441492906788537,
+      "grad_norm": 3.8298993914302004,
+      "learning_rate": 2e-05,
+      "loss": 0.1891,
+      "step": 9008
+    },
+    {
+      "epoch": 0.21443873179091688,
+      "grad_norm": 3.5705873202022147,
+      "learning_rate": 2e-05,
+      "loss": 0.4216,
+      "step": 9009
+    },
+    {
+      "epoch": 0.2144625345139484,
+      "grad_norm": 8.612621346888268,
+      "learning_rate": 2e-05,
+      "loss": 0.9471,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2144863372369799,
+      "grad_norm": 5.546751776579494,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 9011
+    },
+    {
+      "epoch": 0.21451013996001142,
+      "grad_norm": 3.7505346467589886,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 9012
+    },
+    {
+      "epoch": 0.21453394268304293,
+      "grad_norm": 9.098182991786937,
+      "learning_rate": 2e-05,
+      "loss": 0.3707,
+      "step": 9013
+    },
+    {
+      "epoch": 0.21455774540607445,
+      "grad_norm": 3.0549961207085157,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 9014
+    },
+    {
+      "epoch": 0.21458154812910596,
+      "grad_norm": 5.395581475464947,
+      "learning_rate": 2e-05,
+      "loss": 1.1012,
+      "step": 9015
+    },
+    {
+      "epoch": 0.21460535085213747,
+      "grad_norm": 31.658507809811148,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 9016
+    },
+    {
+      "epoch": 0.214629153575169,
+      "grad_norm": 3.141046009076022,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 9017
+    },
+    {
+      "epoch": 0.21465295629820053,
+      "grad_norm": 2.064681225249307,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 9018
+    },
+    {
+      "epoch": 0.21467675902123204,
+      "grad_norm": 3.4171116679401745,
+      "learning_rate": 2e-05,
+      "loss": 0.3356,
+      "step": 9019
+    },
+    {
+      "epoch": 0.21470056174426355,
+      "grad_norm": 3.196439230786387,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 9020
+    },
+    {
+      "epoch": 0.21472436446729506,
+      "grad_norm": 2.6094195509098426,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 9021
+    },
+    {
+      "epoch": 0.21474816719032658,
+      "grad_norm": 4.538226117453762,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 9022
+    },
+    {
+      "epoch": 0.2147719699133581,
+      "grad_norm": 7.58682933680886,
+      "learning_rate": 2e-05,
+      "loss": 0.3651,
+      "step": 9023
+    },
+    {
+      "epoch": 0.2147957726363896,
+      "grad_norm": 5.177471269051958,
+      "learning_rate": 2e-05,
+      "loss": 0.4943,
+      "step": 9024
+    },
+    {
+      "epoch": 0.21481957535942112,
+      "grad_norm": 6.590432963458169,
+      "learning_rate": 2e-05,
+      "loss": 1.3546,
+      "step": 9025
+    },
+    {
+      "epoch": 0.21484337808245263,
+      "grad_norm": 9.113441477832072,
+      "learning_rate": 2e-05,
+      "loss": 1.044,
+      "step": 9026
+    },
+    {
+      "epoch": 0.21486718080548414,
+      "grad_norm": 8.275632523685724,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 9027
+    },
+    {
+      "epoch": 0.21489098352851566,
+      "grad_norm": 6.238254955038806,
+      "learning_rate": 2e-05,
+      "loss": 0.7197,
+      "step": 9028
+    },
+    {
+      "epoch": 0.21491478625154717,
+      "grad_norm": 4.172728384308184,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 9029
+    },
+    {
+      "epoch": 0.21493858897457868,
+      "grad_norm": 6.040389917043434,
+      "learning_rate": 2e-05,
+      "loss": 0.9698,
+      "step": 9030
+    },
+    {
+      "epoch": 0.2149623916976102,
+      "grad_norm": 3.420537370144971,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 9031
+    },
+    {
+      "epoch": 0.21498619442064174,
+      "grad_norm": 0.8454547698247808,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 9032
+    },
+    {
+      "epoch": 0.21500999714367325,
+      "grad_norm": 6.214244090324237,
+      "learning_rate": 2e-05,
+      "loss": 0.4814,
+      "step": 9033
+    },
+    {
+      "epoch": 0.21503379986670476,
+      "grad_norm": 4.085244849627632,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 9034
+    },
+    {
+      "epoch": 0.21505760258973627,
+      "grad_norm": 7.579551549137345,
+      "learning_rate": 2e-05,
+      "loss": 1.1273,
+      "step": 9035
+    },
+    {
+      "epoch": 0.2150814053127678,
+      "grad_norm": 1.9311732883758583,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 9036
+    },
+    {
+      "epoch": 0.2151052080357993,
+      "grad_norm": 4.294389521643633,
+      "learning_rate": 2e-05,
+      "loss": 0.7128,
+      "step": 9037
+    },
+    {
+      "epoch": 0.2151290107588308,
+      "grad_norm": 5.888117251600662,
+      "learning_rate": 2e-05,
+      "loss": 0.6789,
+      "step": 9038
+    },
+    {
+      "epoch": 0.21515281348186233,
+      "grad_norm": 8.205877884914527,
+      "learning_rate": 2e-05,
+      "loss": 1.9955,
+      "step": 9039
+    },
+    {
+      "epoch": 0.21517661620489384,
+      "grad_norm": 7.6730204233668715,
+      "learning_rate": 2e-05,
+      "loss": 0.2499,
+      "step": 9040
+    },
+    {
+      "epoch": 0.21520041892792535,
+      "grad_norm": 37.545687504329514,
+      "learning_rate": 2e-05,
+      "loss": 0.3221,
+      "step": 9041
+    },
+    {
+      "epoch": 0.21522422165095687,
+      "grad_norm": 4.323464061653616,
+      "learning_rate": 2e-05,
+      "loss": 0.7931,
+      "step": 9042
+    },
+    {
+      "epoch": 0.21524802437398838,
+      "grad_norm": 6.998536638721116,
+      "learning_rate": 2e-05,
+      "loss": 1.0433,
+      "step": 9043
+    },
+    {
+      "epoch": 0.2152718270970199,
+      "grad_norm": 4.444229387369357,
+      "learning_rate": 2e-05,
+      "loss": 0.2102,
+      "step": 9044
+    },
+    {
+      "epoch": 0.2152956298200514,
+      "grad_norm": 5.491851710747062,
+      "learning_rate": 2e-05,
+      "loss": 0.6748,
+      "step": 9045
+    },
+    {
+      "epoch": 0.21531943254308292,
+      "grad_norm": 8.3730856919124,
+      "learning_rate": 2e-05,
+      "loss": 0.2499,
+      "step": 9046
+    },
+    {
+      "epoch": 0.21534323526611443,
+      "grad_norm": 3.9642407846114596,
+      "learning_rate": 2e-05,
+      "loss": 0.7082,
+      "step": 9047
+    },
+    {
+      "epoch": 0.21536703798914597,
+      "grad_norm": 3.6155653515834736,
+      "learning_rate": 2e-05,
+      "loss": 0.366,
+      "step": 9048
+    },
+    {
+      "epoch": 0.21539084071217748,
+      "grad_norm": 2.6453137497552373,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 9049
+    },
+    {
+      "epoch": 0.215414643435209,
+      "grad_norm": 3.8493944413155283,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2154384461582405,
+      "grad_norm": 5.097595403209767,
+      "learning_rate": 2e-05,
+      "loss": 0.6268,
+      "step": 9051
+    },
+    {
+      "epoch": 0.21546224888127202,
+      "grad_norm": 5.670568025065626,
+      "learning_rate": 2e-05,
+      "loss": 1.0112,
+      "step": 9052
+    },
+    {
+      "epoch": 0.21548605160430354,
+      "grad_norm": 3.757922193931534,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 9053
+    },
+    {
+      "epoch": 0.21550985432733505,
+      "grad_norm": 2.2624895304347974,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 9054
+    },
+    {
+      "epoch": 0.21553365705036656,
+      "grad_norm": 1.8777530939823825,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 9055
+    },
+    {
+      "epoch": 0.21555745977339807,
+      "grad_norm": 3.595293100793298,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 9056
+    },
+    {
+      "epoch": 0.2155812624964296,
+      "grad_norm": 3.443625276371685,
+      "learning_rate": 2e-05,
+      "loss": 0.2157,
+      "step": 9057
+    },
+    {
+      "epoch": 0.2156050652194611,
+      "grad_norm": 5.8918884548951755,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 9058
+    },
+    {
+      "epoch": 0.2156288679424926,
+      "grad_norm": 3.8539613477838146,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 9059
+    },
+    {
+      "epoch": 0.21565267066552413,
+      "grad_norm": 2.6298834716227257,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 9060
+    },
+    {
+      "epoch": 0.21567647338855564,
+      "grad_norm": 5.528462766079175,
+      "learning_rate": 2e-05,
+      "loss": 0.6353,
+      "step": 9061
+    },
+    {
+      "epoch": 0.21570027611158715,
+      "grad_norm": 15.358593792470652,
+      "learning_rate": 2e-05,
+      "loss": 0.7875,
+      "step": 9062
+    },
+    {
+      "epoch": 0.2157240788346187,
+      "grad_norm": 1.4642205846388598,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 9063
+    },
+    {
+      "epoch": 0.2157478815576502,
+      "grad_norm": 3.023025052930479,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 9064
+    },
+    {
+      "epoch": 0.21577168428068172,
+      "grad_norm": 31.547896846445553,
+      "learning_rate": 2e-05,
+      "loss": 0.7403,
+      "step": 9065
+    },
+    {
+      "epoch": 0.21579548700371323,
+      "grad_norm": 6.3433260273996765,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 9066
+    },
+    {
+      "epoch": 0.21581928972674475,
+      "grad_norm": 4.002704648858885,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 9067
+    },
+    {
+      "epoch": 0.21584309244977626,
+      "grad_norm": 6.832633175420633,
+      "learning_rate": 2e-05,
+      "loss": 0.6064,
+      "step": 9068
+    },
+    {
+      "epoch": 0.21586689517280777,
+      "grad_norm": 4.849324452402962,
+      "learning_rate": 2e-05,
+      "loss": 0.7332,
+      "step": 9069
+    },
+    {
+      "epoch": 0.21589069789583928,
+      "grad_norm": 9.002517352176826,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 9070
+    },
+    {
+      "epoch": 0.2159145006188708,
+      "grad_norm": 1.9247502894008595,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 9071
+    },
+    {
+      "epoch": 0.2159383033419023,
+      "grad_norm": 10.9302878511483,
+      "learning_rate": 2e-05,
+      "loss": 0.423,
+      "step": 9072
+    },
+    {
+      "epoch": 0.21596210606493382,
+      "grad_norm": 2.9986360537979047,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 9073
+    },
+    {
+      "epoch": 0.21598590878796534,
+      "grad_norm": 86.27094569896171,
+      "learning_rate": 2e-05,
+      "loss": 0.3513,
+      "step": 9074
+    },
+    {
+      "epoch": 0.21600971151099685,
+      "grad_norm": 3.393873692960189,
+      "learning_rate": 2e-05,
+      "loss": 0.7014,
+      "step": 9075
+    },
+    {
+      "epoch": 0.21603351423402836,
+      "grad_norm": 8.214516806307243,
+      "learning_rate": 2e-05,
+      "loss": 1.3598,
+      "step": 9076
+    },
+    {
+      "epoch": 0.21605731695705988,
+      "grad_norm": 2.400774241384695,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 9077
+    },
+    {
+      "epoch": 0.21608111968009142,
+      "grad_norm": 10.240320852704226,
+      "learning_rate": 2e-05,
+      "loss": 1.1636,
+      "step": 9078
+    },
+    {
+      "epoch": 0.21610492240312293,
+      "grad_norm": 5.515695126746119,
+      "learning_rate": 2e-05,
+      "loss": 0.7884,
+      "step": 9079
+    },
+    {
+      "epoch": 0.21612872512615444,
+      "grad_norm": 21.50147240967617,
+      "learning_rate": 2e-05,
+      "loss": 0.3253,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21615252784918595,
+      "grad_norm": 13.728903864993136,
+      "learning_rate": 2e-05,
+      "loss": 0.8347,
+      "step": 9081
+    },
+    {
+      "epoch": 0.21617633057221747,
+      "grad_norm": 2.636691973735465,
+      "learning_rate": 2e-05,
+      "loss": 0.118,
+      "step": 9082
+    },
+    {
+      "epoch": 0.21620013329524898,
+      "grad_norm": 4.640986336137473,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 9083
+    },
+    {
+      "epoch": 0.2162239360182805,
+      "grad_norm": 5.327047118769834,
+      "learning_rate": 2e-05,
+      "loss": 0.6963,
+      "step": 9084
+    },
+    {
+      "epoch": 0.216247738741312,
+      "grad_norm": 7.325601381914339,
+      "learning_rate": 2e-05,
+      "loss": 0.4349,
+      "step": 9085
+    },
+    {
+      "epoch": 0.21627154146434352,
+      "grad_norm": 3.1600941849112036,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 9086
+    },
+    {
+      "epoch": 0.21629534418737503,
+      "grad_norm": 6.1647300164875265,
+      "learning_rate": 2e-05,
+      "loss": 1.0423,
+      "step": 9087
+    },
+    {
+      "epoch": 0.21631914691040655,
+      "grad_norm": 5.51406515569763,
+      "learning_rate": 2e-05,
+      "loss": 1.531,
+      "step": 9088
+    },
+    {
+      "epoch": 0.21634294963343806,
+      "grad_norm": 5.451102457661744,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 9089
+    },
+    {
+      "epoch": 0.21636675235646957,
+      "grad_norm": 3.6987671513636853,
+      "learning_rate": 2e-05,
+      "loss": 0.188,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21639055507950108,
+      "grad_norm": 11.026073593355202,
+      "learning_rate": 2e-05,
+      "loss": 0.4542,
+      "step": 9091
+    },
+    {
+      "epoch": 0.2164143578025326,
+      "grad_norm": 8.746104819520975,
+      "learning_rate": 2e-05,
+      "loss": 0.3354,
+      "step": 9092
+    },
+    {
+      "epoch": 0.21643816052556414,
+      "grad_norm": 8.411757773721662,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 9093
+    },
+    {
+      "epoch": 0.21646196324859565,
+      "grad_norm": 8.94897487504244,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 9094
+    },
+    {
+      "epoch": 0.21648576597162716,
+      "grad_norm": 4.0592655288857085,
+      "learning_rate": 2e-05,
+      "loss": 0.2486,
+      "step": 9095
+    },
+    {
+      "epoch": 0.21650956869465868,
+      "grad_norm": 6.457674804240205,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 9096
+    },
+    {
+      "epoch": 0.2165333714176902,
+      "grad_norm": 44.97698449973333,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 9097
+    },
+    {
+      "epoch": 0.2165571741407217,
+      "grad_norm": 16.87962621753359,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 9098
+    },
+    {
+      "epoch": 0.21658097686375322,
+      "grad_norm": 7.401290457531605,
+      "learning_rate": 2e-05,
+      "loss": 1.8397,
+      "step": 9099
+    },
+    {
+      "epoch": 0.21660477958678473,
+      "grad_norm": 8.69581158267041,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 9100
+    },
+    {
+      "epoch": 0.21662858230981624,
+      "grad_norm": 1.7821929973765027,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 9101
+    },
+    {
+      "epoch": 0.21665238503284776,
+      "grad_norm": 0.9456356623130251,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 9102
+    },
+    {
+      "epoch": 0.21667618775587927,
+      "grad_norm": 2.369414344539133,
+      "learning_rate": 2e-05,
+      "loss": 0.1793,
+      "step": 9103
+    },
+    {
+      "epoch": 0.21669999047891078,
+      "grad_norm": 2.5523193376217757,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 9104
+    },
+    {
+      "epoch": 0.2167237932019423,
+      "grad_norm": 6.1499680658031775,
+      "learning_rate": 2e-05,
+      "loss": 0.9243,
+      "step": 9105
+    },
+    {
+      "epoch": 0.2167475959249738,
+      "grad_norm": 4.288624608740956,
+      "learning_rate": 2e-05,
+      "loss": 0.6509,
+      "step": 9106
+    },
+    {
+      "epoch": 0.21677139864800532,
+      "grad_norm": 7.427452990611297,
+      "learning_rate": 2e-05,
+      "loss": 0.4099,
+      "step": 9107
+    },
+    {
+      "epoch": 0.21679520137103683,
+      "grad_norm": 12.273013865653745,
+      "learning_rate": 2e-05,
+      "loss": 0.6322,
+      "step": 9108
+    },
+    {
+      "epoch": 0.21681900409406837,
+      "grad_norm": 1.5845178095646306,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 9109
+    },
+    {
+      "epoch": 0.2168428068170999,
+      "grad_norm": 3.9652361391022577,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 9110
+    },
+    {
+      "epoch": 0.2168666095401314,
+      "grad_norm": 2.5212525197930846,
+      "learning_rate": 2e-05,
+      "loss": 0.1319,
+      "step": 9111
+    },
+    {
+      "epoch": 0.2168904122631629,
+      "grad_norm": 1.6751469100417842,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 9112
+    },
+    {
+      "epoch": 0.21691421498619443,
+      "grad_norm": 3.117098115798801,
+      "learning_rate": 2e-05,
+      "loss": 0.2625,
+      "step": 9113
+    },
+    {
+      "epoch": 0.21693801770922594,
+      "grad_norm": 7.608407034065569,
+      "learning_rate": 2e-05,
+      "loss": 0.7653,
+      "step": 9114
+    },
+    {
+      "epoch": 0.21696182043225745,
+      "grad_norm": 16.34539353066861,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 9115
+    },
+    {
+      "epoch": 0.21698562315528896,
+      "grad_norm": 3.0103348609310396,
+      "learning_rate": 2e-05,
+      "loss": 0.1614,
+      "step": 9116
+    },
+    {
+      "epoch": 0.21700942587832048,
+      "grad_norm": 1.8709362765451298,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 9117
+    },
+    {
+      "epoch": 0.217033228601352,
+      "grad_norm": 8.417952452633374,
+      "learning_rate": 2e-05,
+      "loss": 0.3336,
+      "step": 9118
+    },
+    {
+      "epoch": 0.2170570313243835,
+      "grad_norm": 2.426215844991731,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 9119
+    },
+    {
+      "epoch": 0.21708083404741502,
+      "grad_norm": 6.827768447696257,
+      "learning_rate": 2e-05,
+      "loss": 0.2364,
+      "step": 9120
+    },
+    {
+      "epoch": 0.21710463677044653,
+      "grad_norm": 1.497561420603452,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 9121
+    },
+    {
+      "epoch": 0.21712843949347804,
+      "grad_norm": 13.685670993210818,
+      "learning_rate": 2e-05,
+      "loss": 0.8817,
+      "step": 9122
+    },
+    {
+      "epoch": 0.21715224221650956,
+      "grad_norm": 65.50144512706751,
+      "learning_rate": 2e-05,
+      "loss": 0.8223,
+      "step": 9123
+    },
+    {
+      "epoch": 0.2171760449395411,
+      "grad_norm": 5.9981482793387695,
+      "learning_rate": 2e-05,
+      "loss": 0.4177,
+      "step": 9124
+    },
+    {
+      "epoch": 0.2171998476625726,
+      "grad_norm": 2.9344567504781454,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 9125
+    },
+    {
+      "epoch": 0.21722365038560412,
+      "grad_norm": 12.83332917953152,
+      "learning_rate": 2e-05,
+      "loss": 1.2867,
+      "step": 9126
+    },
+    {
+      "epoch": 0.21724745310863564,
+      "grad_norm": 6.475964413162662,
+      "learning_rate": 2e-05,
+      "loss": 1.0114,
+      "step": 9127
+    },
+    {
+      "epoch": 0.21727125583166715,
+      "grad_norm": 6.882119160701789,
+      "learning_rate": 2e-05,
+      "loss": 0.3308,
+      "step": 9128
+    },
+    {
+      "epoch": 0.21729505855469866,
+      "grad_norm": 11.120243312592239,
+      "learning_rate": 2e-05,
+      "loss": 0.9044,
+      "step": 9129
+    },
+    {
+      "epoch": 0.21731886127773017,
+      "grad_norm": 8.973133320703203,
+      "learning_rate": 2e-05,
+      "loss": 0.9,
+      "step": 9130
+    },
+    {
+      "epoch": 0.2173426640007617,
+      "grad_norm": 3.551729531244365,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 9131
+    },
+    {
+      "epoch": 0.2173664667237932,
+      "grad_norm": 15.906631924574326,
+      "learning_rate": 2e-05,
+      "loss": 0.8689,
+      "step": 9132
+    },
+    {
+      "epoch": 0.2173902694468247,
+      "grad_norm": 2.346341540340414,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 9133
+    },
+    {
+      "epoch": 0.21741407216985623,
+      "grad_norm": 19.104197568316604,
+      "learning_rate": 2e-05,
+      "loss": 0.443,
+      "step": 9134
+    },
+    {
+      "epoch": 0.21743787489288774,
+      "grad_norm": 8.082296386506465,
+      "learning_rate": 2e-05,
+      "loss": 0.2426,
+      "step": 9135
+    },
+    {
+      "epoch": 0.21746167761591925,
+      "grad_norm": 3.548771703005267,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 9136
+    },
+    {
+      "epoch": 0.21748548033895077,
+      "grad_norm": 2.2950501706376447,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 9137
+    },
+    {
+      "epoch": 0.21750928306198228,
+      "grad_norm": 9.327273984171324,
+      "learning_rate": 2e-05,
+      "loss": 0.4198,
+      "step": 9138
+    },
+    {
+      "epoch": 0.21753308578501382,
+      "grad_norm": 5.63732307578722,
+      "learning_rate": 2e-05,
+      "loss": 1.3486,
+      "step": 9139
+    },
+    {
+      "epoch": 0.21755688850804533,
+      "grad_norm": 1.43709723790047,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 9140
+    },
+    {
+      "epoch": 0.21758069123107684,
+      "grad_norm": 2.3997228645588846,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 9141
+    },
+    {
+      "epoch": 0.21760449395410836,
+      "grad_norm": 7.270769005409487,
+      "learning_rate": 2e-05,
+      "loss": 0.6662,
+      "step": 9142
+    },
+    {
+      "epoch": 0.21762829667713987,
+      "grad_norm": 1.7311698085595086,
+      "learning_rate": 2e-05,
+      "loss": 0.1461,
+      "step": 9143
+    },
+    {
+      "epoch": 0.21765209940017138,
+      "grad_norm": 3.3181113489428347,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 9144
+    },
+    {
+      "epoch": 0.2176759021232029,
+      "grad_norm": 2.5289994751942135,
+      "learning_rate": 2e-05,
+      "loss": 0.1868,
+      "step": 9145
+    },
+    {
+      "epoch": 0.2176997048462344,
+      "grad_norm": 1.8520640406948419,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 9146
+    },
+    {
+      "epoch": 0.21772350756926592,
+      "grad_norm": 4.302931780615348,
+      "learning_rate": 2e-05,
+      "loss": 0.7402,
+      "step": 9147
+    },
+    {
+      "epoch": 0.21774731029229744,
+      "grad_norm": 5.337172631878874,
+      "learning_rate": 2e-05,
+      "loss": 0.5325,
+      "step": 9148
+    },
+    {
+      "epoch": 0.21777111301532895,
+      "grad_norm": 7.016122346064862,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 9149
+    },
+    {
+      "epoch": 0.21779491573836046,
+      "grad_norm": 2.418219129631899,
+      "learning_rate": 2e-05,
+      "loss": 0.1581,
+      "step": 9150
+    },
+    {
+      "epoch": 0.21781871846139197,
+      "grad_norm": 6.067325801079615,
+      "learning_rate": 2e-05,
+      "loss": 0.873,
+      "step": 9151
+    },
+    {
+      "epoch": 0.2178425211844235,
+      "grad_norm": 7.911215911871692,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 9152
+    },
+    {
+      "epoch": 0.217866323907455,
+      "grad_norm": 9.54683871665308,
+      "learning_rate": 2e-05,
+      "loss": 1.3634,
+      "step": 9153
+    },
+    {
+      "epoch": 0.21789012663048654,
+      "grad_norm": 17.714407502581004,
+      "learning_rate": 2e-05,
+      "loss": 0.5345,
+      "step": 9154
+    },
+    {
+      "epoch": 0.21791392935351805,
+      "grad_norm": 1.4515793650777218,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 9155
+    },
+    {
+      "epoch": 0.21793773207654957,
+      "grad_norm": 3.021995446070577,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 9156
+    },
+    {
+      "epoch": 0.21796153479958108,
+      "grad_norm": 6.006564481400028,
+      "learning_rate": 2e-05,
+      "loss": 0.8029,
+      "step": 9157
+    },
+    {
+      "epoch": 0.2179853375226126,
+      "grad_norm": 18.335535258663686,
+      "learning_rate": 2e-05,
+      "loss": 0.508,
+      "step": 9158
+    },
+    {
+      "epoch": 0.2180091402456441,
+      "grad_norm": 3.0026182858840573,
+      "learning_rate": 2e-05,
+      "loss": 0.4866,
+      "step": 9159
+    },
+    {
+      "epoch": 0.21803294296867562,
+      "grad_norm": 4.709350847322372,
+      "learning_rate": 2e-05,
+      "loss": 0.632,
+      "step": 9160
+    },
+    {
+      "epoch": 0.21805674569170713,
+      "grad_norm": 4.976927820354124,
+      "learning_rate": 2e-05,
+      "loss": 1.0489,
+      "step": 9161
+    },
+    {
+      "epoch": 0.21808054841473865,
+      "grad_norm": 4.233987948627406,
+      "learning_rate": 2e-05,
+      "loss": 0.6377,
+      "step": 9162
+    },
+    {
+      "epoch": 0.21810435113777016,
+      "grad_norm": 4.545267491966331,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 9163
+    },
+    {
+      "epoch": 0.21812815386080167,
+      "grad_norm": 5.821039684912804,
+      "learning_rate": 2e-05,
+      "loss": 0.7215,
+      "step": 9164
+    },
+    {
+      "epoch": 0.21815195658383318,
+      "grad_norm": 15.25881056383746,
+      "learning_rate": 2e-05,
+      "loss": 0.6155,
+      "step": 9165
+    },
+    {
+      "epoch": 0.2181757593068647,
+      "grad_norm": 4.4849098933284886,
+      "learning_rate": 2e-05,
+      "loss": 0.5941,
+      "step": 9166
+    },
+    {
+      "epoch": 0.2181995620298962,
+      "grad_norm": 0.28884444569782686,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 9167
+    },
+    {
+      "epoch": 0.21822336475292772,
+      "grad_norm": 10.21450045971471,
+      "learning_rate": 2e-05,
+      "loss": 0.3798,
+      "step": 9168
+    },
+    {
+      "epoch": 0.21824716747595924,
+      "grad_norm": 4.9681773071179824,
+      "learning_rate": 2e-05,
+      "loss": 0.1625,
+      "step": 9169
+    },
+    {
+      "epoch": 0.21827097019899078,
+      "grad_norm": 7.471718317730367,
+      "learning_rate": 2e-05,
+      "loss": 0.4953,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2182947729220223,
+      "grad_norm": 5.756181841420976,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 9171
+    },
+    {
+      "epoch": 0.2183185756450538,
+      "grad_norm": 8.161479930634322,
+      "learning_rate": 2e-05,
+      "loss": 0.8038,
+      "step": 9172
+    },
+    {
+      "epoch": 0.21834237836808532,
+      "grad_norm": 5.9076802568279705,
+      "learning_rate": 2e-05,
+      "loss": 0.8976,
+      "step": 9173
+    },
+    {
+      "epoch": 0.21836618109111683,
+      "grad_norm": 3.9533556945395376,
+      "learning_rate": 2e-05,
+      "loss": 0.5369,
+      "step": 9174
+    },
+    {
+      "epoch": 0.21838998381414834,
+      "grad_norm": 5.473695912592574,
+      "learning_rate": 2e-05,
+      "loss": 0.552,
+      "step": 9175
+    },
+    {
+      "epoch": 0.21841378653717985,
+      "grad_norm": 1.7418976043482783,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 9176
+    },
+    {
+      "epoch": 0.21843758926021137,
+      "grad_norm": 5.597325823676092,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 9177
+    },
+    {
+      "epoch": 0.21846139198324288,
+      "grad_norm": 6.28920505275177,
+      "learning_rate": 2e-05,
+      "loss": 1.1206,
+      "step": 9178
+    },
+    {
+      "epoch": 0.2184851947062744,
+      "grad_norm": 7.7591052579508695,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 9179
+    },
+    {
+      "epoch": 0.2185089974293059,
+      "grad_norm": 10.184647688606784,
+      "learning_rate": 2e-05,
+      "loss": 0.5627,
+      "step": 9180
+    },
+    {
+      "epoch": 0.21853280015233742,
+      "grad_norm": 2.477147785733827,
+      "learning_rate": 2e-05,
+      "loss": 0.2666,
+      "step": 9181
+    },
+    {
+      "epoch": 0.21855660287536893,
+      "grad_norm": 3.9426007326070915,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 9182
+    },
+    {
+      "epoch": 0.21858040559840045,
+      "grad_norm": 3.381300261023862,
+      "learning_rate": 2e-05,
+      "loss": 0.2038,
+      "step": 9183
+    },
+    {
+      "epoch": 0.21860420832143196,
+      "grad_norm": 6.90222620182804,
+      "learning_rate": 2e-05,
+      "loss": 0.2454,
+      "step": 9184
+    },
+    {
+      "epoch": 0.2186280110444635,
+      "grad_norm": 9.65148196021634,
+      "learning_rate": 2e-05,
+      "loss": 1.0867,
+      "step": 9185
+    },
+    {
+      "epoch": 0.218651813767495,
+      "grad_norm": 8.773744095645942,
+      "learning_rate": 2e-05,
+      "loss": 2.2103,
+      "step": 9186
+    },
+    {
+      "epoch": 0.21867561649052653,
+      "grad_norm": 8.898410376362083,
+      "learning_rate": 2e-05,
+      "loss": 0.6271,
+      "step": 9187
+    },
+    {
+      "epoch": 0.21869941921355804,
+      "grad_norm": 3.365407682682816,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 9188
+    },
+    {
+      "epoch": 0.21872322193658955,
+      "grad_norm": 2.256279944547478,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 9189
+    },
+    {
+      "epoch": 0.21874702465962106,
+      "grad_norm": 10.864213993194076,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 9190
+    },
+    {
+      "epoch": 0.21877082738265258,
+      "grad_norm": 5.103937237485337,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 9191
+    },
+    {
+      "epoch": 0.2187946301056841,
+      "grad_norm": 4.517591196891123,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 9192
+    },
+    {
+      "epoch": 0.2188184328287156,
+      "grad_norm": 4.9937783214167695,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 9193
+    },
+    {
+      "epoch": 0.21884223555174712,
+      "grad_norm": 6.903508333730782,
+      "learning_rate": 2e-05,
+      "loss": 0.9338,
+      "step": 9194
+    },
+    {
+      "epoch": 0.21886603827477863,
+      "grad_norm": 3.3456511420861155,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 9195
+    },
+    {
+      "epoch": 0.21888984099781014,
+      "grad_norm": 2.0774786824549585,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 9196
+    },
+    {
+      "epoch": 0.21891364372084166,
+      "grad_norm": 3.9470744587371978,
+      "learning_rate": 2e-05,
+      "loss": 0.2796,
+      "step": 9197
+    },
+    {
+      "epoch": 0.21893744644387317,
+      "grad_norm": 5.962078715078997,
+      "learning_rate": 2e-05,
+      "loss": 1.0017,
+      "step": 9198
+    },
+    {
+      "epoch": 0.21896124916690468,
+      "grad_norm": 3.0555050176524636,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 9199
+    },
+    {
+      "epoch": 0.21898505188993622,
+      "grad_norm": 5.9355184313071065,
+      "learning_rate": 2e-05,
+      "loss": 0.9577,
+      "step": 9200
+    },
+    {
+      "epoch": 0.21900885461296773,
+      "grad_norm": 4.033660087926153,
+      "learning_rate": 2e-05,
+      "loss": 0.1748,
+      "step": 9201
+    },
+    {
+      "epoch": 0.21903265733599925,
+      "grad_norm": 8.14315461803988,
+      "learning_rate": 2e-05,
+      "loss": 0.9708,
+      "step": 9202
+    },
+    {
+      "epoch": 0.21905646005903076,
+      "grad_norm": 1.1329259671685818,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 9203
+    },
+    {
+      "epoch": 0.21908026278206227,
+      "grad_norm": 1.548103039710399,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 9204
+    },
+    {
+      "epoch": 0.2191040655050938,
+      "grad_norm": 11.285488228974359,
+      "learning_rate": 2e-05,
+      "loss": 0.9712,
+      "step": 9205
+    },
+    {
+      "epoch": 0.2191278682281253,
+      "grad_norm": 4.103930741130917,
+      "learning_rate": 2e-05,
+      "loss": 0.2327,
+      "step": 9206
+    },
+    {
+      "epoch": 0.2191516709511568,
+      "grad_norm": 3.543026351618836,
+      "learning_rate": 2e-05,
+      "loss": 0.3691,
+      "step": 9207
+    },
+    {
+      "epoch": 0.21917547367418833,
+      "grad_norm": 2.421671924251515,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 9208
+    },
+    {
+      "epoch": 0.21919927639721984,
+      "grad_norm": 4.78231388853687,
+      "learning_rate": 2e-05,
+      "loss": 0.2221,
+      "step": 9209
+    },
+    {
+      "epoch": 0.21922307912025135,
+      "grad_norm": 9.412005302537494,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 9210
+    },
+    {
+      "epoch": 0.21924688184328286,
+      "grad_norm": 5.739952539908081,
+      "learning_rate": 2e-05,
+      "loss": 1.1533,
+      "step": 9211
+    },
+    {
+      "epoch": 0.21927068456631438,
+      "grad_norm": 9.142202497605346,
+      "learning_rate": 2e-05,
+      "loss": 0.2404,
+      "step": 9212
+    },
+    {
+      "epoch": 0.2192944872893459,
+      "grad_norm": 6.102126620882211,
+      "learning_rate": 2e-05,
+      "loss": 0.3264,
+      "step": 9213
+    },
+    {
+      "epoch": 0.2193182900123774,
+      "grad_norm": 6.041024033981629,
+      "learning_rate": 2e-05,
+      "loss": 0.2281,
+      "step": 9214
+    },
+    {
+      "epoch": 0.21934209273540894,
+      "grad_norm": 6.0475006475453235,
+      "learning_rate": 2e-05,
+      "loss": 1.2233,
+      "step": 9215
+    },
+    {
+      "epoch": 0.21936589545844046,
+      "grad_norm": 7.489032402302284,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 9216
+    },
+    {
+      "epoch": 0.21938969818147197,
+      "grad_norm": 3.7913613163292674,
+      "learning_rate": 2e-05,
+      "loss": 0.2914,
+      "step": 9217
+    },
+    {
+      "epoch": 0.21941350090450348,
+      "grad_norm": 2.5636462031166816,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 9218
+    },
+    {
+      "epoch": 0.219437303627535,
+      "grad_norm": 6.712241549129951,
+      "learning_rate": 2e-05,
+      "loss": 1.1287,
+      "step": 9219
+    },
+    {
+      "epoch": 0.2194611063505665,
+      "grad_norm": 10.407778290425153,
+      "learning_rate": 2e-05,
+      "loss": 0.4373,
+      "step": 9220
+    },
+    {
+      "epoch": 0.21948490907359802,
+      "grad_norm": 1.7285015835015267,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 9221
+    },
+    {
+      "epoch": 0.21950871179662954,
+      "grad_norm": 3.6977237876737385,
+      "learning_rate": 2e-05,
+      "loss": 0.6929,
+      "step": 9222
+    },
+    {
+      "epoch": 0.21953251451966105,
+      "grad_norm": 5.006407669649835,
+      "learning_rate": 2e-05,
+      "loss": 0.6851,
+      "step": 9223
+    },
+    {
+      "epoch": 0.21955631724269256,
+      "grad_norm": 4.160836240753634,
+      "learning_rate": 2e-05,
+      "loss": 0.5684,
+      "step": 9224
+    },
+    {
+      "epoch": 0.21958011996572407,
+      "grad_norm": 4.693280789532142,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 9225
+    },
+    {
+      "epoch": 0.2196039226887556,
+      "grad_norm": 3.1780906723607734,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 9226
+    },
+    {
+      "epoch": 0.2196277254117871,
+      "grad_norm": 6.534513095946905,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 9227
+    },
+    {
+      "epoch": 0.2196515281348186,
+      "grad_norm": 23.20660788950122,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 9228
+    },
+    {
+      "epoch": 0.21967533085785013,
+      "grad_norm": 5.73752571691023,
+      "learning_rate": 2e-05,
+      "loss": 0.2964,
+      "step": 9229
+    },
+    {
+      "epoch": 0.21969913358088164,
+      "grad_norm": 4.1455467156190196,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 9230
+    },
+    {
+      "epoch": 0.21972293630391318,
+      "grad_norm": 2.258643758181227,
+      "learning_rate": 2e-05,
+      "loss": 0.1331,
+      "step": 9231
+    },
+    {
+      "epoch": 0.2197467390269447,
+      "grad_norm": 5.519572147006365,
+      "learning_rate": 2e-05,
+      "loss": 0.2899,
+      "step": 9232
+    },
+    {
+      "epoch": 0.2197705417499762,
+      "grad_norm": 1.5542563938830372,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 9233
+    },
+    {
+      "epoch": 0.21979434447300772,
+      "grad_norm": 3.934505075628574,
+      "learning_rate": 2e-05,
+      "loss": 0.2857,
+      "step": 9234
+    },
+    {
+      "epoch": 0.21981814719603923,
+      "grad_norm": 2.999855502697223,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 9235
+    },
+    {
+      "epoch": 0.21984194991907074,
+      "grad_norm": 1.4065198752088293,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 9236
+    },
+    {
+      "epoch": 0.21986575264210226,
+      "grad_norm": 23.13286142006799,
+      "learning_rate": 2e-05,
+      "loss": 0.4573,
+      "step": 9237
+    },
+    {
+      "epoch": 0.21988955536513377,
+      "grad_norm": 5.5589272028654095,
+      "learning_rate": 2e-05,
+      "loss": 0.6635,
+      "step": 9238
+    },
+    {
+      "epoch": 0.21991335808816528,
+      "grad_norm": 5.817583170715041,
+      "learning_rate": 2e-05,
+      "loss": 1.0482,
+      "step": 9239
+    },
+    {
+      "epoch": 0.2199371608111968,
+      "grad_norm": 14.340797150436574,
+      "learning_rate": 2e-05,
+      "loss": 0.5871,
+      "step": 9240
+    },
+    {
+      "epoch": 0.2199609635342283,
+      "grad_norm": 5.6272320511939,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 9241
+    },
+    {
+      "epoch": 0.21998476625725982,
+      "grad_norm": 6.285744026545105,
+      "learning_rate": 2e-05,
+      "loss": 0.3629,
+      "step": 9242
+    },
+    {
+      "epoch": 0.22000856898029134,
+      "grad_norm": 7.164249863228292,
+      "learning_rate": 2e-05,
+      "loss": 0.8257,
+      "step": 9243
+    },
+    {
+      "epoch": 0.22003237170332285,
+      "grad_norm": 2.9214192344448557,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 9244
+    },
+    {
+      "epoch": 0.22005617442635436,
+      "grad_norm": 42.79875642792871,
+      "learning_rate": 2e-05,
+      "loss": 1.3758,
+      "step": 9245
+    },
+    {
+      "epoch": 0.2200799771493859,
+      "grad_norm": 11.870413614290532,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 9246
+    },
+    {
+      "epoch": 0.22010377987241742,
+      "grad_norm": 5.106208540509897,
+      "learning_rate": 2e-05,
+      "loss": 0.4579,
+      "step": 9247
+    },
+    {
+      "epoch": 0.22012758259544893,
+      "grad_norm": 2.719793595296606,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 9248
+    },
+    {
+      "epoch": 0.22015138531848044,
+      "grad_norm": 3.437058991222038,
+      "learning_rate": 2e-05,
+      "loss": 0.2196,
+      "step": 9249
+    },
+    {
+      "epoch": 0.22017518804151195,
+      "grad_norm": 4.348914902379812,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 9250
+    },
+    {
+      "epoch": 0.22019899076454347,
+      "grad_norm": 2.718754003941663,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 9251
+    },
+    {
+      "epoch": 0.22022279348757498,
+      "grad_norm": 5.722103694616124,
+      "learning_rate": 2e-05,
+      "loss": 1.03,
+      "step": 9252
+    },
+    {
+      "epoch": 0.2202465962106065,
+      "grad_norm": 17.303949969074395,
+      "learning_rate": 2e-05,
+      "loss": 0.9418,
+      "step": 9253
+    },
+    {
+      "epoch": 0.220270398933638,
+      "grad_norm": 4.7287449678530455,
+      "learning_rate": 2e-05,
+      "loss": 0.7132,
+      "step": 9254
+    },
+    {
+      "epoch": 0.22029420165666952,
+      "grad_norm": 3.0378041345922435,
+      "learning_rate": 2e-05,
+      "loss": 0.1506,
+      "step": 9255
+    },
+    {
+      "epoch": 0.22031800437970103,
+      "grad_norm": 4.201449883535181,
+      "learning_rate": 2e-05,
+      "loss": 0.1277,
+      "step": 9256
+    },
+    {
+      "epoch": 0.22034180710273255,
+      "grad_norm": 5.4125089107269355,
+      "learning_rate": 2e-05,
+      "loss": 0.5841,
+      "step": 9257
+    },
+    {
+      "epoch": 0.22036560982576406,
+      "grad_norm": 5.022185046784903,
+      "learning_rate": 2e-05,
+      "loss": 0.8244,
+      "step": 9258
+    },
+    {
+      "epoch": 0.22038941254879557,
+      "grad_norm": 22.540067456794464,
+      "learning_rate": 2e-05,
+      "loss": 0.7852,
+      "step": 9259
+    },
+    {
+      "epoch": 0.22041321527182708,
+      "grad_norm": 18.531526420688387,
+      "learning_rate": 2e-05,
+      "loss": 1.5334,
+      "step": 9260
+    },
+    {
+      "epoch": 0.22043701799485863,
+      "grad_norm": 8.988139057922613,
+      "learning_rate": 2e-05,
+      "loss": 0.4526,
+      "step": 9261
+    },
+    {
+      "epoch": 0.22046082071789014,
+      "grad_norm": 9.81848492232436,
+      "learning_rate": 2e-05,
+      "loss": 1.162,
+      "step": 9262
+    },
+    {
+      "epoch": 0.22048462344092165,
+      "grad_norm": 6.156490146436015,
+      "learning_rate": 2e-05,
+      "loss": 0.3645,
+      "step": 9263
+    },
+    {
+      "epoch": 0.22050842616395316,
+      "grad_norm": 7.5959365184730805,
+      "learning_rate": 2e-05,
+      "loss": 0.3723,
+      "step": 9264
+    },
+    {
+      "epoch": 0.22053222888698468,
+      "grad_norm": 3.664320153820524,
+      "learning_rate": 2e-05,
+      "loss": 0.3297,
+      "step": 9265
+    },
+    {
+      "epoch": 0.2205560316100162,
+      "grad_norm": 5.725688012481956,
+      "learning_rate": 2e-05,
+      "loss": 0.7779,
+      "step": 9266
+    },
+    {
+      "epoch": 0.2205798343330477,
+      "grad_norm": 2.49393486975217,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 9267
+    },
+    {
+      "epoch": 0.22060363705607922,
+      "grad_norm": 3.836469767424578,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 9268
+    },
+    {
+      "epoch": 0.22062743977911073,
+      "grad_norm": 5.579442110780747,
+      "learning_rate": 2e-05,
+      "loss": 0.5286,
+      "step": 9269
+    },
+    {
+      "epoch": 0.22065124250214224,
+      "grad_norm": 8.561661814154993,
+      "learning_rate": 2e-05,
+      "loss": 1.2335,
+      "step": 9270
+    },
+    {
+      "epoch": 0.22067504522517375,
+      "grad_norm": 4.13345441214748,
+      "learning_rate": 2e-05,
+      "loss": 0.1698,
+      "step": 9271
+    },
+    {
+      "epoch": 0.22069884794820527,
+      "grad_norm": 4.993565906041943,
+      "learning_rate": 2e-05,
+      "loss": 1.0159,
+      "step": 9272
+    },
+    {
+      "epoch": 0.22072265067123678,
+      "grad_norm": 6.818614133143525,
+      "learning_rate": 2e-05,
+      "loss": 1.2673,
+      "step": 9273
+    },
+    {
+      "epoch": 0.2207464533942683,
+      "grad_norm": 2.337554003605295,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 9274
+    },
+    {
+      "epoch": 0.2207702561172998,
+      "grad_norm": 3.5448927285164262,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 9275
+    },
+    {
+      "epoch": 0.22079405884033135,
+      "grad_norm": 3.1082441465237522,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 9276
+    },
+    {
+      "epoch": 0.22081786156336286,
+      "grad_norm": 2.963826179494825,
+      "learning_rate": 2e-05,
+      "loss": 0.1431,
+      "step": 9277
+    },
+    {
+      "epoch": 0.22084166428639437,
+      "grad_norm": 5.522082489813046,
+      "learning_rate": 2e-05,
+      "loss": 0.3182,
+      "step": 9278
+    },
+    {
+      "epoch": 0.2208654670094259,
+      "grad_norm": 2.1678049677800075,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 9279
+    },
+    {
+      "epoch": 0.2208892697324574,
+      "grad_norm": 3.402940147456283,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2209130724554889,
+      "grad_norm": 4.524808990182539,
+      "learning_rate": 2e-05,
+      "loss": 0.6072,
+      "step": 9281
+    },
+    {
+      "epoch": 0.22093687517852043,
+      "grad_norm": 11.572819397258266,
+      "learning_rate": 2e-05,
+      "loss": 0.3764,
+      "step": 9282
+    },
+    {
+      "epoch": 0.22096067790155194,
+      "grad_norm": 7.455045432670181,
+      "learning_rate": 2e-05,
+      "loss": 1.6885,
+      "step": 9283
+    },
+    {
+      "epoch": 0.22098448062458345,
+      "grad_norm": 5.7282908926247655,
+      "learning_rate": 2e-05,
+      "loss": 0.7066,
+      "step": 9284
+    },
+    {
+      "epoch": 0.22100828334761496,
+      "grad_norm": 9.727360776216381,
+      "learning_rate": 2e-05,
+      "loss": 0.5101,
+      "step": 9285
+    },
+    {
+      "epoch": 0.22103208607064648,
+      "grad_norm": 2.086932571162359,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 9286
+    },
+    {
+      "epoch": 0.221055888793678,
+      "grad_norm": 3.493587300410978,
+      "learning_rate": 2e-05,
+      "loss": 0.1799,
+      "step": 9287
+    },
+    {
+      "epoch": 0.2210796915167095,
+      "grad_norm": 6.481501398220084,
+      "learning_rate": 2e-05,
+      "loss": 0.7914,
+      "step": 9288
+    },
+    {
+      "epoch": 0.22110349423974102,
+      "grad_norm": 11.841283242728752,
+      "learning_rate": 2e-05,
+      "loss": 0.5047,
+      "step": 9289
+    },
+    {
+      "epoch": 0.22112729696277253,
+      "grad_norm": 2.5840583003820257,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 9290
+    },
+    {
+      "epoch": 0.22115109968580404,
+      "grad_norm": 4.41927699987353,
+      "learning_rate": 2e-05,
+      "loss": 0.2731,
+      "step": 9291
+    },
+    {
+      "epoch": 0.22117490240883558,
+      "grad_norm": 6.957723299205847,
+      "learning_rate": 2e-05,
+      "loss": 0.6076,
+      "step": 9292
+    },
+    {
+      "epoch": 0.2211987051318671,
+      "grad_norm": 12.964147093897077,
+      "learning_rate": 2e-05,
+      "loss": 0.3328,
+      "step": 9293
+    },
+    {
+      "epoch": 0.2212225078548986,
+      "grad_norm": 5.634442233826975,
+      "learning_rate": 2e-05,
+      "loss": 0.7247,
+      "step": 9294
+    },
+    {
+      "epoch": 0.22124631057793012,
+      "grad_norm": 3.6908651046954413,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 9295
+    },
+    {
+      "epoch": 0.22127011330096163,
+      "grad_norm": 5.591502533593843,
+      "learning_rate": 2e-05,
+      "loss": 0.2591,
+      "step": 9296
+    },
+    {
+      "epoch": 0.22129391602399315,
+      "grad_norm": 1.88488482578702,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 9297
+    },
+    {
+      "epoch": 0.22131771874702466,
+      "grad_norm": 1.7688321735090855,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 9298
+    },
+    {
+      "epoch": 0.22134152147005617,
+      "grad_norm": 3.687750356409038,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 9299
+    },
+    {
+      "epoch": 0.2213653241930877,
+      "grad_norm": 2.6195381534434916,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 9300
+    },
+    {
+      "epoch": 0.2213891269161192,
+      "grad_norm": 4.936491493081963,
+      "learning_rate": 2e-05,
+      "loss": 1.1543,
+      "step": 9301
+    },
+    {
+      "epoch": 0.2214129296391507,
+      "grad_norm": 7.2528478332863,
+      "learning_rate": 2e-05,
+      "loss": 1.1477,
+      "step": 9302
+    },
+    {
+      "epoch": 0.22143673236218223,
+      "grad_norm": 6.784748328632717,
+      "learning_rate": 2e-05,
+      "loss": 0.2115,
+      "step": 9303
+    },
+    {
+      "epoch": 0.22146053508521374,
+      "grad_norm": 2.0023550755007316,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 9304
+    },
+    {
+      "epoch": 0.22148433780824525,
+      "grad_norm": 1.0349362998475866,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 9305
+    },
+    {
+      "epoch": 0.22150814053127676,
+      "grad_norm": 14.57262446221145,
+      "learning_rate": 2e-05,
+      "loss": 1.4289,
+      "step": 9306
+    },
+    {
+      "epoch": 0.2215319432543083,
+      "grad_norm": 6.253003835113935,
+      "learning_rate": 2e-05,
+      "loss": 0.7956,
+      "step": 9307
+    },
+    {
+      "epoch": 0.22155574597733982,
+      "grad_norm": 9.346840481233503,
+      "learning_rate": 2e-05,
+      "loss": 0.513,
+      "step": 9308
+    },
+    {
+      "epoch": 0.22157954870037133,
+      "grad_norm": 3.3396256401737285,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 9309
+    },
+    {
+      "epoch": 0.22160335142340284,
+      "grad_norm": 9.884042852038167,
+      "learning_rate": 2e-05,
+      "loss": 1.3957,
+      "step": 9310
+    },
+    {
+      "epoch": 0.22162715414643436,
+      "grad_norm": 5.8525586068088975,
+      "learning_rate": 2e-05,
+      "loss": 0.3327,
+      "step": 9311
+    },
+    {
+      "epoch": 0.22165095686946587,
+      "grad_norm": 4.281808557340688,
+      "learning_rate": 2e-05,
+      "loss": 0.1175,
+      "step": 9312
+    },
+    {
+      "epoch": 0.22167475959249738,
+      "grad_norm": 6.616908178066094,
+      "learning_rate": 2e-05,
+      "loss": 0.2459,
+      "step": 9313
+    },
+    {
+      "epoch": 0.2216985623155289,
+      "grad_norm": 2.700370169638153,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 9314
+    },
+    {
+      "epoch": 0.2217223650385604,
+      "grad_norm": 6.096014056884958,
+      "learning_rate": 2e-05,
+      "loss": 0.7199,
+      "step": 9315
+    },
+    {
+      "epoch": 0.22174616776159192,
+      "grad_norm": 4.481915536997183,
+      "learning_rate": 2e-05,
+      "loss": 0.5827,
+      "step": 9316
+    },
+    {
+      "epoch": 0.22176997048462344,
+      "grad_norm": 5.264918707879802,
+      "learning_rate": 2e-05,
+      "loss": 0.5077,
+      "step": 9317
+    },
+    {
+      "epoch": 0.22179377320765495,
+      "grad_norm": 1.2416658087896197,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 9318
+    },
+    {
+      "epoch": 0.22181757593068646,
+      "grad_norm": 3.8407280868908606,
+      "learning_rate": 2e-05,
+      "loss": 0.2114,
+      "step": 9319
+    },
+    {
+      "epoch": 0.22184137865371797,
+      "grad_norm": 10.802519700074376,
+      "learning_rate": 2e-05,
+      "loss": 0.6217,
+      "step": 9320
+    },
+    {
+      "epoch": 0.2218651813767495,
+      "grad_norm": 8.851470736370986,
+      "learning_rate": 2e-05,
+      "loss": 0.907,
+      "step": 9321
+    },
+    {
+      "epoch": 0.22188898409978103,
+      "grad_norm": 4.303326164943059,
+      "learning_rate": 2e-05,
+      "loss": 0.6602,
+      "step": 9322
+    },
+    {
+      "epoch": 0.22191278682281254,
+      "grad_norm": 11.590329901955712,
+      "learning_rate": 2e-05,
+      "loss": 1.3444,
+      "step": 9323
+    },
+    {
+      "epoch": 0.22193658954584405,
+      "grad_norm": 9.39777671018022,
+      "learning_rate": 2e-05,
+      "loss": 1.1489,
+      "step": 9324
+    },
+    {
+      "epoch": 0.22196039226887557,
+      "grad_norm": 9.707439897667944,
+      "learning_rate": 2e-05,
+      "loss": 0.6798,
+      "step": 9325
+    },
+    {
+      "epoch": 0.22198419499190708,
+      "grad_norm": 9.30847515510017,
+      "learning_rate": 2e-05,
+      "loss": 0.8339,
+      "step": 9326
+    },
+    {
+      "epoch": 0.2220079977149386,
+      "grad_norm": 2.040454294715865,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 9327
+    },
+    {
+      "epoch": 0.2220318004379701,
+      "grad_norm": 7.522311130479002,
+      "learning_rate": 2e-05,
+      "loss": 0.5993,
+      "step": 9328
+    },
+    {
+      "epoch": 0.22205560316100162,
+      "grad_norm": 3.0841192169847043,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 9329
+    },
+    {
+      "epoch": 0.22207940588403313,
+      "grad_norm": 6.209482306577013,
+      "learning_rate": 2e-05,
+      "loss": 1.0155,
+      "step": 9330
+    },
+    {
+      "epoch": 0.22210320860706464,
+      "grad_norm": 8.01164875616016,
+      "learning_rate": 2e-05,
+      "loss": 0.3236,
+      "step": 9331
+    },
+    {
+      "epoch": 0.22212701133009616,
+      "grad_norm": 7.93383236911376,
+      "learning_rate": 2e-05,
+      "loss": 0.4501,
+      "step": 9332
+    },
+    {
+      "epoch": 0.22215081405312767,
+      "grad_norm": 4.534226620698889,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 9333
+    },
+    {
+      "epoch": 0.22217461677615918,
+      "grad_norm": 9.21343879997276,
+      "learning_rate": 2e-05,
+      "loss": 1.2822,
+      "step": 9334
+    },
+    {
+      "epoch": 0.2221984194991907,
+      "grad_norm": 6.928575632455533,
+      "learning_rate": 2e-05,
+      "loss": 0.4457,
+      "step": 9335
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 2.288665950596951,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 9336
+    },
+    {
+      "epoch": 0.22224602494525375,
+      "grad_norm": 5.92801139000567,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 9337
+    },
+    {
+      "epoch": 0.22226982766828526,
+      "grad_norm": 27.18274882928978,
+      "learning_rate": 2e-05,
+      "loss": 0.9933,
+      "step": 9338
+    },
+    {
+      "epoch": 0.22229363039131678,
+      "grad_norm": 7.967105389080205,
+      "learning_rate": 2e-05,
+      "loss": 0.8378,
+      "step": 9339
+    },
+    {
+      "epoch": 0.2223174331143483,
+      "grad_norm": 6.130210053001333,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2223412358373798,
+      "grad_norm": 1.8950600975481728,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 9341
+    },
+    {
+      "epoch": 0.22236503856041132,
+      "grad_norm": 5.3123926720999854,
+      "learning_rate": 2e-05,
+      "loss": 0.8602,
+      "step": 9342
+    },
+    {
+      "epoch": 0.22238884128344283,
+      "grad_norm": 5.494415479297195,
+      "learning_rate": 2e-05,
+      "loss": 0.781,
+      "step": 9343
+    },
+    {
+      "epoch": 0.22241264400647434,
+      "grad_norm": 8.721656783679414,
+      "learning_rate": 2e-05,
+      "loss": 0.7608,
+      "step": 9344
+    },
+    {
+      "epoch": 0.22243644672950585,
+      "grad_norm": 6.3659168673062165,
+      "learning_rate": 2e-05,
+      "loss": 0.3797,
+      "step": 9345
+    },
+    {
+      "epoch": 0.22246024945253737,
+      "grad_norm": 2.6107123018682334,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 9346
+    },
+    {
+      "epoch": 0.22248405217556888,
+      "grad_norm": 4.24356482707563,
+      "learning_rate": 2e-05,
+      "loss": 0.6187,
+      "step": 9347
+    },
+    {
+      "epoch": 0.2225078548986004,
+      "grad_norm": 3.775821046115118,
+      "learning_rate": 2e-05,
+      "loss": 0.6425,
+      "step": 9348
+    },
+    {
+      "epoch": 0.2225316576216319,
+      "grad_norm": 3.0976109707350807,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 9349
+    },
+    {
+      "epoch": 0.22255546034466342,
+      "grad_norm": 5.893726861463623,
+      "learning_rate": 2e-05,
+      "loss": 0.3179,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22257926306769493,
+      "grad_norm": 5.000726332861866,
+      "learning_rate": 2e-05,
+      "loss": 0.5794,
+      "step": 9351
+    },
+    {
+      "epoch": 0.22260306579072645,
+      "grad_norm": 7.092615547265955,
+      "learning_rate": 2e-05,
+      "loss": 0.2289,
+      "step": 9352
+    },
+    {
+      "epoch": 0.222626868513758,
+      "grad_norm": 4.401005311482442,
+      "learning_rate": 2e-05,
+      "loss": 0.5832,
+      "step": 9353
+    },
+    {
+      "epoch": 0.2226506712367895,
+      "grad_norm": 9.78310919895005,
+      "learning_rate": 2e-05,
+      "loss": 1.0577,
+      "step": 9354
+    },
+    {
+      "epoch": 0.222674473959821,
+      "grad_norm": 4.80524386806689,
+      "learning_rate": 2e-05,
+      "loss": 0.6793,
+      "step": 9355
+    },
+    {
+      "epoch": 0.22269827668285253,
+      "grad_norm": 1.7761858055294915,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 9356
+    },
+    {
+      "epoch": 0.22272207940588404,
+      "grad_norm": 5.202625891485068,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 9357
+    },
+    {
+      "epoch": 0.22274588212891555,
+      "grad_norm": 1.115755672192668,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 9358
+    },
+    {
+      "epoch": 0.22276968485194706,
+      "grad_norm": 4.281272506806463,
+      "learning_rate": 2e-05,
+      "loss": 0.343,
+      "step": 9359
+    },
+    {
+      "epoch": 0.22279348757497858,
+      "grad_norm": 4.132612011882025,
+      "learning_rate": 2e-05,
+      "loss": 0.5898,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2228172902980101,
+      "grad_norm": 5.067571863136761,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 9361
+    },
+    {
+      "epoch": 0.2228410930210416,
+      "grad_norm": 11.920548258065063,
+      "learning_rate": 2e-05,
+      "loss": 0.8741,
+      "step": 9362
+    },
+    {
+      "epoch": 0.22286489574407312,
+      "grad_norm": 4.728552900326577,
+      "learning_rate": 2e-05,
+      "loss": 0.1893,
+      "step": 9363
+    },
+    {
+      "epoch": 0.22288869846710463,
+      "grad_norm": 2.8785733298621294,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 9364
+    },
+    {
+      "epoch": 0.22291250119013614,
+      "grad_norm": 7.878839487825683,
+      "learning_rate": 2e-05,
+      "loss": 1.1992,
+      "step": 9365
+    },
+    {
+      "epoch": 0.22293630391316765,
+      "grad_norm": 4.818797244033894,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 9366
+    },
+    {
+      "epoch": 0.22296010663619917,
+      "grad_norm": 5.344377202254461,
+      "learning_rate": 2e-05,
+      "loss": 0.7343,
+      "step": 9367
+    },
+    {
+      "epoch": 0.2229839093592307,
+      "grad_norm": 4.532347600536317,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 9368
+    },
+    {
+      "epoch": 0.22300771208226222,
+      "grad_norm": 3.0028064031058412,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 9369
+    },
+    {
+      "epoch": 0.22303151480529373,
+      "grad_norm": 6.330050200634951,
+      "learning_rate": 2e-05,
+      "loss": 1.0211,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22305531752832525,
+      "grad_norm": 2.0948966203448314,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 9371
+    },
+    {
+      "epoch": 0.22307912025135676,
+      "grad_norm": 7.802751125221752,
+      "learning_rate": 2e-05,
+      "loss": 1.1343,
+      "step": 9372
+    },
+    {
+      "epoch": 0.22310292297438827,
+      "grad_norm": 6.0638409378152485,
+      "learning_rate": 2e-05,
+      "loss": 1.0233,
+      "step": 9373
+    },
+    {
+      "epoch": 0.2231267256974198,
+      "grad_norm": 1.238221735501709,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 9374
+    },
+    {
+      "epoch": 0.2231505284204513,
+      "grad_norm": 1.721346851981914,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 9375
+    },
+    {
+      "epoch": 0.2231743311434828,
+      "grad_norm": 7.5408282703612395,
+      "learning_rate": 2e-05,
+      "loss": 0.7524,
+      "step": 9376
+    },
+    {
+      "epoch": 0.22319813386651433,
+      "grad_norm": 9.430087654753365,
+      "learning_rate": 2e-05,
+      "loss": 0.6728,
+      "step": 9377
+    },
+    {
+      "epoch": 0.22322193658954584,
+      "grad_norm": 4.162269019277165,
+      "learning_rate": 2e-05,
+      "loss": 0.6383,
+      "step": 9378
+    },
+    {
+      "epoch": 0.22324573931257735,
+      "grad_norm": 5.787895769051602,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 9379
+    },
+    {
+      "epoch": 0.22326954203560886,
+      "grad_norm": 4.0985684591029425,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22329334475864038,
+      "grad_norm": 6.4543570918166395,
+      "learning_rate": 2e-05,
+      "loss": 1.6379,
+      "step": 9381
+    },
+    {
+      "epoch": 0.2233171474816719,
+      "grad_norm": 6.537408831641343,
+      "learning_rate": 2e-05,
+      "loss": 0.3239,
+      "step": 9382
+    },
+    {
+      "epoch": 0.22334095020470343,
+      "grad_norm": 6.8125700626365955,
+      "learning_rate": 2e-05,
+      "loss": 0.6463,
+      "step": 9383
+    },
+    {
+      "epoch": 0.22336475292773494,
+      "grad_norm": 3.846765063531516,
+      "learning_rate": 2e-05,
+      "loss": 0.1785,
+      "step": 9384
+    },
+    {
+      "epoch": 0.22338855565076646,
+      "grad_norm": 2.1919799623453056,
+      "learning_rate": 2e-05,
+      "loss": 0.1415,
+      "step": 9385
+    },
+    {
+      "epoch": 0.22341235837379797,
+      "grad_norm": 1.7167344723329352,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 9386
+    },
+    {
+      "epoch": 0.22343616109682948,
+      "grad_norm": 10.523983696052452,
+      "learning_rate": 2e-05,
+      "loss": 0.5794,
+      "step": 9387
+    },
+    {
+      "epoch": 0.223459963819861,
+      "grad_norm": 6.964104577962277,
+      "learning_rate": 2e-05,
+      "loss": 0.9418,
+      "step": 9388
+    },
+    {
+      "epoch": 0.2234837665428925,
+      "grad_norm": 8.158333386613126,
+      "learning_rate": 2e-05,
+      "loss": 1.4003,
+      "step": 9389
+    },
+    {
+      "epoch": 0.22350756926592402,
+      "grad_norm": 6.875066493026005,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22353137198895554,
+      "grad_norm": 6.494446152960387,
+      "learning_rate": 2e-05,
+      "loss": 0.4146,
+      "step": 9391
+    },
+    {
+      "epoch": 0.22355517471198705,
+      "grad_norm": 6.6846043005086475,
+      "learning_rate": 2e-05,
+      "loss": 0.3142,
+      "step": 9392
+    },
+    {
+      "epoch": 0.22357897743501856,
+      "grad_norm": 1.2716717974496359,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 9393
+    },
+    {
+      "epoch": 0.22360278015805007,
+      "grad_norm": 5.540048439741116,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 9394
+    },
+    {
+      "epoch": 0.2236265828810816,
+      "grad_norm": 7.122401916210293,
+      "learning_rate": 2e-05,
+      "loss": 0.2601,
+      "step": 9395
+    },
+    {
+      "epoch": 0.2236503856041131,
+      "grad_norm": 9.126942619890649,
+      "learning_rate": 2e-05,
+      "loss": 1.4001,
+      "step": 9396
+    },
+    {
+      "epoch": 0.2236741883271446,
+      "grad_norm": 5.682430905669514,
+      "learning_rate": 2e-05,
+      "loss": 0.37,
+      "step": 9397
+    },
+    {
+      "epoch": 0.22369799105017615,
+      "grad_norm": 3.150648304690772,
+      "learning_rate": 2e-05,
+      "loss": 0.151,
+      "step": 9398
+    },
+    {
+      "epoch": 0.22372179377320767,
+      "grad_norm": 3.2005267643508395,
+      "learning_rate": 2e-05,
+      "loss": 0.4226,
+      "step": 9399
+    },
+    {
+      "epoch": 0.22374559649623918,
+      "grad_norm": 2.902316471366838,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 9400
+    },
+    {
+      "epoch": 0.2237693992192707,
+      "grad_norm": 10.204708949296048,
+      "learning_rate": 2e-05,
+      "loss": 0.7714,
+      "step": 9401
+    },
+    {
+      "epoch": 0.2237932019423022,
+      "grad_norm": 4.42471127037217,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 9402
+    },
+    {
+      "epoch": 0.22381700466533372,
+      "grad_norm": 4.902491208369738,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 9403
+    },
+    {
+      "epoch": 0.22384080738836523,
+      "grad_norm": 1.378551523843144,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 9404
+    },
+    {
+      "epoch": 0.22386461011139674,
+      "grad_norm": 2.120776211180105,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 9405
+    },
+    {
+      "epoch": 0.22388841283442826,
+      "grad_norm": 8.372164597757306,
+      "learning_rate": 2e-05,
+      "loss": 0.9729,
+      "step": 9406
+    },
+    {
+      "epoch": 0.22391221555745977,
+      "grad_norm": 3.6851026156296007,
+      "learning_rate": 2e-05,
+      "loss": 0.5587,
+      "step": 9407
+    },
+    {
+      "epoch": 0.22393601828049128,
+      "grad_norm": 6.394910789046901,
+      "learning_rate": 2e-05,
+      "loss": 0.3037,
+      "step": 9408
+    },
+    {
+      "epoch": 0.2239598210035228,
+      "grad_norm": 6.0366482928085485,
+      "learning_rate": 2e-05,
+      "loss": 1.1883,
+      "step": 9409
+    },
+    {
+      "epoch": 0.2239836237265543,
+      "grad_norm": 3.043607218077324,
+      "learning_rate": 2e-05,
+      "loss": 0.1867,
+      "step": 9410
+    },
+    {
+      "epoch": 0.22400742644958582,
+      "grad_norm": 4.061324809463281,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 9411
+    },
+    {
+      "epoch": 0.22403122917261734,
+      "grad_norm": 1.4337198009387218,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 9412
+    },
+    {
+      "epoch": 0.22405503189564885,
+      "grad_norm": 5.49644836236473,
+      "learning_rate": 2e-05,
+      "loss": 0.1984,
+      "step": 9413
+    },
+    {
+      "epoch": 0.2240788346186804,
+      "grad_norm": 2.016914782992675,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 9414
+    },
+    {
+      "epoch": 0.2241026373417119,
+      "grad_norm": 3.4590702361078653,
+      "learning_rate": 2e-05,
+      "loss": 0.3376,
+      "step": 9415
+    },
+    {
+      "epoch": 0.22412644006474342,
+      "grad_norm": 4.269129389089989,
+      "learning_rate": 2e-05,
+      "loss": 0.2762,
+      "step": 9416
+    },
+    {
+      "epoch": 0.22415024278777493,
+      "grad_norm": 2.9843445295258864,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 9417
+    },
+    {
+      "epoch": 0.22417404551080644,
+      "grad_norm": 5.213015335144265,
+      "learning_rate": 2e-05,
+      "loss": 0.1794,
+      "step": 9418
+    },
+    {
+      "epoch": 0.22419784823383795,
+      "grad_norm": 4.469632137793048,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 9419
+    },
+    {
+      "epoch": 0.22422165095686947,
+      "grad_norm": 7.282202714226415,
+      "learning_rate": 2e-05,
+      "loss": 0.2398,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22424545367990098,
+      "grad_norm": 3.8374521568980255,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 9421
+    },
+    {
+      "epoch": 0.2242692564029325,
+      "grad_norm": 5.28337251245336,
+      "learning_rate": 2e-05,
+      "loss": 1.0011,
+      "step": 9422
+    },
+    {
+      "epoch": 0.224293059125964,
+      "grad_norm": 1.974377799291156,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 9423
+    },
+    {
+      "epoch": 0.22431686184899552,
+      "grad_norm": 7.796543771514978,
+      "learning_rate": 2e-05,
+      "loss": 1.1291,
+      "step": 9424
+    },
+    {
+      "epoch": 0.22434066457202703,
+      "grad_norm": 4.336279336077328,
+      "learning_rate": 2e-05,
+      "loss": 0.2478,
+      "step": 9425
+    },
+    {
+      "epoch": 0.22436446729505855,
+      "grad_norm": 4.560750969604584,
+      "learning_rate": 2e-05,
+      "loss": 0.4793,
+      "step": 9426
+    },
+    {
+      "epoch": 0.22438827001809006,
+      "grad_norm": 5.234086385627883,
+      "learning_rate": 2e-05,
+      "loss": 0.9342,
+      "step": 9427
+    },
+    {
+      "epoch": 0.22441207274112157,
+      "grad_norm": 7.039239447377369,
+      "learning_rate": 2e-05,
+      "loss": 0.3245,
+      "step": 9428
+    },
+    {
+      "epoch": 0.2244358754641531,
+      "grad_norm": 4.492076413779023,
+      "learning_rate": 2e-05,
+      "loss": 0.5157,
+      "step": 9429
+    },
+    {
+      "epoch": 0.22445967818718462,
+      "grad_norm": 6.57024529874543,
+      "learning_rate": 2e-05,
+      "loss": 0.6788,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22448348091021614,
+      "grad_norm": 2.867236564865635,
+      "learning_rate": 2e-05,
+      "loss": 0.1864,
+      "step": 9431
+    },
+    {
+      "epoch": 0.22450728363324765,
+      "grad_norm": 6.118658550603622,
+      "learning_rate": 2e-05,
+      "loss": 0.6449,
+      "step": 9432
+    },
+    {
+      "epoch": 0.22453108635627916,
+      "grad_norm": 5.119245447178267,
+      "learning_rate": 2e-05,
+      "loss": 0.5964,
+      "step": 9433
+    },
+    {
+      "epoch": 0.22455488907931068,
+      "grad_norm": 7.138257035460392,
+      "learning_rate": 2e-05,
+      "loss": 0.4596,
+      "step": 9434
+    },
+    {
+      "epoch": 0.2245786918023422,
+      "grad_norm": 5.593363577402494,
+      "learning_rate": 2e-05,
+      "loss": 0.8094,
+      "step": 9435
+    },
+    {
+      "epoch": 0.2246024945253737,
+      "grad_norm": 5.4747228112935655,
+      "learning_rate": 2e-05,
+      "loss": 0.9793,
+      "step": 9436
+    },
+    {
+      "epoch": 0.22462629724840522,
+      "grad_norm": 9.985455932611599,
+      "learning_rate": 2e-05,
+      "loss": 0.5154,
+      "step": 9437
+    },
+    {
+      "epoch": 0.22465009997143673,
+      "grad_norm": 2.740909684373835,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 9438
+    },
+    {
+      "epoch": 0.22467390269446824,
+      "grad_norm": 2.3127960350833443,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 9439
+    },
+    {
+      "epoch": 0.22469770541749975,
+      "grad_norm": 5.652261940865218,
+      "learning_rate": 2e-05,
+      "loss": 0.891,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22472150814053127,
+      "grad_norm": 4.156257817096896,
+      "learning_rate": 2e-05,
+      "loss": 0.3086,
+      "step": 9441
+    },
+    {
+      "epoch": 0.22474531086356278,
+      "grad_norm": 4.525750781942471,
+      "learning_rate": 2e-05,
+      "loss": 0.7625,
+      "step": 9442
+    },
+    {
+      "epoch": 0.2247691135865943,
+      "grad_norm": 6.731358480014902,
+      "learning_rate": 2e-05,
+      "loss": 0.3744,
+      "step": 9443
+    },
+    {
+      "epoch": 0.22479291630962583,
+      "grad_norm": 4.873855589940729,
+      "learning_rate": 2e-05,
+      "loss": 0.9008,
+      "step": 9444
+    },
+    {
+      "epoch": 0.22481671903265735,
+      "grad_norm": 4.994376893399382,
+      "learning_rate": 2e-05,
+      "loss": 0.7588,
+      "step": 9445
+    },
+    {
+      "epoch": 0.22484052175568886,
+      "grad_norm": 3.0008258742688154,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 9446
+    },
+    {
+      "epoch": 0.22486432447872037,
+      "grad_norm": 4.338252796796181,
+      "learning_rate": 2e-05,
+      "loss": 0.2714,
+      "step": 9447
+    },
+    {
+      "epoch": 0.2248881272017519,
+      "grad_norm": 4.32892510575786,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 9448
+    },
+    {
+      "epoch": 0.2249119299247834,
+      "grad_norm": 6.5766088233092725,
+      "learning_rate": 2e-05,
+      "loss": 0.3133,
+      "step": 9449
+    },
+    {
+      "epoch": 0.2249357326478149,
+      "grad_norm": 3.4609190439620496,
+      "learning_rate": 2e-05,
+      "loss": 0.5609,
+      "step": 9450
+    },
+    {
+      "epoch": 0.22495953537084643,
+      "grad_norm": 5.48572169729774,
+      "learning_rate": 2e-05,
+      "loss": 0.8575,
+      "step": 9451
+    },
+    {
+      "epoch": 0.22498333809387794,
+      "grad_norm": 2.934368786217182,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 9452
+    },
+    {
+      "epoch": 0.22500714081690945,
+      "grad_norm": 2.0259173113026794,
+      "learning_rate": 2e-05,
+      "loss": 0.1156,
+      "step": 9453
+    },
+    {
+      "epoch": 0.22503094353994096,
+      "grad_norm": 1.0711319803362018,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 9454
+    },
+    {
+      "epoch": 0.22505474626297248,
+      "grad_norm": 3.99496130731075,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 9455
+    },
+    {
+      "epoch": 0.225078548986004,
+      "grad_norm": 6.017090943119713,
+      "learning_rate": 2e-05,
+      "loss": 0.1528,
+      "step": 9456
+    },
+    {
+      "epoch": 0.2251023517090355,
+      "grad_norm": 1.7154631281523607,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 9457
+    },
+    {
+      "epoch": 0.22512615443206702,
+      "grad_norm": 7.170981702974083,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 9458
+    },
+    {
+      "epoch": 0.22514995715509856,
+      "grad_norm": 6.413181816539093,
+      "learning_rate": 2e-05,
+      "loss": 0.965,
+      "step": 9459
+    },
+    {
+      "epoch": 0.22517375987813007,
+      "grad_norm": 3.369392527207135,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 9460
+    },
+    {
+      "epoch": 0.22519756260116158,
+      "grad_norm": 2.8646495392915283,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 9461
+    },
+    {
+      "epoch": 0.2252213653241931,
+      "grad_norm": 5.654793619611316,
+      "learning_rate": 2e-05,
+      "loss": 0.6887,
+      "step": 9462
+    },
+    {
+      "epoch": 0.2252451680472246,
+      "grad_norm": 1.381847287621693,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 9463
+    },
+    {
+      "epoch": 0.22526897077025612,
+      "grad_norm": 13.128824652911971,
+      "learning_rate": 2e-05,
+      "loss": 1.2781,
+      "step": 9464
+    },
+    {
+      "epoch": 0.22529277349328763,
+      "grad_norm": 3.528062352991287,
+      "learning_rate": 2e-05,
+      "loss": 0.443,
+      "step": 9465
+    },
+    {
+      "epoch": 0.22531657621631915,
+      "grad_norm": 6.1706933692845025,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 9466
+    },
+    {
+      "epoch": 0.22534037893935066,
+      "grad_norm": 3.394867228786981,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 9467
+    },
+    {
+      "epoch": 0.22536418166238217,
+      "grad_norm": 5.053168419664101,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 9468
+    },
+    {
+      "epoch": 0.2253879843854137,
+      "grad_norm": 3.789428661345403,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 9469
+    },
+    {
+      "epoch": 0.2254117871084452,
+      "grad_norm": 13.23400689950211,
+      "learning_rate": 2e-05,
+      "loss": 0.7055,
+      "step": 9470
+    },
+    {
+      "epoch": 0.2254355898314767,
+      "grad_norm": 5.402055545691871,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 9471
+    },
+    {
+      "epoch": 0.22545939255450823,
+      "grad_norm": 8.42747572994449,
+      "learning_rate": 2e-05,
+      "loss": 0.3697,
+      "step": 9472
+    },
+    {
+      "epoch": 0.22548319527753974,
+      "grad_norm": 8.755507168884861,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 9473
+    },
+    {
+      "epoch": 0.22550699800057125,
+      "grad_norm": 21.619712228832583,
+      "learning_rate": 2e-05,
+      "loss": 0.1604,
+      "step": 9474
+    },
+    {
+      "epoch": 0.2255308007236028,
+      "grad_norm": 10.549446574858662,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 9475
+    },
+    {
+      "epoch": 0.2255546034466343,
+      "grad_norm": 1.9679014845148748,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 9476
+    },
+    {
+      "epoch": 0.22557840616966582,
+      "grad_norm": 4.784902821074081,
+      "learning_rate": 2e-05,
+      "loss": 0.7109,
+      "step": 9477
+    },
+    {
+      "epoch": 0.22560220889269733,
+      "grad_norm": 6.753709652424266,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 9478
+    },
+    {
+      "epoch": 0.22562601161572884,
+      "grad_norm": 9.975560626794172,
+      "learning_rate": 2e-05,
+      "loss": 0.2187,
+      "step": 9479
+    },
+    {
+      "epoch": 0.22564981433876036,
+      "grad_norm": 2.334588384031532,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22567361706179187,
+      "grad_norm": 30.84868346478998,
+      "learning_rate": 2e-05,
+      "loss": 1.3641,
+      "step": 9481
+    },
+    {
+      "epoch": 0.22569741978482338,
+      "grad_norm": 10.169015946188233,
+      "learning_rate": 2e-05,
+      "loss": 0.9431,
+      "step": 9482
+    },
+    {
+      "epoch": 0.2257212225078549,
+      "grad_norm": 64.15784931014171,
+      "learning_rate": 2e-05,
+      "loss": 0.77,
+      "step": 9483
+    },
+    {
+      "epoch": 0.2257450252308864,
+      "grad_norm": 56.56445892119919,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 9484
+    },
+    {
+      "epoch": 0.22576882795391792,
+      "grad_norm": 3.1019197473668827,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 9485
+    },
+    {
+      "epoch": 0.22579263067694944,
+      "grad_norm": 8.330230804635708,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 9486
+    },
+    {
+      "epoch": 0.22581643339998095,
+      "grad_norm": 6.150861349349793,
+      "learning_rate": 2e-05,
+      "loss": 1.6248,
+      "step": 9487
+    },
+    {
+      "epoch": 0.22584023612301246,
+      "grad_norm": 1.4326949823387987,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 9488
+    },
+    {
+      "epoch": 0.22586403884604397,
+      "grad_norm": 3.549651887934873,
+      "learning_rate": 2e-05,
+      "loss": 0.1431,
+      "step": 9489
+    },
+    {
+      "epoch": 0.22588784156907551,
+      "grad_norm": 2.172754114117131,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 9490
+    },
+    {
+      "epoch": 0.22591164429210703,
+      "grad_norm": 4.361859318220804,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 9491
+    },
+    {
+      "epoch": 0.22593544701513854,
+      "grad_norm": 2.1272242737395692,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 9492
+    },
+    {
+      "epoch": 0.22595924973817005,
+      "grad_norm": 15.50700062183142,
+      "learning_rate": 2e-05,
+      "loss": 1.2132,
+      "step": 9493
+    },
+    {
+      "epoch": 0.22598305246120157,
+      "grad_norm": 5.26380108052513,
+      "learning_rate": 2e-05,
+      "loss": 1.0227,
+      "step": 9494
+    },
+    {
+      "epoch": 0.22600685518423308,
+      "grad_norm": 5.3176942511948555,
+      "learning_rate": 2e-05,
+      "loss": 0.716,
+      "step": 9495
+    },
+    {
+      "epoch": 0.2260306579072646,
+      "grad_norm": 8.980396194997125,
+      "learning_rate": 2e-05,
+      "loss": 0.7875,
+      "step": 9496
+    },
+    {
+      "epoch": 0.2260544606302961,
+      "grad_norm": 9.622593812120368,
+      "learning_rate": 2e-05,
+      "loss": 0.4861,
+      "step": 9497
+    },
+    {
+      "epoch": 0.22607826335332762,
+      "grad_norm": 2.876482216025156,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 9498
+    },
+    {
+      "epoch": 0.22610206607635913,
+      "grad_norm": 4.160741097482611,
+      "learning_rate": 2e-05,
+      "loss": 0.2672,
+      "step": 9499
+    },
+    {
+      "epoch": 0.22612586879939064,
+      "grad_norm": 7.828582597420878,
+      "learning_rate": 2e-05,
+      "loss": 0.2928,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22614967152242216,
+      "grad_norm": 14.477166808569319,
+      "learning_rate": 2e-05,
+      "loss": 0.1799,
+      "step": 9501
+    },
+    {
+      "epoch": 0.22617347424545367,
+      "grad_norm": 4.870006904846537,
+      "learning_rate": 2e-05,
+      "loss": 0.3165,
+      "step": 9502
+    },
+    {
+      "epoch": 0.22619727696848518,
+      "grad_norm": 4.044629502204771,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 9503
+    },
+    {
+      "epoch": 0.2262210796915167,
+      "grad_norm": 4.12363426103904,
+      "learning_rate": 2e-05,
+      "loss": 0.298,
+      "step": 9504
+    },
+    {
+      "epoch": 0.22624488241454824,
+      "grad_norm": 10.164606581677882,
+      "learning_rate": 2e-05,
+      "loss": 0.442,
+      "step": 9505
+    },
+    {
+      "epoch": 0.22626868513757975,
+      "grad_norm": 2.1229132439439002,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 9506
+    },
+    {
+      "epoch": 0.22629248786061126,
+      "grad_norm": 4.806527982251484,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 9507
+    },
+    {
+      "epoch": 0.22631629058364278,
+      "grad_norm": 5.0667323351115945,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 9508
+    },
+    {
+      "epoch": 0.2263400933066743,
+      "grad_norm": 2.790653434259259,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 9509
+    },
+    {
+      "epoch": 0.2263638960297058,
+      "grad_norm": 3.9355180328092207,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22638769875273732,
+      "grad_norm": 4.245129465311279,
+      "learning_rate": 2e-05,
+      "loss": 0.8361,
+      "step": 9511
+    },
+    {
+      "epoch": 0.22641150147576883,
+      "grad_norm": 9.727544292579738,
+      "learning_rate": 2e-05,
+      "loss": 0.8398,
+      "step": 9512
+    },
+    {
+      "epoch": 0.22643530419880034,
+      "grad_norm": 8.47692451149059,
+      "learning_rate": 2e-05,
+      "loss": 0.2285,
+      "step": 9513
+    },
+    {
+      "epoch": 0.22645910692183185,
+      "grad_norm": 2.5272974663497774,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 9514
+    },
+    {
+      "epoch": 0.22648290964486337,
+      "grad_norm": 6.105268480000877,
+      "learning_rate": 2e-05,
+      "loss": 0.2617,
+      "step": 9515
+    },
+    {
+      "epoch": 0.22650671236789488,
+      "grad_norm": 4.704652799008381,
+      "learning_rate": 2e-05,
+      "loss": 0.8837,
+      "step": 9516
+    },
+    {
+      "epoch": 0.2265305150909264,
+      "grad_norm": 3.3865964094125016,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 9517
+    },
+    {
+      "epoch": 0.2265543178139579,
+      "grad_norm": 1.8868708924491657,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 9518
+    },
+    {
+      "epoch": 0.22657812053698942,
+      "grad_norm": 1.430303408749068,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 9519
+    },
+    {
+      "epoch": 0.22660192326002096,
+      "grad_norm": 6.866795568076423,
+      "learning_rate": 2e-05,
+      "loss": 0.3361,
+      "step": 9520
+    },
+    {
+      "epoch": 0.22662572598305247,
+      "grad_norm": 11.192391993806282,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 9521
+    },
+    {
+      "epoch": 0.22664952870608399,
+      "grad_norm": 3.335311265896136,
+      "learning_rate": 2e-05,
+      "loss": 0.4126,
+      "step": 9522
+    },
+    {
+      "epoch": 0.2266733314291155,
+      "grad_norm": 8.763285448136406,
+      "learning_rate": 2e-05,
+      "loss": 1.3215,
+      "step": 9523
+    },
+    {
+      "epoch": 0.226697134152147,
+      "grad_norm": 5.611035427205062,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 9524
+    },
+    {
+      "epoch": 0.22672093687517852,
+      "grad_norm": 2.482636195979046,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 9525
+    },
+    {
+      "epoch": 0.22674473959821004,
+      "grad_norm": 2.146014407141889,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 9526
+    },
+    {
+      "epoch": 0.22676854232124155,
+      "grad_norm": 19.75784055886474,
+      "learning_rate": 2e-05,
+      "loss": 0.3013,
+      "step": 9527
+    },
+    {
+      "epoch": 0.22679234504427306,
+      "grad_norm": 2.5436451396847586,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 9528
+    },
+    {
+      "epoch": 0.22681614776730458,
+      "grad_norm": 15.971004199334594,
+      "learning_rate": 2e-05,
+      "loss": 0.3827,
+      "step": 9529
+    },
+    {
+      "epoch": 0.2268399504903361,
+      "grad_norm": 22.575513526993294,
+      "learning_rate": 2e-05,
+      "loss": 0.5196,
+      "step": 9530
+    },
+    {
+      "epoch": 0.2268637532133676,
+      "grad_norm": 11.83501984773544,
+      "learning_rate": 2e-05,
+      "loss": 1.2295,
+      "step": 9531
+    },
+    {
+      "epoch": 0.22688755593639912,
+      "grad_norm": 3.0691838308787163,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 9532
+    },
+    {
+      "epoch": 0.22691135865943063,
+      "grad_norm": 2.928831456418966,
+      "learning_rate": 2e-05,
+      "loss": 0.1653,
+      "step": 9533
+    },
+    {
+      "epoch": 0.22693516138246214,
+      "grad_norm": 7.064248837932697,
+      "learning_rate": 2e-05,
+      "loss": 1.3468,
+      "step": 9534
+    },
+    {
+      "epoch": 0.22695896410549365,
+      "grad_norm": 13.721251739007336,
+      "learning_rate": 2e-05,
+      "loss": 0.3959,
+      "step": 9535
+    },
+    {
+      "epoch": 0.2269827668285252,
+      "grad_norm": 6.173870026581868,
+      "learning_rate": 2e-05,
+      "loss": 1.048,
+      "step": 9536
+    },
+    {
+      "epoch": 0.2270065695515567,
+      "grad_norm": 5.912489541093802,
+      "learning_rate": 2e-05,
+      "loss": 0.5459,
+      "step": 9537
+    },
+    {
+      "epoch": 0.22703037227458822,
+      "grad_norm": 3.7215960907751295,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 9538
+    },
+    {
+      "epoch": 0.22705417499761973,
+      "grad_norm": 7.181148722686434,
+      "learning_rate": 2e-05,
+      "loss": 0.6875,
+      "step": 9539
+    },
+    {
+      "epoch": 0.22707797772065125,
+      "grad_norm": 4.109299900040462,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 9540
+    },
+    {
+      "epoch": 0.22710178044368276,
+      "grad_norm": 1.694833192854976,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 9541
+    },
+    {
+      "epoch": 0.22712558316671427,
+      "grad_norm": 14.681499525249826,
+      "learning_rate": 2e-05,
+      "loss": 0.2174,
+      "step": 9542
+    },
+    {
+      "epoch": 0.2271493858897458,
+      "grad_norm": 4.1329828494554866,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 9543
+    },
+    {
+      "epoch": 0.2271731886127773,
+      "grad_norm": 3.883135241354625,
+      "learning_rate": 2e-05,
+      "loss": 0.2047,
+      "step": 9544
+    },
+    {
+      "epoch": 0.2271969913358088,
+      "grad_norm": 1.1807730018928186,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 9545
+    },
+    {
+      "epoch": 0.22722079405884033,
+      "grad_norm": 2.3508276058717232,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 9546
+    },
+    {
+      "epoch": 0.22724459678187184,
+      "grad_norm": 5.494949932278127,
+      "learning_rate": 2e-05,
+      "loss": 0.7766,
+      "step": 9547
+    },
+    {
+      "epoch": 0.22726839950490335,
+      "grad_norm": 2.988076379383456,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 9548
+    },
+    {
+      "epoch": 0.22729220222793486,
+      "grad_norm": 1.318268644368025,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 9549
+    },
+    {
+      "epoch": 0.22731600495096638,
+      "grad_norm": 8.155120449384151,
+      "learning_rate": 2e-05,
+      "loss": 0.7593,
+      "step": 9550
+    },
+    {
+      "epoch": 0.22733980767399792,
+      "grad_norm": 2.2004149558539248,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 9551
+    },
+    {
+      "epoch": 0.22736361039702943,
+      "grad_norm": 3.9026465893856703,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 9552
+    },
+    {
+      "epoch": 0.22738741312006094,
+      "grad_norm": 10.34973656100839,
+      "learning_rate": 2e-05,
+      "loss": 0.5298,
+      "step": 9553
+    },
+    {
+      "epoch": 0.22741121584309246,
+      "grad_norm": 4.19993591321158,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 9554
+    },
+    {
+      "epoch": 0.22743501856612397,
+      "grad_norm": 12.542415984765006,
+      "learning_rate": 2e-05,
+      "loss": 0.3654,
+      "step": 9555
+    },
+    {
+      "epoch": 0.22745882128915548,
+      "grad_norm": 4.984075902825035,
+      "learning_rate": 2e-05,
+      "loss": 0.7166,
+      "step": 9556
+    },
+    {
+      "epoch": 0.227482624012187,
+      "grad_norm": 0.9872689930194556,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 9557
+    },
+    {
+      "epoch": 0.2275064267352185,
+      "grad_norm": 6.448532820042312,
+      "learning_rate": 2e-05,
+      "loss": 0.1712,
+      "step": 9558
+    },
+    {
+      "epoch": 0.22753022945825002,
+      "grad_norm": 13.370046145369185,
+      "learning_rate": 2e-05,
+      "loss": 1.0654,
+      "step": 9559
+    },
+    {
+      "epoch": 0.22755403218128153,
+      "grad_norm": 5.773445864397772,
+      "learning_rate": 2e-05,
+      "loss": 0.9735,
+      "step": 9560
+    },
+    {
+      "epoch": 0.22757783490431305,
+      "grad_norm": 3.7777053138023673,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 9561
+    },
+    {
+      "epoch": 0.22760163762734456,
+      "grad_norm": 5.274317174255092,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 9562
+    },
+    {
+      "epoch": 0.22762544035037607,
+      "grad_norm": 3.3736821018044547,
+      "learning_rate": 2e-05,
+      "loss": 0.3066,
+      "step": 9563
+    },
+    {
+      "epoch": 0.2276492430734076,
+      "grad_norm": 9.352940226483048,
+      "learning_rate": 2e-05,
+      "loss": 1.3958,
+      "step": 9564
+    },
+    {
+      "epoch": 0.2276730457964391,
+      "grad_norm": 2.6578939430202597,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 9565
+    },
+    {
+      "epoch": 0.22769684851947064,
+      "grad_norm": 9.778361774281427,
+      "learning_rate": 2e-05,
+      "loss": 1.0801,
+      "step": 9566
+    },
+    {
+      "epoch": 0.22772065124250215,
+      "grad_norm": 8.225365604495424,
+      "learning_rate": 2e-05,
+      "loss": 0.4419,
+      "step": 9567
+    },
+    {
+      "epoch": 0.22774445396553367,
+      "grad_norm": 22.968196950982474,
+      "learning_rate": 2e-05,
+      "loss": 0.31,
+      "step": 9568
+    },
+    {
+      "epoch": 0.22776825668856518,
+      "grad_norm": 6.615751739434096,
+      "learning_rate": 2e-05,
+      "loss": 0.9541,
+      "step": 9569
+    },
+    {
+      "epoch": 0.2277920594115967,
+      "grad_norm": 6.548768993535241,
+      "learning_rate": 2e-05,
+      "loss": 0.7847,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2278158621346282,
+      "grad_norm": 2.8750222980757103,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 9571
+    },
+    {
+      "epoch": 0.22783966485765972,
+      "grad_norm": 11.398503444871066,
+      "learning_rate": 2e-05,
+      "loss": 0.4064,
+      "step": 9572
+    },
+    {
+      "epoch": 0.22786346758069123,
+      "grad_norm": 2.4772386126323687,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 9573
+    },
+    {
+      "epoch": 0.22788727030372274,
+      "grad_norm": 2.387790682263466,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 9574
+    },
+    {
+      "epoch": 0.22791107302675426,
+      "grad_norm": 3.736149959020619,
+      "learning_rate": 2e-05,
+      "loss": 0.259,
+      "step": 9575
+    },
+    {
+      "epoch": 0.22793487574978577,
+      "grad_norm": 13.96228573639045,
+      "learning_rate": 2e-05,
+      "loss": 1.0887,
+      "step": 9576
+    },
+    {
+      "epoch": 0.22795867847281728,
+      "grad_norm": 2.709122657817588,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 9577
+    },
+    {
+      "epoch": 0.2279824811958488,
+      "grad_norm": 2.214769576705605,
+      "learning_rate": 2e-05,
+      "loss": 0.1203,
+      "step": 9578
+    },
+    {
+      "epoch": 0.2280062839188803,
+      "grad_norm": 15.374774498936167,
+      "learning_rate": 2e-05,
+      "loss": 0.2526,
+      "step": 9579
+    },
+    {
+      "epoch": 0.22803008664191182,
+      "grad_norm": 3.566763609294492,
+      "learning_rate": 2e-05,
+      "loss": 0.1311,
+      "step": 9580
+    },
+    {
+      "epoch": 0.22805388936494336,
+      "grad_norm": 4.79074654159678,
+      "learning_rate": 2e-05,
+      "loss": 0.2719,
+      "step": 9581
+    },
+    {
+      "epoch": 0.22807769208797488,
+      "grad_norm": 3.2421829674150593,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 9582
+    },
+    {
+      "epoch": 0.2281014948110064,
+      "grad_norm": 2.135090753691825,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 9583
+    },
+    {
+      "epoch": 0.2281252975340379,
+      "grad_norm": 4.4086076604873945,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 9584
+    },
+    {
+      "epoch": 0.22814910025706941,
+      "grad_norm": 3.712304372635291,
+      "learning_rate": 2e-05,
+      "loss": 0.5241,
+      "step": 9585
+    },
+    {
+      "epoch": 0.22817290298010093,
+      "grad_norm": 3.3709295851464973,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 9586
+    },
+    {
+      "epoch": 0.22819670570313244,
+      "grad_norm": 2.4192137181382556,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 9587
+    },
+    {
+      "epoch": 0.22822050842616395,
+      "grad_norm": 5.657331702290981,
+      "learning_rate": 2e-05,
+      "loss": 0.6349,
+      "step": 9588
+    },
+    {
+      "epoch": 0.22824431114919547,
+      "grad_norm": 2.4483971744240387,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 9589
+    },
+    {
+      "epoch": 0.22826811387222698,
+      "grad_norm": 2.0076799195977766,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 9590
+    },
+    {
+      "epoch": 0.2282919165952585,
+      "grad_norm": 2.9755014894515353,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 9591
+    },
+    {
+      "epoch": 0.22831571931829,
+      "grad_norm": 1.1633715159086644,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 9592
+    },
+    {
+      "epoch": 0.22833952204132152,
+      "grad_norm": 4.534067793958215,
+      "learning_rate": 2e-05,
+      "loss": 0.8387,
+      "step": 9593
+    },
+    {
+      "epoch": 0.22836332476435303,
+      "grad_norm": 3.533099133271867,
+      "learning_rate": 2e-05,
+      "loss": 0.5604,
+      "step": 9594
+    },
+    {
+      "epoch": 0.22838712748738454,
+      "grad_norm": 2.3914622020481575,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 9595
+    },
+    {
+      "epoch": 0.22841093021041609,
+      "grad_norm": 26.219126288397767,
+      "learning_rate": 2e-05,
+      "loss": 1.2479,
+      "step": 9596
+    },
+    {
+      "epoch": 0.2284347329334476,
+      "grad_norm": 2.486406599094038,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 9597
+    },
+    {
+      "epoch": 0.2284585356564791,
+      "grad_norm": 0.7562627616789214,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 9598
+    },
+    {
+      "epoch": 0.22848233837951062,
+      "grad_norm": 2.5570098052351433,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 9599
+    },
+    {
+      "epoch": 0.22850614110254214,
+      "grad_norm": 7.688413622410793,
+      "learning_rate": 2e-05,
+      "loss": 1.3006,
+      "step": 9600
+    },
+    {
+      "epoch": 0.22852994382557365,
+      "grad_norm": 5.808599243414075,
+      "learning_rate": 2e-05,
+      "loss": 0.5725,
+      "step": 9601
+    },
+    {
+      "epoch": 0.22855374654860516,
+      "grad_norm": 4.67972725306067,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 9602
+    },
+    {
+      "epoch": 0.22857754927163668,
+      "grad_norm": 33.073540061428254,
+      "learning_rate": 2e-05,
+      "loss": 0.3717,
+      "step": 9603
+    },
+    {
+      "epoch": 0.2286013519946682,
+      "grad_norm": 2.107089915506382,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 9604
+    },
+    {
+      "epoch": 0.2286251547176997,
+      "grad_norm": 2.2193261723009368,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 9605
+    },
+    {
+      "epoch": 0.22864895744073122,
+      "grad_norm": 0.6753898286863207,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 9606
+    },
+    {
+      "epoch": 0.22867276016376273,
+      "grad_norm": 1.894910050271345,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 9607
+    },
+    {
+      "epoch": 0.22869656288679424,
+      "grad_norm": 7.180093821475948,
+      "learning_rate": 2e-05,
+      "loss": 0.9371,
+      "step": 9608
+    },
+    {
+      "epoch": 0.22872036560982575,
+      "grad_norm": 6.546832161875086,
+      "learning_rate": 2e-05,
+      "loss": 0.9739,
+      "step": 9609
+    },
+    {
+      "epoch": 0.22874416833285727,
+      "grad_norm": 13.784609570983399,
+      "learning_rate": 2e-05,
+      "loss": 1.1115,
+      "step": 9610
+    },
+    {
+      "epoch": 0.22876797105588878,
+      "grad_norm": 12.791871530609102,
+      "learning_rate": 2e-05,
+      "loss": 0.8242,
+      "step": 9611
+    },
+    {
+      "epoch": 0.22879177377892032,
+      "grad_norm": 5.8037133733345225,
+      "learning_rate": 2e-05,
+      "loss": 0.3832,
+      "step": 9612
+    },
+    {
+      "epoch": 0.22881557650195183,
+      "grad_norm": 11.571171813294717,
+      "learning_rate": 2e-05,
+      "loss": 0.9968,
+      "step": 9613
+    },
+    {
+      "epoch": 0.22883937922498335,
+      "grad_norm": 2.227133389102929,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 9614
+    },
+    {
+      "epoch": 0.22886318194801486,
+      "grad_norm": 1.272462219550346,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 9615
+    },
+    {
+      "epoch": 0.22888698467104637,
+      "grad_norm": 11.40402107630687,
+      "learning_rate": 2e-05,
+      "loss": 1.415,
+      "step": 9616
+    },
+    {
+      "epoch": 0.22891078739407789,
+      "grad_norm": 7.21496493133383,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 9617
+    },
+    {
+      "epoch": 0.2289345901171094,
+      "grad_norm": 19.52390313957763,
+      "learning_rate": 2e-05,
+      "loss": 0.54,
+      "step": 9618
+    },
+    {
+      "epoch": 0.2289583928401409,
+      "grad_norm": 19.271422359275544,
+      "learning_rate": 2e-05,
+      "loss": 0.2992,
+      "step": 9619
+    },
+    {
+      "epoch": 0.22898219556317242,
+      "grad_norm": 4.181847299581136,
+      "learning_rate": 2e-05,
+      "loss": 0.1411,
+      "step": 9620
+    },
+    {
+      "epoch": 0.22900599828620394,
+      "grad_norm": 16.86055546667661,
+      "learning_rate": 2e-05,
+      "loss": 1.1556,
+      "step": 9621
+    },
+    {
+      "epoch": 0.22902980100923545,
+      "grad_norm": 5.8839065660500935,
+      "learning_rate": 2e-05,
+      "loss": 0.3588,
+      "step": 9622
+    },
+    {
+      "epoch": 0.22905360373226696,
+      "grad_norm": 6.242525103976511,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 9623
+    },
+    {
+      "epoch": 0.22907740645529848,
+      "grad_norm": 0.9508910944294293,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 9624
+    },
+    {
+      "epoch": 0.22910120917833,
+      "grad_norm": 1.2991066645398561,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 9625
+    },
+    {
+      "epoch": 0.2291250119013615,
+      "grad_norm": 3.7470155992250214,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 9626
+    },
+    {
+      "epoch": 0.22914881462439304,
+      "grad_norm": 5.0108919842982464,
+      "learning_rate": 2e-05,
+      "loss": 0.2706,
+      "step": 9627
+    },
+    {
+      "epoch": 0.22917261734742456,
+      "grad_norm": 2.8947619434623193,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 9628
+    },
+    {
+      "epoch": 0.22919642007045607,
+      "grad_norm": 7.310800111643198,
+      "learning_rate": 2e-05,
+      "loss": 0.9187,
+      "step": 9629
+    },
+    {
+      "epoch": 0.22922022279348758,
+      "grad_norm": 12.0933564052251,
+      "learning_rate": 2e-05,
+      "loss": 2.2089,
+      "step": 9630
+    },
+    {
+      "epoch": 0.2292440255165191,
+      "grad_norm": 7.4907999459340475,
+      "learning_rate": 2e-05,
+      "loss": 1.2177,
+      "step": 9631
+    },
+    {
+      "epoch": 0.2292678282395506,
+      "grad_norm": 0.6896747237280038,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 9632
+    },
+    {
+      "epoch": 0.22929163096258212,
+      "grad_norm": 4.369414974539549,
+      "learning_rate": 2e-05,
+      "loss": 0.5122,
+      "step": 9633
+    },
+    {
+      "epoch": 0.22931543368561363,
+      "grad_norm": 7.976692350551995,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 9634
+    },
+    {
+      "epoch": 0.22933923640864515,
+      "grad_norm": 3.2693916373026948,
+      "learning_rate": 2e-05,
+      "loss": 0.2569,
+      "step": 9635
+    },
+    {
+      "epoch": 0.22936303913167666,
+      "grad_norm": 3.7065813603294817,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 9636
+    },
+    {
+      "epoch": 0.22938684185470817,
+      "grad_norm": 2.6486385048278773,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 9637
+    },
+    {
+      "epoch": 0.2294106445777397,
+      "grad_norm": 1.4609707295742005,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 9638
+    },
+    {
+      "epoch": 0.2294344473007712,
+      "grad_norm": 9.05625694379429,
+      "learning_rate": 2e-05,
+      "loss": 0.8567,
+      "step": 9639
+    },
+    {
+      "epoch": 0.2294582500238027,
+      "grad_norm": 11.783685794160444,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 9640
+    },
+    {
+      "epoch": 0.22948205274683423,
+      "grad_norm": 3.060822938863928,
+      "learning_rate": 2e-05,
+      "loss": 0.2834,
+      "step": 9641
+    },
+    {
+      "epoch": 0.22950585546986577,
+      "grad_norm": 7.714410035873174,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 9642
+    },
+    {
+      "epoch": 0.22952965819289728,
+      "grad_norm": 10.515443502108687,
+      "learning_rate": 2e-05,
+      "loss": 1.0341,
+      "step": 9643
+    },
+    {
+      "epoch": 0.2295534609159288,
+      "grad_norm": 1.5424442449115385,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 9644
+    },
+    {
+      "epoch": 0.2295772636389603,
+      "grad_norm": 2.480278622368074,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 9645
+    },
+    {
+      "epoch": 0.22960106636199182,
+      "grad_norm": 4.521667779429096,
+      "learning_rate": 2e-05,
+      "loss": 0.2674,
+      "step": 9646
+    },
+    {
+      "epoch": 0.22962486908502333,
+      "grad_norm": 4.989657786018957,
+      "learning_rate": 2e-05,
+      "loss": 0.5612,
+      "step": 9647
+    },
+    {
+      "epoch": 0.22964867180805484,
+      "grad_norm": 15.500408197782171,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 9648
+    },
+    {
+      "epoch": 0.22967247453108636,
+      "grad_norm": 10.639231331048391,
+      "learning_rate": 2e-05,
+      "loss": 0.3979,
+      "step": 9649
+    },
+    {
+      "epoch": 0.22969627725411787,
+      "grad_norm": 5.265535652879701,
+      "learning_rate": 2e-05,
+      "loss": 1.1563,
+      "step": 9650
+    },
+    {
+      "epoch": 0.22972007997714938,
+      "grad_norm": 0.5943204395074126,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 9651
+    },
+    {
+      "epoch": 0.2297438827001809,
+      "grad_norm": 1.7138880637009282,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 9652
+    },
+    {
+      "epoch": 0.2297676854232124,
+      "grad_norm": 29.02855281895477,
+      "learning_rate": 2e-05,
+      "loss": 1.3196,
+      "step": 9653
+    },
+    {
+      "epoch": 0.22979148814624392,
+      "grad_norm": 7.7325572931770665,
+      "learning_rate": 2e-05,
+      "loss": 0.535,
+      "step": 9654
+    },
+    {
+      "epoch": 0.22981529086927543,
+      "grad_norm": 8.880417751946243,
+      "learning_rate": 2e-05,
+      "loss": 2.2601,
+      "step": 9655
+    },
+    {
+      "epoch": 0.22983909359230695,
+      "grad_norm": 2.9379763559142997,
+      "learning_rate": 2e-05,
+      "loss": 0.1708,
+      "step": 9656
+    },
+    {
+      "epoch": 0.2298628963153385,
+      "grad_norm": 0.8848455861514999,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 9657
+    },
+    {
+      "epoch": 0.22988669903837,
+      "grad_norm": 5.933837236681505,
+      "learning_rate": 2e-05,
+      "loss": 0.6564,
+      "step": 9658
+    },
+    {
+      "epoch": 0.22991050176140151,
+      "grad_norm": 5.3866335718124745,
+      "learning_rate": 2e-05,
+      "loss": 1.0909,
+      "step": 9659
+    },
+    {
+      "epoch": 0.22993430448443303,
+      "grad_norm": 6.867483734988599,
+      "learning_rate": 2e-05,
+      "loss": 0.3777,
+      "step": 9660
+    },
+    {
+      "epoch": 0.22995810720746454,
+      "grad_norm": 6.456477435089132,
+      "learning_rate": 2e-05,
+      "loss": 0.1727,
+      "step": 9661
+    },
+    {
+      "epoch": 0.22998190993049605,
+      "grad_norm": 3.92595239072423,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 9662
+    },
+    {
+      "epoch": 0.23000571265352757,
+      "grad_norm": 4.73252740107517,
+      "learning_rate": 2e-05,
+      "loss": 0.616,
+      "step": 9663
+    },
+    {
+      "epoch": 0.23002951537655908,
+      "grad_norm": 3.6989740058418366,
+      "learning_rate": 2e-05,
+      "loss": 0.469,
+      "step": 9664
+    },
+    {
+      "epoch": 0.2300533180995906,
+      "grad_norm": 4.162641771968289,
+      "learning_rate": 2e-05,
+      "loss": 0.2068,
+      "step": 9665
+    },
+    {
+      "epoch": 0.2300771208226221,
+      "grad_norm": 4.404981894972508,
+      "learning_rate": 2e-05,
+      "loss": 0.6089,
+      "step": 9666
+    },
+    {
+      "epoch": 0.23010092354565362,
+      "grad_norm": 6.636422149716335,
+      "learning_rate": 2e-05,
+      "loss": 0.3501,
+      "step": 9667
+    },
+    {
+      "epoch": 0.23012472626868513,
+      "grad_norm": 4.4459149444066846,
+      "learning_rate": 2e-05,
+      "loss": 0.2493,
+      "step": 9668
+    },
+    {
+      "epoch": 0.23014852899171664,
+      "grad_norm": 3.530836063603226,
+      "learning_rate": 2e-05,
+      "loss": 0.1745,
+      "step": 9669
+    },
+    {
+      "epoch": 0.23017233171474816,
+      "grad_norm": 7.188944111739091,
+      "learning_rate": 2e-05,
+      "loss": 1.4286,
+      "step": 9670
+    },
+    {
+      "epoch": 0.23019613443777967,
+      "grad_norm": 12.847295801452429,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 9671
+    },
+    {
+      "epoch": 0.23021993716081118,
+      "grad_norm": 4.66602524242167,
+      "learning_rate": 2e-05,
+      "loss": 0.7041,
+      "step": 9672
+    },
+    {
+      "epoch": 0.23024373988384272,
+      "grad_norm": 0.8216490796751861,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 9673
+    },
+    {
+      "epoch": 0.23026754260687424,
+      "grad_norm": 1.2038970790342167,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 9674
+    },
+    {
+      "epoch": 0.23029134532990575,
+      "grad_norm": 3.4995092696979193,
+      "learning_rate": 2e-05,
+      "loss": 0.2184,
+      "step": 9675
+    },
+    {
+      "epoch": 0.23031514805293726,
+      "grad_norm": 2.4988246049467358,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 9676
+    },
+    {
+      "epoch": 0.23033895077596878,
+      "grad_norm": 9.35705567600897,
+      "learning_rate": 2e-05,
+      "loss": 0.322,
+      "step": 9677
+    },
+    {
+      "epoch": 0.2303627534990003,
+      "grad_norm": 9.146736320061938,
+      "learning_rate": 2e-05,
+      "loss": 0.7644,
+      "step": 9678
+    },
+    {
+      "epoch": 0.2303865562220318,
+      "grad_norm": 3.738768335033108,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 9679
+    },
+    {
+      "epoch": 0.23041035894506331,
+      "grad_norm": 3.771769359869376,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 9680
+    },
+    {
+      "epoch": 0.23043416166809483,
+      "grad_norm": 5.598942723250217,
+      "learning_rate": 2e-05,
+      "loss": 0.1181,
+      "step": 9681
+    },
+    {
+      "epoch": 0.23045796439112634,
+      "grad_norm": 3.0107639429952635,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 9682
+    },
+    {
+      "epoch": 0.23048176711415785,
+      "grad_norm": 4.309253715569236,
+      "learning_rate": 2e-05,
+      "loss": 0.1462,
+      "step": 9683
+    },
+    {
+      "epoch": 0.23050556983718937,
+      "grad_norm": 2.246676206931224,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 9684
+    },
+    {
+      "epoch": 0.23052937256022088,
+      "grad_norm": 2.220881843010721,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 9685
+    },
+    {
+      "epoch": 0.2305531752832524,
+      "grad_norm": 2.5529430013294703,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 9686
+    },
+    {
+      "epoch": 0.2305769780062839,
+      "grad_norm": 10.449619200377489,
+      "learning_rate": 2e-05,
+      "loss": 1.1107,
+      "step": 9687
+    },
+    {
+      "epoch": 0.23060078072931545,
+      "grad_norm": 56.709553270891206,
+      "learning_rate": 2e-05,
+      "loss": 0.5184,
+      "step": 9688
+    },
+    {
+      "epoch": 0.23062458345234696,
+      "grad_norm": 4.332955183017469,
+      "learning_rate": 2e-05,
+      "loss": 0.2999,
+      "step": 9689
+    },
+    {
+      "epoch": 0.23064838617537847,
+      "grad_norm": 3.394044568701158,
+      "learning_rate": 2e-05,
+      "loss": 0.2567,
+      "step": 9690
+    },
+    {
+      "epoch": 0.23067218889840999,
+      "grad_norm": 11.95735994743772,
+      "learning_rate": 2e-05,
+      "loss": 0.7833,
+      "step": 9691
+    },
+    {
+      "epoch": 0.2306959916214415,
+      "grad_norm": 3.413963415126476,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 9692
+    },
+    {
+      "epoch": 0.230719794344473,
+      "grad_norm": 5.8053671728356715,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 9693
+    },
+    {
+      "epoch": 0.23074359706750452,
+      "grad_norm": 1.6310804678162978,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 9694
+    },
+    {
+      "epoch": 0.23076739979053604,
+      "grad_norm": 8.658427763530327,
+      "learning_rate": 2e-05,
+      "loss": 0.3785,
+      "step": 9695
+    },
+    {
+      "epoch": 0.23079120251356755,
+      "grad_norm": 4.8581114651864885,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 9696
+    },
+    {
+      "epoch": 0.23081500523659906,
+      "grad_norm": 1.258285356039682,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 9697
+    },
+    {
+      "epoch": 0.23083880795963058,
+      "grad_norm": 1.0811603188438537,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 9698
+    },
+    {
+      "epoch": 0.2308626106826621,
+      "grad_norm": 14.51591354644706,
+      "learning_rate": 2e-05,
+      "loss": 0.6492,
+      "step": 9699
+    },
+    {
+      "epoch": 0.2308864134056936,
+      "grad_norm": 8.450851367751541,
+      "learning_rate": 2e-05,
+      "loss": 0.9078,
+      "step": 9700
+    },
+    {
+      "epoch": 0.23091021612872512,
+      "grad_norm": 2.204226407385953,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 9701
+    },
+    {
+      "epoch": 0.23093401885175663,
+      "grad_norm": 2.3833563453031457,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 9702
+    },
+    {
+      "epoch": 0.23095782157478817,
+      "grad_norm": 2.4182755184600695,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 9703
+    },
+    {
+      "epoch": 0.23098162429781968,
+      "grad_norm": 7.009887531413109,
+      "learning_rate": 2e-05,
+      "loss": 0.3369,
+      "step": 9704
+    },
+    {
+      "epoch": 0.2310054270208512,
+      "grad_norm": 3.1286665056748686,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 9705
+    },
+    {
+      "epoch": 0.2310292297438827,
+      "grad_norm": 13.589885960749086,
+      "learning_rate": 2e-05,
+      "loss": 0.992,
+      "step": 9706
+    },
+    {
+      "epoch": 0.23105303246691422,
+      "grad_norm": 10.40456876845658,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 9707
+    },
+    {
+      "epoch": 0.23107683518994573,
+      "grad_norm": 2.463513811059117,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 9708
+    },
+    {
+      "epoch": 0.23110063791297725,
+      "grad_norm": 5.5653153226144285,
+      "learning_rate": 2e-05,
+      "loss": 0.8316,
+      "step": 9709
+    },
+    {
+      "epoch": 0.23112444063600876,
+      "grad_norm": 3.7733601173263893,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 9710
+    },
+    {
+      "epoch": 0.23114824335904027,
+      "grad_norm": 0.7006391483241896,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 9711
+    },
+    {
+      "epoch": 0.23117204608207179,
+      "grad_norm": 27.93195962429023,
+      "learning_rate": 2e-05,
+      "loss": 1.6192,
+      "step": 9712
+    },
+    {
+      "epoch": 0.2311958488051033,
+      "grad_norm": 1.9841040361464917,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 9713
+    },
+    {
+      "epoch": 0.2312196515281348,
+      "grad_norm": 4.921359630388563,
+      "learning_rate": 2e-05,
+      "loss": 0.5727,
+      "step": 9714
+    },
+    {
+      "epoch": 0.23124345425116632,
+      "grad_norm": 7.7267260447089035,
+      "learning_rate": 2e-05,
+      "loss": 0.2196,
+      "step": 9715
+    },
+    {
+      "epoch": 0.23126725697419784,
+      "grad_norm": 3.059428874767886,
+      "learning_rate": 2e-05,
+      "loss": 0.2123,
+      "step": 9716
+    },
+    {
+      "epoch": 0.23129105969722935,
+      "grad_norm": 6.267344005295689,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 9717
+    },
+    {
+      "epoch": 0.2313148624202609,
+      "grad_norm": 4.398263024447505,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 9718
+    },
+    {
+      "epoch": 0.2313386651432924,
+      "grad_norm": 6.176714358155578,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 9719
+    },
+    {
+      "epoch": 0.23136246786632392,
+      "grad_norm": 7.6737157005161425,
+      "learning_rate": 2e-05,
+      "loss": 0.3503,
+      "step": 9720
+    },
+    {
+      "epoch": 0.23138627058935543,
+      "grad_norm": 2.371705222643073,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 9721
+    },
+    {
+      "epoch": 0.23141007331238694,
+      "grad_norm": 5.584250111204502,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 9722
+    },
+    {
+      "epoch": 0.23143387603541846,
+      "grad_norm": 1.866252987640877,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 9723
+    },
+    {
+      "epoch": 0.23145767875844997,
+      "grad_norm": 5.42325071820809,
+      "learning_rate": 2e-05,
+      "loss": 0.5058,
+      "step": 9724
+    },
+    {
+      "epoch": 0.23148148148148148,
+      "grad_norm": 3.2957118445650684,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 9725
+    },
+    {
+      "epoch": 0.231505284204513,
+      "grad_norm": 6.084570593507645,
+      "learning_rate": 2e-05,
+      "loss": 0.3841,
+      "step": 9726
+    },
+    {
+      "epoch": 0.2315290869275445,
+      "grad_norm": 2.156659467096261,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 9727
+    },
+    {
+      "epoch": 0.23155288965057602,
+      "grad_norm": 1.0706348716663958,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 9728
+    },
+    {
+      "epoch": 0.23157669237360753,
+      "grad_norm": 1.3963979737352832,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 9729
+    },
+    {
+      "epoch": 0.23160049509663905,
+      "grad_norm": 5.836071843944711,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 9730
+    },
+    {
+      "epoch": 0.23162429781967056,
+      "grad_norm": 3.2352759968351417,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 9731
+    },
+    {
+      "epoch": 0.23164810054270207,
+      "grad_norm": 5.524061292397284,
+      "learning_rate": 2e-05,
+      "loss": 0.4826,
+      "step": 9732
+    },
+    {
+      "epoch": 0.2316719032657336,
+      "grad_norm": 8.157615862258108,
+      "learning_rate": 2e-05,
+      "loss": 1.3313,
+      "step": 9733
+    },
+    {
+      "epoch": 0.23169570598876513,
+      "grad_norm": 4.510502625894955,
+      "learning_rate": 2e-05,
+      "loss": 0.7543,
+      "step": 9734
+    },
+    {
+      "epoch": 0.23171950871179664,
+      "grad_norm": 5.501430620744715,
+      "learning_rate": 2e-05,
+      "loss": 0.499,
+      "step": 9735
+    },
+    {
+      "epoch": 0.23174331143482815,
+      "grad_norm": 17.582551656880725,
+      "learning_rate": 2e-05,
+      "loss": 0.4708,
+      "step": 9736
+    },
+    {
+      "epoch": 0.23176711415785967,
+      "grad_norm": 7.1771779578989,
+      "learning_rate": 2e-05,
+      "loss": 1.7433,
+      "step": 9737
+    },
+    {
+      "epoch": 0.23179091688089118,
+      "grad_norm": 5.258094532831738,
+      "learning_rate": 2e-05,
+      "loss": 1.2575,
+      "step": 9738
+    },
+    {
+      "epoch": 0.2318147196039227,
+      "grad_norm": 7.858057257472591,
+      "learning_rate": 2e-05,
+      "loss": 0.6613,
+      "step": 9739
+    },
+    {
+      "epoch": 0.2318385223269542,
+      "grad_norm": 1.2098644973593857,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 9740
+    },
+    {
+      "epoch": 0.23186232504998572,
+      "grad_norm": 1.7279807198233896,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 9741
+    },
+    {
+      "epoch": 0.23188612777301723,
+      "grad_norm": 14.170157362017466,
+      "learning_rate": 2e-05,
+      "loss": 0.6277,
+      "step": 9742
+    },
+    {
+      "epoch": 0.23190993049604874,
+      "grad_norm": 19.26566358095217,
+      "learning_rate": 2e-05,
+      "loss": 1.6957,
+      "step": 9743
+    },
+    {
+      "epoch": 0.23193373321908026,
+      "grad_norm": 3.7182935059762015,
+      "learning_rate": 2e-05,
+      "loss": 0.3343,
+      "step": 9744
+    },
+    {
+      "epoch": 0.23195753594211177,
+      "grad_norm": 5.272535584100286,
+      "learning_rate": 2e-05,
+      "loss": 0.8388,
+      "step": 9745
+    },
+    {
+      "epoch": 0.23198133866514328,
+      "grad_norm": 3.623541860398807,
+      "learning_rate": 2e-05,
+      "loss": 0.1784,
+      "step": 9746
+    },
+    {
+      "epoch": 0.2320051413881748,
+      "grad_norm": 5.583892811762755,
+      "learning_rate": 2e-05,
+      "loss": 0.2962,
+      "step": 9747
+    },
+    {
+      "epoch": 0.2320289441112063,
+      "grad_norm": 6.890803556460083,
+      "learning_rate": 2e-05,
+      "loss": 0.8592,
+      "step": 9748
+    },
+    {
+      "epoch": 0.23205274683423785,
+      "grad_norm": 12.93510740440931,
+      "learning_rate": 2e-05,
+      "loss": 0.3707,
+      "step": 9749
+    },
+    {
+      "epoch": 0.23207654955726936,
+      "grad_norm": 19.528036618700362,
+      "learning_rate": 2e-05,
+      "loss": 1.0473,
+      "step": 9750
+    },
+    {
+      "epoch": 0.23210035228030088,
+      "grad_norm": 3.2019185290990015,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 9751
+    },
+    {
+      "epoch": 0.2321241550033324,
+      "grad_norm": 10.009767442709,
+      "learning_rate": 2e-05,
+      "loss": 1.089,
+      "step": 9752
+    },
+    {
+      "epoch": 0.2321479577263639,
+      "grad_norm": 6.187164486980582,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 9753
+    },
+    {
+      "epoch": 0.23217176044939541,
+      "grad_norm": 12.914014029326086,
+      "learning_rate": 2e-05,
+      "loss": 1.2378,
+      "step": 9754
+    },
+    {
+      "epoch": 0.23219556317242693,
+      "grad_norm": 4.632571350764852,
+      "learning_rate": 2e-05,
+      "loss": 0.4723,
+      "step": 9755
+    },
+    {
+      "epoch": 0.23221936589545844,
+      "grad_norm": 4.099818877956517,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 9756
+    },
+    {
+      "epoch": 0.23224316861848995,
+      "grad_norm": 1.9419289631134533,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 9757
+    },
+    {
+      "epoch": 0.23226697134152147,
+      "grad_norm": 1.4718994433801273,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 9758
+    },
+    {
+      "epoch": 0.23229077406455298,
+      "grad_norm": 2.251847076105268,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 9759
+    },
+    {
+      "epoch": 0.2323145767875845,
+      "grad_norm": 4.411691048604553,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 9760
+    },
+    {
+      "epoch": 0.232338379510616,
+      "grad_norm": 3.9447289111058526,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 9761
+    },
+    {
+      "epoch": 0.23236218223364752,
+      "grad_norm": 13.52761557534861,
+      "learning_rate": 2e-05,
+      "loss": 1.975,
+      "step": 9762
+    },
+    {
+      "epoch": 0.23238598495667903,
+      "grad_norm": 4.544809507184357,
+      "learning_rate": 2e-05,
+      "loss": 0.4152,
+      "step": 9763
+    },
+    {
+      "epoch": 0.23240978767971057,
+      "grad_norm": 3.6707062004190245,
+      "learning_rate": 2e-05,
+      "loss": 0.4854,
+      "step": 9764
+    },
+    {
+      "epoch": 0.23243359040274209,
+      "grad_norm": 8.827720945087414,
+      "learning_rate": 2e-05,
+      "loss": 1.0232,
+      "step": 9765
+    },
+    {
+      "epoch": 0.2324573931257736,
+      "grad_norm": 4.619007580603277,
+      "learning_rate": 2e-05,
+      "loss": 0.7934,
+      "step": 9766
+    },
+    {
+      "epoch": 0.2324811958488051,
+      "grad_norm": 0.6502857751538809,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 9767
+    },
+    {
+      "epoch": 0.23250499857183662,
+      "grad_norm": 4.977935368422371,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 9768
+    },
+    {
+      "epoch": 0.23252880129486814,
+      "grad_norm": 3.7897437814328656,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 9769
+    },
+    {
+      "epoch": 0.23255260401789965,
+      "grad_norm": 7.100332015416068,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23257640674093116,
+      "grad_norm": 1.7088391649243357,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 9771
+    },
+    {
+      "epoch": 0.23260020946396268,
+      "grad_norm": 3.536074092754693,
+      "learning_rate": 2e-05,
+      "loss": 0.225,
+      "step": 9772
+    },
+    {
+      "epoch": 0.2326240121869942,
+      "grad_norm": 1.5933708117816185,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 9773
+    },
+    {
+      "epoch": 0.2326478149100257,
+      "grad_norm": 5.001394323100133,
+      "learning_rate": 2e-05,
+      "loss": 0.169,
+      "step": 9774
+    },
+    {
+      "epoch": 0.23267161763305722,
+      "grad_norm": 5.284645388694366,
+      "learning_rate": 2e-05,
+      "loss": 0.6447,
+      "step": 9775
+    },
+    {
+      "epoch": 0.23269542035608873,
+      "grad_norm": 6.9559096040385695,
+      "learning_rate": 2e-05,
+      "loss": 1.0016,
+      "step": 9776
+    },
+    {
+      "epoch": 0.23271922307912024,
+      "grad_norm": 8.017482320998772,
+      "learning_rate": 2e-05,
+      "loss": 2.0456,
+      "step": 9777
+    },
+    {
+      "epoch": 0.23274302580215175,
+      "grad_norm": 12.79922793352656,
+      "learning_rate": 2e-05,
+      "loss": 1.1135,
+      "step": 9778
+    },
+    {
+      "epoch": 0.2327668285251833,
+      "grad_norm": 10.204159520405407,
+      "learning_rate": 2e-05,
+      "loss": 0.4004,
+      "step": 9779
+    },
+    {
+      "epoch": 0.2327906312482148,
+      "grad_norm": 14.796197637659729,
+      "learning_rate": 2e-05,
+      "loss": 1.6945,
+      "step": 9780
+    },
+    {
+      "epoch": 0.23281443397124632,
+      "grad_norm": 4.749293469517998,
+      "learning_rate": 2e-05,
+      "loss": 0.1863,
+      "step": 9781
+    },
+    {
+      "epoch": 0.23283823669427783,
+      "grad_norm": 3.4824175473798,
+      "learning_rate": 2e-05,
+      "loss": 0.0982,
+      "step": 9782
+    },
+    {
+      "epoch": 0.23286203941730935,
+      "grad_norm": 3.8750894958769,
+      "learning_rate": 2e-05,
+      "loss": 0.2083,
+      "step": 9783
+    },
+    {
+      "epoch": 0.23288584214034086,
+      "grad_norm": 5.168671436703316,
+      "learning_rate": 2e-05,
+      "loss": 0.1883,
+      "step": 9784
+    },
+    {
+      "epoch": 0.23290964486337237,
+      "grad_norm": 8.097765147996785,
+      "learning_rate": 2e-05,
+      "loss": 1.2637,
+      "step": 9785
+    },
+    {
+      "epoch": 0.23293344758640389,
+      "grad_norm": 4.056490487612006,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 9786
+    },
+    {
+      "epoch": 0.2329572503094354,
+      "grad_norm": 11.312795669129022,
+      "learning_rate": 2e-05,
+      "loss": 0.3113,
+      "step": 9787
+    },
+    {
+      "epoch": 0.2329810530324669,
+      "grad_norm": 2.6950782706674796,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 9788
+    },
+    {
+      "epoch": 0.23300485575549842,
+      "grad_norm": 3.3589096802320824,
+      "learning_rate": 2e-05,
+      "loss": 0.2431,
+      "step": 9789
+    },
+    {
+      "epoch": 0.23302865847852994,
+      "grad_norm": 2.9125795997551176,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 9790
+    },
+    {
+      "epoch": 0.23305246120156145,
+      "grad_norm": 1.8505210249827801,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 9791
+    },
+    {
+      "epoch": 0.23307626392459296,
+      "grad_norm": 12.881255125815985,
+      "learning_rate": 2e-05,
+      "loss": 0.8054,
+      "step": 9792
+    },
+    {
+      "epoch": 0.23310006664762448,
+      "grad_norm": 2.476342715681131,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 9793
+    },
+    {
+      "epoch": 0.233123869370656,
+      "grad_norm": 4.498242584436699,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 9794
+    },
+    {
+      "epoch": 0.23314767209368753,
+      "grad_norm": 2.5008224680740154,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 9795
+    },
+    {
+      "epoch": 0.23317147481671904,
+      "grad_norm": 4.4144421159916645,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 9796
+    },
+    {
+      "epoch": 0.23319527753975056,
+      "grad_norm": 2.965874080086928,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 9797
+    },
+    {
+      "epoch": 0.23321908026278207,
+      "grad_norm": 9.573644032403894,
+      "learning_rate": 2e-05,
+      "loss": 1.1944,
+      "step": 9798
+    },
+    {
+      "epoch": 0.23324288298581358,
+      "grad_norm": 7.726930443659633,
+      "learning_rate": 2e-05,
+      "loss": 0.3572,
+      "step": 9799
+    },
+    {
+      "epoch": 0.2332666857088451,
+      "grad_norm": 3.744176033452663,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 9800
+    },
+    {
+      "epoch": 0.2332904884318766,
+      "grad_norm": 2.178274855379829,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 9801
+    },
+    {
+      "epoch": 0.23331429115490812,
+      "grad_norm": 4.3245238177930965,
+      "learning_rate": 2e-05,
+      "loss": 0.3372,
+      "step": 9802
+    },
+    {
+      "epoch": 0.23333809387793963,
+      "grad_norm": 2.8323315028107348,
+      "learning_rate": 2e-05,
+      "loss": 0.2252,
+      "step": 9803
+    },
+    {
+      "epoch": 0.23336189660097115,
+      "grad_norm": 2.8962933014805636,
+      "learning_rate": 2e-05,
+      "loss": 0.2525,
+      "step": 9804
+    },
+    {
+      "epoch": 0.23338569932400266,
+      "grad_norm": 4.694629973368744,
+      "learning_rate": 2e-05,
+      "loss": 0.9862,
+      "step": 9805
+    },
+    {
+      "epoch": 0.23340950204703417,
+      "grad_norm": 2.4082097862444147,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 9806
+    },
+    {
+      "epoch": 0.23343330477006569,
+      "grad_norm": 4.935932769787872,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 9807
+    },
+    {
+      "epoch": 0.2334571074930972,
+      "grad_norm": 2.9009207114172098,
+      "learning_rate": 2e-05,
+      "loss": 0.1423,
+      "step": 9808
+    },
+    {
+      "epoch": 0.2334809102161287,
+      "grad_norm": 6.212614119088812,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 9809
+    },
+    {
+      "epoch": 0.23350471293916025,
+      "grad_norm": 2.341652704760411,
+      "learning_rate": 2e-05,
+      "loss": 0.1465,
+      "step": 9810
+    },
+    {
+      "epoch": 0.23352851566219177,
+      "grad_norm": 5.804899272780351,
+      "learning_rate": 2e-05,
+      "loss": 1.4688,
+      "step": 9811
+    },
+    {
+      "epoch": 0.23355231838522328,
+      "grad_norm": 4.597350770171291,
+      "learning_rate": 2e-05,
+      "loss": 0.85,
+      "step": 9812
+    },
+    {
+      "epoch": 0.2335761211082548,
+      "grad_norm": 5.008790840557464,
+      "learning_rate": 2e-05,
+      "loss": 0.8501,
+      "step": 9813
+    },
+    {
+      "epoch": 0.2335999238312863,
+      "grad_norm": 8.763076358569146,
+      "learning_rate": 2e-05,
+      "loss": 0.3368,
+      "step": 9814
+    },
+    {
+      "epoch": 0.23362372655431782,
+      "grad_norm": 2.4930070670806974,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 9815
+    },
+    {
+      "epoch": 0.23364752927734933,
+      "grad_norm": 2.3390662612030626,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 9816
+    },
+    {
+      "epoch": 0.23367133200038084,
+      "grad_norm": 2.4548441079570877,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 9817
+    },
+    {
+      "epoch": 0.23369513472341236,
+      "grad_norm": 2.9006807779031716,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 9818
+    },
+    {
+      "epoch": 0.23371893744644387,
+      "grad_norm": 2.084411183360357,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 9819
+    },
+    {
+      "epoch": 0.23374274016947538,
+      "grad_norm": 6.9163055815811765,
+      "learning_rate": 2e-05,
+      "loss": 0.3252,
+      "step": 9820
+    },
+    {
+      "epoch": 0.2337665428925069,
+      "grad_norm": 1.8859134682179353,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 9821
+    },
+    {
+      "epoch": 0.2337903456155384,
+      "grad_norm": 4.585369318072788,
+      "learning_rate": 2e-05,
+      "loss": 1.0355,
+      "step": 9822
+    },
+    {
+      "epoch": 0.23381414833856992,
+      "grad_norm": 6.747397906787987,
+      "learning_rate": 2e-05,
+      "loss": 0.4021,
+      "step": 9823
+    },
+    {
+      "epoch": 0.23383795106160143,
+      "grad_norm": 7.76817996035336,
+      "learning_rate": 2e-05,
+      "loss": 0.3356,
+      "step": 9824
+    },
+    {
+      "epoch": 0.23386175378463298,
+      "grad_norm": 4.271755388916336,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 9825
+    },
+    {
+      "epoch": 0.2338855565076645,
+      "grad_norm": 8.042947121130105,
+      "learning_rate": 2e-05,
+      "loss": 1.8052,
+      "step": 9826
+    },
+    {
+      "epoch": 0.233909359230696,
+      "grad_norm": 5.857313885357979,
+      "learning_rate": 2e-05,
+      "loss": 0.2789,
+      "step": 9827
+    },
+    {
+      "epoch": 0.23393316195372751,
+      "grad_norm": 2.1473269919230815,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 9828
+    },
+    {
+      "epoch": 0.23395696467675903,
+      "grad_norm": 10.37221849024658,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 9829
+    },
+    {
+      "epoch": 0.23398076739979054,
+      "grad_norm": 3.4876435829507466,
+      "learning_rate": 2e-05,
+      "loss": 0.636,
+      "step": 9830
+    },
+    {
+      "epoch": 0.23400457012282205,
+      "grad_norm": 10.638673279549357,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 9831
+    },
+    {
+      "epoch": 0.23402837284585357,
+      "grad_norm": 21.065579680539173,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 9832
+    },
+    {
+      "epoch": 0.23405217556888508,
+      "grad_norm": 3.417352744811548,
+      "learning_rate": 2e-05,
+      "loss": 0.6731,
+      "step": 9833
+    },
+    {
+      "epoch": 0.2340759782919166,
+      "grad_norm": 4.080028352660795,
+      "learning_rate": 2e-05,
+      "loss": 0.6158,
+      "step": 9834
+    },
+    {
+      "epoch": 0.2340997810149481,
+      "grad_norm": 18.038286687411823,
+      "learning_rate": 2e-05,
+      "loss": 0.7148,
+      "step": 9835
+    },
+    {
+      "epoch": 0.23412358373797962,
+      "grad_norm": 22.062080369346848,
+      "learning_rate": 2e-05,
+      "loss": 0.3789,
+      "step": 9836
+    },
+    {
+      "epoch": 0.23414738646101113,
+      "grad_norm": 3.393710492888937,
+      "learning_rate": 2e-05,
+      "loss": 0.151,
+      "step": 9837
+    },
+    {
+      "epoch": 0.23417118918404264,
+      "grad_norm": 5.068828419386981,
+      "learning_rate": 2e-05,
+      "loss": 0.5667,
+      "step": 9838
+    },
+    {
+      "epoch": 0.23419499190707416,
+      "grad_norm": 5.964159909725309,
+      "learning_rate": 2e-05,
+      "loss": 0.2704,
+      "step": 9839
+    },
+    {
+      "epoch": 0.2342187946301057,
+      "grad_norm": 3.38696404984303,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 9840
+    },
+    {
+      "epoch": 0.2342425973531372,
+      "grad_norm": 4.101846043375054,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 9841
+    },
+    {
+      "epoch": 0.23426640007616872,
+      "grad_norm": 1.0039461213305398,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 9842
+    },
+    {
+      "epoch": 0.23429020279920024,
+      "grad_norm": 0.791312506775162,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 9843
+    },
+    {
+      "epoch": 0.23431400552223175,
+      "grad_norm": 6.336040117949207,
+      "learning_rate": 2e-05,
+      "loss": 0.7392,
+      "step": 9844
+    },
+    {
+      "epoch": 0.23433780824526326,
+      "grad_norm": 3.7796682886528092,
+      "learning_rate": 2e-05,
+      "loss": 0.2075,
+      "step": 9845
+    },
+    {
+      "epoch": 0.23436161096829478,
+      "grad_norm": 0.8769855557413926,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 9846
+    },
+    {
+      "epoch": 0.2343854136913263,
+      "grad_norm": 4.441766091132055,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 9847
+    },
+    {
+      "epoch": 0.2344092164143578,
+      "grad_norm": 4.276599246638745,
+      "learning_rate": 2e-05,
+      "loss": 0.7128,
+      "step": 9848
+    },
+    {
+      "epoch": 0.23443301913738931,
+      "grad_norm": 12.751637900698618,
+      "learning_rate": 2e-05,
+      "loss": 0.7343,
+      "step": 9849
+    },
+    {
+      "epoch": 0.23445682186042083,
+      "grad_norm": 5.125816675165102,
+      "learning_rate": 2e-05,
+      "loss": 1.0256,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23448062458345234,
+      "grad_norm": 2.7884348456835624,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 9851
+    },
+    {
+      "epoch": 0.23450442730648385,
+      "grad_norm": 5.321580784644442,
+      "learning_rate": 2e-05,
+      "loss": 1.0058,
+      "step": 9852
+    },
+    {
+      "epoch": 0.23452823002951537,
+      "grad_norm": 19.4986284851891,
+      "learning_rate": 2e-05,
+      "loss": 1.2972,
+      "step": 9853
+    },
+    {
+      "epoch": 0.23455203275254688,
+      "grad_norm": 2.8282498333132717,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 9854
+    },
+    {
+      "epoch": 0.2345758354755784,
+      "grad_norm": 4.7136422190407945,
+      "learning_rate": 2e-05,
+      "loss": 0.6941,
+      "step": 9855
+    },
+    {
+      "epoch": 0.23459963819860993,
+      "grad_norm": 6.047082635287852,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 9856
+    },
+    {
+      "epoch": 0.23462344092164145,
+      "grad_norm": 4.232031652086799,
+      "learning_rate": 2e-05,
+      "loss": 0.8124,
+      "step": 9857
+    },
+    {
+      "epoch": 0.23464724364467296,
+      "grad_norm": 4.6294981791134076,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 9858
+    },
+    {
+      "epoch": 0.23467104636770447,
+      "grad_norm": 2.5131907920930137,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 9859
+    },
+    {
+      "epoch": 0.23469484909073599,
+      "grad_norm": 1.6406271347295707,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 9860
+    },
+    {
+      "epoch": 0.2347186518137675,
+      "grad_norm": 7.116209407267321,
+      "learning_rate": 2e-05,
+      "loss": 0.7684,
+      "step": 9861
+    },
+    {
+      "epoch": 0.234742454536799,
+      "grad_norm": 2.933859035880722,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 9862
+    },
+    {
+      "epoch": 0.23476625725983052,
+      "grad_norm": 5.906558832755264,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 9863
+    },
+    {
+      "epoch": 0.23479005998286204,
+      "grad_norm": 1.5606507028366423,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 9864
+    },
+    {
+      "epoch": 0.23481386270589355,
+      "grad_norm": 4.68364054576626,
+      "learning_rate": 2e-05,
+      "loss": 0.57,
+      "step": 9865
+    },
+    {
+      "epoch": 0.23483766542892506,
+      "grad_norm": 4.274198154476648,
+      "learning_rate": 2e-05,
+      "loss": 0.2697,
+      "step": 9866
+    },
+    {
+      "epoch": 0.23486146815195658,
+      "grad_norm": 5.949055748764581,
+      "learning_rate": 2e-05,
+      "loss": 0.9702,
+      "step": 9867
+    },
+    {
+      "epoch": 0.2348852708749881,
+      "grad_norm": 3.747012308106694,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 9868
+    },
+    {
+      "epoch": 0.2349090735980196,
+      "grad_norm": 5.474694224744325,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 9869
+    },
+    {
+      "epoch": 0.23493287632105112,
+      "grad_norm": 4.736571894113525,
+      "learning_rate": 2e-05,
+      "loss": 0.6335,
+      "step": 9870
+    },
+    {
+      "epoch": 0.23495667904408266,
+      "grad_norm": 5.772494302889139,
+      "learning_rate": 2e-05,
+      "loss": 0.8607,
+      "step": 9871
+    },
+    {
+      "epoch": 0.23498048176711417,
+      "grad_norm": 4.435199716855027,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 9872
+    },
+    {
+      "epoch": 0.23500428449014568,
+      "grad_norm": 4.148473213830138,
+      "learning_rate": 2e-05,
+      "loss": 0.5266,
+      "step": 9873
+    },
+    {
+      "epoch": 0.2350280872131772,
+      "grad_norm": 37.52387996486819,
+      "learning_rate": 2e-05,
+      "loss": 0.553,
+      "step": 9874
+    },
+    {
+      "epoch": 0.2350518899362087,
+      "grad_norm": 4.546511221501612,
+      "learning_rate": 2e-05,
+      "loss": 0.8481,
+      "step": 9875
+    },
+    {
+      "epoch": 0.23507569265924022,
+      "grad_norm": 12.795672906802976,
+      "learning_rate": 2e-05,
+      "loss": 1.1108,
+      "step": 9876
+    },
+    {
+      "epoch": 0.23509949538227173,
+      "grad_norm": 1.856005037030009,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 9877
+    },
+    {
+      "epoch": 0.23512329810530325,
+      "grad_norm": 5.172385618918321,
+      "learning_rate": 2e-05,
+      "loss": 0.7415,
+      "step": 9878
+    },
+    {
+      "epoch": 0.23514710082833476,
+      "grad_norm": 13.172475079132271,
+      "learning_rate": 2e-05,
+      "loss": 1.6211,
+      "step": 9879
+    },
+    {
+      "epoch": 0.23517090355136627,
+      "grad_norm": 5.366713892386562,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23519470627439779,
+      "grad_norm": 7.065273869461342,
+      "learning_rate": 2e-05,
+      "loss": 0.2536,
+      "step": 9881
+    },
+    {
+      "epoch": 0.2352185089974293,
+      "grad_norm": 9.01849325578013,
+      "learning_rate": 2e-05,
+      "loss": 0.5899,
+      "step": 9882
+    },
+    {
+      "epoch": 0.2352423117204608,
+      "grad_norm": 5.531370541000579,
+      "learning_rate": 2e-05,
+      "loss": 1.9044,
+      "step": 9883
+    },
+    {
+      "epoch": 0.23526611444349232,
+      "grad_norm": 22.973395213462577,
+      "learning_rate": 2e-05,
+      "loss": 0.3562,
+      "step": 9884
+    },
+    {
+      "epoch": 0.23528991716652384,
+      "grad_norm": 4.432254214249484,
+      "learning_rate": 2e-05,
+      "loss": 0.6259,
+      "step": 9885
+    },
+    {
+      "epoch": 0.23531371988955538,
+      "grad_norm": 1.9919581072829098,
+      "learning_rate": 2e-05,
+      "loss": 0.1031,
+      "step": 9886
+    },
+    {
+      "epoch": 0.2353375226125869,
+      "grad_norm": 9.251743754642616,
+      "learning_rate": 2e-05,
+      "loss": 0.4937,
+      "step": 9887
+    },
+    {
+      "epoch": 0.2353613253356184,
+      "grad_norm": 5.013623782556426,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 9888
+    },
+    {
+      "epoch": 0.23538512805864992,
+      "grad_norm": 7.847934032472038,
+      "learning_rate": 2e-05,
+      "loss": 0.3026,
+      "step": 9889
+    },
+    {
+      "epoch": 0.23540893078168143,
+      "grad_norm": 23.467304370638384,
+      "learning_rate": 2e-05,
+      "loss": 0.2133,
+      "step": 9890
+    },
+    {
+      "epoch": 0.23543273350471294,
+      "grad_norm": 1.7598897525527992,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 9891
+    },
+    {
+      "epoch": 0.23545653622774446,
+      "grad_norm": 3.7063305722799207,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 9892
+    },
+    {
+      "epoch": 0.23548033895077597,
+      "grad_norm": 1.6714058589360523,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 9893
+    },
+    {
+      "epoch": 0.23550414167380748,
+      "grad_norm": 4.278224091566265,
+      "learning_rate": 2e-05,
+      "loss": 0.1958,
+      "step": 9894
+    },
+    {
+      "epoch": 0.235527944396839,
+      "grad_norm": 4.5360308685854545,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 9895
+    },
+    {
+      "epoch": 0.2355517471198705,
+      "grad_norm": 3.2524728205303743,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 9896
+    },
+    {
+      "epoch": 0.23557554984290202,
+      "grad_norm": 1.2112635349598313,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 9897
+    },
+    {
+      "epoch": 0.23559935256593353,
+      "grad_norm": 8.547896642132295,
+      "learning_rate": 2e-05,
+      "loss": 0.9851,
+      "step": 9898
+    },
+    {
+      "epoch": 0.23562315528896505,
+      "grad_norm": 4.077278392728522,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 9899
+    },
+    {
+      "epoch": 0.23564695801199656,
+      "grad_norm": 3.60240619798673,
+      "learning_rate": 2e-05,
+      "loss": 0.4579,
+      "step": 9900
+    },
+    {
+      "epoch": 0.2356707607350281,
+      "grad_norm": 1.7244181308017272,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 9901
+    },
+    {
+      "epoch": 0.2356945634580596,
+      "grad_norm": 6.135333407447248,
+      "learning_rate": 2e-05,
+      "loss": 0.7469,
+      "step": 9902
+    },
+    {
+      "epoch": 0.23571836618109113,
+      "grad_norm": 2.5277174001054727,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 9903
+    },
+    {
+      "epoch": 0.23574216890412264,
+      "grad_norm": 4.049906793485375,
+      "learning_rate": 2e-05,
+      "loss": 0.2677,
+      "step": 9904
+    },
+    {
+      "epoch": 0.23576597162715415,
+      "grad_norm": 3.184550975006464,
+      "learning_rate": 2e-05,
+      "loss": 0.3155,
+      "step": 9905
+    },
+    {
+      "epoch": 0.23578977435018567,
+      "grad_norm": 3.686355654726171,
+      "learning_rate": 2e-05,
+      "loss": 0.2168,
+      "step": 9906
+    },
+    {
+      "epoch": 0.23581357707321718,
+      "grad_norm": 5.869188267314023,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 9907
+    },
+    {
+      "epoch": 0.2358373797962487,
+      "grad_norm": 0.6928200165798181,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 9908
+    },
+    {
+      "epoch": 0.2358611825192802,
+      "grad_norm": 1.6189952847997422,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 9909
+    },
+    {
+      "epoch": 0.23588498524231172,
+      "grad_norm": 0.7657905970166143,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23590878796534323,
+      "grad_norm": 2.853070827325011,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 9911
+    },
+    {
+      "epoch": 0.23593259068837474,
+      "grad_norm": 2.0878577873594226,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 9912
+    },
+    {
+      "epoch": 0.23595639341140626,
+      "grad_norm": 5.2189548899878995,
+      "learning_rate": 2e-05,
+      "loss": 0.2248,
+      "step": 9913
+    },
+    {
+      "epoch": 0.23598019613443777,
+      "grad_norm": 10.32001483838017,
+      "learning_rate": 2e-05,
+      "loss": 0.5637,
+      "step": 9914
+    },
+    {
+      "epoch": 0.23600399885746928,
+      "grad_norm": 4.544167201660522,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 9915
+    },
+    {
+      "epoch": 0.2360278015805008,
+      "grad_norm": 3.6733662943591683,
+      "learning_rate": 2e-05,
+      "loss": 0.5883,
+      "step": 9916
+    },
+    {
+      "epoch": 0.23605160430353234,
+      "grad_norm": 3.2403586981802386,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 9917
+    },
+    {
+      "epoch": 0.23607540702656385,
+      "grad_norm": 4.9834562147326045,
+      "learning_rate": 2e-05,
+      "loss": 0.9666,
+      "step": 9918
+    },
+    {
+      "epoch": 0.23609920974959536,
+      "grad_norm": 6.238045803702202,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 9919
+    },
+    {
+      "epoch": 0.23612301247262688,
+      "grad_norm": 2.97800558512258,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 9920
+    },
+    {
+      "epoch": 0.2361468151956584,
+      "grad_norm": 3.4858166188453135,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 9921
+    },
+    {
+      "epoch": 0.2361706179186899,
+      "grad_norm": 8.051776718814056,
+      "learning_rate": 2e-05,
+      "loss": 1.2921,
+      "step": 9922
+    },
+    {
+      "epoch": 0.23619442064172141,
+      "grad_norm": 5.005768195781688,
+      "learning_rate": 2e-05,
+      "loss": 0.3004,
+      "step": 9923
+    },
+    {
+      "epoch": 0.23621822336475293,
+      "grad_norm": 16.258915688248457,
+      "learning_rate": 2e-05,
+      "loss": 0.796,
+      "step": 9924
+    },
+    {
+      "epoch": 0.23624202608778444,
+      "grad_norm": 16.012921837410595,
+      "learning_rate": 2e-05,
+      "loss": 0.9599,
+      "step": 9925
+    },
+    {
+      "epoch": 0.23626582881081595,
+      "grad_norm": 5.93770809181876,
+      "learning_rate": 2e-05,
+      "loss": 1.2966,
+      "step": 9926
+    },
+    {
+      "epoch": 0.23628963153384747,
+      "grad_norm": 1.0897722603915792,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 9927
+    },
+    {
+      "epoch": 0.23631343425687898,
+      "grad_norm": 15.007287025582267,
+      "learning_rate": 2e-05,
+      "loss": 0.3596,
+      "step": 9928
+    },
+    {
+      "epoch": 0.2363372369799105,
+      "grad_norm": 6.549021062994045,
+      "learning_rate": 2e-05,
+      "loss": 0.8378,
+      "step": 9929
+    },
+    {
+      "epoch": 0.236361039702942,
+      "grad_norm": 1.5005908217763821,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23638484242597352,
+      "grad_norm": 2.2594479263633227,
+      "learning_rate": 2e-05,
+      "loss": 0.1532,
+      "step": 9931
+    },
+    {
+      "epoch": 0.23640864514900506,
+      "grad_norm": 10.07421127973934,
+      "learning_rate": 2e-05,
+      "loss": 1.218,
+      "step": 9932
+    },
+    {
+      "epoch": 0.23643244787203657,
+      "grad_norm": 8.343787324969753,
+      "learning_rate": 2e-05,
+      "loss": 0.5336,
+      "step": 9933
+    },
+    {
+      "epoch": 0.23645625059506808,
+      "grad_norm": 2.753618766279132,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 9934
+    },
+    {
+      "epoch": 0.2364800533180996,
+      "grad_norm": 3.6327630422029813,
+      "learning_rate": 2e-05,
+      "loss": 0.3896,
+      "step": 9935
+    },
+    {
+      "epoch": 0.2365038560411311,
+      "grad_norm": 3.101413055093787,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 9936
+    },
+    {
+      "epoch": 0.23652765876416262,
+      "grad_norm": 4.203795172492097,
+      "learning_rate": 2e-05,
+      "loss": 0.6433,
+      "step": 9937
+    },
+    {
+      "epoch": 0.23655146148719414,
+      "grad_norm": 2.174086250068279,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 9938
+    },
+    {
+      "epoch": 0.23657526421022565,
+      "grad_norm": 8.992683696062818,
+      "learning_rate": 2e-05,
+      "loss": 0.917,
+      "step": 9939
+    },
+    {
+      "epoch": 0.23659906693325716,
+      "grad_norm": 6.681849135418063,
+      "learning_rate": 2e-05,
+      "loss": 1.6538,
+      "step": 9940
+    },
+    {
+      "epoch": 0.23662286965628868,
+      "grad_norm": 2.871314209643567,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 9941
+    },
+    {
+      "epoch": 0.2366466723793202,
+      "grad_norm": 3.693524621145293,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 9942
+    },
+    {
+      "epoch": 0.2366704751023517,
+      "grad_norm": 5.438147494156701,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 9943
+    },
+    {
+      "epoch": 0.23669427782538321,
+      "grad_norm": 1.638046953965268,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 9944
+    },
+    {
+      "epoch": 0.23671808054841473,
+      "grad_norm": 2.551057975768333,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 9945
+    },
+    {
+      "epoch": 0.23674188327144624,
+      "grad_norm": 4.112334483818425,
+      "learning_rate": 2e-05,
+      "loss": 0.1228,
+      "step": 9946
+    },
+    {
+      "epoch": 0.23676568599447778,
+      "grad_norm": 5.84499845542705,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 9947
+    },
+    {
+      "epoch": 0.2367894887175093,
+      "grad_norm": 5.136648921562534,
+      "learning_rate": 2e-05,
+      "loss": 0.2585,
+      "step": 9948
+    },
+    {
+      "epoch": 0.2368132914405408,
+      "grad_norm": 3.7389238737656045,
+      "learning_rate": 2e-05,
+      "loss": 0.1931,
+      "step": 9949
+    },
+    {
+      "epoch": 0.23683709416357232,
+      "grad_norm": 3.3972515679093593,
+      "learning_rate": 2e-05,
+      "loss": 0.328,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23686089688660383,
+      "grad_norm": 4.771099014576597,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 9951
+    },
+    {
+      "epoch": 0.23688469960963535,
+      "grad_norm": 5.899989630013796,
+      "learning_rate": 2e-05,
+      "loss": 0.8435,
+      "step": 9952
+    },
+    {
+      "epoch": 0.23690850233266686,
+      "grad_norm": 7.9876419728699,
+      "learning_rate": 2e-05,
+      "loss": 1.8878,
+      "step": 9953
+    },
+    {
+      "epoch": 0.23693230505569837,
+      "grad_norm": 6.225465971080275,
+      "learning_rate": 2e-05,
+      "loss": 0.9909,
+      "step": 9954
+    },
+    {
+      "epoch": 0.23695610777872989,
+      "grad_norm": 3.55068309413892,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 9955
+    },
+    {
+      "epoch": 0.2369799105017614,
+      "grad_norm": 12.081445517301342,
+      "learning_rate": 2e-05,
+      "loss": 1.0559,
+      "step": 9956
+    },
+    {
+      "epoch": 0.2370037132247929,
+      "grad_norm": 7.782879356545282,
+      "learning_rate": 2e-05,
+      "loss": 0.5742,
+      "step": 9957
+    },
+    {
+      "epoch": 0.23702751594782442,
+      "grad_norm": 2.349365000489302,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 9958
+    },
+    {
+      "epoch": 0.23705131867085594,
+      "grad_norm": 8.205186609566123,
+      "learning_rate": 2e-05,
+      "loss": 0.467,
+      "step": 9959
+    },
+    {
+      "epoch": 0.23707512139388745,
+      "grad_norm": 1.4333743789519025,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 9960
+    },
+    {
+      "epoch": 0.23709892411691896,
+      "grad_norm": 7.070605764944952,
+      "learning_rate": 2e-05,
+      "loss": 0.3315,
+      "step": 9961
+    },
+    {
+      "epoch": 0.2371227268399505,
+      "grad_norm": 2.792042226915726,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 9962
+    },
+    {
+      "epoch": 0.23714652956298202,
+      "grad_norm": 6.767333429999892,
+      "learning_rate": 2e-05,
+      "loss": 0.3798,
+      "step": 9963
+    },
+    {
+      "epoch": 0.23717033228601353,
+      "grad_norm": 4.0239631085511185,
+      "learning_rate": 2e-05,
+      "loss": 0.7128,
+      "step": 9964
+    },
+    {
+      "epoch": 0.23719413500904504,
+      "grad_norm": 7.422707148073649,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 9965
+    },
+    {
+      "epoch": 0.23721793773207656,
+      "grad_norm": 4.642019724822086,
+      "learning_rate": 2e-05,
+      "loss": 0.2773,
+      "step": 9966
+    },
+    {
+      "epoch": 0.23724174045510807,
+      "grad_norm": 4.914587737559364,
+      "learning_rate": 2e-05,
+      "loss": 0.5342,
+      "step": 9967
+    },
+    {
+      "epoch": 0.23726554317813958,
+      "grad_norm": 4.830538450461741,
+      "learning_rate": 2e-05,
+      "loss": 0.5057,
+      "step": 9968
+    },
+    {
+      "epoch": 0.2372893459011711,
+      "grad_norm": 8.07106471787473,
+      "learning_rate": 2e-05,
+      "loss": 0.3112,
+      "step": 9969
+    },
+    {
+      "epoch": 0.2373131486242026,
+      "grad_norm": 12.403322751133146,
+      "learning_rate": 2e-05,
+      "loss": 0.6287,
+      "step": 9970
+    },
+    {
+      "epoch": 0.23733695134723412,
+      "grad_norm": 1.5567187429903386,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 9971
+    },
+    {
+      "epoch": 0.23736075407026563,
+      "grad_norm": 1.741224660180196,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 9972
+    },
+    {
+      "epoch": 0.23738455679329715,
+      "grad_norm": 2.4527539369836955,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 9973
+    },
+    {
+      "epoch": 0.23740835951632866,
+      "grad_norm": 5.495112148358975,
+      "learning_rate": 2e-05,
+      "loss": 0.9381,
+      "step": 9974
+    },
+    {
+      "epoch": 0.23743216223936017,
+      "grad_norm": 6.7834060454356955,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 9975
+    },
+    {
+      "epoch": 0.23745596496239169,
+      "grad_norm": 1.7797270898139004,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 9976
+    },
+    {
+      "epoch": 0.2374797676854232,
+      "grad_norm": 0.42850100591118473,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 9977
+    },
+    {
+      "epoch": 0.23750357040845474,
+      "grad_norm": 5.720077455893404,
+      "learning_rate": 2e-05,
+      "loss": 0.9084,
+      "step": 9978
+    },
+    {
+      "epoch": 0.23752737313148625,
+      "grad_norm": 5.350770633126773,
+      "learning_rate": 2e-05,
+      "loss": 0.6825,
+      "step": 9979
+    },
+    {
+      "epoch": 0.23755117585451777,
+      "grad_norm": 4.732723339902365,
+      "learning_rate": 2e-05,
+      "loss": 0.8395,
+      "step": 9980
+    },
+    {
+      "epoch": 0.23757497857754928,
+      "grad_norm": 6.66207831506907,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 9981
+    },
+    {
+      "epoch": 0.2375987813005808,
+      "grad_norm": 4.437721034858498,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 9982
+    },
+    {
+      "epoch": 0.2376225840236123,
+      "grad_norm": 2.0813547969780877,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 9983
+    },
+    {
+      "epoch": 0.23764638674664382,
+      "grad_norm": 12.885796277345548,
+      "learning_rate": 2e-05,
+      "loss": 0.2969,
+      "step": 9984
+    },
+    {
+      "epoch": 0.23767018946967533,
+      "grad_norm": 4.368077671022311,
+      "learning_rate": 2e-05,
+      "loss": 0.6159,
+      "step": 9985
+    },
+    {
+      "epoch": 0.23769399219270684,
+      "grad_norm": 3.711995268484604,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 9986
+    },
+    {
+      "epoch": 0.23771779491573836,
+      "grad_norm": 4.204388421384874,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 9987
+    },
+    {
+      "epoch": 0.23774159763876987,
+      "grad_norm": 3.5046859528000107,
+      "learning_rate": 2e-05,
+      "loss": 0.6391,
+      "step": 9988
+    },
+    {
+      "epoch": 0.23776540036180138,
+      "grad_norm": 7.7066864689875665,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 9989
+    },
+    {
+      "epoch": 0.2377892030848329,
+      "grad_norm": 12.99288368753823,
+      "learning_rate": 2e-05,
+      "loss": 0.9207,
+      "step": 9990
+    },
+    {
+      "epoch": 0.2378130058078644,
+      "grad_norm": 3.26013627034849,
+      "learning_rate": 2e-05,
+      "loss": 0.3875,
+      "step": 9991
+    },
+    {
+      "epoch": 0.23783680853089592,
+      "grad_norm": 2.2295344268220263,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 9992
+    },
+    {
+      "epoch": 0.23786061125392746,
+      "grad_norm": 3.4993659472441836,
+      "learning_rate": 2e-05,
+      "loss": 0.4519,
+      "step": 9993
+    },
+    {
+      "epoch": 0.23788441397695897,
+      "grad_norm": 7.446316772603121,
+      "learning_rate": 2e-05,
+      "loss": 0.6398,
+      "step": 9994
+    },
+    {
+      "epoch": 0.2379082166999905,
+      "grad_norm": 1.892071802857465,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 9995
+    },
+    {
+      "epoch": 0.237932019423022,
+      "grad_norm": 1.7954668529989795,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 9996
+    },
+    {
+      "epoch": 0.23795582214605351,
+      "grad_norm": 3.265486102026222,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 9997
+    },
+    {
+      "epoch": 0.23797962486908503,
+      "grad_norm": 2.5342455535766493,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 9998
+    },
+    {
+      "epoch": 0.23800342759211654,
+      "grad_norm": 5.628046224707814,
+      "learning_rate": 2e-05,
+      "loss": 0.2867,
+      "step": 9999
+    },
+    {
+      "epoch": 0.23802723031514805,
+      "grad_norm": 13.898669529670586,
+      "learning_rate": 2e-05,
+      "loss": 1.0465,
+      "step": 10000
+    },
+    {
+      "epoch": 0.23805103303817957,
+      "grad_norm": 2.595081001478918,
+      "learning_rate": 2e-05,
+      "loss": 0.1522,
+      "step": 10001
+    },
+    {
+      "epoch": 0.23807483576121108,
+      "grad_norm": 2.8332260888804086,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 10002
+    },
+    {
+      "epoch": 0.2380986384842426,
+      "grad_norm": 2.565941344027251,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 10003
+    },
+    {
+      "epoch": 0.2381224412072741,
+      "grad_norm": 8.288274783590117,
+      "learning_rate": 2e-05,
+      "loss": 0.6992,
+      "step": 10004
+    },
+    {
+      "epoch": 0.23814624393030562,
+      "grad_norm": 3.2859696612253546,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 10005
+    },
+    {
+      "epoch": 0.23817004665333713,
+      "grad_norm": 19.27035671440318,
+      "learning_rate": 2e-05,
+      "loss": 1.0449,
+      "step": 10006
+    },
+    {
+      "epoch": 0.23819384937636864,
+      "grad_norm": 14.485510654475252,
+      "learning_rate": 2e-05,
+      "loss": 0.2446,
+      "step": 10007
+    },
+    {
+      "epoch": 0.23821765209940018,
+      "grad_norm": 3.3194953282884043,
+      "learning_rate": 2e-05,
+      "loss": 0.2278,
+      "step": 10008
+    },
+    {
+      "epoch": 0.2382414548224317,
+      "grad_norm": 3.6624092724401045,
+      "learning_rate": 2e-05,
+      "loss": 0.3626,
+      "step": 10009
+    },
+    {
+      "epoch": 0.2382652575454632,
+      "grad_norm": 7.838773619201024,
+      "learning_rate": 2e-05,
+      "loss": 0.7806,
+      "step": 10010
+    },
+    {
+      "epoch": 0.23828906026849472,
+      "grad_norm": 2.51416172255916,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 10011
+    },
+    {
+      "epoch": 0.23831286299152624,
+      "grad_norm": 5.595319978711279,
+      "learning_rate": 2e-05,
+      "loss": 1.1009,
+      "step": 10012
+    },
+    {
+      "epoch": 0.23833666571455775,
+      "grad_norm": 3.4043979251216325,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 10013
+    },
+    {
+      "epoch": 0.23836046843758926,
+      "grad_norm": 11.604128931309894,
+      "learning_rate": 2e-05,
+      "loss": 1.1501,
+      "step": 10014
+    },
+    {
+      "epoch": 0.23838427116062078,
+      "grad_norm": 2.3930883186226866,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 10015
+    },
+    {
+      "epoch": 0.2384080738836523,
+      "grad_norm": 6.3258095935188905,
+      "learning_rate": 2e-05,
+      "loss": 0.9159,
+      "step": 10016
+    },
+    {
+      "epoch": 0.2384318766066838,
+      "grad_norm": 12.342272196943407,
+      "learning_rate": 2e-05,
+      "loss": 0.3182,
+      "step": 10017
+    },
+    {
+      "epoch": 0.23845567932971531,
+      "grad_norm": 1.9359117153266934,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 10018
+    },
+    {
+      "epoch": 0.23847948205274683,
+      "grad_norm": 2.023826193466926,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 10019
+    },
+    {
+      "epoch": 0.23850328477577834,
+      "grad_norm": 11.738447294490449,
+      "learning_rate": 2e-05,
+      "loss": 0.5093,
+      "step": 10020
+    },
+    {
+      "epoch": 0.23852708749880985,
+      "grad_norm": 5.713052208271627,
+      "learning_rate": 2e-05,
+      "loss": 0.7607,
+      "step": 10021
+    },
+    {
+      "epoch": 0.23855089022184137,
+      "grad_norm": 3.8948181338130534,
+      "learning_rate": 2e-05,
+      "loss": 0.2767,
+      "step": 10022
+    },
+    {
+      "epoch": 0.2385746929448729,
+      "grad_norm": 13.111047617246571,
+      "learning_rate": 2e-05,
+      "loss": 0.4404,
+      "step": 10023
+    },
+    {
+      "epoch": 0.23859849566790442,
+      "grad_norm": 4.789281291477388,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 10024
+    },
+    {
+      "epoch": 0.23862229839093593,
+      "grad_norm": 12.372205129474535,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 10025
+    },
+    {
+      "epoch": 0.23864610111396745,
+      "grad_norm": 1.6084781762641445,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 10026
+    },
+    {
+      "epoch": 0.23866990383699896,
+      "grad_norm": 23.467785285727487,
+      "learning_rate": 2e-05,
+      "loss": 1.0968,
+      "step": 10027
+    },
+    {
+      "epoch": 0.23869370656003047,
+      "grad_norm": 1.9840898156507483,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 10028
+    },
+    {
+      "epoch": 0.23871750928306198,
+      "grad_norm": 11.231043292797498,
+      "learning_rate": 2e-05,
+      "loss": 2.5453,
+      "step": 10029
+    },
+    {
+      "epoch": 0.2387413120060935,
+      "grad_norm": 1.8625608887359715,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 10030
+    },
+    {
+      "epoch": 0.238765114729125,
+      "grad_norm": 3.9880257147332716,
+      "learning_rate": 2e-05,
+      "loss": 0.5375,
+      "step": 10031
+    },
+    {
+      "epoch": 0.23878891745215652,
+      "grad_norm": 9.038863732899593,
+      "learning_rate": 2e-05,
+      "loss": 0.3084,
+      "step": 10032
+    },
+    {
+      "epoch": 0.23881272017518804,
+      "grad_norm": 4.152042028944744,
+      "learning_rate": 2e-05,
+      "loss": 0.5975,
+      "step": 10033
+    },
+    {
+      "epoch": 0.23883652289821955,
+      "grad_norm": 5.6320238760207975,
+      "learning_rate": 2e-05,
+      "loss": 0.5798,
+      "step": 10034
+    },
+    {
+      "epoch": 0.23886032562125106,
+      "grad_norm": 5.611752512973526,
+      "learning_rate": 2e-05,
+      "loss": 0.9178,
+      "step": 10035
+    },
+    {
+      "epoch": 0.23888412834428258,
+      "grad_norm": 6.599867210974876,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 10036
+    },
+    {
+      "epoch": 0.2389079310673141,
+      "grad_norm": 36.76890487657806,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 10037
+    },
+    {
+      "epoch": 0.2389317337903456,
+      "grad_norm": 7.616115740400365,
+      "learning_rate": 2e-05,
+      "loss": 1.2585,
+      "step": 10038
+    },
+    {
+      "epoch": 0.23895553651337714,
+      "grad_norm": 2.0594568322459756,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 10039
+    },
+    {
+      "epoch": 0.23897933923640866,
+      "grad_norm": 4.524824564572836,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 10040
+    },
+    {
+      "epoch": 0.23900314195944017,
+      "grad_norm": 43.3564939578175,
+      "learning_rate": 2e-05,
+      "loss": 0.4966,
+      "step": 10041
+    },
+    {
+      "epoch": 0.23902694468247168,
+      "grad_norm": 11.785248324855633,
+      "learning_rate": 2e-05,
+      "loss": 0.3212,
+      "step": 10042
+    },
+    {
+      "epoch": 0.2390507474055032,
+      "grad_norm": 3.2854168011572726,
+      "learning_rate": 2e-05,
+      "loss": 0.1307,
+      "step": 10043
+    },
+    {
+      "epoch": 0.2390745501285347,
+      "grad_norm": 3.127972694625341,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 10044
+    },
+    {
+      "epoch": 0.23909835285156622,
+      "grad_norm": 5.4309757009556865,
+      "learning_rate": 2e-05,
+      "loss": 0.8659,
+      "step": 10045
+    },
+    {
+      "epoch": 0.23912215557459773,
+      "grad_norm": 4.105502703456678,
+      "learning_rate": 2e-05,
+      "loss": 0.7765,
+      "step": 10046
+    },
+    {
+      "epoch": 0.23914595829762925,
+      "grad_norm": 4.2679955764698745,
+      "learning_rate": 2e-05,
+      "loss": 0.6772,
+      "step": 10047
+    },
+    {
+      "epoch": 0.23916976102066076,
+      "grad_norm": 10.753089890799895,
+      "learning_rate": 2e-05,
+      "loss": 0.4987,
+      "step": 10048
+    },
+    {
+      "epoch": 0.23919356374369227,
+      "grad_norm": 7.696464686270608,
+      "learning_rate": 2e-05,
+      "loss": 0.5627,
+      "step": 10049
+    },
+    {
+      "epoch": 0.23921736646672379,
+      "grad_norm": 7.780692206640547,
+      "learning_rate": 2e-05,
+      "loss": 1.325,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2392411691897553,
+      "grad_norm": 11.468893763931062,
+      "learning_rate": 2e-05,
+      "loss": 1.1346,
+      "step": 10051
+    },
+    {
+      "epoch": 0.2392649719127868,
+      "grad_norm": 7.318855101866439,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 10052
+    },
+    {
+      "epoch": 0.23928877463581832,
+      "grad_norm": 5.140518593896744,
+      "learning_rate": 2e-05,
+      "loss": 0.6055,
+      "step": 10053
+    },
+    {
+      "epoch": 0.23931257735884987,
+      "grad_norm": 2.1671685172336552,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 10054
+    },
+    {
+      "epoch": 0.23933638008188138,
+      "grad_norm": 10.264079672580113,
+      "learning_rate": 2e-05,
+      "loss": 0.6046,
+      "step": 10055
+    },
+    {
+      "epoch": 0.2393601828049129,
+      "grad_norm": 2.812792957987188,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 10056
+    },
+    {
+      "epoch": 0.2393839855279444,
+      "grad_norm": 6.425609440256681,
+      "learning_rate": 2e-05,
+      "loss": 1.0247,
+      "step": 10057
+    },
+    {
+      "epoch": 0.23940778825097592,
+      "grad_norm": 2.561139314258295,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 10058
+    },
+    {
+      "epoch": 0.23943159097400743,
+      "grad_norm": 11.470856237798388,
+      "learning_rate": 2e-05,
+      "loss": 2.1354,
+      "step": 10059
+    },
+    {
+      "epoch": 0.23945539369703894,
+      "grad_norm": 7.561434944248472,
+      "learning_rate": 2e-05,
+      "loss": 0.8969,
+      "step": 10060
+    },
+    {
+      "epoch": 0.23947919642007046,
+      "grad_norm": 7.5564561557242325,
+      "learning_rate": 2e-05,
+      "loss": 0.311,
+      "step": 10061
+    },
+    {
+      "epoch": 0.23950299914310197,
+      "grad_norm": 11.070905720939301,
+      "learning_rate": 2e-05,
+      "loss": 1.1681,
+      "step": 10062
+    },
+    {
+      "epoch": 0.23952680186613348,
+      "grad_norm": 7.713290443465922,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 10063
+    },
+    {
+      "epoch": 0.239550604589165,
+      "grad_norm": 28.727674045905623,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 10064
+    },
+    {
+      "epoch": 0.2395744073121965,
+      "grad_norm": 6.389519256744544,
+      "learning_rate": 2e-05,
+      "loss": 1.2513,
+      "step": 10065
+    },
+    {
+      "epoch": 0.23959821003522802,
+      "grad_norm": 7.04189586412032,
+      "learning_rate": 2e-05,
+      "loss": 0.4151,
+      "step": 10066
+    },
+    {
+      "epoch": 0.23962201275825953,
+      "grad_norm": 2.260471111448473,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 10067
+    },
+    {
+      "epoch": 0.23964581548129105,
+      "grad_norm": 9.167728276952364,
+      "learning_rate": 2e-05,
+      "loss": 1.0551,
+      "step": 10068
+    },
+    {
+      "epoch": 0.2396696182043226,
+      "grad_norm": 10.080323418758312,
+      "learning_rate": 2e-05,
+      "loss": 0.7794,
+      "step": 10069
+    },
+    {
+      "epoch": 0.2396934209273541,
+      "grad_norm": 5.6465524884558995,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 10070
+    },
+    {
+      "epoch": 0.2397172236503856,
+      "grad_norm": 4.021979397803885,
+      "learning_rate": 2e-05,
+      "loss": 0.5211,
+      "step": 10071
+    },
+    {
+      "epoch": 0.23974102637341713,
+      "grad_norm": 1.9526672588474587,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 10072
+    },
+    {
+      "epoch": 0.23976482909644864,
+      "grad_norm": 6.7674630987502224,
+      "learning_rate": 2e-05,
+      "loss": 0.3551,
+      "step": 10073
+    },
+    {
+      "epoch": 0.23978863181948015,
+      "grad_norm": 6.7181413158256555,
+      "learning_rate": 2e-05,
+      "loss": 1.5682,
+      "step": 10074
+    },
+    {
+      "epoch": 0.23981243454251167,
+      "grad_norm": 6.146092292217513,
+      "learning_rate": 2e-05,
+      "loss": 0.953,
+      "step": 10075
+    },
+    {
+      "epoch": 0.23983623726554318,
+      "grad_norm": 6.273716632516616,
+      "learning_rate": 2e-05,
+      "loss": 0.5848,
+      "step": 10076
+    },
+    {
+      "epoch": 0.2398600399885747,
+      "grad_norm": 4.361211938141511,
+      "learning_rate": 2e-05,
+      "loss": 0.2244,
+      "step": 10077
+    },
+    {
+      "epoch": 0.2398838427116062,
+      "grad_norm": 1.743986214310313,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 10078
+    },
+    {
+      "epoch": 0.23990764543463772,
+      "grad_norm": 7.4077378011823045,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 10079
+    },
+    {
+      "epoch": 0.23993144815766923,
+      "grad_norm": 4.531947951517937,
+      "learning_rate": 2e-05,
+      "loss": 0.7412,
+      "step": 10080
+    },
+    {
+      "epoch": 0.23995525088070074,
+      "grad_norm": 5.325532976668049,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 10081
+    },
+    {
+      "epoch": 0.23997905360373226,
+      "grad_norm": 4.136591488718073,
+      "learning_rate": 2e-05,
+      "loss": 0.3473,
+      "step": 10082
+    },
+    {
+      "epoch": 0.24000285632676377,
+      "grad_norm": 6.391318082206075,
+      "learning_rate": 2e-05,
+      "loss": 0.2112,
+      "step": 10083
+    },
+    {
+      "epoch": 0.2400266590497953,
+      "grad_norm": 4.3636713181495335,
+      "learning_rate": 2e-05,
+      "loss": 0.9104,
+      "step": 10084
+    },
+    {
+      "epoch": 0.24005046177282682,
+      "grad_norm": 5.302177224544641,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 10085
+    },
+    {
+      "epoch": 0.24007426449585834,
+      "grad_norm": 2.40060187949071,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 10086
+    },
+    {
+      "epoch": 0.24009806721888985,
+      "grad_norm": 7.656103371371398,
+      "learning_rate": 2e-05,
+      "loss": 0.3252,
+      "step": 10087
+    },
+    {
+      "epoch": 0.24012186994192136,
+      "grad_norm": 7.178500407577741,
+      "learning_rate": 2e-05,
+      "loss": 0.6973,
+      "step": 10088
+    },
+    {
+      "epoch": 0.24014567266495288,
+      "grad_norm": 6.775597402041957,
+      "learning_rate": 2e-05,
+      "loss": 0.3299,
+      "step": 10089
+    },
+    {
+      "epoch": 0.2401694753879844,
+      "grad_norm": 2.3773959965853253,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 10090
+    },
+    {
+      "epoch": 0.2401932781110159,
+      "grad_norm": 1.2288683756599337,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 10091
+    },
+    {
+      "epoch": 0.24021708083404741,
+      "grad_norm": 5.68113918324594,
+      "learning_rate": 2e-05,
+      "loss": 0.9477,
+      "step": 10092
+    },
+    {
+      "epoch": 0.24024088355707893,
+      "grad_norm": 1.4338472293296234,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 10093
+    },
+    {
+      "epoch": 0.24026468628011044,
+      "grad_norm": 2.057823692520514,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 10094
+    },
+    {
+      "epoch": 0.24028848900314195,
+      "grad_norm": 2.0626080867767085,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 10095
+    },
+    {
+      "epoch": 0.24031229172617347,
+      "grad_norm": 3.496368312118026,
+      "learning_rate": 2e-05,
+      "loss": 0.7318,
+      "step": 10096
+    },
+    {
+      "epoch": 0.24033609444920498,
+      "grad_norm": 0.7353771492963752,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 10097
+    },
+    {
+      "epoch": 0.2403598971722365,
+      "grad_norm": 11.629040764863491,
+      "learning_rate": 2e-05,
+      "loss": 0.5384,
+      "step": 10098
+    },
+    {
+      "epoch": 0.240383699895268,
+      "grad_norm": 3.3498225488510394,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 10099
+    },
+    {
+      "epoch": 0.24040750261829955,
+      "grad_norm": 8.683355556511998,
+      "learning_rate": 2e-05,
+      "loss": 0.6544,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24043130534133106,
+      "grad_norm": 4.760628714761831,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 10101
+    },
+    {
+      "epoch": 0.24045510806436257,
+      "grad_norm": 6.486490956665729,
+      "learning_rate": 2e-05,
+      "loss": 0.2333,
+      "step": 10102
+    },
+    {
+      "epoch": 0.24047891078739408,
+      "grad_norm": 7.650641806658933,
+      "learning_rate": 2e-05,
+      "loss": 0.4251,
+      "step": 10103
+    },
+    {
+      "epoch": 0.2405027135104256,
+      "grad_norm": 5.8790174059473355,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 10104
+    },
+    {
+      "epoch": 0.2405265162334571,
+      "grad_norm": 0.43963824518328365,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 10105
+    },
+    {
+      "epoch": 0.24055031895648862,
+      "grad_norm": 4.6344835010460175,
+      "learning_rate": 2e-05,
+      "loss": 0.7557,
+      "step": 10106
+    },
+    {
+      "epoch": 0.24057412167952014,
+      "grad_norm": 4.917254213845892,
+      "learning_rate": 2e-05,
+      "loss": 0.7247,
+      "step": 10107
+    },
+    {
+      "epoch": 0.24059792440255165,
+      "grad_norm": 3.167590597151768,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 10108
+    },
+    {
+      "epoch": 0.24062172712558316,
+      "grad_norm": 2.772097032768253,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 10109
+    },
+    {
+      "epoch": 0.24064552984861468,
+      "grad_norm": 4.8875372631174185,
+      "learning_rate": 2e-05,
+      "loss": 0.6835,
+      "step": 10110
+    },
+    {
+      "epoch": 0.2406693325716462,
+      "grad_norm": 2.495072475692785,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 10111
+    },
+    {
+      "epoch": 0.2406931352946777,
+      "grad_norm": 2.9519206572631047,
+      "learning_rate": 2e-05,
+      "loss": 0.1302,
+      "step": 10112
+    },
+    {
+      "epoch": 0.24071693801770921,
+      "grad_norm": 5.12747202766159,
+      "learning_rate": 2e-05,
+      "loss": 0.176,
+      "step": 10113
+    },
+    {
+      "epoch": 0.24074074074074073,
+      "grad_norm": 4.929847693909087,
+      "learning_rate": 2e-05,
+      "loss": 0.4102,
+      "step": 10114
+    },
+    {
+      "epoch": 0.24076454346377227,
+      "grad_norm": 6.667643348094674,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 10115
+    },
+    {
+      "epoch": 0.24078834618680378,
+      "grad_norm": 5.2196235124955495,
+      "learning_rate": 2e-05,
+      "loss": 0.5899,
+      "step": 10116
+    },
+    {
+      "epoch": 0.2408121489098353,
+      "grad_norm": 7.372872707682166,
+      "learning_rate": 2e-05,
+      "loss": 0.8739,
+      "step": 10117
+    },
+    {
+      "epoch": 0.2408359516328668,
+      "grad_norm": 5.776262431603391,
+      "learning_rate": 2e-05,
+      "loss": 0.3471,
+      "step": 10118
+    },
+    {
+      "epoch": 0.24085975435589832,
+      "grad_norm": 1.3658307826880987,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 10119
+    },
+    {
+      "epoch": 0.24088355707892983,
+      "grad_norm": 5.351367209688058,
+      "learning_rate": 2e-05,
+      "loss": 0.7409,
+      "step": 10120
+    },
+    {
+      "epoch": 0.24090735980196135,
+      "grad_norm": 4.402997506761359,
+      "learning_rate": 2e-05,
+      "loss": 0.1836,
+      "step": 10121
+    },
+    {
+      "epoch": 0.24093116252499286,
+      "grad_norm": 3.804310105756987,
+      "learning_rate": 2e-05,
+      "loss": 0.3609,
+      "step": 10122
+    },
+    {
+      "epoch": 0.24095496524802437,
+      "grad_norm": 6.217161660842101,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 10123
+    },
+    {
+      "epoch": 0.24097876797105588,
+      "grad_norm": 7.53652712742559,
+      "learning_rate": 2e-05,
+      "loss": 1.2377,
+      "step": 10124
+    },
+    {
+      "epoch": 0.2410025706940874,
+      "grad_norm": 6.943724850337797,
+      "learning_rate": 2e-05,
+      "loss": 0.8882,
+      "step": 10125
+    },
+    {
+      "epoch": 0.2410263734171189,
+      "grad_norm": 3.010324132341494,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 10126
+    },
+    {
+      "epoch": 0.24105017614015042,
+      "grad_norm": 17.289062099176274,
+      "learning_rate": 2e-05,
+      "loss": 0.6286,
+      "step": 10127
+    },
+    {
+      "epoch": 0.24107397886318194,
+      "grad_norm": 5.638647362608521,
+      "learning_rate": 2e-05,
+      "loss": 0.7326,
+      "step": 10128
+    },
+    {
+      "epoch": 0.24109778158621345,
+      "grad_norm": 1.5663380897838952,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 10129
+    },
+    {
+      "epoch": 0.241121584309245,
+      "grad_norm": 8.172463251558945,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 10130
+    },
+    {
+      "epoch": 0.2411453870322765,
+      "grad_norm": 7.930569591752099,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 10131
+    },
+    {
+      "epoch": 0.24116918975530802,
+      "grad_norm": 4.445940990908837,
+      "learning_rate": 2e-05,
+      "loss": 0.605,
+      "step": 10132
+    },
+    {
+      "epoch": 0.24119299247833953,
+      "grad_norm": 3.4342041762407423,
+      "learning_rate": 2e-05,
+      "loss": 0.1537,
+      "step": 10133
+    },
+    {
+      "epoch": 0.24121679520137104,
+      "grad_norm": 1.5425752845832206,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 10134
+    },
+    {
+      "epoch": 0.24124059792440256,
+      "grad_norm": 3.1615429996384794,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 10135
+    },
+    {
+      "epoch": 0.24126440064743407,
+      "grad_norm": 2.102386436396337,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 10136
+    },
+    {
+      "epoch": 0.24128820337046558,
+      "grad_norm": 1.726493425081929,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 10137
+    },
+    {
+      "epoch": 0.2413120060934971,
+      "grad_norm": 2.461586048560489,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 10138
+    },
+    {
+      "epoch": 0.2413358088165286,
+      "grad_norm": 5.464087737395198,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 10139
+    },
+    {
+      "epoch": 0.24135961153956012,
+      "grad_norm": 9.519364562411523,
+      "learning_rate": 2e-05,
+      "loss": 0.5803,
+      "step": 10140
+    },
+    {
+      "epoch": 0.24138341426259163,
+      "grad_norm": 11.735483655943536,
+      "learning_rate": 2e-05,
+      "loss": 0.8531,
+      "step": 10141
+    },
+    {
+      "epoch": 0.24140721698562315,
+      "grad_norm": 2.9261442097854773,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 10142
+    },
+    {
+      "epoch": 0.24143101970865466,
+      "grad_norm": 1.2453593774225036,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 10143
+    },
+    {
+      "epoch": 0.24145482243168617,
+      "grad_norm": 3.8405514143650294,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 10144
+    },
+    {
+      "epoch": 0.2414786251547177,
+      "grad_norm": 3.3958699026438874,
+      "learning_rate": 2e-05,
+      "loss": 0.1519,
+      "step": 10145
+    },
+    {
+      "epoch": 0.24150242787774923,
+      "grad_norm": 1.3453522510313956,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 10146
+    },
+    {
+      "epoch": 0.24152623060078074,
+      "grad_norm": 6.6635199663574936,
+      "learning_rate": 2e-05,
+      "loss": 0.9093,
+      "step": 10147
+    },
+    {
+      "epoch": 0.24155003332381225,
+      "grad_norm": 5.751587798244529,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 10148
+    },
+    {
+      "epoch": 0.24157383604684377,
+      "grad_norm": 3.490079500332762,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 10149
+    },
+    {
+      "epoch": 0.24159763876987528,
+      "grad_norm": 7.763558670005956,
+      "learning_rate": 2e-05,
+      "loss": 1.4316,
+      "step": 10150
+    },
+    {
+      "epoch": 0.2416214414929068,
+      "grad_norm": 6.26068617011823,
+      "learning_rate": 2e-05,
+      "loss": 0.932,
+      "step": 10151
+    },
+    {
+      "epoch": 0.2416452442159383,
+      "grad_norm": 2.7783323400215734,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 10152
+    },
+    {
+      "epoch": 0.24166904693896982,
+      "grad_norm": 3.827514169171281,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 10153
+    },
+    {
+      "epoch": 0.24169284966200133,
+      "grad_norm": 1.512209677050076,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 10154
+    },
+    {
+      "epoch": 0.24171665238503284,
+      "grad_norm": 4.492402456754391,
+      "learning_rate": 2e-05,
+      "loss": 1.0322,
+      "step": 10155
+    },
+    {
+      "epoch": 0.24174045510806436,
+      "grad_norm": 10.94986853533337,
+      "learning_rate": 2e-05,
+      "loss": 0.7931,
+      "step": 10156
+    },
+    {
+      "epoch": 0.24176425783109587,
+      "grad_norm": 6.874367118331749,
+      "learning_rate": 2e-05,
+      "loss": 0.4448,
+      "step": 10157
+    },
+    {
+      "epoch": 0.24178806055412738,
+      "grad_norm": 3.2217627525447328,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 10158
+    },
+    {
+      "epoch": 0.2418118632771589,
+      "grad_norm": 4.950025936167602,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 10159
+    },
+    {
+      "epoch": 0.2418356660001904,
+      "grad_norm": 30.9093877605482,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 10160
+    },
+    {
+      "epoch": 0.24185946872322195,
+      "grad_norm": 2.0684150861120734,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 10161
+    },
+    {
+      "epoch": 0.24188327144625346,
+      "grad_norm": 3.915626811224296,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 10162
+    },
+    {
+      "epoch": 0.24190707416928497,
+      "grad_norm": 2.632366817729007,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 10163
+    },
+    {
+      "epoch": 0.2419308768923165,
+      "grad_norm": 3.7431385272464874,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 10164
+    },
+    {
+      "epoch": 0.241954679615348,
+      "grad_norm": 2.8026434519355448,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 10165
+    },
+    {
+      "epoch": 0.2419784823383795,
+      "grad_norm": 2.6296668430921826,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 10166
+    },
+    {
+      "epoch": 0.24200228506141103,
+      "grad_norm": 2.157450030634504,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 10167
+    },
+    {
+      "epoch": 0.24202608778444254,
+      "grad_norm": 3.6009698503426697,
+      "learning_rate": 2e-05,
+      "loss": 0.145,
+      "step": 10168
+    },
+    {
+      "epoch": 0.24204989050747405,
+      "grad_norm": 4.175920769888079,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 10169
+    },
+    {
+      "epoch": 0.24207369323050557,
+      "grad_norm": 1.2962984569668272,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 10170
+    },
+    {
+      "epoch": 0.24209749595353708,
+      "grad_norm": 3.7178217354952614,
+      "learning_rate": 2e-05,
+      "loss": 0.2908,
+      "step": 10171
+    },
+    {
+      "epoch": 0.2421212986765686,
+      "grad_norm": 4.516173026709346,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 10172
+    },
+    {
+      "epoch": 0.2421451013996001,
+      "grad_norm": 1.5149007005849184,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 10173
+    },
+    {
+      "epoch": 0.24216890412263162,
+      "grad_norm": 5.587366869934485,
+      "learning_rate": 2e-05,
+      "loss": 0.3236,
+      "step": 10174
+    },
+    {
+      "epoch": 0.24219270684566313,
+      "grad_norm": 4.066295114111068,
+      "learning_rate": 2e-05,
+      "loss": 0.588,
+      "step": 10175
+    },
+    {
+      "epoch": 0.24221650956869467,
+      "grad_norm": 7.72122088783568,
+      "learning_rate": 2e-05,
+      "loss": 1.2572,
+      "step": 10176
+    },
+    {
+      "epoch": 0.24224031229172618,
+      "grad_norm": 4.311068910381527,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 10177
+    },
+    {
+      "epoch": 0.2422641150147577,
+      "grad_norm": 9.739587295162162,
+      "learning_rate": 2e-05,
+      "loss": 1.2194,
+      "step": 10178
+    },
+    {
+      "epoch": 0.2422879177377892,
+      "grad_norm": 9.242217868551057,
+      "learning_rate": 2e-05,
+      "loss": 1.3833,
+      "step": 10179
+    },
+    {
+      "epoch": 0.24231172046082072,
+      "grad_norm": 9.986713697686609,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 10180
+    },
+    {
+      "epoch": 0.24233552318385224,
+      "grad_norm": 2.8648073062667385,
+      "learning_rate": 2e-05,
+      "loss": 0.134,
+      "step": 10181
+    },
+    {
+      "epoch": 0.24235932590688375,
+      "grad_norm": 2.1483499182260046,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 10182
+    },
+    {
+      "epoch": 0.24238312862991526,
+      "grad_norm": 14.571899209280478,
+      "learning_rate": 2e-05,
+      "loss": 0.4746,
+      "step": 10183
+    },
+    {
+      "epoch": 0.24240693135294678,
+      "grad_norm": 4.478975777633674,
+      "learning_rate": 2e-05,
+      "loss": 0.4442,
+      "step": 10184
+    },
+    {
+      "epoch": 0.2424307340759783,
+      "grad_norm": 2.4254900461558213,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 10185
+    },
+    {
+      "epoch": 0.2424545367990098,
+      "grad_norm": 2.949233604539727,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 10186
+    },
+    {
+      "epoch": 0.24247833952204131,
+      "grad_norm": 25.470674798308483,
+      "learning_rate": 2e-05,
+      "loss": 0.3801,
+      "step": 10187
+    },
+    {
+      "epoch": 0.24250214224507283,
+      "grad_norm": 2.5613213990537727,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 10188
+    },
+    {
+      "epoch": 0.24252594496810434,
+      "grad_norm": 5.7097531078310535,
+      "learning_rate": 2e-05,
+      "loss": 0.5298,
+      "step": 10189
+    },
+    {
+      "epoch": 0.24254974769113585,
+      "grad_norm": 9.09664601248479,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 10190
+    },
+    {
+      "epoch": 0.2425735504141674,
+      "grad_norm": 1.7270990597183746,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 10191
+    },
+    {
+      "epoch": 0.2425973531371989,
+      "grad_norm": 0.7764423818945996,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 10192
+    },
+    {
+      "epoch": 0.24262115586023042,
+      "grad_norm": 5.834497959970644,
+      "learning_rate": 2e-05,
+      "loss": 1.1794,
+      "step": 10193
+    },
+    {
+      "epoch": 0.24264495858326193,
+      "grad_norm": 2.4961154265174197,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 10194
+    },
+    {
+      "epoch": 0.24266876130629345,
+      "grad_norm": 6.164258886426267,
+      "learning_rate": 2e-05,
+      "loss": 0.5677,
+      "step": 10195
+    },
+    {
+      "epoch": 0.24269256402932496,
+      "grad_norm": 8.142727089774414,
+      "learning_rate": 2e-05,
+      "loss": 0.2221,
+      "step": 10196
+    },
+    {
+      "epoch": 0.24271636675235647,
+      "grad_norm": 2.2074492210821335,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 10197
+    },
+    {
+      "epoch": 0.24274016947538798,
+      "grad_norm": 1.4022556542957603,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 10198
+    },
+    {
+      "epoch": 0.2427639721984195,
+      "grad_norm": 9.282207946332615,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 10199
+    },
+    {
+      "epoch": 0.242787774921451,
+      "grad_norm": 3.0376066982702246,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24281157764448252,
+      "grad_norm": 2.2121428329716006,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 10201
+    },
+    {
+      "epoch": 0.24283538036751404,
+      "grad_norm": 7.524868168602154,
+      "learning_rate": 2e-05,
+      "loss": 0.6663,
+      "step": 10202
+    },
+    {
+      "epoch": 0.24285918309054555,
+      "grad_norm": 2.977687276179472,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 10203
+    },
+    {
+      "epoch": 0.24288298581357706,
+      "grad_norm": 1.0882786269802662,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 10204
+    },
+    {
+      "epoch": 0.24290678853660858,
+      "grad_norm": 1.3014379828082543,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 10205
+    },
+    {
+      "epoch": 0.24293059125964012,
+      "grad_norm": 4.578741436397371,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 10206
+    },
+    {
+      "epoch": 0.24295439398267163,
+      "grad_norm": 4.84656743780493,
+      "learning_rate": 2e-05,
+      "loss": 0.6645,
+      "step": 10207
+    },
+    {
+      "epoch": 0.24297819670570314,
+      "grad_norm": 1.922085528054493,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 10208
+    },
+    {
+      "epoch": 0.24300199942873466,
+      "grad_norm": 4.1460899723116835,
+      "learning_rate": 2e-05,
+      "loss": 0.6747,
+      "step": 10209
+    },
+    {
+      "epoch": 0.24302580215176617,
+      "grad_norm": 4.906618261467777,
+      "learning_rate": 2e-05,
+      "loss": 0.609,
+      "step": 10210
+    },
+    {
+      "epoch": 0.24304960487479768,
+      "grad_norm": 2.9139113559408916,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 10211
+    },
+    {
+      "epoch": 0.2430734075978292,
+      "grad_norm": 3.012886947668084,
+      "learning_rate": 2e-05,
+      "loss": 0.1748,
+      "step": 10212
+    },
+    {
+      "epoch": 0.2430972103208607,
+      "grad_norm": 5.180479861259828,
+      "learning_rate": 2e-05,
+      "loss": 0.6056,
+      "step": 10213
+    },
+    {
+      "epoch": 0.24312101304389222,
+      "grad_norm": 11.401464679087189,
+      "learning_rate": 2e-05,
+      "loss": 0.8546,
+      "step": 10214
+    },
+    {
+      "epoch": 0.24314481576692373,
+      "grad_norm": 3.5904346219192678,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 10215
+    },
+    {
+      "epoch": 0.24316861848995525,
+      "grad_norm": 5.495205001586632,
+      "learning_rate": 2e-05,
+      "loss": 0.3257,
+      "step": 10216
+    },
+    {
+      "epoch": 0.24319242121298676,
+      "grad_norm": 3.213428838664223,
+      "learning_rate": 2e-05,
+      "loss": 0.3177,
+      "step": 10217
+    },
+    {
+      "epoch": 0.24321622393601827,
+      "grad_norm": 5.1891267975884725,
+      "learning_rate": 2e-05,
+      "loss": 0.3342,
+      "step": 10218
+    },
+    {
+      "epoch": 0.24324002665904979,
+      "grad_norm": 19.986981773382823,
+      "learning_rate": 2e-05,
+      "loss": 0.7352,
+      "step": 10219
+    },
+    {
+      "epoch": 0.2432638293820813,
+      "grad_norm": 4.0880642023468665,
+      "learning_rate": 2e-05,
+      "loss": 0.5674,
+      "step": 10220
+    },
+    {
+      "epoch": 0.24328763210511284,
+      "grad_norm": 1.5033245934506858,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 10221
+    },
+    {
+      "epoch": 0.24331143482814435,
+      "grad_norm": 1.6928910474812309,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 10222
+    },
+    {
+      "epoch": 0.24333523755117586,
+      "grad_norm": 6.656014470790237,
+      "learning_rate": 2e-05,
+      "loss": 0.5028,
+      "step": 10223
+    },
+    {
+      "epoch": 0.24335904027420738,
+      "grad_norm": 0.6874721642677469,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 10224
+    },
+    {
+      "epoch": 0.2433828429972389,
+      "grad_norm": 2.9883300594265347,
+      "learning_rate": 2e-05,
+      "loss": 0.171,
+      "step": 10225
+    },
+    {
+      "epoch": 0.2434066457202704,
+      "grad_norm": 4.949384166466044,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 10226
+    },
+    {
+      "epoch": 0.24343044844330192,
+      "grad_norm": 6.233594662191866,
+      "learning_rate": 2e-05,
+      "loss": 0.6278,
+      "step": 10227
+    },
+    {
+      "epoch": 0.24345425116633343,
+      "grad_norm": 1.534308135951355,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 10228
+    },
+    {
+      "epoch": 0.24347805388936494,
+      "grad_norm": 2.4338757946724994,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 10229
+    },
+    {
+      "epoch": 0.24350185661239646,
+      "grad_norm": 5.271982854104065,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24352565933542797,
+      "grad_norm": 2.684163787435742,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 10231
+    },
+    {
+      "epoch": 0.24354946205845948,
+      "grad_norm": 12.88921972220241,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 10232
+    },
+    {
+      "epoch": 0.243573264781491,
+      "grad_norm": 24.766373369508187,
+      "learning_rate": 2e-05,
+      "loss": 0.4488,
+      "step": 10233
+    },
+    {
+      "epoch": 0.2435970675045225,
+      "grad_norm": 5.416545311713271,
+      "learning_rate": 2e-05,
+      "loss": 0.5319,
+      "step": 10234
+    },
+    {
+      "epoch": 0.24362087022755402,
+      "grad_norm": 5.7876753944064285,
+      "learning_rate": 2e-05,
+      "loss": 0.8257,
+      "step": 10235
+    },
+    {
+      "epoch": 0.24364467295058553,
+      "grad_norm": 1.5812486777994004,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 10236
+    },
+    {
+      "epoch": 0.24366847567361707,
+      "grad_norm": 1.2078602647039336,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 10237
+    },
+    {
+      "epoch": 0.2436922783966486,
+      "grad_norm": 20.52020200167088,
+      "learning_rate": 2e-05,
+      "loss": 1.5954,
+      "step": 10238
+    },
+    {
+      "epoch": 0.2437160811196801,
+      "grad_norm": 2.5667978456142313,
+      "learning_rate": 2e-05,
+      "loss": 0.1871,
+      "step": 10239
+    },
+    {
+      "epoch": 0.2437398838427116,
+      "grad_norm": 1.9055604298007238,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24376368656574313,
+      "grad_norm": 4.138239162875857,
+      "learning_rate": 2e-05,
+      "loss": 0.1327,
+      "step": 10241
+    },
+    {
+      "epoch": 0.24378748928877464,
+      "grad_norm": 4.60682651766084,
+      "learning_rate": 2e-05,
+      "loss": 0.1937,
+      "step": 10242
+    },
+    {
+      "epoch": 0.24381129201180615,
+      "grad_norm": 11.989510907471558,
+      "learning_rate": 2e-05,
+      "loss": 1.595,
+      "step": 10243
+    },
+    {
+      "epoch": 0.24383509473483767,
+      "grad_norm": 2.6826556404364643,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 10244
+    },
+    {
+      "epoch": 0.24385889745786918,
+      "grad_norm": 3.6923881332532402,
+      "learning_rate": 2e-05,
+      "loss": 0.6724,
+      "step": 10245
+    },
+    {
+      "epoch": 0.2438827001809007,
+      "grad_norm": 12.307209439399562,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 10246
+    },
+    {
+      "epoch": 0.2439065029039322,
+      "grad_norm": 3.688783984971714,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 10247
+    },
+    {
+      "epoch": 0.24393030562696372,
+      "grad_norm": 12.219468563212303,
+      "learning_rate": 2e-05,
+      "loss": 0.9302,
+      "step": 10248
+    },
+    {
+      "epoch": 0.24395410834999523,
+      "grad_norm": 6.350236079673995,
+      "learning_rate": 2e-05,
+      "loss": 1.0494,
+      "step": 10249
+    },
+    {
+      "epoch": 0.24397791107302674,
+      "grad_norm": 8.808363562015849,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24400171379605826,
+      "grad_norm": 8.41112192272698,
+      "learning_rate": 2e-05,
+      "loss": 1.097,
+      "step": 10251
+    },
+    {
+      "epoch": 0.2440255165190898,
+      "grad_norm": 5.4393237809596355,
+      "learning_rate": 2e-05,
+      "loss": 0.7541,
+      "step": 10252
+    },
+    {
+      "epoch": 0.2440493192421213,
+      "grad_norm": 1.7887027546668908,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 10253
+    },
+    {
+      "epoch": 0.24407312196515282,
+      "grad_norm": 4.67887546957363,
+      "learning_rate": 2e-05,
+      "loss": 0.9339,
+      "step": 10254
+    },
+    {
+      "epoch": 0.24409692468818434,
+      "grad_norm": 12.192644368127452,
+      "learning_rate": 2e-05,
+      "loss": 1.1954,
+      "step": 10255
+    },
+    {
+      "epoch": 0.24412072741121585,
+      "grad_norm": 9.072410359561925,
+      "learning_rate": 2e-05,
+      "loss": 1.1667,
+      "step": 10256
+    },
+    {
+      "epoch": 0.24414453013424736,
+      "grad_norm": 4.811643072751528,
+      "learning_rate": 2e-05,
+      "loss": 0.6071,
+      "step": 10257
+    },
+    {
+      "epoch": 0.24416833285727887,
+      "grad_norm": 4.978765708337248,
+      "learning_rate": 2e-05,
+      "loss": 0.8531,
+      "step": 10258
+    },
+    {
+      "epoch": 0.2441921355803104,
+      "grad_norm": 4.940028427724514,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 10259
+    },
+    {
+      "epoch": 0.2442159383033419,
+      "grad_norm": 1.9275170092703195,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 10260
+    },
+    {
+      "epoch": 0.2442397410263734,
+      "grad_norm": 3.327820653962899,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 10261
+    },
+    {
+      "epoch": 0.24426354374940493,
+      "grad_norm": 0.36344983474893544,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 10262
+    },
+    {
+      "epoch": 0.24428734647243644,
+      "grad_norm": 4.415407013972564,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 10263
+    },
+    {
+      "epoch": 0.24431114919546795,
+      "grad_norm": 4.730807362372209,
+      "learning_rate": 2e-05,
+      "loss": 0.6437,
+      "step": 10264
+    },
+    {
+      "epoch": 0.24433495191849947,
+      "grad_norm": 5.009310031421708,
+      "learning_rate": 2e-05,
+      "loss": 0.8732,
+      "step": 10265
+    },
+    {
+      "epoch": 0.24435875464153098,
+      "grad_norm": 1.94595985537718,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 10266
+    },
+    {
+      "epoch": 0.24438255736456252,
+      "grad_norm": 0.9710758238487133,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 10267
+    },
+    {
+      "epoch": 0.24440636008759403,
+      "grad_norm": 8.776701603963728,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 10268
+    },
+    {
+      "epoch": 0.24443016281062555,
+      "grad_norm": 3.9361072937411614,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 10269
+    },
+    {
+      "epoch": 0.24445396553365706,
+      "grad_norm": 8.005355756537663,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24447776825668857,
+      "grad_norm": 7.499854320365985,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 10271
+    },
+    {
+      "epoch": 0.24450157097972008,
+      "grad_norm": 3.7398493853823958,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 10272
+    },
+    {
+      "epoch": 0.2445253737027516,
+      "grad_norm": 2.8773054994795713,
+      "learning_rate": 2e-05,
+      "loss": 0.3369,
+      "step": 10273
+    },
+    {
+      "epoch": 0.2445491764257831,
+      "grad_norm": 8.219334753672067,
+      "learning_rate": 2e-05,
+      "loss": 1.1422,
+      "step": 10274
+    },
+    {
+      "epoch": 0.24457297914881462,
+      "grad_norm": 4.116205461681769,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 10275
+    },
+    {
+      "epoch": 0.24459678187184614,
+      "grad_norm": 2.4814975724538484,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 10276
+    },
+    {
+      "epoch": 0.24462058459487765,
+      "grad_norm": 5.74336281081906,
+      "learning_rate": 2e-05,
+      "loss": 0.3322,
+      "step": 10277
+    },
+    {
+      "epoch": 0.24464438731790916,
+      "grad_norm": 4.463732980026443,
+      "learning_rate": 2e-05,
+      "loss": 0.2876,
+      "step": 10278
+    },
+    {
+      "epoch": 0.24466819004094068,
+      "grad_norm": 4.544567915000913,
+      "learning_rate": 2e-05,
+      "loss": 0.5025,
+      "step": 10279
+    },
+    {
+      "epoch": 0.2446919927639722,
+      "grad_norm": 5.03888378479605,
+      "learning_rate": 2e-05,
+      "loss": 0.6769,
+      "step": 10280
+    },
+    {
+      "epoch": 0.2447157954870037,
+      "grad_norm": 3.2977453360081634,
+      "learning_rate": 2e-05,
+      "loss": 0.1822,
+      "step": 10281
+    },
+    {
+      "epoch": 0.24473959821003524,
+      "grad_norm": 1.2595155422992088,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 10282
+    },
+    {
+      "epoch": 0.24476340093306675,
+      "grad_norm": 7.481627333638578,
+      "learning_rate": 2e-05,
+      "loss": 0.2889,
+      "step": 10283
+    },
+    {
+      "epoch": 0.24478720365609827,
+      "grad_norm": 4.178916263107396,
+      "learning_rate": 2e-05,
+      "loss": 0.8621,
+      "step": 10284
+    },
+    {
+      "epoch": 0.24481100637912978,
+      "grad_norm": 5.074595696202635,
+      "learning_rate": 2e-05,
+      "loss": 1.3171,
+      "step": 10285
+    },
+    {
+      "epoch": 0.2448348091021613,
+      "grad_norm": 4.968472614491379,
+      "learning_rate": 2e-05,
+      "loss": 0.8485,
+      "step": 10286
+    },
+    {
+      "epoch": 0.2448586118251928,
+      "grad_norm": 4.837673736967257,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 10287
+    },
+    {
+      "epoch": 0.24488241454822432,
+      "grad_norm": 6.504664632419368,
+      "learning_rate": 2e-05,
+      "loss": 0.8237,
+      "step": 10288
+    },
+    {
+      "epoch": 0.24490621727125583,
+      "grad_norm": 12.625563446111013,
+      "learning_rate": 2e-05,
+      "loss": 0.7527,
+      "step": 10289
+    },
+    {
+      "epoch": 0.24493001999428735,
+      "grad_norm": 5.945262756979537,
+      "learning_rate": 2e-05,
+      "loss": 1.0617,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24495382271731886,
+      "grad_norm": 1.796964997736016,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 10291
+    },
+    {
+      "epoch": 0.24497762544035037,
+      "grad_norm": 3.5364554673739486,
+      "learning_rate": 2e-05,
+      "loss": 0.1585,
+      "step": 10292
+    },
+    {
+      "epoch": 0.24500142816338188,
+      "grad_norm": 2.5524974781378944,
+      "learning_rate": 2e-05,
+      "loss": 0.1804,
+      "step": 10293
+    },
+    {
+      "epoch": 0.2450252308864134,
+      "grad_norm": 5.55054805350695,
+      "learning_rate": 2e-05,
+      "loss": 0.6561,
+      "step": 10294
+    },
+    {
+      "epoch": 0.2450490336094449,
+      "grad_norm": 3.308593294818308,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 10295
+    },
+    {
+      "epoch": 0.24507283633247642,
+      "grad_norm": 3.3196760807885024,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 10296
+    },
+    {
+      "epoch": 0.24509663905550794,
+      "grad_norm": 5.175178800745277,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 10297
+    },
+    {
+      "epoch": 0.24512044177853948,
+      "grad_norm": 6.175047253582831,
+      "learning_rate": 2e-05,
+      "loss": 0.6471,
+      "step": 10298
+    },
+    {
+      "epoch": 0.245144244501571,
+      "grad_norm": 2.725421464499417,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 10299
+    },
+    {
+      "epoch": 0.2451680472246025,
+      "grad_norm": 11.342563456510703,
+      "learning_rate": 2e-05,
+      "loss": 0.6543,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24519184994763402,
+      "grad_norm": 0.9017494902057166,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 10301
+    },
+    {
+      "epoch": 0.24521565267066553,
+      "grad_norm": 2.6178376414115014,
+      "learning_rate": 2e-05,
+      "loss": 0.199,
+      "step": 10302
+    },
+    {
+      "epoch": 0.24523945539369704,
+      "grad_norm": 7.285112407424802,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 10303
+    },
+    {
+      "epoch": 0.24526325811672856,
+      "grad_norm": 4.287358835602221,
+      "learning_rate": 2e-05,
+      "loss": 0.6711,
+      "step": 10304
+    },
+    {
+      "epoch": 0.24528706083976007,
+      "grad_norm": 6.956436496702621,
+      "learning_rate": 2e-05,
+      "loss": 0.4566,
+      "step": 10305
+    },
+    {
+      "epoch": 0.24531086356279158,
+      "grad_norm": 1.6911476911607983,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 10306
+    },
+    {
+      "epoch": 0.2453346662858231,
+      "grad_norm": 2.4185621615294393,
+      "learning_rate": 2e-05,
+      "loss": 0.2324,
+      "step": 10307
+    },
+    {
+      "epoch": 0.2453584690088546,
+      "grad_norm": 15.789381049834644,
+      "learning_rate": 2e-05,
+      "loss": 0.3643,
+      "step": 10308
+    },
+    {
+      "epoch": 0.24538227173188612,
+      "grad_norm": 4.02451528613913,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 10309
+    },
+    {
+      "epoch": 0.24540607445491763,
+      "grad_norm": 5.900404103025737,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 10310
+    },
+    {
+      "epoch": 0.24542987717794915,
+      "grad_norm": 5.289157495417671,
+      "learning_rate": 2e-05,
+      "loss": 0.3091,
+      "step": 10311
+    },
+    {
+      "epoch": 0.24545367990098066,
+      "grad_norm": 2.1468987527105328,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 10312
+    },
+    {
+      "epoch": 0.2454774826240122,
+      "grad_norm": 3.067643795380961,
+      "learning_rate": 2e-05,
+      "loss": 0.179,
+      "step": 10313
+    },
+    {
+      "epoch": 0.2455012853470437,
+      "grad_norm": 2.381946013092289,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 10314
+    },
+    {
+      "epoch": 0.24552508807007523,
+      "grad_norm": 6.261069218582776,
+      "learning_rate": 2e-05,
+      "loss": 1.0949,
+      "step": 10315
+    },
+    {
+      "epoch": 0.24554889079310674,
+      "grad_norm": 2.511251422065132,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 10316
+    },
+    {
+      "epoch": 0.24557269351613825,
+      "grad_norm": 0.6617066405328617,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 10317
+    },
+    {
+      "epoch": 0.24559649623916976,
+      "grad_norm": 7.397707580879113,
+      "learning_rate": 2e-05,
+      "loss": 1.2649,
+      "step": 10318
+    },
+    {
+      "epoch": 0.24562029896220128,
+      "grad_norm": 3.5802765334082713,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 10319
+    },
+    {
+      "epoch": 0.2456441016852328,
+      "grad_norm": 0.9075685042600729,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 10320
+    },
+    {
+      "epoch": 0.2456679044082643,
+      "grad_norm": 8.059077659155568,
+      "learning_rate": 2e-05,
+      "loss": 1.6015,
+      "step": 10321
+    },
+    {
+      "epoch": 0.24569170713129582,
+      "grad_norm": 0.4354801608560839,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 10322
+    },
+    {
+      "epoch": 0.24571550985432733,
+      "grad_norm": 9.402112693820856,
+      "learning_rate": 2e-05,
+      "loss": 0.6381,
+      "step": 10323
+    },
+    {
+      "epoch": 0.24573931257735884,
+      "grad_norm": 3.7133942810246476,
+      "learning_rate": 2e-05,
+      "loss": 0.6173,
+      "step": 10324
+    },
+    {
+      "epoch": 0.24576311530039036,
+      "grad_norm": 2.982951102275487,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 10325
+    },
+    {
+      "epoch": 0.24578691802342187,
+      "grad_norm": 2.380898929763488,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 10326
+    },
+    {
+      "epoch": 0.24581072074645338,
+      "grad_norm": 8.508003270249684,
+      "learning_rate": 2e-05,
+      "loss": 0.9113,
+      "step": 10327
+    },
+    {
+      "epoch": 0.24583452346948492,
+      "grad_norm": 8.039988366963685,
+      "learning_rate": 2e-05,
+      "loss": 0.555,
+      "step": 10328
+    },
+    {
+      "epoch": 0.24585832619251644,
+      "grad_norm": 1.1429793787099345,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 10329
+    },
+    {
+      "epoch": 0.24588212891554795,
+      "grad_norm": 8.551235733289401,
+      "learning_rate": 2e-05,
+      "loss": 1.8069,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24590593163857946,
+      "grad_norm": 4.677666141350127,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 10331
+    },
+    {
+      "epoch": 0.24592973436161097,
+      "grad_norm": 2.4935449576968507,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 10332
+    },
+    {
+      "epoch": 0.2459535370846425,
+      "grad_norm": 3.0897547238342895,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 10333
+    },
+    {
+      "epoch": 0.245977339807674,
+      "grad_norm": 2.3305801913893416,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 10334
+    },
+    {
+      "epoch": 0.2460011425307055,
+      "grad_norm": 2.4337415526400066,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 10335
+    },
+    {
+      "epoch": 0.24602494525373703,
+      "grad_norm": 0.8109909619948246,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 10336
+    },
+    {
+      "epoch": 0.24604874797676854,
+      "grad_norm": 5.024254782356936,
+      "learning_rate": 2e-05,
+      "loss": 0.4285,
+      "step": 10337
+    },
+    {
+      "epoch": 0.24607255069980005,
+      "grad_norm": 4.7682289578812735,
+      "learning_rate": 2e-05,
+      "loss": 0.3227,
+      "step": 10338
+    },
+    {
+      "epoch": 0.24609635342283157,
+      "grad_norm": 9.23873938829775,
+      "learning_rate": 2e-05,
+      "loss": 0.9349,
+      "step": 10339
+    },
+    {
+      "epoch": 0.24612015614586308,
+      "grad_norm": 3.098280810546734,
+      "learning_rate": 2e-05,
+      "loss": 0.2272,
+      "step": 10340
+    },
+    {
+      "epoch": 0.2461439588688946,
+      "grad_norm": 6.6766589654105095,
+      "learning_rate": 2e-05,
+      "loss": 0.2514,
+      "step": 10341
+    },
+    {
+      "epoch": 0.2461677615919261,
+      "grad_norm": 5.866763162868542,
+      "learning_rate": 2e-05,
+      "loss": 0.7214,
+      "step": 10342
+    },
+    {
+      "epoch": 0.24619156431495764,
+      "grad_norm": 2.4580336718123728,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 10343
+    },
+    {
+      "epoch": 0.24621536703798916,
+      "grad_norm": 3.18566232420127,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 10344
+    },
+    {
+      "epoch": 0.24623916976102067,
+      "grad_norm": 2.8440287054278577,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 10345
+    },
+    {
+      "epoch": 0.24626297248405218,
+      "grad_norm": 3.7912377307611664,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 10346
+    },
+    {
+      "epoch": 0.2462867752070837,
+      "grad_norm": 3.054421651702213,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 10347
+    },
+    {
+      "epoch": 0.2463105779301152,
+      "grad_norm": 2.3775058634337114,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 10348
+    },
+    {
+      "epoch": 0.24633438065314672,
+      "grad_norm": 2.636073475153126,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 10349
+    },
+    {
+      "epoch": 0.24635818337617824,
+      "grad_norm": 10.39930963244158,
+      "learning_rate": 2e-05,
+      "loss": 0.5027,
+      "step": 10350
+    },
+    {
+      "epoch": 0.24638198609920975,
+      "grad_norm": 2.437506982674941,
+      "learning_rate": 2e-05,
+      "loss": 0.1293,
+      "step": 10351
+    },
+    {
+      "epoch": 0.24640578882224126,
+      "grad_norm": 0.5550078394728412,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 10352
+    },
+    {
+      "epoch": 0.24642959154527277,
+      "grad_norm": 3.587534515192526,
+      "learning_rate": 2e-05,
+      "loss": 0.2524,
+      "step": 10353
+    },
+    {
+      "epoch": 0.2464533942683043,
+      "grad_norm": 5.058619967538443,
+      "learning_rate": 2e-05,
+      "loss": 0.5643,
+      "step": 10354
+    },
+    {
+      "epoch": 0.2464771969913358,
+      "grad_norm": 6.534651933458562,
+      "learning_rate": 2e-05,
+      "loss": 0.6158,
+      "step": 10355
+    },
+    {
+      "epoch": 0.2465009997143673,
+      "grad_norm": 4.115913405784468,
+      "learning_rate": 2e-05,
+      "loss": 0.7211,
+      "step": 10356
+    },
+    {
+      "epoch": 0.24652480243739883,
+      "grad_norm": 7.192531912833693,
+      "learning_rate": 2e-05,
+      "loss": 0.3676,
+      "step": 10357
+    },
+    {
+      "epoch": 0.24654860516043034,
+      "grad_norm": 3.695374516376894,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 10358
+    },
+    {
+      "epoch": 0.24657240788346188,
+      "grad_norm": 4.705989114346679,
+      "learning_rate": 2e-05,
+      "loss": 0.5587,
+      "step": 10359
+    },
+    {
+      "epoch": 0.2465962106064934,
+      "grad_norm": 4.744813125033997,
+      "learning_rate": 2e-05,
+      "loss": 0.3209,
+      "step": 10360
+    },
+    {
+      "epoch": 0.2466200133295249,
+      "grad_norm": 5.484093477069281,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 10361
+    },
+    {
+      "epoch": 0.24664381605255642,
+      "grad_norm": 4.556325871416358,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 10362
+    },
+    {
+      "epoch": 0.24666761877558793,
+      "grad_norm": 2.531174520489526,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 10363
+    },
+    {
+      "epoch": 0.24669142149861945,
+      "grad_norm": 2.011523221236294,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 10364
+    },
+    {
+      "epoch": 0.24671522422165096,
+      "grad_norm": 4.6478753834257045,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 10365
+    },
+    {
+      "epoch": 0.24673902694468247,
+      "grad_norm": 6.801479507441617,
+      "learning_rate": 2e-05,
+      "loss": 0.872,
+      "step": 10366
+    },
+    {
+      "epoch": 0.24676282966771398,
+      "grad_norm": 2.950598912520247,
+      "learning_rate": 2e-05,
+      "loss": 0.1704,
+      "step": 10367
+    },
+    {
+      "epoch": 0.2467866323907455,
+      "grad_norm": 3.6098843958771623,
+      "learning_rate": 2e-05,
+      "loss": 0.1858,
+      "step": 10368
+    },
+    {
+      "epoch": 0.246810435113777,
+      "grad_norm": 4.491303087627623,
+      "learning_rate": 2e-05,
+      "loss": 0.3927,
+      "step": 10369
+    },
+    {
+      "epoch": 0.24683423783680852,
+      "grad_norm": 3.1468159689838666,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 10370
+    },
+    {
+      "epoch": 0.24685804055984004,
+      "grad_norm": 5.028428498348317,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 10371
+    },
+    {
+      "epoch": 0.24688184328287155,
+      "grad_norm": 3.5471854482693526,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 10372
+    },
+    {
+      "epoch": 0.24690564600590306,
+      "grad_norm": 3.3735406408661897,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 10373
+    },
+    {
+      "epoch": 0.2469294487289346,
+      "grad_norm": 12.81257101160543,
+      "learning_rate": 2e-05,
+      "loss": 0.9409,
+      "step": 10374
+    },
+    {
+      "epoch": 0.24695325145196612,
+      "grad_norm": 2.7944104716588627,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 10375
+    },
+    {
+      "epoch": 0.24697705417499763,
+      "grad_norm": 16.817976296028693,
+      "learning_rate": 2e-05,
+      "loss": 1.2769,
+      "step": 10376
+    },
+    {
+      "epoch": 0.24700085689802914,
+      "grad_norm": 2.3350787011561454,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 10377
+    },
+    {
+      "epoch": 0.24702465962106065,
+      "grad_norm": 26.109274311022606,
+      "learning_rate": 2e-05,
+      "loss": 0.559,
+      "step": 10378
+    },
+    {
+      "epoch": 0.24704846234409217,
+      "grad_norm": 5.802134786746928,
+      "learning_rate": 2e-05,
+      "loss": 0.7596,
+      "step": 10379
+    },
+    {
+      "epoch": 0.24707226506712368,
+      "grad_norm": 7.326693300392135,
+      "learning_rate": 2e-05,
+      "loss": 0.9228,
+      "step": 10380
+    },
+    {
+      "epoch": 0.2470960677901552,
+      "grad_norm": 6.15560774275016,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 10381
+    },
+    {
+      "epoch": 0.2471198705131867,
+      "grad_norm": 3.479101639940859,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 10382
+    },
+    {
+      "epoch": 0.24714367323621822,
+      "grad_norm": 3.1256587353399716,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 10383
+    },
+    {
+      "epoch": 0.24716747595924973,
+      "grad_norm": 2.089977280521049,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 10384
+    },
+    {
+      "epoch": 0.24719127868228125,
+      "grad_norm": 4.065102491814586,
+      "learning_rate": 2e-05,
+      "loss": 0.5097,
+      "step": 10385
+    },
+    {
+      "epoch": 0.24721508140531276,
+      "grad_norm": 2.2431680531589437,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 10386
+    },
+    {
+      "epoch": 0.24723888412834427,
+      "grad_norm": 6.373071047620311,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 10387
+    },
+    {
+      "epoch": 0.24726268685137578,
+      "grad_norm": 4.358991133903159,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 10388
+    },
+    {
+      "epoch": 0.24728648957440733,
+      "grad_norm": 2.599221545425915,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 10389
+    },
+    {
+      "epoch": 0.24731029229743884,
+      "grad_norm": 10.659114782896964,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 10390
+    },
+    {
+      "epoch": 0.24733409502047035,
+      "grad_norm": 4.207907094259197,
+      "learning_rate": 2e-05,
+      "loss": 0.5981,
+      "step": 10391
+    },
+    {
+      "epoch": 0.24735789774350186,
+      "grad_norm": 1.6127794575966858,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 10392
+    },
+    {
+      "epoch": 0.24738170046653338,
+      "grad_norm": 5.7898424961681485,
+      "learning_rate": 2e-05,
+      "loss": 0.5114,
+      "step": 10393
+    },
+    {
+      "epoch": 0.2474055031895649,
+      "grad_norm": 9.863167655674655,
+      "learning_rate": 2e-05,
+      "loss": 1.1841,
+      "step": 10394
+    },
+    {
+      "epoch": 0.2474293059125964,
+      "grad_norm": 7.770610348145945,
+      "learning_rate": 2e-05,
+      "loss": 1.1675,
+      "step": 10395
+    },
+    {
+      "epoch": 0.24745310863562792,
+      "grad_norm": 0.7111641648036776,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 10396
+    },
+    {
+      "epoch": 0.24747691135865943,
+      "grad_norm": 5.885428392580317,
+      "learning_rate": 2e-05,
+      "loss": 1.0298,
+      "step": 10397
+    },
+    {
+      "epoch": 0.24750071408169094,
+      "grad_norm": 2.714822147480616,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 10398
+    },
+    {
+      "epoch": 0.24752451680472246,
+      "grad_norm": 0.6600358159456062,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 10399
+    },
+    {
+      "epoch": 0.24754831952775397,
+      "grad_norm": 3.12592469704513,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 10400
+    },
+    {
+      "epoch": 0.24757212225078548,
+      "grad_norm": 5.200106196486366,
+      "learning_rate": 2e-05,
+      "loss": 0.7813,
+      "step": 10401
+    },
+    {
+      "epoch": 0.247595924973817,
+      "grad_norm": 4.289844137377401,
+      "learning_rate": 2e-05,
+      "loss": 0.5645,
+      "step": 10402
+    },
+    {
+      "epoch": 0.2476197276968485,
+      "grad_norm": 5.173516625751181,
+      "learning_rate": 2e-05,
+      "loss": 0.7458,
+      "step": 10403
+    },
+    {
+      "epoch": 0.24764353041988005,
+      "grad_norm": 3.6738454149205424,
+      "learning_rate": 2e-05,
+      "loss": 0.4602,
+      "step": 10404
+    },
+    {
+      "epoch": 0.24766733314291156,
+      "grad_norm": 4.72887211780693,
+      "learning_rate": 2e-05,
+      "loss": 0.8413,
+      "step": 10405
+    },
+    {
+      "epoch": 0.24769113586594307,
+      "grad_norm": 6.5918999409919605,
+      "learning_rate": 2e-05,
+      "loss": 0.2829,
+      "step": 10406
+    },
+    {
+      "epoch": 0.2477149385889746,
+      "grad_norm": 1.534799039012339,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 10407
+    },
+    {
+      "epoch": 0.2477387413120061,
+      "grad_norm": 0.49905491336939223,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 10408
+    },
+    {
+      "epoch": 0.2477625440350376,
+      "grad_norm": 6.211098175811566,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 10409
+    },
+    {
+      "epoch": 0.24778634675806913,
+      "grad_norm": 11.572927377958749,
+      "learning_rate": 2e-05,
+      "loss": 0.2101,
+      "step": 10410
+    },
+    {
+      "epoch": 0.24781014948110064,
+      "grad_norm": 1.6660643051989898,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 10411
+    },
+    {
+      "epoch": 0.24783395220413215,
+      "grad_norm": 3.3930416101612684,
+      "learning_rate": 2e-05,
+      "loss": 0.4503,
+      "step": 10412
+    },
+    {
+      "epoch": 0.24785775492716366,
+      "grad_norm": 3.9116638513546067,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 10413
+    },
+    {
+      "epoch": 0.24788155765019518,
+      "grad_norm": 5.6134833823473675,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 10414
+    },
+    {
+      "epoch": 0.2479053603732267,
+      "grad_norm": 5.869245418120961,
+      "learning_rate": 2e-05,
+      "loss": 0.604,
+      "step": 10415
+    },
+    {
+      "epoch": 0.2479291630962582,
+      "grad_norm": 0.4786209355993946,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 10416
+    },
+    {
+      "epoch": 0.24795296581928972,
+      "grad_norm": 5.690665461236819,
+      "learning_rate": 2e-05,
+      "loss": 1.0378,
+      "step": 10417
+    },
+    {
+      "epoch": 0.24797676854232123,
+      "grad_norm": 10.033921659204111,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 10418
+    },
+    {
+      "epoch": 0.24800057126535274,
+      "grad_norm": 5.334459900722541,
+      "learning_rate": 2e-05,
+      "loss": 0.6237,
+      "step": 10419
+    },
+    {
+      "epoch": 0.24802437398838428,
+      "grad_norm": 3.709317902688754,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 10420
+    },
+    {
+      "epoch": 0.2480481767114158,
+      "grad_norm": 8.112287142265496,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 10421
+    },
+    {
+      "epoch": 0.2480719794344473,
+      "grad_norm": 5.1602842100688635,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 10422
+    },
+    {
+      "epoch": 0.24809578215747882,
+      "grad_norm": 14.152119583600443,
+      "learning_rate": 2e-05,
+      "loss": 0.377,
+      "step": 10423
+    },
+    {
+      "epoch": 0.24811958488051034,
+      "grad_norm": 4.322450551327932,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 10424
+    },
+    {
+      "epoch": 0.24814338760354185,
+      "grad_norm": 12.99381777585048,
+      "learning_rate": 2e-05,
+      "loss": 0.562,
+      "step": 10425
+    },
+    {
+      "epoch": 0.24816719032657336,
+      "grad_norm": 1.8084380748217679,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 10426
+    },
+    {
+      "epoch": 0.24819099304960487,
+      "grad_norm": 8.246888291343216,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 10427
+    },
+    {
+      "epoch": 0.2482147957726364,
+      "grad_norm": 5.989211769488297,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 10428
+    },
+    {
+      "epoch": 0.2482385984956679,
+      "grad_norm": 4.615362281163148,
+      "learning_rate": 2e-05,
+      "loss": 0.5536,
+      "step": 10429
+    },
+    {
+      "epoch": 0.2482624012186994,
+      "grad_norm": 6.8780136355112065,
+      "learning_rate": 2e-05,
+      "loss": 0.6156,
+      "step": 10430
+    },
+    {
+      "epoch": 0.24828620394173093,
+      "grad_norm": 2.816986104309367,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 10431
+    },
+    {
+      "epoch": 0.24831000666476244,
+      "grad_norm": 1.689954509346569,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 10432
+    },
+    {
+      "epoch": 0.24833380938779395,
+      "grad_norm": 2.176767493205917,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 10433
+    },
+    {
+      "epoch": 0.24835761211082547,
+      "grad_norm": 3.27342799546715,
+      "learning_rate": 2e-05,
+      "loss": 0.2151,
+      "step": 10434
+    },
+    {
+      "epoch": 0.248381414833857,
+      "grad_norm": 1.3054089209638657,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 10435
+    },
+    {
+      "epoch": 0.24840521755688852,
+      "grad_norm": 6.025569496070947,
+      "learning_rate": 2e-05,
+      "loss": 0.4137,
+      "step": 10436
+    },
+    {
+      "epoch": 0.24842902027992003,
+      "grad_norm": 3.887989065876115,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 10437
+    },
+    {
+      "epoch": 0.24845282300295154,
+      "grad_norm": 3.0705530946817468,
+      "learning_rate": 2e-05,
+      "loss": 0.3485,
+      "step": 10438
+    },
+    {
+      "epoch": 0.24847662572598306,
+      "grad_norm": 2.675440030378871,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 10439
+    },
+    {
+      "epoch": 0.24850042844901457,
+      "grad_norm": 4.116216707264579,
+      "learning_rate": 2e-05,
+      "loss": 0.4668,
+      "step": 10440
+    },
+    {
+      "epoch": 0.24852423117204608,
+      "grad_norm": 13.46495296895689,
+      "learning_rate": 2e-05,
+      "loss": 0.6144,
+      "step": 10441
+    },
+    {
+      "epoch": 0.2485480338950776,
+      "grad_norm": 5.82301976852228,
+      "learning_rate": 2e-05,
+      "loss": 0.2165,
+      "step": 10442
+    },
+    {
+      "epoch": 0.2485718366181091,
+      "grad_norm": 8.788449134285363,
+      "learning_rate": 2e-05,
+      "loss": 0.3754,
+      "step": 10443
+    },
+    {
+      "epoch": 0.24859563934114062,
+      "grad_norm": 4.843057296437578,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 10444
+    },
+    {
+      "epoch": 0.24861944206417214,
+      "grad_norm": 2.9520250772866543,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 10445
+    },
+    {
+      "epoch": 0.24864324478720365,
+      "grad_norm": 6.867340839095765,
+      "learning_rate": 2e-05,
+      "loss": 0.9281,
+      "step": 10446
+    },
+    {
+      "epoch": 0.24866704751023516,
+      "grad_norm": 2.9843580093869786,
+      "learning_rate": 2e-05,
+      "loss": 0.2648,
+      "step": 10447
+    },
+    {
+      "epoch": 0.24869085023326667,
+      "grad_norm": 4.171866144759716,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 10448
+    },
+    {
+      "epoch": 0.2487146529562982,
+      "grad_norm": 4.299872998306739,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 10449
+    },
+    {
+      "epoch": 0.24873845567932973,
+      "grad_norm": 2.9812770527069383,
+      "learning_rate": 2e-05,
+      "loss": 0.2335,
+      "step": 10450
+    },
+    {
+      "epoch": 0.24876225840236124,
+      "grad_norm": 5.064920213518445,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 10451
+    },
+    {
+      "epoch": 0.24878606112539275,
+      "grad_norm": 2.5604594296106304,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 10452
+    },
+    {
+      "epoch": 0.24880986384842427,
+      "grad_norm": 4.360714017159821,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 10453
+    },
+    {
+      "epoch": 0.24883366657145578,
+      "grad_norm": 2.0916239330807023,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 10454
+    },
+    {
+      "epoch": 0.2488574692944873,
+      "grad_norm": 1.4648859425407865,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 10455
+    },
+    {
+      "epoch": 0.2488812720175188,
+      "grad_norm": 18.589345720597876,
+      "learning_rate": 2e-05,
+      "loss": 0.723,
+      "step": 10456
+    },
+    {
+      "epoch": 0.24890507474055032,
+      "grad_norm": 3.329430731091434,
+      "learning_rate": 2e-05,
+      "loss": 0.2231,
+      "step": 10457
+    },
+    {
+      "epoch": 0.24892887746358183,
+      "grad_norm": 4.271714437216502,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 10458
+    },
+    {
+      "epoch": 0.24895268018661335,
+      "grad_norm": 2.913214562012243,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 10459
+    },
+    {
+      "epoch": 0.24897648290964486,
+      "grad_norm": 4.37357215492685,
+      "learning_rate": 2e-05,
+      "loss": 0.1915,
+      "step": 10460
+    },
+    {
+      "epoch": 0.24900028563267637,
+      "grad_norm": 3.671494941518083,
+      "learning_rate": 2e-05,
+      "loss": 0.1833,
+      "step": 10461
+    },
+    {
+      "epoch": 0.24902408835570788,
+      "grad_norm": 2.6666522837988937,
+      "learning_rate": 2e-05,
+      "loss": 0.1809,
+      "step": 10462
+    },
+    {
+      "epoch": 0.2490478910787394,
+      "grad_norm": 10.141405470339278,
+      "learning_rate": 2e-05,
+      "loss": 0.9945,
+      "step": 10463
+    },
+    {
+      "epoch": 0.2490716938017709,
+      "grad_norm": 3.829172638455686,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 10464
+    },
+    {
+      "epoch": 0.24909549652480245,
+      "grad_norm": 6.996703300735046,
+      "learning_rate": 2e-05,
+      "loss": 1.322,
+      "step": 10465
+    },
+    {
+      "epoch": 0.24911929924783396,
+      "grad_norm": 3.0528458134669356,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 10466
+    },
+    {
+      "epoch": 0.24914310197086548,
+      "grad_norm": 2.98171131456197,
+      "learning_rate": 2e-05,
+      "loss": 0.1771,
+      "step": 10467
+    },
+    {
+      "epoch": 0.249166904693897,
+      "grad_norm": 5.386628292704574,
+      "learning_rate": 2e-05,
+      "loss": 0.6812,
+      "step": 10468
+    },
+    {
+      "epoch": 0.2491907074169285,
+      "grad_norm": 9.9781437980823,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 10469
+    },
+    {
+      "epoch": 0.24921451013996002,
+      "grad_norm": 27.434324333536498,
+      "learning_rate": 2e-05,
+      "loss": 1.0891,
+      "step": 10470
+    },
+    {
+      "epoch": 0.24923831286299153,
+      "grad_norm": 3.1422570863932773,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 10471
+    },
+    {
+      "epoch": 0.24926211558602304,
+      "grad_norm": 6.798982783986059,
+      "learning_rate": 2e-05,
+      "loss": 0.6781,
+      "step": 10472
+    },
+    {
+      "epoch": 0.24928591830905455,
+      "grad_norm": 3.9428912143400505,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 10473
+    },
+    {
+      "epoch": 0.24930972103208607,
+      "grad_norm": 4.253631854982826,
+      "learning_rate": 2e-05,
+      "loss": 0.6959,
+      "step": 10474
+    },
+    {
+      "epoch": 0.24933352375511758,
+      "grad_norm": 15.028834812920872,
+      "learning_rate": 2e-05,
+      "loss": 0.3567,
+      "step": 10475
+    },
+    {
+      "epoch": 0.2493573264781491,
+      "grad_norm": 5.639709620223986,
+      "learning_rate": 2e-05,
+      "loss": 0.5683,
+      "step": 10476
+    },
+    {
+      "epoch": 0.2493811292011806,
+      "grad_norm": 1.801314023306533,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 10477
+    },
+    {
+      "epoch": 0.24940493192421212,
+      "grad_norm": 5.250149214225466,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 10478
+    },
+    {
+      "epoch": 0.24942873464724363,
+      "grad_norm": 39.578108400097726,
+      "learning_rate": 2e-05,
+      "loss": 1.1842,
+      "step": 10479
+    },
+    {
+      "epoch": 0.24945253737027515,
+      "grad_norm": 6.825171863399661,
+      "learning_rate": 2e-05,
+      "loss": 0.8567,
+      "step": 10480
+    },
+    {
+      "epoch": 0.2494763400933067,
+      "grad_norm": 5.511278151269711,
+      "learning_rate": 2e-05,
+      "loss": 0.7786,
+      "step": 10481
+    },
+    {
+      "epoch": 0.2495001428163382,
+      "grad_norm": 2.1224297918877224,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 10482
+    },
+    {
+      "epoch": 0.2495239455393697,
+      "grad_norm": 10.251780320497485,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 10483
+    },
+    {
+      "epoch": 0.24954774826240123,
+      "grad_norm": 4.809443587931297,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 10484
+    },
+    {
+      "epoch": 0.24957155098543274,
+      "grad_norm": 4.512391650889383,
+      "learning_rate": 2e-05,
+      "loss": 0.8222,
+      "step": 10485
+    },
+    {
+      "epoch": 0.24959535370846425,
+      "grad_norm": 4.2886232416892565,
+      "learning_rate": 2e-05,
+      "loss": 0.6042,
+      "step": 10486
+    },
+    {
+      "epoch": 0.24961915643149576,
+      "grad_norm": 7.916859301632129,
+      "learning_rate": 2e-05,
+      "loss": 1.5588,
+      "step": 10487
+    },
+    {
+      "epoch": 0.24964295915452728,
+      "grad_norm": 5.381605369057118,
+      "learning_rate": 2e-05,
+      "loss": 0.5559,
+      "step": 10488
+    },
+    {
+      "epoch": 0.2496667618775588,
+      "grad_norm": 2.1970816580149277,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 10489
+    },
+    {
+      "epoch": 0.2496905646005903,
+      "grad_norm": 9.645973290559988,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 10490
+    },
+    {
+      "epoch": 0.24971436732362182,
+      "grad_norm": 10.026190065617842,
+      "learning_rate": 2e-05,
+      "loss": 1.5334,
+      "step": 10491
+    },
+    {
+      "epoch": 0.24973817004665333,
+      "grad_norm": 4.126824061890822,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 10492
+    },
+    {
+      "epoch": 0.24976197276968484,
+      "grad_norm": 5.743361746820272,
+      "learning_rate": 2e-05,
+      "loss": 0.8074,
+      "step": 10493
+    },
+    {
+      "epoch": 0.24978577549271636,
+      "grad_norm": 5.138095264365039,
+      "learning_rate": 2e-05,
+      "loss": 0.859,
+      "step": 10494
+    },
+    {
+      "epoch": 0.24980957821574787,
+      "grad_norm": 0.8838602819430166,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 10495
+    },
+    {
+      "epoch": 0.2498333809387794,
+      "grad_norm": 4.622137364301854,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 10496
+    },
+    {
+      "epoch": 0.24985718366181092,
+      "grad_norm": 3.8556132888433305,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 10497
+    },
+    {
+      "epoch": 0.24988098638484244,
+      "grad_norm": 16.65665332174807,
+      "learning_rate": 2e-05,
+      "loss": 0.9916,
+      "step": 10498
+    },
+    {
+      "epoch": 0.24990478910787395,
+      "grad_norm": 3.3714634895490714,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 10499
+    },
+    {
+      "epoch": 0.24992859183090546,
+      "grad_norm": 3.5699527045980313,
+      "learning_rate": 2e-05,
+      "loss": 0.1596,
+      "step": 10500
+    },
+    {
+      "epoch": 0.24995239455393697,
+      "grad_norm": 4.212937354197949,
+      "learning_rate": 2e-05,
+      "loss": 0.4956,
+      "step": 10501
+    },
+    {
+      "epoch": 0.2499761972769685,
+      "grad_norm": 4.601349174433522,
+      "learning_rate": 2e-05,
+      "loss": 0.9137,
+      "step": 10502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.5054583872149685,
+      "learning_rate": 2e-05,
+      "loss": 0.8247,
+      "step": 10503
+    },
+    {
+      "epoch": 0.25002380272303154,
+      "grad_norm": 2.351758344653818,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 10504
+    },
+    {
+      "epoch": 0.250047605446063,
+      "grad_norm": 3.6005510754429584,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 10505
+    },
+    {
+      "epoch": 0.25007140816909457,
+      "grad_norm": 1.3845461500791045,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 10506
+    },
+    {
+      "epoch": 0.25009521089212605,
+      "grad_norm": 6.549409584793541,
+      "learning_rate": 2e-05,
+      "loss": 0.7597,
+      "step": 10507
+    },
+    {
+      "epoch": 0.2501190136151576,
+      "grad_norm": 8.033982264181406,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 10508
+    },
+    {
+      "epoch": 0.2501428163381891,
+      "grad_norm": 2.4694293417578206,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 10509
+    },
+    {
+      "epoch": 0.2501666190612206,
+      "grad_norm": 3.6384992111305343,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 10510
+    },
+    {
+      "epoch": 0.2501904217842521,
+      "grad_norm": 2.073188181328625,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 10511
+    },
+    {
+      "epoch": 0.25021422450728364,
+      "grad_norm": 1.6981886483108206,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 10512
+    },
+    {
+      "epoch": 0.25023802723031513,
+      "grad_norm": 1.8098124406150857,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 10513
+    },
+    {
+      "epoch": 0.25026182995334667,
+      "grad_norm": 1.3583379821955575,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 10514
+    },
+    {
+      "epoch": 0.25028563267637816,
+      "grad_norm": 6.008252854041059,
+      "learning_rate": 2e-05,
+      "loss": 1.0102,
+      "step": 10515
+    },
+    {
+      "epoch": 0.2503094353994097,
+      "grad_norm": 2.4984866496096734,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 10516
+    },
+    {
+      "epoch": 0.2503332381224412,
+      "grad_norm": 1.5088633292510156,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 10517
+    },
+    {
+      "epoch": 0.2503570408454727,
+      "grad_norm": 5.457641673399213,
+      "learning_rate": 2e-05,
+      "loss": 0.8598,
+      "step": 10518
+    },
+    {
+      "epoch": 0.25038084356850426,
+      "grad_norm": 3.8119781090871996,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 10519
+    },
+    {
+      "epoch": 0.25040464629153575,
+      "grad_norm": 10.62861071088944,
+      "learning_rate": 2e-05,
+      "loss": 0.9597,
+      "step": 10520
+    },
+    {
+      "epoch": 0.2504284490145673,
+      "grad_norm": 1.8815155257308538,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 10521
+    },
+    {
+      "epoch": 0.2504522517375988,
+      "grad_norm": 9.14582367564608,
+      "learning_rate": 2e-05,
+      "loss": 1.1945,
+      "step": 10522
+    },
+    {
+      "epoch": 0.2504760544606303,
+      "grad_norm": 13.27592135737455,
+      "learning_rate": 2e-05,
+      "loss": 0.3235,
+      "step": 10523
+    },
+    {
+      "epoch": 0.2504998571836618,
+      "grad_norm": 3.530893467084084,
+      "learning_rate": 2e-05,
+      "loss": 0.712,
+      "step": 10524
+    },
+    {
+      "epoch": 0.25052365990669334,
+      "grad_norm": 3.3768996248089667,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 10525
+    },
+    {
+      "epoch": 0.2505474626297248,
+      "grad_norm": 3.135072970209881,
+      "learning_rate": 2e-05,
+      "loss": 0.4234,
+      "step": 10526
+    },
+    {
+      "epoch": 0.25057126535275637,
+      "grad_norm": 2.429208560746887,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 10527
+    },
+    {
+      "epoch": 0.25059506807578785,
+      "grad_norm": 4.0163682815844215,
+      "learning_rate": 2e-05,
+      "loss": 0.1861,
+      "step": 10528
+    },
+    {
+      "epoch": 0.2506188707988194,
+      "grad_norm": 16.543504122503872,
+      "learning_rate": 2e-05,
+      "loss": 0.7125,
+      "step": 10529
+    },
+    {
+      "epoch": 0.2506426735218509,
+      "grad_norm": 5.1973221035889186,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 10530
+    },
+    {
+      "epoch": 0.2506664762448824,
+      "grad_norm": 5.811529584867616,
+      "learning_rate": 2e-05,
+      "loss": 0.1578,
+      "step": 10531
+    },
+    {
+      "epoch": 0.2506902789679139,
+      "grad_norm": 2.910310045105804,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 10532
+    },
+    {
+      "epoch": 0.25071408169094545,
+      "grad_norm": 4.749519867998814,
+      "learning_rate": 2e-05,
+      "loss": 0.6708,
+      "step": 10533
+    },
+    {
+      "epoch": 0.250737884413977,
+      "grad_norm": 5.7435297774485194,
+      "learning_rate": 2e-05,
+      "loss": 1.2119,
+      "step": 10534
+    },
+    {
+      "epoch": 0.25076168713700847,
+      "grad_norm": 2.5927169991343315,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 10535
+    },
+    {
+      "epoch": 0.25078548986004,
+      "grad_norm": 3.256354201787015,
+      "learning_rate": 2e-05,
+      "loss": 0.3494,
+      "step": 10536
+    },
+    {
+      "epoch": 0.2508092925830715,
+      "grad_norm": 2.0590352434541725,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 10537
+    },
+    {
+      "epoch": 0.25083309530610304,
+      "grad_norm": 1.7791883284556338,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 10538
+    },
+    {
+      "epoch": 0.2508568980291345,
+      "grad_norm": 3.584439529702264,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 10539
+    },
+    {
+      "epoch": 0.25088070075216606,
+      "grad_norm": 3.4880662056677263,
+      "learning_rate": 2e-05,
+      "loss": 0.2488,
+      "step": 10540
+    },
+    {
+      "epoch": 0.25090450347519755,
+      "grad_norm": 1.380568294019043,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 10541
+    },
+    {
+      "epoch": 0.2509283061982291,
+      "grad_norm": 5.034814234406128,
+      "learning_rate": 2e-05,
+      "loss": 0.4436,
+      "step": 10542
+    },
+    {
+      "epoch": 0.2509521089212606,
+      "grad_norm": 4.044362666660856,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 10543
+    },
+    {
+      "epoch": 0.2509759116442921,
+      "grad_norm": 4.519626123145667,
+      "learning_rate": 2e-05,
+      "loss": 0.819,
+      "step": 10544
+    },
+    {
+      "epoch": 0.2509997143673236,
+      "grad_norm": 7.317535959417651,
+      "learning_rate": 2e-05,
+      "loss": 0.7139,
+      "step": 10545
+    },
+    {
+      "epoch": 0.25102351709035514,
+      "grad_norm": 1.6556896745836955,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 10546
+    },
+    {
+      "epoch": 0.2510473198133866,
+      "grad_norm": 2.7743897489237774,
+      "learning_rate": 2e-05,
+      "loss": 0.1756,
+      "step": 10547
+    },
+    {
+      "epoch": 0.25107112253641817,
+      "grad_norm": 2.478483099170746,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 10548
+    },
+    {
+      "epoch": 0.2510949252594497,
+      "grad_norm": 5.6185344982589145,
+      "learning_rate": 2e-05,
+      "loss": 1.2897,
+      "step": 10549
+    },
+    {
+      "epoch": 0.2511187279824812,
+      "grad_norm": 12.812852309669934,
+      "learning_rate": 2e-05,
+      "loss": 0.9212,
+      "step": 10550
+    },
+    {
+      "epoch": 0.25114253070551273,
+      "grad_norm": 6.269471068607666,
+      "learning_rate": 2e-05,
+      "loss": 1.2258,
+      "step": 10551
+    },
+    {
+      "epoch": 0.2511663334285442,
+      "grad_norm": 3.754962270618289,
+      "learning_rate": 2e-05,
+      "loss": 0.2998,
+      "step": 10552
+    },
+    {
+      "epoch": 0.25119013615157576,
+      "grad_norm": 5.592102064867023,
+      "learning_rate": 2e-05,
+      "loss": 0.8808,
+      "step": 10553
+    },
+    {
+      "epoch": 0.25121393887460725,
+      "grad_norm": 7.431935535874191,
+      "learning_rate": 2e-05,
+      "loss": 1.5909,
+      "step": 10554
+    },
+    {
+      "epoch": 0.2512377415976388,
+      "grad_norm": 4.848935359306523,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 10555
+    },
+    {
+      "epoch": 0.25126154432067027,
+      "grad_norm": 1.9064291796924968,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 10556
+    },
+    {
+      "epoch": 0.2512853470437018,
+      "grad_norm": 1.5102051540299983,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 10557
+    },
+    {
+      "epoch": 0.2513091497667333,
+      "grad_norm": 6.64901777307009,
+      "learning_rate": 2e-05,
+      "loss": 1.4974,
+      "step": 10558
+    },
+    {
+      "epoch": 0.25133295248976484,
+      "grad_norm": 1.5724513476809368,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 10559
+    },
+    {
+      "epoch": 0.2513567552127963,
+      "grad_norm": 6.488635642213852,
+      "learning_rate": 2e-05,
+      "loss": 0.3968,
+      "step": 10560
+    },
+    {
+      "epoch": 0.25138055793582786,
+      "grad_norm": 3.1079424243218843,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 10561
+    },
+    {
+      "epoch": 0.25140436065885935,
+      "grad_norm": 6.981221442332185,
+      "learning_rate": 2e-05,
+      "loss": 0.219,
+      "step": 10562
+    },
+    {
+      "epoch": 0.2514281633818909,
+      "grad_norm": 12.982225248076428,
+      "learning_rate": 2e-05,
+      "loss": 0.7333,
+      "step": 10563
+    },
+    {
+      "epoch": 0.25145196610492243,
+      "grad_norm": 1.553113061913985,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 10564
+    },
+    {
+      "epoch": 0.2514757688279539,
+      "grad_norm": 4.7384502773554695,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 10565
+    },
+    {
+      "epoch": 0.25149957155098546,
+      "grad_norm": 6.320369944334864,
+      "learning_rate": 2e-05,
+      "loss": 1.0464,
+      "step": 10566
+    },
+    {
+      "epoch": 0.25152337427401694,
+      "grad_norm": 19.148333805471246,
+      "learning_rate": 2e-05,
+      "loss": 0.365,
+      "step": 10567
+    },
+    {
+      "epoch": 0.2515471769970485,
+      "grad_norm": 6.039375733050331,
+      "learning_rate": 2e-05,
+      "loss": 1.0078,
+      "step": 10568
+    },
+    {
+      "epoch": 0.25157097972007997,
+      "grad_norm": 3.8855188074591283,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 10569
+    },
+    {
+      "epoch": 0.2515947824431115,
+      "grad_norm": 3.6048536582420625,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 10570
+    },
+    {
+      "epoch": 0.251618585166143,
+      "grad_norm": 1.1272731731973993,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 10571
+    },
+    {
+      "epoch": 0.25164238788917453,
+      "grad_norm": 4.285492111302906,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 10572
+    },
+    {
+      "epoch": 0.251666190612206,
+      "grad_norm": 3.1045931803898896,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 10573
+    },
+    {
+      "epoch": 0.25168999333523756,
+      "grad_norm": 2.6099704157457984,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 10574
+    },
+    {
+      "epoch": 0.25171379605826905,
+      "grad_norm": 4.791459407852522,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 10575
+    },
+    {
+      "epoch": 0.2517375987813006,
+      "grad_norm": 1.8016170636871165,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 10576
+    },
+    {
+      "epoch": 0.25176140150433207,
+      "grad_norm": 2.6157730386550475,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 10577
+    },
+    {
+      "epoch": 0.2517852042273636,
+      "grad_norm": 4.332895051545017,
+      "learning_rate": 2e-05,
+      "loss": 0.1806,
+      "step": 10578
+    },
+    {
+      "epoch": 0.2518090069503951,
+      "grad_norm": 2.5541399227492496,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 10579
+    },
+    {
+      "epoch": 0.25183280967342664,
+      "grad_norm": 5.706467102297336,
+      "learning_rate": 2e-05,
+      "loss": 0.4724,
+      "step": 10580
+    },
+    {
+      "epoch": 0.2518566123964582,
+      "grad_norm": 5.484223028184426,
+      "learning_rate": 2e-05,
+      "loss": 0.6234,
+      "step": 10581
+    },
+    {
+      "epoch": 0.25188041511948966,
+      "grad_norm": 1.4212204056644808,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 10582
+    },
+    {
+      "epoch": 0.2519042178425212,
+      "grad_norm": 3.2066498537708847,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 10583
+    },
+    {
+      "epoch": 0.2519280205655527,
+      "grad_norm": 3.205493152913877,
+      "learning_rate": 2e-05,
+      "loss": 0.2744,
+      "step": 10584
+    },
+    {
+      "epoch": 0.25195182328858423,
+      "grad_norm": 2.132533368120779,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 10585
+    },
+    {
+      "epoch": 0.2519756260116157,
+      "grad_norm": 5.114214353291553,
+      "learning_rate": 2e-05,
+      "loss": 0.5182,
+      "step": 10586
+    },
+    {
+      "epoch": 0.25199942873464726,
+      "grad_norm": 5.872913975400678,
+      "learning_rate": 2e-05,
+      "loss": 0.2313,
+      "step": 10587
+    },
+    {
+      "epoch": 0.25202323145767874,
+      "grad_norm": 4.614400484907146,
+      "learning_rate": 2e-05,
+      "loss": 0.7747,
+      "step": 10588
+    },
+    {
+      "epoch": 0.2520470341807103,
+      "grad_norm": 5.802783389375672,
+      "learning_rate": 2e-05,
+      "loss": 1.3775,
+      "step": 10589
+    },
+    {
+      "epoch": 0.25207083690374177,
+      "grad_norm": 4.56052924500137,
+      "learning_rate": 2e-05,
+      "loss": 0.8185,
+      "step": 10590
+    },
+    {
+      "epoch": 0.2520946396267733,
+      "grad_norm": 3.887346129943517,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 10591
+    },
+    {
+      "epoch": 0.2521184423498048,
+      "grad_norm": 20.381356772471317,
+      "learning_rate": 2e-05,
+      "loss": 0.3349,
+      "step": 10592
+    },
+    {
+      "epoch": 0.25214224507283634,
+      "grad_norm": 10.230426136047772,
+      "learning_rate": 2e-05,
+      "loss": 0.5452,
+      "step": 10593
+    },
+    {
+      "epoch": 0.2521660477958678,
+      "grad_norm": 8.301576915761277,
+      "learning_rate": 2e-05,
+      "loss": 0.8978,
+      "step": 10594
+    },
+    {
+      "epoch": 0.25218985051889936,
+      "grad_norm": 5.011414802986647,
+      "learning_rate": 2e-05,
+      "loss": 0.6822,
+      "step": 10595
+    },
+    {
+      "epoch": 0.2522136532419309,
+      "grad_norm": 2.260596819504601,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 10596
+    },
+    {
+      "epoch": 0.2522374559649624,
+      "grad_norm": 4.169135930770746,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 10597
+    },
+    {
+      "epoch": 0.25226125868799393,
+      "grad_norm": 4.49891437266781,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 10598
+    },
+    {
+      "epoch": 0.2522850614110254,
+      "grad_norm": 17.279258360420457,
+      "learning_rate": 2e-05,
+      "loss": 0.7201,
+      "step": 10599
+    },
+    {
+      "epoch": 0.25230886413405695,
+      "grad_norm": 1.1225521769213174,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 10600
+    },
+    {
+      "epoch": 0.25233266685708844,
+      "grad_norm": 10.6957696741568,
+      "learning_rate": 2e-05,
+      "loss": 0.6706,
+      "step": 10601
+    },
+    {
+      "epoch": 0.25235646958012,
+      "grad_norm": 1.9366675513321652,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 10602
+    },
+    {
+      "epoch": 0.25238027230315146,
+      "grad_norm": 4.21754976878891,
+      "learning_rate": 2e-05,
+      "loss": 0.8133,
+      "step": 10603
+    },
+    {
+      "epoch": 0.252404075026183,
+      "grad_norm": 4.346755989915675,
+      "learning_rate": 2e-05,
+      "loss": 0.2125,
+      "step": 10604
+    },
+    {
+      "epoch": 0.2524278777492145,
+      "grad_norm": 1.371380094644859,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 10605
+    },
+    {
+      "epoch": 0.25245168047224603,
+      "grad_norm": 4.958884760674826,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 10606
+    },
+    {
+      "epoch": 0.2524754831952775,
+      "grad_norm": 3.2006056974001758,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 10607
+    },
+    {
+      "epoch": 0.25249928591830906,
+      "grad_norm": 9.051294490464795,
+      "learning_rate": 2e-05,
+      "loss": 0.6862,
+      "step": 10608
+    },
+    {
+      "epoch": 0.25252308864134054,
+      "grad_norm": 4.072861761091691,
+      "learning_rate": 2e-05,
+      "loss": 0.6068,
+      "step": 10609
+    },
+    {
+      "epoch": 0.2525468913643721,
+      "grad_norm": 10.568157414177414,
+      "learning_rate": 2e-05,
+      "loss": 1.6337,
+      "step": 10610
+    },
+    {
+      "epoch": 0.2525706940874036,
+      "grad_norm": 4.631511114025346,
+      "learning_rate": 2e-05,
+      "loss": 0.7428,
+      "step": 10611
+    },
+    {
+      "epoch": 0.2525944968104351,
+      "grad_norm": 4.864146430874314,
+      "learning_rate": 2e-05,
+      "loss": 0.9204,
+      "step": 10612
+    },
+    {
+      "epoch": 0.25261829953346665,
+      "grad_norm": 6.01417276742654,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 10613
+    },
+    {
+      "epoch": 0.25264210225649814,
+      "grad_norm": 2.2080164622800718,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 10614
+    },
+    {
+      "epoch": 0.2526659049795297,
+      "grad_norm": 6.876157347283896,
+      "learning_rate": 2e-05,
+      "loss": 0.7969,
+      "step": 10615
+    },
+    {
+      "epoch": 0.25268970770256116,
+      "grad_norm": 2.7005269407671326,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 10616
+    },
+    {
+      "epoch": 0.2527135104255927,
+      "grad_norm": 4.509031345736032,
+      "learning_rate": 2e-05,
+      "loss": 0.3454,
+      "step": 10617
+    },
+    {
+      "epoch": 0.2527373131486242,
+      "grad_norm": 16.88389858500815,
+      "learning_rate": 2e-05,
+      "loss": 0.5872,
+      "step": 10618
+    },
+    {
+      "epoch": 0.25276111587165573,
+      "grad_norm": 15.473369166185972,
+      "learning_rate": 2e-05,
+      "loss": 0.6734,
+      "step": 10619
+    },
+    {
+      "epoch": 0.2527849185946872,
+      "grad_norm": 4.315332022506718,
+      "learning_rate": 2e-05,
+      "loss": 0.172,
+      "step": 10620
+    },
+    {
+      "epoch": 0.25280872131771875,
+      "grad_norm": 5.247908205188268,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 10621
+    },
+    {
+      "epoch": 0.25283252404075024,
+      "grad_norm": 4.36750748583199,
+      "learning_rate": 2e-05,
+      "loss": 0.9052,
+      "step": 10622
+    },
+    {
+      "epoch": 0.2528563267637818,
+      "grad_norm": 3.58896209846341,
+      "learning_rate": 2e-05,
+      "loss": 0.5555,
+      "step": 10623
+    },
+    {
+      "epoch": 0.25288012948681327,
+      "grad_norm": 1.446262370053672,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 10624
+    },
+    {
+      "epoch": 0.2529039322098448,
+      "grad_norm": 1.9669584321734004,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 10625
+    },
+    {
+      "epoch": 0.25292773493287635,
+      "grad_norm": 6.427418653497093,
+      "learning_rate": 2e-05,
+      "loss": 0.7994,
+      "step": 10626
+    },
+    {
+      "epoch": 0.25295153765590783,
+      "grad_norm": 1.606417263721543,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 10627
+    },
+    {
+      "epoch": 0.2529753403789394,
+      "grad_norm": 9.611171313417652,
+      "learning_rate": 2e-05,
+      "loss": 0.5313,
+      "step": 10628
+    },
+    {
+      "epoch": 0.25299914310197086,
+      "grad_norm": 6.865811258211417,
+      "learning_rate": 2e-05,
+      "loss": 0.5796,
+      "step": 10629
+    },
+    {
+      "epoch": 0.2530229458250024,
+      "grad_norm": 5.763873503750453,
+      "learning_rate": 2e-05,
+      "loss": 0.6558,
+      "step": 10630
+    },
+    {
+      "epoch": 0.2530467485480339,
+      "grad_norm": 3.8680552436407254,
+      "learning_rate": 2e-05,
+      "loss": 0.257,
+      "step": 10631
+    },
+    {
+      "epoch": 0.2530705512710654,
+      "grad_norm": 4.865312526579146,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 10632
+    },
+    {
+      "epoch": 0.2530943539940969,
+      "grad_norm": 1.4223229995987203,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 10633
+    },
+    {
+      "epoch": 0.25311815671712845,
+      "grad_norm": 1.0817882333365632,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 10634
+    },
+    {
+      "epoch": 0.25314195944015994,
+      "grad_norm": 8.319969990294,
+      "learning_rate": 2e-05,
+      "loss": 0.7402,
+      "step": 10635
+    },
+    {
+      "epoch": 0.2531657621631915,
+      "grad_norm": 2.623674305220944,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 10636
+    },
+    {
+      "epoch": 0.25318956488622296,
+      "grad_norm": 1.6898669430753903,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 10637
+    },
+    {
+      "epoch": 0.2532133676092545,
+      "grad_norm": 3.0294688812280395,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 10638
+    },
+    {
+      "epoch": 0.253237170332286,
+      "grad_norm": 3.761516434480953,
+      "learning_rate": 2e-05,
+      "loss": 0.4926,
+      "step": 10639
+    },
+    {
+      "epoch": 0.25326097305531753,
+      "grad_norm": 7.826624466225676,
+      "learning_rate": 2e-05,
+      "loss": 1.091,
+      "step": 10640
+    },
+    {
+      "epoch": 0.25328477577834907,
+      "grad_norm": 1.5196939293424179,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 10641
+    },
+    {
+      "epoch": 0.25330857850138055,
+      "grad_norm": 3.2812534252223973,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 10642
+    },
+    {
+      "epoch": 0.2533323812244121,
+      "grad_norm": 14.901219235726076,
+      "learning_rate": 2e-05,
+      "loss": 1.2364,
+      "step": 10643
+    },
+    {
+      "epoch": 0.2533561839474436,
+      "grad_norm": 5.481505408569232,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 10644
+    },
+    {
+      "epoch": 0.2533799866704751,
+      "grad_norm": 6.309671175228813,
+      "learning_rate": 2e-05,
+      "loss": 1.1209,
+      "step": 10645
+    },
+    {
+      "epoch": 0.2534037893935066,
+      "grad_norm": 0.5306087628497567,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 10646
+    },
+    {
+      "epoch": 0.25342759211653815,
+      "grad_norm": 1.9208726069762312,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 10647
+    },
+    {
+      "epoch": 0.25345139483956963,
+      "grad_norm": 1.2385114569671414,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 10648
+    },
+    {
+      "epoch": 0.2534751975626012,
+      "grad_norm": 15.399857727443344,
+      "learning_rate": 2e-05,
+      "loss": 1.3175,
+      "step": 10649
+    },
+    {
+      "epoch": 0.25349900028563266,
+      "grad_norm": 2.759787524576053,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 10650
+    },
+    {
+      "epoch": 0.2535228030086642,
+      "grad_norm": 2.0257327868513006,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 10651
+    },
+    {
+      "epoch": 0.2535466057316957,
+      "grad_norm": 8.991106588707103,
+      "learning_rate": 2e-05,
+      "loss": 0.3399,
+      "step": 10652
+    },
+    {
+      "epoch": 0.2535704084547272,
+      "grad_norm": 11.481796498210493,
+      "learning_rate": 2e-05,
+      "loss": 1.1251,
+      "step": 10653
+    },
+    {
+      "epoch": 0.2535942111777587,
+      "grad_norm": 7.604810618226702,
+      "learning_rate": 2e-05,
+      "loss": 0.8945,
+      "step": 10654
+    },
+    {
+      "epoch": 0.25361801390079025,
+      "grad_norm": 6.524244927975194,
+      "learning_rate": 2e-05,
+      "loss": 1.0796,
+      "step": 10655
+    },
+    {
+      "epoch": 0.2536418166238218,
+      "grad_norm": 4.354160208121065,
+      "learning_rate": 2e-05,
+      "loss": 0.2317,
+      "step": 10656
+    },
+    {
+      "epoch": 0.2536656193468533,
+      "grad_norm": 3.1149560535794363,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 10657
+    },
+    {
+      "epoch": 0.2536894220698848,
+      "grad_norm": 4.4165807349324355,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 10658
+    },
+    {
+      "epoch": 0.2537132247929163,
+      "grad_norm": 3.445858231504526,
+      "learning_rate": 2e-05,
+      "loss": 0.3058,
+      "step": 10659
+    },
+    {
+      "epoch": 0.25373702751594784,
+      "grad_norm": 4.810242358813462,
+      "learning_rate": 2e-05,
+      "loss": 1.2522,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25376083023897933,
+      "grad_norm": 9.348616908916702,
+      "learning_rate": 2e-05,
+      "loss": 0.6974,
+      "step": 10661
+    },
+    {
+      "epoch": 0.25378463296201087,
+      "grad_norm": 2.879995952360178,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 10662
+    },
+    {
+      "epoch": 0.25380843568504236,
+      "grad_norm": 7.062290495571006,
+      "learning_rate": 2e-05,
+      "loss": 1.0089,
+      "step": 10663
+    },
+    {
+      "epoch": 0.2538322384080739,
+      "grad_norm": 5.59447157706141,
+      "learning_rate": 2e-05,
+      "loss": 1.1343,
+      "step": 10664
+    },
+    {
+      "epoch": 0.2538560411311054,
+      "grad_norm": 5.257195010459659,
+      "learning_rate": 2e-05,
+      "loss": 0.7225,
+      "step": 10665
+    },
+    {
+      "epoch": 0.2538798438541369,
+      "grad_norm": 2.5112466274328544,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 10666
+    },
+    {
+      "epoch": 0.2539036465771684,
+      "grad_norm": 6.198919361899002,
+      "learning_rate": 2e-05,
+      "loss": 0.2777,
+      "step": 10667
+    },
+    {
+      "epoch": 0.25392744930019995,
+      "grad_norm": 3.679627576815173,
+      "learning_rate": 2e-05,
+      "loss": 0.3637,
+      "step": 10668
+    },
+    {
+      "epoch": 0.25395125202323143,
+      "grad_norm": 3.016435295716997,
+      "learning_rate": 2e-05,
+      "loss": 0.3561,
+      "step": 10669
+    },
+    {
+      "epoch": 0.253975054746263,
+      "grad_norm": 2.9773898664481235,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 10670
+    },
+    {
+      "epoch": 0.2539988574692945,
+      "grad_norm": 7.951081279767956,
+      "learning_rate": 2e-05,
+      "loss": 0.4215,
+      "step": 10671
+    },
+    {
+      "epoch": 0.254022660192326,
+      "grad_norm": 38.296953940148825,
+      "learning_rate": 2e-05,
+      "loss": 0.6432,
+      "step": 10672
+    },
+    {
+      "epoch": 0.25404646291535754,
+      "grad_norm": 2.292364818149068,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 10673
+    },
+    {
+      "epoch": 0.254070265638389,
+      "grad_norm": 7.20424453321879,
+      "learning_rate": 2e-05,
+      "loss": 0.8592,
+      "step": 10674
+    },
+    {
+      "epoch": 0.25409406836142057,
+      "grad_norm": 3.2886800576969355,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 10675
+    },
+    {
+      "epoch": 0.25411787108445205,
+      "grad_norm": 2.4445002071575836,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 10676
+    },
+    {
+      "epoch": 0.2541416738074836,
+      "grad_norm": 19.41955375728405,
+      "learning_rate": 2e-05,
+      "loss": 1.6747,
+      "step": 10677
+    },
+    {
+      "epoch": 0.2541654765305151,
+      "grad_norm": 5.223110907170522,
+      "learning_rate": 2e-05,
+      "loss": 0.3161,
+      "step": 10678
+    },
+    {
+      "epoch": 0.2541892792535466,
+      "grad_norm": 0.656514852519572,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 10679
+    },
+    {
+      "epoch": 0.2542130819765781,
+      "grad_norm": 5.4338507286757745,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 10680
+    },
+    {
+      "epoch": 0.25423688469960964,
+      "grad_norm": 5.687633939158412,
+      "learning_rate": 2e-05,
+      "loss": 0.8155,
+      "step": 10681
+    },
+    {
+      "epoch": 0.25426068742264113,
+      "grad_norm": 2.848349394289421,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 10682
+    },
+    {
+      "epoch": 0.25428449014567267,
+      "grad_norm": 3.098722550231194,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 10683
+    },
+    {
+      "epoch": 0.25430829286870416,
+      "grad_norm": 3.826964435370045,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 10684
+    },
+    {
+      "epoch": 0.2543320955917357,
+      "grad_norm": 2.020903308054092,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 10685
+    },
+    {
+      "epoch": 0.25435589831476724,
+      "grad_norm": 10.379487817276836,
+      "learning_rate": 2e-05,
+      "loss": 0.6101,
+      "step": 10686
+    },
+    {
+      "epoch": 0.2543797010377987,
+      "grad_norm": 6.955888265113414,
+      "learning_rate": 2e-05,
+      "loss": 1.2827,
+      "step": 10687
+    },
+    {
+      "epoch": 0.25440350376083026,
+      "grad_norm": 1.0886290229815039,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 10688
+    },
+    {
+      "epoch": 0.25442730648386175,
+      "grad_norm": 2.8457626338060558,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 10689
+    },
+    {
+      "epoch": 0.2544511092068933,
+      "grad_norm": 6.28134627570147,
+      "learning_rate": 2e-05,
+      "loss": 0.4381,
+      "step": 10690
+    },
+    {
+      "epoch": 0.2544749119299248,
+      "grad_norm": 5.328945399969567,
+      "learning_rate": 2e-05,
+      "loss": 0.3364,
+      "step": 10691
+    },
+    {
+      "epoch": 0.2544987146529563,
+      "grad_norm": 7.6578024182779165,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 10692
+    },
+    {
+      "epoch": 0.2545225173759878,
+      "grad_norm": 4.653492898244848,
+      "learning_rate": 2e-05,
+      "loss": 0.1494,
+      "step": 10693
+    },
+    {
+      "epoch": 0.25454632009901934,
+      "grad_norm": 3.670257129111244,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 10694
+    },
+    {
+      "epoch": 0.2545701228220508,
+      "grad_norm": 2.6927393613875923,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 10695
+    },
+    {
+      "epoch": 0.25459392554508237,
+      "grad_norm": 3.1273097639351444,
+      "learning_rate": 2e-05,
+      "loss": 0.2158,
+      "step": 10696
+    },
+    {
+      "epoch": 0.25461772826811385,
+      "grad_norm": 6.739886277495473,
+      "learning_rate": 2e-05,
+      "loss": 1.1492,
+      "step": 10697
+    },
+    {
+      "epoch": 0.2546415309911454,
+      "grad_norm": 6.480755358451073,
+      "learning_rate": 2e-05,
+      "loss": 1.0468,
+      "step": 10698
+    },
+    {
+      "epoch": 0.2546653337141769,
+      "grad_norm": 5.817175587506278,
+      "learning_rate": 2e-05,
+      "loss": 0.6666,
+      "step": 10699
+    },
+    {
+      "epoch": 0.2546891364372084,
+      "grad_norm": 7.530924484001707,
+      "learning_rate": 2e-05,
+      "loss": 0.9414,
+      "step": 10700
+    },
+    {
+      "epoch": 0.2547129391602399,
+      "grad_norm": 3.7991170658666076,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 10701
+    },
+    {
+      "epoch": 0.25473674188327144,
+      "grad_norm": 5.545033657987308,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 10702
+    },
+    {
+      "epoch": 0.254760544606303,
+      "grad_norm": 2.668088355299342,
+      "learning_rate": 2e-05,
+      "loss": 0.1235,
+      "step": 10703
+    },
+    {
+      "epoch": 0.25478434732933447,
+      "grad_norm": 1.1516929808095544,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 10704
+    },
+    {
+      "epoch": 0.254808150052366,
+      "grad_norm": 1.5965292174827,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 10705
+    },
+    {
+      "epoch": 0.2548319527753975,
+      "grad_norm": 0.6943551783825327,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 10706
+    },
+    {
+      "epoch": 0.25485575549842904,
+      "grad_norm": 8.519468483059367,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 10707
+    },
+    {
+      "epoch": 0.2548795582214605,
+      "grad_norm": 4.055229184571647,
+      "learning_rate": 2e-05,
+      "loss": 0.7123,
+      "step": 10708
+    },
+    {
+      "epoch": 0.25490336094449206,
+      "grad_norm": 4.970001598302383,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 10709
+    },
+    {
+      "epoch": 0.25492716366752355,
+      "grad_norm": 4.244139888601522,
+      "learning_rate": 2e-05,
+      "loss": 0.8227,
+      "step": 10710
+    },
+    {
+      "epoch": 0.2549509663905551,
+      "grad_norm": 10.673676800850078,
+      "learning_rate": 2e-05,
+      "loss": 1.2939,
+      "step": 10711
+    },
+    {
+      "epoch": 0.2549747691135866,
+      "grad_norm": 7.341863918491041,
+      "learning_rate": 2e-05,
+      "loss": 0.8037,
+      "step": 10712
+    },
+    {
+      "epoch": 0.2549985718366181,
+      "grad_norm": 10.008352182941053,
+      "learning_rate": 2e-05,
+      "loss": 1.3136,
+      "step": 10713
+    },
+    {
+      "epoch": 0.2550223745596496,
+      "grad_norm": 0.5274848226296066,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 10714
+    },
+    {
+      "epoch": 0.25504617728268114,
+      "grad_norm": 2.915821710629231,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 10715
+    },
+    {
+      "epoch": 0.2550699800057126,
+      "grad_norm": 2.4254903693505727,
+      "learning_rate": 2e-05,
+      "loss": 0.1249,
+      "step": 10716
+    },
+    {
+      "epoch": 0.25509378272874417,
+      "grad_norm": 1.5827036034925543,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 10717
+    },
+    {
+      "epoch": 0.2551175854517757,
+      "grad_norm": 5.462942143435428,
+      "learning_rate": 2e-05,
+      "loss": 0.5208,
+      "step": 10718
+    },
+    {
+      "epoch": 0.2551413881748072,
+      "grad_norm": 1.2730874715037142,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 10719
+    },
+    {
+      "epoch": 0.25516519089783873,
+      "grad_norm": 5.412545139149654,
+      "learning_rate": 2e-05,
+      "loss": 1.1542,
+      "step": 10720
+    },
+    {
+      "epoch": 0.2551889936208702,
+      "grad_norm": 33.85042558980414,
+      "learning_rate": 2e-05,
+      "loss": 2.0181,
+      "step": 10721
+    },
+    {
+      "epoch": 0.25521279634390176,
+      "grad_norm": 0.8697663911930821,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 10722
+    },
+    {
+      "epoch": 0.25523659906693325,
+      "grad_norm": 3.9934138437854685,
+      "learning_rate": 2e-05,
+      "loss": 0.314,
+      "step": 10723
+    },
+    {
+      "epoch": 0.2552604017899648,
+      "grad_norm": 1.035018265307137,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 10724
+    },
+    {
+      "epoch": 0.25528420451299627,
+      "grad_norm": 4.933748861085765,
+      "learning_rate": 2e-05,
+      "loss": 0.4502,
+      "step": 10725
+    },
+    {
+      "epoch": 0.2553080072360278,
+      "grad_norm": 7.654699154707427,
+      "learning_rate": 2e-05,
+      "loss": 1.0075,
+      "step": 10726
+    },
+    {
+      "epoch": 0.2553318099590593,
+      "grad_norm": 6.198207953097516,
+      "learning_rate": 2e-05,
+      "loss": 0.7443,
+      "step": 10727
+    },
+    {
+      "epoch": 0.25535561268209084,
+      "grad_norm": 15.919479110094029,
+      "learning_rate": 2e-05,
+      "loss": 1.1847,
+      "step": 10728
+    },
+    {
+      "epoch": 0.2553794154051223,
+      "grad_norm": 3.755545283001642,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 10729
+    },
+    {
+      "epoch": 0.25540321812815386,
+      "grad_norm": 6.031456718182044,
+      "learning_rate": 2e-05,
+      "loss": 0.253,
+      "step": 10730
+    },
+    {
+      "epoch": 0.25542702085118535,
+      "grad_norm": 17.28690130811098,
+      "learning_rate": 2e-05,
+      "loss": 1.0745,
+      "step": 10731
+    },
+    {
+      "epoch": 0.2554508235742169,
+      "grad_norm": 3.5517344872920655,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 10732
+    },
+    {
+      "epoch": 0.25547462629724843,
+      "grad_norm": 1.0217313952706708,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 10733
+    },
+    {
+      "epoch": 0.2554984290202799,
+      "grad_norm": 0.4496341240488648,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 10734
+    },
+    {
+      "epoch": 0.25552223174331146,
+      "grad_norm": 7.322662325875406,
+      "learning_rate": 2e-05,
+      "loss": 0.9625,
+      "step": 10735
+    },
+    {
+      "epoch": 0.25554603446634294,
+      "grad_norm": 3.4798289472045063,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 10736
+    },
+    {
+      "epoch": 0.2555698371893745,
+      "grad_norm": 4.200608560654516,
+      "learning_rate": 2e-05,
+      "loss": 0.5101,
+      "step": 10737
+    },
+    {
+      "epoch": 0.25559363991240597,
+      "grad_norm": 42.183887661455906,
+      "learning_rate": 2e-05,
+      "loss": 1.2998,
+      "step": 10738
+    },
+    {
+      "epoch": 0.2556174426354375,
+      "grad_norm": 0.6787216997946146,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 10739
+    },
+    {
+      "epoch": 0.255641245358469,
+      "grad_norm": 2.146140704933912,
+      "learning_rate": 2e-05,
+      "loss": 0.133,
+      "step": 10740
+    },
+    {
+      "epoch": 0.25566504808150053,
+      "grad_norm": 2.7967982554785635,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 10741
+    },
+    {
+      "epoch": 0.255688850804532,
+      "grad_norm": 3.218719955227931,
+      "learning_rate": 2e-05,
+      "loss": 0.1781,
+      "step": 10742
+    },
+    {
+      "epoch": 0.25571265352756356,
+      "grad_norm": 5.767286702872603,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 10743
+    },
+    {
+      "epoch": 0.25573645625059505,
+      "grad_norm": 5.900251797368314,
+      "learning_rate": 2e-05,
+      "loss": 1.0079,
+      "step": 10744
+    },
+    {
+      "epoch": 0.2557602589736266,
+      "grad_norm": 41.78533759342911,
+      "learning_rate": 2e-05,
+      "loss": 0.5795,
+      "step": 10745
+    },
+    {
+      "epoch": 0.25578406169665807,
+      "grad_norm": 5.92540441975684,
+      "learning_rate": 2e-05,
+      "loss": 0.7156,
+      "step": 10746
+    },
+    {
+      "epoch": 0.2558078644196896,
+      "grad_norm": 4.233718574338577,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 10747
+    },
+    {
+      "epoch": 0.25583166714272115,
+      "grad_norm": 5.145297623150496,
+      "learning_rate": 2e-05,
+      "loss": 0.5446,
+      "step": 10748
+    },
+    {
+      "epoch": 0.25585546986575264,
+      "grad_norm": 3.2134216123711585,
+      "learning_rate": 2e-05,
+      "loss": 0.2993,
+      "step": 10749
+    },
+    {
+      "epoch": 0.2558792725887842,
+      "grad_norm": 1.8590695356982259,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25590307531181566,
+      "grad_norm": 3.6448810847669146,
+      "learning_rate": 2e-05,
+      "loss": 0.637,
+      "step": 10751
+    },
+    {
+      "epoch": 0.2559268780348472,
+      "grad_norm": 6.376113123567774,
+      "learning_rate": 2e-05,
+      "loss": 1.3054,
+      "step": 10752
+    },
+    {
+      "epoch": 0.2559506807578787,
+      "grad_norm": 9.613272061415348,
+      "learning_rate": 2e-05,
+      "loss": 1.0057,
+      "step": 10753
+    },
+    {
+      "epoch": 0.25597448348091023,
+      "grad_norm": 3.826112032156432,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 10754
+    },
+    {
+      "epoch": 0.2559982862039417,
+      "grad_norm": 3.1097827148830692,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 10755
+    },
+    {
+      "epoch": 0.25602208892697326,
+      "grad_norm": 3.421407845746706,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 10756
+    },
+    {
+      "epoch": 0.25604589165000474,
+      "grad_norm": 2.3730807290295837,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 10757
+    },
+    {
+      "epoch": 0.2560696943730363,
+      "grad_norm": 2.235615592794272,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 10758
+    },
+    {
+      "epoch": 0.25609349709606777,
+      "grad_norm": 3.855804141072742,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 10759
+    },
+    {
+      "epoch": 0.2561172998190993,
+      "grad_norm": 4.386581422249787,
+      "learning_rate": 2e-05,
+      "loss": 0.1794,
+      "step": 10760
+    },
+    {
+      "epoch": 0.2561411025421308,
+      "grad_norm": 2.919010991527941,
+      "learning_rate": 2e-05,
+      "loss": 0.3482,
+      "step": 10761
+    },
+    {
+      "epoch": 0.25616490526516233,
+      "grad_norm": 6.563007586193031,
+      "learning_rate": 2e-05,
+      "loss": 0.2084,
+      "step": 10762
+    },
+    {
+      "epoch": 0.2561887079881939,
+      "grad_norm": 7.907211585325972,
+      "learning_rate": 2e-05,
+      "loss": 0.3324,
+      "step": 10763
+    },
+    {
+      "epoch": 0.25621251071122536,
+      "grad_norm": 6.940826805950208,
+      "learning_rate": 2e-05,
+      "loss": 0.291,
+      "step": 10764
+    },
+    {
+      "epoch": 0.2562363134342569,
+      "grad_norm": 2.04560391868525,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 10765
+    },
+    {
+      "epoch": 0.2562601161572884,
+      "grad_norm": 8.990543071088908,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 10766
+    },
+    {
+      "epoch": 0.2562839188803199,
+      "grad_norm": 2.5765311427184536,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 10767
+    },
+    {
+      "epoch": 0.2563077216033514,
+      "grad_norm": 2.510908249561724,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 10768
+    },
+    {
+      "epoch": 0.25633152432638295,
+      "grad_norm": 6.887302125868229,
+      "learning_rate": 2e-05,
+      "loss": 0.8579,
+      "step": 10769
+    },
+    {
+      "epoch": 0.25635532704941444,
+      "grad_norm": 2.316650823624416,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 10770
+    },
+    {
+      "epoch": 0.256379129772446,
+      "grad_norm": 1.9961656316057355,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 10771
+    },
+    {
+      "epoch": 0.25640293249547746,
+      "grad_norm": 13.136942356505667,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 10772
+    },
+    {
+      "epoch": 0.256426735218509,
+      "grad_norm": 28.719518113555814,
+      "learning_rate": 2e-05,
+      "loss": 1.6845,
+      "step": 10773
+    },
+    {
+      "epoch": 0.2564505379415405,
+      "grad_norm": 6.859382679452156,
+      "learning_rate": 2e-05,
+      "loss": 0.9111,
+      "step": 10774
+    },
+    {
+      "epoch": 0.25647434066457203,
+      "grad_norm": 4.263094662046845,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 10775
+    },
+    {
+      "epoch": 0.2564981433876035,
+      "grad_norm": 3.4367746623683866,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 10776
+    },
+    {
+      "epoch": 0.25652194611063506,
+      "grad_norm": 8.908319820354944,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 10777
+    },
+    {
+      "epoch": 0.2565457488336666,
+      "grad_norm": 5.286831590135081,
+      "learning_rate": 2e-05,
+      "loss": 0.3532,
+      "step": 10778
+    },
+    {
+      "epoch": 0.2565695515566981,
+      "grad_norm": 6.038260353710542,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 10779
+    },
+    {
+      "epoch": 0.2565933542797296,
+      "grad_norm": 2.429918433670545,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 10780
+    },
+    {
+      "epoch": 0.2566171570027611,
+      "grad_norm": 6.316805912796239,
+      "learning_rate": 2e-05,
+      "loss": 0.9219,
+      "step": 10781
+    },
+    {
+      "epoch": 0.25664095972579265,
+      "grad_norm": 5.977444310643307,
+      "learning_rate": 2e-05,
+      "loss": 0.8267,
+      "step": 10782
+    },
+    {
+      "epoch": 0.25666476244882414,
+      "grad_norm": 4.602179545459554,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 10783
+    },
+    {
+      "epoch": 0.2566885651718557,
+      "grad_norm": 2.293506110468444,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 10784
+    },
+    {
+      "epoch": 0.25671236789488716,
+      "grad_norm": 3.005659083282059,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 10785
+    },
+    {
+      "epoch": 0.2567361706179187,
+      "grad_norm": 2.2712489961093727,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 10786
+    },
+    {
+      "epoch": 0.2567599733409502,
+      "grad_norm": 2.1333504666745697,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 10787
+    },
+    {
+      "epoch": 0.25678377606398173,
+      "grad_norm": 3.482449384379317,
+      "learning_rate": 2e-05,
+      "loss": 0.2442,
+      "step": 10788
+    },
+    {
+      "epoch": 0.2568075787870132,
+      "grad_norm": 3.323087591255585,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 10789
+    },
+    {
+      "epoch": 0.25683138151004475,
+      "grad_norm": 3.256899365007207,
+      "learning_rate": 2e-05,
+      "loss": 0.1273,
+      "step": 10790
+    },
+    {
+      "epoch": 0.25685518423307624,
+      "grad_norm": 3.420997288062616,
+      "learning_rate": 2e-05,
+      "loss": 0.4224,
+      "step": 10791
+    },
+    {
+      "epoch": 0.2568789869561078,
+      "grad_norm": 13.363282454858929,
+      "learning_rate": 2e-05,
+      "loss": 1.7068,
+      "step": 10792
+    },
+    {
+      "epoch": 0.2569027896791393,
+      "grad_norm": 7.164155826356301,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 10793
+    },
+    {
+      "epoch": 0.2569265924021708,
+      "grad_norm": 2.3346823480820196,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 10794
+    },
+    {
+      "epoch": 0.25695039512520235,
+      "grad_norm": 1.8737142534848168,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 10795
+    },
+    {
+      "epoch": 0.25697419784823383,
+      "grad_norm": 2.451670173260494,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 10796
+    },
+    {
+      "epoch": 0.2569980005712654,
+      "grad_norm": 2.853030409327408,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 10797
+    },
+    {
+      "epoch": 0.25702180329429686,
+      "grad_norm": 2.8063461042436617,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 10798
+    },
+    {
+      "epoch": 0.2570456060173284,
+      "grad_norm": 3.747888511491805,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 10799
+    },
+    {
+      "epoch": 0.2570694087403599,
+      "grad_norm": 4.595880812549756,
+      "learning_rate": 2e-05,
+      "loss": 0.5813,
+      "step": 10800
+    },
+    {
+      "epoch": 0.2570932114633914,
+      "grad_norm": 5.13190397498083,
+      "learning_rate": 2e-05,
+      "loss": 0.9994,
+      "step": 10801
+    },
+    {
+      "epoch": 0.2571170141864229,
+      "grad_norm": 1.9696385574740432,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 10802
+    },
+    {
+      "epoch": 0.25714081690945445,
+      "grad_norm": 10.734839819190762,
+      "learning_rate": 2e-05,
+      "loss": 0.2685,
+      "step": 10803
+    },
+    {
+      "epoch": 0.25716461963248594,
+      "grad_norm": 14.12948232761925,
+      "learning_rate": 2e-05,
+      "loss": 1.1215,
+      "step": 10804
+    },
+    {
+      "epoch": 0.2571884223555175,
+      "grad_norm": 2.7167986348073376,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 10805
+    },
+    {
+      "epoch": 0.25721222507854896,
+      "grad_norm": 7.697323703717234,
+      "learning_rate": 2e-05,
+      "loss": 1.2048,
+      "step": 10806
+    },
+    {
+      "epoch": 0.2572360278015805,
+      "grad_norm": 1.0071196693191464,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 10807
+    },
+    {
+      "epoch": 0.25725983052461204,
+      "grad_norm": 3.0096545370652676,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 10808
+    },
+    {
+      "epoch": 0.25728363324764353,
+      "grad_norm": 0.5157159764680325,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 10809
+    },
+    {
+      "epoch": 0.25730743597067507,
+      "grad_norm": 0.9114763820401414,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 10810
+    },
+    {
+      "epoch": 0.25733123869370655,
+      "grad_norm": 6.834991010656435,
+      "learning_rate": 2e-05,
+      "loss": 0.2094,
+      "step": 10811
+    },
+    {
+      "epoch": 0.2573550414167381,
+      "grad_norm": 8.91502356698724,
+      "learning_rate": 2e-05,
+      "loss": 1.2544,
+      "step": 10812
+    },
+    {
+      "epoch": 0.2573788441397696,
+      "grad_norm": 1.904599502824695,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 10813
+    },
+    {
+      "epoch": 0.2574026468628011,
+      "grad_norm": 30.207694785205916,
+      "learning_rate": 2e-05,
+      "loss": 0.8364,
+      "step": 10814
+    },
+    {
+      "epoch": 0.2574264495858326,
+      "grad_norm": 38.48483972628477,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 10815
+    },
+    {
+      "epoch": 0.25745025230886415,
+      "grad_norm": 2.2584733358759497,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 10816
+    },
+    {
+      "epoch": 0.25747405503189563,
+      "grad_norm": 11.273352519687167,
+      "learning_rate": 2e-05,
+      "loss": 1.073,
+      "step": 10817
+    },
+    {
+      "epoch": 0.2574978577549272,
+      "grad_norm": 3.4929462772589397,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 10818
+    },
+    {
+      "epoch": 0.25752166047795866,
+      "grad_norm": 8.656236127738335,
+      "learning_rate": 2e-05,
+      "loss": 0.3665,
+      "step": 10819
+    },
+    {
+      "epoch": 0.2575454632009902,
+      "grad_norm": 10.197896360247004,
+      "learning_rate": 2e-05,
+      "loss": 0.5332,
+      "step": 10820
+    },
+    {
+      "epoch": 0.2575692659240217,
+      "grad_norm": 4.837331031753431,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 10821
+    },
+    {
+      "epoch": 0.2575930686470532,
+      "grad_norm": 13.8447994408699,
+      "learning_rate": 2e-05,
+      "loss": 0.7206,
+      "step": 10822
+    },
+    {
+      "epoch": 0.2576168713700847,
+      "grad_norm": 10.744282920737165,
+      "learning_rate": 2e-05,
+      "loss": 1.1096,
+      "step": 10823
+    },
+    {
+      "epoch": 0.25764067409311625,
+      "grad_norm": 4.814931616733441,
+      "learning_rate": 2e-05,
+      "loss": 0.2664,
+      "step": 10824
+    },
+    {
+      "epoch": 0.2576644768161478,
+      "grad_norm": 7.311379145037241,
+      "learning_rate": 2e-05,
+      "loss": 1.1462,
+      "step": 10825
+    },
+    {
+      "epoch": 0.2576882795391793,
+      "grad_norm": 3.148117006499398,
+      "learning_rate": 2e-05,
+      "loss": 0.309,
+      "step": 10826
+    },
+    {
+      "epoch": 0.2577120822622108,
+      "grad_norm": 8.108911350214285,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 10827
+    },
+    {
+      "epoch": 0.2577358849852423,
+      "grad_norm": 6.198908733983822,
+      "learning_rate": 2e-05,
+      "loss": 0.6714,
+      "step": 10828
+    },
+    {
+      "epoch": 0.25775968770827384,
+      "grad_norm": 8.153824345251513,
+      "learning_rate": 2e-05,
+      "loss": 0.9802,
+      "step": 10829
+    },
+    {
+      "epoch": 0.25778349043130533,
+      "grad_norm": 15.50491264239206,
+      "learning_rate": 2e-05,
+      "loss": 0.668,
+      "step": 10830
+    },
+    {
+      "epoch": 0.25780729315433687,
+      "grad_norm": 12.225475199515671,
+      "learning_rate": 2e-05,
+      "loss": 0.713,
+      "step": 10831
+    },
+    {
+      "epoch": 0.25783109587736835,
+      "grad_norm": 2.8497511572866205,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 10832
+    },
+    {
+      "epoch": 0.2578548986003999,
+      "grad_norm": 8.166367965398356,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 10833
+    },
+    {
+      "epoch": 0.2578787013234314,
+      "grad_norm": 2.9312530300593687,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 10834
+    },
+    {
+      "epoch": 0.2579025040464629,
+      "grad_norm": 4.415157956163808,
+      "learning_rate": 2e-05,
+      "loss": 0.6538,
+      "step": 10835
+    },
+    {
+      "epoch": 0.2579263067694944,
+      "grad_norm": 1.4666767454084955,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 10836
+    },
+    {
+      "epoch": 0.25795010949252595,
+      "grad_norm": 1.2191776443287723,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 10837
+    },
+    {
+      "epoch": 0.25797391221555743,
+      "grad_norm": 4.103690735301718,
+      "learning_rate": 2e-05,
+      "loss": 0.7375,
+      "step": 10838
+    },
+    {
+      "epoch": 0.257997714938589,
+      "grad_norm": 2.0119327602196755,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 10839
+    },
+    {
+      "epoch": 0.2580215176616205,
+      "grad_norm": 19.908623796103083,
+      "learning_rate": 2e-05,
+      "loss": 1.2283,
+      "step": 10840
+    },
+    {
+      "epoch": 0.258045320384652,
+      "grad_norm": 19.981799383686578,
+      "learning_rate": 2e-05,
+      "loss": 0.9719,
+      "step": 10841
+    },
+    {
+      "epoch": 0.25806912310768354,
+      "grad_norm": 6.8814394295238985,
+      "learning_rate": 2e-05,
+      "loss": 0.9693,
+      "step": 10842
+    },
+    {
+      "epoch": 0.258092925830715,
+      "grad_norm": 2.7488175802161527,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 10843
+    },
+    {
+      "epoch": 0.25811672855374657,
+      "grad_norm": 11.874050827413413,
+      "learning_rate": 2e-05,
+      "loss": 0.1864,
+      "step": 10844
+    },
+    {
+      "epoch": 0.25814053127677805,
+      "grad_norm": 2.1576194956749863,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 10845
+    },
+    {
+      "epoch": 0.2581643339998096,
+      "grad_norm": 8.337649489384187,
+      "learning_rate": 2e-05,
+      "loss": 1.1064,
+      "step": 10846
+    },
+    {
+      "epoch": 0.2581881367228411,
+      "grad_norm": 7.325151897102461,
+      "learning_rate": 2e-05,
+      "loss": 0.8105,
+      "step": 10847
+    },
+    {
+      "epoch": 0.2582119394458726,
+      "grad_norm": 2.6900789641060325,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 10848
+    },
+    {
+      "epoch": 0.2582357421689041,
+      "grad_norm": 4.151359604792979,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 10849
+    },
+    {
+      "epoch": 0.25825954489193564,
+      "grad_norm": 3.9469385939125954,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 10850
+    },
+    {
+      "epoch": 0.25828334761496713,
+      "grad_norm": 1.3945803204729936,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 10851
+    },
+    {
+      "epoch": 0.25830715033799867,
+      "grad_norm": 4.849828364066134,
+      "learning_rate": 2e-05,
+      "loss": 0.1828,
+      "step": 10852
+    },
+    {
+      "epoch": 0.25833095306103016,
+      "grad_norm": 3.6997262027506395,
+      "learning_rate": 2e-05,
+      "loss": 0.1519,
+      "step": 10853
+    },
+    {
+      "epoch": 0.2583547557840617,
+      "grad_norm": 3.472985773236161,
+      "learning_rate": 2e-05,
+      "loss": 0.1773,
+      "step": 10854
+    },
+    {
+      "epoch": 0.25837855850709324,
+      "grad_norm": 6.405258542260222,
+      "learning_rate": 2e-05,
+      "loss": 0.7806,
+      "step": 10855
+    },
+    {
+      "epoch": 0.2584023612301247,
+      "grad_norm": 3.1702570936503154,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 10856
+    },
+    {
+      "epoch": 0.25842616395315626,
+      "grad_norm": 3.486228124033281,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 10857
+    },
+    {
+      "epoch": 0.25844996667618775,
+      "grad_norm": 1.765325157724857,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 10858
+    },
+    {
+      "epoch": 0.2584737693992193,
+      "grad_norm": 1.4373431477546956,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 10859
+    },
+    {
+      "epoch": 0.2584975721222508,
+      "grad_norm": 2.771362908967284,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 10860
+    },
+    {
+      "epoch": 0.2585213748452823,
+      "grad_norm": 7.385843076079911,
+      "learning_rate": 2e-05,
+      "loss": 0.4881,
+      "step": 10861
+    },
+    {
+      "epoch": 0.2585451775683138,
+      "grad_norm": 1.234073458407282,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 10862
+    },
+    {
+      "epoch": 0.25856898029134534,
+      "grad_norm": 6.680473379241099,
+      "learning_rate": 2e-05,
+      "loss": 0.6774,
+      "step": 10863
+    },
+    {
+      "epoch": 0.2585927830143768,
+      "grad_norm": 4.50449539583326,
+      "learning_rate": 2e-05,
+      "loss": 0.3563,
+      "step": 10864
+    },
+    {
+      "epoch": 0.25861658573740837,
+      "grad_norm": 5.282279724937187,
+      "learning_rate": 2e-05,
+      "loss": 0.3624,
+      "step": 10865
+    },
+    {
+      "epoch": 0.25864038846043985,
+      "grad_norm": 6.057741307293732,
+      "learning_rate": 2e-05,
+      "loss": 0.8888,
+      "step": 10866
+    },
+    {
+      "epoch": 0.2586641911834714,
+      "grad_norm": 0.660165018439155,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 10867
+    },
+    {
+      "epoch": 0.2586879939065029,
+      "grad_norm": 3.009723817190735,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 10868
+    },
+    {
+      "epoch": 0.2587117966295344,
+      "grad_norm": 0.7039194939530553,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 10869
+    },
+    {
+      "epoch": 0.25873559935256596,
+      "grad_norm": 5.343090316046503,
+      "learning_rate": 2e-05,
+      "loss": 0.53,
+      "step": 10870
+    },
+    {
+      "epoch": 0.25875940207559744,
+      "grad_norm": 5.273561046153556,
+      "learning_rate": 2e-05,
+      "loss": 0.926,
+      "step": 10871
+    },
+    {
+      "epoch": 0.258783204798629,
+      "grad_norm": 1.6484064746650793,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 10872
+    },
+    {
+      "epoch": 0.25880700752166047,
+      "grad_norm": 15.77875867549669,
+      "learning_rate": 2e-05,
+      "loss": 0.5156,
+      "step": 10873
+    },
+    {
+      "epoch": 0.258830810244692,
+      "grad_norm": 3.1261461979726803,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 10874
+    },
+    {
+      "epoch": 0.2588546129677235,
+      "grad_norm": 2.4272630149965853,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 10875
+    },
+    {
+      "epoch": 0.25887841569075504,
+      "grad_norm": 21.966129249236285,
+      "learning_rate": 2e-05,
+      "loss": 1.4194,
+      "step": 10876
+    },
+    {
+      "epoch": 0.2589022184137865,
+      "grad_norm": 1.3249477479959404,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 10877
+    },
+    {
+      "epoch": 0.25892602113681806,
+      "grad_norm": 6.118910641149433,
+      "learning_rate": 2e-05,
+      "loss": 0.1956,
+      "step": 10878
+    },
+    {
+      "epoch": 0.25894982385984955,
+      "grad_norm": 6.577765124838205,
+      "learning_rate": 2e-05,
+      "loss": 1.2902,
+      "step": 10879
+    },
+    {
+      "epoch": 0.2589736265828811,
+      "grad_norm": 2.3105398810047078,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 10880
+    },
+    {
+      "epoch": 0.2589974293059126,
+      "grad_norm": 4.546006059264495,
+      "learning_rate": 2e-05,
+      "loss": 0.3108,
+      "step": 10881
+    },
+    {
+      "epoch": 0.2590212320289441,
+      "grad_norm": 8.432267016168348,
+      "learning_rate": 2e-05,
+      "loss": 1.334,
+      "step": 10882
+    },
+    {
+      "epoch": 0.2590450347519756,
+      "grad_norm": 3.256864824614797,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 10883
+    },
+    {
+      "epoch": 0.25906883747500714,
+      "grad_norm": 5.561214037816271,
+      "learning_rate": 2e-05,
+      "loss": 0.6946,
+      "step": 10884
+    },
+    {
+      "epoch": 0.2590926401980387,
+      "grad_norm": 12.053637605176514,
+      "learning_rate": 2e-05,
+      "loss": 1.1958,
+      "step": 10885
+    },
+    {
+      "epoch": 0.25911644292107017,
+      "grad_norm": 7.989931893596226,
+      "learning_rate": 2e-05,
+      "loss": 1.7521,
+      "step": 10886
+    },
+    {
+      "epoch": 0.2591402456441017,
+      "grad_norm": 5.195367218678853,
+      "learning_rate": 2e-05,
+      "loss": 0.3529,
+      "step": 10887
+    },
+    {
+      "epoch": 0.2591640483671332,
+      "grad_norm": 5.994179049234054,
+      "learning_rate": 2e-05,
+      "loss": 1.0532,
+      "step": 10888
+    },
+    {
+      "epoch": 0.25918785109016473,
+      "grad_norm": 0.4001258935457338,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 10889
+    },
+    {
+      "epoch": 0.2592116538131962,
+      "grad_norm": 9.356820398169647,
+      "learning_rate": 2e-05,
+      "loss": 1.1185,
+      "step": 10890
+    },
+    {
+      "epoch": 0.25923545653622776,
+      "grad_norm": 8.206626738957759,
+      "learning_rate": 2e-05,
+      "loss": 1.2322,
+      "step": 10891
+    },
+    {
+      "epoch": 0.25925925925925924,
+      "grad_norm": 2.247508252907484,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 10892
+    },
+    {
+      "epoch": 0.2592830619822908,
+      "grad_norm": 7.290686657036747,
+      "learning_rate": 2e-05,
+      "loss": 1.1774,
+      "step": 10893
+    },
+    {
+      "epoch": 0.25930686470532227,
+      "grad_norm": 2.7123519199966846,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 10894
+    },
+    {
+      "epoch": 0.2593306674283538,
+      "grad_norm": 3.897211900749154,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 10895
+    },
+    {
+      "epoch": 0.2593544701513853,
+      "grad_norm": 1.5565806250783523,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 10896
+    },
+    {
+      "epoch": 0.25937827287441684,
+      "grad_norm": 13.293339110406707,
+      "learning_rate": 2e-05,
+      "loss": 0.6543,
+      "step": 10897
+    },
+    {
+      "epoch": 0.2594020755974483,
+      "grad_norm": 4.243532819578403,
+      "learning_rate": 2e-05,
+      "loss": 0.7458,
+      "step": 10898
+    },
+    {
+      "epoch": 0.25942587832047986,
+      "grad_norm": 2.6945401813282963,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 10899
+    },
+    {
+      "epoch": 0.2594496810435114,
+      "grad_norm": 2.1925071130381673,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 10900
+    },
+    {
+      "epoch": 0.2594734837665429,
+      "grad_norm": 6.697605562549842,
+      "learning_rate": 2e-05,
+      "loss": 0.2413,
+      "step": 10901
+    },
+    {
+      "epoch": 0.25949728648957443,
+      "grad_norm": 7.363778277132294,
+      "learning_rate": 2e-05,
+      "loss": 0.6127,
+      "step": 10902
+    },
+    {
+      "epoch": 0.2595210892126059,
+      "grad_norm": 7.0338083052503935,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 10903
+    },
+    {
+      "epoch": 0.25954489193563746,
+      "grad_norm": 3.895117405980226,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 10904
+    },
+    {
+      "epoch": 0.25956869465866894,
+      "grad_norm": 1.599094102944286,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 10905
+    },
+    {
+      "epoch": 0.2595924973817005,
+      "grad_norm": 4.126283017316913,
+      "learning_rate": 2e-05,
+      "loss": 0.7639,
+      "step": 10906
+    },
+    {
+      "epoch": 0.25961630010473197,
+      "grad_norm": 5.133385769086765,
+      "learning_rate": 2e-05,
+      "loss": 0.3378,
+      "step": 10907
+    },
+    {
+      "epoch": 0.2596401028277635,
+      "grad_norm": 11.434939022388837,
+      "learning_rate": 2e-05,
+      "loss": 0.5171,
+      "step": 10908
+    },
+    {
+      "epoch": 0.259663905550795,
+      "grad_norm": 4.380002731678323,
+      "learning_rate": 2e-05,
+      "loss": 0.5714,
+      "step": 10909
+    },
+    {
+      "epoch": 0.25968770827382653,
+      "grad_norm": 2.020149447809107,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 10910
+    },
+    {
+      "epoch": 0.259711510996858,
+      "grad_norm": 9.667115063411511,
+      "learning_rate": 2e-05,
+      "loss": 0.588,
+      "step": 10911
+    },
+    {
+      "epoch": 0.25973531371988956,
+      "grad_norm": 6.77931692173266,
+      "learning_rate": 2e-05,
+      "loss": 1.4083,
+      "step": 10912
+    },
+    {
+      "epoch": 0.25975911644292105,
+      "grad_norm": 4.314315007664234,
+      "learning_rate": 2e-05,
+      "loss": 0.7106,
+      "step": 10913
+    },
+    {
+      "epoch": 0.2597829191659526,
+      "grad_norm": 7.069549465044231,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 10914
+    },
+    {
+      "epoch": 0.2598067218889841,
+      "grad_norm": 6.866729659740736,
+      "learning_rate": 2e-05,
+      "loss": 1.2323,
+      "step": 10915
+    },
+    {
+      "epoch": 0.2598305246120156,
+      "grad_norm": 9.148373587830413,
+      "learning_rate": 2e-05,
+      "loss": 0.3032,
+      "step": 10916
+    },
+    {
+      "epoch": 0.25985432733504715,
+      "grad_norm": 11.381157463150014,
+      "learning_rate": 2e-05,
+      "loss": 0.8476,
+      "step": 10917
+    },
+    {
+      "epoch": 0.25987813005807864,
+      "grad_norm": 2.6477838423413607,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 10918
+    },
+    {
+      "epoch": 0.2599019327811102,
+      "grad_norm": 8.282821226919724,
+      "learning_rate": 2e-05,
+      "loss": 0.3921,
+      "step": 10919
+    },
+    {
+      "epoch": 0.25992573550414166,
+      "grad_norm": 5.203105883747255,
+      "learning_rate": 2e-05,
+      "loss": 0.8266,
+      "step": 10920
+    },
+    {
+      "epoch": 0.2599495382271732,
+      "grad_norm": 4.1053054204451955,
+      "learning_rate": 2e-05,
+      "loss": 0.8999,
+      "step": 10921
+    },
+    {
+      "epoch": 0.2599733409502047,
+      "grad_norm": 3.6182638694277474,
+      "learning_rate": 2e-05,
+      "loss": 0.2688,
+      "step": 10922
+    },
+    {
+      "epoch": 0.25999714367323623,
+      "grad_norm": 1.1573394978756644,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 10923
+    },
+    {
+      "epoch": 0.2600209463962677,
+      "grad_norm": 5.276467019970001,
+      "learning_rate": 2e-05,
+      "loss": 1.0937,
+      "step": 10924
+    },
+    {
+      "epoch": 0.26004474911929926,
+      "grad_norm": 1.9634760789789987,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 10925
+    },
+    {
+      "epoch": 0.26006855184233074,
+      "grad_norm": 1.9009506759416657,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 10926
+    },
+    {
+      "epoch": 0.2600923545653623,
+      "grad_norm": 5.071688412106921,
+      "learning_rate": 2e-05,
+      "loss": 1.0326,
+      "step": 10927
+    },
+    {
+      "epoch": 0.26011615728839377,
+      "grad_norm": 5.795002002317489,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 10928
+    },
+    {
+      "epoch": 0.2601399600114253,
+      "grad_norm": 1.3037250144173398,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 10929
+    },
+    {
+      "epoch": 0.26016376273445685,
+      "grad_norm": 2.5060573313204366,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 10930
+    },
+    {
+      "epoch": 0.26018756545748833,
+      "grad_norm": 2.007139095142761,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 10931
+    },
+    {
+      "epoch": 0.2602113681805199,
+      "grad_norm": 2.1135378612652675,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 10932
+    },
+    {
+      "epoch": 0.26023517090355136,
+      "grad_norm": 6.823869101119163,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 10933
+    },
+    {
+      "epoch": 0.2602589736265829,
+      "grad_norm": 4.383538879578836,
+      "learning_rate": 2e-05,
+      "loss": 0.7759,
+      "step": 10934
+    },
+    {
+      "epoch": 0.2602827763496144,
+      "grad_norm": 2.009931130071259,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 10935
+    },
+    {
+      "epoch": 0.2603065790726459,
+      "grad_norm": 1.5163738557167261,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 10936
+    },
+    {
+      "epoch": 0.2603303817956774,
+      "grad_norm": 5.8426992758293395,
+      "learning_rate": 2e-05,
+      "loss": 0.8775,
+      "step": 10937
+    },
+    {
+      "epoch": 0.26035418451870895,
+      "grad_norm": 5.617776137892044,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 10938
+    },
+    {
+      "epoch": 0.26037798724174044,
+      "grad_norm": 5.974771869090001,
+      "learning_rate": 2e-05,
+      "loss": 1.1006,
+      "step": 10939
+    },
+    {
+      "epoch": 0.260401789964772,
+      "grad_norm": 8.595357021556907,
+      "learning_rate": 2e-05,
+      "loss": 0.9617,
+      "step": 10940
+    },
+    {
+      "epoch": 0.26042559268780346,
+      "grad_norm": 3.996799696514421,
+      "learning_rate": 2e-05,
+      "loss": 0.2646,
+      "step": 10941
+    },
+    {
+      "epoch": 0.260449395410835,
+      "grad_norm": 5.389746380811985,
+      "learning_rate": 2e-05,
+      "loss": 0.9705,
+      "step": 10942
+    },
+    {
+      "epoch": 0.2604731981338665,
+      "grad_norm": 3.5030810332906044,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 10943
+    },
+    {
+      "epoch": 0.26049700085689803,
+      "grad_norm": 1.7012095704435188,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 10944
+    },
+    {
+      "epoch": 0.2605208035799295,
+      "grad_norm": 3.7180674019603677,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 10945
+    },
+    {
+      "epoch": 0.26054460630296106,
+      "grad_norm": 4.533875617937217,
+      "learning_rate": 2e-05,
+      "loss": 0.229,
+      "step": 10946
+    },
+    {
+      "epoch": 0.2605684090259926,
+      "grad_norm": 2.5085598314824757,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 10947
+    },
+    {
+      "epoch": 0.2605922117490241,
+      "grad_norm": 3.370297551917097,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 10948
+    },
+    {
+      "epoch": 0.2606160144720556,
+      "grad_norm": 3.5940859786247663,
+      "learning_rate": 2e-05,
+      "loss": 0.2556,
+      "step": 10949
+    },
+    {
+      "epoch": 0.2606398171950871,
+      "grad_norm": 8.471086488981184,
+      "learning_rate": 2e-05,
+      "loss": 0.7793,
+      "step": 10950
+    },
+    {
+      "epoch": 0.26066361991811865,
+      "grad_norm": 3.873472981139091,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 10951
+    },
+    {
+      "epoch": 0.26068742264115013,
+      "grad_norm": 4.159658579626231,
+      "learning_rate": 2e-05,
+      "loss": 0.1931,
+      "step": 10952
+    },
+    {
+      "epoch": 0.2607112253641817,
+      "grad_norm": 3.887289883144461,
+      "learning_rate": 2e-05,
+      "loss": 0.132,
+      "step": 10953
+    },
+    {
+      "epoch": 0.26073502808721316,
+      "grad_norm": 5.732510963428866,
+      "learning_rate": 2e-05,
+      "loss": 1.0044,
+      "step": 10954
+    },
+    {
+      "epoch": 0.2607588308102447,
+      "grad_norm": 1.1747320180964687,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 10955
+    },
+    {
+      "epoch": 0.2607826335332762,
+      "grad_norm": 4.217722633188216,
+      "learning_rate": 2e-05,
+      "loss": 0.3063,
+      "step": 10956
+    },
+    {
+      "epoch": 0.2608064362563077,
+      "grad_norm": 4.890752023989247,
+      "learning_rate": 2e-05,
+      "loss": 0.71,
+      "step": 10957
+    },
+    {
+      "epoch": 0.2608302389793392,
+      "grad_norm": 3.20014525645334,
+      "learning_rate": 2e-05,
+      "loss": 0.1528,
+      "step": 10958
+    },
+    {
+      "epoch": 0.26085404170237075,
+      "grad_norm": 4.53929748059585,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 10959
+    },
+    {
+      "epoch": 0.26087784442540224,
+      "grad_norm": 11.251711039975405,
+      "learning_rate": 2e-05,
+      "loss": 0.9368,
+      "step": 10960
+    },
+    {
+      "epoch": 0.2609016471484338,
+      "grad_norm": 16.211933773511404,
+      "learning_rate": 2e-05,
+      "loss": 0.4965,
+      "step": 10961
+    },
+    {
+      "epoch": 0.2609254498714653,
+      "grad_norm": 4.765257814286025,
+      "learning_rate": 2e-05,
+      "loss": 0.766,
+      "step": 10962
+    },
+    {
+      "epoch": 0.2609492525944968,
+      "grad_norm": 4.10423577835465,
+      "learning_rate": 2e-05,
+      "loss": 0.6586,
+      "step": 10963
+    },
+    {
+      "epoch": 0.26097305531752835,
+      "grad_norm": 4.40875065538636,
+      "learning_rate": 2e-05,
+      "loss": 0.6174,
+      "step": 10964
+    },
+    {
+      "epoch": 0.26099685804055983,
+      "grad_norm": 2.4301333886850425,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 10965
+    },
+    {
+      "epoch": 0.2610206607635914,
+      "grad_norm": 3.7456383928764785,
+      "learning_rate": 2e-05,
+      "loss": 0.2145,
+      "step": 10966
+    },
+    {
+      "epoch": 0.26104446348662286,
+      "grad_norm": 4.261040695950028,
+      "learning_rate": 2e-05,
+      "loss": 0.3859,
+      "step": 10967
+    },
+    {
+      "epoch": 0.2610682662096544,
+      "grad_norm": 6.547133860665371,
+      "learning_rate": 2e-05,
+      "loss": 1.5803,
+      "step": 10968
+    },
+    {
+      "epoch": 0.2610920689326859,
+      "grad_norm": 9.942924305432774,
+      "learning_rate": 2e-05,
+      "loss": 0.3703,
+      "step": 10969
+    },
+    {
+      "epoch": 0.2611158716557174,
+      "grad_norm": 1.3708547936156454,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 10970
+    },
+    {
+      "epoch": 0.2611396743787489,
+      "grad_norm": 1.503555470922533,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 10971
+    },
+    {
+      "epoch": 0.26116347710178045,
+      "grad_norm": 3.020312620751161,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 10972
+    },
+    {
+      "epoch": 0.26118727982481194,
+      "grad_norm": 2.05582667880321,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 10973
+    },
+    {
+      "epoch": 0.2612110825478435,
+      "grad_norm": 0.9723554764621214,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 10974
+    },
+    {
+      "epoch": 0.26123488527087496,
+      "grad_norm": 14.993837951138582,
+      "learning_rate": 2e-05,
+      "loss": 1.8176,
+      "step": 10975
+    },
+    {
+      "epoch": 0.2612586879939065,
+      "grad_norm": 2.0551054280498895,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 10976
+    },
+    {
+      "epoch": 0.26128249071693804,
+      "grad_norm": 3.7459472051770533,
+      "learning_rate": 2e-05,
+      "loss": 0.1837,
+      "step": 10977
+    },
+    {
+      "epoch": 0.26130629343996953,
+      "grad_norm": 3.5893704210746815,
+      "learning_rate": 2e-05,
+      "loss": 0.1959,
+      "step": 10978
+    },
+    {
+      "epoch": 0.26133009616300107,
+      "grad_norm": 2.0671147704032924,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 10979
+    },
+    {
+      "epoch": 0.26135389888603255,
+      "grad_norm": 1.9775299522870777,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 10980
+    },
+    {
+      "epoch": 0.2613777016090641,
+      "grad_norm": 7.558621056277632,
+      "learning_rate": 2e-05,
+      "loss": 0.6514,
+      "step": 10981
+    },
+    {
+      "epoch": 0.2614015043320956,
+      "grad_norm": 3.238336231370237,
+      "learning_rate": 2e-05,
+      "loss": 0.331,
+      "step": 10982
+    },
+    {
+      "epoch": 0.2614253070551271,
+      "grad_norm": 7.350054625631482,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 10983
+    },
+    {
+      "epoch": 0.2614491097781586,
+      "grad_norm": 3.1402776484131505,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 10984
+    },
+    {
+      "epoch": 0.26147291250119015,
+      "grad_norm": 5.546515306836419,
+      "learning_rate": 2e-05,
+      "loss": 0.7993,
+      "step": 10985
+    },
+    {
+      "epoch": 0.26149671522422163,
+      "grad_norm": 5.624934908895934,
+      "learning_rate": 2e-05,
+      "loss": 0.3068,
+      "step": 10986
+    },
+    {
+      "epoch": 0.2615205179472532,
+      "grad_norm": 5.2125713085671315,
+      "learning_rate": 2e-05,
+      "loss": 0.2285,
+      "step": 10987
+    },
+    {
+      "epoch": 0.26154432067028466,
+      "grad_norm": 4.858618486985801,
+      "learning_rate": 2e-05,
+      "loss": 0.3477,
+      "step": 10988
+    },
+    {
+      "epoch": 0.2615681233933162,
+      "grad_norm": 1.12822203696359,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 10989
+    },
+    {
+      "epoch": 0.2615919261163477,
+      "grad_norm": 8.501151405922277,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 10990
+    },
+    {
+      "epoch": 0.2616157288393792,
+      "grad_norm": 5.403421440102962,
+      "learning_rate": 2e-05,
+      "loss": 0.4072,
+      "step": 10991
+    },
+    {
+      "epoch": 0.26163953156241077,
+      "grad_norm": 7.480336224149052,
+      "learning_rate": 2e-05,
+      "loss": 1.2383,
+      "step": 10992
+    },
+    {
+      "epoch": 0.26166333428544225,
+      "grad_norm": 9.840728228987492,
+      "learning_rate": 2e-05,
+      "loss": 1.6253,
+      "step": 10993
+    },
+    {
+      "epoch": 0.2616871370084738,
+      "grad_norm": 6.7611289433625,
+      "learning_rate": 2e-05,
+      "loss": 0.9058,
+      "step": 10994
+    },
+    {
+      "epoch": 0.2617109397315053,
+      "grad_norm": 3.827311472345144,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 10995
+    },
+    {
+      "epoch": 0.2617347424545368,
+      "grad_norm": 8.479003284574542,
+      "learning_rate": 2e-05,
+      "loss": 0.3267,
+      "step": 10996
+    },
+    {
+      "epoch": 0.2617585451775683,
+      "grad_norm": 2.28904735538402,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 10997
+    },
+    {
+      "epoch": 0.26178234790059984,
+      "grad_norm": 0.8847891238855786,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 10998
+    },
+    {
+      "epoch": 0.26180615062363133,
+      "grad_norm": 12.780060991328131,
+      "learning_rate": 2e-05,
+      "loss": 0.653,
+      "step": 10999
+    },
+    {
+      "epoch": 0.26182995334666287,
+      "grad_norm": 2.391081832218374,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26185375606969435,
+      "grad_norm": 0.8441346465067229,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 11001
+    },
+    {
+      "epoch": 0.2618775587927259,
+      "grad_norm": 3.9092545591574526,
+      "learning_rate": 2e-05,
+      "loss": 0.1787,
+      "step": 11002
+    },
+    {
+      "epoch": 0.2619013615157574,
+      "grad_norm": 2.412332409223673,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 11003
+    },
+    {
+      "epoch": 0.2619251642387889,
+      "grad_norm": 2.1924489329349393,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 11004
+    },
+    {
+      "epoch": 0.2619489669618204,
+      "grad_norm": 2.5449792101423587,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 11005
+    },
+    {
+      "epoch": 0.26197276968485195,
+      "grad_norm": 3.9250433199393346,
+      "learning_rate": 2e-05,
+      "loss": 0.4838,
+      "step": 11006
+    },
+    {
+      "epoch": 0.2619965724078835,
+      "grad_norm": 8.711717988334515,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 11007
+    },
+    {
+      "epoch": 0.262020375130915,
+      "grad_norm": 4.632968135142036,
+      "learning_rate": 2e-05,
+      "loss": 0.5436,
+      "step": 11008
+    },
+    {
+      "epoch": 0.2620441778539465,
+      "grad_norm": 19.757958869516003,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 11009
+    },
+    {
+      "epoch": 0.262067980576978,
+      "grad_norm": 13.302725424541443,
+      "learning_rate": 2e-05,
+      "loss": 1.1844,
+      "step": 11010
+    },
+    {
+      "epoch": 0.26209178330000954,
+      "grad_norm": 1.668078699526675,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 11011
+    },
+    {
+      "epoch": 0.262115586023041,
+      "grad_norm": 14.022287619761428,
+      "learning_rate": 2e-05,
+      "loss": 0.9285,
+      "step": 11012
+    },
+    {
+      "epoch": 0.26213938874607257,
+      "grad_norm": 1.0937318424625997,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 11013
+    },
+    {
+      "epoch": 0.26216319146910405,
+      "grad_norm": 3.5773381517376137,
+      "learning_rate": 2e-05,
+      "loss": 0.1759,
+      "step": 11014
+    },
+    {
+      "epoch": 0.2621869941921356,
+      "grad_norm": 14.388864623534179,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 11015
+    },
+    {
+      "epoch": 0.2622107969151671,
+      "grad_norm": 3.5563281434164002,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 11016
+    },
+    {
+      "epoch": 0.2622345996381986,
+      "grad_norm": 7.295431805837456,
+      "learning_rate": 2e-05,
+      "loss": 0.6117,
+      "step": 11017
+    },
+    {
+      "epoch": 0.2622584023612301,
+      "grad_norm": 4.981064656287021,
+      "learning_rate": 2e-05,
+      "loss": 0.3161,
+      "step": 11018
+    },
+    {
+      "epoch": 0.26228220508426164,
+      "grad_norm": 4.459282278041379,
+      "learning_rate": 2e-05,
+      "loss": 0.7057,
+      "step": 11019
+    },
+    {
+      "epoch": 0.26230600780729313,
+      "grad_norm": 15.683612924761674,
+      "learning_rate": 2e-05,
+      "loss": 1.1614,
+      "step": 11020
+    },
+    {
+      "epoch": 0.26232981053032467,
+      "grad_norm": 8.654442555771567,
+      "learning_rate": 2e-05,
+      "loss": 0.6894,
+      "step": 11021
+    },
+    {
+      "epoch": 0.2623536132533562,
+      "grad_norm": 0.7912338020000887,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 11022
+    },
+    {
+      "epoch": 0.2623774159763877,
+      "grad_norm": 4.531426070973697,
+      "learning_rate": 2e-05,
+      "loss": 0.4464,
+      "step": 11023
+    },
+    {
+      "epoch": 0.26240121869941924,
+      "grad_norm": 3.591517535117464,
+      "learning_rate": 2e-05,
+      "loss": 0.2989,
+      "step": 11024
+    },
+    {
+      "epoch": 0.2624250214224507,
+      "grad_norm": 13.514391890361095,
+      "learning_rate": 2e-05,
+      "loss": 0.882,
+      "step": 11025
+    },
+    {
+      "epoch": 0.26244882414548226,
+      "grad_norm": 5.539527519945279,
+      "learning_rate": 2e-05,
+      "loss": 0.9027,
+      "step": 11026
+    },
+    {
+      "epoch": 0.26247262686851375,
+      "grad_norm": 4.890370485347912,
+      "learning_rate": 2e-05,
+      "loss": 0.8409,
+      "step": 11027
+    },
+    {
+      "epoch": 0.2624964295915453,
+      "grad_norm": 3.3038992501398594,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 11028
+    },
+    {
+      "epoch": 0.2625202323145768,
+      "grad_norm": 5.841804137547143,
+      "learning_rate": 2e-05,
+      "loss": 0.8866,
+      "step": 11029
+    },
+    {
+      "epoch": 0.2625440350376083,
+      "grad_norm": 7.954326823186632,
+      "learning_rate": 2e-05,
+      "loss": 0.4778,
+      "step": 11030
+    },
+    {
+      "epoch": 0.2625678377606398,
+      "grad_norm": 0.7809263186452116,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 11031
+    },
+    {
+      "epoch": 0.26259164048367134,
+      "grad_norm": 4.183911987276552,
+      "learning_rate": 2e-05,
+      "loss": 0.8916,
+      "step": 11032
+    },
+    {
+      "epoch": 0.2626154432067028,
+      "grad_norm": 3.072966093029238,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 11033
+    },
+    {
+      "epoch": 0.26263924592973437,
+      "grad_norm": 2.656620677062103,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 11034
+    },
+    {
+      "epoch": 0.26266304865276585,
+      "grad_norm": 9.279964862015715,
+      "learning_rate": 2e-05,
+      "loss": 0.3257,
+      "step": 11035
+    },
+    {
+      "epoch": 0.2626868513757974,
+      "grad_norm": 4.724959004129287,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 11036
+    },
+    {
+      "epoch": 0.26271065409882893,
+      "grad_norm": 7.01124200907212,
+      "learning_rate": 2e-05,
+      "loss": 0.2548,
+      "step": 11037
+    },
+    {
+      "epoch": 0.2627344568218604,
+      "grad_norm": 7.087194556905783,
+      "learning_rate": 2e-05,
+      "loss": 0.3016,
+      "step": 11038
+    },
+    {
+      "epoch": 0.26275825954489196,
+      "grad_norm": 4.186797537527675,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 11039
+    },
+    {
+      "epoch": 0.26278206226792344,
+      "grad_norm": 0.9068124772708188,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 11040
+    },
+    {
+      "epoch": 0.262805864990955,
+      "grad_norm": 5.0376465474234005,
+      "learning_rate": 2e-05,
+      "loss": 0.5783,
+      "step": 11041
+    },
+    {
+      "epoch": 0.26282966771398647,
+      "grad_norm": 4.102944426241032,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 11042
+    },
+    {
+      "epoch": 0.262853470437018,
+      "grad_norm": 7.86890793425355,
+      "learning_rate": 2e-05,
+      "loss": 0.5148,
+      "step": 11043
+    },
+    {
+      "epoch": 0.2628772731600495,
+      "grad_norm": 6.534370563233652,
+      "learning_rate": 2e-05,
+      "loss": 1.4509,
+      "step": 11044
+    },
+    {
+      "epoch": 0.26290107588308104,
+      "grad_norm": 6.801160687945194,
+      "learning_rate": 2e-05,
+      "loss": 0.2955,
+      "step": 11045
+    },
+    {
+      "epoch": 0.2629248786061125,
+      "grad_norm": 5.5983672850445085,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 11046
+    },
+    {
+      "epoch": 0.26294868132914406,
+      "grad_norm": 8.954702909111173,
+      "learning_rate": 2e-05,
+      "loss": 0.4029,
+      "step": 11047
+    },
+    {
+      "epoch": 0.26297248405217555,
+      "grad_norm": 4.3432283520871104,
+      "learning_rate": 2e-05,
+      "loss": 0.4521,
+      "step": 11048
+    },
+    {
+      "epoch": 0.2629962867752071,
+      "grad_norm": 2.1923249232699527,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 11049
+    },
+    {
+      "epoch": 0.2630200894982386,
+      "grad_norm": 1.2242697914169012,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 11050
+    },
+    {
+      "epoch": 0.2630438922212701,
+      "grad_norm": 4.403833756320459,
+      "learning_rate": 2e-05,
+      "loss": 1.2248,
+      "step": 11051
+    },
+    {
+      "epoch": 0.26306769494430166,
+      "grad_norm": 3.100538964015362,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 11052
+    },
+    {
+      "epoch": 0.26309149766733314,
+      "grad_norm": 4.599181242532915,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 11053
+    },
+    {
+      "epoch": 0.2631153003903647,
+      "grad_norm": 8.53634435026414,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 11054
+    },
+    {
+      "epoch": 0.26313910311339617,
+      "grad_norm": 1.682866701152625,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 11055
+    },
+    {
+      "epoch": 0.2631629058364277,
+      "grad_norm": 4.043674084955777,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 11056
+    },
+    {
+      "epoch": 0.2631867085594592,
+      "grad_norm": 1.7550043624997242,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 11057
+    },
+    {
+      "epoch": 0.26321051128249073,
+      "grad_norm": 1.4528920030108792,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 11058
+    },
+    {
+      "epoch": 0.2632343140055222,
+      "grad_norm": 6.027526455981879,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 11059
+    },
+    {
+      "epoch": 0.26325811672855376,
+      "grad_norm": 6.569524651371292,
+      "learning_rate": 2e-05,
+      "loss": 1.2372,
+      "step": 11060
+    },
+    {
+      "epoch": 0.26328191945158524,
+      "grad_norm": 6.645905722560469,
+      "learning_rate": 2e-05,
+      "loss": 0.8072,
+      "step": 11061
+    },
+    {
+      "epoch": 0.2633057221746168,
+      "grad_norm": 4.055292343532824,
+      "learning_rate": 2e-05,
+      "loss": 0.4532,
+      "step": 11062
+    },
+    {
+      "epoch": 0.26332952489764827,
+      "grad_norm": 3.9396859179038315,
+      "learning_rate": 2e-05,
+      "loss": 0.2042,
+      "step": 11063
+    },
+    {
+      "epoch": 0.2633533276206798,
+      "grad_norm": 4.645663533633716,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 11064
+    },
+    {
+      "epoch": 0.2633771303437113,
+      "grad_norm": 1.9955334428218066,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 11065
+    },
+    {
+      "epoch": 0.26340093306674284,
+      "grad_norm": 9.139207126877366,
+      "learning_rate": 2e-05,
+      "loss": 0.3701,
+      "step": 11066
+    },
+    {
+      "epoch": 0.2634247357897743,
+      "grad_norm": 1.5671519529774618,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 11067
+    },
+    {
+      "epoch": 0.26344853851280586,
+      "grad_norm": 8.06733663153684,
+      "learning_rate": 2e-05,
+      "loss": 0.7441,
+      "step": 11068
+    },
+    {
+      "epoch": 0.2634723412358374,
+      "grad_norm": 3.3856478049776957,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 11069
+    },
+    {
+      "epoch": 0.2634961439588689,
+      "grad_norm": 8.18775662550401,
+      "learning_rate": 2e-05,
+      "loss": 1.1129,
+      "step": 11070
+    },
+    {
+      "epoch": 0.26351994668190043,
+      "grad_norm": 1.8919388287552437,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 11071
+    },
+    {
+      "epoch": 0.2635437494049319,
+      "grad_norm": 2.853790796232645,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 11072
+    },
+    {
+      "epoch": 0.26356755212796346,
+      "grad_norm": 4.007450935506205,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 11073
+    },
+    {
+      "epoch": 0.26359135485099494,
+      "grad_norm": 9.380657657870948,
+      "learning_rate": 2e-05,
+      "loss": 0.2008,
+      "step": 11074
+    },
+    {
+      "epoch": 0.2636151575740265,
+      "grad_norm": 3.5050918921738985,
+      "learning_rate": 2e-05,
+      "loss": 0.5403,
+      "step": 11075
+    },
+    {
+      "epoch": 0.26363896029705797,
+      "grad_norm": 3.9148768281795094,
+      "learning_rate": 2e-05,
+      "loss": 0.6072,
+      "step": 11076
+    },
+    {
+      "epoch": 0.2636627630200895,
+      "grad_norm": 4.319149226361313,
+      "learning_rate": 2e-05,
+      "loss": 0.622,
+      "step": 11077
+    },
+    {
+      "epoch": 0.263686565743121,
+      "grad_norm": 9.035880678776467,
+      "learning_rate": 2e-05,
+      "loss": 0.6623,
+      "step": 11078
+    },
+    {
+      "epoch": 0.26371036846615253,
+      "grad_norm": 8.09013710559372,
+      "learning_rate": 2e-05,
+      "loss": 0.3071,
+      "step": 11079
+    },
+    {
+      "epoch": 0.263734171189184,
+      "grad_norm": 4.366896142186413,
+      "learning_rate": 2e-05,
+      "loss": 0.8182,
+      "step": 11080
+    },
+    {
+      "epoch": 0.26375797391221556,
+      "grad_norm": 11.411319189825196,
+      "learning_rate": 2e-05,
+      "loss": 0.4961,
+      "step": 11081
+    },
+    {
+      "epoch": 0.26378177663524704,
+      "grad_norm": 6.841950369973988,
+      "learning_rate": 2e-05,
+      "loss": 0.7415,
+      "step": 11082
+    },
+    {
+      "epoch": 0.2638055793582786,
+      "grad_norm": 4.860400714892501,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 11083
+    },
+    {
+      "epoch": 0.2638293820813101,
+      "grad_norm": 3.0259342750931677,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 11084
+    },
+    {
+      "epoch": 0.2638531848043416,
+      "grad_norm": 5.58481842070901,
+      "learning_rate": 2e-05,
+      "loss": 0.5152,
+      "step": 11085
+    },
+    {
+      "epoch": 0.26387698752737315,
+      "grad_norm": 6.861186099200632,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 11086
+    },
+    {
+      "epoch": 0.26390079025040464,
+      "grad_norm": 4.021865976895545,
+      "learning_rate": 2e-05,
+      "loss": 0.2579,
+      "step": 11087
+    },
+    {
+      "epoch": 0.2639245929734362,
+      "grad_norm": 4.161771224799763,
+      "learning_rate": 2e-05,
+      "loss": 0.5725,
+      "step": 11088
+    },
+    {
+      "epoch": 0.26394839569646766,
+      "grad_norm": 6.286322996687396,
+      "learning_rate": 2e-05,
+      "loss": 0.3273,
+      "step": 11089
+    },
+    {
+      "epoch": 0.2639721984194992,
+      "grad_norm": 0.9206382146975802,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 11090
+    },
+    {
+      "epoch": 0.2639960011425307,
+      "grad_norm": 4.430848697841284,
+      "learning_rate": 2e-05,
+      "loss": 0.2737,
+      "step": 11091
+    },
+    {
+      "epoch": 0.26401980386556223,
+      "grad_norm": 4.726402306847706,
+      "learning_rate": 2e-05,
+      "loss": 0.6319,
+      "step": 11092
+    },
+    {
+      "epoch": 0.2640436065885937,
+      "grad_norm": 3.5352251371521803,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 11093
+    },
+    {
+      "epoch": 0.26406740931162526,
+      "grad_norm": 4.496857256699063,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 11094
+    },
+    {
+      "epoch": 0.26409121203465674,
+      "grad_norm": 7.718016383230533,
+      "learning_rate": 2e-05,
+      "loss": 0.8149,
+      "step": 11095
+    },
+    {
+      "epoch": 0.2641150147576883,
+      "grad_norm": 14.83363353416569,
+      "learning_rate": 2e-05,
+      "loss": 1.1496,
+      "step": 11096
+    },
+    {
+      "epoch": 0.26413881748071977,
+      "grad_norm": 1.1560576283027761,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 11097
+    },
+    {
+      "epoch": 0.2641626202037513,
+      "grad_norm": 2.985550008792259,
+      "learning_rate": 2e-05,
+      "loss": 0.2338,
+      "step": 11098
+    },
+    {
+      "epoch": 0.26418642292678285,
+      "grad_norm": 17.43980563063874,
+      "learning_rate": 2e-05,
+      "loss": 1.2634,
+      "step": 11099
+    },
+    {
+      "epoch": 0.26421022564981433,
+      "grad_norm": 5.0448866789825,
+      "learning_rate": 2e-05,
+      "loss": 0.1462,
+      "step": 11100
+    },
+    {
+      "epoch": 0.2642340283728459,
+      "grad_norm": 8.013776594694109,
+      "learning_rate": 2e-05,
+      "loss": 0.9147,
+      "step": 11101
+    },
+    {
+      "epoch": 0.26425783109587736,
+      "grad_norm": 5.636727078278143,
+      "learning_rate": 2e-05,
+      "loss": 0.201,
+      "step": 11102
+    },
+    {
+      "epoch": 0.2642816338189089,
+      "grad_norm": 7.214228656754315,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 11103
+    },
+    {
+      "epoch": 0.2643054365419404,
+      "grad_norm": 3.9825585293265937,
+      "learning_rate": 2e-05,
+      "loss": 0.1629,
+      "step": 11104
+    },
+    {
+      "epoch": 0.2643292392649719,
+      "grad_norm": 8.523655215510823,
+      "learning_rate": 2e-05,
+      "loss": 1.287,
+      "step": 11105
+    },
+    {
+      "epoch": 0.2643530419880034,
+      "grad_norm": 4.710874707359942,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 11106
+    },
+    {
+      "epoch": 0.26437684471103495,
+      "grad_norm": 2.6731282596034647,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 11107
+    },
+    {
+      "epoch": 0.26440064743406644,
+      "grad_norm": 6.341484878834299,
+      "learning_rate": 2e-05,
+      "loss": 1.0066,
+      "step": 11108
+    },
+    {
+      "epoch": 0.264424450157098,
+      "grad_norm": 12.958793672335693,
+      "learning_rate": 2e-05,
+      "loss": 0.4342,
+      "step": 11109
+    },
+    {
+      "epoch": 0.26444825288012946,
+      "grad_norm": 2.799255618066026,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 11110
+    },
+    {
+      "epoch": 0.264472055603161,
+      "grad_norm": 3.205568742565226,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 11111
+    },
+    {
+      "epoch": 0.2644958583261925,
+      "grad_norm": 1.1385460425566833,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 11112
+    },
+    {
+      "epoch": 0.26451966104922403,
+      "grad_norm": 2.6549313914027075,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 11113
+    },
+    {
+      "epoch": 0.26454346377225557,
+      "grad_norm": 5.03621739154717,
+      "learning_rate": 2e-05,
+      "loss": 0.4653,
+      "step": 11114
+    },
+    {
+      "epoch": 0.26456726649528706,
+      "grad_norm": 1.8427325999183113,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 11115
+    },
+    {
+      "epoch": 0.2645910692183186,
+      "grad_norm": 6.08670324513849,
+      "learning_rate": 2e-05,
+      "loss": 0.3045,
+      "step": 11116
+    },
+    {
+      "epoch": 0.2646148719413501,
+      "grad_norm": 2.6208352029651025,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 11117
+    },
+    {
+      "epoch": 0.2646386746643816,
+      "grad_norm": 9.982176600720722,
+      "learning_rate": 2e-05,
+      "loss": 1.6725,
+      "step": 11118
+    },
+    {
+      "epoch": 0.2646624773874131,
+      "grad_norm": 6.791256304620149,
+      "learning_rate": 2e-05,
+      "loss": 0.8515,
+      "step": 11119
+    },
+    {
+      "epoch": 0.26468628011044465,
+      "grad_norm": 2.624025080411227,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 11120
+    },
+    {
+      "epoch": 0.26471008283347613,
+      "grad_norm": 5.611456142812838,
+      "learning_rate": 2e-05,
+      "loss": 1.0752,
+      "step": 11121
+    },
+    {
+      "epoch": 0.2647338855565077,
+      "grad_norm": 9.46348701141997,
+      "learning_rate": 2e-05,
+      "loss": 0.5007,
+      "step": 11122
+    },
+    {
+      "epoch": 0.26475768827953916,
+      "grad_norm": 3.527840914411758,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 11123
+    },
+    {
+      "epoch": 0.2647814910025707,
+      "grad_norm": 4.320236703184796,
+      "learning_rate": 2e-05,
+      "loss": 0.671,
+      "step": 11124
+    },
+    {
+      "epoch": 0.2648052937256022,
+      "grad_norm": 4.524688379281285,
+      "learning_rate": 2e-05,
+      "loss": 0.7264,
+      "step": 11125
+    },
+    {
+      "epoch": 0.2648290964486337,
+      "grad_norm": 1.7443470158771832,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 11126
+    },
+    {
+      "epoch": 0.2648528991716652,
+      "grad_norm": 5.733362680858034,
+      "learning_rate": 2e-05,
+      "loss": 0.3282,
+      "step": 11127
+    },
+    {
+      "epoch": 0.26487670189469675,
+      "grad_norm": 2.185904376545117,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 11128
+    },
+    {
+      "epoch": 0.2649005046177283,
+      "grad_norm": 1.7148295091926744,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 11129
+    },
+    {
+      "epoch": 0.2649243073407598,
+      "grad_norm": 1.4520184050120943,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 11130
+    },
+    {
+      "epoch": 0.2649481100637913,
+      "grad_norm": 5.400782350099486,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 11131
+    },
+    {
+      "epoch": 0.2649719127868228,
+      "grad_norm": 5.343416988541051,
+      "learning_rate": 2e-05,
+      "loss": 1.0062,
+      "step": 11132
+    },
+    {
+      "epoch": 0.26499571550985435,
+      "grad_norm": 2.3100466594076963,
+      "learning_rate": 2e-05,
+      "loss": 0.1712,
+      "step": 11133
+    },
+    {
+      "epoch": 0.26501951823288583,
+      "grad_norm": 3.7069938061146095,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 11134
+    },
+    {
+      "epoch": 0.26504332095591737,
+      "grad_norm": 4.404907089254023,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 11135
+    },
+    {
+      "epoch": 0.26506712367894886,
+      "grad_norm": 8.724468466652572,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 11136
+    },
+    {
+      "epoch": 0.2650909264019804,
+      "grad_norm": 5.987461148449978,
+      "learning_rate": 2e-05,
+      "loss": 0.579,
+      "step": 11137
+    },
+    {
+      "epoch": 0.2651147291250119,
+      "grad_norm": 6.321580166066887,
+      "learning_rate": 2e-05,
+      "loss": 0.6566,
+      "step": 11138
+    },
+    {
+      "epoch": 0.2651385318480434,
+      "grad_norm": 2.6976517254123276,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 11139
+    },
+    {
+      "epoch": 0.2651623345710749,
+      "grad_norm": 8.227949430747689,
+      "learning_rate": 2e-05,
+      "loss": 1.0779,
+      "step": 11140
+    },
+    {
+      "epoch": 0.26518613729410645,
+      "grad_norm": 4.002798865802193,
+      "learning_rate": 2e-05,
+      "loss": 0.6683,
+      "step": 11141
+    },
+    {
+      "epoch": 0.26520994001713794,
+      "grad_norm": 9.599881886571728,
+      "learning_rate": 2e-05,
+      "loss": 0.3004,
+      "step": 11142
+    },
+    {
+      "epoch": 0.2652337427401695,
+      "grad_norm": 5.5931325553314695,
+      "learning_rate": 2e-05,
+      "loss": 1.206,
+      "step": 11143
+    },
+    {
+      "epoch": 0.265257545463201,
+      "grad_norm": 4.716847555493687,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 11144
+    },
+    {
+      "epoch": 0.2652813481862325,
+      "grad_norm": 5.739763354154199,
+      "learning_rate": 2e-05,
+      "loss": 0.3154,
+      "step": 11145
+    },
+    {
+      "epoch": 0.26530515090926404,
+      "grad_norm": 6.669845912469889,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 11146
+    },
+    {
+      "epoch": 0.26532895363229553,
+      "grad_norm": 9.664744254227788,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 11147
+    },
+    {
+      "epoch": 0.26535275635532707,
+      "grad_norm": 1.4139380585122192,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 11148
+    },
+    {
+      "epoch": 0.26537655907835855,
+      "grad_norm": 6.067431664305773,
+      "learning_rate": 2e-05,
+      "loss": 0.569,
+      "step": 11149
+    },
+    {
+      "epoch": 0.2654003618013901,
+      "grad_norm": 2.642023950062255,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 11150
+    },
+    {
+      "epoch": 0.2654241645244216,
+      "grad_norm": 5.363099264627551,
+      "learning_rate": 2e-05,
+      "loss": 0.9144,
+      "step": 11151
+    },
+    {
+      "epoch": 0.2654479672474531,
+      "grad_norm": 3.1891735201372793,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 11152
+    },
+    {
+      "epoch": 0.2654717699704846,
+      "grad_norm": 4.4353262890438545,
+      "learning_rate": 2e-05,
+      "loss": 0.4827,
+      "step": 11153
+    },
+    {
+      "epoch": 0.26549557269351615,
+      "grad_norm": 6.129259581769866,
+      "learning_rate": 2e-05,
+      "loss": 0.8578,
+      "step": 11154
+    },
+    {
+      "epoch": 0.26551937541654763,
+      "grad_norm": 5.465602308857727,
+      "learning_rate": 2e-05,
+      "loss": 0.6583,
+      "step": 11155
+    },
+    {
+      "epoch": 0.2655431781395792,
+      "grad_norm": 2.022237445926264,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 11156
+    },
+    {
+      "epoch": 0.26556698086261066,
+      "grad_norm": 5.026068462069759,
+      "learning_rate": 2e-05,
+      "loss": 1.0637,
+      "step": 11157
+    },
+    {
+      "epoch": 0.2655907835856422,
+      "grad_norm": 7.873402274599292,
+      "learning_rate": 2e-05,
+      "loss": 0.326,
+      "step": 11158
+    },
+    {
+      "epoch": 0.26561458630867374,
+      "grad_norm": 6.954621732417206,
+      "learning_rate": 2e-05,
+      "loss": 1.6541,
+      "step": 11159
+    },
+    {
+      "epoch": 0.2656383890317052,
+      "grad_norm": 11.538820398186878,
+      "learning_rate": 2e-05,
+      "loss": 1.0694,
+      "step": 11160
+    },
+    {
+      "epoch": 0.26566219175473677,
+      "grad_norm": 4.072879667915247,
+      "learning_rate": 2e-05,
+      "loss": 0.6023,
+      "step": 11161
+    },
+    {
+      "epoch": 0.26568599447776825,
+      "grad_norm": 5.423386182773297,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 11162
+    },
+    {
+      "epoch": 0.2657097972007998,
+      "grad_norm": 1.320309428596101,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 11163
+    },
+    {
+      "epoch": 0.2657335999238313,
+      "grad_norm": 15.564486516096784,
+      "learning_rate": 2e-05,
+      "loss": 1.1289,
+      "step": 11164
+    },
+    {
+      "epoch": 0.2657574026468628,
+      "grad_norm": 3.1138294301379124,
+      "learning_rate": 2e-05,
+      "loss": 0.1651,
+      "step": 11165
+    },
+    {
+      "epoch": 0.2657812053698943,
+      "grad_norm": 6.432691171776413,
+      "learning_rate": 2e-05,
+      "loss": 0.3633,
+      "step": 11166
+    },
+    {
+      "epoch": 0.26580500809292584,
+      "grad_norm": 2.346012102299665,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 11167
+    },
+    {
+      "epoch": 0.26582881081595733,
+      "grad_norm": 3.713807925027615,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 11168
+    },
+    {
+      "epoch": 0.26585261353898887,
+      "grad_norm": 3.206846063523187,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 11169
+    },
+    {
+      "epoch": 0.26587641626202035,
+      "grad_norm": 9.611193230667705,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 11170
+    },
+    {
+      "epoch": 0.2659002189850519,
+      "grad_norm": 1.7895446467950595,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 11171
+    },
+    {
+      "epoch": 0.2659240217080834,
+      "grad_norm": 3.312701642043764,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 11172
+    },
+    {
+      "epoch": 0.2659478244311149,
+      "grad_norm": 5.5620092936027605,
+      "learning_rate": 2e-05,
+      "loss": 1.2943,
+      "step": 11173
+    },
+    {
+      "epoch": 0.26597162715414646,
+      "grad_norm": 2.5123045306918907,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 11174
+    },
+    {
+      "epoch": 0.26599542987717795,
+      "grad_norm": 3.0615517568986395,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 11175
+    },
+    {
+      "epoch": 0.2660192326002095,
+      "grad_norm": 2.7561850980700573,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 11176
+    },
+    {
+      "epoch": 0.266043035323241,
+      "grad_norm": 5.321148273275194,
+      "learning_rate": 2e-05,
+      "loss": 1.2513,
+      "step": 11177
+    },
+    {
+      "epoch": 0.2660668380462725,
+      "grad_norm": 2.2931358120652274,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 11178
+    },
+    {
+      "epoch": 0.266090640769304,
+      "grad_norm": 0.990672148476506,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 11179
+    },
+    {
+      "epoch": 0.26611444349233554,
+      "grad_norm": 3.993691364996223,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 11180
+    },
+    {
+      "epoch": 0.266138246215367,
+      "grad_norm": 5.550026826048462,
+      "learning_rate": 2e-05,
+      "loss": 1.5896,
+      "step": 11181
+    },
+    {
+      "epoch": 0.26616204893839857,
+      "grad_norm": 6.130779065360839,
+      "learning_rate": 2e-05,
+      "loss": 0.8676,
+      "step": 11182
+    },
+    {
+      "epoch": 0.26618585166143005,
+      "grad_norm": 4.148110513422333,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 11183
+    },
+    {
+      "epoch": 0.2662096543844616,
+      "grad_norm": 5.2828072590783375,
+      "learning_rate": 2e-05,
+      "loss": 0.8956,
+      "step": 11184
+    },
+    {
+      "epoch": 0.2662334571074931,
+      "grad_norm": 4.876751407218139,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 11185
+    },
+    {
+      "epoch": 0.2662572598305246,
+      "grad_norm": 1.1703746194773461,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 11186
+    },
+    {
+      "epoch": 0.2662810625535561,
+      "grad_norm": 5.444988726494996,
+      "learning_rate": 2e-05,
+      "loss": 1.147,
+      "step": 11187
+    },
+    {
+      "epoch": 0.26630486527658764,
+      "grad_norm": 3.4058264508104816,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 11188
+    },
+    {
+      "epoch": 0.2663286679996192,
+      "grad_norm": 1.301246799377759,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 11189
+    },
+    {
+      "epoch": 0.26635247072265067,
+      "grad_norm": 1.580695860325053,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 11190
+    },
+    {
+      "epoch": 0.2663762734456822,
+      "grad_norm": 2.0350258819634104,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 11191
+    },
+    {
+      "epoch": 0.2664000761687137,
+      "grad_norm": 4.6261655726205095,
+      "learning_rate": 2e-05,
+      "loss": 0.8465,
+      "step": 11192
+    },
+    {
+      "epoch": 0.26642387889174524,
+      "grad_norm": 8.300806853141998,
+      "learning_rate": 2e-05,
+      "loss": 0.6149,
+      "step": 11193
+    },
+    {
+      "epoch": 0.2664476816147767,
+      "grad_norm": 6.218422320909903,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 11194
+    },
+    {
+      "epoch": 0.26647148433780826,
+      "grad_norm": 3.891808738373367,
+      "learning_rate": 2e-05,
+      "loss": 0.3787,
+      "step": 11195
+    },
+    {
+      "epoch": 0.26649528706083975,
+      "grad_norm": 5.1508141433688746,
+      "learning_rate": 2e-05,
+      "loss": 0.871,
+      "step": 11196
+    },
+    {
+      "epoch": 0.2665190897838713,
+      "grad_norm": 1.6568173701906677,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 11197
+    },
+    {
+      "epoch": 0.2665428925069028,
+      "grad_norm": 5.30596076927495,
+      "learning_rate": 2e-05,
+      "loss": 0.1781,
+      "step": 11198
+    },
+    {
+      "epoch": 0.2665666952299343,
+      "grad_norm": 0.7281206160695565,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11199
+    },
+    {
+      "epoch": 0.2665904979529658,
+      "grad_norm": 4.777423375145085,
+      "learning_rate": 2e-05,
+      "loss": 1.0231,
+      "step": 11200
+    },
+    {
+      "epoch": 0.26661430067599734,
+      "grad_norm": 2.7176210684866415,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 11201
+    },
+    {
+      "epoch": 0.2666381033990288,
+      "grad_norm": 2.3641690537197744,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 11202
+    },
+    {
+      "epoch": 0.26666190612206037,
+      "grad_norm": 2.3270804142314456,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 11203
+    },
+    {
+      "epoch": 0.26668570884509185,
+      "grad_norm": 4.987531107568586,
+      "learning_rate": 2e-05,
+      "loss": 0.2244,
+      "step": 11204
+    },
+    {
+      "epoch": 0.2667095115681234,
+      "grad_norm": 4.521100934569938,
+      "learning_rate": 2e-05,
+      "loss": 0.6281,
+      "step": 11205
+    },
+    {
+      "epoch": 0.26673331429115493,
+      "grad_norm": 1.3549005577049136,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 11206
+    },
+    {
+      "epoch": 0.2667571170141864,
+      "grad_norm": 2.2720831680084945,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 11207
+    },
+    {
+      "epoch": 0.26678091973721796,
+      "grad_norm": 6.552002581459431,
+      "learning_rate": 2e-05,
+      "loss": 0.7263,
+      "step": 11208
+    },
+    {
+      "epoch": 0.26680472246024944,
+      "grad_norm": 3.833893265001432,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 11209
+    },
+    {
+      "epoch": 0.266828525183281,
+      "grad_norm": 1.2475883648646424,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 11210
+    },
+    {
+      "epoch": 0.26685232790631247,
+      "grad_norm": 3.804957949145965,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 11211
+    },
+    {
+      "epoch": 0.266876130629344,
+      "grad_norm": 1.6919813509417339,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 11212
+    },
+    {
+      "epoch": 0.2668999333523755,
+      "grad_norm": 6.778636672325326,
+      "learning_rate": 2e-05,
+      "loss": 0.7504,
+      "step": 11213
+    },
+    {
+      "epoch": 0.26692373607540704,
+      "grad_norm": 5.745511308959104,
+      "learning_rate": 2e-05,
+      "loss": 0.7492,
+      "step": 11214
+    },
+    {
+      "epoch": 0.2669475387984385,
+      "grad_norm": 6.29644292634367,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 11215
+    },
+    {
+      "epoch": 0.26697134152147006,
+      "grad_norm": 1.250082978377788,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 11216
+    },
+    {
+      "epoch": 0.26699514424450155,
+      "grad_norm": 5.93132217735157,
+      "learning_rate": 2e-05,
+      "loss": 0.609,
+      "step": 11217
+    },
+    {
+      "epoch": 0.2670189469675331,
+      "grad_norm": 5.8083521773931395,
+      "learning_rate": 2e-05,
+      "loss": 0.1721,
+      "step": 11218
+    },
+    {
+      "epoch": 0.2670427496905646,
+      "grad_norm": 3.2012673165667205,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 11219
+    },
+    {
+      "epoch": 0.2670665524135961,
+      "grad_norm": 5.317225577701042,
+      "learning_rate": 2e-05,
+      "loss": 0.9645,
+      "step": 11220
+    },
+    {
+      "epoch": 0.26709035513662766,
+      "grad_norm": 2.3458762348079856,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 11221
+    },
+    {
+      "epoch": 0.26711415785965914,
+      "grad_norm": 6.919684086045705,
+      "learning_rate": 2e-05,
+      "loss": 0.5887,
+      "step": 11222
+    },
+    {
+      "epoch": 0.2671379605826907,
+      "grad_norm": 5.641183104978161,
+      "learning_rate": 2e-05,
+      "loss": 0.8145,
+      "step": 11223
+    },
+    {
+      "epoch": 0.26716176330572217,
+      "grad_norm": 4.529665240420692,
+      "learning_rate": 2e-05,
+      "loss": 0.2702,
+      "step": 11224
+    },
+    {
+      "epoch": 0.2671855660287537,
+      "grad_norm": 18.033262979704464,
+      "learning_rate": 2e-05,
+      "loss": 0.9724,
+      "step": 11225
+    },
+    {
+      "epoch": 0.2672093687517852,
+      "grad_norm": 3.22834614507084,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 11226
+    },
+    {
+      "epoch": 0.26723317147481673,
+      "grad_norm": 3.439288810916718,
+      "learning_rate": 2e-05,
+      "loss": 0.698,
+      "step": 11227
+    },
+    {
+      "epoch": 0.2672569741978482,
+      "grad_norm": 9.046684382897572,
+      "learning_rate": 2e-05,
+      "loss": 0.224,
+      "step": 11228
+    },
+    {
+      "epoch": 0.26728077692087976,
+      "grad_norm": 3.8566817659938226,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 11229
+    },
+    {
+      "epoch": 0.26730457964391124,
+      "grad_norm": 7.71997938839187,
+      "learning_rate": 2e-05,
+      "loss": 0.4354,
+      "step": 11230
+    },
+    {
+      "epoch": 0.2673283823669428,
+      "grad_norm": 4.8268419515165535,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 11231
+    },
+    {
+      "epoch": 0.26735218508997427,
+      "grad_norm": 1.9720246202122418,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 11232
+    },
+    {
+      "epoch": 0.2673759878130058,
+      "grad_norm": 1.9485744443792907,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 11233
+    },
+    {
+      "epoch": 0.2673997905360373,
+      "grad_norm": 0.5518001880834991,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 11234
+    },
+    {
+      "epoch": 0.26742359325906884,
+      "grad_norm": 0.8699374663868331,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 11235
+    },
+    {
+      "epoch": 0.2674473959821004,
+      "grad_norm": 1.738175856051931,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 11236
+    },
+    {
+      "epoch": 0.26747119870513186,
+      "grad_norm": 1.735628434628672,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 11237
+    },
+    {
+      "epoch": 0.2674950014281634,
+      "grad_norm": 0.9644592546142443,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 11238
+    },
+    {
+      "epoch": 0.2675188041511949,
+      "grad_norm": 9.802071634345321,
+      "learning_rate": 2e-05,
+      "loss": 1.0081,
+      "step": 11239
+    },
+    {
+      "epoch": 0.26754260687422643,
+      "grad_norm": 1.8518025762093449,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 11240
+    },
+    {
+      "epoch": 0.2675664095972579,
+      "grad_norm": 3.7631191807607967,
+      "learning_rate": 2e-05,
+      "loss": 0.6753,
+      "step": 11241
+    },
+    {
+      "epoch": 0.26759021232028946,
+      "grad_norm": 5.011695267363996,
+      "learning_rate": 2e-05,
+      "loss": 0.9917,
+      "step": 11242
+    },
+    {
+      "epoch": 0.26761401504332094,
+      "grad_norm": 5.31390357790671,
+      "learning_rate": 2e-05,
+      "loss": 0.5054,
+      "step": 11243
+    },
+    {
+      "epoch": 0.2676378177663525,
+      "grad_norm": 3.6697848269822857,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 11244
+    },
+    {
+      "epoch": 0.26766162048938397,
+      "grad_norm": 4.874997115855409,
+      "learning_rate": 2e-05,
+      "loss": 0.6745,
+      "step": 11245
+    },
+    {
+      "epoch": 0.2676854232124155,
+      "grad_norm": 5.735124862385644,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 11246
+    },
+    {
+      "epoch": 0.267709225935447,
+      "grad_norm": 17.723246301636156,
+      "learning_rate": 2e-05,
+      "loss": 1.1191,
+      "step": 11247
+    },
+    {
+      "epoch": 0.26773302865847853,
+      "grad_norm": 11.608768287774904,
+      "learning_rate": 2e-05,
+      "loss": 0.2831,
+      "step": 11248
+    },
+    {
+      "epoch": 0.26775683138151,
+      "grad_norm": 2.6012104420840014,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 11249
+    },
+    {
+      "epoch": 0.26778063410454156,
+      "grad_norm": 2.751857762254367,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 11250
+    },
+    {
+      "epoch": 0.2678044368275731,
+      "grad_norm": 6.512151706242089,
+      "learning_rate": 2e-05,
+      "loss": 0.2174,
+      "step": 11251
+    },
+    {
+      "epoch": 0.2678282395506046,
+      "grad_norm": 5.850866757197943,
+      "learning_rate": 2e-05,
+      "loss": 0.8603,
+      "step": 11252
+    },
+    {
+      "epoch": 0.2678520422736361,
+      "grad_norm": 3.5478451609664576,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 11253
+    },
+    {
+      "epoch": 0.2678758449966676,
+      "grad_norm": 3.4776238324560147,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 11254
+    },
+    {
+      "epoch": 0.26789964771969915,
+      "grad_norm": 5.602419362724646,
+      "learning_rate": 2e-05,
+      "loss": 0.7034,
+      "step": 11255
+    },
+    {
+      "epoch": 0.26792345044273064,
+      "grad_norm": 3.018811972846903,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 11256
+    },
+    {
+      "epoch": 0.2679472531657622,
+      "grad_norm": 6.155424067071989,
+      "learning_rate": 2e-05,
+      "loss": 0.7311,
+      "step": 11257
+    },
+    {
+      "epoch": 0.26797105588879366,
+      "grad_norm": 5.020052419103554,
+      "learning_rate": 2e-05,
+      "loss": 0.7949,
+      "step": 11258
+    },
+    {
+      "epoch": 0.2679948586118252,
+      "grad_norm": 2.7893833174608167,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 11259
+    },
+    {
+      "epoch": 0.2680186613348567,
+      "grad_norm": 8.368232092113358,
+      "learning_rate": 2e-05,
+      "loss": 0.7537,
+      "step": 11260
+    },
+    {
+      "epoch": 0.26804246405788823,
+      "grad_norm": 8.006932684932528,
+      "learning_rate": 2e-05,
+      "loss": 1.8214,
+      "step": 11261
+    },
+    {
+      "epoch": 0.2680662667809197,
+      "grad_norm": 4.633945159800352,
+      "learning_rate": 2e-05,
+      "loss": 0.7467,
+      "step": 11262
+    },
+    {
+      "epoch": 0.26809006950395126,
+      "grad_norm": 4.824059658310058,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 11263
+    },
+    {
+      "epoch": 0.26811387222698274,
+      "grad_norm": 1.7854581059429644,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 11264
+    },
+    {
+      "epoch": 0.2681376749500143,
+      "grad_norm": 6.0562879029319,
+      "learning_rate": 2e-05,
+      "loss": 1.2779,
+      "step": 11265
+    },
+    {
+      "epoch": 0.2681614776730458,
+      "grad_norm": 6.625449687997547,
+      "learning_rate": 2e-05,
+      "loss": 0.6454,
+      "step": 11266
+    },
+    {
+      "epoch": 0.2681852803960773,
+      "grad_norm": 2.415721775146074,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 11267
+    },
+    {
+      "epoch": 0.26820908311910885,
+      "grad_norm": 9.374058708393006,
+      "learning_rate": 2e-05,
+      "loss": 1.1894,
+      "step": 11268
+    },
+    {
+      "epoch": 0.26823288584214033,
+      "grad_norm": 3.2417456636230773,
+      "learning_rate": 2e-05,
+      "loss": 0.433,
+      "step": 11269
+    },
+    {
+      "epoch": 0.2682566885651719,
+      "grad_norm": 5.176551016557778,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 11270
+    },
+    {
+      "epoch": 0.26828049128820336,
+      "grad_norm": 9.262984489702538,
+      "learning_rate": 2e-05,
+      "loss": 0.4249,
+      "step": 11271
+    },
+    {
+      "epoch": 0.2683042940112349,
+      "grad_norm": 4.558625005128687,
+      "learning_rate": 2e-05,
+      "loss": 0.3446,
+      "step": 11272
+    },
+    {
+      "epoch": 0.2683280967342664,
+      "grad_norm": 8.503665861751417,
+      "learning_rate": 2e-05,
+      "loss": 0.8172,
+      "step": 11273
+    },
+    {
+      "epoch": 0.2683518994572979,
+      "grad_norm": 4.690203786483604,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 11274
+    },
+    {
+      "epoch": 0.2683757021803294,
+      "grad_norm": 21.94053675714972,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 11275
+    },
+    {
+      "epoch": 0.26839950490336095,
+      "grad_norm": 3.877429107743765,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 11276
+    },
+    {
+      "epoch": 0.26842330762639244,
+      "grad_norm": 4.418400864142767,
+      "learning_rate": 2e-05,
+      "loss": 0.7451,
+      "step": 11277
+    },
+    {
+      "epoch": 0.268447110349424,
+      "grad_norm": 7.708963878851223,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 11278
+    },
+    {
+      "epoch": 0.26847091307245546,
+      "grad_norm": 3.117091360415551,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 11279
+    },
+    {
+      "epoch": 0.268494715795487,
+      "grad_norm": 4.988249223836748,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 11280
+    },
+    {
+      "epoch": 0.26851851851851855,
+      "grad_norm": 6.356814740361354,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 11281
+    },
+    {
+      "epoch": 0.26854232124155003,
+      "grad_norm": 7.919855017940047,
+      "learning_rate": 2e-05,
+      "loss": 1.3044,
+      "step": 11282
+    },
+    {
+      "epoch": 0.26856612396458157,
+      "grad_norm": 2.261843397140661,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 11283
+    },
+    {
+      "epoch": 0.26858992668761306,
+      "grad_norm": 5.416400029004039,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 11284
+    },
+    {
+      "epoch": 0.2686137294106446,
+      "grad_norm": 3.172502814349265,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 11285
+    },
+    {
+      "epoch": 0.2686375321336761,
+      "grad_norm": 5.377281327723788,
+      "learning_rate": 2e-05,
+      "loss": 0.9601,
+      "step": 11286
+    },
+    {
+      "epoch": 0.2686613348567076,
+      "grad_norm": 14.714103161503242,
+      "learning_rate": 2e-05,
+      "loss": 0.99,
+      "step": 11287
+    },
+    {
+      "epoch": 0.2686851375797391,
+      "grad_norm": 4.6406868827942285,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 11288
+    },
+    {
+      "epoch": 0.26870894030277065,
+      "grad_norm": 1.3139204097003836,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 11289
+    },
+    {
+      "epoch": 0.26873274302580213,
+      "grad_norm": 21.481648213154457,
+      "learning_rate": 2e-05,
+      "loss": 0.5787,
+      "step": 11290
+    },
+    {
+      "epoch": 0.2687565457488337,
+      "grad_norm": 4.779094504624482,
+      "learning_rate": 2e-05,
+      "loss": 0.5401,
+      "step": 11291
+    },
+    {
+      "epoch": 0.26878034847186516,
+      "grad_norm": 3.9168818641081127,
+      "learning_rate": 2e-05,
+      "loss": 1.0155,
+      "step": 11292
+    },
+    {
+      "epoch": 0.2688041511948967,
+      "grad_norm": 4.283701560692781,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 11293
+    },
+    {
+      "epoch": 0.2688279539179282,
+      "grad_norm": 4.709597496102177,
+      "learning_rate": 2e-05,
+      "loss": 0.3519,
+      "step": 11294
+    },
+    {
+      "epoch": 0.2688517566409597,
+      "grad_norm": 2.484517012888488,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 11295
+    },
+    {
+      "epoch": 0.26887555936399127,
+      "grad_norm": 5.7512928500494045,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 11296
+    },
+    {
+      "epoch": 0.26889936208702275,
+      "grad_norm": 7.485272649665313,
+      "learning_rate": 2e-05,
+      "loss": 1.5642,
+      "step": 11297
+    },
+    {
+      "epoch": 0.2689231648100543,
+      "grad_norm": 1.4875695764583876,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 11298
+    },
+    {
+      "epoch": 0.2689469675330858,
+      "grad_norm": 4.032884929104541,
+      "learning_rate": 2e-05,
+      "loss": 0.4313,
+      "step": 11299
+    },
+    {
+      "epoch": 0.2689707702561173,
+      "grad_norm": 3.5533121234343485,
+      "learning_rate": 2e-05,
+      "loss": 0.1418,
+      "step": 11300
+    },
+    {
+      "epoch": 0.2689945729791488,
+      "grad_norm": 5.7284138858818965,
+      "learning_rate": 2e-05,
+      "loss": 0.8723,
+      "step": 11301
+    },
+    {
+      "epoch": 0.26901837570218035,
+      "grad_norm": 7.742391226224848,
+      "learning_rate": 2e-05,
+      "loss": 0.562,
+      "step": 11302
+    },
+    {
+      "epoch": 0.26904217842521183,
+      "grad_norm": 4.437113599513995,
+      "learning_rate": 2e-05,
+      "loss": 0.5068,
+      "step": 11303
+    },
+    {
+      "epoch": 0.26906598114824337,
+      "grad_norm": 5.479905286308048,
+      "learning_rate": 2e-05,
+      "loss": 0.9445,
+      "step": 11304
+    },
+    {
+      "epoch": 0.26908978387127486,
+      "grad_norm": 6.32284483961833,
+      "learning_rate": 2e-05,
+      "loss": 1.1468,
+      "step": 11305
+    },
+    {
+      "epoch": 0.2691135865943064,
+      "grad_norm": 3.2448570628932782,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 11306
+    },
+    {
+      "epoch": 0.2691373893173379,
+      "grad_norm": 15.03613904562087,
+      "learning_rate": 2e-05,
+      "loss": 1.8304,
+      "step": 11307
+    },
+    {
+      "epoch": 0.2691611920403694,
+      "grad_norm": 1.1753422612099933,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 11308
+    },
+    {
+      "epoch": 0.2691849947634009,
+      "grad_norm": 2.496866612109467,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 11309
+    },
+    {
+      "epoch": 0.26920879748643245,
+      "grad_norm": 13.941689956155344,
+      "learning_rate": 2e-05,
+      "loss": 0.3169,
+      "step": 11310
+    },
+    {
+      "epoch": 0.269232600209464,
+      "grad_norm": 0.48966147154170575,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 11311
+    },
+    {
+      "epoch": 0.2692564029324955,
+      "grad_norm": 1.7888196540308818,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 11312
+    },
+    {
+      "epoch": 0.269280205655527,
+      "grad_norm": 3.539173046208678,
+      "learning_rate": 2e-05,
+      "loss": 0.6222,
+      "step": 11313
+    },
+    {
+      "epoch": 0.2693040083785585,
+      "grad_norm": 6.404244709196364,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 11314
+    },
+    {
+      "epoch": 0.26932781110159004,
+      "grad_norm": 4.130522381842992,
+      "learning_rate": 2e-05,
+      "loss": 0.6421,
+      "step": 11315
+    },
+    {
+      "epoch": 0.2693516138246215,
+      "grad_norm": 5.474342009882907,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 11316
+    },
+    {
+      "epoch": 0.26937541654765307,
+      "grad_norm": 5.669013473063207,
+      "learning_rate": 2e-05,
+      "loss": 0.1827,
+      "step": 11317
+    },
+    {
+      "epoch": 0.26939921927068455,
+      "grad_norm": 1.322816538067759,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 11318
+    },
+    {
+      "epoch": 0.2694230219937161,
+      "grad_norm": 11.295698017995536,
+      "learning_rate": 2e-05,
+      "loss": 0.5308,
+      "step": 11319
+    },
+    {
+      "epoch": 0.2694468247167476,
+      "grad_norm": 4.74693733490323,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 11320
+    },
+    {
+      "epoch": 0.2694706274397791,
+      "grad_norm": 4.076488751033759,
+      "learning_rate": 2e-05,
+      "loss": 0.5221,
+      "step": 11321
+    },
+    {
+      "epoch": 0.2694944301628106,
+      "grad_norm": 4.104493500397092,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 11322
+    },
+    {
+      "epoch": 0.26951823288584215,
+      "grad_norm": 5.554210501450364,
+      "learning_rate": 2e-05,
+      "loss": 0.7463,
+      "step": 11323
+    },
+    {
+      "epoch": 0.26954203560887363,
+      "grad_norm": 1.6474583534308007,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 11324
+    },
+    {
+      "epoch": 0.26956583833190517,
+      "grad_norm": 1.5012416119173448,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 11325
+    },
+    {
+      "epoch": 0.26958964105493666,
+      "grad_norm": 6.826581484832515,
+      "learning_rate": 2e-05,
+      "loss": 0.8002,
+      "step": 11326
+    },
+    {
+      "epoch": 0.2696134437779682,
+      "grad_norm": 6.9058928078855635,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 11327
+    },
+    {
+      "epoch": 0.26963724650099974,
+      "grad_norm": 2.725518633062639,
+      "learning_rate": 2e-05,
+      "loss": 0.2964,
+      "step": 11328
+    },
+    {
+      "epoch": 0.2696610492240312,
+      "grad_norm": 4.419572171711557,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 11329
+    },
+    {
+      "epoch": 0.26968485194706276,
+      "grad_norm": 4.162608137425578,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 11330
+    },
+    {
+      "epoch": 0.26970865467009425,
+      "grad_norm": 2.147635331773262,
+      "learning_rate": 2e-05,
+      "loss": 0.1499,
+      "step": 11331
+    },
+    {
+      "epoch": 0.2697324573931258,
+      "grad_norm": 6.733968441277021,
+      "learning_rate": 2e-05,
+      "loss": 0.2606,
+      "step": 11332
+    },
+    {
+      "epoch": 0.2697562601161573,
+      "grad_norm": 2.004720263500333,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 11333
+    },
+    {
+      "epoch": 0.2697800628391888,
+      "grad_norm": 4.076569715274931,
+      "learning_rate": 2e-05,
+      "loss": 0.4225,
+      "step": 11334
+    },
+    {
+      "epoch": 0.2698038655622203,
+      "grad_norm": 3.3578189757558974,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 11335
+    },
+    {
+      "epoch": 0.26982766828525184,
+      "grad_norm": 5.082959952508231,
+      "learning_rate": 2e-05,
+      "loss": 0.8213,
+      "step": 11336
+    },
+    {
+      "epoch": 0.26985147100828333,
+      "grad_norm": 4.026643606125328,
+      "learning_rate": 2e-05,
+      "loss": 0.227,
+      "step": 11337
+    },
+    {
+      "epoch": 0.26987527373131487,
+      "grad_norm": 3.1519225406163995,
+      "learning_rate": 2e-05,
+      "loss": 0.1562,
+      "step": 11338
+    },
+    {
+      "epoch": 0.26989907645434635,
+      "grad_norm": 1.4658272295687116,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 11339
+    },
+    {
+      "epoch": 0.2699228791773779,
+      "grad_norm": 4.900145068399242,
+      "learning_rate": 2e-05,
+      "loss": 0.8041,
+      "step": 11340
+    },
+    {
+      "epoch": 0.2699466819004094,
+      "grad_norm": 4.054397031078805,
+      "learning_rate": 2e-05,
+      "loss": 0.7596,
+      "step": 11341
+    },
+    {
+      "epoch": 0.2699704846234409,
+      "grad_norm": 7.157502306432246,
+      "learning_rate": 2e-05,
+      "loss": 1.2696,
+      "step": 11342
+    },
+    {
+      "epoch": 0.26999428734647246,
+      "grad_norm": 3.4230414804603893,
+      "learning_rate": 2e-05,
+      "loss": 0.2782,
+      "step": 11343
+    },
+    {
+      "epoch": 0.27001809006950395,
+      "grad_norm": 9.464733768757652,
+      "learning_rate": 2e-05,
+      "loss": 1.0293,
+      "step": 11344
+    },
+    {
+      "epoch": 0.2700418927925355,
+      "grad_norm": 6.394293255057449,
+      "learning_rate": 2e-05,
+      "loss": 0.7264,
+      "step": 11345
+    },
+    {
+      "epoch": 0.270065695515567,
+      "grad_norm": 9.563360596665083,
+      "learning_rate": 2e-05,
+      "loss": 1.1158,
+      "step": 11346
+    },
+    {
+      "epoch": 0.2700894982385985,
+      "grad_norm": 2.138214032604963,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 11347
+    },
+    {
+      "epoch": 0.27011330096163,
+      "grad_norm": 0.6932473023429085,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 11348
+    },
+    {
+      "epoch": 0.27013710368466154,
+      "grad_norm": 2.4063270987972887,
+      "learning_rate": 2e-05,
+      "loss": 0.3086,
+      "step": 11349
+    },
+    {
+      "epoch": 0.270160906407693,
+      "grad_norm": 6.139906120603337,
+      "learning_rate": 2e-05,
+      "loss": 0.509,
+      "step": 11350
+    },
+    {
+      "epoch": 0.27018470913072457,
+      "grad_norm": 11.974070685775667,
+      "learning_rate": 2e-05,
+      "loss": 0.3793,
+      "step": 11351
+    },
+    {
+      "epoch": 0.27020851185375605,
+      "grad_norm": 3.5899295194817396,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 11352
+    },
+    {
+      "epoch": 0.2702323145767876,
+      "grad_norm": 1.6746334685457387,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 11353
+    },
+    {
+      "epoch": 0.2702561172998191,
+      "grad_norm": 5.592790973560816,
+      "learning_rate": 2e-05,
+      "loss": 1.3615,
+      "step": 11354
+    },
+    {
+      "epoch": 0.2702799200228506,
+      "grad_norm": 2.949984645371312,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 11355
+    },
+    {
+      "epoch": 0.2703037227458821,
+      "grad_norm": 1.8163278658444735,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 11356
+    },
+    {
+      "epoch": 0.27032752546891364,
+      "grad_norm": 0.9698464704039808,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 11357
+    },
+    {
+      "epoch": 0.2703513281919452,
+      "grad_norm": 3.332588359974717,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 11358
+    },
+    {
+      "epoch": 0.27037513091497667,
+      "grad_norm": 1.5221840609264592,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 11359
+    },
+    {
+      "epoch": 0.2703989336380082,
+      "grad_norm": 1.8794706906479426,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 11360
+    },
+    {
+      "epoch": 0.2704227363610397,
+      "grad_norm": 3.4633618808307736,
+      "learning_rate": 2e-05,
+      "loss": 0.4686,
+      "step": 11361
+    },
+    {
+      "epoch": 0.27044653908407124,
+      "grad_norm": 4.645972053181408,
+      "learning_rate": 2e-05,
+      "loss": 0.7685,
+      "step": 11362
+    },
+    {
+      "epoch": 0.2704703418071027,
+      "grad_norm": 5.232925871883316,
+      "learning_rate": 2e-05,
+      "loss": 0.8456,
+      "step": 11363
+    },
+    {
+      "epoch": 0.27049414453013426,
+      "grad_norm": 2.3545628998458645,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 11364
+    },
+    {
+      "epoch": 0.27051794725316575,
+      "grad_norm": 2.0605922230037135,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 11365
+    },
+    {
+      "epoch": 0.2705417499761973,
+      "grad_norm": 0.6372145368195415,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 11366
+    },
+    {
+      "epoch": 0.2705655526992288,
+      "grad_norm": 35.542756667328845,
+      "learning_rate": 2e-05,
+      "loss": 0.3992,
+      "step": 11367
+    },
+    {
+      "epoch": 0.2705893554222603,
+      "grad_norm": 7.3946490037796435,
+      "learning_rate": 2e-05,
+      "loss": 2.2697,
+      "step": 11368
+    },
+    {
+      "epoch": 0.2706131581452918,
+      "grad_norm": 10.67995986049578,
+      "learning_rate": 2e-05,
+      "loss": 0.8383,
+      "step": 11369
+    },
+    {
+      "epoch": 0.27063696086832334,
+      "grad_norm": 3.6536448565477864,
+      "learning_rate": 2e-05,
+      "loss": 0.4815,
+      "step": 11370
+    },
+    {
+      "epoch": 0.2706607635913548,
+      "grad_norm": 13.780639847002051,
+      "learning_rate": 2e-05,
+      "loss": 1.6072,
+      "step": 11371
+    },
+    {
+      "epoch": 0.27068456631438637,
+      "grad_norm": 5.1343848830945165,
+      "learning_rate": 2e-05,
+      "loss": 0.1824,
+      "step": 11372
+    },
+    {
+      "epoch": 0.2707083690374179,
+      "grad_norm": 3.9538678548929793,
+      "learning_rate": 2e-05,
+      "loss": 0.4832,
+      "step": 11373
+    },
+    {
+      "epoch": 0.2707321717604494,
+      "grad_norm": 2.777194335773363,
+      "learning_rate": 2e-05,
+      "loss": 0.1821,
+      "step": 11374
+    },
+    {
+      "epoch": 0.27075597448348093,
+      "grad_norm": 3.4083089497353454,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 11375
+    },
+    {
+      "epoch": 0.2707797772065124,
+      "grad_norm": 4.271733346063034,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 11376
+    },
+    {
+      "epoch": 0.27080357992954396,
+      "grad_norm": 5.929955437759077,
+      "learning_rate": 2e-05,
+      "loss": 0.8084,
+      "step": 11377
+    },
+    {
+      "epoch": 0.27082738265257544,
+      "grad_norm": 5.393945148282866,
+      "learning_rate": 2e-05,
+      "loss": 0.3231,
+      "step": 11378
+    },
+    {
+      "epoch": 0.270851185375607,
+      "grad_norm": 2.015285818038613,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 11379
+    },
+    {
+      "epoch": 0.27087498809863847,
+      "grad_norm": 1.2885510713900907,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 11380
+    },
+    {
+      "epoch": 0.27089879082167,
+      "grad_norm": 3.770364807446555,
+      "learning_rate": 2e-05,
+      "loss": 0.6153,
+      "step": 11381
+    },
+    {
+      "epoch": 0.2709225935447015,
+      "grad_norm": 6.807582982927736,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 11382
+    },
+    {
+      "epoch": 0.27094639626773304,
+      "grad_norm": 11.71695938611299,
+      "learning_rate": 2e-05,
+      "loss": 0.3346,
+      "step": 11383
+    },
+    {
+      "epoch": 0.2709701989907645,
+      "grad_norm": 2.697840954216315,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 11384
+    },
+    {
+      "epoch": 0.27099400171379606,
+      "grad_norm": 3.3440645388468595,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 11385
+    },
+    {
+      "epoch": 0.27101780443682755,
+      "grad_norm": 2.3828573109407114,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 11386
+    },
+    {
+      "epoch": 0.2710416071598591,
+      "grad_norm": 3.7827900132338006,
+      "learning_rate": 2e-05,
+      "loss": 0.5369,
+      "step": 11387
+    },
+    {
+      "epoch": 0.27106540988289063,
+      "grad_norm": 4.709102428794587,
+      "learning_rate": 2e-05,
+      "loss": 0.633,
+      "step": 11388
+    },
+    {
+      "epoch": 0.2710892126059221,
+      "grad_norm": 10.73487281776204,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 11389
+    },
+    {
+      "epoch": 0.27111301532895365,
+      "grad_norm": 0.6888029353124401,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 11390
+    },
+    {
+      "epoch": 0.27113681805198514,
+      "grad_norm": 4.249417874717239,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 11391
+    },
+    {
+      "epoch": 0.2711606207750167,
+      "grad_norm": 13.364288392435546,
+      "learning_rate": 2e-05,
+      "loss": 0.2281,
+      "step": 11392
+    },
+    {
+      "epoch": 0.27118442349804817,
+      "grad_norm": 3.9632421360625263,
+      "learning_rate": 2e-05,
+      "loss": 0.2482,
+      "step": 11393
+    },
+    {
+      "epoch": 0.2712082262210797,
+      "grad_norm": 1.359575117249579,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 11394
+    },
+    {
+      "epoch": 0.2712320289441112,
+      "grad_norm": 8.826865754473346,
+      "learning_rate": 2e-05,
+      "loss": 0.7275,
+      "step": 11395
+    },
+    {
+      "epoch": 0.27125583166714273,
+      "grad_norm": 6.855866292204377,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 11396
+    },
+    {
+      "epoch": 0.2712796343901742,
+      "grad_norm": 4.686574267288075,
+      "learning_rate": 2e-05,
+      "loss": 0.486,
+      "step": 11397
+    },
+    {
+      "epoch": 0.27130343711320576,
+      "grad_norm": 2.4131505823806085,
+      "learning_rate": 2e-05,
+      "loss": 0.149,
+      "step": 11398
+    },
+    {
+      "epoch": 0.27132723983623724,
+      "grad_norm": 3.361662425372981,
+      "learning_rate": 2e-05,
+      "loss": 0.2823,
+      "step": 11399
+    },
+    {
+      "epoch": 0.2713510425592688,
+      "grad_norm": 1.272891688908244,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27137484528230027,
+      "grad_norm": 13.642234284081853,
+      "learning_rate": 2e-05,
+      "loss": 0.7653,
+      "step": 11401
+    },
+    {
+      "epoch": 0.2713986480053318,
+      "grad_norm": 6.31803257897611,
+      "learning_rate": 2e-05,
+      "loss": 1.926,
+      "step": 11402
+    },
+    {
+      "epoch": 0.27142245072836335,
+      "grad_norm": 4.8709190365411,
+      "learning_rate": 2e-05,
+      "loss": 0.7928,
+      "step": 11403
+    },
+    {
+      "epoch": 0.27144625345139484,
+      "grad_norm": 6.384149891259293,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 11404
+    },
+    {
+      "epoch": 0.2714700561744264,
+      "grad_norm": 6.467934423062599,
+      "learning_rate": 2e-05,
+      "loss": 0.9636,
+      "step": 11405
+    },
+    {
+      "epoch": 0.27149385889745786,
+      "grad_norm": 16.09885837134412,
+      "learning_rate": 2e-05,
+      "loss": 1.4222,
+      "step": 11406
+    },
+    {
+      "epoch": 0.2715176616204894,
+      "grad_norm": 4.490671968172796,
+      "learning_rate": 2e-05,
+      "loss": 0.2255,
+      "step": 11407
+    },
+    {
+      "epoch": 0.2715414643435209,
+      "grad_norm": 18.804891644669475,
+      "learning_rate": 2e-05,
+      "loss": 0.4,
+      "step": 11408
+    },
+    {
+      "epoch": 0.27156526706655243,
+      "grad_norm": 2.4367457215892183,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 11409
+    },
+    {
+      "epoch": 0.2715890697895839,
+      "grad_norm": 2.271598302402077,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 11410
+    },
+    {
+      "epoch": 0.27161287251261546,
+      "grad_norm": 5.827290344911075,
+      "learning_rate": 2e-05,
+      "loss": 0.1555,
+      "step": 11411
+    },
+    {
+      "epoch": 0.27163667523564694,
+      "grad_norm": 9.41958650029281,
+      "learning_rate": 2e-05,
+      "loss": 0.8807,
+      "step": 11412
+    },
+    {
+      "epoch": 0.2716604779586785,
+      "grad_norm": 4.791577448185296,
+      "learning_rate": 2e-05,
+      "loss": 0.8277,
+      "step": 11413
+    },
+    {
+      "epoch": 0.27168428068170997,
+      "grad_norm": 5.895253674520674,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 11414
+    },
+    {
+      "epoch": 0.2717080834047415,
+      "grad_norm": 2.7813210430086257,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 11415
+    },
+    {
+      "epoch": 0.271731886127773,
+      "grad_norm": 3.5382766628423608,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 11416
+    },
+    {
+      "epoch": 0.27175568885080453,
+      "grad_norm": 7.528275979822095,
+      "learning_rate": 2e-05,
+      "loss": 0.2372,
+      "step": 11417
+    },
+    {
+      "epoch": 0.2717794915738361,
+      "grad_norm": 4.313755486640671,
+      "learning_rate": 2e-05,
+      "loss": 0.251,
+      "step": 11418
+    },
+    {
+      "epoch": 0.27180329429686756,
+      "grad_norm": 14.550848017091958,
+      "learning_rate": 2e-05,
+      "loss": 0.6079,
+      "step": 11419
+    },
+    {
+      "epoch": 0.2718270970198991,
+      "grad_norm": 2.6799054160850972,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 11420
+    },
+    {
+      "epoch": 0.2718508997429306,
+      "grad_norm": 2.8112123598609973,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 11421
+    },
+    {
+      "epoch": 0.2718747024659621,
+      "grad_norm": 3.918909424303229,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 11422
+    },
+    {
+      "epoch": 0.2718985051889936,
+      "grad_norm": 4.478158359095508,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 11423
+    },
+    {
+      "epoch": 0.27192230791202515,
+      "grad_norm": 6.1905592780513565,
+      "learning_rate": 2e-05,
+      "loss": 0.4465,
+      "step": 11424
+    },
+    {
+      "epoch": 0.27194611063505664,
+      "grad_norm": 5.488655433379848,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 11425
+    },
+    {
+      "epoch": 0.2719699133580882,
+      "grad_norm": 3.090462300577356,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 11426
+    },
+    {
+      "epoch": 0.27199371608111966,
+      "grad_norm": 4.892374245362692,
+      "learning_rate": 2e-05,
+      "loss": 0.6005,
+      "step": 11427
+    },
+    {
+      "epoch": 0.2720175188041512,
+      "grad_norm": 0.5821008633763113,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 11428
+    },
+    {
+      "epoch": 0.2720413215271827,
+      "grad_norm": 13.51750244582911,
+      "learning_rate": 2e-05,
+      "loss": 0.6032,
+      "step": 11429
+    },
+    {
+      "epoch": 0.27206512425021423,
+      "grad_norm": 1.8643732163894111,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 11430
+    },
+    {
+      "epoch": 0.2720889269732457,
+      "grad_norm": 2.215659443677423,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 11431
+    },
+    {
+      "epoch": 0.27211272969627726,
+      "grad_norm": 5.254324613462806,
+      "learning_rate": 2e-05,
+      "loss": 0.4595,
+      "step": 11432
+    },
+    {
+      "epoch": 0.2721365324193088,
+      "grad_norm": 0.735079983711632,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 11433
+    },
+    {
+      "epoch": 0.2721603351423403,
+      "grad_norm": 2.541451755699795,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 11434
+    },
+    {
+      "epoch": 0.2721841378653718,
+      "grad_norm": 13.32017316155887,
+      "learning_rate": 2e-05,
+      "loss": 0.1705,
+      "step": 11435
+    },
+    {
+      "epoch": 0.2722079405884033,
+      "grad_norm": 14.730827508989943,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 11436
+    },
+    {
+      "epoch": 0.27223174331143485,
+      "grad_norm": 4.640029286934781,
+      "learning_rate": 2e-05,
+      "loss": 0.3099,
+      "step": 11437
+    },
+    {
+      "epoch": 0.27225554603446633,
+      "grad_norm": 1.030772746184593,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 11438
+    },
+    {
+      "epoch": 0.2722793487574979,
+      "grad_norm": 7.276467766484152,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 11439
+    },
+    {
+      "epoch": 0.27230315148052936,
+      "grad_norm": 2.805489234477603,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 11440
+    },
+    {
+      "epoch": 0.2723269542035609,
+      "grad_norm": 2.763864510787467,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 11441
+    },
+    {
+      "epoch": 0.2723507569265924,
+      "grad_norm": 8.491316857284891,
+      "learning_rate": 2e-05,
+      "loss": 1.8845,
+      "step": 11442
+    },
+    {
+      "epoch": 0.2723745596496239,
+      "grad_norm": 11.535742594947104,
+      "learning_rate": 2e-05,
+      "loss": 0.1819,
+      "step": 11443
+    },
+    {
+      "epoch": 0.2723983623726554,
+      "grad_norm": 3.0225780454753775,
+      "learning_rate": 2e-05,
+      "loss": 0.1583,
+      "step": 11444
+    },
+    {
+      "epoch": 0.27242216509568695,
+      "grad_norm": 3.150964688925141,
+      "learning_rate": 2e-05,
+      "loss": 0.2266,
+      "step": 11445
+    },
+    {
+      "epoch": 0.27244596781871844,
+      "grad_norm": 6.085823753527256,
+      "learning_rate": 2e-05,
+      "loss": 0.7294,
+      "step": 11446
+    },
+    {
+      "epoch": 0.27246977054175,
+      "grad_norm": 4.272114816842004,
+      "learning_rate": 2e-05,
+      "loss": 0.245,
+      "step": 11447
+    },
+    {
+      "epoch": 0.27249357326478146,
+      "grad_norm": 4.712746943574151,
+      "learning_rate": 2e-05,
+      "loss": 0.2781,
+      "step": 11448
+    },
+    {
+      "epoch": 0.272517375987813,
+      "grad_norm": 6.959912015511363,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 11449
+    },
+    {
+      "epoch": 0.27254117871084454,
+      "grad_norm": 3.387381047970399,
+      "learning_rate": 2e-05,
+      "loss": 0.2709,
+      "step": 11450
+    },
+    {
+      "epoch": 0.27256498143387603,
+      "grad_norm": 10.970328305573881,
+      "learning_rate": 2e-05,
+      "loss": 1.766,
+      "step": 11451
+    },
+    {
+      "epoch": 0.27258878415690757,
+      "grad_norm": 5.498098502050286,
+      "learning_rate": 2e-05,
+      "loss": 1.1394,
+      "step": 11452
+    },
+    {
+      "epoch": 0.27261258687993906,
+      "grad_norm": 2.5205364167303093,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 11453
+    },
+    {
+      "epoch": 0.2726363896029706,
+      "grad_norm": 1.2333584971782496,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 11454
+    },
+    {
+      "epoch": 0.2726601923260021,
+      "grad_norm": 1.0424036490433002,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 11455
+    },
+    {
+      "epoch": 0.2726839950490336,
+      "grad_norm": 1.3928726714865918,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 11456
+    },
+    {
+      "epoch": 0.2727077977720651,
+      "grad_norm": 4.31240742330187,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 11457
+    },
+    {
+      "epoch": 0.27273160049509665,
+      "grad_norm": 25.624254276689637,
+      "learning_rate": 2e-05,
+      "loss": 1.0586,
+      "step": 11458
+    },
+    {
+      "epoch": 0.27275540321812813,
+      "grad_norm": 4.573343628161526,
+      "learning_rate": 2e-05,
+      "loss": 0.7023,
+      "step": 11459
+    },
+    {
+      "epoch": 0.2727792059411597,
+      "grad_norm": 3.759355837594755,
+      "learning_rate": 2e-05,
+      "loss": 0.5019,
+      "step": 11460
+    },
+    {
+      "epoch": 0.27280300866419116,
+      "grad_norm": 5.991048097932054,
+      "learning_rate": 2e-05,
+      "loss": 1.2225,
+      "step": 11461
+    },
+    {
+      "epoch": 0.2728268113872227,
+      "grad_norm": 4.698321426546395,
+      "learning_rate": 2e-05,
+      "loss": 0.135,
+      "step": 11462
+    },
+    {
+      "epoch": 0.2728506141102542,
+      "grad_norm": 3.682978159453558,
+      "learning_rate": 2e-05,
+      "loss": 0.132,
+      "step": 11463
+    },
+    {
+      "epoch": 0.2728744168332857,
+      "grad_norm": 2.9626054943855933,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 11464
+    },
+    {
+      "epoch": 0.27289821955631727,
+      "grad_norm": 0.6136900620467718,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 11465
+    },
+    {
+      "epoch": 0.27292202227934875,
+      "grad_norm": 1.852504346342522,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 11466
+    },
+    {
+      "epoch": 0.2729458250023803,
+      "grad_norm": 3.5215638409667314,
+      "learning_rate": 2e-05,
+      "loss": 0.1984,
+      "step": 11467
+    },
+    {
+      "epoch": 0.2729696277254118,
+      "grad_norm": 0.8775345967396341,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 11468
+    },
+    {
+      "epoch": 0.2729934304484433,
+      "grad_norm": 3.852630144765092,
+      "learning_rate": 2e-05,
+      "loss": 0.5301,
+      "step": 11469
+    },
+    {
+      "epoch": 0.2730172331714748,
+      "grad_norm": 3.8088741263754424,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 11470
+    },
+    {
+      "epoch": 0.27304103589450635,
+      "grad_norm": 2.1365745367183706,
+      "learning_rate": 2e-05,
+      "loss": 0.1745,
+      "step": 11471
+    },
+    {
+      "epoch": 0.27306483861753783,
+      "grad_norm": 0.8606368647489038,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 11472
+    },
+    {
+      "epoch": 0.27308864134056937,
+      "grad_norm": 6.100388300809603,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 11473
+    },
+    {
+      "epoch": 0.27311244406360086,
+      "grad_norm": 8.852865466476027,
+      "learning_rate": 2e-05,
+      "loss": 0.451,
+      "step": 11474
+    },
+    {
+      "epoch": 0.2731362467866324,
+      "grad_norm": 11.683361796130072,
+      "learning_rate": 2e-05,
+      "loss": 0.6061,
+      "step": 11475
+    },
+    {
+      "epoch": 0.2731600495096639,
+      "grad_norm": 4.492102974185642,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 11476
+    },
+    {
+      "epoch": 0.2731838522326954,
+      "grad_norm": 0.09656879874540497,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 11477
+    },
+    {
+      "epoch": 0.2732076549557269,
+      "grad_norm": 4.073315915421588,
+      "learning_rate": 2e-05,
+      "loss": 0.1431,
+      "step": 11478
+    },
+    {
+      "epoch": 0.27323145767875845,
+      "grad_norm": 6.6451179581312525,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 11479
+    },
+    {
+      "epoch": 0.27325526040179,
+      "grad_norm": 2.1632619223027665,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 11480
+    },
+    {
+      "epoch": 0.2732790631248215,
+      "grad_norm": 15.005357011619548,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 11481
+    },
+    {
+      "epoch": 0.273302865847853,
+      "grad_norm": 0.2287096876081832,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 11482
+    },
+    {
+      "epoch": 0.2733266685708845,
+      "grad_norm": 6.657086520536774,
+      "learning_rate": 2e-05,
+      "loss": 1.3721,
+      "step": 11483
+    },
+    {
+      "epoch": 0.27335047129391604,
+      "grad_norm": 9.356672239225864,
+      "learning_rate": 2e-05,
+      "loss": 1.7547,
+      "step": 11484
+    },
+    {
+      "epoch": 0.2733742740169475,
+      "grad_norm": 2.680638980941103,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 11485
+    },
+    {
+      "epoch": 0.27339807673997907,
+      "grad_norm": 5.35428261183439,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 11486
+    },
+    {
+      "epoch": 0.27342187946301055,
+      "grad_norm": 4.77828597073729,
+      "learning_rate": 2e-05,
+      "loss": 0.7711,
+      "step": 11487
+    },
+    {
+      "epoch": 0.2734456821860421,
+      "grad_norm": 4.447753662048106,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 11488
+    },
+    {
+      "epoch": 0.2734694849090736,
+      "grad_norm": 6.525957441396254,
+      "learning_rate": 2e-05,
+      "loss": 0.5608,
+      "step": 11489
+    },
+    {
+      "epoch": 0.2734932876321051,
+      "grad_norm": 3.7918375835071205,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 11490
+    },
+    {
+      "epoch": 0.2735170903551366,
+      "grad_norm": 1.3297706685442727,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 11491
+    },
+    {
+      "epoch": 0.27354089307816815,
+      "grad_norm": 4.036075832758715,
+      "learning_rate": 2e-05,
+      "loss": 0.5369,
+      "step": 11492
+    },
+    {
+      "epoch": 0.27356469580119963,
+      "grad_norm": 6.429897213335889,
+      "learning_rate": 2e-05,
+      "loss": 1.4262,
+      "step": 11493
+    },
+    {
+      "epoch": 0.27358849852423117,
+      "grad_norm": 11.529387326464514,
+      "learning_rate": 2e-05,
+      "loss": 1.0887,
+      "step": 11494
+    },
+    {
+      "epoch": 0.2736123012472627,
+      "grad_norm": 5.803169984721314,
+      "learning_rate": 2e-05,
+      "loss": 0.2141,
+      "step": 11495
+    },
+    {
+      "epoch": 0.2736361039702942,
+      "grad_norm": 11.460535685269766,
+      "learning_rate": 2e-05,
+      "loss": 1.2623,
+      "step": 11496
+    },
+    {
+      "epoch": 0.27365990669332574,
+      "grad_norm": 5.7489715762148395,
+      "learning_rate": 2e-05,
+      "loss": 0.4002,
+      "step": 11497
+    },
+    {
+      "epoch": 0.2736837094163572,
+      "grad_norm": 5.260351561575411,
+      "learning_rate": 2e-05,
+      "loss": 0.6295,
+      "step": 11498
+    },
+    {
+      "epoch": 0.27370751213938876,
+      "grad_norm": 12.549156401240412,
+      "learning_rate": 2e-05,
+      "loss": 0.4855,
+      "step": 11499
+    },
+    {
+      "epoch": 0.27373131486242025,
+      "grad_norm": 7.179019734916514,
+      "learning_rate": 2e-05,
+      "loss": 1.2939,
+      "step": 11500
+    },
+    {
+      "epoch": 0.2737551175854518,
+      "grad_norm": 10.684245672962836,
+      "learning_rate": 2e-05,
+      "loss": 0.804,
+      "step": 11501
+    },
+    {
+      "epoch": 0.2737789203084833,
+      "grad_norm": 6.052057839965208,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 11502
+    },
+    {
+      "epoch": 0.2738027230315148,
+      "grad_norm": 3.3201861443446763,
+      "learning_rate": 2e-05,
+      "loss": 0.4825,
+      "step": 11503
+    },
+    {
+      "epoch": 0.2738265257545463,
+      "grad_norm": 3.2640358900587008,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 11504
+    },
+    {
+      "epoch": 0.27385032847757784,
+      "grad_norm": 3.6327477394801577,
+      "learning_rate": 2e-05,
+      "loss": 0.3376,
+      "step": 11505
+    },
+    {
+      "epoch": 0.2738741312006093,
+      "grad_norm": 5.467508300642792,
+      "learning_rate": 2e-05,
+      "loss": 0.5156,
+      "step": 11506
+    },
+    {
+      "epoch": 0.27389793392364087,
+      "grad_norm": 3.4506172499884062,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 11507
+    },
+    {
+      "epoch": 0.27392173664667235,
+      "grad_norm": 15.311318249745362,
+      "learning_rate": 2e-05,
+      "loss": 0.2549,
+      "step": 11508
+    },
+    {
+      "epoch": 0.2739455393697039,
+      "grad_norm": 5.740185987274512,
+      "learning_rate": 2e-05,
+      "loss": 0.3475,
+      "step": 11509
+    },
+    {
+      "epoch": 0.27396934209273544,
+      "grad_norm": 4.0478138194276205,
+      "learning_rate": 2e-05,
+      "loss": 0.7091,
+      "step": 11510
+    },
+    {
+      "epoch": 0.2739931448157669,
+      "grad_norm": 3.6910353351426193,
+      "learning_rate": 2e-05,
+      "loss": 0.3904,
+      "step": 11511
+    },
+    {
+      "epoch": 0.27401694753879846,
+      "grad_norm": 10.103972798783015,
+      "learning_rate": 2e-05,
+      "loss": 1.0568,
+      "step": 11512
+    },
+    {
+      "epoch": 0.27404075026182995,
+      "grad_norm": 4.389944208299323,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 11513
+    },
+    {
+      "epoch": 0.2740645529848615,
+      "grad_norm": 4.874157430824293,
+      "learning_rate": 2e-05,
+      "loss": 0.3209,
+      "step": 11514
+    },
+    {
+      "epoch": 0.27408835570789297,
+      "grad_norm": 4.533511579908403,
+      "learning_rate": 2e-05,
+      "loss": 0.667,
+      "step": 11515
+    },
+    {
+      "epoch": 0.2741121584309245,
+      "grad_norm": 13.274093052059548,
+      "learning_rate": 2e-05,
+      "loss": 1.1027,
+      "step": 11516
+    },
+    {
+      "epoch": 0.274135961153956,
+      "grad_norm": 2.6922843139414754,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 11517
+    },
+    {
+      "epoch": 0.27415976387698754,
+      "grad_norm": 8.0007231142337,
+      "learning_rate": 2e-05,
+      "loss": 1.6401,
+      "step": 11518
+    },
+    {
+      "epoch": 0.274183566600019,
+      "grad_norm": 1.8471444706220908,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 11519
+    },
+    {
+      "epoch": 0.27420736932305056,
+      "grad_norm": 3.8484202219779,
+      "learning_rate": 2e-05,
+      "loss": 0.2834,
+      "step": 11520
+    },
+    {
+      "epoch": 0.27423117204608205,
+      "grad_norm": 1.9951704536325472,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 11521
+    },
+    {
+      "epoch": 0.2742549747691136,
+      "grad_norm": 2.9512879071056686,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 11522
+    },
+    {
+      "epoch": 0.2742787774921451,
+      "grad_norm": 5.516272178714555,
+      "learning_rate": 2e-05,
+      "loss": 0.2423,
+      "step": 11523
+    },
+    {
+      "epoch": 0.2743025802151766,
+      "grad_norm": 2.880144352962447,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 11524
+    },
+    {
+      "epoch": 0.27432638293820816,
+      "grad_norm": 2.447133420083511,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 11525
+    },
+    {
+      "epoch": 0.27435018566123964,
+      "grad_norm": 4.430857870489303,
+      "learning_rate": 2e-05,
+      "loss": 0.9603,
+      "step": 11526
+    },
+    {
+      "epoch": 0.2743739883842712,
+      "grad_norm": 4.586134284156631,
+      "learning_rate": 2e-05,
+      "loss": 0.1088,
+      "step": 11527
+    },
+    {
+      "epoch": 0.27439779110730267,
+      "grad_norm": 3.6876233693192915,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 11528
+    },
+    {
+      "epoch": 0.2744215938303342,
+      "grad_norm": 3.1621554938241636,
+      "learning_rate": 2e-05,
+      "loss": 0.126,
+      "step": 11529
+    },
+    {
+      "epoch": 0.2744453965533657,
+      "grad_norm": 3.3976376428241046,
+      "learning_rate": 2e-05,
+      "loss": 0.4329,
+      "step": 11530
+    },
+    {
+      "epoch": 0.27446919927639724,
+      "grad_norm": 2.236141040238639,
+      "learning_rate": 2e-05,
+      "loss": 0.2119,
+      "step": 11531
+    },
+    {
+      "epoch": 0.2744930019994287,
+      "grad_norm": 0.8547847550407638,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 11532
+    },
+    {
+      "epoch": 0.27451680472246026,
+      "grad_norm": 7.0978904770517355,
+      "learning_rate": 2e-05,
+      "loss": 0.2773,
+      "step": 11533
+    },
+    {
+      "epoch": 0.27454060744549175,
+      "grad_norm": 3.2775789714093495,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 11534
+    },
+    {
+      "epoch": 0.2745644101685233,
+      "grad_norm": 3.794045444255127,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 11535
+    },
+    {
+      "epoch": 0.2745882128915548,
+      "grad_norm": 3.2035953561217743,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 11536
+    },
+    {
+      "epoch": 0.2746120156145863,
+      "grad_norm": 10.213929687368266,
+      "learning_rate": 2e-05,
+      "loss": 0.7106,
+      "step": 11537
+    },
+    {
+      "epoch": 0.2746358183376178,
+      "grad_norm": 2.3799560142961544,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 11538
+    },
+    {
+      "epoch": 0.27465962106064934,
+      "grad_norm": 9.22853287553411,
+      "learning_rate": 2e-05,
+      "loss": 0.9952,
+      "step": 11539
+    },
+    {
+      "epoch": 0.2746834237836809,
+      "grad_norm": 6.734404341658045,
+      "learning_rate": 2e-05,
+      "loss": 0.6959,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27470722650671237,
+      "grad_norm": 4.9060256792504715,
+      "learning_rate": 2e-05,
+      "loss": 0.6933,
+      "step": 11541
+    },
+    {
+      "epoch": 0.2747310292297439,
+      "grad_norm": 3.358592369620241,
+      "learning_rate": 2e-05,
+      "loss": 0.172,
+      "step": 11542
+    },
+    {
+      "epoch": 0.2747548319527754,
+      "grad_norm": 24.80771517101313,
+      "learning_rate": 2e-05,
+      "loss": 0.5321,
+      "step": 11543
+    },
+    {
+      "epoch": 0.27477863467580693,
+      "grad_norm": 3.5772743031357197,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 11544
+    },
+    {
+      "epoch": 0.2748024373988384,
+      "grad_norm": 4.0614226879691895,
+      "learning_rate": 2e-05,
+      "loss": 0.6193,
+      "step": 11545
+    },
+    {
+      "epoch": 0.27482624012186996,
+      "grad_norm": 4.215616811285998,
+      "learning_rate": 2e-05,
+      "loss": 0.718,
+      "step": 11546
+    },
+    {
+      "epoch": 0.27485004284490144,
+      "grad_norm": 3.980441041794567,
+      "learning_rate": 2e-05,
+      "loss": 0.4422,
+      "step": 11547
+    },
+    {
+      "epoch": 0.274873845567933,
+      "grad_norm": 2.022342640526666,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 11548
+    },
+    {
+      "epoch": 0.27489764829096447,
+      "grad_norm": 3.2441569451310555,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 11549
+    },
+    {
+      "epoch": 0.274921451013996,
+      "grad_norm": 10.469955147999245,
+      "learning_rate": 2e-05,
+      "loss": 0.5406,
+      "step": 11550
+    },
+    {
+      "epoch": 0.2749452537370275,
+      "grad_norm": 1.3417986494958019,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 11551
+    },
+    {
+      "epoch": 0.27496905646005904,
+      "grad_norm": 3.643493624126023,
+      "learning_rate": 2e-05,
+      "loss": 0.6882,
+      "step": 11552
+    },
+    {
+      "epoch": 0.2749928591830905,
+      "grad_norm": 7.4874553531005,
+      "learning_rate": 2e-05,
+      "loss": 0.8696,
+      "step": 11553
+    },
+    {
+      "epoch": 0.27501666190612206,
+      "grad_norm": 1.8643348562318112,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 11554
+    },
+    {
+      "epoch": 0.2750404646291536,
+      "grad_norm": 5.335016067585397,
+      "learning_rate": 2e-05,
+      "loss": 0.6942,
+      "step": 11555
+    },
+    {
+      "epoch": 0.2750642673521851,
+      "grad_norm": 5.0867444919536275,
+      "learning_rate": 2e-05,
+      "loss": 0.6286,
+      "step": 11556
+    },
+    {
+      "epoch": 0.27508807007521663,
+      "grad_norm": 6.16574563218399,
+      "learning_rate": 2e-05,
+      "loss": 1.5614,
+      "step": 11557
+    },
+    {
+      "epoch": 0.2751118727982481,
+      "grad_norm": 2.140248947286559,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 11558
+    },
+    {
+      "epoch": 0.27513567552127965,
+      "grad_norm": 3.639366434096883,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 11559
+    },
+    {
+      "epoch": 0.27515947824431114,
+      "grad_norm": 6.047192139674037,
+      "learning_rate": 2e-05,
+      "loss": 0.7316,
+      "step": 11560
+    },
+    {
+      "epoch": 0.2751832809673427,
+      "grad_norm": 1.5557880183677395,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 11561
+    },
+    {
+      "epoch": 0.27520708369037417,
+      "grad_norm": 2.7641209293910243,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 11562
+    },
+    {
+      "epoch": 0.2752308864134057,
+      "grad_norm": 0.9422621766404312,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 11563
+    },
+    {
+      "epoch": 0.2752546891364372,
+      "grad_norm": 1.1122076552725773,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 11564
+    },
+    {
+      "epoch": 0.27527849185946873,
+      "grad_norm": 5.971517592584747,
+      "learning_rate": 2e-05,
+      "loss": 1.2295,
+      "step": 11565
+    },
+    {
+      "epoch": 0.2753022945825002,
+      "grad_norm": 3.291834742268886,
+      "learning_rate": 2e-05,
+      "loss": 0.2068,
+      "step": 11566
+    },
+    {
+      "epoch": 0.27532609730553176,
+      "grad_norm": 1.4253332060599966,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 11567
+    },
+    {
+      "epoch": 0.27534990002856324,
+      "grad_norm": 13.999409085608642,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 11568
+    },
+    {
+      "epoch": 0.2753737027515948,
+      "grad_norm": 2.3379339045535885,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 11569
+    },
+    {
+      "epoch": 0.27539750547462627,
+      "grad_norm": 9.557621430995283,
+      "learning_rate": 2e-05,
+      "loss": 1.3417,
+      "step": 11570
+    },
+    {
+      "epoch": 0.2754213081976578,
+      "grad_norm": 14.019564110692503,
+      "learning_rate": 2e-05,
+      "loss": 0.8225,
+      "step": 11571
+    },
+    {
+      "epoch": 0.27544511092068935,
+      "grad_norm": 3.637239752355365,
+      "learning_rate": 2e-05,
+      "loss": 0.1289,
+      "step": 11572
+    },
+    {
+      "epoch": 0.27546891364372084,
+      "grad_norm": 5.067699682599819,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 11573
+    },
+    {
+      "epoch": 0.2754927163667524,
+      "grad_norm": 2.1511186748719755,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 11574
+    },
+    {
+      "epoch": 0.27551651908978386,
+      "grad_norm": 8.500497180958453,
+      "learning_rate": 2e-05,
+      "loss": 1.1288,
+      "step": 11575
+    },
+    {
+      "epoch": 0.2755403218128154,
+      "grad_norm": 9.249930860964374,
+      "learning_rate": 2e-05,
+      "loss": 0.2352,
+      "step": 11576
+    },
+    {
+      "epoch": 0.2755641245358469,
+      "grad_norm": 8.540558117456413,
+      "learning_rate": 2e-05,
+      "loss": 1.5359,
+      "step": 11577
+    },
+    {
+      "epoch": 0.27558792725887843,
+      "grad_norm": 11.100017884480241,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 11578
+    },
+    {
+      "epoch": 0.2756117299819099,
+      "grad_norm": 3.32843130585733,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 11579
+    },
+    {
+      "epoch": 0.27563553270494145,
+      "grad_norm": 2.9878738743819993,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 11580
+    },
+    {
+      "epoch": 0.27565933542797294,
+      "grad_norm": 2.2933882798478504,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 11581
+    },
+    {
+      "epoch": 0.2756831381510045,
+      "grad_norm": 7.938438453821813,
+      "learning_rate": 2e-05,
+      "loss": 1.0548,
+      "step": 11582
+    },
+    {
+      "epoch": 0.27570694087403597,
+      "grad_norm": 16.819896192621037,
+      "learning_rate": 2e-05,
+      "loss": 1.1042,
+      "step": 11583
+    },
+    {
+      "epoch": 0.2757307435970675,
+      "grad_norm": 7.11755121214912,
+      "learning_rate": 2e-05,
+      "loss": 0.2796,
+      "step": 11584
+    },
+    {
+      "epoch": 0.275754546320099,
+      "grad_norm": 2.6816165650281167,
+      "learning_rate": 2e-05,
+      "loss": 0.1475,
+      "step": 11585
+    },
+    {
+      "epoch": 0.27577834904313053,
+      "grad_norm": 1.7190575122129845,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 11586
+    },
+    {
+      "epoch": 0.2758021517661621,
+      "grad_norm": 6.335552316901511,
+      "learning_rate": 2e-05,
+      "loss": 0.3622,
+      "step": 11587
+    },
+    {
+      "epoch": 0.27582595448919356,
+      "grad_norm": 10.136894192359453,
+      "learning_rate": 2e-05,
+      "loss": 0.4574,
+      "step": 11588
+    },
+    {
+      "epoch": 0.2758497572122251,
+      "grad_norm": 0.3595068236224715,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 11589
+    },
+    {
+      "epoch": 0.2758735599352566,
+      "grad_norm": 4.056884055436607,
+      "learning_rate": 2e-05,
+      "loss": 0.6178,
+      "step": 11590
+    },
+    {
+      "epoch": 0.2758973626582881,
+      "grad_norm": 3.9213809645168514,
+      "learning_rate": 2e-05,
+      "loss": 0.1996,
+      "step": 11591
+    },
+    {
+      "epoch": 0.2759211653813196,
+      "grad_norm": 12.016741740644777,
+      "learning_rate": 2e-05,
+      "loss": 2.3003,
+      "step": 11592
+    },
+    {
+      "epoch": 0.27594496810435115,
+      "grad_norm": 5.946173073261606,
+      "learning_rate": 2e-05,
+      "loss": 0.6927,
+      "step": 11593
+    },
+    {
+      "epoch": 0.27596877082738264,
+      "grad_norm": 4.378463877984905,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 11594
+    },
+    {
+      "epoch": 0.2759925735504142,
+      "grad_norm": 4.238521479334295,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 11595
+    },
+    {
+      "epoch": 0.27601637627344566,
+      "grad_norm": 4.270990646528466,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 11596
+    },
+    {
+      "epoch": 0.2760401789964772,
+      "grad_norm": 1.9838790063993428,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 11597
+    },
+    {
+      "epoch": 0.2760639817195087,
+      "grad_norm": 2.188591128158693,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 11598
+    },
+    {
+      "epoch": 0.27608778444254023,
+      "grad_norm": 4.123318892257099,
+      "learning_rate": 2e-05,
+      "loss": 0.4632,
+      "step": 11599
+    },
+    {
+      "epoch": 0.2761115871655717,
+      "grad_norm": 4.0758579297470865,
+      "learning_rate": 2e-05,
+      "loss": 0.674,
+      "step": 11600
+    },
+    {
+      "epoch": 0.27613538988860326,
+      "grad_norm": 3.952316359272771,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 11601
+    },
+    {
+      "epoch": 0.2761591926116348,
+      "grad_norm": 3.4286038798820213,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 11602
+    },
+    {
+      "epoch": 0.2761829953346663,
+      "grad_norm": 8.513945503386262,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 11603
+    },
+    {
+      "epoch": 0.2762067980576978,
+      "grad_norm": 3.829507170697135,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 11604
+    },
+    {
+      "epoch": 0.2762306007807293,
+      "grad_norm": 4.1232394925851,
+      "learning_rate": 2e-05,
+      "loss": 0.7243,
+      "step": 11605
+    },
+    {
+      "epoch": 0.27625440350376085,
+      "grad_norm": 1.903228618293793,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 11606
+    },
+    {
+      "epoch": 0.27627820622679233,
+      "grad_norm": 2.5183372254586023,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 11607
+    },
+    {
+      "epoch": 0.2763020089498239,
+      "grad_norm": 2.970055629833478,
+      "learning_rate": 2e-05,
+      "loss": 0.4567,
+      "step": 11608
+    },
+    {
+      "epoch": 0.27632581167285536,
+      "grad_norm": 6.601534693536326,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 11609
+    },
+    {
+      "epoch": 0.2763496143958869,
+      "grad_norm": 8.73799815523429,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 11610
+    },
+    {
+      "epoch": 0.2763734171189184,
+      "grad_norm": 5.550641400502647,
+      "learning_rate": 2e-05,
+      "loss": 1.3499,
+      "step": 11611
+    },
+    {
+      "epoch": 0.2763972198419499,
+      "grad_norm": 1.5820603675865883,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 11612
+    },
+    {
+      "epoch": 0.2764210225649814,
+      "grad_norm": 4.702260459348747,
+      "learning_rate": 2e-05,
+      "loss": 0.7969,
+      "step": 11613
+    },
+    {
+      "epoch": 0.27644482528801295,
+      "grad_norm": 3.165495934727525,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 11614
+    },
+    {
+      "epoch": 0.27646862801104444,
+      "grad_norm": 5.477297227223605,
+      "learning_rate": 2e-05,
+      "loss": 0.1924,
+      "step": 11615
+    },
+    {
+      "epoch": 0.276492430734076,
+      "grad_norm": 16.889579727506565,
+      "learning_rate": 2e-05,
+      "loss": 0.6743,
+      "step": 11616
+    },
+    {
+      "epoch": 0.2765162334571075,
+      "grad_norm": 10.815989092262067,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 11617
+    },
+    {
+      "epoch": 0.276540036180139,
+      "grad_norm": 4.78608689878006,
+      "learning_rate": 2e-05,
+      "loss": 0.7646,
+      "step": 11618
+    },
+    {
+      "epoch": 0.27656383890317054,
+      "grad_norm": 5.954368584073014,
+      "learning_rate": 2e-05,
+      "loss": 0.2822,
+      "step": 11619
+    },
+    {
+      "epoch": 0.27658764162620203,
+      "grad_norm": 6.499374275766682,
+      "learning_rate": 2e-05,
+      "loss": 0.6805,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27661144434923357,
+      "grad_norm": 2.4493751826731414,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 11621
+    },
+    {
+      "epoch": 0.27663524707226506,
+      "grad_norm": 15.262140576975296,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 11622
+    },
+    {
+      "epoch": 0.2766590497952966,
+      "grad_norm": 2.657020974145134,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 11623
+    },
+    {
+      "epoch": 0.2766828525183281,
+      "grad_norm": 1.6480522209378934,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 11624
+    },
+    {
+      "epoch": 0.2767066552413596,
+      "grad_norm": 13.613915472652913,
+      "learning_rate": 2e-05,
+      "loss": 0.7944,
+      "step": 11625
+    },
+    {
+      "epoch": 0.2767304579643911,
+      "grad_norm": 2.949521947904022,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 11626
+    },
+    {
+      "epoch": 0.27675426068742265,
+      "grad_norm": 5.606744813266148,
+      "learning_rate": 2e-05,
+      "loss": 0.7102,
+      "step": 11627
+    },
+    {
+      "epoch": 0.27677806341045413,
+      "grad_norm": 19.531615564719672,
+      "learning_rate": 2e-05,
+      "loss": 0.4647,
+      "step": 11628
+    },
+    {
+      "epoch": 0.2768018661334857,
+      "grad_norm": 2.9954466077953366,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 11629
+    },
+    {
+      "epoch": 0.27682566885651716,
+      "grad_norm": 2.8925007240376344,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 11630
+    },
+    {
+      "epoch": 0.2768494715795487,
+      "grad_norm": 14.881033629202797,
+      "learning_rate": 2e-05,
+      "loss": 0.4445,
+      "step": 11631
+    },
+    {
+      "epoch": 0.27687327430258024,
+      "grad_norm": 5.8535279179432465,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 11632
+    },
+    {
+      "epoch": 0.2768970770256117,
+      "grad_norm": 8.255609164961616,
+      "learning_rate": 2e-05,
+      "loss": 0.3912,
+      "step": 11633
+    },
+    {
+      "epoch": 0.27692087974864327,
+      "grad_norm": 6.9703462704616586,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 11634
+    },
+    {
+      "epoch": 0.27694468247167475,
+      "grad_norm": 7.063095712319218,
+      "learning_rate": 2e-05,
+      "loss": 1.8546,
+      "step": 11635
+    },
+    {
+      "epoch": 0.2769684851947063,
+      "grad_norm": 20.284375819745353,
+      "learning_rate": 2e-05,
+      "loss": 0.3964,
+      "step": 11636
+    },
+    {
+      "epoch": 0.2769922879177378,
+      "grad_norm": 2.6677763334074713,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 11637
+    },
+    {
+      "epoch": 0.2770160906407693,
+      "grad_norm": 7.521411386778742,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 11638
+    },
+    {
+      "epoch": 0.2770398933638008,
+      "grad_norm": 3.326756406755832,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 11639
+    },
+    {
+      "epoch": 0.27706369608683235,
+      "grad_norm": 9.511053006037328,
+      "learning_rate": 2e-05,
+      "loss": 0.3545,
+      "step": 11640
+    },
+    {
+      "epoch": 0.27708749880986383,
+      "grad_norm": 2.701115018681462,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 11641
+    },
+    {
+      "epoch": 0.27711130153289537,
+      "grad_norm": 5.141957531729055,
+      "learning_rate": 2e-05,
+      "loss": 0.8944,
+      "step": 11642
+    },
+    {
+      "epoch": 0.27713510425592686,
+      "grad_norm": 5.64850695395008,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 11643
+    },
+    {
+      "epoch": 0.2771589069789584,
+      "grad_norm": 1.769221373291094,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 11644
+    },
+    {
+      "epoch": 0.2771827097019899,
+      "grad_norm": 3.3149634027817387,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 11645
+    },
+    {
+      "epoch": 0.2772065124250214,
+      "grad_norm": 3.467730774221773,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 11646
+    },
+    {
+      "epoch": 0.27723031514805296,
+      "grad_norm": 9.730323361083732,
+      "learning_rate": 2e-05,
+      "loss": 1.1472,
+      "step": 11647
+    },
+    {
+      "epoch": 0.27725411787108445,
+      "grad_norm": 7.827729413474657,
+      "learning_rate": 2e-05,
+      "loss": 1.7027,
+      "step": 11648
+    },
+    {
+      "epoch": 0.277277920594116,
+      "grad_norm": 2.898708680080954,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 11649
+    },
+    {
+      "epoch": 0.2773017233171475,
+      "grad_norm": 2.8310848326781586,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 11650
+    },
+    {
+      "epoch": 0.277325526040179,
+      "grad_norm": 2.2048228423821774,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 11651
+    },
+    {
+      "epoch": 0.2773493287632105,
+      "grad_norm": 9.290118255001907,
+      "learning_rate": 2e-05,
+      "loss": 0.3994,
+      "step": 11652
+    },
+    {
+      "epoch": 0.27737313148624204,
+      "grad_norm": 1.796747225168059,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 11653
+    },
+    {
+      "epoch": 0.2773969342092735,
+      "grad_norm": 9.39922941882717,
+      "learning_rate": 2e-05,
+      "loss": 0.2103,
+      "step": 11654
+    },
+    {
+      "epoch": 0.27742073693230507,
+      "grad_norm": 1.4458543899098604,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 11655
+    },
+    {
+      "epoch": 0.27744453965533655,
+      "grad_norm": 3.9299387608964547,
+      "learning_rate": 2e-05,
+      "loss": 0.3182,
+      "step": 11656
+    },
+    {
+      "epoch": 0.2774683423783681,
+      "grad_norm": 1.6419066058143446,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 11657
+    },
+    {
+      "epoch": 0.2774921451013996,
+      "grad_norm": 1.1677908200430045,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 11658
+    },
+    {
+      "epoch": 0.2775159478244311,
+      "grad_norm": 2.269521849660167,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 11659
+    },
+    {
+      "epoch": 0.2775397505474626,
+      "grad_norm": 1.3169393109752816,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 11660
+    },
+    {
+      "epoch": 0.27756355327049415,
+      "grad_norm": 8.420450226581321,
+      "learning_rate": 2e-05,
+      "loss": 0.9529,
+      "step": 11661
+    },
+    {
+      "epoch": 0.2775873559935257,
+      "grad_norm": 1.7320780123266863,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 11662
+    },
+    {
+      "epoch": 0.27761115871655717,
+      "grad_norm": 2.474544230150491,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 11663
+    },
+    {
+      "epoch": 0.2776349614395887,
+      "grad_norm": 6.575541210381915,
+      "learning_rate": 2e-05,
+      "loss": 0.82,
+      "step": 11664
+    },
+    {
+      "epoch": 0.2776587641626202,
+      "grad_norm": 2.6057558431196925,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 11665
+    },
+    {
+      "epoch": 0.27768256688565174,
+      "grad_norm": 6.119047320763696,
+      "learning_rate": 2e-05,
+      "loss": 0.6207,
+      "step": 11666
+    },
+    {
+      "epoch": 0.2777063696086832,
+      "grad_norm": 5.9465350558541505,
+      "learning_rate": 2e-05,
+      "loss": 0.1193,
+      "step": 11667
+    },
+    {
+      "epoch": 0.27773017233171476,
+      "grad_norm": 5.214888755723232,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 11668
+    },
+    {
+      "epoch": 0.27775397505474625,
+      "grad_norm": 2.6341732764901984,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 11669
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 3.721477508069747,
+      "learning_rate": 2e-05,
+      "loss": 0.5075,
+      "step": 11670
+    },
+    {
+      "epoch": 0.2778015805008093,
+      "grad_norm": 3.8483211159887154,
+      "learning_rate": 2e-05,
+      "loss": 0.1809,
+      "step": 11671
+    },
+    {
+      "epoch": 0.2778253832238408,
+      "grad_norm": 3.2246559056555673,
+      "learning_rate": 2e-05,
+      "loss": 0.1646,
+      "step": 11672
+    },
+    {
+      "epoch": 0.2778491859468723,
+      "grad_norm": 2.485381020927756,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 11673
+    },
+    {
+      "epoch": 0.27787298866990384,
+      "grad_norm": 5.575486803694293,
+      "learning_rate": 2e-05,
+      "loss": 0.6636,
+      "step": 11674
+    },
+    {
+      "epoch": 0.2778967913929353,
+      "grad_norm": 1.792791472741448,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 11675
+    },
+    {
+      "epoch": 0.27792059411596687,
+      "grad_norm": 1.500643635365473,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 11676
+    },
+    {
+      "epoch": 0.2779443968389984,
+      "grad_norm": 1.7939200526464196,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 11677
+    },
+    {
+      "epoch": 0.2779681995620299,
+      "grad_norm": 5.815287700227923,
+      "learning_rate": 2e-05,
+      "loss": 1.2382,
+      "step": 11678
+    },
+    {
+      "epoch": 0.27799200228506143,
+      "grad_norm": 4.619752900064198,
+      "learning_rate": 2e-05,
+      "loss": 0.2565,
+      "step": 11679
+    },
+    {
+      "epoch": 0.2780158050080929,
+      "grad_norm": 1.8231971459245906,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 11680
+    },
+    {
+      "epoch": 0.27803960773112446,
+      "grad_norm": 9.197805877159807,
+      "learning_rate": 2e-05,
+      "loss": 0.4556,
+      "step": 11681
+    },
+    {
+      "epoch": 0.27806341045415595,
+      "grad_norm": 2.4792746304687276,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 11682
+    },
+    {
+      "epoch": 0.2780872131771875,
+      "grad_norm": 5.442297741338608,
+      "learning_rate": 2e-05,
+      "loss": 1.3084,
+      "step": 11683
+    },
+    {
+      "epoch": 0.27811101590021897,
+      "grad_norm": 4.521357714758738,
+      "learning_rate": 2e-05,
+      "loss": 0.497,
+      "step": 11684
+    },
+    {
+      "epoch": 0.2781348186232505,
+      "grad_norm": 2.784694139150057,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 11685
+    },
+    {
+      "epoch": 0.278158621346282,
+      "grad_norm": 3.174018119767521,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 11686
+    },
+    {
+      "epoch": 0.27818242406931354,
+      "grad_norm": 7.537474606204,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 11687
+    },
+    {
+      "epoch": 0.278206226792345,
+      "grad_norm": 2.8445769765742486,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 11688
+    },
+    {
+      "epoch": 0.27823002951537656,
+      "grad_norm": 8.078423463459053,
+      "learning_rate": 2e-05,
+      "loss": 0.4817,
+      "step": 11689
+    },
+    {
+      "epoch": 0.27825383223840805,
+      "grad_norm": 4.588394061166517,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 11690
+    },
+    {
+      "epoch": 0.2782776349614396,
+      "grad_norm": 5.800893270973238,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 11691
+    },
+    {
+      "epoch": 0.2783014376844711,
+      "grad_norm": 7.4848474424892,
+      "learning_rate": 2e-05,
+      "loss": 0.6307,
+      "step": 11692
+    },
+    {
+      "epoch": 0.2783252404075026,
+      "grad_norm": 16.0658321541726,
+      "learning_rate": 2e-05,
+      "loss": 0.5221,
+      "step": 11693
+    },
+    {
+      "epoch": 0.27834904313053416,
+      "grad_norm": 6.14065566253209,
+      "learning_rate": 2e-05,
+      "loss": 0.87,
+      "step": 11694
+    },
+    {
+      "epoch": 0.27837284585356564,
+      "grad_norm": 7.494979608206797,
+      "learning_rate": 2e-05,
+      "loss": 1.0843,
+      "step": 11695
+    },
+    {
+      "epoch": 0.2783966485765972,
+      "grad_norm": 4.816164615704815,
+      "learning_rate": 2e-05,
+      "loss": 0.7045,
+      "step": 11696
+    },
+    {
+      "epoch": 0.27842045129962867,
+      "grad_norm": 4.037750385594302,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 11697
+    },
+    {
+      "epoch": 0.2784442540226602,
+      "grad_norm": 5.379513964413568,
+      "learning_rate": 2e-05,
+      "loss": 0.8628,
+      "step": 11698
+    },
+    {
+      "epoch": 0.2784680567456917,
+      "grad_norm": 4.48350582560501,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 11699
+    },
+    {
+      "epoch": 0.27849185946872324,
+      "grad_norm": 4.994005188071142,
+      "learning_rate": 2e-05,
+      "loss": 0.9312,
+      "step": 11700
+    },
+    {
+      "epoch": 0.2785156621917547,
+      "grad_norm": 3.0715320176079657,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 11701
+    },
+    {
+      "epoch": 0.27853946491478626,
+      "grad_norm": 1.8768102985434119,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 11702
+    },
+    {
+      "epoch": 0.27856326763781775,
+      "grad_norm": 3.0615191516681812,
+      "learning_rate": 2e-05,
+      "loss": 0.3562,
+      "step": 11703
+    },
+    {
+      "epoch": 0.2785870703608493,
+      "grad_norm": 8.62395050999088,
+      "learning_rate": 2e-05,
+      "loss": 0.5524,
+      "step": 11704
+    },
+    {
+      "epoch": 0.2786108730838808,
+      "grad_norm": 4.405129783258904,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 11705
+    },
+    {
+      "epoch": 0.2786346758069123,
+      "grad_norm": 5.917803611016695,
+      "learning_rate": 2e-05,
+      "loss": 0.965,
+      "step": 11706
+    },
+    {
+      "epoch": 0.2786584785299438,
+      "grad_norm": 6.5368902393845945,
+      "learning_rate": 2e-05,
+      "loss": 0.8943,
+      "step": 11707
+    },
+    {
+      "epoch": 0.27868228125297534,
+      "grad_norm": 7.559696204567332,
+      "learning_rate": 2e-05,
+      "loss": 1.962,
+      "step": 11708
+    },
+    {
+      "epoch": 0.2787060839760069,
+      "grad_norm": 1.9020545899814436,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 11709
+    },
+    {
+      "epoch": 0.27872988669903837,
+      "grad_norm": 1.6819499075214337,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 11710
+    },
+    {
+      "epoch": 0.2787536894220699,
+      "grad_norm": 13.492762352891297,
+      "learning_rate": 2e-05,
+      "loss": 0.6051,
+      "step": 11711
+    },
+    {
+      "epoch": 0.2787774921451014,
+      "grad_norm": 9.877188620985752,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 11712
+    },
+    {
+      "epoch": 0.27880129486813293,
+      "grad_norm": 2.6964387218726724,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 11713
+    },
+    {
+      "epoch": 0.2788250975911644,
+      "grad_norm": 5.124432695882886,
+      "learning_rate": 2e-05,
+      "loss": 1.0567,
+      "step": 11714
+    },
+    {
+      "epoch": 0.27884890031419596,
+      "grad_norm": 4.57497058121552,
+      "learning_rate": 2e-05,
+      "loss": 0.2047,
+      "step": 11715
+    },
+    {
+      "epoch": 0.27887270303722744,
+      "grad_norm": 3.7045205237582,
+      "learning_rate": 2e-05,
+      "loss": 0.6721,
+      "step": 11716
+    },
+    {
+      "epoch": 0.278896505760259,
+      "grad_norm": 9.337165835688282,
+      "learning_rate": 2e-05,
+      "loss": 0.6921,
+      "step": 11717
+    },
+    {
+      "epoch": 0.27892030848329047,
+      "grad_norm": 3.174317359380413,
+      "learning_rate": 2e-05,
+      "loss": 0.4971,
+      "step": 11718
+    },
+    {
+      "epoch": 0.278944111206322,
+      "grad_norm": 4.461647852407653,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 11719
+    },
+    {
+      "epoch": 0.2789679139293535,
+      "grad_norm": 5.346099992841142,
+      "learning_rate": 2e-05,
+      "loss": 0.5396,
+      "step": 11720
+    },
+    {
+      "epoch": 0.27899171665238504,
+      "grad_norm": 2.8104557715963514,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 11721
+    },
+    {
+      "epoch": 0.2790155193754165,
+      "grad_norm": 6.918582531671597,
+      "learning_rate": 2e-05,
+      "loss": 1.1684,
+      "step": 11722
+    },
+    {
+      "epoch": 0.27903932209844806,
+      "grad_norm": 2.347632869065846,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 11723
+    },
+    {
+      "epoch": 0.2790631248214796,
+      "grad_norm": 2.9445396766968406,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 11724
+    },
+    {
+      "epoch": 0.2790869275445111,
+      "grad_norm": 8.235135351075154,
+      "learning_rate": 2e-05,
+      "loss": 0.5173,
+      "step": 11725
+    },
+    {
+      "epoch": 0.27911073026754263,
+      "grad_norm": 3.975528777037553,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 11726
+    },
+    {
+      "epoch": 0.2791345329905741,
+      "grad_norm": 4.344930833893764,
+      "learning_rate": 2e-05,
+      "loss": 0.4572,
+      "step": 11727
+    },
+    {
+      "epoch": 0.27915833571360565,
+      "grad_norm": 2.5368541647117535,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 11728
+    },
+    {
+      "epoch": 0.27918213843663714,
+      "grad_norm": 2.835360701086813,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 11729
+    },
+    {
+      "epoch": 0.2792059411596687,
+      "grad_norm": 4.447378952040811,
+      "learning_rate": 2e-05,
+      "loss": 0.6527,
+      "step": 11730
+    },
+    {
+      "epoch": 0.27922974388270017,
+      "grad_norm": 1.5922791968180554,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 11731
+    },
+    {
+      "epoch": 0.2792535466057317,
+      "grad_norm": 3.518472947980463,
+      "learning_rate": 2e-05,
+      "loss": 0.4077,
+      "step": 11732
+    },
+    {
+      "epoch": 0.2792773493287632,
+      "grad_norm": 4.1752317242965065,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 11733
+    },
+    {
+      "epoch": 0.27930115205179473,
+      "grad_norm": 0.7968989016016486,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 11734
+    },
+    {
+      "epoch": 0.2793249547748262,
+      "grad_norm": 5.587853316933621,
+      "learning_rate": 2e-05,
+      "loss": 0.8155,
+      "step": 11735
+    },
+    {
+      "epoch": 0.27934875749785776,
+      "grad_norm": 8.375439413179567,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 11736
+    },
+    {
+      "epoch": 0.27937256022088924,
+      "grad_norm": 5.70833901543725,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 11737
+    },
+    {
+      "epoch": 0.2793963629439208,
+      "grad_norm": 1.5385636127511766,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 11738
+    },
+    {
+      "epoch": 0.2794201656669523,
+      "grad_norm": 1.6953393914597898,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 11739
+    },
+    {
+      "epoch": 0.2794439683899838,
+      "grad_norm": 2.9835316035873243,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 11740
+    },
+    {
+      "epoch": 0.27946777111301535,
+      "grad_norm": 4.028059212497849,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 11741
+    },
+    {
+      "epoch": 0.27949157383604684,
+      "grad_norm": 7.091100581390565,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 11742
+    },
+    {
+      "epoch": 0.2795153765590784,
+      "grad_norm": 16.70997448728324,
+      "learning_rate": 2e-05,
+      "loss": 0.2921,
+      "step": 11743
+    },
+    {
+      "epoch": 0.27953917928210986,
+      "grad_norm": 4.61380059623939,
+      "learning_rate": 2e-05,
+      "loss": 0.1465,
+      "step": 11744
+    },
+    {
+      "epoch": 0.2795629820051414,
+      "grad_norm": 0.897382801962329,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 11745
+    },
+    {
+      "epoch": 0.2795867847281729,
+      "grad_norm": 4.986467296700152,
+      "learning_rate": 2e-05,
+      "loss": 0.9,
+      "step": 11746
+    },
+    {
+      "epoch": 0.27961058745120443,
+      "grad_norm": 3.162834112295713,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 11747
+    },
+    {
+      "epoch": 0.2796343901742359,
+      "grad_norm": 3.941433676513891,
+      "learning_rate": 2e-05,
+      "loss": 0.7198,
+      "step": 11748
+    },
+    {
+      "epoch": 0.27965819289726745,
+      "grad_norm": 4.310045073395953,
+      "learning_rate": 2e-05,
+      "loss": 0.4772,
+      "step": 11749
+    },
+    {
+      "epoch": 0.27968199562029894,
+      "grad_norm": 6.32986001948142,
+      "learning_rate": 2e-05,
+      "loss": 0.8222,
+      "step": 11750
+    },
+    {
+      "epoch": 0.2797057983433305,
+      "grad_norm": 1.9775434830938707,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 11751
+    },
+    {
+      "epoch": 0.27972960106636197,
+      "grad_norm": 1.8951360904273815,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 11752
+    },
+    {
+      "epoch": 0.2797534037893935,
+      "grad_norm": 9.021639357371265,
+      "learning_rate": 2e-05,
+      "loss": 0.9773,
+      "step": 11753
+    },
+    {
+      "epoch": 0.27977720651242505,
+      "grad_norm": 3.367561579914312,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 11754
+    },
+    {
+      "epoch": 0.27980100923545653,
+      "grad_norm": 4.762879470262028,
+      "learning_rate": 2e-05,
+      "loss": 0.3163,
+      "step": 11755
+    },
+    {
+      "epoch": 0.2798248119584881,
+      "grad_norm": 3.53174338584083,
+      "learning_rate": 2e-05,
+      "loss": 0.2211,
+      "step": 11756
+    },
+    {
+      "epoch": 0.27984861468151956,
+      "grad_norm": 6.32119146727092,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 11757
+    },
+    {
+      "epoch": 0.2798724174045511,
+      "grad_norm": 2.8067417958382794,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 11758
+    },
+    {
+      "epoch": 0.2798962201275826,
+      "grad_norm": 10.805188624419191,
+      "learning_rate": 2e-05,
+      "loss": 0.4711,
+      "step": 11759
+    },
+    {
+      "epoch": 0.2799200228506141,
+      "grad_norm": 1.8943547956827074,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 11760
+    },
+    {
+      "epoch": 0.2799438255736456,
+      "grad_norm": 9.173841023969642,
+      "learning_rate": 2e-05,
+      "loss": 1.0102,
+      "step": 11761
+    },
+    {
+      "epoch": 0.27996762829667715,
+      "grad_norm": 6.271413624991028,
+      "learning_rate": 2e-05,
+      "loss": 0.525,
+      "step": 11762
+    },
+    {
+      "epoch": 0.27999143101970864,
+      "grad_norm": 5.129675219181966,
+      "learning_rate": 2e-05,
+      "loss": 0.8208,
+      "step": 11763
+    },
+    {
+      "epoch": 0.2800152337427402,
+      "grad_norm": 6.46633615471133,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 11764
+    },
+    {
+      "epoch": 0.28003903646577166,
+      "grad_norm": 6.0349532667977845,
+      "learning_rate": 2e-05,
+      "loss": 1.0701,
+      "step": 11765
+    },
+    {
+      "epoch": 0.2800628391888032,
+      "grad_norm": 2.870803798980369,
+      "learning_rate": 2e-05,
+      "loss": 0.0952,
+      "step": 11766
+    },
+    {
+      "epoch": 0.2800866419118347,
+      "grad_norm": 4.836323380931722,
+      "learning_rate": 2e-05,
+      "loss": 0.411,
+      "step": 11767
+    },
+    {
+      "epoch": 0.28011044463486623,
+      "grad_norm": 1.3404527249585212,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 11768
+    },
+    {
+      "epoch": 0.28013424735789777,
+      "grad_norm": 4.913392842906687,
+      "learning_rate": 2e-05,
+      "loss": 0.4245,
+      "step": 11769
+    },
+    {
+      "epoch": 0.28015805008092926,
+      "grad_norm": 5.170431157690231,
+      "learning_rate": 2e-05,
+      "loss": 0.2574,
+      "step": 11770
+    },
+    {
+      "epoch": 0.2801818528039608,
+      "grad_norm": 11.753181307892895,
+      "learning_rate": 2e-05,
+      "loss": 0.6362,
+      "step": 11771
+    },
+    {
+      "epoch": 0.2802056555269923,
+      "grad_norm": 3.1988272490134704,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 11772
+    },
+    {
+      "epoch": 0.2802294582500238,
+      "grad_norm": 1.3607706664678174,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 11773
+    },
+    {
+      "epoch": 0.2802532609730553,
+      "grad_norm": 5.919133588844354,
+      "learning_rate": 2e-05,
+      "loss": 0.1506,
+      "step": 11774
+    },
+    {
+      "epoch": 0.28027706369608685,
+      "grad_norm": 0.5702995566562601,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 11775
+    },
+    {
+      "epoch": 0.28030086641911833,
+      "grad_norm": 7.2071725695435145,
+      "learning_rate": 2e-05,
+      "loss": 0.4943,
+      "step": 11776
+    },
+    {
+      "epoch": 0.2803246691421499,
+      "grad_norm": 3.9883789311211584,
+      "learning_rate": 2e-05,
+      "loss": 0.4827,
+      "step": 11777
+    },
+    {
+      "epoch": 0.28034847186518136,
+      "grad_norm": 0.6593869013356753,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 11778
+    },
+    {
+      "epoch": 0.2803722745882129,
+      "grad_norm": 4.1701074164650365,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 11779
+    },
+    {
+      "epoch": 0.2803960773112444,
+      "grad_norm": 6.522321659784812,
+      "learning_rate": 2e-05,
+      "loss": 0.4038,
+      "step": 11780
+    },
+    {
+      "epoch": 0.2804198800342759,
+      "grad_norm": 6.712984671743066,
+      "learning_rate": 2e-05,
+      "loss": 0.8431,
+      "step": 11781
+    },
+    {
+      "epoch": 0.2804436827573074,
+      "grad_norm": 1.3703935066848756,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 11782
+    },
+    {
+      "epoch": 0.28046748548033895,
+      "grad_norm": 2.7216168812084582,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 11783
+    },
+    {
+      "epoch": 0.2804912882033705,
+      "grad_norm": 2.7259613742117166,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 11784
+    },
+    {
+      "epoch": 0.280515090926402,
+      "grad_norm": 3.1849048163340865,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 11785
+    },
+    {
+      "epoch": 0.2805388936494335,
+      "grad_norm": 3.158153490459864,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 11786
+    },
+    {
+      "epoch": 0.280562696372465,
+      "grad_norm": 2.697034146771451,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 11787
+    },
+    {
+      "epoch": 0.28058649909549654,
+      "grad_norm": 8.096776250911175,
+      "learning_rate": 2e-05,
+      "loss": 0.5406,
+      "step": 11788
+    },
+    {
+      "epoch": 0.28061030181852803,
+      "grad_norm": 6.21154818819635,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 11789
+    },
+    {
+      "epoch": 0.28063410454155957,
+      "grad_norm": 4.859671309175734,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 11790
+    },
+    {
+      "epoch": 0.28065790726459106,
+      "grad_norm": 26.07556397385667,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 11791
+    },
+    {
+      "epoch": 0.2806817099876226,
+      "grad_norm": 4.795065355167974,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 11792
+    },
+    {
+      "epoch": 0.2807055127106541,
+      "grad_norm": 1.803948053396794,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 11793
+    },
+    {
+      "epoch": 0.2807293154336856,
+      "grad_norm": 13.07932301430022,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 11794
+    },
+    {
+      "epoch": 0.2807531181567171,
+      "grad_norm": 3.38984696262428,
+      "learning_rate": 2e-05,
+      "loss": 0.5337,
+      "step": 11795
+    },
+    {
+      "epoch": 0.28077692087974865,
+      "grad_norm": 6.429186066183324,
+      "learning_rate": 2e-05,
+      "loss": 1.2765,
+      "step": 11796
+    },
+    {
+      "epoch": 0.28080072360278013,
+      "grad_norm": 21.115936603035138,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 11797
+    },
+    {
+      "epoch": 0.2808245263258117,
+      "grad_norm": 2.2123744458576193,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 11798
+    },
+    {
+      "epoch": 0.2808483290488432,
+      "grad_norm": 1.653503702657757,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11799
+    },
+    {
+      "epoch": 0.2808721317718747,
+      "grad_norm": 3.559417091633572,
+      "learning_rate": 2e-05,
+      "loss": 0.2087,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28089593449490624,
+      "grad_norm": 3.130717252338976,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 11801
+    },
+    {
+      "epoch": 0.2809197372179377,
+      "grad_norm": 4.712006763965521,
+      "learning_rate": 2e-05,
+      "loss": 0.6987,
+      "step": 11802
+    },
+    {
+      "epoch": 0.28094353994096927,
+      "grad_norm": 2.8165375227517964,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 11803
+    },
+    {
+      "epoch": 0.28096734266400075,
+      "grad_norm": 73.98550935338203,
+      "learning_rate": 2e-05,
+      "loss": 1.1591,
+      "step": 11804
+    },
+    {
+      "epoch": 0.2809911453870323,
+      "grad_norm": 5.640805322823129,
+      "learning_rate": 2e-05,
+      "loss": 0.5446,
+      "step": 11805
+    },
+    {
+      "epoch": 0.2810149481100638,
+      "grad_norm": 2.781919350005493,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 11806
+    },
+    {
+      "epoch": 0.2810387508330953,
+      "grad_norm": 1.5721963907738286,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 11807
+    },
+    {
+      "epoch": 0.2810625535561268,
+      "grad_norm": 5.844555846026361,
+      "learning_rate": 2e-05,
+      "loss": 0.7899,
+      "step": 11808
+    },
+    {
+      "epoch": 0.28108635627915834,
+      "grad_norm": 5.612121898285073,
+      "learning_rate": 2e-05,
+      "loss": 0.9926,
+      "step": 11809
+    },
+    {
+      "epoch": 0.28111015900218983,
+      "grad_norm": 3.3122921251649875,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 11810
+    },
+    {
+      "epoch": 0.28113396172522137,
+      "grad_norm": 4.630293726727167,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 11811
+    },
+    {
+      "epoch": 0.28115776444825286,
+      "grad_norm": 4.693645580777988,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 11812
+    },
+    {
+      "epoch": 0.2811815671712844,
+      "grad_norm": 4.984442335954895,
+      "learning_rate": 2e-05,
+      "loss": 0.6112,
+      "step": 11813
+    },
+    {
+      "epoch": 0.28120536989431594,
+      "grad_norm": 5.407186714336903,
+      "learning_rate": 2e-05,
+      "loss": 0.957,
+      "step": 11814
+    },
+    {
+      "epoch": 0.2812291726173474,
+      "grad_norm": 4.736833191188566,
+      "learning_rate": 2e-05,
+      "loss": 0.639,
+      "step": 11815
+    },
+    {
+      "epoch": 0.28125297534037896,
+      "grad_norm": 7.778676123843351,
+      "learning_rate": 2e-05,
+      "loss": 1.5484,
+      "step": 11816
+    },
+    {
+      "epoch": 0.28127677806341045,
+      "grad_norm": 2.6998692322995366,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 11817
+    },
+    {
+      "epoch": 0.281300580786442,
+      "grad_norm": 8.403472365520216,
+      "learning_rate": 2e-05,
+      "loss": 0.3432,
+      "step": 11818
+    },
+    {
+      "epoch": 0.2813243835094735,
+      "grad_norm": 1.4394810259558628,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 11819
+    },
+    {
+      "epoch": 0.281348186232505,
+      "grad_norm": 6.42500632754959,
+      "learning_rate": 2e-05,
+      "loss": 0.4141,
+      "step": 11820
+    },
+    {
+      "epoch": 0.2813719889555365,
+      "grad_norm": 2.7201706477145646,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 11821
+    },
+    {
+      "epoch": 0.28139579167856804,
+      "grad_norm": 12.721059611154214,
+      "learning_rate": 2e-05,
+      "loss": 0.284,
+      "step": 11822
+    },
+    {
+      "epoch": 0.2814195944015995,
+      "grad_norm": 2.722734309449401,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 11823
+    },
+    {
+      "epoch": 0.28144339712463107,
+      "grad_norm": 1.1499014677796628,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 11824
+    },
+    {
+      "epoch": 0.28146719984766255,
+      "grad_norm": 5.562468965379059,
+      "learning_rate": 2e-05,
+      "loss": 0.8904,
+      "step": 11825
+    },
+    {
+      "epoch": 0.2814910025706941,
+      "grad_norm": 3.9772179419773135,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 11826
+    },
+    {
+      "epoch": 0.2815148052937256,
+      "grad_norm": 6.444344703154745,
+      "learning_rate": 2e-05,
+      "loss": 0.5079,
+      "step": 11827
+    },
+    {
+      "epoch": 0.2815386080167571,
+      "grad_norm": 2.6775477993729013,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 11828
+    },
+    {
+      "epoch": 0.2815624107397886,
+      "grad_norm": 5.688675345909082,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 11829
+    },
+    {
+      "epoch": 0.28158621346282015,
+      "grad_norm": 2.071704957620559,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 11830
+    },
+    {
+      "epoch": 0.2816100161858517,
+      "grad_norm": 2.5641964979317415,
+      "learning_rate": 2e-05,
+      "loss": 0.4338,
+      "step": 11831
+    },
+    {
+      "epoch": 0.28163381890888317,
+      "grad_norm": 2.096136025488606,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 11832
+    },
+    {
+      "epoch": 0.2816576216319147,
+      "grad_norm": 2.3938997280412067,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 11833
+    },
+    {
+      "epoch": 0.2816814243549462,
+      "grad_norm": 2.0306612751338013,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 11834
+    },
+    {
+      "epoch": 0.28170522707797774,
+      "grad_norm": 4.710653610992949,
+      "learning_rate": 2e-05,
+      "loss": 0.5824,
+      "step": 11835
+    },
+    {
+      "epoch": 0.2817290298010092,
+      "grad_norm": 2.7244952919507712,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 11836
+    },
+    {
+      "epoch": 0.28175283252404076,
+      "grad_norm": 5.211954070182399,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 11837
+    },
+    {
+      "epoch": 0.28177663524707225,
+      "grad_norm": 2.007894124360809,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 11838
+    },
+    {
+      "epoch": 0.2818004379701038,
+      "grad_norm": 1.821405345568767,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 11839
+    },
+    {
+      "epoch": 0.2818242406931353,
+      "grad_norm": 6.492816499019974,
+      "learning_rate": 2e-05,
+      "loss": 0.5922,
+      "step": 11840
+    },
+    {
+      "epoch": 0.2818480434161668,
+      "grad_norm": 14.647851651896621,
+      "learning_rate": 2e-05,
+      "loss": 1.2899,
+      "step": 11841
+    },
+    {
+      "epoch": 0.2818718461391983,
+      "grad_norm": 7.59607079455922,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 11842
+    },
+    {
+      "epoch": 0.28189564886222984,
+      "grad_norm": 3.089916028195044,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 11843
+    },
+    {
+      "epoch": 0.2819194515852613,
+      "grad_norm": 5.006728488912989,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 11844
+    },
+    {
+      "epoch": 0.28194325430829287,
+      "grad_norm": 0.8647089282974183,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 11845
+    },
+    {
+      "epoch": 0.2819670570313244,
+      "grad_norm": 2.3170868645156255,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 11846
+    },
+    {
+      "epoch": 0.2819908597543559,
+      "grad_norm": 6.290023759838628,
+      "learning_rate": 2e-05,
+      "loss": 1.4319,
+      "step": 11847
+    },
+    {
+      "epoch": 0.28201466247738743,
+      "grad_norm": 0.650944372474666,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 11848
+    },
+    {
+      "epoch": 0.2820384652004189,
+      "grad_norm": 4.02712270682829,
+      "learning_rate": 2e-05,
+      "loss": 0.831,
+      "step": 11849
+    },
+    {
+      "epoch": 0.28206226792345046,
+      "grad_norm": 21.551764478379177,
+      "learning_rate": 2e-05,
+      "loss": 1.1762,
+      "step": 11850
+    },
+    {
+      "epoch": 0.28208607064648195,
+      "grad_norm": 2.2522526868636437,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 11851
+    },
+    {
+      "epoch": 0.2821098733695135,
+      "grad_norm": 9.66066319075296,
+      "learning_rate": 2e-05,
+      "loss": 0.2739,
+      "step": 11852
+    },
+    {
+      "epoch": 0.28213367609254497,
+      "grad_norm": 4.5671503932548285,
+      "learning_rate": 2e-05,
+      "loss": 0.8347,
+      "step": 11853
+    },
+    {
+      "epoch": 0.2821574788155765,
+      "grad_norm": 3.972919304495159,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 11854
+    },
+    {
+      "epoch": 0.282181281538608,
+      "grad_norm": 0.8221609794674731,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 11855
+    },
+    {
+      "epoch": 0.28220508426163954,
+      "grad_norm": 1.870396747920041,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 11856
+    },
+    {
+      "epoch": 0.282228886984671,
+      "grad_norm": 2.170661065947705,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 11857
+    },
+    {
+      "epoch": 0.28225268970770256,
+      "grad_norm": 4.4432023425731035,
+      "learning_rate": 2e-05,
+      "loss": 0.766,
+      "step": 11858
+    },
+    {
+      "epoch": 0.28227649243073405,
+      "grad_norm": 3.6121242168745487,
+      "learning_rate": 2e-05,
+      "loss": 0.4636,
+      "step": 11859
+    },
+    {
+      "epoch": 0.2823002951537656,
+      "grad_norm": 0.1735228811290801,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 11860
+    },
+    {
+      "epoch": 0.28232409787679713,
+      "grad_norm": 0.9768199557205842,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 11861
+    },
+    {
+      "epoch": 0.2823479005998286,
+      "grad_norm": 5.4034515714135996,
+      "learning_rate": 2e-05,
+      "loss": 1.1,
+      "step": 11862
+    },
+    {
+      "epoch": 0.28237170332286016,
+      "grad_norm": 3.464193590493287,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 11863
+    },
+    {
+      "epoch": 0.28239550604589164,
+      "grad_norm": 5.6137472771952215,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 11864
+    },
+    {
+      "epoch": 0.2824193087689232,
+      "grad_norm": 3.981393001749283,
+      "learning_rate": 2e-05,
+      "loss": 0.2872,
+      "step": 11865
+    },
+    {
+      "epoch": 0.28244311149195467,
+      "grad_norm": 3.7475238277740424,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 11866
+    },
+    {
+      "epoch": 0.2824669142149862,
+      "grad_norm": 2.6143433838042602,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 11867
+    },
+    {
+      "epoch": 0.2824907169380177,
+      "grad_norm": 2.330047806579746,
+      "learning_rate": 2e-05,
+      "loss": 0.0951,
+      "step": 11868
+    },
+    {
+      "epoch": 0.28251451966104923,
+      "grad_norm": 3.3246579062441266,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 11869
+    },
+    {
+      "epoch": 0.2825383223840807,
+      "grad_norm": 4.129512634728285,
+      "learning_rate": 2e-05,
+      "loss": 0.126,
+      "step": 11870
+    },
+    {
+      "epoch": 0.28256212510711226,
+      "grad_norm": 7.809996427400473,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 11871
+    },
+    {
+      "epoch": 0.28258592783014375,
+      "grad_norm": 4.694421348441694,
+      "learning_rate": 2e-05,
+      "loss": 0.1492,
+      "step": 11872
+    },
+    {
+      "epoch": 0.2826097305531753,
+      "grad_norm": 4.084249848568233,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 11873
+    },
+    {
+      "epoch": 0.28263353327620677,
+      "grad_norm": 2.3171792837537755,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 11874
+    },
+    {
+      "epoch": 0.2826573359992383,
+      "grad_norm": 1.5714071795164342,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 11875
+    },
+    {
+      "epoch": 0.28268113872226985,
+      "grad_norm": 3.8897615502550162,
+      "learning_rate": 2e-05,
+      "loss": 0.2166,
+      "step": 11876
+    },
+    {
+      "epoch": 0.28270494144530134,
+      "grad_norm": 2.5273034442772992,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 11877
+    },
+    {
+      "epoch": 0.2827287441683329,
+      "grad_norm": 3.923743699674348,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 11878
+    },
+    {
+      "epoch": 0.28275254689136436,
+      "grad_norm": 2.940013961075999,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 11879
+    },
+    {
+      "epoch": 0.2827763496143959,
+      "grad_norm": 1.0904975834478956,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 11880
+    },
+    {
+      "epoch": 0.2828001523374274,
+      "grad_norm": 9.147259603999498,
+      "learning_rate": 2e-05,
+      "loss": 0.9164,
+      "step": 11881
+    },
+    {
+      "epoch": 0.28282395506045893,
+      "grad_norm": 3.3201276442701046,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 11882
+    },
+    {
+      "epoch": 0.2828477577834904,
+      "grad_norm": 2.3662299405590845,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 11883
+    },
+    {
+      "epoch": 0.28287156050652196,
+      "grad_norm": 5.389808017248612,
+      "learning_rate": 2e-05,
+      "loss": 0.6836,
+      "step": 11884
+    },
+    {
+      "epoch": 0.28289536322955344,
+      "grad_norm": 3.180084050413976,
+      "learning_rate": 2e-05,
+      "loss": 0.5377,
+      "step": 11885
+    },
+    {
+      "epoch": 0.282919165952585,
+      "grad_norm": 10.511038998788624,
+      "learning_rate": 2e-05,
+      "loss": 1.5963,
+      "step": 11886
+    },
+    {
+      "epoch": 0.28294296867561647,
+      "grad_norm": 0.3042368763045904,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 11887
+    },
+    {
+      "epoch": 0.282966771398648,
+      "grad_norm": 9.185134740151764,
+      "learning_rate": 2e-05,
+      "loss": 0.5066,
+      "step": 11888
+    },
+    {
+      "epoch": 0.2829905741216795,
+      "grad_norm": 0.885111862437145,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 11889
+    },
+    {
+      "epoch": 0.28301437684471104,
+      "grad_norm": 5.282629824468837,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 11890
+    },
+    {
+      "epoch": 0.2830381795677426,
+      "grad_norm": 0.6315227197470781,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 11891
+    },
+    {
+      "epoch": 0.28306198229077406,
+      "grad_norm": 1.5405486121794465,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 11892
+    },
+    {
+      "epoch": 0.2830857850138056,
+      "grad_norm": 4.337948179207669,
+      "learning_rate": 2e-05,
+      "loss": 0.135,
+      "step": 11893
+    },
+    {
+      "epoch": 0.2831095877368371,
+      "grad_norm": 4.9054209423335475,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 11894
+    },
+    {
+      "epoch": 0.28313339045986863,
+      "grad_norm": 0.47775901344948657,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 11895
+    },
+    {
+      "epoch": 0.2831571931829001,
+      "grad_norm": 2.8262322603021444,
+      "learning_rate": 2e-05,
+      "loss": 0.206,
+      "step": 11896
+    },
+    {
+      "epoch": 0.28318099590593165,
+      "grad_norm": 3.3980799461525537,
+      "learning_rate": 2e-05,
+      "loss": 0.175,
+      "step": 11897
+    },
+    {
+      "epoch": 0.28320479862896314,
+      "grad_norm": 6.648077511965483,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 11898
+    },
+    {
+      "epoch": 0.2832286013519947,
+      "grad_norm": 15.311654115769514,
+      "learning_rate": 2e-05,
+      "loss": 0.7819,
+      "step": 11899
+    },
+    {
+      "epoch": 0.28325240407502617,
+      "grad_norm": 1.49907800004992,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 11900
+    },
+    {
+      "epoch": 0.2832762067980577,
+      "grad_norm": 8.905478720114674,
+      "learning_rate": 2e-05,
+      "loss": 1.2931,
+      "step": 11901
+    },
+    {
+      "epoch": 0.2833000095210892,
+      "grad_norm": 2.7824238639556187,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 11902
+    },
+    {
+      "epoch": 0.28332381224412073,
+      "grad_norm": 12.571964585073728,
+      "learning_rate": 2e-05,
+      "loss": 1.1778,
+      "step": 11903
+    },
+    {
+      "epoch": 0.2833476149671522,
+      "grad_norm": 7.08011748056226,
+      "learning_rate": 2e-05,
+      "loss": 0.8025,
+      "step": 11904
+    },
+    {
+      "epoch": 0.28337141769018376,
+      "grad_norm": 12.38313569527553,
+      "learning_rate": 2e-05,
+      "loss": 0.5603,
+      "step": 11905
+    },
+    {
+      "epoch": 0.2833952204132153,
+      "grad_norm": 3.5098814327041414,
+      "learning_rate": 2e-05,
+      "loss": 0.3811,
+      "step": 11906
+    },
+    {
+      "epoch": 0.2834190231362468,
+      "grad_norm": 3.3777976353682146,
+      "learning_rate": 2e-05,
+      "loss": 0.1305,
+      "step": 11907
+    },
+    {
+      "epoch": 0.2834428258592783,
+      "grad_norm": 4.597424441550305,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 11908
+    },
+    {
+      "epoch": 0.2834666285823098,
+      "grad_norm": 8.069515454379172,
+      "learning_rate": 2e-05,
+      "loss": 0.8347,
+      "step": 11909
+    },
+    {
+      "epoch": 0.28349043130534135,
+      "grad_norm": 4.03266431149955,
+      "learning_rate": 2e-05,
+      "loss": 0.6339,
+      "step": 11910
+    },
+    {
+      "epoch": 0.28351423402837284,
+      "grad_norm": 5.049938875171866,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 11911
+    },
+    {
+      "epoch": 0.2835380367514044,
+      "grad_norm": 6.981700638315576,
+      "learning_rate": 2e-05,
+      "loss": 0.1834,
+      "step": 11912
+    },
+    {
+      "epoch": 0.28356183947443586,
+      "grad_norm": 3.218047137813175,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 11913
+    },
+    {
+      "epoch": 0.2835856421974674,
+      "grad_norm": 5.176535552237116,
+      "learning_rate": 2e-05,
+      "loss": 0.1833,
+      "step": 11914
+    },
+    {
+      "epoch": 0.2836094449204989,
+      "grad_norm": 2.4942790635532144,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 11915
+    },
+    {
+      "epoch": 0.28363324764353043,
+      "grad_norm": 4.280610403820035,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 11916
+    },
+    {
+      "epoch": 0.2836570503665619,
+      "grad_norm": 7.970457327812798,
+      "learning_rate": 2e-05,
+      "loss": 0.4116,
+      "step": 11917
+    },
+    {
+      "epoch": 0.28368085308959345,
+      "grad_norm": 3.93602553997143,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 11918
+    },
+    {
+      "epoch": 0.28370465581262494,
+      "grad_norm": 1.5026078757436518,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 11919
+    },
+    {
+      "epoch": 0.2837284585356565,
+      "grad_norm": 3.9465622303168217,
+      "learning_rate": 2e-05,
+      "loss": 0.2056,
+      "step": 11920
+    },
+    {
+      "epoch": 0.283752261258688,
+      "grad_norm": 3.6884484639738755,
+      "learning_rate": 2e-05,
+      "loss": 0.5744,
+      "step": 11921
+    },
+    {
+      "epoch": 0.2837760639817195,
+      "grad_norm": 1.3802557411729985,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 11922
+    },
+    {
+      "epoch": 0.28379986670475105,
+      "grad_norm": 6.284976081877956,
+      "learning_rate": 2e-05,
+      "loss": 1.6199,
+      "step": 11923
+    },
+    {
+      "epoch": 0.28382366942778253,
+      "grad_norm": 7.491933097615997,
+      "learning_rate": 2e-05,
+      "loss": 1.0228,
+      "step": 11924
+    },
+    {
+      "epoch": 0.2838474721508141,
+      "grad_norm": 2.948056085717254,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 11925
+    },
+    {
+      "epoch": 0.28387127487384556,
+      "grad_norm": 5.67934187849192,
+      "learning_rate": 2e-05,
+      "loss": 0.2772,
+      "step": 11926
+    },
+    {
+      "epoch": 0.2838950775968771,
+      "grad_norm": 14.801711950796362,
+      "learning_rate": 2e-05,
+      "loss": 0.639,
+      "step": 11927
+    },
+    {
+      "epoch": 0.2839188803199086,
+      "grad_norm": 5.775194430916174,
+      "learning_rate": 2e-05,
+      "loss": 1.2516,
+      "step": 11928
+    },
+    {
+      "epoch": 0.2839426830429401,
+      "grad_norm": 1.6145861138217075,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 11929
+    },
+    {
+      "epoch": 0.2839664857659716,
+      "grad_norm": 84.08040685940107,
+      "learning_rate": 2e-05,
+      "loss": 1.6984,
+      "step": 11930
+    },
+    {
+      "epoch": 0.28399028848900315,
+      "grad_norm": 3.58910259453847,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 11931
+    },
+    {
+      "epoch": 0.28401409121203464,
+      "grad_norm": 5.210827661300013,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 11932
+    },
+    {
+      "epoch": 0.2840378939350662,
+      "grad_norm": 3.20365871481255,
+      "learning_rate": 2e-05,
+      "loss": 0.3607,
+      "step": 11933
+    },
+    {
+      "epoch": 0.28406169665809766,
+      "grad_norm": 3.6909836561391063,
+      "learning_rate": 2e-05,
+      "loss": 0.4871,
+      "step": 11934
+    },
+    {
+      "epoch": 0.2840854993811292,
+      "grad_norm": 7.529487509085523,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 11935
+    },
+    {
+      "epoch": 0.28410930210416074,
+      "grad_norm": 4.7084087253965965,
+      "learning_rate": 2e-05,
+      "loss": 0.9175,
+      "step": 11936
+    },
+    {
+      "epoch": 0.28413310482719223,
+      "grad_norm": 2.9049085569580817,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 11937
+    },
+    {
+      "epoch": 0.28415690755022377,
+      "grad_norm": 8.1179407124681,
+      "learning_rate": 2e-05,
+      "loss": 1.4032,
+      "step": 11938
+    },
+    {
+      "epoch": 0.28418071027325525,
+      "grad_norm": 23.43598036590418,
+      "learning_rate": 2e-05,
+      "loss": 1.0422,
+      "step": 11939
+    },
+    {
+      "epoch": 0.2842045129962868,
+      "grad_norm": 12.553458080213105,
+      "learning_rate": 2e-05,
+      "loss": 0.5441,
+      "step": 11940
+    },
+    {
+      "epoch": 0.2842283157193183,
+      "grad_norm": 4.508218131717259,
+      "learning_rate": 2e-05,
+      "loss": 0.6913,
+      "step": 11941
+    },
+    {
+      "epoch": 0.2842521184423498,
+      "grad_norm": 2.503179155393895,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 11942
+    },
+    {
+      "epoch": 0.2842759211653813,
+      "grad_norm": 1.9987148743744372,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 11943
+    },
+    {
+      "epoch": 0.28429972388841285,
+      "grad_norm": 3.9164015548815097,
+      "learning_rate": 2e-05,
+      "loss": 0.3413,
+      "step": 11944
+    },
+    {
+      "epoch": 0.28432352661144433,
+      "grad_norm": 3.1644516266456053,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 11945
+    },
+    {
+      "epoch": 0.2843473293344759,
+      "grad_norm": 4.98376250781979,
+      "learning_rate": 2e-05,
+      "loss": 0.2672,
+      "step": 11946
+    },
+    {
+      "epoch": 0.28437113205750736,
+      "grad_norm": 2.8949309392836304,
+      "learning_rate": 2e-05,
+      "loss": 0.0966,
+      "step": 11947
+    },
+    {
+      "epoch": 0.2843949347805389,
+      "grad_norm": 5.028657587270716,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 11948
+    },
+    {
+      "epoch": 0.2844187375035704,
+      "grad_norm": 4.876870250119005,
+      "learning_rate": 2e-05,
+      "loss": 0.9572,
+      "step": 11949
+    },
+    {
+      "epoch": 0.2844425402266019,
+      "grad_norm": 1.2209860318341559,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 11950
+    },
+    {
+      "epoch": 0.2844663429496334,
+      "grad_norm": 2.66671752356354,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 11951
+    },
+    {
+      "epoch": 0.28449014567266495,
+      "grad_norm": 3.7605508022799907,
+      "learning_rate": 2e-05,
+      "loss": 0.1618,
+      "step": 11952
+    },
+    {
+      "epoch": 0.2845139483956965,
+      "grad_norm": 6.66081374210736,
+      "learning_rate": 2e-05,
+      "loss": 1.1278,
+      "step": 11953
+    },
+    {
+      "epoch": 0.284537751118728,
+      "grad_norm": 5.653733556514,
+      "learning_rate": 2e-05,
+      "loss": 0.2187,
+      "step": 11954
+    },
+    {
+      "epoch": 0.2845615538417595,
+      "grad_norm": 2.0126572524372235,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 11955
+    },
+    {
+      "epoch": 0.284585356564791,
+      "grad_norm": 2.6235414164710056,
+      "learning_rate": 2e-05,
+      "loss": 0.1822,
+      "step": 11956
+    },
+    {
+      "epoch": 0.28460915928782254,
+      "grad_norm": 2.3128807411084624,
+      "learning_rate": 2e-05,
+      "loss": 0.1742,
+      "step": 11957
+    },
+    {
+      "epoch": 0.28463296201085403,
+      "grad_norm": 0.8089986541682281,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 11958
+    },
+    {
+      "epoch": 0.28465676473388557,
+      "grad_norm": 7.200311516334291,
+      "learning_rate": 2e-05,
+      "loss": 0.6376,
+      "step": 11959
+    },
+    {
+      "epoch": 0.28468056745691706,
+      "grad_norm": 6.4263432318868805,
+      "learning_rate": 2e-05,
+      "loss": 1.3713,
+      "step": 11960
+    },
+    {
+      "epoch": 0.2847043701799486,
+      "grad_norm": 1.7586800301821535,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 11961
+    },
+    {
+      "epoch": 0.2847281729029801,
+      "grad_norm": 3.592845832732884,
+      "learning_rate": 2e-05,
+      "loss": 0.8008,
+      "step": 11962
+    },
+    {
+      "epoch": 0.2847519756260116,
+      "grad_norm": 5.885708530088117,
+      "learning_rate": 2e-05,
+      "loss": 0.779,
+      "step": 11963
+    },
+    {
+      "epoch": 0.2847757783490431,
+      "grad_norm": 8.131170628943428,
+      "learning_rate": 2e-05,
+      "loss": 1.4455,
+      "step": 11964
+    },
+    {
+      "epoch": 0.28479958107207465,
+      "grad_norm": 15.167214786422042,
+      "learning_rate": 2e-05,
+      "loss": 0.4949,
+      "step": 11965
+    },
+    {
+      "epoch": 0.28482338379510613,
+      "grad_norm": 24.938505289438552,
+      "learning_rate": 2e-05,
+      "loss": 0.2958,
+      "step": 11966
+    },
+    {
+      "epoch": 0.2848471865181377,
+      "grad_norm": 2.7144927571023008,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 11967
+    },
+    {
+      "epoch": 0.2848709892411692,
+      "grad_norm": 4.691914709900104,
+      "learning_rate": 2e-05,
+      "loss": 1.0869,
+      "step": 11968
+    },
+    {
+      "epoch": 0.2848947919642007,
+      "grad_norm": 5.159503379083397,
+      "learning_rate": 2e-05,
+      "loss": 0.5637,
+      "step": 11969
+    },
+    {
+      "epoch": 0.28491859468723224,
+      "grad_norm": 6.723735455498076,
+      "learning_rate": 2e-05,
+      "loss": 0.8003,
+      "step": 11970
+    },
+    {
+      "epoch": 0.2849423974102637,
+      "grad_norm": 10.228643983030244,
+      "learning_rate": 2e-05,
+      "loss": 0.4371,
+      "step": 11971
+    },
+    {
+      "epoch": 0.28496620013329527,
+      "grad_norm": 3.3398861988522874,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 11972
+    },
+    {
+      "epoch": 0.28499000285632675,
+      "grad_norm": 4.589868019479437,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 11973
+    },
+    {
+      "epoch": 0.2850138055793583,
+      "grad_norm": 2.125491902646647,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 11974
+    },
+    {
+      "epoch": 0.2850376083023898,
+      "grad_norm": 4.764173475394065,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 11975
+    },
+    {
+      "epoch": 0.2850614110254213,
+      "grad_norm": 6.657347820485296,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 11976
+    },
+    {
+      "epoch": 0.2850852137484528,
+      "grad_norm": 9.668209941239489,
+      "learning_rate": 2e-05,
+      "loss": 0.6015,
+      "step": 11977
+    },
+    {
+      "epoch": 0.28510901647148434,
+      "grad_norm": 3.2283931236234658,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 11978
+    },
+    {
+      "epoch": 0.28513281919451583,
+      "grad_norm": 3.2951796592163336,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 11979
+    },
+    {
+      "epoch": 0.28515662191754737,
+      "grad_norm": 9.831846321259864,
+      "learning_rate": 2e-05,
+      "loss": 0.5712,
+      "step": 11980
+    },
+    {
+      "epoch": 0.28518042464057886,
+      "grad_norm": 5.049294888350551,
+      "learning_rate": 2e-05,
+      "loss": 0.8539,
+      "step": 11981
+    },
+    {
+      "epoch": 0.2852042273636104,
+      "grad_norm": 10.086006544500055,
+      "learning_rate": 2e-05,
+      "loss": 2.2058,
+      "step": 11982
+    },
+    {
+      "epoch": 0.28522803008664194,
+      "grad_norm": 4.402250578965753,
+      "learning_rate": 2e-05,
+      "loss": 0.7025,
+      "step": 11983
+    },
+    {
+      "epoch": 0.2852518328096734,
+      "grad_norm": 5.115378694399307,
+      "learning_rate": 2e-05,
+      "loss": 0.6633,
+      "step": 11984
+    },
+    {
+      "epoch": 0.28527563553270496,
+      "grad_norm": 4.280127489000399,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 11985
+    },
+    {
+      "epoch": 0.28529943825573645,
+      "grad_norm": 1.314934871984868,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 11986
+    },
+    {
+      "epoch": 0.285323240978768,
+      "grad_norm": 9.113123708001536,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 11987
+    },
+    {
+      "epoch": 0.2853470437017995,
+      "grad_norm": 4.689194319440101,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 11988
+    },
+    {
+      "epoch": 0.285370846424831,
+      "grad_norm": 3.8872819623143484,
+      "learning_rate": 2e-05,
+      "loss": 0.5253,
+      "step": 11989
+    },
+    {
+      "epoch": 0.2853946491478625,
+      "grad_norm": 9.922598484610692,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 11990
+    },
+    {
+      "epoch": 0.28541845187089404,
+      "grad_norm": 3.5061903920083055,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 11991
+    },
+    {
+      "epoch": 0.2854422545939255,
+      "grad_norm": 3.0940753473659766,
+      "learning_rate": 2e-05,
+      "loss": 0.1608,
+      "step": 11992
+    },
+    {
+      "epoch": 0.28546605731695707,
+      "grad_norm": 4.838195137540941,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 11993
+    },
+    {
+      "epoch": 0.28548986003998855,
+      "grad_norm": 2.233595542353763,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 11994
+    },
+    {
+      "epoch": 0.2855136627630201,
+      "grad_norm": 1.647471225773737,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 11995
+    },
+    {
+      "epoch": 0.2855374654860516,
+      "grad_norm": 0.947571811040971,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 11996
+    },
+    {
+      "epoch": 0.2855612682090831,
+      "grad_norm": 10.828870736180924,
+      "learning_rate": 2e-05,
+      "loss": 0.3559,
+      "step": 11997
+    },
+    {
+      "epoch": 0.28558507093211466,
+      "grad_norm": 3.1812030003340475,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 11998
+    },
+    {
+      "epoch": 0.28560887365514614,
+      "grad_norm": 8.577385464127643,
+      "learning_rate": 2e-05,
+      "loss": 1.9311,
+      "step": 11999
+    },
+    {
+      "epoch": 0.2856326763781777,
+      "grad_norm": 5.464011132879335,
+      "learning_rate": 2e-05,
+      "loss": 0.8503,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28565647910120917,
+      "grad_norm": 2.90378426556803,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 12001
+    },
+    {
+      "epoch": 0.2856802818242407,
+      "grad_norm": 6.273338094634289,
+      "learning_rate": 2e-05,
+      "loss": 0.2872,
+      "step": 12002
+    },
+    {
+      "epoch": 0.2857040845472722,
+      "grad_norm": 11.461400956132557,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 12003
+    },
+    {
+      "epoch": 0.28572788727030374,
+      "grad_norm": 4.918908686338549,
+      "learning_rate": 2e-05,
+      "loss": 0.8972,
+      "step": 12004
+    },
+    {
+      "epoch": 0.2857516899933352,
+      "grad_norm": 7.770977983512071,
+      "learning_rate": 2e-05,
+      "loss": 0.3193,
+      "step": 12005
+    },
+    {
+      "epoch": 0.28577549271636676,
+      "grad_norm": 3.4152264488209902,
+      "learning_rate": 2e-05,
+      "loss": 0.2824,
+      "step": 12006
+    },
+    {
+      "epoch": 0.28579929543939825,
+      "grad_norm": 4.594881330202638,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 12007
+    },
+    {
+      "epoch": 0.2858230981624298,
+      "grad_norm": 1.2865112237471976,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 12008
+    },
+    {
+      "epoch": 0.2858469008854613,
+      "grad_norm": 4.334840867506973,
+      "learning_rate": 2e-05,
+      "loss": 0.1813,
+      "step": 12009
+    },
+    {
+      "epoch": 0.2858707036084928,
+      "grad_norm": 4.731650340145353,
+      "learning_rate": 2e-05,
+      "loss": 0.6034,
+      "step": 12010
+    },
+    {
+      "epoch": 0.2858945063315243,
+      "grad_norm": 4.299314664405014,
+      "learning_rate": 2e-05,
+      "loss": 0.8731,
+      "step": 12011
+    },
+    {
+      "epoch": 0.28591830905455584,
+      "grad_norm": 3.049757418081465,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 12012
+    },
+    {
+      "epoch": 0.2859421117775874,
+      "grad_norm": 3.73986227930257,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 12013
+    },
+    {
+      "epoch": 0.28596591450061887,
+      "grad_norm": 4.038501687675588,
+      "learning_rate": 2e-05,
+      "loss": 0.1645,
+      "step": 12014
+    },
+    {
+      "epoch": 0.2859897172236504,
+      "grad_norm": 15.657322360704967,
+      "learning_rate": 2e-05,
+      "loss": 0.5006,
+      "step": 12015
+    },
+    {
+      "epoch": 0.2860135199466819,
+      "grad_norm": 5.462281962928581,
+      "learning_rate": 2e-05,
+      "loss": 0.9278,
+      "step": 12016
+    },
+    {
+      "epoch": 0.28603732266971343,
+      "grad_norm": 7.247139272819888,
+      "learning_rate": 2e-05,
+      "loss": 1.1393,
+      "step": 12017
+    },
+    {
+      "epoch": 0.2860611253927449,
+      "grad_norm": 5.397094218066903,
+      "learning_rate": 2e-05,
+      "loss": 0.5458,
+      "step": 12018
+    },
+    {
+      "epoch": 0.28608492811577646,
+      "grad_norm": 3.310269784480724,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 12019
+    },
+    {
+      "epoch": 0.28610873083880795,
+      "grad_norm": 5.486228684563201,
+      "learning_rate": 2e-05,
+      "loss": 0.7612,
+      "step": 12020
+    },
+    {
+      "epoch": 0.2861325335618395,
+      "grad_norm": 1.8477574887878954,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 12021
+    },
+    {
+      "epoch": 0.28615633628487097,
+      "grad_norm": 4.42383820018991,
+      "learning_rate": 2e-05,
+      "loss": 0.3618,
+      "step": 12022
+    },
+    {
+      "epoch": 0.2861801390079025,
+      "grad_norm": 1.6027308199757226,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 12023
+    },
+    {
+      "epoch": 0.286203941730934,
+      "grad_norm": 3.281092372201027,
+      "learning_rate": 2e-05,
+      "loss": 0.6289,
+      "step": 12024
+    },
+    {
+      "epoch": 0.28622774445396554,
+      "grad_norm": 4.4841689410875665,
+      "learning_rate": 2e-05,
+      "loss": 0.61,
+      "step": 12025
+    },
+    {
+      "epoch": 0.286251547176997,
+      "grad_norm": 1.7229818667530452,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 12026
+    },
+    {
+      "epoch": 0.28627534990002856,
+      "grad_norm": 2.751762637763852,
+      "learning_rate": 2e-05,
+      "loss": 0.3865,
+      "step": 12027
+    },
+    {
+      "epoch": 0.2862991526230601,
+      "grad_norm": 5.458636997873173,
+      "learning_rate": 2e-05,
+      "loss": 0.9963,
+      "step": 12028
+    },
+    {
+      "epoch": 0.2863229553460916,
+      "grad_norm": 2.0636596776205964,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 12029
+    },
+    {
+      "epoch": 0.28634675806912313,
+      "grad_norm": 4.824283732126633,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 12030
+    },
+    {
+      "epoch": 0.2863705607921546,
+      "grad_norm": 5.679648385424733,
+      "learning_rate": 2e-05,
+      "loss": 0.4476,
+      "step": 12031
+    },
+    {
+      "epoch": 0.28639436351518616,
+      "grad_norm": 1.557302377785246,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 12032
+    },
+    {
+      "epoch": 0.28641816623821764,
+      "grad_norm": 2.5143900635409087,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 12033
+    },
+    {
+      "epoch": 0.2864419689612492,
+      "grad_norm": 3.0614999312871434,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 12034
+    },
+    {
+      "epoch": 0.28646577168428067,
+      "grad_norm": 3.1802494402354693,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 12035
+    },
+    {
+      "epoch": 0.2864895744073122,
+      "grad_norm": 5.5412589917492605,
+      "learning_rate": 2e-05,
+      "loss": 0.4793,
+      "step": 12036
+    },
+    {
+      "epoch": 0.2865133771303437,
+      "grad_norm": 2.2205163139123827,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 12037
+    },
+    {
+      "epoch": 0.28653717985337523,
+      "grad_norm": 1.4108178823501738,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 12038
+    },
+    {
+      "epoch": 0.2865609825764067,
+      "grad_norm": 4.515253794895962,
+      "learning_rate": 2e-05,
+      "loss": 0.6739,
+      "step": 12039
+    },
+    {
+      "epoch": 0.28658478529943826,
+      "grad_norm": 4.823802126570264,
+      "learning_rate": 2e-05,
+      "loss": 0.7553,
+      "step": 12040
+    },
+    {
+      "epoch": 0.28660858802246975,
+      "grad_norm": 0.5547658667016545,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12041
+    },
+    {
+      "epoch": 0.2866323907455013,
+      "grad_norm": 11.180626860364368,
+      "learning_rate": 2e-05,
+      "loss": 0.3644,
+      "step": 12042
+    },
+    {
+      "epoch": 0.2866561934685328,
+      "grad_norm": 11.504092060011931,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 12043
+    },
+    {
+      "epoch": 0.2866799961915643,
+      "grad_norm": 3.0325465564043874,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 12044
+    },
+    {
+      "epoch": 0.28670379891459585,
+      "grad_norm": 5.228064002961594,
+      "learning_rate": 2e-05,
+      "loss": 1.0541,
+      "step": 12045
+    },
+    {
+      "epoch": 0.28672760163762734,
+      "grad_norm": 11.744944804650773,
+      "learning_rate": 2e-05,
+      "loss": 0.4634,
+      "step": 12046
+    },
+    {
+      "epoch": 0.2867514043606589,
+      "grad_norm": 5.124575766345885,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 12047
+    },
+    {
+      "epoch": 0.28677520708369036,
+      "grad_norm": 8.520270994713862,
+      "learning_rate": 2e-05,
+      "loss": 0.8353,
+      "step": 12048
+    },
+    {
+      "epoch": 0.2867990098067219,
+      "grad_norm": 7.0809272156733165,
+      "learning_rate": 2e-05,
+      "loss": 1.5354,
+      "step": 12049
+    },
+    {
+      "epoch": 0.2868228125297534,
+      "grad_norm": 3.839648355806252,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 12050
+    },
+    {
+      "epoch": 0.28684661525278493,
+      "grad_norm": 7.325536736055905,
+      "learning_rate": 2e-05,
+      "loss": 0.8727,
+      "step": 12051
+    },
+    {
+      "epoch": 0.2868704179758164,
+      "grad_norm": 3.8955024761265715,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 12052
+    },
+    {
+      "epoch": 0.28689422069884796,
+      "grad_norm": 7.75238479757009,
+      "learning_rate": 2e-05,
+      "loss": 0.7234,
+      "step": 12053
+    },
+    {
+      "epoch": 0.28691802342187944,
+      "grad_norm": 5.319294666227626,
+      "learning_rate": 2e-05,
+      "loss": 1.0096,
+      "step": 12054
+    },
+    {
+      "epoch": 0.286941826144911,
+      "grad_norm": 4.188201052526242,
+      "learning_rate": 2e-05,
+      "loss": 0.598,
+      "step": 12055
+    },
+    {
+      "epoch": 0.28696562886794247,
+      "grad_norm": 6.384307290413454,
+      "learning_rate": 2e-05,
+      "loss": 0.8214,
+      "step": 12056
+    },
+    {
+      "epoch": 0.286989431590974,
+      "grad_norm": 7.403361271121736,
+      "learning_rate": 2e-05,
+      "loss": 1.4889,
+      "step": 12057
+    },
+    {
+      "epoch": 0.28701323431400555,
+      "grad_norm": 4.243167317798738,
+      "learning_rate": 2e-05,
+      "loss": 0.2842,
+      "step": 12058
+    },
+    {
+      "epoch": 0.28703703703703703,
+      "grad_norm": 4.798588747377063,
+      "learning_rate": 2e-05,
+      "loss": 0.6866,
+      "step": 12059
+    },
+    {
+      "epoch": 0.2870608397600686,
+      "grad_norm": 1.8335317876021302,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 12060
+    },
+    {
+      "epoch": 0.28708464248310006,
+      "grad_norm": 5.844216372862684,
+      "learning_rate": 2e-05,
+      "loss": 1.4305,
+      "step": 12061
+    },
+    {
+      "epoch": 0.2871084452061316,
+      "grad_norm": 4.674999294376316,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 12062
+    },
+    {
+      "epoch": 0.2871322479291631,
+      "grad_norm": 4.681805850886738,
+      "learning_rate": 2e-05,
+      "loss": 0.5106,
+      "step": 12063
+    },
+    {
+      "epoch": 0.2871560506521946,
+      "grad_norm": 7.462634322295233,
+      "learning_rate": 2e-05,
+      "loss": 0.2536,
+      "step": 12064
+    },
+    {
+      "epoch": 0.2871798533752261,
+      "grad_norm": 6.553134110815243,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 12065
+    },
+    {
+      "epoch": 0.28720365609825765,
+      "grad_norm": 2.915855614147865,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 12066
+    },
+    {
+      "epoch": 0.28722745882128914,
+      "grad_norm": 0.5126720420292625,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 12067
+    },
+    {
+      "epoch": 0.2872512615443207,
+      "grad_norm": 3.766240477607733,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 12068
+    },
+    {
+      "epoch": 0.28727506426735216,
+      "grad_norm": 4.126850176832384,
+      "learning_rate": 2e-05,
+      "loss": 0.6873,
+      "step": 12069
+    },
+    {
+      "epoch": 0.2872988669903837,
+      "grad_norm": 4.4333346111376315,
+      "learning_rate": 2e-05,
+      "loss": 0.553,
+      "step": 12070
+    },
+    {
+      "epoch": 0.2873226697134152,
+      "grad_norm": 4.547198442537721,
+      "learning_rate": 2e-05,
+      "loss": 0.4454,
+      "step": 12071
+    },
+    {
+      "epoch": 0.28734647243644673,
+      "grad_norm": 6.344947361290504,
+      "learning_rate": 2e-05,
+      "loss": 0.6914,
+      "step": 12072
+    },
+    {
+      "epoch": 0.2873702751594782,
+      "grad_norm": 3.367871766583035,
+      "learning_rate": 2e-05,
+      "loss": 0.3204,
+      "step": 12073
+    },
+    {
+      "epoch": 0.28739407788250976,
+      "grad_norm": 4.81756136979604,
+      "learning_rate": 2e-05,
+      "loss": 0.8313,
+      "step": 12074
+    },
+    {
+      "epoch": 0.2874178806055413,
+      "grad_norm": 5.389761305460558,
+      "learning_rate": 2e-05,
+      "loss": 0.6433,
+      "step": 12075
+    },
+    {
+      "epoch": 0.2874416833285728,
+      "grad_norm": 6.091003995000974,
+      "learning_rate": 2e-05,
+      "loss": 0.6789,
+      "step": 12076
+    },
+    {
+      "epoch": 0.2874654860516043,
+      "grad_norm": 4.5853789636295295,
+      "learning_rate": 2e-05,
+      "loss": 0.3077,
+      "step": 12077
+    },
+    {
+      "epoch": 0.2874892887746358,
+      "grad_norm": 3.2156684794877806,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 12078
+    },
+    {
+      "epoch": 0.28751309149766735,
+      "grad_norm": 4.225230565579495,
+      "learning_rate": 2e-05,
+      "loss": 0.784,
+      "step": 12079
+    },
+    {
+      "epoch": 0.28753689422069884,
+      "grad_norm": 3.875627861186175,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 12080
+    },
+    {
+      "epoch": 0.2875606969437304,
+      "grad_norm": 24.999301843618678,
+      "learning_rate": 2e-05,
+      "loss": 0.7389,
+      "step": 12081
+    },
+    {
+      "epoch": 0.28758449966676186,
+      "grad_norm": 5.000065529599356,
+      "learning_rate": 2e-05,
+      "loss": 0.5191,
+      "step": 12082
+    },
+    {
+      "epoch": 0.2876083023897934,
+      "grad_norm": 27.067882866315117,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 12083
+    },
+    {
+      "epoch": 0.2876321051128249,
+      "grad_norm": 4.62474914841651,
+      "learning_rate": 2e-05,
+      "loss": 0.8944,
+      "step": 12084
+    },
+    {
+      "epoch": 0.28765590783585643,
+      "grad_norm": 1.7626439921810457,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 12085
+    },
+    {
+      "epoch": 0.2876797105588879,
+      "grad_norm": 6.525062785460011,
+      "learning_rate": 2e-05,
+      "loss": 1.2286,
+      "step": 12086
+    },
+    {
+      "epoch": 0.28770351328191945,
+      "grad_norm": 4.053606939323172,
+      "learning_rate": 2e-05,
+      "loss": 0.5986,
+      "step": 12087
+    },
+    {
+      "epoch": 0.28772731600495094,
+      "grad_norm": 62.22005402826272,
+      "learning_rate": 2e-05,
+      "loss": 0.6483,
+      "step": 12088
+    },
+    {
+      "epoch": 0.2877511187279825,
+      "grad_norm": 4.715546403941644,
+      "learning_rate": 2e-05,
+      "loss": 0.2196,
+      "step": 12089
+    },
+    {
+      "epoch": 0.287774921451014,
+      "grad_norm": 22.927480426642866,
+      "learning_rate": 2e-05,
+      "loss": 1.0849,
+      "step": 12090
+    },
+    {
+      "epoch": 0.2877987241740455,
+      "grad_norm": 6.626083325447659,
+      "learning_rate": 2e-05,
+      "loss": 0.8333,
+      "step": 12091
+    },
+    {
+      "epoch": 0.28782252689707705,
+      "grad_norm": 0.8625545109451144,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 12092
+    },
+    {
+      "epoch": 0.28784632962010853,
+      "grad_norm": 5.684919585990045,
+      "learning_rate": 2e-05,
+      "loss": 1.1167,
+      "step": 12093
+    },
+    {
+      "epoch": 0.2878701323431401,
+      "grad_norm": 3.1313557407809265,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 12094
+    },
+    {
+      "epoch": 0.28789393506617156,
+      "grad_norm": 5.631619779182217,
+      "learning_rate": 2e-05,
+      "loss": 0.1878,
+      "step": 12095
+    },
+    {
+      "epoch": 0.2879177377892031,
+      "grad_norm": 2.6333836583219465,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 12096
+    },
+    {
+      "epoch": 0.2879415405122346,
+      "grad_norm": 2.231446670705775,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 12097
+    },
+    {
+      "epoch": 0.2879653432352661,
+      "grad_norm": 2.7795193296160283,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 12098
+    },
+    {
+      "epoch": 0.2879891459582976,
+      "grad_norm": 2.368333486647679,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 12099
+    },
+    {
+      "epoch": 0.28801294868132915,
+      "grad_norm": 4.546315157812509,
+      "learning_rate": 2e-05,
+      "loss": 0.5376,
+      "step": 12100
+    },
+    {
+      "epoch": 0.28803675140436064,
+      "grad_norm": 2.0461150130049237,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 12101
+    },
+    {
+      "epoch": 0.2880605541273922,
+      "grad_norm": 4.472149486304001,
+      "learning_rate": 2e-05,
+      "loss": 0.2076,
+      "step": 12102
+    },
+    {
+      "epoch": 0.28808435685042366,
+      "grad_norm": 18.985995504841217,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 12103
+    },
+    {
+      "epoch": 0.2881081595734552,
+      "grad_norm": 4.125083821227356,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 12104
+    },
+    {
+      "epoch": 0.28813196229648674,
+      "grad_norm": 4.130090404118667,
+      "learning_rate": 2e-05,
+      "loss": 0.7246,
+      "step": 12105
+    },
+    {
+      "epoch": 0.28815576501951823,
+      "grad_norm": 4.7497322554953225,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 12106
+    },
+    {
+      "epoch": 0.28817956774254977,
+      "grad_norm": 3.029931233825969,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 12107
+    },
+    {
+      "epoch": 0.28820337046558125,
+      "grad_norm": 8.48818827652687,
+      "learning_rate": 2e-05,
+      "loss": 0.4337,
+      "step": 12108
+    },
+    {
+      "epoch": 0.2882271731886128,
+      "grad_norm": 2.3985551082022227,
+      "learning_rate": 2e-05,
+      "loss": 0.157,
+      "step": 12109
+    },
+    {
+      "epoch": 0.2882509759116443,
+      "grad_norm": 7.704501865979843,
+      "learning_rate": 2e-05,
+      "loss": 0.4691,
+      "step": 12110
+    },
+    {
+      "epoch": 0.2882747786346758,
+      "grad_norm": 6.6186221193436685,
+      "learning_rate": 2e-05,
+      "loss": 0.5896,
+      "step": 12111
+    },
+    {
+      "epoch": 0.2882985813577073,
+      "grad_norm": 2.6313406616570614,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 12112
+    },
+    {
+      "epoch": 0.28832238408073885,
+      "grad_norm": 7.4259568509153455,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 12113
+    },
+    {
+      "epoch": 0.28834618680377033,
+      "grad_norm": 27.27290870034069,
+      "learning_rate": 2e-05,
+      "loss": 0.3052,
+      "step": 12114
+    },
+    {
+      "epoch": 0.2883699895268019,
+      "grad_norm": 7.0581890435961085,
+      "learning_rate": 2e-05,
+      "loss": 0.3046,
+      "step": 12115
+    },
+    {
+      "epoch": 0.28839379224983336,
+      "grad_norm": 1.4379349604039156,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 12116
+    },
+    {
+      "epoch": 0.2884175949728649,
+      "grad_norm": 3.09566796792604,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 12117
+    },
+    {
+      "epoch": 0.2884413976958964,
+      "grad_norm": 4.591514978510395,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 12118
+    },
+    {
+      "epoch": 0.2884652004189279,
+      "grad_norm": 5.362004910477047,
+      "learning_rate": 2e-05,
+      "loss": 0.9251,
+      "step": 12119
+    },
+    {
+      "epoch": 0.28848900314195947,
+      "grad_norm": 0.9439221168824814,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 12120
+    },
+    {
+      "epoch": 0.28851280586499095,
+      "grad_norm": 0.40639696803609426,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 12121
+    },
+    {
+      "epoch": 0.2885366085880225,
+      "grad_norm": 3.1968186357503163,
+      "learning_rate": 2e-05,
+      "loss": 0.4548,
+      "step": 12122
+    },
+    {
+      "epoch": 0.288560411311054,
+      "grad_norm": 4.957652827718739,
+      "learning_rate": 2e-05,
+      "loss": 0.9226,
+      "step": 12123
+    },
+    {
+      "epoch": 0.2885842140340855,
+      "grad_norm": 3.335883584592152,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 12124
+    },
+    {
+      "epoch": 0.288608016757117,
+      "grad_norm": 3.243644260643718,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 12125
+    },
+    {
+      "epoch": 0.28863181948014854,
+      "grad_norm": 5.519522644237711,
+      "learning_rate": 2e-05,
+      "loss": 1.0152,
+      "step": 12126
+    },
+    {
+      "epoch": 0.28865562220318003,
+      "grad_norm": 8.542970841246017,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 12127
+    },
+    {
+      "epoch": 0.28867942492621157,
+      "grad_norm": 3.1372516121293472,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 12128
+    },
+    {
+      "epoch": 0.28870322764924305,
+      "grad_norm": 3.3806335264393295,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 12129
+    },
+    {
+      "epoch": 0.2887270303722746,
+      "grad_norm": 4.780950299565773,
+      "learning_rate": 2e-05,
+      "loss": 0.1868,
+      "step": 12130
+    },
+    {
+      "epoch": 0.2887508330953061,
+      "grad_norm": 10.542598597882078,
+      "learning_rate": 2e-05,
+      "loss": 0.5745,
+      "step": 12131
+    },
+    {
+      "epoch": 0.2887746358183376,
+      "grad_norm": 6.478923591990003,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 12132
+    },
+    {
+      "epoch": 0.2887984385413691,
+      "grad_norm": 0.2788622805683156,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 12133
+    },
+    {
+      "epoch": 0.28882224126440065,
+      "grad_norm": 4.697078597296672,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 12134
+    },
+    {
+      "epoch": 0.2888460439874322,
+      "grad_norm": 6.149628578766171,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 12135
+    },
+    {
+      "epoch": 0.2888698467104637,
+      "grad_norm": 7.093352724539685,
+      "learning_rate": 2e-05,
+      "loss": 0.5115,
+      "step": 12136
+    },
+    {
+      "epoch": 0.2888936494334952,
+      "grad_norm": 3.55607871156122,
+      "learning_rate": 2e-05,
+      "loss": 0.4385,
+      "step": 12137
+    },
+    {
+      "epoch": 0.2889174521565267,
+      "grad_norm": 5.809675566075367,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 12138
+    },
+    {
+      "epoch": 0.28894125487955824,
+      "grad_norm": 3.6325552808344272,
+      "learning_rate": 2e-05,
+      "loss": 0.3882,
+      "step": 12139
+    },
+    {
+      "epoch": 0.2889650576025897,
+      "grad_norm": 2.6494110723591136,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 12140
+    },
+    {
+      "epoch": 0.28898886032562127,
+      "grad_norm": 0.8719583951680687,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 12141
+    },
+    {
+      "epoch": 0.28901266304865275,
+      "grad_norm": 4.310194564813092,
+      "learning_rate": 2e-05,
+      "loss": 0.7821,
+      "step": 12142
+    },
+    {
+      "epoch": 0.2890364657716843,
+      "grad_norm": 17.16976482342267,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 12143
+    },
+    {
+      "epoch": 0.2890602684947158,
+      "grad_norm": 4.094027744183887,
+      "learning_rate": 2e-05,
+      "loss": 0.7898,
+      "step": 12144
+    },
+    {
+      "epoch": 0.2890840712177473,
+      "grad_norm": 1.0460599493432239,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 12145
+    },
+    {
+      "epoch": 0.2891078739407788,
+      "grad_norm": 3.356168334413294,
+      "learning_rate": 2e-05,
+      "loss": 0.3902,
+      "step": 12146
+    },
+    {
+      "epoch": 0.28913167666381034,
+      "grad_norm": 10.857900093271057,
+      "learning_rate": 2e-05,
+      "loss": 1.6555,
+      "step": 12147
+    },
+    {
+      "epoch": 0.28915547938684183,
+      "grad_norm": 6.619290303125557,
+      "learning_rate": 2e-05,
+      "loss": 1.1844,
+      "step": 12148
+    },
+    {
+      "epoch": 0.28917928210987337,
+      "grad_norm": 5.523771995632771,
+      "learning_rate": 2e-05,
+      "loss": 0.5137,
+      "step": 12149
+    },
+    {
+      "epoch": 0.2892030848329049,
+      "grad_norm": 7.927804757265261,
+      "learning_rate": 2e-05,
+      "loss": 0.2517,
+      "step": 12150
+    },
+    {
+      "epoch": 0.2892268875559364,
+      "grad_norm": 3.885866933042249,
+      "learning_rate": 2e-05,
+      "loss": 0.4681,
+      "step": 12151
+    },
+    {
+      "epoch": 0.28925069027896794,
+      "grad_norm": 7.238246491462252,
+      "learning_rate": 2e-05,
+      "loss": 1.4695,
+      "step": 12152
+    },
+    {
+      "epoch": 0.2892744930019994,
+      "grad_norm": 1.7869760221227324,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 12153
+    },
+    {
+      "epoch": 0.28929829572503096,
+      "grad_norm": 2.6020753636532614,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 12154
+    },
+    {
+      "epoch": 0.28932209844806245,
+      "grad_norm": 5.603815810509072,
+      "learning_rate": 2e-05,
+      "loss": 0.5059,
+      "step": 12155
+    },
+    {
+      "epoch": 0.289345901171094,
+      "grad_norm": 1.7916750823151961,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 12156
+    },
+    {
+      "epoch": 0.2893697038941255,
+      "grad_norm": 11.62107680242299,
+      "learning_rate": 2e-05,
+      "loss": 0.7574,
+      "step": 12157
+    },
+    {
+      "epoch": 0.289393506617157,
+      "grad_norm": 1.1829252957049858,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 12158
+    },
+    {
+      "epoch": 0.2894173093401885,
+      "grad_norm": 3.9664791033412223,
+      "learning_rate": 2e-05,
+      "loss": 0.5581,
+      "step": 12159
+    },
+    {
+      "epoch": 0.28944111206322004,
+      "grad_norm": 5.5286221601573615,
+      "learning_rate": 2e-05,
+      "loss": 1.1524,
+      "step": 12160
+    },
+    {
+      "epoch": 0.2894649147862515,
+      "grad_norm": 12.543111474765023,
+      "learning_rate": 2e-05,
+      "loss": 1.109,
+      "step": 12161
+    },
+    {
+      "epoch": 0.28948871750928307,
+      "grad_norm": 2.1909117968404916,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 12162
+    },
+    {
+      "epoch": 0.28951252023231455,
+      "grad_norm": 3.3799649917941403,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 12163
+    },
+    {
+      "epoch": 0.2895363229553461,
+      "grad_norm": 2.3956832401346624,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 12164
+    },
+    {
+      "epoch": 0.28956012567837763,
+      "grad_norm": 4.801588402876046,
+      "learning_rate": 2e-05,
+      "loss": 0.6566,
+      "step": 12165
+    },
+    {
+      "epoch": 0.2895839284014091,
+      "grad_norm": 7.861918256226013,
+      "learning_rate": 2e-05,
+      "loss": 0.5209,
+      "step": 12166
+    },
+    {
+      "epoch": 0.28960773112444066,
+      "grad_norm": 2.0475346776381067,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 12167
+    },
+    {
+      "epoch": 0.28963153384747214,
+      "grad_norm": 5.103168409056913,
+      "learning_rate": 2e-05,
+      "loss": 0.8798,
+      "step": 12168
+    },
+    {
+      "epoch": 0.2896553365705037,
+      "grad_norm": 4.484482123637833,
+      "learning_rate": 2e-05,
+      "loss": 0.5519,
+      "step": 12169
+    },
+    {
+      "epoch": 0.28967913929353517,
+      "grad_norm": 3.4483024008370347,
+      "learning_rate": 2e-05,
+      "loss": 0.3708,
+      "step": 12170
+    },
+    {
+      "epoch": 0.2897029420165667,
+      "grad_norm": 9.821077540749588,
+      "learning_rate": 2e-05,
+      "loss": 0.1334,
+      "step": 12171
+    },
+    {
+      "epoch": 0.2897267447395982,
+      "grad_norm": 5.846872542162468,
+      "learning_rate": 2e-05,
+      "loss": 0.4084,
+      "step": 12172
+    },
+    {
+      "epoch": 0.28975054746262974,
+      "grad_norm": 3.9909337179046735,
+      "learning_rate": 2e-05,
+      "loss": 0.4457,
+      "step": 12173
+    },
+    {
+      "epoch": 0.2897743501856612,
+      "grad_norm": 1.4324439958585407,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 12174
+    },
+    {
+      "epoch": 0.28979815290869276,
+      "grad_norm": 37.66293779862574,
+      "learning_rate": 2e-05,
+      "loss": 2.1016,
+      "step": 12175
+    },
+    {
+      "epoch": 0.28982195563172425,
+      "grad_norm": 2.9346296893988644,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 12176
+    },
+    {
+      "epoch": 0.2898457583547558,
+      "grad_norm": 2.7752242394988613,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 12177
+    },
+    {
+      "epoch": 0.2898695610777873,
+      "grad_norm": 2.0136195815096527,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 12178
+    },
+    {
+      "epoch": 0.2898933638008188,
+      "grad_norm": 0.37983626737667847,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 12179
+    },
+    {
+      "epoch": 0.28991716652385036,
+      "grad_norm": 9.936222049324968,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 12180
+    },
+    {
+      "epoch": 0.28994096924688184,
+      "grad_norm": 2.001978309493753,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 12181
+    },
+    {
+      "epoch": 0.2899647719699134,
+      "grad_norm": 10.420805967717984,
+      "learning_rate": 2e-05,
+      "loss": 1.2145,
+      "step": 12182
+    },
+    {
+      "epoch": 0.28998857469294487,
+      "grad_norm": 26.955642896812478,
+      "learning_rate": 2e-05,
+      "loss": 0.5034,
+      "step": 12183
+    },
+    {
+      "epoch": 0.2900123774159764,
+      "grad_norm": 146.2156356136437,
+      "learning_rate": 2e-05,
+      "loss": 0.6596,
+      "step": 12184
+    },
+    {
+      "epoch": 0.2900361801390079,
+      "grad_norm": 2.1472821878319572,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 12185
+    },
+    {
+      "epoch": 0.29005998286203943,
+      "grad_norm": 2.9606201584452343,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 12186
+    },
+    {
+      "epoch": 0.2900837855850709,
+      "grad_norm": 6.029001929264598,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 12187
+    },
+    {
+      "epoch": 0.29010758830810246,
+      "grad_norm": 5.216052729615479,
+      "learning_rate": 2e-05,
+      "loss": 0.5929,
+      "step": 12188
+    },
+    {
+      "epoch": 0.29013139103113395,
+      "grad_norm": 2.88601938522832,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 12189
+    },
+    {
+      "epoch": 0.2901551937541655,
+      "grad_norm": 11.75955237273921,
+      "learning_rate": 2e-05,
+      "loss": 0.638,
+      "step": 12190
+    },
+    {
+      "epoch": 0.29017899647719697,
+      "grad_norm": 0.72642774952607,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 12191
+    },
+    {
+      "epoch": 0.2902027992002285,
+      "grad_norm": 3.785328696107514,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 12192
+    },
+    {
+      "epoch": 0.29022660192326,
+      "grad_norm": 4.607281682191526,
+      "learning_rate": 2e-05,
+      "loss": 0.4349,
+      "step": 12193
+    },
+    {
+      "epoch": 0.29025040464629154,
+      "grad_norm": 5.822594788508345,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 12194
+    },
+    {
+      "epoch": 0.290274207369323,
+      "grad_norm": 6.410763560402356,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 12195
+    },
+    {
+      "epoch": 0.29029801009235456,
+      "grad_norm": 4.168141645543988,
+      "learning_rate": 2e-05,
+      "loss": 0.3234,
+      "step": 12196
+    },
+    {
+      "epoch": 0.2903218128153861,
+      "grad_norm": 4.203341574214985,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 12197
+    },
+    {
+      "epoch": 0.2903456155384176,
+      "grad_norm": 14.504687469324189,
+      "learning_rate": 2e-05,
+      "loss": 0.6917,
+      "step": 12198
+    },
+    {
+      "epoch": 0.29036941826144913,
+      "grad_norm": 43.90414946326434,
+      "learning_rate": 2e-05,
+      "loss": 1.0397,
+      "step": 12199
+    },
+    {
+      "epoch": 0.2903932209844806,
+      "grad_norm": 10.91578817890281,
+      "learning_rate": 2e-05,
+      "loss": 0.6114,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29041702370751216,
+      "grad_norm": 2.9151631657791754,
+      "learning_rate": 2e-05,
+      "loss": 0.1319,
+      "step": 12201
+    },
+    {
+      "epoch": 0.29044082643054364,
+      "grad_norm": 12.17623249343706,
+      "learning_rate": 2e-05,
+      "loss": 1.5073,
+      "step": 12202
+    },
+    {
+      "epoch": 0.2904646291535752,
+      "grad_norm": 1.8709329104878292,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 12203
+    },
+    {
+      "epoch": 0.29048843187660667,
+      "grad_norm": 4.543326102184337,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 12204
+    },
+    {
+      "epoch": 0.2905122345996382,
+      "grad_norm": 11.831441964106318,
+      "learning_rate": 2e-05,
+      "loss": 0.3499,
+      "step": 12205
+    },
+    {
+      "epoch": 0.2905360373226697,
+      "grad_norm": 3.4169945778688953,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 12206
+    },
+    {
+      "epoch": 0.29055984004570123,
+      "grad_norm": 3.55618015763461,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 12207
+    },
+    {
+      "epoch": 0.2905836427687327,
+      "grad_norm": 1.4086184923577267,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 12208
+    },
+    {
+      "epoch": 0.29060744549176426,
+      "grad_norm": 1.5918246410660852,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 12209
+    },
+    {
+      "epoch": 0.29063124821479575,
+      "grad_norm": 4.290599657099159,
+      "learning_rate": 2e-05,
+      "loss": 0.4413,
+      "step": 12210
+    },
+    {
+      "epoch": 0.2906550509378273,
+      "grad_norm": 4.716845041980409,
+      "learning_rate": 2e-05,
+      "loss": 0.7513,
+      "step": 12211
+    },
+    {
+      "epoch": 0.2906788536608588,
+      "grad_norm": 4.02902531822464,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 12212
+    },
+    {
+      "epoch": 0.2907026563838903,
+      "grad_norm": 18.766565980941152,
+      "learning_rate": 2e-05,
+      "loss": 2.322,
+      "step": 12213
+    },
+    {
+      "epoch": 0.29072645910692185,
+      "grad_norm": 2.4971345262075806,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 12214
+    },
+    {
+      "epoch": 0.29075026182995334,
+      "grad_norm": 3.710428788241158,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 12215
+    },
+    {
+      "epoch": 0.2907740645529849,
+      "grad_norm": 2.687878936190623,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 12216
+    },
+    {
+      "epoch": 0.29079786727601636,
+      "grad_norm": 2.5307227643986794,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 12217
+    },
+    {
+      "epoch": 0.2908216699990479,
+      "grad_norm": 9.251562294000415,
+      "learning_rate": 2e-05,
+      "loss": 0.4975,
+      "step": 12218
+    },
+    {
+      "epoch": 0.2908454727220794,
+      "grad_norm": 5.547042390328072,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 12219
+    },
+    {
+      "epoch": 0.29086927544511093,
+      "grad_norm": 1.368576876287689,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 12220
+    },
+    {
+      "epoch": 0.2908930781681424,
+      "grad_norm": 12.255057538112881,
+      "learning_rate": 2e-05,
+      "loss": 1.1027,
+      "step": 12221
+    },
+    {
+      "epoch": 0.29091688089117396,
+      "grad_norm": 4.442590761902889,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 12222
+    },
+    {
+      "epoch": 0.29094068361420544,
+      "grad_norm": 3.9854535804478073,
+      "learning_rate": 2e-05,
+      "loss": 0.1721,
+      "step": 12223
+    },
+    {
+      "epoch": 0.290964486337237,
+      "grad_norm": 6.84404910023293,
+      "learning_rate": 2e-05,
+      "loss": 2.2531,
+      "step": 12224
+    },
+    {
+      "epoch": 0.29098828906026847,
+      "grad_norm": 4.596919825160325,
+      "learning_rate": 2e-05,
+      "loss": 0.409,
+      "step": 12225
+    },
+    {
+      "epoch": 0.2910120917833,
+      "grad_norm": 3.006869073920232,
+      "learning_rate": 2e-05,
+      "loss": 0.2268,
+      "step": 12226
+    },
+    {
+      "epoch": 0.29103589450633155,
+      "grad_norm": 8.847614348414906,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 12227
+    },
+    {
+      "epoch": 0.29105969722936303,
+      "grad_norm": 1.3010778508070933,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 12228
+    },
+    {
+      "epoch": 0.2910834999523946,
+      "grad_norm": 2.623999831945143,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 12229
+    },
+    {
+      "epoch": 0.29110730267542606,
+      "grad_norm": 2.9586901004357733,
+      "learning_rate": 2e-05,
+      "loss": 0.3691,
+      "step": 12230
+    },
+    {
+      "epoch": 0.2911311053984576,
+      "grad_norm": 8.157676323167854,
+      "learning_rate": 2e-05,
+      "loss": 0.6417,
+      "step": 12231
+    },
+    {
+      "epoch": 0.2911549081214891,
+      "grad_norm": 1.6310883756516485,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 12232
+    },
+    {
+      "epoch": 0.2911787108445206,
+      "grad_norm": 2.637370990563297,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 12233
+    },
+    {
+      "epoch": 0.2912025135675521,
+      "grad_norm": 7.327284545249554,
+      "learning_rate": 2e-05,
+      "loss": 0.2963,
+      "step": 12234
+    },
+    {
+      "epoch": 0.29122631629058365,
+      "grad_norm": 4.43031674410665,
+      "learning_rate": 2e-05,
+      "loss": 0.1976,
+      "step": 12235
+    },
+    {
+      "epoch": 0.29125011901361514,
+      "grad_norm": 3.4351116340225687,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 12236
+    },
+    {
+      "epoch": 0.2912739217366467,
+      "grad_norm": 6.435780564319943,
+      "learning_rate": 2e-05,
+      "loss": 0.8163,
+      "step": 12237
+    },
+    {
+      "epoch": 0.29129772445967816,
+      "grad_norm": 3.8333494636123806,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 12238
+    },
+    {
+      "epoch": 0.2913215271827097,
+      "grad_norm": 1.9544197405764847,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 12239
+    },
+    {
+      "epoch": 0.2913453299057412,
+      "grad_norm": 7.7479802009654275,
+      "learning_rate": 2e-05,
+      "loss": 1.2766,
+      "step": 12240
+    },
+    {
+      "epoch": 0.29136913262877273,
+      "grad_norm": 2.2283143859507506,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 12241
+    },
+    {
+      "epoch": 0.29139293535180427,
+      "grad_norm": 5.015731921690985,
+      "learning_rate": 2e-05,
+      "loss": 0.5928,
+      "step": 12242
+    },
+    {
+      "epoch": 0.29141673807483576,
+      "grad_norm": 1.9138057921456562,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 12243
+    },
+    {
+      "epoch": 0.2914405407978673,
+      "grad_norm": 2.2560552028552663,
+      "learning_rate": 2e-05,
+      "loss": 0.2433,
+      "step": 12244
+    },
+    {
+      "epoch": 0.2914643435208988,
+      "grad_norm": 3.993604909888202,
+      "learning_rate": 2e-05,
+      "loss": 0.0952,
+      "step": 12245
+    },
+    {
+      "epoch": 0.2914881462439303,
+      "grad_norm": 4.177382078361939,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 12246
+    },
+    {
+      "epoch": 0.2915119489669618,
+      "grad_norm": 7.382173788971998,
+      "learning_rate": 2e-05,
+      "loss": 1.6217,
+      "step": 12247
+    },
+    {
+      "epoch": 0.29153575168999335,
+      "grad_norm": 2.7217587647480004,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 12248
+    },
+    {
+      "epoch": 0.29155955441302484,
+      "grad_norm": 10.195011493913528,
+      "learning_rate": 2e-05,
+      "loss": 0.3399,
+      "step": 12249
+    },
+    {
+      "epoch": 0.2915833571360564,
+      "grad_norm": 4.4292265085198395,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 12250
+    },
+    {
+      "epoch": 0.29160715985908786,
+      "grad_norm": 1.6109716691679377,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 12251
+    },
+    {
+      "epoch": 0.2916309625821194,
+      "grad_norm": 1.9643110893779998,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 12252
+    },
+    {
+      "epoch": 0.2916547653051509,
+      "grad_norm": 8.327876226974189,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 12253
+    },
+    {
+      "epoch": 0.29167856802818243,
+      "grad_norm": 7.85306331647293,
+      "learning_rate": 2e-05,
+      "loss": 1.7473,
+      "step": 12254
+    },
+    {
+      "epoch": 0.2917023707512139,
+      "grad_norm": 11.875265601374158,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 12255
+    },
+    {
+      "epoch": 0.29172617347424545,
+      "grad_norm": 6.893146238488362,
+      "learning_rate": 2e-05,
+      "loss": 1.6091,
+      "step": 12256
+    },
+    {
+      "epoch": 0.291749976197277,
+      "grad_norm": 1.934235478937473,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 12257
+    },
+    {
+      "epoch": 0.2917737789203085,
+      "grad_norm": 5.30449783105045,
+      "learning_rate": 2e-05,
+      "loss": 0.183,
+      "step": 12258
+    },
+    {
+      "epoch": 0.29179758164334,
+      "grad_norm": 2.1543609550226077,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 12259
+    },
+    {
+      "epoch": 0.2918213843663715,
+      "grad_norm": 4.782831554471754,
+      "learning_rate": 2e-05,
+      "loss": 0.8234,
+      "step": 12260
+    },
+    {
+      "epoch": 0.29184518708940305,
+      "grad_norm": 5.3253101161226155,
+      "learning_rate": 2e-05,
+      "loss": 0.4314,
+      "step": 12261
+    },
+    {
+      "epoch": 0.29186898981243453,
+      "grad_norm": 1.6277770317868394,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 12262
+    },
+    {
+      "epoch": 0.2918927925354661,
+      "grad_norm": 1.0629144834843787,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 12263
+    },
+    {
+      "epoch": 0.29191659525849756,
+      "grad_norm": 5.855484031647245,
+      "learning_rate": 2e-05,
+      "loss": 0.212,
+      "step": 12264
+    },
+    {
+      "epoch": 0.2919403979815291,
+      "grad_norm": 1.9017664476076226,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 12265
+    },
+    {
+      "epoch": 0.2919642007045606,
+      "grad_norm": 3.5599513368132576,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 12266
+    },
+    {
+      "epoch": 0.2919880034275921,
+      "grad_norm": 5.547488525707618,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 12267
+    },
+    {
+      "epoch": 0.2920118061506236,
+      "grad_norm": 4.455193578796918,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 12268
+    },
+    {
+      "epoch": 0.29203560887365515,
+      "grad_norm": 5.1612725826688735,
+      "learning_rate": 2e-05,
+      "loss": 0.2157,
+      "step": 12269
+    },
+    {
+      "epoch": 0.29205941159668664,
+      "grad_norm": 4.910014410018029,
+      "learning_rate": 2e-05,
+      "loss": 0.1698,
+      "step": 12270
+    },
+    {
+      "epoch": 0.2920832143197182,
+      "grad_norm": 7.809187610644821,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 12271
+    },
+    {
+      "epoch": 0.2921070170427497,
+      "grad_norm": 0.8371058249444,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 12272
+    },
+    {
+      "epoch": 0.2921308197657812,
+      "grad_norm": 5.15686538653153,
+      "learning_rate": 2e-05,
+      "loss": 0.9169,
+      "step": 12273
+    },
+    {
+      "epoch": 0.29215462248881274,
+      "grad_norm": 6.091516874515608,
+      "learning_rate": 2e-05,
+      "loss": 0.6658,
+      "step": 12274
+    },
+    {
+      "epoch": 0.29217842521184423,
+      "grad_norm": 19.853929437765572,
+      "learning_rate": 2e-05,
+      "loss": 0.7407,
+      "step": 12275
+    },
+    {
+      "epoch": 0.29220222793487577,
+      "grad_norm": 4.23139010662161,
+      "learning_rate": 2e-05,
+      "loss": 0.3389,
+      "step": 12276
+    },
+    {
+      "epoch": 0.29222603065790725,
+      "grad_norm": 5.243673554856454,
+      "learning_rate": 2e-05,
+      "loss": 0.984,
+      "step": 12277
+    },
+    {
+      "epoch": 0.2922498333809388,
+      "grad_norm": 3.047483071210149,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 12278
+    },
+    {
+      "epoch": 0.2922736361039703,
+      "grad_norm": 1.9707319875768758,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 12279
+    },
+    {
+      "epoch": 0.2922974388270018,
+      "grad_norm": 4.606730553278481,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 12280
+    },
+    {
+      "epoch": 0.2923212415500333,
+      "grad_norm": 1.4403181817696749,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 12281
+    },
+    {
+      "epoch": 0.29234504427306485,
+      "grad_norm": 16.151105715250512,
+      "learning_rate": 2e-05,
+      "loss": 1.4744,
+      "step": 12282
+    },
+    {
+      "epoch": 0.29236884699609633,
+      "grad_norm": 2.4214027710582156,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 12283
+    },
+    {
+      "epoch": 0.2923926497191279,
+      "grad_norm": 2.8194053136707145,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 12284
+    },
+    {
+      "epoch": 0.29241645244215936,
+      "grad_norm": 2.4882591953184825,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 12285
+    },
+    {
+      "epoch": 0.2924402551651909,
+      "grad_norm": 3.321680050087449,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 12286
+    },
+    {
+      "epoch": 0.29246405788822244,
+      "grad_norm": 10.850382868588378,
+      "learning_rate": 2e-05,
+      "loss": 1.0039,
+      "step": 12287
+    },
+    {
+      "epoch": 0.2924878606112539,
+      "grad_norm": 4.14817634176349,
+      "learning_rate": 2e-05,
+      "loss": 0.6559,
+      "step": 12288
+    },
+    {
+      "epoch": 0.29251166333428547,
+      "grad_norm": 12.769016887473265,
+      "learning_rate": 2e-05,
+      "loss": 0.7869,
+      "step": 12289
+    },
+    {
+      "epoch": 0.29253546605731695,
+      "grad_norm": 3.0200269640320014,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 12290
+    },
+    {
+      "epoch": 0.2925592687803485,
+      "grad_norm": 3.9726660334629624,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 12291
+    },
+    {
+      "epoch": 0.29258307150338,
+      "grad_norm": 2.7844574762179315,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 12292
+    },
+    {
+      "epoch": 0.2926068742264115,
+      "grad_norm": 5.237553333377671,
+      "learning_rate": 2e-05,
+      "loss": 0.696,
+      "step": 12293
+    },
+    {
+      "epoch": 0.292630676949443,
+      "grad_norm": 6.794081084425174,
+      "learning_rate": 2e-05,
+      "loss": 0.7822,
+      "step": 12294
+    },
+    {
+      "epoch": 0.29265447967247454,
+      "grad_norm": 2.4071142433190724,
+      "learning_rate": 2e-05,
+      "loss": 0.1118,
+      "step": 12295
+    },
+    {
+      "epoch": 0.29267828239550603,
+      "grad_norm": 8.420425474121657,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 12296
+    },
+    {
+      "epoch": 0.29270208511853757,
+      "grad_norm": 13.219754444978362,
+      "learning_rate": 2e-05,
+      "loss": 0.9342,
+      "step": 12297
+    },
+    {
+      "epoch": 0.29272588784156905,
+      "grad_norm": 17.94290459261037,
+      "learning_rate": 2e-05,
+      "loss": 0.9021,
+      "step": 12298
+    },
+    {
+      "epoch": 0.2927496905646006,
+      "grad_norm": 2.5983468907017224,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 12299
+    },
+    {
+      "epoch": 0.2927734932876321,
+      "grad_norm": 2.1023381477591303,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 12300
+    },
+    {
+      "epoch": 0.2927972960106636,
+      "grad_norm": 5.986525298732536,
+      "learning_rate": 2e-05,
+      "loss": 0.2963,
+      "step": 12301
+    },
+    {
+      "epoch": 0.29282109873369516,
+      "grad_norm": 4.567508351323876,
+      "learning_rate": 2e-05,
+      "loss": 0.8859,
+      "step": 12302
+    },
+    {
+      "epoch": 0.29284490145672665,
+      "grad_norm": 15.856063298184715,
+      "learning_rate": 2e-05,
+      "loss": 1.3877,
+      "step": 12303
+    },
+    {
+      "epoch": 0.2928687041797582,
+      "grad_norm": 2.8591052486000477,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 12304
+    },
+    {
+      "epoch": 0.2928925069027897,
+      "grad_norm": 7.130093303346692,
+      "learning_rate": 2e-05,
+      "loss": 0.7662,
+      "step": 12305
+    },
+    {
+      "epoch": 0.2929163096258212,
+      "grad_norm": 3.877417064273813,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 12306
+    },
+    {
+      "epoch": 0.2929401123488527,
+      "grad_norm": 7.574520780660509,
+      "learning_rate": 2e-05,
+      "loss": 0.2653,
+      "step": 12307
+    },
+    {
+      "epoch": 0.29296391507188424,
+      "grad_norm": 1.4287517960465312,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 12308
+    },
+    {
+      "epoch": 0.2929877177949157,
+      "grad_norm": 3.5071794447990277,
+      "learning_rate": 2e-05,
+      "loss": 0.3545,
+      "step": 12309
+    },
+    {
+      "epoch": 0.29301152051794727,
+      "grad_norm": 4.2880126026808005,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 12310
+    },
+    {
+      "epoch": 0.29303532324097875,
+      "grad_norm": 5.102420119164222,
+      "learning_rate": 2e-05,
+      "loss": 0.6229,
+      "step": 12311
+    },
+    {
+      "epoch": 0.2930591259640103,
+      "grad_norm": 3.0355273728512358,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 12312
+    },
+    {
+      "epoch": 0.2930829286870418,
+      "grad_norm": 3.4512989870345776,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 12313
+    },
+    {
+      "epoch": 0.2931067314100733,
+      "grad_norm": 1.0899563126991307,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 12314
+    },
+    {
+      "epoch": 0.2931305341331048,
+      "grad_norm": 10.454367337635261,
+      "learning_rate": 2e-05,
+      "loss": 0.497,
+      "step": 12315
+    },
+    {
+      "epoch": 0.29315433685613634,
+      "grad_norm": 4.310295006838718,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 12316
+    },
+    {
+      "epoch": 0.29317813957916783,
+      "grad_norm": 6.88320319458439,
+      "learning_rate": 2e-05,
+      "loss": 0.723,
+      "step": 12317
+    },
+    {
+      "epoch": 0.29320194230219937,
+      "grad_norm": 3.8724594226971343,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 12318
+    },
+    {
+      "epoch": 0.2932257450252309,
+      "grad_norm": 4.05448273647799,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 12319
+    },
+    {
+      "epoch": 0.2932495477482624,
+      "grad_norm": 2.839115448502951,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 12320
+    },
+    {
+      "epoch": 0.29327335047129394,
+      "grad_norm": 5.34791079763165,
+      "learning_rate": 2e-05,
+      "loss": 0.6538,
+      "step": 12321
+    },
+    {
+      "epoch": 0.2932971531943254,
+      "grad_norm": 1.7257803200164938,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 12322
+    },
+    {
+      "epoch": 0.29332095591735696,
+      "grad_norm": 7.817340627283633,
+      "learning_rate": 2e-05,
+      "loss": 0.5856,
+      "step": 12323
+    },
+    {
+      "epoch": 0.29334475864038845,
+      "grad_norm": 7.083171416181168,
+      "learning_rate": 2e-05,
+      "loss": 0.9711,
+      "step": 12324
+    },
+    {
+      "epoch": 0.29336856136342,
+      "grad_norm": 3.8834628575298606,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 12325
+    },
+    {
+      "epoch": 0.2933923640864515,
+      "grad_norm": 5.460801479492033,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 12326
+    },
+    {
+      "epoch": 0.293416166809483,
+      "grad_norm": 5.821835326265566,
+      "learning_rate": 2e-05,
+      "loss": 1.4601,
+      "step": 12327
+    },
+    {
+      "epoch": 0.2934399695325145,
+      "grad_norm": 30.73900298402793,
+      "learning_rate": 2e-05,
+      "loss": 1.3375,
+      "step": 12328
+    },
+    {
+      "epoch": 0.29346377225554604,
+      "grad_norm": 5.4675079634875745,
+      "learning_rate": 2e-05,
+      "loss": 0.9063,
+      "step": 12329
+    },
+    {
+      "epoch": 0.2934875749785775,
+      "grad_norm": 4.353949158883181,
+      "learning_rate": 2e-05,
+      "loss": 0.235,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29351137770160907,
+      "grad_norm": 3.8118041722306537,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 12331
+    },
+    {
+      "epoch": 0.29353518042464055,
+      "grad_norm": 3.965065279249192,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 12332
+    },
+    {
+      "epoch": 0.2935589831476721,
+      "grad_norm": 3.290197285540633,
+      "learning_rate": 2e-05,
+      "loss": 0.4328,
+      "step": 12333
+    },
+    {
+      "epoch": 0.29358278587070363,
+      "grad_norm": 5.330610585257796,
+      "learning_rate": 2e-05,
+      "loss": 0.6123,
+      "step": 12334
+    },
+    {
+      "epoch": 0.2936065885937351,
+      "grad_norm": 9.666712109263516,
+      "learning_rate": 2e-05,
+      "loss": 1.1961,
+      "step": 12335
+    },
+    {
+      "epoch": 0.29363039131676666,
+      "grad_norm": 8.16286010303848,
+      "learning_rate": 2e-05,
+      "loss": 0.3275,
+      "step": 12336
+    },
+    {
+      "epoch": 0.29365419403979814,
+      "grad_norm": 1.5839884666462485,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 12337
+    },
+    {
+      "epoch": 0.2936779967628297,
+      "grad_norm": 4.1449342297518195,
+      "learning_rate": 2e-05,
+      "loss": 0.2662,
+      "step": 12338
+    },
+    {
+      "epoch": 0.29370179948586117,
+      "grad_norm": 2.7363844327598916,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 12339
+    },
+    {
+      "epoch": 0.2937256022088927,
+      "grad_norm": 5.5262971154710065,
+      "learning_rate": 2e-05,
+      "loss": 0.3044,
+      "step": 12340
+    },
+    {
+      "epoch": 0.2937494049319242,
+      "grad_norm": 1.8991262209121875,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 12341
+    },
+    {
+      "epoch": 0.29377320765495574,
+      "grad_norm": 3.1956050807435656,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 12342
+    },
+    {
+      "epoch": 0.2937970103779872,
+      "grad_norm": 8.609151581952894,
+      "learning_rate": 2e-05,
+      "loss": 0.3891,
+      "step": 12343
+    },
+    {
+      "epoch": 0.29382081310101876,
+      "grad_norm": 3.521079992076892,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 12344
+    },
+    {
+      "epoch": 0.29384461582405025,
+      "grad_norm": 1.9300337390451319,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 12345
+    },
+    {
+      "epoch": 0.2938684185470818,
+      "grad_norm": 2.8311044239416012,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 12346
+    },
+    {
+      "epoch": 0.2938922212701133,
+      "grad_norm": 1.3850875346877574,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 12347
+    },
+    {
+      "epoch": 0.2939160239931448,
+      "grad_norm": 9.029934737531791,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 12348
+    },
+    {
+      "epoch": 0.29393982671617636,
+      "grad_norm": 3.8701648184089206,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 12349
+    },
+    {
+      "epoch": 0.29396362943920784,
+      "grad_norm": 4.086415987342684,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 12350
+    },
+    {
+      "epoch": 0.2939874321622394,
+      "grad_norm": 10.230616353224086,
+      "learning_rate": 2e-05,
+      "loss": 0.4499,
+      "step": 12351
+    },
+    {
+      "epoch": 0.29401123488527087,
+      "grad_norm": 6.954013441235883,
+      "learning_rate": 2e-05,
+      "loss": 1.064,
+      "step": 12352
+    },
+    {
+      "epoch": 0.2940350376083024,
+      "grad_norm": 12.061120878403965,
+      "learning_rate": 2e-05,
+      "loss": 1.2644,
+      "step": 12353
+    },
+    {
+      "epoch": 0.2940588403313339,
+      "grad_norm": 6.169236955304224,
+      "learning_rate": 2e-05,
+      "loss": 0.7429,
+      "step": 12354
+    },
+    {
+      "epoch": 0.29408264305436543,
+      "grad_norm": 7.969135525899818,
+      "learning_rate": 2e-05,
+      "loss": 0.4438,
+      "step": 12355
+    },
+    {
+      "epoch": 0.2941064457773969,
+      "grad_norm": 2.7891265576100417,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 12356
+    },
+    {
+      "epoch": 0.29413024850042846,
+      "grad_norm": 8.346364334521391,
+      "learning_rate": 2e-05,
+      "loss": 0.3263,
+      "step": 12357
+    },
+    {
+      "epoch": 0.29415405122345994,
+      "grad_norm": 7.565012996438606,
+      "learning_rate": 2e-05,
+      "loss": 0.2987,
+      "step": 12358
+    },
+    {
+      "epoch": 0.2941778539464915,
+      "grad_norm": 2.260394555250346,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 12359
+    },
+    {
+      "epoch": 0.29420165666952297,
+      "grad_norm": 5.194514261773092,
+      "learning_rate": 2e-05,
+      "loss": 0.8031,
+      "step": 12360
+    },
+    {
+      "epoch": 0.2942254593925545,
+      "grad_norm": 2.240129621173681,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 12361
+    },
+    {
+      "epoch": 0.294249262115586,
+      "grad_norm": 4.540449504705103,
+      "learning_rate": 2e-05,
+      "loss": 0.1415,
+      "step": 12362
+    },
+    {
+      "epoch": 0.29427306483861754,
+      "grad_norm": 6.083001632208708,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 12363
+    },
+    {
+      "epoch": 0.2942968675616491,
+      "grad_norm": 1.7264446563760203,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 12364
+    },
+    {
+      "epoch": 0.29432067028468056,
+      "grad_norm": 1.5822600768135973,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 12365
+    },
+    {
+      "epoch": 0.2943444730077121,
+      "grad_norm": 9.421806929771291,
+      "learning_rate": 2e-05,
+      "loss": 0.537,
+      "step": 12366
+    },
+    {
+      "epoch": 0.2943682757307436,
+      "grad_norm": 1.7177918836479336,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 12367
+    },
+    {
+      "epoch": 0.29439207845377513,
+      "grad_norm": 2.4849290178629584,
+      "learning_rate": 2e-05,
+      "loss": 0.2448,
+      "step": 12368
+    },
+    {
+      "epoch": 0.2944158811768066,
+      "grad_norm": 1.5311332744593977,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 12369
+    },
+    {
+      "epoch": 0.29443968389983816,
+      "grad_norm": 2.4575545220069848,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 12370
+    },
+    {
+      "epoch": 0.29446348662286964,
+      "grad_norm": 4.315975216091813,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 12371
+    },
+    {
+      "epoch": 0.2944872893459012,
+      "grad_norm": 3.812652488579631,
+      "learning_rate": 2e-05,
+      "loss": 0.4785,
+      "step": 12372
+    },
+    {
+      "epoch": 0.29451109206893267,
+      "grad_norm": 1.851653691736054,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 12373
+    },
+    {
+      "epoch": 0.2945348947919642,
+      "grad_norm": 2.542364815750434,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 12374
+    },
+    {
+      "epoch": 0.2945586975149957,
+      "grad_norm": 4.619991449854027,
+      "learning_rate": 2e-05,
+      "loss": 0.7042,
+      "step": 12375
+    },
+    {
+      "epoch": 0.29458250023802723,
+      "grad_norm": 6.414964843401942,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 12376
+    },
+    {
+      "epoch": 0.2946063029610587,
+      "grad_norm": 2.9555782652253804,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 12377
+    },
+    {
+      "epoch": 0.29463010568409026,
+      "grad_norm": 3.9269261583291724,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 12378
+    },
+    {
+      "epoch": 0.2946539084071218,
+      "grad_norm": 5.2105684841410245,
+      "learning_rate": 2e-05,
+      "loss": 0.8437,
+      "step": 12379
+    },
+    {
+      "epoch": 0.2946777111301533,
+      "grad_norm": 2.2553342818572113,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 12380
+    },
+    {
+      "epoch": 0.2947015138531848,
+      "grad_norm": 5.029966403283451,
+      "learning_rate": 2e-05,
+      "loss": 0.6766,
+      "step": 12381
+    },
+    {
+      "epoch": 0.2947253165762163,
+      "grad_norm": 5.973174338367206,
+      "learning_rate": 2e-05,
+      "loss": 1.375,
+      "step": 12382
+    },
+    {
+      "epoch": 0.29474911929924785,
+      "grad_norm": 7.638515115410367,
+      "learning_rate": 2e-05,
+      "loss": 0.321,
+      "step": 12383
+    },
+    {
+      "epoch": 0.29477292202227934,
+      "grad_norm": 0.9951450707459992,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 12384
+    },
+    {
+      "epoch": 0.2947967247453109,
+      "grad_norm": 2.3371251407364544,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 12385
+    },
+    {
+      "epoch": 0.29482052746834236,
+      "grad_norm": 8.378914563788781,
+      "learning_rate": 2e-05,
+      "loss": 1.1189,
+      "step": 12386
+    },
+    {
+      "epoch": 0.2948443301913739,
+      "grad_norm": 4.8380552678248865,
+      "learning_rate": 2e-05,
+      "loss": 0.2355,
+      "step": 12387
+    },
+    {
+      "epoch": 0.2948681329144054,
+      "grad_norm": 7.276467266429306,
+      "learning_rate": 2e-05,
+      "loss": 0.3055,
+      "step": 12388
+    },
+    {
+      "epoch": 0.29489193563743693,
+      "grad_norm": 5.253658976589528,
+      "learning_rate": 2e-05,
+      "loss": 0.5188,
+      "step": 12389
+    },
+    {
+      "epoch": 0.2949157383604684,
+      "grad_norm": 2.769795105174027,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29493954108349996,
+      "grad_norm": 6.1307700145318735,
+      "learning_rate": 2e-05,
+      "loss": 0.1773,
+      "step": 12391
+    },
+    {
+      "epoch": 0.29496334380653144,
+      "grad_norm": 6.795494642890559,
+      "learning_rate": 2e-05,
+      "loss": 1.2693,
+      "step": 12392
+    },
+    {
+      "epoch": 0.294987146529563,
+      "grad_norm": 3.198834374087881,
+      "learning_rate": 2e-05,
+      "loss": 0.3736,
+      "step": 12393
+    },
+    {
+      "epoch": 0.2950109492525945,
+      "grad_norm": 4.7130347284540655,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 12394
+    },
+    {
+      "epoch": 0.295034751975626,
+      "grad_norm": 4.178278629532492,
+      "learning_rate": 2e-05,
+      "loss": 0.5355,
+      "step": 12395
+    },
+    {
+      "epoch": 0.29505855469865755,
+      "grad_norm": 7.092261459491183,
+      "learning_rate": 2e-05,
+      "loss": 1.3477,
+      "step": 12396
+    },
+    {
+      "epoch": 0.29508235742168903,
+      "grad_norm": 1.8764039303361613,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 12397
+    },
+    {
+      "epoch": 0.2951061601447206,
+      "grad_norm": 3.4362721346181693,
+      "learning_rate": 2e-05,
+      "loss": 0.4512,
+      "step": 12398
+    },
+    {
+      "epoch": 0.29512996286775206,
+      "grad_norm": 2.600450023421729,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 12399
+    },
+    {
+      "epoch": 0.2951537655907836,
+      "grad_norm": 2.955672577853586,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 12400
+    },
+    {
+      "epoch": 0.2951775683138151,
+      "grad_norm": 1.738302116409229,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 12401
+    },
+    {
+      "epoch": 0.2952013710368466,
+      "grad_norm": 3.4910109099548703,
+      "learning_rate": 2e-05,
+      "loss": 0.5096,
+      "step": 12402
+    },
+    {
+      "epoch": 0.2952251737598781,
+      "grad_norm": 4.976458136285968,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 12403
+    },
+    {
+      "epoch": 0.29524897648290965,
+      "grad_norm": 5.054536832188636,
+      "learning_rate": 2e-05,
+      "loss": 0.8331,
+      "step": 12404
+    },
+    {
+      "epoch": 0.29527277920594114,
+      "grad_norm": 4.6671134742033775,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 12405
+    },
+    {
+      "epoch": 0.2952965819289727,
+      "grad_norm": 8.675136310983408,
+      "learning_rate": 2e-05,
+      "loss": 0.8793,
+      "step": 12406
+    },
+    {
+      "epoch": 0.29532038465200416,
+      "grad_norm": 1.9908383787976216,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 12407
+    },
+    {
+      "epoch": 0.2953441873750357,
+      "grad_norm": 1.4672154520960963,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 12408
+    },
+    {
+      "epoch": 0.29536799009806725,
+      "grad_norm": 1.0662560552949696,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 12409
+    },
+    {
+      "epoch": 0.29539179282109873,
+      "grad_norm": 6.509797846495738,
+      "learning_rate": 2e-05,
+      "loss": 1.3804,
+      "step": 12410
+    },
+    {
+      "epoch": 0.29541559554413027,
+      "grad_norm": 4.240100073296895,
+      "learning_rate": 2e-05,
+      "loss": 0.9135,
+      "step": 12411
+    },
+    {
+      "epoch": 0.29543939826716176,
+      "grad_norm": 1.6304164360820321,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 12412
+    },
+    {
+      "epoch": 0.2954632009901933,
+      "grad_norm": 2.188521841411001,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 12413
+    },
+    {
+      "epoch": 0.2954870037132248,
+      "grad_norm": 4.247078057033041,
+      "learning_rate": 2e-05,
+      "loss": 0.1354,
+      "step": 12414
+    },
+    {
+      "epoch": 0.2955108064362563,
+      "grad_norm": 4.60019877325794,
+      "learning_rate": 2e-05,
+      "loss": 0.5326,
+      "step": 12415
+    },
+    {
+      "epoch": 0.2955346091592878,
+      "grad_norm": 2.759482175576414,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 12416
+    },
+    {
+      "epoch": 0.29555841188231935,
+      "grad_norm": 5.619256492577355,
+      "learning_rate": 2e-05,
+      "loss": 0.8486,
+      "step": 12417
+    },
+    {
+      "epoch": 0.29558221460535083,
+      "grad_norm": 3.6793495757742143,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 12418
+    },
+    {
+      "epoch": 0.2956060173283824,
+      "grad_norm": 6.765983496482973,
+      "learning_rate": 2e-05,
+      "loss": 1.2413,
+      "step": 12419
+    },
+    {
+      "epoch": 0.29562982005141386,
+      "grad_norm": 5.997778566264666,
+      "learning_rate": 2e-05,
+      "loss": 0.3511,
+      "step": 12420
+    },
+    {
+      "epoch": 0.2956536227744454,
+      "grad_norm": 0.655259940211429,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 12421
+    },
+    {
+      "epoch": 0.2956774254974769,
+      "grad_norm": 13.485041176719369,
+      "learning_rate": 2e-05,
+      "loss": 0.2927,
+      "step": 12422
+    },
+    {
+      "epoch": 0.2957012282205084,
+      "grad_norm": 2.4931148492788346,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 12423
+    },
+    {
+      "epoch": 0.29572503094353997,
+      "grad_norm": 2.278399843956497,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 12424
+    },
+    {
+      "epoch": 0.29574883366657145,
+      "grad_norm": 7.376018617614496,
+      "learning_rate": 2e-05,
+      "loss": 0.7947,
+      "step": 12425
+    },
+    {
+      "epoch": 0.295772636389603,
+      "grad_norm": 18.42141289771657,
+      "learning_rate": 2e-05,
+      "loss": 2.5325,
+      "step": 12426
+    },
+    {
+      "epoch": 0.2957964391126345,
+      "grad_norm": 5.1492152358554515,
+      "learning_rate": 2e-05,
+      "loss": 0.3397,
+      "step": 12427
+    },
+    {
+      "epoch": 0.295820241835666,
+      "grad_norm": 4.07102549768182,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 12428
+    },
+    {
+      "epoch": 0.2958440445586975,
+      "grad_norm": 4.752133601703059,
+      "learning_rate": 2e-05,
+      "loss": 0.769,
+      "step": 12429
+    },
+    {
+      "epoch": 0.29586784728172905,
+      "grad_norm": 5.495140539614379,
+      "learning_rate": 2e-05,
+      "loss": 0.7492,
+      "step": 12430
+    },
+    {
+      "epoch": 0.29589165000476053,
+      "grad_norm": 3.544470612101844,
+      "learning_rate": 2e-05,
+      "loss": 0.5769,
+      "step": 12431
+    },
+    {
+      "epoch": 0.29591545272779207,
+      "grad_norm": 3.2560275880272407,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 12432
+    },
+    {
+      "epoch": 0.29593925545082356,
+      "grad_norm": 2.6986111864020974,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 12433
+    },
+    {
+      "epoch": 0.2959630581738551,
+      "grad_norm": 4.40882310874969,
+      "learning_rate": 2e-05,
+      "loss": 0.8039,
+      "step": 12434
+    },
+    {
+      "epoch": 0.2959868608968866,
+      "grad_norm": 9.693702547033372,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 12435
+    },
+    {
+      "epoch": 0.2960106636199181,
+      "grad_norm": 3.841943681811293,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 12436
+    },
+    {
+      "epoch": 0.2960344663429496,
+      "grad_norm": 3.9591156432520953,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 12437
+    },
+    {
+      "epoch": 0.29605826906598115,
+      "grad_norm": 4.110131709176861,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 12438
+    },
+    {
+      "epoch": 0.2960820717890127,
+      "grad_norm": 8.70165856031354,
+      "learning_rate": 2e-05,
+      "loss": 0.4051,
+      "step": 12439
+    },
+    {
+      "epoch": 0.2961058745120442,
+      "grad_norm": 1.7534168389213503,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 12440
+    },
+    {
+      "epoch": 0.2961296772350757,
+      "grad_norm": 18.3345714738255,
+      "learning_rate": 2e-05,
+      "loss": 2.1497,
+      "step": 12441
+    },
+    {
+      "epoch": 0.2961534799581072,
+      "grad_norm": 6.279176779749814,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 12442
+    },
+    {
+      "epoch": 0.29617728268113874,
+      "grad_norm": 5.493430484981501,
+      "learning_rate": 2e-05,
+      "loss": 0.2688,
+      "step": 12443
+    },
+    {
+      "epoch": 0.29620108540417023,
+      "grad_norm": 2.996760288729488,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 12444
+    },
+    {
+      "epoch": 0.29622488812720177,
+      "grad_norm": 1.522089375904679,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 12445
+    },
+    {
+      "epoch": 0.29624869085023325,
+      "grad_norm": 3.7578423286638416,
+      "learning_rate": 2e-05,
+      "loss": 0.5168,
+      "step": 12446
+    },
+    {
+      "epoch": 0.2962724935732648,
+      "grad_norm": 10.967010885873943,
+      "learning_rate": 2e-05,
+      "loss": 0.6139,
+      "step": 12447
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 7.9797545318305305,
+      "learning_rate": 2e-05,
+      "loss": 0.4559,
+      "step": 12448
+    },
+    {
+      "epoch": 0.2963200990193278,
+      "grad_norm": 9.124894713510907,
+      "learning_rate": 2e-05,
+      "loss": 0.8341,
+      "step": 12449
+    },
+    {
+      "epoch": 0.2963439017423593,
+      "grad_norm": 1.6076935505471943,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 12450
+    },
+    {
+      "epoch": 0.29636770446539085,
+      "grad_norm": 7.181078918374759,
+      "learning_rate": 2e-05,
+      "loss": 1.1473,
+      "step": 12451
+    },
+    {
+      "epoch": 0.29639150718842233,
+      "grad_norm": 4.7740340390908385,
+      "learning_rate": 2e-05,
+      "loss": 0.6224,
+      "step": 12452
+    },
+    {
+      "epoch": 0.2964153099114539,
+      "grad_norm": 0.6947927324396902,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 12453
+    },
+    {
+      "epoch": 0.29643911263448536,
+      "grad_norm": 10.857260120100335,
+      "learning_rate": 2e-05,
+      "loss": 0.5247,
+      "step": 12454
+    },
+    {
+      "epoch": 0.2964629153575169,
+      "grad_norm": 2.5505768665759714,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 12455
+    },
+    {
+      "epoch": 0.29648671808054844,
+      "grad_norm": 4.505506074158059,
+      "learning_rate": 2e-05,
+      "loss": 0.4737,
+      "step": 12456
+    },
+    {
+      "epoch": 0.2965105208035799,
+      "grad_norm": 7.494901781301308,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 12457
+    },
+    {
+      "epoch": 0.29653432352661147,
+      "grad_norm": 4.097017412836666,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 12458
+    },
+    {
+      "epoch": 0.29655812624964295,
+      "grad_norm": 6.7150277753575,
+      "learning_rate": 2e-05,
+      "loss": 0.4768,
+      "step": 12459
+    },
+    {
+      "epoch": 0.2965819289726745,
+      "grad_norm": 4.785838375155575,
+      "learning_rate": 2e-05,
+      "loss": 0.6241,
+      "step": 12460
+    },
+    {
+      "epoch": 0.296605731695706,
+      "grad_norm": 3.0021140276319835,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 12461
+    },
+    {
+      "epoch": 0.2966295344187375,
+      "grad_norm": 9.797170600655306,
+      "learning_rate": 2e-05,
+      "loss": 1.9969,
+      "step": 12462
+    },
+    {
+      "epoch": 0.296653337141769,
+      "grad_norm": 0.7918555671465436,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 12463
+    },
+    {
+      "epoch": 0.29667713986480054,
+      "grad_norm": 1.8852395689972046,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 12464
+    },
+    {
+      "epoch": 0.29670094258783203,
+      "grad_norm": 2.3431164991472206,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 12465
+    },
+    {
+      "epoch": 0.29672474531086357,
+      "grad_norm": 1.6266332774256083,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 12466
+    },
+    {
+      "epoch": 0.29674854803389505,
+      "grad_norm": 3.5210147534506335,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 12467
+    },
+    {
+      "epoch": 0.2967723507569266,
+      "grad_norm": 4.108797449541471,
+      "learning_rate": 2e-05,
+      "loss": 0.8365,
+      "step": 12468
+    },
+    {
+      "epoch": 0.2967961534799581,
+      "grad_norm": 4.739929694011213,
+      "learning_rate": 2e-05,
+      "loss": 0.6502,
+      "step": 12469
+    },
+    {
+      "epoch": 0.2968199562029896,
+      "grad_norm": 6.404163376976626,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 12470
+    },
+    {
+      "epoch": 0.29684375892602116,
+      "grad_norm": 0.1360032101767026,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 12471
+    },
+    {
+      "epoch": 0.29686756164905265,
+      "grad_norm": 0.6121080773417051,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12472
+    },
+    {
+      "epoch": 0.2968913643720842,
+      "grad_norm": 4.511485403255308,
+      "learning_rate": 2e-05,
+      "loss": 0.2695,
+      "step": 12473
+    },
+    {
+      "epoch": 0.2969151670951157,
+      "grad_norm": 7.533455907805434,
+      "learning_rate": 2e-05,
+      "loss": 0.3038,
+      "step": 12474
+    },
+    {
+      "epoch": 0.2969389698181472,
+      "grad_norm": 11.051093012715276,
+      "learning_rate": 2e-05,
+      "loss": 0.8565,
+      "step": 12475
+    },
+    {
+      "epoch": 0.2969627725411787,
+      "grad_norm": 0.10499830371617817,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 12476
+    },
+    {
+      "epoch": 0.29698657526421024,
+      "grad_norm": 3.8985685359657043,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 12477
+    },
+    {
+      "epoch": 0.2970103779872417,
+      "grad_norm": 2.512921279990077,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 12478
+    },
+    {
+      "epoch": 0.29703418071027327,
+      "grad_norm": 2.9892636498758924,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 12479
+    },
+    {
+      "epoch": 0.29705798343330475,
+      "grad_norm": 5.8528810011498855,
+      "learning_rate": 2e-05,
+      "loss": 1.1,
+      "step": 12480
+    },
+    {
+      "epoch": 0.2970817861563363,
+      "grad_norm": 6.202221421770675,
+      "learning_rate": 2e-05,
+      "loss": 0.542,
+      "step": 12481
+    },
+    {
+      "epoch": 0.2971055888793678,
+      "grad_norm": 2.4993110209949956,
+      "learning_rate": 2e-05,
+      "loss": 0.1336,
+      "step": 12482
+    },
+    {
+      "epoch": 0.2971293916023993,
+      "grad_norm": 13.59579283861308,
+      "learning_rate": 2e-05,
+      "loss": 1.3534,
+      "step": 12483
+    },
+    {
+      "epoch": 0.2971531943254308,
+      "grad_norm": 0.36414007039489443,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 12484
+    },
+    {
+      "epoch": 0.29717699704846234,
+      "grad_norm": 2.6128142774577827,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 12485
+    },
+    {
+      "epoch": 0.2972007997714939,
+      "grad_norm": 1.935694351103537,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 12486
+    },
+    {
+      "epoch": 0.29722460249452537,
+      "grad_norm": 23.960561607996507,
+      "learning_rate": 2e-05,
+      "loss": 0.852,
+      "step": 12487
+    },
+    {
+      "epoch": 0.2972484052175569,
+      "grad_norm": 12.925068827562471,
+      "learning_rate": 2e-05,
+      "loss": 0.6059,
+      "step": 12488
+    },
+    {
+      "epoch": 0.2972722079405884,
+      "grad_norm": 2.0543751610049568,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 12489
+    },
+    {
+      "epoch": 0.29729601066361994,
+      "grad_norm": 4.201222228325933,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 12490
+    },
+    {
+      "epoch": 0.2973198133866514,
+      "grad_norm": 1.580248023794046,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 12491
+    },
+    {
+      "epoch": 0.29734361610968296,
+      "grad_norm": 2.3516990701287934,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 12492
+    },
+    {
+      "epoch": 0.29736741883271445,
+      "grad_norm": 4.299700366807354,
+      "learning_rate": 2e-05,
+      "loss": 0.6003,
+      "step": 12493
+    },
+    {
+      "epoch": 0.297391221555746,
+      "grad_norm": 1.6723089451166253,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 12494
+    },
+    {
+      "epoch": 0.2974150242787775,
+      "grad_norm": 8.090647046306486,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 12495
+    },
+    {
+      "epoch": 0.297438827001809,
+      "grad_norm": 2.2255805562744064,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 12496
+    },
+    {
+      "epoch": 0.2974626297248405,
+      "grad_norm": 3.8678520816776203,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 12497
+    },
+    {
+      "epoch": 0.29748643244787204,
+      "grad_norm": 2.2126579237240063,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 12498
+    },
+    {
+      "epoch": 0.2975102351709035,
+      "grad_norm": 7.491643369438367,
+      "learning_rate": 2e-05,
+      "loss": 0.839,
+      "step": 12499
+    },
+    {
+      "epoch": 0.29753403789393507,
+      "grad_norm": 0.43025328355370707,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 12500
+    },
+    {
+      "epoch": 0.2975578406169666,
+      "grad_norm": 11.304313456110197,
+      "learning_rate": 2e-05,
+      "loss": 0.4599,
+      "step": 12501
+    },
+    {
+      "epoch": 0.2975816433399981,
+      "grad_norm": 4.627319065701306,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 12502
+    },
+    {
+      "epoch": 0.29760544606302963,
+      "grad_norm": 4.994894712229723,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 12503
+    },
+    {
+      "epoch": 0.2976292487860611,
+      "grad_norm": 15.257897364762446,
+      "learning_rate": 2e-05,
+      "loss": 0.4089,
+      "step": 12504
+    },
+    {
+      "epoch": 0.29765305150909266,
+      "grad_norm": 1.6782107903458658,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 12505
+    },
+    {
+      "epoch": 0.29767685423212414,
+      "grad_norm": 5.75315176827131,
+      "learning_rate": 2e-05,
+      "loss": 0.9773,
+      "step": 12506
+    },
+    {
+      "epoch": 0.2977006569551557,
+      "grad_norm": 2.843644584521933,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 12507
+    },
+    {
+      "epoch": 0.29772445967818717,
+      "grad_norm": 2.841787186831314,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 12508
+    },
+    {
+      "epoch": 0.2977482624012187,
+      "grad_norm": 3.6910407623055264,
+      "learning_rate": 2e-05,
+      "loss": 0.559,
+      "step": 12509
+    },
+    {
+      "epoch": 0.2977720651242502,
+      "grad_norm": 6.930011746111125,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 12510
+    },
+    {
+      "epoch": 0.29779586784728174,
+      "grad_norm": 6.012559893383622,
+      "learning_rate": 2e-05,
+      "loss": 0.5288,
+      "step": 12511
+    },
+    {
+      "epoch": 0.2978196705703132,
+      "grad_norm": 3.4693286146796436,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 12512
+    },
+    {
+      "epoch": 0.29784347329334476,
+      "grad_norm": 2.979577363854608,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 12513
+    },
+    {
+      "epoch": 0.29786727601637625,
+      "grad_norm": 4.058607190969775,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 12514
+    },
+    {
+      "epoch": 0.2978910787394078,
+      "grad_norm": 5.590364480316672,
+      "learning_rate": 2e-05,
+      "loss": 0.2622,
+      "step": 12515
+    },
+    {
+      "epoch": 0.29791488146243933,
+      "grad_norm": 3.6981005702032537,
+      "learning_rate": 2e-05,
+      "loss": 0.6454,
+      "step": 12516
+    },
+    {
+      "epoch": 0.2979386841854708,
+      "grad_norm": 2.9199284850858778,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 12517
+    },
+    {
+      "epoch": 0.29796248690850236,
+      "grad_norm": 2.5357930978422236,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 12518
+    },
+    {
+      "epoch": 0.29798628963153384,
+      "grad_norm": 11.773941455936576,
+      "learning_rate": 2e-05,
+      "loss": 1.6006,
+      "step": 12519
+    },
+    {
+      "epoch": 0.2980100923545654,
+      "grad_norm": 1.1209230289463439,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 12520
+    },
+    {
+      "epoch": 0.29803389507759687,
+      "grad_norm": 2.46459710401986,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 12521
+    },
+    {
+      "epoch": 0.2980576978006284,
+      "grad_norm": 4.332065815412807,
+      "learning_rate": 2e-05,
+      "loss": 0.3168,
+      "step": 12522
+    },
+    {
+      "epoch": 0.2980815005236599,
+      "grad_norm": 7.706161585854942,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 12523
+    },
+    {
+      "epoch": 0.29810530324669143,
+      "grad_norm": 4.385658477895872,
+      "learning_rate": 2e-05,
+      "loss": 0.4743,
+      "step": 12524
+    },
+    {
+      "epoch": 0.2981291059697229,
+      "grad_norm": 6.25979644236386,
+      "learning_rate": 2e-05,
+      "loss": 1.2352,
+      "step": 12525
+    },
+    {
+      "epoch": 0.29815290869275446,
+      "grad_norm": 2.402645685689051,
+      "learning_rate": 2e-05,
+      "loss": 0.2596,
+      "step": 12526
+    },
+    {
+      "epoch": 0.29817671141578594,
+      "grad_norm": 1.2697135027209767,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 12527
+    },
+    {
+      "epoch": 0.2982005141388175,
+      "grad_norm": 4.5421124747949095,
+      "learning_rate": 2e-05,
+      "loss": 0.6707,
+      "step": 12528
+    },
+    {
+      "epoch": 0.29822431686184897,
+      "grad_norm": 3.334166126913072,
+      "learning_rate": 2e-05,
+      "loss": 0.5221,
+      "step": 12529
+    },
+    {
+      "epoch": 0.2982481195848805,
+      "grad_norm": 2.141955237881112,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 12530
+    },
+    {
+      "epoch": 0.29827192230791205,
+      "grad_norm": 15.589187371748563,
+      "learning_rate": 2e-05,
+      "loss": 0.7748,
+      "step": 12531
+    },
+    {
+      "epoch": 0.29829572503094354,
+      "grad_norm": 0.7199388199690804,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 12532
+    },
+    {
+      "epoch": 0.2983195277539751,
+      "grad_norm": 1.811176913336088,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 12533
+    },
+    {
+      "epoch": 0.29834333047700656,
+      "grad_norm": 9.678453545423162,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 12534
+    },
+    {
+      "epoch": 0.2983671332000381,
+      "grad_norm": 2.1270567517322103,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 12535
+    },
+    {
+      "epoch": 0.2983909359230696,
+      "grad_norm": 1.6453861499635876,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 12536
+    },
+    {
+      "epoch": 0.29841473864610113,
+      "grad_norm": 14.68271972580606,
+      "learning_rate": 2e-05,
+      "loss": 1.3177,
+      "step": 12537
+    },
+    {
+      "epoch": 0.2984385413691326,
+      "grad_norm": 5.126782838157788,
+      "learning_rate": 2e-05,
+      "loss": 0.552,
+      "step": 12538
+    },
+    {
+      "epoch": 0.29846234409216416,
+      "grad_norm": 2.1034130158614763,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 12539
+    },
+    {
+      "epoch": 0.29848614681519564,
+      "grad_norm": 2.8186378082367574,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 12540
+    },
+    {
+      "epoch": 0.2985099495382272,
+      "grad_norm": 1.2177053903400208,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 12541
+    },
+    {
+      "epoch": 0.29853375226125867,
+      "grad_norm": 8.226802204858485,
+      "learning_rate": 2e-05,
+      "loss": 1.4449,
+      "step": 12542
+    },
+    {
+      "epoch": 0.2985575549842902,
+      "grad_norm": 5.866365056221292,
+      "learning_rate": 2e-05,
+      "loss": 1.2429,
+      "step": 12543
+    },
+    {
+      "epoch": 0.2985813577073217,
+      "grad_norm": 13.104449473511373,
+      "learning_rate": 2e-05,
+      "loss": 0.9954,
+      "step": 12544
+    },
+    {
+      "epoch": 0.29860516043035323,
+      "grad_norm": 5.997320896472439,
+      "learning_rate": 2e-05,
+      "loss": 0.7314,
+      "step": 12545
+    },
+    {
+      "epoch": 0.2986289631533848,
+      "grad_norm": 2.601287555611924,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 12546
+    },
+    {
+      "epoch": 0.29865276587641626,
+      "grad_norm": 4.707132080426773,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 12547
+    },
+    {
+      "epoch": 0.2986765685994478,
+      "grad_norm": 14.32217812846991,
+      "learning_rate": 2e-05,
+      "loss": 1.6993,
+      "step": 12548
+    },
+    {
+      "epoch": 0.2987003713224793,
+      "grad_norm": 0.12436953918970553,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 12549
+    },
+    {
+      "epoch": 0.2987241740455108,
+      "grad_norm": 2.879271852351288,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 12550
+    },
+    {
+      "epoch": 0.2987479767685423,
+      "grad_norm": 3.6762845966606275,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 12551
+    },
+    {
+      "epoch": 0.29877177949157385,
+      "grad_norm": 15.676586673513876,
+      "learning_rate": 2e-05,
+      "loss": 0.6123,
+      "step": 12552
+    },
+    {
+      "epoch": 0.29879558221460534,
+      "grad_norm": 2.6501712383070473,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 12553
+    },
+    {
+      "epoch": 0.2988193849376369,
+      "grad_norm": 4.582096488707918,
+      "learning_rate": 2e-05,
+      "loss": 0.3454,
+      "step": 12554
+    },
+    {
+      "epoch": 0.29884318766066836,
+      "grad_norm": 4.5021839627601095,
+      "learning_rate": 2e-05,
+      "loss": 0.432,
+      "step": 12555
+    },
+    {
+      "epoch": 0.2988669903836999,
+      "grad_norm": 3.7287955358975555,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 12556
+    },
+    {
+      "epoch": 0.2988907931067314,
+      "grad_norm": 3.360156561023615,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 12557
+    },
+    {
+      "epoch": 0.29891459582976293,
+      "grad_norm": 2.064880592300771,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 12558
+    },
+    {
+      "epoch": 0.2989383985527944,
+      "grad_norm": 2.439109659015719,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 12559
+    },
+    {
+      "epoch": 0.29896220127582596,
+      "grad_norm": 3.841775370770136,
+      "learning_rate": 2e-05,
+      "loss": 0.2653,
+      "step": 12560
+    },
+    {
+      "epoch": 0.2989860039988575,
+      "grad_norm": 1.8694931142391646,
+      "learning_rate": 2e-05,
+      "loss": 0.114,
+      "step": 12561
+    },
+    {
+      "epoch": 0.299009806721889,
+      "grad_norm": 1.2268482511638046,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 12562
+    },
+    {
+      "epoch": 0.2990336094449205,
+      "grad_norm": 10.187987971100888,
+      "learning_rate": 2e-05,
+      "loss": 0.4929,
+      "step": 12563
+    },
+    {
+      "epoch": 0.299057412167952,
+      "grad_norm": 4.496196685271561,
+      "learning_rate": 2e-05,
+      "loss": 0.5781,
+      "step": 12564
+    },
+    {
+      "epoch": 0.29908121489098355,
+      "grad_norm": 3.3188194674594835,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 12565
+    },
+    {
+      "epoch": 0.29910501761401503,
+      "grad_norm": 15.292455027263692,
+      "learning_rate": 2e-05,
+      "loss": 0.8602,
+      "step": 12566
+    },
+    {
+      "epoch": 0.2991288203370466,
+      "grad_norm": 8.349208312732062,
+      "learning_rate": 2e-05,
+      "loss": 0.692,
+      "step": 12567
+    },
+    {
+      "epoch": 0.29915262306007806,
+      "grad_norm": 2.0511737057088744,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 12568
+    },
+    {
+      "epoch": 0.2991764257831096,
+      "grad_norm": 2.1141907394724497,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 12569
+    },
+    {
+      "epoch": 0.2992002285061411,
+      "grad_norm": 5.9517405425934315,
+      "learning_rate": 2e-05,
+      "loss": 0.8918,
+      "step": 12570
+    },
+    {
+      "epoch": 0.2992240312291726,
+      "grad_norm": 3.5435001825851713,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 12571
+    },
+    {
+      "epoch": 0.2992478339522041,
+      "grad_norm": 2.0780679385746996,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 12572
+    },
+    {
+      "epoch": 0.29927163667523565,
+      "grad_norm": 4.07958822768751,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 12573
+    },
+    {
+      "epoch": 0.29929543939826714,
+      "grad_norm": 7.642203521910614,
+      "learning_rate": 2e-05,
+      "loss": 1.3713,
+      "step": 12574
+    },
+    {
+      "epoch": 0.2993192421212987,
+      "grad_norm": 1.386939371687012,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 12575
+    },
+    {
+      "epoch": 0.29934304484433016,
+      "grad_norm": 1.0071663156405608,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 12576
+    },
+    {
+      "epoch": 0.2993668475673617,
+      "grad_norm": 1.1990418637993856,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 12577
+    },
+    {
+      "epoch": 0.29939065029039325,
+      "grad_norm": 2.4192708437629062,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 12578
+    },
+    {
+      "epoch": 0.29941445301342473,
+      "grad_norm": 4.328532545918061,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 12579
+    },
+    {
+      "epoch": 0.29943825573645627,
+      "grad_norm": 8.863707890547731,
+      "learning_rate": 2e-05,
+      "loss": 1.287,
+      "step": 12580
+    },
+    {
+      "epoch": 0.29946205845948776,
+      "grad_norm": 0.5154752928898286,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 12581
+    },
+    {
+      "epoch": 0.2994858611825193,
+      "grad_norm": 3.0291509247087136,
+      "learning_rate": 2e-05,
+      "loss": 0.1465,
+      "step": 12582
+    },
+    {
+      "epoch": 0.2995096639055508,
+      "grad_norm": 3.9251784116456294,
+      "learning_rate": 2e-05,
+      "loss": 0.1973,
+      "step": 12583
+    },
+    {
+      "epoch": 0.2995334666285823,
+      "grad_norm": 10.806071505024152,
+      "learning_rate": 2e-05,
+      "loss": 1.0232,
+      "step": 12584
+    },
+    {
+      "epoch": 0.2995572693516138,
+      "grad_norm": 6.566271973775776,
+      "learning_rate": 2e-05,
+      "loss": 0.2572,
+      "step": 12585
+    },
+    {
+      "epoch": 0.29958107207464535,
+      "grad_norm": 8.70451290527972,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 12586
+    },
+    {
+      "epoch": 0.29960487479767683,
+      "grad_norm": 1.527158662195362,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 12587
+    },
+    {
+      "epoch": 0.2996286775207084,
+      "grad_norm": 1.9356868553313775,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 12588
+    },
+    {
+      "epoch": 0.29965248024373986,
+      "grad_norm": 10.631309816623062,
+      "learning_rate": 2e-05,
+      "loss": 0.3718,
+      "step": 12589
+    },
+    {
+      "epoch": 0.2996762829667714,
+      "grad_norm": 1.3702346614788787,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 12590
+    },
+    {
+      "epoch": 0.2997000856898029,
+      "grad_norm": 4.069600973736745,
+      "learning_rate": 2e-05,
+      "loss": 0.5937,
+      "step": 12591
+    },
+    {
+      "epoch": 0.2997238884128344,
+      "grad_norm": 1.6568596536617342,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 12592
+    },
+    {
+      "epoch": 0.29974769113586597,
+      "grad_norm": 4.336916998061697,
+      "learning_rate": 2e-05,
+      "loss": 0.5384,
+      "step": 12593
+    },
+    {
+      "epoch": 0.29977149385889745,
+      "grad_norm": 2.38458449973066,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 12594
+    },
+    {
+      "epoch": 0.299795296581929,
+      "grad_norm": 17.37183882047088,
+      "learning_rate": 2e-05,
+      "loss": 0.2319,
+      "step": 12595
+    },
+    {
+      "epoch": 0.2998190993049605,
+      "grad_norm": 3.1410312442993993,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 12596
+    },
+    {
+      "epoch": 0.299842902027992,
+      "grad_norm": 4.577033214585283,
+      "learning_rate": 2e-05,
+      "loss": 0.2329,
+      "step": 12597
+    },
+    {
+      "epoch": 0.2998667047510235,
+      "grad_norm": 0.625571113213675,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 12598
+    },
+    {
+      "epoch": 0.29989050747405505,
+      "grad_norm": 12.294101195773576,
+      "learning_rate": 2e-05,
+      "loss": 0.4952,
+      "step": 12599
+    },
+    {
+      "epoch": 0.29991431019708653,
+      "grad_norm": 7.574307159227927,
+      "learning_rate": 2e-05,
+      "loss": 1.9394,
+      "step": 12600
+    },
+    {
+      "epoch": 0.29993811292011807,
+      "grad_norm": 6.3905546893830385,
+      "learning_rate": 2e-05,
+      "loss": 0.8397,
+      "step": 12601
+    },
+    {
+      "epoch": 0.29996191564314956,
+      "grad_norm": 3.264314242124045,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 12602
+    },
+    {
+      "epoch": 0.2999857183661811,
+      "grad_norm": 3.45985025223018,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 12603
+    },
+    {
+      "epoch": 0.3000095210892126,
+      "grad_norm": 8.571062282559357,
+      "learning_rate": 2e-05,
+      "loss": 1.1911,
+      "step": 12604
+    },
+    {
+      "epoch": 0.3000333238122441,
+      "grad_norm": 1.1449301343119196,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 12605
+    },
+    {
+      "epoch": 0.3000571265352756,
+      "grad_norm": 3.0207974398070916,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 12606
+    },
+    {
+      "epoch": 0.30008092925830715,
+      "grad_norm": 4.307335196751416,
+      "learning_rate": 2e-05,
+      "loss": 0.5926,
+      "step": 12607
+    },
+    {
+      "epoch": 0.3001047319813387,
+      "grad_norm": 10.385099179557994,
+      "learning_rate": 2e-05,
+      "loss": 0.6972,
+      "step": 12608
+    },
+    {
+      "epoch": 0.3001285347043702,
+      "grad_norm": 0.8675491647953953,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 12609
+    },
+    {
+      "epoch": 0.3001523374274017,
+      "grad_norm": 16.219375978304928,
+      "learning_rate": 2e-05,
+      "loss": 0.6738,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3001761401504332,
+      "grad_norm": 5.333595605640602,
+      "learning_rate": 2e-05,
+      "loss": 0.9059,
+      "step": 12611
+    },
+    {
+      "epoch": 0.30019994287346474,
+      "grad_norm": 1.4946679949030266,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 12612
+    },
+    {
+      "epoch": 0.3002237455964962,
+      "grad_norm": 4.037857516919237,
+      "learning_rate": 2e-05,
+      "loss": 0.4696,
+      "step": 12613
+    },
+    {
+      "epoch": 0.30024754831952777,
+      "grad_norm": 4.796925983247651,
+      "learning_rate": 2e-05,
+      "loss": 0.7204,
+      "step": 12614
+    },
+    {
+      "epoch": 0.30027135104255925,
+      "grad_norm": 2.525128950529036,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 12615
+    },
+    {
+      "epoch": 0.3002951537655908,
+      "grad_norm": 2.2046911971159586,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 12616
+    },
+    {
+      "epoch": 0.3003189564886223,
+      "grad_norm": 2.37105896031622,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 12617
+    },
+    {
+      "epoch": 0.3003427592116538,
+      "grad_norm": 5.010599101997515,
+      "learning_rate": 2e-05,
+      "loss": 0.9396,
+      "step": 12618
+    },
+    {
+      "epoch": 0.3003665619346853,
+      "grad_norm": 4.5071179641464765,
+      "learning_rate": 2e-05,
+      "loss": 0.2899,
+      "step": 12619
+    },
+    {
+      "epoch": 0.30039036465771685,
+      "grad_norm": 5.916921469125708,
+      "learning_rate": 2e-05,
+      "loss": 0.8841,
+      "step": 12620
+    },
+    {
+      "epoch": 0.30041416738074833,
+      "grad_norm": 2.770837103195445,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 12621
+    },
+    {
+      "epoch": 0.30043797010377987,
+      "grad_norm": 22.848245301930493,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 12622
+    },
+    {
+      "epoch": 0.3004617728268114,
+      "grad_norm": 4.2663724599827955,
+      "learning_rate": 2e-05,
+      "loss": 0.563,
+      "step": 12623
+    },
+    {
+      "epoch": 0.3004855755498429,
+      "grad_norm": 2.2869124455853997,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 12624
+    },
+    {
+      "epoch": 0.30050937827287444,
+      "grad_norm": 2.044940465542758,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 12625
+    },
+    {
+      "epoch": 0.3005331809959059,
+      "grad_norm": 3.3259153212256867,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 12626
+    },
+    {
+      "epoch": 0.30055698371893746,
+      "grad_norm": 5.963063217340499,
+      "learning_rate": 2e-05,
+      "loss": 0.1334,
+      "step": 12627
+    },
+    {
+      "epoch": 0.30058078644196895,
+      "grad_norm": 3.288382768422792,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 12628
+    },
+    {
+      "epoch": 0.3006045891650005,
+      "grad_norm": 8.039311959876501,
+      "learning_rate": 2e-05,
+      "loss": 1.0157,
+      "step": 12629
+    },
+    {
+      "epoch": 0.300628391888032,
+      "grad_norm": 2.237955671495256,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 12630
+    },
+    {
+      "epoch": 0.3006521946110635,
+      "grad_norm": 11.62499080445974,
+      "learning_rate": 2e-05,
+      "loss": 0.3538,
+      "step": 12631
+    },
+    {
+      "epoch": 0.300675997334095,
+      "grad_norm": 5.346869257804766,
+      "learning_rate": 2e-05,
+      "loss": 0.6992,
+      "step": 12632
+    },
+    {
+      "epoch": 0.30069980005712654,
+      "grad_norm": 1.674127909685947,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 12633
+    },
+    {
+      "epoch": 0.30072360278015803,
+      "grad_norm": 3.4225617909597066,
+      "learning_rate": 2e-05,
+      "loss": 0.1289,
+      "step": 12634
+    },
+    {
+      "epoch": 0.30074740550318957,
+      "grad_norm": 0.6054257333615728,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 12635
+    },
+    {
+      "epoch": 0.30077120822622105,
+      "grad_norm": 2.6232314832172112,
+      "learning_rate": 2e-05,
+      "loss": 0.1516,
+      "step": 12636
+    },
+    {
+      "epoch": 0.3007950109492526,
+      "grad_norm": 4.051009731889746,
+      "learning_rate": 2e-05,
+      "loss": 0.4445,
+      "step": 12637
+    },
+    {
+      "epoch": 0.30081881367228414,
+      "grad_norm": 20.85875380290747,
+      "learning_rate": 2e-05,
+      "loss": 0.5535,
+      "step": 12638
+    },
+    {
+      "epoch": 0.3008426163953156,
+      "grad_norm": 0.41864418982804746,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 12639
+    },
+    {
+      "epoch": 0.30086641911834716,
+      "grad_norm": 1.3497368548737008,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 12640
+    },
+    {
+      "epoch": 0.30089022184137865,
+      "grad_norm": 3.2804661852686596,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 12641
+    },
+    {
+      "epoch": 0.3009140245644102,
+      "grad_norm": 5.503913663529038,
+      "learning_rate": 2e-05,
+      "loss": 1.0601,
+      "step": 12642
+    },
+    {
+      "epoch": 0.3009378272874417,
+      "grad_norm": 2.6001227027451117,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 12643
+    },
+    {
+      "epoch": 0.3009616300104732,
+      "grad_norm": 7.85158929608344,
+      "learning_rate": 2e-05,
+      "loss": 1.3764,
+      "step": 12644
+    },
+    {
+      "epoch": 0.3009854327335047,
+      "grad_norm": 2.8042996275910212,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 12645
+    },
+    {
+      "epoch": 0.30100923545653624,
+      "grad_norm": 3.496165371303306,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 12646
+    },
+    {
+      "epoch": 0.3010330381795677,
+      "grad_norm": 2.7171809622341456,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 12647
+    },
+    {
+      "epoch": 0.30105684090259927,
+      "grad_norm": 4.079460905473185,
+      "learning_rate": 2e-05,
+      "loss": 0.4505,
+      "step": 12648
+    },
+    {
+      "epoch": 0.30108064362563075,
+      "grad_norm": 11.403084760330932,
+      "learning_rate": 2e-05,
+      "loss": 0.377,
+      "step": 12649
+    },
+    {
+      "epoch": 0.3011044463486623,
+      "grad_norm": 7.967643482269289,
+      "learning_rate": 2e-05,
+      "loss": 0.2161,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3011282490716938,
+      "grad_norm": 1.6349022380183642,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 12651
+    },
+    {
+      "epoch": 0.3011520517947253,
+      "grad_norm": 3.136966510709885,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 12652
+    },
+    {
+      "epoch": 0.30117585451775686,
+      "grad_norm": 12.500037621724928,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 12653
+    },
+    {
+      "epoch": 0.30119965724078834,
+      "grad_norm": 9.10646526388942,
+      "learning_rate": 2e-05,
+      "loss": 0.3658,
+      "step": 12654
+    },
+    {
+      "epoch": 0.3012234599638199,
+      "grad_norm": 3.755210676738833,
+      "learning_rate": 2e-05,
+      "loss": 0.5307,
+      "step": 12655
+    },
+    {
+      "epoch": 0.30124726268685137,
+      "grad_norm": 0.8026985222574443,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 12656
+    },
+    {
+      "epoch": 0.3012710654098829,
+      "grad_norm": 3.7741040495367035,
+      "learning_rate": 2e-05,
+      "loss": 0.177,
+      "step": 12657
+    },
+    {
+      "epoch": 0.3012948681329144,
+      "grad_norm": 5.780151811013532,
+      "learning_rate": 2e-05,
+      "loss": 0.58,
+      "step": 12658
+    },
+    {
+      "epoch": 0.30131867085594594,
+      "grad_norm": 2.06115846132786,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 12659
+    },
+    {
+      "epoch": 0.3013424735789774,
+      "grad_norm": 4.476969864407092,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 12660
+    },
+    {
+      "epoch": 0.30136627630200896,
+      "grad_norm": 4.322418996035777,
+      "learning_rate": 2e-05,
+      "loss": 0.7395,
+      "step": 12661
+    },
+    {
+      "epoch": 0.30139007902504045,
+      "grad_norm": 4.796637185414099,
+      "learning_rate": 2e-05,
+      "loss": 0.6216,
+      "step": 12662
+    },
+    {
+      "epoch": 0.301413881748072,
+      "grad_norm": 13.630927307158803,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 12663
+    },
+    {
+      "epoch": 0.3014376844711035,
+      "grad_norm": 5.416153829211385,
+      "learning_rate": 2e-05,
+      "loss": 0.7255,
+      "step": 12664
+    },
+    {
+      "epoch": 0.301461487194135,
+      "grad_norm": 16.6175219670628,
+      "learning_rate": 2e-05,
+      "loss": 0.5926,
+      "step": 12665
+    },
+    {
+      "epoch": 0.3014852899171665,
+      "grad_norm": 4.8231709628468975,
+      "learning_rate": 2e-05,
+      "loss": 0.6171,
+      "step": 12666
+    },
+    {
+      "epoch": 0.30150909264019804,
+      "grad_norm": 7.162939167365752,
+      "learning_rate": 2e-05,
+      "loss": 0.7729,
+      "step": 12667
+    },
+    {
+      "epoch": 0.3015328953632296,
+      "grad_norm": 3.0141689971366117,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 12668
+    },
+    {
+      "epoch": 0.30155669808626107,
+      "grad_norm": 1.8966422088087596,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 12669
+    },
+    {
+      "epoch": 0.3015805008092926,
+      "grad_norm": 2.3605708713792084,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3016043035323241,
+      "grad_norm": 8.76583380089527,
+      "learning_rate": 2e-05,
+      "loss": 0.2435,
+      "step": 12671
+    },
+    {
+      "epoch": 0.30162810625535563,
+      "grad_norm": 6.616781132021048,
+      "learning_rate": 2e-05,
+      "loss": 1.1342,
+      "step": 12672
+    },
+    {
+      "epoch": 0.3016519089783871,
+      "grad_norm": 0.2002409963622314,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 12673
+    },
+    {
+      "epoch": 0.30167571170141866,
+      "grad_norm": 0.6368734200798438,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 12674
+    },
+    {
+      "epoch": 0.30169951442445014,
+      "grad_norm": 3.9788794791757804,
+      "learning_rate": 2e-05,
+      "loss": 0.3168,
+      "step": 12675
+    },
+    {
+      "epoch": 0.3017233171474817,
+      "grad_norm": 2.7314922459959936,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 12676
+    },
+    {
+      "epoch": 0.30174711987051317,
+      "grad_norm": 3.160196771119684,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 12677
+    },
+    {
+      "epoch": 0.3017709225935447,
+      "grad_norm": 11.709881312260329,
+      "learning_rate": 2e-05,
+      "loss": 0.4036,
+      "step": 12678
+    },
+    {
+      "epoch": 0.3017947253165762,
+      "grad_norm": 14.990614849421648,
+      "learning_rate": 2e-05,
+      "loss": 0.975,
+      "step": 12679
+    },
+    {
+      "epoch": 0.30181852803960774,
+      "grad_norm": 11.677323189449877,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3018423307626392,
+      "grad_norm": 3.9859531143053046,
+      "learning_rate": 2e-05,
+      "loss": 0.2655,
+      "step": 12681
+    },
+    {
+      "epoch": 0.30186613348567076,
+      "grad_norm": 6.110011537674864,
+      "learning_rate": 2e-05,
+      "loss": 0.3592,
+      "step": 12682
+    },
+    {
+      "epoch": 0.3018899362087023,
+      "grad_norm": 1.474780609637891,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 12683
+    },
+    {
+      "epoch": 0.3019137389317338,
+      "grad_norm": 5.432162412459299,
+      "learning_rate": 2e-05,
+      "loss": 0.6922,
+      "step": 12684
+    },
+    {
+      "epoch": 0.30193754165476533,
+      "grad_norm": 5.595783048195695,
+      "learning_rate": 2e-05,
+      "loss": 0.2843,
+      "step": 12685
+    },
+    {
+      "epoch": 0.3019613443777968,
+      "grad_norm": 6.360765536782894,
+      "learning_rate": 2e-05,
+      "loss": 0.6134,
+      "step": 12686
+    },
+    {
+      "epoch": 0.30198514710082835,
+      "grad_norm": 3.3692050522390122,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 12687
+    },
+    {
+      "epoch": 0.30200894982385984,
+      "grad_norm": 3.9527120798990207,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 12688
+    },
+    {
+      "epoch": 0.3020327525468914,
+      "grad_norm": 2.8641753593673758,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 12689
+    },
+    {
+      "epoch": 0.30205655526992287,
+      "grad_norm": 2.1670472978224975,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3020803579929544,
+      "grad_norm": 4.088631595630882,
+      "learning_rate": 2e-05,
+      "loss": 0.6546,
+      "step": 12691
+    },
+    {
+      "epoch": 0.3021041607159859,
+      "grad_norm": 5.949835199056564,
+      "learning_rate": 2e-05,
+      "loss": 0.2426,
+      "step": 12692
+    },
+    {
+      "epoch": 0.30212796343901743,
+      "grad_norm": 3.874949702063251,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 12693
+    },
+    {
+      "epoch": 0.3021517661620489,
+      "grad_norm": 4.268371693858586,
+      "learning_rate": 2e-05,
+      "loss": 0.6185,
+      "step": 12694
+    },
+    {
+      "epoch": 0.30217556888508046,
+      "grad_norm": 0.7865111281301765,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 12695
+    },
+    {
+      "epoch": 0.30219937160811194,
+      "grad_norm": 1.6953740563905102,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 12696
+    },
+    {
+      "epoch": 0.3022231743311435,
+      "grad_norm": 5.1494868704207315,
+      "learning_rate": 2e-05,
+      "loss": 1.1766,
+      "step": 12697
+    },
+    {
+      "epoch": 0.30224697705417497,
+      "grad_norm": 7.72433398049548,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 12698
+    },
+    {
+      "epoch": 0.3022707797772065,
+      "grad_norm": 2.8234109354189107,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 12699
+    },
+    {
+      "epoch": 0.30229458250023805,
+      "grad_norm": 5.25304424922036,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 12700
+    },
+    {
+      "epoch": 0.30231838522326954,
+      "grad_norm": 1.7786586550900902,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 12701
+    },
+    {
+      "epoch": 0.3023421879463011,
+      "grad_norm": 3.8781548829176833,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 12702
+    },
+    {
+      "epoch": 0.30236599066933256,
+      "grad_norm": 0.8165029565410046,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 12703
+    },
+    {
+      "epoch": 0.3023897933923641,
+      "grad_norm": 5.395626583773606,
+      "learning_rate": 2e-05,
+      "loss": 0.3134,
+      "step": 12704
+    },
+    {
+      "epoch": 0.3024135961153956,
+      "grad_norm": 5.44282902733574,
+      "learning_rate": 2e-05,
+      "loss": 0.1837,
+      "step": 12705
+    },
+    {
+      "epoch": 0.30243739883842713,
+      "grad_norm": 0.4461405832203592,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 12706
+    },
+    {
+      "epoch": 0.3024612015614586,
+      "grad_norm": 3.594574289953953,
+      "learning_rate": 2e-05,
+      "loss": 0.5287,
+      "step": 12707
+    },
+    {
+      "epoch": 0.30248500428449016,
+      "grad_norm": 5.133581787631197,
+      "learning_rate": 2e-05,
+      "loss": 0.6274,
+      "step": 12708
+    },
+    {
+      "epoch": 0.30250880700752164,
+      "grad_norm": 4.6746405019650945,
+      "learning_rate": 2e-05,
+      "loss": 0.5667,
+      "step": 12709
+    },
+    {
+      "epoch": 0.3025326097305532,
+      "grad_norm": 6.527240350302794,
+      "learning_rate": 2e-05,
+      "loss": 1.094,
+      "step": 12710
+    },
+    {
+      "epoch": 0.30255641245358467,
+      "grad_norm": 6.828835750405047,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 12711
+    },
+    {
+      "epoch": 0.3025802151766162,
+      "grad_norm": 4.078258218203635,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 12712
+    },
+    {
+      "epoch": 0.3026040178996477,
+      "grad_norm": 2.8311087270371815,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 12713
+    },
+    {
+      "epoch": 0.30262782062267923,
+      "grad_norm": 6.006678019562935,
+      "learning_rate": 2e-05,
+      "loss": 0.8846,
+      "step": 12714
+    },
+    {
+      "epoch": 0.3026516233457108,
+      "grad_norm": 11.464610061383201,
+      "learning_rate": 2e-05,
+      "loss": 0.1976,
+      "step": 12715
+    },
+    {
+      "epoch": 0.30267542606874226,
+      "grad_norm": 0.8932282124767195,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 12716
+    },
+    {
+      "epoch": 0.3026992287917738,
+      "grad_norm": 2.6664846736774668,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 12717
+    },
+    {
+      "epoch": 0.3027230315148053,
+      "grad_norm": 3.0336294307843628,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 12718
+    },
+    {
+      "epoch": 0.3027468342378368,
+      "grad_norm": 9.068576254093898,
+      "learning_rate": 2e-05,
+      "loss": 0.6016,
+      "step": 12719
+    },
+    {
+      "epoch": 0.3027706369608683,
+      "grad_norm": 6.013489723302669,
+      "learning_rate": 2e-05,
+      "loss": 1.022,
+      "step": 12720
+    },
+    {
+      "epoch": 0.30279443968389985,
+      "grad_norm": 6.587574160549052,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 12721
+    },
+    {
+      "epoch": 0.30281824240693134,
+      "grad_norm": 5.7850483102346555,
+      "learning_rate": 2e-05,
+      "loss": 0.7335,
+      "step": 12722
+    },
+    {
+      "epoch": 0.3028420451299629,
+      "grad_norm": 2.0168832283642577,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 12723
+    },
+    {
+      "epoch": 0.30286584785299436,
+      "grad_norm": 1.7671228819861955,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 12724
+    },
+    {
+      "epoch": 0.3028896505760259,
+      "grad_norm": 4.137426354691724,
+      "learning_rate": 2e-05,
+      "loss": 0.6013,
+      "step": 12725
+    },
+    {
+      "epoch": 0.3029134532990574,
+      "grad_norm": 3.8625190918156616,
+      "learning_rate": 2e-05,
+      "loss": 0.4977,
+      "step": 12726
+    },
+    {
+      "epoch": 0.30293725602208893,
+      "grad_norm": 0.5516655492742885,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 12727
+    },
+    {
+      "epoch": 0.3029610587451204,
+      "grad_norm": 2.5854486267263166,
+      "learning_rate": 2e-05,
+      "loss": 0.2327,
+      "step": 12728
+    },
+    {
+      "epoch": 0.30298486146815196,
+      "grad_norm": 0.32639313131239533,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 12729
+    },
+    {
+      "epoch": 0.3030086641911835,
+      "grad_norm": 0.7417053583386527,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 12730
+    },
+    {
+      "epoch": 0.303032466914215,
+      "grad_norm": 4.63365838367922,
+      "learning_rate": 2e-05,
+      "loss": 0.5979,
+      "step": 12731
+    },
+    {
+      "epoch": 0.3030562696372465,
+      "grad_norm": 5.1630643947342145,
+      "learning_rate": 2e-05,
+      "loss": 0.8238,
+      "step": 12732
+    },
+    {
+      "epoch": 0.303080072360278,
+      "grad_norm": 13.278596023937004,
+      "learning_rate": 2e-05,
+      "loss": 1.1547,
+      "step": 12733
+    },
+    {
+      "epoch": 0.30310387508330955,
+      "grad_norm": 0.7723147176767271,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 12734
+    },
+    {
+      "epoch": 0.30312767780634103,
+      "grad_norm": 3.447491369483594,
+      "learning_rate": 2e-05,
+      "loss": 0.6991,
+      "step": 12735
+    },
+    {
+      "epoch": 0.3031514805293726,
+      "grad_norm": 5.867077823765783,
+      "learning_rate": 2e-05,
+      "loss": 1.0063,
+      "step": 12736
+    },
+    {
+      "epoch": 0.30317528325240406,
+      "grad_norm": 0.28666680525204685,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 12737
+    },
+    {
+      "epoch": 0.3031990859754356,
+      "grad_norm": 10.65599745851903,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 12738
+    },
+    {
+      "epoch": 0.3032228886984671,
+      "grad_norm": 1.579143254821282,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 12739
+    },
+    {
+      "epoch": 0.3032466914214986,
+      "grad_norm": 3.784297955152443,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3032704941445301,
+      "grad_norm": 1.7554602115174411,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 12741
+    },
+    {
+      "epoch": 0.30329429686756165,
+      "grad_norm": 9.629699970743822,
+      "learning_rate": 2e-05,
+      "loss": 1.8441,
+      "step": 12742
+    },
+    {
+      "epoch": 0.30331809959059314,
+      "grad_norm": 35.92625094900329,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 12743
+    },
+    {
+      "epoch": 0.3033419023136247,
+      "grad_norm": 4.035298085284401,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 12744
+    },
+    {
+      "epoch": 0.3033657050366562,
+      "grad_norm": 10.519465907372485,
+      "learning_rate": 2e-05,
+      "loss": 0.4808,
+      "step": 12745
+    },
+    {
+      "epoch": 0.3033895077596877,
+      "grad_norm": 9.399461122126104,
+      "learning_rate": 2e-05,
+      "loss": 0.8453,
+      "step": 12746
+    },
+    {
+      "epoch": 0.30341331048271925,
+      "grad_norm": 1.6127816866361664,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 12747
+    },
+    {
+      "epoch": 0.30343711320575073,
+      "grad_norm": 4.339075572045028,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 12748
+    },
+    {
+      "epoch": 0.30346091592878227,
+      "grad_norm": 13.438780170286762,
+      "learning_rate": 2e-05,
+      "loss": 0.9441,
+      "step": 12749
+    },
+    {
+      "epoch": 0.30348471865181376,
+      "grad_norm": 3.0126285296045068,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3035085213748453,
+      "grad_norm": 4.572975208619349,
+      "learning_rate": 2e-05,
+      "loss": 0.6952,
+      "step": 12751
+    },
+    {
+      "epoch": 0.3035323240978768,
+      "grad_norm": 6.497987943399411,
+      "learning_rate": 2e-05,
+      "loss": 0.7411,
+      "step": 12752
+    },
+    {
+      "epoch": 0.3035561268209083,
+      "grad_norm": 5.663331833972584,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 12753
+    },
+    {
+      "epoch": 0.3035799295439398,
+      "grad_norm": 4.510409476521374,
+      "learning_rate": 2e-05,
+      "loss": 0.5802,
+      "step": 12754
+    },
+    {
+      "epoch": 0.30360373226697135,
+      "grad_norm": 4.794500069070384,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 12755
+    },
+    {
+      "epoch": 0.30362753499000283,
+      "grad_norm": 6.558884144046871,
+      "learning_rate": 2e-05,
+      "loss": 0.2838,
+      "step": 12756
+    },
+    {
+      "epoch": 0.3036513377130344,
+      "grad_norm": 10.699016870825325,
+      "learning_rate": 2e-05,
+      "loss": 0.6267,
+      "step": 12757
+    },
+    {
+      "epoch": 0.30367514043606586,
+      "grad_norm": 14.238546412861838,
+      "learning_rate": 2e-05,
+      "loss": 1.2495,
+      "step": 12758
+    },
+    {
+      "epoch": 0.3036989431590974,
+      "grad_norm": 2.046483282958636,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 12759
+    },
+    {
+      "epoch": 0.30372274588212894,
+      "grad_norm": 4.008680483080347,
+      "learning_rate": 2e-05,
+      "loss": 0.7782,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3037465486051604,
+      "grad_norm": 2.6164574909102964,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 12761
+    },
+    {
+      "epoch": 0.30377035132819197,
+      "grad_norm": 2.0702481941090563,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 12762
+    },
+    {
+      "epoch": 0.30379415405122345,
+      "grad_norm": 3.510692102284442,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 12763
+    },
+    {
+      "epoch": 0.303817956774255,
+      "grad_norm": 2.8390415453601765,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 12764
+    },
+    {
+      "epoch": 0.3038417594972865,
+      "grad_norm": 3.1849046086541244,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 12765
+    },
+    {
+      "epoch": 0.303865562220318,
+      "grad_norm": 7.5686449587901805,
+      "learning_rate": 2e-05,
+      "loss": 0.508,
+      "step": 12766
+    },
+    {
+      "epoch": 0.3038893649433495,
+      "grad_norm": 14.081595154481597,
+      "learning_rate": 2e-05,
+      "loss": 0.4432,
+      "step": 12767
+    },
+    {
+      "epoch": 0.30391316766638105,
+      "grad_norm": 4.1380534898571,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 12768
+    },
+    {
+      "epoch": 0.30393697038941253,
+      "grad_norm": 4.7209321870754355,
+      "learning_rate": 2e-05,
+      "loss": 0.6383,
+      "step": 12769
+    },
+    {
+      "epoch": 0.30396077311244407,
+      "grad_norm": 4.175764819629072,
+      "learning_rate": 2e-05,
+      "loss": 0.1577,
+      "step": 12770
+    },
+    {
+      "epoch": 0.30398457583547556,
+      "grad_norm": 16.484737226379725,
+      "learning_rate": 2e-05,
+      "loss": 0.8919,
+      "step": 12771
+    },
+    {
+      "epoch": 0.3040083785585071,
+      "grad_norm": 7.0284203640273875,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 12772
+    },
+    {
+      "epoch": 0.3040321812815386,
+      "grad_norm": 8.324340684440115,
+      "learning_rate": 2e-05,
+      "loss": 0.3019,
+      "step": 12773
+    },
+    {
+      "epoch": 0.3040559840045701,
+      "grad_norm": 2.6389112978615894,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 12774
+    },
+    {
+      "epoch": 0.30407978672760166,
+      "grad_norm": 5.73251945700114,
+      "learning_rate": 2e-05,
+      "loss": 0.5707,
+      "step": 12775
+    },
+    {
+      "epoch": 0.30410358945063315,
+      "grad_norm": 8.818773407095136,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 12776
+    },
+    {
+      "epoch": 0.3041273921736647,
+      "grad_norm": 5.679903940918956,
+      "learning_rate": 2e-05,
+      "loss": 0.2014,
+      "step": 12777
+    },
+    {
+      "epoch": 0.3041511948966962,
+      "grad_norm": 9.296438577724265,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 12778
+    },
+    {
+      "epoch": 0.3041749976197277,
+      "grad_norm": 2.014894475058159,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 12779
+    },
+    {
+      "epoch": 0.3041988003427592,
+      "grad_norm": 0.35150764720030075,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 12780
+    },
+    {
+      "epoch": 0.30422260306579074,
+      "grad_norm": 4.343209456257298,
+      "learning_rate": 2e-05,
+      "loss": 0.1345,
+      "step": 12781
+    },
+    {
+      "epoch": 0.3042464057888222,
+      "grad_norm": 6.127789604279704,
+      "learning_rate": 2e-05,
+      "loss": 1.3382,
+      "step": 12782
+    },
+    {
+      "epoch": 0.30427020851185377,
+      "grad_norm": 0.5892351050329753,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 12783
+    },
+    {
+      "epoch": 0.30429401123488525,
+      "grad_norm": 3.139783363736221,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 12784
+    },
+    {
+      "epoch": 0.3043178139579168,
+      "grad_norm": 0.26917538909447725,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 12785
+    },
+    {
+      "epoch": 0.3043416166809483,
+      "grad_norm": 5.285292316636162,
+      "learning_rate": 2e-05,
+      "loss": 0.6386,
+      "step": 12786
+    },
+    {
+      "epoch": 0.3043654194039798,
+      "grad_norm": 4.443983030741797,
+      "learning_rate": 2e-05,
+      "loss": 0.2974,
+      "step": 12787
+    },
+    {
+      "epoch": 0.3043892221270113,
+      "grad_norm": 4.126008140266584,
+      "learning_rate": 2e-05,
+      "loss": 0.6597,
+      "step": 12788
+    },
+    {
+      "epoch": 0.30441302485004285,
+      "grad_norm": 8.348716809816384,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 12789
+    },
+    {
+      "epoch": 0.3044368275730744,
+      "grad_norm": 4.884915483672075,
+      "learning_rate": 2e-05,
+      "loss": 0.7647,
+      "step": 12790
+    },
+    {
+      "epoch": 0.30446063029610587,
+      "grad_norm": 1.2629125465297442,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 12791
+    },
+    {
+      "epoch": 0.3044844330191374,
+      "grad_norm": 4.305059297138527,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 12792
+    },
+    {
+      "epoch": 0.3045082357421689,
+      "grad_norm": 2.103424140049377,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 12793
+    },
+    {
+      "epoch": 0.30453203846520044,
+      "grad_norm": 3.3573530814656594,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 12794
+    },
+    {
+      "epoch": 0.3045558411882319,
+      "grad_norm": 0.8744913167123172,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 12795
+    },
+    {
+      "epoch": 0.30457964391126346,
+      "grad_norm": 0.8473942414022791,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 12796
+    },
+    {
+      "epoch": 0.30460344663429495,
+      "grad_norm": 0.57892738692195,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 12797
+    },
+    {
+      "epoch": 0.3046272493573265,
+      "grad_norm": 4.319430136314287,
+      "learning_rate": 2e-05,
+      "loss": 0.3009,
+      "step": 12798
+    },
+    {
+      "epoch": 0.304651052080358,
+      "grad_norm": 9.106624223878235,
+      "learning_rate": 2e-05,
+      "loss": 0.3577,
+      "step": 12799
+    },
+    {
+      "epoch": 0.3046748548033895,
+      "grad_norm": 2.8115035993998063,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 12800
+    },
+    {
+      "epoch": 0.304698657526421,
+      "grad_norm": 7.339267141586349,
+      "learning_rate": 2e-05,
+      "loss": 0.9739,
+      "step": 12801
+    },
+    {
+      "epoch": 0.30472246024945254,
+      "grad_norm": 8.590685291380604,
+      "learning_rate": 2e-05,
+      "loss": 0.455,
+      "step": 12802
+    },
+    {
+      "epoch": 0.304746262972484,
+      "grad_norm": 4.883602050888714,
+      "learning_rate": 2e-05,
+      "loss": 0.201,
+      "step": 12803
+    },
+    {
+      "epoch": 0.30477006569551557,
+      "grad_norm": 1.9520057611369626,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 12804
+    },
+    {
+      "epoch": 0.3047938684185471,
+      "grad_norm": 2.8674542293166514,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 12805
+    },
+    {
+      "epoch": 0.3048176711415786,
+      "grad_norm": 0.06574526142009883,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 12806
+    },
+    {
+      "epoch": 0.30484147386461014,
+      "grad_norm": 5.19826099985157,
+      "learning_rate": 2e-05,
+      "loss": 0.8279,
+      "step": 12807
+    },
+    {
+      "epoch": 0.3048652765876416,
+      "grad_norm": 10.32817715182275,
+      "learning_rate": 2e-05,
+      "loss": 0.7395,
+      "step": 12808
+    },
+    {
+      "epoch": 0.30488907931067316,
+      "grad_norm": 3.023890717685703,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 12809
+    },
+    {
+      "epoch": 0.30491288203370465,
+      "grad_norm": 3.9986669290703154,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 12810
+    },
+    {
+      "epoch": 0.3049366847567362,
+      "grad_norm": 2.572492317269147,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 12811
+    },
+    {
+      "epoch": 0.3049604874797677,
+      "grad_norm": 3.920343621207752,
+      "learning_rate": 2e-05,
+      "loss": 0.6089,
+      "step": 12812
+    },
+    {
+      "epoch": 0.3049842902027992,
+      "grad_norm": 9.300903548414304,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 12813
+    },
+    {
+      "epoch": 0.3050080929258307,
+      "grad_norm": 9.74090685871255,
+      "learning_rate": 2e-05,
+      "loss": 0.4822,
+      "step": 12814
+    },
+    {
+      "epoch": 0.30503189564886224,
+      "grad_norm": 0.7692085697145246,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 12815
+    },
+    {
+      "epoch": 0.3050556983718937,
+      "grad_norm": 2.1386619136704175,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 12816
+    },
+    {
+      "epoch": 0.30507950109492527,
+      "grad_norm": 2.9031463663608257,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 12817
+    },
+    {
+      "epoch": 0.30510330381795675,
+      "grad_norm": 0.573349611877889,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 12818
+    },
+    {
+      "epoch": 0.3051271065409883,
+      "grad_norm": 4.660247188874274,
+      "learning_rate": 2e-05,
+      "loss": 0.9403,
+      "step": 12819
+    },
+    {
+      "epoch": 0.3051509092640198,
+      "grad_norm": 3.9958215692273455,
+      "learning_rate": 2e-05,
+      "loss": 0.4027,
+      "step": 12820
+    },
+    {
+      "epoch": 0.3051747119870513,
+      "grad_norm": 4.373168902754441,
+      "learning_rate": 2e-05,
+      "loss": 0.8122,
+      "step": 12821
+    },
+    {
+      "epoch": 0.30519851471008286,
+      "grad_norm": 9.169809841532663,
+      "learning_rate": 2e-05,
+      "loss": 1.4267,
+      "step": 12822
+    },
+    {
+      "epoch": 0.30522231743311434,
+      "grad_norm": 2.325166127447874,
+      "learning_rate": 2e-05,
+      "loss": 0.1126,
+      "step": 12823
+    },
+    {
+      "epoch": 0.3052461201561459,
+      "grad_norm": 3.936755333151441,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 12824
+    },
+    {
+      "epoch": 0.30526992287917737,
+      "grad_norm": 13.007325410662519,
+      "learning_rate": 2e-05,
+      "loss": 1.4606,
+      "step": 12825
+    },
+    {
+      "epoch": 0.3052937256022089,
+      "grad_norm": 4.35496268367294,
+      "learning_rate": 2e-05,
+      "loss": 0.8694,
+      "step": 12826
+    },
+    {
+      "epoch": 0.3053175283252404,
+      "grad_norm": 10.085325333892413,
+      "learning_rate": 2e-05,
+      "loss": 2.1923,
+      "step": 12827
+    },
+    {
+      "epoch": 0.30534133104827194,
+      "grad_norm": 5.382837492218066,
+      "learning_rate": 2e-05,
+      "loss": 0.5174,
+      "step": 12828
+    },
+    {
+      "epoch": 0.3053651337713034,
+      "grad_norm": 1.1523808069695016,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 12829
+    },
+    {
+      "epoch": 0.30538893649433496,
+      "grad_norm": 5.591782722239179,
+      "learning_rate": 2e-05,
+      "loss": 0.8219,
+      "step": 12830
+    },
+    {
+      "epoch": 0.30541273921736645,
+      "grad_norm": 4.844082572685352,
+      "learning_rate": 2e-05,
+      "loss": 0.5654,
+      "step": 12831
+    },
+    {
+      "epoch": 0.305436541940398,
+      "grad_norm": 7.074173149263713,
+      "learning_rate": 2e-05,
+      "loss": 0.5753,
+      "step": 12832
+    },
+    {
+      "epoch": 0.3054603446634295,
+      "grad_norm": 6.996235408489724,
+      "learning_rate": 2e-05,
+      "loss": 0.3453,
+      "step": 12833
+    },
+    {
+      "epoch": 0.305484147386461,
+      "grad_norm": 3.31952069672338,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 12834
+    },
+    {
+      "epoch": 0.3055079501094925,
+      "grad_norm": 0.697478242446909,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 12835
+    },
+    {
+      "epoch": 0.30553175283252404,
+      "grad_norm": 1.7923941843101696,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 12836
+    },
+    {
+      "epoch": 0.3055555555555556,
+      "grad_norm": 2.4757784268514054,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 12837
+    },
+    {
+      "epoch": 0.30557935827858707,
+      "grad_norm": 3.5337152974324897,
+      "learning_rate": 2e-05,
+      "loss": 0.5518,
+      "step": 12838
+    },
+    {
+      "epoch": 0.3056031610016186,
+      "grad_norm": 4.164047344746438,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 12839
+    },
+    {
+      "epoch": 0.3056269637246501,
+      "grad_norm": 12.35005652692792,
+      "learning_rate": 2e-05,
+      "loss": 1.2363,
+      "step": 12840
+    },
+    {
+      "epoch": 0.30565076644768163,
+      "grad_norm": 3.4102808638851707,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 12841
+    },
+    {
+      "epoch": 0.3056745691707131,
+      "grad_norm": 3.8648699087862637,
+      "learning_rate": 2e-05,
+      "loss": 0.3989,
+      "step": 12842
+    },
+    {
+      "epoch": 0.30569837189374466,
+      "grad_norm": 8.326117063202927,
+      "learning_rate": 2e-05,
+      "loss": 1.2347,
+      "step": 12843
+    },
+    {
+      "epoch": 0.30572217461677614,
+      "grad_norm": 7.8801599330811385,
+      "learning_rate": 2e-05,
+      "loss": 1.7608,
+      "step": 12844
+    },
+    {
+      "epoch": 0.3057459773398077,
+      "grad_norm": 3.899750052039258,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 12845
+    },
+    {
+      "epoch": 0.30576978006283917,
+      "grad_norm": 2.543919744944345,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 12846
+    },
+    {
+      "epoch": 0.3057935827858707,
+      "grad_norm": 2.8395548221054545,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 12847
+    },
+    {
+      "epoch": 0.3058173855089022,
+      "grad_norm": 8.001975161537016,
+      "learning_rate": 2e-05,
+      "loss": 0.2626,
+      "step": 12848
+    },
+    {
+      "epoch": 0.30584118823193374,
+      "grad_norm": 2.834522274084813,
+      "learning_rate": 2e-05,
+      "loss": 0.3393,
+      "step": 12849
+    },
+    {
+      "epoch": 0.3058649909549652,
+      "grad_norm": 9.397434713118347,
+      "learning_rate": 2e-05,
+      "loss": 1.5587,
+      "step": 12850
+    },
+    {
+      "epoch": 0.30588879367799676,
+      "grad_norm": 3.575448751570637,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 12851
+    },
+    {
+      "epoch": 0.3059125964010283,
+      "grad_norm": 4.181464926521813,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 12852
+    },
+    {
+      "epoch": 0.3059363991240598,
+      "grad_norm": 2.208283308278166,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 12853
+    },
+    {
+      "epoch": 0.30596020184709133,
+      "grad_norm": 2.898219632030574,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 12854
+    },
+    {
+      "epoch": 0.3059840045701228,
+      "grad_norm": 3.364164189895957,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 12855
+    },
+    {
+      "epoch": 0.30600780729315435,
+      "grad_norm": 2.488121412162114,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 12856
+    },
+    {
+      "epoch": 0.30603161001618584,
+      "grad_norm": 6.882535926157731,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 12857
+    },
+    {
+      "epoch": 0.3060554127392174,
+      "grad_norm": 2.464614878438455,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 12858
+    },
+    {
+      "epoch": 0.30607921546224887,
+      "grad_norm": 3.5980731565747766,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 12859
+    },
+    {
+      "epoch": 0.3061030181852804,
+      "grad_norm": 4.939553533601128,
+      "learning_rate": 2e-05,
+      "loss": 0.5094,
+      "step": 12860
+    },
+    {
+      "epoch": 0.3061268209083119,
+      "grad_norm": 3.4853137144284574,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 12861
+    },
+    {
+      "epoch": 0.30615062363134343,
+      "grad_norm": 0.9029197840904931,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 12862
+    },
+    {
+      "epoch": 0.3061744263543749,
+      "grad_norm": 4.3480933881759025,
+      "learning_rate": 2e-05,
+      "loss": 0.9268,
+      "step": 12863
+    },
+    {
+      "epoch": 0.30619822907740646,
+      "grad_norm": 6.837072893402318,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 12864
+    },
+    {
+      "epoch": 0.30622203180043794,
+      "grad_norm": 5.609848233240594,
+      "learning_rate": 2e-05,
+      "loss": 0.8575,
+      "step": 12865
+    },
+    {
+      "epoch": 0.3062458345234695,
+      "grad_norm": 2.813852496704329,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 12866
+    },
+    {
+      "epoch": 0.306269637246501,
+      "grad_norm": 4.528295973095726,
+      "learning_rate": 2e-05,
+      "loss": 0.2768,
+      "step": 12867
+    },
+    {
+      "epoch": 0.3062934399695325,
+      "grad_norm": 4.208905283121277,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 12868
+    },
+    {
+      "epoch": 0.30631724269256405,
+      "grad_norm": 11.507584651271097,
+      "learning_rate": 2e-05,
+      "loss": 0.3495,
+      "step": 12869
+    },
+    {
+      "epoch": 0.30634104541559554,
+      "grad_norm": 4.952451118604737,
+      "learning_rate": 2e-05,
+      "loss": 0.3111,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3063648481386271,
+      "grad_norm": 7.062403207070471,
+      "learning_rate": 2e-05,
+      "loss": 0.8048,
+      "step": 12871
+    },
+    {
+      "epoch": 0.30638865086165856,
+      "grad_norm": 4.589307051104116,
+      "learning_rate": 2e-05,
+      "loss": 1.0008,
+      "step": 12872
+    },
+    {
+      "epoch": 0.3064124535846901,
+      "grad_norm": 14.76012573642848,
+      "learning_rate": 2e-05,
+      "loss": 0.5142,
+      "step": 12873
+    },
+    {
+      "epoch": 0.3064362563077216,
+      "grad_norm": 1.7682036156272531,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 12874
+    },
+    {
+      "epoch": 0.30646005903075313,
+      "grad_norm": 3.482441499182296,
+      "learning_rate": 2e-05,
+      "loss": 0.1993,
+      "step": 12875
+    },
+    {
+      "epoch": 0.3064838617537846,
+      "grad_norm": 1.4287524596149714,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 12876
+    },
+    {
+      "epoch": 0.30650766447681616,
+      "grad_norm": 1.2452935447152809,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 12877
+    },
+    {
+      "epoch": 0.30653146719984764,
+      "grad_norm": 8.448985509855412,
+      "learning_rate": 2e-05,
+      "loss": 0.8965,
+      "step": 12878
+    },
+    {
+      "epoch": 0.3065552699228792,
+      "grad_norm": 10.970794985744181,
+      "learning_rate": 2e-05,
+      "loss": 0.3628,
+      "step": 12879
+    },
+    {
+      "epoch": 0.30657907264591067,
+      "grad_norm": 4.120068227043072,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3066028753689422,
+      "grad_norm": 4.431265228013844,
+      "learning_rate": 2e-05,
+      "loss": 0.8984,
+      "step": 12881
+    },
+    {
+      "epoch": 0.30662667809197375,
+      "grad_norm": 4.617412643008992,
+      "learning_rate": 2e-05,
+      "loss": 0.9358,
+      "step": 12882
+    },
+    {
+      "epoch": 0.30665048081500523,
+      "grad_norm": 2.3223114725332574,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 12883
+    },
+    {
+      "epoch": 0.3066742835380368,
+      "grad_norm": 6.708504831804915,
+      "learning_rate": 2e-05,
+      "loss": 0.3646,
+      "step": 12884
+    },
+    {
+      "epoch": 0.30669808626106826,
+      "grad_norm": 1.260485189316377,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 12885
+    },
+    {
+      "epoch": 0.3067218889840998,
+      "grad_norm": 5.441641507713506,
+      "learning_rate": 2e-05,
+      "loss": 0.2622,
+      "step": 12886
+    },
+    {
+      "epoch": 0.3067456917071313,
+      "grad_norm": 1.1694195956780626,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 12887
+    },
+    {
+      "epoch": 0.3067694944301628,
+      "grad_norm": 3.3588800806057924,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 12888
+    },
+    {
+      "epoch": 0.3067932971531943,
+      "grad_norm": 5.42905186721062,
+      "learning_rate": 2e-05,
+      "loss": 0.7647,
+      "step": 12889
+    },
+    {
+      "epoch": 0.30681709987622585,
+      "grad_norm": 2.8128399506305586,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 12890
+    },
+    {
+      "epoch": 0.30684090259925734,
+      "grad_norm": 6.474329061322946,
+      "learning_rate": 2e-05,
+      "loss": 1.1011,
+      "step": 12891
+    },
+    {
+      "epoch": 0.3068647053222889,
+      "grad_norm": 4.1545384991368355,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 12892
+    },
+    {
+      "epoch": 0.30688850804532036,
+      "grad_norm": 4.0178880853591945,
+      "learning_rate": 2e-05,
+      "loss": 0.295,
+      "step": 12893
+    },
+    {
+      "epoch": 0.3069123107683519,
+      "grad_norm": 3.552490429610162,
+      "learning_rate": 2e-05,
+      "loss": 0.4404,
+      "step": 12894
+    },
+    {
+      "epoch": 0.3069361134913834,
+      "grad_norm": 1.8559070101790236,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 12895
+    },
+    {
+      "epoch": 0.30695991621441493,
+      "grad_norm": 2.18257893399652,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 12896
+    },
+    {
+      "epoch": 0.30698371893744647,
+      "grad_norm": 2.3660319099907046,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 12897
+    },
+    {
+      "epoch": 0.30700752166047796,
+      "grad_norm": 1.2710677346126256,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 12898
+    },
+    {
+      "epoch": 0.3070313243835095,
+      "grad_norm": 2.833288472908901,
+      "learning_rate": 2e-05,
+      "loss": 0.1594,
+      "step": 12899
+    },
+    {
+      "epoch": 0.307055127106541,
+      "grad_norm": 0.8495833073315793,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3070789298295725,
+      "grad_norm": 1.5330988561250392,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 12901
+    },
+    {
+      "epoch": 0.307102732552604,
+      "grad_norm": 10.421329342965599,
+      "learning_rate": 2e-05,
+      "loss": 0.9287,
+      "step": 12902
+    },
+    {
+      "epoch": 0.30712653527563555,
+      "grad_norm": 7.264583432179443,
+      "learning_rate": 2e-05,
+      "loss": 0.9363,
+      "step": 12903
+    },
+    {
+      "epoch": 0.30715033799866703,
+      "grad_norm": 1.9060473768870094,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 12904
+    },
+    {
+      "epoch": 0.3071741407216986,
+      "grad_norm": 2.840210174478234,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 12905
+    },
+    {
+      "epoch": 0.30719794344473006,
+      "grad_norm": 1.868648158643605,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 12906
+    },
+    {
+      "epoch": 0.3072217461677616,
+      "grad_norm": 9.069759453635076,
+      "learning_rate": 2e-05,
+      "loss": 0.9277,
+      "step": 12907
+    },
+    {
+      "epoch": 0.3072455488907931,
+      "grad_norm": 5.099009496320947,
+      "learning_rate": 2e-05,
+      "loss": 1.0585,
+      "step": 12908
+    },
+    {
+      "epoch": 0.3072693516138246,
+      "grad_norm": 0.7372529044967158,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 12909
+    },
+    {
+      "epoch": 0.3072931543368561,
+      "grad_norm": 8.246225788084214,
+      "learning_rate": 2e-05,
+      "loss": 0.657,
+      "step": 12910
+    },
+    {
+      "epoch": 0.30731695705988765,
+      "grad_norm": 4.330279690781075,
+      "learning_rate": 2e-05,
+      "loss": 0.6939,
+      "step": 12911
+    },
+    {
+      "epoch": 0.3073407597829192,
+      "grad_norm": 5.458720029061777,
+      "learning_rate": 2e-05,
+      "loss": 0.7731,
+      "step": 12912
+    },
+    {
+      "epoch": 0.3073645625059507,
+      "grad_norm": 1.900870785431342,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 12913
+    },
+    {
+      "epoch": 0.3073883652289822,
+      "grad_norm": 5.665199358254508,
+      "learning_rate": 2e-05,
+      "loss": 0.6655,
+      "step": 12914
+    },
+    {
+      "epoch": 0.3074121679520137,
+      "grad_norm": 0.6683311628143737,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 12915
+    },
+    {
+      "epoch": 0.30743597067504524,
+      "grad_norm": 11.985316095468455,
+      "learning_rate": 2e-05,
+      "loss": 0.6016,
+      "step": 12916
+    },
+    {
+      "epoch": 0.30745977339807673,
+      "grad_norm": 1.2803424492503113,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 12917
+    },
+    {
+      "epoch": 0.30748357612110827,
+      "grad_norm": 4.363618729914322,
+      "learning_rate": 2e-05,
+      "loss": 1.0351,
+      "step": 12918
+    },
+    {
+      "epoch": 0.30750737884413976,
+      "grad_norm": 12.12870591925291,
+      "learning_rate": 2e-05,
+      "loss": 0.4039,
+      "step": 12919
+    },
+    {
+      "epoch": 0.3075311815671713,
+      "grad_norm": 2.9425196537964102,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 12920
+    },
+    {
+      "epoch": 0.3075549842902028,
+      "grad_norm": 1.6706218196182934,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 12921
+    },
+    {
+      "epoch": 0.3075787870132343,
+      "grad_norm": 3.352152889307279,
+      "learning_rate": 2e-05,
+      "loss": 0.2035,
+      "step": 12922
+    },
+    {
+      "epoch": 0.3076025897362658,
+      "grad_norm": 6.068449110289314,
+      "learning_rate": 2e-05,
+      "loss": 1.545,
+      "step": 12923
+    },
+    {
+      "epoch": 0.30762639245929735,
+      "grad_norm": 4.651512097836698,
+      "learning_rate": 2e-05,
+      "loss": 0.7418,
+      "step": 12924
+    },
+    {
+      "epoch": 0.30765019518232883,
+      "grad_norm": 2.156690080834365,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 12925
+    },
+    {
+      "epoch": 0.3076739979053604,
+      "grad_norm": 4.182773726499404,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 12926
+    },
+    {
+      "epoch": 0.3076978006283919,
+      "grad_norm": 3.559839888211729,
+      "learning_rate": 2e-05,
+      "loss": 0.4684,
+      "step": 12927
+    },
+    {
+      "epoch": 0.3077216033514234,
+      "grad_norm": 6.998747125821224,
+      "learning_rate": 2e-05,
+      "loss": 0.2803,
+      "step": 12928
+    },
+    {
+      "epoch": 0.30774540607445494,
+      "grad_norm": 3.469994537110409,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 12929
+    },
+    {
+      "epoch": 0.3077692087974864,
+      "grad_norm": 3.3279716958893593,
+      "learning_rate": 2e-05,
+      "loss": 0.2496,
+      "step": 12930
+    },
+    {
+      "epoch": 0.30779301152051797,
+      "grad_norm": 1.841305422930775,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 12931
+    },
+    {
+      "epoch": 0.30781681424354945,
+      "grad_norm": 3.4153968445885696,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 12932
+    },
+    {
+      "epoch": 0.307840616966581,
+      "grad_norm": 3.7642092064392574,
+      "learning_rate": 2e-05,
+      "loss": 0.4341,
+      "step": 12933
+    },
+    {
+      "epoch": 0.3078644196896125,
+      "grad_norm": 12.473959432690096,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 12934
+    },
+    {
+      "epoch": 0.307888222412644,
+      "grad_norm": 3.0003593189103186,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 12935
+    },
+    {
+      "epoch": 0.3079120251356755,
+      "grad_norm": 2.0594034278722115,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 12936
+    },
+    {
+      "epoch": 0.30793582785870705,
+      "grad_norm": 3.247372063886515,
+      "learning_rate": 2e-05,
+      "loss": 0.1745,
+      "step": 12937
+    },
+    {
+      "epoch": 0.30795963058173853,
+      "grad_norm": 1.3960489556317035,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 12938
+    },
+    {
+      "epoch": 0.30798343330477007,
+      "grad_norm": 38.910299519784644,
+      "learning_rate": 2e-05,
+      "loss": 0.3096,
+      "step": 12939
+    },
+    {
+      "epoch": 0.30800723602780156,
+      "grad_norm": 2.5214395118759687,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3080310387508331,
+      "grad_norm": 4.726692414539878,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 12941
+    },
+    {
+      "epoch": 0.3080548414738646,
+      "grad_norm": 2.438895751214954,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 12942
+    },
+    {
+      "epoch": 0.3080786441968961,
+      "grad_norm": 4.457274923967615,
+      "learning_rate": 2e-05,
+      "loss": 0.6191,
+      "step": 12943
+    },
+    {
+      "epoch": 0.30810244691992766,
+      "grad_norm": 1.171790393971063,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 12944
+    },
+    {
+      "epoch": 0.30812624964295915,
+      "grad_norm": 0.40465839478837023,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 12945
+    },
+    {
+      "epoch": 0.3081500523659907,
+      "grad_norm": 2.3248947480436835,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 12946
+    },
+    {
+      "epoch": 0.3081738550890222,
+      "grad_norm": 2.5791691475537704,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 12947
+    },
+    {
+      "epoch": 0.3081976578120537,
+      "grad_norm": 4.966768396308892,
+      "learning_rate": 2e-05,
+      "loss": 0.255,
+      "step": 12948
+    },
+    {
+      "epoch": 0.3082214605350852,
+      "grad_norm": 3.5494233556944264,
+      "learning_rate": 2e-05,
+      "loss": 0.5324,
+      "step": 12949
+    },
+    {
+      "epoch": 0.30824526325811674,
+      "grad_norm": 7.670601795464566,
+      "learning_rate": 2e-05,
+      "loss": 0.7598,
+      "step": 12950
+    },
+    {
+      "epoch": 0.3082690659811482,
+      "grad_norm": 6.32127351002845,
+      "learning_rate": 2e-05,
+      "loss": 0.7903,
+      "step": 12951
+    },
+    {
+      "epoch": 0.30829286870417977,
+      "grad_norm": 3.936342217158921,
+      "learning_rate": 2e-05,
+      "loss": 0.7823,
+      "step": 12952
+    },
+    {
+      "epoch": 0.30831667142721125,
+      "grad_norm": 3.206255949967549,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 12953
+    },
+    {
+      "epoch": 0.3083404741502428,
+      "grad_norm": 4.8625464518397,
+      "learning_rate": 2e-05,
+      "loss": 0.8622,
+      "step": 12954
+    },
+    {
+      "epoch": 0.3083642768732743,
+      "grad_norm": 3.1345067766232058,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 12955
+    },
+    {
+      "epoch": 0.3083880795963058,
+      "grad_norm": 2.1254960916941235,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 12956
+    },
+    {
+      "epoch": 0.3084118823193373,
+      "grad_norm": 1.1654080966681222,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 12957
+    },
+    {
+      "epoch": 0.30843568504236885,
+      "grad_norm": 0.30731905620466615,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 12958
+    },
+    {
+      "epoch": 0.3084594877654004,
+      "grad_norm": 4.5555060214199425,
+      "learning_rate": 2e-05,
+      "loss": 0.4925,
+      "step": 12959
+    },
+    {
+      "epoch": 0.30848329048843187,
+      "grad_norm": 7.098671250761333,
+      "learning_rate": 2e-05,
+      "loss": 0.7142,
+      "step": 12960
+    },
+    {
+      "epoch": 0.3085070932114634,
+      "grad_norm": 4.188348116644954,
+      "learning_rate": 2e-05,
+      "loss": 0.7311,
+      "step": 12961
+    },
+    {
+      "epoch": 0.3085308959344949,
+      "grad_norm": 3.162749689208375,
+      "learning_rate": 2e-05,
+      "loss": 0.6236,
+      "step": 12962
+    },
+    {
+      "epoch": 0.30855469865752644,
+      "grad_norm": 4.4018346227083684,
+      "learning_rate": 2e-05,
+      "loss": 0.4097,
+      "step": 12963
+    },
+    {
+      "epoch": 0.3085785013805579,
+      "grad_norm": 4.133252148036837,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 12964
+    },
+    {
+      "epoch": 0.30860230410358946,
+      "grad_norm": 5.632105147188983,
+      "learning_rate": 2e-05,
+      "loss": 0.8979,
+      "step": 12965
+    },
+    {
+      "epoch": 0.30862610682662095,
+      "grad_norm": 5.487794088218929,
+      "learning_rate": 2e-05,
+      "loss": 1.1327,
+      "step": 12966
+    },
+    {
+      "epoch": 0.3086499095496525,
+      "grad_norm": 3.384060572966077,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 12967
+    },
+    {
+      "epoch": 0.308673712272684,
+      "grad_norm": 1.1664295622641077,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 12968
+    },
+    {
+      "epoch": 0.3086975149957155,
+      "grad_norm": 2.2888303762443654,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 12969
+    },
+    {
+      "epoch": 0.308721317718747,
+      "grad_norm": 2.6122027412230087,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 12970
+    },
+    {
+      "epoch": 0.30874512044177854,
+      "grad_norm": 2.1742015781796273,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 12971
+    },
+    {
+      "epoch": 0.30876892316481,
+      "grad_norm": 1.3545426541749874,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 12972
+    },
+    {
+      "epoch": 0.30879272588784157,
+      "grad_norm": 6.728325108242964,
+      "learning_rate": 2e-05,
+      "loss": 1.1775,
+      "step": 12973
+    },
+    {
+      "epoch": 0.3088165286108731,
+      "grad_norm": 6.323140779501377,
+      "learning_rate": 2e-05,
+      "loss": 1.373,
+      "step": 12974
+    },
+    {
+      "epoch": 0.3088403313339046,
+      "grad_norm": 9.825975113979473,
+      "learning_rate": 2e-05,
+      "loss": 0.5323,
+      "step": 12975
+    },
+    {
+      "epoch": 0.30886413405693613,
+      "grad_norm": 10.270816335307156,
+      "learning_rate": 2e-05,
+      "loss": 1.2405,
+      "step": 12976
+    },
+    {
+      "epoch": 0.3088879367799676,
+      "grad_norm": 7.514942271510781,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 12977
+    },
+    {
+      "epoch": 0.30891173950299916,
+      "grad_norm": 0.3770060660191759,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 12978
+    },
+    {
+      "epoch": 0.30893554222603065,
+      "grad_norm": 3.209383465735767,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 12979
+    },
+    {
+      "epoch": 0.3089593449490622,
+      "grad_norm": 7.08397702999503,
+      "learning_rate": 2e-05,
+      "loss": 1.4475,
+      "step": 12980
+    },
+    {
+      "epoch": 0.30898314767209367,
+      "grad_norm": 9.303576358453839,
+      "learning_rate": 2e-05,
+      "loss": 0.6587,
+      "step": 12981
+    },
+    {
+      "epoch": 0.3090069503951252,
+      "grad_norm": 8.791202723787961,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 12982
+    },
+    {
+      "epoch": 0.3090307531181567,
+      "grad_norm": 11.959549978574397,
+      "learning_rate": 2e-05,
+      "loss": 0.7609,
+      "step": 12983
+    },
+    {
+      "epoch": 0.30905455584118824,
+      "grad_norm": 2.7954755669574416,
+      "learning_rate": 2e-05,
+      "loss": 0.46,
+      "step": 12984
+    },
+    {
+      "epoch": 0.3090783585642197,
+      "grad_norm": 8.555072769399807,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 12985
+    },
+    {
+      "epoch": 0.30910216128725126,
+      "grad_norm": 2.356270455510463,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 12986
+    },
+    {
+      "epoch": 0.30912596401028275,
+      "grad_norm": 2.7176688555634088,
+      "learning_rate": 2e-05,
+      "loss": 0.4049,
+      "step": 12987
+    },
+    {
+      "epoch": 0.3091497667333143,
+      "grad_norm": 3.0267459769718608,
+      "learning_rate": 2e-05,
+      "loss": 0.6294,
+      "step": 12988
+    },
+    {
+      "epoch": 0.30917356945634583,
+      "grad_norm": 4.205916320692391,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 12989
+    },
+    {
+      "epoch": 0.3091973721793773,
+      "grad_norm": 3.9943930577695665,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 12990
+    },
+    {
+      "epoch": 0.30922117490240886,
+      "grad_norm": 3.3840025314137527,
+      "learning_rate": 2e-05,
+      "loss": 0.673,
+      "step": 12991
+    },
+    {
+      "epoch": 0.30924497762544034,
+      "grad_norm": 4.461185759501589,
+      "learning_rate": 2e-05,
+      "loss": 0.2773,
+      "step": 12992
+    },
+    {
+      "epoch": 0.3092687803484719,
+      "grad_norm": 0.8387172107483325,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 12993
+    },
+    {
+      "epoch": 0.30929258307150337,
+      "grad_norm": 1.3982008243874335,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 12994
+    },
+    {
+      "epoch": 0.3093163857945349,
+      "grad_norm": 1.0038291129822436,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 12995
+    },
+    {
+      "epoch": 0.3093401885175664,
+      "grad_norm": 6.697793000681085,
+      "learning_rate": 2e-05,
+      "loss": 1.3931,
+      "step": 12996
+    },
+    {
+      "epoch": 0.30936399124059794,
+      "grad_norm": 14.602957363548594,
+      "learning_rate": 2e-05,
+      "loss": 1.3333,
+      "step": 12997
+    },
+    {
+      "epoch": 0.3093877939636294,
+      "grad_norm": 4.452236907239815,
+      "learning_rate": 2e-05,
+      "loss": 0.617,
+      "step": 12998
+    },
+    {
+      "epoch": 0.30941159668666096,
+      "grad_norm": 3.6260851394162885,
+      "learning_rate": 2e-05,
+      "loss": 0.4476,
+      "step": 12999
+    },
+    {
+      "epoch": 0.30943539940969245,
+      "grad_norm": 0.8254944770771141,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 13000
+    },
+    {
+      "epoch": 0.309459202132724,
+      "grad_norm": 4.876650016794009,
+      "learning_rate": 2e-05,
+      "loss": 0.2617,
+      "step": 13001
+    },
+    {
+      "epoch": 0.3094830048557555,
+      "grad_norm": 0.6548879785260258,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 13002
+    },
+    {
+      "epoch": 0.309506807578787,
+      "grad_norm": 0.5636392025218531,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 13003
+    },
+    {
+      "epoch": 0.30953061030181855,
+      "grad_norm": 6.303486045677687,
+      "learning_rate": 2e-05,
+      "loss": 1.15,
+      "step": 13004
+    },
+    {
+      "epoch": 0.30955441302485004,
+      "grad_norm": 0.36739943609749554,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13005
+    },
+    {
+      "epoch": 0.3095782157478816,
+      "grad_norm": 30.225272966548896,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 13006
+    },
+    {
+      "epoch": 0.30960201847091307,
+      "grad_norm": 6.373282693142321,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 13007
+    },
+    {
+      "epoch": 0.3096258211939446,
+      "grad_norm": 1.908643799912265,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 13008
+    },
+    {
+      "epoch": 0.3096496239169761,
+      "grad_norm": 1.2719618710517302,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 13009
+    },
+    {
+      "epoch": 0.30967342664000763,
+      "grad_norm": 3.6907226233511112,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 13010
+    },
+    {
+      "epoch": 0.3096972293630391,
+      "grad_norm": 0.9445461379390145,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 13011
+    },
+    {
+      "epoch": 0.30972103208607066,
+      "grad_norm": 0.6452790975791124,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 13012
+    },
+    {
+      "epoch": 0.30974483480910214,
+      "grad_norm": 3.5893462880385623,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 13013
+    },
+    {
+      "epoch": 0.3097686375321337,
+      "grad_norm": 18.770560092500748,
+      "learning_rate": 2e-05,
+      "loss": 0.2835,
+      "step": 13014
+    },
+    {
+      "epoch": 0.30979244025516517,
+      "grad_norm": 3.101634382722931,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 13015
+    },
+    {
+      "epoch": 0.3098162429781967,
+      "grad_norm": 1.062257818516551,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 13016
+    },
+    {
+      "epoch": 0.3098400457012282,
+      "grad_norm": 16.374583434921973,
+      "learning_rate": 2e-05,
+      "loss": 0.4362,
+      "step": 13017
+    },
+    {
+      "epoch": 0.30986384842425974,
+      "grad_norm": 2.3366681461650805,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 13018
+    },
+    {
+      "epoch": 0.3098876511472913,
+      "grad_norm": 4.192168655426618,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 13019
+    },
+    {
+      "epoch": 0.30991145387032276,
+      "grad_norm": 2.51547336894344,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 13020
+    },
+    {
+      "epoch": 0.3099352565933543,
+      "grad_norm": 7.324415386667835,
+      "learning_rate": 2e-05,
+      "loss": 0.7334,
+      "step": 13021
+    },
+    {
+      "epoch": 0.3099590593163858,
+      "grad_norm": 5.332112139957136,
+      "learning_rate": 2e-05,
+      "loss": 0.5884,
+      "step": 13022
+    },
+    {
+      "epoch": 0.30998286203941733,
+      "grad_norm": 3.825680388832461,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 13023
+    },
+    {
+      "epoch": 0.3100066647624488,
+      "grad_norm": 2.3460888612087127,
+      "learning_rate": 2e-05,
+      "loss": 0.1031,
+      "step": 13024
+    },
+    {
+      "epoch": 0.31003046748548035,
+      "grad_norm": 9.5227336697073,
+      "learning_rate": 2e-05,
+      "loss": 0.7813,
+      "step": 13025
+    },
+    {
+      "epoch": 0.31005427020851184,
+      "grad_norm": 2.3920259175889695,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 13026
+    },
+    {
+      "epoch": 0.3100780729315434,
+      "grad_norm": 3.3161390498612193,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 13027
+    },
+    {
+      "epoch": 0.31010187565457487,
+      "grad_norm": 6.663891863562233,
+      "learning_rate": 2e-05,
+      "loss": 0.2855,
+      "step": 13028
+    },
+    {
+      "epoch": 0.3101256783776064,
+      "grad_norm": 8.301099595341578,
+      "learning_rate": 2e-05,
+      "loss": 0.4634,
+      "step": 13029
+    },
+    {
+      "epoch": 0.3101494811006379,
+      "grad_norm": 3.745861423422696,
+      "learning_rate": 2e-05,
+      "loss": 0.6825,
+      "step": 13030
+    },
+    {
+      "epoch": 0.31017328382366943,
+      "grad_norm": 0.2532613057771385,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 13031
+    },
+    {
+      "epoch": 0.3101970865467009,
+      "grad_norm": 2.083940226119874,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 13032
+    },
+    {
+      "epoch": 0.31022088926973246,
+      "grad_norm": 0.5106024136447096,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 13033
+    },
+    {
+      "epoch": 0.310244691992764,
+      "grad_norm": 7.830157404846556,
+      "learning_rate": 2e-05,
+      "loss": 0.8694,
+      "step": 13034
+    },
+    {
+      "epoch": 0.3102684947157955,
+      "grad_norm": 0.49631540711612404,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 13035
+    },
+    {
+      "epoch": 0.310292297438827,
+      "grad_norm": 3.5023854680577258,
+      "learning_rate": 2e-05,
+      "loss": 0.4451,
+      "step": 13036
+    },
+    {
+      "epoch": 0.3103161001618585,
+      "grad_norm": 9.959071541783677,
+      "learning_rate": 2e-05,
+      "loss": 0.651,
+      "step": 13037
+    },
+    {
+      "epoch": 0.31033990288489005,
+      "grad_norm": 1.4022970584969898,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 13038
+    },
+    {
+      "epoch": 0.31036370560792154,
+      "grad_norm": 14.731473596450973,
+      "learning_rate": 2e-05,
+      "loss": 1.3965,
+      "step": 13039
+    },
+    {
+      "epoch": 0.3103875083309531,
+      "grad_norm": 3.535916963407496,
+      "learning_rate": 2e-05,
+      "loss": 0.3782,
+      "step": 13040
+    },
+    {
+      "epoch": 0.31041131105398456,
+      "grad_norm": 4.693732142145427,
+      "learning_rate": 2e-05,
+      "loss": 0.5877,
+      "step": 13041
+    },
+    {
+      "epoch": 0.3104351137770161,
+      "grad_norm": 8.422191817553475,
+      "learning_rate": 2e-05,
+      "loss": 0.897,
+      "step": 13042
+    },
+    {
+      "epoch": 0.3104589165000476,
+      "grad_norm": 6.603635011238346,
+      "learning_rate": 2e-05,
+      "loss": 1.1384,
+      "step": 13043
+    },
+    {
+      "epoch": 0.31048271922307913,
+      "grad_norm": 8.096226163931371,
+      "learning_rate": 2e-05,
+      "loss": 0.4222,
+      "step": 13044
+    },
+    {
+      "epoch": 0.3105065219461106,
+      "grad_norm": 4.267514081045677,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 13045
+    },
+    {
+      "epoch": 0.31053032466914215,
+      "grad_norm": 7.035679927430544,
+      "learning_rate": 2e-05,
+      "loss": 1.0088,
+      "step": 13046
+    },
+    {
+      "epoch": 0.31055412739217364,
+      "grad_norm": 0.8234301627236631,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 13047
+    },
+    {
+      "epoch": 0.3105779301152052,
+      "grad_norm": 5.389750775141076,
+      "learning_rate": 2e-05,
+      "loss": 0.1873,
+      "step": 13048
+    },
+    {
+      "epoch": 0.3106017328382367,
+      "grad_norm": 0.5141954628690535,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 13049
+    },
+    {
+      "epoch": 0.3106255355612682,
+      "grad_norm": 2.5379499561787693,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 13050
+    },
+    {
+      "epoch": 0.31064933828429975,
+      "grad_norm": 4.1017339704204865,
+      "learning_rate": 2e-05,
+      "loss": 0.6361,
+      "step": 13051
+    },
+    {
+      "epoch": 0.31067314100733123,
+      "grad_norm": 4.207278139605814,
+      "learning_rate": 2e-05,
+      "loss": 0.3586,
+      "step": 13052
+    },
+    {
+      "epoch": 0.3106969437303628,
+      "grad_norm": 5.534520047577759,
+      "learning_rate": 2e-05,
+      "loss": 0.4516,
+      "step": 13053
+    },
+    {
+      "epoch": 0.31072074645339426,
+      "grad_norm": 3.9840314416452873,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 13054
+    },
+    {
+      "epoch": 0.3107445491764258,
+      "grad_norm": 8.837616205880861,
+      "learning_rate": 2e-05,
+      "loss": 0.9376,
+      "step": 13055
+    },
+    {
+      "epoch": 0.3107683518994573,
+      "grad_norm": 6.995880149961044,
+      "learning_rate": 2e-05,
+      "loss": 0.974,
+      "step": 13056
+    },
+    {
+      "epoch": 0.3107921546224888,
+      "grad_norm": 2.1915772247771974,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 13057
+    },
+    {
+      "epoch": 0.3108159573455203,
+      "grad_norm": 3.4365289988971335,
+      "learning_rate": 2e-05,
+      "loss": 0.3839,
+      "step": 13058
+    },
+    {
+      "epoch": 0.31083976006855185,
+      "grad_norm": 2.227530510673123,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 13059
+    },
+    {
+      "epoch": 0.31086356279158334,
+      "grad_norm": 3.081778844663155,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 13060
+    },
+    {
+      "epoch": 0.3108873655146149,
+      "grad_norm": 1.7011896038919887,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 13061
+    },
+    {
+      "epoch": 0.31091116823764636,
+      "grad_norm": 9.121121672772068,
+      "learning_rate": 2e-05,
+      "loss": 1.3713,
+      "step": 13062
+    },
+    {
+      "epoch": 0.3109349709606779,
+      "grad_norm": 6.694633286751819,
+      "learning_rate": 2e-05,
+      "loss": 0.3721,
+      "step": 13063
+    },
+    {
+      "epoch": 0.31095877368370944,
+      "grad_norm": 3.0692821474643366,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 13064
+    },
+    {
+      "epoch": 0.31098257640674093,
+      "grad_norm": 4.507208463608577,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 13065
+    },
+    {
+      "epoch": 0.31100637912977247,
+      "grad_norm": 4.82107718638751,
+      "learning_rate": 2e-05,
+      "loss": 0.8397,
+      "step": 13066
+    },
+    {
+      "epoch": 0.31103018185280396,
+      "grad_norm": 1.7522425614911592,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 13067
+    },
+    {
+      "epoch": 0.3110539845758355,
+      "grad_norm": 2.4877580507457266,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 13068
+    },
+    {
+      "epoch": 0.311077787298867,
+      "grad_norm": 15.47938742296623,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 13069
+    },
+    {
+      "epoch": 0.3111015900218985,
+      "grad_norm": 5.546751882232235,
+      "learning_rate": 2e-05,
+      "loss": 1.0613,
+      "step": 13070
+    },
+    {
+      "epoch": 0.31112539274493,
+      "grad_norm": 3.1274446505967473,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 13071
+    },
+    {
+      "epoch": 0.31114919546796155,
+      "grad_norm": 5.884536196780016,
+      "learning_rate": 2e-05,
+      "loss": 1.129,
+      "step": 13072
+    },
+    {
+      "epoch": 0.31117299819099303,
+      "grad_norm": 7.0476314957722135,
+      "learning_rate": 2e-05,
+      "loss": 1.4894,
+      "step": 13073
+    },
+    {
+      "epoch": 0.3111968009140246,
+      "grad_norm": 0.8360129199058526,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 13074
+    },
+    {
+      "epoch": 0.31122060363705606,
+      "grad_norm": 2.2361307673040764,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 13075
+    },
+    {
+      "epoch": 0.3112444063600876,
+      "grad_norm": 13.767050645405563,
+      "learning_rate": 2e-05,
+      "loss": 0.629,
+      "step": 13076
+    },
+    {
+      "epoch": 0.3112682090831191,
+      "grad_norm": 11.3195884071285,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 13077
+    },
+    {
+      "epoch": 0.3112920118061506,
+      "grad_norm": 7.468073689454454,
+      "learning_rate": 2e-05,
+      "loss": 0.9025,
+      "step": 13078
+    },
+    {
+      "epoch": 0.3113158145291821,
+      "grad_norm": 3.4793140185641294,
+      "learning_rate": 2e-05,
+      "loss": 0.5356,
+      "step": 13079
+    },
+    {
+      "epoch": 0.31133961725221365,
+      "grad_norm": 3.268394468471645,
+      "learning_rate": 2e-05,
+      "loss": 0.5454,
+      "step": 13080
+    },
+    {
+      "epoch": 0.3113634199752452,
+      "grad_norm": 2.5564420396117096,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 13081
+    },
+    {
+      "epoch": 0.3113872226982767,
+      "grad_norm": 7.703062284320292,
+      "learning_rate": 2e-05,
+      "loss": 0.724,
+      "step": 13082
+    },
+    {
+      "epoch": 0.3114110254213082,
+      "grad_norm": 3.3633495276557888,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 13083
+    },
+    {
+      "epoch": 0.3114348281443397,
+      "grad_norm": 21.07104306789516,
+      "learning_rate": 2e-05,
+      "loss": 2.164,
+      "step": 13084
+    },
+    {
+      "epoch": 0.31145863086737124,
+      "grad_norm": 2.5460455850821204,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 13085
+    },
+    {
+      "epoch": 0.31148243359040273,
+      "grad_norm": 9.819037983899259,
+      "learning_rate": 2e-05,
+      "loss": 1.6703,
+      "step": 13086
+    },
+    {
+      "epoch": 0.31150623631343427,
+      "grad_norm": 13.479506257388962,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 13087
+    },
+    {
+      "epoch": 0.31153003903646576,
+      "grad_norm": 1.57818772115972,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 13088
+    },
+    {
+      "epoch": 0.3115538417594973,
+      "grad_norm": 5.203871675605622,
+      "learning_rate": 2e-05,
+      "loss": 1.2119,
+      "step": 13089
+    },
+    {
+      "epoch": 0.3115776444825288,
+      "grad_norm": 1.7453752624425187,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 13090
+    },
+    {
+      "epoch": 0.3116014472055603,
+      "grad_norm": 3.985226909906183,
+      "learning_rate": 2e-05,
+      "loss": 0.6371,
+      "step": 13091
+    },
+    {
+      "epoch": 0.3116252499285918,
+      "grad_norm": 6.123792616039544,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 13092
+    },
+    {
+      "epoch": 0.31164905265162335,
+      "grad_norm": 6.380764698512492,
+      "learning_rate": 2e-05,
+      "loss": 1.1901,
+      "step": 13093
+    },
+    {
+      "epoch": 0.31167285537465483,
+      "grad_norm": 3.055605023814318,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 13094
+    },
+    {
+      "epoch": 0.3116966580976864,
+      "grad_norm": 4.615799956212269,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 13095
+    },
+    {
+      "epoch": 0.3117204608207179,
+      "grad_norm": 2.674756452913969,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 13096
+    },
+    {
+      "epoch": 0.3117442635437494,
+      "grad_norm": 4.758252610759227,
+      "learning_rate": 2e-05,
+      "loss": 0.4052,
+      "step": 13097
+    },
+    {
+      "epoch": 0.31176806626678094,
+      "grad_norm": 4.406322919503192,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 13098
+    },
+    {
+      "epoch": 0.3117918689898124,
+      "grad_norm": 6.377131356231522,
+      "learning_rate": 2e-05,
+      "loss": 1.2701,
+      "step": 13099
+    },
+    {
+      "epoch": 0.31181567171284397,
+      "grad_norm": 5.995763914489485,
+      "learning_rate": 2e-05,
+      "loss": 1.0636,
+      "step": 13100
+    },
+    {
+      "epoch": 0.31183947443587545,
+      "grad_norm": 3.8864584922684675,
+      "learning_rate": 2e-05,
+      "loss": 0.5362,
+      "step": 13101
+    },
+    {
+      "epoch": 0.311863277158907,
+      "grad_norm": 3.642581696798102,
+      "learning_rate": 2e-05,
+      "loss": 0.4597,
+      "step": 13102
+    },
+    {
+      "epoch": 0.3118870798819385,
+      "grad_norm": 4.407333010482366,
+      "learning_rate": 2e-05,
+      "loss": 0.5433,
+      "step": 13103
+    },
+    {
+      "epoch": 0.31191088260497,
+      "grad_norm": 8.032057094068545,
+      "learning_rate": 2e-05,
+      "loss": 1.344,
+      "step": 13104
+    },
+    {
+      "epoch": 0.3119346853280015,
+      "grad_norm": 4.971182835690929,
+      "learning_rate": 2e-05,
+      "loss": 0.5562,
+      "step": 13105
+    },
+    {
+      "epoch": 0.31195848805103304,
+      "grad_norm": 1.5809156239024913,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 13106
+    },
+    {
+      "epoch": 0.31198229077406453,
+      "grad_norm": 13.87860529968784,
+      "learning_rate": 2e-05,
+      "loss": 0.6291,
+      "step": 13107
+    },
+    {
+      "epoch": 0.31200609349709607,
+      "grad_norm": 2.2804286866135977,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 13108
+    },
+    {
+      "epoch": 0.31202989622012756,
+      "grad_norm": 15.63081295509265,
+      "learning_rate": 2e-05,
+      "loss": 0.9129,
+      "step": 13109
+    },
+    {
+      "epoch": 0.3120536989431591,
+      "grad_norm": 5.650600554215076,
+      "learning_rate": 2e-05,
+      "loss": 1.0811,
+      "step": 13110
+    },
+    {
+      "epoch": 0.31207750166619064,
+      "grad_norm": 2.247830325439797,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 13111
+    },
+    {
+      "epoch": 0.3121013043892221,
+      "grad_norm": 3.987628868059838,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 13112
+    },
+    {
+      "epoch": 0.31212510711225366,
+      "grad_norm": 4.434286384485741,
+      "learning_rate": 2e-05,
+      "loss": 0.7884,
+      "step": 13113
+    },
+    {
+      "epoch": 0.31214890983528515,
+      "grad_norm": 2.741613354078758,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 13114
+    },
+    {
+      "epoch": 0.3121727125583167,
+      "grad_norm": 17.68483355758212,
+      "learning_rate": 2e-05,
+      "loss": 0.9612,
+      "step": 13115
+    },
+    {
+      "epoch": 0.3121965152813482,
+      "grad_norm": 7.062943957423289,
+      "learning_rate": 2e-05,
+      "loss": 1.3054,
+      "step": 13116
+    },
+    {
+      "epoch": 0.3122203180043797,
+      "grad_norm": 5.5098737416581045,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 13117
+    },
+    {
+      "epoch": 0.3122441207274112,
+      "grad_norm": 5.601804997347041,
+      "learning_rate": 2e-05,
+      "loss": 0.9879,
+      "step": 13118
+    },
+    {
+      "epoch": 0.31226792345044274,
+      "grad_norm": 3.7507537559668767,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 13119
+    },
+    {
+      "epoch": 0.3122917261734742,
+      "grad_norm": 9.748124287704227,
+      "learning_rate": 2e-05,
+      "loss": 0.3546,
+      "step": 13120
+    },
+    {
+      "epoch": 0.31231552889650577,
+      "grad_norm": 2.917368682430612,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 13121
+    },
+    {
+      "epoch": 0.31233933161953725,
+      "grad_norm": 7.654561482006495,
+      "learning_rate": 2e-05,
+      "loss": 0.9336,
+      "step": 13122
+    },
+    {
+      "epoch": 0.3123631343425688,
+      "grad_norm": 9.207217050176794,
+      "learning_rate": 2e-05,
+      "loss": 0.4587,
+      "step": 13123
+    },
+    {
+      "epoch": 0.3123869370656003,
+      "grad_norm": 4.983223135294721,
+      "learning_rate": 2e-05,
+      "loss": 0.5598,
+      "step": 13124
+    },
+    {
+      "epoch": 0.3124107397886318,
+      "grad_norm": 9.869125021629806,
+      "learning_rate": 2e-05,
+      "loss": 0.2958,
+      "step": 13125
+    },
+    {
+      "epoch": 0.31243454251166336,
+      "grad_norm": 15.305913576988281,
+      "learning_rate": 2e-05,
+      "loss": 1.8292,
+      "step": 13126
+    },
+    {
+      "epoch": 0.31245834523469485,
+      "grad_norm": 4.935327128436323,
+      "learning_rate": 2e-05,
+      "loss": 0.2337,
+      "step": 13127
+    },
+    {
+      "epoch": 0.3124821479577264,
+      "grad_norm": 3.73226710009982,
+      "learning_rate": 2e-05,
+      "loss": 0.2232,
+      "step": 13128
+    },
+    {
+      "epoch": 0.31250595068075787,
+      "grad_norm": 2.391119524125844,
+      "learning_rate": 2e-05,
+      "loss": 0.268,
+      "step": 13129
+    },
+    {
+      "epoch": 0.3125297534037894,
+      "grad_norm": 3.22536803328575,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 13130
+    },
+    {
+      "epoch": 0.3125535561268209,
+      "grad_norm": 1.7905007562380437,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 13131
+    },
+    {
+      "epoch": 0.31257735884985244,
+      "grad_norm": 8.888756011052594,
+      "learning_rate": 2e-05,
+      "loss": 0.8089,
+      "step": 13132
+    },
+    {
+      "epoch": 0.3126011615728839,
+      "grad_norm": 6.079994798928963,
+      "learning_rate": 2e-05,
+      "loss": 1.0893,
+      "step": 13133
+    },
+    {
+      "epoch": 0.31262496429591546,
+      "grad_norm": 3.335470393772837,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 13134
+    },
+    {
+      "epoch": 0.31264876701894695,
+      "grad_norm": 4.944149469563567,
+      "learning_rate": 2e-05,
+      "loss": 1.0842,
+      "step": 13135
+    },
+    {
+      "epoch": 0.3126725697419785,
+      "grad_norm": 2.3554782313022966,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 13136
+    },
+    {
+      "epoch": 0.31269637246501,
+      "grad_norm": 4.560160018847487,
+      "learning_rate": 2e-05,
+      "loss": 0.27,
+      "step": 13137
+    },
+    {
+      "epoch": 0.3127201751880415,
+      "grad_norm": 3.6094411283888372,
+      "learning_rate": 2e-05,
+      "loss": 0.438,
+      "step": 13138
+    },
+    {
+      "epoch": 0.312743977911073,
+      "grad_norm": 5.80110254835857,
+      "learning_rate": 2e-05,
+      "loss": 0.5031,
+      "step": 13139
+    },
+    {
+      "epoch": 0.31276778063410454,
+      "grad_norm": 4.890148194933355,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 13140
+    },
+    {
+      "epoch": 0.3127915833571361,
+      "grad_norm": 6.982898641090873,
+      "learning_rate": 2e-05,
+      "loss": 1.3892,
+      "step": 13141
+    },
+    {
+      "epoch": 0.31281538608016757,
+      "grad_norm": 3.6498698914875183,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 13142
+    },
+    {
+      "epoch": 0.3128391888031991,
+      "grad_norm": 1.4360911971121313,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 13143
+    },
+    {
+      "epoch": 0.3128629915262306,
+      "grad_norm": 1.0500290045929528,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 13144
+    },
+    {
+      "epoch": 0.31288679424926213,
+      "grad_norm": 5.171147202045403,
+      "learning_rate": 2e-05,
+      "loss": 0.6579,
+      "step": 13145
+    },
+    {
+      "epoch": 0.3129105969722936,
+      "grad_norm": 6.0173891101450545,
+      "learning_rate": 2e-05,
+      "loss": 0.7519,
+      "step": 13146
+    },
+    {
+      "epoch": 0.31293439969532516,
+      "grad_norm": 1.2029788365191962,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 13147
+    },
+    {
+      "epoch": 0.31295820241835665,
+      "grad_norm": 1.8502638045440862,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 13148
+    },
+    {
+      "epoch": 0.3129820051413882,
+      "grad_norm": 7.341809860235337,
+      "learning_rate": 2e-05,
+      "loss": 0.7711,
+      "step": 13149
+    },
+    {
+      "epoch": 0.31300580786441967,
+      "grad_norm": 4.945668113707848,
+      "learning_rate": 2e-05,
+      "loss": 0.7004,
+      "step": 13150
+    },
+    {
+      "epoch": 0.3130296105874512,
+      "grad_norm": 0.5070151335679085,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 13151
+    },
+    {
+      "epoch": 0.3130534133104827,
+      "grad_norm": 0.9576111395140757,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 13152
+    },
+    {
+      "epoch": 0.31307721603351424,
+      "grad_norm": 4.31407594595055,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 13153
+    },
+    {
+      "epoch": 0.3131010187565457,
+      "grad_norm": 1.0380644024018528,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 13154
+    },
+    {
+      "epoch": 0.31312482147957726,
+      "grad_norm": 5.572283301263178,
+      "learning_rate": 2e-05,
+      "loss": 0.7689,
+      "step": 13155
+    },
+    {
+      "epoch": 0.3131486242026088,
+      "grad_norm": 15.212180886621814,
+      "learning_rate": 2e-05,
+      "loss": 0.5886,
+      "step": 13156
+    },
+    {
+      "epoch": 0.3131724269256403,
+      "grad_norm": 4.108662075743345,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 13157
+    },
+    {
+      "epoch": 0.31319622964867183,
+      "grad_norm": 4.082262629404008,
+      "learning_rate": 2e-05,
+      "loss": 0.4423,
+      "step": 13158
+    },
+    {
+      "epoch": 0.3132200323717033,
+      "grad_norm": 0.9700880069166435,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 13159
+    },
+    {
+      "epoch": 0.31324383509473486,
+      "grad_norm": 5.587210530872863,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 13160
+    },
+    {
+      "epoch": 0.31326763781776634,
+      "grad_norm": 5.021123918616717,
+      "learning_rate": 2e-05,
+      "loss": 0.7705,
+      "step": 13161
+    },
+    {
+      "epoch": 0.3132914405407979,
+      "grad_norm": 6.1056605574018965,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 13162
+    },
+    {
+      "epoch": 0.31331524326382937,
+      "grad_norm": 8.906278022840189,
+      "learning_rate": 2e-05,
+      "loss": 2.3866,
+      "step": 13163
+    },
+    {
+      "epoch": 0.3133390459868609,
+      "grad_norm": 5.698757078144735,
+      "learning_rate": 2e-05,
+      "loss": 0.7486,
+      "step": 13164
+    },
+    {
+      "epoch": 0.3133628487098924,
+      "grad_norm": 5.3306502850597255,
+      "learning_rate": 2e-05,
+      "loss": 0.5826,
+      "step": 13165
+    },
+    {
+      "epoch": 0.31338665143292394,
+      "grad_norm": 3.376396537728915,
+      "learning_rate": 2e-05,
+      "loss": 0.3123,
+      "step": 13166
+    },
+    {
+      "epoch": 0.3134104541559554,
+      "grad_norm": 5.187976423743922,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 13167
+    },
+    {
+      "epoch": 0.31343425687898696,
+      "grad_norm": 12.357771818060268,
+      "learning_rate": 2e-05,
+      "loss": 0.6103,
+      "step": 13168
+    },
+    {
+      "epoch": 0.31345805960201845,
+      "grad_norm": 1.4507258907979432,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 13169
+    },
+    {
+      "epoch": 0.31348186232505,
+      "grad_norm": 1.4857824010109217,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 13170
+    },
+    {
+      "epoch": 0.3135056650480815,
+      "grad_norm": 5.17284201665448,
+      "learning_rate": 2e-05,
+      "loss": 0.6668,
+      "step": 13171
+    },
+    {
+      "epoch": 0.313529467771113,
+      "grad_norm": 7.217842786126545,
+      "learning_rate": 2e-05,
+      "loss": 0.9691,
+      "step": 13172
+    },
+    {
+      "epoch": 0.31355327049414455,
+      "grad_norm": 0.49887364277745916,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 13173
+    },
+    {
+      "epoch": 0.31357707321717604,
+      "grad_norm": 6.306625547887239,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 13174
+    },
+    {
+      "epoch": 0.3136008759402076,
+      "grad_norm": 2.3656572671981766,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 13175
+    },
+    {
+      "epoch": 0.31362467866323906,
+      "grad_norm": 5.642370313283819,
+      "learning_rate": 2e-05,
+      "loss": 1.2025,
+      "step": 13176
+    },
+    {
+      "epoch": 0.3136484813862706,
+      "grad_norm": 4.04674870103199,
+      "learning_rate": 2e-05,
+      "loss": 0.4845,
+      "step": 13177
+    },
+    {
+      "epoch": 0.3136722841093021,
+      "grad_norm": 3.127144192893964,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 13178
+    },
+    {
+      "epoch": 0.31369608683233363,
+      "grad_norm": 6.889667218402135,
+      "learning_rate": 2e-05,
+      "loss": 0.4184,
+      "step": 13179
+    },
+    {
+      "epoch": 0.3137198895553651,
+      "grad_norm": 5.7562017314728005,
+      "learning_rate": 2e-05,
+      "loss": 0.9791,
+      "step": 13180
+    },
+    {
+      "epoch": 0.31374369227839666,
+      "grad_norm": 1.828951515852055,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 13181
+    },
+    {
+      "epoch": 0.31376749500142814,
+      "grad_norm": 4.360838647387873,
+      "learning_rate": 2e-05,
+      "loss": 0.6589,
+      "step": 13182
+    },
+    {
+      "epoch": 0.3137912977244597,
+      "grad_norm": 4.268073998710021,
+      "learning_rate": 2e-05,
+      "loss": 0.2512,
+      "step": 13183
+    },
+    {
+      "epoch": 0.31381510044749117,
+      "grad_norm": 11.708026220155947,
+      "learning_rate": 2e-05,
+      "loss": 0.2245,
+      "step": 13184
+    },
+    {
+      "epoch": 0.3138389031705227,
+      "grad_norm": 3.0114142110084505,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 13185
+    },
+    {
+      "epoch": 0.31386270589355425,
+      "grad_norm": 6.205315990311454,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 13186
+    },
+    {
+      "epoch": 0.31388650861658574,
+      "grad_norm": 1.7928563769468402,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 13187
+    },
+    {
+      "epoch": 0.3139103113396173,
+      "grad_norm": 5.25010485910487,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 13188
+    },
+    {
+      "epoch": 0.31393411406264876,
+      "grad_norm": 5.5946896652887625,
+      "learning_rate": 2e-05,
+      "loss": 0.8134,
+      "step": 13189
+    },
+    {
+      "epoch": 0.3139579167856803,
+      "grad_norm": 12.718947992021363,
+      "learning_rate": 2e-05,
+      "loss": 0.804,
+      "step": 13190
+    },
+    {
+      "epoch": 0.3139817195087118,
+      "grad_norm": 2.6974089792400844,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 13191
+    },
+    {
+      "epoch": 0.31400552223174333,
+      "grad_norm": 6.925457913758823,
+      "learning_rate": 2e-05,
+      "loss": 0.6046,
+      "step": 13192
+    },
+    {
+      "epoch": 0.3140293249547748,
+      "grad_norm": 3.9318263143212215,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 13193
+    },
+    {
+      "epoch": 0.31405312767780635,
+      "grad_norm": 4.042700219799379,
+      "learning_rate": 2e-05,
+      "loss": 0.7274,
+      "step": 13194
+    },
+    {
+      "epoch": 0.31407693040083784,
+      "grad_norm": 5.853559165510553,
+      "learning_rate": 2e-05,
+      "loss": 0.4785,
+      "step": 13195
+    },
+    {
+      "epoch": 0.3141007331238694,
+      "grad_norm": 4.522603850464809,
+      "learning_rate": 2e-05,
+      "loss": 0.5316,
+      "step": 13196
+    },
+    {
+      "epoch": 0.31412453584690087,
+      "grad_norm": 3.505242603386455,
+      "learning_rate": 2e-05,
+      "loss": 0.6852,
+      "step": 13197
+    },
+    {
+      "epoch": 0.3141483385699324,
+      "grad_norm": 4.995429294312986,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 13198
+    },
+    {
+      "epoch": 0.3141721412929639,
+      "grad_norm": 0.7211241674257916,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 13199
+    },
+    {
+      "epoch": 0.31419594401599543,
+      "grad_norm": 1.8757784475851254,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 13200
+    },
+    {
+      "epoch": 0.3142197467390269,
+      "grad_norm": 4.721741033378745,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 13201
+    },
+    {
+      "epoch": 0.31424354946205846,
+      "grad_norm": 0.8181229735914901,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 13202
+    },
+    {
+      "epoch": 0.31426735218509,
+      "grad_norm": 6.540309298191324,
+      "learning_rate": 2e-05,
+      "loss": 0.2779,
+      "step": 13203
+    },
+    {
+      "epoch": 0.3142911549081215,
+      "grad_norm": 1.9970610412775094,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 13204
+    },
+    {
+      "epoch": 0.314314957631153,
+      "grad_norm": 3.1364465029231186,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 13205
+    },
+    {
+      "epoch": 0.3143387603541845,
+      "grad_norm": 17.325635140190634,
+      "learning_rate": 2e-05,
+      "loss": 0.2877,
+      "step": 13206
+    },
+    {
+      "epoch": 0.31436256307721605,
+      "grad_norm": 7.652984606899567,
+      "learning_rate": 2e-05,
+      "loss": 1.9258,
+      "step": 13207
+    },
+    {
+      "epoch": 0.31438636580024754,
+      "grad_norm": 3.1715117257942342,
+      "learning_rate": 2e-05,
+      "loss": 0.5783,
+      "step": 13208
+    },
+    {
+      "epoch": 0.3144101685232791,
+      "grad_norm": 2.312620739204382,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 13209
+    },
+    {
+      "epoch": 0.31443397124631056,
+      "grad_norm": 0.6149970644128204,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 13210
+    },
+    {
+      "epoch": 0.3144577739693421,
+      "grad_norm": 3.1528780922255533,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 13211
+    },
+    {
+      "epoch": 0.3144815766923736,
+      "grad_norm": 13.978163621878245,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 13212
+    },
+    {
+      "epoch": 0.31450537941540513,
+      "grad_norm": 5.381331096877487,
+      "learning_rate": 2e-05,
+      "loss": 0.1977,
+      "step": 13213
+    },
+    {
+      "epoch": 0.3145291821384366,
+      "grad_norm": 3.7522277827826933,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 13214
+    },
+    {
+      "epoch": 0.31455298486146815,
+      "grad_norm": 8.401256419271908,
+      "learning_rate": 2e-05,
+      "loss": 0.4048,
+      "step": 13215
+    },
+    {
+      "epoch": 0.31457678758449964,
+      "grad_norm": 4.466658838251353,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 13216
+    },
+    {
+      "epoch": 0.3146005903075312,
+      "grad_norm": 0.32533754312704716,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 13217
+    },
+    {
+      "epoch": 0.3146243930305627,
+      "grad_norm": 0.22676546680008985,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 13218
+    },
+    {
+      "epoch": 0.3146481957535942,
+      "grad_norm": 1.0315606062654328,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 13219
+    },
+    {
+      "epoch": 0.31467199847662575,
+      "grad_norm": 2.759647431850478,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 13220
+    },
+    {
+      "epoch": 0.31469580119965723,
+      "grad_norm": 5.678797179534137,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 13221
+    },
+    {
+      "epoch": 0.3147196039226888,
+      "grad_norm": 7.983303264974569,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 13222
+    },
+    {
+      "epoch": 0.31474340664572026,
+      "grad_norm": 1.0249992357682385,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 13223
+    },
+    {
+      "epoch": 0.3147672093687518,
+      "grad_norm": 4.089261481106662,
+      "learning_rate": 2e-05,
+      "loss": 0.7291,
+      "step": 13224
+    },
+    {
+      "epoch": 0.3147910120917833,
+      "grad_norm": 3.3072579089653615,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 13225
+    },
+    {
+      "epoch": 0.3148148148148148,
+      "grad_norm": 4.158689075035785,
+      "learning_rate": 2e-05,
+      "loss": 0.7881,
+      "step": 13226
+    },
+    {
+      "epoch": 0.3148386175378463,
+      "grad_norm": 4.897020182253951,
+      "learning_rate": 2e-05,
+      "loss": 0.955,
+      "step": 13227
+    },
+    {
+      "epoch": 0.31486242026087785,
+      "grad_norm": 5.4233726092322305,
+      "learning_rate": 2e-05,
+      "loss": 0.3245,
+      "step": 13228
+    },
+    {
+      "epoch": 0.31488622298390934,
+      "grad_norm": 4.643395543984851,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 13229
+    },
+    {
+      "epoch": 0.3149100257069409,
+      "grad_norm": 0.48557337525400535,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 13230
+    },
+    {
+      "epoch": 0.31493382842997236,
+      "grad_norm": 2.9773070420120256,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 13231
+    },
+    {
+      "epoch": 0.3149576311530039,
+      "grad_norm": 0.495597507714623,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 13232
+    },
+    {
+      "epoch": 0.31498143387603544,
+      "grad_norm": 0.44437362853623086,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 13233
+    },
+    {
+      "epoch": 0.31500523659906693,
+      "grad_norm": 1.970637548140078,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 13234
+    },
+    {
+      "epoch": 0.31502903932209847,
+      "grad_norm": 5.903632559179478,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 13235
+    },
+    {
+      "epoch": 0.31505284204512995,
+      "grad_norm": 0.9079550817509066,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13236
+    },
+    {
+      "epoch": 0.3150766447681615,
+      "grad_norm": 4.419024065832351,
+      "learning_rate": 2e-05,
+      "loss": 0.5195,
+      "step": 13237
+    },
+    {
+      "epoch": 0.315100447491193,
+      "grad_norm": 20.507271948612413,
+      "learning_rate": 2e-05,
+      "loss": 0.2277,
+      "step": 13238
+    },
+    {
+      "epoch": 0.3151242502142245,
+      "grad_norm": 1.3897694978237969,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 13239
+    },
+    {
+      "epoch": 0.315148052937256,
+      "grad_norm": 4.036672045292543,
+      "learning_rate": 2e-05,
+      "loss": 0.4833,
+      "step": 13240
+    },
+    {
+      "epoch": 0.31517185566028755,
+      "grad_norm": 1.6042294973514561,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 13241
+    },
+    {
+      "epoch": 0.31519565838331903,
+      "grad_norm": 4.859909860570085,
+      "learning_rate": 2e-05,
+      "loss": 0.2451,
+      "step": 13242
+    },
+    {
+      "epoch": 0.3152194611063506,
+      "grad_norm": 3.8449360003713253,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 13243
+    },
+    {
+      "epoch": 0.31524326382938206,
+      "grad_norm": 3.067777773629524,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 13244
+    },
+    {
+      "epoch": 0.3152670665524136,
+      "grad_norm": 67.51382726382558,
+      "learning_rate": 2e-05,
+      "loss": 0.9034,
+      "step": 13245
+    },
+    {
+      "epoch": 0.3152908692754451,
+      "grad_norm": 8.065704265925746,
+      "learning_rate": 2e-05,
+      "loss": 1.2459,
+      "step": 13246
+    },
+    {
+      "epoch": 0.3153146719984766,
+      "grad_norm": 4.128746088391465,
+      "learning_rate": 2e-05,
+      "loss": 0.4487,
+      "step": 13247
+    },
+    {
+      "epoch": 0.31533847472150817,
+      "grad_norm": 8.494983700826042,
+      "learning_rate": 2e-05,
+      "loss": 0.6826,
+      "step": 13248
+    },
+    {
+      "epoch": 0.31536227744453965,
+      "grad_norm": 3.0410589600889333,
+      "learning_rate": 2e-05,
+      "loss": 0.2923,
+      "step": 13249
+    },
+    {
+      "epoch": 0.3153860801675712,
+      "grad_norm": 2.7491763555319872,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 13250
+    },
+    {
+      "epoch": 0.3154098828906027,
+      "grad_norm": 2.6538245091482366,
+      "learning_rate": 2e-05,
+      "loss": 0.1389,
+      "step": 13251
+    },
+    {
+      "epoch": 0.3154336856136342,
+      "grad_norm": 2.1298145416654024,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 13252
+    },
+    {
+      "epoch": 0.3154574883366657,
+      "grad_norm": 4.8070533266179485,
+      "learning_rate": 2e-05,
+      "loss": 0.1533,
+      "step": 13253
+    },
+    {
+      "epoch": 0.31548129105969724,
+      "grad_norm": 0.7310566262929948,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 13254
+    },
+    {
+      "epoch": 0.31550509378272873,
+      "grad_norm": 7.05676438781504,
+      "learning_rate": 2e-05,
+      "loss": 0.3314,
+      "step": 13255
+    },
+    {
+      "epoch": 0.31552889650576027,
+      "grad_norm": 2.1599643990653368,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 13256
+    },
+    {
+      "epoch": 0.31555269922879176,
+      "grad_norm": 4.476210015002093,
+      "learning_rate": 2e-05,
+      "loss": 0.7586,
+      "step": 13257
+    },
+    {
+      "epoch": 0.3155765019518233,
+      "grad_norm": 2.8000865758058446,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 13258
+    },
+    {
+      "epoch": 0.3156003046748548,
+      "grad_norm": 5.171343381973131,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 13259
+    },
+    {
+      "epoch": 0.3156241073978863,
+      "grad_norm": 15.965620628280844,
+      "learning_rate": 2e-05,
+      "loss": 0.9643,
+      "step": 13260
+    },
+    {
+      "epoch": 0.3156479101209178,
+      "grad_norm": 23.674787113161603,
+      "learning_rate": 2e-05,
+      "loss": 0.2471,
+      "step": 13261
+    },
+    {
+      "epoch": 0.31567171284394935,
+      "grad_norm": 1.3431331848576056,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 13262
+    },
+    {
+      "epoch": 0.3156955155669809,
+      "grad_norm": 2.375610434420202,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 13263
+    },
+    {
+      "epoch": 0.3157193182900124,
+      "grad_norm": 5.501202015605262,
+      "learning_rate": 2e-05,
+      "loss": 0.9143,
+      "step": 13264
+    },
+    {
+      "epoch": 0.3157431210130439,
+      "grad_norm": 2.0067540178968635,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 13265
+    },
+    {
+      "epoch": 0.3157669237360754,
+      "grad_norm": 33.776953293054156,
+      "learning_rate": 2e-05,
+      "loss": 0.6737,
+      "step": 13266
+    },
+    {
+      "epoch": 0.31579072645910694,
+      "grad_norm": 3.435452036026913,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 13267
+    },
+    {
+      "epoch": 0.3158145291821384,
+      "grad_norm": 5.637554630806454,
+      "learning_rate": 2e-05,
+      "loss": 0.8409,
+      "step": 13268
+    },
+    {
+      "epoch": 0.31583833190516997,
+      "grad_norm": 0.8871028380670664,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 13269
+    },
+    {
+      "epoch": 0.31586213462820145,
+      "grad_norm": 3.001050870542782,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 13270
+    },
+    {
+      "epoch": 0.315885937351233,
+      "grad_norm": 19.49051533850413,
+      "learning_rate": 2e-05,
+      "loss": 0.6011,
+      "step": 13271
+    },
+    {
+      "epoch": 0.3159097400742645,
+      "grad_norm": 2.79169506048951,
+      "learning_rate": 2e-05,
+      "loss": 0.1834,
+      "step": 13272
+    },
+    {
+      "epoch": 0.315933542797296,
+      "grad_norm": 13.88719568933096,
+      "learning_rate": 2e-05,
+      "loss": 0.6508,
+      "step": 13273
+    },
+    {
+      "epoch": 0.3159573455203275,
+      "grad_norm": 6.885008511817393,
+      "learning_rate": 2e-05,
+      "loss": 1.1705,
+      "step": 13274
+    },
+    {
+      "epoch": 0.31598114824335904,
+      "grad_norm": 2.365270097433382,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 13275
+    },
+    {
+      "epoch": 0.31600495096639053,
+      "grad_norm": 11.302159821019035,
+      "learning_rate": 2e-05,
+      "loss": 0.9245,
+      "step": 13276
+    },
+    {
+      "epoch": 0.31602875368942207,
+      "grad_norm": 3.8642568596279783,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 13277
+    },
+    {
+      "epoch": 0.3160525564124536,
+      "grad_norm": 11.354798187803883,
+      "learning_rate": 2e-05,
+      "loss": 0.6155,
+      "step": 13278
+    },
+    {
+      "epoch": 0.3160763591354851,
+      "grad_norm": 4.9387248182644425,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 13279
+    },
+    {
+      "epoch": 0.31610016185851664,
+      "grad_norm": 8.413289691826051,
+      "learning_rate": 2e-05,
+      "loss": 0.8838,
+      "step": 13280
+    },
+    {
+      "epoch": 0.3161239645815481,
+      "grad_norm": 1.4147870129224673,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 13281
+    },
+    {
+      "epoch": 0.31614776730457966,
+      "grad_norm": 4.519612966296033,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 13282
+    },
+    {
+      "epoch": 0.31617157002761115,
+      "grad_norm": 3.5536558969031202,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 13283
+    },
+    {
+      "epoch": 0.3161953727506427,
+      "grad_norm": 6.796428119764559,
+      "learning_rate": 2e-05,
+      "loss": 1.2494,
+      "step": 13284
+    },
+    {
+      "epoch": 0.3162191754736742,
+      "grad_norm": 2.7316852747502214,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 13285
+    },
+    {
+      "epoch": 0.3162429781967057,
+      "grad_norm": 0.21377805185736987,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 13286
+    },
+    {
+      "epoch": 0.3162667809197372,
+      "grad_norm": 4.4019616654024265,
+      "learning_rate": 2e-05,
+      "loss": 0.3289,
+      "step": 13287
+    },
+    {
+      "epoch": 0.31629058364276874,
+      "grad_norm": 2.722409904918532,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 13288
+    },
+    {
+      "epoch": 0.3163143863658002,
+      "grad_norm": 3.641596051297153,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 13289
+    },
+    {
+      "epoch": 0.31633818908883177,
+      "grad_norm": 2.786903848208161,
+      "learning_rate": 2e-05,
+      "loss": 0.1813,
+      "step": 13290
+    },
+    {
+      "epoch": 0.31636199181186325,
+      "grad_norm": 3.792023415187771,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 13291
+    },
+    {
+      "epoch": 0.3163857945348948,
+      "grad_norm": 7.9715112888979425,
+      "learning_rate": 2e-05,
+      "loss": 1.2599,
+      "step": 13292
+    },
+    {
+      "epoch": 0.31640959725792633,
+      "grad_norm": 5.746353651026401,
+      "learning_rate": 2e-05,
+      "loss": 1.3611,
+      "step": 13293
+    },
+    {
+      "epoch": 0.3164333999809578,
+      "grad_norm": 3.912360535185226,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 13294
+    },
+    {
+      "epoch": 0.31645720270398936,
+      "grad_norm": 13.080857835949153,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 13295
+    },
+    {
+      "epoch": 0.31648100542702085,
+      "grad_norm": 3.6315437285188334,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 13296
+    },
+    {
+      "epoch": 0.3165048081500524,
+      "grad_norm": 0.6223827091105479,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 13297
+    },
+    {
+      "epoch": 0.31652861087308387,
+      "grad_norm": 5.94449691313055,
+      "learning_rate": 2e-05,
+      "loss": 0.7348,
+      "step": 13298
+    },
+    {
+      "epoch": 0.3165524135961154,
+      "grad_norm": 2.6610351614960037,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 13299
+    },
+    {
+      "epoch": 0.3165762163191469,
+      "grad_norm": 2.3210876616348206,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 13300
+    },
+    {
+      "epoch": 0.31660001904217844,
+      "grad_norm": 3.990605850904389,
+      "learning_rate": 2e-05,
+      "loss": 0.8272,
+      "step": 13301
+    },
+    {
+      "epoch": 0.3166238217652099,
+      "grad_norm": 11.138738457248433,
+      "learning_rate": 2e-05,
+      "loss": 0.2342,
+      "step": 13302
+    },
+    {
+      "epoch": 0.31664762448824146,
+      "grad_norm": 6.154553807187657,
+      "learning_rate": 2e-05,
+      "loss": 0.7262,
+      "step": 13303
+    },
+    {
+      "epoch": 0.31667142721127295,
+      "grad_norm": 2.9581249285080324,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 13304
+    },
+    {
+      "epoch": 0.3166952299343045,
+      "grad_norm": 15.767822576537641,
+      "learning_rate": 2e-05,
+      "loss": 1.228,
+      "step": 13305
+    },
+    {
+      "epoch": 0.316719032657336,
+      "grad_norm": 3.415342692099187,
+      "learning_rate": 2e-05,
+      "loss": 0.2508,
+      "step": 13306
+    },
+    {
+      "epoch": 0.3167428353803675,
+      "grad_norm": 2.633359034923857,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 13307
+    },
+    {
+      "epoch": 0.31676663810339906,
+      "grad_norm": 4.2340012784382886,
+      "learning_rate": 2e-05,
+      "loss": 0.4256,
+      "step": 13308
+    },
+    {
+      "epoch": 0.31679044082643054,
+      "grad_norm": 8.666399911367858,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 13309
+    },
+    {
+      "epoch": 0.3168142435494621,
+      "grad_norm": 6.490924707969781,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 13310
+    },
+    {
+      "epoch": 0.31683804627249357,
+      "grad_norm": 11.266761311038618,
+      "learning_rate": 2e-05,
+      "loss": 1.3685,
+      "step": 13311
+    },
+    {
+      "epoch": 0.3168618489955251,
+      "grad_norm": 1.886679609218184,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 13312
+    },
+    {
+      "epoch": 0.3168856517185566,
+      "grad_norm": 13.648532861521423,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 13313
+    },
+    {
+      "epoch": 0.31690945444158813,
+      "grad_norm": 4.500705483562995,
+      "learning_rate": 2e-05,
+      "loss": 0.6008,
+      "step": 13314
+    },
+    {
+      "epoch": 0.3169332571646196,
+      "grad_norm": 3.7575373415892352,
+      "learning_rate": 2e-05,
+      "loss": 0.3799,
+      "step": 13315
+    },
+    {
+      "epoch": 0.31695705988765116,
+      "grad_norm": 6.950014087196554,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 13316
+    },
+    {
+      "epoch": 0.31698086261068265,
+      "grad_norm": 6.617290478528259,
+      "learning_rate": 2e-05,
+      "loss": 0.7602,
+      "step": 13317
+    },
+    {
+      "epoch": 0.3170046653337142,
+      "grad_norm": 2.4431838182341337,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 13318
+    },
+    {
+      "epoch": 0.31702846805674567,
+      "grad_norm": 11.855616148465984,
+      "learning_rate": 2e-05,
+      "loss": 0.642,
+      "step": 13319
+    },
+    {
+      "epoch": 0.3170522707797772,
+      "grad_norm": 5.831154273719251,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 13320
+    },
+    {
+      "epoch": 0.3170760735028087,
+      "grad_norm": 17.903540188340916,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 13321
+    },
+    {
+      "epoch": 0.31709987622584024,
+      "grad_norm": 7.165738230310089,
+      "learning_rate": 2e-05,
+      "loss": 0.7737,
+      "step": 13322
+    },
+    {
+      "epoch": 0.3171236789488717,
+      "grad_norm": 4.110735823436893,
+      "learning_rate": 2e-05,
+      "loss": 0.8281,
+      "step": 13323
+    },
+    {
+      "epoch": 0.31714748167190326,
+      "grad_norm": 0.6419587614058124,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 13324
+    },
+    {
+      "epoch": 0.3171712843949348,
+      "grad_norm": 6.132628813909216,
+      "learning_rate": 2e-05,
+      "loss": 0.5815,
+      "step": 13325
+    },
+    {
+      "epoch": 0.3171950871179663,
+      "grad_norm": 4.625702527790488,
+      "learning_rate": 2e-05,
+      "loss": 0.6227,
+      "step": 13326
+    },
+    {
+      "epoch": 0.31721888984099783,
+      "grad_norm": 7.681012690916501,
+      "learning_rate": 2e-05,
+      "loss": 0.6054,
+      "step": 13327
+    },
+    {
+      "epoch": 0.3172426925640293,
+      "grad_norm": 4.770208991900878,
+      "learning_rate": 2e-05,
+      "loss": 0.9084,
+      "step": 13328
+    },
+    {
+      "epoch": 0.31726649528706086,
+      "grad_norm": 1.219989339898635,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 13329
+    },
+    {
+      "epoch": 0.31729029801009234,
+      "grad_norm": 3.3426900516430647,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 13330
+    },
+    {
+      "epoch": 0.3173141007331239,
+      "grad_norm": 4.01687877768558,
+      "learning_rate": 2e-05,
+      "loss": 0.6914,
+      "step": 13331
+    },
+    {
+      "epoch": 0.31733790345615537,
+      "grad_norm": 3.00723352717336,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 13332
+    },
+    {
+      "epoch": 0.3173617061791869,
+      "grad_norm": 8.45449211835635,
+      "learning_rate": 2e-05,
+      "loss": 0.3654,
+      "step": 13333
+    },
+    {
+      "epoch": 0.3173855089022184,
+      "grad_norm": 3.659482930222107,
+      "learning_rate": 2e-05,
+      "loss": 0.4245,
+      "step": 13334
+    },
+    {
+      "epoch": 0.31740931162524993,
+      "grad_norm": 4.083137511112337,
+      "learning_rate": 2e-05,
+      "loss": 0.4906,
+      "step": 13335
+    },
+    {
+      "epoch": 0.3174331143482814,
+      "grad_norm": 2.9212679253226055,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 13336
+    },
+    {
+      "epoch": 0.31745691707131296,
+      "grad_norm": 7.995373311117112,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 13337
+    },
+    {
+      "epoch": 0.31748071979434445,
+      "grad_norm": 7.942252701618457,
+      "learning_rate": 2e-05,
+      "loss": 0.7495,
+      "step": 13338
+    },
+    {
+      "epoch": 0.317504522517376,
+      "grad_norm": 2.747293662274117,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 13339
+    },
+    {
+      "epoch": 0.3175283252404075,
+      "grad_norm": 2.1920274607202286,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 13340
+    },
+    {
+      "epoch": 0.317552127963439,
+      "grad_norm": 5.675110118571155,
+      "learning_rate": 2e-05,
+      "loss": 1.2588,
+      "step": 13341
+    },
+    {
+      "epoch": 0.31757593068647055,
+      "grad_norm": 3.3103538107115402,
+      "learning_rate": 2e-05,
+      "loss": 0.1181,
+      "step": 13342
+    },
+    {
+      "epoch": 0.31759973340950204,
+      "grad_norm": 5.171227899697726,
+      "learning_rate": 2e-05,
+      "loss": 0.7862,
+      "step": 13343
+    },
+    {
+      "epoch": 0.3176235361325336,
+      "grad_norm": 2.3096920973694033,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 13344
+    },
+    {
+      "epoch": 0.31764733885556506,
+      "grad_norm": 3.289120289307175,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 13345
+    },
+    {
+      "epoch": 0.3176711415785966,
+      "grad_norm": 13.476047874888085,
+      "learning_rate": 2e-05,
+      "loss": 0.4502,
+      "step": 13346
+    },
+    {
+      "epoch": 0.3176949443016281,
+      "grad_norm": 1.0811737683364087,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 13347
+    },
+    {
+      "epoch": 0.31771874702465963,
+      "grad_norm": 15.738916239223922,
+      "learning_rate": 2e-05,
+      "loss": 0.4225,
+      "step": 13348
+    },
+    {
+      "epoch": 0.3177425497476911,
+      "grad_norm": 2.2402754128560076,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 13349
+    },
+    {
+      "epoch": 0.31776635247072266,
+      "grad_norm": 3.6882454117604024,
+      "learning_rate": 2e-05,
+      "loss": 0.385,
+      "step": 13350
+    },
+    {
+      "epoch": 0.31779015519375414,
+      "grad_norm": 1.1596469577762123,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 13351
+    },
+    {
+      "epoch": 0.3178139579167857,
+      "grad_norm": 1.3080167095720763,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 13352
+    },
+    {
+      "epoch": 0.31783776063981717,
+      "grad_norm": 5.50541478255557,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 13353
+    },
+    {
+      "epoch": 0.3178615633628487,
+      "grad_norm": 1.7075667830521983,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 13354
+    },
+    {
+      "epoch": 0.31788536608588025,
+      "grad_norm": 4.612408169141737,
+      "learning_rate": 2e-05,
+      "loss": 0.5481,
+      "step": 13355
+    },
+    {
+      "epoch": 0.31790916880891174,
+      "grad_norm": 1.331252486008768,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 13356
+    },
+    {
+      "epoch": 0.3179329715319433,
+      "grad_norm": 5.801640946218633,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 13357
+    },
+    {
+      "epoch": 0.31795677425497476,
+      "grad_norm": 1.672856769755445,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 13358
+    },
+    {
+      "epoch": 0.3179805769780063,
+      "grad_norm": 4.910539945235195,
+      "learning_rate": 2e-05,
+      "loss": 0.9768,
+      "step": 13359
+    },
+    {
+      "epoch": 0.3180043797010378,
+      "grad_norm": 0.41311324175123754,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13360
+    },
+    {
+      "epoch": 0.31802818242406933,
+      "grad_norm": 4.396786244625053,
+      "learning_rate": 2e-05,
+      "loss": 0.7792,
+      "step": 13361
+    },
+    {
+      "epoch": 0.3180519851471008,
+      "grad_norm": 3.354036809929288,
+      "learning_rate": 2e-05,
+      "loss": 0.0513,
+      "step": 13362
+    },
+    {
+      "epoch": 0.31807578787013235,
+      "grad_norm": 7.082213829826151,
+      "learning_rate": 2e-05,
+      "loss": 1.4377,
+      "step": 13363
+    },
+    {
+      "epoch": 0.31809959059316384,
+      "grad_norm": 6.39767719836891,
+      "learning_rate": 2e-05,
+      "loss": 1.12,
+      "step": 13364
+    },
+    {
+      "epoch": 0.3181233933161954,
+      "grad_norm": 3.2154007488848917,
+      "learning_rate": 2e-05,
+      "loss": 0.3151,
+      "step": 13365
+    },
+    {
+      "epoch": 0.31814719603922686,
+      "grad_norm": 5.399393395814506,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 13366
+    },
+    {
+      "epoch": 0.3181709987622584,
+      "grad_norm": 13.07593007347557,
+      "learning_rate": 2e-05,
+      "loss": 0.7003,
+      "step": 13367
+    },
+    {
+      "epoch": 0.3181948014852899,
+      "grad_norm": 3.690383344796967,
+      "learning_rate": 2e-05,
+      "loss": 0.5333,
+      "step": 13368
+    },
+    {
+      "epoch": 0.31821860420832143,
+      "grad_norm": 8.270438193283905,
+      "learning_rate": 2e-05,
+      "loss": 0.3267,
+      "step": 13369
+    },
+    {
+      "epoch": 0.318242406931353,
+      "grad_norm": 2.7600653960170196,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 13370
+    },
+    {
+      "epoch": 0.31826620965438446,
+      "grad_norm": 4.177904685158704,
+      "learning_rate": 2e-05,
+      "loss": 0.2028,
+      "step": 13371
+    },
+    {
+      "epoch": 0.318290012377416,
+      "grad_norm": 0.3970854392327592,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 13372
+    },
+    {
+      "epoch": 0.3183138151004475,
+      "grad_norm": 0.8920606765086889,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 13373
+    },
+    {
+      "epoch": 0.318337617823479,
+      "grad_norm": 6.135553605552384,
+      "learning_rate": 2e-05,
+      "loss": 0.3505,
+      "step": 13374
+    },
+    {
+      "epoch": 0.3183614205465105,
+      "grad_norm": 5.3352550219935795,
+      "learning_rate": 2e-05,
+      "loss": 0.7738,
+      "step": 13375
+    },
+    {
+      "epoch": 0.31838522326954205,
+      "grad_norm": 8.237399999958942,
+      "learning_rate": 2e-05,
+      "loss": 0.711,
+      "step": 13376
+    },
+    {
+      "epoch": 0.31840902599257354,
+      "grad_norm": 2.3488024238657497,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 13377
+    },
+    {
+      "epoch": 0.3184328287156051,
+      "grad_norm": 10.104939805268163,
+      "learning_rate": 2e-05,
+      "loss": 0.761,
+      "step": 13378
+    },
+    {
+      "epoch": 0.31845663143863656,
+      "grad_norm": 3.215372036315253,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 13379
+    },
+    {
+      "epoch": 0.3184804341616681,
+      "grad_norm": 5.076335451534991,
+      "learning_rate": 2e-05,
+      "loss": 0.4569,
+      "step": 13380
+    },
+    {
+      "epoch": 0.3185042368846996,
+      "grad_norm": 1.816713571081374,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 13381
+    },
+    {
+      "epoch": 0.31852803960773113,
+      "grad_norm": 3.74113690366434,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 13382
+    },
+    {
+      "epoch": 0.3185518423307626,
+      "grad_norm": 1.037263139650094,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 13383
+    },
+    {
+      "epoch": 0.31857564505379415,
+      "grad_norm": 6.030455556927966,
+      "learning_rate": 2e-05,
+      "loss": 0.9118,
+      "step": 13384
+    },
+    {
+      "epoch": 0.3185994477768257,
+      "grad_norm": 6.8161784804126375,
+      "learning_rate": 2e-05,
+      "loss": 1.1106,
+      "step": 13385
+    },
+    {
+      "epoch": 0.3186232504998572,
+      "grad_norm": 3.567036950076067,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 13386
+    },
+    {
+      "epoch": 0.3186470532228887,
+      "grad_norm": 6.48299424727314,
+      "learning_rate": 2e-05,
+      "loss": 0.9245,
+      "step": 13387
+    },
+    {
+      "epoch": 0.3186708559459202,
+      "grad_norm": 3.476802466603216,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 13388
+    },
+    {
+      "epoch": 0.31869465866895175,
+      "grad_norm": 5.658315244079765,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 13389
+    },
+    {
+      "epoch": 0.31871846139198323,
+      "grad_norm": 1.3198799898643718,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 13390
+    },
+    {
+      "epoch": 0.3187422641150148,
+      "grad_norm": 1.3133161260450101,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 13391
+    },
+    {
+      "epoch": 0.31876606683804626,
+      "grad_norm": 6.23070083495262,
+      "learning_rate": 2e-05,
+      "loss": 1.3387,
+      "step": 13392
+    },
+    {
+      "epoch": 0.3187898695610778,
+      "grad_norm": 3.7148440380696655,
+      "learning_rate": 2e-05,
+      "loss": 0.5515,
+      "step": 13393
+    },
+    {
+      "epoch": 0.3188136722841093,
+      "grad_norm": 3.446200824918565,
+      "learning_rate": 2e-05,
+      "loss": 0.5321,
+      "step": 13394
+    },
+    {
+      "epoch": 0.3188374750071408,
+      "grad_norm": 2.0964143224354412,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 13395
+    },
+    {
+      "epoch": 0.3188612777301723,
+      "grad_norm": 0.6683686508057817,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 13396
+    },
+    {
+      "epoch": 0.31888508045320385,
+      "grad_norm": 0.4447847444417171,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 13397
+    },
+    {
+      "epoch": 0.31890888317623534,
+      "grad_norm": 20.31615667019048,
+      "learning_rate": 2e-05,
+      "loss": 0.7459,
+      "step": 13398
+    },
+    {
+      "epoch": 0.3189326858992669,
+      "grad_norm": 2.474667980309516,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 13399
+    },
+    {
+      "epoch": 0.3189564886222984,
+      "grad_norm": 1.8667037244660885,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 13400
+    },
+    {
+      "epoch": 0.3189802913453299,
+      "grad_norm": 8.01244308878162,
+      "learning_rate": 2e-05,
+      "loss": 0.989,
+      "step": 13401
+    },
+    {
+      "epoch": 0.31900409406836144,
+      "grad_norm": 4.4028355221268,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 13402
+    },
+    {
+      "epoch": 0.31902789679139293,
+      "grad_norm": 10.956928988937191,
+      "learning_rate": 2e-05,
+      "loss": 0.8651,
+      "step": 13403
+    },
+    {
+      "epoch": 0.31905169951442447,
+      "grad_norm": 22.39505984949309,
+      "learning_rate": 2e-05,
+      "loss": 0.6264,
+      "step": 13404
+    },
+    {
+      "epoch": 0.31907550223745595,
+      "grad_norm": 5.291407373671033,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 13405
+    },
+    {
+      "epoch": 0.3190993049604875,
+      "grad_norm": 2.7958459706904044,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 13406
+    },
+    {
+      "epoch": 0.319123107683519,
+      "grad_norm": 5.375506025398597,
+      "learning_rate": 2e-05,
+      "loss": 0.5805,
+      "step": 13407
+    },
+    {
+      "epoch": 0.3191469104065505,
+      "grad_norm": 2.9269437619534546,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 13408
+    },
+    {
+      "epoch": 0.319170713129582,
+      "grad_norm": 5.85198881354174,
+      "learning_rate": 2e-05,
+      "loss": 0.3972,
+      "step": 13409
+    },
+    {
+      "epoch": 0.31919451585261355,
+      "grad_norm": 4.346238390017049,
+      "learning_rate": 2e-05,
+      "loss": 0.2726,
+      "step": 13410
+    },
+    {
+      "epoch": 0.31921831857564503,
+      "grad_norm": 4.037089985391876,
+      "learning_rate": 2e-05,
+      "loss": 0.6596,
+      "step": 13411
+    },
+    {
+      "epoch": 0.3192421212986766,
+      "grad_norm": 4.770588442990559,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 13412
+    },
+    {
+      "epoch": 0.31926592402170806,
+      "grad_norm": 1.825909872483825,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 13413
+    },
+    {
+      "epoch": 0.3192897267447396,
+      "grad_norm": 4.512872706066832,
+      "learning_rate": 2e-05,
+      "loss": 0.7658,
+      "step": 13414
+    },
+    {
+      "epoch": 0.31931352946777114,
+      "grad_norm": 2.8454490416111984,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 13415
+    },
+    {
+      "epoch": 0.3193373321908026,
+      "grad_norm": 4.599825087631185,
+      "learning_rate": 2e-05,
+      "loss": 0.8192,
+      "step": 13416
+    },
+    {
+      "epoch": 0.31936113491383417,
+      "grad_norm": 3.922290608501414,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 13417
+    },
+    {
+      "epoch": 0.31938493763686565,
+      "grad_norm": 2.5558146608154177,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 13418
+    },
+    {
+      "epoch": 0.3194087403598972,
+      "grad_norm": 1.8297810539004782,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 13419
+    },
+    {
+      "epoch": 0.3194325430829287,
+      "grad_norm": 9.02736638958771,
+      "learning_rate": 2e-05,
+      "loss": 0.8561,
+      "step": 13420
+    },
+    {
+      "epoch": 0.3194563458059602,
+      "grad_norm": 6.012375522083639,
+      "learning_rate": 2e-05,
+      "loss": 1.5594,
+      "step": 13421
+    },
+    {
+      "epoch": 0.3194801485289917,
+      "grad_norm": 4.342422246202067,
+      "learning_rate": 2e-05,
+      "loss": 0.3204,
+      "step": 13422
+    },
+    {
+      "epoch": 0.31950395125202324,
+      "grad_norm": 4.46705723364696,
+      "learning_rate": 2e-05,
+      "loss": 0.343,
+      "step": 13423
+    },
+    {
+      "epoch": 0.31952775397505473,
+      "grad_norm": 5.47461296854926,
+      "learning_rate": 2e-05,
+      "loss": 0.6289,
+      "step": 13424
+    },
+    {
+      "epoch": 0.31955155669808627,
+      "grad_norm": 0.5455187819579657,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 13425
+    },
+    {
+      "epoch": 0.31957535942111776,
+      "grad_norm": 7.371766663543183,
+      "learning_rate": 2e-05,
+      "loss": 0.962,
+      "step": 13426
+    },
+    {
+      "epoch": 0.3195991621441493,
+      "grad_norm": 1.6735349504392136,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 13427
+    },
+    {
+      "epoch": 0.3196229648671808,
+      "grad_norm": 4.2045200513563525,
+      "learning_rate": 2e-05,
+      "loss": 0.6406,
+      "step": 13428
+    },
+    {
+      "epoch": 0.3196467675902123,
+      "grad_norm": 2.7746431577089994,
+      "learning_rate": 2e-05,
+      "loss": 0.1807,
+      "step": 13429
+    },
+    {
+      "epoch": 0.31967057031324386,
+      "grad_norm": 0.7434744399113178,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 13430
+    },
+    {
+      "epoch": 0.31969437303627535,
+      "grad_norm": 13.187255092759683,
+      "learning_rate": 2e-05,
+      "loss": 2.0509,
+      "step": 13431
+    },
+    {
+      "epoch": 0.3197181757593069,
+      "grad_norm": 7.241463650098168,
+      "learning_rate": 2e-05,
+      "loss": 0.5053,
+      "step": 13432
+    },
+    {
+      "epoch": 0.3197419784823384,
+      "grad_norm": 4.230589352050828,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 13433
+    },
+    {
+      "epoch": 0.3197657812053699,
+      "grad_norm": 0.8971547580616275,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 13434
+    },
+    {
+      "epoch": 0.3197895839284014,
+      "grad_norm": 13.256425334652437,
+      "learning_rate": 2e-05,
+      "loss": 0.8409,
+      "step": 13435
+    },
+    {
+      "epoch": 0.31981338665143294,
+      "grad_norm": 3.053530429391799,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 13436
+    },
+    {
+      "epoch": 0.3198371893744644,
+      "grad_norm": 13.961051390670207,
+      "learning_rate": 2e-05,
+      "loss": 1.4965,
+      "step": 13437
+    },
+    {
+      "epoch": 0.31986099209749597,
+      "grad_norm": 7.100227941191294,
+      "learning_rate": 2e-05,
+      "loss": 1.6548,
+      "step": 13438
+    },
+    {
+      "epoch": 0.31988479482052745,
+      "grad_norm": 5.527560390548198,
+      "learning_rate": 2e-05,
+      "loss": 0.3727,
+      "step": 13439
+    },
+    {
+      "epoch": 0.319908597543559,
+      "grad_norm": 1.5590576503715323,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 13440
+    },
+    {
+      "epoch": 0.3199324002665905,
+      "grad_norm": 13.976137612658102,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 13441
+    },
+    {
+      "epoch": 0.319956202989622,
+      "grad_norm": 4.678009335917177,
+      "learning_rate": 2e-05,
+      "loss": 0.3038,
+      "step": 13442
+    },
+    {
+      "epoch": 0.3199800057126535,
+      "grad_norm": 1.5222610966523904,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 13443
+    },
+    {
+      "epoch": 0.32000380843568504,
+      "grad_norm": 4.937296538479594,
+      "learning_rate": 2e-05,
+      "loss": 0.5301,
+      "step": 13444
+    },
+    {
+      "epoch": 0.32002761115871653,
+      "grad_norm": 3.7816296982724373,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 13445
+    },
+    {
+      "epoch": 0.32005141388174807,
+      "grad_norm": 3.278793129794279,
+      "learning_rate": 2e-05,
+      "loss": 0.1628,
+      "step": 13446
+    },
+    {
+      "epoch": 0.3200752166047796,
+      "grad_norm": 4.212437506603395,
+      "learning_rate": 2e-05,
+      "loss": 0.4825,
+      "step": 13447
+    },
+    {
+      "epoch": 0.3200990193278111,
+      "grad_norm": 20.008694225221234,
+      "learning_rate": 2e-05,
+      "loss": 1.4188,
+      "step": 13448
+    },
+    {
+      "epoch": 0.32012282205084264,
+      "grad_norm": 6.98408187094152,
+      "learning_rate": 2e-05,
+      "loss": 1.4035,
+      "step": 13449
+    },
+    {
+      "epoch": 0.3201466247738741,
+      "grad_norm": 1.852579165942086,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 13450
+    },
+    {
+      "epoch": 0.32017042749690566,
+      "grad_norm": 1.7403997295508016,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 13451
+    },
+    {
+      "epoch": 0.32019423021993715,
+      "grad_norm": 2.483731952447728,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 13452
+    },
+    {
+      "epoch": 0.3202180329429687,
+      "grad_norm": 1.823948326906876,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 13453
+    },
+    {
+      "epoch": 0.3202418356660002,
+      "grad_norm": 1.6977364665883672,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 13454
+    },
+    {
+      "epoch": 0.3202656383890317,
+      "grad_norm": 0.9834169921409953,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 13455
+    },
+    {
+      "epoch": 0.3202894411120632,
+      "grad_norm": 5.982307781266999,
+      "learning_rate": 2e-05,
+      "loss": 1.0795,
+      "step": 13456
+    },
+    {
+      "epoch": 0.32031324383509474,
+      "grad_norm": 1.2158801024250137,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13457
+    },
+    {
+      "epoch": 0.3203370465581262,
+      "grad_norm": 5.8936824226362665,
+      "learning_rate": 2e-05,
+      "loss": 0.1212,
+      "step": 13458
+    },
+    {
+      "epoch": 0.32036084928115777,
+      "grad_norm": 1.475394157503412,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 13459
+    },
+    {
+      "epoch": 0.32038465200418925,
+      "grad_norm": 2.5085830584363906,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 13460
+    },
+    {
+      "epoch": 0.3204084547272208,
+      "grad_norm": 1.4237009222498918,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 13461
+    },
+    {
+      "epoch": 0.32043225745025233,
+      "grad_norm": 0.6564086903447877,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 13462
+    },
+    {
+      "epoch": 0.3204560601732838,
+      "grad_norm": 7.39957212506043,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 13463
+    },
+    {
+      "epoch": 0.32047986289631536,
+      "grad_norm": 1.603594156653946,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 13464
+    },
+    {
+      "epoch": 0.32050366561934684,
+      "grad_norm": 12.920288573666708,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 13465
+    },
+    {
+      "epoch": 0.3205274683423784,
+      "grad_norm": 2.673268970374846,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 13466
+    },
+    {
+      "epoch": 0.32055127106540987,
+      "grad_norm": 0.7233912956245278,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 13467
+    },
+    {
+      "epoch": 0.3205750737884414,
+      "grad_norm": 4.322872968883359,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 13468
+    },
+    {
+      "epoch": 0.3205988765114729,
+      "grad_norm": 3.8007287086026453,
+      "learning_rate": 2e-05,
+      "loss": 0.5228,
+      "step": 13469
+    },
+    {
+      "epoch": 0.32062267923450444,
+      "grad_norm": 7.194964645041351,
+      "learning_rate": 2e-05,
+      "loss": 1.1665,
+      "step": 13470
+    },
+    {
+      "epoch": 0.3206464819575359,
+      "grad_norm": 4.316171600703451,
+      "learning_rate": 2e-05,
+      "loss": 0.5306,
+      "step": 13471
+    },
+    {
+      "epoch": 0.32067028468056746,
+      "grad_norm": 10.724565404261197,
+      "learning_rate": 2e-05,
+      "loss": 0.4075,
+      "step": 13472
+    },
+    {
+      "epoch": 0.32069408740359895,
+      "grad_norm": 3.9660564954160917,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 13473
+    },
+    {
+      "epoch": 0.3207178901266305,
+      "grad_norm": 3.3779869955963706,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 13474
+    },
+    {
+      "epoch": 0.320741692849662,
+      "grad_norm": 3.6407637582449137,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 13475
+    },
+    {
+      "epoch": 0.3207654955726935,
+      "grad_norm": 4.990979747459595,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 13476
+    },
+    {
+      "epoch": 0.32078929829572506,
+      "grad_norm": 4.702852158989141,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 13477
+    },
+    {
+      "epoch": 0.32081310101875654,
+      "grad_norm": 2.6918627620102233,
+      "learning_rate": 2e-05,
+      "loss": 0.1161,
+      "step": 13478
+    },
+    {
+      "epoch": 0.3208369037417881,
+      "grad_norm": 7.822245771792719,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 13479
+    },
+    {
+      "epoch": 0.32086070646481957,
+      "grad_norm": 0.1543887918345261,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 13480
+    },
+    {
+      "epoch": 0.3208845091878511,
+      "grad_norm": 5.072270495671792,
+      "learning_rate": 2e-05,
+      "loss": 0.872,
+      "step": 13481
+    },
+    {
+      "epoch": 0.3209083119108826,
+      "grad_norm": 13.699043136818565,
+      "learning_rate": 2e-05,
+      "loss": 1.5562,
+      "step": 13482
+    },
+    {
+      "epoch": 0.32093211463391413,
+      "grad_norm": 3.6956258718151336,
+      "learning_rate": 2e-05,
+      "loss": 0.1208,
+      "step": 13483
+    },
+    {
+      "epoch": 0.3209559173569456,
+      "grad_norm": 13.179796479693971,
+      "learning_rate": 2e-05,
+      "loss": 0.6524,
+      "step": 13484
+    },
+    {
+      "epoch": 0.32097972007997716,
+      "grad_norm": 3.663685038905925,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 13485
+    },
+    {
+      "epoch": 0.32100352280300865,
+      "grad_norm": 1.9746962746210113,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 13486
+    },
+    {
+      "epoch": 0.3210273255260402,
+      "grad_norm": 4.544578608820124,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 13487
+    },
+    {
+      "epoch": 0.32105112824907167,
+      "grad_norm": 4.702341900517404,
+      "learning_rate": 2e-05,
+      "loss": 0.7083,
+      "step": 13488
+    },
+    {
+      "epoch": 0.3210749309721032,
+      "grad_norm": 2.2506835838192387,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 13489
+    },
+    {
+      "epoch": 0.3210987336951347,
+      "grad_norm": 10.48948811834581,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 13490
+    },
+    {
+      "epoch": 0.32112253641816624,
+      "grad_norm": 5.216885527112166,
+      "learning_rate": 2e-05,
+      "loss": 0.8064,
+      "step": 13491
+    },
+    {
+      "epoch": 0.3211463391411978,
+      "grad_norm": 9.224178311368942,
+      "learning_rate": 2e-05,
+      "loss": 0.7087,
+      "step": 13492
+    },
+    {
+      "epoch": 0.32117014186422926,
+      "grad_norm": 11.803832507818951,
+      "learning_rate": 2e-05,
+      "loss": 1.0443,
+      "step": 13493
+    },
+    {
+      "epoch": 0.3211939445872608,
+      "grad_norm": 6.429932947753894,
+      "learning_rate": 2e-05,
+      "loss": 1.0862,
+      "step": 13494
+    },
+    {
+      "epoch": 0.3212177473102923,
+      "grad_norm": 1.2499130924177728,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 13495
+    },
+    {
+      "epoch": 0.32124155003332383,
+      "grad_norm": 0.6125912279459025,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 13496
+    },
+    {
+      "epoch": 0.3212653527563553,
+      "grad_norm": 5.224491312875483,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 13497
+    },
+    {
+      "epoch": 0.32128915547938686,
+      "grad_norm": 0.6258203205929577,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 13498
+    },
+    {
+      "epoch": 0.32131295820241834,
+      "grad_norm": 4.317972462636004,
+      "learning_rate": 2e-05,
+      "loss": 0.3942,
+      "step": 13499
+    },
+    {
+      "epoch": 0.3213367609254499,
+      "grad_norm": 6.04901923078583,
+      "learning_rate": 2e-05,
+      "loss": 1.047,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32136056364848137,
+      "grad_norm": 2.446623465781362,
+      "learning_rate": 2e-05,
+      "loss": 0.1658,
+      "step": 13501
+    },
+    {
+      "epoch": 0.3213843663715129,
+      "grad_norm": 5.5219489038694585,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 13502
+    },
+    {
+      "epoch": 0.3214081690945444,
+      "grad_norm": 5.500130757914944,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 13503
+    },
+    {
+      "epoch": 0.32143197181757593,
+      "grad_norm": 17.848529153077795,
+      "learning_rate": 2e-05,
+      "loss": 0.3058,
+      "step": 13504
+    },
+    {
+      "epoch": 0.3214557745406074,
+      "grad_norm": 0.3964776104621829,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 13505
+    },
+    {
+      "epoch": 0.32147957726363896,
+      "grad_norm": 4.777825065015135,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 13506
+    },
+    {
+      "epoch": 0.3215033799866705,
+      "grad_norm": 32.33149387437573,
+      "learning_rate": 2e-05,
+      "loss": 0.7433,
+      "step": 13507
+    },
+    {
+      "epoch": 0.321527182709702,
+      "grad_norm": 2.046656960807398,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 13508
+    },
+    {
+      "epoch": 0.3215509854327335,
+      "grad_norm": 2.8924213501821474,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 13509
+    },
+    {
+      "epoch": 0.321574788155765,
+      "grad_norm": 5.041513674591953,
+      "learning_rate": 2e-05,
+      "loss": 0.1823,
+      "step": 13510
+    },
+    {
+      "epoch": 0.32159859087879655,
+      "grad_norm": 2.7118732466239184,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 13511
+    },
+    {
+      "epoch": 0.32162239360182804,
+      "grad_norm": 3.726095285202032,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 13512
+    },
+    {
+      "epoch": 0.3216461963248596,
+      "grad_norm": 3.113977213635995,
+      "learning_rate": 2e-05,
+      "loss": 0.474,
+      "step": 13513
+    },
+    {
+      "epoch": 0.32166999904789106,
+      "grad_norm": 10.792821721391482,
+      "learning_rate": 2e-05,
+      "loss": 0.7696,
+      "step": 13514
+    },
+    {
+      "epoch": 0.3216938017709226,
+      "grad_norm": 4.720219995285387,
+      "learning_rate": 2e-05,
+      "loss": 0.3306,
+      "step": 13515
+    },
+    {
+      "epoch": 0.3217176044939541,
+      "grad_norm": 3.732140582044,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 13516
+    },
+    {
+      "epoch": 0.32174140721698563,
+      "grad_norm": 4.162515697508844,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 13517
+    },
+    {
+      "epoch": 0.3217652099400171,
+      "grad_norm": 6.311758069603749,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 13518
+    },
+    {
+      "epoch": 0.32178901266304866,
+      "grad_norm": 5.62831785940046,
+      "learning_rate": 2e-05,
+      "loss": 0.7825,
+      "step": 13519
+    },
+    {
+      "epoch": 0.32181281538608014,
+      "grad_norm": 10.254963415837503,
+      "learning_rate": 2e-05,
+      "loss": 0.4642,
+      "step": 13520
+    },
+    {
+      "epoch": 0.3218366181091117,
+      "grad_norm": 4.021121153156466,
+      "learning_rate": 2e-05,
+      "loss": 0.1738,
+      "step": 13521
+    },
+    {
+      "epoch": 0.3218604208321432,
+      "grad_norm": 8.433593932321202,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 13522
+    },
+    {
+      "epoch": 0.3218842235551747,
+      "grad_norm": 3.287823605447273,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 13523
+    },
+    {
+      "epoch": 0.32190802627820625,
+      "grad_norm": 5.204312750876175,
+      "learning_rate": 2e-05,
+      "loss": 1.0064,
+      "step": 13524
+    },
+    {
+      "epoch": 0.32193182900123773,
+      "grad_norm": 2.907294371254995,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 13525
+    },
+    {
+      "epoch": 0.3219556317242693,
+      "grad_norm": 7.343653159550438,
+      "learning_rate": 2e-05,
+      "loss": 1.3595,
+      "step": 13526
+    },
+    {
+      "epoch": 0.32197943444730076,
+      "grad_norm": 1.9395972615455348,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 13527
+    },
+    {
+      "epoch": 0.3220032371703323,
+      "grad_norm": 4.418207666085916,
+      "learning_rate": 2e-05,
+      "loss": 0.488,
+      "step": 13528
+    },
+    {
+      "epoch": 0.3220270398933638,
+      "grad_norm": 1.8393133313607704,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 13529
+    },
+    {
+      "epoch": 0.3220508426163953,
+      "grad_norm": 0.3536536315305732,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 13530
+    },
+    {
+      "epoch": 0.3220746453394268,
+      "grad_norm": 2.8747140832714164,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 13531
+    },
+    {
+      "epoch": 0.32209844806245835,
+      "grad_norm": 4.164673561675422,
+      "learning_rate": 2e-05,
+      "loss": 0.9633,
+      "step": 13532
+    },
+    {
+      "epoch": 0.32212225078548984,
+      "grad_norm": 4.865954063867311,
+      "learning_rate": 2e-05,
+      "loss": 0.659,
+      "step": 13533
+    },
+    {
+      "epoch": 0.3221460535085214,
+      "grad_norm": 4.588082687094081,
+      "learning_rate": 2e-05,
+      "loss": 1.0762,
+      "step": 13534
+    },
+    {
+      "epoch": 0.32216985623155286,
+      "grad_norm": 1.9359440847427023,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 13535
+    },
+    {
+      "epoch": 0.3221936589545844,
+      "grad_norm": 5.756774744469465,
+      "learning_rate": 2e-05,
+      "loss": 0.628,
+      "step": 13536
+    },
+    {
+      "epoch": 0.32221746167761595,
+      "grad_norm": 6.123052683212171,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 13537
+    },
+    {
+      "epoch": 0.32224126440064743,
+      "grad_norm": 31.809902000240395,
+      "learning_rate": 2e-05,
+      "loss": 1.3792,
+      "step": 13538
+    },
+    {
+      "epoch": 0.32226506712367897,
+      "grad_norm": 4.555788758212927,
+      "learning_rate": 2e-05,
+      "loss": 0.4178,
+      "step": 13539
+    },
+    {
+      "epoch": 0.32228886984671046,
+      "grad_norm": 4.697974010218923,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 13540
+    },
+    {
+      "epoch": 0.322312672569742,
+      "grad_norm": 14.8504253883851,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 13541
+    },
+    {
+      "epoch": 0.3223364752927735,
+      "grad_norm": 3.265291436922816,
+      "learning_rate": 2e-05,
+      "loss": 0.1131,
+      "step": 13542
+    },
+    {
+      "epoch": 0.322360278015805,
+      "grad_norm": 0.6935359357211036,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 13543
+    },
+    {
+      "epoch": 0.3223840807388365,
+      "grad_norm": 1.0036470420536054,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 13544
+    },
+    {
+      "epoch": 0.32240788346186805,
+      "grad_norm": 10.973084699679246,
+      "learning_rate": 2e-05,
+      "loss": 0.6658,
+      "step": 13545
+    },
+    {
+      "epoch": 0.32243168618489954,
+      "grad_norm": 24.069880833479075,
+      "learning_rate": 2e-05,
+      "loss": 1.5983,
+      "step": 13546
+    },
+    {
+      "epoch": 0.3224554889079311,
+      "grad_norm": 2.6197702209814535,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 13547
+    },
+    {
+      "epoch": 0.32247929163096256,
+      "grad_norm": 4.347458713895609,
+      "learning_rate": 2e-05,
+      "loss": 0.6943,
+      "step": 13548
+    },
+    {
+      "epoch": 0.3225030943539941,
+      "grad_norm": 1.6407958523043205,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 13549
+    },
+    {
+      "epoch": 0.3225268970770256,
+      "grad_norm": 1.852049676223217,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 13550
+    },
+    {
+      "epoch": 0.32255069980005713,
+      "grad_norm": 6.513340140373785,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 13551
+    },
+    {
+      "epoch": 0.32257450252308867,
+      "grad_norm": 0.6850582834756782,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 13552
+    },
+    {
+      "epoch": 0.32259830524612015,
+      "grad_norm": 4.101820654963145,
+      "learning_rate": 2e-05,
+      "loss": 0.7006,
+      "step": 13553
+    },
+    {
+      "epoch": 0.3226221079691517,
+      "grad_norm": 3.98879830645695,
+      "learning_rate": 2e-05,
+      "loss": 0.2521,
+      "step": 13554
+    },
+    {
+      "epoch": 0.3226459106921832,
+      "grad_norm": 1.9304465060682232,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 13555
+    },
+    {
+      "epoch": 0.3226697134152147,
+      "grad_norm": 0.6523761935222377,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13556
+    },
+    {
+      "epoch": 0.3226935161382462,
+      "grad_norm": 5.733230816831119,
+      "learning_rate": 2e-05,
+      "loss": 0.3159,
+      "step": 13557
+    },
+    {
+      "epoch": 0.32271731886127775,
+      "grad_norm": 7.64468671702852,
+      "learning_rate": 2e-05,
+      "loss": 0.6539,
+      "step": 13558
+    },
+    {
+      "epoch": 0.32274112158430923,
+      "grad_norm": 3.826748593910873,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 13559
+    },
+    {
+      "epoch": 0.3227649243073408,
+      "grad_norm": 3.248046677766563,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 13560
+    },
+    {
+      "epoch": 0.32278872703037226,
+      "grad_norm": 6.343878209311849,
+      "learning_rate": 2e-05,
+      "loss": 1.2506,
+      "step": 13561
+    },
+    {
+      "epoch": 0.3228125297534038,
+      "grad_norm": 4.421390574316291,
+      "learning_rate": 2e-05,
+      "loss": 0.84,
+      "step": 13562
+    },
+    {
+      "epoch": 0.3228363324764353,
+      "grad_norm": 4.867207665940956,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 13563
+    },
+    {
+      "epoch": 0.3228601351994668,
+      "grad_norm": 4.2875146025891775,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 13564
+    },
+    {
+      "epoch": 0.3228839379224983,
+      "grad_norm": 6.974415345533848,
+      "learning_rate": 2e-05,
+      "loss": 1.057,
+      "step": 13565
+    },
+    {
+      "epoch": 0.32290774064552985,
+      "grad_norm": 4.100723579953433,
+      "learning_rate": 2e-05,
+      "loss": 0.7917,
+      "step": 13566
+    },
+    {
+      "epoch": 0.32293154336856134,
+      "grad_norm": 2.3666902928122147,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 13567
+    },
+    {
+      "epoch": 0.3229553460915929,
+      "grad_norm": 4.213496562553561,
+      "learning_rate": 2e-05,
+      "loss": 0.1986,
+      "step": 13568
+    },
+    {
+      "epoch": 0.3229791488146244,
+      "grad_norm": 5.183265854818109,
+      "learning_rate": 2e-05,
+      "loss": 0.2194,
+      "step": 13569
+    },
+    {
+      "epoch": 0.3230029515376559,
+      "grad_norm": 4.136179061123561,
+      "learning_rate": 2e-05,
+      "loss": 0.7391,
+      "step": 13570
+    },
+    {
+      "epoch": 0.32302675426068744,
+      "grad_norm": 12.458203996029056,
+      "learning_rate": 2e-05,
+      "loss": 0.938,
+      "step": 13571
+    },
+    {
+      "epoch": 0.32305055698371893,
+      "grad_norm": 2.0148432789697504,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 13572
+    },
+    {
+      "epoch": 0.32307435970675047,
+      "grad_norm": 1.5191701518833665,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 13573
+    },
+    {
+      "epoch": 0.32309816242978195,
+      "grad_norm": 5.89039133127604,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 13574
+    },
+    {
+      "epoch": 0.3231219651528135,
+      "grad_norm": 2.0372277260334632,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 13575
+    },
+    {
+      "epoch": 0.323145767875845,
+      "grad_norm": 1.0705767265785566,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 13576
+    },
+    {
+      "epoch": 0.3231695705988765,
+      "grad_norm": 0.07836155318191848,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 13577
+    },
+    {
+      "epoch": 0.323193373321908,
+      "grad_norm": 1.0546703273656668,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 13578
+    },
+    {
+      "epoch": 0.32321717604493955,
+      "grad_norm": 5.890997101037896,
+      "learning_rate": 2e-05,
+      "loss": 0.6886,
+      "step": 13579
+    },
+    {
+      "epoch": 0.32324097876797103,
+      "grad_norm": 4.25010309711099,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 13580
+    },
+    {
+      "epoch": 0.3232647814910026,
+      "grad_norm": 9.120543062543687,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 13581
+    },
+    {
+      "epoch": 0.32328858421403406,
+      "grad_norm": 6.6826277886047745,
+      "learning_rate": 2e-05,
+      "loss": 1.1866,
+      "step": 13582
+    },
+    {
+      "epoch": 0.3233123869370656,
+      "grad_norm": 5.794695734328299,
+      "learning_rate": 2e-05,
+      "loss": 0.3593,
+      "step": 13583
+    },
+    {
+      "epoch": 0.32333618966009714,
+      "grad_norm": 10.646067847118251,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 13584
+    },
+    {
+      "epoch": 0.3233599923831286,
+      "grad_norm": 9.007163517704347,
+      "learning_rate": 2e-05,
+      "loss": 0.5228,
+      "step": 13585
+    },
+    {
+      "epoch": 0.32338379510616017,
+      "grad_norm": 9.217374305781698,
+      "learning_rate": 2e-05,
+      "loss": 1.1143,
+      "step": 13586
+    },
+    {
+      "epoch": 0.32340759782919165,
+      "grad_norm": 14.720298695980961,
+      "learning_rate": 2e-05,
+      "loss": 0.257,
+      "step": 13587
+    },
+    {
+      "epoch": 0.3234314005522232,
+      "grad_norm": 8.362537094884942,
+      "learning_rate": 2e-05,
+      "loss": 1.4035,
+      "step": 13588
+    },
+    {
+      "epoch": 0.3234552032752547,
+      "grad_norm": 3.502719400331593,
+      "learning_rate": 2e-05,
+      "loss": 0.3989,
+      "step": 13589
+    },
+    {
+      "epoch": 0.3234790059982862,
+      "grad_norm": 2.0897278079696195,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 13590
+    },
+    {
+      "epoch": 0.3235028087213177,
+      "grad_norm": 6.939695492813038,
+      "learning_rate": 2e-05,
+      "loss": 0.4554,
+      "step": 13591
+    },
+    {
+      "epoch": 0.32352661144434924,
+      "grad_norm": 2.3327556963147416,
+      "learning_rate": 2e-05,
+      "loss": 0.2369,
+      "step": 13592
+    },
+    {
+      "epoch": 0.32355041416738073,
+      "grad_norm": 0.9183033788658903,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13593
+    },
+    {
+      "epoch": 0.32357421689041227,
+      "grad_norm": 4.5950298736769675,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 13594
+    },
+    {
+      "epoch": 0.32359801961344375,
+      "grad_norm": 1.1512507350542638,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 13595
+    },
+    {
+      "epoch": 0.3236218223364753,
+      "grad_norm": 3.7710313403794045,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 13596
+    },
+    {
+      "epoch": 0.3236456250595068,
+      "grad_norm": 2.0109190634462846,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 13597
+    },
+    {
+      "epoch": 0.3236694277825383,
+      "grad_norm": 0.887176261622789,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 13598
+    },
+    {
+      "epoch": 0.32369323050556986,
+      "grad_norm": 4.422774308473023,
+      "learning_rate": 2e-05,
+      "loss": 0.6933,
+      "step": 13599
+    },
+    {
+      "epoch": 0.32371703322860135,
+      "grad_norm": 1.022951938443775,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 13600
+    },
+    {
+      "epoch": 0.3237408359516329,
+      "grad_norm": 2.690546899613649,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 13601
+    },
+    {
+      "epoch": 0.3237646386746644,
+      "grad_norm": 3.648048899694342,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 13602
+    },
+    {
+      "epoch": 0.3237884413976959,
+      "grad_norm": 3.356706985518384,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 13603
+    },
+    {
+      "epoch": 0.3238122441207274,
+      "grad_norm": 5.41751763524409,
+      "learning_rate": 2e-05,
+      "loss": 1.0758,
+      "step": 13604
+    },
+    {
+      "epoch": 0.32383604684375894,
+      "grad_norm": 3.470311117837646,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 13605
+    },
+    {
+      "epoch": 0.3238598495667904,
+      "grad_norm": 2.802384472515252,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 13606
+    },
+    {
+      "epoch": 0.32388365228982197,
+      "grad_norm": 4.404968239631842,
+      "learning_rate": 2e-05,
+      "loss": 0.7635,
+      "step": 13607
+    },
+    {
+      "epoch": 0.32390745501285345,
+      "grad_norm": 2.2911030780645776,
+      "learning_rate": 2e-05,
+      "loss": 0.1591,
+      "step": 13608
+    },
+    {
+      "epoch": 0.323931257735885,
+      "grad_norm": 2.372631221715057,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 13609
+    },
+    {
+      "epoch": 0.3239550604589165,
+      "grad_norm": 4.963124651736331,
+      "learning_rate": 2e-05,
+      "loss": 0.8145,
+      "step": 13610
+    },
+    {
+      "epoch": 0.323978863181948,
+      "grad_norm": 6.716931857593223,
+      "learning_rate": 2e-05,
+      "loss": 0.9918,
+      "step": 13611
+    },
+    {
+      "epoch": 0.3240026659049795,
+      "grad_norm": 4.756841874143909,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 13612
+    },
+    {
+      "epoch": 0.32402646862801104,
+      "grad_norm": 6.276197260752032,
+      "learning_rate": 2e-05,
+      "loss": 1.0033,
+      "step": 13613
+    },
+    {
+      "epoch": 0.3240502713510426,
+      "grad_norm": 4.7206585548281845,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 13614
+    },
+    {
+      "epoch": 0.32407407407407407,
+      "grad_norm": 15.097578146271113,
+      "learning_rate": 2e-05,
+      "loss": 0.8469,
+      "step": 13615
+    },
+    {
+      "epoch": 0.3240978767971056,
+      "grad_norm": 4.550831457609298,
+      "learning_rate": 2e-05,
+      "loss": 0.4976,
+      "step": 13616
+    },
+    {
+      "epoch": 0.3241216795201371,
+      "grad_norm": 4.9266453516335025,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 13617
+    },
+    {
+      "epoch": 0.32414548224316864,
+      "grad_norm": 0.4135981305169928,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 13618
+    },
+    {
+      "epoch": 0.3241692849662001,
+      "grad_norm": 6.254021122320391,
+      "learning_rate": 2e-05,
+      "loss": 1.3994,
+      "step": 13619
+    },
+    {
+      "epoch": 0.32419308768923166,
+      "grad_norm": 2.6168517980640096,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 13620
+    },
+    {
+      "epoch": 0.32421689041226315,
+      "grad_norm": 6.371629206201315,
+      "learning_rate": 2e-05,
+      "loss": 0.687,
+      "step": 13621
+    },
+    {
+      "epoch": 0.3242406931352947,
+      "grad_norm": 3.3296458139069607,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 13622
+    },
+    {
+      "epoch": 0.3242644958583262,
+      "grad_norm": 10.992247813298995,
+      "learning_rate": 2e-05,
+      "loss": 0.6015,
+      "step": 13623
+    },
+    {
+      "epoch": 0.3242882985813577,
+      "grad_norm": 3.345740386554555,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 13624
+    },
+    {
+      "epoch": 0.3243121013043892,
+      "grad_norm": 11.409601142504561,
+      "learning_rate": 2e-05,
+      "loss": 1.2028,
+      "step": 13625
+    },
+    {
+      "epoch": 0.32433590402742074,
+      "grad_norm": 4.951390385701354,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 13626
+    },
+    {
+      "epoch": 0.3243597067504522,
+      "grad_norm": 5.939908510644139,
+      "learning_rate": 2e-05,
+      "loss": 1.0442,
+      "step": 13627
+    },
+    {
+      "epoch": 0.32438350947348377,
+      "grad_norm": 0.9063267021362263,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 13628
+    },
+    {
+      "epoch": 0.3244073121965153,
+      "grad_norm": 14.690638159119308,
+      "learning_rate": 2e-05,
+      "loss": 1.0303,
+      "step": 13629
+    },
+    {
+      "epoch": 0.3244311149195468,
+      "grad_norm": 63.44477048277427,
+      "learning_rate": 2e-05,
+      "loss": 0.9391,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32445491764257833,
+      "grad_norm": 9.889946610855597,
+      "learning_rate": 2e-05,
+      "loss": 1.2289,
+      "step": 13631
+    },
+    {
+      "epoch": 0.3244787203656098,
+      "grad_norm": 4.678896659199068,
+      "learning_rate": 2e-05,
+      "loss": 0.1962,
+      "step": 13632
+    },
+    {
+      "epoch": 0.32450252308864136,
+      "grad_norm": 0.9625287682199212,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 13633
+    },
+    {
+      "epoch": 0.32452632581167284,
+      "grad_norm": 3.4235607932151697,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 13634
+    },
+    {
+      "epoch": 0.3245501285347044,
+      "grad_norm": 3.6931372816572803,
+      "learning_rate": 2e-05,
+      "loss": 0.2461,
+      "step": 13635
+    },
+    {
+      "epoch": 0.32457393125773587,
+      "grad_norm": 3.406523958328308,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 13636
+    },
+    {
+      "epoch": 0.3245977339807674,
+      "grad_norm": 4.158056754888025,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 13637
+    },
+    {
+      "epoch": 0.3246215367037989,
+      "grad_norm": 3.988951907129686,
+      "learning_rate": 2e-05,
+      "loss": 0.9509,
+      "step": 13638
+    },
+    {
+      "epoch": 0.32464533942683044,
+      "grad_norm": 12.007341606877056,
+      "learning_rate": 2e-05,
+      "loss": 0.143,
+      "step": 13639
+    },
+    {
+      "epoch": 0.3246691421498619,
+      "grad_norm": 10.019428611205125,
+      "learning_rate": 2e-05,
+      "loss": 0.999,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32469294487289346,
+      "grad_norm": 1.598043266450748,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 13641
+    },
+    {
+      "epoch": 0.32471674759592495,
+      "grad_norm": 2.2112436989523188,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 13642
+    },
+    {
+      "epoch": 0.3247405503189565,
+      "grad_norm": 8.47182003122823,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 13643
+    },
+    {
+      "epoch": 0.32476435304198803,
+      "grad_norm": 7.673744435188709,
+      "learning_rate": 2e-05,
+      "loss": 0.4923,
+      "step": 13644
+    },
+    {
+      "epoch": 0.3247881557650195,
+      "grad_norm": 2.505915891009841,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 13645
+    },
+    {
+      "epoch": 0.32481195848805106,
+      "grad_norm": 0.7794873249095492,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 13646
+    },
+    {
+      "epoch": 0.32483576121108254,
+      "grad_norm": 5.484305190356816,
+      "learning_rate": 2e-05,
+      "loss": 1.7212,
+      "step": 13647
+    },
+    {
+      "epoch": 0.3248595639341141,
+      "grad_norm": 7.7494658232667515,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 13648
+    },
+    {
+      "epoch": 0.32488336665714557,
+      "grad_norm": 4.192669003266132,
+      "learning_rate": 2e-05,
+      "loss": 0.7155,
+      "step": 13649
+    },
+    {
+      "epoch": 0.3249071693801771,
+      "grad_norm": 1.3878475086368545,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 13650
+    },
+    {
+      "epoch": 0.3249309721032086,
+      "grad_norm": 0.653374766507204,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 13651
+    },
+    {
+      "epoch": 0.32495477482624013,
+      "grad_norm": 4.0751559446084045,
+      "learning_rate": 2e-05,
+      "loss": 0.1311,
+      "step": 13652
+    },
+    {
+      "epoch": 0.3249785775492716,
+      "grad_norm": 3.174761118271914,
+      "learning_rate": 2e-05,
+      "loss": 0.6546,
+      "step": 13653
+    },
+    {
+      "epoch": 0.32500238027230316,
+      "grad_norm": 3.9621599963114686,
+      "learning_rate": 2e-05,
+      "loss": 0.5815,
+      "step": 13654
+    },
+    {
+      "epoch": 0.32502618299533464,
+      "grad_norm": 6.214679400404341,
+      "learning_rate": 2e-05,
+      "loss": 1.5042,
+      "step": 13655
+    },
+    {
+      "epoch": 0.3250499857183662,
+      "grad_norm": 3.9328224748555387,
+      "learning_rate": 2e-05,
+      "loss": 0.4561,
+      "step": 13656
+    },
+    {
+      "epoch": 0.32507378844139767,
+      "grad_norm": 1.6212149149704613,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 13657
+    },
+    {
+      "epoch": 0.3250975911644292,
+      "grad_norm": 2.415288120735385,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 13658
+    },
+    {
+      "epoch": 0.32512139388746075,
+      "grad_norm": 4.765105655830338,
+      "learning_rate": 2e-05,
+      "loss": 0.9139,
+      "step": 13659
+    },
+    {
+      "epoch": 0.32514519661049224,
+      "grad_norm": 3.3882707052816836,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 13660
+    },
+    {
+      "epoch": 0.3251689993335238,
+      "grad_norm": 5.622506049820949,
+      "learning_rate": 2e-05,
+      "loss": 1.0883,
+      "step": 13661
+    },
+    {
+      "epoch": 0.32519280205655526,
+      "grad_norm": 11.550637281016726,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 13662
+    },
+    {
+      "epoch": 0.3252166047795868,
+      "grad_norm": 2.3699755120241788,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 13663
+    },
+    {
+      "epoch": 0.3252404075026183,
+      "grad_norm": 1.910734676082009,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 13664
+    },
+    {
+      "epoch": 0.32526421022564983,
+      "grad_norm": 5.520518846959367,
+      "learning_rate": 2e-05,
+      "loss": 1.3104,
+      "step": 13665
+    },
+    {
+      "epoch": 0.3252880129486813,
+      "grad_norm": 2.1113478081314123,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 13666
+    },
+    {
+      "epoch": 0.32531181567171286,
+      "grad_norm": 3.687759994066413,
+      "learning_rate": 2e-05,
+      "loss": 0.1911,
+      "step": 13667
+    },
+    {
+      "epoch": 0.32533561839474434,
+      "grad_norm": 9.735198298069196,
+      "learning_rate": 2e-05,
+      "loss": 0.5014,
+      "step": 13668
+    },
+    {
+      "epoch": 0.3253594211177759,
+      "grad_norm": 0.7844688324179213,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 13669
+    },
+    {
+      "epoch": 0.32538322384080737,
+      "grad_norm": 2.2634560847991425,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 13670
+    },
+    {
+      "epoch": 0.3254070265638389,
+      "grad_norm": 3.8468095190674467,
+      "learning_rate": 2e-05,
+      "loss": 0.7937,
+      "step": 13671
+    },
+    {
+      "epoch": 0.3254308292868704,
+      "grad_norm": 2.320884858671267,
+      "learning_rate": 2e-05,
+      "loss": 0.1436,
+      "step": 13672
+    },
+    {
+      "epoch": 0.32545463200990193,
+      "grad_norm": 6.740052542597856,
+      "learning_rate": 2e-05,
+      "loss": 1.1619,
+      "step": 13673
+    },
+    {
+      "epoch": 0.3254784347329335,
+      "grad_norm": 8.88155736215412,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 13674
+    },
+    {
+      "epoch": 0.32550223745596496,
+      "grad_norm": 3.5607409526052742,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 13675
+    },
+    {
+      "epoch": 0.3255260401789965,
+      "grad_norm": 7.843025846161417,
+      "learning_rate": 2e-05,
+      "loss": 0.6553,
+      "step": 13676
+    },
+    {
+      "epoch": 0.325549842902028,
+      "grad_norm": 5.184610489583109,
+      "learning_rate": 2e-05,
+      "loss": 0.9624,
+      "step": 13677
+    },
+    {
+      "epoch": 0.3255736456250595,
+      "grad_norm": 0.05636040122342072,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 13678
+    },
+    {
+      "epoch": 0.325597448348091,
+      "grad_norm": 12.635063415121822,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 13679
+    },
+    {
+      "epoch": 0.32562125107112255,
+      "grad_norm": 4.121122451176246,
+      "learning_rate": 2e-05,
+      "loss": 0.7477,
+      "step": 13680
+    },
+    {
+      "epoch": 0.32564505379415404,
+      "grad_norm": 4.1318910977245835,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 13681
+    },
+    {
+      "epoch": 0.3256688565171856,
+      "grad_norm": 4.548712046748503,
+      "learning_rate": 2e-05,
+      "loss": 0.3322,
+      "step": 13682
+    },
+    {
+      "epoch": 0.32569265924021706,
+      "grad_norm": 5.671497650458658,
+      "learning_rate": 2e-05,
+      "loss": 1.2889,
+      "step": 13683
+    },
+    {
+      "epoch": 0.3257164619632486,
+      "grad_norm": 6.828224980310579,
+      "learning_rate": 2e-05,
+      "loss": 0.8707,
+      "step": 13684
+    },
+    {
+      "epoch": 0.3257402646862801,
+      "grad_norm": 4.579125178992738,
+      "learning_rate": 2e-05,
+      "loss": 0.7572,
+      "step": 13685
+    },
+    {
+      "epoch": 0.32576406740931163,
+      "grad_norm": 2.6294313919830232,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 13686
+    },
+    {
+      "epoch": 0.3257878701323431,
+      "grad_norm": 11.229639736571015,
+      "learning_rate": 2e-05,
+      "loss": 0.5858,
+      "step": 13687
+    },
+    {
+      "epoch": 0.32581167285537466,
+      "grad_norm": 1.7221946115522213,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 13688
+    },
+    {
+      "epoch": 0.3258354755784062,
+      "grad_norm": 5.817604558777632,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 13689
+    },
+    {
+      "epoch": 0.3258592783014377,
+      "grad_norm": 6.3944348626247605,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 13690
+    },
+    {
+      "epoch": 0.3258830810244692,
+      "grad_norm": 7.31786838058461,
+      "learning_rate": 2e-05,
+      "loss": 0.7675,
+      "step": 13691
+    },
+    {
+      "epoch": 0.3259068837475007,
+      "grad_norm": 7.015094534994144,
+      "learning_rate": 2e-05,
+      "loss": 0.5248,
+      "step": 13692
+    },
+    {
+      "epoch": 0.32593068647053225,
+      "grad_norm": 2.5879353649211696,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 13693
+    },
+    {
+      "epoch": 0.32595448919356373,
+      "grad_norm": 5.553167794324742,
+      "learning_rate": 2e-05,
+      "loss": 0.6756,
+      "step": 13694
+    },
+    {
+      "epoch": 0.3259782919165953,
+      "grad_norm": 4.549705390334146,
+      "learning_rate": 2e-05,
+      "loss": 0.856,
+      "step": 13695
+    },
+    {
+      "epoch": 0.32600209463962676,
+      "grad_norm": 5.217619160377788,
+      "learning_rate": 2e-05,
+      "loss": 0.9616,
+      "step": 13696
+    },
+    {
+      "epoch": 0.3260258973626583,
+      "grad_norm": 8.132044871400534,
+      "learning_rate": 2e-05,
+      "loss": 0.7401,
+      "step": 13697
+    },
+    {
+      "epoch": 0.3260497000856898,
+      "grad_norm": 5.297048354687139,
+      "learning_rate": 2e-05,
+      "loss": 0.4435,
+      "step": 13698
+    },
+    {
+      "epoch": 0.3260735028087213,
+      "grad_norm": 3.76438195920449,
+      "learning_rate": 2e-05,
+      "loss": 0.5071,
+      "step": 13699
+    },
+    {
+      "epoch": 0.3260973055317528,
+      "grad_norm": 1.893093213128705,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 13700
+    },
+    {
+      "epoch": 0.32612110825478435,
+      "grad_norm": 3.265509000937768,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 13701
+    },
+    {
+      "epoch": 0.32614491097781584,
+      "grad_norm": 4.552949747263639,
+      "learning_rate": 2e-05,
+      "loss": 0.1272,
+      "step": 13702
+    },
+    {
+      "epoch": 0.3261687137008474,
+      "grad_norm": 3.8173090075537996,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 13703
+    },
+    {
+      "epoch": 0.32619251642387886,
+      "grad_norm": 1.9117565051838064,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 13704
+    },
+    {
+      "epoch": 0.3262163191469104,
+      "grad_norm": 5.025573840515151,
+      "learning_rate": 2e-05,
+      "loss": 1.074,
+      "step": 13705
+    },
+    {
+      "epoch": 0.32624012186994195,
+      "grad_norm": 1.6722263920875038,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 13706
+    },
+    {
+      "epoch": 0.32626392459297343,
+      "grad_norm": 17.473805224872304,
+      "learning_rate": 2e-05,
+      "loss": 0.5103,
+      "step": 13707
+    },
+    {
+      "epoch": 0.32628772731600497,
+      "grad_norm": 4.206453906763172,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 13708
+    },
+    {
+      "epoch": 0.32631153003903646,
+      "grad_norm": 9.297104936821068,
+      "learning_rate": 2e-05,
+      "loss": 0.4484,
+      "step": 13709
+    },
+    {
+      "epoch": 0.326335332762068,
+      "grad_norm": 41.53817461406164,
+      "learning_rate": 2e-05,
+      "loss": 1.0185,
+      "step": 13710
+    },
+    {
+      "epoch": 0.3263591354850995,
+      "grad_norm": 5.652057096550158,
+      "learning_rate": 2e-05,
+      "loss": 1.3056,
+      "step": 13711
+    },
+    {
+      "epoch": 0.326382938208131,
+      "grad_norm": 3.7093847756545655,
+      "learning_rate": 2e-05,
+      "loss": 0.5032,
+      "step": 13712
+    },
+    {
+      "epoch": 0.3264067409311625,
+      "grad_norm": 3.269483212354434,
+      "learning_rate": 2e-05,
+      "loss": 0.2673,
+      "step": 13713
+    },
+    {
+      "epoch": 0.32643054365419405,
+      "grad_norm": 5.941632118621166,
+      "learning_rate": 2e-05,
+      "loss": 1.0321,
+      "step": 13714
+    },
+    {
+      "epoch": 0.32645434637722553,
+      "grad_norm": 4.124170605735082,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 13715
+    },
+    {
+      "epoch": 0.3264781491002571,
+      "grad_norm": 2.0958839740958837,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 13716
+    },
+    {
+      "epoch": 0.32650195182328856,
+      "grad_norm": 6.181405498132458,
+      "learning_rate": 2e-05,
+      "loss": 0.6636,
+      "step": 13717
+    },
+    {
+      "epoch": 0.3265257545463201,
+      "grad_norm": 5.32023784858766,
+      "learning_rate": 2e-05,
+      "loss": 0.2767,
+      "step": 13718
+    },
+    {
+      "epoch": 0.3265495572693516,
+      "grad_norm": 2.0637186829341836,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 13719
+    },
+    {
+      "epoch": 0.3265733599923831,
+      "grad_norm": 4.497934523919481,
+      "learning_rate": 2e-05,
+      "loss": 0.2761,
+      "step": 13720
+    },
+    {
+      "epoch": 0.32659716271541467,
+      "grad_norm": 0.6195117533612448,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 13721
+    },
+    {
+      "epoch": 0.32662096543844615,
+      "grad_norm": 5.615518795978636,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 13722
+    },
+    {
+      "epoch": 0.3266447681614777,
+      "grad_norm": 1.4330435676374702,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 13723
+    },
+    {
+      "epoch": 0.3266685708845092,
+      "grad_norm": 5.14743838928845,
+      "learning_rate": 2e-05,
+      "loss": 0.2918,
+      "step": 13724
+    },
+    {
+      "epoch": 0.3266923736075407,
+      "grad_norm": 5.022806183881943,
+      "learning_rate": 2e-05,
+      "loss": 1.1788,
+      "step": 13725
+    },
+    {
+      "epoch": 0.3267161763305722,
+      "grad_norm": 0.48561183630564064,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 13726
+    },
+    {
+      "epoch": 0.32673997905360375,
+      "grad_norm": 25.65206488120544,
+      "learning_rate": 2e-05,
+      "loss": 0.7988,
+      "step": 13727
+    },
+    {
+      "epoch": 0.32676378177663523,
+      "grad_norm": 3.1619159099193554,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 13728
+    },
+    {
+      "epoch": 0.32678758449966677,
+      "grad_norm": 5.331033091856087,
+      "learning_rate": 2e-05,
+      "loss": 0.7996,
+      "step": 13729
+    },
+    {
+      "epoch": 0.32681138722269826,
+      "grad_norm": 5.835638178413667,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 13730
+    },
+    {
+      "epoch": 0.3268351899457298,
+      "grad_norm": 4.179625937967404,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 13731
+    },
+    {
+      "epoch": 0.3268589926687613,
+      "grad_norm": 3.955003302729412,
+      "learning_rate": 2e-05,
+      "loss": 0.8125,
+      "step": 13732
+    },
+    {
+      "epoch": 0.3268827953917928,
+      "grad_norm": 6.087334801681492,
+      "learning_rate": 2e-05,
+      "loss": 1.1275,
+      "step": 13733
+    },
+    {
+      "epoch": 0.3269065981148243,
+      "grad_norm": 2.6861588941163816,
+      "learning_rate": 2e-05,
+      "loss": 0.2256,
+      "step": 13734
+    },
+    {
+      "epoch": 0.32693040083785585,
+      "grad_norm": 3.258725426971355,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 13735
+    },
+    {
+      "epoch": 0.3269542035608874,
+      "grad_norm": 3.9155383033413433,
+      "learning_rate": 2e-05,
+      "loss": 0.1788,
+      "step": 13736
+    },
+    {
+      "epoch": 0.3269780062839189,
+      "grad_norm": 8.708623354750326,
+      "learning_rate": 2e-05,
+      "loss": 0.419,
+      "step": 13737
+    },
+    {
+      "epoch": 0.3270018090069504,
+      "grad_norm": 2.2132524749354925,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 13738
+    },
+    {
+      "epoch": 0.3270256117299819,
+      "grad_norm": 3.863869260247936,
+      "learning_rate": 2e-05,
+      "loss": 0.829,
+      "step": 13739
+    },
+    {
+      "epoch": 0.32704941445301344,
+      "grad_norm": 6.279320701575553,
+      "learning_rate": 2e-05,
+      "loss": 1.0972,
+      "step": 13740
+    },
+    {
+      "epoch": 0.32707321717604493,
+      "grad_norm": 1.1683223212450717,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 13741
+    },
+    {
+      "epoch": 0.32709701989907647,
+      "grad_norm": 4.15469854836579,
+      "learning_rate": 2e-05,
+      "loss": 0.5519,
+      "step": 13742
+    },
+    {
+      "epoch": 0.32712082262210795,
+      "grad_norm": 5.431659431963576,
+      "learning_rate": 2e-05,
+      "loss": 0.4976,
+      "step": 13743
+    },
+    {
+      "epoch": 0.3271446253451395,
+      "grad_norm": 3.0101048935753965,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 13744
+    },
+    {
+      "epoch": 0.327168428068171,
+      "grad_norm": 5.59427391967166,
+      "learning_rate": 2e-05,
+      "loss": 0.3254,
+      "step": 13745
+    },
+    {
+      "epoch": 0.3271922307912025,
+      "grad_norm": 9.425749652097322,
+      "learning_rate": 2e-05,
+      "loss": 0.734,
+      "step": 13746
+    },
+    {
+      "epoch": 0.327216033514234,
+      "grad_norm": 2.7049995201851584,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 13747
+    },
+    {
+      "epoch": 0.32723983623726555,
+      "grad_norm": 1.5421815067198061,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 13748
+    },
+    {
+      "epoch": 0.32726363896029703,
+      "grad_norm": 1.6105596414690184,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 13749
+    },
+    {
+      "epoch": 0.3272874416833286,
+      "grad_norm": 6.251374912244417,
+      "learning_rate": 2e-05,
+      "loss": 0.2793,
+      "step": 13750
+    },
+    {
+      "epoch": 0.3273112444063601,
+      "grad_norm": 2.340027284623549,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 13751
+    },
+    {
+      "epoch": 0.3273350471293916,
+      "grad_norm": 11.490226443539806,
+      "learning_rate": 2e-05,
+      "loss": 1.5287,
+      "step": 13752
+    },
+    {
+      "epoch": 0.32735884985242314,
+      "grad_norm": 3.1710075593904348,
+      "learning_rate": 2e-05,
+      "loss": 0.3958,
+      "step": 13753
+    },
+    {
+      "epoch": 0.3273826525754546,
+      "grad_norm": 5.698514260126146,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 13754
+    },
+    {
+      "epoch": 0.32740645529848617,
+      "grad_norm": 5.68218847177465,
+      "learning_rate": 2e-05,
+      "loss": 1.1062,
+      "step": 13755
+    },
+    {
+      "epoch": 0.32743025802151765,
+      "grad_norm": 3.7785744093199845,
+      "learning_rate": 2e-05,
+      "loss": 0.3888,
+      "step": 13756
+    },
+    {
+      "epoch": 0.3274540607445492,
+      "grad_norm": 10.061062119854062,
+      "learning_rate": 2e-05,
+      "loss": 1.3142,
+      "step": 13757
+    },
+    {
+      "epoch": 0.3274778634675807,
+      "grad_norm": 2.9325430476305083,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 13758
+    },
+    {
+      "epoch": 0.3275016661906122,
+      "grad_norm": 2.7045793192820655,
+      "learning_rate": 2e-05,
+      "loss": 0.3197,
+      "step": 13759
+    },
+    {
+      "epoch": 0.3275254689136437,
+      "grad_norm": 2.0194175945631745,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 13760
+    },
+    {
+      "epoch": 0.32754927163667524,
+      "grad_norm": 6.739282176783539,
+      "learning_rate": 2e-05,
+      "loss": 0.6926,
+      "step": 13761
+    },
+    {
+      "epoch": 0.32757307435970673,
+      "grad_norm": 6.061410996399997,
+      "learning_rate": 2e-05,
+      "loss": 0.9705,
+      "step": 13762
+    },
+    {
+      "epoch": 0.32759687708273827,
+      "grad_norm": 14.807067065270754,
+      "learning_rate": 2e-05,
+      "loss": 1.2198,
+      "step": 13763
+    },
+    {
+      "epoch": 0.32762067980576975,
+      "grad_norm": 0.04068348209424067,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 13764
+    },
+    {
+      "epoch": 0.3276444825288013,
+      "grad_norm": 13.877337070959664,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 13765
+    },
+    {
+      "epoch": 0.32766828525183284,
+      "grad_norm": 3.4465009497276773,
+      "learning_rate": 2e-05,
+      "loss": 0.1866,
+      "step": 13766
+    },
+    {
+      "epoch": 0.3276920879748643,
+      "grad_norm": 7.469043295022126,
+      "learning_rate": 2e-05,
+      "loss": 0.3608,
+      "step": 13767
+    },
+    {
+      "epoch": 0.32771589069789586,
+      "grad_norm": 2.3481817893171546,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 13768
+    },
+    {
+      "epoch": 0.32773969342092735,
+      "grad_norm": 1.2023330772430756,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 13769
+    },
+    {
+      "epoch": 0.3277634961439589,
+      "grad_norm": 9.869601941722403,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 13770
+    },
+    {
+      "epoch": 0.3277872988669904,
+      "grad_norm": 2.423118321421628,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 13771
+    },
+    {
+      "epoch": 0.3278111015900219,
+      "grad_norm": 7.431274739705451,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 13772
+    },
+    {
+      "epoch": 0.3278349043130534,
+      "grad_norm": 0.8513136189223718,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 13773
+    },
+    {
+      "epoch": 0.32785870703608494,
+      "grad_norm": 1.4636375024717896,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 13774
+    },
+    {
+      "epoch": 0.3278825097591164,
+      "grad_norm": 3.212078545028098,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 13775
+    },
+    {
+      "epoch": 0.32790631248214797,
+      "grad_norm": 6.3090073518475664,
+      "learning_rate": 2e-05,
+      "loss": 1.0704,
+      "step": 13776
+    },
+    {
+      "epoch": 0.32793011520517945,
+      "grad_norm": 5.013292487716287,
+      "learning_rate": 2e-05,
+      "loss": 0.8952,
+      "step": 13777
+    },
+    {
+      "epoch": 0.327953917928211,
+      "grad_norm": 2.594420421851897,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 13778
+    },
+    {
+      "epoch": 0.3279777206512425,
+      "grad_norm": 5.403739268020745,
+      "learning_rate": 2e-05,
+      "loss": 1.0948,
+      "step": 13779
+    },
+    {
+      "epoch": 0.328001523374274,
+      "grad_norm": 5.074153596277479,
+      "learning_rate": 2e-05,
+      "loss": 0.8042,
+      "step": 13780
+    },
+    {
+      "epoch": 0.32802532609730556,
+      "grad_norm": 1.8215910738649943,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 13781
+    },
+    {
+      "epoch": 0.32804912882033704,
+      "grad_norm": 2.2252425584517748,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 13782
+    },
+    {
+      "epoch": 0.3280729315433686,
+      "grad_norm": 2.0265885172285665,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 13783
+    },
+    {
+      "epoch": 0.32809673426640007,
+      "grad_norm": 6.946591436098986,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 13784
+    },
+    {
+      "epoch": 0.3281205369894316,
+      "grad_norm": 2.321412136589934,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 13785
+    },
+    {
+      "epoch": 0.3281443397124631,
+      "grad_norm": 1.5851193444235068,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 13786
+    },
+    {
+      "epoch": 0.32816814243549464,
+      "grad_norm": 5.29417290351397,
+      "learning_rate": 2e-05,
+      "loss": 0.7302,
+      "step": 13787
+    },
+    {
+      "epoch": 0.3281919451585261,
+      "grad_norm": 7.285470148161865,
+      "learning_rate": 2e-05,
+      "loss": 0.2959,
+      "step": 13788
+    },
+    {
+      "epoch": 0.32821574788155766,
+      "grad_norm": 3.538106726449458,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 13789
+    },
+    {
+      "epoch": 0.32823955060458915,
+      "grad_norm": 9.405411609656925,
+      "learning_rate": 2e-05,
+      "loss": 0.2646,
+      "step": 13790
+    },
+    {
+      "epoch": 0.3282633533276207,
+      "grad_norm": 1.3868689210944827,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 13791
+    },
+    {
+      "epoch": 0.3282871560506522,
+      "grad_norm": 2.782490081415466,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 13792
+    },
+    {
+      "epoch": 0.3283109587736837,
+      "grad_norm": 6.142422633116519,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 13793
+    },
+    {
+      "epoch": 0.3283347614967152,
+      "grad_norm": 3.9461882420595984,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 13794
+    },
+    {
+      "epoch": 0.32835856421974674,
+      "grad_norm": 0.42373219452142297,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13795
+    },
+    {
+      "epoch": 0.3283823669427783,
+      "grad_norm": 2.989610103730306,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 13796
+    },
+    {
+      "epoch": 0.32840616966580977,
+      "grad_norm": 6.148396131494096,
+      "learning_rate": 2e-05,
+      "loss": 0.6858,
+      "step": 13797
+    },
+    {
+      "epoch": 0.3284299723888413,
+      "grad_norm": 10.751793523212147,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 13798
+    },
+    {
+      "epoch": 0.3284537751118728,
+      "grad_norm": 4.942022277016055,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 13799
+    },
+    {
+      "epoch": 0.32847757783490433,
+      "grad_norm": 2.101721633634883,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 13800
+    },
+    {
+      "epoch": 0.3285013805579358,
+      "grad_norm": 1.9920966138837695,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 13801
+    },
+    {
+      "epoch": 0.32852518328096736,
+      "grad_norm": 3.286570434693491,
+      "learning_rate": 2e-05,
+      "loss": 0.6414,
+      "step": 13802
+    },
+    {
+      "epoch": 0.32854898600399884,
+      "grad_norm": 6.08213383610764,
+      "learning_rate": 2e-05,
+      "loss": 0.7301,
+      "step": 13803
+    },
+    {
+      "epoch": 0.3285727887270304,
+      "grad_norm": 6.624174198210413,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 13804
+    },
+    {
+      "epoch": 0.32859659145006187,
+      "grad_norm": 4.309018225471996,
+      "learning_rate": 2e-05,
+      "loss": 0.7055,
+      "step": 13805
+    },
+    {
+      "epoch": 0.3286203941730934,
+      "grad_norm": 10.035901497525835,
+      "learning_rate": 2e-05,
+      "loss": 0.451,
+      "step": 13806
+    },
+    {
+      "epoch": 0.3286441968961249,
+      "grad_norm": 4.274794225490523,
+      "learning_rate": 2e-05,
+      "loss": 0.767,
+      "step": 13807
+    },
+    {
+      "epoch": 0.32866799961915644,
+      "grad_norm": 6.155043554684905,
+      "learning_rate": 2e-05,
+      "loss": 1.2337,
+      "step": 13808
+    },
+    {
+      "epoch": 0.3286918023421879,
+      "grad_norm": 5.101899045698609,
+      "learning_rate": 2e-05,
+      "loss": 0.5077,
+      "step": 13809
+    },
+    {
+      "epoch": 0.32871560506521946,
+      "grad_norm": 2.5495551733990838,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 13810
+    },
+    {
+      "epoch": 0.328739407788251,
+      "grad_norm": 10.81983198874608,
+      "learning_rate": 2e-05,
+      "loss": 0.9996,
+      "step": 13811
+    },
+    {
+      "epoch": 0.3287632105112825,
+      "grad_norm": 8.50676306390175,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 13812
+    },
+    {
+      "epoch": 0.32878701323431403,
+      "grad_norm": 4.6655412755501855,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 13813
+    },
+    {
+      "epoch": 0.3288108159573455,
+      "grad_norm": 2.6145792093358096,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 13814
+    },
+    {
+      "epoch": 0.32883461868037706,
+      "grad_norm": 2.964839564433854,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 13815
+    },
+    {
+      "epoch": 0.32885842140340854,
+      "grad_norm": 6.968641140599841,
+      "learning_rate": 2e-05,
+      "loss": 0.3367,
+      "step": 13816
+    },
+    {
+      "epoch": 0.3288822241264401,
+      "grad_norm": 1.23315533323871,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 13817
+    },
+    {
+      "epoch": 0.32890602684947157,
+      "grad_norm": 3.920059588046777,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 13818
+    },
+    {
+      "epoch": 0.3289298295725031,
+      "grad_norm": 6.227297433534602,
+      "learning_rate": 2e-05,
+      "loss": 0.6052,
+      "step": 13819
+    },
+    {
+      "epoch": 0.3289536322955346,
+      "grad_norm": 2.586938781783402,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 13820
+    },
+    {
+      "epoch": 0.32897743501856613,
+      "grad_norm": 5.647359166839393,
+      "learning_rate": 2e-05,
+      "loss": 0.664,
+      "step": 13821
+    },
+    {
+      "epoch": 0.3290012377415976,
+      "grad_norm": 0.22026124857020088,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 13822
+    },
+    {
+      "epoch": 0.32902504046462916,
+      "grad_norm": 4.681719151241637,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 13823
+    },
+    {
+      "epoch": 0.32904884318766064,
+      "grad_norm": 6.842746274706633,
+      "learning_rate": 2e-05,
+      "loss": 0.8242,
+      "step": 13824
+    },
+    {
+      "epoch": 0.3290726459106922,
+      "grad_norm": 1.2662761145949741,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 13825
+    },
+    {
+      "epoch": 0.32909644863372367,
+      "grad_norm": 5.070326184336654,
+      "learning_rate": 2e-05,
+      "loss": 1.2941,
+      "step": 13826
+    },
+    {
+      "epoch": 0.3291202513567552,
+      "grad_norm": 5.550945254787545,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 13827
+    },
+    {
+      "epoch": 0.32914405407978675,
+      "grad_norm": 0.8878115563486401,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 13828
+    },
+    {
+      "epoch": 0.32916785680281824,
+      "grad_norm": 12.775051721385207,
+      "learning_rate": 2e-05,
+      "loss": 0.7552,
+      "step": 13829
+    },
+    {
+      "epoch": 0.3291916595258498,
+      "grad_norm": 1.1035475310225282,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 13830
+    },
+    {
+      "epoch": 0.32921546224888126,
+      "grad_norm": 5.776595462255457,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 13831
+    },
+    {
+      "epoch": 0.3292392649719128,
+      "grad_norm": 1.9538836523031053,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 13832
+    },
+    {
+      "epoch": 0.3292630676949443,
+      "grad_norm": 7.955176931409762,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 13833
+    },
+    {
+      "epoch": 0.32928687041797583,
+      "grad_norm": 3.526674467894377,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 13834
+    },
+    {
+      "epoch": 0.3293106731410073,
+      "grad_norm": 5.918832322448483,
+      "learning_rate": 2e-05,
+      "loss": 0.4146,
+      "step": 13835
+    },
+    {
+      "epoch": 0.32933447586403886,
+      "grad_norm": 5.177282562964679,
+      "learning_rate": 2e-05,
+      "loss": 1.0006,
+      "step": 13836
+    },
+    {
+      "epoch": 0.32935827858707034,
+      "grad_norm": 9.003168376514179,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 13837
+    },
+    {
+      "epoch": 0.3293820813101019,
+      "grad_norm": 6.224416297975879,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 13838
+    },
+    {
+      "epoch": 0.32940588403313337,
+      "grad_norm": 1.1077625273108058,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 13839
+    },
+    {
+      "epoch": 0.3294296867561649,
+      "grad_norm": 4.671607452615621,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 13840
+    },
+    {
+      "epoch": 0.3294534894791964,
+      "grad_norm": 6.811868761318264,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 13841
+    },
+    {
+      "epoch": 0.32947729220222793,
+      "grad_norm": 5.300716509833284,
+      "learning_rate": 2e-05,
+      "loss": 0.7236,
+      "step": 13842
+    },
+    {
+      "epoch": 0.3295010949252595,
+      "grad_norm": 4.649716545556165,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 13843
+    },
+    {
+      "epoch": 0.32952489764829096,
+      "grad_norm": 10.145594309722979,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 13844
+    },
+    {
+      "epoch": 0.3295487003713225,
+      "grad_norm": 0.43784428788138696,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 13845
+    },
+    {
+      "epoch": 0.329572503094354,
+      "grad_norm": 9.504172994768837,
+      "learning_rate": 2e-05,
+      "loss": 1.3551,
+      "step": 13846
+    },
+    {
+      "epoch": 0.3295963058173855,
+      "grad_norm": 1.3218347437788482,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 13847
+    },
+    {
+      "epoch": 0.329620108540417,
+      "grad_norm": 4.352178441232611,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 13848
+    },
+    {
+      "epoch": 0.32964391126344855,
+      "grad_norm": 4.8107319758625735,
+      "learning_rate": 2e-05,
+      "loss": 0.4577,
+      "step": 13849
+    },
+    {
+      "epoch": 0.32966771398648004,
+      "grad_norm": 2.5146022356427036,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 13850
+    },
+    {
+      "epoch": 0.3296915167095116,
+      "grad_norm": 7.989859561051355,
+      "learning_rate": 2e-05,
+      "loss": 1.1225,
+      "step": 13851
+    },
+    {
+      "epoch": 0.32971531943254306,
+      "grad_norm": 8.395360492339247,
+      "learning_rate": 2e-05,
+      "loss": 0.701,
+      "step": 13852
+    },
+    {
+      "epoch": 0.3297391221555746,
+      "grad_norm": 3.0838859775375336,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 13853
+    },
+    {
+      "epoch": 0.3297629248786061,
+      "grad_norm": 4.318003459484947,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 13854
+    },
+    {
+      "epoch": 0.32978672760163763,
+      "grad_norm": 8.842529667651307,
+      "learning_rate": 2e-05,
+      "loss": 1.107,
+      "step": 13855
+    },
+    {
+      "epoch": 0.3298105303246691,
+      "grad_norm": 3.8631412321262952,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 13856
+    },
+    {
+      "epoch": 0.32983433304770066,
+      "grad_norm": 5.883473391551275,
+      "learning_rate": 2e-05,
+      "loss": 1.1178,
+      "step": 13857
+    },
+    {
+      "epoch": 0.3298581357707322,
+      "grad_norm": 4.163078880057785,
+      "learning_rate": 2e-05,
+      "loss": 0.5734,
+      "step": 13858
+    },
+    {
+      "epoch": 0.3298819384937637,
+      "grad_norm": 2.3275132820868545,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 13859
+    },
+    {
+      "epoch": 0.3299057412167952,
+      "grad_norm": 6.812811752261833,
+      "learning_rate": 2e-05,
+      "loss": 1.2522,
+      "step": 13860
+    },
+    {
+      "epoch": 0.3299295439398267,
+      "grad_norm": 10.63845758723048,
+      "learning_rate": 2e-05,
+      "loss": 0.6297,
+      "step": 13861
+    },
+    {
+      "epoch": 0.32995334666285825,
+      "grad_norm": 9.173969826323686,
+      "learning_rate": 2e-05,
+      "loss": 0.8895,
+      "step": 13862
+    },
+    {
+      "epoch": 0.32997714938588973,
+      "grad_norm": 10.576397655533164,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 13863
+    },
+    {
+      "epoch": 0.3300009521089213,
+      "grad_norm": 5.667689524050673,
+      "learning_rate": 2e-05,
+      "loss": 0.9858,
+      "step": 13864
+    },
+    {
+      "epoch": 0.33002475483195276,
+      "grad_norm": 7.247295843805553,
+      "learning_rate": 2e-05,
+      "loss": 0.308,
+      "step": 13865
+    },
+    {
+      "epoch": 0.3300485575549843,
+      "grad_norm": 3.680541931863084,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 13866
+    },
+    {
+      "epoch": 0.3300723602780158,
+      "grad_norm": 3.443364638083981,
+      "learning_rate": 2e-05,
+      "loss": 0.1731,
+      "step": 13867
+    },
+    {
+      "epoch": 0.3300961630010473,
+      "grad_norm": 6.495968728338328,
+      "learning_rate": 2e-05,
+      "loss": 0.9748,
+      "step": 13868
+    },
+    {
+      "epoch": 0.3301199657240788,
+      "grad_norm": 7.131919247562532,
+      "learning_rate": 2e-05,
+      "loss": 0.7774,
+      "step": 13869
+    },
+    {
+      "epoch": 0.33014376844711035,
+      "grad_norm": 6.430132606893967,
+      "learning_rate": 2e-05,
+      "loss": 1.0656,
+      "step": 13870
+    },
+    {
+      "epoch": 0.33016757117014184,
+      "grad_norm": 3.006903392428747,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 13871
+    },
+    {
+      "epoch": 0.3301913738931734,
+      "grad_norm": 4.13141057697536,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 13872
+    },
+    {
+      "epoch": 0.3302151766162049,
+      "grad_norm": 0.9083470013738768,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 13873
+    },
+    {
+      "epoch": 0.3302389793392364,
+      "grad_norm": 10.409185534186099,
+      "learning_rate": 2e-05,
+      "loss": 0.3585,
+      "step": 13874
+    },
+    {
+      "epoch": 0.33026278206226795,
+      "grad_norm": 5.4180049720706975,
+      "learning_rate": 2e-05,
+      "loss": 0.6364,
+      "step": 13875
+    },
+    {
+      "epoch": 0.33028658478529943,
+      "grad_norm": 8.00780620044117,
+      "learning_rate": 2e-05,
+      "loss": 0.329,
+      "step": 13876
+    },
+    {
+      "epoch": 0.33031038750833097,
+      "grad_norm": 5.160905361932592,
+      "learning_rate": 2e-05,
+      "loss": 0.9817,
+      "step": 13877
+    },
+    {
+      "epoch": 0.33033419023136246,
+      "grad_norm": 5.776172721279915,
+      "learning_rate": 2e-05,
+      "loss": 0.2107,
+      "step": 13878
+    },
+    {
+      "epoch": 0.330357992954394,
+      "grad_norm": 15.31259122319494,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 13879
+    },
+    {
+      "epoch": 0.3303817956774255,
+      "grad_norm": 5.684627068927944,
+      "learning_rate": 2e-05,
+      "loss": 1.2193,
+      "step": 13880
+    },
+    {
+      "epoch": 0.330405598400457,
+      "grad_norm": 6.15039319290181,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 13881
+    },
+    {
+      "epoch": 0.3304294011234885,
+      "grad_norm": 1.5367116637838962,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 13882
+    },
+    {
+      "epoch": 0.33045320384652005,
+      "grad_norm": 1.3412835343211478,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 13883
+    },
+    {
+      "epoch": 0.33047700656955153,
+      "grad_norm": 2.0793583581429496,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 13884
+    },
+    {
+      "epoch": 0.3305008092925831,
+      "grad_norm": 2.6045355893158053,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 13885
+    },
+    {
+      "epoch": 0.33052461201561456,
+      "grad_norm": 2.8092831644400116,
+      "learning_rate": 2e-05,
+      "loss": 0.3415,
+      "step": 13886
+    },
+    {
+      "epoch": 0.3305484147386461,
+      "grad_norm": 2.5998822751855872,
+      "learning_rate": 2e-05,
+      "loss": 0.1128,
+      "step": 13887
+    },
+    {
+      "epoch": 0.33057221746167764,
+      "grad_norm": 1.9376336809142363,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 13888
+    },
+    {
+      "epoch": 0.3305960201847091,
+      "grad_norm": 11.178076367399957,
+      "learning_rate": 2e-05,
+      "loss": 0.9751,
+      "step": 13889
+    },
+    {
+      "epoch": 0.33061982290774067,
+      "grad_norm": 8.551904256732394,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 13890
+    },
+    {
+      "epoch": 0.33064362563077215,
+      "grad_norm": 7.4605927103103005,
+      "learning_rate": 2e-05,
+      "loss": 0.357,
+      "step": 13891
+    },
+    {
+      "epoch": 0.3306674283538037,
+      "grad_norm": 2.8490621848493256,
+      "learning_rate": 2e-05,
+      "loss": 0.3539,
+      "step": 13892
+    },
+    {
+      "epoch": 0.3306912310768352,
+      "grad_norm": 6.565793164197253,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 13893
+    },
+    {
+      "epoch": 0.3307150337998667,
+      "grad_norm": 1.8201913934368537,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 13894
+    },
+    {
+      "epoch": 0.3307388365228982,
+      "grad_norm": 14.297276016994644,
+      "learning_rate": 2e-05,
+      "loss": 0.9511,
+      "step": 13895
+    },
+    {
+      "epoch": 0.33076263924592975,
+      "grad_norm": 6.199429500720089,
+      "learning_rate": 2e-05,
+      "loss": 0.882,
+      "step": 13896
+    },
+    {
+      "epoch": 0.33078644196896123,
+      "grad_norm": 1.0141559122357684,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 13897
+    },
+    {
+      "epoch": 0.33081024469199277,
+      "grad_norm": 5.924980726006036,
+      "learning_rate": 2e-05,
+      "loss": 1.0227,
+      "step": 13898
+    },
+    {
+      "epoch": 0.33083404741502426,
+      "grad_norm": 2.1536785983406688,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 13899
+    },
+    {
+      "epoch": 0.3308578501380558,
+      "grad_norm": 5.513857531035751,
+      "learning_rate": 2e-05,
+      "loss": 0.1916,
+      "step": 13900
+    },
+    {
+      "epoch": 0.3308816528610873,
+      "grad_norm": 1.4373471984520074,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 13901
+    },
+    {
+      "epoch": 0.3309054555841188,
+      "grad_norm": 2.8719919324597516,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 13902
+    },
+    {
+      "epoch": 0.33092925830715036,
+      "grad_norm": 4.990246787172737,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 13903
+    },
+    {
+      "epoch": 0.33095306103018185,
+      "grad_norm": 2.6424698735592735,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 13904
+    },
+    {
+      "epoch": 0.3309768637532134,
+      "grad_norm": 0.4402145666493777,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 13905
+    },
+    {
+      "epoch": 0.3310006664762449,
+      "grad_norm": 5.155283042993791,
+      "learning_rate": 2e-05,
+      "loss": 0.948,
+      "step": 13906
+    },
+    {
+      "epoch": 0.3310244691992764,
+      "grad_norm": 2.7652773542432656,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 13907
+    },
+    {
+      "epoch": 0.3310482719223079,
+      "grad_norm": 9.093712444067018,
+      "learning_rate": 2e-05,
+      "loss": 1.3195,
+      "step": 13908
+    },
+    {
+      "epoch": 0.33107207464533944,
+      "grad_norm": 5.669254826090774,
+      "learning_rate": 2e-05,
+      "loss": 0.5158,
+      "step": 13909
+    },
+    {
+      "epoch": 0.3310958773683709,
+      "grad_norm": 1.4752249180707735,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 13910
+    },
+    {
+      "epoch": 0.33111968009140247,
+      "grad_norm": 5.633762489366989,
+      "learning_rate": 2e-05,
+      "loss": 0.3534,
+      "step": 13911
+    },
+    {
+      "epoch": 0.33114348281443395,
+      "grad_norm": 7.111169334102606,
+      "learning_rate": 2e-05,
+      "loss": 0.5885,
+      "step": 13912
+    },
+    {
+      "epoch": 0.3311672855374655,
+      "grad_norm": 5.5359487161263,
+      "learning_rate": 2e-05,
+      "loss": 0.1944,
+      "step": 13913
+    },
+    {
+      "epoch": 0.331191088260497,
+      "grad_norm": 1.4567305475475576,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 13914
+    },
+    {
+      "epoch": 0.3312148909835285,
+      "grad_norm": 4.372593401304419,
+      "learning_rate": 2e-05,
+      "loss": 0.8343,
+      "step": 13915
+    },
+    {
+      "epoch": 0.33123869370656,
+      "grad_norm": 2.7977362477388197,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 13916
+    },
+    {
+      "epoch": 0.33126249642959155,
+      "grad_norm": 0.6281698197973058,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 13917
+    },
+    {
+      "epoch": 0.3312862991526231,
+      "grad_norm": 5.499381961253585,
+      "learning_rate": 2e-05,
+      "loss": 1.0351,
+      "step": 13918
+    },
+    {
+      "epoch": 0.3313101018756546,
+      "grad_norm": 3.305330088555354,
+      "learning_rate": 2e-05,
+      "loss": 0.1296,
+      "step": 13919
+    },
+    {
+      "epoch": 0.3313339045986861,
+      "grad_norm": 2.567190516136497,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 13920
+    },
+    {
+      "epoch": 0.3313577073217176,
+      "grad_norm": 4.911072591090264,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 13921
+    },
+    {
+      "epoch": 0.33138151004474914,
+      "grad_norm": 6.559305388912295,
+      "learning_rate": 2e-05,
+      "loss": 0.5063,
+      "step": 13922
+    },
+    {
+      "epoch": 0.3314053127677806,
+      "grad_norm": 3.1114947060108564,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 13923
+    },
+    {
+      "epoch": 0.33142911549081217,
+      "grad_norm": 4.207641331166325,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 13924
+    },
+    {
+      "epoch": 0.33145291821384365,
+      "grad_norm": 9.457015885501322,
+      "learning_rate": 2e-05,
+      "loss": 1.2961,
+      "step": 13925
+    },
+    {
+      "epoch": 0.3314767209368752,
+      "grad_norm": 4.908341080506923,
+      "learning_rate": 2e-05,
+      "loss": 0.1479,
+      "step": 13926
+    },
+    {
+      "epoch": 0.3315005236599067,
+      "grad_norm": 3.6294293340293304,
+      "learning_rate": 2e-05,
+      "loss": 0.3264,
+      "step": 13927
+    },
+    {
+      "epoch": 0.3315243263829382,
+      "grad_norm": 0.6654653067581883,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 13928
+    },
+    {
+      "epoch": 0.3315481291059697,
+      "grad_norm": 3.5979157834506514,
+      "learning_rate": 2e-05,
+      "loss": 0.1402,
+      "step": 13929
+    },
+    {
+      "epoch": 0.33157193182900124,
+      "grad_norm": 2.870890039185765,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 13930
+    },
+    {
+      "epoch": 0.33159573455203273,
+      "grad_norm": 2.01199002909186,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 13931
+    },
+    {
+      "epoch": 0.33161953727506427,
+      "grad_norm": 2.354311797274541,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 13932
+    },
+    {
+      "epoch": 0.3316433399980958,
+      "grad_norm": 10.1254323343214,
+      "learning_rate": 2e-05,
+      "loss": 0.9757,
+      "step": 13933
+    },
+    {
+      "epoch": 0.3316671427211273,
+      "grad_norm": 4.869015332659947,
+      "learning_rate": 2e-05,
+      "loss": 0.4926,
+      "step": 13934
+    },
+    {
+      "epoch": 0.33169094544415884,
+      "grad_norm": 9.019085911486567,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 13935
+    },
+    {
+      "epoch": 0.3317147481671903,
+      "grad_norm": 5.831960122968377,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 13936
+    },
+    {
+      "epoch": 0.33173855089022186,
+      "grad_norm": 3.556245746811464,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 13937
+    },
+    {
+      "epoch": 0.33176235361325335,
+      "grad_norm": 0.9840928535611028,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 13938
+    },
+    {
+      "epoch": 0.3317861563362849,
+      "grad_norm": 3.4368645882905207,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 13939
+    },
+    {
+      "epoch": 0.3318099590593164,
+      "grad_norm": 3.886461625705018,
+      "learning_rate": 2e-05,
+      "loss": 0.1796,
+      "step": 13940
+    },
+    {
+      "epoch": 0.3318337617823479,
+      "grad_norm": 4.512549725802736,
+      "learning_rate": 2e-05,
+      "loss": 0.5437,
+      "step": 13941
+    },
+    {
+      "epoch": 0.3318575645053794,
+      "grad_norm": 4.288348211694601,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 13942
+    },
+    {
+      "epoch": 0.33188136722841094,
+      "grad_norm": 5.563017995998519,
+      "learning_rate": 2e-05,
+      "loss": 0.2868,
+      "step": 13943
+    },
+    {
+      "epoch": 0.3319051699514424,
+      "grad_norm": 4.842699183341851,
+      "learning_rate": 2e-05,
+      "loss": 0.7635,
+      "step": 13944
+    },
+    {
+      "epoch": 0.33192897267447397,
+      "grad_norm": 0.8726927602817519,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 13945
+    },
+    {
+      "epoch": 0.33195277539750545,
+      "grad_norm": 0.8242258489577443,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 13946
+    },
+    {
+      "epoch": 0.331976578120537,
+      "grad_norm": 4.667427045772278,
+      "learning_rate": 2e-05,
+      "loss": 0.589,
+      "step": 13947
+    },
+    {
+      "epoch": 0.3320003808435685,
+      "grad_norm": 3.8288150649862986,
+      "learning_rate": 2e-05,
+      "loss": 0.1626,
+      "step": 13948
+    },
+    {
+      "epoch": 0.3320241835666,
+      "grad_norm": 4.287872609440791,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 13949
+    },
+    {
+      "epoch": 0.33204798628963156,
+      "grad_norm": 0.7273545640244642,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 13950
+    },
+    {
+      "epoch": 0.33207178901266304,
+      "grad_norm": 0.6809926649893622,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 13951
+    },
+    {
+      "epoch": 0.3320955917356946,
+      "grad_norm": 3.57095694901723,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 13952
+    },
+    {
+      "epoch": 0.33211939445872607,
+      "grad_norm": 2.964735242525552,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 13953
+    },
+    {
+      "epoch": 0.3321431971817576,
+      "grad_norm": 4.658627471694098,
+      "learning_rate": 2e-05,
+      "loss": 0.3946,
+      "step": 13954
+    },
+    {
+      "epoch": 0.3321669999047891,
+      "grad_norm": 8.294120660840113,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 13955
+    },
+    {
+      "epoch": 0.33219080262782064,
+      "grad_norm": 4.845011481887197,
+      "learning_rate": 2e-05,
+      "loss": 0.6226,
+      "step": 13956
+    },
+    {
+      "epoch": 0.3322146053508521,
+      "grad_norm": 6.23491459390025,
+      "learning_rate": 2e-05,
+      "loss": 1.2193,
+      "step": 13957
+    },
+    {
+      "epoch": 0.33223840807388366,
+      "grad_norm": 3.378370581679054,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 13958
+    },
+    {
+      "epoch": 0.33226221079691515,
+      "grad_norm": 1.9413269448719164,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 13959
+    },
+    {
+      "epoch": 0.3322860135199467,
+      "grad_norm": 0.7643520285419451,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 13960
+    },
+    {
+      "epoch": 0.3323098162429782,
+      "grad_norm": 8.763480010782105,
+      "learning_rate": 2e-05,
+      "loss": 0.1473,
+      "step": 13961
+    },
+    {
+      "epoch": 0.3323336189660097,
+      "grad_norm": 4.361483027159625,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 13962
+    },
+    {
+      "epoch": 0.3323574216890412,
+      "grad_norm": 2.1097212946301704,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 13963
+    },
+    {
+      "epoch": 0.33238122441207274,
+      "grad_norm": 6.5847511023674885,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 13964
+    },
+    {
+      "epoch": 0.3324050271351043,
+      "grad_norm": 5.712580159772502,
+      "learning_rate": 2e-05,
+      "loss": 1.1787,
+      "step": 13965
+    },
+    {
+      "epoch": 0.33242882985813577,
+      "grad_norm": 0.4277889766065971,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 13966
+    },
+    {
+      "epoch": 0.3324526325811673,
+      "grad_norm": 3.6459487736970444,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 13967
+    },
+    {
+      "epoch": 0.3324764353041988,
+      "grad_norm": 3.7275140704291694,
+      "learning_rate": 2e-05,
+      "loss": 0.3911,
+      "step": 13968
+    },
+    {
+      "epoch": 0.33250023802723033,
+      "grad_norm": 3.3526459643302706,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 13969
+    },
+    {
+      "epoch": 0.3325240407502618,
+      "grad_norm": 6.515948382974083,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 13970
+    },
+    {
+      "epoch": 0.33254784347329336,
+      "grad_norm": 4.757717786618754,
+      "learning_rate": 2e-05,
+      "loss": 0.5634,
+      "step": 13971
+    },
+    {
+      "epoch": 0.33257164619632484,
+      "grad_norm": 4.43879462657854,
+      "learning_rate": 2e-05,
+      "loss": 0.3935,
+      "step": 13972
+    },
+    {
+      "epoch": 0.3325954489193564,
+      "grad_norm": 4.076655856927857,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 13973
+    },
+    {
+      "epoch": 0.33261925164238787,
+      "grad_norm": 4.468866827986678,
+      "learning_rate": 2e-05,
+      "loss": 0.4721,
+      "step": 13974
+    },
+    {
+      "epoch": 0.3326430543654194,
+      "grad_norm": 2.3695762205109316,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 13975
+    },
+    {
+      "epoch": 0.3326668570884509,
+      "grad_norm": 5.443763342469302,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 13976
+    },
+    {
+      "epoch": 0.33269065981148244,
+      "grad_norm": 3.9986863948179323,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 13977
+    },
+    {
+      "epoch": 0.3327144625345139,
+      "grad_norm": 5.112566614328481,
+      "learning_rate": 2e-05,
+      "loss": 0.5765,
+      "step": 13978
+    },
+    {
+      "epoch": 0.33273826525754546,
+      "grad_norm": 6.056158172262645,
+      "learning_rate": 2e-05,
+      "loss": 0.1302,
+      "step": 13979
+    },
+    {
+      "epoch": 0.332762067980577,
+      "grad_norm": 3.501489398115927,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 13980
+    },
+    {
+      "epoch": 0.3327858707036085,
+      "grad_norm": 7.070276463174571,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 13981
+    },
+    {
+      "epoch": 0.33280967342664003,
+      "grad_norm": 5.006035032775821,
+      "learning_rate": 2e-05,
+      "loss": 0.6962,
+      "step": 13982
+    },
+    {
+      "epoch": 0.3328334761496715,
+      "grad_norm": 5.713877897637108,
+      "learning_rate": 2e-05,
+      "loss": 0.2911,
+      "step": 13983
+    },
+    {
+      "epoch": 0.33285727887270306,
+      "grad_norm": 4.091688341303269,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 13984
+    },
+    {
+      "epoch": 0.33288108159573454,
+      "grad_norm": 7.8894804195096855,
+      "learning_rate": 2e-05,
+      "loss": 0.4262,
+      "step": 13985
+    },
+    {
+      "epoch": 0.3329048843187661,
+      "grad_norm": 4.871407307037411,
+      "learning_rate": 2e-05,
+      "loss": 1.0428,
+      "step": 13986
+    },
+    {
+      "epoch": 0.33292868704179757,
+      "grad_norm": 6.055421815353565,
+      "learning_rate": 2e-05,
+      "loss": 0.851,
+      "step": 13987
+    },
+    {
+      "epoch": 0.3329524897648291,
+      "grad_norm": 1.7554596727408673,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 13988
+    },
+    {
+      "epoch": 0.3329762924878606,
+      "grad_norm": 5.723182707971678,
+      "learning_rate": 2e-05,
+      "loss": 0.8328,
+      "step": 13989
+    },
+    {
+      "epoch": 0.33300009521089213,
+      "grad_norm": 2.4200042757030302,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 13990
+    },
+    {
+      "epoch": 0.3330238979339236,
+      "grad_norm": 8.493988141521045,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 13991
+    },
+    {
+      "epoch": 0.33304770065695516,
+      "grad_norm": 2.087573090397061,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 13992
+    },
+    {
+      "epoch": 0.33307150337998664,
+      "grad_norm": 0.9875121146987313,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 13993
+    },
+    {
+      "epoch": 0.3330953061030182,
+      "grad_norm": 1.52579693784751,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 13994
+    },
+    {
+      "epoch": 0.3331191088260497,
+      "grad_norm": 6.584092737902046,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 13995
+    },
+    {
+      "epoch": 0.3331429115490812,
+      "grad_norm": 2.134611246752009,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 13996
+    },
+    {
+      "epoch": 0.33316671427211275,
+      "grad_norm": 2.356851310378862,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 13997
+    },
+    {
+      "epoch": 0.33319051699514424,
+      "grad_norm": 2.8075730003177672,
+      "learning_rate": 2e-05,
+      "loss": 0.1972,
+      "step": 13998
+    },
+    {
+      "epoch": 0.3332143197181758,
+      "grad_norm": 5.344427352973014,
+      "learning_rate": 2e-05,
+      "loss": 0.7502,
+      "step": 13999
+    },
+    {
+      "epoch": 0.33323812244120726,
+      "grad_norm": 8.337918494117387,
+      "learning_rate": 2e-05,
+      "loss": 1.4883,
+      "step": 14000
+    },
+    {
+      "epoch": 0.3332619251642388,
+      "grad_norm": 2.783486259026745,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 14001
+    },
+    {
+      "epoch": 0.3332857278872703,
+      "grad_norm": 6.449586661215998,
+      "learning_rate": 2e-05,
+      "loss": 0.586,
+      "step": 14002
+    },
+    {
+      "epoch": 0.33330953061030183,
+      "grad_norm": 3.151535031791939,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 14003
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 12.261305851783796,
+      "learning_rate": 2e-05,
+      "loss": 0.6679,
+      "step": 14004
+    },
+    {
+      "epoch": 0.33335713605636486,
+      "grad_norm": 3.293191079955754,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 14005
+    },
+    {
+      "epoch": 0.33338093877939634,
+      "grad_norm": 1.5758759080162155,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 14006
+    },
+    {
+      "epoch": 0.3334047415024279,
+      "grad_norm": 11.075528960191923,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 14007
+    },
+    {
+      "epoch": 0.33342854422545937,
+      "grad_norm": 0.5776700460062969,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 14008
+    },
+    {
+      "epoch": 0.3334523469484909,
+      "grad_norm": 0.6499610244227046,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 14009
+    },
+    {
+      "epoch": 0.33347614967152245,
+      "grad_norm": 14.50012538304307,
+      "learning_rate": 2e-05,
+      "loss": 0.8051,
+      "step": 14010
+    },
+    {
+      "epoch": 0.33349995239455393,
+      "grad_norm": 0.3003992442149169,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 14011
+    },
+    {
+      "epoch": 0.3335237551175855,
+      "grad_norm": 7.445759610381916,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 14012
+    },
+    {
+      "epoch": 0.33354755784061696,
+      "grad_norm": 5.8349851071069,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 14013
+    },
+    {
+      "epoch": 0.3335713605636485,
+      "grad_norm": 4.729044270267532,
+      "learning_rate": 2e-05,
+      "loss": 0.1941,
+      "step": 14014
+    },
+    {
+      "epoch": 0.33359516328668,
+      "grad_norm": 1.6822452637560434,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 14015
+    },
+    {
+      "epoch": 0.3336189660097115,
+      "grad_norm": 5.253640327836691,
+      "learning_rate": 2e-05,
+      "loss": 0.7436,
+      "step": 14016
+    },
+    {
+      "epoch": 0.333642768732743,
+      "grad_norm": 1.895628355031701,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 14017
+    },
+    {
+      "epoch": 0.33366657145577455,
+      "grad_norm": 10.24348400968821,
+      "learning_rate": 2e-05,
+      "loss": 0.4794,
+      "step": 14018
+    },
+    {
+      "epoch": 0.33369037417880604,
+      "grad_norm": 7.099194372497503,
+      "learning_rate": 2e-05,
+      "loss": 0.232,
+      "step": 14019
+    },
+    {
+      "epoch": 0.3337141769018376,
+      "grad_norm": 5.873246255412742,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 14020
+    },
+    {
+      "epoch": 0.33373797962486906,
+      "grad_norm": 5.831866985662312,
+      "learning_rate": 2e-05,
+      "loss": 1.0692,
+      "step": 14021
+    },
+    {
+      "epoch": 0.3337617823479006,
+      "grad_norm": 4.809954759714449,
+      "learning_rate": 2e-05,
+      "loss": 0.2555,
+      "step": 14022
+    },
+    {
+      "epoch": 0.3337855850709321,
+      "grad_norm": 10.577717926922961,
+      "learning_rate": 2e-05,
+      "loss": 0.3867,
+      "step": 14023
+    },
+    {
+      "epoch": 0.33380938779396363,
+      "grad_norm": 0.8169035161751099,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 14024
+    },
+    {
+      "epoch": 0.33383319051699517,
+      "grad_norm": 9.921892122179628,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 14025
+    },
+    {
+      "epoch": 0.33385699324002666,
+      "grad_norm": 6.277452267785032,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 14026
+    },
+    {
+      "epoch": 0.3338807959630582,
+      "grad_norm": 3.8282109599029397,
+      "learning_rate": 2e-05,
+      "loss": 0.7141,
+      "step": 14027
+    },
+    {
+      "epoch": 0.3339045986860897,
+      "grad_norm": 8.902291798969356,
+      "learning_rate": 2e-05,
+      "loss": 0.5736,
+      "step": 14028
+    },
+    {
+      "epoch": 0.3339284014091212,
+      "grad_norm": 4.509922481683078,
+      "learning_rate": 2e-05,
+      "loss": 0.5935,
+      "step": 14029
+    },
+    {
+      "epoch": 0.3339522041321527,
+      "grad_norm": 0.7767871616405414,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 14030
+    },
+    {
+      "epoch": 0.33397600685518425,
+      "grad_norm": 5.588211064278602,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 14031
+    },
+    {
+      "epoch": 0.33399980957821573,
+      "grad_norm": 4.584947387370916,
+      "learning_rate": 2e-05,
+      "loss": 0.6027,
+      "step": 14032
+    },
+    {
+      "epoch": 0.3340236123012473,
+      "grad_norm": 4.802966719692863,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 14033
+    },
+    {
+      "epoch": 0.33404741502427876,
+      "grad_norm": 0.966124448776925,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 14034
+    },
+    {
+      "epoch": 0.3340712177473103,
+      "grad_norm": 7.349811406385223,
+      "learning_rate": 2e-05,
+      "loss": 0.702,
+      "step": 14035
+    },
+    {
+      "epoch": 0.3340950204703418,
+      "grad_norm": 4.80644073798298,
+      "learning_rate": 2e-05,
+      "loss": 0.3525,
+      "step": 14036
+    },
+    {
+      "epoch": 0.3341188231933733,
+      "grad_norm": 1.1178691420895237,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 14037
+    },
+    {
+      "epoch": 0.3341426259164048,
+      "grad_norm": 2.2849479413656466,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 14038
+    },
+    {
+      "epoch": 0.33416642863943635,
+      "grad_norm": 5.964986586966225,
+      "learning_rate": 2e-05,
+      "loss": 1.358,
+      "step": 14039
+    },
+    {
+      "epoch": 0.3341902313624679,
+      "grad_norm": 2.8538540053952923,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 14040
+    },
+    {
+      "epoch": 0.3342140340854994,
+      "grad_norm": 6.689089753093162,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 14041
+    },
+    {
+      "epoch": 0.3342378368085309,
+      "grad_norm": 2.114768874580101,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 14042
+    },
+    {
+      "epoch": 0.3342616395315624,
+      "grad_norm": 4.859561817851169,
+      "learning_rate": 2e-05,
+      "loss": 0.4583,
+      "step": 14043
+    },
+    {
+      "epoch": 0.33428544225459395,
+      "grad_norm": 2.260128779301926,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 14044
+    },
+    {
+      "epoch": 0.33430924497762543,
+      "grad_norm": 4.553302175686151,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 14045
+    },
+    {
+      "epoch": 0.33433304770065697,
+      "grad_norm": 5.827157278485613,
+      "learning_rate": 2e-05,
+      "loss": 1.3608,
+      "step": 14046
+    },
+    {
+      "epoch": 0.33435685042368846,
+      "grad_norm": 3.893954546645848,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 14047
+    },
+    {
+      "epoch": 0.33438065314672,
+      "grad_norm": 2.682357158146726,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 14048
+    },
+    {
+      "epoch": 0.3344044558697515,
+      "grad_norm": 3.376831068749085,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 14049
+    },
+    {
+      "epoch": 0.334428258592783,
+      "grad_norm": 3.7166729083005294,
+      "learning_rate": 2e-05,
+      "loss": 0.2168,
+      "step": 14050
+    },
+    {
+      "epoch": 0.3344520613158145,
+      "grad_norm": 4.1924002945478795,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 14051
+    },
+    {
+      "epoch": 0.33447586403884605,
+      "grad_norm": 7.738509156400031,
+      "learning_rate": 2e-05,
+      "loss": 0.8138,
+      "step": 14052
+    },
+    {
+      "epoch": 0.33449966676187753,
+      "grad_norm": 2.08504373382838,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 14053
+    },
+    {
+      "epoch": 0.3345234694849091,
+      "grad_norm": 7.267590557170292,
+      "learning_rate": 2e-05,
+      "loss": 0.4122,
+      "step": 14054
+    },
+    {
+      "epoch": 0.3345472722079406,
+      "grad_norm": 1.4167189426233286,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 14055
+    },
+    {
+      "epoch": 0.3345710749309721,
+      "grad_norm": 5.596191648983531,
+      "learning_rate": 2e-05,
+      "loss": 0.5599,
+      "step": 14056
+    },
+    {
+      "epoch": 0.33459487765400364,
+      "grad_norm": 3.395371558109503,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 14057
+    },
+    {
+      "epoch": 0.3346186803770351,
+      "grad_norm": 7.4349455921319585,
+      "learning_rate": 2e-05,
+      "loss": 0.5493,
+      "step": 14058
+    },
+    {
+      "epoch": 0.33464248310006667,
+      "grad_norm": 7.292397537196749,
+      "learning_rate": 2e-05,
+      "loss": 0.2336,
+      "step": 14059
+    },
+    {
+      "epoch": 0.33466628582309815,
+      "grad_norm": 1.732053138721193,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 14060
+    },
+    {
+      "epoch": 0.3346900885461297,
+      "grad_norm": 3.1066284160058557,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 14061
+    },
+    {
+      "epoch": 0.3347138912691612,
+      "grad_norm": 3.5075114757156776,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 14062
+    },
+    {
+      "epoch": 0.3347376939921927,
+      "grad_norm": 2.3159948794943324,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 14063
+    },
+    {
+      "epoch": 0.3347614967152242,
+      "grad_norm": 3.795166016585903,
+      "learning_rate": 2e-05,
+      "loss": 0.5615,
+      "step": 14064
+    },
+    {
+      "epoch": 0.33478529943825575,
+      "grad_norm": 5.3137237855907475,
+      "learning_rate": 2e-05,
+      "loss": 0.9575,
+      "step": 14065
+    },
+    {
+      "epoch": 0.33480910216128723,
+      "grad_norm": 4.879701522384797,
+      "learning_rate": 2e-05,
+      "loss": 0.5515,
+      "step": 14066
+    },
+    {
+      "epoch": 0.33483290488431877,
+      "grad_norm": 4.947310410279018,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 14067
+    },
+    {
+      "epoch": 0.33485670760735026,
+      "grad_norm": 1.8814771011259854,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 14068
+    },
+    {
+      "epoch": 0.3348805103303818,
+      "grad_norm": 19.603801251395446,
+      "learning_rate": 2e-05,
+      "loss": 0.3864,
+      "step": 14069
+    },
+    {
+      "epoch": 0.3349043130534133,
+      "grad_norm": 3.385089528050905,
+      "learning_rate": 2e-05,
+      "loss": 0.2283,
+      "step": 14070
+    },
+    {
+      "epoch": 0.3349281157764448,
+      "grad_norm": 1.8777733280761755,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 14071
+    },
+    {
+      "epoch": 0.33495191849947636,
+      "grad_norm": 0.6341337037894108,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 14072
+    },
+    {
+      "epoch": 0.33497572122250785,
+      "grad_norm": 4.834778735094565,
+      "learning_rate": 2e-05,
+      "loss": 0.5838,
+      "step": 14073
+    },
+    {
+      "epoch": 0.3349995239455394,
+      "grad_norm": 12.891624639026197,
+      "learning_rate": 2e-05,
+      "loss": 0.9495,
+      "step": 14074
+    },
+    {
+      "epoch": 0.3350233266685709,
+      "grad_norm": 2.286154655101716,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 14075
+    },
+    {
+      "epoch": 0.3350471293916024,
+      "grad_norm": 8.591330419325956,
+      "learning_rate": 2e-05,
+      "loss": 1.1076,
+      "step": 14076
+    },
+    {
+      "epoch": 0.3350709321146339,
+      "grad_norm": 4.726289966677246,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 14077
+    },
+    {
+      "epoch": 0.33509473483766544,
+      "grad_norm": 4.648047074772306,
+      "learning_rate": 2e-05,
+      "loss": 0.5516,
+      "step": 14078
+    },
+    {
+      "epoch": 0.3351185375606969,
+      "grad_norm": 1.4911579879670274,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 14079
+    },
+    {
+      "epoch": 0.33514234028372847,
+      "grad_norm": 5.275144286384728,
+      "learning_rate": 2e-05,
+      "loss": 0.6583,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33516614300675995,
+      "grad_norm": 4.30432151623619,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 14081
+    },
+    {
+      "epoch": 0.3351899457297915,
+      "grad_norm": 2.3062808977575266,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 14082
+    },
+    {
+      "epoch": 0.335213748452823,
+      "grad_norm": 3.534664418935155,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 14083
+    },
+    {
+      "epoch": 0.3352375511758545,
+      "grad_norm": 2.243660751226305,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 14084
+    },
+    {
+      "epoch": 0.335261353898886,
+      "grad_norm": 9.004007916556379,
+      "learning_rate": 2e-05,
+      "loss": 0.9601,
+      "step": 14085
+    },
+    {
+      "epoch": 0.33528515662191755,
+      "grad_norm": 6.348528830089973,
+      "learning_rate": 2e-05,
+      "loss": 1.3961,
+      "step": 14086
+    },
+    {
+      "epoch": 0.3353089593449491,
+      "grad_norm": 5.746694517477059,
+      "learning_rate": 2e-05,
+      "loss": 0.9709,
+      "step": 14087
+    },
+    {
+      "epoch": 0.33533276206798057,
+      "grad_norm": 5.347320472299453,
+      "learning_rate": 2e-05,
+      "loss": 1.2036,
+      "step": 14088
+    },
+    {
+      "epoch": 0.3353565647910121,
+      "grad_norm": 3.7227825474129226,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 14089
+    },
+    {
+      "epoch": 0.3353803675140436,
+      "grad_norm": 5.232538158428705,
+      "learning_rate": 2e-05,
+      "loss": 0.8226,
+      "step": 14090
+    },
+    {
+      "epoch": 0.33540417023707514,
+      "grad_norm": 5.067283732245764,
+      "learning_rate": 2e-05,
+      "loss": 0.1248,
+      "step": 14091
+    },
+    {
+      "epoch": 0.3354279729601066,
+      "grad_norm": 7.1867250536120375,
+      "learning_rate": 2e-05,
+      "loss": 0.7064,
+      "step": 14092
+    },
+    {
+      "epoch": 0.33545177568313816,
+      "grad_norm": 3.4978472883929306,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 14093
+    },
+    {
+      "epoch": 0.33547557840616965,
+      "grad_norm": 4.470091126821873,
+      "learning_rate": 2e-05,
+      "loss": 0.569,
+      "step": 14094
+    },
+    {
+      "epoch": 0.3354993811292012,
+      "grad_norm": 3.262244774592394,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 14095
+    },
+    {
+      "epoch": 0.3355231838522327,
+      "grad_norm": 4.855429823361319,
+      "learning_rate": 2e-05,
+      "loss": 0.2084,
+      "step": 14096
+    },
+    {
+      "epoch": 0.3355469865752642,
+      "grad_norm": 2.2961610636241536,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 14097
+    },
+    {
+      "epoch": 0.3355707892982957,
+      "grad_norm": 3.6756370381638286,
+      "learning_rate": 2e-05,
+      "loss": 0.5153,
+      "step": 14098
+    },
+    {
+      "epoch": 0.33559459202132724,
+      "grad_norm": 4.151351051654834,
+      "learning_rate": 2e-05,
+      "loss": 0.6928,
+      "step": 14099
+    },
+    {
+      "epoch": 0.33561839474435873,
+      "grad_norm": 0.5167003603799194,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 14100
+    },
+    {
+      "epoch": 0.33564219746739027,
+      "grad_norm": 1.9697809524292278,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 14101
+    },
+    {
+      "epoch": 0.3356660001904218,
+      "grad_norm": 1.260727072913537,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 14102
+    },
+    {
+      "epoch": 0.3356898029134533,
+      "grad_norm": 2.2298908693197963,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 14103
+    },
+    {
+      "epoch": 0.33571360563648484,
+      "grad_norm": 4.201213717734452,
+      "learning_rate": 2e-05,
+      "loss": 0.5422,
+      "step": 14104
+    },
+    {
+      "epoch": 0.3357374083595163,
+      "grad_norm": 3.257492924741085,
+      "learning_rate": 2e-05,
+      "loss": 0.5613,
+      "step": 14105
+    },
+    {
+      "epoch": 0.33576121108254786,
+      "grad_norm": 6.713547132419327,
+      "learning_rate": 2e-05,
+      "loss": 1.8872,
+      "step": 14106
+    },
+    {
+      "epoch": 0.33578501380557935,
+      "grad_norm": 1.0000547442529248,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14107
+    },
+    {
+      "epoch": 0.3358088165286109,
+      "grad_norm": 5.024970411383084,
+      "learning_rate": 2e-05,
+      "loss": 0.8195,
+      "step": 14108
+    },
+    {
+      "epoch": 0.3358326192516424,
+      "grad_norm": 1.9979369635658089,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 14109
+    },
+    {
+      "epoch": 0.3358564219746739,
+      "grad_norm": 3.4655310190149025,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 14110
+    },
+    {
+      "epoch": 0.3358802246977054,
+      "grad_norm": 4.719445310289538,
+      "learning_rate": 2e-05,
+      "loss": 0.8105,
+      "step": 14111
+    },
+    {
+      "epoch": 0.33590402742073694,
+      "grad_norm": 5.701077423996575,
+      "learning_rate": 2e-05,
+      "loss": 1.1121,
+      "step": 14112
+    },
+    {
+      "epoch": 0.3359278301437684,
+      "grad_norm": 11.500381089526126,
+      "learning_rate": 2e-05,
+      "loss": 1.3396,
+      "step": 14113
+    },
+    {
+      "epoch": 0.33595163286679997,
+      "grad_norm": 3.4520844360121057,
+      "learning_rate": 2e-05,
+      "loss": 0.1685,
+      "step": 14114
+    },
+    {
+      "epoch": 0.33597543558983145,
+      "grad_norm": 5.113495887479925,
+      "learning_rate": 2e-05,
+      "loss": 0.4678,
+      "step": 14115
+    },
+    {
+      "epoch": 0.335999238312863,
+      "grad_norm": 3.0978670708248526,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 14116
+    },
+    {
+      "epoch": 0.33602304103589453,
+      "grad_norm": 6.405168425345403,
+      "learning_rate": 2e-05,
+      "loss": 0.6457,
+      "step": 14117
+    },
+    {
+      "epoch": 0.336046843758926,
+      "grad_norm": 3.8915666878406103,
+      "learning_rate": 2e-05,
+      "loss": 0.6346,
+      "step": 14118
+    },
+    {
+      "epoch": 0.33607064648195756,
+      "grad_norm": 3.7404581830868704,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 14119
+    },
+    {
+      "epoch": 0.33609444920498904,
+      "grad_norm": 1.927070486987837,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 14120
+    },
+    {
+      "epoch": 0.3361182519280206,
+      "grad_norm": 8.344100236066184,
+      "learning_rate": 2e-05,
+      "loss": 1.6551,
+      "step": 14121
+    },
+    {
+      "epoch": 0.33614205465105207,
+      "grad_norm": 5.405408947052731,
+      "learning_rate": 2e-05,
+      "loss": 0.6977,
+      "step": 14122
+    },
+    {
+      "epoch": 0.3361658573740836,
+      "grad_norm": 8.636389779488656,
+      "learning_rate": 2e-05,
+      "loss": 0.3591,
+      "step": 14123
+    },
+    {
+      "epoch": 0.3361896600971151,
+      "grad_norm": 16.990805771367146,
+      "learning_rate": 2e-05,
+      "loss": 0.7215,
+      "step": 14124
+    },
+    {
+      "epoch": 0.33621346282014664,
+      "grad_norm": 4.501014595123542,
+      "learning_rate": 2e-05,
+      "loss": 0.9061,
+      "step": 14125
+    },
+    {
+      "epoch": 0.3362372655431781,
+      "grad_norm": 0.3474233942457846,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 14126
+    },
+    {
+      "epoch": 0.33626106826620966,
+      "grad_norm": 13.513993456027011,
+      "learning_rate": 2e-05,
+      "loss": 0.8286,
+      "step": 14127
+    },
+    {
+      "epoch": 0.33628487098924115,
+      "grad_norm": 4.04759460803666,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 14128
+    },
+    {
+      "epoch": 0.3363086737122727,
+      "grad_norm": 5.142350432831724,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 14129
+    },
+    {
+      "epoch": 0.3363324764353042,
+      "grad_norm": 8.821324034459497,
+      "learning_rate": 2e-05,
+      "loss": 0.6615,
+      "step": 14130
+    },
+    {
+      "epoch": 0.3363562791583357,
+      "grad_norm": 7.974412034431168,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 14131
+    },
+    {
+      "epoch": 0.33638008188136725,
+      "grad_norm": 6.135125840569311,
+      "learning_rate": 2e-05,
+      "loss": 0.275,
+      "step": 14132
+    },
+    {
+      "epoch": 0.33640388460439874,
+      "grad_norm": 4.990537571711282,
+      "learning_rate": 2e-05,
+      "loss": 0.8286,
+      "step": 14133
+    },
+    {
+      "epoch": 0.3364276873274303,
+      "grad_norm": 2.268236252690855,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 14134
+    },
+    {
+      "epoch": 0.33645149005046177,
+      "grad_norm": 5.785178593585562,
+      "learning_rate": 2e-05,
+      "loss": 0.9853,
+      "step": 14135
+    },
+    {
+      "epoch": 0.3364752927734933,
+      "grad_norm": 4.378888178511987,
+      "learning_rate": 2e-05,
+      "loss": 0.603,
+      "step": 14136
+    },
+    {
+      "epoch": 0.3364990954965248,
+      "grad_norm": 5.888328499381664,
+      "learning_rate": 2e-05,
+      "loss": 0.2834,
+      "step": 14137
+    },
+    {
+      "epoch": 0.33652289821955633,
+      "grad_norm": 1.7587816109666523,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 14138
+    },
+    {
+      "epoch": 0.3365467009425878,
+      "grad_norm": 5.858732643317879,
+      "learning_rate": 2e-05,
+      "loss": 1.2567,
+      "step": 14139
+    },
+    {
+      "epoch": 0.33657050366561936,
+      "grad_norm": 11.14379371217708,
+      "learning_rate": 2e-05,
+      "loss": 0.3727,
+      "step": 14140
+    },
+    {
+      "epoch": 0.33659430638865084,
+      "grad_norm": 3.030871440715008,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 14141
+    },
+    {
+      "epoch": 0.3366181091116824,
+      "grad_norm": 4.446415846967364,
+      "learning_rate": 2e-05,
+      "loss": 0.1983,
+      "step": 14142
+    },
+    {
+      "epoch": 0.33664191183471387,
+      "grad_norm": 4.503350707994396,
+      "learning_rate": 2e-05,
+      "loss": 0.7868,
+      "step": 14143
+    },
+    {
+      "epoch": 0.3366657145577454,
+      "grad_norm": 11.05997107139046,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 14144
+    },
+    {
+      "epoch": 0.3366895172807769,
+      "grad_norm": 2.4784842693624696,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 14145
+    },
+    {
+      "epoch": 0.33671332000380844,
+      "grad_norm": 4.849947327216983,
+      "learning_rate": 2e-05,
+      "loss": 0.5216,
+      "step": 14146
+    },
+    {
+      "epoch": 0.33673712272684,
+      "grad_norm": 3.6343089184728763,
+      "learning_rate": 2e-05,
+      "loss": 0.1794,
+      "step": 14147
+    },
+    {
+      "epoch": 0.33676092544987146,
+      "grad_norm": 3.335985468093908,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 14148
+    },
+    {
+      "epoch": 0.336784728172903,
+      "grad_norm": 4.711979122223448,
+      "learning_rate": 2e-05,
+      "loss": 0.2922,
+      "step": 14149
+    },
+    {
+      "epoch": 0.3368085308959345,
+      "grad_norm": 2.217320428042117,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 14150
+    },
+    {
+      "epoch": 0.33683233361896603,
+      "grad_norm": 3.1700854744680935,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 14151
+    },
+    {
+      "epoch": 0.3368561363419975,
+      "grad_norm": 0.6769010479830728,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 14152
+    },
+    {
+      "epoch": 0.33687993906502905,
+      "grad_norm": 5.228722567774698,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 14153
+    },
+    {
+      "epoch": 0.33690374178806054,
+      "grad_norm": 6.996356070671661,
+      "learning_rate": 2e-05,
+      "loss": 1.2513,
+      "step": 14154
+    },
+    {
+      "epoch": 0.3369275445110921,
+      "grad_norm": 8.068556577768796,
+      "learning_rate": 2e-05,
+      "loss": 1.0665,
+      "step": 14155
+    },
+    {
+      "epoch": 0.33695134723412357,
+      "grad_norm": 2.399980013902845,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 14156
+    },
+    {
+      "epoch": 0.3369751499571551,
+      "grad_norm": 3.6219120590738414,
+      "learning_rate": 2e-05,
+      "loss": 0.5766,
+      "step": 14157
+    },
+    {
+      "epoch": 0.3369989526801866,
+      "grad_norm": 4.666334786701668,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 14158
+    },
+    {
+      "epoch": 0.33702275540321813,
+      "grad_norm": 14.515654073890303,
+      "learning_rate": 2e-05,
+      "loss": 0.6067,
+      "step": 14159
+    },
+    {
+      "epoch": 0.3370465581262496,
+      "grad_norm": 4.273859615920449,
+      "learning_rate": 2e-05,
+      "loss": 0.6184,
+      "step": 14160
+    },
+    {
+      "epoch": 0.33707036084928116,
+      "grad_norm": 5.099635665784742,
+      "learning_rate": 2e-05,
+      "loss": 0.6826,
+      "step": 14161
+    },
+    {
+      "epoch": 0.3370941635723127,
+      "grad_norm": 6.793297026656851,
+      "learning_rate": 2e-05,
+      "loss": 0.7126,
+      "step": 14162
+    },
+    {
+      "epoch": 0.3371179662953442,
+      "grad_norm": 3.2149838702943185,
+      "learning_rate": 2e-05,
+      "loss": 0.1592,
+      "step": 14163
+    },
+    {
+      "epoch": 0.3371417690183757,
+      "grad_norm": 0.9006396691543667,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14164
+    },
+    {
+      "epoch": 0.3371655717414072,
+      "grad_norm": 5.784502777817832,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 14165
+    },
+    {
+      "epoch": 0.33718937446443875,
+      "grad_norm": 4.805369346282235,
+      "learning_rate": 2e-05,
+      "loss": 0.8607,
+      "step": 14166
+    },
+    {
+      "epoch": 0.33721317718747024,
+      "grad_norm": 0.6963652904407388,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 14167
+    },
+    {
+      "epoch": 0.3372369799105018,
+      "grad_norm": 7.208488363354882,
+      "learning_rate": 2e-05,
+      "loss": 0.2333,
+      "step": 14168
+    },
+    {
+      "epoch": 0.33726078263353326,
+      "grad_norm": 12.739653484689832,
+      "learning_rate": 2e-05,
+      "loss": 0.948,
+      "step": 14169
+    },
+    {
+      "epoch": 0.3372845853565648,
+      "grad_norm": 10.642521389795087,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 14170
+    },
+    {
+      "epoch": 0.3373083880795963,
+      "grad_norm": 4.6494589113823555,
+      "learning_rate": 2e-05,
+      "loss": 0.4154,
+      "step": 14171
+    },
+    {
+      "epoch": 0.33733219080262783,
+      "grad_norm": 2.4997554520251386,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 14172
+    },
+    {
+      "epoch": 0.3373559935256593,
+      "grad_norm": 2.7362649476581455,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 14173
+    },
+    {
+      "epoch": 0.33737979624869086,
+      "grad_norm": 2.998946314128282,
+      "learning_rate": 2e-05,
+      "loss": 0.3688,
+      "step": 14174
+    },
+    {
+      "epoch": 0.33740359897172234,
+      "grad_norm": 4.587829370508888,
+      "learning_rate": 2e-05,
+      "loss": 0.7454,
+      "step": 14175
+    },
+    {
+      "epoch": 0.3374274016947539,
+      "grad_norm": 3.263999501105645,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 14176
+    },
+    {
+      "epoch": 0.3374512044177854,
+      "grad_norm": 0.16726705996821928,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 14177
+    },
+    {
+      "epoch": 0.3374750071408169,
+      "grad_norm": 0.5577569563284935,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 14178
+    },
+    {
+      "epoch": 0.33749880986384845,
+      "grad_norm": 1.3828427467087507,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 14179
+    },
+    {
+      "epoch": 0.33752261258687993,
+      "grad_norm": 6.855247971687757,
+      "learning_rate": 2e-05,
+      "loss": 0.6497,
+      "step": 14180
+    },
+    {
+      "epoch": 0.3375464153099115,
+      "grad_norm": 2.101963659160056,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 14181
+    },
+    {
+      "epoch": 0.33757021803294296,
+      "grad_norm": 6.61997255001665,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 14182
+    },
+    {
+      "epoch": 0.3375940207559745,
+      "grad_norm": 4.106967457685286,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 14183
+    },
+    {
+      "epoch": 0.337617823479006,
+      "grad_norm": 5.889550183596376,
+      "learning_rate": 2e-05,
+      "loss": 0.9279,
+      "step": 14184
+    },
+    {
+      "epoch": 0.3376416262020375,
+      "grad_norm": 16.79032136620732,
+      "learning_rate": 2e-05,
+      "loss": 0.4611,
+      "step": 14185
+    },
+    {
+      "epoch": 0.337665428925069,
+      "grad_norm": 3.2404443557903386,
+      "learning_rate": 2e-05,
+      "loss": 0.3563,
+      "step": 14186
+    },
+    {
+      "epoch": 0.33768923164810055,
+      "grad_norm": 5.243111066456938,
+      "learning_rate": 2e-05,
+      "loss": 0.3115,
+      "step": 14187
+    },
+    {
+      "epoch": 0.33771303437113204,
+      "grad_norm": 8.995065699052624,
+      "learning_rate": 2e-05,
+      "loss": 0.7753,
+      "step": 14188
+    },
+    {
+      "epoch": 0.3377368370941636,
+      "grad_norm": 2.3343590991350402,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 14189
+    },
+    {
+      "epoch": 0.33776063981719506,
+      "grad_norm": 1.7352572680354277,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 14190
+    },
+    {
+      "epoch": 0.3377844425402266,
+      "grad_norm": 0.907921422403408,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 14191
+    },
+    {
+      "epoch": 0.3378082452632581,
+      "grad_norm": 20.71235750353227,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 14192
+    },
+    {
+      "epoch": 0.33783204798628963,
+      "grad_norm": 0.1532261124622441,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 14193
+    },
+    {
+      "epoch": 0.33785585070932117,
+      "grad_norm": 2.4339445182817654,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 14194
+    },
+    {
+      "epoch": 0.33787965343235266,
+      "grad_norm": 6.311437255944877,
+      "learning_rate": 2e-05,
+      "loss": 1.0664,
+      "step": 14195
+    },
+    {
+      "epoch": 0.3379034561553842,
+      "grad_norm": 2.9671218677489466,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 14196
+    },
+    {
+      "epoch": 0.3379272588784157,
+      "grad_norm": 4.349921919434977,
+      "learning_rate": 2e-05,
+      "loss": 0.7328,
+      "step": 14197
+    },
+    {
+      "epoch": 0.3379510616014472,
+      "grad_norm": 5.003923686787123,
+      "learning_rate": 2e-05,
+      "loss": 0.7514,
+      "step": 14198
+    },
+    {
+      "epoch": 0.3379748643244787,
+      "grad_norm": 10.828796283635398,
+      "learning_rate": 2e-05,
+      "loss": 0.9984,
+      "step": 14199
+    },
+    {
+      "epoch": 0.33799866704751025,
+      "grad_norm": 5.937454345027348,
+      "learning_rate": 2e-05,
+      "loss": 1.1093,
+      "step": 14200
+    },
+    {
+      "epoch": 0.33802246977054173,
+      "grad_norm": 2.031658144402342,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 14201
+    },
+    {
+      "epoch": 0.3380462724935733,
+      "grad_norm": 9.854610536384627,
+      "learning_rate": 2e-05,
+      "loss": 1.1581,
+      "step": 14202
+    },
+    {
+      "epoch": 0.33807007521660476,
+      "grad_norm": 3.460248791432326,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 14203
+    },
+    {
+      "epoch": 0.3380938779396363,
+      "grad_norm": 4.745770553676226,
+      "learning_rate": 2e-05,
+      "loss": 0.5065,
+      "step": 14204
+    },
+    {
+      "epoch": 0.3381176806626678,
+      "grad_norm": 5.7045462840031815,
+      "learning_rate": 2e-05,
+      "loss": 0.8308,
+      "step": 14205
+    },
+    {
+      "epoch": 0.3381414833856993,
+      "grad_norm": 15.441815300488075,
+      "learning_rate": 2e-05,
+      "loss": 0.5754,
+      "step": 14206
+    },
+    {
+      "epoch": 0.3381652861087308,
+      "grad_norm": 1.558465950856167,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 14207
+    },
+    {
+      "epoch": 0.33818908883176235,
+      "grad_norm": 4.065656920439136,
+      "learning_rate": 2e-05,
+      "loss": 0.6525,
+      "step": 14208
+    },
+    {
+      "epoch": 0.3382128915547939,
+      "grad_norm": 6.069086859585791,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 14209
+    },
+    {
+      "epoch": 0.3382366942778254,
+      "grad_norm": 8.40175307874862,
+      "learning_rate": 2e-05,
+      "loss": 1.7056,
+      "step": 14210
+    },
+    {
+      "epoch": 0.3382604970008569,
+      "grad_norm": 3.0508644262511213,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 14211
+    },
+    {
+      "epoch": 0.3382842997238884,
+      "grad_norm": 3.3643700002674,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 14212
+    },
+    {
+      "epoch": 0.33830810244691994,
+      "grad_norm": 3.3025408989050535,
+      "learning_rate": 2e-05,
+      "loss": 0.5891,
+      "step": 14213
+    },
+    {
+      "epoch": 0.33833190516995143,
+      "grad_norm": 3.4750433190270402,
+      "learning_rate": 2e-05,
+      "loss": 0.2362,
+      "step": 14214
+    },
+    {
+      "epoch": 0.33835570789298297,
+      "grad_norm": 3.3372593477135504,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 14215
+    },
+    {
+      "epoch": 0.33837951061601446,
+      "grad_norm": 9.536276120215636,
+      "learning_rate": 2e-05,
+      "loss": 1.3639,
+      "step": 14216
+    },
+    {
+      "epoch": 0.338403313339046,
+      "grad_norm": 4.050979694812088,
+      "learning_rate": 2e-05,
+      "loss": 0.549,
+      "step": 14217
+    },
+    {
+      "epoch": 0.3384271160620775,
+      "grad_norm": 6.057557398288085,
+      "learning_rate": 2e-05,
+      "loss": 0.8047,
+      "step": 14218
+    },
+    {
+      "epoch": 0.338450918785109,
+      "grad_norm": 3.798974177007415,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 14219
+    },
+    {
+      "epoch": 0.3384747215081405,
+      "grad_norm": 1.086727748245186,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 14220
+    },
+    {
+      "epoch": 0.33849852423117205,
+      "grad_norm": 3.1928495243277633,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 14221
+    },
+    {
+      "epoch": 0.33852232695420353,
+      "grad_norm": 9.478461392171184,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 14222
+    },
+    {
+      "epoch": 0.3385461296772351,
+      "grad_norm": 3.013113184101676,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 14223
+    },
+    {
+      "epoch": 0.3385699324002666,
+      "grad_norm": 1.3294229911369895,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 14224
+    },
+    {
+      "epoch": 0.3385937351232981,
+      "grad_norm": 3.4284935993235854,
+      "learning_rate": 2e-05,
+      "loss": 0.2025,
+      "step": 14225
+    },
+    {
+      "epoch": 0.33861753784632964,
+      "grad_norm": 2.472113237688649,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 14226
+    },
+    {
+      "epoch": 0.3386413405693611,
+      "grad_norm": 3.2986195691570463,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 14227
+    },
+    {
+      "epoch": 0.33866514329239267,
+      "grad_norm": 4.821761281584423,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 14228
+    },
+    {
+      "epoch": 0.33868894601542415,
+      "grad_norm": 4.340913926453477,
+      "learning_rate": 2e-05,
+      "loss": 0.7943,
+      "step": 14229
+    },
+    {
+      "epoch": 0.3387127487384557,
+      "grad_norm": 0.9733967273208523,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 14230
+    },
+    {
+      "epoch": 0.3387365514614872,
+      "grad_norm": 3.735392402206094,
+      "learning_rate": 2e-05,
+      "loss": 0.4113,
+      "step": 14231
+    },
+    {
+      "epoch": 0.3387603541845187,
+      "grad_norm": 8.172028117720759,
+      "learning_rate": 2e-05,
+      "loss": 1.3713,
+      "step": 14232
+    },
+    {
+      "epoch": 0.3387841569075502,
+      "grad_norm": 3.352824022921375,
+      "learning_rate": 2e-05,
+      "loss": 0.2256,
+      "step": 14233
+    },
+    {
+      "epoch": 0.33880795963058175,
+      "grad_norm": 4.558264020470839,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 14234
+    },
+    {
+      "epoch": 0.33883176235361323,
+      "grad_norm": 3.966672362775698,
+      "learning_rate": 2e-05,
+      "loss": 0.5805,
+      "step": 14235
+    },
+    {
+      "epoch": 0.33885556507664477,
+      "grad_norm": 4.511192824471342,
+      "learning_rate": 2e-05,
+      "loss": 0.2068,
+      "step": 14236
+    },
+    {
+      "epoch": 0.33887936779967626,
+      "grad_norm": 1.4826669640696197,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 14237
+    },
+    {
+      "epoch": 0.3389031705227078,
+      "grad_norm": 4.423813071133294,
+      "learning_rate": 2e-05,
+      "loss": 0.6818,
+      "step": 14238
+    },
+    {
+      "epoch": 0.33892697324573934,
+      "grad_norm": 6.714686790823116,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 14239
+    },
+    {
+      "epoch": 0.3389507759687708,
+      "grad_norm": 1.3419408962131327,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 14240
+    },
+    {
+      "epoch": 0.33897457869180236,
+      "grad_norm": 3.331957175916743,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 14241
+    },
+    {
+      "epoch": 0.33899838141483385,
+      "grad_norm": 5.632587694059466,
+      "learning_rate": 2e-05,
+      "loss": 0.3831,
+      "step": 14242
+    },
+    {
+      "epoch": 0.3390221841378654,
+      "grad_norm": 11.870420450104142,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 14243
+    },
+    {
+      "epoch": 0.3390459868608969,
+      "grad_norm": 6.0161333429802,
+      "learning_rate": 2e-05,
+      "loss": 1.0394,
+      "step": 14244
+    },
+    {
+      "epoch": 0.3390697895839284,
+      "grad_norm": 7.572132740998886,
+      "learning_rate": 2e-05,
+      "loss": 1.0555,
+      "step": 14245
+    },
+    {
+      "epoch": 0.3390935923069599,
+      "grad_norm": 5.444343910749047,
+      "learning_rate": 2e-05,
+      "loss": 0.7772,
+      "step": 14246
+    },
+    {
+      "epoch": 0.33911739502999144,
+      "grad_norm": 15.26948821777459,
+      "learning_rate": 2e-05,
+      "loss": 0.5632,
+      "step": 14247
+    },
+    {
+      "epoch": 0.3391411977530229,
+      "grad_norm": 6.933080228287366,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 14248
+    },
+    {
+      "epoch": 0.33916500047605447,
+      "grad_norm": 13.994477423574372,
+      "learning_rate": 2e-05,
+      "loss": 0.4379,
+      "step": 14249
+    },
+    {
+      "epoch": 0.33918880319908595,
+      "grad_norm": 3.833155990206913,
+      "learning_rate": 2e-05,
+      "loss": 0.7453,
+      "step": 14250
+    },
+    {
+      "epoch": 0.3392126059221175,
+      "grad_norm": 4.340023789548702,
+      "learning_rate": 2e-05,
+      "loss": 0.5291,
+      "step": 14251
+    },
+    {
+      "epoch": 0.339236408645149,
+      "grad_norm": 2.6709929310256673,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 14252
+    },
+    {
+      "epoch": 0.3392602113681805,
+      "grad_norm": 5.569222733992789,
+      "learning_rate": 2e-05,
+      "loss": 0.6607,
+      "step": 14253
+    },
+    {
+      "epoch": 0.33928401409121206,
+      "grad_norm": 0.6867539212234288,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 14254
+    },
+    {
+      "epoch": 0.33930781681424355,
+      "grad_norm": 4.243320899234768,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 14255
+    },
+    {
+      "epoch": 0.3393316195372751,
+      "grad_norm": 1.5265528860267996,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 14256
+    },
+    {
+      "epoch": 0.33935542226030657,
+      "grad_norm": 2.0344191210648734,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 14257
+    },
+    {
+      "epoch": 0.3393792249833381,
+      "grad_norm": 2.3385282102186085,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 14258
+    },
+    {
+      "epoch": 0.3394030277063696,
+      "grad_norm": 1.3463278024290182,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 14259
+    },
+    {
+      "epoch": 0.33942683042940114,
+      "grad_norm": 4.260708317659932,
+      "learning_rate": 2e-05,
+      "loss": 0.966,
+      "step": 14260
+    },
+    {
+      "epoch": 0.3394506331524326,
+      "grad_norm": 7.825287813102669,
+      "learning_rate": 2e-05,
+      "loss": 0.3583,
+      "step": 14261
+    },
+    {
+      "epoch": 0.33947443587546416,
+      "grad_norm": 3.5827624206899884,
+      "learning_rate": 2e-05,
+      "loss": 0.2315,
+      "step": 14262
+    },
+    {
+      "epoch": 0.33949823859849565,
+      "grad_norm": 1.0525500906003766,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 14263
+    },
+    {
+      "epoch": 0.3395220413215272,
+      "grad_norm": 2.87309079417315,
+      "learning_rate": 2e-05,
+      "loss": 0.2522,
+      "step": 14264
+    },
+    {
+      "epoch": 0.3395458440445587,
+      "grad_norm": 6.027169248339943,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 14265
+    },
+    {
+      "epoch": 0.3395696467675902,
+      "grad_norm": 5.154919625490065,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 14266
+    },
+    {
+      "epoch": 0.3395934494906217,
+      "grad_norm": 4.014028363474218,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 14267
+    },
+    {
+      "epoch": 0.33961725221365324,
+      "grad_norm": 11.057152374435539,
+      "learning_rate": 2e-05,
+      "loss": 0.5413,
+      "step": 14268
+    },
+    {
+      "epoch": 0.3396410549366848,
+      "grad_norm": 4.67948602582001,
+      "learning_rate": 2e-05,
+      "loss": 0.8241,
+      "step": 14269
+    },
+    {
+      "epoch": 0.33966485765971627,
+      "grad_norm": 9.127145475423031,
+      "learning_rate": 2e-05,
+      "loss": 1.005,
+      "step": 14270
+    },
+    {
+      "epoch": 0.3396886603827478,
+      "grad_norm": 4.683944561697979,
+      "learning_rate": 2e-05,
+      "loss": 0.6898,
+      "step": 14271
+    },
+    {
+      "epoch": 0.3397124631057793,
+      "grad_norm": 1.2115887166241759,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 14272
+    },
+    {
+      "epoch": 0.33973626582881084,
+      "grad_norm": 3.6283666344716483,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 14273
+    },
+    {
+      "epoch": 0.3397600685518423,
+      "grad_norm": 9.900786887015402,
+      "learning_rate": 2e-05,
+      "loss": 1.444,
+      "step": 14274
+    },
+    {
+      "epoch": 0.33978387127487386,
+      "grad_norm": 1.6329657514098082,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 14275
+    },
+    {
+      "epoch": 0.33980767399790535,
+      "grad_norm": 3.6896307718105934,
+      "learning_rate": 2e-05,
+      "loss": 0.4915,
+      "step": 14276
+    },
+    {
+      "epoch": 0.3398314767209369,
+      "grad_norm": 19.045531146167146,
+      "learning_rate": 2e-05,
+      "loss": 0.7718,
+      "step": 14277
+    },
+    {
+      "epoch": 0.33985527944396837,
+      "grad_norm": 4.694873019049861,
+      "learning_rate": 2e-05,
+      "loss": 0.5976,
+      "step": 14278
+    },
+    {
+      "epoch": 0.3398790821669999,
+      "grad_norm": 4.194646149652189,
+      "learning_rate": 2e-05,
+      "loss": 0.8303,
+      "step": 14279
+    },
+    {
+      "epoch": 0.3399028848900314,
+      "grad_norm": 1.8041323129972735,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 14280
+    },
+    {
+      "epoch": 0.33992668761306294,
+      "grad_norm": 6.475934338531204,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 14281
+    },
+    {
+      "epoch": 0.3399504903360944,
+      "grad_norm": 3.7186127783958325,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 14282
+    },
+    {
+      "epoch": 0.33997429305912596,
+      "grad_norm": 3.003500773569188,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 14283
+    },
+    {
+      "epoch": 0.3399980957821575,
+      "grad_norm": 10.012667313996875,
+      "learning_rate": 2e-05,
+      "loss": 0.74,
+      "step": 14284
+    },
+    {
+      "epoch": 0.340021898505189,
+      "grad_norm": 3.48946128964399,
+      "learning_rate": 2e-05,
+      "loss": 0.3755,
+      "step": 14285
+    },
+    {
+      "epoch": 0.34004570122822053,
+      "grad_norm": 3.4676529776230853,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 14286
+    },
+    {
+      "epoch": 0.340069503951252,
+      "grad_norm": 0.6324610054788409,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 14287
+    },
+    {
+      "epoch": 0.34009330667428356,
+      "grad_norm": 1.9343783337399696,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 14288
+    },
+    {
+      "epoch": 0.34011710939731504,
+      "grad_norm": 5.411782464591615,
+      "learning_rate": 2e-05,
+      "loss": 0.6308,
+      "step": 14289
+    },
+    {
+      "epoch": 0.3401409121203466,
+      "grad_norm": 5.891002058435499,
+      "learning_rate": 2e-05,
+      "loss": 0.8231,
+      "step": 14290
+    },
+    {
+      "epoch": 0.34016471484337807,
+      "grad_norm": 1.6361878494180875,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 14291
+    },
+    {
+      "epoch": 0.3401885175664096,
+      "grad_norm": 5.394563297176229,
+      "learning_rate": 2e-05,
+      "loss": 0.8124,
+      "step": 14292
+    },
+    {
+      "epoch": 0.3402123202894411,
+      "grad_norm": 10.862793098225191,
+      "learning_rate": 2e-05,
+      "loss": 0.7984,
+      "step": 14293
+    },
+    {
+      "epoch": 0.34023612301247264,
+      "grad_norm": 7.47318482965293,
+      "learning_rate": 2e-05,
+      "loss": 0.6979,
+      "step": 14294
+    },
+    {
+      "epoch": 0.3402599257355041,
+      "grad_norm": 3.1675085491360995,
+      "learning_rate": 2e-05,
+      "loss": 0.5834,
+      "step": 14295
+    },
+    {
+      "epoch": 0.34028372845853566,
+      "grad_norm": 2.9874204228698478,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 14296
+    },
+    {
+      "epoch": 0.34030753118156715,
+      "grad_norm": 4.459992507490253,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 14297
+    },
+    {
+      "epoch": 0.3403313339045987,
+      "grad_norm": 3.2524242531617604,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 14298
+    },
+    {
+      "epoch": 0.34035513662763023,
+      "grad_norm": 6.630281529079569,
+      "learning_rate": 2e-05,
+      "loss": 0.4703,
+      "step": 14299
+    },
+    {
+      "epoch": 0.3403789393506617,
+      "grad_norm": 2.7622770620635664,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 14300
+    },
+    {
+      "epoch": 0.34040274207369325,
+      "grad_norm": 3.729233320988564,
+      "learning_rate": 2e-05,
+      "loss": 0.7207,
+      "step": 14301
+    },
+    {
+      "epoch": 0.34042654479672474,
+      "grad_norm": 25.134060420104287,
+      "learning_rate": 2e-05,
+      "loss": 0.8756,
+      "step": 14302
+    },
+    {
+      "epoch": 0.3404503475197563,
+      "grad_norm": 2.522194348987811,
+      "learning_rate": 2e-05,
+      "loss": 0.1683,
+      "step": 14303
+    },
+    {
+      "epoch": 0.34047415024278777,
+      "grad_norm": 2.534114679832585,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 14304
+    },
+    {
+      "epoch": 0.3404979529658193,
+      "grad_norm": 5.872994073437201,
+      "learning_rate": 2e-05,
+      "loss": 0.2925,
+      "step": 14305
+    },
+    {
+      "epoch": 0.3405217556888508,
+      "grad_norm": 4.0161007222798055,
+      "learning_rate": 2e-05,
+      "loss": 0.1874,
+      "step": 14306
+    },
+    {
+      "epoch": 0.34054555841188233,
+      "grad_norm": 4.438470286970188,
+      "learning_rate": 2e-05,
+      "loss": 0.9182,
+      "step": 14307
+    },
+    {
+      "epoch": 0.3405693611349138,
+      "grad_norm": 6.800464067389984,
+      "learning_rate": 2e-05,
+      "loss": 0.7548,
+      "step": 14308
+    },
+    {
+      "epoch": 0.34059316385794536,
+      "grad_norm": 10.827323130413738,
+      "learning_rate": 2e-05,
+      "loss": 0.6046,
+      "step": 14309
+    },
+    {
+      "epoch": 0.34061696658097684,
+      "grad_norm": 13.199312418068578,
+      "learning_rate": 2e-05,
+      "loss": 0.9314,
+      "step": 14310
+    },
+    {
+      "epoch": 0.3406407693040084,
+      "grad_norm": 1.034794941257117,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 14311
+    },
+    {
+      "epoch": 0.34066457202703987,
+      "grad_norm": 7.599028574228848,
+      "learning_rate": 2e-05,
+      "loss": 0.9523,
+      "step": 14312
+    },
+    {
+      "epoch": 0.3406883747500714,
+      "grad_norm": 0.7749363159203838,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 14313
+    },
+    {
+      "epoch": 0.34071217747310295,
+      "grad_norm": 6.459849496558884,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 14314
+    },
+    {
+      "epoch": 0.34073598019613444,
+      "grad_norm": 5.459165722445264,
+      "learning_rate": 2e-05,
+      "loss": 0.6929,
+      "step": 14315
+    },
+    {
+      "epoch": 0.340759782919166,
+      "grad_norm": 6.037332474240764,
+      "learning_rate": 2e-05,
+      "loss": 0.213,
+      "step": 14316
+    },
+    {
+      "epoch": 0.34078358564219746,
+      "grad_norm": 2.6273731090294192,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 14317
+    },
+    {
+      "epoch": 0.340807388365229,
+      "grad_norm": 6.7925018607934415,
+      "learning_rate": 2e-05,
+      "loss": 0.296,
+      "step": 14318
+    },
+    {
+      "epoch": 0.3408311910882605,
+      "grad_norm": 10.800587645860114,
+      "learning_rate": 2e-05,
+      "loss": 0.9543,
+      "step": 14319
+    },
+    {
+      "epoch": 0.34085499381129203,
+      "grad_norm": 8.73564173227999,
+      "learning_rate": 2e-05,
+      "loss": 0.9137,
+      "step": 14320
+    },
+    {
+      "epoch": 0.3408787965343235,
+      "grad_norm": 1.63725239461631,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 14321
+    },
+    {
+      "epoch": 0.34090259925735505,
+      "grad_norm": 3.209299669085477,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 14322
+    },
+    {
+      "epoch": 0.34092640198038654,
+      "grad_norm": 5.060635349359921,
+      "learning_rate": 2e-05,
+      "loss": 0.5189,
+      "step": 14323
+    },
+    {
+      "epoch": 0.3409502047034181,
+      "grad_norm": 5.845115141698772,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 14324
+    },
+    {
+      "epoch": 0.34097400742644957,
+      "grad_norm": 6.594204204666578,
+      "learning_rate": 2e-05,
+      "loss": 0.2836,
+      "step": 14325
+    },
+    {
+      "epoch": 0.3409978101494811,
+      "grad_norm": 1.7852620185652406,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 14326
+    },
+    {
+      "epoch": 0.3410216128725126,
+      "grad_norm": 5.727094156576576,
+      "learning_rate": 2e-05,
+      "loss": 0.6586,
+      "step": 14327
+    },
+    {
+      "epoch": 0.34104541559554413,
+      "grad_norm": 2.565015157860602,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 14328
+    },
+    {
+      "epoch": 0.3410692183185756,
+      "grad_norm": 1.5107915446924298,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 14329
+    },
+    {
+      "epoch": 0.34109302104160716,
+      "grad_norm": 3.8993903975839825,
+      "learning_rate": 2e-05,
+      "loss": 0.1636,
+      "step": 14330
+    },
+    {
+      "epoch": 0.3411168237646387,
+      "grad_norm": 0.3920510670704945,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 14331
+    },
+    {
+      "epoch": 0.3411406264876702,
+      "grad_norm": 4.525324587739059,
+      "learning_rate": 2e-05,
+      "loss": 0.2484,
+      "step": 14332
+    },
+    {
+      "epoch": 0.3411644292107017,
+      "grad_norm": 4.338068628474466,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 14333
+    },
+    {
+      "epoch": 0.3411882319337332,
+      "grad_norm": 5.956851991611263,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 14334
+    },
+    {
+      "epoch": 0.34121203465676475,
+      "grad_norm": 1.618066698366794,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 14335
+    },
+    {
+      "epoch": 0.34123583737979624,
+      "grad_norm": 4.323167514397405,
+      "learning_rate": 2e-05,
+      "loss": 0.8375,
+      "step": 14336
+    },
+    {
+      "epoch": 0.3412596401028278,
+      "grad_norm": 2.827003015511564,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 14337
+    },
+    {
+      "epoch": 0.34128344282585926,
+      "grad_norm": 6.802091177627869,
+      "learning_rate": 2e-05,
+      "loss": 0.3034,
+      "step": 14338
+    },
+    {
+      "epoch": 0.3413072455488908,
+      "grad_norm": 7.796173472445037,
+      "learning_rate": 2e-05,
+      "loss": 0.3418,
+      "step": 14339
+    },
+    {
+      "epoch": 0.3413310482719223,
+      "grad_norm": 0.8498499785235359,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14340
+    },
+    {
+      "epoch": 0.34135485099495383,
+      "grad_norm": 4.9428007918026875,
+      "learning_rate": 2e-05,
+      "loss": 0.4118,
+      "step": 14341
+    },
+    {
+      "epoch": 0.3413786537179853,
+      "grad_norm": 6.578042210772343,
+      "learning_rate": 2e-05,
+      "loss": 0.279,
+      "step": 14342
+    },
+    {
+      "epoch": 0.34140245644101685,
+      "grad_norm": 1.301366769566844,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 14343
+    },
+    {
+      "epoch": 0.34142625916404834,
+      "grad_norm": 3.4987803433261293,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 14344
+    },
+    {
+      "epoch": 0.3414500618870799,
+      "grad_norm": 4.690460011918396,
+      "learning_rate": 2e-05,
+      "loss": 0.534,
+      "step": 14345
+    },
+    {
+      "epoch": 0.3414738646101114,
+      "grad_norm": 4.896328404381847,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 14346
+    },
+    {
+      "epoch": 0.3414976673331429,
+      "grad_norm": 4.210487209934236,
+      "learning_rate": 2e-05,
+      "loss": 0.7696,
+      "step": 14347
+    },
+    {
+      "epoch": 0.34152147005617445,
+      "grad_norm": 4.3486163628096515,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 14348
+    },
+    {
+      "epoch": 0.34154527277920593,
+      "grad_norm": 0.5175993901400809,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 14349
+    },
+    {
+      "epoch": 0.3415690755022375,
+      "grad_norm": 2.66376879802665,
+      "learning_rate": 2e-05,
+      "loss": 0.5835,
+      "step": 14350
+    },
+    {
+      "epoch": 0.34159287822526896,
+      "grad_norm": 8.099695652329423,
+      "learning_rate": 2e-05,
+      "loss": 0.379,
+      "step": 14351
+    },
+    {
+      "epoch": 0.3416166809483005,
+      "grad_norm": 2.359977114946174,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 14352
+    },
+    {
+      "epoch": 0.341640483671332,
+      "grad_norm": 4.265958306628257,
+      "learning_rate": 2e-05,
+      "loss": 0.2462,
+      "step": 14353
+    },
+    {
+      "epoch": 0.3416642863943635,
+      "grad_norm": 10.558238328274982,
+      "learning_rate": 2e-05,
+      "loss": 1.1482,
+      "step": 14354
+    },
+    {
+      "epoch": 0.341688089117395,
+      "grad_norm": 7.999981018282161,
+      "learning_rate": 2e-05,
+      "loss": 0.9773,
+      "step": 14355
+    },
+    {
+      "epoch": 0.34171189184042655,
+      "grad_norm": 3.6220540626551263,
+      "learning_rate": 2e-05,
+      "loss": 0.4973,
+      "step": 14356
+    },
+    {
+      "epoch": 0.34173569456345804,
+      "grad_norm": 4.764210077701206,
+      "learning_rate": 2e-05,
+      "loss": 0.8059,
+      "step": 14357
+    },
+    {
+      "epoch": 0.3417594972864896,
+      "grad_norm": 5.0992470776851055,
+      "learning_rate": 2e-05,
+      "loss": 0.6233,
+      "step": 14358
+    },
+    {
+      "epoch": 0.34178330000952106,
+      "grad_norm": 0.7857539767656603,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 14359
+    },
+    {
+      "epoch": 0.3418071027325526,
+      "grad_norm": 3.102792943502312,
+      "learning_rate": 2e-05,
+      "loss": 0.5437,
+      "step": 14360
+    },
+    {
+      "epoch": 0.34183090545558414,
+      "grad_norm": 8.351411364908921,
+      "learning_rate": 2e-05,
+      "loss": 1.0956,
+      "step": 14361
+    },
+    {
+      "epoch": 0.34185470817861563,
+      "grad_norm": 6.1173129091187235,
+      "learning_rate": 2e-05,
+      "loss": 0.7622,
+      "step": 14362
+    },
+    {
+      "epoch": 0.34187851090164717,
+      "grad_norm": 3.8002453932693494,
+      "learning_rate": 2e-05,
+      "loss": 0.8616,
+      "step": 14363
+    },
+    {
+      "epoch": 0.34190231362467866,
+      "grad_norm": 6.939954455155325,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 14364
+    },
+    {
+      "epoch": 0.3419261163477102,
+      "grad_norm": 4.099521729917452,
+      "learning_rate": 2e-05,
+      "loss": 0.4046,
+      "step": 14365
+    },
+    {
+      "epoch": 0.3419499190707417,
+      "grad_norm": 4.510289170980126,
+      "learning_rate": 2e-05,
+      "loss": 0.6108,
+      "step": 14366
+    },
+    {
+      "epoch": 0.3419737217937732,
+      "grad_norm": 4.839394432110614,
+      "learning_rate": 2e-05,
+      "loss": 0.5549,
+      "step": 14367
+    },
+    {
+      "epoch": 0.3419975245168047,
+      "grad_norm": 5.30914157931682,
+      "learning_rate": 2e-05,
+      "loss": 0.2137,
+      "step": 14368
+    },
+    {
+      "epoch": 0.34202132723983625,
+      "grad_norm": 0.7394763027513057,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 14369
+    },
+    {
+      "epoch": 0.34204512996286773,
+      "grad_norm": 2.5057349766785024,
+      "learning_rate": 2e-05,
+      "loss": 0.2479,
+      "step": 14370
+    },
+    {
+      "epoch": 0.3420689326858993,
+      "grad_norm": 6.570105799050901,
+      "learning_rate": 2e-05,
+      "loss": 0.693,
+      "step": 14371
+    },
+    {
+      "epoch": 0.34209273540893076,
+      "grad_norm": 4.034336833591619,
+      "learning_rate": 2e-05,
+      "loss": 0.5263,
+      "step": 14372
+    },
+    {
+      "epoch": 0.3421165381319623,
+      "grad_norm": 5.9436843179901535,
+      "learning_rate": 2e-05,
+      "loss": 0.2359,
+      "step": 14373
+    },
+    {
+      "epoch": 0.3421403408549938,
+      "grad_norm": 1.8062410584748676,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 14374
+    },
+    {
+      "epoch": 0.3421641435780253,
+      "grad_norm": 2.9529912386148904,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 14375
+    },
+    {
+      "epoch": 0.34218794630105687,
+      "grad_norm": 1.286139375711384,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 14376
+    },
+    {
+      "epoch": 0.34221174902408835,
+      "grad_norm": 7.1941894619676345,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 14377
+    },
+    {
+      "epoch": 0.3422355517471199,
+      "grad_norm": 9.644328485494244,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 14378
+    },
+    {
+      "epoch": 0.3422593544701514,
+      "grad_norm": 6.2044205879108265,
+      "learning_rate": 2e-05,
+      "loss": 1.1358,
+      "step": 14379
+    },
+    {
+      "epoch": 0.3422831571931829,
+      "grad_norm": 4.323912152787653,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 14380
+    },
+    {
+      "epoch": 0.3423069599162144,
+      "grad_norm": 5.713163784459081,
+      "learning_rate": 2e-05,
+      "loss": 0.3173,
+      "step": 14381
+    },
+    {
+      "epoch": 0.34233076263924594,
+      "grad_norm": 3.2858999516246667,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 14382
+    },
+    {
+      "epoch": 0.34235456536227743,
+      "grad_norm": 5.811209209370194,
+      "learning_rate": 2e-05,
+      "loss": 1.0124,
+      "step": 14383
+    },
+    {
+      "epoch": 0.34237836808530897,
+      "grad_norm": 6.031870069665033,
+      "learning_rate": 2e-05,
+      "loss": 0.3113,
+      "step": 14384
+    },
+    {
+      "epoch": 0.34240217080834046,
+      "grad_norm": 0.9504011795885007,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 14385
+    },
+    {
+      "epoch": 0.342425973531372,
+      "grad_norm": 2.402698357608539,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 14386
+    },
+    {
+      "epoch": 0.3424497762544035,
+      "grad_norm": 6.724056547679917,
+      "learning_rate": 2e-05,
+      "loss": 1.1003,
+      "step": 14387
+    },
+    {
+      "epoch": 0.342473578977435,
+      "grad_norm": 2.639326512102448,
+      "learning_rate": 2e-05,
+      "loss": 0.4322,
+      "step": 14388
+    },
+    {
+      "epoch": 0.3424973817004665,
+      "grad_norm": 5.676081874622279,
+      "learning_rate": 2e-05,
+      "loss": 1.0684,
+      "step": 14389
+    },
+    {
+      "epoch": 0.34252118442349805,
+      "grad_norm": 6.19736758010597,
+      "learning_rate": 2e-05,
+      "loss": 0.7899,
+      "step": 14390
+    },
+    {
+      "epoch": 0.3425449871465296,
+      "grad_norm": 2.504414882751347,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 14391
+    },
+    {
+      "epoch": 0.3425687898695611,
+      "grad_norm": 6.670519996531511,
+      "learning_rate": 2e-05,
+      "loss": 1.2551,
+      "step": 14392
+    },
+    {
+      "epoch": 0.3425925925925926,
+      "grad_norm": 1.8829115030445565,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 14393
+    },
+    {
+      "epoch": 0.3426163953156241,
+      "grad_norm": 3.6162771706534462,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 14394
+    },
+    {
+      "epoch": 0.34264019803865564,
+      "grad_norm": 3.746868528213659,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 14395
+    },
+    {
+      "epoch": 0.3426640007616871,
+      "grad_norm": 0.630859921271551,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 14396
+    },
+    {
+      "epoch": 0.34268780348471867,
+      "grad_norm": 2.726028710215074,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 14397
+    },
+    {
+      "epoch": 0.34271160620775015,
+      "grad_norm": 5.659760170932597,
+      "learning_rate": 2e-05,
+      "loss": 0.3927,
+      "step": 14398
+    },
+    {
+      "epoch": 0.3427354089307817,
+      "grad_norm": 5.446097763514871,
+      "learning_rate": 2e-05,
+      "loss": 1.1796,
+      "step": 14399
+    },
+    {
+      "epoch": 0.3427592116538132,
+      "grad_norm": 6.0919655489824995,
+      "learning_rate": 2e-05,
+      "loss": 0.9727,
+      "step": 14400
+    },
+    {
+      "epoch": 0.3427830143768447,
+      "grad_norm": 2.188985435916808,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 14401
+    },
+    {
+      "epoch": 0.3428068170998762,
+      "grad_norm": 5.0656005008926,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 14402
+    },
+    {
+      "epoch": 0.34283061982290775,
+      "grad_norm": 9.991028518136961,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 14403
+    },
+    {
+      "epoch": 0.34285442254593923,
+      "grad_norm": 2.087888814758669,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 14404
+    },
+    {
+      "epoch": 0.34287822526897077,
+      "grad_norm": 5.383457707424385,
+      "learning_rate": 2e-05,
+      "loss": 0.516,
+      "step": 14405
+    },
+    {
+      "epoch": 0.3429020279920023,
+      "grad_norm": 6.967831952258948,
+      "learning_rate": 2e-05,
+      "loss": 1.1636,
+      "step": 14406
+    },
+    {
+      "epoch": 0.3429258307150338,
+      "grad_norm": 6.884767371621791,
+      "learning_rate": 2e-05,
+      "loss": 0.3377,
+      "step": 14407
+    },
+    {
+      "epoch": 0.34294963343806534,
+      "grad_norm": 1.3883010591841078,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 14408
+    },
+    {
+      "epoch": 0.3429734361610968,
+      "grad_norm": 7.01231117851198,
+      "learning_rate": 2e-05,
+      "loss": 0.8221,
+      "step": 14409
+    },
+    {
+      "epoch": 0.34299723888412836,
+      "grad_norm": 8.375529841639565,
+      "learning_rate": 2e-05,
+      "loss": 0.2899,
+      "step": 14410
+    },
+    {
+      "epoch": 0.34302104160715985,
+      "grad_norm": 2.212238307877066,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 14411
+    },
+    {
+      "epoch": 0.3430448443301914,
+      "grad_norm": 3.8595827149994166,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 14412
+    },
+    {
+      "epoch": 0.3430686470532229,
+      "grad_norm": 3.803531306160833,
+      "learning_rate": 2e-05,
+      "loss": 0.3446,
+      "step": 14413
+    },
+    {
+      "epoch": 0.3430924497762544,
+      "grad_norm": 3.699847938342717,
+      "learning_rate": 2e-05,
+      "loss": 0.7663,
+      "step": 14414
+    },
+    {
+      "epoch": 0.3431162524992859,
+      "grad_norm": 2.3450672920568305,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 14415
+    },
+    {
+      "epoch": 0.34314005522231744,
+      "grad_norm": 5.8872500599747895,
+      "learning_rate": 2e-05,
+      "loss": 0.1465,
+      "step": 14416
+    },
+    {
+      "epoch": 0.3431638579453489,
+      "grad_norm": 7.203786814109833,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 14417
+    },
+    {
+      "epoch": 0.34318766066838047,
+      "grad_norm": 11.044030602237864,
+      "learning_rate": 2e-05,
+      "loss": 1.3449,
+      "step": 14418
+    },
+    {
+      "epoch": 0.34321146339141195,
+      "grad_norm": 2.8938205677633726,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 14419
+    },
+    {
+      "epoch": 0.3432352661144435,
+      "grad_norm": 23.71496293561554,
+      "learning_rate": 2e-05,
+      "loss": 0.6045,
+      "step": 14420
+    },
+    {
+      "epoch": 0.34325906883747503,
+      "grad_norm": 0.6506937432260984,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 14421
+    },
+    {
+      "epoch": 0.3432828715605065,
+      "grad_norm": 1.0577866932484574,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 14422
+    },
+    {
+      "epoch": 0.34330667428353806,
+      "grad_norm": 5.32467452035484,
+      "learning_rate": 2e-05,
+      "loss": 0.3214,
+      "step": 14423
+    },
+    {
+      "epoch": 0.34333047700656955,
+      "grad_norm": 1.7774257358370535,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 14424
+    },
+    {
+      "epoch": 0.3433542797296011,
+      "grad_norm": 0.8175846074520349,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 14425
+    },
+    {
+      "epoch": 0.34337808245263257,
+      "grad_norm": 1.6273875067993198,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 14426
+    },
+    {
+      "epoch": 0.3434018851756641,
+      "grad_norm": 4.685211208229464,
+      "learning_rate": 2e-05,
+      "loss": 0.5833,
+      "step": 14427
+    },
+    {
+      "epoch": 0.3434256878986956,
+      "grad_norm": 6.9733668609329715,
+      "learning_rate": 2e-05,
+      "loss": 1.4181,
+      "step": 14428
+    },
+    {
+      "epoch": 0.34344949062172714,
+      "grad_norm": 1.3876957720869694,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 14429
+    },
+    {
+      "epoch": 0.3434732933447586,
+      "grad_norm": 10.371065198345912,
+      "learning_rate": 2e-05,
+      "loss": 1.0367,
+      "step": 14430
+    },
+    {
+      "epoch": 0.34349709606779016,
+      "grad_norm": 5.679866882469451,
+      "learning_rate": 2e-05,
+      "loss": 0.2538,
+      "step": 14431
+    },
+    {
+      "epoch": 0.34352089879082165,
+      "grad_norm": 4.141003891524345,
+      "learning_rate": 2e-05,
+      "loss": 0.2567,
+      "step": 14432
+    },
+    {
+      "epoch": 0.3435447015138532,
+      "grad_norm": 3.9689267364833434,
+      "learning_rate": 2e-05,
+      "loss": 0.7062,
+      "step": 14433
+    },
+    {
+      "epoch": 0.3435685042368847,
+      "grad_norm": 4.786519177411101,
+      "learning_rate": 2e-05,
+      "loss": 0.2205,
+      "step": 14434
+    },
+    {
+      "epoch": 0.3435923069599162,
+      "grad_norm": 5.948436030505957,
+      "learning_rate": 2e-05,
+      "loss": 0.3453,
+      "step": 14435
+    },
+    {
+      "epoch": 0.34361610968294776,
+      "grad_norm": 5.707999878176385,
+      "learning_rate": 2e-05,
+      "loss": 0.3578,
+      "step": 14436
+    },
+    {
+      "epoch": 0.34363991240597924,
+      "grad_norm": 0.17451960842678318,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 14437
+    },
+    {
+      "epoch": 0.3436637151290108,
+      "grad_norm": 1.5362133278830759,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 14438
+    },
+    {
+      "epoch": 0.34368751785204227,
+      "grad_norm": 4.024686692319518,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 14439
+    },
+    {
+      "epoch": 0.3437113205750738,
+      "grad_norm": 3.823286517860564,
+      "learning_rate": 2e-05,
+      "loss": 0.6885,
+      "step": 14440
+    },
+    {
+      "epoch": 0.3437351232981053,
+      "grad_norm": 5.490635285921082,
+      "learning_rate": 2e-05,
+      "loss": 0.4865,
+      "step": 14441
+    },
+    {
+      "epoch": 0.34375892602113683,
+      "grad_norm": 2.6531962605670176,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 14442
+    },
+    {
+      "epoch": 0.3437827287441683,
+      "grad_norm": 7.71096026462646,
+      "learning_rate": 2e-05,
+      "loss": 0.2653,
+      "step": 14443
+    },
+    {
+      "epoch": 0.34380653146719986,
+      "grad_norm": 4.558239037078468,
+      "learning_rate": 2e-05,
+      "loss": 0.7217,
+      "step": 14444
+    },
+    {
+      "epoch": 0.34383033419023135,
+      "grad_norm": 11.55175044738961,
+      "learning_rate": 2e-05,
+      "loss": 1.0787,
+      "step": 14445
+    },
+    {
+      "epoch": 0.3438541369132629,
+      "grad_norm": 10.217737028681723,
+      "learning_rate": 2e-05,
+      "loss": 0.4581,
+      "step": 14446
+    },
+    {
+      "epoch": 0.34387793963629437,
+      "grad_norm": 2.995594638834343,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 14447
+    },
+    {
+      "epoch": 0.3439017423593259,
+      "grad_norm": 2.0829055601127644,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 14448
+    },
+    {
+      "epoch": 0.3439255450823574,
+      "grad_norm": 8.197116016040756,
+      "learning_rate": 2e-05,
+      "loss": 1.769,
+      "step": 14449
+    },
+    {
+      "epoch": 0.34394934780538894,
+      "grad_norm": 7.831822046225257,
+      "learning_rate": 2e-05,
+      "loss": 1.3461,
+      "step": 14450
+    },
+    {
+      "epoch": 0.3439731505284204,
+      "grad_norm": 2.2636631080138536,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 14451
+    },
+    {
+      "epoch": 0.34399695325145196,
+      "grad_norm": 8.165294790133121,
+      "learning_rate": 2e-05,
+      "loss": 0.4029,
+      "step": 14452
+    },
+    {
+      "epoch": 0.3440207559744835,
+      "grad_norm": 3.5998349672402825,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 14453
+    },
+    {
+      "epoch": 0.344044558697515,
+      "grad_norm": 4.493870521137645,
+      "learning_rate": 2e-05,
+      "loss": 0.3902,
+      "step": 14454
+    },
+    {
+      "epoch": 0.34406836142054653,
+      "grad_norm": 3.7951588774471157,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 14455
+    },
+    {
+      "epoch": 0.344092164143578,
+      "grad_norm": 3.6361096954920957,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 14456
+    },
+    {
+      "epoch": 0.34411596686660956,
+      "grad_norm": 1.9506003715235745,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 14457
+    },
+    {
+      "epoch": 0.34413976958964104,
+      "grad_norm": 8.071970661076715,
+      "learning_rate": 2e-05,
+      "loss": 0.4747,
+      "step": 14458
+    },
+    {
+      "epoch": 0.3441635723126726,
+      "grad_norm": 2.3404806432155696,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 14459
+    },
+    {
+      "epoch": 0.34418737503570407,
+      "grad_norm": 6.649159815948443,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 14460
+    },
+    {
+      "epoch": 0.3442111777587356,
+      "grad_norm": 7.149360717362478,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 14461
+    },
+    {
+      "epoch": 0.3442349804817671,
+      "grad_norm": 8.147769246849673,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 14462
+    },
+    {
+      "epoch": 0.34425878320479864,
+      "grad_norm": 16.534095155425167,
+      "learning_rate": 2e-05,
+      "loss": 0.8225,
+      "step": 14463
+    },
+    {
+      "epoch": 0.3442825859278301,
+      "grad_norm": 3.3699548762036327,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 14464
+    },
+    {
+      "epoch": 0.34430638865086166,
+      "grad_norm": 5.919731756830048,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 14465
+    },
+    {
+      "epoch": 0.34433019137389315,
+      "grad_norm": 2.9791998632764156,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 14466
+    },
+    {
+      "epoch": 0.3443539940969247,
+      "grad_norm": 7.823508681110066,
+      "learning_rate": 2e-05,
+      "loss": 0.9407,
+      "step": 14467
+    },
+    {
+      "epoch": 0.34437779681995623,
+      "grad_norm": 4.599380325918959,
+      "learning_rate": 2e-05,
+      "loss": 0.9738,
+      "step": 14468
+    },
+    {
+      "epoch": 0.3444015995429877,
+      "grad_norm": 6.260668386198106,
+      "learning_rate": 2e-05,
+      "loss": 0.2822,
+      "step": 14469
+    },
+    {
+      "epoch": 0.34442540226601925,
+      "grad_norm": 1.5221441813969983,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 14470
+    },
+    {
+      "epoch": 0.34444920498905074,
+      "grad_norm": 5.0645735372020395,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 14471
+    },
+    {
+      "epoch": 0.3444730077120823,
+      "grad_norm": 6.786915343468591,
+      "learning_rate": 2e-05,
+      "loss": 0.3803,
+      "step": 14472
+    },
+    {
+      "epoch": 0.34449681043511376,
+      "grad_norm": 4.3934504339231815,
+      "learning_rate": 2e-05,
+      "loss": 0.7579,
+      "step": 14473
+    },
+    {
+      "epoch": 0.3445206131581453,
+      "grad_norm": 2.9278244327198957,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 14474
+    },
+    {
+      "epoch": 0.3445444158811768,
+      "grad_norm": 2.1850931636478466,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 14475
+    },
+    {
+      "epoch": 0.34456821860420833,
+      "grad_norm": 5.595188195679823,
+      "learning_rate": 2e-05,
+      "loss": 0.2108,
+      "step": 14476
+    },
+    {
+      "epoch": 0.3445920213272398,
+      "grad_norm": 3.0670600443805296,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 14477
+    },
+    {
+      "epoch": 0.34461582405027136,
+      "grad_norm": 2.960843775907569,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 14478
+    },
+    {
+      "epoch": 0.34463962677330284,
+      "grad_norm": 6.317418422329189,
+      "learning_rate": 2e-05,
+      "loss": 0.2292,
+      "step": 14479
+    },
+    {
+      "epoch": 0.3446634294963344,
+      "grad_norm": 1.536126460560864,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34468723221936587,
+      "grad_norm": 5.607469842216135,
+      "learning_rate": 2e-05,
+      "loss": 0.9437,
+      "step": 14481
+    },
+    {
+      "epoch": 0.3447110349423974,
+      "grad_norm": 4.603980630210518,
+      "learning_rate": 2e-05,
+      "loss": 0.471,
+      "step": 14482
+    },
+    {
+      "epoch": 0.34473483766542895,
+      "grad_norm": 3.7543772343538957,
+      "learning_rate": 2e-05,
+      "loss": 0.6097,
+      "step": 14483
+    },
+    {
+      "epoch": 0.34475864038846044,
+      "grad_norm": 0.6010697893992846,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 14484
+    },
+    {
+      "epoch": 0.344782443111492,
+      "grad_norm": 1.6169660644287558,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 14485
+    },
+    {
+      "epoch": 0.34480624583452346,
+      "grad_norm": 5.267280179296164,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 14486
+    },
+    {
+      "epoch": 0.344830048557555,
+      "grad_norm": 13.189010600785071,
+      "learning_rate": 2e-05,
+      "loss": 0.464,
+      "step": 14487
+    },
+    {
+      "epoch": 0.3448538512805865,
+      "grad_norm": 2.2587636054479923,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 14488
+    },
+    {
+      "epoch": 0.34487765400361803,
+      "grad_norm": 11.400174061686743,
+      "learning_rate": 2e-05,
+      "loss": 0.3864,
+      "step": 14489
+    },
+    {
+      "epoch": 0.3449014567266495,
+      "grad_norm": 1.3796555673299449,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 14490
+    },
+    {
+      "epoch": 0.34492525944968105,
+      "grad_norm": 0.38345187643471873,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 14491
+    },
+    {
+      "epoch": 0.34494906217271254,
+      "grad_norm": 6.1684075063265755,
+      "learning_rate": 2e-05,
+      "loss": 0.5099,
+      "step": 14492
+    },
+    {
+      "epoch": 0.3449728648957441,
+      "grad_norm": 2.917268348037137,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 14493
+    },
+    {
+      "epoch": 0.34499666761877557,
+      "grad_norm": 4.944493439225668,
+      "learning_rate": 2e-05,
+      "loss": 0.5842,
+      "step": 14494
+    },
+    {
+      "epoch": 0.3450204703418071,
+      "grad_norm": 4.4581657919379065,
+      "learning_rate": 2e-05,
+      "loss": 0.455,
+      "step": 14495
+    },
+    {
+      "epoch": 0.3450442730648386,
+      "grad_norm": 6.194792198915714,
+      "learning_rate": 2e-05,
+      "loss": 1.0017,
+      "step": 14496
+    },
+    {
+      "epoch": 0.34506807578787013,
+      "grad_norm": 4.4955096617729815,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 14497
+    },
+    {
+      "epoch": 0.3450918785109017,
+      "grad_norm": 2.3293727688934966,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 14498
+    },
+    {
+      "epoch": 0.34511568123393316,
+      "grad_norm": 2.5481965170262497,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 14499
+    },
+    {
+      "epoch": 0.3451394839569647,
+      "grad_norm": 1.131344227637549,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 14500
+    },
+    {
+      "epoch": 0.3451632866799962,
+      "grad_norm": 2.7498992037126673,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 14501
+    },
+    {
+      "epoch": 0.3451870894030277,
+      "grad_norm": 6.413716302776379,
+      "learning_rate": 2e-05,
+      "loss": 0.96,
+      "step": 14502
+    },
+    {
+      "epoch": 0.3452108921260592,
+      "grad_norm": 0.023072670525795893,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 14503
+    },
+    {
+      "epoch": 0.34523469484909075,
+      "grad_norm": 33.52605595897837,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 14504
+    },
+    {
+      "epoch": 0.34525849757212224,
+      "grad_norm": 4.983482355344538,
+      "learning_rate": 2e-05,
+      "loss": 0.4059,
+      "step": 14505
+    },
+    {
+      "epoch": 0.3452823002951538,
+      "grad_norm": 15.613191708582821,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 14506
+    },
+    {
+      "epoch": 0.34530610301818526,
+      "grad_norm": 3.787425159572303,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 14507
+    },
+    {
+      "epoch": 0.3453299057412168,
+      "grad_norm": 46.05988174113294,
+      "learning_rate": 2e-05,
+      "loss": 0.9145,
+      "step": 14508
+    },
+    {
+      "epoch": 0.3453537084642483,
+      "grad_norm": 13.533711160502158,
+      "learning_rate": 2e-05,
+      "loss": 0.9222,
+      "step": 14509
+    },
+    {
+      "epoch": 0.34537751118727983,
+      "grad_norm": 2.256387782684179,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 14510
+    },
+    {
+      "epoch": 0.3454013139103113,
+      "grad_norm": 1.9474983576112932,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 14511
+    },
+    {
+      "epoch": 0.34542511663334285,
+      "grad_norm": 9.05473584408686,
+      "learning_rate": 2e-05,
+      "loss": 0.8924,
+      "step": 14512
+    },
+    {
+      "epoch": 0.3454489193563744,
+      "grad_norm": 2.276489472865095,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 14513
+    },
+    {
+      "epoch": 0.3454727220794059,
+      "grad_norm": 7.867219925256929,
+      "learning_rate": 2e-05,
+      "loss": 0.4172,
+      "step": 14514
+    },
+    {
+      "epoch": 0.3454965248024374,
+      "grad_norm": 0.16031648275490765,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 14515
+    },
+    {
+      "epoch": 0.3455203275254689,
+      "grad_norm": 6.400471847585806,
+      "learning_rate": 2e-05,
+      "loss": 1.1212,
+      "step": 14516
+    },
+    {
+      "epoch": 0.34554413024850045,
+      "grad_norm": 9.584164597889227,
+      "learning_rate": 2e-05,
+      "loss": 1.7601,
+      "step": 14517
+    },
+    {
+      "epoch": 0.34556793297153193,
+      "grad_norm": 11.813282004067668,
+      "learning_rate": 2e-05,
+      "loss": 0.9187,
+      "step": 14518
+    },
+    {
+      "epoch": 0.3455917356945635,
+      "grad_norm": 1.6508622050019957,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 14519
+    },
+    {
+      "epoch": 0.34561553841759496,
+      "grad_norm": 0.7677857267888835,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 14520
+    },
+    {
+      "epoch": 0.3456393411406265,
+      "grad_norm": 17.157376062561447,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 14521
+    },
+    {
+      "epoch": 0.345663143863658,
+      "grad_norm": 16.413892651270444,
+      "learning_rate": 2e-05,
+      "loss": 0.4756,
+      "step": 14522
+    },
+    {
+      "epoch": 0.3456869465866895,
+      "grad_norm": 1.9873441094941546,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 14523
+    },
+    {
+      "epoch": 0.345710749309721,
+      "grad_norm": 2.499847092248951,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 14524
+    },
+    {
+      "epoch": 0.34573455203275255,
+      "grad_norm": 7.863831026883918,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 14525
+    },
+    {
+      "epoch": 0.34575835475578404,
+      "grad_norm": 3.2204222457569736,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 14526
+    },
+    {
+      "epoch": 0.3457821574788156,
+      "grad_norm": 11.872786300380422,
+      "learning_rate": 2e-05,
+      "loss": 0.8835,
+      "step": 14527
+    },
+    {
+      "epoch": 0.3458059602018471,
+      "grad_norm": 2.594352869845692,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 14528
+    },
+    {
+      "epoch": 0.3458297629248786,
+      "grad_norm": 0.3454166718511546,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 14529
+    },
+    {
+      "epoch": 0.34585356564791014,
+      "grad_norm": 14.799174069693112,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 14530
+    },
+    {
+      "epoch": 0.34587736837094163,
+      "grad_norm": 6.861638734785231,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 14531
+    },
+    {
+      "epoch": 0.34590117109397317,
+      "grad_norm": 11.663952628955247,
+      "learning_rate": 2e-05,
+      "loss": 0.591,
+      "step": 14532
+    },
+    {
+      "epoch": 0.34592497381700466,
+      "grad_norm": 8.424271950710702,
+      "learning_rate": 2e-05,
+      "loss": 0.9303,
+      "step": 14533
+    },
+    {
+      "epoch": 0.3459487765400362,
+      "grad_norm": 7.520916642355245,
+      "learning_rate": 2e-05,
+      "loss": 0.9835,
+      "step": 14534
+    },
+    {
+      "epoch": 0.3459725792630677,
+      "grad_norm": 1.911367361673269,
+      "learning_rate": 2e-05,
+      "loss": 0.1374,
+      "step": 14535
+    },
+    {
+      "epoch": 0.3459963819860992,
+      "grad_norm": 7.97555211766143,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 14536
+    },
+    {
+      "epoch": 0.3460201847091307,
+      "grad_norm": 6.436828776584825,
+      "learning_rate": 2e-05,
+      "loss": 0.3183,
+      "step": 14537
+    },
+    {
+      "epoch": 0.34604398743216225,
+      "grad_norm": 5.668273527015188,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 14538
+    },
+    {
+      "epoch": 0.34606779015519373,
+      "grad_norm": 1.154861992745544,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 14539
+    },
+    {
+      "epoch": 0.3460915928782253,
+      "grad_norm": 0.5512376176651953,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 14540
+    },
+    {
+      "epoch": 0.34611539560125676,
+      "grad_norm": 4.1999635195037515,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 14541
+    },
+    {
+      "epoch": 0.3461391983242883,
+      "grad_norm": 5.006297863309805,
+      "learning_rate": 2e-05,
+      "loss": 0.7204,
+      "step": 14542
+    },
+    {
+      "epoch": 0.34616300104731984,
+      "grad_norm": 1.3967660514668456,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 14543
+    },
+    {
+      "epoch": 0.3461868037703513,
+      "grad_norm": 4.888099057349146,
+      "learning_rate": 2e-05,
+      "loss": 1.0502,
+      "step": 14544
+    },
+    {
+      "epoch": 0.34621060649338287,
+      "grad_norm": 6.312960176373652,
+      "learning_rate": 2e-05,
+      "loss": 1.0525,
+      "step": 14545
+    },
+    {
+      "epoch": 0.34623440921641435,
+      "grad_norm": 0.939404158429254,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 14546
+    },
+    {
+      "epoch": 0.3462582119394459,
+      "grad_norm": 6.091071984649203,
+      "learning_rate": 2e-05,
+      "loss": 1.5298,
+      "step": 14547
+    },
+    {
+      "epoch": 0.3462820146624774,
+      "grad_norm": 3.7465432824049993,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 14548
+    },
+    {
+      "epoch": 0.3463058173855089,
+      "grad_norm": 5.499055339056155,
+      "learning_rate": 2e-05,
+      "loss": 0.6165,
+      "step": 14549
+    },
+    {
+      "epoch": 0.3463296201085404,
+      "grad_norm": 2.471714258559287,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 14550
+    },
+    {
+      "epoch": 0.34635342283157194,
+      "grad_norm": 13.07868609435809,
+      "learning_rate": 2e-05,
+      "loss": 0.4547,
+      "step": 14551
+    },
+    {
+      "epoch": 0.34637722555460343,
+      "grad_norm": 4.131910855104667,
+      "learning_rate": 2e-05,
+      "loss": 0.6304,
+      "step": 14552
+    },
+    {
+      "epoch": 0.34640102827763497,
+      "grad_norm": 5.350654754275951,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 14553
+    },
+    {
+      "epoch": 0.34642483100066646,
+      "grad_norm": 5.765098406561264,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 14554
+    },
+    {
+      "epoch": 0.346448633723698,
+      "grad_norm": 1.2316473203619427,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 14555
+    },
+    {
+      "epoch": 0.3464724364467295,
+      "grad_norm": 4.324153111175862,
+      "learning_rate": 2e-05,
+      "loss": 0.5619,
+      "step": 14556
+    },
+    {
+      "epoch": 0.346496239169761,
+      "grad_norm": 4.168822756461247,
+      "learning_rate": 2e-05,
+      "loss": 0.9294,
+      "step": 14557
+    },
+    {
+      "epoch": 0.34652004189279256,
+      "grad_norm": 6.99741195291104,
+      "learning_rate": 2e-05,
+      "loss": 1.1459,
+      "step": 14558
+    },
+    {
+      "epoch": 0.34654384461582405,
+      "grad_norm": 6.510003345415034,
+      "learning_rate": 2e-05,
+      "loss": 0.6558,
+      "step": 14559
+    },
+    {
+      "epoch": 0.3465676473388556,
+      "grad_norm": 3.4152148102026785,
+      "learning_rate": 2e-05,
+      "loss": 0.1626,
+      "step": 14560
+    },
+    {
+      "epoch": 0.3465914500618871,
+      "grad_norm": 0.7925724831364847,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 14561
+    },
+    {
+      "epoch": 0.3466152527849186,
+      "grad_norm": 3.7347975748741304,
+      "learning_rate": 2e-05,
+      "loss": 0.703,
+      "step": 14562
+    },
+    {
+      "epoch": 0.3466390555079501,
+      "grad_norm": 5.191814373407833,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 14563
+    },
+    {
+      "epoch": 0.34666285823098164,
+      "grad_norm": 3.4862307713087612,
+      "learning_rate": 2e-05,
+      "loss": 0.5698,
+      "step": 14564
+    },
+    {
+      "epoch": 0.3466866609540131,
+      "grad_norm": 2.7170646742990674,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 14565
+    },
+    {
+      "epoch": 0.34671046367704467,
+      "grad_norm": 1.725646672746256,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 14566
+    },
+    {
+      "epoch": 0.34673426640007615,
+      "grad_norm": 1.6608473788756664,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 14567
+    },
+    {
+      "epoch": 0.3467580691231077,
+      "grad_norm": 5.648780655327861,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 14568
+    },
+    {
+      "epoch": 0.3467818718461392,
+      "grad_norm": 10.130148796217592,
+      "learning_rate": 2e-05,
+      "loss": 0.9155,
+      "step": 14569
+    },
+    {
+      "epoch": 0.3468056745691707,
+      "grad_norm": 8.268664879328915,
+      "learning_rate": 2e-05,
+      "loss": 1.678,
+      "step": 14570
+    },
+    {
+      "epoch": 0.3468294772922022,
+      "grad_norm": 1.8332004709387133,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 14571
+    },
+    {
+      "epoch": 0.34685328001523374,
+      "grad_norm": 2.6736801066658913,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 14572
+    },
+    {
+      "epoch": 0.34687708273826523,
+      "grad_norm": 3.371383960369993,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 14573
+    },
+    {
+      "epoch": 0.34690088546129677,
+      "grad_norm": 2.225103062058131,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 14574
+    },
+    {
+      "epoch": 0.3469246881843283,
+      "grad_norm": 6.124519760867068,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 14575
+    },
+    {
+      "epoch": 0.3469484909073598,
+      "grad_norm": 0.9368556201424226,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14576
+    },
+    {
+      "epoch": 0.34697229363039134,
+      "grad_norm": 5.8995079617231525,
+      "learning_rate": 2e-05,
+      "loss": 0.7169,
+      "step": 14577
+    },
+    {
+      "epoch": 0.3469960963534228,
+      "grad_norm": 1.201593580750831,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 14578
+    },
+    {
+      "epoch": 0.34701989907645436,
+      "grad_norm": 1.3461168373139543,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 14579
+    },
+    {
+      "epoch": 0.34704370179948585,
+      "grad_norm": 4.894757203082201,
+      "learning_rate": 2e-05,
+      "loss": 0.5315,
+      "step": 14580
+    },
+    {
+      "epoch": 0.3470675045225174,
+      "grad_norm": 9.320337282223099,
+      "learning_rate": 2e-05,
+      "loss": 0.9277,
+      "step": 14581
+    },
+    {
+      "epoch": 0.3470913072455489,
+      "grad_norm": 4.545339615525304,
+      "learning_rate": 2e-05,
+      "loss": 0.297,
+      "step": 14582
+    },
+    {
+      "epoch": 0.3471151099685804,
+      "grad_norm": 2.1693092094319324,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 14583
+    },
+    {
+      "epoch": 0.3471389126916119,
+      "grad_norm": 7.083487328488788,
+      "learning_rate": 2e-05,
+      "loss": 0.6463,
+      "step": 14584
+    },
+    {
+      "epoch": 0.34716271541464344,
+      "grad_norm": 2.147307280958674,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 14585
+    },
+    {
+      "epoch": 0.3471865181376749,
+      "grad_norm": 5.029587892491107,
+      "learning_rate": 2e-05,
+      "loss": 0.8523,
+      "step": 14586
+    },
+    {
+      "epoch": 0.34721032086070647,
+      "grad_norm": 3.0557431825232637,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 14587
+    },
+    {
+      "epoch": 0.34723412358373795,
+      "grad_norm": 5.131536996985191,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 14588
+    },
+    {
+      "epoch": 0.3472579263067695,
+      "grad_norm": 4.609469242744813,
+      "learning_rate": 2e-05,
+      "loss": 0.5965,
+      "step": 14589
+    },
+    {
+      "epoch": 0.34728172902980103,
+      "grad_norm": 5.06963685643247,
+      "learning_rate": 2e-05,
+      "loss": 0.2658,
+      "step": 14590
+    },
+    {
+      "epoch": 0.3473055317528325,
+      "grad_norm": 5.318064661724405,
+      "learning_rate": 2e-05,
+      "loss": 1.0469,
+      "step": 14591
+    },
+    {
+      "epoch": 0.34732933447586406,
+      "grad_norm": 4.827624755943802,
+      "learning_rate": 2e-05,
+      "loss": 0.6729,
+      "step": 14592
+    },
+    {
+      "epoch": 0.34735313719889555,
+      "grad_norm": 3.791141739522202,
+      "learning_rate": 2e-05,
+      "loss": 0.1235,
+      "step": 14593
+    },
+    {
+      "epoch": 0.3473769399219271,
+      "grad_norm": 2.4503597527882475,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 14594
+    },
+    {
+      "epoch": 0.34740074264495857,
+      "grad_norm": 2.0515715774711203,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 14595
+    },
+    {
+      "epoch": 0.3474245453679901,
+      "grad_norm": 4.05757450875365,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 14596
+    },
+    {
+      "epoch": 0.3474483480910216,
+      "grad_norm": 1.406291990635029,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 14597
+    },
+    {
+      "epoch": 0.34747215081405314,
+      "grad_norm": 11.587269799358383,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 14598
+    },
+    {
+      "epoch": 0.3474959535370846,
+      "grad_norm": 2.335392091940739,
+      "learning_rate": 2e-05,
+      "loss": 0.1815,
+      "step": 14599
+    },
+    {
+      "epoch": 0.34751975626011616,
+      "grad_norm": 3.8892633400573122,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 14600
+    },
+    {
+      "epoch": 0.34754355898314765,
+      "grad_norm": 4.6969490597877375,
+      "learning_rate": 2e-05,
+      "loss": 0.7763,
+      "step": 14601
+    },
+    {
+      "epoch": 0.3475673617061792,
+      "grad_norm": 3.1828332158114128,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 14602
+    },
+    {
+      "epoch": 0.3475911644292107,
+      "grad_norm": 1.149352851943564,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14603
+    },
+    {
+      "epoch": 0.3476149671522422,
+      "grad_norm": 3.0275968638280557,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 14604
+    },
+    {
+      "epoch": 0.34763876987527376,
+      "grad_norm": 0.560335242545916,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 14605
+    },
+    {
+      "epoch": 0.34766257259830524,
+      "grad_norm": 1.032779243552068,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 14606
+    },
+    {
+      "epoch": 0.3476863753213368,
+      "grad_norm": 3.02595671735316,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 14607
+    },
+    {
+      "epoch": 0.34771017804436827,
+      "grad_norm": 6.529621363757079,
+      "learning_rate": 2e-05,
+      "loss": 0.4558,
+      "step": 14608
+    },
+    {
+      "epoch": 0.3477339807673998,
+      "grad_norm": 5.093024066022264,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 14609
+    },
+    {
+      "epoch": 0.3477577834904313,
+      "grad_norm": 1.0728923172031406,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 14610
+    },
+    {
+      "epoch": 0.34778158621346283,
+      "grad_norm": 4.146593402918697,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 14611
+    },
+    {
+      "epoch": 0.3478053889364943,
+      "grad_norm": 4.8649821546362855,
+      "learning_rate": 2e-05,
+      "loss": 0.2933,
+      "step": 14612
+    },
+    {
+      "epoch": 0.34782919165952586,
+      "grad_norm": 1.848968171489043,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 14613
+    },
+    {
+      "epoch": 0.34785299438255735,
+      "grad_norm": 0.7812173496720783,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 14614
+    },
+    {
+      "epoch": 0.3478767971055889,
+      "grad_norm": 5.19203044225256,
+      "learning_rate": 2e-05,
+      "loss": 0.456,
+      "step": 14615
+    },
+    {
+      "epoch": 0.34790059982862037,
+      "grad_norm": 18.197592612182817,
+      "learning_rate": 2e-05,
+      "loss": 0.6402,
+      "step": 14616
+    },
+    {
+      "epoch": 0.3479244025516519,
+      "grad_norm": 3.9332424816440623,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 14617
+    },
+    {
+      "epoch": 0.3479482052746834,
+      "grad_norm": 3.555137576757215,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 14618
+    },
+    {
+      "epoch": 0.34797200799771494,
+      "grad_norm": 5.547111328035181,
+      "learning_rate": 2e-05,
+      "loss": 1.1725,
+      "step": 14619
+    },
+    {
+      "epoch": 0.3479958107207465,
+      "grad_norm": 5.844518038864233,
+      "learning_rate": 2e-05,
+      "loss": 0.1508,
+      "step": 14620
+    },
+    {
+      "epoch": 0.34801961344377796,
+      "grad_norm": 0.48824288662084614,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 14621
+    },
+    {
+      "epoch": 0.3480434161668095,
+      "grad_norm": 0.5440304015897085,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 14622
+    },
+    {
+      "epoch": 0.348067218889841,
+      "grad_norm": 7.192251303032062,
+      "learning_rate": 2e-05,
+      "loss": 0.8202,
+      "step": 14623
+    },
+    {
+      "epoch": 0.34809102161287253,
+      "grad_norm": 6.063444333233286,
+      "learning_rate": 2e-05,
+      "loss": 0.7196,
+      "step": 14624
+    },
+    {
+      "epoch": 0.348114824335904,
+      "grad_norm": 3.441119303148551,
+      "learning_rate": 2e-05,
+      "loss": 0.593,
+      "step": 14625
+    },
+    {
+      "epoch": 0.34813862705893556,
+      "grad_norm": 3.9883731461099408,
+      "learning_rate": 2e-05,
+      "loss": 0.5636,
+      "step": 14626
+    },
+    {
+      "epoch": 0.34816242978196704,
+      "grad_norm": 5.691644335134066,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 14627
+    },
+    {
+      "epoch": 0.3481862325049986,
+      "grad_norm": 4.348278824051703,
+      "learning_rate": 2e-05,
+      "loss": 0.1428,
+      "step": 14628
+    },
+    {
+      "epoch": 0.34821003522803007,
+      "grad_norm": 6.04684157777321,
+      "learning_rate": 2e-05,
+      "loss": 0.6375,
+      "step": 14629
+    },
+    {
+      "epoch": 0.3482338379510616,
+      "grad_norm": 2.1905835458189773,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 14630
+    },
+    {
+      "epoch": 0.3482576406740931,
+      "grad_norm": 4.237340158548962,
+      "learning_rate": 2e-05,
+      "loss": 0.4949,
+      "step": 14631
+    },
+    {
+      "epoch": 0.34828144339712463,
+      "grad_norm": 2.5159887502569016,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 14632
+    },
+    {
+      "epoch": 0.3483052461201561,
+      "grad_norm": 6.071399522565395,
+      "learning_rate": 2e-05,
+      "loss": 0.7745,
+      "step": 14633
+    },
+    {
+      "epoch": 0.34832904884318766,
+      "grad_norm": 6.308552702847295,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 14634
+    },
+    {
+      "epoch": 0.3483528515662192,
+      "grad_norm": 4.614508840350516,
+      "learning_rate": 2e-05,
+      "loss": 0.726,
+      "step": 14635
+    },
+    {
+      "epoch": 0.3483766542892507,
+      "grad_norm": 3.0183586063349486,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 14636
+    },
+    {
+      "epoch": 0.3484004570122822,
+      "grad_norm": 5.001258826403422,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 14637
+    },
+    {
+      "epoch": 0.3484242597353137,
+      "grad_norm": 4.780372454965898,
+      "learning_rate": 2e-05,
+      "loss": 0.2955,
+      "step": 14638
+    },
+    {
+      "epoch": 0.34844806245834525,
+      "grad_norm": 6.53366790727223,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 14639
+    },
+    {
+      "epoch": 0.34847186518137674,
+      "grad_norm": 6.949752522481233,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 14640
+    },
+    {
+      "epoch": 0.3484956679044083,
+      "grad_norm": 1.3371979349234195,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 14641
+    },
+    {
+      "epoch": 0.34851947062743976,
+      "grad_norm": 4.803731503278085,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 14642
+    },
+    {
+      "epoch": 0.3485432733504713,
+      "grad_norm": 6.319458742653581,
+      "learning_rate": 2e-05,
+      "loss": 0.3097,
+      "step": 14643
+    },
+    {
+      "epoch": 0.3485670760735028,
+      "grad_norm": 2.8092874991481915,
+      "learning_rate": 2e-05,
+      "loss": 0.2478,
+      "step": 14644
+    },
+    {
+      "epoch": 0.34859087879653433,
+      "grad_norm": 8.919502894169241,
+      "learning_rate": 2e-05,
+      "loss": 1.2045,
+      "step": 14645
+    },
+    {
+      "epoch": 0.3486146815195658,
+      "grad_norm": 1.9634395793001773,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 14646
+    },
+    {
+      "epoch": 0.34863848424259736,
+      "grad_norm": 17.395412635437513,
+      "learning_rate": 2e-05,
+      "loss": 1.2102,
+      "step": 14647
+    },
+    {
+      "epoch": 0.34866228696562884,
+      "grad_norm": 1.3560327362521172,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 14648
+    },
+    {
+      "epoch": 0.3486860896886604,
+      "grad_norm": 2.811955460409845,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 14649
+    },
+    {
+      "epoch": 0.3487098924116919,
+      "grad_norm": 1.7012737796895907,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 14650
+    },
+    {
+      "epoch": 0.3487336951347234,
+      "grad_norm": 3.244598259662393,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 14651
+    },
+    {
+      "epoch": 0.34875749785775495,
+      "grad_norm": 5.6526032330547356,
+      "learning_rate": 2e-05,
+      "loss": 1.0658,
+      "step": 14652
+    },
+    {
+      "epoch": 0.34878130058078644,
+      "grad_norm": 4.645775277957902,
+      "learning_rate": 2e-05,
+      "loss": 0.7068,
+      "step": 14653
+    },
+    {
+      "epoch": 0.348805103303818,
+      "grad_norm": 10.89155998906843,
+      "learning_rate": 2e-05,
+      "loss": 0.3432,
+      "step": 14654
+    },
+    {
+      "epoch": 0.34882890602684946,
+      "grad_norm": 5.044747471411678,
+      "learning_rate": 2e-05,
+      "loss": 0.3207,
+      "step": 14655
+    },
+    {
+      "epoch": 0.348852708749881,
+      "grad_norm": 3.624526310341468,
+      "learning_rate": 2e-05,
+      "loss": 0.4604,
+      "step": 14656
+    },
+    {
+      "epoch": 0.3488765114729125,
+      "grad_norm": 1.0123757050960316,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 14657
+    },
+    {
+      "epoch": 0.34890031419594403,
+      "grad_norm": 7.857066609067083,
+      "learning_rate": 2e-05,
+      "loss": 0.2621,
+      "step": 14658
+    },
+    {
+      "epoch": 0.3489241169189755,
+      "grad_norm": 6.305197560099984,
+      "learning_rate": 2e-05,
+      "loss": 0.9707,
+      "step": 14659
+    },
+    {
+      "epoch": 0.34894791964200705,
+      "grad_norm": 5.411286833527905,
+      "learning_rate": 2e-05,
+      "loss": 0.4622,
+      "step": 14660
+    },
+    {
+      "epoch": 0.34897172236503854,
+      "grad_norm": 2.804329934906863,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 14661
+    },
+    {
+      "epoch": 0.3489955250880701,
+      "grad_norm": 0.36125722018649975,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 14662
+    },
+    {
+      "epoch": 0.34901932781110157,
+      "grad_norm": 2.9705190269455617,
+      "learning_rate": 2e-05,
+      "loss": 0.1356,
+      "step": 14663
+    },
+    {
+      "epoch": 0.3490431305341331,
+      "grad_norm": 9.980880287257753,
+      "learning_rate": 2e-05,
+      "loss": 2.2604,
+      "step": 14664
+    },
+    {
+      "epoch": 0.34906693325716465,
+      "grad_norm": 5.124698368350442,
+      "learning_rate": 2e-05,
+      "loss": 0.2324,
+      "step": 14665
+    },
+    {
+      "epoch": 0.34909073598019613,
+      "grad_norm": 2.5954435421321698,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 14666
+    },
+    {
+      "epoch": 0.3491145387032277,
+      "grad_norm": 1.9582350296379636,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 14667
+    },
+    {
+      "epoch": 0.34913834142625916,
+      "grad_norm": 2.7163574645916633,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 14668
+    },
+    {
+      "epoch": 0.3491621441492907,
+      "grad_norm": 1.970350871740441,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 14669
+    },
+    {
+      "epoch": 0.3491859468723222,
+      "grad_norm": 5.59794662291776,
+      "learning_rate": 2e-05,
+      "loss": 1.1285,
+      "step": 14670
+    },
+    {
+      "epoch": 0.3492097495953537,
+      "grad_norm": 6.362243545885975,
+      "learning_rate": 2e-05,
+      "loss": 0.4212,
+      "step": 14671
+    },
+    {
+      "epoch": 0.3492335523183852,
+      "grad_norm": 5.051130384361379,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 14672
+    },
+    {
+      "epoch": 0.34925735504141675,
+      "grad_norm": 5.52170295476203,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 14673
+    },
+    {
+      "epoch": 0.34928115776444824,
+      "grad_norm": 1.8664549539974205,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 14674
+    },
+    {
+      "epoch": 0.3493049604874798,
+      "grad_norm": 4.272233668512509,
+      "learning_rate": 2e-05,
+      "loss": 0.1889,
+      "step": 14675
+    },
+    {
+      "epoch": 0.34932876321051126,
+      "grad_norm": 1.489401931846771,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 14676
+    },
+    {
+      "epoch": 0.3493525659335428,
+      "grad_norm": 3.855048539675568,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 14677
+    },
+    {
+      "epoch": 0.3493763686565743,
+      "grad_norm": 6.800327429170579,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 14678
+    },
+    {
+      "epoch": 0.34940017137960583,
+      "grad_norm": 4.691706073109704,
+      "learning_rate": 2e-05,
+      "loss": 0.8075,
+      "step": 14679
+    },
+    {
+      "epoch": 0.34942397410263737,
+      "grad_norm": 2.1969005902861207,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 14680
+    },
+    {
+      "epoch": 0.34944777682566885,
+      "grad_norm": 9.376199201183214,
+      "learning_rate": 2e-05,
+      "loss": 1.9895,
+      "step": 14681
+    },
+    {
+      "epoch": 0.3494715795487004,
+      "grad_norm": 3.4226156393263216,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 14682
+    },
+    {
+      "epoch": 0.3494953822717319,
+      "grad_norm": 3.36553790483399,
+      "learning_rate": 2e-05,
+      "loss": 0.5848,
+      "step": 14683
+    },
+    {
+      "epoch": 0.3495191849947634,
+      "grad_norm": 3.442239337956031,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 14684
+    },
+    {
+      "epoch": 0.3495429877177949,
+      "grad_norm": 2.849367593214484,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 14685
+    },
+    {
+      "epoch": 0.34956679044082645,
+      "grad_norm": 3.578045917648145,
+      "learning_rate": 2e-05,
+      "loss": 0.5238,
+      "step": 14686
+    },
+    {
+      "epoch": 0.34959059316385793,
+      "grad_norm": 2.485041305917093,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 14687
+    },
+    {
+      "epoch": 0.3496143958868895,
+      "grad_norm": 8.552406846535836,
+      "learning_rate": 2e-05,
+      "loss": 0.3639,
+      "step": 14688
+    },
+    {
+      "epoch": 0.34963819860992096,
+      "grad_norm": 3.492352241870825,
+      "learning_rate": 2e-05,
+      "loss": 0.2269,
+      "step": 14689
+    },
+    {
+      "epoch": 0.3496620013329525,
+      "grad_norm": 6.7632395360463455,
+      "learning_rate": 2e-05,
+      "loss": 2.0722,
+      "step": 14690
+    },
+    {
+      "epoch": 0.349685804055984,
+      "grad_norm": 3.4280469842958317,
+      "learning_rate": 2e-05,
+      "loss": 0.393,
+      "step": 14691
+    },
+    {
+      "epoch": 0.3497096067790155,
+      "grad_norm": 1.8986543302789387,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 14692
+    },
+    {
+      "epoch": 0.349733409502047,
+      "grad_norm": 3.6268412594153134,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 14693
+    },
+    {
+      "epoch": 0.34975721222507855,
+      "grad_norm": 2.16171814648681,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 14694
+    },
+    {
+      "epoch": 0.34978101494811004,
+      "grad_norm": 3.077214247949118,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 14695
+    },
+    {
+      "epoch": 0.3498048176711416,
+      "grad_norm": 3.8988470919902225,
+      "learning_rate": 2e-05,
+      "loss": 0.6373,
+      "step": 14696
+    },
+    {
+      "epoch": 0.3498286203941731,
+      "grad_norm": 4.484026027384753,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 14697
+    },
+    {
+      "epoch": 0.3498524231172046,
+      "grad_norm": 10.93273911219994,
+      "learning_rate": 2e-05,
+      "loss": 0.24,
+      "step": 14698
+    },
+    {
+      "epoch": 0.34987622584023614,
+      "grad_norm": 4.268755453784777,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 14699
+    },
+    {
+      "epoch": 0.34990002856326763,
+      "grad_norm": 4.096945171745834,
+      "learning_rate": 2e-05,
+      "loss": 0.4782,
+      "step": 14700
+    },
+    {
+      "epoch": 0.34992383128629917,
+      "grad_norm": 7.436754056038991,
+      "learning_rate": 2e-05,
+      "loss": 0.5328,
+      "step": 14701
+    },
+    {
+      "epoch": 0.34994763400933065,
+      "grad_norm": 8.096788985168255,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 14702
+    },
+    {
+      "epoch": 0.3499714367323622,
+      "grad_norm": 5.380823652779582,
+      "learning_rate": 2e-05,
+      "loss": 0.4351,
+      "step": 14703
+    },
+    {
+      "epoch": 0.3499952394553937,
+      "grad_norm": 4.318989801201767,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 14704
+    },
+    {
+      "epoch": 0.3500190421784252,
+      "grad_norm": 7.921381440590523,
+      "learning_rate": 2e-05,
+      "loss": 0.6201,
+      "step": 14705
+    },
+    {
+      "epoch": 0.3500428449014567,
+      "grad_norm": 2.8895763026955588,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 14706
+    },
+    {
+      "epoch": 0.35006664762448825,
+      "grad_norm": 0.6173058973641995,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 14707
+    },
+    {
+      "epoch": 0.35009045034751973,
+      "grad_norm": 6.878388325024517,
+      "learning_rate": 2e-05,
+      "loss": 0.8432,
+      "step": 14708
+    },
+    {
+      "epoch": 0.3501142530705513,
+      "grad_norm": 2.579045889680659,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 14709
+    },
+    {
+      "epoch": 0.35013805579358276,
+      "grad_norm": 3.8007014388267284,
+      "learning_rate": 2e-05,
+      "loss": 0.7337,
+      "step": 14710
+    },
+    {
+      "epoch": 0.3501618585166143,
+      "grad_norm": 3.379849419996292,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 14711
+    },
+    {
+      "epoch": 0.35018566123964584,
+      "grad_norm": 5.863517210513637,
+      "learning_rate": 2e-05,
+      "loss": 1.3361,
+      "step": 14712
+    },
+    {
+      "epoch": 0.3502094639626773,
+      "grad_norm": 5.000938841190674,
+      "learning_rate": 2e-05,
+      "loss": 0.8892,
+      "step": 14713
+    },
+    {
+      "epoch": 0.35023326668570887,
+      "grad_norm": 1.121723268858899,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 14714
+    },
+    {
+      "epoch": 0.35025706940874035,
+      "grad_norm": 2.0873680811611957,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 14715
+    },
+    {
+      "epoch": 0.3502808721317719,
+      "grad_norm": 5.968019481357401,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 14716
+    },
+    {
+      "epoch": 0.3503046748548034,
+      "grad_norm": 7.7311637079145425,
+      "learning_rate": 2e-05,
+      "loss": 1.0289,
+      "step": 14717
+    },
+    {
+      "epoch": 0.3503284775778349,
+      "grad_norm": 0.845146548091661,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 14718
+    },
+    {
+      "epoch": 0.3503522803008664,
+      "grad_norm": 3.56105230939848,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 14719
+    },
+    {
+      "epoch": 0.35037608302389794,
+      "grad_norm": 3.7370012044212,
+      "learning_rate": 2e-05,
+      "loss": 0.471,
+      "step": 14720
+    },
+    {
+      "epoch": 0.35039988574692943,
+      "grad_norm": 2.528217875079,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 14721
+    },
+    {
+      "epoch": 0.35042368846996097,
+      "grad_norm": 2.7316613701816967,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 14722
+    },
+    {
+      "epoch": 0.35044749119299246,
+      "grad_norm": 4.666579255803522,
+      "learning_rate": 2e-05,
+      "loss": 1.165,
+      "step": 14723
+    },
+    {
+      "epoch": 0.350471293916024,
+      "grad_norm": 5.737238516244961,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 14724
+    },
+    {
+      "epoch": 0.3504950966390555,
+      "grad_norm": 2.254875567852806,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 14725
+    },
+    {
+      "epoch": 0.350518899362087,
+      "grad_norm": 0.5286026793416164,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 14726
+    },
+    {
+      "epoch": 0.35054270208511856,
+      "grad_norm": 7.338689576416116,
+      "learning_rate": 2e-05,
+      "loss": 0.6912,
+      "step": 14727
+    },
+    {
+      "epoch": 0.35056650480815005,
+      "grad_norm": 5.418822950163426,
+      "learning_rate": 2e-05,
+      "loss": 0.3825,
+      "step": 14728
+    },
+    {
+      "epoch": 0.3505903075311816,
+      "grad_norm": 4.161506852948553,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 14729
+    },
+    {
+      "epoch": 0.3506141102542131,
+      "grad_norm": 9.237476158577792,
+      "learning_rate": 2e-05,
+      "loss": 0.4216,
+      "step": 14730
+    },
+    {
+      "epoch": 0.3506379129772446,
+      "grad_norm": 11.607782993838617,
+      "learning_rate": 2e-05,
+      "loss": 0.743,
+      "step": 14731
+    },
+    {
+      "epoch": 0.3506617157002761,
+      "grad_norm": 3.151654204290691,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 14732
+    },
+    {
+      "epoch": 0.35068551842330764,
+      "grad_norm": 7.50734064986733,
+      "learning_rate": 2e-05,
+      "loss": 0.3327,
+      "step": 14733
+    },
+    {
+      "epoch": 0.3507093211463391,
+      "grad_norm": 4.062536512695638,
+      "learning_rate": 2e-05,
+      "loss": 0.4306,
+      "step": 14734
+    },
+    {
+      "epoch": 0.35073312386937067,
+      "grad_norm": 2.7263301589430333,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 14735
+    },
+    {
+      "epoch": 0.35075692659240215,
+      "grad_norm": 3.0657059589850815,
+      "learning_rate": 2e-05,
+      "loss": 0.1868,
+      "step": 14736
+    },
+    {
+      "epoch": 0.3507807293154337,
+      "grad_norm": 7.843376740866593,
+      "learning_rate": 2e-05,
+      "loss": 1.2033,
+      "step": 14737
+    },
+    {
+      "epoch": 0.3508045320384652,
+      "grad_norm": 3.022021669803767,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 14738
+    },
+    {
+      "epoch": 0.3508283347614967,
+      "grad_norm": 8.577254662910114,
+      "learning_rate": 2e-05,
+      "loss": 1.1089,
+      "step": 14739
+    },
+    {
+      "epoch": 0.3508521374845282,
+      "grad_norm": 2.150793046348626,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 14740
+    },
+    {
+      "epoch": 0.35087594020755974,
+      "grad_norm": 2.9519979932944045,
+      "learning_rate": 2e-05,
+      "loss": 0.1856,
+      "step": 14741
+    },
+    {
+      "epoch": 0.3508997429305913,
+      "grad_norm": 6.048693084743315,
+      "learning_rate": 2e-05,
+      "loss": 0.4473,
+      "step": 14742
+    },
+    {
+      "epoch": 0.35092354565362277,
+      "grad_norm": 3.851234585511366,
+      "learning_rate": 2e-05,
+      "loss": 0.7138,
+      "step": 14743
+    },
+    {
+      "epoch": 0.3509473483766543,
+      "grad_norm": 9.329103958341856,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 14744
+    },
+    {
+      "epoch": 0.3509711510996858,
+      "grad_norm": 4.1242096665024786,
+      "learning_rate": 2e-05,
+      "loss": 0.5715,
+      "step": 14745
+    },
+    {
+      "epoch": 0.35099495382271734,
+      "grad_norm": 2.30593673139524,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 14746
+    },
+    {
+      "epoch": 0.3510187565457488,
+      "grad_norm": 1.3621077556053238,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 14747
+    },
+    {
+      "epoch": 0.35104255926878036,
+      "grad_norm": 3.7546563356627414,
+      "learning_rate": 2e-05,
+      "loss": 0.1321,
+      "step": 14748
+    },
+    {
+      "epoch": 0.35106636199181185,
+      "grad_norm": 4.284713115023686,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 14749
+    },
+    {
+      "epoch": 0.3510901647148434,
+      "grad_norm": 2.727351393696416,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 14750
+    },
+    {
+      "epoch": 0.3511139674378749,
+      "grad_norm": 3.880009719870198,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 14751
+    },
+    {
+      "epoch": 0.3511377701609064,
+      "grad_norm": 2.7005022464392487,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 14752
+    },
+    {
+      "epoch": 0.3511615728839379,
+      "grad_norm": 2.8064692212489812,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 14753
+    },
+    {
+      "epoch": 0.35118537560696944,
+      "grad_norm": 7.671357132950529,
+      "learning_rate": 2e-05,
+      "loss": 0.544,
+      "step": 14754
+    },
+    {
+      "epoch": 0.3512091783300009,
+      "grad_norm": 2.3807514373173415,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 14755
+    },
+    {
+      "epoch": 0.35123298105303247,
+      "grad_norm": 6.185766544666784,
+      "learning_rate": 2e-05,
+      "loss": 0.2882,
+      "step": 14756
+    },
+    {
+      "epoch": 0.351256783776064,
+      "grad_norm": 1.2568964931779212,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 14757
+    },
+    {
+      "epoch": 0.3512805864990955,
+      "grad_norm": 3.256398784999365,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 14758
+    },
+    {
+      "epoch": 0.35130438922212703,
+      "grad_norm": 5.551287538080116,
+      "learning_rate": 2e-05,
+      "loss": 0.7085,
+      "step": 14759
+    },
+    {
+      "epoch": 0.3513281919451585,
+      "grad_norm": 3.3539283843763834,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 14760
+    },
+    {
+      "epoch": 0.35135199466819006,
+      "grad_norm": 0.3422440439917548,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 14761
+    },
+    {
+      "epoch": 0.35137579739122154,
+      "grad_norm": 5.170597910155721,
+      "learning_rate": 2e-05,
+      "loss": 1.0618,
+      "step": 14762
+    },
+    {
+      "epoch": 0.3513996001142531,
+      "grad_norm": 2.201583652244338,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 14763
+    },
+    {
+      "epoch": 0.35142340283728457,
+      "grad_norm": 5.739532017882056,
+      "learning_rate": 2e-05,
+      "loss": 0.4582,
+      "step": 14764
+    },
+    {
+      "epoch": 0.3514472055603161,
+      "grad_norm": 2.6233468308620562,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 14765
+    },
+    {
+      "epoch": 0.3514710082833476,
+      "grad_norm": 2.2712614370938966,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 14766
+    },
+    {
+      "epoch": 0.35149481100637914,
+      "grad_norm": 4.753380449625873,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 14767
+    },
+    {
+      "epoch": 0.3515186137294106,
+      "grad_norm": 5.867214134894159,
+      "learning_rate": 2e-05,
+      "loss": 1.4285,
+      "step": 14768
+    },
+    {
+      "epoch": 0.35154241645244216,
+      "grad_norm": 3.5843525535531477,
+      "learning_rate": 2e-05,
+      "loss": 0.3291,
+      "step": 14769
+    },
+    {
+      "epoch": 0.35156621917547365,
+      "grad_norm": 1.5554243463108286,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 14770
+    },
+    {
+      "epoch": 0.3515900218985052,
+      "grad_norm": 0.7964027748689984,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 14771
+    },
+    {
+      "epoch": 0.35161382462153673,
+      "grad_norm": 6.963990339755882,
+      "learning_rate": 2e-05,
+      "loss": 0.4386,
+      "step": 14772
+    },
+    {
+      "epoch": 0.3516376273445682,
+      "grad_norm": 7.4748287368997754,
+      "learning_rate": 2e-05,
+      "loss": 1.3411,
+      "step": 14773
+    },
+    {
+      "epoch": 0.35166143006759976,
+      "grad_norm": 5.7190385178838525,
+      "learning_rate": 2e-05,
+      "loss": 0.6154,
+      "step": 14774
+    },
+    {
+      "epoch": 0.35168523279063124,
+      "grad_norm": 4.60214132136734,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 14775
+    },
+    {
+      "epoch": 0.3517090355136628,
+      "grad_norm": 1.0577050134040804,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 14776
+    },
+    {
+      "epoch": 0.35173283823669427,
+      "grad_norm": 7.437706173509203,
+      "learning_rate": 2e-05,
+      "loss": 0.7799,
+      "step": 14777
+    },
+    {
+      "epoch": 0.3517566409597258,
+      "grad_norm": 1.800178180840907,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 14778
+    },
+    {
+      "epoch": 0.3517804436827573,
+      "grad_norm": 19.91392443312543,
+      "learning_rate": 2e-05,
+      "loss": 0.5098,
+      "step": 14779
+    },
+    {
+      "epoch": 0.35180424640578883,
+      "grad_norm": 7.372450738744554,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 14780
+    },
+    {
+      "epoch": 0.3518280491288203,
+      "grad_norm": 2.3616034557456875,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 14781
+    },
+    {
+      "epoch": 0.35185185185185186,
+      "grad_norm": 4.327764362456783,
+      "learning_rate": 2e-05,
+      "loss": 0.8087,
+      "step": 14782
+    },
+    {
+      "epoch": 0.35187565457488335,
+      "grad_norm": 1.5141302264363696,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 14783
+    },
+    {
+      "epoch": 0.3518994572979149,
+      "grad_norm": 5.480253823843159,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 14784
+    },
+    {
+      "epoch": 0.35192326002094637,
+      "grad_norm": 5.575481308441396,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 14785
+    },
+    {
+      "epoch": 0.3519470627439779,
+      "grad_norm": 8.11473140333306,
+      "learning_rate": 2e-05,
+      "loss": 1.0944,
+      "step": 14786
+    },
+    {
+      "epoch": 0.35197086546700945,
+      "grad_norm": 2.640970476867252,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 14787
+    },
+    {
+      "epoch": 0.35199466819004094,
+      "grad_norm": 3.9337413449232246,
+      "learning_rate": 2e-05,
+      "loss": 0.6562,
+      "step": 14788
+    },
+    {
+      "epoch": 0.3520184709130725,
+      "grad_norm": 11.33610506419963,
+      "learning_rate": 2e-05,
+      "loss": 1.0679,
+      "step": 14789
+    },
+    {
+      "epoch": 0.35204227363610396,
+      "grad_norm": 4.502664413826928,
+      "learning_rate": 2e-05,
+      "loss": 0.8243,
+      "step": 14790
+    },
+    {
+      "epoch": 0.3520660763591355,
+      "grad_norm": 4.055794561640549,
+      "learning_rate": 2e-05,
+      "loss": 0.577,
+      "step": 14791
+    },
+    {
+      "epoch": 0.352089879082167,
+      "grad_norm": 1.8663769813322395,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 14792
+    },
+    {
+      "epoch": 0.35211368180519853,
+      "grad_norm": 3.1935000711065222,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 14793
+    },
+    {
+      "epoch": 0.35213748452823,
+      "grad_norm": 11.019288998060896,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 14794
+    },
+    {
+      "epoch": 0.35216128725126156,
+      "grad_norm": 1.2671061624945072,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 14795
+    },
+    {
+      "epoch": 0.35218508997429304,
+      "grad_norm": 2.2939054526275333,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 14796
+    },
+    {
+      "epoch": 0.3522088926973246,
+      "grad_norm": 1.1022100492885507,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 14797
+    },
+    {
+      "epoch": 0.35223269542035607,
+      "grad_norm": 4.4236042293234386,
+      "learning_rate": 2e-05,
+      "loss": 0.3188,
+      "step": 14798
+    },
+    {
+      "epoch": 0.3522564981433876,
+      "grad_norm": 5.997604197853146,
+      "learning_rate": 2e-05,
+      "loss": 1.3516,
+      "step": 14799
+    },
+    {
+      "epoch": 0.3522803008664191,
+      "grad_norm": 1.4746496314129882,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 14800
+    },
+    {
+      "epoch": 0.35230410358945063,
+      "grad_norm": 3.080591818826765,
+      "learning_rate": 2e-05,
+      "loss": 0.1302,
+      "step": 14801
+    },
+    {
+      "epoch": 0.3523279063124822,
+      "grad_norm": 11.201574863323085,
+      "learning_rate": 2e-05,
+      "loss": 0.9463,
+      "step": 14802
+    },
+    {
+      "epoch": 0.35235170903551366,
+      "grad_norm": 0.8309917597000328,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 14803
+    },
+    {
+      "epoch": 0.3523755117585452,
+      "grad_norm": 4.524273986866666,
+      "learning_rate": 2e-05,
+      "loss": 0.9217,
+      "step": 14804
+    },
+    {
+      "epoch": 0.3523993144815767,
+      "grad_norm": 4.711145001027085,
+      "learning_rate": 2e-05,
+      "loss": 0.7213,
+      "step": 14805
+    },
+    {
+      "epoch": 0.3524231172046082,
+      "grad_norm": 4.209870132343747,
+      "learning_rate": 2e-05,
+      "loss": 0.5666,
+      "step": 14806
+    },
+    {
+      "epoch": 0.3524469199276397,
+      "grad_norm": 3.7049264550852916,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 14807
+    },
+    {
+      "epoch": 0.35247072265067125,
+      "grad_norm": 2.0638947546767987,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 14808
+    },
+    {
+      "epoch": 0.35249452537370274,
+      "grad_norm": 3.404480417339602,
+      "learning_rate": 2e-05,
+      "loss": 0.7772,
+      "step": 14809
+    },
+    {
+      "epoch": 0.3525183280967343,
+      "grad_norm": 0.5542543980589775,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 14810
+    },
+    {
+      "epoch": 0.35254213081976576,
+      "grad_norm": 20.46610934059052,
+      "learning_rate": 2e-05,
+      "loss": 2.8185,
+      "step": 14811
+    },
+    {
+      "epoch": 0.3525659335427973,
+      "grad_norm": 18.887986476666615,
+      "learning_rate": 2e-05,
+      "loss": 1.4884,
+      "step": 14812
+    },
+    {
+      "epoch": 0.3525897362658288,
+      "grad_norm": 4.199905109602147,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 14813
+    },
+    {
+      "epoch": 0.35261353898886033,
+      "grad_norm": 0.4622847974137136,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 14814
+    },
+    {
+      "epoch": 0.3526373417118918,
+      "grad_norm": 2.837999429137783,
+      "learning_rate": 2e-05,
+      "loss": 0.3768,
+      "step": 14815
+    },
+    {
+      "epoch": 0.35266114443492336,
+      "grad_norm": 8.83697318453414,
+      "learning_rate": 2e-05,
+      "loss": 0.4769,
+      "step": 14816
+    },
+    {
+      "epoch": 0.35268494715795484,
+      "grad_norm": 2.8530733496094567,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 14817
+    },
+    {
+      "epoch": 0.3527087498809864,
+      "grad_norm": 5.0975197447493015,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 14818
+    },
+    {
+      "epoch": 0.3527325526040179,
+      "grad_norm": 6.967686176130067,
+      "learning_rate": 2e-05,
+      "loss": 0.5045,
+      "step": 14819
+    },
+    {
+      "epoch": 0.3527563553270494,
+      "grad_norm": 5.798067489790293,
+      "learning_rate": 2e-05,
+      "loss": 0.7456,
+      "step": 14820
+    },
+    {
+      "epoch": 0.35278015805008095,
+      "grad_norm": 0.6391977440573803,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 14821
+    },
+    {
+      "epoch": 0.35280396077311243,
+      "grad_norm": 4.482126833215094,
+      "learning_rate": 2e-05,
+      "loss": 0.5525,
+      "step": 14822
+    },
+    {
+      "epoch": 0.352827763496144,
+      "grad_norm": 9.931888289294374,
+      "learning_rate": 2e-05,
+      "loss": 0.4829,
+      "step": 14823
+    },
+    {
+      "epoch": 0.35285156621917546,
+      "grad_norm": 5.905817913297451,
+      "learning_rate": 2e-05,
+      "loss": 0.7475,
+      "step": 14824
+    },
+    {
+      "epoch": 0.352875368942207,
+      "grad_norm": 2.9542324966621574,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 14825
+    },
+    {
+      "epoch": 0.3528991716652385,
+      "grad_norm": 8.949857783986092,
+      "learning_rate": 2e-05,
+      "loss": 0.4882,
+      "step": 14826
+    },
+    {
+      "epoch": 0.35292297438827,
+      "grad_norm": 9.310264884329763,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 14827
+    },
+    {
+      "epoch": 0.3529467771113015,
+      "grad_norm": 6.9076363016982745,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 14828
+    },
+    {
+      "epoch": 0.35297057983433305,
+      "grad_norm": 4.500179760323942,
+      "learning_rate": 2e-05,
+      "loss": 0.1909,
+      "step": 14829
+    },
+    {
+      "epoch": 0.35299438255736454,
+      "grad_norm": 16.94110794229165,
+      "learning_rate": 2e-05,
+      "loss": 0.8531,
+      "step": 14830
+    },
+    {
+      "epoch": 0.3530181852803961,
+      "grad_norm": 5.899673523799611,
+      "learning_rate": 2e-05,
+      "loss": 0.8714,
+      "step": 14831
+    },
+    {
+      "epoch": 0.35304198800342756,
+      "grad_norm": 3.3348915772581713,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 14832
+    },
+    {
+      "epoch": 0.3530657907264591,
+      "grad_norm": 3.2041565180080402,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 14833
+    },
+    {
+      "epoch": 0.35308959344949065,
+      "grad_norm": 4.913660014903453,
+      "learning_rate": 2e-05,
+      "loss": 0.1681,
+      "step": 14834
+    },
+    {
+      "epoch": 0.35311339617252213,
+      "grad_norm": 3.4750272035606593,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 14835
+    },
+    {
+      "epoch": 0.3531371988955537,
+      "grad_norm": 4.501180325731107,
+      "learning_rate": 2e-05,
+      "loss": 0.2998,
+      "step": 14836
+    },
+    {
+      "epoch": 0.35316100161858516,
+      "grad_norm": 2.400942626242387,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 14837
+    },
+    {
+      "epoch": 0.3531848043416167,
+      "grad_norm": 1.8276080320122967,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 14838
+    },
+    {
+      "epoch": 0.3532086070646482,
+      "grad_norm": 12.863045259180476,
+      "learning_rate": 2e-05,
+      "loss": 1.0646,
+      "step": 14839
+    },
+    {
+      "epoch": 0.3532324097876797,
+      "grad_norm": 7.7568944132517395,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 14840
+    },
+    {
+      "epoch": 0.3532562125107112,
+      "grad_norm": 2.2863156512769423,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 14841
+    },
+    {
+      "epoch": 0.35328001523374275,
+      "grad_norm": 4.2351285910473875,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 14842
+    },
+    {
+      "epoch": 0.35330381795677424,
+      "grad_norm": 1.1728407388584507,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 14843
+    },
+    {
+      "epoch": 0.3533276206798058,
+      "grad_norm": 2.2767481293647434,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 14844
+    },
+    {
+      "epoch": 0.35335142340283726,
+      "grad_norm": 7.454802933982428,
+      "learning_rate": 2e-05,
+      "loss": 1.695,
+      "step": 14845
+    },
+    {
+      "epoch": 0.3533752261258688,
+      "grad_norm": 2.15136636641437,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 14846
+    },
+    {
+      "epoch": 0.3533990288489003,
+      "grad_norm": 5.920899966156419,
+      "learning_rate": 2e-05,
+      "loss": 0.2871,
+      "step": 14847
+    },
+    {
+      "epoch": 0.35342283157193183,
+      "grad_norm": 5.185181716695623,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 14848
+    },
+    {
+      "epoch": 0.35344663429496337,
+      "grad_norm": 2.956017269626208,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 14849
+    },
+    {
+      "epoch": 0.35347043701799485,
+      "grad_norm": 1.8003376672395464,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 14850
+    },
+    {
+      "epoch": 0.3534942397410264,
+      "grad_norm": 5.320551290634136,
+      "learning_rate": 2e-05,
+      "loss": 0.4349,
+      "step": 14851
+    },
+    {
+      "epoch": 0.3535180424640579,
+      "grad_norm": 3.8563183183878826,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 14852
+    },
+    {
+      "epoch": 0.3535418451870894,
+      "grad_norm": 5.081086596634944,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 14853
+    },
+    {
+      "epoch": 0.3535656479101209,
+      "grad_norm": 1.5325329828908238,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 14854
+    },
+    {
+      "epoch": 0.35358945063315245,
+      "grad_norm": 6.685423201303927,
+      "learning_rate": 2e-05,
+      "loss": 1.0881,
+      "step": 14855
+    },
+    {
+      "epoch": 0.35361325335618393,
+      "grad_norm": 2.7665904462360182,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 14856
+    },
+    {
+      "epoch": 0.3536370560792155,
+      "grad_norm": 7.585456347818611,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 14857
+    },
+    {
+      "epoch": 0.35366085880224696,
+      "grad_norm": 5.186120869164594,
+      "learning_rate": 2e-05,
+      "loss": 0.4976,
+      "step": 14858
+    },
+    {
+      "epoch": 0.3536846615252785,
+      "grad_norm": 12.834695238864267,
+      "learning_rate": 2e-05,
+      "loss": 0.6223,
+      "step": 14859
+    },
+    {
+      "epoch": 0.35370846424831,
+      "grad_norm": 3.004050977597216,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 14860
+    },
+    {
+      "epoch": 0.3537322669713415,
+      "grad_norm": 4.482527765627345,
+      "learning_rate": 2e-05,
+      "loss": 0.6146,
+      "step": 14861
+    },
+    {
+      "epoch": 0.353756069694373,
+      "grad_norm": 9.250829307003812,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 14862
+    },
+    {
+      "epoch": 0.35377987241740455,
+      "grad_norm": 5.145107908356869,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 14863
+    },
+    {
+      "epoch": 0.3538036751404361,
+      "grad_norm": 7.928724414193917,
+      "learning_rate": 2e-05,
+      "loss": 0.5621,
+      "step": 14864
+    },
+    {
+      "epoch": 0.3538274778634676,
+      "grad_norm": 7.869837060043589,
+      "learning_rate": 2e-05,
+      "loss": 0.8298,
+      "step": 14865
+    },
+    {
+      "epoch": 0.3538512805864991,
+      "grad_norm": 5.029127519269763,
+      "learning_rate": 2e-05,
+      "loss": 0.1123,
+      "step": 14866
+    },
+    {
+      "epoch": 0.3538750833095306,
+      "grad_norm": 8.015536464744738,
+      "learning_rate": 2e-05,
+      "loss": 0.1569,
+      "step": 14867
+    },
+    {
+      "epoch": 0.35389888603256214,
+      "grad_norm": 2.422541616881017,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 14868
+    },
+    {
+      "epoch": 0.35392268875559363,
+      "grad_norm": 1.5926939687086294,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 14869
+    },
+    {
+      "epoch": 0.35394649147862517,
+      "grad_norm": 0.7712938993455009,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35397029420165665,
+      "grad_norm": 2.6448828877893424,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 14871
+    },
+    {
+      "epoch": 0.3539940969246882,
+      "grad_norm": 2.3046181754664596,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 14872
+    },
+    {
+      "epoch": 0.3540178996477197,
+      "grad_norm": 11.4475041903709,
+      "learning_rate": 2e-05,
+      "loss": 0.6105,
+      "step": 14873
+    },
+    {
+      "epoch": 0.3540417023707512,
+      "grad_norm": 5.667439015460466,
+      "learning_rate": 2e-05,
+      "loss": 0.4861,
+      "step": 14874
+    },
+    {
+      "epoch": 0.3540655050937827,
+      "grad_norm": 2.2298072159947497,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 14875
+    },
+    {
+      "epoch": 0.35408930781681425,
+      "grad_norm": 5.583698566038993,
+      "learning_rate": 2e-05,
+      "loss": 0.7313,
+      "step": 14876
+    },
+    {
+      "epoch": 0.35411311053984573,
+      "grad_norm": 7.663946809791746,
+      "learning_rate": 2e-05,
+      "loss": 0.3625,
+      "step": 14877
+    },
+    {
+      "epoch": 0.3541369132628773,
+      "grad_norm": 72.22127894052595,
+      "learning_rate": 2e-05,
+      "loss": 0.5302,
+      "step": 14878
+    },
+    {
+      "epoch": 0.3541607159859088,
+      "grad_norm": 1.9559804285664637,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 14879
+    },
+    {
+      "epoch": 0.3541845187089403,
+      "grad_norm": 6.728969178608326,
+      "learning_rate": 2e-05,
+      "loss": 0.9808,
+      "step": 14880
+    },
+    {
+      "epoch": 0.35420832143197184,
+      "grad_norm": 0.27153684523424787,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 14881
+    },
+    {
+      "epoch": 0.3542321241550033,
+      "grad_norm": 5.539998813247812,
+      "learning_rate": 2e-05,
+      "loss": 1.3052,
+      "step": 14882
+    },
+    {
+      "epoch": 0.35425592687803487,
+      "grad_norm": 2.2332579566471273,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 14883
+    },
+    {
+      "epoch": 0.35427972960106635,
+      "grad_norm": 5.3486191353953885,
+      "learning_rate": 2e-05,
+      "loss": 0.7785,
+      "step": 14884
+    },
+    {
+      "epoch": 0.3543035323240979,
+      "grad_norm": 3.015203195850541,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 14885
+    },
+    {
+      "epoch": 0.3543273350471294,
+      "grad_norm": 12.28986033592616,
+      "learning_rate": 2e-05,
+      "loss": 1.1519,
+      "step": 14886
+    },
+    {
+      "epoch": 0.3543511377701609,
+      "grad_norm": 7.683577812753437,
+      "learning_rate": 2e-05,
+      "loss": 0.3224,
+      "step": 14887
+    },
+    {
+      "epoch": 0.3543749404931924,
+      "grad_norm": 2.4084319685718465,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 14888
+    },
+    {
+      "epoch": 0.35439874321622394,
+      "grad_norm": 6.934636960785544,
+      "learning_rate": 2e-05,
+      "loss": 0.1364,
+      "step": 14889
+    },
+    {
+      "epoch": 0.35442254593925543,
+      "grad_norm": 5.608975526181507,
+      "learning_rate": 2e-05,
+      "loss": 0.6702,
+      "step": 14890
+    },
+    {
+      "epoch": 0.35444634866228697,
+      "grad_norm": 4.640257875427424,
+      "learning_rate": 2e-05,
+      "loss": 0.8978,
+      "step": 14891
+    },
+    {
+      "epoch": 0.35447015138531845,
+      "grad_norm": 4.77519984051623,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 14892
+    },
+    {
+      "epoch": 0.35449395410835,
+      "grad_norm": 3.725452983849088,
+      "learning_rate": 2e-05,
+      "loss": 0.5777,
+      "step": 14893
+    },
+    {
+      "epoch": 0.35451775683138154,
+      "grad_norm": 2.601582771679916,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 14894
+    },
+    {
+      "epoch": 0.354541559554413,
+      "grad_norm": 2.5270686359390635,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 14895
+    },
+    {
+      "epoch": 0.35456536227744456,
+      "grad_norm": 2.6402828862573187,
+      "learning_rate": 2e-05,
+      "loss": 0.137,
+      "step": 14896
+    },
+    {
+      "epoch": 0.35458916500047605,
+      "grad_norm": 3.414419528671193,
+      "learning_rate": 2e-05,
+      "loss": 0.2001,
+      "step": 14897
+    },
+    {
+      "epoch": 0.3546129677235076,
+      "grad_norm": 1.2084835194835253,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 14898
+    },
+    {
+      "epoch": 0.3546367704465391,
+      "grad_norm": 2.30101757468934,
+      "learning_rate": 2e-05,
+      "loss": 0.1307,
+      "step": 14899
+    },
+    {
+      "epoch": 0.3546605731695706,
+      "grad_norm": 5.264822723094008,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 14900
+    },
+    {
+      "epoch": 0.3546843758926021,
+      "grad_norm": 1.1323548081563997,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 14901
+    },
+    {
+      "epoch": 0.35470817861563364,
+      "grad_norm": 2.766924390226127,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 14902
+    },
+    {
+      "epoch": 0.3547319813386651,
+      "grad_norm": 3.658674081729795,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 14903
+    },
+    {
+      "epoch": 0.35475578406169667,
+      "grad_norm": 3.940908337067807,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 14904
+    },
+    {
+      "epoch": 0.35477958678472815,
+      "grad_norm": 2.97687509160788,
+      "learning_rate": 2e-05,
+      "loss": 0.6716,
+      "step": 14905
+    },
+    {
+      "epoch": 0.3548033895077597,
+      "grad_norm": 3.852687818511012,
+      "learning_rate": 2e-05,
+      "loss": 0.2096,
+      "step": 14906
+    },
+    {
+      "epoch": 0.3548271922307912,
+      "grad_norm": 3.523868826431609,
+      "learning_rate": 2e-05,
+      "loss": 0.6066,
+      "step": 14907
+    },
+    {
+      "epoch": 0.3548509949538227,
+      "grad_norm": 2.0504808857807904,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 14908
+    },
+    {
+      "epoch": 0.35487479767685426,
+      "grad_norm": 3.4673975147976317,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 14909
+    },
+    {
+      "epoch": 0.35489860039988574,
+      "grad_norm": 4.570818096303651,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 14910
+    },
+    {
+      "epoch": 0.3549224031229173,
+      "grad_norm": 4.79470929261419,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 14911
+    },
+    {
+      "epoch": 0.35494620584594877,
+      "grad_norm": 8.118146535316388,
+      "learning_rate": 2e-05,
+      "loss": 0.7262,
+      "step": 14912
+    },
+    {
+      "epoch": 0.3549700085689803,
+      "grad_norm": 1.1093029719281566,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 14913
+    },
+    {
+      "epoch": 0.3549938112920118,
+      "grad_norm": 5.585546313255232,
+      "learning_rate": 2e-05,
+      "loss": 0.1977,
+      "step": 14914
+    },
+    {
+      "epoch": 0.35501761401504334,
+      "grad_norm": 4.5464528504985635,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 14915
+    },
+    {
+      "epoch": 0.3550414167380748,
+      "grad_norm": 10.377735234418244,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 14916
+    },
+    {
+      "epoch": 0.35506521946110636,
+      "grad_norm": 1.056940082493548,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 14917
+    },
+    {
+      "epoch": 0.35508902218413785,
+      "grad_norm": 4.809182527759392,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 14918
+    },
+    {
+      "epoch": 0.3551128249071694,
+      "grad_norm": 18.65610305903548,
+      "learning_rate": 2e-05,
+      "loss": 2.36,
+      "step": 14919
+    },
+    {
+      "epoch": 0.3551366276302009,
+      "grad_norm": 1.7006770916744198,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 14920
+    },
+    {
+      "epoch": 0.3551604303532324,
+      "grad_norm": 4.808058417863498,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 14921
+    },
+    {
+      "epoch": 0.3551842330762639,
+      "grad_norm": 1.6150576574300004,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 14922
+    },
+    {
+      "epoch": 0.35520803579929544,
+      "grad_norm": 5.529480582534429,
+      "learning_rate": 2e-05,
+      "loss": 0.6743,
+      "step": 14923
+    },
+    {
+      "epoch": 0.355231838522327,
+      "grad_norm": 3.0733518943930562,
+      "learning_rate": 2e-05,
+      "loss": 0.5396,
+      "step": 14924
+    },
+    {
+      "epoch": 0.35525564124535847,
+      "grad_norm": 4.792399908502501,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 14925
+    },
+    {
+      "epoch": 0.35527944396839,
+      "grad_norm": 0.8380399967052985,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 14926
+    },
+    {
+      "epoch": 0.3553032466914215,
+      "grad_norm": 3.5033170843161123,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 14927
+    },
+    {
+      "epoch": 0.35532704941445303,
+      "grad_norm": 4.880162238368949,
+      "learning_rate": 2e-05,
+      "loss": 0.8186,
+      "step": 14928
+    },
+    {
+      "epoch": 0.3553508521374845,
+      "grad_norm": 1.1058182368363036,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 14929
+    },
+    {
+      "epoch": 0.35537465486051606,
+      "grad_norm": 1.3654652509266993,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35539845758354754,
+      "grad_norm": 3.628496177911362,
+      "learning_rate": 2e-05,
+      "loss": 0.7788,
+      "step": 14931
+    },
+    {
+      "epoch": 0.3554222603065791,
+      "grad_norm": 3.7681343936279825,
+      "learning_rate": 2e-05,
+      "loss": 0.3866,
+      "step": 14932
+    },
+    {
+      "epoch": 0.35544606302961057,
+      "grad_norm": 3.1947379130058184,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 14933
+    },
+    {
+      "epoch": 0.3554698657526421,
+      "grad_norm": 2.7805709186949645,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 14934
+    },
+    {
+      "epoch": 0.3554936684756736,
+      "grad_norm": 1.7661589825462514,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 14935
+    },
+    {
+      "epoch": 0.35551747119870514,
+      "grad_norm": 9.885636326383596,
+      "learning_rate": 2e-05,
+      "loss": 0.9302,
+      "step": 14936
+    },
+    {
+      "epoch": 0.3555412739217366,
+      "grad_norm": 5.445578258122017,
+      "learning_rate": 2e-05,
+      "loss": 0.9904,
+      "step": 14937
+    },
+    {
+      "epoch": 0.35556507664476816,
+      "grad_norm": 2.882841453349942,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 14938
+    },
+    {
+      "epoch": 0.3555888793677997,
+      "grad_norm": 4.550214754709906,
+      "learning_rate": 2e-05,
+      "loss": 0.6196,
+      "step": 14939
+    },
+    {
+      "epoch": 0.3556126820908312,
+      "grad_norm": 3.896912590494785,
+      "learning_rate": 2e-05,
+      "loss": 0.7145,
+      "step": 14940
+    },
+    {
+      "epoch": 0.35563648481386273,
+      "grad_norm": 4.061851338127341,
+      "learning_rate": 2e-05,
+      "loss": 0.3138,
+      "step": 14941
+    },
+    {
+      "epoch": 0.3556602875368942,
+      "grad_norm": 3.9453017577134912,
+      "learning_rate": 2e-05,
+      "loss": 0.687,
+      "step": 14942
+    },
+    {
+      "epoch": 0.35568409025992576,
+      "grad_norm": 8.105551061559675,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 14943
+    },
+    {
+      "epoch": 0.35570789298295724,
+      "grad_norm": 6.860428934329482,
+      "learning_rate": 2e-05,
+      "loss": 0.2302,
+      "step": 14944
+    },
+    {
+      "epoch": 0.3557316957059888,
+      "grad_norm": 5.861595564228697,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 14945
+    },
+    {
+      "epoch": 0.35575549842902027,
+      "grad_norm": 6.124903297390973,
+      "learning_rate": 2e-05,
+      "loss": 0.6212,
+      "step": 14946
+    },
+    {
+      "epoch": 0.3557793011520518,
+      "grad_norm": 7.576005160736971,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 14947
+    },
+    {
+      "epoch": 0.3558031038750833,
+      "grad_norm": 1.8787027489908712,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 14948
+    },
+    {
+      "epoch": 0.35582690659811483,
+      "grad_norm": 2.7464001983643924,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 14949
+    },
+    {
+      "epoch": 0.3558507093211463,
+      "grad_norm": 4.382340731816808,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 14950
+    },
+    {
+      "epoch": 0.35587451204417786,
+      "grad_norm": 4.529238128052419,
+      "learning_rate": 2e-05,
+      "loss": 0.5428,
+      "step": 14951
+    },
+    {
+      "epoch": 0.35589831476720934,
+      "grad_norm": 3.094557348234888,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 14952
+    },
+    {
+      "epoch": 0.3559221174902409,
+      "grad_norm": 2.8397262988257266,
+      "learning_rate": 2e-05,
+      "loss": 0.5128,
+      "step": 14953
+    },
+    {
+      "epoch": 0.35594592021327237,
+      "grad_norm": 5.512989536129957,
+      "learning_rate": 2e-05,
+      "loss": 0.3637,
+      "step": 14954
+    },
+    {
+      "epoch": 0.3559697229363039,
+      "grad_norm": 0.7951548649385449,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 14955
+    },
+    {
+      "epoch": 0.35599352565933545,
+      "grad_norm": 1.7724117536478423,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 14956
+    },
+    {
+      "epoch": 0.35601732838236694,
+      "grad_norm": 2.1269450140062123,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 14957
+    },
+    {
+      "epoch": 0.3560411311053985,
+      "grad_norm": 6.153552548714198,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 14958
+    },
+    {
+      "epoch": 0.35606493382842996,
+      "grad_norm": 6.539542986379896,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 14959
+    },
+    {
+      "epoch": 0.3560887365514615,
+      "grad_norm": 1.138234348662368,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 14960
+    },
+    {
+      "epoch": 0.356112539274493,
+      "grad_norm": 2.9406256754730893,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 14961
+    },
+    {
+      "epoch": 0.35613634199752453,
+      "grad_norm": 4.518986502068385,
+      "learning_rate": 2e-05,
+      "loss": 0.305,
+      "step": 14962
+    },
+    {
+      "epoch": 0.356160144720556,
+      "grad_norm": 2.057178309116451,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 14963
+    },
+    {
+      "epoch": 0.35618394744358756,
+      "grad_norm": 6.913735212856376,
+      "learning_rate": 2e-05,
+      "loss": 0.3459,
+      "step": 14964
+    },
+    {
+      "epoch": 0.35620775016661904,
+      "grad_norm": 4.703975613604976,
+      "learning_rate": 2e-05,
+      "loss": 0.5745,
+      "step": 14965
+    },
+    {
+      "epoch": 0.3562315528896506,
+      "grad_norm": 2.05470536575996,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 14966
+    },
+    {
+      "epoch": 0.35625535561268207,
+      "grad_norm": 5.036863910178232,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 14967
+    },
+    {
+      "epoch": 0.3562791583357136,
+      "grad_norm": 5.748963967201365,
+      "learning_rate": 2e-05,
+      "loss": 0.6234,
+      "step": 14968
+    },
+    {
+      "epoch": 0.3563029610587451,
+      "grad_norm": 14.277813559605896,
+      "learning_rate": 2e-05,
+      "loss": 0.3401,
+      "step": 14969
+    },
+    {
+      "epoch": 0.35632676378177663,
+      "grad_norm": 2.23453774154281,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 14970
+    },
+    {
+      "epoch": 0.3563505665048082,
+      "grad_norm": 3.1655862169382827,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 14971
+    },
+    {
+      "epoch": 0.35637436922783966,
+      "grad_norm": 4.247133852099305,
+      "learning_rate": 2e-05,
+      "loss": 0.1584,
+      "step": 14972
+    },
+    {
+      "epoch": 0.3563981719508712,
+      "grad_norm": 4.702198709693508,
+      "learning_rate": 2e-05,
+      "loss": 0.6065,
+      "step": 14973
+    },
+    {
+      "epoch": 0.3564219746739027,
+      "grad_norm": 10.387588408714526,
+      "learning_rate": 2e-05,
+      "loss": 0.8583,
+      "step": 14974
+    },
+    {
+      "epoch": 0.3564457773969342,
+      "grad_norm": 1.6144922350866173,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 14975
+    },
+    {
+      "epoch": 0.3564695801199657,
+      "grad_norm": 0.45675959480428224,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 14976
+    },
+    {
+      "epoch": 0.35649338284299725,
+      "grad_norm": 2.845045646059957,
+      "learning_rate": 2e-05,
+      "loss": 0.3992,
+      "step": 14977
+    },
+    {
+      "epoch": 0.35651718556602874,
+      "grad_norm": 1.3989234976330487,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 14978
+    },
+    {
+      "epoch": 0.3565409882890603,
+      "grad_norm": 7.379916903387184,
+      "learning_rate": 2e-05,
+      "loss": 0.7102,
+      "step": 14979
+    },
+    {
+      "epoch": 0.35656479101209176,
+      "grad_norm": 0.8412955217849948,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 14980
+    },
+    {
+      "epoch": 0.3565885937351233,
+      "grad_norm": 2.1187783757419063,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 14981
+    },
+    {
+      "epoch": 0.3566123964581548,
+      "grad_norm": 4.226377580035813,
+      "learning_rate": 2e-05,
+      "loss": 0.5777,
+      "step": 14982
+    },
+    {
+      "epoch": 0.35663619918118633,
+      "grad_norm": 14.282279637950339,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 14983
+    },
+    {
+      "epoch": 0.3566600019042178,
+      "grad_norm": 4.555056517889762,
+      "learning_rate": 2e-05,
+      "loss": 0.4255,
+      "step": 14984
+    },
+    {
+      "epoch": 0.35668380462724936,
+      "grad_norm": 4.5531806644161374,
+      "learning_rate": 2e-05,
+      "loss": 0.5555,
+      "step": 14985
+    },
+    {
+      "epoch": 0.3567076073502809,
+      "grad_norm": 6.764807733858228,
+      "learning_rate": 2e-05,
+      "loss": 1.3991,
+      "step": 14986
+    },
+    {
+      "epoch": 0.3567314100733124,
+      "grad_norm": 1.008957365892496,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 14987
+    },
+    {
+      "epoch": 0.3567552127963439,
+      "grad_norm": 5.1064484311200165,
+      "learning_rate": 2e-05,
+      "loss": 0.3792,
+      "step": 14988
+    },
+    {
+      "epoch": 0.3567790155193754,
+      "grad_norm": 0.1852726647387532,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 14989
+    },
+    {
+      "epoch": 0.35680281824240695,
+      "grad_norm": 26.68531249040798,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 14990
+    },
+    {
+      "epoch": 0.35682662096543843,
+      "grad_norm": 4.290937353560896,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 14991
+    },
+    {
+      "epoch": 0.35685042368847,
+      "grad_norm": 0.5888438556462627,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 14992
+    },
+    {
+      "epoch": 0.35687422641150146,
+      "grad_norm": 9.1822306279998,
+      "learning_rate": 2e-05,
+      "loss": 0.1669,
+      "step": 14993
+    },
+    {
+      "epoch": 0.356898029134533,
+      "grad_norm": 4.853569992592126,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 14994
+    },
+    {
+      "epoch": 0.3569218318575645,
+      "grad_norm": 4.751465966909361,
+      "learning_rate": 2e-05,
+      "loss": 0.8527,
+      "step": 14995
+    },
+    {
+      "epoch": 0.356945634580596,
+      "grad_norm": 6.652833365808084,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 14996
+    },
+    {
+      "epoch": 0.3569694373036275,
+      "grad_norm": 1.6126554415900238,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 14997
+    },
+    {
+      "epoch": 0.35699324002665905,
+      "grad_norm": 2.046995263430253,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 14998
+    },
+    {
+      "epoch": 0.35701704274969054,
+      "grad_norm": 4.604689880530445,
+      "learning_rate": 2e-05,
+      "loss": 0.4325,
+      "step": 14999
+    },
+    {
+      "epoch": 0.3570408454727221,
+      "grad_norm": 5.832731748530879,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 15000
+    },
+    {
+      "epoch": 0.3570646481957536,
+      "grad_norm": 10.051504767559694,
+      "learning_rate": 2e-05,
+      "loss": 1.1887,
+      "step": 15001
+    },
+    {
+      "epoch": 0.3570884509187851,
+      "grad_norm": 4.315521942331959,
+      "learning_rate": 2e-05,
+      "loss": 0.4203,
+      "step": 15002
+    },
+    {
+      "epoch": 0.35711225364181665,
+      "grad_norm": 0.8279236746114728,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 15003
+    },
+    {
+      "epoch": 0.35713605636484813,
+      "grad_norm": 4.417880528449141,
+      "learning_rate": 2e-05,
+      "loss": 0.591,
+      "step": 15004
+    },
+    {
+      "epoch": 0.35715985908787967,
+      "grad_norm": 33.42687537612195,
+      "learning_rate": 2e-05,
+      "loss": 1.0622,
+      "step": 15005
+    },
+    {
+      "epoch": 0.35718366181091116,
+      "grad_norm": 3.5469193416838634,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 15006
+    },
+    {
+      "epoch": 0.3572074645339427,
+      "grad_norm": 2.9467394267984943,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 15007
+    },
+    {
+      "epoch": 0.3572312672569742,
+      "grad_norm": 3.813450005499208,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 15008
+    },
+    {
+      "epoch": 0.3572550699800057,
+      "grad_norm": 1.979450708178698,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 15009
+    },
+    {
+      "epoch": 0.3572788727030372,
+      "grad_norm": 4.622053108176614,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 15010
+    },
+    {
+      "epoch": 0.35730267542606875,
+      "grad_norm": 2.868019122504078,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 15011
+    },
+    {
+      "epoch": 0.35732647814910024,
+      "grad_norm": 8.707438930154387,
+      "learning_rate": 2e-05,
+      "loss": 1.8455,
+      "step": 15012
+    },
+    {
+      "epoch": 0.3573502808721318,
+      "grad_norm": 2.306238521120742,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 15013
+    },
+    {
+      "epoch": 0.35737408359516326,
+      "grad_norm": 19.237611682739235,
+      "learning_rate": 2e-05,
+      "loss": 0.476,
+      "step": 15014
+    },
+    {
+      "epoch": 0.3573978863181948,
+      "grad_norm": 22.84320904588554,
+      "learning_rate": 2e-05,
+      "loss": 0.4529,
+      "step": 15015
+    },
+    {
+      "epoch": 0.35742168904122634,
+      "grad_norm": 14.732074098863032,
+      "learning_rate": 2e-05,
+      "loss": 0.8255,
+      "step": 15016
+    },
+    {
+      "epoch": 0.35744549176425783,
+      "grad_norm": 0.5206927495364446,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 15017
+    },
+    {
+      "epoch": 0.35746929448728937,
+      "grad_norm": 12.893167802827707,
+      "learning_rate": 2e-05,
+      "loss": 0.8857,
+      "step": 15018
+    },
+    {
+      "epoch": 0.35749309721032085,
+      "grad_norm": 8.6605184048545,
+      "learning_rate": 2e-05,
+      "loss": 0.5828,
+      "step": 15019
+    },
+    {
+      "epoch": 0.3575168999333524,
+      "grad_norm": 9.021236676333091,
+      "learning_rate": 2e-05,
+      "loss": 1.2214,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3575407026563839,
+      "grad_norm": 5.6712329000875625,
+      "learning_rate": 2e-05,
+      "loss": 0.7055,
+      "step": 15021
+    },
+    {
+      "epoch": 0.3575645053794154,
+      "grad_norm": 1.1881121762923637,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 15022
+    },
+    {
+      "epoch": 0.3575883081024469,
+      "grad_norm": 2.3267070466137323,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 15023
+    },
+    {
+      "epoch": 0.35761211082547845,
+      "grad_norm": 4.368686146775539,
+      "learning_rate": 2e-05,
+      "loss": 0.7235,
+      "step": 15024
+    },
+    {
+      "epoch": 0.35763591354850993,
+      "grad_norm": 11.33732264893661,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 15025
+    },
+    {
+      "epoch": 0.3576597162715415,
+      "grad_norm": 4.993584505735662,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 15026
+    },
+    {
+      "epoch": 0.35768351899457296,
+      "grad_norm": 7.4865339168168,
+      "learning_rate": 2e-05,
+      "loss": 1.2492,
+      "step": 15027
+    },
+    {
+      "epoch": 0.3577073217176045,
+      "grad_norm": 27.71010498934004,
+      "learning_rate": 2e-05,
+      "loss": 1.128,
+      "step": 15028
+    },
+    {
+      "epoch": 0.357731124440636,
+      "grad_norm": 4.3770421982180645,
+      "learning_rate": 2e-05,
+      "loss": 0.4664,
+      "step": 15029
+    },
+    {
+      "epoch": 0.3577549271636675,
+      "grad_norm": 4.435397882887587,
+      "learning_rate": 2e-05,
+      "loss": 0.5541,
+      "step": 15030
+    },
+    {
+      "epoch": 0.35777872988669907,
+      "grad_norm": 1.0507825480784945,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 15031
+    },
+    {
+      "epoch": 0.35780253260973055,
+      "grad_norm": 7.689038750538178,
+      "learning_rate": 2e-05,
+      "loss": 0.8025,
+      "step": 15032
+    },
+    {
+      "epoch": 0.3578263353327621,
+      "grad_norm": 10.803960458036025,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 15033
+    },
+    {
+      "epoch": 0.3578501380557936,
+      "grad_norm": 1.6635447581326448,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 15034
+    },
+    {
+      "epoch": 0.3578739407788251,
+      "grad_norm": 4.015687184023133,
+      "learning_rate": 2e-05,
+      "loss": 0.6479,
+      "step": 15035
+    },
+    {
+      "epoch": 0.3578977435018566,
+      "grad_norm": 4.544772456652363,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 15036
+    },
+    {
+      "epoch": 0.35792154622488814,
+      "grad_norm": 2.668872464660625,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 15037
+    },
+    {
+      "epoch": 0.35794534894791963,
+      "grad_norm": 6.031579092867822,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 15038
+    },
+    {
+      "epoch": 0.35796915167095117,
+      "grad_norm": 2.970154400790856,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 15039
+    },
+    {
+      "epoch": 0.35799295439398265,
+      "grad_norm": 6.07460533127566,
+      "learning_rate": 2e-05,
+      "loss": 0.8132,
+      "step": 15040
+    },
+    {
+      "epoch": 0.3580167571170142,
+      "grad_norm": 5.764688397533166,
+      "learning_rate": 2e-05,
+      "loss": 1.4403,
+      "step": 15041
+    },
+    {
+      "epoch": 0.3580405598400457,
+      "grad_norm": 4.6961868686714485,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 15042
+    },
+    {
+      "epoch": 0.3580643625630772,
+      "grad_norm": 2.813676709754298,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 15043
+    },
+    {
+      "epoch": 0.3580881652861087,
+      "grad_norm": 4.006090677872431,
+      "learning_rate": 2e-05,
+      "loss": 0.6732,
+      "step": 15044
+    },
+    {
+      "epoch": 0.35811196800914025,
+      "grad_norm": 5.083115746760049,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 15045
+    },
+    {
+      "epoch": 0.3581357707321718,
+      "grad_norm": 0.6201962750872613,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 15046
+    },
+    {
+      "epoch": 0.3581595734552033,
+      "grad_norm": 14.999682298476985,
+      "learning_rate": 2e-05,
+      "loss": 0.5971,
+      "step": 15047
+    },
+    {
+      "epoch": 0.3581833761782348,
+      "grad_norm": 3.258098256675557,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 15048
+    },
+    {
+      "epoch": 0.3582071789012663,
+      "grad_norm": 1.213738324404733,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 15049
+    },
+    {
+      "epoch": 0.35823098162429784,
+      "grad_norm": 5.443758583201389,
+      "learning_rate": 2e-05,
+      "loss": 0.3127,
+      "step": 15050
+    },
+    {
+      "epoch": 0.3582547843473293,
+      "grad_norm": 3.3238178948455137,
+      "learning_rate": 2e-05,
+      "loss": 0.1963,
+      "step": 15051
+    },
+    {
+      "epoch": 0.35827858707036087,
+      "grad_norm": 1.238914324758141,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 15052
+    },
+    {
+      "epoch": 0.35830238979339235,
+      "grad_norm": 1.1333543978681675,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 15053
+    },
+    {
+      "epoch": 0.3583261925164239,
+      "grad_norm": 5.17114080037074,
+      "learning_rate": 2e-05,
+      "loss": 1.0475,
+      "step": 15054
+    },
+    {
+      "epoch": 0.3583499952394554,
+      "grad_norm": 4.025272585837211,
+      "learning_rate": 2e-05,
+      "loss": 0.4147,
+      "step": 15055
+    },
+    {
+      "epoch": 0.3583737979624869,
+      "grad_norm": 0.6772002153081984,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 15056
+    },
+    {
+      "epoch": 0.3583976006855184,
+      "grad_norm": 1.2654116266971391,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 15057
+    },
+    {
+      "epoch": 0.35842140340854994,
+      "grad_norm": 4.223253995196336,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 15058
+    },
+    {
+      "epoch": 0.35844520613158143,
+      "grad_norm": 7.009859918457651,
+      "learning_rate": 2e-05,
+      "loss": 0.3056,
+      "step": 15059
+    },
+    {
+      "epoch": 0.35846900885461297,
+      "grad_norm": 9.828788273534025,
+      "learning_rate": 2e-05,
+      "loss": 0.4568,
+      "step": 15060
+    },
+    {
+      "epoch": 0.3584928115776445,
+      "grad_norm": 4.308537363523904,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 15061
+    },
+    {
+      "epoch": 0.358516614300676,
+      "grad_norm": 6.905709456253661,
+      "learning_rate": 2e-05,
+      "loss": 1.1001,
+      "step": 15062
+    },
+    {
+      "epoch": 0.35854041702370754,
+      "grad_norm": 7.315183363213172,
+      "learning_rate": 2e-05,
+      "loss": 2.3021,
+      "step": 15063
+    },
+    {
+      "epoch": 0.358564219746739,
+      "grad_norm": 2.9195704608415847,
+      "learning_rate": 2e-05,
+      "loss": 0.1243,
+      "step": 15064
+    },
+    {
+      "epoch": 0.35858802246977056,
+      "grad_norm": 2.29155130051846,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 15065
+    },
+    {
+      "epoch": 0.35861182519280205,
+      "grad_norm": 1.1582514411093578,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 15066
+    },
+    {
+      "epoch": 0.3586356279158336,
+      "grad_norm": 1.830874146958654,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 15067
+    },
+    {
+      "epoch": 0.3586594306388651,
+      "grad_norm": 1.2135334214652393,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 15068
+    },
+    {
+      "epoch": 0.3586832333618966,
+      "grad_norm": 4.103643350395614,
+      "learning_rate": 2e-05,
+      "loss": 0.3736,
+      "step": 15069
+    },
+    {
+      "epoch": 0.3587070360849281,
+      "grad_norm": 2.8454652754902283,
+      "learning_rate": 2e-05,
+      "loss": 0.125,
+      "step": 15070
+    },
+    {
+      "epoch": 0.35873083880795964,
+      "grad_norm": 3.0952813669137687,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 15071
+    },
+    {
+      "epoch": 0.3587546415309911,
+      "grad_norm": 2.7810574473952445,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 15072
+    },
+    {
+      "epoch": 0.35877844425402267,
+      "grad_norm": 2.669372661291488,
+      "learning_rate": 2e-05,
+      "loss": 0.2681,
+      "step": 15073
+    },
+    {
+      "epoch": 0.35880224697705415,
+      "grad_norm": 4.72652240110255,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 15074
+    },
+    {
+      "epoch": 0.3588260497000857,
+      "grad_norm": 5.279640001400455,
+      "learning_rate": 2e-05,
+      "loss": 1.0922,
+      "step": 15075
+    },
+    {
+      "epoch": 0.3588498524231172,
+      "grad_norm": 1.4630648972709501,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 15076
+    },
+    {
+      "epoch": 0.3588736551461487,
+      "grad_norm": 1.4740650042626406,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 15077
+    },
+    {
+      "epoch": 0.35889745786918026,
+      "grad_norm": 3.263686818196325,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 15078
+    },
+    {
+      "epoch": 0.35892126059221174,
+      "grad_norm": 4.466373586147613,
+      "learning_rate": 2e-05,
+      "loss": 0.6472,
+      "step": 15079
+    },
+    {
+      "epoch": 0.3589450633152433,
+      "grad_norm": 4.837811831309134,
+      "learning_rate": 2e-05,
+      "loss": 1.022,
+      "step": 15080
+    },
+    {
+      "epoch": 0.35896886603827477,
+      "grad_norm": 5.264875806071422,
+      "learning_rate": 2e-05,
+      "loss": 0.1346,
+      "step": 15081
+    },
+    {
+      "epoch": 0.3589926687613063,
+      "grad_norm": 5.597757573868205,
+      "learning_rate": 2e-05,
+      "loss": 0.5098,
+      "step": 15082
+    },
+    {
+      "epoch": 0.3590164714843378,
+      "grad_norm": 3.1425259823508185,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 15083
+    },
+    {
+      "epoch": 0.35904027420736934,
+      "grad_norm": 5.180103047701133,
+      "learning_rate": 2e-05,
+      "loss": 0.9978,
+      "step": 15084
+    },
+    {
+      "epoch": 0.3590640769304008,
+      "grad_norm": 4.74958430496341,
+      "learning_rate": 2e-05,
+      "loss": 0.4726,
+      "step": 15085
+    },
+    {
+      "epoch": 0.35908787965343236,
+      "grad_norm": 9.23187751975111,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 15086
+    },
+    {
+      "epoch": 0.35911168237646385,
+      "grad_norm": 0.14259213813496324,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15087
+    },
+    {
+      "epoch": 0.3591354850994954,
+      "grad_norm": 4.048851117785542,
+      "learning_rate": 2e-05,
+      "loss": 0.2936,
+      "step": 15088
+    },
+    {
+      "epoch": 0.3591592878225269,
+      "grad_norm": 4.038712596209683,
+      "learning_rate": 2e-05,
+      "loss": 0.1031,
+      "step": 15089
+    },
+    {
+      "epoch": 0.3591830905455584,
+      "grad_norm": 2.5385916687681984,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 15090
+    },
+    {
+      "epoch": 0.3592068932685899,
+      "grad_norm": 2.084437290866862,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 15091
+    },
+    {
+      "epoch": 0.35923069599162144,
+      "grad_norm": 6.53693419612165,
+      "learning_rate": 2e-05,
+      "loss": 0.6086,
+      "step": 15092
+    },
+    {
+      "epoch": 0.359254498714653,
+      "grad_norm": 3.401118439554808,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 15093
+    },
+    {
+      "epoch": 0.35927830143768447,
+      "grad_norm": 2.4918395430557254,
+      "learning_rate": 2e-05,
+      "loss": 0.0893,
+      "step": 15094
+    },
+    {
+      "epoch": 0.359302104160716,
+      "grad_norm": 5.015338551544657,
+      "learning_rate": 2e-05,
+      "loss": 0.7326,
+      "step": 15095
+    },
+    {
+      "epoch": 0.3593259068837475,
+      "grad_norm": 1.6004283228894223,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 15096
+    },
+    {
+      "epoch": 0.35934970960677903,
+      "grad_norm": 3.2503626400421743,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 15097
+    },
+    {
+      "epoch": 0.3593735123298105,
+      "grad_norm": 5.869906065199145,
+      "learning_rate": 2e-05,
+      "loss": 0.6362,
+      "step": 15098
+    },
+    {
+      "epoch": 0.35939731505284206,
+      "grad_norm": 1.6896492398672225,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 15099
+    },
+    {
+      "epoch": 0.35942111777587354,
+      "grad_norm": 2.3356887085979254,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 15100
+    },
+    {
+      "epoch": 0.3594449204989051,
+      "grad_norm": 4.511554928827386,
+      "learning_rate": 2e-05,
+      "loss": 0.6245,
+      "step": 15101
+    },
+    {
+      "epoch": 0.35946872322193657,
+      "grad_norm": 5.463629436920664,
+      "learning_rate": 2e-05,
+      "loss": 0.3472,
+      "step": 15102
+    },
+    {
+      "epoch": 0.3594925259449681,
+      "grad_norm": 4.7226789508398985,
+      "learning_rate": 2e-05,
+      "loss": 0.9874,
+      "step": 15103
+    },
+    {
+      "epoch": 0.3595163286679996,
+      "grad_norm": 13.980651436299537,
+      "learning_rate": 2e-05,
+      "loss": 1.0637,
+      "step": 15104
+    },
+    {
+      "epoch": 0.35954013139103114,
+      "grad_norm": 5.6191321113182084,
+      "learning_rate": 2e-05,
+      "loss": 0.8148,
+      "step": 15105
+    },
+    {
+      "epoch": 0.3595639341140626,
+      "grad_norm": 6.01686815002336,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 15106
+    },
+    {
+      "epoch": 0.35958773683709416,
+      "grad_norm": 4.764125969996663,
+      "learning_rate": 2e-05,
+      "loss": 0.6657,
+      "step": 15107
+    },
+    {
+      "epoch": 0.3596115395601257,
+      "grad_norm": 4.689423072316084,
+      "learning_rate": 2e-05,
+      "loss": 0.5678,
+      "step": 15108
+    },
+    {
+      "epoch": 0.3596353422831572,
+      "grad_norm": 5.587710347747819,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 15109
+    },
+    {
+      "epoch": 0.35965914500618873,
+      "grad_norm": 6.854390153462126,
+      "learning_rate": 2e-05,
+      "loss": 0.5453,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3596829477292202,
+      "grad_norm": 2.7159367999449033,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 15111
+    },
+    {
+      "epoch": 0.35970675045225176,
+      "grad_norm": 2.2478822231847135,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 15112
+    },
+    {
+      "epoch": 0.35973055317528324,
+      "grad_norm": 0.3339694443818847,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 15113
+    },
+    {
+      "epoch": 0.3597543558983148,
+      "grad_norm": 9.309967914916948,
+      "learning_rate": 2e-05,
+      "loss": 0.584,
+      "step": 15114
+    },
+    {
+      "epoch": 0.35977815862134627,
+      "grad_norm": 5.049270446041979,
+      "learning_rate": 2e-05,
+      "loss": 1.0286,
+      "step": 15115
+    },
+    {
+      "epoch": 0.3598019613443778,
+      "grad_norm": 3.851045775786058,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 15116
+    },
+    {
+      "epoch": 0.3598257640674093,
+      "grad_norm": 2.9267658855863137,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 15117
+    },
+    {
+      "epoch": 0.35984956679044083,
+      "grad_norm": 3.718793838357333,
+      "learning_rate": 2e-05,
+      "loss": 0.1138,
+      "step": 15118
+    },
+    {
+      "epoch": 0.3598733695134723,
+      "grad_norm": 2.7431286432610666,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 15119
+    },
+    {
+      "epoch": 0.35989717223650386,
+      "grad_norm": 2.116269201553432,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 15120
+    },
+    {
+      "epoch": 0.35992097495953534,
+      "grad_norm": 5.984343885648029,
+      "learning_rate": 2e-05,
+      "loss": 0.2867,
+      "step": 15121
+    },
+    {
+      "epoch": 0.3599447776825669,
+      "grad_norm": 13.642216764606092,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 15122
+    },
+    {
+      "epoch": 0.3599685804055984,
+      "grad_norm": 10.998863104165778,
+      "learning_rate": 2e-05,
+      "loss": 0.8416,
+      "step": 15123
+    },
+    {
+      "epoch": 0.3599923831286299,
+      "grad_norm": 1.0142676318995234,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 15124
+    },
+    {
+      "epoch": 0.36001618585166145,
+      "grad_norm": 6.054994739773553,
+      "learning_rate": 2e-05,
+      "loss": 1.2911,
+      "step": 15125
+    },
+    {
+      "epoch": 0.36003998857469294,
+      "grad_norm": 11.251444464983736,
+      "learning_rate": 2e-05,
+      "loss": 0.4053,
+      "step": 15126
+    },
+    {
+      "epoch": 0.3600637912977245,
+      "grad_norm": 6.096330609529573,
+      "learning_rate": 2e-05,
+      "loss": 0.6277,
+      "step": 15127
+    },
+    {
+      "epoch": 0.36008759402075596,
+      "grad_norm": 8.433748120806802,
+      "learning_rate": 2e-05,
+      "loss": 1.1095,
+      "step": 15128
+    },
+    {
+      "epoch": 0.3601113967437875,
+      "grad_norm": 4.921794232993144,
+      "learning_rate": 2e-05,
+      "loss": 0.6813,
+      "step": 15129
+    },
+    {
+      "epoch": 0.360135199466819,
+      "grad_norm": 4.234133973815123,
+      "learning_rate": 2e-05,
+      "loss": 0.5985,
+      "step": 15130
+    },
+    {
+      "epoch": 0.36015900218985053,
+      "grad_norm": 0.48114324592155805,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 15131
+    },
+    {
+      "epoch": 0.360182804912882,
+      "grad_norm": 4.198765444866865,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 15132
+    },
+    {
+      "epoch": 0.36020660763591356,
+      "grad_norm": 2.340688776652513,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 15133
+    },
+    {
+      "epoch": 0.36023041035894504,
+      "grad_norm": 0.6405754457447901,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 15134
+    },
+    {
+      "epoch": 0.3602542130819766,
+      "grad_norm": 4.185307543930275,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 15135
+    },
+    {
+      "epoch": 0.36027801580500807,
+      "grad_norm": 4.64412269014232,
+      "learning_rate": 2e-05,
+      "loss": 0.4153,
+      "step": 15136
+    },
+    {
+      "epoch": 0.3603018185280396,
+      "grad_norm": 5.979464697156654,
+      "learning_rate": 2e-05,
+      "loss": 0.8071,
+      "step": 15137
+    },
+    {
+      "epoch": 0.36032562125107115,
+      "grad_norm": 1.1094553905753934,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 15138
+    },
+    {
+      "epoch": 0.36034942397410263,
+      "grad_norm": 3.8637423609192134,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 15139
+    },
+    {
+      "epoch": 0.3603732266971342,
+      "grad_norm": 4.970748218533612,
+      "learning_rate": 2e-05,
+      "loss": 0.7322,
+      "step": 15140
+    },
+    {
+      "epoch": 0.36039702942016566,
+      "grad_norm": 4.244260400987566,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 15141
+    },
+    {
+      "epoch": 0.3604208321431972,
+      "grad_norm": 1.1571522359039785,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 15142
+    },
+    {
+      "epoch": 0.3604446348662287,
+      "grad_norm": 2.979979339794165,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 15143
+    },
+    {
+      "epoch": 0.3604684375892602,
+      "grad_norm": 13.762733127960397,
+      "learning_rate": 2e-05,
+      "loss": 1.3352,
+      "step": 15144
+    },
+    {
+      "epoch": 0.3604922403122917,
+      "grad_norm": 4.997515688051258,
+      "learning_rate": 2e-05,
+      "loss": 0.7678,
+      "step": 15145
+    },
+    {
+      "epoch": 0.36051604303532325,
+      "grad_norm": 3.655316503582621,
+      "learning_rate": 2e-05,
+      "loss": 0.4182,
+      "step": 15146
+    },
+    {
+      "epoch": 0.36053984575835474,
+      "grad_norm": 7.313894809458203,
+      "learning_rate": 2e-05,
+      "loss": 1.4748,
+      "step": 15147
+    },
+    {
+      "epoch": 0.3605636484813863,
+      "grad_norm": 1.9510108224469964,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 15148
+    },
+    {
+      "epoch": 0.36058745120441776,
+      "grad_norm": 4.169386456514248,
+      "learning_rate": 2e-05,
+      "loss": 0.2537,
+      "step": 15149
+    },
+    {
+      "epoch": 0.3606112539274493,
+      "grad_norm": 1.752065879351608,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 15150
+    },
+    {
+      "epoch": 0.3606350566504808,
+      "grad_norm": 1.0766215768285228,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 15151
+    },
+    {
+      "epoch": 0.36065885937351233,
+      "grad_norm": 1.8321484996962583,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 15152
+    },
+    {
+      "epoch": 0.36068266209654387,
+      "grad_norm": 5.636613974416122,
+      "learning_rate": 2e-05,
+      "loss": 0.5048,
+      "step": 15153
+    },
+    {
+      "epoch": 0.36070646481957536,
+      "grad_norm": 4.618329560507403,
+      "learning_rate": 2e-05,
+      "loss": 0.7585,
+      "step": 15154
+    },
+    {
+      "epoch": 0.3607302675426069,
+      "grad_norm": 4.4622915991649155,
+      "learning_rate": 2e-05,
+      "loss": 0.5628,
+      "step": 15155
+    },
+    {
+      "epoch": 0.3607540702656384,
+      "grad_norm": 5.434485248450813,
+      "learning_rate": 2e-05,
+      "loss": 0.8667,
+      "step": 15156
+    },
+    {
+      "epoch": 0.3607778729886699,
+      "grad_norm": 4.2640620698242,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 15157
+    },
+    {
+      "epoch": 0.3608016757117014,
+      "grad_norm": 5.607104020458859,
+      "learning_rate": 2e-05,
+      "loss": 1.0055,
+      "step": 15158
+    },
+    {
+      "epoch": 0.36082547843473295,
+      "grad_norm": 4.433927868911038,
+      "learning_rate": 2e-05,
+      "loss": 0.5294,
+      "step": 15159
+    },
+    {
+      "epoch": 0.36084928115776443,
+      "grad_norm": 8.587648258022545,
+      "learning_rate": 2e-05,
+      "loss": 1.8172,
+      "step": 15160
+    },
+    {
+      "epoch": 0.360873083880796,
+      "grad_norm": 8.823114318848708,
+      "learning_rate": 2e-05,
+      "loss": 0.4786,
+      "step": 15161
+    },
+    {
+      "epoch": 0.36089688660382746,
+      "grad_norm": 3.446644015069852,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 15162
+    },
+    {
+      "epoch": 0.360920689326859,
+      "grad_norm": 2.872083661770208,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 15163
+    },
+    {
+      "epoch": 0.3609444920498905,
+      "grad_norm": 1.3588018935312374,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 15164
+    },
+    {
+      "epoch": 0.360968294772922,
+      "grad_norm": 2.5284724509704657,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 15165
+    },
+    {
+      "epoch": 0.3609920974959535,
+      "grad_norm": 5.178105187055646,
+      "learning_rate": 2e-05,
+      "loss": 0.6768,
+      "step": 15166
+    },
+    {
+      "epoch": 0.36101590021898505,
+      "grad_norm": 5.119994334591895,
+      "learning_rate": 2e-05,
+      "loss": 0.7571,
+      "step": 15167
+    },
+    {
+      "epoch": 0.3610397029420166,
+      "grad_norm": 1.283403777069141,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 15168
+    },
+    {
+      "epoch": 0.3610635056650481,
+      "grad_norm": 1.0434439978874006,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 15169
+    },
+    {
+      "epoch": 0.3610873083880796,
+      "grad_norm": 7.65657513901264,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 15170
+    },
+    {
+      "epoch": 0.3611111111111111,
+      "grad_norm": 0.1633389700816909,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 15171
+    },
+    {
+      "epoch": 0.36113491383414265,
+      "grad_norm": 5.240686340162474,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 15172
+    },
+    {
+      "epoch": 0.36115871655717413,
+      "grad_norm": 5.136437844487283,
+      "learning_rate": 2e-05,
+      "loss": 0.5573,
+      "step": 15173
+    },
+    {
+      "epoch": 0.36118251928020567,
+      "grad_norm": 4.302661650632263,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 15174
+    },
+    {
+      "epoch": 0.36120632200323716,
+      "grad_norm": 3.306735721092574,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 15175
+    },
+    {
+      "epoch": 0.3612301247262687,
+      "grad_norm": 0.4543355850229277,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 15176
+    },
+    {
+      "epoch": 0.3612539274493002,
+      "grad_norm": 3.9840346698248434,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 15177
+    },
+    {
+      "epoch": 0.3612777301723317,
+      "grad_norm": 9.999321623378313,
+      "learning_rate": 2e-05,
+      "loss": 0.5631,
+      "step": 15178
+    },
+    {
+      "epoch": 0.3613015328953632,
+      "grad_norm": 2.4046593589019767,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 15179
+    },
+    {
+      "epoch": 0.36132533561839475,
+      "grad_norm": 0.6637777884799058,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 15180
+    },
+    {
+      "epoch": 0.36134913834142623,
+      "grad_norm": 3.900644974785346,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 15181
+    },
+    {
+      "epoch": 0.3613729410644578,
+      "grad_norm": 4.490647542383848,
+      "learning_rate": 2e-05,
+      "loss": 0.4674,
+      "step": 15182
+    },
+    {
+      "epoch": 0.3613967437874893,
+      "grad_norm": 0.22063006928173615,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 15183
+    },
+    {
+      "epoch": 0.3614205465105208,
+      "grad_norm": 0.9901238283938256,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 15184
+    },
+    {
+      "epoch": 0.36144434923355234,
+      "grad_norm": 0.1875460779088129,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15185
+    },
+    {
+      "epoch": 0.3614681519565838,
+      "grad_norm": 3.6375894798038635,
+      "learning_rate": 2e-05,
+      "loss": 0.3081,
+      "step": 15186
+    },
+    {
+      "epoch": 0.36149195467961537,
+      "grad_norm": 2.4496395462233025,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 15187
+    },
+    {
+      "epoch": 0.36151575740264685,
+      "grad_norm": 1.5043234702984005,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 15188
+    },
+    {
+      "epoch": 0.3615395601256784,
+      "grad_norm": 0.31430056655941563,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 15189
+    },
+    {
+      "epoch": 0.3615633628487099,
+      "grad_norm": 3.4410216866199455,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 15190
+    },
+    {
+      "epoch": 0.3615871655717414,
+      "grad_norm": 2.8720030747426954,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 15191
+    },
+    {
+      "epoch": 0.3616109682947729,
+      "grad_norm": 2.275872802367858,
+      "learning_rate": 2e-05,
+      "loss": 0.2201,
+      "step": 15192
+    },
+    {
+      "epoch": 0.36163477101780445,
+      "grad_norm": 0.5388927412514958,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15193
+    },
+    {
+      "epoch": 0.36165857374083593,
+      "grad_norm": 1.2166664434356982,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 15194
+    },
+    {
+      "epoch": 0.36168237646386747,
+      "grad_norm": 5.453920708253591,
+      "learning_rate": 2e-05,
+      "loss": 0.6847,
+      "step": 15195
+    },
+    {
+      "epoch": 0.36170617918689896,
+      "grad_norm": 6.461781636185242,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 15196
+    },
+    {
+      "epoch": 0.3617299819099305,
+      "grad_norm": 13.887752589616715,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 15197
+    },
+    {
+      "epoch": 0.361753784632962,
+      "grad_norm": 1.920261222249107,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 15198
+    },
+    {
+      "epoch": 0.3617775873559935,
+      "grad_norm": 11.327632692364986,
+      "learning_rate": 2e-05,
+      "loss": 0.294,
+      "step": 15199
+    },
+    {
+      "epoch": 0.36180139007902506,
+      "grad_norm": 11.118630461174071,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 15200
+    },
+    {
+      "epoch": 0.36182519280205655,
+      "grad_norm": 0.9654698116859053,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 15201
+    },
+    {
+      "epoch": 0.3618489955250881,
+      "grad_norm": 6.918689082762154,
+      "learning_rate": 2e-05,
+      "loss": 0.9552,
+      "step": 15202
+    },
+    {
+      "epoch": 0.3618727982481196,
+      "grad_norm": 2.18012914209275,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 15203
+    },
+    {
+      "epoch": 0.3618966009711511,
+      "grad_norm": 5.352326817819383,
+      "learning_rate": 2e-05,
+      "loss": 0.6813,
+      "step": 15204
+    },
+    {
+      "epoch": 0.3619204036941826,
+      "grad_norm": 9.292622350056014,
+      "learning_rate": 2e-05,
+      "loss": 1.0506,
+      "step": 15205
+    },
+    {
+      "epoch": 0.36194420641721414,
+      "grad_norm": 0.6970195948118897,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 15206
+    },
+    {
+      "epoch": 0.36196800914024563,
+      "grad_norm": 0.34343048183059177,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 15207
+    },
+    {
+      "epoch": 0.36199181186327717,
+      "grad_norm": 3.8687283185172765,
+      "learning_rate": 2e-05,
+      "loss": 0.1562,
+      "step": 15208
+    },
+    {
+      "epoch": 0.36201561458630865,
+      "grad_norm": 6.360385655626223,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 15209
+    },
+    {
+      "epoch": 0.3620394173093402,
+      "grad_norm": 0.9144339343207434,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 15210
+    },
+    {
+      "epoch": 0.3620632200323717,
+      "grad_norm": 9.972595394493359,
+      "learning_rate": 2e-05,
+      "loss": 0.4511,
+      "step": 15211
+    },
+    {
+      "epoch": 0.3620870227554032,
+      "grad_norm": 0.017782133351548542,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 15212
+    },
+    {
+      "epoch": 0.3621108254784347,
+      "grad_norm": 1.3446940614978362,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 15213
+    },
+    {
+      "epoch": 0.36213462820146625,
+      "grad_norm": 5.163570234248027,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 15214
+    },
+    {
+      "epoch": 0.3621584309244978,
+      "grad_norm": 0.25624108719610383,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 15215
+    },
+    {
+      "epoch": 0.3621822336475293,
+      "grad_norm": 2.508033752457057,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 15216
+    },
+    {
+      "epoch": 0.3622060363705608,
+      "grad_norm": 3.5382522662208,
+      "learning_rate": 2e-05,
+      "loss": 0.7268,
+      "step": 15217
+    },
+    {
+      "epoch": 0.3622298390935923,
+      "grad_norm": 3.474218018339831,
+      "learning_rate": 2e-05,
+      "loss": 0.6427,
+      "step": 15218
+    },
+    {
+      "epoch": 0.36225364181662384,
+      "grad_norm": 5.599124964704758,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 15219
+    },
+    {
+      "epoch": 0.3622774445396553,
+      "grad_norm": 5.049881408907423,
+      "learning_rate": 2e-05,
+      "loss": 0.6184,
+      "step": 15220
+    },
+    {
+      "epoch": 0.36230124726268687,
+      "grad_norm": 5.46629475235268,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 15221
+    },
+    {
+      "epoch": 0.36232504998571835,
+      "grad_norm": 6.29658755546314,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 15222
+    },
+    {
+      "epoch": 0.3623488527087499,
+      "grad_norm": 14.16760535734908,
+      "learning_rate": 2e-05,
+      "loss": 0.9445,
+      "step": 15223
+    },
+    {
+      "epoch": 0.3623726554317814,
+      "grad_norm": 57.56672184372217,
+      "learning_rate": 2e-05,
+      "loss": 0.317,
+      "step": 15224
+    },
+    {
+      "epoch": 0.3623964581548129,
+      "grad_norm": 0.4736988050800761,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 15225
+    },
+    {
+      "epoch": 0.3624202608778444,
+      "grad_norm": 10.254112726314528,
+      "learning_rate": 2e-05,
+      "loss": 0.4685,
+      "step": 15226
+    },
+    {
+      "epoch": 0.36244406360087594,
+      "grad_norm": 2.1114282304889893,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 15227
+    },
+    {
+      "epoch": 0.36246786632390743,
+      "grad_norm": 4.012933170066453,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 15228
+    },
+    {
+      "epoch": 0.36249166904693897,
+      "grad_norm": 2.468689370106264,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 15229
+    },
+    {
+      "epoch": 0.3625154717699705,
+      "grad_norm": 7.460513743836155,
+      "learning_rate": 2e-05,
+      "loss": 0.1793,
+      "step": 15230
+    },
+    {
+      "epoch": 0.362539274493002,
+      "grad_norm": 9.088075272468107,
+      "learning_rate": 2e-05,
+      "loss": 0.247,
+      "step": 15231
+    },
+    {
+      "epoch": 0.36256307721603354,
+      "grad_norm": 5.370601196277418,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 15232
+    },
+    {
+      "epoch": 0.362586879939065,
+      "grad_norm": 2.3728748369428776,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 15233
+    },
+    {
+      "epoch": 0.36261068266209656,
+      "grad_norm": 16.161743737884024,
+      "learning_rate": 2e-05,
+      "loss": 0.6615,
+      "step": 15234
+    },
+    {
+      "epoch": 0.36263448538512805,
+      "grad_norm": 14.455451636756628,
+      "learning_rate": 2e-05,
+      "loss": 1.0969,
+      "step": 15235
+    },
+    {
+      "epoch": 0.3626582881081596,
+      "grad_norm": 1.5242898239681044,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 15236
+    },
+    {
+      "epoch": 0.3626820908311911,
+      "grad_norm": 8.38004599667716,
+      "learning_rate": 2e-05,
+      "loss": 0.4587,
+      "step": 15237
+    },
+    {
+      "epoch": 0.3627058935542226,
+      "grad_norm": 5.163603242013422,
+      "learning_rate": 2e-05,
+      "loss": 1.2503,
+      "step": 15238
+    },
+    {
+      "epoch": 0.3627296962772541,
+      "grad_norm": 3.9336924573254146,
+      "learning_rate": 2e-05,
+      "loss": 0.2659,
+      "step": 15239
+    },
+    {
+      "epoch": 0.36275349900028564,
+      "grad_norm": 6.289787501388024,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 15240
+    },
+    {
+      "epoch": 0.3627773017233171,
+      "grad_norm": 2.5970981802622823,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 15241
+    },
+    {
+      "epoch": 0.36280110444634867,
+      "grad_norm": 5.392970434180685,
+      "learning_rate": 2e-05,
+      "loss": 0.7431,
+      "step": 15242
+    },
+    {
+      "epoch": 0.36282490716938015,
+      "grad_norm": 3.8755626479062624,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 15243
+    },
+    {
+      "epoch": 0.3628487098924117,
+      "grad_norm": 3.0402578988645743,
+      "learning_rate": 2e-05,
+      "loss": 0.3811,
+      "step": 15244
+    },
+    {
+      "epoch": 0.36287251261544323,
+      "grad_norm": 1.675638464569423,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 15245
+    },
+    {
+      "epoch": 0.3628963153384747,
+      "grad_norm": 5.63086762909773,
+      "learning_rate": 2e-05,
+      "loss": 0.8395,
+      "step": 15246
+    },
+    {
+      "epoch": 0.36292011806150626,
+      "grad_norm": 2.2591088112950066,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 15247
+    },
+    {
+      "epoch": 0.36294392078453774,
+      "grad_norm": 6.949479646950059,
+      "learning_rate": 2e-05,
+      "loss": 0.9862,
+      "step": 15248
+    },
+    {
+      "epoch": 0.3629677235075693,
+      "grad_norm": 3.240753481808472,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 15249
+    },
+    {
+      "epoch": 0.36299152623060077,
+      "grad_norm": 9.992298239570198,
+      "learning_rate": 2e-05,
+      "loss": 0.3965,
+      "step": 15250
+    },
+    {
+      "epoch": 0.3630153289536323,
+      "grad_norm": 0.2214915993676751,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 15251
+    },
+    {
+      "epoch": 0.3630391316766638,
+      "grad_norm": 2.508181938459677,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 15252
+    },
+    {
+      "epoch": 0.36306293439969534,
+      "grad_norm": 2.7245869974689434,
+      "learning_rate": 2e-05,
+      "loss": 0.173,
+      "step": 15253
+    },
+    {
+      "epoch": 0.3630867371227268,
+      "grad_norm": 4.845555624867609,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 15254
+    },
+    {
+      "epoch": 0.36311053984575836,
+      "grad_norm": 7.045091885546304,
+      "learning_rate": 2e-05,
+      "loss": 0.3897,
+      "step": 15255
+    },
+    {
+      "epoch": 0.36313434256878985,
+      "grad_norm": 2.3038536251077177,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 15256
+    },
+    {
+      "epoch": 0.3631581452918214,
+      "grad_norm": 3.1055297107231348,
+      "learning_rate": 2e-05,
+      "loss": 0.4837,
+      "step": 15257
+    },
+    {
+      "epoch": 0.3631819480148529,
+      "grad_norm": 4.567173000835932,
+      "learning_rate": 2e-05,
+      "loss": 0.6208,
+      "step": 15258
+    },
+    {
+      "epoch": 0.3632057507378844,
+      "grad_norm": 2.3025536058170477,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 15259
+    },
+    {
+      "epoch": 0.36322955346091595,
+      "grad_norm": 7.29853392009972,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 15260
+    },
+    {
+      "epoch": 0.36325335618394744,
+      "grad_norm": 2.353757331789864,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 15261
+    },
+    {
+      "epoch": 0.363277158906979,
+      "grad_norm": 3.840240692621422,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 15262
+    },
+    {
+      "epoch": 0.36330096163001047,
+      "grad_norm": 3.634439215523532,
+      "learning_rate": 2e-05,
+      "loss": 0.1589,
+      "step": 15263
+    },
+    {
+      "epoch": 0.363324764353042,
+      "grad_norm": 6.271473873256489,
+      "learning_rate": 2e-05,
+      "loss": 0.6709,
+      "step": 15264
+    },
+    {
+      "epoch": 0.3633485670760735,
+      "grad_norm": 0.5034559237115981,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 15265
+    },
+    {
+      "epoch": 0.36337236979910503,
+      "grad_norm": 7.739815782335617,
+      "learning_rate": 2e-05,
+      "loss": 0.9008,
+      "step": 15266
+    },
+    {
+      "epoch": 0.3633961725221365,
+      "grad_norm": 0.580091429724435,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 15267
+    },
+    {
+      "epoch": 0.36341997524516806,
+      "grad_norm": 3.2353167670154876,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 15268
+    },
+    {
+      "epoch": 0.36344377796819954,
+      "grad_norm": 4.106431525873862,
+      "learning_rate": 2e-05,
+      "loss": 0.4413,
+      "step": 15269
+    },
+    {
+      "epoch": 0.3634675806912311,
+      "grad_norm": 1.8907189618666949,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 15270
+    },
+    {
+      "epoch": 0.36349138341426257,
+      "grad_norm": 5.844725597339201,
+      "learning_rate": 2e-05,
+      "loss": 0.6188,
+      "step": 15271
+    },
+    {
+      "epoch": 0.3635151861372941,
+      "grad_norm": 3.735807828270647,
+      "learning_rate": 2e-05,
+      "loss": 0.4528,
+      "step": 15272
+    },
+    {
+      "epoch": 0.3635389888603256,
+      "grad_norm": 2.7377012573866892,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 15273
+    },
+    {
+      "epoch": 0.36356279158335714,
+      "grad_norm": 6.569527207648021,
+      "learning_rate": 2e-05,
+      "loss": 1.137,
+      "step": 15274
+    },
+    {
+      "epoch": 0.3635865943063887,
+      "grad_norm": 4.512119561812376,
+      "learning_rate": 2e-05,
+      "loss": 0.1839,
+      "step": 15275
+    },
+    {
+      "epoch": 0.36361039702942016,
+      "grad_norm": 7.063072279610547,
+      "learning_rate": 2e-05,
+      "loss": 0.7346,
+      "step": 15276
+    },
+    {
+      "epoch": 0.3636341997524517,
+      "grad_norm": 4.789733773545887,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 15277
+    },
+    {
+      "epoch": 0.3636580024754832,
+      "grad_norm": 2.4067110103630354,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 15278
+    },
+    {
+      "epoch": 0.36368180519851473,
+      "grad_norm": 5.429796168448131,
+      "learning_rate": 2e-05,
+      "loss": 1.3412,
+      "step": 15279
+    },
+    {
+      "epoch": 0.3637056079215462,
+      "grad_norm": 5.128492257514658,
+      "learning_rate": 2e-05,
+      "loss": 1.2554,
+      "step": 15280
+    },
+    {
+      "epoch": 0.36372941064457776,
+      "grad_norm": 4.618300459725975,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 15281
+    },
+    {
+      "epoch": 0.36375321336760924,
+      "grad_norm": 4.703822206811395,
+      "learning_rate": 2e-05,
+      "loss": 0.6178,
+      "step": 15282
+    },
+    {
+      "epoch": 0.3637770160906408,
+      "grad_norm": 0.5724609599878349,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 15283
+    },
+    {
+      "epoch": 0.36380081881367227,
+      "grad_norm": 1.3435768895084277,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 15284
+    },
+    {
+      "epoch": 0.3638246215367038,
+      "grad_norm": 3.555980141956471,
+      "learning_rate": 2e-05,
+      "loss": 0.3143,
+      "step": 15285
+    },
+    {
+      "epoch": 0.3638484242597353,
+      "grad_norm": 11.12131411558472,
+      "learning_rate": 2e-05,
+      "loss": 1.2492,
+      "step": 15286
+    },
+    {
+      "epoch": 0.36387222698276683,
+      "grad_norm": 6.7369437027434405,
+      "learning_rate": 2e-05,
+      "loss": 0.8881,
+      "step": 15287
+    },
+    {
+      "epoch": 0.3638960297057983,
+      "grad_norm": 3.421145182939867,
+      "learning_rate": 2e-05,
+      "loss": 0.5004,
+      "step": 15288
+    },
+    {
+      "epoch": 0.36391983242882986,
+      "grad_norm": 6.818820638560615,
+      "learning_rate": 2e-05,
+      "loss": 1.0605,
+      "step": 15289
+    },
+    {
+      "epoch": 0.3639436351518614,
+      "grad_norm": 5.355847826995975,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 15290
+    },
+    {
+      "epoch": 0.3639674378748929,
+      "grad_norm": 8.78030803977387,
+      "learning_rate": 2e-05,
+      "loss": 0.5342,
+      "step": 15291
+    },
+    {
+      "epoch": 0.3639912405979244,
+      "grad_norm": 2.994906457313559,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 15292
+    },
+    {
+      "epoch": 0.3640150433209559,
+      "grad_norm": 1.8353997671282032,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 15293
+    },
+    {
+      "epoch": 0.36403884604398745,
+      "grad_norm": 3.8563948953258884,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 15294
+    },
+    {
+      "epoch": 0.36406264876701894,
+      "grad_norm": 1.7138361134286444,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 15295
+    },
+    {
+      "epoch": 0.3640864514900505,
+      "grad_norm": 1.7889894810566085,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 15296
+    },
+    {
+      "epoch": 0.36411025421308196,
+      "grad_norm": 7.337536893600939,
+      "learning_rate": 2e-05,
+      "loss": 0.3249,
+      "step": 15297
+    },
+    {
+      "epoch": 0.3641340569361135,
+      "grad_norm": 2.203139457678313,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 15298
+    },
+    {
+      "epoch": 0.364157859659145,
+      "grad_norm": 6.076301734800288,
+      "learning_rate": 2e-05,
+      "loss": 0.8764,
+      "step": 15299
+    },
+    {
+      "epoch": 0.36418166238217653,
+      "grad_norm": 2.548514318362854,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 15300
+    },
+    {
+      "epoch": 0.364205465105208,
+      "grad_norm": 2.216123750333669,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 15301
+    },
+    {
+      "epoch": 0.36422926782823956,
+      "grad_norm": 2.6585230660872243,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 15302
+    },
+    {
+      "epoch": 0.36425307055127104,
+      "grad_norm": 4.568873039578075,
+      "learning_rate": 2e-05,
+      "loss": 0.6913,
+      "step": 15303
+    },
+    {
+      "epoch": 0.3642768732743026,
+      "grad_norm": 10.674323770202683,
+      "learning_rate": 2e-05,
+      "loss": 0.4317,
+      "step": 15304
+    },
+    {
+      "epoch": 0.3643006759973341,
+      "grad_norm": 16.258499890328306,
+      "learning_rate": 2e-05,
+      "loss": 0.2462,
+      "step": 15305
+    },
+    {
+      "epoch": 0.3643244787203656,
+      "grad_norm": 1.8664047652226303,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 15306
+    },
+    {
+      "epoch": 0.36434828144339715,
+      "grad_norm": 5.85328285727534,
+      "learning_rate": 2e-05,
+      "loss": 1.24,
+      "step": 15307
+    },
+    {
+      "epoch": 0.36437208416642863,
+      "grad_norm": 6.8701046676336635,
+      "learning_rate": 2e-05,
+      "loss": 1.3286,
+      "step": 15308
+    },
+    {
+      "epoch": 0.3643958868894602,
+      "grad_norm": 6.939953480113021,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 15309
+    },
+    {
+      "epoch": 0.36441968961249166,
+      "grad_norm": 5.721829280866155,
+      "learning_rate": 2e-05,
+      "loss": 1.1663,
+      "step": 15310
+    },
+    {
+      "epoch": 0.3644434923355232,
+      "grad_norm": 5.980737482099278,
+      "learning_rate": 2e-05,
+      "loss": 1.0495,
+      "step": 15311
+    },
+    {
+      "epoch": 0.3644672950585547,
+      "grad_norm": 3.8821647269395894,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 15312
+    },
+    {
+      "epoch": 0.3644910977815862,
+      "grad_norm": 7.139691898768541,
+      "learning_rate": 2e-05,
+      "loss": 0.8337,
+      "step": 15313
+    },
+    {
+      "epoch": 0.3645149005046177,
+      "grad_norm": 2.016628310135292,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 15314
+    },
+    {
+      "epoch": 0.36453870322764925,
+      "grad_norm": 4.570588608710907,
+      "learning_rate": 2e-05,
+      "loss": 0.7236,
+      "step": 15315
+    },
+    {
+      "epoch": 0.36456250595068074,
+      "grad_norm": 5.774007127901427,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 15316
+    },
+    {
+      "epoch": 0.3645863086737123,
+      "grad_norm": 3.6595558199493867,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 15317
+    },
+    {
+      "epoch": 0.36461011139674376,
+      "grad_norm": 4.310090184799141,
+      "learning_rate": 2e-05,
+      "loss": 0.5359,
+      "step": 15318
+    },
+    {
+      "epoch": 0.3646339141197753,
+      "grad_norm": 14.09632874718415,
+      "learning_rate": 2e-05,
+      "loss": 0.387,
+      "step": 15319
+    },
+    {
+      "epoch": 0.3646577168428068,
+      "grad_norm": 4.337134873394475,
+      "learning_rate": 2e-05,
+      "loss": 0.5335,
+      "step": 15320
+    },
+    {
+      "epoch": 0.36468151956583833,
+      "grad_norm": 1.6879448714786978,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 15321
+    },
+    {
+      "epoch": 0.36470532228886987,
+      "grad_norm": 3.3747932417604702,
+      "learning_rate": 2e-05,
+      "loss": 0.4521,
+      "step": 15322
+    },
+    {
+      "epoch": 0.36472912501190136,
+      "grad_norm": 8.543040013602978,
+      "learning_rate": 2e-05,
+      "loss": 0.179,
+      "step": 15323
+    },
+    {
+      "epoch": 0.3647529277349329,
+      "grad_norm": 4.080083075219045,
+      "learning_rate": 2e-05,
+      "loss": 0.459,
+      "step": 15324
+    },
+    {
+      "epoch": 0.3647767304579644,
+      "grad_norm": 2.8851389766970437,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 15325
+    },
+    {
+      "epoch": 0.3648005331809959,
+      "grad_norm": 0.5294241004774205,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 15326
+    },
+    {
+      "epoch": 0.3648243359040274,
+      "grad_norm": 1.768476142181162,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 15327
+    },
+    {
+      "epoch": 0.36484813862705895,
+      "grad_norm": 4.2094322171053085,
+      "learning_rate": 2e-05,
+      "loss": 0.626,
+      "step": 15328
+    },
+    {
+      "epoch": 0.36487194135009043,
+      "grad_norm": 2.3353729381813597,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 15329
+    },
+    {
+      "epoch": 0.364895744073122,
+      "grad_norm": 3.804331054274412,
+      "learning_rate": 2e-05,
+      "loss": 0.5289,
+      "step": 15330
+    },
+    {
+      "epoch": 0.36491954679615346,
+      "grad_norm": 8.669667762746496,
+      "learning_rate": 2e-05,
+      "loss": 0.4758,
+      "step": 15331
+    },
+    {
+      "epoch": 0.364943349519185,
+      "grad_norm": 2.7415370129799395,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 15332
+    },
+    {
+      "epoch": 0.3649671522422165,
+      "grad_norm": 1.9590921855045835,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 15333
+    },
+    {
+      "epoch": 0.364990954965248,
+      "grad_norm": 4.737008600392785,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 15334
+    },
+    {
+      "epoch": 0.3650147576882795,
+      "grad_norm": 14.758177480236386,
+      "learning_rate": 2e-05,
+      "loss": 1.0467,
+      "step": 15335
+    },
+    {
+      "epoch": 0.36503856041131105,
+      "grad_norm": 6.584649956863812,
+      "learning_rate": 2e-05,
+      "loss": 1.5543,
+      "step": 15336
+    },
+    {
+      "epoch": 0.3650623631343426,
+      "grad_norm": 7.38653781657581,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 15337
+    },
+    {
+      "epoch": 0.3650861658573741,
+      "grad_norm": 6.826776301567202,
+      "learning_rate": 2e-05,
+      "loss": 0.7536,
+      "step": 15338
+    },
+    {
+      "epoch": 0.3651099685804056,
+      "grad_norm": 0.30572582069148413,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 15339
+    },
+    {
+      "epoch": 0.3651337713034371,
+      "grad_norm": 12.449412699069267,
+      "learning_rate": 2e-05,
+      "loss": 0.265,
+      "step": 15340
+    },
+    {
+      "epoch": 0.36515757402646865,
+      "grad_norm": 9.817814694217578,
+      "learning_rate": 2e-05,
+      "loss": 0.718,
+      "step": 15341
+    },
+    {
+      "epoch": 0.36518137674950013,
+      "grad_norm": 4.222550078121318,
+      "learning_rate": 2e-05,
+      "loss": 0.7747,
+      "step": 15342
+    },
+    {
+      "epoch": 0.36520517947253167,
+      "grad_norm": 2.9552971018836263,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 15343
+    },
+    {
+      "epoch": 0.36522898219556316,
+      "grad_norm": 2.702197767918464,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 15344
+    },
+    {
+      "epoch": 0.3652527849185947,
+      "grad_norm": 4.299458224488853,
+      "learning_rate": 2e-05,
+      "loss": 0.771,
+      "step": 15345
+    },
+    {
+      "epoch": 0.3652765876416262,
+      "grad_norm": 7.937306241256101,
+      "learning_rate": 2e-05,
+      "loss": 0.3777,
+      "step": 15346
+    },
+    {
+      "epoch": 0.3653003903646577,
+      "grad_norm": 0.5295709914902362,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 15347
+    },
+    {
+      "epoch": 0.3653241930876892,
+      "grad_norm": 1.5168382493295098,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 15348
+    },
+    {
+      "epoch": 0.36534799581072075,
+      "grad_norm": 4.739103238683194,
+      "learning_rate": 2e-05,
+      "loss": 0.9139,
+      "step": 15349
+    },
+    {
+      "epoch": 0.36537179853375223,
+      "grad_norm": 6.4859776721302325,
+      "learning_rate": 2e-05,
+      "loss": 0.2176,
+      "step": 15350
+    },
+    {
+      "epoch": 0.3653956012567838,
+      "grad_norm": 2.177073614676405,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 15351
+    },
+    {
+      "epoch": 0.3654194039798153,
+      "grad_norm": 5.2855531865547265,
+      "learning_rate": 2e-05,
+      "loss": 0.6426,
+      "step": 15352
+    },
+    {
+      "epoch": 0.3654432067028468,
+      "grad_norm": 7.831799031496139,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 15353
+    },
+    {
+      "epoch": 0.36546700942587834,
+      "grad_norm": 0.9359480075154929,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 15354
+    },
+    {
+      "epoch": 0.3654908121489098,
+      "grad_norm": 3.011305606493619,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 15355
+    },
+    {
+      "epoch": 0.36551461487194137,
+      "grad_norm": 13.554048899345934,
+      "learning_rate": 2e-05,
+      "loss": 0.9929,
+      "step": 15356
+    },
+    {
+      "epoch": 0.36553841759497285,
+      "grad_norm": 0.4981209074352658,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 15357
+    },
+    {
+      "epoch": 0.3655622203180044,
+      "grad_norm": 4.319664850491004,
+      "learning_rate": 2e-05,
+      "loss": 0.6981,
+      "step": 15358
+    },
+    {
+      "epoch": 0.3655860230410359,
+      "grad_norm": 4.172688306356068,
+      "learning_rate": 2e-05,
+      "loss": 0.2145,
+      "step": 15359
+    },
+    {
+      "epoch": 0.3656098257640674,
+      "grad_norm": 2.0342848763532215,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 15360
+    },
+    {
+      "epoch": 0.3656336284870989,
+      "grad_norm": 3.9096739135600447,
+      "learning_rate": 2e-05,
+      "loss": 0.5068,
+      "step": 15361
+    },
+    {
+      "epoch": 0.36565743121013045,
+      "grad_norm": 6.981529790723327,
+      "learning_rate": 2e-05,
+      "loss": 0.2143,
+      "step": 15362
+    },
+    {
+      "epoch": 0.36568123393316193,
+      "grad_norm": 2.021556885132629,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 15363
+    },
+    {
+      "epoch": 0.36570503665619347,
+      "grad_norm": 11.630658530900615,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 15364
+    },
+    {
+      "epoch": 0.36572883937922496,
+      "grad_norm": 2.8260709093241796,
+      "learning_rate": 2e-05,
+      "loss": 0.1457,
+      "step": 15365
+    },
+    {
+      "epoch": 0.3657526421022565,
+      "grad_norm": 9.456134969346481,
+      "learning_rate": 2e-05,
+      "loss": 1.773,
+      "step": 15366
+    },
+    {
+      "epoch": 0.36577644482528804,
+      "grad_norm": 2.674269870398915,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 15367
+    },
+    {
+      "epoch": 0.3658002475483195,
+      "grad_norm": 1.8220322273971574,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 15368
+    },
+    {
+      "epoch": 0.36582405027135106,
+      "grad_norm": 8.871187540040834,
+      "learning_rate": 2e-05,
+      "loss": 0.7622,
+      "step": 15369
+    },
+    {
+      "epoch": 0.36584785299438255,
+      "grad_norm": 2.4538639988336204,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 15370
+    },
+    {
+      "epoch": 0.3658716557174141,
+      "grad_norm": 8.124930299050668,
+      "learning_rate": 2e-05,
+      "loss": 1.1677,
+      "step": 15371
+    },
+    {
+      "epoch": 0.3658954584404456,
+      "grad_norm": 3.428388455699919,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 15372
+    },
+    {
+      "epoch": 0.3659192611634771,
+      "grad_norm": 1.4930147362978576,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 15373
+    },
+    {
+      "epoch": 0.3659430638865086,
+      "grad_norm": 4.481713083739862,
+      "learning_rate": 2e-05,
+      "loss": 0.8846,
+      "step": 15374
+    },
+    {
+      "epoch": 0.36596686660954014,
+      "grad_norm": 3.3118088739829257,
+      "learning_rate": 2e-05,
+      "loss": 0.3592,
+      "step": 15375
+    },
+    {
+      "epoch": 0.3659906693325716,
+      "grad_norm": 2.7224726711019476,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 15376
+    },
+    {
+      "epoch": 0.36601447205560317,
+      "grad_norm": 4.053402883531281,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 15377
+    },
+    {
+      "epoch": 0.36603827477863465,
+      "grad_norm": 0.2067831630819372,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 15378
+    },
+    {
+      "epoch": 0.3660620775016662,
+      "grad_norm": 6.11726329585758,
+      "learning_rate": 2e-05,
+      "loss": 0.6746,
+      "step": 15379
+    },
+    {
+      "epoch": 0.3660858802246977,
+      "grad_norm": 3.69856850775818,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 15380
+    },
+    {
+      "epoch": 0.3661096829477292,
+      "grad_norm": 5.884039052206391,
+      "learning_rate": 2e-05,
+      "loss": 1.0321,
+      "step": 15381
+    },
+    {
+      "epoch": 0.36613348567076076,
+      "grad_norm": 0.5751220314092459,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 15382
+    },
+    {
+      "epoch": 0.36615728839379225,
+      "grad_norm": 1.915675402328711,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 15383
+    },
+    {
+      "epoch": 0.3661810911168238,
+      "grad_norm": 0.33218968409462946,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 15384
+    },
+    {
+      "epoch": 0.36620489383985527,
+      "grad_norm": 5.6617246245005735,
+      "learning_rate": 2e-05,
+      "loss": 0.8313,
+      "step": 15385
+    },
+    {
+      "epoch": 0.3662286965628868,
+      "grad_norm": 8.480426606586548,
+      "learning_rate": 2e-05,
+      "loss": 0.2714,
+      "step": 15386
+    },
+    {
+      "epoch": 0.3662524992859183,
+      "grad_norm": 6.185074413396428,
+      "learning_rate": 2e-05,
+      "loss": 0.9892,
+      "step": 15387
+    },
+    {
+      "epoch": 0.36627630200894984,
+      "grad_norm": 8.006110376032979,
+      "learning_rate": 2e-05,
+      "loss": 0.9456,
+      "step": 15388
+    },
+    {
+      "epoch": 0.3663001047319813,
+      "grad_norm": 0.5165533481559175,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 15389
+    },
+    {
+      "epoch": 0.36632390745501286,
+      "grad_norm": 3.622429241456353,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 15390
+    },
+    {
+      "epoch": 0.36634771017804435,
+      "grad_norm": 9.585731977165363,
+      "learning_rate": 2e-05,
+      "loss": 0.9026,
+      "step": 15391
+    },
+    {
+      "epoch": 0.3663715129010759,
+      "grad_norm": 6.1612860393970275,
+      "learning_rate": 2e-05,
+      "loss": 0.6519,
+      "step": 15392
+    },
+    {
+      "epoch": 0.3663953156241074,
+      "grad_norm": 2.9576203595529678,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 15393
+    },
+    {
+      "epoch": 0.3664191183471389,
+      "grad_norm": 0.3686511930378895,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15394
+    },
+    {
+      "epoch": 0.3664429210701704,
+      "grad_norm": 1.1523472021924035,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 15395
+    },
+    {
+      "epoch": 0.36646672379320194,
+      "grad_norm": 0.9600865978271348,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 15396
+    },
+    {
+      "epoch": 0.3664905265162335,
+      "grad_norm": 20.776430655685143,
+      "learning_rate": 2e-05,
+      "loss": 0.2723,
+      "step": 15397
+    },
+    {
+      "epoch": 0.36651432923926497,
+      "grad_norm": 2.5125399354166507,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 15398
+    },
+    {
+      "epoch": 0.3665381319622965,
+      "grad_norm": 5.936896845971985,
+      "learning_rate": 2e-05,
+      "loss": 1.3718,
+      "step": 15399
+    },
+    {
+      "epoch": 0.366561934685328,
+      "grad_norm": 6.3056619421654805,
+      "learning_rate": 2e-05,
+      "loss": 1.3265,
+      "step": 15400
+    },
+    {
+      "epoch": 0.36658573740835954,
+      "grad_norm": 5.462314079698747,
+      "learning_rate": 2e-05,
+      "loss": 0.5499,
+      "step": 15401
+    },
+    {
+      "epoch": 0.366609540131391,
+      "grad_norm": 5.813394199295915,
+      "learning_rate": 2e-05,
+      "loss": 0.3437,
+      "step": 15402
+    },
+    {
+      "epoch": 0.36663334285442256,
+      "grad_norm": 0.02447662013393278,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 15403
+    },
+    {
+      "epoch": 0.36665714557745405,
+      "grad_norm": 6.221638697649962,
+      "learning_rate": 2e-05,
+      "loss": 0.8337,
+      "step": 15404
+    },
+    {
+      "epoch": 0.3666809483004856,
+      "grad_norm": 2.4854028266143144,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 15405
+    },
+    {
+      "epoch": 0.3667047510235171,
+      "grad_norm": 6.386161202214746,
+      "learning_rate": 2e-05,
+      "loss": 0.6745,
+      "step": 15406
+    },
+    {
+      "epoch": 0.3667285537465486,
+      "grad_norm": 1.1562092084687032,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 15407
+    },
+    {
+      "epoch": 0.3667523564695801,
+      "grad_norm": 2.0786377639106823,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 15408
+    },
+    {
+      "epoch": 0.36677615919261164,
+      "grad_norm": 3.3944301607782275,
+      "learning_rate": 2e-05,
+      "loss": 0.2277,
+      "step": 15409
+    },
+    {
+      "epoch": 0.3667999619156431,
+      "grad_norm": 4.963597584049868,
+      "learning_rate": 2e-05,
+      "loss": 0.4219,
+      "step": 15410
+    },
+    {
+      "epoch": 0.36682376463867467,
+      "grad_norm": 1.80719021454751,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 15411
+    },
+    {
+      "epoch": 0.3668475673617062,
+      "grad_norm": 2.077364826132835,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 15412
+    },
+    {
+      "epoch": 0.3668713700847377,
+      "grad_norm": 0.10348865108894142,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 15413
+    },
+    {
+      "epoch": 0.36689517280776923,
+      "grad_norm": 4.0105212751740815,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 15414
+    },
+    {
+      "epoch": 0.3669189755308007,
+      "grad_norm": 3.3673962520904657,
+      "learning_rate": 2e-05,
+      "loss": 0.5254,
+      "step": 15415
+    },
+    {
+      "epoch": 0.36694277825383226,
+      "grad_norm": 12.670216067308965,
+      "learning_rate": 2e-05,
+      "loss": 0.6778,
+      "step": 15416
+    },
+    {
+      "epoch": 0.36696658097686374,
+      "grad_norm": 2.8295273139552912,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 15417
+    },
+    {
+      "epoch": 0.3669903836998953,
+      "grad_norm": 1.974647280920019,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 15418
+    },
+    {
+      "epoch": 0.36701418642292677,
+      "grad_norm": 4.23431973323738,
+      "learning_rate": 2e-05,
+      "loss": 0.2664,
+      "step": 15419
+    },
+    {
+      "epoch": 0.3670379891459583,
+      "grad_norm": 3.860707388018347,
+      "learning_rate": 2e-05,
+      "loss": 0.5493,
+      "step": 15420
+    },
+    {
+      "epoch": 0.3670617918689898,
+      "grad_norm": 17.220988293049885,
+      "learning_rate": 2e-05,
+      "loss": 0.232,
+      "step": 15421
+    },
+    {
+      "epoch": 0.36708559459202134,
+      "grad_norm": 3.640759532840541,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 15422
+    },
+    {
+      "epoch": 0.3671093973150528,
+      "grad_norm": 7.903377551516636,
+      "learning_rate": 2e-05,
+      "loss": 0.4204,
+      "step": 15423
+    },
+    {
+      "epoch": 0.36713320003808436,
+      "grad_norm": 29.739170805071964,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 15424
+    },
+    {
+      "epoch": 0.36715700276111585,
+      "grad_norm": 8.544720075881013,
+      "learning_rate": 2e-05,
+      "loss": 0.368,
+      "step": 15425
+    },
+    {
+      "epoch": 0.3671808054841474,
+      "grad_norm": 15.019590407586627,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 15426
+    },
+    {
+      "epoch": 0.36720460820717893,
+      "grad_norm": 3.055713008627733,
+      "learning_rate": 2e-05,
+      "loss": 0.4979,
+      "step": 15427
+    },
+    {
+      "epoch": 0.3672284109302104,
+      "grad_norm": 2.697081425925699,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 15428
+    },
+    {
+      "epoch": 0.36725221365324195,
+      "grad_norm": 0.816197172024321,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 15429
+    },
+    {
+      "epoch": 0.36727601637627344,
+      "grad_norm": 2.9079150334070736,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 15430
+    },
+    {
+      "epoch": 0.367299819099305,
+      "grad_norm": 5.325190707474791,
+      "learning_rate": 2e-05,
+      "loss": 0.6651,
+      "step": 15431
+    },
+    {
+      "epoch": 0.36732362182233647,
+      "grad_norm": 10.267980683883607,
+      "learning_rate": 2e-05,
+      "loss": 0.5371,
+      "step": 15432
+    },
+    {
+      "epoch": 0.367347424545368,
+      "grad_norm": 4.115552649896894,
+      "learning_rate": 2e-05,
+      "loss": 0.4849,
+      "step": 15433
+    },
+    {
+      "epoch": 0.3673712272683995,
+      "grad_norm": 3.5453193237819827,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 15434
+    },
+    {
+      "epoch": 0.36739502999143103,
+      "grad_norm": 4.310113466219928,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 15435
+    },
+    {
+      "epoch": 0.3674188327144625,
+      "grad_norm": 6.0409608757657836,
+      "learning_rate": 2e-05,
+      "loss": 0.671,
+      "step": 15436
+    },
+    {
+      "epoch": 0.36744263543749406,
+      "grad_norm": 4.076509988552471,
+      "learning_rate": 2e-05,
+      "loss": 0.1957,
+      "step": 15437
+    },
+    {
+      "epoch": 0.36746643816052554,
+      "grad_norm": 3.362447913468393,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 15438
+    },
+    {
+      "epoch": 0.3674902408835571,
+      "grad_norm": 4.607531200853299,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 15439
+    },
+    {
+      "epoch": 0.36751404360658857,
+      "grad_norm": 8.792982438228119,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 15440
+    },
+    {
+      "epoch": 0.3675378463296201,
+      "grad_norm": 4.246944871930601,
+      "learning_rate": 2e-05,
+      "loss": 0.3669,
+      "step": 15441
+    },
+    {
+      "epoch": 0.3675616490526516,
+      "grad_norm": 1.9401806777903159,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 15442
+    },
+    {
+      "epoch": 0.36758545177568314,
+      "grad_norm": 4.319598700415398,
+      "learning_rate": 2e-05,
+      "loss": 0.8358,
+      "step": 15443
+    },
+    {
+      "epoch": 0.3676092544987147,
+      "grad_norm": 1.5983518556651284,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 15444
+    },
+    {
+      "epoch": 0.36763305722174616,
+      "grad_norm": 4.87053724216091,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 15445
+    },
+    {
+      "epoch": 0.3676568599447777,
+      "grad_norm": 1.1758302960979292,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 15446
+    },
+    {
+      "epoch": 0.3676806626678092,
+      "grad_norm": 0.6722083849983631,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 15447
+    },
+    {
+      "epoch": 0.36770446539084073,
+      "grad_norm": 13.37516217565821,
+      "learning_rate": 2e-05,
+      "loss": 1.4298,
+      "step": 15448
+    },
+    {
+      "epoch": 0.3677282681138722,
+      "grad_norm": 3.317455558732836,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 15449
+    },
+    {
+      "epoch": 0.36775207083690375,
+      "grad_norm": 0.36255143602669926,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 15450
+    },
+    {
+      "epoch": 0.36777587355993524,
+      "grad_norm": 1.4293243484111728,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 15451
+    },
+    {
+      "epoch": 0.3677996762829668,
+      "grad_norm": 3.970159362579197,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 15452
+    },
+    {
+      "epoch": 0.36782347900599827,
+      "grad_norm": 4.265580447843182,
+      "learning_rate": 2e-05,
+      "loss": 0.1138,
+      "step": 15453
+    },
+    {
+      "epoch": 0.3678472817290298,
+      "grad_norm": 26.88224807173079,
+      "learning_rate": 2e-05,
+      "loss": 0.6291,
+      "step": 15454
+    },
+    {
+      "epoch": 0.3678710844520613,
+      "grad_norm": 6.103389161960644,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 15455
+    },
+    {
+      "epoch": 0.36789488717509283,
+      "grad_norm": 0.4793666235286375,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 15456
+    },
+    {
+      "epoch": 0.3679186898981243,
+      "grad_norm": 5.761079279125234,
+      "learning_rate": 2e-05,
+      "loss": 1.0211,
+      "step": 15457
+    },
+    {
+      "epoch": 0.36794249262115586,
+      "grad_norm": 4.806766175995873,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 15458
+    },
+    {
+      "epoch": 0.3679662953441874,
+      "grad_norm": 4.6056670676669675,
+      "learning_rate": 2e-05,
+      "loss": 0.4492,
+      "step": 15459
+    },
+    {
+      "epoch": 0.3679900980672189,
+      "grad_norm": 4.495055333633834,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 15460
+    },
+    {
+      "epoch": 0.3680139007902504,
+      "grad_norm": 2.4672968464457092,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 15461
+    },
+    {
+      "epoch": 0.3680377035132819,
+      "grad_norm": 4.401902590422859,
+      "learning_rate": 2e-05,
+      "loss": 0.7345,
+      "step": 15462
+    },
+    {
+      "epoch": 0.36806150623631345,
+      "grad_norm": 1.4110190706836239,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 15463
+    },
+    {
+      "epoch": 0.36808530895934494,
+      "grad_norm": 6.311253105043031,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 15464
+    },
+    {
+      "epoch": 0.3681091116823765,
+      "grad_norm": 7.323597636502791,
+      "learning_rate": 2e-05,
+      "loss": 1.1905,
+      "step": 15465
+    },
+    {
+      "epoch": 0.36813291440540796,
+      "grad_norm": 3.972459832984316,
+      "learning_rate": 2e-05,
+      "loss": 0.6129,
+      "step": 15466
+    },
+    {
+      "epoch": 0.3681567171284395,
+      "grad_norm": 9.292858176846716,
+      "learning_rate": 2e-05,
+      "loss": 0.3093,
+      "step": 15467
+    },
+    {
+      "epoch": 0.368180519851471,
+      "grad_norm": 7.0994160838333675,
+      "learning_rate": 2e-05,
+      "loss": 0.8161,
+      "step": 15468
+    },
+    {
+      "epoch": 0.36820432257450253,
+      "grad_norm": 0.3541289317758625,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15469
+    },
+    {
+      "epoch": 0.368228125297534,
+      "grad_norm": 2.904514945429378,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 15470
+    },
+    {
+      "epoch": 0.36825192802056556,
+      "grad_norm": 3.1006159612113433,
+      "learning_rate": 2e-05,
+      "loss": 0.3569,
+      "step": 15471
+    },
+    {
+      "epoch": 0.36827573074359704,
+      "grad_norm": 4.007691178490322,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 15472
+    },
+    {
+      "epoch": 0.3682995334666286,
+      "grad_norm": 5.519182263530529,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 15473
+    },
+    {
+      "epoch": 0.3683233361896601,
+      "grad_norm": 12.230357374267578,
+      "learning_rate": 2e-05,
+      "loss": 0.5508,
+      "step": 15474
+    },
+    {
+      "epoch": 0.3683471389126916,
+      "grad_norm": 0.40280568090287266,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 15475
+    },
+    {
+      "epoch": 0.36837094163572315,
+      "grad_norm": 7.340586022250443,
+      "learning_rate": 2e-05,
+      "loss": 0.7435,
+      "step": 15476
+    },
+    {
+      "epoch": 0.36839474435875463,
+      "grad_norm": 1.7946570881839607,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 15477
+    },
+    {
+      "epoch": 0.3684185470817862,
+      "grad_norm": 15.472574739375627,
+      "learning_rate": 2e-05,
+      "loss": 0.5481,
+      "step": 15478
+    },
+    {
+      "epoch": 0.36844234980481766,
+      "grad_norm": 8.223623796633717,
+      "learning_rate": 2e-05,
+      "loss": 0.7951,
+      "step": 15479
+    },
+    {
+      "epoch": 0.3684661525278492,
+      "grad_norm": 6.951683458049894,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 15480
+    },
+    {
+      "epoch": 0.3684899552508807,
+      "grad_norm": 6.190429829329702,
+      "learning_rate": 2e-05,
+      "loss": 0.9869,
+      "step": 15481
+    },
+    {
+      "epoch": 0.3685137579739122,
+      "grad_norm": 5.612329662313718,
+      "learning_rate": 2e-05,
+      "loss": 0.8065,
+      "step": 15482
+    },
+    {
+      "epoch": 0.3685375606969437,
+      "grad_norm": 2.0257031273844186,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 15483
+    },
+    {
+      "epoch": 0.36856136341997525,
+      "grad_norm": 14.945911762783732,
+      "learning_rate": 2e-05,
+      "loss": 0.5977,
+      "step": 15484
+    },
+    {
+      "epoch": 0.36858516614300674,
+      "grad_norm": 1.866127959283946,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 15485
+    },
+    {
+      "epoch": 0.3686089688660383,
+      "grad_norm": 3.9017017541907877,
+      "learning_rate": 2e-05,
+      "loss": 0.2805,
+      "step": 15486
+    },
+    {
+      "epoch": 0.36863277158906976,
+      "grad_norm": 5.344208000058225,
+      "learning_rate": 2e-05,
+      "loss": 0.9812,
+      "step": 15487
+    },
+    {
+      "epoch": 0.3686565743121013,
+      "grad_norm": 2.9539334931810326,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 15488
+    },
+    {
+      "epoch": 0.36868037703513284,
+      "grad_norm": 1.8525626647671944,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 15489
+    },
+    {
+      "epoch": 0.36870417975816433,
+      "grad_norm": 3.90410560319324,
+      "learning_rate": 2e-05,
+      "loss": 0.2941,
+      "step": 15490
+    },
+    {
+      "epoch": 0.36872798248119587,
+      "grad_norm": 5.068208678408362,
+      "learning_rate": 2e-05,
+      "loss": 1.1277,
+      "step": 15491
+    },
+    {
+      "epoch": 0.36875178520422736,
+      "grad_norm": 3.0664458154987577,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 15492
+    },
+    {
+      "epoch": 0.3687755879272589,
+      "grad_norm": 5.8845535081782785,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 15493
+    },
+    {
+      "epoch": 0.3687993906502904,
+      "grad_norm": 9.093257095033584,
+      "learning_rate": 2e-05,
+      "loss": 0.4449,
+      "step": 15494
+    },
+    {
+      "epoch": 0.3688231933733219,
+      "grad_norm": 14.903165852910638,
+      "learning_rate": 2e-05,
+      "loss": 0.9818,
+      "step": 15495
+    },
+    {
+      "epoch": 0.3688469960963534,
+      "grad_norm": 5.474629718335354,
+      "learning_rate": 2e-05,
+      "loss": 1.142,
+      "step": 15496
+    },
+    {
+      "epoch": 0.36887079881938495,
+      "grad_norm": 2.717124185795252,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 15497
+    },
+    {
+      "epoch": 0.36889460154241643,
+      "grad_norm": 3.116225416658414,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 15498
+    },
+    {
+      "epoch": 0.368918404265448,
+      "grad_norm": 1.7561786420183547,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 15499
+    },
+    {
+      "epoch": 0.36894220698847946,
+      "grad_norm": 5.065647489050407,
+      "learning_rate": 2e-05,
+      "loss": 0.4573,
+      "step": 15500
+    },
+    {
+      "epoch": 0.368966009711511,
+      "grad_norm": 0.27867496305285905,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 15501
+    },
+    {
+      "epoch": 0.3689898124345425,
+      "grad_norm": 9.948654027324986,
+      "learning_rate": 2e-05,
+      "loss": 0.6882,
+      "step": 15502
+    },
+    {
+      "epoch": 0.369013615157574,
+      "grad_norm": 2.6736240864734597,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 15503
+    },
+    {
+      "epoch": 0.36903741788060557,
+      "grad_norm": 4.396095825262648,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 15504
+    },
+    {
+      "epoch": 0.36906122060363705,
+      "grad_norm": 8.369011052942401,
+      "learning_rate": 2e-05,
+      "loss": 0.1374,
+      "step": 15505
+    },
+    {
+      "epoch": 0.3690850233266686,
+      "grad_norm": 3.5179000904702433,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 15506
+    },
+    {
+      "epoch": 0.3691088260497001,
+      "grad_norm": 10.608276808995887,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 15507
+    },
+    {
+      "epoch": 0.3691326287727316,
+      "grad_norm": 4.478385916881014,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 15508
+    },
+    {
+      "epoch": 0.3691564314957631,
+      "grad_norm": 3.6414948099489344,
+      "learning_rate": 2e-05,
+      "loss": 0.4375,
+      "step": 15509
+    },
+    {
+      "epoch": 0.36918023421879465,
+      "grad_norm": 3.4277099060092695,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 15510
+    },
+    {
+      "epoch": 0.36920403694182613,
+      "grad_norm": 4.263105896828991,
+      "learning_rate": 2e-05,
+      "loss": 0.5663,
+      "step": 15511
+    },
+    {
+      "epoch": 0.36922783966485767,
+      "grad_norm": 3.21508720372148,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 15512
+    },
+    {
+      "epoch": 0.36925164238788916,
+      "grad_norm": 4.643260359702519,
+      "learning_rate": 2e-05,
+      "loss": 0.6003,
+      "step": 15513
+    },
+    {
+      "epoch": 0.3692754451109207,
+      "grad_norm": 6.201231230018953,
+      "learning_rate": 2e-05,
+      "loss": 1.9015,
+      "step": 15514
+    },
+    {
+      "epoch": 0.3692992478339522,
+      "grad_norm": 1.111403597800128,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 15515
+    },
+    {
+      "epoch": 0.3693230505569837,
+      "grad_norm": 11.727346570258401,
+      "learning_rate": 2e-05,
+      "loss": 0.4267,
+      "step": 15516
+    },
+    {
+      "epoch": 0.3693468532800152,
+      "grad_norm": 9.138590524691235,
+      "learning_rate": 2e-05,
+      "loss": 1.1044,
+      "step": 15517
+    },
+    {
+      "epoch": 0.36937065600304675,
+      "grad_norm": 4.857871076616853,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 15518
+    },
+    {
+      "epoch": 0.3693944587260783,
+      "grad_norm": 1.4076353746439199,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 15519
+    },
+    {
+      "epoch": 0.3694182614491098,
+      "grad_norm": 4.965234278094075,
+      "learning_rate": 2e-05,
+      "loss": 0.7239,
+      "step": 15520
+    },
+    {
+      "epoch": 0.3694420641721413,
+      "grad_norm": 5.418580477459597,
+      "learning_rate": 2e-05,
+      "loss": 0.1356,
+      "step": 15521
+    },
+    {
+      "epoch": 0.3694658668951728,
+      "grad_norm": 17.848974444526487,
+      "learning_rate": 2e-05,
+      "loss": 0.753,
+      "step": 15522
+    },
+    {
+      "epoch": 0.36948966961820434,
+      "grad_norm": 0.8072119289104953,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 15523
+    },
+    {
+      "epoch": 0.3695134723412358,
+      "grad_norm": 7.94460889989627,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 15524
+    },
+    {
+      "epoch": 0.36953727506426737,
+      "grad_norm": 1.4525365930761316,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 15525
+    },
+    {
+      "epoch": 0.36956107778729885,
+      "grad_norm": 2.8335255376100346,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 15526
+    },
+    {
+      "epoch": 0.3695848805103304,
+      "grad_norm": 5.704771921002427,
+      "learning_rate": 2e-05,
+      "loss": 0.3761,
+      "step": 15527
+    },
+    {
+      "epoch": 0.3696086832333619,
+      "grad_norm": 0.4530381332396022,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 15528
+    },
+    {
+      "epoch": 0.3696324859563934,
+      "grad_norm": 4.601044095109266,
+      "learning_rate": 2e-05,
+      "loss": 0.6119,
+      "step": 15529
+    },
+    {
+      "epoch": 0.3696562886794249,
+      "grad_norm": 0.9557433212863555,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 15530
+    },
+    {
+      "epoch": 0.36968009140245645,
+      "grad_norm": 13.045001586755674,
+      "learning_rate": 2e-05,
+      "loss": 0.8107,
+      "step": 15531
+    },
+    {
+      "epoch": 0.36970389412548793,
+      "grad_norm": 4.928910859915422,
+      "learning_rate": 2e-05,
+      "loss": 0.5405,
+      "step": 15532
+    },
+    {
+      "epoch": 0.36972769684851947,
+      "grad_norm": 3.914897387367425,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 15533
+    },
+    {
+      "epoch": 0.369751499571551,
+      "grad_norm": 6.564555443105227,
+      "learning_rate": 2e-05,
+      "loss": 1.1459,
+      "step": 15534
+    },
+    {
+      "epoch": 0.3697753022945825,
+      "grad_norm": 13.787259960516987,
+      "learning_rate": 2e-05,
+      "loss": 0.4764,
+      "step": 15535
+    },
+    {
+      "epoch": 0.36979910501761404,
+      "grad_norm": 2.3702798606023725,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 15536
+    },
+    {
+      "epoch": 0.3698229077406455,
+      "grad_norm": 4.011333500302266,
+      "learning_rate": 2e-05,
+      "loss": 0.6281,
+      "step": 15537
+    },
+    {
+      "epoch": 0.36984671046367706,
+      "grad_norm": 3.3589918437309274,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 15538
+    },
+    {
+      "epoch": 0.36987051318670855,
+      "grad_norm": 3.4676493759603897,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 15539
+    },
+    {
+      "epoch": 0.3698943159097401,
+      "grad_norm": 1.1606106863206307,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 15540
+    },
+    {
+      "epoch": 0.3699181186327716,
+      "grad_norm": 3.2728426860088695,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 15541
+    },
+    {
+      "epoch": 0.3699419213558031,
+      "grad_norm": 6.103004013043357,
+      "learning_rate": 2e-05,
+      "loss": 1.3485,
+      "step": 15542
+    },
+    {
+      "epoch": 0.3699657240788346,
+      "grad_norm": 4.537791891624216,
+      "learning_rate": 2e-05,
+      "loss": 0.5568,
+      "step": 15543
+    },
+    {
+      "epoch": 0.36998952680186614,
+      "grad_norm": 4.429515562074506,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 15544
+    },
+    {
+      "epoch": 0.3700133295248976,
+      "grad_norm": 3.8263155886033964,
+      "learning_rate": 2e-05,
+      "loss": 0.483,
+      "step": 15545
+    },
+    {
+      "epoch": 0.37003713224792917,
+      "grad_norm": 2.0983813597237653,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 15546
+    },
+    {
+      "epoch": 0.37006093497096065,
+      "grad_norm": 5.271505013507159,
+      "learning_rate": 2e-05,
+      "loss": 0.7766,
+      "step": 15547
+    },
+    {
+      "epoch": 0.3700847376939922,
+      "grad_norm": 6.958527928717982,
+      "learning_rate": 2e-05,
+      "loss": 0.6866,
+      "step": 15548
+    },
+    {
+      "epoch": 0.37010854041702373,
+      "grad_norm": 13.841732798433894,
+      "learning_rate": 2e-05,
+      "loss": 0.446,
+      "step": 15549
+    },
+    {
+      "epoch": 0.3701323431400552,
+      "grad_norm": 5.278834495371719,
+      "learning_rate": 2e-05,
+      "loss": 0.6763,
+      "step": 15550
+    },
+    {
+      "epoch": 0.37015614586308676,
+      "grad_norm": 1.256889428811503,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 15551
+    },
+    {
+      "epoch": 0.37017994858611825,
+      "grad_norm": 5.203114070760019,
+      "learning_rate": 2e-05,
+      "loss": 0.9389,
+      "step": 15552
+    },
+    {
+      "epoch": 0.3702037513091498,
+      "grad_norm": 2.7464361129737616,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 15553
+    },
+    {
+      "epoch": 0.37022755403218127,
+      "grad_norm": 11.375815674916339,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 15554
+    },
+    {
+      "epoch": 0.3702513567552128,
+      "grad_norm": 0.37903527906670903,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 15555
+    },
+    {
+      "epoch": 0.3702751594782443,
+      "grad_norm": 1.3422876454806374,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 15556
+    },
+    {
+      "epoch": 0.37029896220127584,
+      "grad_norm": 2.38205429882656,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 15557
+    },
+    {
+      "epoch": 0.3703227649243073,
+      "grad_norm": 3.044580291469161,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 15558
+    },
+    {
+      "epoch": 0.37034656764733886,
+      "grad_norm": 4.141739923107232,
+      "learning_rate": 2e-05,
+      "loss": 0.6676,
+      "step": 15559
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 3.331950619823879,
+      "learning_rate": 2e-05,
+      "loss": 0.3157,
+      "step": 15560
+    },
+    {
+      "epoch": 0.3703941730934019,
+      "grad_norm": 4.70677251727305,
+      "learning_rate": 2e-05,
+      "loss": 0.7201,
+      "step": 15561
+    },
+    {
+      "epoch": 0.3704179758164334,
+      "grad_norm": 1.5447212429724244,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 15562
+    },
+    {
+      "epoch": 0.3704417785394649,
+      "grad_norm": 29.446030381489265,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 15563
+    },
+    {
+      "epoch": 0.37046558126249646,
+      "grad_norm": 4.524302735176372,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 15564
+    },
+    {
+      "epoch": 0.37048938398552794,
+      "grad_norm": 3.4471751002646522,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 15565
+    },
+    {
+      "epoch": 0.3705131867085595,
+      "grad_norm": 5.847390481998326,
+      "learning_rate": 2e-05,
+      "loss": 0.9047,
+      "step": 15566
+    },
+    {
+      "epoch": 0.37053698943159097,
+      "grad_norm": 109.49045412411519,
+      "learning_rate": 2e-05,
+      "loss": 0.7935,
+      "step": 15567
+    },
+    {
+      "epoch": 0.3705607921546225,
+      "grad_norm": 2.1781432901824753,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 15568
+    },
+    {
+      "epoch": 0.370584594877654,
+      "grad_norm": 2.4594678117082376,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 15569
+    },
+    {
+      "epoch": 0.37060839760068554,
+      "grad_norm": 4.354262582842197,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 15570
+    },
+    {
+      "epoch": 0.370632200323717,
+      "grad_norm": 0.13131070256040475,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 15571
+    },
+    {
+      "epoch": 0.37065600304674856,
+      "grad_norm": 1.8503874237825149,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 15572
+    },
+    {
+      "epoch": 0.37067980576978005,
+      "grad_norm": 5.675430384010278,
+      "learning_rate": 2e-05,
+      "loss": 0.3643,
+      "step": 15573
+    },
+    {
+      "epoch": 0.3707036084928116,
+      "grad_norm": 3.7634359692676376,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 15574
+    },
+    {
+      "epoch": 0.3707274112158431,
+      "grad_norm": 3.4898258303554877,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 15575
+    },
+    {
+      "epoch": 0.3707512139388746,
+      "grad_norm": 3.1037974627604017,
+      "learning_rate": 2e-05,
+      "loss": 0.3001,
+      "step": 15576
+    },
+    {
+      "epoch": 0.3707750166619061,
+      "grad_norm": 0.6270256282144627,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 15577
+    },
+    {
+      "epoch": 0.37079881938493764,
+      "grad_norm": 8.324550656057601,
+      "learning_rate": 2e-05,
+      "loss": 0.6223,
+      "step": 15578
+    },
+    {
+      "epoch": 0.3708226221079691,
+      "grad_norm": 4.692017144148747,
+      "learning_rate": 2e-05,
+      "loss": 0.6947,
+      "step": 15579
+    },
+    {
+      "epoch": 0.37084642483100067,
+      "grad_norm": 27.742794578105038,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 15580
+    },
+    {
+      "epoch": 0.3708702275540322,
+      "grad_norm": 1.3159257252722643,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 15581
+    },
+    {
+      "epoch": 0.3708940302770637,
+      "grad_norm": 16.42448066720252,
+      "learning_rate": 2e-05,
+      "loss": 0.8697,
+      "step": 15582
+    },
+    {
+      "epoch": 0.37091783300009523,
+      "grad_norm": 46.01979699370162,
+      "learning_rate": 2e-05,
+      "loss": 0.3852,
+      "step": 15583
+    },
+    {
+      "epoch": 0.3709416357231267,
+      "grad_norm": 2.2049354672803885,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 15584
+    },
+    {
+      "epoch": 0.37096543844615826,
+      "grad_norm": 1.2069716617985211,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 15585
+    },
+    {
+      "epoch": 0.37098924116918974,
+      "grad_norm": 17.72486030428871,
+      "learning_rate": 2e-05,
+      "loss": 1.0755,
+      "step": 15586
+    },
+    {
+      "epoch": 0.3710130438922213,
+      "grad_norm": 3.538564245649191,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 15587
+    },
+    {
+      "epoch": 0.37103684661525277,
+      "grad_norm": 6.282368642206619,
+      "learning_rate": 2e-05,
+      "loss": 0.5983,
+      "step": 15588
+    },
+    {
+      "epoch": 0.3710606493382843,
+      "grad_norm": 4.847468822370022,
+      "learning_rate": 2e-05,
+      "loss": 0.3911,
+      "step": 15589
+    },
+    {
+      "epoch": 0.3710844520613158,
+      "grad_norm": 91.58778140247583,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 15590
+    },
+    {
+      "epoch": 0.37110825478434734,
+      "grad_norm": 0.36081615239959514,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 15591
+    },
+    {
+      "epoch": 0.3711320575073788,
+      "grad_norm": 4.679768731386682,
+      "learning_rate": 2e-05,
+      "loss": 0.9318,
+      "step": 15592
+    },
+    {
+      "epoch": 0.37115586023041036,
+      "grad_norm": 1.3144597612582138,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15593
+    },
+    {
+      "epoch": 0.37117966295344185,
+      "grad_norm": 2.839720216257574,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 15594
+    },
+    {
+      "epoch": 0.3712034656764734,
+      "grad_norm": 0.13785096349369524,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 15595
+    },
+    {
+      "epoch": 0.37122726839950493,
+      "grad_norm": 4.451324566316733,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 15596
+    },
+    {
+      "epoch": 0.3712510711225364,
+      "grad_norm": 4.093354912178372,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 15597
+    },
+    {
+      "epoch": 0.37127487384556795,
+      "grad_norm": 8.998125500609095,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 15598
+    },
+    {
+      "epoch": 0.37129867656859944,
+      "grad_norm": 8.271493530088561,
+      "learning_rate": 2e-05,
+      "loss": 0.3014,
+      "step": 15599
+    },
+    {
+      "epoch": 0.371322479291631,
+      "grad_norm": 3.2666028038073414,
+      "learning_rate": 2e-05,
+      "loss": 0.4445,
+      "step": 15600
+    },
+    {
+      "epoch": 0.37134628201466247,
+      "grad_norm": 1.4219568981543864,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 15601
+    },
+    {
+      "epoch": 0.371370084737694,
+      "grad_norm": 2.734018979000586,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 15602
+    },
+    {
+      "epoch": 0.3713938874607255,
+      "grad_norm": 7.811166828820367,
+      "learning_rate": 2e-05,
+      "loss": 0.2408,
+      "step": 15603
+    },
+    {
+      "epoch": 0.37141769018375703,
+      "grad_norm": 0.21824591047123093,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 15604
+    },
+    {
+      "epoch": 0.3714414929067885,
+      "grad_norm": 2.341563296826225,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 15605
+    },
+    {
+      "epoch": 0.37146529562982006,
+      "grad_norm": 2.67196840002129,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 15606
+    },
+    {
+      "epoch": 0.37148909835285154,
+      "grad_norm": 10.301004190516577,
+      "learning_rate": 2e-05,
+      "loss": 1.1086,
+      "step": 15607
+    },
+    {
+      "epoch": 0.3715129010758831,
+      "grad_norm": 3.3863241478459574,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 15608
+    },
+    {
+      "epoch": 0.37153670379891457,
+      "grad_norm": 1.3642365643399175,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15609
+    },
+    {
+      "epoch": 0.3715605065219461,
+      "grad_norm": 0.6140853185093486,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 15610
+    },
+    {
+      "epoch": 0.37158430924497765,
+      "grad_norm": 9.317401904558052,
+      "learning_rate": 2e-05,
+      "loss": 0.2544,
+      "step": 15611
+    },
+    {
+      "epoch": 0.37160811196800914,
+      "grad_norm": 4.149976308679262,
+      "learning_rate": 2e-05,
+      "loss": 0.3942,
+      "step": 15612
+    },
+    {
+      "epoch": 0.3716319146910407,
+      "grad_norm": 13.666209182605918,
+      "learning_rate": 2e-05,
+      "loss": 0.8525,
+      "step": 15613
+    },
+    {
+      "epoch": 0.37165571741407216,
+      "grad_norm": 5.1830011222520715,
+      "learning_rate": 2e-05,
+      "loss": 0.7488,
+      "step": 15614
+    },
+    {
+      "epoch": 0.3716795201371037,
+      "grad_norm": 5.570554382227688,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 15615
+    },
+    {
+      "epoch": 0.3717033228601352,
+      "grad_norm": 3.2379630178663925,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 15616
+    },
+    {
+      "epoch": 0.37172712558316673,
+      "grad_norm": 1.226168999416376,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 15617
+    },
+    {
+      "epoch": 0.3717509283061982,
+      "grad_norm": 3.906866322838099,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 15618
+    },
+    {
+      "epoch": 0.37177473102922975,
+      "grad_norm": 13.538501569225605,
+      "learning_rate": 2e-05,
+      "loss": 0.2509,
+      "step": 15619
+    },
+    {
+      "epoch": 0.37179853375226124,
+      "grad_norm": 2.886859721615798,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 15620
+    },
+    {
+      "epoch": 0.3718223364752928,
+      "grad_norm": 3.368621140664813,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 15621
+    },
+    {
+      "epoch": 0.37184613919832427,
+      "grad_norm": 1.5694503891915927,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 15622
+    },
+    {
+      "epoch": 0.3718699419213558,
+      "grad_norm": 18.634925314865793,
+      "learning_rate": 2e-05,
+      "loss": 1.1978,
+      "step": 15623
+    },
+    {
+      "epoch": 0.3718937446443873,
+      "grad_norm": 18.756237973052162,
+      "learning_rate": 2e-05,
+      "loss": 0.5316,
+      "step": 15624
+    },
+    {
+      "epoch": 0.37191754736741883,
+      "grad_norm": 10.706984841333332,
+      "learning_rate": 2e-05,
+      "loss": 0.4725,
+      "step": 15625
+    },
+    {
+      "epoch": 0.3719413500904504,
+      "grad_norm": 4.819060638994931,
+      "learning_rate": 2e-05,
+      "loss": 0.641,
+      "step": 15626
+    },
+    {
+      "epoch": 0.37196515281348186,
+      "grad_norm": 57.83502568832763,
+      "learning_rate": 2e-05,
+      "loss": 0.3153,
+      "step": 15627
+    },
+    {
+      "epoch": 0.3719889555365134,
+      "grad_norm": 3.370836914756732,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 15628
+    },
+    {
+      "epoch": 0.3720127582595449,
+      "grad_norm": 155.0309682610424,
+      "learning_rate": 2e-05,
+      "loss": 0.5299,
+      "step": 15629
+    },
+    {
+      "epoch": 0.3720365609825764,
+      "grad_norm": 12.24010160994891,
+      "learning_rate": 2e-05,
+      "loss": 0.5253,
+      "step": 15630
+    },
+    {
+      "epoch": 0.3720603637056079,
+      "grad_norm": 8.814560670292725,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 15631
+    },
+    {
+      "epoch": 0.37208416642863945,
+      "grad_norm": 2.827851763367169,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 15632
+    },
+    {
+      "epoch": 0.37210796915167094,
+      "grad_norm": 5.108222413683304,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 15633
+    },
+    {
+      "epoch": 0.3721317718747025,
+      "grad_norm": 5.867771613419474,
+      "learning_rate": 2e-05,
+      "loss": 0.6846,
+      "step": 15634
+    },
+    {
+      "epoch": 0.37215557459773396,
+      "grad_norm": 6.320620827635166,
+      "learning_rate": 2e-05,
+      "loss": 1.1107,
+      "step": 15635
+    },
+    {
+      "epoch": 0.3721793773207655,
+      "grad_norm": 1.0786587764851525,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 15636
+    },
+    {
+      "epoch": 0.372203180043797,
+      "grad_norm": 1.2713576706988443,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 15637
+    },
+    {
+      "epoch": 0.37222698276682853,
+      "grad_norm": 0.8371201067413518,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 15638
+    },
+    {
+      "epoch": 0.37225078548986,
+      "grad_norm": 10.130056199491602,
+      "learning_rate": 2e-05,
+      "loss": 0.5819,
+      "step": 15639
+    },
+    {
+      "epoch": 0.37227458821289156,
+      "grad_norm": 15.803905412310883,
+      "learning_rate": 2e-05,
+      "loss": 0.4124,
+      "step": 15640
+    },
+    {
+      "epoch": 0.3722983909359231,
+      "grad_norm": 2.667912375684788,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 15641
+    },
+    {
+      "epoch": 0.3723221936589546,
+      "grad_norm": 5.875638649546529,
+      "learning_rate": 2e-05,
+      "loss": 1.2755,
+      "step": 15642
+    },
+    {
+      "epoch": 0.3723459963819861,
+      "grad_norm": 13.647132439937481,
+      "learning_rate": 2e-05,
+      "loss": 0.8636,
+      "step": 15643
+    },
+    {
+      "epoch": 0.3723697991050176,
+      "grad_norm": 7.910325788318325,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 15644
+    },
+    {
+      "epoch": 0.37239360182804915,
+      "grad_norm": 0.05216049204011755,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 15645
+    },
+    {
+      "epoch": 0.37241740455108063,
+      "grad_norm": 0.9317118386710238,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 15646
+    },
+    {
+      "epoch": 0.3724412072741122,
+      "grad_norm": 6.440381961908279,
+      "learning_rate": 2e-05,
+      "loss": 1.0889,
+      "step": 15647
+    },
+    {
+      "epoch": 0.37246500999714366,
+      "grad_norm": 3.0568634720845846,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 15648
+    },
+    {
+      "epoch": 0.3724888127201752,
+      "grad_norm": 0.27229511424547426,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 15649
+    },
+    {
+      "epoch": 0.3725126154432067,
+      "grad_norm": 7.562529420750614,
+      "learning_rate": 2e-05,
+      "loss": 1.2126,
+      "step": 15650
+    },
+    {
+      "epoch": 0.3725364181662382,
+      "grad_norm": 3.94248670028287,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 15651
+    },
+    {
+      "epoch": 0.3725602208892697,
+      "grad_norm": 8.068204735491303,
+      "learning_rate": 2e-05,
+      "loss": 0.2829,
+      "step": 15652
+    },
+    {
+      "epoch": 0.37258402361230125,
+      "grad_norm": 2.038913769987323,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 15653
+    },
+    {
+      "epoch": 0.37260782633533274,
+      "grad_norm": 5.32118685572127,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 15654
+    },
+    {
+      "epoch": 0.3726316290583643,
+      "grad_norm": 5.6415092635235125,
+      "learning_rate": 2e-05,
+      "loss": 0.2906,
+      "step": 15655
+    },
+    {
+      "epoch": 0.3726554317813958,
+      "grad_norm": 7.788629427456827,
+      "learning_rate": 2e-05,
+      "loss": 2.072,
+      "step": 15656
+    },
+    {
+      "epoch": 0.3726792345044273,
+      "grad_norm": 0.7693891110031617,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 15657
+    },
+    {
+      "epoch": 0.37270303722745884,
+      "grad_norm": 8.181205203504067,
+      "learning_rate": 2e-05,
+      "loss": 0.5487,
+      "step": 15658
+    },
+    {
+      "epoch": 0.37272683995049033,
+      "grad_norm": 3.7186557863118748,
+      "learning_rate": 2e-05,
+      "loss": 0.6723,
+      "step": 15659
+    },
+    {
+      "epoch": 0.37275064267352187,
+      "grad_norm": 2.9945681912217474,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 15660
+    },
+    {
+      "epoch": 0.37277444539655336,
+      "grad_norm": 1.7195872847956166,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 15661
+    },
+    {
+      "epoch": 0.3727982481195849,
+      "grad_norm": 9.576698573235927,
+      "learning_rate": 2e-05,
+      "loss": 0.3594,
+      "step": 15662
+    },
+    {
+      "epoch": 0.3728220508426164,
+      "grad_norm": 5.558114317309817,
+      "learning_rate": 2e-05,
+      "loss": 0.822,
+      "step": 15663
+    },
+    {
+      "epoch": 0.3728458535656479,
+      "grad_norm": 5.334569936456185,
+      "learning_rate": 2e-05,
+      "loss": 0.5244,
+      "step": 15664
+    },
+    {
+      "epoch": 0.3728696562886794,
+      "grad_norm": 8.758908737101047,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 15665
+    },
+    {
+      "epoch": 0.37289345901171095,
+      "grad_norm": 1.7250273899529622,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 15666
+    },
+    {
+      "epoch": 0.37291726173474243,
+      "grad_norm": 3.151851139921698,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 15667
+    },
+    {
+      "epoch": 0.372941064457774,
+      "grad_norm": 3.390405438234939,
+      "learning_rate": 2e-05,
+      "loss": 0.5623,
+      "step": 15668
+    },
+    {
+      "epoch": 0.37296486718080546,
+      "grad_norm": 2.0211295891245853,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 15669
+    },
+    {
+      "epoch": 0.372988669903837,
+      "grad_norm": 3.1608192686154037,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 15670
+    },
+    {
+      "epoch": 0.37301247262686854,
+      "grad_norm": 4.840692705928113,
+      "learning_rate": 2e-05,
+      "loss": 0.9505,
+      "step": 15671
+    },
+    {
+      "epoch": 0.3730362753499,
+      "grad_norm": 36.55673319454581,
+      "learning_rate": 2e-05,
+      "loss": 1.0837,
+      "step": 15672
+    },
+    {
+      "epoch": 0.37306007807293157,
+      "grad_norm": 13.081086668079882,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 15673
+    },
+    {
+      "epoch": 0.37308388079596305,
+      "grad_norm": 2.939234257610221,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 15674
+    },
+    {
+      "epoch": 0.3731076835189946,
+      "grad_norm": 4.118467606789106,
+      "learning_rate": 2e-05,
+      "loss": 0.5262,
+      "step": 15675
+    },
+    {
+      "epoch": 0.3731314862420261,
+      "grad_norm": 2.7911414414994984,
+      "learning_rate": 2e-05,
+      "loss": 0.4147,
+      "step": 15676
+    },
+    {
+      "epoch": 0.3731552889650576,
+      "grad_norm": 2.4106467724124494,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 15677
+    },
+    {
+      "epoch": 0.3731790916880891,
+      "grad_norm": 1.438097060672024,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 15678
+    },
+    {
+      "epoch": 0.37320289441112064,
+      "grad_norm": 3.11863928580944,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 15679
+    },
+    {
+      "epoch": 0.37322669713415213,
+      "grad_norm": 1.7834630201606219,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 15680
+    },
+    {
+      "epoch": 0.37325049985718367,
+      "grad_norm": 1.230526279620074,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15681
+    },
+    {
+      "epoch": 0.37327430258021516,
+      "grad_norm": 10.778788809894548,
+      "learning_rate": 2e-05,
+      "loss": 0.4107,
+      "step": 15682
+    },
+    {
+      "epoch": 0.3732981053032467,
+      "grad_norm": 3.7118651013443538,
+      "learning_rate": 2e-05,
+      "loss": 0.7181,
+      "step": 15683
+    },
+    {
+      "epoch": 0.3733219080262782,
+      "grad_norm": 1.146511970579895,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 15684
+    },
+    {
+      "epoch": 0.3733457107493097,
+      "grad_norm": 5.713938747500083,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 15685
+    },
+    {
+      "epoch": 0.37336951347234126,
+      "grad_norm": 14.006749665980871,
+      "learning_rate": 2e-05,
+      "loss": 0.6428,
+      "step": 15686
+    },
+    {
+      "epoch": 0.37339331619537275,
+      "grad_norm": 7.1013402445550895,
+      "learning_rate": 2e-05,
+      "loss": 0.2923,
+      "step": 15687
+    },
+    {
+      "epoch": 0.3734171189184043,
+      "grad_norm": 7.717803825134527,
+      "learning_rate": 2e-05,
+      "loss": 1.2547,
+      "step": 15688
+    },
+    {
+      "epoch": 0.3734409216414358,
+      "grad_norm": 4.854611280526343,
+      "learning_rate": 2e-05,
+      "loss": 0.8548,
+      "step": 15689
+    },
+    {
+      "epoch": 0.3734647243644673,
+      "grad_norm": 3.6022686003548436,
+      "learning_rate": 2e-05,
+      "loss": 0.5098,
+      "step": 15690
+    },
+    {
+      "epoch": 0.3734885270874988,
+      "grad_norm": 1.7325289883042092,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 15691
+    },
+    {
+      "epoch": 0.37351232981053034,
+      "grad_norm": 2.3129015006754146,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 15692
+    },
+    {
+      "epoch": 0.3735361325335618,
+      "grad_norm": 827.205009179538,
+      "learning_rate": 2e-05,
+      "loss": 0.913,
+      "step": 15693
+    },
+    {
+      "epoch": 0.37355993525659337,
+      "grad_norm": 6.347030002604748,
+      "learning_rate": 2e-05,
+      "loss": 0.292,
+      "step": 15694
+    },
+    {
+      "epoch": 0.37358373797962485,
+      "grad_norm": 2.2363094334060545,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 15695
+    },
+    {
+      "epoch": 0.3736075407026564,
+      "grad_norm": 4.257964901292419,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 15696
+    },
+    {
+      "epoch": 0.3736313434256879,
+      "grad_norm": 5.227670615586482,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 15697
+    },
+    {
+      "epoch": 0.3736551461487194,
+      "grad_norm": 0.8983515882588163,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 15698
+    },
+    {
+      "epoch": 0.3736789488717509,
+      "grad_norm": 5.494748168413347,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 15699
+    },
+    {
+      "epoch": 0.37370275159478245,
+      "grad_norm": 2.9064519741768198,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 15700
+    },
+    {
+      "epoch": 0.37372655431781393,
+      "grad_norm": 1.5354623192994208,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 15701
+    },
+    {
+      "epoch": 0.37375035704084547,
+      "grad_norm": 1.8408771540192845,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 15702
+    },
+    {
+      "epoch": 0.373774159763877,
+      "grad_norm": 1.6376550966989054,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 15703
+    },
+    {
+      "epoch": 0.3737979624869085,
+      "grad_norm": 5.391025233417916,
+      "learning_rate": 2e-05,
+      "loss": 0.5422,
+      "step": 15704
+    },
+    {
+      "epoch": 0.37382176520994004,
+      "grad_norm": 2.7206649964869976,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 15705
+    },
+    {
+      "epoch": 0.3738455679329715,
+      "grad_norm": 7.391370379395411,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 15706
+    },
+    {
+      "epoch": 0.37386937065600306,
+      "grad_norm": 4.070203336170965,
+      "learning_rate": 2e-05,
+      "loss": 0.8803,
+      "step": 15707
+    },
+    {
+      "epoch": 0.37389317337903455,
+      "grad_norm": 6.066270858501287,
+      "learning_rate": 2e-05,
+      "loss": 0.8643,
+      "step": 15708
+    },
+    {
+      "epoch": 0.3739169761020661,
+      "grad_norm": 4.176318738456514,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 15709
+    },
+    {
+      "epoch": 0.3739407788250976,
+      "grad_norm": 4.874287643001483,
+      "learning_rate": 2e-05,
+      "loss": 0.157,
+      "step": 15710
+    },
+    {
+      "epoch": 0.3739645815481291,
+      "grad_norm": 2.0194180207660803,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 15711
+    },
+    {
+      "epoch": 0.3739883842711606,
+      "grad_norm": 6.3144613759146955,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 15712
+    },
+    {
+      "epoch": 0.37401218699419214,
+      "grad_norm": 3.40978916031992,
+      "learning_rate": 2e-05,
+      "loss": 0.4739,
+      "step": 15713
+    },
+    {
+      "epoch": 0.3740359897172236,
+      "grad_norm": 6.185633901432736,
+      "learning_rate": 2e-05,
+      "loss": 0.9255,
+      "step": 15714
+    },
+    {
+      "epoch": 0.37405979244025517,
+      "grad_norm": 3.8718722686477487,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 15715
+    },
+    {
+      "epoch": 0.37408359516328665,
+      "grad_norm": 3.8352881643163013,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 15716
+    },
+    {
+      "epoch": 0.3741073978863182,
+      "grad_norm": 4.587088195587093,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 15717
+    },
+    {
+      "epoch": 0.37413120060934973,
+      "grad_norm": 8.704950824894274,
+      "learning_rate": 2e-05,
+      "loss": 0.1775,
+      "step": 15718
+    },
+    {
+      "epoch": 0.3741550033323812,
+      "grad_norm": 3.6304446849046266,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 15719
+    },
+    {
+      "epoch": 0.37417880605541276,
+      "grad_norm": 3.3656464893694173,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 15720
+    },
+    {
+      "epoch": 0.37420260877844425,
+      "grad_norm": 1.0855487450279828,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 15721
+    },
+    {
+      "epoch": 0.3742264115014758,
+      "grad_norm": 10.346332286353077,
+      "learning_rate": 2e-05,
+      "loss": 0.473,
+      "step": 15722
+    },
+    {
+      "epoch": 0.37425021422450727,
+      "grad_norm": 4.527815047741237,
+      "learning_rate": 2e-05,
+      "loss": 1.0283,
+      "step": 15723
+    },
+    {
+      "epoch": 0.3742740169475388,
+      "grad_norm": 1.5102438760950117,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 15724
+    },
+    {
+      "epoch": 0.3742978196705703,
+      "grad_norm": 2.098561125794148,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 15725
+    },
+    {
+      "epoch": 0.37432162239360184,
+      "grad_norm": 3.4193680911330384,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 15726
+    },
+    {
+      "epoch": 0.3743454251166333,
+      "grad_norm": 0.48533478255568024,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 15727
+    },
+    {
+      "epoch": 0.37436922783966486,
+      "grad_norm": 8.423904193232294,
+      "learning_rate": 2e-05,
+      "loss": 0.2746,
+      "step": 15728
+    },
+    {
+      "epoch": 0.37439303056269635,
+      "grad_norm": 1.5277270571134698,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 15729
+    },
+    {
+      "epoch": 0.3744168332857279,
+      "grad_norm": 3.894397643850278,
+      "learning_rate": 2e-05,
+      "loss": 0.1444,
+      "step": 15730
+    },
+    {
+      "epoch": 0.3744406360087594,
+      "grad_norm": 0.1816784720065262,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 15731
+    },
+    {
+      "epoch": 0.3744644387317909,
+      "grad_norm": 4.6768917041937526,
+      "learning_rate": 2e-05,
+      "loss": 0.2568,
+      "step": 15732
+    },
+    {
+      "epoch": 0.37448824145482246,
+      "grad_norm": 7.9047132519289995,
+      "learning_rate": 2e-05,
+      "loss": 1.0966,
+      "step": 15733
+    },
+    {
+      "epoch": 0.37451204417785394,
+      "grad_norm": 2.894466220766232,
+      "learning_rate": 2e-05,
+      "loss": 0.2748,
+      "step": 15734
+    },
+    {
+      "epoch": 0.3745358469008855,
+      "grad_norm": 6.185862346402519,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 15735
+    },
+    {
+      "epoch": 0.37455964962391697,
+      "grad_norm": 2.7650677992645663,
+      "learning_rate": 2e-05,
+      "loss": 0.5964,
+      "step": 15736
+    },
+    {
+      "epoch": 0.3745834523469485,
+      "grad_norm": 0.7517552687627197,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 15737
+    },
+    {
+      "epoch": 0.37460725506998,
+      "grad_norm": 5.483669655563995,
+      "learning_rate": 2e-05,
+      "loss": 1.106,
+      "step": 15738
+    },
+    {
+      "epoch": 0.37463105779301153,
+      "grad_norm": 2.2499258107662508,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 15739
+    },
+    {
+      "epoch": 0.374654860516043,
+      "grad_norm": 0.6421417845782958,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37467866323907456,
+      "grad_norm": 5.016960078217689,
+      "learning_rate": 2e-05,
+      "loss": 0.8424,
+      "step": 15741
+    },
+    {
+      "epoch": 0.37470246596210605,
+      "grad_norm": 4.380159445041776,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 15742
+    },
+    {
+      "epoch": 0.3747262686851376,
+      "grad_norm": 4.24815653357944,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 15743
+    },
+    {
+      "epoch": 0.37475007140816907,
+      "grad_norm": 6.713439617803856,
+      "learning_rate": 2e-05,
+      "loss": 0.2831,
+      "step": 15744
+    },
+    {
+      "epoch": 0.3747738741312006,
+      "grad_norm": 5.598820589430518,
+      "learning_rate": 2e-05,
+      "loss": 0.5087,
+      "step": 15745
+    },
+    {
+      "epoch": 0.3747976768542321,
+      "grad_norm": 4.2090610777505795,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 15746
+    },
+    {
+      "epoch": 0.37482147957726364,
+      "grad_norm": 8.586976087462387,
+      "learning_rate": 2e-05,
+      "loss": 1.5545,
+      "step": 15747
+    },
+    {
+      "epoch": 0.3748452823002952,
+      "grad_norm": 3.157312686567082,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 15748
+    },
+    {
+      "epoch": 0.37486908502332666,
+      "grad_norm": 1.0545241857197696,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 15749
+    },
+    {
+      "epoch": 0.3748928877463582,
+      "grad_norm": 0.9538301643021844,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 15750
+    },
+    {
+      "epoch": 0.3749166904693897,
+      "grad_norm": 5.413267503298712,
+      "learning_rate": 2e-05,
+      "loss": 1.4528,
+      "step": 15751
+    },
+    {
+      "epoch": 0.37494049319242123,
+      "grad_norm": 0.7283465161799418,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 15752
+    },
+    {
+      "epoch": 0.3749642959154527,
+      "grad_norm": 11.584344664214242,
+      "learning_rate": 2e-05,
+      "loss": 0.3882,
+      "step": 15753
+    },
+    {
+      "epoch": 0.37498809863848426,
+      "grad_norm": 18.40637183720053,
+      "learning_rate": 2e-05,
+      "loss": 0.651,
+      "step": 15754
+    },
+    {
+      "epoch": 0.37501190136151574,
+      "grad_norm": 2.0013952618329767,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 15755
+    },
+    {
+      "epoch": 0.3750357040845473,
+      "grad_norm": 2.4960212933613817,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 15756
+    },
+    {
+      "epoch": 0.37505950680757877,
+      "grad_norm": 17.450143789075614,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 15757
+    },
+    {
+      "epoch": 0.3750833095306103,
+      "grad_norm": 3.615305864907284,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 15758
+    },
+    {
+      "epoch": 0.3751071122536418,
+      "grad_norm": 0.4394981983232437,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 15759
+    },
+    {
+      "epoch": 0.37513091497667334,
+      "grad_norm": 3.7314122859937395,
+      "learning_rate": 2e-05,
+      "loss": 0.4731,
+      "step": 15760
+    },
+    {
+      "epoch": 0.3751547176997048,
+      "grad_norm": 5.586725487919717,
+      "learning_rate": 2e-05,
+      "loss": 0.917,
+      "step": 15761
+    },
+    {
+      "epoch": 0.37517852042273636,
+      "grad_norm": 5.916812286732469,
+      "learning_rate": 2e-05,
+      "loss": 1.3164,
+      "step": 15762
+    },
+    {
+      "epoch": 0.3752023231457679,
+      "grad_norm": 4.678819508517318,
+      "learning_rate": 2e-05,
+      "loss": 0.8424,
+      "step": 15763
+    },
+    {
+      "epoch": 0.3752261258687994,
+      "grad_norm": 3.3876881109556387,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 15764
+    },
+    {
+      "epoch": 0.37524992859183093,
+      "grad_norm": 6.636056866306881,
+      "learning_rate": 2e-05,
+      "loss": 1.0295,
+      "step": 15765
+    },
+    {
+      "epoch": 0.3752737313148624,
+      "grad_norm": 78.90663844383124,
+      "learning_rate": 2e-05,
+      "loss": 1.0692,
+      "step": 15766
+    },
+    {
+      "epoch": 0.37529753403789395,
+      "grad_norm": 3.8599255494615066,
+      "learning_rate": 2e-05,
+      "loss": 0.7829,
+      "step": 15767
+    },
+    {
+      "epoch": 0.37532133676092544,
+      "grad_norm": 3.229119095766469,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 15768
+    },
+    {
+      "epoch": 0.375345139483957,
+      "grad_norm": 2.00301785208239,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 15769
+    },
+    {
+      "epoch": 0.37536894220698847,
+      "grad_norm": 0.32531803504553564,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 15770
+    },
+    {
+      "epoch": 0.37539274493002,
+      "grad_norm": 3.1787935985965623,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 15771
+    },
+    {
+      "epoch": 0.3754165476530515,
+      "grad_norm": 3.4671637076046165,
+      "learning_rate": 2e-05,
+      "loss": 0.5337,
+      "step": 15772
+    },
+    {
+      "epoch": 0.37544035037608303,
+      "grad_norm": 0.9417268401772002,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 15773
+    },
+    {
+      "epoch": 0.3754641530991145,
+      "grad_norm": 9.117038229334767,
+      "learning_rate": 2e-05,
+      "loss": 1.0162,
+      "step": 15774
+    },
+    {
+      "epoch": 0.37548795582214606,
+      "grad_norm": 3.722568641226467,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 15775
+    },
+    {
+      "epoch": 0.37551175854517754,
+      "grad_norm": 1.7019351543146195,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 15776
+    },
+    {
+      "epoch": 0.3755355612682091,
+      "grad_norm": 5.977251095812776,
+      "learning_rate": 2e-05,
+      "loss": 0.805,
+      "step": 15777
+    },
+    {
+      "epoch": 0.3755593639912406,
+      "grad_norm": 3.0390428965190055,
+      "learning_rate": 2e-05,
+      "loss": 0.1783,
+      "step": 15778
+    },
+    {
+      "epoch": 0.3755831667142721,
+      "grad_norm": 50.57766809004795,
+      "learning_rate": 2e-05,
+      "loss": 0.5446,
+      "step": 15779
+    },
+    {
+      "epoch": 0.37560696943730365,
+      "grad_norm": 1.3297290148215823,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 15780
+    },
+    {
+      "epoch": 0.37563077216033514,
+      "grad_norm": 1.1495754134573724,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 15781
+    },
+    {
+      "epoch": 0.3756545748833667,
+      "grad_norm": 4.668480918923046,
+      "learning_rate": 2e-05,
+      "loss": 0.5404,
+      "step": 15782
+    },
+    {
+      "epoch": 0.37567837760639816,
+      "grad_norm": 2.6456154394666886,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 15783
+    },
+    {
+      "epoch": 0.3757021803294297,
+      "grad_norm": 3.6657278989088584,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 15784
+    },
+    {
+      "epoch": 0.3757259830524612,
+      "grad_norm": 3.1011725575860867,
+      "learning_rate": 2e-05,
+      "loss": 0.4075,
+      "step": 15785
+    },
+    {
+      "epoch": 0.37574978577549273,
+      "grad_norm": 2.240272502172784,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 15786
+    },
+    {
+      "epoch": 0.3757735884985242,
+      "grad_norm": 2.8932079213260806,
+      "learning_rate": 2e-05,
+      "loss": 0.4967,
+      "step": 15787
+    },
+    {
+      "epoch": 0.37579739122155575,
+      "grad_norm": 4.646855644562528,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 15788
+    },
+    {
+      "epoch": 0.37582119394458724,
+      "grad_norm": 4.155210775996429,
+      "learning_rate": 2e-05,
+      "loss": 0.6244,
+      "step": 15789
+    },
+    {
+      "epoch": 0.3758449966676188,
+      "grad_norm": 12.343785068488186,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 15790
+    },
+    {
+      "epoch": 0.37586879939065027,
+      "grad_norm": 0.943690911403046,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 15791
+    },
+    {
+      "epoch": 0.3758926021136818,
+      "grad_norm": 0.4948242440067188,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 15792
+    },
+    {
+      "epoch": 0.37591640483671335,
+      "grad_norm": 5.5224611376265464,
+      "learning_rate": 2e-05,
+      "loss": 0.7993,
+      "step": 15793
+    },
+    {
+      "epoch": 0.37594020755974483,
+      "grad_norm": 3.449400760214862,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 15794
+    },
+    {
+      "epoch": 0.3759640102827764,
+      "grad_norm": 1.1328676212718791,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 15795
+    },
+    {
+      "epoch": 0.37598781300580786,
+      "grad_norm": 8.162269471151191,
+      "learning_rate": 2e-05,
+      "loss": 0.6856,
+      "step": 15796
+    },
+    {
+      "epoch": 0.3760116157288394,
+      "grad_norm": 6.991174316102775,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 15797
+    },
+    {
+      "epoch": 0.3760354184518709,
+      "grad_norm": 13.479701698388766,
+      "learning_rate": 2e-05,
+      "loss": 0.9901,
+      "step": 15798
+    },
+    {
+      "epoch": 0.3760592211749024,
+      "grad_norm": 1.8444157994989299,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 15799
+    },
+    {
+      "epoch": 0.3760830238979339,
+      "grad_norm": 2.925114897228675,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 15800
+    },
+    {
+      "epoch": 0.37610682662096545,
+      "grad_norm": 1.6871643041861653,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 15801
+    },
+    {
+      "epoch": 0.37613062934399694,
+      "grad_norm": 95.0809126740159,
+      "learning_rate": 2e-05,
+      "loss": 1.0486,
+      "step": 15802
+    },
+    {
+      "epoch": 0.3761544320670285,
+      "grad_norm": 3.035732743193099,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 15803
+    },
+    {
+      "epoch": 0.37617823479005996,
+      "grad_norm": 5.804581804517559,
+      "learning_rate": 2e-05,
+      "loss": 0.4101,
+      "step": 15804
+    },
+    {
+      "epoch": 0.3762020375130915,
+      "grad_norm": 5.296763114128248,
+      "learning_rate": 2e-05,
+      "loss": 0.5734,
+      "step": 15805
+    },
+    {
+      "epoch": 0.376225840236123,
+      "grad_norm": 5.399974830041115,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 15806
+    },
+    {
+      "epoch": 0.37624964295915453,
+      "grad_norm": 3.143537766519676,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 15807
+    },
+    {
+      "epoch": 0.37627344568218607,
+      "grad_norm": 4.677172359439733,
+      "learning_rate": 2e-05,
+      "loss": 0.7639,
+      "step": 15808
+    },
+    {
+      "epoch": 0.37629724840521755,
+      "grad_norm": 5.3133974233833285,
+      "learning_rate": 2e-05,
+      "loss": 0.7233,
+      "step": 15809
+    },
+    {
+      "epoch": 0.3763210511282491,
+      "grad_norm": 2.6694756733045564,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 15810
+    },
+    {
+      "epoch": 0.3763448538512806,
+      "grad_norm": 1.054117597343882,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 15811
+    },
+    {
+      "epoch": 0.3763686565743121,
+      "grad_norm": 0.8760273233769165,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 15812
+    },
+    {
+      "epoch": 0.3763924592973436,
+      "grad_norm": 0.8976040223661266,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 15813
+    },
+    {
+      "epoch": 0.37641626202037515,
+      "grad_norm": 4.781278216186497,
+      "learning_rate": 2e-05,
+      "loss": 0.234,
+      "step": 15814
+    },
+    {
+      "epoch": 0.37644006474340663,
+      "grad_norm": 6.831098834847852,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 15815
+    },
+    {
+      "epoch": 0.3764638674664382,
+      "grad_norm": 7.612441806940482,
+      "learning_rate": 2e-05,
+      "loss": 0.2,
+      "step": 15816
+    },
+    {
+      "epoch": 0.37648767018946966,
+      "grad_norm": 7.629823029361821,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 15817
+    },
+    {
+      "epoch": 0.3765114729125012,
+      "grad_norm": 25.088750698005406,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 15818
+    },
+    {
+      "epoch": 0.3765352756355327,
+      "grad_norm": 0.18306877747041284,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 15819
+    },
+    {
+      "epoch": 0.3765590783585642,
+      "grad_norm": 0.5829112689929928,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 15820
+    },
+    {
+      "epoch": 0.3765828810815957,
+      "grad_norm": 4.694846186220883,
+      "learning_rate": 2e-05,
+      "loss": 0.6944,
+      "step": 15821
+    },
+    {
+      "epoch": 0.37660668380462725,
+      "grad_norm": 7.414444491189711,
+      "learning_rate": 2e-05,
+      "loss": 0.7521,
+      "step": 15822
+    },
+    {
+      "epoch": 0.37663048652765874,
+      "grad_norm": 3.3780187833876747,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 15823
+    },
+    {
+      "epoch": 0.3766542892506903,
+      "grad_norm": 40.73927750227064,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 15824
+    },
+    {
+      "epoch": 0.3766780919737218,
+      "grad_norm": 18.45589935426972,
+      "learning_rate": 2e-05,
+      "loss": 0.9893,
+      "step": 15825
+    },
+    {
+      "epoch": 0.3767018946967533,
+      "grad_norm": 4.896785468124346,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 15826
+    },
+    {
+      "epoch": 0.37672569741978484,
+      "grad_norm": 5.740988415382276,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 15827
+    },
+    {
+      "epoch": 0.37674950014281633,
+      "grad_norm": 2.5780047330106624,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 15828
+    },
+    {
+      "epoch": 0.37677330286584787,
+      "grad_norm": 1.7970441748492574,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 15829
+    },
+    {
+      "epoch": 0.37679710558887936,
+      "grad_norm": 3.0087106038819926,
+      "learning_rate": 2e-05,
+      "loss": 0.3807,
+      "step": 15830
+    },
+    {
+      "epoch": 0.3768209083119109,
+      "grad_norm": 0.19213967572088916,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 15831
+    },
+    {
+      "epoch": 0.3768447110349424,
+      "grad_norm": 3.6753008611550047,
+      "learning_rate": 2e-05,
+      "loss": 0.4505,
+      "step": 15832
+    },
+    {
+      "epoch": 0.3768685137579739,
+      "grad_norm": 3.810674139602292,
+      "learning_rate": 2e-05,
+      "loss": 0.2926,
+      "step": 15833
+    },
+    {
+      "epoch": 0.3768923164810054,
+      "grad_norm": 2.930091494602156,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 15834
+    },
+    {
+      "epoch": 0.37691611920403695,
+      "grad_norm": 3.0300568583755094,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 15835
+    },
+    {
+      "epoch": 0.37693992192706843,
+      "grad_norm": 9.98027415382065,
+      "learning_rate": 2e-05,
+      "loss": 0.6741,
+      "step": 15836
+    },
+    {
+      "epoch": 0.3769637246501,
+      "grad_norm": 13.146151905553225,
+      "learning_rate": 2e-05,
+      "loss": 0.9002,
+      "step": 15837
+    },
+    {
+      "epoch": 0.37698752737313146,
+      "grad_norm": 2.549327097659356,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 15838
+    },
+    {
+      "epoch": 0.377011330096163,
+      "grad_norm": 6.542408145786596,
+      "learning_rate": 2e-05,
+      "loss": 0.8071,
+      "step": 15839
+    },
+    {
+      "epoch": 0.37703513281919454,
+      "grad_norm": 2.5633524681852022,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 15840
+    },
+    {
+      "epoch": 0.377058935542226,
+      "grad_norm": 2.5735218531061665,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 15841
+    },
+    {
+      "epoch": 0.37708273826525757,
+      "grad_norm": 9.996944930916463,
+      "learning_rate": 2e-05,
+      "loss": 0.2937,
+      "step": 15842
+    },
+    {
+      "epoch": 0.37710654098828905,
+      "grad_norm": 23.776680662808133,
+      "learning_rate": 2e-05,
+      "loss": 1.0619,
+      "step": 15843
+    },
+    {
+      "epoch": 0.3771303437113206,
+      "grad_norm": 0.25102223585966654,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 15844
+    },
+    {
+      "epoch": 0.3771541464343521,
+      "grad_norm": 4.6987679962666995,
+      "learning_rate": 2e-05,
+      "loss": 0.4273,
+      "step": 15845
+    },
+    {
+      "epoch": 0.3771779491573836,
+      "grad_norm": 5.104426300483059,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 15846
+    },
+    {
+      "epoch": 0.3772017518804151,
+      "grad_norm": 10.589329677697219,
+      "learning_rate": 2e-05,
+      "loss": 0.2882,
+      "step": 15847
+    },
+    {
+      "epoch": 0.37722555460344664,
+      "grad_norm": 5.876718071955901,
+      "learning_rate": 2e-05,
+      "loss": 0.5871,
+      "step": 15848
+    },
+    {
+      "epoch": 0.37724935732647813,
+      "grad_norm": 6.24487863962517,
+      "learning_rate": 2e-05,
+      "loss": 0.6643,
+      "step": 15849
+    },
+    {
+      "epoch": 0.37727316004950967,
+      "grad_norm": 8.442725363952967,
+      "learning_rate": 2e-05,
+      "loss": 1.0602,
+      "step": 15850
+    },
+    {
+      "epoch": 0.37729696277254116,
+      "grad_norm": 1.2293354976439623,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 15851
+    },
+    {
+      "epoch": 0.3773207654955727,
+      "grad_norm": 2.0150222966262508,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 15852
+    },
+    {
+      "epoch": 0.3773445682186042,
+      "grad_norm": 1.5655519472789956,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 15853
+    },
+    {
+      "epoch": 0.3773683709416357,
+      "grad_norm": 3.266846488497194,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 15854
+    },
+    {
+      "epoch": 0.37739217366466726,
+      "grad_norm": 11.885465328932273,
+      "learning_rate": 2e-05,
+      "loss": 1.6314,
+      "step": 15855
+    },
+    {
+      "epoch": 0.37741597638769875,
+      "grad_norm": 2.7497168190033614,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 15856
+    },
+    {
+      "epoch": 0.3774397791107303,
+      "grad_norm": 5.106688243749079,
+      "learning_rate": 2e-05,
+      "loss": 0.6396,
+      "step": 15857
+    },
+    {
+      "epoch": 0.3774635818337618,
+      "grad_norm": 6.5139069241757594,
+      "learning_rate": 2e-05,
+      "loss": 1.4236,
+      "step": 15858
+    },
+    {
+      "epoch": 0.3774873845567933,
+      "grad_norm": 5.27711336066453,
+      "learning_rate": 2e-05,
+      "loss": 0.9324,
+      "step": 15859
+    },
+    {
+      "epoch": 0.3775111872798248,
+      "grad_norm": 0.6979711061715761,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 15860
+    },
+    {
+      "epoch": 0.37753499000285634,
+      "grad_norm": 4.952026243249342,
+      "learning_rate": 2e-05,
+      "loss": 0.561,
+      "step": 15861
+    },
+    {
+      "epoch": 0.3775587927258878,
+      "grad_norm": 1.1628219597602645,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 15862
+    },
+    {
+      "epoch": 0.37758259544891937,
+      "grad_norm": 4.335644426327754,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 15863
+    },
+    {
+      "epoch": 0.37760639817195085,
+      "grad_norm": 8.256174937055523,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 15864
+    },
+    {
+      "epoch": 0.3776302008949824,
+      "grad_norm": 14.011530663721237,
+      "learning_rate": 2e-05,
+      "loss": 0.7065,
+      "step": 15865
+    },
+    {
+      "epoch": 0.3776540036180139,
+      "grad_norm": 4.45821050128401,
+      "learning_rate": 2e-05,
+      "loss": 0.7072,
+      "step": 15866
+    },
+    {
+      "epoch": 0.3776778063410454,
+      "grad_norm": 14.036653840650438,
+      "learning_rate": 2e-05,
+      "loss": 1.3113,
+      "step": 15867
+    },
+    {
+      "epoch": 0.3777016090640769,
+      "grad_norm": 12.380041137492913,
+      "learning_rate": 2e-05,
+      "loss": 1.2253,
+      "step": 15868
+    },
+    {
+      "epoch": 0.37772541178710844,
+      "grad_norm": 3.2071920744671623,
+      "learning_rate": 2e-05,
+      "loss": 0.6823,
+      "step": 15869
+    },
+    {
+      "epoch": 0.37774921451014,
+      "grad_norm": 15.505094250990448,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 15870
+    },
+    {
+      "epoch": 0.37777301723317147,
+      "grad_norm": 1.1078989027963488,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 15871
+    },
+    {
+      "epoch": 0.377796819956203,
+      "grad_norm": 3.8402637601004335,
+      "learning_rate": 2e-05,
+      "loss": 0.4176,
+      "step": 15872
+    },
+    {
+      "epoch": 0.3778206226792345,
+      "grad_norm": 5.936828690739428,
+      "learning_rate": 2e-05,
+      "loss": 1.0645,
+      "step": 15873
+    },
+    {
+      "epoch": 0.37784442540226604,
+      "grad_norm": 0.3714554770577327,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 15874
+    },
+    {
+      "epoch": 0.3778682281252975,
+      "grad_norm": 1.552557529312466,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 15875
+    },
+    {
+      "epoch": 0.37789203084832906,
+      "grad_norm": 0.8938309289361374,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 15876
+    },
+    {
+      "epoch": 0.37791583357136055,
+      "grad_norm": 18.721857943322327,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 15877
+    },
+    {
+      "epoch": 0.3779396362943921,
+      "grad_norm": 0.9588634599417717,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 15878
+    },
+    {
+      "epoch": 0.3779634390174236,
+      "grad_norm": 3.4914861587189043,
+      "learning_rate": 2e-05,
+      "loss": 0.422,
+      "step": 15879
+    },
+    {
+      "epoch": 0.3779872417404551,
+      "grad_norm": 4.433519132003354,
+      "learning_rate": 2e-05,
+      "loss": 0.6771,
+      "step": 15880
+    },
+    {
+      "epoch": 0.3780110444634866,
+      "grad_norm": 4.106713986869926,
+      "learning_rate": 2e-05,
+      "loss": 0.7643,
+      "step": 15881
+    },
+    {
+      "epoch": 0.37803484718651814,
+      "grad_norm": 4.268808415413333,
+      "learning_rate": 2e-05,
+      "loss": 0.4577,
+      "step": 15882
+    },
+    {
+      "epoch": 0.3780586499095496,
+      "grad_norm": 0.5570368568874227,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 15883
+    },
+    {
+      "epoch": 0.37808245263258117,
+      "grad_norm": 1.5508775426194141,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 15884
+    },
+    {
+      "epoch": 0.3781062553556127,
+      "grad_norm": 3.570320923433201,
+      "learning_rate": 2e-05,
+      "loss": 0.5634,
+      "step": 15885
+    },
+    {
+      "epoch": 0.3781300580786442,
+      "grad_norm": 8.291431194706124,
+      "learning_rate": 2e-05,
+      "loss": 0.792,
+      "step": 15886
+    },
+    {
+      "epoch": 0.37815386080167573,
+      "grad_norm": 3.969705229288894,
+      "learning_rate": 2e-05,
+      "loss": 0.7188,
+      "step": 15887
+    },
+    {
+      "epoch": 0.3781776635247072,
+      "grad_norm": 1.5983490338671391,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 15888
+    },
+    {
+      "epoch": 0.37820146624773876,
+      "grad_norm": 0.19768840010284042,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 15889
+    },
+    {
+      "epoch": 0.37822526897077025,
+      "grad_norm": 8.969534707292024,
+      "learning_rate": 2e-05,
+      "loss": 0.555,
+      "step": 15890
+    },
+    {
+      "epoch": 0.3782490716938018,
+      "grad_norm": 5.890361115804202,
+      "learning_rate": 2e-05,
+      "loss": 0.4063,
+      "step": 15891
+    },
+    {
+      "epoch": 0.37827287441683327,
+      "grad_norm": 0.19610421390780278,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 15892
+    },
+    {
+      "epoch": 0.3782966771398648,
+      "grad_norm": 4.743968941067606,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 15893
+    },
+    {
+      "epoch": 0.3783204798628963,
+      "grad_norm": 2.3711839617962496,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 15894
+    },
+    {
+      "epoch": 0.37834428258592784,
+      "grad_norm": 3.5943972706245453,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 15895
+    },
+    {
+      "epoch": 0.3783680853089593,
+      "grad_norm": 5.419305738842222,
+      "learning_rate": 2e-05,
+      "loss": 1.1776,
+      "step": 15896
+    },
+    {
+      "epoch": 0.37839188803199086,
+      "grad_norm": 1.633867935734764,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 15897
+    },
+    {
+      "epoch": 0.37841569075502235,
+      "grad_norm": 2.802514294640783,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 15898
+    },
+    {
+      "epoch": 0.3784394934780539,
+      "grad_norm": 3.3460943819222995,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 15899
+    },
+    {
+      "epoch": 0.37846329620108543,
+      "grad_norm": 3.1016661189803627,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 15900
+    },
+    {
+      "epoch": 0.3784870989241169,
+      "grad_norm": 4.925941966820855,
+      "learning_rate": 2e-05,
+      "loss": 0.6318,
+      "step": 15901
+    },
+    {
+      "epoch": 0.37851090164714846,
+      "grad_norm": 6.7988703778603155,
+      "learning_rate": 2e-05,
+      "loss": 1.5262,
+      "step": 15902
+    },
+    {
+      "epoch": 0.37853470437017994,
+      "grad_norm": 5.113749006270245,
+      "learning_rate": 2e-05,
+      "loss": 0.8145,
+      "step": 15903
+    },
+    {
+      "epoch": 0.3785585070932115,
+      "grad_norm": 13.465379542764973,
+      "learning_rate": 2e-05,
+      "loss": 0.5256,
+      "step": 15904
+    },
+    {
+      "epoch": 0.37858230981624297,
+      "grad_norm": 1.8179251385771988,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 15905
+    },
+    {
+      "epoch": 0.3786061125392745,
+      "grad_norm": 7.788106929957878,
+      "learning_rate": 2e-05,
+      "loss": 1.12,
+      "step": 15906
+    },
+    {
+      "epoch": 0.378629915262306,
+      "grad_norm": 2.32418172226188,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 15907
+    },
+    {
+      "epoch": 0.37865371798533753,
+      "grad_norm": 18.230931508560378,
+      "learning_rate": 2e-05,
+      "loss": 0.6585,
+      "step": 15908
+    },
+    {
+      "epoch": 0.378677520708369,
+      "grad_norm": 9.120340770993733,
+      "learning_rate": 2e-05,
+      "loss": 0.8149,
+      "step": 15909
+    },
+    {
+      "epoch": 0.37870132343140056,
+      "grad_norm": 2.2686238853880694,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 15910
+    },
+    {
+      "epoch": 0.37872512615443205,
+      "grad_norm": 0.7602460129659179,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 15911
+    },
+    {
+      "epoch": 0.3787489288774636,
+      "grad_norm": 5.106326711236027,
+      "learning_rate": 2e-05,
+      "loss": 0.1807,
+      "step": 15912
+    },
+    {
+      "epoch": 0.37877273160049507,
+      "grad_norm": 3.6170069078729448,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 15913
+    },
+    {
+      "epoch": 0.3787965343235266,
+      "grad_norm": 4.885664387966339,
+      "learning_rate": 2e-05,
+      "loss": 0.7015,
+      "step": 15914
+    },
+    {
+      "epoch": 0.37882033704655815,
+      "grad_norm": 2.8903714060027843,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 15915
+    },
+    {
+      "epoch": 0.37884413976958964,
+      "grad_norm": 8.541101987536502,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 15916
+    },
+    {
+      "epoch": 0.3788679424926212,
+      "grad_norm": 4.148233926535013,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 15917
+    },
+    {
+      "epoch": 0.37889174521565266,
+      "grad_norm": 5.938458260080512,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 15918
+    },
+    {
+      "epoch": 0.3789155479386842,
+      "grad_norm": 13.273025904308623,
+      "learning_rate": 2e-05,
+      "loss": 0.336,
+      "step": 15919
+    },
+    {
+      "epoch": 0.3789393506617157,
+      "grad_norm": 5.263421250768821,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 15920
+    },
+    {
+      "epoch": 0.37896315338474723,
+      "grad_norm": 0.8067577582711665,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 15921
+    },
+    {
+      "epoch": 0.3789869561077787,
+      "grad_norm": 3.640142954865553,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 15922
+    },
+    {
+      "epoch": 0.37901075883081026,
+      "grad_norm": 13.372417394113546,
+      "learning_rate": 2e-05,
+      "loss": 2.3777,
+      "step": 15923
+    },
+    {
+      "epoch": 0.37903456155384174,
+      "grad_norm": 2.305456707126647,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 15924
+    },
+    {
+      "epoch": 0.3790583642768733,
+      "grad_norm": 6.49177945651969,
+      "learning_rate": 2e-05,
+      "loss": 0.3596,
+      "step": 15925
+    },
+    {
+      "epoch": 0.37908216699990477,
+      "grad_norm": 1.7275513490670809,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 15926
+    },
+    {
+      "epoch": 0.3791059697229363,
+      "grad_norm": 2.38564213583695,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 15927
+    },
+    {
+      "epoch": 0.3791297724459678,
+      "grad_norm": 0.3166010652739403,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 15928
+    },
+    {
+      "epoch": 0.37915357516899933,
+      "grad_norm": 3.0057745188908025,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 15929
+    },
+    {
+      "epoch": 0.3791773778920309,
+      "grad_norm": 5.417306415882865,
+      "learning_rate": 2e-05,
+      "loss": 0.5137,
+      "step": 15930
+    },
+    {
+      "epoch": 0.37920118061506236,
+      "grad_norm": 1.91353465915987,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 15931
+    },
+    {
+      "epoch": 0.3792249833380939,
+      "grad_norm": 2.2259994101348908,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 15932
+    },
+    {
+      "epoch": 0.3792487860611254,
+      "grad_norm": 6.733947277513369,
+      "learning_rate": 2e-05,
+      "loss": 1.371,
+      "step": 15933
+    },
+    {
+      "epoch": 0.3792725887841569,
+      "grad_norm": 6.32660145477049,
+      "learning_rate": 2e-05,
+      "loss": 0.4772,
+      "step": 15934
+    },
+    {
+      "epoch": 0.3792963915071884,
+      "grad_norm": 3.671892798327327,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 15935
+    },
+    {
+      "epoch": 0.37932019423021995,
+      "grad_norm": 3.506749751300328,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 15936
+    },
+    {
+      "epoch": 0.37934399695325144,
+      "grad_norm": 4.055956854421358,
+      "learning_rate": 2e-05,
+      "loss": 0.8099,
+      "step": 15937
+    },
+    {
+      "epoch": 0.379367799676283,
+      "grad_norm": 6.439048013146953,
+      "learning_rate": 2e-05,
+      "loss": 1.6722,
+      "step": 15938
+    },
+    {
+      "epoch": 0.37939160239931446,
+      "grad_norm": 5.916297539084611,
+      "learning_rate": 2e-05,
+      "loss": 0.8969,
+      "step": 15939
+    },
+    {
+      "epoch": 0.379415405122346,
+      "grad_norm": 6.591910837022424,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 15940
+    },
+    {
+      "epoch": 0.3794392078453775,
+      "grad_norm": 0.5995187512881096,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 15941
+    },
+    {
+      "epoch": 0.37946301056840903,
+      "grad_norm": 1.275836969823489,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 15942
+    },
+    {
+      "epoch": 0.3794868132914405,
+      "grad_norm": 1.9476878501321016,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 15943
+    },
+    {
+      "epoch": 0.37951061601447206,
+      "grad_norm": 4.913059459278806,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 15944
+    },
+    {
+      "epoch": 0.37953441873750354,
+      "grad_norm": 20.155190335124036,
+      "learning_rate": 2e-05,
+      "loss": 0.4747,
+      "step": 15945
+    },
+    {
+      "epoch": 0.3795582214605351,
+      "grad_norm": 9.530196505451485,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 15946
+    },
+    {
+      "epoch": 0.3795820241835666,
+      "grad_norm": 1.8978557628340866,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 15947
+    },
+    {
+      "epoch": 0.3796058269065981,
+      "grad_norm": 18.998220138077276,
+      "learning_rate": 2e-05,
+      "loss": 0.432,
+      "step": 15948
+    },
+    {
+      "epoch": 0.37962962962962965,
+      "grad_norm": 8.389394627512837,
+      "learning_rate": 2e-05,
+      "loss": 1.8728,
+      "step": 15949
+    },
+    {
+      "epoch": 0.37965343235266114,
+      "grad_norm": 4.942564950449394,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 15950
+    },
+    {
+      "epoch": 0.3796772350756927,
+      "grad_norm": 4.300835160729571,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 15951
+    },
+    {
+      "epoch": 0.37970103779872416,
+      "grad_norm": 6.388270088402291,
+      "learning_rate": 2e-05,
+      "loss": 1.1392,
+      "step": 15952
+    },
+    {
+      "epoch": 0.3797248405217557,
+      "grad_norm": 2.1446513990712766,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 15953
+    },
+    {
+      "epoch": 0.3797486432447872,
+      "grad_norm": 6.517626590349889,
+      "learning_rate": 2e-05,
+      "loss": 0.7169,
+      "step": 15954
+    },
+    {
+      "epoch": 0.37977244596781873,
+      "grad_norm": 5.3309021351036945,
+      "learning_rate": 2e-05,
+      "loss": 0.8088,
+      "step": 15955
+    },
+    {
+      "epoch": 0.3797962486908502,
+      "grad_norm": 12.513160246905937,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 15956
+    },
+    {
+      "epoch": 0.37982005141388175,
+      "grad_norm": 2.872590729834508,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 15957
+    },
+    {
+      "epoch": 0.37984385413691324,
+      "grad_norm": 2.361191067998352,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 15958
+    },
+    {
+      "epoch": 0.3798676568599448,
+      "grad_norm": 3.156972140678395,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 15959
+    },
+    {
+      "epoch": 0.37989145958297627,
+      "grad_norm": 0.656480981420622,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 15960
+    },
+    {
+      "epoch": 0.3799152623060078,
+      "grad_norm": 2.779197459026225,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 15961
+    },
+    {
+      "epoch": 0.37993906502903935,
+      "grad_norm": 2.576900067573729,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 15962
+    },
+    {
+      "epoch": 0.37996286775207083,
+      "grad_norm": 0.11134512344797175,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 15963
+    },
+    {
+      "epoch": 0.3799866704751024,
+      "grad_norm": 0.9371375124993463,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 15964
+    },
+    {
+      "epoch": 0.38001047319813386,
+      "grad_norm": 0.6414565580225718,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 15965
+    },
+    {
+      "epoch": 0.3800342759211654,
+      "grad_norm": 3.2801541819861013,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 15966
+    },
+    {
+      "epoch": 0.3800580786441969,
+      "grad_norm": 2.4805349339621583,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 15967
+    },
+    {
+      "epoch": 0.3800818813672284,
+      "grad_norm": 5.630874760990321,
+      "learning_rate": 2e-05,
+      "loss": 0.6984,
+      "step": 15968
+    },
+    {
+      "epoch": 0.3801056840902599,
+      "grad_norm": 7.144205203921121,
+      "learning_rate": 2e-05,
+      "loss": 0.4283,
+      "step": 15969
+    },
+    {
+      "epoch": 0.38012948681329145,
+      "grad_norm": 9.558438226390157,
+      "learning_rate": 2e-05,
+      "loss": 0.5695,
+      "step": 15970
+    },
+    {
+      "epoch": 0.38015328953632294,
+      "grad_norm": 2.5982360727277305,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 15971
+    },
+    {
+      "epoch": 0.3801770922593545,
+      "grad_norm": 9.73634815239532,
+      "learning_rate": 2e-05,
+      "loss": 0.2011,
+      "step": 15972
+    },
+    {
+      "epoch": 0.38020089498238596,
+      "grad_norm": 5.068396403538522,
+      "learning_rate": 2e-05,
+      "loss": 0.657,
+      "step": 15973
+    },
+    {
+      "epoch": 0.3802246977054175,
+      "grad_norm": 2.9685691831828778,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 15974
+    },
+    {
+      "epoch": 0.380248500428449,
+      "grad_norm": 2.237051949289196,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 15975
+    },
+    {
+      "epoch": 0.38027230315148053,
+      "grad_norm": 2.918285411621352,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 15976
+    },
+    {
+      "epoch": 0.38029610587451207,
+      "grad_norm": 4.198929997152061,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 15977
+    },
+    {
+      "epoch": 0.38031990859754355,
+      "grad_norm": 2.918936680663165,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 15978
+    },
+    {
+      "epoch": 0.3803437113205751,
+      "grad_norm": 0.8319578114932503,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 15979
+    },
+    {
+      "epoch": 0.3803675140436066,
+      "grad_norm": 2.947498372549278,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 15980
+    },
+    {
+      "epoch": 0.3803913167666381,
+      "grad_norm": 5.490970001798544,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 15981
+    },
+    {
+      "epoch": 0.3804151194896696,
+      "grad_norm": 2.871434636639242,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 15982
+    },
+    {
+      "epoch": 0.38043892221270115,
+      "grad_norm": 3.0967262753027884,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 15983
+    },
+    {
+      "epoch": 0.38046272493573263,
+      "grad_norm": 12.833162229067428,
+      "learning_rate": 2e-05,
+      "loss": 1.288,
+      "step": 15984
+    },
+    {
+      "epoch": 0.3804865276587642,
+      "grad_norm": 25.441396716944944,
+      "learning_rate": 2e-05,
+      "loss": 1.0575,
+      "step": 15985
+    },
+    {
+      "epoch": 0.38051033038179566,
+      "grad_norm": 3.1797691638358927,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 15986
+    },
+    {
+      "epoch": 0.3805341331048272,
+      "grad_norm": 5.505573379289126,
+      "learning_rate": 2e-05,
+      "loss": 0.6542,
+      "step": 15987
+    },
+    {
+      "epoch": 0.3805579358278587,
+      "grad_norm": 3.711946293301275,
+      "learning_rate": 2e-05,
+      "loss": 0.1493,
+      "step": 15988
+    },
+    {
+      "epoch": 0.3805817385508902,
+      "grad_norm": 2.787329267316697,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 15989
+    },
+    {
+      "epoch": 0.3806055412739217,
+      "grad_norm": 3.698380911739527,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 15990
+    },
+    {
+      "epoch": 0.38062934399695325,
+      "grad_norm": 5.517457245994056,
+      "learning_rate": 2e-05,
+      "loss": 0.6214,
+      "step": 15991
+    },
+    {
+      "epoch": 0.3806531467199848,
+      "grad_norm": 14.834735851431804,
+      "learning_rate": 2e-05,
+      "loss": 1.4141,
+      "step": 15992
+    },
+    {
+      "epoch": 0.3806769494430163,
+      "grad_norm": 8.646518215556489,
+      "learning_rate": 2e-05,
+      "loss": 0.3159,
+      "step": 15993
+    },
+    {
+      "epoch": 0.3807007521660478,
+      "grad_norm": 3.7885857450557565,
+      "learning_rate": 2e-05,
+      "loss": 0.7472,
+      "step": 15994
+    },
+    {
+      "epoch": 0.3807245548890793,
+      "grad_norm": 10.361668168895395,
+      "learning_rate": 2e-05,
+      "loss": 1.1821,
+      "step": 15995
+    },
+    {
+      "epoch": 0.38074835761211084,
+      "grad_norm": 2.3292343768631976,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 15996
+    },
+    {
+      "epoch": 0.38077216033514233,
+      "grad_norm": 2.709872988443246,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 15997
+    },
+    {
+      "epoch": 0.38079596305817387,
+      "grad_norm": 45.35597565935531,
+      "learning_rate": 2e-05,
+      "loss": 1.3743,
+      "step": 15998
+    },
+    {
+      "epoch": 0.38081976578120535,
+      "grad_norm": 4.425977009225767,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 15999
+    },
+    {
+      "epoch": 0.3808435685042369,
+      "grad_norm": 6.136587226923672,
+      "learning_rate": 2e-05,
+      "loss": 1.5171,
+      "step": 16000
+    },
+    {
+      "epoch": 0.3808673712272684,
+      "grad_norm": 1.852626392376793,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 16001
+    },
+    {
+      "epoch": 0.3808911739502999,
+      "grad_norm": 0.0663142744499307,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 16002
+    },
+    {
+      "epoch": 0.3809149766733314,
+      "grad_norm": 6.639861213416351,
+      "learning_rate": 2e-05,
+      "loss": 0.1896,
+      "step": 16003
+    },
+    {
+      "epoch": 0.38093877939636295,
+      "grad_norm": 5.832622961827301,
+      "learning_rate": 2e-05,
+      "loss": 0.5944,
+      "step": 16004
+    },
+    {
+      "epoch": 0.38096258211939443,
+      "grad_norm": 4.070767431884015,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 16005
+    },
+    {
+      "epoch": 0.380986384842426,
+      "grad_norm": 3.567579905088526,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 16006
+    },
+    {
+      "epoch": 0.3810101875654575,
+      "grad_norm": 2.762430742270109,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 16007
+    },
+    {
+      "epoch": 0.381033990288489,
+      "grad_norm": 4.159053131182468,
+      "learning_rate": 2e-05,
+      "loss": 0.7567,
+      "step": 16008
+    },
+    {
+      "epoch": 0.38105779301152054,
+      "grad_norm": 0.10997784946708898,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16009
+    },
+    {
+      "epoch": 0.381081595734552,
+      "grad_norm": 5.038618189500455,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 16010
+    },
+    {
+      "epoch": 0.38110539845758357,
+      "grad_norm": 10.272952076380834,
+      "learning_rate": 2e-05,
+      "loss": 0.4095,
+      "step": 16011
+    },
+    {
+      "epoch": 0.38112920118061505,
+      "grad_norm": 3.1744450413208027,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 16012
+    },
+    {
+      "epoch": 0.3811530039036466,
+      "grad_norm": 2.123866304083339,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 16013
+    },
+    {
+      "epoch": 0.3811768066266781,
+      "grad_norm": 8.447505575133915,
+      "learning_rate": 2e-05,
+      "loss": 1.374,
+      "step": 16014
+    },
+    {
+      "epoch": 0.3812006093497096,
+      "grad_norm": 0.13678802473622867,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 16015
+    },
+    {
+      "epoch": 0.3812244120727411,
+      "grad_norm": 5.64814227342207,
+      "learning_rate": 2e-05,
+      "loss": 1.3183,
+      "step": 16016
+    },
+    {
+      "epoch": 0.38124821479577264,
+      "grad_norm": 1.0663927104562667,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 16017
+    },
+    {
+      "epoch": 0.38127201751880413,
+      "grad_norm": 11.575504206013717,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 16018
+    },
+    {
+      "epoch": 0.38129582024183567,
+      "grad_norm": 2.416269100865441,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 16019
+    },
+    {
+      "epoch": 0.38131962296486716,
+      "grad_norm": 3.5560597148569197,
+      "learning_rate": 2e-05,
+      "loss": 0.6107,
+      "step": 16020
+    },
+    {
+      "epoch": 0.3813434256878987,
+      "grad_norm": 10.265282688570505,
+      "learning_rate": 2e-05,
+      "loss": 0.1805,
+      "step": 16021
+    },
+    {
+      "epoch": 0.38136722841093024,
+      "grad_norm": 6.0628091201535685,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 16022
+    },
+    {
+      "epoch": 0.3813910311339617,
+      "grad_norm": 6.577945090428776,
+      "learning_rate": 2e-05,
+      "loss": 0.2712,
+      "step": 16023
+    },
+    {
+      "epoch": 0.38141483385699326,
+      "grad_norm": 3.6784596203195767,
+      "learning_rate": 2e-05,
+      "loss": 0.6472,
+      "step": 16024
+    },
+    {
+      "epoch": 0.38143863658002475,
+      "grad_norm": 3.0230499829171684,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 16025
+    },
+    {
+      "epoch": 0.3814624393030563,
+      "grad_norm": 6.757607660084347,
+      "learning_rate": 2e-05,
+      "loss": 0.9514,
+      "step": 16026
+    },
+    {
+      "epoch": 0.3814862420260878,
+      "grad_norm": 7.205333868513976,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 16027
+    },
+    {
+      "epoch": 0.3815100447491193,
+      "grad_norm": 0.5222816470231422,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 16028
+    },
+    {
+      "epoch": 0.3815338474721508,
+      "grad_norm": 2.8134184518968945,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 16029
+    },
+    {
+      "epoch": 0.38155765019518234,
+      "grad_norm": 1.0723299908206778,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 16030
+    },
+    {
+      "epoch": 0.3815814529182138,
+      "grad_norm": 0.110862811141058,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16031
+    },
+    {
+      "epoch": 0.38160525564124537,
+      "grad_norm": 4.431002366073542,
+      "learning_rate": 2e-05,
+      "loss": 0.8165,
+      "step": 16032
+    },
+    {
+      "epoch": 0.38162905836427685,
+      "grad_norm": 1.9070792523891062,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 16033
+    },
+    {
+      "epoch": 0.3816528610873084,
+      "grad_norm": 6.396761104404623,
+      "learning_rate": 2e-05,
+      "loss": 0.3088,
+      "step": 16034
+    },
+    {
+      "epoch": 0.3816766638103399,
+      "grad_norm": 4.251612585381947,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 16035
+    },
+    {
+      "epoch": 0.3817004665333714,
+      "grad_norm": 1.8934266386777385,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 16036
+    },
+    {
+      "epoch": 0.38172426925640296,
+      "grad_norm": 2.1104853273958706,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 16037
+    },
+    {
+      "epoch": 0.38174807197943444,
+      "grad_norm": 2.696567226098452,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 16038
+    },
+    {
+      "epoch": 0.381771874702466,
+      "grad_norm": 43.735146018106875,
+      "learning_rate": 2e-05,
+      "loss": 0.6216,
+      "step": 16039
+    },
+    {
+      "epoch": 0.38179567742549747,
+      "grad_norm": 5.183195916383831,
+      "learning_rate": 2e-05,
+      "loss": 0.6454,
+      "step": 16040
+    },
+    {
+      "epoch": 0.381819480148529,
+      "grad_norm": 6.334254848523052,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 16041
+    },
+    {
+      "epoch": 0.3818432828715605,
+      "grad_norm": 9.355055758453046,
+      "learning_rate": 2e-05,
+      "loss": 0.5292,
+      "step": 16042
+    },
+    {
+      "epoch": 0.38186708559459204,
+      "grad_norm": 1.6286838028370085,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 16043
+    },
+    {
+      "epoch": 0.3818908883176235,
+      "grad_norm": 5.0438247667177425,
+      "learning_rate": 2e-05,
+      "loss": 0.9298,
+      "step": 16044
+    },
+    {
+      "epoch": 0.38191469104065506,
+      "grad_norm": 6.723281779363867,
+      "learning_rate": 2e-05,
+      "loss": 1.3148,
+      "step": 16045
+    },
+    {
+      "epoch": 0.38193849376368655,
+      "grad_norm": 10.909331474694039,
+      "learning_rate": 2e-05,
+      "loss": 2.0634,
+      "step": 16046
+    },
+    {
+      "epoch": 0.3819622964867181,
+      "grad_norm": 0.9363226713315899,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 16047
+    },
+    {
+      "epoch": 0.3819860992097496,
+      "grad_norm": 4.007700249537216,
+      "learning_rate": 2e-05,
+      "loss": 0.8462,
+      "step": 16048
+    },
+    {
+      "epoch": 0.3820099019327811,
+      "grad_norm": 0.712198149690824,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16049
+    },
+    {
+      "epoch": 0.3820337046558126,
+      "grad_norm": 3.676696084643511,
+      "learning_rate": 2e-05,
+      "loss": 0.387,
+      "step": 16050
+    },
+    {
+      "epoch": 0.38205750737884414,
+      "grad_norm": 0.44690084474086333,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 16051
+    },
+    {
+      "epoch": 0.3820813101018757,
+      "grad_norm": 4.112226538507056,
+      "learning_rate": 2e-05,
+      "loss": 0.6289,
+      "step": 16052
+    },
+    {
+      "epoch": 0.38210511282490717,
+      "grad_norm": 5.385283709998699,
+      "learning_rate": 2e-05,
+      "loss": 0.6328,
+      "step": 16053
+    },
+    {
+      "epoch": 0.3821289155479387,
+      "grad_norm": 1.776287583242177,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 16054
+    },
+    {
+      "epoch": 0.3821527182709702,
+      "grad_norm": 3.1387588899414696,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 16055
+    },
+    {
+      "epoch": 0.38217652099400173,
+      "grad_norm": 0.5308062289273726,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 16056
+    },
+    {
+      "epoch": 0.3822003237170332,
+      "grad_norm": 3.8605106649445955,
+      "learning_rate": 2e-05,
+      "loss": 0.515,
+      "step": 16057
+    },
+    {
+      "epoch": 0.38222412644006476,
+      "grad_norm": 1.681982266921603,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 16058
+    },
+    {
+      "epoch": 0.38224792916309625,
+      "grad_norm": 10.522686046754421,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 16059
+    },
+    {
+      "epoch": 0.3822717318861278,
+      "grad_norm": 8.62037899633295,
+      "learning_rate": 2e-05,
+      "loss": 0.728,
+      "step": 16060
+    },
+    {
+      "epoch": 0.38229553460915927,
+      "grad_norm": 9.666490255055276,
+      "learning_rate": 2e-05,
+      "loss": 1.7895,
+      "step": 16061
+    },
+    {
+      "epoch": 0.3823193373321908,
+      "grad_norm": 4.726243657398606,
+      "learning_rate": 2e-05,
+      "loss": 0.7319,
+      "step": 16062
+    },
+    {
+      "epoch": 0.3823431400552223,
+      "grad_norm": 4.772359079585628,
+      "learning_rate": 2e-05,
+      "loss": 0.707,
+      "step": 16063
+    },
+    {
+      "epoch": 0.38236694277825384,
+      "grad_norm": 5.980139017807156,
+      "learning_rate": 2e-05,
+      "loss": 0.3478,
+      "step": 16064
+    },
+    {
+      "epoch": 0.3823907455012853,
+      "grad_norm": 12.29406026283921,
+      "learning_rate": 2e-05,
+      "loss": 0.3995,
+      "step": 16065
+    },
+    {
+      "epoch": 0.38241454822431686,
+      "grad_norm": 6.684693202386268,
+      "learning_rate": 2e-05,
+      "loss": 1.3196,
+      "step": 16066
+    },
+    {
+      "epoch": 0.3824383509473484,
+      "grad_norm": 2.8999321454191307,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 16067
+    },
+    {
+      "epoch": 0.3824621536703799,
+      "grad_norm": 3.556580860290715,
+      "learning_rate": 2e-05,
+      "loss": 0.5223,
+      "step": 16068
+    },
+    {
+      "epoch": 0.38248595639341143,
+      "grad_norm": 2.757124188882118,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 16069
+    },
+    {
+      "epoch": 0.3825097591164429,
+      "grad_norm": 9.59578248335011,
+      "learning_rate": 2e-05,
+      "loss": 1.51,
+      "step": 16070
+    },
+    {
+      "epoch": 0.38253356183947446,
+      "grad_norm": 7.340878448509151,
+      "learning_rate": 2e-05,
+      "loss": 0.2442,
+      "step": 16071
+    },
+    {
+      "epoch": 0.38255736456250594,
+      "grad_norm": 0.6807407766679252,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 16072
+    },
+    {
+      "epoch": 0.3825811672855375,
+      "grad_norm": 1.9685524126108314,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 16073
+    },
+    {
+      "epoch": 0.38260497000856897,
+      "grad_norm": 2.2144676737328446,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 16074
+    },
+    {
+      "epoch": 0.3826287727316005,
+      "grad_norm": 1.9086748054390406,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 16075
+    },
+    {
+      "epoch": 0.382652575454632,
+      "grad_norm": 5.570715259703198,
+      "learning_rate": 2e-05,
+      "loss": 1.1668,
+      "step": 16076
+    },
+    {
+      "epoch": 0.38267637817766353,
+      "grad_norm": 15.510759792166306,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 16077
+    },
+    {
+      "epoch": 0.382700180900695,
+      "grad_norm": 4.5240774730512845,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 16078
+    },
+    {
+      "epoch": 0.38272398362372656,
+      "grad_norm": 7.056823740831832,
+      "learning_rate": 2e-05,
+      "loss": 1.4202,
+      "step": 16079
+    },
+    {
+      "epoch": 0.38274778634675805,
+      "grad_norm": 4.029026806265255,
+      "learning_rate": 2e-05,
+      "loss": 0.7713,
+      "step": 16080
+    },
+    {
+      "epoch": 0.3827715890697896,
+      "grad_norm": 0.5750115192168499,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 16081
+    },
+    {
+      "epoch": 0.38279539179282107,
+      "grad_norm": 5.056310229304097,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 16082
+    },
+    {
+      "epoch": 0.3828191945158526,
+      "grad_norm": 11.814290400541033,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 16083
+    },
+    {
+      "epoch": 0.38284299723888415,
+      "grad_norm": 3.948278163161421,
+      "learning_rate": 2e-05,
+      "loss": 0.4963,
+      "step": 16084
+    },
+    {
+      "epoch": 0.38286679996191564,
+      "grad_norm": 4.439771659218748,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 16085
+    },
+    {
+      "epoch": 0.3828906026849472,
+      "grad_norm": 1.7766915303852104,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 16086
+    },
+    {
+      "epoch": 0.38291440540797866,
+      "grad_norm": 1.8018345026347562,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 16087
+    },
+    {
+      "epoch": 0.3829382081310102,
+      "grad_norm": 3.6796754672982384,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 16088
+    },
+    {
+      "epoch": 0.3829620108540417,
+      "grad_norm": 9.784092640636565,
+      "learning_rate": 2e-05,
+      "loss": 0.9291,
+      "step": 16089
+    },
+    {
+      "epoch": 0.38298581357707323,
+      "grad_norm": 2.0543147299932265,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 16090
+    },
+    {
+      "epoch": 0.3830096163001047,
+      "grad_norm": 4.233257618072522,
+      "learning_rate": 2e-05,
+      "loss": 0.383,
+      "step": 16091
+    },
+    {
+      "epoch": 0.38303341902313626,
+      "grad_norm": 1.521380367692656,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 16092
+    },
+    {
+      "epoch": 0.38305722174616774,
+      "grad_norm": 10.292474523820632,
+      "learning_rate": 2e-05,
+      "loss": 1.1412,
+      "step": 16093
+    },
+    {
+      "epoch": 0.3830810244691993,
+      "grad_norm": 5.073124284090451,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 16094
+    },
+    {
+      "epoch": 0.38310482719223077,
+      "grad_norm": 0.37015194236521465,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 16095
+    },
+    {
+      "epoch": 0.3831286299152623,
+      "grad_norm": 9.01241542229771,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 16096
+    },
+    {
+      "epoch": 0.3831524326382938,
+      "grad_norm": 9.246261276347425,
+      "learning_rate": 2e-05,
+      "loss": 0.2887,
+      "step": 16097
+    },
+    {
+      "epoch": 0.38317623536132533,
+      "grad_norm": 0.2663018347312242,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 16098
+    },
+    {
+      "epoch": 0.3832000380843569,
+      "grad_norm": 4.100303167187821,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 16099
+    },
+    {
+      "epoch": 0.38322384080738836,
+      "grad_norm": 3.3966747700328805,
+      "learning_rate": 2e-05,
+      "loss": 0.5028,
+      "step": 16100
+    },
+    {
+      "epoch": 0.3832476435304199,
+      "grad_norm": 49.843828198830614,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 16101
+    },
+    {
+      "epoch": 0.3832714462534514,
+      "grad_norm": 6.082195030905312,
+      "learning_rate": 2e-05,
+      "loss": 0.7433,
+      "step": 16102
+    },
+    {
+      "epoch": 0.3832952489764829,
+      "grad_norm": 9.203393524108865,
+      "learning_rate": 2e-05,
+      "loss": 1.1281,
+      "step": 16103
+    },
+    {
+      "epoch": 0.3833190516995144,
+      "grad_norm": 7.637803970290123,
+      "learning_rate": 2e-05,
+      "loss": 1.5857,
+      "step": 16104
+    },
+    {
+      "epoch": 0.38334285442254595,
+      "grad_norm": 6.799364062761336,
+      "learning_rate": 2e-05,
+      "loss": 0.2925,
+      "step": 16105
+    },
+    {
+      "epoch": 0.38336665714557744,
+      "grad_norm": 5.319257844946404,
+      "learning_rate": 2e-05,
+      "loss": 0.2281,
+      "step": 16106
+    },
+    {
+      "epoch": 0.383390459868609,
+      "grad_norm": 5.981032720765452,
+      "learning_rate": 2e-05,
+      "loss": 1.7089,
+      "step": 16107
+    },
+    {
+      "epoch": 0.38341426259164046,
+      "grad_norm": 3.8730018550778205,
+      "learning_rate": 2e-05,
+      "loss": 0.5105,
+      "step": 16108
+    },
+    {
+      "epoch": 0.383438065314672,
+      "grad_norm": 0.8634272747526769,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 16109
+    },
+    {
+      "epoch": 0.3834618680377035,
+      "grad_norm": 1.9270562883009767,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 16110
+    },
+    {
+      "epoch": 0.38348567076073503,
+      "grad_norm": 5.648927820310501,
+      "learning_rate": 2e-05,
+      "loss": 0.1926,
+      "step": 16111
+    },
+    {
+      "epoch": 0.3835094734837665,
+      "grad_norm": 3.8563299107053877,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 16112
+    },
+    {
+      "epoch": 0.38353327620679806,
+      "grad_norm": 5.660438020267089,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 16113
+    },
+    {
+      "epoch": 0.3835570789298296,
+      "grad_norm": 3.3510012846618142,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 16114
+    },
+    {
+      "epoch": 0.3835808816528611,
+      "grad_norm": 3.79484646333946,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 16115
+    },
+    {
+      "epoch": 0.3836046843758926,
+      "grad_norm": 0.09032891756418714,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 16116
+    },
+    {
+      "epoch": 0.3836284870989241,
+      "grad_norm": 0.22860215791809266,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 16117
+    },
+    {
+      "epoch": 0.38365228982195565,
+      "grad_norm": 2.9650214058688444,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 16118
+    },
+    {
+      "epoch": 0.38367609254498714,
+      "grad_norm": 1.1352020036536303,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 16119
+    },
+    {
+      "epoch": 0.3836998952680187,
+      "grad_norm": 1.1157615021549667,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 16120
+    },
+    {
+      "epoch": 0.38372369799105016,
+      "grad_norm": 2.1769928757966106,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 16121
+    },
+    {
+      "epoch": 0.3837475007140817,
+      "grad_norm": 1.2682890787134218,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 16122
+    },
+    {
+      "epoch": 0.3837713034371132,
+      "grad_norm": 5.583414111415605,
+      "learning_rate": 2e-05,
+      "loss": 1.3399,
+      "step": 16123
+    },
+    {
+      "epoch": 0.38379510616014473,
+      "grad_norm": 5.8166265027979,
+      "learning_rate": 2e-05,
+      "loss": 0.6476,
+      "step": 16124
+    },
+    {
+      "epoch": 0.3838189088831762,
+      "grad_norm": 9.566236369747505,
+      "learning_rate": 2e-05,
+      "loss": 0.4412,
+      "step": 16125
+    },
+    {
+      "epoch": 0.38384271160620775,
+      "grad_norm": 2.8591180757802928,
+      "learning_rate": 2e-05,
+      "loss": 0.3829,
+      "step": 16126
+    },
+    {
+      "epoch": 0.38386651432923924,
+      "grad_norm": 3.3233777180209962,
+      "learning_rate": 2e-05,
+      "loss": 0.3119,
+      "step": 16127
+    },
+    {
+      "epoch": 0.3838903170522708,
+      "grad_norm": 1.3774906991722065,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 16128
+    },
+    {
+      "epoch": 0.3839141197753023,
+      "grad_norm": 3.4706995379581858,
+      "learning_rate": 2e-05,
+      "loss": 0.5496,
+      "step": 16129
+    },
+    {
+      "epoch": 0.3839379224983338,
+      "grad_norm": 0.8142360449517536,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 16130
+    },
+    {
+      "epoch": 0.38396172522136535,
+      "grad_norm": 4.267094933328414,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 16131
+    },
+    {
+      "epoch": 0.38398552794439683,
+      "grad_norm": 6.871782641592561,
+      "learning_rate": 2e-05,
+      "loss": 0.6036,
+      "step": 16132
+    },
+    {
+      "epoch": 0.3840093306674284,
+      "grad_norm": 3.9817787481846576,
+      "learning_rate": 2e-05,
+      "loss": 0.8792,
+      "step": 16133
+    },
+    {
+      "epoch": 0.38403313339045986,
+      "grad_norm": 3.512582711964188,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 16134
+    },
+    {
+      "epoch": 0.3840569361134914,
+      "grad_norm": 3.268495395759772,
+      "learning_rate": 2e-05,
+      "loss": 0.2781,
+      "step": 16135
+    },
+    {
+      "epoch": 0.3840807388365229,
+      "grad_norm": 1.890738180557599,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 16136
+    },
+    {
+      "epoch": 0.3841045415595544,
+      "grad_norm": 15.393183747079526,
+      "learning_rate": 2e-05,
+      "loss": 1.1779,
+      "step": 16137
+    },
+    {
+      "epoch": 0.3841283442825859,
+      "grad_norm": 0.5420794421936515,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 16138
+    },
+    {
+      "epoch": 0.38415214700561745,
+      "grad_norm": 5.838805197024001,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 16139
+    },
+    {
+      "epoch": 0.38417594972864894,
+      "grad_norm": 2.1666213625742095,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 16140
+    },
+    {
+      "epoch": 0.3841997524516805,
+      "grad_norm": 4.064341884320195,
+      "learning_rate": 2e-05,
+      "loss": 0.7486,
+      "step": 16141
+    },
+    {
+      "epoch": 0.38422355517471196,
+      "grad_norm": 4.436480510762213,
+      "learning_rate": 2e-05,
+      "loss": 0.3467,
+      "step": 16142
+    },
+    {
+      "epoch": 0.3842473578977435,
+      "grad_norm": 2.144921190606874,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 16143
+    },
+    {
+      "epoch": 0.38427116062077504,
+      "grad_norm": 1.6838927446952472,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 16144
+    },
+    {
+      "epoch": 0.38429496334380653,
+      "grad_norm": 5.806434882320456,
+      "learning_rate": 2e-05,
+      "loss": 0.4132,
+      "step": 16145
+    },
+    {
+      "epoch": 0.38431876606683807,
+      "grad_norm": 5.384111513343748,
+      "learning_rate": 2e-05,
+      "loss": 1.0112,
+      "step": 16146
+    },
+    {
+      "epoch": 0.38434256878986955,
+      "grad_norm": 4.2539985759248,
+      "learning_rate": 2e-05,
+      "loss": 0.2669,
+      "step": 16147
+    },
+    {
+      "epoch": 0.3843663715129011,
+      "grad_norm": 3.6279748452041214,
+      "learning_rate": 2e-05,
+      "loss": 0.543,
+      "step": 16148
+    },
+    {
+      "epoch": 0.3843901742359326,
+      "grad_norm": 5.819496722212192,
+      "learning_rate": 2e-05,
+      "loss": 0.7831,
+      "step": 16149
+    },
+    {
+      "epoch": 0.3844139769589641,
+      "grad_norm": 6.267824105960504,
+      "learning_rate": 2e-05,
+      "loss": 0.3551,
+      "step": 16150
+    },
+    {
+      "epoch": 0.3844377796819956,
+      "grad_norm": 5.142675658561803,
+      "learning_rate": 2e-05,
+      "loss": 0.6499,
+      "step": 16151
+    },
+    {
+      "epoch": 0.38446158240502715,
+      "grad_norm": 12.957321619745136,
+      "learning_rate": 2e-05,
+      "loss": 1.7619,
+      "step": 16152
+    },
+    {
+      "epoch": 0.38448538512805863,
+      "grad_norm": 8.85886964010718,
+      "learning_rate": 2e-05,
+      "loss": 0.286,
+      "step": 16153
+    },
+    {
+      "epoch": 0.3845091878510902,
+      "grad_norm": 2.5849521479392674,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 16154
+    },
+    {
+      "epoch": 0.38453299057412166,
+      "grad_norm": 9.555996339856954,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 16155
+    },
+    {
+      "epoch": 0.3845567932971532,
+      "grad_norm": 7.400538108724387,
+      "learning_rate": 2e-05,
+      "loss": 0.7862,
+      "step": 16156
+    },
+    {
+      "epoch": 0.3845805960201847,
+      "grad_norm": 6.6960504347684955,
+      "learning_rate": 2e-05,
+      "loss": 0.6324,
+      "step": 16157
+    },
+    {
+      "epoch": 0.3846043987432162,
+      "grad_norm": 1.7796064251359809,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 16158
+    },
+    {
+      "epoch": 0.38462820146624777,
+      "grad_norm": 1.599142501403587,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 16159
+    },
+    {
+      "epoch": 0.38465200418927925,
+      "grad_norm": 7.027304425915379,
+      "learning_rate": 2e-05,
+      "loss": 1.6223,
+      "step": 16160
+    },
+    {
+      "epoch": 0.3846758069123108,
+      "grad_norm": 3.693087922334858,
+      "learning_rate": 2e-05,
+      "loss": 0.8089,
+      "step": 16161
+    },
+    {
+      "epoch": 0.3846996096353423,
+      "grad_norm": 2.4644316444136587,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 16162
+    },
+    {
+      "epoch": 0.3847234123583738,
+      "grad_norm": 3.7733052222977492,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 16163
+    },
+    {
+      "epoch": 0.3847472150814053,
+      "grad_norm": 4.458835274677928,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 16164
+    },
+    {
+      "epoch": 0.38477101780443684,
+      "grad_norm": 34.24463165960193,
+      "learning_rate": 2e-05,
+      "loss": 0.7339,
+      "step": 16165
+    },
+    {
+      "epoch": 0.38479482052746833,
+      "grad_norm": 7.208893490360123,
+      "learning_rate": 2e-05,
+      "loss": 1.4279,
+      "step": 16166
+    },
+    {
+      "epoch": 0.38481862325049987,
+      "grad_norm": 2.5055870872774726,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 16167
+    },
+    {
+      "epoch": 0.38484242597353135,
+      "grad_norm": 1.0984228921453474,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 16168
+    },
+    {
+      "epoch": 0.3848662286965629,
+      "grad_norm": 1.5579150858552078,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 16169
+    },
+    {
+      "epoch": 0.3848900314195944,
+      "grad_norm": 7.561829967671142,
+      "learning_rate": 2e-05,
+      "loss": 1.1491,
+      "step": 16170
+    },
+    {
+      "epoch": 0.3849138341426259,
+      "grad_norm": 0.48403717795758955,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 16171
+    },
+    {
+      "epoch": 0.3849376368656574,
+      "grad_norm": 1.052092358572362,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 16172
+    },
+    {
+      "epoch": 0.38496143958868895,
+      "grad_norm": 8.540726132974838,
+      "learning_rate": 2e-05,
+      "loss": 0.8142,
+      "step": 16173
+    },
+    {
+      "epoch": 0.3849852423117205,
+      "grad_norm": 0.8566783718515485,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 16174
+    },
+    {
+      "epoch": 0.385009045034752,
+      "grad_norm": 7.93761033883513,
+      "learning_rate": 2e-05,
+      "loss": 0.8534,
+      "step": 16175
+    },
+    {
+      "epoch": 0.3850328477577835,
+      "grad_norm": 1.4249254482043978,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 16176
+    },
+    {
+      "epoch": 0.385056650480815,
+      "grad_norm": 12.934109304625002,
+      "learning_rate": 2e-05,
+      "loss": 0.8173,
+      "step": 16177
+    },
+    {
+      "epoch": 0.38508045320384654,
+      "grad_norm": 9.091893131139097,
+      "learning_rate": 2e-05,
+      "loss": 1.1402,
+      "step": 16178
+    },
+    {
+      "epoch": 0.385104255926878,
+      "grad_norm": 8.586988383331809,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 16179
+    },
+    {
+      "epoch": 0.38512805864990957,
+      "grad_norm": 7.9608118723984695,
+      "learning_rate": 2e-05,
+      "loss": 0.9811,
+      "step": 16180
+    },
+    {
+      "epoch": 0.38515186137294105,
+      "grad_norm": 4.056160816157019,
+      "learning_rate": 2e-05,
+      "loss": 0.6478,
+      "step": 16181
+    },
+    {
+      "epoch": 0.3851756640959726,
+      "grad_norm": 3.1166665929472956,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 16182
+    },
+    {
+      "epoch": 0.3851994668190041,
+      "grad_norm": 6.349213935378418,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 16183
+    },
+    {
+      "epoch": 0.3852232695420356,
+      "grad_norm": 4.588480831634619,
+      "learning_rate": 2e-05,
+      "loss": 0.724,
+      "step": 16184
+    },
+    {
+      "epoch": 0.3852470722650671,
+      "grad_norm": 1.8571285355227924,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 16185
+    },
+    {
+      "epoch": 0.38527087498809864,
+      "grad_norm": 1.0162934469415057,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 16186
+    },
+    {
+      "epoch": 0.38529467771113013,
+      "grad_norm": 4.134117166657644,
+      "learning_rate": 2e-05,
+      "loss": 0.495,
+      "step": 16187
+    },
+    {
+      "epoch": 0.38531848043416167,
+      "grad_norm": 2.927551040056671,
+      "learning_rate": 2e-05,
+      "loss": 0.1603,
+      "step": 16188
+    },
+    {
+      "epoch": 0.3853422831571932,
+      "grad_norm": 0.22863350515465827,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 16189
+    },
+    {
+      "epoch": 0.3853660858802247,
+      "grad_norm": 7.193606788535965,
+      "learning_rate": 2e-05,
+      "loss": 1.476,
+      "step": 16190
+    },
+    {
+      "epoch": 0.38538988860325624,
+      "grad_norm": 3.8139700036978272,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 16191
+    },
+    {
+      "epoch": 0.3854136913262877,
+      "grad_norm": 4.5475560935936015,
+      "learning_rate": 2e-05,
+      "loss": 0.7684,
+      "step": 16192
+    },
+    {
+      "epoch": 0.38543749404931926,
+      "grad_norm": 25.538462364082356,
+      "learning_rate": 2e-05,
+      "loss": 1.2388,
+      "step": 16193
+    },
+    {
+      "epoch": 0.38546129677235075,
+      "grad_norm": 1.980770806070341,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 16194
+    },
+    {
+      "epoch": 0.3854850994953823,
+      "grad_norm": 1.2921236476967586,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 16195
+    },
+    {
+      "epoch": 0.3855089022184138,
+      "grad_norm": 6.064571928531889,
+      "learning_rate": 2e-05,
+      "loss": 0.2146,
+      "step": 16196
+    },
+    {
+      "epoch": 0.3855327049414453,
+      "grad_norm": 4.040229812015673,
+      "learning_rate": 2e-05,
+      "loss": 0.6235,
+      "step": 16197
+    },
+    {
+      "epoch": 0.3855565076644768,
+      "grad_norm": 2.087841042139465,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 16198
+    },
+    {
+      "epoch": 0.38558031038750834,
+      "grad_norm": 3.6303059979671617,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 16199
+    },
+    {
+      "epoch": 0.3856041131105398,
+      "grad_norm": 1.5050515969769291,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 16200
+    },
+    {
+      "epoch": 0.38562791583357137,
+      "grad_norm": 2.4586432583372666,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 16201
+    },
+    {
+      "epoch": 0.38565171855660285,
+      "grad_norm": 0.35693353033225406,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 16202
+    },
+    {
+      "epoch": 0.3856755212796344,
+      "grad_norm": 2.757994651712532,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 16203
+    },
+    {
+      "epoch": 0.3856993240026659,
+      "grad_norm": 12.899326037398223,
+      "learning_rate": 2e-05,
+      "loss": 2.0626,
+      "step": 16204
+    },
+    {
+      "epoch": 0.3857231267256974,
+      "grad_norm": 2.871054713427025,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 16205
+    },
+    {
+      "epoch": 0.38574692944872896,
+      "grad_norm": 2.5786692456282423,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 16206
+    },
+    {
+      "epoch": 0.38577073217176044,
+      "grad_norm": 4.257484151442344,
+      "learning_rate": 2e-05,
+      "loss": 0.6769,
+      "step": 16207
+    },
+    {
+      "epoch": 0.385794534894792,
+      "grad_norm": 3.6582208785265435,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 16208
+    },
+    {
+      "epoch": 0.38581833761782347,
+      "grad_norm": 5.0866858923267095,
+      "learning_rate": 2e-05,
+      "loss": 0.21,
+      "step": 16209
+    },
+    {
+      "epoch": 0.385842140340855,
+      "grad_norm": 0.6345663565879078,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16210
+    },
+    {
+      "epoch": 0.3858659430638865,
+      "grad_norm": 2.5233361819043316,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 16211
+    },
+    {
+      "epoch": 0.38588974578691804,
+      "grad_norm": 6.171955164412508,
+      "learning_rate": 2e-05,
+      "loss": 1.4541,
+      "step": 16212
+    },
+    {
+      "epoch": 0.3859135485099495,
+      "grad_norm": 0.832747570508487,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 16213
+    },
+    {
+      "epoch": 0.38593735123298106,
+      "grad_norm": 9.473285690672556,
+      "learning_rate": 2e-05,
+      "loss": 1.0,
+      "step": 16214
+    },
+    {
+      "epoch": 0.38596115395601255,
+      "grad_norm": 2.9120227597360917,
+      "learning_rate": 2e-05,
+      "loss": 0.2247,
+      "step": 16215
+    },
+    {
+      "epoch": 0.3859849566790441,
+      "grad_norm": 5.545914867731773,
+      "learning_rate": 2e-05,
+      "loss": 0.8171,
+      "step": 16216
+    },
+    {
+      "epoch": 0.3860087594020756,
+      "grad_norm": 4.490607728270254,
+      "learning_rate": 2e-05,
+      "loss": 0.1662,
+      "step": 16217
+    },
+    {
+      "epoch": 0.3860325621251071,
+      "grad_norm": 1.1692002423913554,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 16218
+    },
+    {
+      "epoch": 0.3860563648481386,
+      "grad_norm": 5.356682204845633,
+      "learning_rate": 2e-05,
+      "loss": 0.7552,
+      "step": 16219
+    },
+    {
+      "epoch": 0.38608016757117014,
+      "grad_norm": 1.278511673554982,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 16220
+    },
+    {
+      "epoch": 0.3861039702942017,
+      "grad_norm": 4.696123856280095,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 16221
+    },
+    {
+      "epoch": 0.38612777301723317,
+      "grad_norm": 1.2114590070754425,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 16222
+    },
+    {
+      "epoch": 0.3861515757402647,
+      "grad_norm": 6.001137587624065,
+      "learning_rate": 2e-05,
+      "loss": 0.6631,
+      "step": 16223
+    },
+    {
+      "epoch": 0.3861753784632962,
+      "grad_norm": 9.488296008546284,
+      "learning_rate": 2e-05,
+      "loss": 0.3481,
+      "step": 16224
+    },
+    {
+      "epoch": 0.38619918118632773,
+      "grad_norm": 3.0557192035701566,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 16225
+    },
+    {
+      "epoch": 0.3862229839093592,
+      "grad_norm": 2.467467066284165,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 16226
+    },
+    {
+      "epoch": 0.38624678663239076,
+      "grad_norm": 0.7157508085790627,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 16227
+    },
+    {
+      "epoch": 0.38627058935542224,
+      "grad_norm": 0.8484421854268229,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16228
+    },
+    {
+      "epoch": 0.3862943920784538,
+      "grad_norm": 4.84135894691895,
+      "learning_rate": 2e-05,
+      "loss": 0.8558,
+      "step": 16229
+    },
+    {
+      "epoch": 0.38631819480148527,
+      "grad_norm": 8.240512450473945,
+      "learning_rate": 2e-05,
+      "loss": 0.3915,
+      "step": 16230
+    },
+    {
+      "epoch": 0.3863419975245168,
+      "grad_norm": 2.192382152786519,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 16231
+    },
+    {
+      "epoch": 0.3863658002475483,
+      "grad_norm": 5.8125861360953195,
+      "learning_rate": 2e-05,
+      "loss": 1.5536,
+      "step": 16232
+    },
+    {
+      "epoch": 0.38638960297057984,
+      "grad_norm": 0.3873472364130248,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 16233
+    },
+    {
+      "epoch": 0.3864134056936113,
+      "grad_norm": 3.0381004742484885,
+      "learning_rate": 2e-05,
+      "loss": 0.333,
+      "step": 16234
+    },
+    {
+      "epoch": 0.38643720841664286,
+      "grad_norm": 3.1295866726105546,
+      "learning_rate": 2e-05,
+      "loss": 0.1418,
+      "step": 16235
+    },
+    {
+      "epoch": 0.3864610111396744,
+      "grad_norm": 1.2243716497759622,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 16236
+    },
+    {
+      "epoch": 0.3864848138627059,
+      "grad_norm": 2.8008206983990367,
+      "learning_rate": 2e-05,
+      "loss": 0.2269,
+      "step": 16237
+    },
+    {
+      "epoch": 0.38650861658573743,
+      "grad_norm": 1.1854370987976204,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 16238
+    },
+    {
+      "epoch": 0.3865324193087689,
+      "grad_norm": 12.265208581588602,
+      "learning_rate": 2e-05,
+      "loss": 1.1776,
+      "step": 16239
+    },
+    {
+      "epoch": 0.38655622203180046,
+      "grad_norm": 0.5574874227479224,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 16240
+    },
+    {
+      "epoch": 0.38658002475483194,
+      "grad_norm": 6.436160142000852,
+      "learning_rate": 2e-05,
+      "loss": 0.2685,
+      "step": 16241
+    },
+    {
+      "epoch": 0.3866038274778635,
+      "grad_norm": 1.6035272326872327,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 16242
+    },
+    {
+      "epoch": 0.38662763020089497,
+      "grad_norm": 0.22266761374505187,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 16243
+    },
+    {
+      "epoch": 0.3866514329239265,
+      "grad_norm": 2.927379519305347,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 16244
+    },
+    {
+      "epoch": 0.386675235646958,
+      "grad_norm": 1.1237196719928644,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 16245
+    },
+    {
+      "epoch": 0.38669903836998953,
+      "grad_norm": 4.702939901224574,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 16246
+    },
+    {
+      "epoch": 0.386722841093021,
+      "grad_norm": 8.298254865875597,
+      "learning_rate": 2e-05,
+      "loss": 0.5003,
+      "step": 16247
+    },
+    {
+      "epoch": 0.38674664381605256,
+      "grad_norm": 0.682899640043727,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16248
+    },
+    {
+      "epoch": 0.38677044653908405,
+      "grad_norm": 11.291077019714768,
+      "learning_rate": 2e-05,
+      "loss": 1.3848,
+      "step": 16249
+    },
+    {
+      "epoch": 0.3867942492621156,
+      "grad_norm": 1.2208711081687027,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 16250
+    },
+    {
+      "epoch": 0.3868180519851471,
+      "grad_norm": 5.241387168055175,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 16251
+    },
+    {
+      "epoch": 0.3868418547081786,
+      "grad_norm": 6.2670238671085565,
+      "learning_rate": 2e-05,
+      "loss": 0.4217,
+      "step": 16252
+    },
+    {
+      "epoch": 0.38686565743121015,
+      "grad_norm": 0.5749607541095235,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16253
+    },
+    {
+      "epoch": 0.38688946015424164,
+      "grad_norm": 5.298638649703968,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 16254
+    },
+    {
+      "epoch": 0.3869132628772732,
+      "grad_norm": 2.053236006618057,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 16255
+    },
+    {
+      "epoch": 0.38693706560030466,
+      "grad_norm": 5.503143825887859,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 16256
+    },
+    {
+      "epoch": 0.3869608683233362,
+      "grad_norm": 12.701504152258003,
+      "learning_rate": 2e-05,
+      "loss": 0.9785,
+      "step": 16257
+    },
+    {
+      "epoch": 0.3869846710463677,
+      "grad_norm": 3.093228220017696,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 16258
+    },
+    {
+      "epoch": 0.38700847376939923,
+      "grad_norm": 0.4951039352237778,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 16259
+    },
+    {
+      "epoch": 0.3870322764924307,
+      "grad_norm": 3.76703106540128,
+      "learning_rate": 2e-05,
+      "loss": 0.5251,
+      "step": 16260
+    },
+    {
+      "epoch": 0.38705607921546226,
+      "grad_norm": 6.008163802318863,
+      "learning_rate": 2e-05,
+      "loss": 0.373,
+      "step": 16261
+    },
+    {
+      "epoch": 0.38707988193849374,
+      "grad_norm": 2.4362605354536395,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 16262
+    },
+    {
+      "epoch": 0.3871036846615253,
+      "grad_norm": 18.240285054651157,
+      "learning_rate": 2e-05,
+      "loss": 0.6069,
+      "step": 16263
+    },
+    {
+      "epoch": 0.38712748738455677,
+      "grad_norm": 0.16720462307789896,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 16264
+    },
+    {
+      "epoch": 0.3871512901075883,
+      "grad_norm": 6.311363043410829,
+      "learning_rate": 2e-05,
+      "loss": 0.752,
+      "step": 16265
+    },
+    {
+      "epoch": 0.38717509283061985,
+      "grad_norm": 8.480923486857955,
+      "learning_rate": 2e-05,
+      "loss": 0.1648,
+      "step": 16266
+    },
+    {
+      "epoch": 0.38719889555365133,
+      "grad_norm": 9.057687098614956,
+      "learning_rate": 2e-05,
+      "loss": 1.1064,
+      "step": 16267
+    },
+    {
+      "epoch": 0.3872226982766829,
+      "grad_norm": 2.065645495664559,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 16268
+    },
+    {
+      "epoch": 0.38724650099971436,
+      "grad_norm": 4.572482374273244,
+      "learning_rate": 2e-05,
+      "loss": 0.624,
+      "step": 16269
+    },
+    {
+      "epoch": 0.3872703037227459,
+      "grad_norm": 6.849108603470857,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 16270
+    },
+    {
+      "epoch": 0.3872941064457774,
+      "grad_norm": 0.791211941969677,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 16271
+    },
+    {
+      "epoch": 0.3873179091688089,
+      "grad_norm": 2.254208288639146,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 16272
+    },
+    {
+      "epoch": 0.3873417118918404,
+      "grad_norm": 6.94853799965,
+      "learning_rate": 2e-05,
+      "loss": 0.6834,
+      "step": 16273
+    },
+    {
+      "epoch": 0.38736551461487195,
+      "grad_norm": 2.5577373311165497,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 16274
+    },
+    {
+      "epoch": 0.38738931733790344,
+      "grad_norm": 6.001618723630165,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 16275
+    },
+    {
+      "epoch": 0.387413120060935,
+      "grad_norm": 20.212426216021026,
+      "learning_rate": 2e-05,
+      "loss": 0.3916,
+      "step": 16276
+    },
+    {
+      "epoch": 0.38743692278396646,
+      "grad_norm": 5.187543027607268,
+      "learning_rate": 2e-05,
+      "loss": 0.7224,
+      "step": 16277
+    },
+    {
+      "epoch": 0.387460725506998,
+      "grad_norm": 3.648818178988195,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 16278
+    },
+    {
+      "epoch": 0.3874845282300295,
+      "grad_norm": 5.344600571929394,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 16279
+    },
+    {
+      "epoch": 0.38750833095306103,
+      "grad_norm": 3.4176101861471166,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 16280
+    },
+    {
+      "epoch": 0.38753213367609257,
+      "grad_norm": 0.2424474317184424,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 16281
+    },
+    {
+      "epoch": 0.38755593639912406,
+      "grad_norm": 2.7841897365085693,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 16282
+    },
+    {
+      "epoch": 0.3875797391221556,
+      "grad_norm": 1.3724865796634043,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 16283
+    },
+    {
+      "epoch": 0.3876035418451871,
+      "grad_norm": 0.31697664543157344,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 16284
+    },
+    {
+      "epoch": 0.3876273445682186,
+      "grad_norm": 0.23163038550891102,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16285
+    },
+    {
+      "epoch": 0.3876511472912501,
+      "grad_norm": 1.15439632367035,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 16286
+    },
+    {
+      "epoch": 0.38767495001428165,
+      "grad_norm": 6.031660373749821,
+      "learning_rate": 2e-05,
+      "loss": 0.6535,
+      "step": 16287
+    },
+    {
+      "epoch": 0.38769875273731313,
+      "grad_norm": 12.722682636867956,
+      "learning_rate": 2e-05,
+      "loss": 0.6388,
+      "step": 16288
+    },
+    {
+      "epoch": 0.3877225554603447,
+      "grad_norm": 5.303923608296396,
+      "learning_rate": 2e-05,
+      "loss": 1.2426,
+      "step": 16289
+    },
+    {
+      "epoch": 0.38774635818337616,
+      "grad_norm": 6.715088856690689,
+      "learning_rate": 2e-05,
+      "loss": 0.6899,
+      "step": 16290
+    },
+    {
+      "epoch": 0.3877701609064077,
+      "grad_norm": 6.018107851563518,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 16291
+    },
+    {
+      "epoch": 0.3877939636294392,
+      "grad_norm": 0.4263494414801996,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 16292
+    },
+    {
+      "epoch": 0.3878177663524707,
+      "grad_norm": 6.912447170613859,
+      "learning_rate": 2e-05,
+      "loss": 0.3841,
+      "step": 16293
+    },
+    {
+      "epoch": 0.3878415690755022,
+      "grad_norm": 16.468815339248255,
+      "learning_rate": 2e-05,
+      "loss": 0.956,
+      "step": 16294
+    },
+    {
+      "epoch": 0.38786537179853375,
+      "grad_norm": 1.2450300752846062,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16295
+    },
+    {
+      "epoch": 0.3878891745215653,
+      "grad_norm": 5.708267976216284,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 16296
+    },
+    {
+      "epoch": 0.3879129772445968,
+      "grad_norm": 4.205854127329911,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 16297
+    },
+    {
+      "epoch": 0.3879367799676283,
+      "grad_norm": 0.7430735601325102,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 16298
+    },
+    {
+      "epoch": 0.3879605826906598,
+      "grad_norm": 2.188583309162187,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 16299
+    },
+    {
+      "epoch": 0.38798438541369135,
+      "grad_norm": 3.2855229115611575,
+      "learning_rate": 2e-05,
+      "loss": 0.3276,
+      "step": 16300
+    },
+    {
+      "epoch": 0.38800818813672283,
+      "grad_norm": 3.409578977705111,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 16301
+    },
+    {
+      "epoch": 0.38803199085975437,
+      "grad_norm": 21.122380319639557,
+      "learning_rate": 2e-05,
+      "loss": 0.3751,
+      "step": 16302
+    },
+    {
+      "epoch": 0.38805579358278586,
+      "grad_norm": 10.575681985306066,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 16303
+    },
+    {
+      "epoch": 0.3880795963058174,
+      "grad_norm": 2.8699029252479393,
+      "learning_rate": 2e-05,
+      "loss": 0.1541,
+      "step": 16304
+    },
+    {
+      "epoch": 0.3881033990288489,
+      "grad_norm": 5.142858274562779,
+      "learning_rate": 2e-05,
+      "loss": 0.4613,
+      "step": 16305
+    },
+    {
+      "epoch": 0.3881272017518804,
+      "grad_norm": 3.3940669115765676,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 16306
+    },
+    {
+      "epoch": 0.3881510044749119,
+      "grad_norm": 11.260029749893366,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 16307
+    },
+    {
+      "epoch": 0.38817480719794345,
+      "grad_norm": 8.508380879299938,
+      "learning_rate": 2e-05,
+      "loss": 0.9003,
+      "step": 16308
+    },
+    {
+      "epoch": 0.38819860992097494,
+      "grad_norm": 1.197211337841287,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 16309
+    },
+    {
+      "epoch": 0.3882224126440065,
+      "grad_norm": 5.9162338507061305,
+      "learning_rate": 2e-05,
+      "loss": 0.9478,
+      "step": 16310
+    },
+    {
+      "epoch": 0.388246215367038,
+      "grad_norm": 3.9110109553926526,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 16311
+    },
+    {
+      "epoch": 0.3882700180900695,
+      "grad_norm": 6.0132915079803375,
+      "learning_rate": 2e-05,
+      "loss": 1.4153,
+      "step": 16312
+    },
+    {
+      "epoch": 0.38829382081310104,
+      "grad_norm": 0.8333546654651653,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 16313
+    },
+    {
+      "epoch": 0.38831762353613253,
+      "grad_norm": 8.236907099284604,
+      "learning_rate": 2e-05,
+      "loss": 0.689,
+      "step": 16314
+    },
+    {
+      "epoch": 0.38834142625916407,
+      "grad_norm": 18.471655055359502,
+      "learning_rate": 2e-05,
+      "loss": 1.1934,
+      "step": 16315
+    },
+    {
+      "epoch": 0.38836522898219555,
+      "grad_norm": 2.949171732801432,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 16316
+    },
+    {
+      "epoch": 0.3883890317052271,
+      "grad_norm": 3.441165670504703,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 16317
+    },
+    {
+      "epoch": 0.3884128344282586,
+      "grad_norm": 2.113819392481507,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 16318
+    },
+    {
+      "epoch": 0.3884366371512901,
+      "grad_norm": 5.900126297652877,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 16319
+    },
+    {
+      "epoch": 0.3884604398743216,
+      "grad_norm": 5.4219952312845585,
+      "learning_rate": 2e-05,
+      "loss": 0.7188,
+      "step": 16320
+    },
+    {
+      "epoch": 0.38848424259735315,
+      "grad_norm": 17.356013582287027,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 16321
+    },
+    {
+      "epoch": 0.38850804532038463,
+      "grad_norm": 3.508897620156914,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 16322
+    },
+    {
+      "epoch": 0.3885318480434162,
+      "grad_norm": 7.594878500137642,
+      "learning_rate": 2e-05,
+      "loss": 0.2243,
+      "step": 16323
+    },
+    {
+      "epoch": 0.38855565076644766,
+      "grad_norm": 0.37611435866931153,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 16324
+    },
+    {
+      "epoch": 0.3885794534894792,
+      "grad_norm": 8.435845637703347,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 16325
+    },
+    {
+      "epoch": 0.3886032562125107,
+      "grad_norm": 7.565157285200917,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 16326
+    },
+    {
+      "epoch": 0.3886270589355422,
+      "grad_norm": 9.523323593255341,
+      "learning_rate": 2e-05,
+      "loss": 0.539,
+      "step": 16327
+    },
+    {
+      "epoch": 0.38865086165857377,
+      "grad_norm": 6.063036093187011,
+      "learning_rate": 2e-05,
+      "loss": 1.0936,
+      "step": 16328
+    },
+    {
+      "epoch": 0.38867466438160525,
+      "grad_norm": 4.5982775358299905,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 16329
+    },
+    {
+      "epoch": 0.3886984671046368,
+      "grad_norm": 1.244317525619268,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 16330
+    },
+    {
+      "epoch": 0.3887222698276683,
+      "grad_norm": 0.4166075671233656,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 16331
+    },
+    {
+      "epoch": 0.3887460725506998,
+      "grad_norm": 6.530453203949911,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 16332
+    },
+    {
+      "epoch": 0.3887698752737313,
+      "grad_norm": 10.52125258305014,
+      "learning_rate": 2e-05,
+      "loss": 0.9132,
+      "step": 16333
+    },
+    {
+      "epoch": 0.38879367799676284,
+      "grad_norm": 4.220454484188999,
+      "learning_rate": 2e-05,
+      "loss": 0.6795,
+      "step": 16334
+    },
+    {
+      "epoch": 0.38881748071979433,
+      "grad_norm": 3.366973111896126,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 16335
+    },
+    {
+      "epoch": 0.38884128344282587,
+      "grad_norm": 2.4621006865036654,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 16336
+    },
+    {
+      "epoch": 0.38886508616585735,
+      "grad_norm": 1.8858801992536403,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 16337
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 1.7242193057676354,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 16338
+    },
+    {
+      "epoch": 0.3889126916119204,
+      "grad_norm": 4.403176008486682,
+      "learning_rate": 2e-05,
+      "loss": 0.646,
+      "step": 16339
+    },
+    {
+      "epoch": 0.3889364943349519,
+      "grad_norm": 4.454753981363101,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 16340
+    },
+    {
+      "epoch": 0.3889602970579834,
+      "grad_norm": 8.413458516844035,
+      "learning_rate": 2e-05,
+      "loss": 0.6642,
+      "step": 16341
+    },
+    {
+      "epoch": 0.38898409978101495,
+      "grad_norm": 2.926178429606558,
+      "learning_rate": 2e-05,
+      "loss": 0.3221,
+      "step": 16342
+    },
+    {
+      "epoch": 0.3890079025040465,
+      "grad_norm": 7.162117076684878,
+      "learning_rate": 2e-05,
+      "loss": 0.8389,
+      "step": 16343
+    },
+    {
+      "epoch": 0.389031705227078,
+      "grad_norm": 3.6235823384376333,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 16344
+    },
+    {
+      "epoch": 0.3890555079501095,
+      "grad_norm": 4.880146063145369,
+      "learning_rate": 2e-05,
+      "loss": 0.7221,
+      "step": 16345
+    },
+    {
+      "epoch": 0.389079310673141,
+      "grad_norm": 4.325121855147337,
+      "learning_rate": 2e-05,
+      "loss": 0.4982,
+      "step": 16346
+    },
+    {
+      "epoch": 0.38910311339617254,
+      "grad_norm": 6.0487876525775155,
+      "learning_rate": 2e-05,
+      "loss": 1.0749,
+      "step": 16347
+    },
+    {
+      "epoch": 0.389126916119204,
+      "grad_norm": 3.893696919080957,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 16348
+    },
+    {
+      "epoch": 0.38915071884223557,
+      "grad_norm": 0.8436934900768742,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16349
+    },
+    {
+      "epoch": 0.38917452156526705,
+      "grad_norm": 1.5658807625623063,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 16350
+    },
+    {
+      "epoch": 0.3891983242882986,
+      "grad_norm": 8.850971022228803,
+      "learning_rate": 2e-05,
+      "loss": 0.2176,
+      "step": 16351
+    },
+    {
+      "epoch": 0.3892221270113301,
+      "grad_norm": 1.3131724707377899,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 16352
+    },
+    {
+      "epoch": 0.3892459297343616,
+      "grad_norm": 3.561937106596015,
+      "learning_rate": 2e-05,
+      "loss": 0.527,
+      "step": 16353
+    },
+    {
+      "epoch": 0.3892697324573931,
+      "grad_norm": 6.393949829756385,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 16354
+    },
+    {
+      "epoch": 0.38929353518042464,
+      "grad_norm": 9.527427406090736,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 16355
+    },
+    {
+      "epoch": 0.38931733790345613,
+      "grad_norm": 2.998872774399459,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 16356
+    },
+    {
+      "epoch": 0.38934114062648767,
+      "grad_norm": 1.5575838607533163,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 16357
+    },
+    {
+      "epoch": 0.3893649433495192,
+      "grad_norm": 1.8491425072554637,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 16358
+    },
+    {
+      "epoch": 0.3893887460725507,
+      "grad_norm": 5.1360966055972135,
+      "learning_rate": 2e-05,
+      "loss": 0.3265,
+      "step": 16359
+    },
+    {
+      "epoch": 0.38941254879558224,
+      "grad_norm": 0.2570085777468945,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16360
+    },
+    {
+      "epoch": 0.3894363515186137,
+      "grad_norm": 4.815983103974409,
+      "learning_rate": 2e-05,
+      "loss": 0.5734,
+      "step": 16361
+    },
+    {
+      "epoch": 0.38946015424164526,
+      "grad_norm": 4.8473396426228055,
+      "learning_rate": 2e-05,
+      "loss": 0.9919,
+      "step": 16362
+    },
+    {
+      "epoch": 0.38948395696467675,
+      "grad_norm": 0.8417476348243302,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 16363
+    },
+    {
+      "epoch": 0.3895077596877083,
+      "grad_norm": 3.8934130766829274,
+      "learning_rate": 2e-05,
+      "loss": 0.6577,
+      "step": 16364
+    },
+    {
+      "epoch": 0.3895315624107398,
+      "grad_norm": 4.793738818174474,
+      "learning_rate": 2e-05,
+      "loss": 0.6836,
+      "step": 16365
+    },
+    {
+      "epoch": 0.3895553651337713,
+      "grad_norm": 1.4221803056158662,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 16366
+    },
+    {
+      "epoch": 0.3895791678568028,
+      "grad_norm": 3.5804000463903276,
+      "learning_rate": 2e-05,
+      "loss": 0.5678,
+      "step": 16367
+    },
+    {
+      "epoch": 0.38960297057983434,
+      "grad_norm": 5.538217794796097,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 16368
+    },
+    {
+      "epoch": 0.3896267733028658,
+      "grad_norm": 4.379286252459767,
+      "learning_rate": 2e-05,
+      "loss": 0.6531,
+      "step": 16369
+    },
+    {
+      "epoch": 0.38965057602589737,
+      "grad_norm": 5.121169358635418,
+      "learning_rate": 2e-05,
+      "loss": 0.2982,
+      "step": 16370
+    },
+    {
+      "epoch": 0.38967437874892885,
+      "grad_norm": 11.784130298979058,
+      "learning_rate": 2e-05,
+      "loss": 0.4066,
+      "step": 16371
+    },
+    {
+      "epoch": 0.3896981814719604,
+      "grad_norm": 2.5790341876427463,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 16372
+    },
+    {
+      "epoch": 0.38972198419499193,
+      "grad_norm": 12.38102673583362,
+      "learning_rate": 2e-05,
+      "loss": 0.5952,
+      "step": 16373
+    },
+    {
+      "epoch": 0.3897457869180234,
+      "grad_norm": 3.391724025640642,
+      "learning_rate": 2e-05,
+      "loss": 0.6595,
+      "step": 16374
+    },
+    {
+      "epoch": 0.38976958964105496,
+      "grad_norm": 3.951098625776948,
+      "learning_rate": 2e-05,
+      "loss": 0.5589,
+      "step": 16375
+    },
+    {
+      "epoch": 0.38979339236408644,
+      "grad_norm": 3.4063630031642624,
+      "learning_rate": 2e-05,
+      "loss": 0.4995,
+      "step": 16376
+    },
+    {
+      "epoch": 0.389817195087118,
+      "grad_norm": 4.951951044148891,
+      "learning_rate": 2e-05,
+      "loss": 1.2403,
+      "step": 16377
+    },
+    {
+      "epoch": 0.38984099781014947,
+      "grad_norm": 7.612595317281915,
+      "learning_rate": 2e-05,
+      "loss": 0.3715,
+      "step": 16378
+    },
+    {
+      "epoch": 0.389864800533181,
+      "grad_norm": 5.378343454035492,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 16379
+    },
+    {
+      "epoch": 0.3898886032562125,
+      "grad_norm": 4.055191448953709,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 16380
+    },
+    {
+      "epoch": 0.38991240597924404,
+      "grad_norm": 5.891604395126521,
+      "learning_rate": 2e-05,
+      "loss": 0.7915,
+      "step": 16381
+    },
+    {
+      "epoch": 0.3899362087022755,
+      "grad_norm": 5.9945514745168165,
+      "learning_rate": 2e-05,
+      "loss": 1.2075,
+      "step": 16382
+    },
+    {
+      "epoch": 0.38996001142530706,
+      "grad_norm": 4.996182044368565,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 16383
+    },
+    {
+      "epoch": 0.38998381414833855,
+      "grad_norm": 10.71345538752526,
+      "learning_rate": 2e-05,
+      "loss": 0.322,
+      "step": 16384
+    },
+    {
+      "epoch": 0.3900076168713701,
+      "grad_norm": 9.780626094269778,
+      "learning_rate": 2e-05,
+      "loss": 1.5136,
+      "step": 16385
+    },
+    {
+      "epoch": 0.3900314195944016,
+      "grad_norm": 2.2508635240682255,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 16386
+    },
+    {
+      "epoch": 0.3900552223174331,
+      "grad_norm": 8.995203246019624,
+      "learning_rate": 2e-05,
+      "loss": 1.1481,
+      "step": 16387
+    },
+    {
+      "epoch": 0.39007902504046466,
+      "grad_norm": 3.2346266939848474,
+      "learning_rate": 2e-05,
+      "loss": 0.2841,
+      "step": 16388
+    },
+    {
+      "epoch": 0.39010282776349614,
+      "grad_norm": 0.13138662247071695,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16389
+    },
+    {
+      "epoch": 0.3901266304865277,
+      "grad_norm": 3.0335314096668076,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 16390
+    },
+    {
+      "epoch": 0.39015043320955917,
+      "grad_norm": 6.919566370696394,
+      "learning_rate": 2e-05,
+      "loss": 0.8365,
+      "step": 16391
+    },
+    {
+      "epoch": 0.3901742359325907,
+      "grad_norm": 5.893672788922728,
+      "learning_rate": 2e-05,
+      "loss": 1.2146,
+      "step": 16392
+    },
+    {
+      "epoch": 0.3901980386556222,
+      "grad_norm": 13.614048533701792,
+      "learning_rate": 2e-05,
+      "loss": 0.2229,
+      "step": 16393
+    },
+    {
+      "epoch": 0.39022184137865373,
+      "grad_norm": 3.4300393291311733,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 16394
+    },
+    {
+      "epoch": 0.3902456441016852,
+      "grad_norm": 15.184775285163036,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 16395
+    },
+    {
+      "epoch": 0.39026944682471676,
+      "grad_norm": 7.185603630006236,
+      "learning_rate": 2e-05,
+      "loss": 0.3922,
+      "step": 16396
+    },
+    {
+      "epoch": 0.39029324954774824,
+      "grad_norm": 21.121970431850507,
+      "learning_rate": 2e-05,
+      "loss": 0.3285,
+      "step": 16397
+    },
+    {
+      "epoch": 0.3903170522707798,
+      "grad_norm": 1.6469556480158836,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 16398
+    },
+    {
+      "epoch": 0.39034085499381127,
+      "grad_norm": 5.357117685929953,
+      "learning_rate": 2e-05,
+      "loss": 0.5466,
+      "step": 16399
+    },
+    {
+      "epoch": 0.3903646577168428,
+      "grad_norm": 13.02042846204657,
+      "learning_rate": 2e-05,
+      "loss": 0.2941,
+      "step": 16400
+    },
+    {
+      "epoch": 0.3903884604398743,
+      "grad_norm": 7.377377852523338,
+      "learning_rate": 2e-05,
+      "loss": 0.6998,
+      "step": 16401
+    },
+    {
+      "epoch": 0.39041226316290584,
+      "grad_norm": 3.1787812503207067,
+      "learning_rate": 2e-05,
+      "loss": 0.603,
+      "step": 16402
+    },
+    {
+      "epoch": 0.3904360658859374,
+      "grad_norm": 3.8508227028904067,
+      "learning_rate": 2e-05,
+      "loss": 0.2987,
+      "step": 16403
+    },
+    {
+      "epoch": 0.39045986860896886,
+      "grad_norm": 7.052538733329222,
+      "learning_rate": 2e-05,
+      "loss": 1.4188,
+      "step": 16404
+    },
+    {
+      "epoch": 0.3904836713320004,
+      "grad_norm": 3.6563197142284944,
+      "learning_rate": 2e-05,
+      "loss": 0.1894,
+      "step": 16405
+    },
+    {
+      "epoch": 0.3905074740550319,
+      "grad_norm": 2.8389529748239672,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 16406
+    },
+    {
+      "epoch": 0.39053127677806343,
+      "grad_norm": 6.14358249800048,
+      "learning_rate": 2e-05,
+      "loss": 1.4484,
+      "step": 16407
+    },
+    {
+      "epoch": 0.3905550795010949,
+      "grad_norm": 10.267275870347097,
+      "learning_rate": 2e-05,
+      "loss": 0.8384,
+      "step": 16408
+    },
+    {
+      "epoch": 0.39057888222412646,
+      "grad_norm": 4.250699221702081,
+      "learning_rate": 2e-05,
+      "loss": 0.6105,
+      "step": 16409
+    },
+    {
+      "epoch": 0.39060268494715794,
+      "grad_norm": 3.8156148566282932,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 16410
+    },
+    {
+      "epoch": 0.3906264876701895,
+      "grad_norm": 10.264063928588687,
+      "learning_rate": 2e-05,
+      "loss": 0.4008,
+      "step": 16411
+    },
+    {
+      "epoch": 0.39065029039322097,
+      "grad_norm": 2.2114931322370315,
+      "learning_rate": 2e-05,
+      "loss": 0.1331,
+      "step": 16412
+    },
+    {
+      "epoch": 0.3906740931162525,
+      "grad_norm": 2.8767242160222275,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 16413
+    },
+    {
+      "epoch": 0.390697895839284,
+      "grad_norm": 5.153476310477366,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 16414
+    },
+    {
+      "epoch": 0.39072169856231553,
+      "grad_norm": 4.853581464662565,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 16415
+    },
+    {
+      "epoch": 0.390745501285347,
+      "grad_norm": 1.8692671470764686,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 16416
+    },
+    {
+      "epoch": 0.39076930400837856,
+      "grad_norm": 15.711834095508577,
+      "learning_rate": 2e-05,
+      "loss": 0.4563,
+      "step": 16417
+    },
+    {
+      "epoch": 0.3907931067314101,
+      "grad_norm": 2.9085558936138307,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 16418
+    },
+    {
+      "epoch": 0.3908169094544416,
+      "grad_norm": 8.560704118657021,
+      "learning_rate": 2e-05,
+      "loss": 0.3208,
+      "step": 16419
+    },
+    {
+      "epoch": 0.3908407121774731,
+      "grad_norm": 6.774593995766304,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 16420
+    },
+    {
+      "epoch": 0.3908645149005046,
+      "grad_norm": 7.802055173997512,
+      "learning_rate": 2e-05,
+      "loss": 0.3291,
+      "step": 16421
+    },
+    {
+      "epoch": 0.39088831762353615,
+      "grad_norm": 0.4417439375298389,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 16422
+    },
+    {
+      "epoch": 0.39091212034656764,
+      "grad_norm": 0.9242866386938383,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 16423
+    },
+    {
+      "epoch": 0.3909359230695992,
+      "grad_norm": 4.229424413252433,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 16424
+    },
+    {
+      "epoch": 0.39095972579263066,
+      "grad_norm": 0.8485617990862843,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 16425
+    },
+    {
+      "epoch": 0.3909835285156622,
+      "grad_norm": 4.604181468094858,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 16426
+    },
+    {
+      "epoch": 0.3910073312386937,
+      "grad_norm": 0.80308703107825,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 16427
+    },
+    {
+      "epoch": 0.39103113396172523,
+      "grad_norm": 0.5411040646954953,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16428
+    },
+    {
+      "epoch": 0.3910549366847567,
+      "grad_norm": 1.7584288871661173,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 16429
+    },
+    {
+      "epoch": 0.39107873940778826,
+      "grad_norm": 3.9252746501198432,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 16430
+    },
+    {
+      "epoch": 0.39110254213081974,
+      "grad_norm": 9.206089058173642,
+      "learning_rate": 2e-05,
+      "loss": 0.9202,
+      "step": 16431
+    },
+    {
+      "epoch": 0.3911263448538513,
+      "grad_norm": 1.9045908815188894,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 16432
+    },
+    {
+      "epoch": 0.3911501475768828,
+      "grad_norm": 6.740800442544979,
+      "learning_rate": 2e-05,
+      "loss": 0.8605,
+      "step": 16433
+    },
+    {
+      "epoch": 0.3911739502999143,
+      "grad_norm": 1.49389012857471,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 16434
+    },
+    {
+      "epoch": 0.39119775302294585,
+      "grad_norm": 6.341283603643581,
+      "learning_rate": 2e-05,
+      "loss": 1.0874,
+      "step": 16435
+    },
+    {
+      "epoch": 0.39122155574597733,
+      "grad_norm": 4.936120269625055,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 16436
+    },
+    {
+      "epoch": 0.3912453584690089,
+      "grad_norm": 15.712222807436536,
+      "learning_rate": 2e-05,
+      "loss": 0.9314,
+      "step": 16437
+    },
+    {
+      "epoch": 0.39126916119204036,
+      "grad_norm": 2.149198381101294,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 16438
+    },
+    {
+      "epoch": 0.3912929639150719,
+      "grad_norm": 5.878459267384394,
+      "learning_rate": 2e-05,
+      "loss": 0.7196,
+      "step": 16439
+    },
+    {
+      "epoch": 0.3913167666381034,
+      "grad_norm": 3.6461715315765217,
+      "learning_rate": 2e-05,
+      "loss": 0.442,
+      "step": 16440
+    },
+    {
+      "epoch": 0.3913405693611349,
+      "grad_norm": 0.31101795093034384,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 16441
+    },
+    {
+      "epoch": 0.3913643720841664,
+      "grad_norm": 7.691051196239703,
+      "learning_rate": 2e-05,
+      "loss": 1.0305,
+      "step": 16442
+    },
+    {
+      "epoch": 0.39138817480719795,
+      "grad_norm": 1.444021593431086,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 16443
+    },
+    {
+      "epoch": 0.39141197753022944,
+      "grad_norm": 6.000717616940722,
+      "learning_rate": 2e-05,
+      "loss": 1.2366,
+      "step": 16444
+    },
+    {
+      "epoch": 0.391435780253261,
+      "grad_norm": 2.205540842852035,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 16445
+    },
+    {
+      "epoch": 0.39145958297629246,
+      "grad_norm": 5.43942165364961,
+      "learning_rate": 2e-05,
+      "loss": 0.6,
+      "step": 16446
+    },
+    {
+      "epoch": 0.391483385699324,
+      "grad_norm": 4.481102083487389,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 16447
+    },
+    {
+      "epoch": 0.3915071884223555,
+      "grad_norm": 39.77994935473694,
+      "learning_rate": 2e-05,
+      "loss": 0.4749,
+      "step": 16448
+    },
+    {
+      "epoch": 0.39153099114538703,
+      "grad_norm": 2.804891088832415,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 16449
+    },
+    {
+      "epoch": 0.39155479386841857,
+      "grad_norm": 4.206254738275344,
+      "learning_rate": 2e-05,
+      "loss": 0.1681,
+      "step": 16450
+    },
+    {
+      "epoch": 0.39157859659145006,
+      "grad_norm": 5.859609384441502,
+      "learning_rate": 2e-05,
+      "loss": 1.128,
+      "step": 16451
+    },
+    {
+      "epoch": 0.3916023993144816,
+      "grad_norm": 2.116160025291701,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 16452
+    },
+    {
+      "epoch": 0.3916262020375131,
+      "grad_norm": 3.946320347757482,
+      "learning_rate": 2e-05,
+      "loss": 0.8955,
+      "step": 16453
+    },
+    {
+      "epoch": 0.3916500047605446,
+      "grad_norm": 4.072148287543722,
+      "learning_rate": 2e-05,
+      "loss": 0.714,
+      "step": 16454
+    },
+    {
+      "epoch": 0.3916738074835761,
+      "grad_norm": 2.4148993931934792,
+      "learning_rate": 2e-05,
+      "loss": 0.2433,
+      "step": 16455
+    },
+    {
+      "epoch": 0.39169761020660765,
+      "grad_norm": 3.3216887851164367,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 16456
+    },
+    {
+      "epoch": 0.39172141292963913,
+      "grad_norm": 1.9801983918714539,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 16457
+    },
+    {
+      "epoch": 0.3917452156526707,
+      "grad_norm": 6.124444294240179,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 16458
+    },
+    {
+      "epoch": 0.39176901837570216,
+      "grad_norm": 9.506274252551522,
+      "learning_rate": 2e-05,
+      "loss": 0.7703,
+      "step": 16459
+    },
+    {
+      "epoch": 0.3917928210987337,
+      "grad_norm": 3.3214581714000873,
+      "learning_rate": 2e-05,
+      "loss": 0.624,
+      "step": 16460
+    },
+    {
+      "epoch": 0.3918166238217652,
+      "grad_norm": 10.756246113131946,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 16461
+    },
+    {
+      "epoch": 0.3918404265447967,
+      "grad_norm": 3.1929065351280124,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 16462
+    },
+    {
+      "epoch": 0.3918642292678282,
+      "grad_norm": 9.92059043736171,
+      "learning_rate": 2e-05,
+      "loss": 0.5493,
+      "step": 16463
+    },
+    {
+      "epoch": 0.39188803199085975,
+      "grad_norm": 5.809339941224207,
+      "learning_rate": 2e-05,
+      "loss": 1.3004,
+      "step": 16464
+    },
+    {
+      "epoch": 0.3919118347138913,
+      "grad_norm": 0.5074885437213715,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 16465
+    },
+    {
+      "epoch": 0.3919356374369228,
+      "grad_norm": 6.445200209631701,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 16466
+    },
+    {
+      "epoch": 0.3919594401599543,
+      "grad_norm": 1.6437088753154279,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 16467
+    },
+    {
+      "epoch": 0.3919832428829858,
+      "grad_norm": 5.495281280426741,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 16468
+    },
+    {
+      "epoch": 0.39200704560601735,
+      "grad_norm": 1.0256602481345019,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 16469
+    },
+    {
+      "epoch": 0.39203084832904883,
+      "grad_norm": 2.3903541319179102,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 16470
+    },
+    {
+      "epoch": 0.39205465105208037,
+      "grad_norm": 2.951153028749566,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 16471
+    },
+    {
+      "epoch": 0.39207845377511186,
+      "grad_norm": 0.49423691628719124,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16472
+    },
+    {
+      "epoch": 0.3921022564981434,
+      "grad_norm": 0.5465235084551889,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 16473
+    },
+    {
+      "epoch": 0.3921260592211749,
+      "grad_norm": 2.2819213553780866,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 16474
+    },
+    {
+      "epoch": 0.3921498619442064,
+      "grad_norm": 4.009280434762831,
+      "learning_rate": 2e-05,
+      "loss": 0.1826,
+      "step": 16475
+    },
+    {
+      "epoch": 0.3921736646672379,
+      "grad_norm": 8.56781311542296,
+      "learning_rate": 2e-05,
+      "loss": 0.7878,
+      "step": 16476
+    },
+    {
+      "epoch": 0.39219746739026945,
+      "grad_norm": 1.8453483243144855,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 16477
+    },
+    {
+      "epoch": 0.39222127011330093,
+      "grad_norm": 10.281783495510805,
+      "learning_rate": 2e-05,
+      "loss": 0.91,
+      "step": 16478
+    },
+    {
+      "epoch": 0.3922450728363325,
+      "grad_norm": 7.525521267366458,
+      "learning_rate": 2e-05,
+      "loss": 0.565,
+      "step": 16479
+    },
+    {
+      "epoch": 0.392268875559364,
+      "grad_norm": 1.6721353463234854,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 16480
+    },
+    {
+      "epoch": 0.3922926782823955,
+      "grad_norm": 2.0516393808369098,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 16481
+    },
+    {
+      "epoch": 0.39231648100542704,
+      "grad_norm": 7.342118580914922,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 16482
+    },
+    {
+      "epoch": 0.3923402837284585,
+      "grad_norm": 2.7322963466196946,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 16483
+    },
+    {
+      "epoch": 0.39236408645149007,
+      "grad_norm": 5.3622576821553665,
+      "learning_rate": 2e-05,
+      "loss": 1.0347,
+      "step": 16484
+    },
+    {
+      "epoch": 0.39238788917452155,
+      "grad_norm": 1.5691232477580246,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 16485
+    },
+    {
+      "epoch": 0.3924116918975531,
+      "grad_norm": 3.452676567307388,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 16486
+    },
+    {
+      "epoch": 0.3924354946205846,
+      "grad_norm": 1.9413287326999569,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 16487
+    },
+    {
+      "epoch": 0.3924592973436161,
+      "grad_norm": 6.831573690638248,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 16488
+    },
+    {
+      "epoch": 0.3924831000666476,
+      "grad_norm": 0.26585464367260764,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 16489
+    },
+    {
+      "epoch": 0.39250690278967915,
+      "grad_norm": 5.803155230892867,
+      "learning_rate": 2e-05,
+      "loss": 1.2096,
+      "step": 16490
+    },
+    {
+      "epoch": 0.39253070551271063,
+      "grad_norm": 6.558204354847019,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 16491
+    },
+    {
+      "epoch": 0.39255450823574217,
+      "grad_norm": 3.03642638168202,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 16492
+    },
+    {
+      "epoch": 0.39257831095877366,
+      "grad_norm": 4.216671199199208,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 16493
+    },
+    {
+      "epoch": 0.3926021136818052,
+      "grad_norm": 1.1606720390433134,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 16494
+    },
+    {
+      "epoch": 0.39262591640483674,
+      "grad_norm": 5.389014162665586,
+      "learning_rate": 2e-05,
+      "loss": 0.362,
+      "step": 16495
+    },
+    {
+      "epoch": 0.3926497191278682,
+      "grad_norm": 13.366593995368584,
+      "learning_rate": 2e-05,
+      "loss": 0.227,
+      "step": 16496
+    },
+    {
+      "epoch": 0.39267352185089976,
+      "grad_norm": 5.5278232660911,
+      "learning_rate": 2e-05,
+      "loss": 0.1658,
+      "step": 16497
+    },
+    {
+      "epoch": 0.39269732457393125,
+      "grad_norm": 3.483574207944292,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 16498
+    },
+    {
+      "epoch": 0.3927211272969628,
+      "grad_norm": 3.7374391289107844,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 16499
+    },
+    {
+      "epoch": 0.3927449300199943,
+      "grad_norm": 2.526437281759567,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 16500
+    },
+    {
+      "epoch": 0.3927687327430258,
+      "grad_norm": 3.3059765613239005,
+      "learning_rate": 2e-05,
+      "loss": 0.1498,
+      "step": 16501
+    },
+    {
+      "epoch": 0.3927925354660573,
+      "grad_norm": 1.405553582405613,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 16502
+    },
+    {
+      "epoch": 0.39281633818908884,
+      "grad_norm": 5.113518863808448,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 16503
+    },
+    {
+      "epoch": 0.39284014091212033,
+      "grad_norm": 11.09110448306042,
+      "learning_rate": 2e-05,
+      "loss": 0.4858,
+      "step": 16504
+    },
+    {
+      "epoch": 0.39286394363515187,
+      "grad_norm": 17.44716993971705,
+      "learning_rate": 2e-05,
+      "loss": 0.2922,
+      "step": 16505
+    },
+    {
+      "epoch": 0.39288774635818335,
+      "grad_norm": 7.803248296133404,
+      "learning_rate": 2e-05,
+      "loss": 1.0763,
+      "step": 16506
+    },
+    {
+      "epoch": 0.3929115490812149,
+      "grad_norm": 3.9313452712747576,
+      "learning_rate": 2e-05,
+      "loss": 0.653,
+      "step": 16507
+    },
+    {
+      "epoch": 0.3929353518042464,
+      "grad_norm": 6.569010156240477,
+      "learning_rate": 2e-05,
+      "loss": 2.1691,
+      "step": 16508
+    },
+    {
+      "epoch": 0.3929591545272779,
+      "grad_norm": 1.3962016498667578,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 16509
+    },
+    {
+      "epoch": 0.39298295725030946,
+      "grad_norm": 26.458598289450716,
+      "learning_rate": 2e-05,
+      "loss": 0.9432,
+      "step": 16510
+    },
+    {
+      "epoch": 0.39300675997334095,
+      "grad_norm": 10.598130862469873,
+      "learning_rate": 2e-05,
+      "loss": 0.57,
+      "step": 16511
+    },
+    {
+      "epoch": 0.3930305626963725,
+      "grad_norm": 12.044055303181375,
+      "learning_rate": 2e-05,
+      "loss": 0.7233,
+      "step": 16512
+    },
+    {
+      "epoch": 0.393054365419404,
+      "grad_norm": 1.2438070068370648,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 16513
+    },
+    {
+      "epoch": 0.3930781681424355,
+      "grad_norm": 1.9450984613815343,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 16514
+    },
+    {
+      "epoch": 0.393101970865467,
+      "grad_norm": 1.1759466137305448,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 16515
+    },
+    {
+      "epoch": 0.39312577358849854,
+      "grad_norm": 0.98127931287464,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 16516
+    },
+    {
+      "epoch": 0.39314957631153,
+      "grad_norm": 7.397427087568994,
+      "learning_rate": 2e-05,
+      "loss": 0.3083,
+      "step": 16517
+    },
+    {
+      "epoch": 0.39317337903456157,
+      "grad_norm": 1.9904311836967605,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 16518
+    },
+    {
+      "epoch": 0.39319718175759305,
+      "grad_norm": 6.540436975461449,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 16519
+    },
+    {
+      "epoch": 0.3932209844806246,
+      "grad_norm": 0.8728477866812774,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 16520
+    },
+    {
+      "epoch": 0.3932447872036561,
+      "grad_norm": 3.625667609367298,
+      "learning_rate": 2e-05,
+      "loss": 0.3337,
+      "step": 16521
+    },
+    {
+      "epoch": 0.3932685899266876,
+      "grad_norm": 4.78849792400469,
+      "learning_rate": 2e-05,
+      "loss": 0.722,
+      "step": 16522
+    },
+    {
+      "epoch": 0.3932923926497191,
+      "grad_norm": 6.502470563471331,
+      "learning_rate": 2e-05,
+      "loss": 1.2674,
+      "step": 16523
+    },
+    {
+      "epoch": 0.39331619537275064,
+      "grad_norm": 1.0746962252681846,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 16524
+    },
+    {
+      "epoch": 0.3933399980957822,
+      "grad_norm": 5.196602589020472,
+      "learning_rate": 2e-05,
+      "loss": 0.8755,
+      "step": 16525
+    },
+    {
+      "epoch": 0.39336380081881367,
+      "grad_norm": 4.490718550758032,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 16526
+    },
+    {
+      "epoch": 0.3933876035418452,
+      "grad_norm": 1.95059095254851,
+      "learning_rate": 2e-05,
+      "loss": 0.234,
+      "step": 16527
+    },
+    {
+      "epoch": 0.3934114062648767,
+      "grad_norm": 0.9471106139234259,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 16528
+    },
+    {
+      "epoch": 0.39343520898790824,
+      "grad_norm": 3.8993354347358182,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 16529
+    },
+    {
+      "epoch": 0.3934590117109397,
+      "grad_norm": 3.975589089249363,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 16530
+    },
+    {
+      "epoch": 0.39348281443397126,
+      "grad_norm": 1.366410203021397,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 16531
+    },
+    {
+      "epoch": 0.39350661715700275,
+      "grad_norm": 7.359960057003576,
+      "learning_rate": 2e-05,
+      "loss": 0.3132,
+      "step": 16532
+    },
+    {
+      "epoch": 0.3935304198800343,
+      "grad_norm": 4.1090607135103,
+      "learning_rate": 2e-05,
+      "loss": 0.655,
+      "step": 16533
+    },
+    {
+      "epoch": 0.3935542226030658,
+      "grad_norm": 4.133491262953085,
+      "learning_rate": 2e-05,
+      "loss": 0.4427,
+      "step": 16534
+    },
+    {
+      "epoch": 0.3935780253260973,
+      "grad_norm": 5.637110421860333,
+      "learning_rate": 2e-05,
+      "loss": 0.9022,
+      "step": 16535
+    },
+    {
+      "epoch": 0.3936018280491288,
+      "grad_norm": 10.887938039043524,
+      "learning_rate": 2e-05,
+      "loss": 0.3499,
+      "step": 16536
+    },
+    {
+      "epoch": 0.39362563077216034,
+      "grad_norm": 4.30064207275377,
+      "learning_rate": 2e-05,
+      "loss": 0.2738,
+      "step": 16537
+    },
+    {
+      "epoch": 0.3936494334951918,
+      "grad_norm": 6.473189251787248,
+      "learning_rate": 2e-05,
+      "loss": 1.1163,
+      "step": 16538
+    },
+    {
+      "epoch": 0.39367323621822337,
+      "grad_norm": 3.39938378724354,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 16539
+    },
+    {
+      "epoch": 0.3936970389412549,
+      "grad_norm": 4.1898043962101355,
+      "learning_rate": 2e-05,
+      "loss": 0.8519,
+      "step": 16540
+    },
+    {
+      "epoch": 0.3937208416642864,
+      "grad_norm": 4.551393224289413,
+      "learning_rate": 2e-05,
+      "loss": 0.2407,
+      "step": 16541
+    },
+    {
+      "epoch": 0.39374464438731793,
+      "grad_norm": 5.639051419609824,
+      "learning_rate": 2e-05,
+      "loss": 0.6092,
+      "step": 16542
+    },
+    {
+      "epoch": 0.3937684471103494,
+      "grad_norm": 2.899567795519499,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 16543
+    },
+    {
+      "epoch": 0.39379224983338096,
+      "grad_norm": 2.717988200171966,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 16544
+    },
+    {
+      "epoch": 0.39381605255641244,
+      "grad_norm": 11.433565664281105,
+      "learning_rate": 2e-05,
+      "loss": 0.6197,
+      "step": 16545
+    },
+    {
+      "epoch": 0.393839855279444,
+      "grad_norm": 1.1633606497608582,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 16546
+    },
+    {
+      "epoch": 0.39386365800247547,
+      "grad_norm": 4.3842386469448105,
+      "learning_rate": 2e-05,
+      "loss": 0.4684,
+      "step": 16547
+    },
+    {
+      "epoch": 0.393887460725507,
+      "grad_norm": 2.0431844805575974,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 16548
+    },
+    {
+      "epoch": 0.3939112634485385,
+      "grad_norm": 5.010337774656584,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 16549
+    },
+    {
+      "epoch": 0.39393506617157004,
+      "grad_norm": 4.1988326528399265,
+      "learning_rate": 2e-05,
+      "loss": 0.7182,
+      "step": 16550
+    },
+    {
+      "epoch": 0.3939588688946015,
+      "grad_norm": 5.033514809533071,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 16551
+    },
+    {
+      "epoch": 0.39398267161763306,
+      "grad_norm": 3.9868071629213238,
+      "learning_rate": 2e-05,
+      "loss": 0.447,
+      "step": 16552
+    },
+    {
+      "epoch": 0.39400647434066455,
+      "grad_norm": 3.839621931794345,
+      "learning_rate": 2e-05,
+      "loss": 0.223,
+      "step": 16553
+    },
+    {
+      "epoch": 0.3940302770636961,
+      "grad_norm": 4.102652063774881,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 16554
+    },
+    {
+      "epoch": 0.39405407978672763,
+      "grad_norm": 3.3882321875131303,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 16555
+    },
+    {
+      "epoch": 0.3940778825097591,
+      "grad_norm": 0.9909574175976021,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 16556
+    },
+    {
+      "epoch": 0.39410168523279066,
+      "grad_norm": 0.14244860204956294,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 16557
+    },
+    {
+      "epoch": 0.39412548795582214,
+      "grad_norm": 3.8659063520610384,
+      "learning_rate": 2e-05,
+      "loss": 0.6698,
+      "step": 16558
+    },
+    {
+      "epoch": 0.3941492906788537,
+      "grad_norm": 10.610180910980386,
+      "learning_rate": 2e-05,
+      "loss": 0.8918,
+      "step": 16559
+    },
+    {
+      "epoch": 0.39417309340188517,
+      "grad_norm": 1.6723332694629092,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 16560
+    },
+    {
+      "epoch": 0.3941968961249167,
+      "grad_norm": 2.939897480626229,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 16561
+    },
+    {
+      "epoch": 0.3942206988479482,
+      "grad_norm": 2.011878541870378,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 16562
+    },
+    {
+      "epoch": 0.39424450157097973,
+      "grad_norm": 0.40416852782827356,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 16563
+    },
+    {
+      "epoch": 0.3942683042940112,
+      "grad_norm": 2.1043205474444058,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 16564
+    },
+    {
+      "epoch": 0.39429210701704276,
+      "grad_norm": 4.964682465585912,
+      "learning_rate": 2e-05,
+      "loss": 0.7672,
+      "step": 16565
+    },
+    {
+      "epoch": 0.39431590974007424,
+      "grad_norm": 2.9454021984329,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 16566
+    },
+    {
+      "epoch": 0.3943397124631058,
+      "grad_norm": 6.114128540566572,
+      "learning_rate": 2e-05,
+      "loss": 0.2385,
+      "step": 16567
+    },
+    {
+      "epoch": 0.39436351518613727,
+      "grad_norm": 5.125180594242858,
+      "learning_rate": 2e-05,
+      "loss": 1.1167,
+      "step": 16568
+    },
+    {
+      "epoch": 0.3943873179091688,
+      "grad_norm": 3.6118192139581318,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 16569
+    },
+    {
+      "epoch": 0.3944111206322003,
+      "grad_norm": 5.477290358885595,
+      "learning_rate": 2e-05,
+      "loss": 0.7566,
+      "step": 16570
+    },
+    {
+      "epoch": 0.39443492335523184,
+      "grad_norm": 2.261319506394894,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 16571
+    },
+    {
+      "epoch": 0.3944587260782634,
+      "grad_norm": 1.114896502889278,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 16572
+    },
+    {
+      "epoch": 0.39448252880129486,
+      "grad_norm": 2.0170400602541183,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 16573
+    },
+    {
+      "epoch": 0.3945063315243264,
+      "grad_norm": 4.6582211136483345,
+      "learning_rate": 2e-05,
+      "loss": 0.4788,
+      "step": 16574
+    },
+    {
+      "epoch": 0.3945301342473579,
+      "grad_norm": 3.7844593737694914,
+      "learning_rate": 2e-05,
+      "loss": 0.5163,
+      "step": 16575
+    },
+    {
+      "epoch": 0.39455393697038943,
+      "grad_norm": 8.685490001591674,
+      "learning_rate": 2e-05,
+      "loss": 0.7233,
+      "step": 16576
+    },
+    {
+      "epoch": 0.3945777396934209,
+      "grad_norm": 6.550696738024453,
+      "learning_rate": 2e-05,
+      "loss": 0.6387,
+      "step": 16577
+    },
+    {
+      "epoch": 0.39460154241645246,
+      "grad_norm": 2.3620721015578274,
+      "learning_rate": 2e-05,
+      "loss": 0.1225,
+      "step": 16578
+    },
+    {
+      "epoch": 0.39462534513948394,
+      "grad_norm": 6.412895526053337,
+      "learning_rate": 2e-05,
+      "loss": 0.914,
+      "step": 16579
+    },
+    {
+      "epoch": 0.3946491478625155,
+      "grad_norm": 1.3564841055264976,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 16580
+    },
+    {
+      "epoch": 0.39467295058554697,
+      "grad_norm": 1.8869104675765243,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 16581
+    },
+    {
+      "epoch": 0.3946967533085785,
+      "grad_norm": 4.964391404912342,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 16582
+    },
+    {
+      "epoch": 0.39472055603161,
+      "grad_norm": 2.3432827492266046,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 16583
+    },
+    {
+      "epoch": 0.39474435875464153,
+      "grad_norm": 1268.798789940187,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 16584
+    },
+    {
+      "epoch": 0.394768161477673,
+      "grad_norm": 12.800750183488873,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 16585
+    },
+    {
+      "epoch": 0.39479196420070456,
+      "grad_norm": 6.12029104256836,
+      "learning_rate": 2e-05,
+      "loss": 0.6471,
+      "step": 16586
+    },
+    {
+      "epoch": 0.3948157669237361,
+      "grad_norm": 1.29824995911599,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 16587
+    },
+    {
+      "epoch": 0.3948395696467676,
+      "grad_norm": 2.421886428793542,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 16588
+    },
+    {
+      "epoch": 0.3948633723697991,
+      "grad_norm": 1.1220947560140715,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 16589
+    },
+    {
+      "epoch": 0.3948871750928306,
+      "grad_norm": 1.1208006559421886,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39491097781586215,
+      "grad_norm": 4.189491120722818,
+      "learning_rate": 2e-05,
+      "loss": 0.2468,
+      "step": 16591
+    },
+    {
+      "epoch": 0.39493478053889364,
+      "grad_norm": 5.01510021974469,
+      "learning_rate": 2e-05,
+      "loss": 0.9275,
+      "step": 16592
+    },
+    {
+      "epoch": 0.3949585832619252,
+      "grad_norm": 18.518461440605858,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 16593
+    },
+    {
+      "epoch": 0.39498238598495666,
+      "grad_norm": 4.948147554238063,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 16594
+    },
+    {
+      "epoch": 0.3950061887079882,
+      "grad_norm": 1.948063686306852,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 16595
+    },
+    {
+      "epoch": 0.3950299914310197,
+      "grad_norm": 0.7178730995876014,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 16596
+    },
+    {
+      "epoch": 0.39505379415405123,
+      "grad_norm": 0.8605098067346987,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 16597
+    },
+    {
+      "epoch": 0.3950775968770827,
+      "grad_norm": 3.8629255388884616,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 16598
+    },
+    {
+      "epoch": 0.39510139960011426,
+      "grad_norm": 4.359736514393117,
+      "learning_rate": 2e-05,
+      "loss": 0.4634,
+      "step": 16599
+    },
+    {
+      "epoch": 0.39512520232314574,
+      "grad_norm": 4.198912602267961,
+      "learning_rate": 2e-05,
+      "loss": 0.9599,
+      "step": 16600
+    },
+    {
+      "epoch": 0.3951490050461773,
+      "grad_norm": 1.6851321287688497,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 16601
+    },
+    {
+      "epoch": 0.3951728077692088,
+      "grad_norm": 14.005961543476126,
+      "learning_rate": 2e-05,
+      "loss": 0.7854,
+      "step": 16602
+    },
+    {
+      "epoch": 0.3951966104922403,
+      "grad_norm": 4.610441285565743,
+      "learning_rate": 2e-05,
+      "loss": 0.5904,
+      "step": 16603
+    },
+    {
+      "epoch": 0.39522041321527185,
+      "grad_norm": 9.562305455992067,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 16604
+    },
+    {
+      "epoch": 0.39524421593830333,
+      "grad_norm": 2.0718542387435726,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 16605
+    },
+    {
+      "epoch": 0.3952680186613349,
+      "grad_norm": 4.056528142060924,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 16606
+    },
+    {
+      "epoch": 0.39529182138436636,
+      "grad_norm": 1.426832484778943,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 16607
+    },
+    {
+      "epoch": 0.3953156241073979,
+      "grad_norm": 3.488576097541299,
+      "learning_rate": 2e-05,
+      "loss": 0.1665,
+      "step": 16608
+    },
+    {
+      "epoch": 0.3953394268304294,
+      "grad_norm": 4.021046707728788,
+      "learning_rate": 2e-05,
+      "loss": 0.5071,
+      "step": 16609
+    },
+    {
+      "epoch": 0.3953632295534609,
+      "grad_norm": 0.15311448105868977,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 16610
+    },
+    {
+      "epoch": 0.3953870322764924,
+      "grad_norm": 8.470800481715354,
+      "learning_rate": 2e-05,
+      "loss": 1.1627,
+      "step": 16611
+    },
+    {
+      "epoch": 0.39541083499952395,
+      "grad_norm": 5.289817431006793,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 16612
+    },
+    {
+      "epoch": 0.39543463772255544,
+      "grad_norm": 6.669490847812891,
+      "learning_rate": 2e-05,
+      "loss": 1.5599,
+      "step": 16613
+    },
+    {
+      "epoch": 0.395458440445587,
+      "grad_norm": 6.581452640704577,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 16614
+    },
+    {
+      "epoch": 0.39548224316861846,
+      "grad_norm": 7.190563507872477,
+      "learning_rate": 2e-05,
+      "loss": 0.6772,
+      "step": 16615
+    },
+    {
+      "epoch": 0.39550604589165,
+      "grad_norm": 1.0371217332638998,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 16616
+    },
+    {
+      "epoch": 0.39552984861468155,
+      "grad_norm": 13.109051716290772,
+      "learning_rate": 2e-05,
+      "loss": 0.6315,
+      "step": 16617
+    },
+    {
+      "epoch": 0.39555365133771303,
+      "grad_norm": 1.2207572022699955,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 16618
+    },
+    {
+      "epoch": 0.39557745406074457,
+      "grad_norm": 6.341104074519262,
+      "learning_rate": 2e-05,
+      "loss": 0.5884,
+      "step": 16619
+    },
+    {
+      "epoch": 0.39560125678377606,
+      "grad_norm": 1.7199494214023636,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 16620
+    },
+    {
+      "epoch": 0.3956250595068076,
+      "grad_norm": 2.2864906274136905,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 16621
+    },
+    {
+      "epoch": 0.3956488622298391,
+      "grad_norm": 3.0472919720816436,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 16622
+    },
+    {
+      "epoch": 0.3956726649528706,
+      "grad_norm": 4.637864356553862,
+      "learning_rate": 2e-05,
+      "loss": 0.1992,
+      "step": 16623
+    },
+    {
+      "epoch": 0.3956964676759021,
+      "grad_norm": 2.2989446345854767,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 16624
+    },
+    {
+      "epoch": 0.39572027039893365,
+      "grad_norm": 5.630038634895152,
+      "learning_rate": 2e-05,
+      "loss": 0.9478,
+      "step": 16625
+    },
+    {
+      "epoch": 0.39574407312196513,
+      "grad_norm": 13.480205447486899,
+      "learning_rate": 2e-05,
+      "loss": 1.2613,
+      "step": 16626
+    },
+    {
+      "epoch": 0.3957678758449967,
+      "grad_norm": 2.330161885445798,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 16627
+    },
+    {
+      "epoch": 0.39579167856802816,
+      "grad_norm": 3.082702775777075,
+      "learning_rate": 2e-05,
+      "loss": 0.3054,
+      "step": 16628
+    },
+    {
+      "epoch": 0.3958154812910597,
+      "grad_norm": 6.106269839924542,
+      "learning_rate": 2e-05,
+      "loss": 1.5695,
+      "step": 16629
+    },
+    {
+      "epoch": 0.3958392840140912,
+      "grad_norm": 2.3036690962455983,
+      "learning_rate": 2e-05,
+      "loss": 0.2279,
+      "step": 16630
+    },
+    {
+      "epoch": 0.3958630867371227,
+      "grad_norm": 4.3493178301167355,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 16631
+    },
+    {
+      "epoch": 0.39588688946015427,
+      "grad_norm": 2.702891331332379,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 16632
+    },
+    {
+      "epoch": 0.39591069218318575,
+      "grad_norm": 0.684996874618658,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 16633
+    },
+    {
+      "epoch": 0.3959344949062173,
+      "grad_norm": 6.883468076473674,
+      "learning_rate": 2e-05,
+      "loss": 0.9308,
+      "step": 16634
+    },
+    {
+      "epoch": 0.3959582976292488,
+      "grad_norm": 4.197616381197077,
+      "learning_rate": 2e-05,
+      "loss": 0.3395,
+      "step": 16635
+    },
+    {
+      "epoch": 0.3959821003522803,
+      "grad_norm": 0.11305528691554528,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 16636
+    },
+    {
+      "epoch": 0.3960059030753118,
+      "grad_norm": 11.693941464473015,
+      "learning_rate": 2e-05,
+      "loss": 0.3179,
+      "step": 16637
+    },
+    {
+      "epoch": 0.39602970579834335,
+      "grad_norm": 0.9706778096262432,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 16638
+    },
+    {
+      "epoch": 0.39605350852137483,
+      "grad_norm": 5.013859523450196,
+      "learning_rate": 2e-05,
+      "loss": 0.8463,
+      "step": 16639
+    },
+    {
+      "epoch": 0.39607731124440637,
+      "grad_norm": 7.268453221308416,
+      "learning_rate": 2e-05,
+      "loss": 0.6418,
+      "step": 16640
+    },
+    {
+      "epoch": 0.39610111396743786,
+      "grad_norm": 0.44530768855033726,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 16641
+    },
+    {
+      "epoch": 0.3961249166904694,
+      "grad_norm": 8.032150535936385,
+      "learning_rate": 2e-05,
+      "loss": 1.4911,
+      "step": 16642
+    },
+    {
+      "epoch": 0.3961487194135009,
+      "grad_norm": 4.460293078962346,
+      "learning_rate": 2e-05,
+      "loss": 0.1242,
+      "step": 16643
+    },
+    {
+      "epoch": 0.3961725221365324,
+      "grad_norm": 0.08274402173702347,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 16644
+    },
+    {
+      "epoch": 0.3961963248595639,
+      "grad_norm": 11.401114110942066,
+      "learning_rate": 2e-05,
+      "loss": 0.6365,
+      "step": 16645
+    },
+    {
+      "epoch": 0.39622012758259545,
+      "grad_norm": 9.401259160319547,
+      "learning_rate": 2e-05,
+      "loss": 0.5233,
+      "step": 16646
+    },
+    {
+      "epoch": 0.396243930305627,
+      "grad_norm": 1.6341025952109656,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 16647
+    },
+    {
+      "epoch": 0.3962677330286585,
+      "grad_norm": 5.191412469695593,
+      "learning_rate": 2e-05,
+      "loss": 0.2712,
+      "step": 16648
+    },
+    {
+      "epoch": 0.39629153575169,
+      "grad_norm": 0.0693660543433656,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 16649
+    },
+    {
+      "epoch": 0.3963153384747215,
+      "grad_norm": 25.27641594431697,
+      "learning_rate": 2e-05,
+      "loss": 1.7364,
+      "step": 16650
+    },
+    {
+      "epoch": 0.39633914119775304,
+      "grad_norm": 2.2468621015509584,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 16651
+    },
+    {
+      "epoch": 0.3963629439207845,
+      "grad_norm": 0.7856814366313517,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 16652
+    },
+    {
+      "epoch": 0.39638674664381607,
+      "grad_norm": 2.3957220611024046,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 16653
+    },
+    {
+      "epoch": 0.39641054936684755,
+      "grad_norm": 0.1368085997542094,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16654
+    },
+    {
+      "epoch": 0.3964343520898791,
+      "grad_norm": 4.609900704341501,
+      "learning_rate": 2e-05,
+      "loss": 0.707,
+      "step": 16655
+    },
+    {
+      "epoch": 0.3964581548129106,
+      "grad_norm": 0.6171829901478787,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 16656
+    },
+    {
+      "epoch": 0.3964819575359421,
+      "grad_norm": 5.885358879920093,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 16657
+    },
+    {
+      "epoch": 0.3965057602589736,
+      "grad_norm": 3.690049387149011,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 16658
+    },
+    {
+      "epoch": 0.39652956298200515,
+      "grad_norm": 2.876691991723184,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 16659
+    },
+    {
+      "epoch": 0.39655336570503663,
+      "grad_norm": 2.7301961005086763,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 16660
+    },
+    {
+      "epoch": 0.39657716842806817,
+      "grad_norm": 3.024329324136286,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 16661
+    },
+    {
+      "epoch": 0.3966009711510997,
+      "grad_norm": 10.013940872883103,
+      "learning_rate": 2e-05,
+      "loss": 1.252,
+      "step": 16662
+    },
+    {
+      "epoch": 0.3966247738741312,
+      "grad_norm": 3.6438779865063515,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 16663
+    },
+    {
+      "epoch": 0.39664857659716274,
+      "grad_norm": 1.002827773457212,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 16664
+    },
+    {
+      "epoch": 0.3966723793201942,
+      "grad_norm": 2.4852297740841474,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 16665
+    },
+    {
+      "epoch": 0.39669618204322576,
+      "grad_norm": 4.077690570196148,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 16666
+    },
+    {
+      "epoch": 0.39671998476625725,
+      "grad_norm": 0.500864766488666,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 16667
+    },
+    {
+      "epoch": 0.3967437874892888,
+      "grad_norm": 4.606295648031068,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 16668
+    },
+    {
+      "epoch": 0.3967675902123203,
+      "grad_norm": 4.5470908990916294,
+      "learning_rate": 2e-05,
+      "loss": 0.8349,
+      "step": 16669
+    },
+    {
+      "epoch": 0.3967913929353518,
+      "grad_norm": 2.598518142295225,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 16670
+    },
+    {
+      "epoch": 0.3968151956583833,
+      "grad_norm": 7.004658138391038,
+      "learning_rate": 2e-05,
+      "loss": 0.1698,
+      "step": 16671
+    },
+    {
+      "epoch": 0.39683899838141484,
+      "grad_norm": 6.101253683059943,
+      "learning_rate": 2e-05,
+      "loss": 1.0788,
+      "step": 16672
+    },
+    {
+      "epoch": 0.3968628011044463,
+      "grad_norm": 6.3872980085966615,
+      "learning_rate": 2e-05,
+      "loss": 1.1417,
+      "step": 16673
+    },
+    {
+      "epoch": 0.39688660382747787,
+      "grad_norm": 4.716143765425121,
+      "learning_rate": 2e-05,
+      "loss": 0.5239,
+      "step": 16674
+    },
+    {
+      "epoch": 0.39691040655050935,
+      "grad_norm": 3.0387235496974654,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 16675
+    },
+    {
+      "epoch": 0.3969342092735409,
+      "grad_norm": 3.292697829675708,
+      "learning_rate": 2e-05,
+      "loss": 0.3745,
+      "step": 16676
+    },
+    {
+      "epoch": 0.39695801199657244,
+      "grad_norm": 4.4915603855410415,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 16677
+    },
+    {
+      "epoch": 0.3969818147196039,
+      "grad_norm": 1.6345044196275562,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 16678
+    },
+    {
+      "epoch": 0.39700561744263546,
+      "grad_norm": 11.941438449026698,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 16679
+    },
+    {
+      "epoch": 0.39702942016566695,
+      "grad_norm": 0.8766468995321011,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 16680
+    },
+    {
+      "epoch": 0.3970532228886985,
+      "grad_norm": 8.483491180543762,
+      "learning_rate": 2e-05,
+      "loss": 0.4532,
+      "step": 16681
+    },
+    {
+      "epoch": 0.39707702561173,
+      "grad_norm": 0.16338955974559147,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 16682
+    },
+    {
+      "epoch": 0.3971008283347615,
+      "grad_norm": 6.424916508753419,
+      "learning_rate": 2e-05,
+      "loss": 1.212,
+      "step": 16683
+    },
+    {
+      "epoch": 0.397124631057793,
+      "grad_norm": 2.0289885029664796,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 16684
+    },
+    {
+      "epoch": 0.39714843378082454,
+      "grad_norm": 0.5395012998258735,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 16685
+    },
+    {
+      "epoch": 0.397172236503856,
+      "grad_norm": 16.04873567594858,
+      "learning_rate": 2e-05,
+      "loss": 0.6154,
+      "step": 16686
+    },
+    {
+      "epoch": 0.39719603922688757,
+      "grad_norm": 2.008712680817932,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 16687
+    },
+    {
+      "epoch": 0.39721984194991905,
+      "grad_norm": 19.81309297168516,
+      "learning_rate": 2e-05,
+      "loss": 2.1184,
+      "step": 16688
+    },
+    {
+      "epoch": 0.3972436446729506,
+      "grad_norm": 5.600110877772073,
+      "learning_rate": 2e-05,
+      "loss": 1.0872,
+      "step": 16689
+    },
+    {
+      "epoch": 0.3972674473959821,
+      "grad_norm": 3.1735218736813833,
+      "learning_rate": 2e-05,
+      "loss": 0.4039,
+      "step": 16690
+    },
+    {
+      "epoch": 0.3972912501190136,
+      "grad_norm": 7.521910110380365,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 16691
+    },
+    {
+      "epoch": 0.39731505284204516,
+      "grad_norm": 4.206524989155287,
+      "learning_rate": 2e-05,
+      "loss": 0.9843,
+      "step": 16692
+    },
+    {
+      "epoch": 0.39733885556507664,
+      "grad_norm": 0.5424480992859484,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16693
+    },
+    {
+      "epoch": 0.3973626582881082,
+      "grad_norm": 6.102240618527793,
+      "learning_rate": 2e-05,
+      "loss": 0.6657,
+      "step": 16694
+    },
+    {
+      "epoch": 0.39738646101113967,
+      "grad_norm": 2.6949496203498584,
+      "learning_rate": 2e-05,
+      "loss": 0.1302,
+      "step": 16695
+    },
+    {
+      "epoch": 0.3974102637341712,
+      "grad_norm": 6.708109990412152,
+      "learning_rate": 2e-05,
+      "loss": 0.7747,
+      "step": 16696
+    },
+    {
+      "epoch": 0.3974340664572027,
+      "grad_norm": 11.91084037639964,
+      "learning_rate": 2e-05,
+      "loss": 1.679,
+      "step": 16697
+    },
+    {
+      "epoch": 0.39745786918023424,
+      "grad_norm": 8.658264216858424,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 16698
+    },
+    {
+      "epoch": 0.3974816719032657,
+      "grad_norm": 13.47987549169215,
+      "learning_rate": 2e-05,
+      "loss": 1.2498,
+      "step": 16699
+    },
+    {
+      "epoch": 0.39750547462629726,
+      "grad_norm": 5.400859010995457,
+      "learning_rate": 2e-05,
+      "loss": 1.1015,
+      "step": 16700
+    },
+    {
+      "epoch": 0.39752927734932875,
+      "grad_norm": 3.1286105532338553,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 16701
+    },
+    {
+      "epoch": 0.3975530800723603,
+      "grad_norm": 1.1864998704511316,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 16702
+    },
+    {
+      "epoch": 0.3975768827953918,
+      "grad_norm": 7.274615958560769,
+      "learning_rate": 2e-05,
+      "loss": 0.2414,
+      "step": 16703
+    },
+    {
+      "epoch": 0.3976006855184233,
+      "grad_norm": 2.470666191398236,
+      "learning_rate": 2e-05,
+      "loss": 0.274,
+      "step": 16704
+    },
+    {
+      "epoch": 0.3976244882414548,
+      "grad_norm": 7.39101116054934,
+      "learning_rate": 2e-05,
+      "loss": 1.4282,
+      "step": 16705
+    },
+    {
+      "epoch": 0.39764829096448634,
+      "grad_norm": 2.1141662152262355,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 16706
+    },
+    {
+      "epoch": 0.3976720936875178,
+      "grad_norm": 3.437298868749237,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 16707
+    },
+    {
+      "epoch": 0.39769589641054937,
+      "grad_norm": 1.1507411675195616,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 16708
+    },
+    {
+      "epoch": 0.3977196991335809,
+      "grad_norm": 4.01245802252974,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 16709
+    },
+    {
+      "epoch": 0.3977435018566124,
+      "grad_norm": 2.4044445241387997,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 16710
+    },
+    {
+      "epoch": 0.39776730457964393,
+      "grad_norm": 5.470169053427628,
+      "learning_rate": 2e-05,
+      "loss": 0.6354,
+      "step": 16711
+    },
+    {
+      "epoch": 0.3977911073026754,
+      "grad_norm": 6.59751924149489,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 16712
+    },
+    {
+      "epoch": 0.39781491002570696,
+      "grad_norm": 2.0623483863279737,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 16713
+    },
+    {
+      "epoch": 0.39783871274873844,
+      "grad_norm": 19.83984461630451,
+      "learning_rate": 2e-05,
+      "loss": 0.2361,
+      "step": 16714
+    },
+    {
+      "epoch": 0.39786251547177,
+      "grad_norm": 2.9940296920872727,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 16715
+    },
+    {
+      "epoch": 0.39788631819480147,
+      "grad_norm": 7.863027381671858,
+      "learning_rate": 2e-05,
+      "loss": 1.1517,
+      "step": 16716
+    },
+    {
+      "epoch": 0.397910120917833,
+      "grad_norm": 4.496989441084256,
+      "learning_rate": 2e-05,
+      "loss": 0.5816,
+      "step": 16717
+    },
+    {
+      "epoch": 0.3979339236408645,
+      "grad_norm": 0.9955454832352482,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 16718
+    },
+    {
+      "epoch": 0.39795772636389604,
+      "grad_norm": 4.003418317312536,
+      "learning_rate": 2e-05,
+      "loss": 0.6658,
+      "step": 16719
+    },
+    {
+      "epoch": 0.3979815290869275,
+      "grad_norm": 2.1407117073134856,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 16720
+    },
+    {
+      "epoch": 0.39800533180995906,
+      "grad_norm": 3.7101868733037864,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 16721
+    },
+    {
+      "epoch": 0.39802913453299055,
+      "grad_norm": 2.597565672234288,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 16722
+    },
+    {
+      "epoch": 0.3980529372560221,
+      "grad_norm": 4.54640889822889,
+      "learning_rate": 2e-05,
+      "loss": 0.2146,
+      "step": 16723
+    },
+    {
+      "epoch": 0.39807673997905363,
+      "grad_norm": 4.873368774054521,
+      "learning_rate": 2e-05,
+      "loss": 0.2356,
+      "step": 16724
+    },
+    {
+      "epoch": 0.3981005427020851,
+      "grad_norm": 0.9393697751651107,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 16725
+    },
+    {
+      "epoch": 0.39812434542511665,
+      "grad_norm": 2.605987693662342,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 16726
+    },
+    {
+      "epoch": 0.39814814814814814,
+      "grad_norm": 4.790094853912132,
+      "learning_rate": 2e-05,
+      "loss": 0.5933,
+      "step": 16727
+    },
+    {
+      "epoch": 0.3981719508711797,
+      "grad_norm": 6.191224358581755,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 16728
+    },
+    {
+      "epoch": 0.39819575359421117,
+      "grad_norm": 6.775463560061387,
+      "learning_rate": 2e-05,
+      "loss": 0.2482,
+      "step": 16729
+    },
+    {
+      "epoch": 0.3982195563172427,
+      "grad_norm": 3.189967190503795,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 16730
+    },
+    {
+      "epoch": 0.3982433590402742,
+      "grad_norm": 3.2430349793261013,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 16731
+    },
+    {
+      "epoch": 0.39826716176330573,
+      "grad_norm": 10.5598285185867,
+      "learning_rate": 2e-05,
+      "loss": 0.8839,
+      "step": 16732
+    },
+    {
+      "epoch": 0.3982909644863372,
+      "grad_norm": 6.313229558472348,
+      "learning_rate": 2e-05,
+      "loss": 1.0015,
+      "step": 16733
+    },
+    {
+      "epoch": 0.39831476720936876,
+      "grad_norm": 4.851669652321706,
+      "learning_rate": 2e-05,
+      "loss": 1.1115,
+      "step": 16734
+    },
+    {
+      "epoch": 0.39833856993240024,
+      "grad_norm": 10.69073573681073,
+      "learning_rate": 2e-05,
+      "loss": 0.2412,
+      "step": 16735
+    },
+    {
+      "epoch": 0.3983623726554318,
+      "grad_norm": 5.546170020856094,
+      "learning_rate": 2e-05,
+      "loss": 0.7627,
+      "step": 16736
+    },
+    {
+      "epoch": 0.39838617537846327,
+      "grad_norm": 4.0529924725151725,
+      "learning_rate": 2e-05,
+      "loss": 0.6466,
+      "step": 16737
+    },
+    {
+      "epoch": 0.3984099781014948,
+      "grad_norm": 2.719902307755631,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 16738
+    },
+    {
+      "epoch": 0.39843378082452635,
+      "grad_norm": 3.474073276383173,
+      "learning_rate": 2e-05,
+      "loss": 0.6293,
+      "step": 16739
+    },
+    {
+      "epoch": 0.39845758354755784,
+      "grad_norm": 5.796917339581876,
+      "learning_rate": 2e-05,
+      "loss": 0.4961,
+      "step": 16740
+    },
+    {
+      "epoch": 0.3984813862705894,
+      "grad_norm": 14.250690495388717,
+      "learning_rate": 2e-05,
+      "loss": 0.342,
+      "step": 16741
+    },
+    {
+      "epoch": 0.39850518899362086,
+      "grad_norm": 4.201734086341756,
+      "learning_rate": 2e-05,
+      "loss": 0.4458,
+      "step": 16742
+    },
+    {
+      "epoch": 0.3985289917166524,
+      "grad_norm": 12.884633883695722,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 16743
+    },
+    {
+      "epoch": 0.3985527944396839,
+      "grad_norm": 3.9532931188051683,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 16744
+    },
+    {
+      "epoch": 0.39857659716271543,
+      "grad_norm": 2.1848816349095403,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 16745
+    },
+    {
+      "epoch": 0.3986003998857469,
+      "grad_norm": 2.143006919663138,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 16746
+    },
+    {
+      "epoch": 0.39862420260877846,
+      "grad_norm": 4.968343703309742,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 16747
+    },
+    {
+      "epoch": 0.39864800533180994,
+      "grad_norm": 1.553504049527215,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 16748
+    },
+    {
+      "epoch": 0.3986718080548415,
+      "grad_norm": 0.11991216840894071,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 16749
+    },
+    {
+      "epoch": 0.39869561077787297,
+      "grad_norm": 4.550075778951076,
+      "learning_rate": 2e-05,
+      "loss": 0.7207,
+      "step": 16750
+    },
+    {
+      "epoch": 0.3987194135009045,
+      "grad_norm": 3.71121504198877,
+      "learning_rate": 2e-05,
+      "loss": 0.125,
+      "step": 16751
+    },
+    {
+      "epoch": 0.398743216223936,
+      "grad_norm": 3.843644715207796,
+      "learning_rate": 2e-05,
+      "loss": 0.4828,
+      "step": 16752
+    },
+    {
+      "epoch": 0.39876701894696753,
+      "grad_norm": 2.1863925092153407,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 16753
+    },
+    {
+      "epoch": 0.3987908216699991,
+      "grad_norm": 5.227484613504849,
+      "learning_rate": 2e-05,
+      "loss": 0.9004,
+      "step": 16754
+    },
+    {
+      "epoch": 0.39881462439303056,
+      "grad_norm": 4.253716978621566,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 16755
+    },
+    {
+      "epoch": 0.3988384271160621,
+      "grad_norm": 4.086125614462984,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 16756
+    },
+    {
+      "epoch": 0.3988622298390936,
+      "grad_norm": 3.8539560448154178,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 16757
+    },
+    {
+      "epoch": 0.3988860325621251,
+      "grad_norm": 8.722204258994916,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 16758
+    },
+    {
+      "epoch": 0.3989098352851566,
+      "grad_norm": 7.006846356443637,
+      "learning_rate": 2e-05,
+      "loss": 2.2008,
+      "step": 16759
+    },
+    {
+      "epoch": 0.39893363800818815,
+      "grad_norm": 4.120168355055444,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 16760
+    },
+    {
+      "epoch": 0.39895744073121964,
+      "grad_norm": 7.133860563802832,
+      "learning_rate": 2e-05,
+      "loss": 0.2488,
+      "step": 16761
+    },
+    {
+      "epoch": 0.3989812434542512,
+      "grad_norm": 0.911728125077802,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 16762
+    },
+    {
+      "epoch": 0.39900504617728266,
+      "grad_norm": 11.844585593289063,
+      "learning_rate": 2e-05,
+      "loss": 0.5884,
+      "step": 16763
+    },
+    {
+      "epoch": 0.3990288489003142,
+      "grad_norm": 4.015139743615277,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 16764
+    },
+    {
+      "epoch": 0.3990526516233457,
+      "grad_norm": 19.091782000220135,
+      "learning_rate": 2e-05,
+      "loss": 1.7007,
+      "step": 16765
+    },
+    {
+      "epoch": 0.39907645434637723,
+      "grad_norm": 6.381870197597361,
+      "learning_rate": 2e-05,
+      "loss": 0.2325,
+      "step": 16766
+    },
+    {
+      "epoch": 0.3991002570694087,
+      "grad_norm": 0.8574746945081383,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 16767
+    },
+    {
+      "epoch": 0.39912405979244026,
+      "grad_norm": 4.46920545008231,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 16768
+    },
+    {
+      "epoch": 0.3991478625154718,
+      "grad_norm": 1.4830793648991591,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 16769
+    },
+    {
+      "epoch": 0.3991716652385033,
+      "grad_norm": 4.4435464910128015,
+      "learning_rate": 2e-05,
+      "loss": 0.6131,
+      "step": 16770
+    },
+    {
+      "epoch": 0.3991954679615348,
+      "grad_norm": 1.399168514624882,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 16771
+    },
+    {
+      "epoch": 0.3992192706845663,
+      "grad_norm": 4.915716571551971,
+      "learning_rate": 2e-05,
+      "loss": 0.5406,
+      "step": 16772
+    },
+    {
+      "epoch": 0.39924307340759785,
+      "grad_norm": 4.913450249913946,
+      "learning_rate": 2e-05,
+      "loss": 0.6178,
+      "step": 16773
+    },
+    {
+      "epoch": 0.39926687613062933,
+      "grad_norm": 5.598649240252198,
+      "learning_rate": 2e-05,
+      "loss": 0.9797,
+      "step": 16774
+    },
+    {
+      "epoch": 0.3992906788536609,
+      "grad_norm": 1.6241702382062146,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 16775
+    },
+    {
+      "epoch": 0.39931448157669236,
+      "grad_norm": 4.789316295636664,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 16776
+    },
+    {
+      "epoch": 0.3993382842997239,
+      "grad_norm": 4.073137148248996,
+      "learning_rate": 2e-05,
+      "loss": 0.4068,
+      "step": 16777
+    },
+    {
+      "epoch": 0.3993620870227554,
+      "grad_norm": 4.532058372380104,
+      "learning_rate": 2e-05,
+      "loss": 0.0926,
+      "step": 16778
+    },
+    {
+      "epoch": 0.3993858897457869,
+      "grad_norm": 0.7925944190559838,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 16779
+    },
+    {
+      "epoch": 0.3994096924688184,
+      "grad_norm": 8.22016675620697,
+      "learning_rate": 2e-05,
+      "loss": 1.2773,
+      "step": 16780
+    },
+    {
+      "epoch": 0.39943349519184995,
+      "grad_norm": 4.429257363047146,
+      "learning_rate": 2e-05,
+      "loss": 0.1606,
+      "step": 16781
+    },
+    {
+      "epoch": 0.39945729791488144,
+      "grad_norm": 2.9793519230650998,
+      "learning_rate": 2e-05,
+      "loss": 0.5776,
+      "step": 16782
+    },
+    {
+      "epoch": 0.399481100637913,
+      "grad_norm": 1.4976495979210205,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 16783
+    },
+    {
+      "epoch": 0.3995049033609445,
+      "grad_norm": 2.86608060873801,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 16784
+    },
+    {
+      "epoch": 0.399528706083976,
+      "grad_norm": 4.107779400502446,
+      "learning_rate": 2e-05,
+      "loss": 0.1725,
+      "step": 16785
+    },
+    {
+      "epoch": 0.39955250880700754,
+      "grad_norm": 0.08527009155478399,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 16786
+    },
+    {
+      "epoch": 0.39957631153003903,
+      "grad_norm": 4.646675769878493,
+      "learning_rate": 2e-05,
+      "loss": 0.7631,
+      "step": 16787
+    },
+    {
+      "epoch": 0.39960011425307057,
+      "grad_norm": 0.9214138650241204,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 16788
+    },
+    {
+      "epoch": 0.39962391697610206,
+      "grad_norm": 1.769772993863665,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 16789
+    },
+    {
+      "epoch": 0.3996477196991336,
+      "grad_norm": 4.386333821346809,
+      "learning_rate": 2e-05,
+      "loss": 0.7864,
+      "step": 16790
+    },
+    {
+      "epoch": 0.3996715224221651,
+      "grad_norm": 20.193049362041634,
+      "learning_rate": 2e-05,
+      "loss": 0.426,
+      "step": 16791
+    },
+    {
+      "epoch": 0.3996953251451966,
+      "grad_norm": 0.7123092385664288,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 16792
+    },
+    {
+      "epoch": 0.3997191278682281,
+      "grad_norm": 0.8041755747861695,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 16793
+    },
+    {
+      "epoch": 0.39974293059125965,
+      "grad_norm": 7.443451912571963,
+      "learning_rate": 2e-05,
+      "loss": 1.5278,
+      "step": 16794
+    },
+    {
+      "epoch": 0.39976673331429113,
+      "grad_norm": 3.875733484528821,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 16795
+    },
+    {
+      "epoch": 0.3997905360373227,
+      "grad_norm": 5.350920531591122,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 16796
+    },
+    {
+      "epoch": 0.39981433876035416,
+      "grad_norm": 8.954216088355187,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 16797
+    },
+    {
+      "epoch": 0.3998381414833857,
+      "grad_norm": 6.203573650304234,
+      "learning_rate": 2e-05,
+      "loss": 0.627,
+      "step": 16798
+    },
+    {
+      "epoch": 0.39986194420641724,
+      "grad_norm": 7.437464615802891,
+      "learning_rate": 2e-05,
+      "loss": 1.1791,
+      "step": 16799
+    },
+    {
+      "epoch": 0.3998857469294487,
+      "grad_norm": 2.8015977693025094,
+      "learning_rate": 2e-05,
+      "loss": 0.4547,
+      "step": 16800
+    },
+    {
+      "epoch": 0.39990954965248027,
+      "grad_norm": 2.3009568544478776,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 16801
+    },
+    {
+      "epoch": 0.39993335237551175,
+      "grad_norm": 9.756262437161402,
+      "learning_rate": 2e-05,
+      "loss": 0.434,
+      "step": 16802
+    },
+    {
+      "epoch": 0.3999571550985433,
+      "grad_norm": 6.613030731863973,
+      "learning_rate": 2e-05,
+      "loss": 1.1066,
+      "step": 16803
+    },
+    {
+      "epoch": 0.3999809578215748,
+      "grad_norm": 0.978103274537172,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 16804
+    },
+    {
+      "epoch": 0.4000047605446063,
+      "grad_norm": 0.7680958181014663,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 16805
+    },
+    {
+      "epoch": 0.4000285632676378,
+      "grad_norm": 4.5631311654846956,
+      "learning_rate": 2e-05,
+      "loss": 0.7685,
+      "step": 16806
+    },
+    {
+      "epoch": 0.40005236599066935,
+      "grad_norm": 6.776666798984902,
+      "learning_rate": 2e-05,
+      "loss": 0.6686,
+      "step": 16807
+    },
+    {
+      "epoch": 0.40007616871370083,
+      "grad_norm": 5.1958902794753,
+      "learning_rate": 2e-05,
+      "loss": 0.8789,
+      "step": 16808
+    },
+    {
+      "epoch": 0.40009997143673237,
+      "grad_norm": 1.1690972389491345,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 16809
+    },
+    {
+      "epoch": 0.40012377415976386,
+      "grad_norm": 16.483207009603884,
+      "learning_rate": 2e-05,
+      "loss": 0.8253,
+      "step": 16810
+    },
+    {
+      "epoch": 0.4001475768827954,
+      "grad_norm": 4.482406997148126,
+      "learning_rate": 2e-05,
+      "loss": 0.4761,
+      "step": 16811
+    },
+    {
+      "epoch": 0.4001713796058269,
+      "grad_norm": 0.40974124741346685,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 16812
+    },
+    {
+      "epoch": 0.4001951823288584,
+      "grad_norm": 5.771392313499058,
+      "learning_rate": 2e-05,
+      "loss": 0.7039,
+      "step": 16813
+    },
+    {
+      "epoch": 0.40021898505188996,
+      "grad_norm": 6.27697046715575,
+      "learning_rate": 2e-05,
+      "loss": 0.4644,
+      "step": 16814
+    },
+    {
+      "epoch": 0.40024278777492145,
+      "grad_norm": 2.524545598384543,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 16815
+    },
+    {
+      "epoch": 0.400266590497953,
+      "grad_norm": 0.5719854106509552,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 16816
+    },
+    {
+      "epoch": 0.4002903932209845,
+      "grad_norm": 5.215508004644252,
+      "learning_rate": 2e-05,
+      "loss": 0.6926,
+      "step": 16817
+    },
+    {
+      "epoch": 0.400314195944016,
+      "grad_norm": 4.84407801720002,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 16818
+    },
+    {
+      "epoch": 0.4003379986670475,
+      "grad_norm": 2.46712593975436,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 16819
+    },
+    {
+      "epoch": 0.40036180139007904,
+      "grad_norm": 4.610046646930954,
+      "learning_rate": 2e-05,
+      "loss": 0.6272,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4003856041131105,
+      "grad_norm": 29.003930793749568,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 16821
+    },
+    {
+      "epoch": 0.40040940683614207,
+      "grad_norm": 0.9425437397964702,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 16822
+    },
+    {
+      "epoch": 0.40043320955917355,
+      "grad_norm": 1.0267185936468826,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 16823
+    },
+    {
+      "epoch": 0.4004570122822051,
+      "grad_norm": 3.4918416093398092,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 16824
+    },
+    {
+      "epoch": 0.4004808150052366,
+      "grad_norm": 3.7596426460672263,
+      "learning_rate": 2e-05,
+      "loss": 0.5548,
+      "step": 16825
+    },
+    {
+      "epoch": 0.4005046177282681,
+      "grad_norm": 19.408658599724873,
+      "learning_rate": 2e-05,
+      "loss": 0.2818,
+      "step": 16826
+    },
+    {
+      "epoch": 0.4005284204512996,
+      "grad_norm": 3.935486451935161,
+      "learning_rate": 2e-05,
+      "loss": 0.1811,
+      "step": 16827
+    },
+    {
+      "epoch": 0.40055222317433115,
+      "grad_norm": 16.455158462308056,
+      "learning_rate": 2e-05,
+      "loss": 1.0718,
+      "step": 16828
+    },
+    {
+      "epoch": 0.40057602589736263,
+      "grad_norm": 2.829413253105101,
+      "learning_rate": 2e-05,
+      "loss": 0.3781,
+      "step": 16829
+    },
+    {
+      "epoch": 0.40059982862039417,
+      "grad_norm": 4.053732429860124,
+      "learning_rate": 2e-05,
+      "loss": 0.4638,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4006236313434257,
+      "grad_norm": 4.025389293701207,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 16831
+    },
+    {
+      "epoch": 0.4006474340664572,
+      "grad_norm": 5.72504549194792,
+      "learning_rate": 2e-05,
+      "loss": 1.3253,
+      "step": 16832
+    },
+    {
+      "epoch": 0.40067123678948874,
+      "grad_norm": 0.5731991625447265,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 16833
+    },
+    {
+      "epoch": 0.4006950395125202,
+      "grad_norm": 5.048578161821462,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 16834
+    },
+    {
+      "epoch": 0.40071884223555176,
+      "grad_norm": 11.663931676309723,
+      "learning_rate": 2e-05,
+      "loss": 0.8896,
+      "step": 16835
+    },
+    {
+      "epoch": 0.40074264495858325,
+      "grad_norm": 18.111972680044133,
+      "learning_rate": 2e-05,
+      "loss": 1.1801,
+      "step": 16836
+    },
+    {
+      "epoch": 0.4007664476816148,
+      "grad_norm": 0.3951862126742277,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 16837
+    },
+    {
+      "epoch": 0.4007902504046463,
+      "grad_norm": 6.579152912860707,
+      "learning_rate": 2e-05,
+      "loss": 0.5922,
+      "step": 16838
+    },
+    {
+      "epoch": 0.4008140531276778,
+      "grad_norm": 6.619281442126789,
+      "learning_rate": 2e-05,
+      "loss": 0.8869,
+      "step": 16839
+    },
+    {
+      "epoch": 0.4008378558507093,
+      "grad_norm": 9.256282327982003,
+      "learning_rate": 2e-05,
+      "loss": 0.286,
+      "step": 16840
+    },
+    {
+      "epoch": 0.40086165857374084,
+      "grad_norm": 5.798178043006554,
+      "learning_rate": 2e-05,
+      "loss": 1.0409,
+      "step": 16841
+    },
+    {
+      "epoch": 0.4008854612967723,
+      "grad_norm": 3.5385841163882765,
+      "learning_rate": 2e-05,
+      "loss": 0.3555,
+      "step": 16842
+    },
+    {
+      "epoch": 0.40090926401980387,
+      "grad_norm": 0.8345641272165742,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16843
+    },
+    {
+      "epoch": 0.40093306674283535,
+      "grad_norm": 0.13419437440370163,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 16844
+    },
+    {
+      "epoch": 0.4009568694658669,
+      "grad_norm": 2.2646950040403984,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 16845
+    },
+    {
+      "epoch": 0.40098067218889843,
+      "grad_norm": 7.11844769232105,
+      "learning_rate": 2e-05,
+      "loss": 0.7749,
+      "step": 16846
+    },
+    {
+      "epoch": 0.4010044749119299,
+      "grad_norm": 11.94119148893571,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 16847
+    },
+    {
+      "epoch": 0.40102827763496146,
+      "grad_norm": 6.734469267816082,
+      "learning_rate": 2e-05,
+      "loss": 0.8032,
+      "step": 16848
+    },
+    {
+      "epoch": 0.40105208035799295,
+      "grad_norm": 1.0200875486190113,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 16849
+    },
+    {
+      "epoch": 0.4010758830810245,
+      "grad_norm": 10.450807819116035,
+      "learning_rate": 2e-05,
+      "loss": 1.0709,
+      "step": 16850
+    },
+    {
+      "epoch": 0.40109968580405597,
+      "grad_norm": 1.4878826284750604,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 16851
+    },
+    {
+      "epoch": 0.4011234885270875,
+      "grad_norm": 1.9047548562090897,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 16852
+    },
+    {
+      "epoch": 0.401147291250119,
+      "grad_norm": 9.553123114456737,
+      "learning_rate": 2e-05,
+      "loss": 0.6195,
+      "step": 16853
+    },
+    {
+      "epoch": 0.40117109397315054,
+      "grad_norm": 8.253244757746884,
+      "learning_rate": 2e-05,
+      "loss": 0.4376,
+      "step": 16854
+    },
+    {
+      "epoch": 0.401194896696182,
+      "grad_norm": 0.6988101534448584,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 16855
+    },
+    {
+      "epoch": 0.40121869941921356,
+      "grad_norm": 2.9551647764172557,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 16856
+    },
+    {
+      "epoch": 0.40124250214224505,
+      "grad_norm": 4.061319426360927,
+      "learning_rate": 2e-05,
+      "loss": 0.907,
+      "step": 16857
+    },
+    {
+      "epoch": 0.4012663048652766,
+      "grad_norm": 4.87944277025558,
+      "learning_rate": 2e-05,
+      "loss": 0.7102,
+      "step": 16858
+    },
+    {
+      "epoch": 0.4012901075883081,
+      "grad_norm": 7.94062380897124,
+      "learning_rate": 2e-05,
+      "loss": 1.0607,
+      "step": 16859
+    },
+    {
+      "epoch": 0.4013139103113396,
+      "grad_norm": 2.335247844114081,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 16860
+    },
+    {
+      "epoch": 0.40133771303437116,
+      "grad_norm": 23.522026933635722,
+      "learning_rate": 2e-05,
+      "loss": 0.3994,
+      "step": 16861
+    },
+    {
+      "epoch": 0.40136151575740264,
+      "grad_norm": 6.6495205397613795,
+      "learning_rate": 2e-05,
+      "loss": 0.2513,
+      "step": 16862
+    },
+    {
+      "epoch": 0.4013853184804342,
+      "grad_norm": 54.19607605113589,
+      "learning_rate": 2e-05,
+      "loss": 2.818,
+      "step": 16863
+    },
+    {
+      "epoch": 0.40140912120346567,
+      "grad_norm": 69.11720355786787,
+      "learning_rate": 2e-05,
+      "loss": 0.959,
+      "step": 16864
+    },
+    {
+      "epoch": 0.4014329239264972,
+      "grad_norm": 5.393233979603805,
+      "learning_rate": 2e-05,
+      "loss": 0.5605,
+      "step": 16865
+    },
+    {
+      "epoch": 0.4014567266495287,
+      "grad_norm": 13.225607502791535,
+      "learning_rate": 2e-05,
+      "loss": 0.6484,
+      "step": 16866
+    },
+    {
+      "epoch": 0.40148052937256024,
+      "grad_norm": 7.720798637145112,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 16867
+    },
+    {
+      "epoch": 0.4015043320955917,
+      "grad_norm": 3.3459724910403583,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 16868
+    },
+    {
+      "epoch": 0.40152813481862326,
+      "grad_norm": 23.395456417603203,
+      "learning_rate": 2e-05,
+      "loss": 2.569,
+      "step": 16869
+    },
+    {
+      "epoch": 0.40155193754165475,
+      "grad_norm": 3.8032368375770207,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 16870
+    },
+    {
+      "epoch": 0.4015757402646863,
+      "grad_norm": 2.4500702905460763,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 16871
+    },
+    {
+      "epoch": 0.4015995429877178,
+      "grad_norm": 1.285885247014,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 16872
+    },
+    {
+      "epoch": 0.4016233457107493,
+      "grad_norm": 3.4677933960231497,
+      "learning_rate": 2e-05,
+      "loss": 0.5041,
+      "step": 16873
+    },
+    {
+      "epoch": 0.4016471484337808,
+      "grad_norm": 2.744553739813686,
+      "learning_rate": 2e-05,
+      "loss": 0.2346,
+      "step": 16874
+    },
+    {
+      "epoch": 0.40167095115681234,
+      "grad_norm": 2.9496369817877612,
+      "learning_rate": 2e-05,
+      "loss": 0.5667,
+      "step": 16875
+    },
+    {
+      "epoch": 0.4016947538798439,
+      "grad_norm": 3.190461120947259,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 16876
+    },
+    {
+      "epoch": 0.40171855660287537,
+      "grad_norm": 1.5549056803132832,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 16877
+    },
+    {
+      "epoch": 0.4017423593259069,
+      "grad_norm": 7.572748441819859,
+      "learning_rate": 2e-05,
+      "loss": 0.3296,
+      "step": 16878
+    },
+    {
+      "epoch": 0.4017661620489384,
+      "grad_norm": 0.916873342283132,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 16879
+    },
+    {
+      "epoch": 0.40178996477196993,
+      "grad_norm": 0.4282504835265869,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4018137674950014,
+      "grad_norm": 18.004557178975876,
+      "learning_rate": 2e-05,
+      "loss": 0.4677,
+      "step": 16881
+    },
+    {
+      "epoch": 0.40183757021803296,
+      "grad_norm": 3.5141411117993773,
+      "learning_rate": 2e-05,
+      "loss": 0.5205,
+      "step": 16882
+    },
+    {
+      "epoch": 0.40186137294106444,
+      "grad_norm": 6.731758150578432,
+      "learning_rate": 2e-05,
+      "loss": 0.2278,
+      "step": 16883
+    },
+    {
+      "epoch": 0.401885175664096,
+      "grad_norm": 7.068480687823527,
+      "learning_rate": 2e-05,
+      "loss": 1.027,
+      "step": 16884
+    },
+    {
+      "epoch": 0.40190897838712747,
+      "grad_norm": 5.566086929516886,
+      "learning_rate": 2e-05,
+      "loss": 0.8246,
+      "step": 16885
+    },
+    {
+      "epoch": 0.401932781110159,
+      "grad_norm": 6.593020060602852,
+      "learning_rate": 2e-05,
+      "loss": 1.3861,
+      "step": 16886
+    },
+    {
+      "epoch": 0.4019565838331905,
+      "grad_norm": 1.8753644624777672,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 16887
+    },
+    {
+      "epoch": 0.40198038655622204,
+      "grad_norm": 4.312008886877774,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 16888
+    },
+    {
+      "epoch": 0.4020041892792535,
+      "grad_norm": 4.556359856066748,
+      "learning_rate": 2e-05,
+      "loss": 0.7216,
+      "step": 16889
+    },
+    {
+      "epoch": 0.40202799200228506,
+      "grad_norm": 11.25888385673724,
+      "learning_rate": 2e-05,
+      "loss": 1.0282,
+      "step": 16890
+    },
+    {
+      "epoch": 0.4020517947253166,
+      "grad_norm": 8.325048019506378,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 16891
+    },
+    {
+      "epoch": 0.4020755974483481,
+      "grad_norm": 2.140547882227268,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 16892
+    },
+    {
+      "epoch": 0.40209940017137963,
+      "grad_norm": 5.696880965105138,
+      "learning_rate": 2e-05,
+      "loss": 0.1972,
+      "step": 16893
+    },
+    {
+      "epoch": 0.4021232028944111,
+      "grad_norm": 7.2171617882063765,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 16894
+    },
+    {
+      "epoch": 0.40214700561744265,
+      "grad_norm": 10.790260552129679,
+      "learning_rate": 2e-05,
+      "loss": 0.4199,
+      "step": 16895
+    },
+    {
+      "epoch": 0.40217080834047414,
+      "grad_norm": 5.291448823768346,
+      "learning_rate": 2e-05,
+      "loss": 1.1202,
+      "step": 16896
+    },
+    {
+      "epoch": 0.4021946110635057,
+      "grad_norm": 3.466691868151584,
+      "learning_rate": 2e-05,
+      "loss": 0.0909,
+      "step": 16897
+    },
+    {
+      "epoch": 0.40221841378653717,
+      "grad_norm": 8.088300932860692,
+      "learning_rate": 2e-05,
+      "loss": 0.3271,
+      "step": 16898
+    },
+    {
+      "epoch": 0.4022422165095687,
+      "grad_norm": 1.4302266911060002,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 16899
+    },
+    {
+      "epoch": 0.4022660192326002,
+      "grad_norm": 3.9293995691823884,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 16900
+    },
+    {
+      "epoch": 0.40228982195563173,
+      "grad_norm": 3.3402555244292484,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 16901
+    },
+    {
+      "epoch": 0.4023136246786632,
+      "grad_norm": 3.8219121270322005,
+      "learning_rate": 2e-05,
+      "loss": 0.1922,
+      "step": 16902
+    },
+    {
+      "epoch": 0.40233742740169476,
+      "grad_norm": 2.9027672892547223,
+      "learning_rate": 2e-05,
+      "loss": 0.5495,
+      "step": 16903
+    },
+    {
+      "epoch": 0.40236123012472624,
+      "grad_norm": 5.723106542116122,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 16904
+    },
+    {
+      "epoch": 0.4023850328477578,
+      "grad_norm": 4.815527062154682,
+      "learning_rate": 2e-05,
+      "loss": 0.2833,
+      "step": 16905
+    },
+    {
+      "epoch": 0.4024088355707893,
+      "grad_norm": 2.7145751124422968,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 16906
+    },
+    {
+      "epoch": 0.4024326382938208,
+      "grad_norm": 0.22649521006607432,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 16907
+    },
+    {
+      "epoch": 0.40245644101685235,
+      "grad_norm": 3.078255951507046,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 16908
+    },
+    {
+      "epoch": 0.40248024373988384,
+      "grad_norm": 3.3271014601486844,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 16909
+    },
+    {
+      "epoch": 0.4025040464629154,
+      "grad_norm": 8.442500714155972,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 16910
+    },
+    {
+      "epoch": 0.40252784918594686,
+      "grad_norm": 3.6990625602487035,
+      "learning_rate": 2e-05,
+      "loss": 0.6844,
+      "step": 16911
+    },
+    {
+      "epoch": 0.4025516519089784,
+      "grad_norm": 4.58421835401498,
+      "learning_rate": 2e-05,
+      "loss": 0.5672,
+      "step": 16912
+    },
+    {
+      "epoch": 0.4025754546320099,
+      "grad_norm": 3.0166099245879376,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 16913
+    },
+    {
+      "epoch": 0.40259925735504143,
+      "grad_norm": 1.9435650992424094,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 16914
+    },
+    {
+      "epoch": 0.4026230600780729,
+      "grad_norm": 7.377724323472959,
+      "learning_rate": 2e-05,
+      "loss": 0.6276,
+      "step": 16915
+    },
+    {
+      "epoch": 0.40264686280110445,
+      "grad_norm": 4.767400575029967,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 16916
+    },
+    {
+      "epoch": 0.40267066552413594,
+      "grad_norm": 2.5746340003843327,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 16917
+    },
+    {
+      "epoch": 0.4026944682471675,
+      "grad_norm": 5.850268713697849,
+      "learning_rate": 2e-05,
+      "loss": 1.2347,
+      "step": 16918
+    },
+    {
+      "epoch": 0.40271827097019897,
+      "grad_norm": 5.598514001396489,
+      "learning_rate": 2e-05,
+      "loss": 0.6739,
+      "step": 16919
+    },
+    {
+      "epoch": 0.4027420736932305,
+      "grad_norm": 25.67251205524118,
+      "learning_rate": 2e-05,
+      "loss": 1.0082,
+      "step": 16920
+    },
+    {
+      "epoch": 0.40276587641626205,
+      "grad_norm": 2.656842035719775,
+      "learning_rate": 2e-05,
+      "loss": 0.0952,
+      "step": 16921
+    },
+    {
+      "epoch": 0.40278967913929353,
+      "grad_norm": 0.17739698085593497,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 16922
+    },
+    {
+      "epoch": 0.4028134818623251,
+      "grad_norm": 4.809105842899408,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 16923
+    },
+    {
+      "epoch": 0.40283728458535656,
+      "grad_norm": 1.146186055416008,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 16924
+    },
+    {
+      "epoch": 0.4028610873083881,
+      "grad_norm": 9.29080138149147,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 16925
+    },
+    {
+      "epoch": 0.4028848900314196,
+      "grad_norm": 11.63330429771853,
+      "learning_rate": 2e-05,
+      "loss": 0.9234,
+      "step": 16926
+    },
+    {
+      "epoch": 0.4029086927544511,
+      "grad_norm": 7.743133800630805,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 16927
+    },
+    {
+      "epoch": 0.4029324954774826,
+      "grad_norm": 4.465586179231508,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 16928
+    },
+    {
+      "epoch": 0.40295629820051415,
+      "grad_norm": 5.8231728274065455,
+      "learning_rate": 2e-05,
+      "loss": 0.8754,
+      "step": 16929
+    },
+    {
+      "epoch": 0.40298010092354564,
+      "grad_norm": 5.978028652101619,
+      "learning_rate": 2e-05,
+      "loss": 0.9093,
+      "step": 16930
+    },
+    {
+      "epoch": 0.4030039036465772,
+      "grad_norm": 0.7392936261527385,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 16931
+    },
+    {
+      "epoch": 0.40302770636960866,
+      "grad_norm": 3.163790813566991,
+      "learning_rate": 2e-05,
+      "loss": 0.1499,
+      "step": 16932
+    },
+    {
+      "epoch": 0.4030515090926402,
+      "grad_norm": 3.47551204806781,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 16933
+    },
+    {
+      "epoch": 0.4030753118156717,
+      "grad_norm": 1.2711768022121512,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 16934
+    },
+    {
+      "epoch": 0.40309911453870323,
+      "grad_norm": 0.22098120556597758,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 16935
+    },
+    {
+      "epoch": 0.40312291726173477,
+      "grad_norm": 8.116907511232306,
+      "learning_rate": 2e-05,
+      "loss": 2.3812,
+      "step": 16936
+    },
+    {
+      "epoch": 0.40314671998476626,
+      "grad_norm": 5.082771760766948,
+      "learning_rate": 2e-05,
+      "loss": 0.6881,
+      "step": 16937
+    },
+    {
+      "epoch": 0.4031705227077978,
+      "grad_norm": 4.65018918911556,
+      "learning_rate": 2e-05,
+      "loss": 0.6942,
+      "step": 16938
+    },
+    {
+      "epoch": 0.4031943254308293,
+      "grad_norm": 4.00784133410624,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 16939
+    },
+    {
+      "epoch": 0.4032181281538608,
+      "grad_norm": 3.4756530728291244,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 16940
+    },
+    {
+      "epoch": 0.4032419308768923,
+      "grad_norm": 31.648737282090785,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 16941
+    },
+    {
+      "epoch": 0.40326573359992385,
+      "grad_norm": 0.2285410778416903,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 16942
+    },
+    {
+      "epoch": 0.40328953632295533,
+      "grad_norm": 2.3320152946505974,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 16943
+    },
+    {
+      "epoch": 0.4033133390459869,
+      "grad_norm": 3.3513170361077327,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 16944
+    },
+    {
+      "epoch": 0.40333714176901836,
+      "grad_norm": 0.3070409820416238,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16945
+    },
+    {
+      "epoch": 0.4033609444920499,
+      "grad_norm": 0.646108432024375,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 16946
+    },
+    {
+      "epoch": 0.4033847472150814,
+      "grad_norm": 3.4345686590782,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 16947
+    },
+    {
+      "epoch": 0.4034085499381129,
+      "grad_norm": 0.21519791308250508,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 16948
+    },
+    {
+      "epoch": 0.4034323526611444,
+      "grad_norm": 2.1105386609411054,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 16949
+    },
+    {
+      "epoch": 0.40345615538417595,
+      "grad_norm": 4.195705166772442,
+      "learning_rate": 2e-05,
+      "loss": 0.6075,
+      "step": 16950
+    },
+    {
+      "epoch": 0.40347995810720744,
+      "grad_norm": 53.35693783299521,
+      "learning_rate": 2e-05,
+      "loss": 0.2659,
+      "step": 16951
+    },
+    {
+      "epoch": 0.403503760830239,
+      "grad_norm": 5.143527498957111,
+      "learning_rate": 2e-05,
+      "loss": 0.7486,
+      "step": 16952
+    },
+    {
+      "epoch": 0.4035275635532705,
+      "grad_norm": 5.872926789312057,
+      "learning_rate": 2e-05,
+      "loss": 0.9646,
+      "step": 16953
+    },
+    {
+      "epoch": 0.403551366276302,
+      "grad_norm": 4.917473218102571,
+      "learning_rate": 2e-05,
+      "loss": 0.4909,
+      "step": 16954
+    },
+    {
+      "epoch": 0.40357516899933354,
+      "grad_norm": 0.6569317394159336,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 16955
+    },
+    {
+      "epoch": 0.40359897172236503,
+      "grad_norm": 3.2738367327253486,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 16956
+    },
+    {
+      "epoch": 0.40362277444539657,
+      "grad_norm": 3.5890239680880796,
+      "learning_rate": 2e-05,
+      "loss": 0.6561,
+      "step": 16957
+    },
+    {
+      "epoch": 0.40364657716842806,
+      "grad_norm": 0.10331693683718829,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 16958
+    },
+    {
+      "epoch": 0.4036703798914596,
+      "grad_norm": 6.063124668850666,
+      "learning_rate": 2e-05,
+      "loss": 0.2707,
+      "step": 16959
+    },
+    {
+      "epoch": 0.4036941826144911,
+      "grad_norm": 1.3331605891970082,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4037179853375226,
+      "grad_norm": 0.7863110621731457,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 16961
+    },
+    {
+      "epoch": 0.4037417880605541,
+      "grad_norm": 4.564239068792497,
+      "learning_rate": 2e-05,
+      "loss": 0.5883,
+      "step": 16962
+    },
+    {
+      "epoch": 0.40376559078358565,
+      "grad_norm": 3.6695661545221596,
+      "learning_rate": 2e-05,
+      "loss": 0.5974,
+      "step": 16963
+    },
+    {
+      "epoch": 0.40378939350661713,
+      "grad_norm": 3.4175240932584905,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 16964
+    },
+    {
+      "epoch": 0.4038131962296487,
+      "grad_norm": 11.551615392824786,
+      "learning_rate": 2e-05,
+      "loss": 0.2885,
+      "step": 16965
+    },
+    {
+      "epoch": 0.40383699895268016,
+      "grad_norm": 7.438198738995598,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 16966
+    },
+    {
+      "epoch": 0.4038608016757117,
+      "grad_norm": 3.5982044434684406,
+      "learning_rate": 2e-05,
+      "loss": 0.2786,
+      "step": 16967
+    },
+    {
+      "epoch": 0.40388460439874324,
+      "grad_norm": 0.992546127468494,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 16968
+    },
+    {
+      "epoch": 0.4039084071217747,
+      "grad_norm": 6.474014475310585,
+      "learning_rate": 2e-05,
+      "loss": 0.2517,
+      "step": 16969
+    },
+    {
+      "epoch": 0.40393220984480627,
+      "grad_norm": 13.473406440008285,
+      "learning_rate": 2e-05,
+      "loss": 0.7562,
+      "step": 16970
+    },
+    {
+      "epoch": 0.40395601256783775,
+      "grad_norm": 1.858882134290083,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 16971
+    },
+    {
+      "epoch": 0.4039798152908693,
+      "grad_norm": 8.795479425596122,
+      "learning_rate": 2e-05,
+      "loss": 0.7446,
+      "step": 16972
+    },
+    {
+      "epoch": 0.4040036180139008,
+      "grad_norm": 2.5386850427731136,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 16973
+    },
+    {
+      "epoch": 0.4040274207369323,
+      "grad_norm": 5.241482709271932,
+      "learning_rate": 2e-05,
+      "loss": 0.3725,
+      "step": 16974
+    },
+    {
+      "epoch": 0.4040512234599638,
+      "grad_norm": 4.355392774528504,
+      "learning_rate": 2e-05,
+      "loss": 0.3129,
+      "step": 16975
+    },
+    {
+      "epoch": 0.40407502618299534,
+      "grad_norm": 3.1864728019476933,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 16976
+    },
+    {
+      "epoch": 0.40409882890602683,
+      "grad_norm": 6.423823163109066,
+      "learning_rate": 2e-05,
+      "loss": 1.4643,
+      "step": 16977
+    },
+    {
+      "epoch": 0.40412263162905837,
+      "grad_norm": 3.314099024544491,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 16978
+    },
+    {
+      "epoch": 0.40414643435208986,
+      "grad_norm": 4.3858379685740125,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 16979
+    },
+    {
+      "epoch": 0.4041702370751214,
+      "grad_norm": 3.4853453317747713,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4041940397981529,
+      "grad_norm": 5.885481146921483,
+      "learning_rate": 2e-05,
+      "loss": 0.6817,
+      "step": 16981
+    },
+    {
+      "epoch": 0.4042178425211844,
+      "grad_norm": 1.0172759931138038,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 16982
+    },
+    {
+      "epoch": 0.40424164524421596,
+      "grad_norm": 0.2997668430227027,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 16983
+    },
+    {
+      "epoch": 0.40426544796724745,
+      "grad_norm": 5.142295434947555,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 16984
+    },
+    {
+      "epoch": 0.404289250690279,
+      "grad_norm": 9.085666683129048,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 16985
+    },
+    {
+      "epoch": 0.4043130534133105,
+      "grad_norm": 0.3349514567821617,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 16986
+    },
+    {
+      "epoch": 0.404336856136342,
+      "grad_norm": 4.7230181367699595,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 16987
+    },
+    {
+      "epoch": 0.4043606588593735,
+      "grad_norm": 2.549771419489961,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 16988
+    },
+    {
+      "epoch": 0.40438446158240504,
+      "grad_norm": 4.83125892942535,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 16989
+    },
+    {
+      "epoch": 0.4044082643054365,
+      "grad_norm": 16.792582538104742,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 16990
+    },
+    {
+      "epoch": 0.40443206702846807,
+      "grad_norm": 0.4961762934153313,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 16991
+    },
+    {
+      "epoch": 0.40445586975149955,
+      "grad_norm": 4.199049388476633,
+      "learning_rate": 2e-05,
+      "loss": 0.1364,
+      "step": 16992
+    },
+    {
+      "epoch": 0.4044796724745311,
+      "grad_norm": 3.3455201866769206,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 16993
+    },
+    {
+      "epoch": 0.4045034751975626,
+      "grad_norm": 4.795229244455748,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 16994
+    },
+    {
+      "epoch": 0.4045272779205941,
+      "grad_norm": 3.734232158477648,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 16995
+    },
+    {
+      "epoch": 0.4045510806436256,
+      "grad_norm": 1.620671466883885,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 16996
+    },
+    {
+      "epoch": 0.40457488336665715,
+      "grad_norm": 6.851085976816848,
+      "learning_rate": 2e-05,
+      "loss": 1.0679,
+      "step": 16997
+    },
+    {
+      "epoch": 0.4045986860896887,
+      "grad_norm": 3.6564620812933493,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 16998
+    },
+    {
+      "epoch": 0.40462248881272017,
+      "grad_norm": 3.803715626865556,
+      "learning_rate": 2e-05,
+      "loss": 0.689,
+      "step": 16999
+    },
+    {
+      "epoch": 0.4046462915357517,
+      "grad_norm": 5.254111286801618,
+      "learning_rate": 2e-05,
+      "loss": 0.8022,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4046700942587832,
+      "grad_norm": 8.14994642141716,
+      "learning_rate": 2e-05,
+      "loss": 0.6033,
+      "step": 17001
+    },
+    {
+      "epoch": 0.40469389698181474,
+      "grad_norm": 2.965525220181356,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 17002
+    },
+    {
+      "epoch": 0.4047176997048462,
+      "grad_norm": 4.8287886070910755,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 17003
+    },
+    {
+      "epoch": 0.40474150242787776,
+      "grad_norm": 2.4658660986654604,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 17004
+    },
+    {
+      "epoch": 0.40476530515090925,
+      "grad_norm": 6.034104266050245,
+      "learning_rate": 2e-05,
+      "loss": 1.06,
+      "step": 17005
+    },
+    {
+      "epoch": 0.4047891078739408,
+      "grad_norm": 9.093622359903472,
+      "learning_rate": 2e-05,
+      "loss": 0.3098,
+      "step": 17006
+    },
+    {
+      "epoch": 0.4048129105969723,
+      "grad_norm": 9.486266653298166,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 17007
+    },
+    {
+      "epoch": 0.4048367133200038,
+      "grad_norm": 7.159495103228362,
+      "learning_rate": 2e-05,
+      "loss": 0.7378,
+      "step": 17008
+    },
+    {
+      "epoch": 0.4048605160430353,
+      "grad_norm": 4.898753745516623,
+      "learning_rate": 2e-05,
+      "loss": 0.2009,
+      "step": 17009
+    },
+    {
+      "epoch": 0.40488431876606684,
+      "grad_norm": 1.5609677218152762,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4049081214890983,
+      "grad_norm": 2.8384535380178653,
+      "learning_rate": 2e-05,
+      "loss": 0.4551,
+      "step": 17011
+    },
+    {
+      "epoch": 0.40493192421212987,
+      "grad_norm": 0.5247714335767252,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 17012
+    },
+    {
+      "epoch": 0.4049557269351614,
+      "grad_norm": 1.7275602165559742,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 17013
+    },
+    {
+      "epoch": 0.4049795296581929,
+      "grad_norm": 16.659690322845854,
+      "learning_rate": 2e-05,
+      "loss": 0.8469,
+      "step": 17014
+    },
+    {
+      "epoch": 0.40500333238122443,
+      "grad_norm": 19.092109220359244,
+      "learning_rate": 2e-05,
+      "loss": 0.4285,
+      "step": 17015
+    },
+    {
+      "epoch": 0.4050271351042559,
+      "grad_norm": 4.452785780256283,
+      "learning_rate": 2e-05,
+      "loss": 0.181,
+      "step": 17016
+    },
+    {
+      "epoch": 0.40505093782728746,
+      "grad_norm": 7.060252517560984,
+      "learning_rate": 2e-05,
+      "loss": 0.7871,
+      "step": 17017
+    },
+    {
+      "epoch": 0.40507474055031895,
+      "grad_norm": 12.56825528884407,
+      "learning_rate": 2e-05,
+      "loss": 1.0042,
+      "step": 17018
+    },
+    {
+      "epoch": 0.4050985432733505,
+      "grad_norm": 3.491971969174512,
+      "learning_rate": 2e-05,
+      "loss": 0.5404,
+      "step": 17019
+    },
+    {
+      "epoch": 0.40512234599638197,
+      "grad_norm": 4.101275887561224,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 17020
+    },
+    {
+      "epoch": 0.4051461487194135,
+      "grad_norm": 4.993271478433975,
+      "learning_rate": 2e-05,
+      "loss": 0.8803,
+      "step": 17021
+    },
+    {
+      "epoch": 0.405169951442445,
+      "grad_norm": 5.45750519055592,
+      "learning_rate": 2e-05,
+      "loss": 0.3299,
+      "step": 17022
+    },
+    {
+      "epoch": 0.40519375416547654,
+      "grad_norm": 2.0749153788695094,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 17023
+    },
+    {
+      "epoch": 0.405217556888508,
+      "grad_norm": 5.833709785769441,
+      "learning_rate": 2e-05,
+      "loss": 0.9265,
+      "step": 17024
+    },
+    {
+      "epoch": 0.40524135961153956,
+      "grad_norm": 1.301449290889395,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 17025
+    },
+    {
+      "epoch": 0.40526516233457105,
+      "grad_norm": 8.241557498220045,
+      "learning_rate": 2e-05,
+      "loss": 0.8143,
+      "step": 17026
+    },
+    {
+      "epoch": 0.4052889650576026,
+      "grad_norm": 1.7728254620444075,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 17027
+    },
+    {
+      "epoch": 0.40531276778063413,
+      "grad_norm": 4.178364677533782,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 17028
+    },
+    {
+      "epoch": 0.4053365705036656,
+      "grad_norm": 0.7863407618530168,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17029
+    },
+    {
+      "epoch": 0.40536037322669716,
+      "grad_norm": 0.5006535407903124,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 17030
+    },
+    {
+      "epoch": 0.40538417594972864,
+      "grad_norm": 5.5859090194168015,
+      "learning_rate": 2e-05,
+      "loss": 0.8618,
+      "step": 17031
+    },
+    {
+      "epoch": 0.4054079786727602,
+      "grad_norm": 4.238806913956749,
+      "learning_rate": 2e-05,
+      "loss": 0.5092,
+      "step": 17032
+    },
+    {
+      "epoch": 0.40543178139579167,
+      "grad_norm": 3.388588305439812,
+      "learning_rate": 2e-05,
+      "loss": 0.3874,
+      "step": 17033
+    },
+    {
+      "epoch": 0.4054555841188232,
+      "grad_norm": 1.112267970742851,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 17034
+    },
+    {
+      "epoch": 0.4054793868418547,
+      "grad_norm": 7.226314212029765,
+      "learning_rate": 2e-05,
+      "loss": 0.8492,
+      "step": 17035
+    },
+    {
+      "epoch": 0.40550318956488624,
+      "grad_norm": 2.0754109612006166,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 17036
+    },
+    {
+      "epoch": 0.4055269922879177,
+      "grad_norm": 5.982439687437009,
+      "learning_rate": 2e-05,
+      "loss": 0.6854,
+      "step": 17037
+    },
+    {
+      "epoch": 0.40555079501094926,
+      "grad_norm": 0.5106723000549052,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17038
+    },
+    {
+      "epoch": 0.40557459773398075,
+      "grad_norm": 8.207493889653788,
+      "learning_rate": 2e-05,
+      "loss": 1.6613,
+      "step": 17039
+    },
+    {
+      "epoch": 0.4055984004570123,
+      "grad_norm": 2.59780984821893,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 17040
+    },
+    {
+      "epoch": 0.40562220318004377,
+      "grad_norm": 4.755201042787419,
+      "learning_rate": 2e-05,
+      "loss": 0.7699,
+      "step": 17041
+    },
+    {
+      "epoch": 0.4056460059030753,
+      "grad_norm": 4.911371642934493,
+      "learning_rate": 2e-05,
+      "loss": 0.2591,
+      "step": 17042
+    },
+    {
+      "epoch": 0.40566980862610685,
+      "grad_norm": 2.974242008557587,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 17043
+    },
+    {
+      "epoch": 0.40569361134913834,
+      "grad_norm": 4.818275346715788,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 17044
+    },
+    {
+      "epoch": 0.4057174140721699,
+      "grad_norm": 16.424786763564885,
+      "learning_rate": 2e-05,
+      "loss": 1.674,
+      "step": 17045
+    },
+    {
+      "epoch": 0.40574121679520136,
+      "grad_norm": 6.357145711670777,
+      "learning_rate": 2e-05,
+      "loss": 0.8167,
+      "step": 17046
+    },
+    {
+      "epoch": 0.4057650195182329,
+      "grad_norm": 1.6932004495708846,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 17047
+    },
+    {
+      "epoch": 0.4057888222412644,
+      "grad_norm": 3.4150346557681033,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 17048
+    },
+    {
+      "epoch": 0.40581262496429593,
+      "grad_norm": 5.23575929175356,
+      "learning_rate": 2e-05,
+      "loss": 0.3985,
+      "step": 17049
+    },
+    {
+      "epoch": 0.4058364276873274,
+      "grad_norm": 2.6740679733094335,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 17050
+    },
+    {
+      "epoch": 0.40586023041035896,
+      "grad_norm": 1.2989827631361672,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 17051
+    },
+    {
+      "epoch": 0.40588403313339044,
+      "grad_norm": 0.10007005828431437,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17052
+    },
+    {
+      "epoch": 0.405907835856422,
+      "grad_norm": 5.261291532559829,
+      "learning_rate": 2e-05,
+      "loss": 0.6759,
+      "step": 17053
+    },
+    {
+      "epoch": 0.40593163857945347,
+      "grad_norm": 2.6484161263533292,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 17054
+    },
+    {
+      "epoch": 0.405955441302485,
+      "grad_norm": 2.734214209624622,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 17055
+    },
+    {
+      "epoch": 0.4059792440255165,
+      "grad_norm": 2.7422879603002563,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 17056
+    },
+    {
+      "epoch": 0.40600304674854804,
+      "grad_norm": 1.640356442275554,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 17057
+    },
+    {
+      "epoch": 0.4060268494715796,
+      "grad_norm": 0.8786205281266336,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 17058
+    },
+    {
+      "epoch": 0.40605065219461106,
+      "grad_norm": 0.9724155718984577,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 17059
+    },
+    {
+      "epoch": 0.4060744549176426,
+      "grad_norm": 4.17389099681953,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4060982576406741,
+      "grad_norm": 5.288602854415519,
+      "learning_rate": 2e-05,
+      "loss": 1.1125,
+      "step": 17061
+    },
+    {
+      "epoch": 0.40612206036370563,
+      "grad_norm": 2.213316403133015,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 17062
+    },
+    {
+      "epoch": 0.4061458630867371,
+      "grad_norm": 6.230102723719239,
+      "learning_rate": 2e-05,
+      "loss": 0.3282,
+      "step": 17063
+    },
+    {
+      "epoch": 0.40616966580976865,
+      "grad_norm": 2.8367739055774615,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 17064
+    },
+    {
+      "epoch": 0.40619346853280014,
+      "grad_norm": 0.5163351503281831,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 17065
+    },
+    {
+      "epoch": 0.4062172712558317,
+      "grad_norm": 5.875073943087285,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 17066
+    },
+    {
+      "epoch": 0.40624107397886317,
+      "grad_norm": 3.451331419967344,
+      "learning_rate": 2e-05,
+      "loss": 0.4289,
+      "step": 17067
+    },
+    {
+      "epoch": 0.4062648767018947,
+      "grad_norm": 29.745813686724098,
+      "learning_rate": 2e-05,
+      "loss": 0.4624,
+      "step": 17068
+    },
+    {
+      "epoch": 0.4062886794249262,
+      "grad_norm": 2.0408579806504537,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 17069
+    },
+    {
+      "epoch": 0.40631248214795773,
+      "grad_norm": 0.09416474562591261,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 17070
+    },
+    {
+      "epoch": 0.4063362848709892,
+      "grad_norm": 3.197163256503532,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 17071
+    },
+    {
+      "epoch": 0.40636008759402076,
+      "grad_norm": 1.634778872560247,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 17072
+    },
+    {
+      "epoch": 0.40638389031705224,
+      "grad_norm": 2.728854535913391,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 17073
+    },
+    {
+      "epoch": 0.4064076930400838,
+      "grad_norm": 0.49354834685580323,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 17074
+    },
+    {
+      "epoch": 0.4064314957631153,
+      "grad_norm": 5.377184760791456,
+      "learning_rate": 2e-05,
+      "loss": 0.6513,
+      "step": 17075
+    },
+    {
+      "epoch": 0.4064552984861468,
+      "grad_norm": 6.766176205209922,
+      "learning_rate": 2e-05,
+      "loss": 1.0038,
+      "step": 17076
+    },
+    {
+      "epoch": 0.40647910120917835,
+      "grad_norm": 3.9381051859706826,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 17077
+    },
+    {
+      "epoch": 0.40650290393220984,
+      "grad_norm": 11.004908323183216,
+      "learning_rate": 2e-05,
+      "loss": 0.6475,
+      "step": 17078
+    },
+    {
+      "epoch": 0.4065267066552414,
+      "grad_norm": 9.96068584513501,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 17079
+    },
+    {
+      "epoch": 0.40655050937827286,
+      "grad_norm": 50.147584780656096,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4065743121013044,
+      "grad_norm": 0.7697736323572039,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17081
+    },
+    {
+      "epoch": 0.4065981148243359,
+      "grad_norm": 3.71457702566033,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 17082
+    },
+    {
+      "epoch": 0.40662191754736743,
+      "grad_norm": 19.993578705092922,
+      "learning_rate": 2e-05,
+      "loss": 0.6525,
+      "step": 17083
+    },
+    {
+      "epoch": 0.4066457202703989,
+      "grad_norm": 3.414871393400424,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 17084
+    },
+    {
+      "epoch": 0.40666952299343045,
+      "grad_norm": 5.46338796626761,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 17085
+    },
+    {
+      "epoch": 0.40669332571646194,
+      "grad_norm": 11.086132844341485,
+      "learning_rate": 2e-05,
+      "loss": 0.5041,
+      "step": 17086
+    },
+    {
+      "epoch": 0.4067171284394935,
+      "grad_norm": 4.976421109685015,
+      "learning_rate": 2e-05,
+      "loss": 0.2146,
+      "step": 17087
+    },
+    {
+      "epoch": 0.40674093116252497,
+      "grad_norm": 4.173473252349144,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 17088
+    },
+    {
+      "epoch": 0.4067647338855565,
+      "grad_norm": 1.3691449236164714,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 17089
+    },
+    {
+      "epoch": 0.40678853660858805,
+      "grad_norm": 6.870165210576049,
+      "learning_rate": 2e-05,
+      "loss": 0.683,
+      "step": 17090
+    },
+    {
+      "epoch": 0.40681233933161953,
+      "grad_norm": 0.13153799828830504,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 17091
+    },
+    {
+      "epoch": 0.4068361420546511,
+      "grad_norm": 0.6884561878943772,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 17092
+    },
+    {
+      "epoch": 0.40685994477768256,
+      "grad_norm": 0.7288790795100331,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 17093
+    },
+    {
+      "epoch": 0.4068837475007141,
+      "grad_norm": 6.130399181086275,
+      "learning_rate": 2e-05,
+      "loss": 0.8043,
+      "step": 17094
+    },
+    {
+      "epoch": 0.4069075502237456,
+      "grad_norm": 3.9150198394420013,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 17095
+    },
+    {
+      "epoch": 0.4069313529467771,
+      "grad_norm": 6.330012993509091,
+      "learning_rate": 2e-05,
+      "loss": 1.2069,
+      "step": 17096
+    },
+    {
+      "epoch": 0.4069551556698086,
+      "grad_norm": 5.069778772270686,
+      "learning_rate": 2e-05,
+      "loss": 0.9022,
+      "step": 17097
+    },
+    {
+      "epoch": 0.40697895839284015,
+      "grad_norm": 3.1259164553263874,
+      "learning_rate": 2e-05,
+      "loss": 0.3422,
+      "step": 17098
+    },
+    {
+      "epoch": 0.40700276111587164,
+      "grad_norm": 11.527141740340918,
+      "learning_rate": 2e-05,
+      "loss": 1.0365,
+      "step": 17099
+    },
+    {
+      "epoch": 0.4070265638389032,
+      "grad_norm": 0.10578834135313983,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 17100
+    },
+    {
+      "epoch": 0.40705036656193466,
+      "grad_norm": 2.442122465522065,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 17101
+    },
+    {
+      "epoch": 0.4070741692849662,
+      "grad_norm": 2.808813101008751,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 17102
+    },
+    {
+      "epoch": 0.4070979720079977,
+      "grad_norm": 5.204936198747735,
+      "learning_rate": 2e-05,
+      "loss": 0.5495,
+      "step": 17103
+    },
+    {
+      "epoch": 0.40712177473102923,
+      "grad_norm": 2.239963344932059,
+      "learning_rate": 2e-05,
+      "loss": 0.1667,
+      "step": 17104
+    },
+    {
+      "epoch": 0.40714557745406077,
+      "grad_norm": 0.7463571183037144,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 17105
+    },
+    {
+      "epoch": 0.40716938017709225,
+      "grad_norm": 8.377410341218605,
+      "learning_rate": 2e-05,
+      "loss": 1.5087,
+      "step": 17106
+    },
+    {
+      "epoch": 0.4071931829001238,
+      "grad_norm": 4.5058181569886155,
+      "learning_rate": 2e-05,
+      "loss": 0.5771,
+      "step": 17107
+    },
+    {
+      "epoch": 0.4072169856231553,
+      "grad_norm": 0.36197440041557083,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17108
+    },
+    {
+      "epoch": 0.4072407883461868,
+      "grad_norm": 5.041999036816977,
+      "learning_rate": 2e-05,
+      "loss": 0.7596,
+      "step": 17109
+    },
+    {
+      "epoch": 0.4072645910692183,
+      "grad_norm": 4.38199922173272,
+      "learning_rate": 2e-05,
+      "loss": 0.228,
+      "step": 17110
+    },
+    {
+      "epoch": 0.40728839379224985,
+      "grad_norm": 4.466857943888787,
+      "learning_rate": 2e-05,
+      "loss": 0.9103,
+      "step": 17111
+    },
+    {
+      "epoch": 0.40731219651528133,
+      "grad_norm": 5.672318608508447,
+      "learning_rate": 2e-05,
+      "loss": 0.742,
+      "step": 17112
+    },
+    {
+      "epoch": 0.4073359992383129,
+      "grad_norm": 11.930153706764822,
+      "learning_rate": 2e-05,
+      "loss": 0.1805,
+      "step": 17113
+    },
+    {
+      "epoch": 0.40735980196134436,
+      "grad_norm": 1.4407392634393323,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 17114
+    },
+    {
+      "epoch": 0.4073836046843759,
+      "grad_norm": 3.368094871707816,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 17115
+    },
+    {
+      "epoch": 0.4074074074074074,
+      "grad_norm": 4.0557274326408725,
+      "learning_rate": 2e-05,
+      "loss": 0.6611,
+      "step": 17116
+    },
+    {
+      "epoch": 0.4074312101304389,
+      "grad_norm": 0.8725953803967113,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 17117
+    },
+    {
+      "epoch": 0.4074550128534704,
+      "grad_norm": 6.251764769258758,
+      "learning_rate": 2e-05,
+      "loss": 0.4422,
+      "step": 17118
+    },
+    {
+      "epoch": 0.40747881557650195,
+      "grad_norm": 2.6337337665230436,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 17119
+    },
+    {
+      "epoch": 0.4075026182995335,
+      "grad_norm": 6.30870957414829,
+      "learning_rate": 2e-05,
+      "loss": 1.2388,
+      "step": 17120
+    },
+    {
+      "epoch": 0.407526421022565,
+      "grad_norm": 2.075613040146885,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 17121
+    },
+    {
+      "epoch": 0.4075502237455965,
+      "grad_norm": 8.961867792660618,
+      "learning_rate": 2e-05,
+      "loss": 0.7667,
+      "step": 17122
+    },
+    {
+      "epoch": 0.407574026468628,
+      "grad_norm": 3.81136878364134,
+      "learning_rate": 2e-05,
+      "loss": 0.376,
+      "step": 17123
+    },
+    {
+      "epoch": 0.40759782919165954,
+      "grad_norm": 0.2569781285800934,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 17124
+    },
+    {
+      "epoch": 0.40762163191469103,
+      "grad_norm": 13.527472049749868,
+      "learning_rate": 2e-05,
+      "loss": 0.9473,
+      "step": 17125
+    },
+    {
+      "epoch": 0.40764543463772257,
+      "grad_norm": 2.851705422546483,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 17126
+    },
+    {
+      "epoch": 0.40766923736075406,
+      "grad_norm": 0.4933626721470647,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17127
+    },
+    {
+      "epoch": 0.4076930400837856,
+      "grad_norm": 5.001239303728059,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 17128
+    },
+    {
+      "epoch": 0.4077168428068171,
+      "grad_norm": 2.732931940778915,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 17129
+    },
+    {
+      "epoch": 0.4077406455298486,
+      "grad_norm": 35.26156561980108,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4077644482528801,
+      "grad_norm": 12.338625777456853,
+      "learning_rate": 2e-05,
+      "loss": 0.907,
+      "step": 17131
+    },
+    {
+      "epoch": 0.40778825097591165,
+      "grad_norm": 2.13691940995972,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 17132
+    },
+    {
+      "epoch": 0.40781205369894313,
+      "grad_norm": 3.2923372600763834,
+      "learning_rate": 2e-05,
+      "loss": 0.6379,
+      "step": 17133
+    },
+    {
+      "epoch": 0.4078358564219747,
+      "grad_norm": 13.472819914748495,
+      "learning_rate": 2e-05,
+      "loss": 0.6491,
+      "step": 17134
+    },
+    {
+      "epoch": 0.4078596591450062,
+      "grad_norm": 0.9325717004557601,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 17135
+    },
+    {
+      "epoch": 0.4078834618680377,
+      "grad_norm": 230.88020622870678,
+      "learning_rate": 2e-05,
+      "loss": 0.5929,
+      "step": 17136
+    },
+    {
+      "epoch": 0.40790726459106924,
+      "grad_norm": 16.29660799990968,
+      "learning_rate": 2e-05,
+      "loss": 0.4828,
+      "step": 17137
+    },
+    {
+      "epoch": 0.4079310673141007,
+      "grad_norm": 5.2742954679120615,
+      "learning_rate": 2e-05,
+      "loss": 0.6387,
+      "step": 17138
+    },
+    {
+      "epoch": 0.40795487003713227,
+      "grad_norm": 2.5314924533429286,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 17139
+    },
+    {
+      "epoch": 0.40797867276016375,
+      "grad_norm": 11.172828814434071,
+      "learning_rate": 2e-05,
+      "loss": 1.3373,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4080024754831953,
+      "grad_norm": 7.02984499671448,
+      "learning_rate": 2e-05,
+      "loss": 1.3276,
+      "step": 17141
+    },
+    {
+      "epoch": 0.4080262782062268,
+      "grad_norm": 2.0702898731983868,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 17142
+    },
+    {
+      "epoch": 0.4080500809292583,
+      "grad_norm": 5.853585536356757,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 17143
+    },
+    {
+      "epoch": 0.4080738836522898,
+      "grad_norm": 3.6714829527984643,
+      "learning_rate": 2e-05,
+      "loss": 0.7435,
+      "step": 17144
+    },
+    {
+      "epoch": 0.40809768637532134,
+      "grad_norm": 0.20374899257199494,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 17145
+    },
+    {
+      "epoch": 0.40812148909835283,
+      "grad_norm": 3.5002917543056338,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 17146
+    },
+    {
+      "epoch": 0.40814529182138437,
+      "grad_norm": 0.2236553770504073,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 17147
+    },
+    {
+      "epoch": 0.40816909454441586,
+      "grad_norm": 4.466383397797569,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 17148
+    },
+    {
+      "epoch": 0.4081928972674474,
+      "grad_norm": 2.2009305082447437,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 17149
+    },
+    {
+      "epoch": 0.40821669999047894,
+      "grad_norm": 1.7226640208137378,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 17150
+    },
+    {
+      "epoch": 0.4082405027135104,
+      "grad_norm": 10.475234468291688,
+      "learning_rate": 2e-05,
+      "loss": 0.3817,
+      "step": 17151
+    },
+    {
+      "epoch": 0.40826430543654196,
+      "grad_norm": 4.644941067491744,
+      "learning_rate": 2e-05,
+      "loss": 0.1788,
+      "step": 17152
+    },
+    {
+      "epoch": 0.40828810815957345,
+      "grad_norm": 4.398608904230118,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 17153
+    },
+    {
+      "epoch": 0.408311910882605,
+      "grad_norm": 6.620826866428234,
+      "learning_rate": 2e-05,
+      "loss": 0.2456,
+      "step": 17154
+    },
+    {
+      "epoch": 0.4083357136056365,
+      "grad_norm": 3.754503873140308,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 17155
+    },
+    {
+      "epoch": 0.408359516328668,
+      "grad_norm": 3.305895179686076,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 17156
+    },
+    {
+      "epoch": 0.4083833190516995,
+      "grad_norm": 7.98446255374208,
+      "learning_rate": 2e-05,
+      "loss": 0.9865,
+      "step": 17157
+    },
+    {
+      "epoch": 0.40840712177473104,
+      "grad_norm": 4.226243188414497,
+      "learning_rate": 2e-05,
+      "loss": 0.2485,
+      "step": 17158
+    },
+    {
+      "epoch": 0.4084309244977625,
+      "grad_norm": 4.921013538588593,
+      "learning_rate": 2e-05,
+      "loss": 0.6619,
+      "step": 17159
+    },
+    {
+      "epoch": 0.40845472722079407,
+      "grad_norm": 46.78406347064837,
+      "learning_rate": 2e-05,
+      "loss": 0.8458,
+      "step": 17160
+    },
+    {
+      "epoch": 0.40847852994382555,
+      "grad_norm": 15.379381653382431,
+      "learning_rate": 2e-05,
+      "loss": 0.6131,
+      "step": 17161
+    },
+    {
+      "epoch": 0.4085023326668571,
+      "grad_norm": 10.17104100120454,
+      "learning_rate": 2e-05,
+      "loss": 1.6299,
+      "step": 17162
+    },
+    {
+      "epoch": 0.4085261353898886,
+      "grad_norm": 0.9889803291668463,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 17163
+    },
+    {
+      "epoch": 0.4085499381129201,
+      "grad_norm": 8.654242548268172,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 17164
+    },
+    {
+      "epoch": 0.40857374083595166,
+      "grad_norm": 2.4179400338756776,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 17165
+    },
+    {
+      "epoch": 0.40859754355898315,
+      "grad_norm": 9.176940199349703,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 17166
+    },
+    {
+      "epoch": 0.4086213462820147,
+      "grad_norm": 2.0001587211929293,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 17167
+    },
+    {
+      "epoch": 0.40864514900504617,
+      "grad_norm": 4.430756877942816,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 17168
+    },
+    {
+      "epoch": 0.4086689517280777,
+      "grad_norm": 13.814170409106405,
+      "learning_rate": 2e-05,
+      "loss": 0.3314,
+      "step": 17169
+    },
+    {
+      "epoch": 0.4086927544511092,
+      "grad_norm": 11.865977255789499,
+      "learning_rate": 2e-05,
+      "loss": 0.2508,
+      "step": 17170
+    },
+    {
+      "epoch": 0.40871655717414074,
+      "grad_norm": 0.2609568881490105,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17171
+    },
+    {
+      "epoch": 0.4087403598971722,
+      "grad_norm": 6.738110187557953,
+      "learning_rate": 2e-05,
+      "loss": 1.0149,
+      "step": 17172
+    },
+    {
+      "epoch": 0.40876416262020376,
+      "grad_norm": 2.566876590951818,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 17173
+    },
+    {
+      "epoch": 0.40878796534323525,
+      "grad_norm": 30.413511414473952,
+      "learning_rate": 2e-05,
+      "loss": 0.282,
+      "step": 17174
+    },
+    {
+      "epoch": 0.4088117680662668,
+      "grad_norm": 16.51807616604537,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 17175
+    },
+    {
+      "epoch": 0.4088355707892983,
+      "grad_norm": 16.58228220296728,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 17176
+    },
+    {
+      "epoch": 0.4088593735123298,
+      "grad_norm": 1.971388671176239,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 17177
+    },
+    {
+      "epoch": 0.4088831762353613,
+      "grad_norm": 36.72290761113479,
+      "learning_rate": 2e-05,
+      "loss": 0.9761,
+      "step": 17178
+    },
+    {
+      "epoch": 0.40890697895839284,
+      "grad_norm": 2.620813348960673,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 17179
+    },
+    {
+      "epoch": 0.4089307816814244,
+      "grad_norm": 21.47189177854019,
+      "learning_rate": 2e-05,
+      "loss": 1.2708,
+      "step": 17180
+    },
+    {
+      "epoch": 0.40895458440445587,
+      "grad_norm": 2.2665690063952693,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 17181
+    },
+    {
+      "epoch": 0.4089783871274874,
+      "grad_norm": 1.8721479412752933,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 17182
+    },
+    {
+      "epoch": 0.4090021898505189,
+      "grad_norm": 4.931695007108337,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 17183
+    },
+    {
+      "epoch": 0.40902599257355043,
+      "grad_norm": 3.0795900924866206,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 17184
+    },
+    {
+      "epoch": 0.4090497952965819,
+      "grad_norm": 15.62985273184224,
+      "learning_rate": 2e-05,
+      "loss": 0.5917,
+      "step": 17185
+    },
+    {
+      "epoch": 0.40907359801961346,
+      "grad_norm": 4.857654385296449,
+      "learning_rate": 2e-05,
+      "loss": 0.7185,
+      "step": 17186
+    },
+    {
+      "epoch": 0.40909740074264495,
+      "grad_norm": 1.6266384665399303,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 17187
+    },
+    {
+      "epoch": 0.4091212034656765,
+      "grad_norm": 1.4109947129115457,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 17188
+    },
+    {
+      "epoch": 0.40914500618870797,
+      "grad_norm": 18.568801412184698,
+      "learning_rate": 2e-05,
+      "loss": 0.3386,
+      "step": 17189
+    },
+    {
+      "epoch": 0.4091688089117395,
+      "grad_norm": 3.5271218876069956,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 17190
+    },
+    {
+      "epoch": 0.409192611634771,
+      "grad_norm": 3.528605036174523,
+      "learning_rate": 2e-05,
+      "loss": 0.506,
+      "step": 17191
+    },
+    {
+      "epoch": 0.40921641435780254,
+      "grad_norm": 3.111343172641379,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 17192
+    },
+    {
+      "epoch": 0.409240217080834,
+      "grad_norm": 0.04516418376676215,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 17193
+    },
+    {
+      "epoch": 0.40926401980386556,
+      "grad_norm": 2.535284414491975,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 17194
+    },
+    {
+      "epoch": 0.40928782252689705,
+      "grad_norm": 7.1108462952844755,
+      "learning_rate": 2e-05,
+      "loss": 1.0331,
+      "step": 17195
+    },
+    {
+      "epoch": 0.4093116252499286,
+      "grad_norm": 8.803436763404079,
+      "learning_rate": 2e-05,
+      "loss": 0.7468,
+      "step": 17196
+    },
+    {
+      "epoch": 0.40933542797296013,
+      "grad_norm": 110.47225652774671,
+      "learning_rate": 2e-05,
+      "loss": 0.5836,
+      "step": 17197
+    },
+    {
+      "epoch": 0.4093592306959916,
+      "grad_norm": 4.082181140346847,
+      "learning_rate": 2e-05,
+      "loss": 0.3871,
+      "step": 17198
+    },
+    {
+      "epoch": 0.40938303341902316,
+      "grad_norm": 3.384872350854354,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 17199
+    },
+    {
+      "epoch": 0.40940683614205464,
+      "grad_norm": 4.375020528044972,
+      "learning_rate": 2e-05,
+      "loss": 0.6057,
+      "step": 17200
+    },
+    {
+      "epoch": 0.4094306388650862,
+      "grad_norm": 6.0954365598572435,
+      "learning_rate": 2e-05,
+      "loss": 0.8159,
+      "step": 17201
+    },
+    {
+      "epoch": 0.40945444158811767,
+      "grad_norm": 1.6671629476223377,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 17202
+    },
+    {
+      "epoch": 0.4094782443111492,
+      "grad_norm": 3.0263948329877204,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 17203
+    },
+    {
+      "epoch": 0.4095020470341807,
+      "grad_norm": 1.5719730418655693,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 17204
+    },
+    {
+      "epoch": 0.40952584975721223,
+      "grad_norm": 11.482721301933678,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 17205
+    },
+    {
+      "epoch": 0.4095496524802437,
+      "grad_norm": 16.098143876833728,
+      "learning_rate": 2e-05,
+      "loss": 1.2454,
+      "step": 17206
+    },
+    {
+      "epoch": 0.40957345520327526,
+      "grad_norm": 5.2957316109699475,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 17207
+    },
+    {
+      "epoch": 0.40959725792630675,
+      "grad_norm": 4.035542174725899,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 17208
+    },
+    {
+      "epoch": 0.4096210606493383,
+      "grad_norm": 8.021958093680535,
+      "learning_rate": 2e-05,
+      "loss": 0.8372,
+      "step": 17209
+    },
+    {
+      "epoch": 0.40964486337236977,
+      "grad_norm": 8.838298122128677,
+      "learning_rate": 2e-05,
+      "loss": 0.3982,
+      "step": 17210
+    },
+    {
+      "epoch": 0.4096686660954013,
+      "grad_norm": 10.092718901369814,
+      "learning_rate": 2e-05,
+      "loss": 0.504,
+      "step": 17211
+    },
+    {
+      "epoch": 0.40969246881843285,
+      "grad_norm": 5.195953480675647,
+      "learning_rate": 2e-05,
+      "loss": 0.8127,
+      "step": 17212
+    },
+    {
+      "epoch": 0.40971627154146434,
+      "grad_norm": 6.844916112875084,
+      "learning_rate": 2e-05,
+      "loss": 0.7732,
+      "step": 17213
+    },
+    {
+      "epoch": 0.4097400742644959,
+      "grad_norm": 5.646422809350492,
+      "learning_rate": 2e-05,
+      "loss": 0.6241,
+      "step": 17214
+    },
+    {
+      "epoch": 0.40976387698752736,
+      "grad_norm": 4.002082543604757,
+      "learning_rate": 2e-05,
+      "loss": 0.1603,
+      "step": 17215
+    },
+    {
+      "epoch": 0.4097876797105589,
+      "grad_norm": 6.053987742961986,
+      "learning_rate": 2e-05,
+      "loss": 0.8536,
+      "step": 17216
+    },
+    {
+      "epoch": 0.4098114824335904,
+      "grad_norm": 2.6725182003156873,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 17217
+    },
+    {
+      "epoch": 0.40983528515662193,
+      "grad_norm": 5.416259650248621,
+      "learning_rate": 2e-05,
+      "loss": 0.7394,
+      "step": 17218
+    },
+    {
+      "epoch": 0.4098590878796534,
+      "grad_norm": 4.785106655795944,
+      "learning_rate": 2e-05,
+      "loss": 0.56,
+      "step": 17219
+    },
+    {
+      "epoch": 0.40988289060268496,
+      "grad_norm": 19.812858447339664,
+      "learning_rate": 2e-05,
+      "loss": 0.6573,
+      "step": 17220
+    },
+    {
+      "epoch": 0.40990669332571644,
+      "grad_norm": 8.986364820044766,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 17221
+    },
+    {
+      "epoch": 0.409930496048748,
+      "grad_norm": 0.8772047896172109,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17222
+    },
+    {
+      "epoch": 0.40995429877177947,
+      "grad_norm": 1.3655326220535393,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 17223
+    },
+    {
+      "epoch": 0.409978101494811,
+      "grad_norm": 5.097804432857309,
+      "learning_rate": 2e-05,
+      "loss": 0.2526,
+      "step": 17224
+    },
+    {
+      "epoch": 0.4100019042178425,
+      "grad_norm": 7.45349434454423,
+      "learning_rate": 2e-05,
+      "loss": 0.8028,
+      "step": 17225
+    },
+    {
+      "epoch": 0.41002570694087404,
+      "grad_norm": 2.9948213510175745,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 17226
+    },
+    {
+      "epoch": 0.4100495096639056,
+      "grad_norm": 0.5984241140299211,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 17227
+    },
+    {
+      "epoch": 0.41007331238693706,
+      "grad_norm": 6.229786906689211,
+      "learning_rate": 2e-05,
+      "loss": 1.3452,
+      "step": 17228
+    },
+    {
+      "epoch": 0.4100971151099686,
+      "grad_norm": 2.4658086025912977,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 17229
+    },
+    {
+      "epoch": 0.4101209178330001,
+      "grad_norm": 5.439664682880289,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 17230
+    },
+    {
+      "epoch": 0.41014472055603163,
+      "grad_norm": 5.012881122397953,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 17231
+    },
+    {
+      "epoch": 0.4101685232790631,
+      "grad_norm": 4.062078927080612,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 17232
+    },
+    {
+      "epoch": 0.41019232600209465,
+      "grad_norm": 2.085581305774468,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 17233
+    },
+    {
+      "epoch": 0.41021612872512614,
+      "grad_norm": 5.430719117694065,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 17234
+    },
+    {
+      "epoch": 0.4102399314481577,
+      "grad_norm": 7.551568405407472,
+      "learning_rate": 2e-05,
+      "loss": 0.3848,
+      "step": 17235
+    },
+    {
+      "epoch": 0.41026373417118916,
+      "grad_norm": 4.682795798383984,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 17236
+    },
+    {
+      "epoch": 0.4102875368942207,
+      "grad_norm": 4.126865422340611,
+      "learning_rate": 2e-05,
+      "loss": 0.2962,
+      "step": 17237
+    },
+    {
+      "epoch": 0.4103113396172522,
+      "grad_norm": 4.068169292513762,
+      "learning_rate": 2e-05,
+      "loss": 0.4433,
+      "step": 17238
+    },
+    {
+      "epoch": 0.41033514234028373,
+      "grad_norm": 25.38931235300119,
+      "learning_rate": 2e-05,
+      "loss": 1.7303,
+      "step": 17239
+    },
+    {
+      "epoch": 0.4103589450633152,
+      "grad_norm": 5.15765064055373,
+      "learning_rate": 2e-05,
+      "loss": 0.4817,
+      "step": 17240
+    },
+    {
+      "epoch": 0.41038274778634676,
+      "grad_norm": 26.397195493938685,
+      "learning_rate": 2e-05,
+      "loss": 1.4058,
+      "step": 17241
+    },
+    {
+      "epoch": 0.4104065505093783,
+      "grad_norm": 199.1757252417488,
+      "learning_rate": 2e-05,
+      "loss": 1.7711,
+      "step": 17242
+    },
+    {
+      "epoch": 0.4104303532324098,
+      "grad_norm": 9.40653730404071,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 17243
+    },
+    {
+      "epoch": 0.4104541559554413,
+      "grad_norm": 2.884722305076766,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 17244
+    },
+    {
+      "epoch": 0.4104779586784728,
+      "grad_norm": 0.4001786653366842,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 17245
+    },
+    {
+      "epoch": 0.41050176140150435,
+      "grad_norm": 307.85993818680157,
+      "learning_rate": 2e-05,
+      "loss": 0.7133,
+      "step": 17246
+    },
+    {
+      "epoch": 0.41052556412453584,
+      "grad_norm": 2.3005484962148333,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 17247
+    },
+    {
+      "epoch": 0.4105493668475674,
+      "grad_norm": 4.712130343724187,
+      "learning_rate": 2e-05,
+      "loss": 0.9017,
+      "step": 17248
+    },
+    {
+      "epoch": 0.41057316957059886,
+      "grad_norm": 3.092116719148343,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 17249
+    },
+    {
+      "epoch": 0.4105969722936304,
+      "grad_norm": 6.553493966021659,
+      "learning_rate": 2e-05,
+      "loss": 1.0557,
+      "step": 17250
+    },
+    {
+      "epoch": 0.4106207750166619,
+      "grad_norm": 0.15213895950428297,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 17251
+    },
+    {
+      "epoch": 0.41064457773969343,
+      "grad_norm": 10.195567030561534,
+      "learning_rate": 2e-05,
+      "loss": 1.1748,
+      "step": 17252
+    },
+    {
+      "epoch": 0.4106683804627249,
+      "grad_norm": 3.8310240219003746,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 17253
+    },
+    {
+      "epoch": 0.41069218318575645,
+      "grad_norm": 5.554637275814389,
+      "learning_rate": 2e-05,
+      "loss": 0.268,
+      "step": 17254
+    },
+    {
+      "epoch": 0.41071598590878794,
+      "grad_norm": 2.9287774500220007,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 17255
+    },
+    {
+      "epoch": 0.4107397886318195,
+      "grad_norm": 2.3899292235444882,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 17256
+    },
+    {
+      "epoch": 0.410763591354851,
+      "grad_norm": 3.714551607664966,
+      "learning_rate": 2e-05,
+      "loss": 0.5142,
+      "step": 17257
+    },
+    {
+      "epoch": 0.4107873940778825,
+      "grad_norm": 2.5672397581768154,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 17258
+    },
+    {
+      "epoch": 0.41081119680091405,
+      "grad_norm": 3.919975072761086,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 17259
+    },
+    {
+      "epoch": 0.41083499952394553,
+      "grad_norm": 8.066239566902096,
+      "learning_rate": 2e-05,
+      "loss": 1.3053,
+      "step": 17260
+    },
+    {
+      "epoch": 0.4108588022469771,
+      "grad_norm": 0.5853827412635516,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17261
+    },
+    {
+      "epoch": 0.41088260497000856,
+      "grad_norm": 10.730186676078977,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 17262
+    },
+    {
+      "epoch": 0.4109064076930401,
+      "grad_norm": 4.63106142274184,
+      "learning_rate": 2e-05,
+      "loss": 0.8903,
+      "step": 17263
+    },
+    {
+      "epoch": 0.4109302104160716,
+      "grad_norm": 6.897772533481958,
+      "learning_rate": 2e-05,
+      "loss": 0.548,
+      "step": 17264
+    },
+    {
+      "epoch": 0.4109540131391031,
+      "grad_norm": 1.010101915975199,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 17265
+    },
+    {
+      "epoch": 0.4109778158621346,
+      "grad_norm": 4.4436986291024345,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 17266
+    },
+    {
+      "epoch": 0.41100161858516615,
+      "grad_norm": 4.667634616155304,
+      "learning_rate": 2e-05,
+      "loss": 0.8855,
+      "step": 17267
+    },
+    {
+      "epoch": 0.41102542130819764,
+      "grad_norm": 5.859891005330422,
+      "learning_rate": 2e-05,
+      "loss": 1.3258,
+      "step": 17268
+    },
+    {
+      "epoch": 0.4110492240312292,
+      "grad_norm": 2.525662895230181,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 17269
+    },
+    {
+      "epoch": 0.41107302675426066,
+      "grad_norm": 10.066393986783366,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 17270
+    },
+    {
+      "epoch": 0.4110968294772922,
+      "grad_norm": 3.6938791145078294,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 17271
+    },
+    {
+      "epoch": 0.41112063220032374,
+      "grad_norm": 10.374546561660267,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 17272
+    },
+    {
+      "epoch": 0.41114443492335523,
+      "grad_norm": 1.975740301043684,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 17273
+    },
+    {
+      "epoch": 0.41116823764638677,
+      "grad_norm": 2.4430769463766837,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 17274
+    },
+    {
+      "epoch": 0.41119204036941825,
+      "grad_norm": 2.5637349013624258,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 17275
+    },
+    {
+      "epoch": 0.4112158430924498,
+      "grad_norm": 4.649212764400704,
+      "learning_rate": 2e-05,
+      "loss": 0.6423,
+      "step": 17276
+    },
+    {
+      "epoch": 0.4112396458154813,
+      "grad_norm": 6.07402958015706,
+      "learning_rate": 2e-05,
+      "loss": 0.4425,
+      "step": 17277
+    },
+    {
+      "epoch": 0.4112634485385128,
+      "grad_norm": 5.166996761696434,
+      "learning_rate": 2e-05,
+      "loss": 0.5866,
+      "step": 17278
+    },
+    {
+      "epoch": 0.4112872512615443,
+      "grad_norm": 4.600232554262866,
+      "learning_rate": 2e-05,
+      "loss": 0.7215,
+      "step": 17279
+    },
+    {
+      "epoch": 0.41131105398457585,
+      "grad_norm": 1.4113384698136864,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 17280
+    },
+    {
+      "epoch": 0.41133485670760733,
+      "grad_norm": 2.2864586933855553,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 17281
+    },
+    {
+      "epoch": 0.4113586594306389,
+      "grad_norm": 6.2551628176133,
+      "learning_rate": 2e-05,
+      "loss": 1.0105,
+      "step": 17282
+    },
+    {
+      "epoch": 0.41138246215367036,
+      "grad_norm": 5.458825511673146,
+      "learning_rate": 2e-05,
+      "loss": 0.5479,
+      "step": 17283
+    },
+    {
+      "epoch": 0.4114062648767019,
+      "grad_norm": 4.008301245017033,
+      "learning_rate": 2e-05,
+      "loss": 0.7521,
+      "step": 17284
+    },
+    {
+      "epoch": 0.4114300675997334,
+      "grad_norm": 3.024769095759379,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 17285
+    },
+    {
+      "epoch": 0.4114538703227649,
+      "grad_norm": 3.2107658177697207,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 17286
+    },
+    {
+      "epoch": 0.41147767304579647,
+      "grad_norm": 2.4713995342624515,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 17287
+    },
+    {
+      "epoch": 0.41150147576882795,
+      "grad_norm": 16.91930626210708,
+      "learning_rate": 2e-05,
+      "loss": 1.2247,
+      "step": 17288
+    },
+    {
+      "epoch": 0.4115252784918595,
+      "grad_norm": 3.95125306732777,
+      "learning_rate": 2e-05,
+      "loss": 0.4803,
+      "step": 17289
+    },
+    {
+      "epoch": 0.411549081214891,
+      "grad_norm": 5.635566582069963,
+      "learning_rate": 2e-05,
+      "loss": 0.1421,
+      "step": 17290
+    },
+    {
+      "epoch": 0.4115728839379225,
+      "grad_norm": 26.676919679998225,
+      "learning_rate": 2e-05,
+      "loss": 0.1579,
+      "step": 17291
+    },
+    {
+      "epoch": 0.411596686660954,
+      "grad_norm": 2.4155360893865887,
+      "learning_rate": 2e-05,
+      "loss": 0.1123,
+      "step": 17292
+    },
+    {
+      "epoch": 0.41162048938398554,
+      "grad_norm": 2.455025763269726,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 17293
+    },
+    {
+      "epoch": 0.41164429210701703,
+      "grad_norm": 1.06888025300719,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 17294
+    },
+    {
+      "epoch": 0.41166809483004857,
+      "grad_norm": 6.603466867539237,
+      "learning_rate": 2e-05,
+      "loss": 1.4292,
+      "step": 17295
+    },
+    {
+      "epoch": 0.41169189755308006,
+      "grad_norm": 14.943993512526706,
+      "learning_rate": 2e-05,
+      "loss": 1.5132,
+      "step": 17296
+    },
+    {
+      "epoch": 0.4117157002761116,
+      "grad_norm": 1.4275376605947312,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 17297
+    },
+    {
+      "epoch": 0.4117395029991431,
+      "grad_norm": 11.66825259082036,
+      "learning_rate": 2e-05,
+      "loss": 0.6747,
+      "step": 17298
+    },
+    {
+      "epoch": 0.4117633057221746,
+      "grad_norm": 2.0063651130333864,
+      "learning_rate": 2e-05,
+      "loss": 0.1894,
+      "step": 17299
+    },
+    {
+      "epoch": 0.4117871084452061,
+      "grad_norm": 4.00915238473364,
+      "learning_rate": 2e-05,
+      "loss": 0.681,
+      "step": 17300
+    },
+    {
+      "epoch": 0.41181091116823765,
+      "grad_norm": 4.479924323993425,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 17301
+    },
+    {
+      "epoch": 0.4118347138912692,
+      "grad_norm": 8.014506681577313,
+      "learning_rate": 2e-05,
+      "loss": 0.8337,
+      "step": 17302
+    },
+    {
+      "epoch": 0.4118585166143007,
+      "grad_norm": 0.7462822105370244,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17303
+    },
+    {
+      "epoch": 0.4118823193373322,
+      "grad_norm": 0.8500280682764333,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 17304
+    },
+    {
+      "epoch": 0.4119061220603637,
+      "grad_norm": 4.034165593151979,
+      "learning_rate": 2e-05,
+      "loss": 0.7737,
+      "step": 17305
+    },
+    {
+      "epoch": 0.41192992478339524,
+      "grad_norm": 1.564213682376923,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 17306
+    },
+    {
+      "epoch": 0.4119537275064267,
+      "grad_norm": 5.607725836264967,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 17307
+    },
+    {
+      "epoch": 0.41197753022945827,
+      "grad_norm": 14.972440966635249,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 17308
+    },
+    {
+      "epoch": 0.41200133295248975,
+      "grad_norm": 2.521840008925239,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 17309
+    },
+    {
+      "epoch": 0.4120251356755213,
+      "grad_norm": 15.446068378540831,
+      "learning_rate": 2e-05,
+      "loss": 0.755,
+      "step": 17310
+    },
+    {
+      "epoch": 0.4120489383985528,
+      "grad_norm": 6.689045218013191,
+      "learning_rate": 2e-05,
+      "loss": 1.158,
+      "step": 17311
+    },
+    {
+      "epoch": 0.4120727411215843,
+      "grad_norm": 0.7138053323477977,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 17312
+    },
+    {
+      "epoch": 0.4120965438446158,
+      "grad_norm": 5.676029934874964,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 17313
+    },
+    {
+      "epoch": 0.41212034656764734,
+      "grad_norm": 11.318847297107812,
+      "learning_rate": 2e-05,
+      "loss": 0.5698,
+      "step": 17314
+    },
+    {
+      "epoch": 0.41214414929067883,
+      "grad_norm": 5.475703960144281,
+      "learning_rate": 2e-05,
+      "loss": 0.6704,
+      "step": 17315
+    },
+    {
+      "epoch": 0.41216795201371037,
+      "grad_norm": 1.444511385878299,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 17316
+    },
+    {
+      "epoch": 0.4121917547367419,
+      "grad_norm": 12.598127520404738,
+      "learning_rate": 2e-05,
+      "loss": 0.767,
+      "step": 17317
+    },
+    {
+      "epoch": 0.4122155574597734,
+      "grad_norm": 4.390854464033773,
+      "learning_rate": 2e-05,
+      "loss": 0.1273,
+      "step": 17318
+    },
+    {
+      "epoch": 0.41223936018280494,
+      "grad_norm": 7.591940883819318,
+      "learning_rate": 2e-05,
+      "loss": 0.9861,
+      "step": 17319
+    },
+    {
+      "epoch": 0.4122631629058364,
+      "grad_norm": 45.39232551868602,
+      "learning_rate": 2e-05,
+      "loss": 0.3689,
+      "step": 17320
+    },
+    {
+      "epoch": 0.41228696562886796,
+      "grad_norm": 0.4838355866759372,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 17321
+    },
+    {
+      "epoch": 0.41231076835189945,
+      "grad_norm": 4.705775750892016,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 17322
+    },
+    {
+      "epoch": 0.412334571074931,
+      "grad_norm": 4.013902324382883,
+      "learning_rate": 2e-05,
+      "loss": 0.3693,
+      "step": 17323
+    },
+    {
+      "epoch": 0.4123583737979625,
+      "grad_norm": 2.7421436106527532,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 17324
+    },
+    {
+      "epoch": 0.412382176520994,
+      "grad_norm": 7.74240042196062,
+      "learning_rate": 2e-05,
+      "loss": 1.1446,
+      "step": 17325
+    },
+    {
+      "epoch": 0.4124059792440255,
+      "grad_norm": 6.362386245716907,
+      "learning_rate": 2e-05,
+      "loss": 0.246,
+      "step": 17326
+    },
+    {
+      "epoch": 0.41242978196705704,
+      "grad_norm": 6.149861442686042,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 17327
+    },
+    {
+      "epoch": 0.4124535846900885,
+      "grad_norm": 3.6202526638392225,
+      "learning_rate": 2e-05,
+      "loss": 0.5072,
+      "step": 17328
+    },
+    {
+      "epoch": 0.41247738741312007,
+      "grad_norm": 10.65553349828503,
+      "learning_rate": 2e-05,
+      "loss": 0.6796,
+      "step": 17329
+    },
+    {
+      "epoch": 0.41250119013615155,
+      "grad_norm": 3.4785202898070757,
+      "learning_rate": 2e-05,
+      "loss": 0.5997,
+      "step": 17330
+    },
+    {
+      "epoch": 0.4125249928591831,
+      "grad_norm": 9.15162626617235,
+      "learning_rate": 2e-05,
+      "loss": 0.4672,
+      "step": 17331
+    },
+    {
+      "epoch": 0.4125487955822146,
+      "grad_norm": 22.014298337918383,
+      "learning_rate": 2e-05,
+      "loss": 0.494,
+      "step": 17332
+    },
+    {
+      "epoch": 0.4125725983052461,
+      "grad_norm": 3.0190311771699894,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 17333
+    },
+    {
+      "epoch": 0.41259640102827766,
+      "grad_norm": 4.33572047209917,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 17334
+    },
+    {
+      "epoch": 0.41262020375130914,
+      "grad_norm": 2.195317540714589,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 17335
+    },
+    {
+      "epoch": 0.4126440064743407,
+      "grad_norm": 3.9404900188898613,
+      "learning_rate": 2e-05,
+      "loss": 0.5505,
+      "step": 17336
+    },
+    {
+      "epoch": 0.41266780919737217,
+      "grad_norm": 5.385173137728865,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 17337
+    },
+    {
+      "epoch": 0.4126916119204037,
+      "grad_norm": 2.0524642107938997,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 17338
+    },
+    {
+      "epoch": 0.4127154146434352,
+      "grad_norm": 3.4171434148685282,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 17339
+    },
+    {
+      "epoch": 0.41273921736646674,
+      "grad_norm": 3.0705141744584803,
+      "learning_rate": 2e-05,
+      "loss": 0.494,
+      "step": 17340
+    },
+    {
+      "epoch": 0.4127630200894982,
+      "grad_norm": 26.80703402827273,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 17341
+    },
+    {
+      "epoch": 0.41278682281252976,
+      "grad_norm": 3.028990799060753,
+      "learning_rate": 2e-05,
+      "loss": 0.5669,
+      "step": 17342
+    },
+    {
+      "epoch": 0.41281062553556125,
+      "grad_norm": 4.239062280380266,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 17343
+    },
+    {
+      "epoch": 0.4128344282585928,
+      "grad_norm": 7.832363006086446,
+      "learning_rate": 2e-05,
+      "loss": 0.2148,
+      "step": 17344
+    },
+    {
+      "epoch": 0.4128582309816243,
+      "grad_norm": 0.5605378106924592,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 17345
+    },
+    {
+      "epoch": 0.4128820337046558,
+      "grad_norm": 8.777135314432567,
+      "learning_rate": 2e-05,
+      "loss": 0.8436,
+      "step": 17346
+    },
+    {
+      "epoch": 0.4129058364276873,
+      "grad_norm": 4.447154248496244,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 17347
+    },
+    {
+      "epoch": 0.41292963915071884,
+      "grad_norm": 3.327645353793728,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 17348
+    },
+    {
+      "epoch": 0.4129534418737504,
+      "grad_norm": 7.703066355688523,
+      "learning_rate": 2e-05,
+      "loss": 1.5796,
+      "step": 17349
+    },
+    {
+      "epoch": 0.41297724459678187,
+      "grad_norm": 3.0878887766007055,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 17350
+    },
+    {
+      "epoch": 0.4130010473198134,
+      "grad_norm": 2.640726351983737,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 17351
+    },
+    {
+      "epoch": 0.4130248500428449,
+      "grad_norm": 5.227011592459847,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 17352
+    },
+    {
+      "epoch": 0.41304865276587643,
+      "grad_norm": 9.224633091780339,
+      "learning_rate": 2e-05,
+      "loss": 0.6581,
+      "step": 17353
+    },
+    {
+      "epoch": 0.4130724554889079,
+      "grad_norm": 6.7423814983991806,
+      "learning_rate": 2e-05,
+      "loss": 0.9728,
+      "step": 17354
+    },
+    {
+      "epoch": 0.41309625821193946,
+      "grad_norm": 2.4663446667700377,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 17355
+    },
+    {
+      "epoch": 0.41312006093497095,
+      "grad_norm": 2.0408476164498084,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 17356
+    },
+    {
+      "epoch": 0.4131438636580025,
+      "grad_norm": 4.9894121458614125,
+      "learning_rate": 2e-05,
+      "loss": 0.563,
+      "step": 17357
+    },
+    {
+      "epoch": 0.41316766638103397,
+      "grad_norm": 10.428215564252634,
+      "learning_rate": 2e-05,
+      "loss": 0.3599,
+      "step": 17358
+    },
+    {
+      "epoch": 0.4131914691040655,
+      "grad_norm": 3.779007506431919,
+      "learning_rate": 2e-05,
+      "loss": 0.6606,
+      "step": 17359
+    },
+    {
+      "epoch": 0.413215271827097,
+      "grad_norm": 4.645858991756843,
+      "learning_rate": 2e-05,
+      "loss": 0.6835,
+      "step": 17360
+    },
+    {
+      "epoch": 0.41323907455012854,
+      "grad_norm": 1.607751972946741,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 17361
+    },
+    {
+      "epoch": 0.41326287727316,
+      "grad_norm": 0.6044717276840776,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 17362
+    },
+    {
+      "epoch": 0.41328667999619156,
+      "grad_norm": 2.3537898732113356,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 17363
+    },
+    {
+      "epoch": 0.4133104827192231,
+      "grad_norm": 1.713375423661553,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 17364
+    },
+    {
+      "epoch": 0.4133342854422546,
+      "grad_norm": 12.153133584758535,
+      "learning_rate": 2e-05,
+      "loss": 0.4071,
+      "step": 17365
+    },
+    {
+      "epoch": 0.41335808816528613,
+      "grad_norm": 3.955020806851507,
+      "learning_rate": 2e-05,
+      "loss": 0.3706,
+      "step": 17366
+    },
+    {
+      "epoch": 0.4133818908883176,
+      "grad_norm": 2.894886185248336,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 17367
+    },
+    {
+      "epoch": 0.41340569361134916,
+      "grad_norm": 0.21445638432633315,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17368
+    },
+    {
+      "epoch": 0.41342949633438064,
+      "grad_norm": 5.448153049526366,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 17369
+    },
+    {
+      "epoch": 0.4134532990574122,
+      "grad_norm": 3.134758941150519,
+      "learning_rate": 2e-05,
+      "loss": 0.589,
+      "step": 17370
+    },
+    {
+      "epoch": 0.41347710178044367,
+      "grad_norm": 3.890599658170927,
+      "learning_rate": 2e-05,
+      "loss": 0.6631,
+      "step": 17371
+    },
+    {
+      "epoch": 0.4135009045034752,
+      "grad_norm": 1.2349718894369552,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 17372
+    },
+    {
+      "epoch": 0.4135247072265067,
+      "grad_norm": 3.005577200736404,
+      "learning_rate": 2e-05,
+      "loss": 0.4734,
+      "step": 17373
+    },
+    {
+      "epoch": 0.41354850994953823,
+      "grad_norm": 1.7364811930795403,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 17374
+    },
+    {
+      "epoch": 0.4135723126725697,
+      "grad_norm": 5.771477618062416,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 17375
+    },
+    {
+      "epoch": 0.41359611539560126,
+      "grad_norm": 6.157069486870395,
+      "learning_rate": 2e-05,
+      "loss": 0.4622,
+      "step": 17376
+    },
+    {
+      "epoch": 0.41361991811863275,
+      "grad_norm": 3.1164563930917377,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 17377
+    },
+    {
+      "epoch": 0.4136437208416643,
+      "grad_norm": 8.870767313951347,
+      "learning_rate": 2e-05,
+      "loss": 0.9201,
+      "step": 17378
+    },
+    {
+      "epoch": 0.4136675235646958,
+      "grad_norm": 3.046737776444733,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 17379
+    },
+    {
+      "epoch": 0.4136913262877273,
+      "grad_norm": 1.8128463582720353,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17380
+    },
+    {
+      "epoch": 0.41371512901075885,
+      "grad_norm": 2.538336702496516,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 17381
+    },
+    {
+      "epoch": 0.41373893173379034,
+      "grad_norm": 0.6470795035047572,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 17382
+    },
+    {
+      "epoch": 0.4137627344568219,
+      "grad_norm": 3.7314351111959554,
+      "learning_rate": 2e-05,
+      "loss": 0.4044,
+      "step": 17383
+    },
+    {
+      "epoch": 0.41378653717985336,
+      "grad_norm": 4.741795948074941,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 17384
+    },
+    {
+      "epoch": 0.4138103399028849,
+      "grad_norm": 2.7917407296925196,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 17385
+    },
+    {
+      "epoch": 0.4138341426259164,
+      "grad_norm": 3.3928593499574844,
+      "learning_rate": 2e-05,
+      "loss": 0.6244,
+      "step": 17386
+    },
+    {
+      "epoch": 0.41385794534894793,
+      "grad_norm": 4.669187252950296,
+      "learning_rate": 2e-05,
+      "loss": 0.6781,
+      "step": 17387
+    },
+    {
+      "epoch": 0.4138817480719794,
+      "grad_norm": 2.9040067026202516,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 17388
+    },
+    {
+      "epoch": 0.41390555079501096,
+      "grad_norm": 5.629651539792389,
+      "learning_rate": 2e-05,
+      "loss": 0.1623,
+      "step": 17389
+    },
+    {
+      "epoch": 0.41392935351804244,
+      "grad_norm": 0.8859912880852304,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17390
+    },
+    {
+      "epoch": 0.413953156241074,
+      "grad_norm": 4.999162282255801,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 17391
+    },
+    {
+      "epoch": 0.41397695896410547,
+      "grad_norm": 3.0047890755594775,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 17392
+    },
+    {
+      "epoch": 0.414000761687137,
+      "grad_norm": 2.6088926490747184,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 17393
+    },
+    {
+      "epoch": 0.41402456441016855,
+      "grad_norm": 10.479692214112966,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 17394
+    },
+    {
+      "epoch": 0.41404836713320003,
+      "grad_norm": 1.1503229164036675,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 17395
+    },
+    {
+      "epoch": 0.4140721698562316,
+      "grad_norm": 4.818515759740215,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 17396
+    },
+    {
+      "epoch": 0.41409597257926306,
+      "grad_norm": 3.61990916883586,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 17397
+    },
+    {
+      "epoch": 0.4141197753022946,
+      "grad_norm": 2.2631096373160053,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 17398
+    },
+    {
+      "epoch": 0.4141435780253261,
+      "grad_norm": 3.823319689431794,
+      "learning_rate": 2e-05,
+      "loss": 0.7925,
+      "step": 17399
+    },
+    {
+      "epoch": 0.4141673807483576,
+      "grad_norm": 6.096125110423422,
+      "learning_rate": 2e-05,
+      "loss": 0.9525,
+      "step": 17400
+    },
+    {
+      "epoch": 0.4141911834713891,
+      "grad_norm": 3.1643205693016268,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 17401
+    },
+    {
+      "epoch": 0.41421498619442065,
+      "grad_norm": 4.784678975759937,
+      "learning_rate": 2e-05,
+      "loss": 0.6424,
+      "step": 17402
+    },
+    {
+      "epoch": 0.41423878891745214,
+      "grad_norm": 1.5733605577201426,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 17403
+    },
+    {
+      "epoch": 0.4142625916404837,
+      "grad_norm": 8.732038873177379,
+      "learning_rate": 2e-05,
+      "loss": 0.2904,
+      "step": 17404
+    },
+    {
+      "epoch": 0.41428639436351516,
+      "grad_norm": 4.018141441428609,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 17405
+    },
+    {
+      "epoch": 0.4143101970865467,
+      "grad_norm": 9.279775086864234,
+      "learning_rate": 2e-05,
+      "loss": 0.3749,
+      "step": 17406
+    },
+    {
+      "epoch": 0.4143339998095782,
+      "grad_norm": 5.043588243132352,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 17407
+    },
+    {
+      "epoch": 0.41435780253260973,
+      "grad_norm": 5.487094556440774,
+      "learning_rate": 2e-05,
+      "loss": 1.0821,
+      "step": 17408
+    },
+    {
+      "epoch": 0.41438160525564127,
+      "grad_norm": 0.169201035178772,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 17409
+    },
+    {
+      "epoch": 0.41440540797867276,
+      "grad_norm": 2.1130907358518196,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 17410
+    },
+    {
+      "epoch": 0.4144292107017043,
+      "grad_norm": 4.525906607465714,
+      "learning_rate": 2e-05,
+      "loss": 0.1827,
+      "step": 17411
+    },
+    {
+      "epoch": 0.4144530134247358,
+      "grad_norm": 1.7126087661854785,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 17412
+    },
+    {
+      "epoch": 0.4144768161477673,
+      "grad_norm": 3.994873991007605,
+      "learning_rate": 2e-05,
+      "loss": 0.637,
+      "step": 17413
+    },
+    {
+      "epoch": 0.4145006188707988,
+      "grad_norm": 2.0960864574165132,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 17414
+    },
+    {
+      "epoch": 0.41452442159383035,
+      "grad_norm": 4.52497123828531,
+      "learning_rate": 2e-05,
+      "loss": 0.8238,
+      "step": 17415
+    },
+    {
+      "epoch": 0.41454822431686184,
+      "grad_norm": 5.506213018505356,
+      "learning_rate": 2e-05,
+      "loss": 0.7197,
+      "step": 17416
+    },
+    {
+      "epoch": 0.4145720270398934,
+      "grad_norm": 3.3108001101018547,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 17417
+    },
+    {
+      "epoch": 0.41459582976292486,
+      "grad_norm": 6.128609496535885,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 17418
+    },
+    {
+      "epoch": 0.4146196324859564,
+      "grad_norm": 4.186913045399835,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 17419
+    },
+    {
+      "epoch": 0.4146434352089879,
+      "grad_norm": 4.402162043815555,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 17420
+    },
+    {
+      "epoch": 0.41466723793201943,
+      "grad_norm": 3.7280657493318277,
+      "learning_rate": 2e-05,
+      "loss": 0.8534,
+      "step": 17421
+    },
+    {
+      "epoch": 0.4146910406550509,
+      "grad_norm": 4.941778217121801,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 17422
+    },
+    {
+      "epoch": 0.41471484337808245,
+      "grad_norm": 3.4713125791881225,
+      "learning_rate": 2e-05,
+      "loss": 0.2313,
+      "step": 17423
+    },
+    {
+      "epoch": 0.414738646101114,
+      "grad_norm": 1.8238982350365414,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 17424
+    },
+    {
+      "epoch": 0.4147624488241455,
+      "grad_norm": 4.403920696909378,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 17425
+    },
+    {
+      "epoch": 0.414786251547177,
+      "grad_norm": 2.266356983924017,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 17426
+    },
+    {
+      "epoch": 0.4148100542702085,
+      "grad_norm": 1.285346682557179,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 17427
+    },
+    {
+      "epoch": 0.41483385699324005,
+      "grad_norm": 4.475442952957348,
+      "learning_rate": 2e-05,
+      "loss": 0.4302,
+      "step": 17428
+    },
+    {
+      "epoch": 0.41485765971627153,
+      "grad_norm": 4.440268420497237,
+      "learning_rate": 2e-05,
+      "loss": 0.5411,
+      "step": 17429
+    },
+    {
+      "epoch": 0.4148814624393031,
+      "grad_norm": 5.480965084358165,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 17430
+    },
+    {
+      "epoch": 0.41490526516233456,
+      "grad_norm": 5.490473118415316,
+      "learning_rate": 2e-05,
+      "loss": 0.5045,
+      "step": 17431
+    },
+    {
+      "epoch": 0.4149290678853661,
+      "grad_norm": 7.3719819150965495,
+      "learning_rate": 2e-05,
+      "loss": 0.4318,
+      "step": 17432
+    },
+    {
+      "epoch": 0.4149528706083976,
+      "grad_norm": 4.551974224249539,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 17433
+    },
+    {
+      "epoch": 0.4149766733314291,
+      "grad_norm": 4.6926116036331145,
+      "learning_rate": 2e-05,
+      "loss": 0.4888,
+      "step": 17434
+    },
+    {
+      "epoch": 0.4150004760544606,
+      "grad_norm": 3.050602761543991,
+      "learning_rate": 2e-05,
+      "loss": 0.3988,
+      "step": 17435
+    },
+    {
+      "epoch": 0.41502427877749215,
+      "grad_norm": 5.971496579789934,
+      "learning_rate": 2e-05,
+      "loss": 1.3622,
+      "step": 17436
+    },
+    {
+      "epoch": 0.41504808150052364,
+      "grad_norm": 12.397105493476364,
+      "learning_rate": 2e-05,
+      "loss": 0.4126,
+      "step": 17437
+    },
+    {
+      "epoch": 0.4150718842235552,
+      "grad_norm": 2.9239061253008534,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 17438
+    },
+    {
+      "epoch": 0.4150956869465867,
+      "grad_norm": 0.5185346013826202,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 17439
+    },
+    {
+      "epoch": 0.4151194896696182,
+      "grad_norm": 0.9138932159778946,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 17440
+    },
+    {
+      "epoch": 0.41514329239264974,
+      "grad_norm": 4.70415080723159,
+      "learning_rate": 2e-05,
+      "loss": 0.5321,
+      "step": 17441
+    },
+    {
+      "epoch": 0.41516709511568123,
+      "grad_norm": 0.2609880883617882,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17442
+    },
+    {
+      "epoch": 0.41519089783871277,
+      "grad_norm": 24.58780431013341,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 17443
+    },
+    {
+      "epoch": 0.41521470056174425,
+      "grad_norm": 6.216567386791932,
+      "learning_rate": 2e-05,
+      "loss": 1.4696,
+      "step": 17444
+    },
+    {
+      "epoch": 0.4152385032847758,
+      "grad_norm": 0.4860323395371507,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 17445
+    },
+    {
+      "epoch": 0.4152623060078073,
+      "grad_norm": 1.4609804414087717,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 17446
+    },
+    {
+      "epoch": 0.4152861087308388,
+      "grad_norm": 3.887279502274401,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 17447
+    },
+    {
+      "epoch": 0.4153099114538703,
+      "grad_norm": 1.823318891949465,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 17448
+    },
+    {
+      "epoch": 0.41533371417690185,
+      "grad_norm": 0.23547746615084414,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 17449
+    },
+    {
+      "epoch": 0.41535751689993333,
+      "grad_norm": 7.327396570447083,
+      "learning_rate": 2e-05,
+      "loss": 0.8256,
+      "step": 17450
+    },
+    {
+      "epoch": 0.4153813196229649,
+      "grad_norm": 6.958983585014048,
+      "learning_rate": 2e-05,
+      "loss": 0.6047,
+      "step": 17451
+    },
+    {
+      "epoch": 0.41540512234599636,
+      "grad_norm": 3.3480867577065214,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 17452
+    },
+    {
+      "epoch": 0.4154289250690279,
+      "grad_norm": 3.8983023352893342,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 17453
+    },
+    {
+      "epoch": 0.4154527277920594,
+      "grad_norm": 4.135800275514278,
+      "learning_rate": 2e-05,
+      "loss": 0.6864,
+      "step": 17454
+    },
+    {
+      "epoch": 0.4154765305150909,
+      "grad_norm": 1.0600441430839596,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 17455
+    },
+    {
+      "epoch": 0.41550033323812247,
+      "grad_norm": 0.4044873853623426,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17456
+    },
+    {
+      "epoch": 0.41552413596115395,
+      "grad_norm": 2.724889689499028,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 17457
+    },
+    {
+      "epoch": 0.4155479386841855,
+      "grad_norm": 1.3186677537300784,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 17458
+    },
+    {
+      "epoch": 0.415571741407217,
+      "grad_norm": 3.343546337917015,
+      "learning_rate": 2e-05,
+      "loss": 0.1874,
+      "step": 17459
+    },
+    {
+      "epoch": 0.4155955441302485,
+      "grad_norm": 14.400606166104101,
+      "learning_rate": 2e-05,
+      "loss": 0.804,
+      "step": 17460
+    },
+    {
+      "epoch": 0.41561934685328,
+      "grad_norm": 10.116651479748215,
+      "learning_rate": 2e-05,
+      "loss": 0.4414,
+      "step": 17461
+    },
+    {
+      "epoch": 0.41564314957631154,
+      "grad_norm": 2.768859350861681,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 17462
+    },
+    {
+      "epoch": 0.41566695229934303,
+      "grad_norm": 3.62552471478038,
+      "learning_rate": 2e-05,
+      "loss": 0.4358,
+      "step": 17463
+    },
+    {
+      "epoch": 0.41569075502237457,
+      "grad_norm": 30.192544736345006,
+      "learning_rate": 2e-05,
+      "loss": 0.6232,
+      "step": 17464
+    },
+    {
+      "epoch": 0.41571455774540605,
+      "grad_norm": 3.490777594460172,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 17465
+    },
+    {
+      "epoch": 0.4157383604684376,
+      "grad_norm": 3.626627809783103,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 17466
+    },
+    {
+      "epoch": 0.4157621631914691,
+      "grad_norm": 4.96363882302348,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 17467
+    },
+    {
+      "epoch": 0.4157859659145006,
+      "grad_norm": 1.7630254771685299,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 17468
+    },
+    {
+      "epoch": 0.4158097686375321,
+      "grad_norm": 4.911759138788303,
+      "learning_rate": 2e-05,
+      "loss": 0.8141,
+      "step": 17469
+    },
+    {
+      "epoch": 0.41583357136056365,
+      "grad_norm": 1.1352151566660877,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 17470
+    },
+    {
+      "epoch": 0.4158573740835952,
+      "grad_norm": 5.965188165068206,
+      "learning_rate": 2e-05,
+      "loss": 0.7619,
+      "step": 17471
+    },
+    {
+      "epoch": 0.4158811768066267,
+      "grad_norm": 4.059098199674821,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 17472
+    },
+    {
+      "epoch": 0.4159049795296582,
+      "grad_norm": 2.2858915834168765,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 17473
+    },
+    {
+      "epoch": 0.4159287822526897,
+      "grad_norm": 4.547100012727119,
+      "learning_rate": 2e-05,
+      "loss": 0.696,
+      "step": 17474
+    },
+    {
+      "epoch": 0.41595258497572124,
+      "grad_norm": 16.247257811613785,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 17475
+    },
+    {
+      "epoch": 0.4159763876987527,
+      "grad_norm": 1.589211445699706,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17476
+    },
+    {
+      "epoch": 0.41600019042178427,
+      "grad_norm": 12.301245447342865,
+      "learning_rate": 2e-05,
+      "loss": 0.5215,
+      "step": 17477
+    },
+    {
+      "epoch": 0.41602399314481575,
+      "grad_norm": 2.0339050395685896,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 17478
+    },
+    {
+      "epoch": 0.4160477958678473,
+      "grad_norm": 8.591335231229113,
+      "learning_rate": 2e-05,
+      "loss": 0.2765,
+      "step": 17479
+    },
+    {
+      "epoch": 0.4160715985908788,
+      "grad_norm": 2.0498124564334286,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 17480
+    },
+    {
+      "epoch": 0.4160954013139103,
+      "grad_norm": 12.406399479124115,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 17481
+    },
+    {
+      "epoch": 0.4161192040369418,
+      "grad_norm": 5.96322783867319,
+      "learning_rate": 2e-05,
+      "loss": 1.3005,
+      "step": 17482
+    },
+    {
+      "epoch": 0.41614300675997334,
+      "grad_norm": 1.8108436557825267,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 17483
+    },
+    {
+      "epoch": 0.41616680948300483,
+      "grad_norm": 7.865664908488403,
+      "learning_rate": 2e-05,
+      "loss": 1.3453,
+      "step": 17484
+    },
+    {
+      "epoch": 0.41619061220603637,
+      "grad_norm": 1.3798019927948857,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 17485
+    },
+    {
+      "epoch": 0.4162144149290679,
+      "grad_norm": 6.364531442525118,
+      "learning_rate": 2e-05,
+      "loss": 0.7819,
+      "step": 17486
+    },
+    {
+      "epoch": 0.4162382176520994,
+      "grad_norm": 6.767194364120678,
+      "learning_rate": 2e-05,
+      "loss": 0.5388,
+      "step": 17487
+    },
+    {
+      "epoch": 0.41626202037513094,
+      "grad_norm": 1.357562306313215,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 17488
+    },
+    {
+      "epoch": 0.4162858230981624,
+      "grad_norm": 2.844979890258961,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 17489
+    },
+    {
+      "epoch": 0.41630962582119396,
+      "grad_norm": 2.078990643431668,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 17490
+    },
+    {
+      "epoch": 0.41633342854422545,
+      "grad_norm": 3.4303030697114325,
+      "learning_rate": 2e-05,
+      "loss": 0.3032,
+      "step": 17491
+    },
+    {
+      "epoch": 0.416357231267257,
+      "grad_norm": 0.6539234670917233,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 17492
+    },
+    {
+      "epoch": 0.4163810339902885,
+      "grad_norm": 4.59408712163031,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 17493
+    },
+    {
+      "epoch": 0.41640483671332,
+      "grad_norm": 0.7440832590673008,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17494
+    },
+    {
+      "epoch": 0.4164286394363515,
+      "grad_norm": 0.16776820222905195,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17495
+    },
+    {
+      "epoch": 0.41645244215938304,
+      "grad_norm": 0.22553303703215302,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 17496
+    },
+    {
+      "epoch": 0.4164762448824145,
+      "grad_norm": 5.210801661732269,
+      "learning_rate": 2e-05,
+      "loss": 0.8672,
+      "step": 17497
+    },
+    {
+      "epoch": 0.41650004760544607,
+      "grad_norm": 12.277825013762314,
+      "learning_rate": 2e-05,
+      "loss": 0.3497,
+      "step": 17498
+    },
+    {
+      "epoch": 0.41652385032847755,
+      "grad_norm": 26.27440053411321,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 17499
+    },
+    {
+      "epoch": 0.4165476530515091,
+      "grad_norm": 5.05367002905693,
+      "learning_rate": 2e-05,
+      "loss": 1.0936,
+      "step": 17500
+    },
+    {
+      "epoch": 0.41657145577454063,
+      "grad_norm": 2.8291876368183506,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 17501
+    },
+    {
+      "epoch": 0.4165952584975721,
+      "grad_norm": 2.6094060538924677,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 17502
+    },
+    {
+      "epoch": 0.41661906122060366,
+      "grad_norm": 8.077549054938617,
+      "learning_rate": 2e-05,
+      "loss": 0.5633,
+      "step": 17503
+    },
+    {
+      "epoch": 0.41664286394363514,
+      "grad_norm": 2.210951931224536,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 17504
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 8.865257535298717,
+      "learning_rate": 2e-05,
+      "loss": 1.4685,
+      "step": 17505
+    },
+    {
+      "epoch": 0.41669046938969817,
+      "grad_norm": 2.0838809550437105,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 17506
+    },
+    {
+      "epoch": 0.4167142721127297,
+      "grad_norm": 3.3329568358694224,
+      "learning_rate": 2e-05,
+      "loss": 0.421,
+      "step": 17507
+    },
+    {
+      "epoch": 0.4167380748357612,
+      "grad_norm": 7.119331589840431,
+      "learning_rate": 2e-05,
+      "loss": 0.9452,
+      "step": 17508
+    },
+    {
+      "epoch": 0.41676187755879274,
+      "grad_norm": 8.91202281178048,
+      "learning_rate": 2e-05,
+      "loss": 0.5531,
+      "step": 17509
+    },
+    {
+      "epoch": 0.4167856802818242,
+      "grad_norm": 4.694745401579601,
+      "learning_rate": 2e-05,
+      "loss": 0.7283,
+      "step": 17510
+    },
+    {
+      "epoch": 0.41680948300485576,
+      "grad_norm": 2.792580132471616,
+      "learning_rate": 2e-05,
+      "loss": 0.4459,
+      "step": 17511
+    },
+    {
+      "epoch": 0.41683328572788725,
+      "grad_norm": 5.010240774115324,
+      "learning_rate": 2e-05,
+      "loss": 0.8286,
+      "step": 17512
+    },
+    {
+      "epoch": 0.4168570884509188,
+      "grad_norm": 3.216107653018246,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 17513
+    },
+    {
+      "epoch": 0.4168808911739503,
+      "grad_norm": 3.8573531866804878,
+      "learning_rate": 2e-05,
+      "loss": 0.4406,
+      "step": 17514
+    },
+    {
+      "epoch": 0.4169046938969818,
+      "grad_norm": 3.531361651096211,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 17515
+    },
+    {
+      "epoch": 0.41692849662001336,
+      "grad_norm": 10.872674488963726,
+      "learning_rate": 2e-05,
+      "loss": 0.8447,
+      "step": 17516
+    },
+    {
+      "epoch": 0.41695229934304484,
+      "grad_norm": 4.1106957825188095,
+      "learning_rate": 2e-05,
+      "loss": 0.6557,
+      "step": 17517
+    },
+    {
+      "epoch": 0.4169761020660764,
+      "grad_norm": 3.766120958136288,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 17518
+    },
+    {
+      "epoch": 0.41699990478910787,
+      "grad_norm": 5.8834471598361455,
+      "learning_rate": 2e-05,
+      "loss": 0.761,
+      "step": 17519
+    },
+    {
+      "epoch": 0.4170237075121394,
+      "grad_norm": 2.5110984005933554,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 17520
+    },
+    {
+      "epoch": 0.4170475102351709,
+      "grad_norm": 6.079681363790627,
+      "learning_rate": 2e-05,
+      "loss": 0.7054,
+      "step": 17521
+    },
+    {
+      "epoch": 0.41707131295820243,
+      "grad_norm": 1.9657828399440287,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 17522
+    },
+    {
+      "epoch": 0.4170951156812339,
+      "grad_norm": 5.936419835370246,
+      "learning_rate": 2e-05,
+      "loss": 1.2846,
+      "step": 17523
+    },
+    {
+      "epoch": 0.41711891840426546,
+      "grad_norm": 4.721416113326705,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 17524
+    },
+    {
+      "epoch": 0.41714272112729694,
+      "grad_norm": 0.254784176658643,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17525
+    },
+    {
+      "epoch": 0.4171665238503285,
+      "grad_norm": 2.576670217898495,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 17526
+    },
+    {
+      "epoch": 0.41719032657335997,
+      "grad_norm": 0.25096715927526175,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 17527
+    },
+    {
+      "epoch": 0.4172141292963915,
+      "grad_norm": 2.4506149177207948,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 17528
+    },
+    {
+      "epoch": 0.417237932019423,
+      "grad_norm": 4.2664208654628775,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 17529
+    },
+    {
+      "epoch": 0.41726173474245454,
+      "grad_norm": 4.6845375842973445,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 17530
+    },
+    {
+      "epoch": 0.4172855374654861,
+      "grad_norm": 0.720894881487724,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 17531
+    },
+    {
+      "epoch": 0.41730934018851756,
+      "grad_norm": 2.0484727282116144,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 17532
+    },
+    {
+      "epoch": 0.4173331429115491,
+      "grad_norm": 6.462991203841826,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 17533
+    },
+    {
+      "epoch": 0.4173569456345806,
+      "grad_norm": 3.6635751099316325,
+      "learning_rate": 2e-05,
+      "loss": 0.6776,
+      "step": 17534
+    },
+    {
+      "epoch": 0.41738074835761213,
+      "grad_norm": 4.6469115333071835,
+      "learning_rate": 2e-05,
+      "loss": 0.9579,
+      "step": 17535
+    },
+    {
+      "epoch": 0.4174045510806436,
+      "grad_norm": 4.973838527487629,
+      "learning_rate": 2e-05,
+      "loss": 0.8212,
+      "step": 17536
+    },
+    {
+      "epoch": 0.41742835380367516,
+      "grad_norm": 4.65943016666766,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 17537
+    },
+    {
+      "epoch": 0.41745215652670664,
+      "grad_norm": 3.250764610671128,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 17538
+    },
+    {
+      "epoch": 0.4174759592497382,
+      "grad_norm": 2.1190293367898843,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 17539
+    },
+    {
+      "epoch": 0.41749976197276967,
+      "grad_norm": 9.951044040998,
+      "learning_rate": 2e-05,
+      "loss": 1.0544,
+      "step": 17540
+    },
+    {
+      "epoch": 0.4175235646958012,
+      "grad_norm": 1.9308691025229365,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 17541
+    },
+    {
+      "epoch": 0.4175473674188327,
+      "grad_norm": 3.549529484115593,
+      "learning_rate": 2e-05,
+      "loss": 0.5358,
+      "step": 17542
+    },
+    {
+      "epoch": 0.41757117014186423,
+      "grad_norm": 3.2361206175618866,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 17543
+    },
+    {
+      "epoch": 0.4175949728648957,
+      "grad_norm": 8.06095535367296,
+      "learning_rate": 2e-05,
+      "loss": 0.2652,
+      "step": 17544
+    },
+    {
+      "epoch": 0.41761877558792726,
+      "grad_norm": 3.8186223426388493,
+      "learning_rate": 2e-05,
+      "loss": 0.6567,
+      "step": 17545
+    },
+    {
+      "epoch": 0.4176425783109588,
+      "grad_norm": 6.302201848390028,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 17546
+    },
+    {
+      "epoch": 0.4176663810339903,
+      "grad_norm": 3.663975900946344,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 17547
+    },
+    {
+      "epoch": 0.4176901837570218,
+      "grad_norm": 0.5837887399749943,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 17548
+    },
+    {
+      "epoch": 0.4177139864800533,
+      "grad_norm": 2.9683715580476275,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 17549
+    },
+    {
+      "epoch": 0.41773778920308485,
+      "grad_norm": 4.484624155712276,
+      "learning_rate": 2e-05,
+      "loss": 0.4967,
+      "step": 17550
+    },
+    {
+      "epoch": 0.41776159192611634,
+      "grad_norm": 11.114404955661602,
+      "learning_rate": 2e-05,
+      "loss": 0.6918,
+      "step": 17551
+    },
+    {
+      "epoch": 0.4177853946491479,
+      "grad_norm": 4.3292147880176435,
+      "learning_rate": 2e-05,
+      "loss": 0.1906,
+      "step": 17552
+    },
+    {
+      "epoch": 0.41780919737217936,
+      "grad_norm": 696.8876450405946,
+      "learning_rate": 2e-05,
+      "loss": 1.3771,
+      "step": 17553
+    },
+    {
+      "epoch": 0.4178330000952109,
+      "grad_norm": 2.9297704273570586,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 17554
+    },
+    {
+      "epoch": 0.4178568028182424,
+      "grad_norm": 5.890556598420361,
+      "learning_rate": 2e-05,
+      "loss": 0.6244,
+      "step": 17555
+    },
+    {
+      "epoch": 0.41788060554127393,
+      "grad_norm": 5.687971000252711,
+      "learning_rate": 2e-05,
+      "loss": 0.4362,
+      "step": 17556
+    },
+    {
+      "epoch": 0.4179044082643054,
+      "grad_norm": 8.976744047929817,
+      "learning_rate": 2e-05,
+      "loss": 0.888,
+      "step": 17557
+    },
+    {
+      "epoch": 0.41792821098733696,
+      "grad_norm": 5.276431089081933,
+      "learning_rate": 2e-05,
+      "loss": 0.1985,
+      "step": 17558
+    },
+    {
+      "epoch": 0.41795201371036844,
+      "grad_norm": 1.7187560145387857,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 17559
+    },
+    {
+      "epoch": 0.4179758164334,
+      "grad_norm": 11.179604518502147,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 17560
+    },
+    {
+      "epoch": 0.4179996191564315,
+      "grad_norm": 0.5406833014002435,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 17561
+    },
+    {
+      "epoch": 0.418023421879463,
+      "grad_norm": 5.836161174530594,
+      "learning_rate": 2e-05,
+      "loss": 1.3236,
+      "step": 17562
+    },
+    {
+      "epoch": 0.41804722460249455,
+      "grad_norm": 1.2568606542094112,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 17563
+    },
+    {
+      "epoch": 0.41807102732552603,
+      "grad_norm": 2.481796142051097,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 17564
+    },
+    {
+      "epoch": 0.4180948300485576,
+      "grad_norm": 14.004366851903097,
+      "learning_rate": 2e-05,
+      "loss": 0.4584,
+      "step": 17565
+    },
+    {
+      "epoch": 0.41811863277158906,
+      "grad_norm": 8.630733595021216,
+      "learning_rate": 2e-05,
+      "loss": 0.6641,
+      "step": 17566
+    },
+    {
+      "epoch": 0.4181424354946206,
+      "grad_norm": 3.826254539877625,
+      "learning_rate": 2e-05,
+      "loss": 0.5213,
+      "step": 17567
+    },
+    {
+      "epoch": 0.4181662382176521,
+      "grad_norm": 15.261308206515945,
+      "learning_rate": 2e-05,
+      "loss": 0.544,
+      "step": 17568
+    },
+    {
+      "epoch": 0.4181900409406836,
+      "grad_norm": 6.998868727025618,
+      "learning_rate": 2e-05,
+      "loss": 1.5962,
+      "step": 17569
+    },
+    {
+      "epoch": 0.4182138436637151,
+      "grad_norm": 5.518906649129103,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 17570
+    },
+    {
+      "epoch": 0.41823764638674665,
+      "grad_norm": 6.211863409426341,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 17571
+    },
+    {
+      "epoch": 0.41826144910977814,
+      "grad_norm": 7.450410860252777,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 17572
+    },
+    {
+      "epoch": 0.4182852518328097,
+      "grad_norm": 2.166232626488009,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 17573
+    },
+    {
+      "epoch": 0.41830905455584116,
+      "grad_norm": 3.1913737840228213,
+      "learning_rate": 2e-05,
+      "loss": 0.1793,
+      "step": 17574
+    },
+    {
+      "epoch": 0.4183328572788727,
+      "grad_norm": 1.3792675655541335,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 17575
+    },
+    {
+      "epoch": 0.4183566600019042,
+      "grad_norm": 4.327561227137543,
+      "learning_rate": 2e-05,
+      "loss": 0.5398,
+      "step": 17576
+    },
+    {
+      "epoch": 0.41838046272493573,
+      "grad_norm": 0.4673571116866127,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 17577
+    },
+    {
+      "epoch": 0.41840426544796727,
+      "grad_norm": 5.876917541006545,
+      "learning_rate": 2e-05,
+      "loss": 0.5875,
+      "step": 17578
+    },
+    {
+      "epoch": 0.41842806817099876,
+      "grad_norm": 3.580977936909282,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 17579
+    },
+    {
+      "epoch": 0.4184518708940303,
+      "grad_norm": 6.9903106777470025,
+      "learning_rate": 2e-05,
+      "loss": 1.436,
+      "step": 17580
+    },
+    {
+      "epoch": 0.4184756736170618,
+      "grad_norm": 2.745021154839792,
+      "learning_rate": 2e-05,
+      "loss": 0.1444,
+      "step": 17581
+    },
+    {
+      "epoch": 0.4184994763400933,
+      "grad_norm": 0.44256830236070666,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 17582
+    },
+    {
+      "epoch": 0.4185232790631248,
+      "grad_norm": 7.871619530682104,
+      "learning_rate": 2e-05,
+      "loss": 1.2617,
+      "step": 17583
+    },
+    {
+      "epoch": 0.41854708178615635,
+      "grad_norm": 4.349440283912939,
+      "learning_rate": 2e-05,
+      "loss": 0.6488,
+      "step": 17584
+    },
+    {
+      "epoch": 0.41857088450918783,
+      "grad_norm": 5.551406394832511,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 17585
+    },
+    {
+      "epoch": 0.4185946872322194,
+      "grad_norm": 0.021606197542559383,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 17586
+    },
+    {
+      "epoch": 0.41861848995525086,
+      "grad_norm": 0.7155405991014432,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 17587
+    },
+    {
+      "epoch": 0.4186422926782824,
+      "grad_norm": 15.845299087795782,
+      "learning_rate": 2e-05,
+      "loss": 0.6404,
+      "step": 17588
+    },
+    {
+      "epoch": 0.4186660954013139,
+      "grad_norm": 13.904705639334885,
+      "learning_rate": 2e-05,
+      "loss": 0.5291,
+      "step": 17589
+    },
+    {
+      "epoch": 0.4186898981243454,
+      "grad_norm": 7.01074778551274,
+      "learning_rate": 2e-05,
+      "loss": 1.1267,
+      "step": 17590
+    },
+    {
+      "epoch": 0.4187137008473769,
+      "grad_norm": 1.015125592627238,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 17591
+    },
+    {
+      "epoch": 0.41873750357040845,
+      "grad_norm": 1.06627512051382,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 17592
+    },
+    {
+      "epoch": 0.41876130629344,
+      "grad_norm": 0.5568101692086487,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 17593
+    },
+    {
+      "epoch": 0.4187851090164715,
+      "grad_norm": 0.3636787004292057,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 17594
+    },
+    {
+      "epoch": 0.418808911739503,
+      "grad_norm": 3.1267125778815097,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 17595
+    },
+    {
+      "epoch": 0.4188327144625345,
+      "grad_norm": 0.9259475268311118,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 17596
+    },
+    {
+      "epoch": 0.41885651718556605,
+      "grad_norm": 6.375293651393976,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 17597
+    },
+    {
+      "epoch": 0.41888031990859753,
+      "grad_norm": 10.855092756807295,
+      "learning_rate": 2e-05,
+      "loss": 1.0418,
+      "step": 17598
+    },
+    {
+      "epoch": 0.4189041226316291,
+      "grad_norm": 2.1961691662394696,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 17599
+    },
+    {
+      "epoch": 0.41892792535466056,
+      "grad_norm": 2.709973889150585,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 17600
+    },
+    {
+      "epoch": 0.4189517280776921,
+      "grad_norm": 4.410347849406027,
+      "learning_rate": 2e-05,
+      "loss": 0.486,
+      "step": 17601
+    },
+    {
+      "epoch": 0.4189755308007236,
+      "grad_norm": 0.13216212385285006,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 17602
+    },
+    {
+      "epoch": 0.4189993335237551,
+      "grad_norm": 2.56501999506038,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 17603
+    },
+    {
+      "epoch": 0.4190231362467866,
+      "grad_norm": 3.0521300757765926,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 17604
+    },
+    {
+      "epoch": 0.41904693896981815,
+      "grad_norm": 5.398010786013815,
+      "learning_rate": 2e-05,
+      "loss": 0.7398,
+      "step": 17605
+    },
+    {
+      "epoch": 0.41907074169284964,
+      "grad_norm": 9.669503646148751,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 17606
+    },
+    {
+      "epoch": 0.4190945444158812,
+      "grad_norm": 9.929189868627839,
+      "learning_rate": 2e-05,
+      "loss": 0.7676,
+      "step": 17607
+    },
+    {
+      "epoch": 0.4191183471389127,
+      "grad_norm": 0.48720441662044706,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 17608
+    },
+    {
+      "epoch": 0.4191421498619442,
+      "grad_norm": 4.893003884919836,
+      "learning_rate": 2e-05,
+      "loss": 0.6374,
+      "step": 17609
+    },
+    {
+      "epoch": 0.41916595258497574,
+      "grad_norm": 3.4771024189320685,
+      "learning_rate": 2e-05,
+      "loss": 0.4614,
+      "step": 17610
+    },
+    {
+      "epoch": 0.41918975530800723,
+      "grad_norm": 4.152434982113129,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 17611
+    },
+    {
+      "epoch": 0.41921355803103877,
+      "grad_norm": 0.766105515065205,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 17612
+    },
+    {
+      "epoch": 0.41923736075407025,
+      "grad_norm": 3.5295459524055035,
+      "learning_rate": 2e-05,
+      "loss": 0.6382,
+      "step": 17613
+    },
+    {
+      "epoch": 0.4192611634771018,
+      "grad_norm": 7.360517855060516,
+      "learning_rate": 2e-05,
+      "loss": 0.5775,
+      "step": 17614
+    },
+    {
+      "epoch": 0.4192849662001333,
+      "grad_norm": 6.164324054100779,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 17615
+    },
+    {
+      "epoch": 0.4193087689231648,
+      "grad_norm": 9.317637558385723,
+      "learning_rate": 2e-05,
+      "loss": 0.4427,
+      "step": 17616
+    },
+    {
+      "epoch": 0.4193325716461963,
+      "grad_norm": 16.007403776432554,
+      "learning_rate": 2e-05,
+      "loss": 0.9465,
+      "step": 17617
+    },
+    {
+      "epoch": 0.41935637436922785,
+      "grad_norm": 2.6672383652461265,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 17618
+    },
+    {
+      "epoch": 0.41938017709225933,
+      "grad_norm": 1.496764676293502,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 17619
+    },
+    {
+      "epoch": 0.4194039798152909,
+      "grad_norm": 3.404279224400011,
+      "learning_rate": 2e-05,
+      "loss": 0.3864,
+      "step": 17620
+    },
+    {
+      "epoch": 0.41942778253832236,
+      "grad_norm": 2.9372334401258904,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 17621
+    },
+    {
+      "epoch": 0.4194515852613539,
+      "grad_norm": 3.045765085421539,
+      "learning_rate": 2e-05,
+      "loss": 0.2748,
+      "step": 17622
+    },
+    {
+      "epoch": 0.41947538798438544,
+      "grad_norm": 2.6728746632792624,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 17623
+    },
+    {
+      "epoch": 0.4194991907074169,
+      "grad_norm": 4.057060682285637,
+      "learning_rate": 2e-05,
+      "loss": 1.3268,
+      "step": 17624
+    },
+    {
+      "epoch": 0.41952299343044847,
+      "grad_norm": 13.028658482463344,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 17625
+    },
+    {
+      "epoch": 0.41954679615347995,
+      "grad_norm": 3.362018275334008,
+      "learning_rate": 2e-05,
+      "loss": 0.114,
+      "step": 17626
+    },
+    {
+      "epoch": 0.4195705988765115,
+      "grad_norm": 7.038237165599584,
+      "learning_rate": 2e-05,
+      "loss": 0.295,
+      "step": 17627
+    },
+    {
+      "epoch": 0.419594401599543,
+      "grad_norm": 3.114936887404352,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 17628
+    },
+    {
+      "epoch": 0.4196182043225745,
+      "grad_norm": 2.260542148889654,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 17629
+    },
+    {
+      "epoch": 0.419642007045606,
+      "grad_norm": 4.266164524357076,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 17630
+    },
+    {
+      "epoch": 0.41966580976863754,
+      "grad_norm": 1.1006520261579844,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 17631
+    },
+    {
+      "epoch": 0.41968961249166903,
+      "grad_norm": 1.1921208306343272,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 17632
+    },
+    {
+      "epoch": 0.41971341521470057,
+      "grad_norm": 1.9447589629329487,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 17633
+    },
+    {
+      "epoch": 0.41973721793773205,
+      "grad_norm": 13.040365135594428,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 17634
+    },
+    {
+      "epoch": 0.4197610206607636,
+      "grad_norm": 5.038766610727035,
+      "learning_rate": 2e-05,
+      "loss": 0.7647,
+      "step": 17635
+    },
+    {
+      "epoch": 0.4197848233837951,
+      "grad_norm": 1.5572749780484947,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 17636
+    },
+    {
+      "epoch": 0.4198086261068266,
+      "grad_norm": 3.983097374399758,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 17637
+    },
+    {
+      "epoch": 0.41983242882985816,
+      "grad_norm": 0.3121304491503998,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 17638
+    },
+    {
+      "epoch": 0.41985623155288965,
+      "grad_norm": 1.1965096420424488,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 17639
+    },
+    {
+      "epoch": 0.4198800342759212,
+      "grad_norm": 1.7452504149231223,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 17640
+    },
+    {
+      "epoch": 0.4199038369989527,
+      "grad_norm": 2.1089768500334203,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 17641
+    },
+    {
+      "epoch": 0.4199276397219842,
+      "grad_norm": 6.288189262684257,
+      "learning_rate": 2e-05,
+      "loss": 0.9925,
+      "step": 17642
+    },
+    {
+      "epoch": 0.4199514424450157,
+      "grad_norm": 5.719799805282191,
+      "learning_rate": 2e-05,
+      "loss": 0.6091,
+      "step": 17643
+    },
+    {
+      "epoch": 0.41997524516804724,
+      "grad_norm": 5.760077005838978,
+      "learning_rate": 2e-05,
+      "loss": 0.7276,
+      "step": 17644
+    },
+    {
+      "epoch": 0.4199990478910787,
+      "grad_norm": 1.2711720930293753,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 17645
+    },
+    {
+      "epoch": 0.42002285061411027,
+      "grad_norm": 8.216136442178888,
+      "learning_rate": 2e-05,
+      "loss": 0.3058,
+      "step": 17646
+    },
+    {
+      "epoch": 0.42004665333714175,
+      "grad_norm": 0.4102557511289516,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 17647
+    },
+    {
+      "epoch": 0.4200704560601733,
+      "grad_norm": 3.842261658963772,
+      "learning_rate": 2e-05,
+      "loss": 0.5082,
+      "step": 17648
+    },
+    {
+      "epoch": 0.4200942587832048,
+      "grad_norm": 11.471413627874409,
+      "learning_rate": 2e-05,
+      "loss": 0.6312,
+      "step": 17649
+    },
+    {
+      "epoch": 0.4201180615062363,
+      "grad_norm": 6.720794591809097,
+      "learning_rate": 2e-05,
+      "loss": 1.4753,
+      "step": 17650
+    },
+    {
+      "epoch": 0.4201418642292678,
+      "grad_norm": 2.617208035826567,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 17651
+    },
+    {
+      "epoch": 0.42016566695229934,
+      "grad_norm": 5.3402307458444955,
+      "learning_rate": 2e-05,
+      "loss": 1.2606,
+      "step": 17652
+    },
+    {
+      "epoch": 0.4201894696753309,
+      "grad_norm": 3.2632064078869094,
+      "learning_rate": 2e-05,
+      "loss": 0.4944,
+      "step": 17653
+    },
+    {
+      "epoch": 0.42021327239836237,
+      "grad_norm": 0.124620228473258,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 17654
+    },
+    {
+      "epoch": 0.4202370751213939,
+      "grad_norm": 3.4314258273308877,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 17655
+    },
+    {
+      "epoch": 0.4202608778444254,
+      "grad_norm": 6.866895019696412,
+      "learning_rate": 2e-05,
+      "loss": 0.2805,
+      "step": 17656
+    },
+    {
+      "epoch": 0.42028468056745694,
+      "grad_norm": 1.998798609347067,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 17657
+    },
+    {
+      "epoch": 0.4203084832904884,
+      "grad_norm": 82.60200612836438,
+      "learning_rate": 2e-05,
+      "loss": 0.1781,
+      "step": 17658
+    },
+    {
+      "epoch": 0.42033228601351996,
+      "grad_norm": 2.1050715172518335,
+      "learning_rate": 2e-05,
+      "loss": 0.0921,
+      "step": 17659
+    },
+    {
+      "epoch": 0.42035608873655145,
+      "grad_norm": 2.4130448005167446,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 17660
+    },
+    {
+      "epoch": 0.420379891459583,
+      "grad_norm": 1.838966098351833,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 17661
+    },
+    {
+      "epoch": 0.4204036941826145,
+      "grad_norm": 4.153705779725445,
+      "learning_rate": 2e-05,
+      "loss": 0.3065,
+      "step": 17662
+    },
+    {
+      "epoch": 0.420427496905646,
+      "grad_norm": 9.698508162499436,
+      "learning_rate": 2e-05,
+      "loss": 1.634,
+      "step": 17663
+    },
+    {
+      "epoch": 0.4204512996286775,
+      "grad_norm": 3.5956266151880394,
+      "learning_rate": 2e-05,
+      "loss": 0.7846,
+      "step": 17664
+    },
+    {
+      "epoch": 0.42047510235170904,
+      "grad_norm": 0.4597103387002602,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 17665
+    },
+    {
+      "epoch": 0.4204989050747405,
+      "grad_norm": 2.925542975516851,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 17666
+    },
+    {
+      "epoch": 0.42052270779777207,
+      "grad_norm": 3.6289893705630734,
+      "learning_rate": 2e-05,
+      "loss": 0.313,
+      "step": 17667
+    },
+    {
+      "epoch": 0.4205465105208036,
+      "grad_norm": 8.72072466369568,
+      "learning_rate": 2e-05,
+      "loss": 0.7095,
+      "step": 17668
+    },
+    {
+      "epoch": 0.4205703132438351,
+      "grad_norm": 5.8541075273290115,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 17669
+    },
+    {
+      "epoch": 0.42059411596686663,
+      "grad_norm": 11.262752428778025,
+      "learning_rate": 2e-05,
+      "loss": 0.9543,
+      "step": 17670
+    },
+    {
+      "epoch": 0.4206179186898981,
+      "grad_norm": 3.0428385039434676,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 17671
+    },
+    {
+      "epoch": 0.42064172141292966,
+      "grad_norm": 10.877564224635012,
+      "learning_rate": 2e-05,
+      "loss": 1.5981,
+      "step": 17672
+    },
+    {
+      "epoch": 0.42066552413596114,
+      "grad_norm": 1.971343649644237,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 17673
+    },
+    {
+      "epoch": 0.4206893268589927,
+      "grad_norm": 2.4598227228419964,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 17674
+    },
+    {
+      "epoch": 0.42071312958202417,
+      "grad_norm": 1.654171386085454,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 17675
+    },
+    {
+      "epoch": 0.4207369323050557,
+      "grad_norm": 3.4223672189239465,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 17676
+    },
+    {
+      "epoch": 0.4207607350280872,
+      "grad_norm": 4.243482818849471,
+      "learning_rate": 2e-05,
+      "loss": 0.1653,
+      "step": 17677
+    },
+    {
+      "epoch": 0.42078453775111874,
+      "grad_norm": 5.18726932357617,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 17678
+    },
+    {
+      "epoch": 0.4208083404741502,
+      "grad_norm": 2.611800437496883,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 17679
+    },
+    {
+      "epoch": 0.42083214319718176,
+      "grad_norm": 10.445057093819912,
+      "learning_rate": 2e-05,
+      "loss": 0.8083,
+      "step": 17680
+    },
+    {
+      "epoch": 0.42085594592021325,
+      "grad_norm": 2.071273128887989,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 17681
+    },
+    {
+      "epoch": 0.4208797486432448,
+      "grad_norm": 8.043886032842908,
+      "learning_rate": 2e-05,
+      "loss": 0.8816,
+      "step": 17682
+    },
+    {
+      "epoch": 0.42090355136627633,
+      "grad_norm": 8.281477044212426,
+      "learning_rate": 2e-05,
+      "loss": 1.6322,
+      "step": 17683
+    },
+    {
+      "epoch": 0.4209273540893078,
+      "grad_norm": 0.6100416402629715,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 17684
+    },
+    {
+      "epoch": 0.42095115681233936,
+      "grad_norm": 2.412801424852331,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 17685
+    },
+    {
+      "epoch": 0.42097495953537084,
+      "grad_norm": 5.4218222947365815,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 17686
+    },
+    {
+      "epoch": 0.4209987622584024,
+      "grad_norm": 10.765916498917202,
+      "learning_rate": 2e-05,
+      "loss": 0.3173,
+      "step": 17687
+    },
+    {
+      "epoch": 0.42102256498143387,
+      "grad_norm": 10.743832796257436,
+      "learning_rate": 2e-05,
+      "loss": 1.1402,
+      "step": 17688
+    },
+    {
+      "epoch": 0.4210463677044654,
+      "grad_norm": 5.39501369979936,
+      "learning_rate": 2e-05,
+      "loss": 0.5831,
+      "step": 17689
+    },
+    {
+      "epoch": 0.4210701704274969,
+      "grad_norm": 1.8885197186888298,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 17690
+    },
+    {
+      "epoch": 0.42109397315052843,
+      "grad_norm": 3.8518600233761515,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 17691
+    },
+    {
+      "epoch": 0.4211177758735599,
+      "grad_norm": 4.25472616290717,
+      "learning_rate": 2e-05,
+      "loss": 0.6478,
+      "step": 17692
+    },
+    {
+      "epoch": 0.42114157859659146,
+      "grad_norm": 3.326648723328764,
+      "learning_rate": 2e-05,
+      "loss": 0.4654,
+      "step": 17693
+    },
+    {
+      "epoch": 0.42116538131962294,
+      "grad_norm": 4.240083078007101,
+      "learning_rate": 2e-05,
+      "loss": 1.1791,
+      "step": 17694
+    },
+    {
+      "epoch": 0.4211891840426545,
+      "grad_norm": 2.386146635641969,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 17695
+    },
+    {
+      "epoch": 0.42121298676568597,
+      "grad_norm": 6.282868263482238,
+      "learning_rate": 2e-05,
+      "loss": 0.4498,
+      "step": 17696
+    },
+    {
+      "epoch": 0.4212367894887175,
+      "grad_norm": 3.3040576609692267,
+      "learning_rate": 2e-05,
+      "loss": 0.1279,
+      "step": 17697
+    },
+    {
+      "epoch": 0.421260592211749,
+      "grad_norm": 6.745795839976254,
+      "learning_rate": 2e-05,
+      "loss": 0.7336,
+      "step": 17698
+    },
+    {
+      "epoch": 0.42128439493478054,
+      "grad_norm": 2.585314989117973,
+      "learning_rate": 2e-05,
+      "loss": 0.2833,
+      "step": 17699
+    },
+    {
+      "epoch": 0.4213081976578121,
+      "grad_norm": 7.108986813482382,
+      "learning_rate": 2e-05,
+      "loss": 0.2465,
+      "step": 17700
+    },
+    {
+      "epoch": 0.42133200038084356,
+      "grad_norm": 3.1158644592748836,
+      "learning_rate": 2e-05,
+      "loss": 0.1444,
+      "step": 17701
+    },
+    {
+      "epoch": 0.4213558031038751,
+      "grad_norm": 2.8976545763227266,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 17702
+    },
+    {
+      "epoch": 0.4213796058269066,
+      "grad_norm": 2.7463136922773392,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 17703
+    },
+    {
+      "epoch": 0.42140340854993813,
+      "grad_norm": 0.44617739053778865,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17704
+    },
+    {
+      "epoch": 0.4214272112729696,
+      "grad_norm": 4.610968669848232,
+      "learning_rate": 2e-05,
+      "loss": 0.9433,
+      "step": 17705
+    },
+    {
+      "epoch": 0.42145101399600116,
+      "grad_norm": 2.5094527791438948,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 17706
+    },
+    {
+      "epoch": 0.42147481671903264,
+      "grad_norm": 2.5426892871616684,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 17707
+    },
+    {
+      "epoch": 0.4214986194420642,
+      "grad_norm": 4.280993121376194,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 17708
+    },
+    {
+      "epoch": 0.42152242216509567,
+      "grad_norm": 1.345030475725821,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 17709
+    },
+    {
+      "epoch": 0.4215462248881272,
+      "grad_norm": 1.7947291154497553,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 17710
+    },
+    {
+      "epoch": 0.4215700276111587,
+      "grad_norm": 9.031913277264447,
+      "learning_rate": 2e-05,
+      "loss": 0.5233,
+      "step": 17711
+    },
+    {
+      "epoch": 0.42159383033419023,
+      "grad_norm": 1.0666673264209154,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 17712
+    },
+    {
+      "epoch": 0.4216176330572217,
+      "grad_norm": 3.8756276025365093,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 17713
+    },
+    {
+      "epoch": 0.42164143578025326,
+      "grad_norm": 2.325171445877207,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 17714
+    },
+    {
+      "epoch": 0.4216652385032848,
+      "grad_norm": 3.759182916375667,
+      "learning_rate": 2e-05,
+      "loss": 0.2127,
+      "step": 17715
+    },
+    {
+      "epoch": 0.4216890412263163,
+      "grad_norm": 4.085815613900849,
+      "learning_rate": 2e-05,
+      "loss": 0.7468,
+      "step": 17716
+    },
+    {
+      "epoch": 0.4217128439493478,
+      "grad_norm": 1.1448500782710234,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 17717
+    },
+    {
+      "epoch": 0.4217366466723793,
+      "grad_norm": 3.219826907674239,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 17718
+    },
+    {
+      "epoch": 0.42176044939541085,
+      "grad_norm": 0.44684357489653137,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17719
+    },
+    {
+      "epoch": 0.42178425211844234,
+      "grad_norm": 3.9079951002476667,
+      "learning_rate": 2e-05,
+      "loss": 0.5403,
+      "step": 17720
+    },
+    {
+      "epoch": 0.4218080548414739,
+      "grad_norm": 3.8254066955774557,
+      "learning_rate": 2e-05,
+      "loss": 0.6761,
+      "step": 17721
+    },
+    {
+      "epoch": 0.42183185756450536,
+      "grad_norm": 2.8507739538311396,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 17722
+    },
+    {
+      "epoch": 0.4218556602875369,
+      "grad_norm": 8.619421276278368,
+      "learning_rate": 2e-05,
+      "loss": 0.847,
+      "step": 17723
+    },
+    {
+      "epoch": 0.4218794630105684,
+      "grad_norm": 2.099567838068709,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 17724
+    },
+    {
+      "epoch": 0.42190326573359993,
+      "grad_norm": 4.159300489611622,
+      "learning_rate": 2e-05,
+      "loss": 0.6416,
+      "step": 17725
+    },
+    {
+      "epoch": 0.4219270684566314,
+      "grad_norm": 1.5862764624490007,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 17726
+    },
+    {
+      "epoch": 0.42195087117966296,
+      "grad_norm": 0.7668865652924536,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 17727
+    },
+    {
+      "epoch": 0.42197467390269444,
+      "grad_norm": 11.244222838807087,
+      "learning_rate": 2e-05,
+      "loss": 0.6491,
+      "step": 17728
+    },
+    {
+      "epoch": 0.421998476625726,
+      "grad_norm": 3.5487629534685623,
+      "learning_rate": 2e-05,
+      "loss": 0.5006,
+      "step": 17729
+    },
+    {
+      "epoch": 0.4220222793487575,
+      "grad_norm": 10.383948744694969,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 17730
+    },
+    {
+      "epoch": 0.422046082071789,
+      "grad_norm": 3.607524645124174,
+      "learning_rate": 2e-05,
+      "loss": 0.7164,
+      "step": 17731
+    },
+    {
+      "epoch": 0.42206988479482055,
+      "grad_norm": 8.877588274592664,
+      "learning_rate": 2e-05,
+      "loss": 0.4186,
+      "step": 17732
+    },
+    {
+      "epoch": 0.42209368751785203,
+      "grad_norm": 3.3326310800083245,
+      "learning_rate": 2e-05,
+      "loss": 0.2,
+      "step": 17733
+    },
+    {
+      "epoch": 0.4221174902408836,
+      "grad_norm": 16.69263837068435,
+      "learning_rate": 2e-05,
+      "loss": 0.3329,
+      "step": 17734
+    },
+    {
+      "epoch": 0.42214129296391506,
+      "grad_norm": 5.13244797611144,
+      "learning_rate": 2e-05,
+      "loss": 0.7686,
+      "step": 17735
+    },
+    {
+      "epoch": 0.4221650956869466,
+      "grad_norm": 5.690879460387938,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 17736
+    },
+    {
+      "epoch": 0.4221888984099781,
+      "grad_norm": 5.8080504854912185,
+      "learning_rate": 2e-05,
+      "loss": 0.3189,
+      "step": 17737
+    },
+    {
+      "epoch": 0.4222127011330096,
+      "grad_norm": 3.3666143698443176,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 17738
+    },
+    {
+      "epoch": 0.4222365038560411,
+      "grad_norm": 6.376164958726869,
+      "learning_rate": 2e-05,
+      "loss": 1.3922,
+      "step": 17739
+    },
+    {
+      "epoch": 0.42226030657907265,
+      "grad_norm": 4.5022659267509715,
+      "learning_rate": 2e-05,
+      "loss": 0.3839,
+      "step": 17740
+    },
+    {
+      "epoch": 0.42228410930210414,
+      "grad_norm": 6.497322975237149,
+      "learning_rate": 2e-05,
+      "loss": 0.3141,
+      "step": 17741
+    },
+    {
+      "epoch": 0.4223079120251357,
+      "grad_norm": 1.8817188375550598,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 17742
+    },
+    {
+      "epoch": 0.42233171474816716,
+      "grad_norm": 5.562234477071984,
+      "learning_rate": 2e-05,
+      "loss": 0.2986,
+      "step": 17743
+    },
+    {
+      "epoch": 0.4223555174711987,
+      "grad_norm": 4.455475940901512,
+      "learning_rate": 2e-05,
+      "loss": 0.6309,
+      "step": 17744
+    },
+    {
+      "epoch": 0.42237932019423025,
+      "grad_norm": 0.48837842443935137,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 17745
+    },
+    {
+      "epoch": 0.42240312291726173,
+      "grad_norm": 0.32343715860758104,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 17746
+    },
+    {
+      "epoch": 0.42242692564029327,
+      "grad_norm": 0.7132488157766612,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 17747
+    },
+    {
+      "epoch": 0.42245072836332476,
+      "grad_norm": 3.8147622804979737,
+      "learning_rate": 2e-05,
+      "loss": 0.4667,
+      "step": 17748
+    },
+    {
+      "epoch": 0.4224745310863563,
+      "grad_norm": 4.529559171457933,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 17749
+    },
+    {
+      "epoch": 0.4224983338093878,
+      "grad_norm": 4.531214462917449,
+      "learning_rate": 2e-05,
+      "loss": 0.1629,
+      "step": 17750
+    },
+    {
+      "epoch": 0.4225221365324193,
+      "grad_norm": 2.0714615164851677,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 17751
+    },
+    {
+      "epoch": 0.4225459392554508,
+      "grad_norm": 47.20009503044082,
+      "learning_rate": 2e-05,
+      "loss": 0.796,
+      "step": 17752
+    },
+    {
+      "epoch": 0.42256974197848235,
+      "grad_norm": 4.824432537673018,
+      "learning_rate": 2e-05,
+      "loss": 0.5643,
+      "step": 17753
+    },
+    {
+      "epoch": 0.42259354470151383,
+      "grad_norm": 1.7343298771260283,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 17754
+    },
+    {
+      "epoch": 0.4226173474245454,
+      "grad_norm": 5.299981743888199,
+      "learning_rate": 2e-05,
+      "loss": 1.0596,
+      "step": 17755
+    },
+    {
+      "epoch": 0.42264115014757686,
+      "grad_norm": 4.039330797159877,
+      "learning_rate": 2e-05,
+      "loss": 0.7441,
+      "step": 17756
+    },
+    {
+      "epoch": 0.4226649528706084,
+      "grad_norm": 7.203122242427904,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 17757
+    },
+    {
+      "epoch": 0.4226887555936399,
+      "grad_norm": 5.779505665448802,
+      "learning_rate": 2e-05,
+      "loss": 0.6371,
+      "step": 17758
+    },
+    {
+      "epoch": 0.4227125583166714,
+      "grad_norm": 7.074390181349285,
+      "learning_rate": 2e-05,
+      "loss": 0.2491,
+      "step": 17759
+    },
+    {
+      "epoch": 0.42273636103970297,
+      "grad_norm": 1.4082814788110767,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 17760
+    },
+    {
+      "epoch": 0.42276016376273445,
+      "grad_norm": 0.6084659308517179,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 17761
+    },
+    {
+      "epoch": 0.422783966485766,
+      "grad_norm": 8.569096239486921,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 17762
+    },
+    {
+      "epoch": 0.4228077692087975,
+      "grad_norm": 3.2158080122711943,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 17763
+    },
+    {
+      "epoch": 0.422831571931829,
+      "grad_norm": 3.8851052345562653,
+      "learning_rate": 2e-05,
+      "loss": 0.6673,
+      "step": 17764
+    },
+    {
+      "epoch": 0.4228553746548605,
+      "grad_norm": 1.982110847092418,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 17765
+    },
+    {
+      "epoch": 0.42287917737789205,
+      "grad_norm": 6.716953603608116,
+      "learning_rate": 2e-05,
+      "loss": 0.8595,
+      "step": 17766
+    },
+    {
+      "epoch": 0.42290298010092353,
+      "grad_norm": 3.7136237072925042,
+      "learning_rate": 2e-05,
+      "loss": 0.4708,
+      "step": 17767
+    },
+    {
+      "epoch": 0.42292678282395507,
+      "grad_norm": 9.80597514234642,
+      "learning_rate": 2e-05,
+      "loss": 0.6231,
+      "step": 17768
+    },
+    {
+      "epoch": 0.42295058554698656,
+      "grad_norm": 4.314861004298471,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 17769
+    },
+    {
+      "epoch": 0.4229743882700181,
+      "grad_norm": 13.428044849181582,
+      "learning_rate": 2e-05,
+      "loss": 0.2776,
+      "step": 17770
+    },
+    {
+      "epoch": 0.4229981909930496,
+      "grad_norm": 3.7390269603610427,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 17771
+    },
+    {
+      "epoch": 0.4230219937160811,
+      "grad_norm": 2.7288963563184443,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 17772
+    },
+    {
+      "epoch": 0.4230457964391126,
+      "grad_norm": 0.1908661778719015,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 17773
+    },
+    {
+      "epoch": 0.42306959916214415,
+      "grad_norm": 4.067055355098967,
+      "learning_rate": 2e-05,
+      "loss": 0.5691,
+      "step": 17774
+    },
+    {
+      "epoch": 0.4230934018851757,
+      "grad_norm": 0.2177474204843092,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 17775
+    },
+    {
+      "epoch": 0.4231172046082072,
+      "grad_norm": 8.1890541065637,
+      "learning_rate": 2e-05,
+      "loss": 1.1271,
+      "step": 17776
+    },
+    {
+      "epoch": 0.4231410073312387,
+      "grad_norm": 4.6249376287891915,
+      "learning_rate": 2e-05,
+      "loss": 0.6881,
+      "step": 17777
+    },
+    {
+      "epoch": 0.4231648100542702,
+      "grad_norm": 2.4012777122005318,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 17778
+    },
+    {
+      "epoch": 0.42318861277730174,
+      "grad_norm": 2.1028494521463785,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 17779
+    },
+    {
+      "epoch": 0.42321241550033323,
+      "grad_norm": 2.0299010586836403,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 17780
+    },
+    {
+      "epoch": 0.42323621822336477,
+      "grad_norm": 2.2843222863083947,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 17781
+    },
+    {
+      "epoch": 0.42326002094639625,
+      "grad_norm": 1.105695846633074,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 17782
+    },
+    {
+      "epoch": 0.4232838236694278,
+      "grad_norm": 9.139968833579248,
+      "learning_rate": 2e-05,
+      "loss": 1.0927,
+      "step": 17783
+    },
+    {
+      "epoch": 0.4233076263924593,
+      "grad_norm": 6.177557925618714,
+      "learning_rate": 2e-05,
+      "loss": 0.8913,
+      "step": 17784
+    },
+    {
+      "epoch": 0.4233314291154908,
+      "grad_norm": 2.4287450439398524,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 17785
+    },
+    {
+      "epoch": 0.4233552318385223,
+      "grad_norm": 5.0812665509816055,
+      "learning_rate": 2e-05,
+      "loss": 0.638,
+      "step": 17786
+    },
+    {
+      "epoch": 0.42337903456155385,
+      "grad_norm": 1.5390104869113543,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 17787
+    },
+    {
+      "epoch": 0.42340283728458533,
+      "grad_norm": 3.216437372668468,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 17788
+    },
+    {
+      "epoch": 0.4234266400076169,
+      "grad_norm": 6.105898902860745,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 17789
+    },
+    {
+      "epoch": 0.4234504427306484,
+      "grad_norm": 6.262893472564844,
+      "learning_rate": 2e-05,
+      "loss": 1.0401,
+      "step": 17790
+    },
+    {
+      "epoch": 0.4234742454536799,
+      "grad_norm": 1.4158873358613948,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 17791
+    },
+    {
+      "epoch": 0.42349804817671144,
+      "grad_norm": 1.9463623513789785,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 17792
+    },
+    {
+      "epoch": 0.4235218508997429,
+      "grad_norm": 3.2275027808224768,
+      "learning_rate": 2e-05,
+      "loss": 0.3414,
+      "step": 17793
+    },
+    {
+      "epoch": 0.42354565362277447,
+      "grad_norm": 0.6009206870007494,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 17794
+    },
+    {
+      "epoch": 0.42356945634580595,
+      "grad_norm": 0.1990787063114137,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 17795
+    },
+    {
+      "epoch": 0.4235932590688375,
+      "grad_norm": 4.852131938855514,
+      "learning_rate": 2e-05,
+      "loss": 0.6187,
+      "step": 17796
+    },
+    {
+      "epoch": 0.423617061791869,
+      "grad_norm": 1.9087822423903733,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 17797
+    },
+    {
+      "epoch": 0.4236408645149005,
+      "grad_norm": 2.8232233232775945,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 17798
+    },
+    {
+      "epoch": 0.423664667237932,
+      "grad_norm": 12.134870342423918,
+      "learning_rate": 2e-05,
+      "loss": 0.8687,
+      "step": 17799
+    },
+    {
+      "epoch": 0.42368846996096354,
+      "grad_norm": 10.523190320722824,
+      "learning_rate": 2e-05,
+      "loss": 0.3701,
+      "step": 17800
+    },
+    {
+      "epoch": 0.42371227268399503,
+      "grad_norm": 21.77242506091384,
+      "learning_rate": 2e-05,
+      "loss": 0.3837,
+      "step": 17801
+    },
+    {
+      "epoch": 0.42373607540702657,
+      "grad_norm": 0.9549728244824236,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 17802
+    },
+    {
+      "epoch": 0.42375987813005805,
+      "grad_norm": 9.943386729115991,
+      "learning_rate": 2e-05,
+      "loss": 1.2668,
+      "step": 17803
+    },
+    {
+      "epoch": 0.4237836808530896,
+      "grad_norm": 0.552391905454471,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 17804
+    },
+    {
+      "epoch": 0.42380748357612114,
+      "grad_norm": 2.0323745796659187,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 17805
+    },
+    {
+      "epoch": 0.4238312862991526,
+      "grad_norm": 10.175663626316187,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 17806
+    },
+    {
+      "epoch": 0.42385508902218416,
+      "grad_norm": 6.997987266137088,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 17807
+    },
+    {
+      "epoch": 0.42387889174521565,
+      "grad_norm": 5.076269976662268,
+      "learning_rate": 2e-05,
+      "loss": 0.8183,
+      "step": 17808
+    },
+    {
+      "epoch": 0.4239026944682472,
+      "grad_norm": 614.7206631897222,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 17809
+    },
+    {
+      "epoch": 0.4239264971912787,
+      "grad_norm": 0.6621319917484333,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 17810
+    },
+    {
+      "epoch": 0.4239502999143102,
+      "grad_norm": 4.841270763006371,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 17811
+    },
+    {
+      "epoch": 0.4239741026373417,
+      "grad_norm": 12.325545083761881,
+      "learning_rate": 2e-05,
+      "loss": 2.034,
+      "step": 17812
+    },
+    {
+      "epoch": 0.42399790536037324,
+      "grad_norm": 46.73079949379977,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 17813
+    },
+    {
+      "epoch": 0.4240217080834047,
+      "grad_norm": 4.702471775570901,
+      "learning_rate": 2e-05,
+      "loss": 0.4439,
+      "step": 17814
+    },
+    {
+      "epoch": 0.42404551080643627,
+      "grad_norm": 8.273545727644676,
+      "learning_rate": 2e-05,
+      "loss": 0.5141,
+      "step": 17815
+    },
+    {
+      "epoch": 0.42406931352946775,
+      "grad_norm": 6.083447411062426,
+      "learning_rate": 2e-05,
+      "loss": 0.5175,
+      "step": 17816
+    },
+    {
+      "epoch": 0.4240931162524993,
+      "grad_norm": 1.0727339968959395,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 17817
+    },
+    {
+      "epoch": 0.4241169189755308,
+      "grad_norm": 2.642008186907456,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 17818
+    },
+    {
+      "epoch": 0.4241407216985623,
+      "grad_norm": 2.2487547756864386,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 17819
+    },
+    {
+      "epoch": 0.4241645244215938,
+      "grad_norm": 4.063734703588445,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 17820
+    },
+    {
+      "epoch": 0.42418832714462534,
+      "grad_norm": 5.281183457058603,
+      "learning_rate": 2e-05,
+      "loss": 0.4162,
+      "step": 17821
+    },
+    {
+      "epoch": 0.4242121298676569,
+      "grad_norm": 2.25450134466494,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 17822
+    },
+    {
+      "epoch": 0.42423593259068837,
+      "grad_norm": 2.0930093132634457,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 17823
+    },
+    {
+      "epoch": 0.4242597353137199,
+      "grad_norm": 3.5805861870908036,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 17824
+    },
+    {
+      "epoch": 0.4242835380367514,
+      "grad_norm": 8.2219261238629,
+      "learning_rate": 2e-05,
+      "loss": 1.2593,
+      "step": 17825
+    },
+    {
+      "epoch": 0.42430734075978294,
+      "grad_norm": 7.135281604650449,
+      "learning_rate": 2e-05,
+      "loss": 1.4107,
+      "step": 17826
+    },
+    {
+      "epoch": 0.4243311434828144,
+      "grad_norm": 4.505272127056204,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 17827
+    },
+    {
+      "epoch": 0.42435494620584596,
+      "grad_norm": 1.8279515089715053,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 17828
+    },
+    {
+      "epoch": 0.42437874892887745,
+      "grad_norm": 0.9742782498516176,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 17829
+    },
+    {
+      "epoch": 0.424402551651909,
+      "grad_norm": 0.25874297400775326,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 17830
+    },
+    {
+      "epoch": 0.4244263543749405,
+      "grad_norm": 0.1194038738214598,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 17831
+    },
+    {
+      "epoch": 0.424450157097972,
+      "grad_norm": 5.274067690460689,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 17832
+    },
+    {
+      "epoch": 0.4244739598210035,
+      "grad_norm": 11.630057103103836,
+      "learning_rate": 2e-05,
+      "loss": 0.3338,
+      "step": 17833
+    },
+    {
+      "epoch": 0.42449776254403504,
+      "grad_norm": 2.4259953823547793,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 17834
+    },
+    {
+      "epoch": 0.4245215652670665,
+      "grad_norm": 9.74748617106468,
+      "learning_rate": 2e-05,
+      "loss": 1.7406,
+      "step": 17835
+    },
+    {
+      "epoch": 0.42454536799009807,
+      "grad_norm": 8.911202004667368,
+      "learning_rate": 2e-05,
+      "loss": 0.7878,
+      "step": 17836
+    },
+    {
+      "epoch": 0.4245691707131296,
+      "grad_norm": 0.02297221531921744,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 17837
+    },
+    {
+      "epoch": 0.4245929734361611,
+      "grad_norm": 4.058878867273417,
+      "learning_rate": 2e-05,
+      "loss": 0.7353,
+      "step": 17838
+    },
+    {
+      "epoch": 0.42461677615919263,
+      "grad_norm": 0.8888262893611972,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 17839
+    },
+    {
+      "epoch": 0.4246405788822241,
+      "grad_norm": 6.140877360703309,
+      "learning_rate": 2e-05,
+      "loss": 0.891,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42466438160525566,
+      "grad_norm": 3.3905713892362632,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 17841
+    },
+    {
+      "epoch": 0.42468818432828714,
+      "grad_norm": 5.466522412454431,
+      "learning_rate": 2e-05,
+      "loss": 0.8108,
+      "step": 17842
+    },
+    {
+      "epoch": 0.4247119870513187,
+      "grad_norm": 9.948415371169819,
+      "learning_rate": 2e-05,
+      "loss": 0.4626,
+      "step": 17843
+    },
+    {
+      "epoch": 0.42473578977435017,
+      "grad_norm": 13.139734398293236,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 17844
+    },
+    {
+      "epoch": 0.4247595924973817,
+      "grad_norm": 0.6674239346413541,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 17845
+    },
+    {
+      "epoch": 0.4247833952204132,
+      "grad_norm": 2.117289239782678,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17846
+    },
+    {
+      "epoch": 0.42480719794344474,
+      "grad_norm": 2.4967060234073415,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 17847
+    },
+    {
+      "epoch": 0.4248310006664762,
+      "grad_norm": 4.962061188645983,
+      "learning_rate": 2e-05,
+      "loss": 1.0493,
+      "step": 17848
+    },
+    {
+      "epoch": 0.42485480338950776,
+      "grad_norm": 10.58606530645759,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 17849
+    },
+    {
+      "epoch": 0.42487860611253925,
+      "grad_norm": 3.442001228487579,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 17850
+    },
+    {
+      "epoch": 0.4249024088355708,
+      "grad_norm": 0.5541504015224119,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 17851
+    },
+    {
+      "epoch": 0.42492621155860233,
+      "grad_norm": 2.6460010267373364,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 17852
+    },
+    {
+      "epoch": 0.4249500142816338,
+      "grad_norm": 3.38891814623728,
+      "learning_rate": 2e-05,
+      "loss": 0.3907,
+      "step": 17853
+    },
+    {
+      "epoch": 0.42497381700466536,
+      "grad_norm": 6.988376497523465,
+      "learning_rate": 2e-05,
+      "loss": 0.3765,
+      "step": 17854
+    },
+    {
+      "epoch": 0.42499761972769684,
+      "grad_norm": 11.325956039841923,
+      "learning_rate": 2e-05,
+      "loss": 0.8349,
+      "step": 17855
+    },
+    {
+      "epoch": 0.4250214224507284,
+      "grad_norm": 1.9469205324993453,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 17856
+    },
+    {
+      "epoch": 0.42504522517375987,
+      "grad_norm": 7.936291438853332,
+      "learning_rate": 2e-05,
+      "loss": 0.8933,
+      "step": 17857
+    },
+    {
+      "epoch": 0.4250690278967914,
+      "grad_norm": 6.415369071362298,
+      "learning_rate": 2e-05,
+      "loss": 0.2368,
+      "step": 17858
+    },
+    {
+      "epoch": 0.4250928306198229,
+      "grad_norm": 1.7123804330459569,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 17859
+    },
+    {
+      "epoch": 0.42511663334285443,
+      "grad_norm": 5.502792703859379,
+      "learning_rate": 2e-05,
+      "loss": 0.7948,
+      "step": 17860
+    },
+    {
+      "epoch": 0.4251404360658859,
+      "grad_norm": 1.4111228477055346,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 17861
+    },
+    {
+      "epoch": 0.42516423878891746,
+      "grad_norm": 6.902181369503908,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 17862
+    },
+    {
+      "epoch": 0.42518804151194894,
+      "grad_norm": 4.677633195271211,
+      "learning_rate": 2e-05,
+      "loss": 0.4366,
+      "step": 17863
+    },
+    {
+      "epoch": 0.4252118442349805,
+      "grad_norm": 27.66311525257976,
+      "learning_rate": 2e-05,
+      "loss": 0.7192,
+      "step": 17864
+    },
+    {
+      "epoch": 0.42523564695801197,
+      "grad_norm": 0.5612377406431255,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 17865
+    },
+    {
+      "epoch": 0.4252594496810435,
+      "grad_norm": 7.377068501290452,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 17866
+    },
+    {
+      "epoch": 0.42528325240407505,
+      "grad_norm": 6.733257372933816,
+      "learning_rate": 2e-05,
+      "loss": 1.2659,
+      "step": 17867
+    },
+    {
+      "epoch": 0.42530705512710654,
+      "grad_norm": 5.547533344388735,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 17868
+    },
+    {
+      "epoch": 0.4253308578501381,
+      "grad_norm": 0.8318909721335409,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 17869
+    },
+    {
+      "epoch": 0.42535466057316956,
+      "grad_norm": 3.155038109522949,
+      "learning_rate": 2e-05,
+      "loss": 0.5692,
+      "step": 17870
+    },
+    {
+      "epoch": 0.4253784632962011,
+      "grad_norm": 5.166664820002918,
+      "learning_rate": 2e-05,
+      "loss": 0.7645,
+      "step": 17871
+    },
+    {
+      "epoch": 0.4254022660192326,
+      "grad_norm": 3.0370382899594643,
+      "learning_rate": 2e-05,
+      "loss": 0.3309,
+      "step": 17872
+    },
+    {
+      "epoch": 0.42542606874226413,
+      "grad_norm": 4.811604679208787,
+      "learning_rate": 2e-05,
+      "loss": 0.467,
+      "step": 17873
+    },
+    {
+      "epoch": 0.4254498714652956,
+      "grad_norm": 4.434505641732706,
+      "learning_rate": 2e-05,
+      "loss": 0.1902,
+      "step": 17874
+    },
+    {
+      "epoch": 0.42547367418832716,
+      "grad_norm": 2.6488988149679833,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 17875
+    },
+    {
+      "epoch": 0.42549747691135864,
+      "grad_norm": 3.1443408600073606,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 17876
+    },
+    {
+      "epoch": 0.4255212796343902,
+      "grad_norm": 5.715874701404239,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 17877
+    },
+    {
+      "epoch": 0.42554508235742167,
+      "grad_norm": 0.08243639855942131,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 17878
+    },
+    {
+      "epoch": 0.4255688850804532,
+      "grad_norm": 4.0438319572750965,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 17879
+    },
+    {
+      "epoch": 0.4255926878034847,
+      "grad_norm": 4.646269948462077,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 17880
+    },
+    {
+      "epoch": 0.42561649052651623,
+      "grad_norm": 3.415371176217529,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 17881
+    },
+    {
+      "epoch": 0.4256402932495478,
+      "grad_norm": 7.376097465133677,
+      "learning_rate": 2e-05,
+      "loss": 1.5243,
+      "step": 17882
+    },
+    {
+      "epoch": 0.42566409597257926,
+      "grad_norm": 3.083843541160381,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 17883
+    },
+    {
+      "epoch": 0.4256878986956108,
+      "grad_norm": 7.293316246548977,
+      "learning_rate": 2e-05,
+      "loss": 1.9613,
+      "step": 17884
+    },
+    {
+      "epoch": 0.4257117014186423,
+      "grad_norm": 3.6598711595979863,
+      "learning_rate": 2e-05,
+      "loss": 0.2549,
+      "step": 17885
+    },
+    {
+      "epoch": 0.4257355041416738,
+      "grad_norm": 5.869212728279528,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 17886
+    },
+    {
+      "epoch": 0.4257593068647053,
+      "grad_norm": 5.697013660763826,
+      "learning_rate": 2e-05,
+      "loss": 0.8629,
+      "step": 17887
+    },
+    {
+      "epoch": 0.42578310958773685,
+      "grad_norm": 0.62105202376839,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 17888
+    },
+    {
+      "epoch": 0.42580691231076834,
+      "grad_norm": 9.368451333416223,
+      "learning_rate": 2e-05,
+      "loss": 0.6982,
+      "step": 17889
+    },
+    {
+      "epoch": 0.4258307150337999,
+      "grad_norm": 5.939396690290032,
+      "learning_rate": 2e-05,
+      "loss": 0.2099,
+      "step": 17890
+    },
+    {
+      "epoch": 0.42585451775683136,
+      "grad_norm": 7.484841042906344,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 17891
+    },
+    {
+      "epoch": 0.4258783204798629,
+      "grad_norm": 2.479486344476704,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 17892
+    },
+    {
+      "epoch": 0.4259021232028944,
+      "grad_norm": 6.062325936605602,
+      "learning_rate": 2e-05,
+      "loss": 0.9209,
+      "step": 17893
+    },
+    {
+      "epoch": 0.42592592592592593,
+      "grad_norm": 5.4433868434195265,
+      "learning_rate": 2e-05,
+      "loss": 0.9952,
+      "step": 17894
+    },
+    {
+      "epoch": 0.4259497286489574,
+      "grad_norm": 5.996737380004534,
+      "learning_rate": 2e-05,
+      "loss": 0.9519,
+      "step": 17895
+    },
+    {
+      "epoch": 0.42597353137198896,
+      "grad_norm": 0.6694825899500819,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 17896
+    },
+    {
+      "epoch": 0.4259973340950205,
+      "grad_norm": 4.61569635711123,
+      "learning_rate": 2e-05,
+      "loss": 0.86,
+      "step": 17897
+    },
+    {
+      "epoch": 0.426021136818052,
+      "grad_norm": 1.9376134339543407,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 17898
+    },
+    {
+      "epoch": 0.4260449395410835,
+      "grad_norm": 7.433351107437483,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 17899
+    },
+    {
+      "epoch": 0.426068742264115,
+      "grad_norm": 0.6520792179691945,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 17900
+    },
+    {
+      "epoch": 0.42609254498714655,
+      "grad_norm": 1.4241657154711078,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 17901
+    },
+    {
+      "epoch": 0.42611634771017803,
+      "grad_norm": 0.07053183274888765,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 17902
+    },
+    {
+      "epoch": 0.4261401504332096,
+      "grad_norm": 0.5807287315346258,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 17903
+    },
+    {
+      "epoch": 0.42616395315624106,
+      "grad_norm": 2.4691376795801885,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 17904
+    },
+    {
+      "epoch": 0.4261877558792726,
+      "grad_norm": 2.021553488169564,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 17905
+    },
+    {
+      "epoch": 0.4262115586023041,
+      "grad_norm": 2.565832957158064,
+      "learning_rate": 2e-05,
+      "loss": 0.5549,
+      "step": 17906
+    },
+    {
+      "epoch": 0.4262353613253356,
+      "grad_norm": 2.610157805193216,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 17907
+    },
+    {
+      "epoch": 0.4262591640483671,
+      "grad_norm": 3.1820541846113537,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 17908
+    },
+    {
+      "epoch": 0.42628296677139865,
+      "grad_norm": 4.645926457161149,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 17909
+    },
+    {
+      "epoch": 0.42630676949443014,
+      "grad_norm": 10.36571579239299,
+      "learning_rate": 2e-05,
+      "loss": 0.651,
+      "step": 17910
+    },
+    {
+      "epoch": 0.4263305722174617,
+      "grad_norm": 0.3039934372943863,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 17911
+    },
+    {
+      "epoch": 0.4263543749404932,
+      "grad_norm": 3.7163975292003495,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 17912
+    },
+    {
+      "epoch": 0.4263781776635247,
+      "grad_norm": 0.4647680610909147,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 17913
+    },
+    {
+      "epoch": 0.42640198038655625,
+      "grad_norm": 14.876949492340282,
+      "learning_rate": 2e-05,
+      "loss": 0.5798,
+      "step": 17914
+    },
+    {
+      "epoch": 0.42642578310958773,
+      "grad_norm": 10.39880694348804,
+      "learning_rate": 2e-05,
+      "loss": 0.4109,
+      "step": 17915
+    },
+    {
+      "epoch": 0.42644958583261927,
+      "grad_norm": 4.7059920176912655,
+      "learning_rate": 2e-05,
+      "loss": 0.2599,
+      "step": 17916
+    },
+    {
+      "epoch": 0.42647338855565076,
+      "grad_norm": 3.381005077997982,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 17917
+    },
+    {
+      "epoch": 0.4264971912786823,
+      "grad_norm": 6.719310159760833,
+      "learning_rate": 2e-05,
+      "loss": 1.1231,
+      "step": 17918
+    },
+    {
+      "epoch": 0.4265209940017138,
+      "grad_norm": 5.006490070751696,
+      "learning_rate": 2e-05,
+      "loss": 0.7763,
+      "step": 17919
+    },
+    {
+      "epoch": 0.4265447967247453,
+      "grad_norm": 2.5416456240069363,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 17920
+    },
+    {
+      "epoch": 0.4265685994477768,
+      "grad_norm": 4.918234555567204,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 17921
+    },
+    {
+      "epoch": 0.42659240217080835,
+      "grad_norm": 7.5875924441830955,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 17922
+    },
+    {
+      "epoch": 0.42661620489383983,
+      "grad_norm": 3.8525686820723903,
+      "learning_rate": 2e-05,
+      "loss": 0.6756,
+      "step": 17923
+    },
+    {
+      "epoch": 0.4266400076168714,
+      "grad_norm": 1.9993321514529574,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 17924
+    },
+    {
+      "epoch": 0.42666381033990286,
+      "grad_norm": 1.4367393962809607,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 17925
+    },
+    {
+      "epoch": 0.4266876130629344,
+      "grad_norm": 8.88947279428747,
+      "learning_rate": 2e-05,
+      "loss": 0.9115,
+      "step": 17926
+    },
+    {
+      "epoch": 0.42671141578596594,
+      "grad_norm": 2.195732901630193,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 17927
+    },
+    {
+      "epoch": 0.4267352185089974,
+      "grad_norm": 3.037997061237699,
+      "learning_rate": 2e-05,
+      "loss": 0.7566,
+      "step": 17928
+    },
+    {
+      "epoch": 0.42675902123202897,
+      "grad_norm": 4.862321146878975,
+      "learning_rate": 2e-05,
+      "loss": 0.5607,
+      "step": 17929
+    },
+    {
+      "epoch": 0.42678282395506045,
+      "grad_norm": 6.023766820095561,
+      "learning_rate": 2e-05,
+      "loss": 0.5962,
+      "step": 17930
+    },
+    {
+      "epoch": 0.426806626678092,
+      "grad_norm": 1.3435614018279012,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 17931
+    },
+    {
+      "epoch": 0.4268304294011235,
+      "grad_norm": 295.3540923959477,
+      "learning_rate": 2e-05,
+      "loss": 2.0799,
+      "step": 17932
+    },
+    {
+      "epoch": 0.426854232124155,
+      "grad_norm": 14.11762753688646,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 17933
+    },
+    {
+      "epoch": 0.4268780348471865,
+      "grad_norm": 0.24735304880917333,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 17934
+    },
+    {
+      "epoch": 0.42690183757021805,
+      "grad_norm": 5.549472551262839,
+      "learning_rate": 2e-05,
+      "loss": 1.0713,
+      "step": 17935
+    },
+    {
+      "epoch": 0.42692564029324953,
+      "grad_norm": 5.676540213934568,
+      "learning_rate": 2e-05,
+      "loss": 0.9134,
+      "step": 17936
+    },
+    {
+      "epoch": 0.42694944301628107,
+      "grad_norm": 6.456578996263449,
+      "learning_rate": 2e-05,
+      "loss": 1.8744,
+      "step": 17937
+    },
+    {
+      "epoch": 0.42697324573931256,
+      "grad_norm": 2.593553897611124,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 17938
+    },
+    {
+      "epoch": 0.4269970484623441,
+      "grad_norm": 3.9785953654855803,
+      "learning_rate": 2e-05,
+      "loss": 0.6936,
+      "step": 17939
+    },
+    {
+      "epoch": 0.4270208511853756,
+      "grad_norm": 11.994478292706289,
+      "learning_rate": 2e-05,
+      "loss": 0.6003,
+      "step": 17940
+    },
+    {
+      "epoch": 0.4270446539084071,
+      "grad_norm": 4.886362294471643,
+      "learning_rate": 2e-05,
+      "loss": 0.422,
+      "step": 17941
+    },
+    {
+      "epoch": 0.42706845663143866,
+      "grad_norm": 1.7990713851731923,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 17942
+    },
+    {
+      "epoch": 0.42709225935447015,
+      "grad_norm": 5.68121297196656,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 17943
+    },
+    {
+      "epoch": 0.4271160620775017,
+      "grad_norm": 0.23430148133955317,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 17944
+    },
+    {
+      "epoch": 0.4271398648005332,
+      "grad_norm": 4.247028994466424,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 17945
+    },
+    {
+      "epoch": 0.4271636675235647,
+      "grad_norm": 0.15978685390784766,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 17946
+    },
+    {
+      "epoch": 0.4271874702465962,
+      "grad_norm": 5.328249974636345,
+      "learning_rate": 2e-05,
+      "loss": 0.3389,
+      "step": 17947
+    },
+    {
+      "epoch": 0.42721127296962774,
+      "grad_norm": 5.933953801894099,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 17948
+    },
+    {
+      "epoch": 0.4272350756926592,
+      "grad_norm": 4.0304311744758134,
+      "learning_rate": 2e-05,
+      "loss": 0.1475,
+      "step": 17949
+    },
+    {
+      "epoch": 0.42725887841569077,
+      "grad_norm": 2.774277510370324,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 17950
+    },
+    {
+      "epoch": 0.42728268113872225,
+      "grad_norm": 7.4951977953467095,
+      "learning_rate": 2e-05,
+      "loss": 0.2831,
+      "step": 17951
+    },
+    {
+      "epoch": 0.4273064838617538,
+      "grad_norm": 4.849440439319264,
+      "learning_rate": 2e-05,
+      "loss": 0.2427,
+      "step": 17952
+    },
+    {
+      "epoch": 0.4273302865847853,
+      "grad_norm": 4.394797564670529,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 17953
+    },
+    {
+      "epoch": 0.4273540893078168,
+      "grad_norm": 5.836863867971875,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 17954
+    },
+    {
+      "epoch": 0.4273778920308483,
+      "grad_norm": 0.7367589958177269,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 17955
+    },
+    {
+      "epoch": 0.42740169475387985,
+      "grad_norm": 3.1931524387842463,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 17956
+    },
+    {
+      "epoch": 0.42742549747691133,
+      "grad_norm": 0.5982369344675217,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 17957
+    },
+    {
+      "epoch": 0.42744930019994287,
+      "grad_norm": 5.356245308882861,
+      "learning_rate": 2e-05,
+      "loss": 0.9603,
+      "step": 17958
+    },
+    {
+      "epoch": 0.4274731029229744,
+      "grad_norm": 7.260411553141484,
+      "learning_rate": 2e-05,
+      "loss": 0.2622,
+      "step": 17959
+    },
+    {
+      "epoch": 0.4274969056460059,
+      "grad_norm": 0.8017849138548887,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 17960
+    },
+    {
+      "epoch": 0.42752070836903744,
+      "grad_norm": 3.9090203531329077,
+      "learning_rate": 2e-05,
+      "loss": 0.3978,
+      "step": 17961
+    },
+    {
+      "epoch": 0.4275445110920689,
+      "grad_norm": 5.932587759243868,
+      "learning_rate": 2e-05,
+      "loss": 0.2347,
+      "step": 17962
+    },
+    {
+      "epoch": 0.42756831381510046,
+      "grad_norm": 1.9388849720371817,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 17963
+    },
+    {
+      "epoch": 0.42759211653813195,
+      "grad_norm": 0.06234402438128184,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 17964
+    },
+    {
+      "epoch": 0.4276159192611635,
+      "grad_norm": 6.487976229859152,
+      "learning_rate": 2e-05,
+      "loss": 0.5017,
+      "step": 17965
+    },
+    {
+      "epoch": 0.427639721984195,
+      "grad_norm": 14.950570991279728,
+      "learning_rate": 2e-05,
+      "loss": 0.443,
+      "step": 17966
+    },
+    {
+      "epoch": 0.4276635247072265,
+      "grad_norm": 4.655913431554189,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 17967
+    },
+    {
+      "epoch": 0.427687327430258,
+      "grad_norm": 4.615446335220098,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 17968
+    },
+    {
+      "epoch": 0.42771113015328954,
+      "grad_norm": 5.437527275655717,
+      "learning_rate": 2e-05,
+      "loss": 1.246,
+      "step": 17969
+    },
+    {
+      "epoch": 0.42773493287632103,
+      "grad_norm": 2.4512883272071133,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 17970
+    },
+    {
+      "epoch": 0.42775873559935257,
+      "grad_norm": 3.3385700431566994,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 17971
+    },
+    {
+      "epoch": 0.42778253832238405,
+      "grad_norm": 0.30940896685061825,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 17972
+    },
+    {
+      "epoch": 0.4278063410454156,
+      "grad_norm": 11.79520942900275,
+      "learning_rate": 2e-05,
+      "loss": 0.4615,
+      "step": 17973
+    },
+    {
+      "epoch": 0.42783014376844714,
+      "grad_norm": 5.759455069247793,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 17974
+    },
+    {
+      "epoch": 0.4278539464914786,
+      "grad_norm": 1.0201964573286584,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 17975
+    },
+    {
+      "epoch": 0.42787774921451016,
+      "grad_norm": 66.91426873851799,
+      "learning_rate": 2e-05,
+      "loss": 0.8923,
+      "step": 17976
+    },
+    {
+      "epoch": 0.42790155193754165,
+      "grad_norm": 3.2346515438786008,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 17977
+    },
+    {
+      "epoch": 0.4279253546605732,
+      "grad_norm": 14.912791499737308,
+      "learning_rate": 2e-05,
+      "loss": 0.3457,
+      "step": 17978
+    },
+    {
+      "epoch": 0.4279491573836047,
+      "grad_norm": 6.9227533018824525,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 17979
+    },
+    {
+      "epoch": 0.4279729601066362,
+      "grad_norm": 2.0854564324492095,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 17980
+    },
+    {
+      "epoch": 0.4279967628296677,
+      "grad_norm": 5.2231566120147095,
+      "learning_rate": 2e-05,
+      "loss": 0.6761,
+      "step": 17981
+    },
+    {
+      "epoch": 0.42802056555269924,
+      "grad_norm": 3.1298572849331694,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 17982
+    },
+    {
+      "epoch": 0.4280443682757307,
+      "grad_norm": 32.908354746355904,
+      "learning_rate": 2e-05,
+      "loss": 0.9669,
+      "step": 17983
+    },
+    {
+      "epoch": 0.42806817099876227,
+      "grad_norm": 5.509205301457974,
+      "learning_rate": 2e-05,
+      "loss": 0.7538,
+      "step": 17984
+    },
+    {
+      "epoch": 0.42809197372179375,
+      "grad_norm": 3.286665315891711,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 17985
+    },
+    {
+      "epoch": 0.4281157764448253,
+      "grad_norm": 203.52447253469066,
+      "learning_rate": 2e-05,
+      "loss": 0.4794,
+      "step": 17986
+    },
+    {
+      "epoch": 0.4281395791678568,
+      "grad_norm": 9.273924140084134,
+      "learning_rate": 2e-05,
+      "loss": 1.1149,
+      "step": 17987
+    },
+    {
+      "epoch": 0.4281633818908883,
+      "grad_norm": 0.5945990021315058,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 17988
+    },
+    {
+      "epoch": 0.42818718461391986,
+      "grad_norm": 10.507041447684514,
+      "learning_rate": 2e-05,
+      "loss": 0.5959,
+      "step": 17989
+    },
+    {
+      "epoch": 0.42821098733695134,
+      "grad_norm": 4.198085808840832,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 17990
+    },
+    {
+      "epoch": 0.4282347900599829,
+      "grad_norm": 1.6483819163723488,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 17991
+    },
+    {
+      "epoch": 0.42825859278301437,
+      "grad_norm": 7.6343521227674325,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 17992
+    },
+    {
+      "epoch": 0.4282823955060459,
+      "grad_norm": 5.127611945950059,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 17993
+    },
+    {
+      "epoch": 0.4283061982290774,
+      "grad_norm": 3.14441906546512,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 17994
+    },
+    {
+      "epoch": 0.42833000095210894,
+      "grad_norm": 0.5942965781905102,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 17995
+    },
+    {
+      "epoch": 0.4283538036751404,
+      "grad_norm": 2.747529447082553,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 17996
+    },
+    {
+      "epoch": 0.42837760639817196,
+      "grad_norm": 4.4566114887863275,
+      "learning_rate": 2e-05,
+      "loss": 0.7315,
+      "step": 17997
+    },
+    {
+      "epoch": 0.42840140912120345,
+      "grad_norm": 0.5183516115941772,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 17998
+    },
+    {
+      "epoch": 0.428425211844235,
+      "grad_norm": 5.586567982086752,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 17999
+    },
+    {
+      "epoch": 0.4284490145672665,
+      "grad_norm": 3.8932933771480958,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 18000
+    },
+    {
+      "epoch": 0.428472817290298,
+      "grad_norm": 3.351867159148407,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 18001
+    },
+    {
+      "epoch": 0.4284966200133295,
+      "grad_norm": 2.840227657265566,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 18002
+    },
+    {
+      "epoch": 0.42852042273636104,
+      "grad_norm": 0.6662540046555776,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 18003
+    },
+    {
+      "epoch": 0.4285442254593926,
+      "grad_norm": 4.230565569521943,
+      "learning_rate": 2e-05,
+      "loss": 0.6351,
+      "step": 18004
+    },
+    {
+      "epoch": 0.42856802818242407,
+      "grad_norm": 5.432506543570088,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 18005
+    },
+    {
+      "epoch": 0.4285918309054556,
+      "grad_norm": 1.6061316038796178,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 18006
+    },
+    {
+      "epoch": 0.4286156336284871,
+      "grad_norm": 2.18042773423881,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 18007
+    },
+    {
+      "epoch": 0.42863943635151863,
+      "grad_norm": 8.291847158196925,
+      "learning_rate": 2e-05,
+      "loss": 0.5977,
+      "step": 18008
+    },
+    {
+      "epoch": 0.4286632390745501,
+      "grad_norm": 6.786839623392459,
+      "learning_rate": 2e-05,
+      "loss": 0.812,
+      "step": 18009
+    },
+    {
+      "epoch": 0.42868704179758166,
+      "grad_norm": 3.9609094466785737,
+      "learning_rate": 2e-05,
+      "loss": 0.6694,
+      "step": 18010
+    },
+    {
+      "epoch": 0.42871084452061314,
+      "grad_norm": 2.6772168018745406,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 18011
+    },
+    {
+      "epoch": 0.4287346472436447,
+      "grad_norm": 4.919090510078559,
+      "learning_rate": 2e-05,
+      "loss": 0.5216,
+      "step": 18012
+    },
+    {
+      "epoch": 0.42875844996667617,
+      "grad_norm": 2.0205691617150814,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 18013
+    },
+    {
+      "epoch": 0.4287822526897077,
+      "grad_norm": 9.04156192747983,
+      "learning_rate": 2e-05,
+      "loss": 1.0742,
+      "step": 18014
+    },
+    {
+      "epoch": 0.4288060554127392,
+      "grad_norm": 2.934752937100799,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 18015
+    },
+    {
+      "epoch": 0.42882985813577074,
+      "grad_norm": 1.6420055205595352,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 18016
+    },
+    {
+      "epoch": 0.4288536608588022,
+      "grad_norm": 1.326506639308504,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 18017
+    },
+    {
+      "epoch": 0.42887746358183376,
+      "grad_norm": 36.6054685464649,
+      "learning_rate": 2e-05,
+      "loss": 1.2649,
+      "step": 18018
+    },
+    {
+      "epoch": 0.4289012663048653,
+      "grad_norm": 1.607935905404975,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 18019
+    },
+    {
+      "epoch": 0.4289250690278968,
+      "grad_norm": 2.5236348739713055,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 18020
+    },
+    {
+      "epoch": 0.42894887175092833,
+      "grad_norm": 12.865898277696768,
+      "learning_rate": 2e-05,
+      "loss": 0.4353,
+      "step": 18021
+    },
+    {
+      "epoch": 0.4289726744739598,
+      "grad_norm": 2.297689549226309,
+      "learning_rate": 2e-05,
+      "loss": 0.0896,
+      "step": 18022
+    },
+    {
+      "epoch": 0.42899647719699135,
+      "grad_norm": 8.91926265221312,
+      "learning_rate": 2e-05,
+      "loss": 0.6014,
+      "step": 18023
+    },
+    {
+      "epoch": 0.42902027992002284,
+      "grad_norm": 5.2871259159478,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 18024
+    },
+    {
+      "epoch": 0.4290440826430544,
+      "grad_norm": 0.36313752863093307,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 18025
+    },
+    {
+      "epoch": 0.42906788536608587,
+      "grad_norm": 12.26010989554066,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 18026
+    },
+    {
+      "epoch": 0.4290916880891174,
+      "grad_norm": 4.00939052910572,
+      "learning_rate": 2e-05,
+      "loss": 0.5455,
+      "step": 18027
+    },
+    {
+      "epoch": 0.4291154908121489,
+      "grad_norm": 3.7291202323831336,
+      "learning_rate": 2e-05,
+      "loss": 0.7951,
+      "step": 18028
+    },
+    {
+      "epoch": 0.42913929353518043,
+      "grad_norm": 8.856462562267406,
+      "learning_rate": 2e-05,
+      "loss": 0.9949,
+      "step": 18029
+    },
+    {
+      "epoch": 0.4291630962582119,
+      "grad_norm": 0.46159100921446167,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18030
+    },
+    {
+      "epoch": 0.42918689898124346,
+      "grad_norm": 3.084289702000911,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 18031
+    },
+    {
+      "epoch": 0.42921070170427494,
+      "grad_norm": 0.9036929769784653,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 18032
+    },
+    {
+      "epoch": 0.4292345044273065,
+      "grad_norm": 5.319679424324305,
+      "learning_rate": 2e-05,
+      "loss": 0.3862,
+      "step": 18033
+    },
+    {
+      "epoch": 0.429258307150338,
+      "grad_norm": 3.410223678555086,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 18034
+    },
+    {
+      "epoch": 0.4292821098733695,
+      "grad_norm": 0.24797772588033276,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18035
+    },
+    {
+      "epoch": 0.42930591259640105,
+      "grad_norm": 3.626999939335139,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 18036
+    },
+    {
+      "epoch": 0.42932971531943254,
+      "grad_norm": 3.8473230588481884,
+      "learning_rate": 2e-05,
+      "loss": 0.6414,
+      "step": 18037
+    },
+    {
+      "epoch": 0.4293535180424641,
+      "grad_norm": 8.007169764323883,
+      "learning_rate": 2e-05,
+      "loss": 0.2978,
+      "step": 18038
+    },
+    {
+      "epoch": 0.42937732076549556,
+      "grad_norm": 2.013197255914335,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 18039
+    },
+    {
+      "epoch": 0.4294011234885271,
+      "grad_norm": 2.046058033433012,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 18040
+    },
+    {
+      "epoch": 0.4294249262115586,
+      "grad_norm": 3.2977759393137185,
+      "learning_rate": 2e-05,
+      "loss": 0.4518,
+      "step": 18041
+    },
+    {
+      "epoch": 0.42944872893459013,
+      "grad_norm": 3.9632241084013486,
+      "learning_rate": 2e-05,
+      "loss": 0.1891,
+      "step": 18042
+    },
+    {
+      "epoch": 0.4294725316576216,
+      "grad_norm": 5.406179797464494,
+      "learning_rate": 2e-05,
+      "loss": 0.8763,
+      "step": 18043
+    },
+    {
+      "epoch": 0.42949633438065316,
+      "grad_norm": 2.787457113456082,
+      "learning_rate": 2e-05,
+      "loss": 0.3436,
+      "step": 18044
+    },
+    {
+      "epoch": 0.42952013710368464,
+      "grad_norm": 8.08442912190112,
+      "learning_rate": 2e-05,
+      "loss": 0.2884,
+      "step": 18045
+    },
+    {
+      "epoch": 0.4295439398267162,
+      "grad_norm": 7.679320231323225,
+      "learning_rate": 2e-05,
+      "loss": 1.0345,
+      "step": 18046
+    },
+    {
+      "epoch": 0.42956774254974767,
+      "grad_norm": 7.855915862217509,
+      "learning_rate": 2e-05,
+      "loss": 1.6141,
+      "step": 18047
+    },
+    {
+      "epoch": 0.4295915452727792,
+      "grad_norm": 1.4001761434245414,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 18048
+    },
+    {
+      "epoch": 0.42961534799581075,
+      "grad_norm": 3.8053516686139295,
+      "learning_rate": 2e-05,
+      "loss": 0.7623,
+      "step": 18049
+    },
+    {
+      "epoch": 0.42963915071884223,
+      "grad_norm": 1.5917674503011392,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 18050
+    },
+    {
+      "epoch": 0.4296629534418738,
+      "grad_norm": 3.5366092361910035,
+      "learning_rate": 2e-05,
+      "loss": 0.4799,
+      "step": 18051
+    },
+    {
+      "epoch": 0.42968675616490526,
+      "grad_norm": 4.214214650095801,
+      "learning_rate": 2e-05,
+      "loss": 0.7853,
+      "step": 18052
+    },
+    {
+      "epoch": 0.4297105588879368,
+      "grad_norm": 0.34540319022177535,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18053
+    },
+    {
+      "epoch": 0.4297343616109683,
+      "grad_norm": 1.99990229834525,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 18054
+    },
+    {
+      "epoch": 0.4297581643339998,
+      "grad_norm": 4.115279397589966,
+      "learning_rate": 2e-05,
+      "loss": 0.7275,
+      "step": 18055
+    },
+    {
+      "epoch": 0.4297819670570313,
+      "grad_norm": 0.37799737389668625,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18056
+    },
+    {
+      "epoch": 0.42980576978006285,
+      "grad_norm": 3.099842125593005,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 18057
+    },
+    {
+      "epoch": 0.42982957250309434,
+      "grad_norm": 1.6559990013159192,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 18058
+    },
+    {
+      "epoch": 0.4298533752261259,
+      "grad_norm": 7.079302657632993,
+      "learning_rate": 2e-05,
+      "loss": 1.3213,
+      "step": 18059
+    },
+    {
+      "epoch": 0.42987717794915736,
+      "grad_norm": 5.827209280867901,
+      "learning_rate": 2e-05,
+      "loss": 0.6786,
+      "step": 18060
+    },
+    {
+      "epoch": 0.4299009806721889,
+      "grad_norm": 5.06227644714984,
+      "learning_rate": 2e-05,
+      "loss": 0.6321,
+      "step": 18061
+    },
+    {
+      "epoch": 0.4299247833952204,
+      "grad_norm": 2.425769832005787,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 18062
+    },
+    {
+      "epoch": 0.42994858611825193,
+      "grad_norm": 1.8153553192308431,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 18063
+    },
+    {
+      "epoch": 0.42997238884128347,
+      "grad_norm": 0.33934449050454296,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 18064
+    },
+    {
+      "epoch": 0.42999619156431496,
+      "grad_norm": 3.4119153227694135,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 18065
+    },
+    {
+      "epoch": 0.4300199942873465,
+      "grad_norm": 4.805489885890145,
+      "learning_rate": 2e-05,
+      "loss": 0.721,
+      "step": 18066
+    },
+    {
+      "epoch": 0.430043797010378,
+      "grad_norm": 6.994779335741262,
+      "learning_rate": 2e-05,
+      "loss": 0.6907,
+      "step": 18067
+    },
+    {
+      "epoch": 0.4300675997334095,
+      "grad_norm": 0.5192833956516485,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 18068
+    },
+    {
+      "epoch": 0.430091402456441,
+      "grad_norm": 2.541840937841585,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 18069
+    },
+    {
+      "epoch": 0.43011520517947255,
+      "grad_norm": 1.3136446064255825,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 18070
+    },
+    {
+      "epoch": 0.43013900790250403,
+      "grad_norm": 0.26373350274115437,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 18071
+    },
+    {
+      "epoch": 0.4301628106255356,
+      "grad_norm": 13.409765059536607,
+      "learning_rate": 2e-05,
+      "loss": 1.4964,
+      "step": 18072
+    },
+    {
+      "epoch": 0.43018661334856706,
+      "grad_norm": 6.771888130748283,
+      "learning_rate": 2e-05,
+      "loss": 1.4167,
+      "step": 18073
+    },
+    {
+      "epoch": 0.4302104160715986,
+      "grad_norm": 3.217665507459777,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 18074
+    },
+    {
+      "epoch": 0.4302342187946301,
+      "grad_norm": 7.8607884859167365,
+      "learning_rate": 2e-05,
+      "loss": 0.8509,
+      "step": 18075
+    },
+    {
+      "epoch": 0.4302580215176616,
+      "grad_norm": 4.20649383023466,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 18076
+    },
+    {
+      "epoch": 0.4302818242406931,
+      "grad_norm": 2.832519849557831,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 18077
+    },
+    {
+      "epoch": 0.43030562696372465,
+      "grad_norm": 4.762710328727274,
+      "learning_rate": 2e-05,
+      "loss": 0.7453,
+      "step": 18078
+    },
+    {
+      "epoch": 0.43032942968675614,
+      "grad_norm": 10.958331301387195,
+      "learning_rate": 2e-05,
+      "loss": 0.7571,
+      "step": 18079
+    },
+    {
+      "epoch": 0.4303532324097877,
+      "grad_norm": 1.1929659213887178,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 18080
+    },
+    {
+      "epoch": 0.4303770351328192,
+      "grad_norm": 6.907886272827697,
+      "learning_rate": 2e-05,
+      "loss": 0.7672,
+      "step": 18081
+    },
+    {
+      "epoch": 0.4304008378558507,
+      "grad_norm": 3.2329520067818747,
+      "learning_rate": 2e-05,
+      "loss": 0.4051,
+      "step": 18082
+    },
+    {
+      "epoch": 0.43042464057888224,
+      "grad_norm": 12.751503274859722,
+      "learning_rate": 2e-05,
+      "loss": 0.6553,
+      "step": 18083
+    },
+    {
+      "epoch": 0.43044844330191373,
+      "grad_norm": 3.0336667229635514,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 18084
+    },
+    {
+      "epoch": 0.43047224602494527,
+      "grad_norm": 5.21503536976664,
+      "learning_rate": 2e-05,
+      "loss": 0.2008,
+      "step": 18085
+    },
+    {
+      "epoch": 0.43049604874797676,
+      "grad_norm": 1.1575098661924925,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18086
+    },
+    {
+      "epoch": 0.4305198514710083,
+      "grad_norm": 2.017673382473895,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 18087
+    },
+    {
+      "epoch": 0.4305436541940398,
+      "grad_norm": 5.066462318096029,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 18088
+    },
+    {
+      "epoch": 0.4305674569170713,
+      "grad_norm": 3.409679438454218,
+      "learning_rate": 2e-05,
+      "loss": 0.5216,
+      "step": 18089
+    },
+    {
+      "epoch": 0.4305912596401028,
+      "grad_norm": 0.3939085805384336,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 18090
+    },
+    {
+      "epoch": 0.43061506236313435,
+      "grad_norm": 0.4598187339694022,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 18091
+    },
+    {
+      "epoch": 0.43063886508616583,
+      "grad_norm": 1.9862472442547028,
+      "learning_rate": 2e-05,
+      "loss": 0.2083,
+      "step": 18092
+    },
+    {
+      "epoch": 0.4306626678091974,
+      "grad_norm": 12.251612979022774,
+      "learning_rate": 2e-05,
+      "loss": 0.3163,
+      "step": 18093
+    },
+    {
+      "epoch": 0.43068647053222886,
+      "grad_norm": 10.472984189172205,
+      "learning_rate": 2e-05,
+      "loss": 0.7402,
+      "step": 18094
+    },
+    {
+      "epoch": 0.4307102732552604,
+      "grad_norm": 5.2582035258881765,
+      "learning_rate": 2e-05,
+      "loss": 0.7885,
+      "step": 18095
+    },
+    {
+      "epoch": 0.43073407597829194,
+      "grad_norm": 3.030361588370188,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 18096
+    },
+    {
+      "epoch": 0.4307578787013234,
+      "grad_norm": 2.572151555575267,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 18097
+    },
+    {
+      "epoch": 0.43078168142435497,
+      "grad_norm": 2.3052825506298924,
+      "learning_rate": 2e-05,
+      "loss": 0.169,
+      "step": 18098
+    },
+    {
+      "epoch": 0.43080548414738645,
+      "grad_norm": 5.804030922917041,
+      "learning_rate": 2e-05,
+      "loss": 0.9068,
+      "step": 18099
+    },
+    {
+      "epoch": 0.430829286870418,
+      "grad_norm": 3.6206371806878432,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 18100
+    },
+    {
+      "epoch": 0.4308530895934495,
+      "grad_norm": 24.433017681151366,
+      "learning_rate": 2e-05,
+      "loss": 0.3716,
+      "step": 18101
+    },
+    {
+      "epoch": 0.430876892316481,
+      "grad_norm": 1.9077423563633318,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 18102
+    },
+    {
+      "epoch": 0.4309006950395125,
+      "grad_norm": 1.927767647780785,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 18103
+    },
+    {
+      "epoch": 0.43092449776254405,
+      "grad_norm": 5.133066114638107,
+      "learning_rate": 2e-05,
+      "loss": 0.2849,
+      "step": 18104
+    },
+    {
+      "epoch": 0.43094830048557553,
+      "grad_norm": 2.66731614842488,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 18105
+    },
+    {
+      "epoch": 0.43097210320860707,
+      "grad_norm": 0.41918862659909556,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 18106
+    },
+    {
+      "epoch": 0.43099590593163856,
+      "grad_norm": 1.6611422905398336,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 18107
+    },
+    {
+      "epoch": 0.4310197086546701,
+      "grad_norm": 3.263264244931058,
+      "learning_rate": 2e-05,
+      "loss": 0.5097,
+      "step": 18108
+    },
+    {
+      "epoch": 0.4310435113777016,
+      "grad_norm": 1.127382269346701,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 18109
+    },
+    {
+      "epoch": 0.4310673141007331,
+      "grad_norm": 1.1369067508028683,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 18110
+    },
+    {
+      "epoch": 0.43109111682376466,
+      "grad_norm": 7.35614615313156,
+      "learning_rate": 2e-05,
+      "loss": 0.3153,
+      "step": 18111
+    },
+    {
+      "epoch": 0.43111491954679615,
+      "grad_norm": 5.260290517827853,
+      "learning_rate": 2e-05,
+      "loss": 0.5857,
+      "step": 18112
+    },
+    {
+      "epoch": 0.4311387222698277,
+      "grad_norm": 2.704424384816289,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 18113
+    },
+    {
+      "epoch": 0.4311625249928592,
+      "grad_norm": 5.09631805175107,
+      "learning_rate": 2e-05,
+      "loss": 0.5507,
+      "step": 18114
+    },
+    {
+      "epoch": 0.4311863277158907,
+      "grad_norm": 3.9823031250151857,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 18115
+    },
+    {
+      "epoch": 0.4312101304389222,
+      "grad_norm": 12.180962049655404,
+      "learning_rate": 2e-05,
+      "loss": 1.1299,
+      "step": 18116
+    },
+    {
+      "epoch": 0.43123393316195374,
+      "grad_norm": 8.173589481146475,
+      "learning_rate": 2e-05,
+      "loss": 1.1051,
+      "step": 18117
+    },
+    {
+      "epoch": 0.4312577358849852,
+      "grad_norm": 0.6598114747541418,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 18118
+    },
+    {
+      "epoch": 0.43128153860801677,
+      "grad_norm": 5.347004468526235,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 18119
+    },
+    {
+      "epoch": 0.43130534133104825,
+      "grad_norm": 7.931613067476962,
+      "learning_rate": 2e-05,
+      "loss": 1.6765,
+      "step": 18120
+    },
+    {
+      "epoch": 0.4313291440540798,
+      "grad_norm": 3.7411239494408033,
+      "learning_rate": 2e-05,
+      "loss": 0.2961,
+      "step": 18121
+    },
+    {
+      "epoch": 0.4313529467771113,
+      "grad_norm": 1.7998776458565144,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 18122
+    },
+    {
+      "epoch": 0.4313767495001428,
+      "grad_norm": 6.805638395699822,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 18123
+    },
+    {
+      "epoch": 0.4314005522231743,
+      "grad_norm": 3.7814368038126016,
+      "learning_rate": 2e-05,
+      "loss": 0.4106,
+      "step": 18124
+    },
+    {
+      "epoch": 0.43142435494620585,
+      "grad_norm": 0.9021583246522934,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 18125
+    },
+    {
+      "epoch": 0.4314481576692374,
+      "grad_norm": 7.521980956024072,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 18126
+    },
+    {
+      "epoch": 0.43147196039226887,
+      "grad_norm": 0.9512207908923259,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 18127
+    },
+    {
+      "epoch": 0.4314957631153004,
+      "grad_norm": 9.930429906680494,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 18128
+    },
+    {
+      "epoch": 0.4315195658383319,
+      "grad_norm": 8.068261022511876,
+      "learning_rate": 2e-05,
+      "loss": 0.3406,
+      "step": 18129
+    },
+    {
+      "epoch": 0.43154336856136344,
+      "grad_norm": 0.6607259700531717,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 18130
+    },
+    {
+      "epoch": 0.4315671712843949,
+      "grad_norm": 10.138242032171332,
+      "learning_rate": 2e-05,
+      "loss": 1.0258,
+      "step": 18131
+    },
+    {
+      "epoch": 0.43159097400742646,
+      "grad_norm": 4.485841164558153,
+      "learning_rate": 2e-05,
+      "loss": 0.7059,
+      "step": 18132
+    },
+    {
+      "epoch": 0.43161477673045795,
+      "grad_norm": 5.820922017539099,
+      "learning_rate": 2e-05,
+      "loss": 0.9775,
+      "step": 18133
+    },
+    {
+      "epoch": 0.4316385794534895,
+      "grad_norm": 0.22901397117281444,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18134
+    },
+    {
+      "epoch": 0.431662382176521,
+      "grad_norm": 12.46143130148283,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 18135
+    },
+    {
+      "epoch": 0.4316861848995525,
+      "grad_norm": 4.6189259716817554,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 18136
+    },
+    {
+      "epoch": 0.431709987622584,
+      "grad_norm": 6.918738633568024,
+      "learning_rate": 2e-05,
+      "loss": 1.2632,
+      "step": 18137
+    },
+    {
+      "epoch": 0.43173379034561554,
+      "grad_norm": 2.9503041154871807,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 18138
+    },
+    {
+      "epoch": 0.431757593068647,
+      "grad_norm": 3.4011292510320303,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 18139
+    },
+    {
+      "epoch": 0.43178139579167857,
+      "grad_norm": 3.433567685722294,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 18140
+    },
+    {
+      "epoch": 0.4318051985147101,
+      "grad_norm": 6.005633650060551,
+      "learning_rate": 2e-05,
+      "loss": 0.9076,
+      "step": 18141
+    },
+    {
+      "epoch": 0.4318290012377416,
+      "grad_norm": 7.58020943534718,
+      "learning_rate": 2e-05,
+      "loss": 1.6498,
+      "step": 18142
+    },
+    {
+      "epoch": 0.43185280396077314,
+      "grad_norm": 4.523546956277636,
+      "learning_rate": 2e-05,
+      "loss": 1.1105,
+      "step": 18143
+    },
+    {
+      "epoch": 0.4318766066838046,
+      "grad_norm": 10.35880287759633,
+      "learning_rate": 2e-05,
+      "loss": 1.56,
+      "step": 18144
+    },
+    {
+      "epoch": 0.43190040940683616,
+      "grad_norm": 3.81838186900132,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 18145
+    },
+    {
+      "epoch": 0.43192421212986765,
+      "grad_norm": 4.383038011082859,
+      "learning_rate": 2e-05,
+      "loss": 0.5068,
+      "step": 18146
+    },
+    {
+      "epoch": 0.4319480148528992,
+      "grad_norm": 3.070067828470613,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 18147
+    },
+    {
+      "epoch": 0.43197181757593067,
+      "grad_norm": 6.7675678406370325,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 18148
+    },
+    {
+      "epoch": 0.4319956202989622,
+      "grad_norm": 1.0954995152577334,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 18149
+    },
+    {
+      "epoch": 0.4320194230219937,
+      "grad_norm": 5.48734183797888,
+      "learning_rate": 2e-05,
+      "loss": 0.1704,
+      "step": 18150
+    },
+    {
+      "epoch": 0.43204322574502524,
+      "grad_norm": 1.8183356068716108,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 18151
+    },
+    {
+      "epoch": 0.4320670284680567,
+      "grad_norm": 4.115535807556814,
+      "learning_rate": 2e-05,
+      "loss": 0.704,
+      "step": 18152
+    },
+    {
+      "epoch": 0.43209083119108826,
+      "grad_norm": 2.872938732443064,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 18153
+    },
+    {
+      "epoch": 0.43211463391411975,
+      "grad_norm": 2.4341054312702526,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 18154
+    },
+    {
+      "epoch": 0.4321384366371513,
+      "grad_norm": 14.567360273518549,
+      "learning_rate": 2e-05,
+      "loss": 0.6943,
+      "step": 18155
+    },
+    {
+      "epoch": 0.43216223936018283,
+      "grad_norm": 4.495455715777243,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 18156
+    },
+    {
+      "epoch": 0.4321860420832143,
+      "grad_norm": 3.624358060051303,
+      "learning_rate": 2e-05,
+      "loss": 1.0037,
+      "step": 18157
+    },
+    {
+      "epoch": 0.43220984480624586,
+      "grad_norm": 5.228587734238186,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 18158
+    },
+    {
+      "epoch": 0.43223364752927734,
+      "grad_norm": 1.5272133907671501,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 18159
+    },
+    {
+      "epoch": 0.4322574502523089,
+      "grad_norm": 4.811867089582909,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 18160
+    },
+    {
+      "epoch": 0.43228125297534037,
+      "grad_norm": 1.4025710908183013,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 18161
+    },
+    {
+      "epoch": 0.4323050556983719,
+      "grad_norm": 1.471566996768824,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 18162
+    },
+    {
+      "epoch": 0.4323288584214034,
+      "grad_norm": 3.7817783255588684,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 18163
+    },
+    {
+      "epoch": 0.43235266114443494,
+      "grad_norm": 9.64663126463297,
+      "learning_rate": 2e-05,
+      "loss": 1.1609,
+      "step": 18164
+    },
+    {
+      "epoch": 0.4323764638674664,
+      "grad_norm": 2.000160256658775,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 18165
+    },
+    {
+      "epoch": 0.43240026659049796,
+      "grad_norm": 0.979741404586303,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 18166
+    },
+    {
+      "epoch": 0.43242406931352945,
+      "grad_norm": 2.9153872748642295,
+      "learning_rate": 2e-05,
+      "loss": 0.6937,
+      "step": 18167
+    },
+    {
+      "epoch": 0.432447872036561,
+      "grad_norm": 1.3431563053366118,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 18168
+    },
+    {
+      "epoch": 0.4324716747595925,
+      "grad_norm": 1.5589042941099545,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 18169
+    },
+    {
+      "epoch": 0.432495477482624,
+      "grad_norm": 4.834619581739847,
+      "learning_rate": 2e-05,
+      "loss": 0.1799,
+      "step": 18170
+    },
+    {
+      "epoch": 0.43251928020565555,
+      "grad_norm": 3.441926863668122,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 18171
+    },
+    {
+      "epoch": 0.43254308292868704,
+      "grad_norm": 8.980843689005749,
+      "learning_rate": 2e-05,
+      "loss": 0.9621,
+      "step": 18172
+    },
+    {
+      "epoch": 0.4325668856517186,
+      "grad_norm": 13.06620964522205,
+      "learning_rate": 2e-05,
+      "loss": 0.3648,
+      "step": 18173
+    },
+    {
+      "epoch": 0.43259068837475007,
+      "grad_norm": 7.8677729511005445,
+      "learning_rate": 2e-05,
+      "loss": 0.6757,
+      "step": 18174
+    },
+    {
+      "epoch": 0.4326144910977816,
+      "grad_norm": 2.7985362815322006,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 18175
+    },
+    {
+      "epoch": 0.4326382938208131,
+      "grad_norm": 6.088455104320861,
+      "learning_rate": 2e-05,
+      "loss": 1.2455,
+      "step": 18176
+    },
+    {
+      "epoch": 0.43266209654384463,
+      "grad_norm": 55.43908549156854,
+      "learning_rate": 2e-05,
+      "loss": 0.2968,
+      "step": 18177
+    },
+    {
+      "epoch": 0.4326858992668761,
+      "grad_norm": 3.703590065008684,
+      "learning_rate": 2e-05,
+      "loss": 0.4781,
+      "step": 18178
+    },
+    {
+      "epoch": 0.43270970198990766,
+      "grad_norm": 17.544305200680455,
+      "learning_rate": 2e-05,
+      "loss": 1.2992,
+      "step": 18179
+    },
+    {
+      "epoch": 0.43273350471293914,
+      "grad_norm": 0.6330851863788898,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18180
+    },
+    {
+      "epoch": 0.4327573074359707,
+      "grad_norm": 6.174583409161402,
+      "learning_rate": 2e-05,
+      "loss": 0.9769,
+      "step": 18181
+    },
+    {
+      "epoch": 0.43278111015900217,
+      "grad_norm": 5.662068411935811,
+      "learning_rate": 2e-05,
+      "loss": 0.8316,
+      "step": 18182
+    },
+    {
+      "epoch": 0.4328049128820337,
+      "grad_norm": 4.724789490823756,
+      "learning_rate": 2e-05,
+      "loss": 0.754,
+      "step": 18183
+    },
+    {
+      "epoch": 0.4328287156050652,
+      "grad_norm": 3.3789077084971395,
+      "learning_rate": 2e-05,
+      "loss": 0.4077,
+      "step": 18184
+    },
+    {
+      "epoch": 0.43285251832809674,
+      "grad_norm": 6.508528656485652,
+      "learning_rate": 2e-05,
+      "loss": 1.0582,
+      "step": 18185
+    },
+    {
+      "epoch": 0.4328763210511283,
+      "grad_norm": 4.601926254250001,
+      "learning_rate": 2e-05,
+      "loss": 1.1506,
+      "step": 18186
+    },
+    {
+      "epoch": 0.43290012377415976,
+      "grad_norm": 3.488984079279525,
+      "learning_rate": 2e-05,
+      "loss": 0.2645,
+      "step": 18187
+    },
+    {
+      "epoch": 0.4329239264971913,
+      "grad_norm": 5.258371495313624,
+      "learning_rate": 2e-05,
+      "loss": 1.4673,
+      "step": 18188
+    },
+    {
+      "epoch": 0.4329477292202228,
+      "grad_norm": 6.810216892978602,
+      "learning_rate": 2e-05,
+      "loss": 0.9377,
+      "step": 18189
+    },
+    {
+      "epoch": 0.43297153194325433,
+      "grad_norm": 1.4136534027804915,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 18190
+    },
+    {
+      "epoch": 0.4329953346662858,
+      "grad_norm": 3.2424020896160215,
+      "learning_rate": 2e-05,
+      "loss": 0.2313,
+      "step": 18191
+    },
+    {
+      "epoch": 0.43301913738931735,
+      "grad_norm": 1.8107808731474417,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 18192
+    },
+    {
+      "epoch": 0.43304294011234884,
+      "grad_norm": 5.2481553022302085,
+      "learning_rate": 2e-05,
+      "loss": 0.4535,
+      "step": 18193
+    },
+    {
+      "epoch": 0.4330667428353804,
+      "grad_norm": 2.443375995397135,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 18194
+    },
+    {
+      "epoch": 0.43309054555841187,
+      "grad_norm": 3.9071022173153063,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 18195
+    },
+    {
+      "epoch": 0.4331143482814434,
+      "grad_norm": 0.24560108164896188,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18196
+    },
+    {
+      "epoch": 0.4331381510044749,
+      "grad_norm": 0.5113059842111333,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 18197
+    },
+    {
+      "epoch": 0.43316195372750643,
+      "grad_norm": 2.690864167311588,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 18198
+    },
+    {
+      "epoch": 0.4331857564505379,
+      "grad_norm": 2.4572245319682215,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 18199
+    },
+    {
+      "epoch": 0.43320955917356946,
+      "grad_norm": 3.2451502353225403,
+      "learning_rate": 2e-05,
+      "loss": 0.1236,
+      "step": 18200
+    },
+    {
+      "epoch": 0.43323336189660094,
+      "grad_norm": 1.8216913625711688,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 18201
+    },
+    {
+      "epoch": 0.4332571646196325,
+      "grad_norm": 6.439727181641586,
+      "learning_rate": 2e-05,
+      "loss": 0.219,
+      "step": 18202
+    },
+    {
+      "epoch": 0.433280967342664,
+      "grad_norm": 5.584330650211809,
+      "learning_rate": 2e-05,
+      "loss": 1.073,
+      "step": 18203
+    },
+    {
+      "epoch": 0.4333047700656955,
+      "grad_norm": 2.667707280953647,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 18204
+    },
+    {
+      "epoch": 0.43332857278872705,
+      "grad_norm": 0.2011107647937116,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18205
+    },
+    {
+      "epoch": 0.43335237551175854,
+      "grad_norm": 0.6883325298229696,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 18206
+    },
+    {
+      "epoch": 0.4333761782347901,
+      "grad_norm": 1.536341084045316,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 18207
+    },
+    {
+      "epoch": 0.43339998095782156,
+      "grad_norm": 5.446700123633541,
+      "learning_rate": 2e-05,
+      "loss": 1.3868,
+      "step": 18208
+    },
+    {
+      "epoch": 0.4334237836808531,
+      "grad_norm": 1.920319552087349,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 18209
+    },
+    {
+      "epoch": 0.4334475864038846,
+      "grad_norm": 7.275796469112536,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 18210
+    },
+    {
+      "epoch": 0.43347138912691613,
+      "grad_norm": 2.593218495202396,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 18211
+    },
+    {
+      "epoch": 0.4334951918499476,
+      "grad_norm": 5.087959140750895,
+      "learning_rate": 2e-05,
+      "loss": 0.275,
+      "step": 18212
+    },
+    {
+      "epoch": 0.43351899457297915,
+      "grad_norm": 2.513700471858717,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 18213
+    },
+    {
+      "epoch": 0.43354279729601064,
+      "grad_norm": 2.7101483601998493,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 18214
+    },
+    {
+      "epoch": 0.4335666000190422,
+      "grad_norm": 5.170267328562314,
+      "learning_rate": 2e-05,
+      "loss": 1.1639,
+      "step": 18215
+    },
+    {
+      "epoch": 0.43359040274207367,
+      "grad_norm": 8.33192016947262,
+      "learning_rate": 2e-05,
+      "loss": 0.9454,
+      "step": 18216
+    },
+    {
+      "epoch": 0.4336142054651052,
+      "grad_norm": 4.132620977439921,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 18217
+    },
+    {
+      "epoch": 0.43363800818813675,
+      "grad_norm": 4.793550614693578,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 18218
+    },
+    {
+      "epoch": 0.43366181091116823,
+      "grad_norm": 3.3250625652359718,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 18219
+    },
+    {
+      "epoch": 0.4336856136341998,
+      "grad_norm": 4.012294627981964,
+      "learning_rate": 2e-05,
+      "loss": 0.1354,
+      "step": 18220
+    },
+    {
+      "epoch": 0.43370941635723126,
+      "grad_norm": 1.24171533254215,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 18221
+    },
+    {
+      "epoch": 0.4337332190802628,
+      "grad_norm": 235.27978221359317,
+      "learning_rate": 2e-05,
+      "loss": 0.2467,
+      "step": 18222
+    },
+    {
+      "epoch": 0.4337570218032943,
+      "grad_norm": 3.596375280557893,
+      "learning_rate": 2e-05,
+      "loss": 0.4214,
+      "step": 18223
+    },
+    {
+      "epoch": 0.4337808245263258,
+      "grad_norm": 4.33462386808924,
+      "learning_rate": 2e-05,
+      "loss": 0.5279,
+      "step": 18224
+    },
+    {
+      "epoch": 0.4338046272493573,
+      "grad_norm": 1.079206090222716,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 18225
+    },
+    {
+      "epoch": 0.43382842997238885,
+      "grad_norm": 7.886591558472999,
+      "learning_rate": 2e-05,
+      "loss": 1.3552,
+      "step": 18226
+    },
+    {
+      "epoch": 0.43385223269542034,
+      "grad_norm": 207.51988707347087,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 18227
+    },
+    {
+      "epoch": 0.4338760354184519,
+      "grad_norm": 5.302981074616207,
+      "learning_rate": 2e-05,
+      "loss": 1.1782,
+      "step": 18228
+    },
+    {
+      "epoch": 0.43389983814148336,
+      "grad_norm": 4.405163231305786,
+      "learning_rate": 2e-05,
+      "loss": 1.1361,
+      "step": 18229
+    },
+    {
+      "epoch": 0.4339236408645149,
+      "grad_norm": 3.765384611896474,
+      "learning_rate": 2e-05,
+      "loss": 0.4251,
+      "step": 18230
+    },
+    {
+      "epoch": 0.4339474435875464,
+      "grad_norm": 6.9697898314399485,
+      "learning_rate": 2e-05,
+      "loss": 0.591,
+      "step": 18231
+    },
+    {
+      "epoch": 0.43397124631057793,
+      "grad_norm": 0.9375955878707702,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 18232
+    },
+    {
+      "epoch": 0.43399504903360947,
+      "grad_norm": 1.3607984412722753,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 18233
+    },
+    {
+      "epoch": 0.43401885175664096,
+      "grad_norm": 1.3324911508475876,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 18234
+    },
+    {
+      "epoch": 0.4340426544796725,
+      "grad_norm": 2.6186040790185037,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 18235
+    },
+    {
+      "epoch": 0.434066457202704,
+      "grad_norm": 3.187814772532367,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 18236
+    },
+    {
+      "epoch": 0.4340902599257355,
+      "grad_norm": 0.950000387533292,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 18237
+    },
+    {
+      "epoch": 0.434114062648767,
+      "grad_norm": 4.053296757631057,
+      "learning_rate": 2e-05,
+      "loss": 0.8899,
+      "step": 18238
+    },
+    {
+      "epoch": 0.43413786537179855,
+      "grad_norm": 49.283999192021845,
+      "learning_rate": 2e-05,
+      "loss": 1.2543,
+      "step": 18239
+    },
+    {
+      "epoch": 0.43416166809483003,
+      "grad_norm": 7.0412390469983865,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 18240
+    },
+    {
+      "epoch": 0.4341854708178616,
+      "grad_norm": 7.110102407306772,
+      "learning_rate": 2e-05,
+      "loss": 0.706,
+      "step": 18241
+    },
+    {
+      "epoch": 0.43420927354089306,
+      "grad_norm": 175.98487565156591,
+      "learning_rate": 2e-05,
+      "loss": 0.7001,
+      "step": 18242
+    },
+    {
+      "epoch": 0.4342330762639246,
+      "grad_norm": 4.0377487737318605,
+      "learning_rate": 2e-05,
+      "loss": 0.2171,
+      "step": 18243
+    },
+    {
+      "epoch": 0.4342568789869561,
+      "grad_norm": 4.450594491340003,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 18244
+    },
+    {
+      "epoch": 0.4342806817099876,
+      "grad_norm": 46.45988282706895,
+      "learning_rate": 2e-05,
+      "loss": 0.999,
+      "step": 18245
+    },
+    {
+      "epoch": 0.4343044844330191,
+      "grad_norm": 2.052088293655204,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 18246
+    },
+    {
+      "epoch": 0.43432828715605065,
+      "grad_norm": 2.9019512524516964,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 18247
+    },
+    {
+      "epoch": 0.4343520898790822,
+      "grad_norm": 1.2682815844358502,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 18248
+    },
+    {
+      "epoch": 0.4343758926021137,
+      "grad_norm": 2.077002175844432,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 18249
+    },
+    {
+      "epoch": 0.4343996953251452,
+      "grad_norm": 3.691193247580361,
+      "learning_rate": 2e-05,
+      "loss": 0.2379,
+      "step": 18250
+    },
+    {
+      "epoch": 0.4344234980481767,
+      "grad_norm": 4.524826961855028,
+      "learning_rate": 2e-05,
+      "loss": 0.6291,
+      "step": 18251
+    },
+    {
+      "epoch": 0.43444730077120824,
+      "grad_norm": 1.398920804725769,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 18252
+    },
+    {
+      "epoch": 0.43447110349423973,
+      "grad_norm": 1.7368898062388836,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 18253
+    },
+    {
+      "epoch": 0.43449490621727127,
+      "grad_norm": 4.826622499299207,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 18254
+    },
+    {
+      "epoch": 0.43451870894030276,
+      "grad_norm": 14.666044025975003,
+      "learning_rate": 2e-05,
+      "loss": 0.7999,
+      "step": 18255
+    },
+    {
+      "epoch": 0.4345425116633343,
+      "grad_norm": 2.5022095680058287,
+      "learning_rate": 2e-05,
+      "loss": 0.3013,
+      "step": 18256
+    },
+    {
+      "epoch": 0.4345663143863658,
+      "grad_norm": 4.392748413784544,
+      "learning_rate": 2e-05,
+      "loss": 0.6829,
+      "step": 18257
+    },
+    {
+      "epoch": 0.4345901171093973,
+      "grad_norm": 13.125675932382146,
+      "learning_rate": 2e-05,
+      "loss": 0.4964,
+      "step": 18258
+    },
+    {
+      "epoch": 0.4346139198324288,
+      "grad_norm": 2.8656918132506144,
+      "learning_rate": 2e-05,
+      "loss": 0.3282,
+      "step": 18259
+    },
+    {
+      "epoch": 0.43463772255546035,
+      "grad_norm": 3.4173155562147457,
+      "learning_rate": 2e-05,
+      "loss": 0.437,
+      "step": 18260
+    },
+    {
+      "epoch": 0.43466152527849183,
+      "grad_norm": 3.392820682250069,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 18261
+    },
+    {
+      "epoch": 0.4346853280015234,
+      "grad_norm": 5.261943422798432,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 18262
+    },
+    {
+      "epoch": 0.4347091307245549,
+      "grad_norm": 2.787054499925091,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 18263
+    },
+    {
+      "epoch": 0.4347329334475864,
+      "grad_norm": 3.332496763158046,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 18264
+    },
+    {
+      "epoch": 0.43475673617061794,
+      "grad_norm": 5.535575187790685,
+      "learning_rate": 2e-05,
+      "loss": 0.8179,
+      "step": 18265
+    },
+    {
+      "epoch": 0.4347805388936494,
+      "grad_norm": 3.4606367306780004,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 18266
+    },
+    {
+      "epoch": 0.43480434161668097,
+      "grad_norm": 5.223795676095175,
+      "learning_rate": 2e-05,
+      "loss": 0.8291,
+      "step": 18267
+    },
+    {
+      "epoch": 0.43482814433971245,
+      "grad_norm": 0.8549030830159017,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 18268
+    },
+    {
+      "epoch": 0.434851947062744,
+      "grad_norm": 3.834528977325841,
+      "learning_rate": 2e-05,
+      "loss": 0.4548,
+      "step": 18269
+    },
+    {
+      "epoch": 0.4348757497857755,
+      "grad_norm": 66.14749542717144,
+      "learning_rate": 2e-05,
+      "loss": 0.3927,
+      "step": 18270
+    },
+    {
+      "epoch": 0.434899552508807,
+      "grad_norm": 3.383252138233481,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 18271
+    },
+    {
+      "epoch": 0.4349233552318385,
+      "grad_norm": 0.6125236164534456,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 18272
+    },
+    {
+      "epoch": 0.43494715795487005,
+      "grad_norm": 58.42869175158255,
+      "learning_rate": 2e-05,
+      "loss": 0.7278,
+      "step": 18273
+    },
+    {
+      "epoch": 0.43497096067790153,
+      "grad_norm": 9.546037672034368,
+      "learning_rate": 2e-05,
+      "loss": 1.5735,
+      "step": 18274
+    },
+    {
+      "epoch": 0.43499476340093307,
+      "grad_norm": 4.649108232841876,
+      "learning_rate": 2e-05,
+      "loss": 0.773,
+      "step": 18275
+    },
+    {
+      "epoch": 0.43501856612396456,
+      "grad_norm": 2.7214023840301773,
+      "learning_rate": 2e-05,
+      "loss": 0.0872,
+      "step": 18276
+    },
+    {
+      "epoch": 0.4350423688469961,
+      "grad_norm": 6.385794344976207,
+      "learning_rate": 2e-05,
+      "loss": 1.1572,
+      "step": 18277
+    },
+    {
+      "epoch": 0.43506617157002764,
+      "grad_norm": 1.3489483348628828,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 18278
+    },
+    {
+      "epoch": 0.4350899742930591,
+      "grad_norm": 7.04733459947715,
+      "learning_rate": 2e-05,
+      "loss": 1.0536,
+      "step": 18279
+    },
+    {
+      "epoch": 0.43511377701609066,
+      "grad_norm": 3.7936278400539956,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 18280
+    },
+    {
+      "epoch": 0.43513757973912215,
+      "grad_norm": 2.7443156173358547,
+      "learning_rate": 2e-05,
+      "loss": 0.2311,
+      "step": 18281
+    },
+    {
+      "epoch": 0.4351613824621537,
+      "grad_norm": 2.389268287006448,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 18282
+    },
+    {
+      "epoch": 0.4351851851851852,
+      "grad_norm": 3.5863867559889817,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 18283
+    },
+    {
+      "epoch": 0.4352089879082167,
+      "grad_norm": 1.490404374021447,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 18284
+    },
+    {
+      "epoch": 0.4352327906312482,
+      "grad_norm": 10.560071795141615,
+      "learning_rate": 2e-05,
+      "loss": 1.097,
+      "step": 18285
+    },
+    {
+      "epoch": 0.43525659335427974,
+      "grad_norm": 5.590632517380709,
+      "learning_rate": 2e-05,
+      "loss": 1.0341,
+      "step": 18286
+    },
+    {
+      "epoch": 0.4352803960773112,
+      "grad_norm": 2.8722875412243707,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 18287
+    },
+    {
+      "epoch": 0.43530419880034277,
+      "grad_norm": 0.20941344849037755,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 18288
+    },
+    {
+      "epoch": 0.43532800152337425,
+      "grad_norm": 1.4357403467988181,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 18289
+    },
+    {
+      "epoch": 0.4353518042464058,
+      "grad_norm": 3.2948170189391552,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 18290
+    },
+    {
+      "epoch": 0.4353756069694373,
+      "grad_norm": 4.9012690330978055,
+      "learning_rate": 2e-05,
+      "loss": 0.2969,
+      "step": 18291
+    },
+    {
+      "epoch": 0.4353994096924688,
+      "grad_norm": 5.12378510045135,
+      "learning_rate": 2e-05,
+      "loss": 0.994,
+      "step": 18292
+    },
+    {
+      "epoch": 0.43542321241550036,
+      "grad_norm": 1.529839225093282,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 18293
+    },
+    {
+      "epoch": 0.43544701513853185,
+      "grad_norm": 3.964072742718942,
+      "learning_rate": 2e-05,
+      "loss": 0.3289,
+      "step": 18294
+    },
+    {
+      "epoch": 0.4354708178615634,
+      "grad_norm": 4.525453003251647,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 18295
+    },
+    {
+      "epoch": 0.43549462058459487,
+      "grad_norm": 1.3732159898611758,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 18296
+    },
+    {
+      "epoch": 0.4355184233076264,
+      "grad_norm": 8.917023465912331,
+      "learning_rate": 2e-05,
+      "loss": 0.986,
+      "step": 18297
+    },
+    {
+      "epoch": 0.4355422260306579,
+      "grad_norm": 6.561531494274473,
+      "learning_rate": 2e-05,
+      "loss": 1.7829,
+      "step": 18298
+    },
+    {
+      "epoch": 0.43556602875368944,
+      "grad_norm": 1.3021515595222688,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 18299
+    },
+    {
+      "epoch": 0.4355898314767209,
+      "grad_norm": 2.6557227715631537,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 18300
+    },
+    {
+      "epoch": 0.43561363419975246,
+      "grad_norm": 0.6955377471333114,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18301
+    },
+    {
+      "epoch": 0.43563743692278395,
+      "grad_norm": 2.5616363036449976,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 18302
+    },
+    {
+      "epoch": 0.4356612396458155,
+      "grad_norm": 1.7006135992472124,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 18303
+    },
+    {
+      "epoch": 0.435685042368847,
+      "grad_norm": 0.8644865437728902,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 18304
+    },
+    {
+      "epoch": 0.4357088450918785,
+      "grad_norm": 9.103792481523456,
+      "learning_rate": 2e-05,
+      "loss": 1.3979,
+      "step": 18305
+    },
+    {
+      "epoch": 0.43573264781491,
+      "grad_norm": 4.0873565839059225,
+      "learning_rate": 2e-05,
+      "loss": 0.7619,
+      "step": 18306
+    },
+    {
+      "epoch": 0.43575645053794154,
+      "grad_norm": 17.797819886494104,
+      "learning_rate": 2e-05,
+      "loss": 1.3686,
+      "step": 18307
+    },
+    {
+      "epoch": 0.4357802532609731,
+      "grad_norm": 3.097980406221678,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 18308
+    },
+    {
+      "epoch": 0.43580405598400457,
+      "grad_norm": 0.2239315520628155,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 18309
+    },
+    {
+      "epoch": 0.4358278587070361,
+      "grad_norm": 1.949923372456764,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 18310
+    },
+    {
+      "epoch": 0.4358516614300676,
+      "grad_norm": 0.4916096039138503,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18311
+    },
+    {
+      "epoch": 0.43587546415309913,
+      "grad_norm": 2.1435184537275136,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 18312
+    },
+    {
+      "epoch": 0.4358992668761306,
+      "grad_norm": 3.121269084935846,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 18313
+    },
+    {
+      "epoch": 0.43592306959916216,
+      "grad_norm": 0.5348550041652035,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 18314
+    },
+    {
+      "epoch": 0.43594687232219365,
+      "grad_norm": 0.9033985438836806,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 18315
+    },
+    {
+      "epoch": 0.4359706750452252,
+      "grad_norm": 3.7373299995835616,
+      "learning_rate": 2e-05,
+      "loss": 0.1551,
+      "step": 18316
+    },
+    {
+      "epoch": 0.43599447776825667,
+      "grad_norm": 4.957773028975513,
+      "learning_rate": 2e-05,
+      "loss": 0.1825,
+      "step": 18317
+    },
+    {
+      "epoch": 0.4360182804912882,
+      "grad_norm": 4.645633524689327,
+      "learning_rate": 2e-05,
+      "loss": 0.7824,
+      "step": 18318
+    },
+    {
+      "epoch": 0.4360420832143197,
+      "grad_norm": 3.425960013487225,
+      "learning_rate": 2e-05,
+      "loss": 0.1307,
+      "step": 18319
+    },
+    {
+      "epoch": 0.43606588593735124,
+      "grad_norm": 0.35655198072065614,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 18320
+    },
+    {
+      "epoch": 0.4360896886603827,
+      "grad_norm": 3.444926173045131,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 18321
+    },
+    {
+      "epoch": 0.43611349138341426,
+      "grad_norm": 6.641354299638857,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 18322
+    },
+    {
+      "epoch": 0.43613729410644575,
+      "grad_norm": 4.708590616785259,
+      "learning_rate": 2e-05,
+      "loss": 0.5559,
+      "step": 18323
+    },
+    {
+      "epoch": 0.4361610968294773,
+      "grad_norm": 3.483508042072712,
+      "learning_rate": 2e-05,
+      "loss": 0.565,
+      "step": 18324
+    },
+    {
+      "epoch": 0.43618489955250883,
+      "grad_norm": 4.503677148699122,
+      "learning_rate": 2e-05,
+      "loss": 0.8552,
+      "step": 18325
+    },
+    {
+      "epoch": 0.4362087022755403,
+      "grad_norm": 6.7371983721452615,
+      "learning_rate": 2e-05,
+      "loss": 1.239,
+      "step": 18326
+    },
+    {
+      "epoch": 0.43623250499857186,
+      "grad_norm": 3.6295126116537646,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 18327
+    },
+    {
+      "epoch": 0.43625630772160334,
+      "grad_norm": 4.5011492085815075,
+      "learning_rate": 2e-05,
+      "loss": 0.2649,
+      "step": 18328
+    },
+    {
+      "epoch": 0.4362801104446349,
+      "grad_norm": 1.5101022571355014,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 18329
+    },
+    {
+      "epoch": 0.43630391316766637,
+      "grad_norm": 0.24033683664485472,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18330
+    },
+    {
+      "epoch": 0.4363277158906979,
+      "grad_norm": 0.963461953440682,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 18331
+    },
+    {
+      "epoch": 0.4363515186137294,
+      "grad_norm": 4.573314316645102,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 18332
+    },
+    {
+      "epoch": 0.43637532133676094,
+      "grad_norm": 3.3839117092741757,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 18333
+    },
+    {
+      "epoch": 0.4363991240597924,
+      "grad_norm": 5.283302638137047,
+      "learning_rate": 2e-05,
+      "loss": 0.9307,
+      "step": 18334
+    },
+    {
+      "epoch": 0.43642292678282396,
+      "grad_norm": 4.236455301353303,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 18335
+    },
+    {
+      "epoch": 0.43644672950585545,
+      "grad_norm": 2.125549968387035,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 18336
+    },
+    {
+      "epoch": 0.436470532228887,
+      "grad_norm": 1.9217350523607473,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 18337
+    },
+    {
+      "epoch": 0.4364943349519185,
+      "grad_norm": 11.062678464901982,
+      "learning_rate": 2e-05,
+      "loss": 0.2586,
+      "step": 18338
+    },
+    {
+      "epoch": 0.43651813767495,
+      "grad_norm": 0.997848255912962,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 18339
+    },
+    {
+      "epoch": 0.43654194039798155,
+      "grad_norm": 1.2156012005403574,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 18340
+    },
+    {
+      "epoch": 0.43656574312101304,
+      "grad_norm": 5.158732917607504,
+      "learning_rate": 2e-05,
+      "loss": 0.7502,
+      "step": 18341
+    },
+    {
+      "epoch": 0.4365895458440446,
+      "grad_norm": 0.7482921028392421,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 18342
+    },
+    {
+      "epoch": 0.43661334856707606,
+      "grad_norm": 5.287847173512929,
+      "learning_rate": 2e-05,
+      "loss": 0.3451,
+      "step": 18343
+    },
+    {
+      "epoch": 0.4366371512901076,
+      "grad_norm": 1.6057278304002085,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 18344
+    },
+    {
+      "epoch": 0.4366609540131391,
+      "grad_norm": 3.5589049301061744,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 18345
+    },
+    {
+      "epoch": 0.43668475673617063,
+      "grad_norm": 1.8272616796764534,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 18346
+    },
+    {
+      "epoch": 0.4367085594592021,
+      "grad_norm": 7.90962172357102,
+      "learning_rate": 2e-05,
+      "loss": 0.2035,
+      "step": 18347
+    },
+    {
+      "epoch": 0.43673236218223366,
+      "grad_norm": 2.1326274884058205,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 18348
+    },
+    {
+      "epoch": 0.43675616490526514,
+      "grad_norm": 2.3840055060853467,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 18349
+    },
+    {
+      "epoch": 0.4367799676282967,
+      "grad_norm": 1.0418877494956875,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 18350
+    },
+    {
+      "epoch": 0.43680377035132817,
+      "grad_norm": 12.301990784156423,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 18351
+    },
+    {
+      "epoch": 0.4368275730743597,
+      "grad_norm": 5.3519584870635715,
+      "learning_rate": 2e-05,
+      "loss": 0.8749,
+      "step": 18352
+    },
+    {
+      "epoch": 0.4368513757973912,
+      "grad_norm": 4.757413487104291,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 18353
+    },
+    {
+      "epoch": 0.43687517852042274,
+      "grad_norm": 5.9436019566875675,
+      "learning_rate": 2e-05,
+      "loss": 0.9224,
+      "step": 18354
+    },
+    {
+      "epoch": 0.4368989812434543,
+      "grad_norm": 1.3997777156801634,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 18355
+    },
+    {
+      "epoch": 0.43692278396648576,
+      "grad_norm": 0.6397773112665117,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18356
+    },
+    {
+      "epoch": 0.4369465866895173,
+      "grad_norm": 5.259336199240801,
+      "learning_rate": 2e-05,
+      "loss": 0.8585,
+      "step": 18357
+    },
+    {
+      "epoch": 0.4369703894125488,
+      "grad_norm": 2.7204824718957923,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 18358
+    },
+    {
+      "epoch": 0.43699419213558033,
+      "grad_norm": 14.084594966197262,
+      "learning_rate": 2e-05,
+      "loss": 1.0054,
+      "step": 18359
+    },
+    {
+      "epoch": 0.4370179948586118,
+      "grad_norm": 3.253907941856031,
+      "learning_rate": 2e-05,
+      "loss": 0.5196,
+      "step": 18360
+    },
+    {
+      "epoch": 0.43704179758164335,
+      "grad_norm": 5.869626377349834,
+      "learning_rate": 2e-05,
+      "loss": 0.9137,
+      "step": 18361
+    },
+    {
+      "epoch": 0.43706560030467484,
+      "grad_norm": 4.5689260783427565,
+      "learning_rate": 2e-05,
+      "loss": 0.5415,
+      "step": 18362
+    },
+    {
+      "epoch": 0.4370894030277064,
+      "grad_norm": 3.491241339227205,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 18363
+    },
+    {
+      "epoch": 0.43711320575073787,
+      "grad_norm": 5.722178334276299,
+      "learning_rate": 2e-05,
+      "loss": 0.2689,
+      "step": 18364
+    },
+    {
+      "epoch": 0.4371370084737694,
+      "grad_norm": 0.5414553734035258,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 18365
+    },
+    {
+      "epoch": 0.4371608111968009,
+      "grad_norm": 2.9453242825441643,
+      "learning_rate": 2e-05,
+      "loss": 0.3957,
+      "step": 18366
+    },
+    {
+      "epoch": 0.43718461391983243,
+      "grad_norm": 6.452435192875521,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 18367
+    },
+    {
+      "epoch": 0.4372084166428639,
+      "grad_norm": 6.171253330891303,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 18368
+    },
+    {
+      "epoch": 0.43723221936589546,
+      "grad_norm": 1.0436896118371468,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 18369
+    },
+    {
+      "epoch": 0.437256022088927,
+      "grad_norm": 3.311525624931971,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 18370
+    },
+    {
+      "epoch": 0.4372798248119585,
+      "grad_norm": 1.3448184394342961,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 18371
+    },
+    {
+      "epoch": 0.43730362753499,
+      "grad_norm": 11.911235091785432,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 18372
+    },
+    {
+      "epoch": 0.4373274302580215,
+      "grad_norm": 6.131083207094403,
+      "learning_rate": 2e-05,
+      "loss": 0.2687,
+      "step": 18373
+    },
+    {
+      "epoch": 0.43735123298105305,
+      "grad_norm": 8.300442750868978,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 18374
+    },
+    {
+      "epoch": 0.43737503570408454,
+      "grad_norm": 2.2765493399752876,
+      "learning_rate": 2e-05,
+      "loss": 0.2,
+      "step": 18375
+    },
+    {
+      "epoch": 0.4373988384271161,
+      "grad_norm": 4.529335883316464,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 18376
+    },
+    {
+      "epoch": 0.43742264115014756,
+      "grad_norm": 2.232724161906066,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 18377
+    },
+    {
+      "epoch": 0.4374464438731791,
+      "grad_norm": 0.5476094663444658,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 18378
+    },
+    {
+      "epoch": 0.4374702465962106,
+      "grad_norm": 4.512442087964775,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 18379
+    },
+    {
+      "epoch": 0.43749404931924213,
+      "grad_norm": 0.4219425221343925,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 18380
+    },
+    {
+      "epoch": 0.4375178520422736,
+      "grad_norm": 4.280574401780845,
+      "learning_rate": 2e-05,
+      "loss": 0.5081,
+      "step": 18381
+    },
+    {
+      "epoch": 0.43754165476530515,
+      "grad_norm": 3.26011609146922,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 18382
+    },
+    {
+      "epoch": 0.43756545748833664,
+      "grad_norm": 3.539003482994629,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 18383
+    },
+    {
+      "epoch": 0.4375892602113682,
+      "grad_norm": 2930.473006144675,
+      "learning_rate": 2e-05,
+      "loss": 0.8495,
+      "step": 18384
+    },
+    {
+      "epoch": 0.4376130629343997,
+      "grad_norm": 2.9607871285972633,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 18385
+    },
+    {
+      "epoch": 0.4376368656574312,
+      "grad_norm": 4.416981093428238,
+      "learning_rate": 2e-05,
+      "loss": 1.0589,
+      "step": 18386
+    },
+    {
+      "epoch": 0.43766066838046275,
+      "grad_norm": 0.1705452146771988,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 18387
+    },
+    {
+      "epoch": 0.43768447110349423,
+      "grad_norm": 4.3895674361000365,
+      "learning_rate": 2e-05,
+      "loss": 0.4861,
+      "step": 18388
+    },
+    {
+      "epoch": 0.4377082738265258,
+      "grad_norm": 2.0040314227127727,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 18389
+    },
+    {
+      "epoch": 0.43773207654955726,
+      "grad_norm": 6.226432141938981,
+      "learning_rate": 2e-05,
+      "loss": 0.939,
+      "step": 18390
+    },
+    {
+      "epoch": 0.4377558792725888,
+      "grad_norm": 28.01886387291696,
+      "learning_rate": 2e-05,
+      "loss": 1.0285,
+      "step": 18391
+    },
+    {
+      "epoch": 0.4377796819956203,
+      "grad_norm": 5.13327371090762,
+      "learning_rate": 2e-05,
+      "loss": 0.7221,
+      "step": 18392
+    },
+    {
+      "epoch": 0.4378034847186518,
+      "grad_norm": 0.1536591853101836,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18393
+    },
+    {
+      "epoch": 0.4378272874416833,
+      "grad_norm": 5.74295013858627,
+      "learning_rate": 2e-05,
+      "loss": 1.5701,
+      "step": 18394
+    },
+    {
+      "epoch": 0.43785109016471485,
+      "grad_norm": 2.8001407621145424,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 18395
+    },
+    {
+      "epoch": 0.43787489288774634,
+      "grad_norm": 4.260068179566073,
+      "learning_rate": 2e-05,
+      "loss": 0.6343,
+      "step": 18396
+    },
+    {
+      "epoch": 0.4378986956107779,
+      "grad_norm": 2.1451174514017692,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 18397
+    },
+    {
+      "epoch": 0.43792249833380936,
+      "grad_norm": 0.09624785389159228,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 18398
+    },
+    {
+      "epoch": 0.4379463010568409,
+      "grad_norm": 9.898222794942658,
+      "learning_rate": 2e-05,
+      "loss": 0.5733,
+      "step": 18399
+    },
+    {
+      "epoch": 0.43797010377987244,
+      "grad_norm": 8.594809956331503,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 18400
+    },
+    {
+      "epoch": 0.43799390650290393,
+      "grad_norm": 3.9891395603990083,
+      "learning_rate": 2e-05,
+      "loss": 0.5172,
+      "step": 18401
+    },
+    {
+      "epoch": 0.43801770922593547,
+      "grad_norm": 5.0768446013801185,
+      "learning_rate": 2e-05,
+      "loss": 0.6431,
+      "step": 18402
+    },
+    {
+      "epoch": 0.43804151194896696,
+      "grad_norm": 0.6278653594161626,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 18403
+    },
+    {
+      "epoch": 0.4380653146719985,
+      "grad_norm": 26.48488915454047,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 18404
+    },
+    {
+      "epoch": 0.43808911739503,
+      "grad_norm": 9.244367117219022,
+      "learning_rate": 2e-05,
+      "loss": 0.2752,
+      "step": 18405
+    },
+    {
+      "epoch": 0.4381129201180615,
+      "grad_norm": 0.4751562754459796,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18406
+    },
+    {
+      "epoch": 0.438136722841093,
+      "grad_norm": 2.041491529152465,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 18407
+    },
+    {
+      "epoch": 0.43816052556412455,
+      "grad_norm": 6.910714849188314,
+      "learning_rate": 2e-05,
+      "loss": 0.9925,
+      "step": 18408
+    },
+    {
+      "epoch": 0.43818432828715603,
+      "grad_norm": 4.4019598734259695,
+      "learning_rate": 2e-05,
+      "loss": 0.543,
+      "step": 18409
+    },
+    {
+      "epoch": 0.4382081310101876,
+      "grad_norm": 4.325788414142824,
+      "learning_rate": 2e-05,
+      "loss": 0.5775,
+      "step": 18410
+    },
+    {
+      "epoch": 0.43823193373321906,
+      "grad_norm": 5.777843611460916,
+      "learning_rate": 2e-05,
+      "loss": 0.7747,
+      "step": 18411
+    },
+    {
+      "epoch": 0.4382557364562506,
+      "grad_norm": 1.9550285331401769,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 18412
+    },
+    {
+      "epoch": 0.4382795391792821,
+      "grad_norm": 4.02293769402732,
+      "learning_rate": 2e-05,
+      "loss": 1.0158,
+      "step": 18413
+    },
+    {
+      "epoch": 0.4383033419023136,
+      "grad_norm": 1.4875290209189949,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 18414
+    },
+    {
+      "epoch": 0.43832714462534517,
+      "grad_norm": 49.84453464213927,
+      "learning_rate": 2e-05,
+      "loss": 0.3915,
+      "step": 18415
+    },
+    {
+      "epoch": 0.43835094734837665,
+      "grad_norm": 4.0038913289520215,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 18416
+    },
+    {
+      "epoch": 0.4383747500714082,
+      "grad_norm": 5.3757902472383785,
+      "learning_rate": 2e-05,
+      "loss": 1.0484,
+      "step": 18417
+    },
+    {
+      "epoch": 0.4383985527944397,
+      "grad_norm": 4.781008216112681,
+      "learning_rate": 2e-05,
+      "loss": 0.3651,
+      "step": 18418
+    },
+    {
+      "epoch": 0.4384223555174712,
+      "grad_norm": 0.10820350222789439,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 18419
+    },
+    {
+      "epoch": 0.4384461582405027,
+      "grad_norm": 2.823712591494069,
+      "learning_rate": 2e-05,
+      "loss": 0.1106,
+      "step": 18420
+    },
+    {
+      "epoch": 0.43846996096353424,
+      "grad_norm": 1.4482508699765766,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 18421
+    },
+    {
+      "epoch": 0.43849376368656573,
+      "grad_norm": 1.5843243178868527,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 18422
+    },
+    {
+      "epoch": 0.43851756640959727,
+      "grad_norm": 1.1609813178291022,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 18423
+    },
+    {
+      "epoch": 0.43854136913262876,
+      "grad_norm": 2.2534992276055847,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 18424
+    },
+    {
+      "epoch": 0.4385651718556603,
+      "grad_norm": 0.6644944490973268,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 18425
+    },
+    {
+      "epoch": 0.4385889745786918,
+      "grad_norm": 3.8232429264692955,
+      "learning_rate": 2e-05,
+      "loss": 0.5999,
+      "step": 18426
+    },
+    {
+      "epoch": 0.4386127773017233,
+      "grad_norm": 4.188675237358517,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 18427
+    },
+    {
+      "epoch": 0.4386365800247548,
+      "grad_norm": 6.429058760341003,
+      "learning_rate": 2e-05,
+      "loss": 1.8922,
+      "step": 18428
+    },
+    {
+      "epoch": 0.43866038274778635,
+      "grad_norm": 14.941432200496596,
+      "learning_rate": 2e-05,
+      "loss": 1.0428,
+      "step": 18429
+    },
+    {
+      "epoch": 0.4386841854708179,
+      "grad_norm": 35.708396461137404,
+      "learning_rate": 2e-05,
+      "loss": 0.857,
+      "step": 18430
+    },
+    {
+      "epoch": 0.4387079881938494,
+      "grad_norm": 1.2794660803457774,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 18431
+    },
+    {
+      "epoch": 0.4387317909168809,
+      "grad_norm": 1.4737023175125659,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 18432
+    },
+    {
+      "epoch": 0.4387555936399124,
+      "grad_norm": 3.2733772314854117,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 18433
+    },
+    {
+      "epoch": 0.43877939636294394,
+      "grad_norm": 1.9977950572189571,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 18434
+    },
+    {
+      "epoch": 0.4388031990859754,
+      "grad_norm": 1.2767146631140647,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 18435
+    },
+    {
+      "epoch": 0.43882700180900697,
+      "grad_norm": 9.830899497362811,
+      "learning_rate": 2e-05,
+      "loss": 0.3157,
+      "step": 18436
+    },
+    {
+      "epoch": 0.43885080453203845,
+      "grad_norm": 10.597757343293887,
+      "learning_rate": 2e-05,
+      "loss": 0.7225,
+      "step": 18437
+    },
+    {
+      "epoch": 0.43887460725507,
+      "grad_norm": 13.91821215899627,
+      "learning_rate": 2e-05,
+      "loss": 0.3732,
+      "step": 18438
+    },
+    {
+      "epoch": 0.4388984099781015,
+      "grad_norm": 2.992746458692247,
+      "learning_rate": 2e-05,
+      "loss": 0.6133,
+      "step": 18439
+    },
+    {
+      "epoch": 0.438922212701133,
+      "grad_norm": 9.857374744330222,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 18440
+    },
+    {
+      "epoch": 0.4389460154241645,
+      "grad_norm": 5.451225957651054,
+      "learning_rate": 2e-05,
+      "loss": 0.5367,
+      "step": 18441
+    },
+    {
+      "epoch": 0.43896981814719604,
+      "grad_norm": 13.594894726000172,
+      "learning_rate": 2e-05,
+      "loss": 1.5792,
+      "step": 18442
+    },
+    {
+      "epoch": 0.43899362087022753,
+      "grad_norm": 1.8460859093556534,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 18443
+    },
+    {
+      "epoch": 0.43901742359325907,
+      "grad_norm": 11.345250633356532,
+      "learning_rate": 2e-05,
+      "loss": 0.4673,
+      "step": 18444
+    },
+    {
+      "epoch": 0.43904122631629056,
+      "grad_norm": 3.93332874514614,
+      "learning_rate": 2e-05,
+      "loss": 0.3699,
+      "step": 18445
+    },
+    {
+      "epoch": 0.4390650290393221,
+      "grad_norm": 9.714073911884766,
+      "learning_rate": 2e-05,
+      "loss": 0.5334,
+      "step": 18446
+    },
+    {
+      "epoch": 0.43908883176235364,
+      "grad_norm": 5.379737188667996,
+      "learning_rate": 2e-05,
+      "loss": 0.8007,
+      "step": 18447
+    },
+    {
+      "epoch": 0.4391126344853851,
+      "grad_norm": 5.62257372450245,
+      "learning_rate": 2e-05,
+      "loss": 0.5517,
+      "step": 18448
+    },
+    {
+      "epoch": 0.43913643720841666,
+      "grad_norm": 3.011665750396206,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 18449
+    },
+    {
+      "epoch": 0.43916023993144815,
+      "grad_norm": 3.9030586613225426,
+      "learning_rate": 2e-05,
+      "loss": 0.4012,
+      "step": 18450
+    },
+    {
+      "epoch": 0.4391840426544797,
+      "grad_norm": 4.317326237844854,
+      "learning_rate": 2e-05,
+      "loss": 1.2282,
+      "step": 18451
+    },
+    {
+      "epoch": 0.4392078453775112,
+      "grad_norm": 10.501876511730076,
+      "learning_rate": 2e-05,
+      "loss": 0.9622,
+      "step": 18452
+    },
+    {
+      "epoch": 0.4392316481005427,
+      "grad_norm": 6.363031247570119,
+      "learning_rate": 2e-05,
+      "loss": 1.14,
+      "step": 18453
+    },
+    {
+      "epoch": 0.4392554508235742,
+      "grad_norm": 19.64033923058078,
+      "learning_rate": 2e-05,
+      "loss": 0.236,
+      "step": 18454
+    },
+    {
+      "epoch": 0.43927925354660574,
+      "grad_norm": 9.037169258454531,
+      "learning_rate": 2e-05,
+      "loss": 0.5609,
+      "step": 18455
+    },
+    {
+      "epoch": 0.4393030562696372,
+      "grad_norm": 3.427988815353481,
+      "learning_rate": 2e-05,
+      "loss": 0.4829,
+      "step": 18456
+    },
+    {
+      "epoch": 0.43932685899266877,
+      "grad_norm": 8.781322307487617,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 18457
+    },
+    {
+      "epoch": 0.43935066171570025,
+      "grad_norm": 0.5727108277751124,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 18458
+    },
+    {
+      "epoch": 0.4393744644387318,
+      "grad_norm": 3.2878989512631778,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 18459
+    },
+    {
+      "epoch": 0.4393982671617633,
+      "grad_norm": 3.8813556876076336,
+      "learning_rate": 2e-05,
+      "loss": 0.6737,
+      "step": 18460
+    },
+    {
+      "epoch": 0.4394220698847948,
+      "grad_norm": 0.6685977083473121,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 18461
+    },
+    {
+      "epoch": 0.43944587260782636,
+      "grad_norm": 4.01782382566049,
+      "learning_rate": 2e-05,
+      "loss": 0.223,
+      "step": 18462
+    },
+    {
+      "epoch": 0.43946967533085785,
+      "grad_norm": 4.680391407720467,
+      "learning_rate": 2e-05,
+      "loss": 0.7125,
+      "step": 18463
+    },
+    {
+      "epoch": 0.4394934780538894,
+      "grad_norm": 5.305705662394993,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 18464
+    },
+    {
+      "epoch": 0.43951728077692087,
+      "grad_norm": 0.987163475956507,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 18465
+    },
+    {
+      "epoch": 0.4395410834999524,
+      "grad_norm": 4.080170852233746,
+      "learning_rate": 2e-05,
+      "loss": 0.1436,
+      "step": 18466
+    },
+    {
+      "epoch": 0.4395648862229839,
+      "grad_norm": 11.30792574126167,
+      "learning_rate": 2e-05,
+      "loss": 0.8619,
+      "step": 18467
+    },
+    {
+      "epoch": 0.43958868894601544,
+      "grad_norm": 2.1440077752135567,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 18468
+    },
+    {
+      "epoch": 0.4396124916690469,
+      "grad_norm": 8.217680526097942,
+      "learning_rate": 2e-05,
+      "loss": 1.951,
+      "step": 18469
+    },
+    {
+      "epoch": 0.43963629439207846,
+      "grad_norm": 1.9716698581097498,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 18470
+    },
+    {
+      "epoch": 0.43966009711510995,
+      "grad_norm": 6.142306041570587,
+      "learning_rate": 2e-05,
+      "loss": 0.2577,
+      "step": 18471
+    },
+    {
+      "epoch": 0.4396838998381415,
+      "grad_norm": 4.29358017179279,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 18472
+    },
+    {
+      "epoch": 0.439707702561173,
+      "grad_norm": 8.160498121169761,
+      "learning_rate": 2e-05,
+      "loss": 2.0927,
+      "step": 18473
+    },
+    {
+      "epoch": 0.4397315052842045,
+      "grad_norm": 7.551363992723108,
+      "learning_rate": 2e-05,
+      "loss": 1.8085,
+      "step": 18474
+    },
+    {
+      "epoch": 0.439755308007236,
+      "grad_norm": 0.29628478005141123,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18475
+    },
+    {
+      "epoch": 0.43977911073026754,
+      "grad_norm": 6.7360425804155275,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 18476
+    },
+    {
+      "epoch": 0.4398029134532991,
+      "grad_norm": 1.2965508879464376,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 18477
+    },
+    {
+      "epoch": 0.43982671617633057,
+      "grad_norm": 2.662358160344174,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 18478
+    },
+    {
+      "epoch": 0.4398505188993621,
+      "grad_norm": 3.9505229152515744,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 18479
+    },
+    {
+      "epoch": 0.4398743216223936,
+      "grad_norm": 3.8457083130926537,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 18480
+    },
+    {
+      "epoch": 0.43989812434542513,
+      "grad_norm": 1.7461951391699584,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 18481
+    },
+    {
+      "epoch": 0.4399219270684566,
+      "grad_norm": 0.8966723169192725,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 18482
+    },
+    {
+      "epoch": 0.43994572979148816,
+      "grad_norm": 0.6174536064007558,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 18483
+    },
+    {
+      "epoch": 0.43996953251451965,
+      "grad_norm": 2.6847460570085917,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 18484
+    },
+    {
+      "epoch": 0.4399933352375512,
+      "grad_norm": 2.3852775611899055,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 18485
+    },
+    {
+      "epoch": 0.44001713796058267,
+      "grad_norm": 14.08670756963147,
+      "learning_rate": 2e-05,
+      "loss": 1.8994,
+      "step": 18486
+    },
+    {
+      "epoch": 0.4400409406836142,
+      "grad_norm": 3.1134287798789355,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 18487
+    },
+    {
+      "epoch": 0.4400647434066457,
+      "grad_norm": 6.663439588909993,
+      "learning_rate": 2e-05,
+      "loss": 0.2608,
+      "step": 18488
+    },
+    {
+      "epoch": 0.44008854612967724,
+      "grad_norm": 16.891992871009034,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 18489
+    },
+    {
+      "epoch": 0.4401123488527087,
+      "grad_norm": 4.045810848948703,
+      "learning_rate": 2e-05,
+      "loss": 0.8139,
+      "step": 18490
+    },
+    {
+      "epoch": 0.44013615157574026,
+      "grad_norm": 2.376362389312263,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 18491
+    },
+    {
+      "epoch": 0.4401599542987718,
+      "grad_norm": 3.2545917868389598,
+      "learning_rate": 2e-05,
+      "loss": 0.2367,
+      "step": 18492
+    },
+    {
+      "epoch": 0.4401837570218033,
+      "grad_norm": 1.3575563387545782,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 18493
+    },
+    {
+      "epoch": 0.44020755974483483,
+      "grad_norm": 0.7878685411828054,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 18494
+    },
+    {
+      "epoch": 0.4402313624678663,
+      "grad_norm": 10.484163220048895,
+      "learning_rate": 2e-05,
+      "loss": 0.8974,
+      "step": 18495
+    },
+    {
+      "epoch": 0.44025516519089786,
+      "grad_norm": 5.666346949255792,
+      "learning_rate": 2e-05,
+      "loss": 0.8547,
+      "step": 18496
+    },
+    {
+      "epoch": 0.44027896791392934,
+      "grad_norm": 4.181905206566195,
+      "learning_rate": 2e-05,
+      "loss": 0.4357,
+      "step": 18497
+    },
+    {
+      "epoch": 0.4403027706369609,
+      "grad_norm": 1.7361115638942162,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 18498
+    },
+    {
+      "epoch": 0.44032657335999237,
+      "grad_norm": 4.242524559240272,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 18499
+    },
+    {
+      "epoch": 0.4403503760830239,
+      "grad_norm": 1.7968216790624936,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 18500
+    },
+    {
+      "epoch": 0.4403741788060554,
+      "grad_norm": 5.972837950548965,
+      "learning_rate": 2e-05,
+      "loss": 1.7092,
+      "step": 18501
+    },
+    {
+      "epoch": 0.44039798152908693,
+      "grad_norm": 4.749115828896411,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 18502
+    },
+    {
+      "epoch": 0.4404217842521184,
+      "grad_norm": 2.4845971654318677,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 18503
+    },
+    {
+      "epoch": 0.44044558697514996,
+      "grad_norm": 1.1670837775225258,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 18504
+    },
+    {
+      "epoch": 0.44046938969818145,
+      "grad_norm": 3.6544303934190343,
+      "learning_rate": 2e-05,
+      "loss": 0.5819,
+      "step": 18505
+    },
+    {
+      "epoch": 0.440493192421213,
+      "grad_norm": 1.4303531953674693,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 18506
+    },
+    {
+      "epoch": 0.4405169951442445,
+      "grad_norm": 1.8686578159058231,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 18507
+    },
+    {
+      "epoch": 0.440540797867276,
+      "grad_norm": 4.328071160198162,
+      "learning_rate": 2e-05,
+      "loss": 0.6099,
+      "step": 18508
+    },
+    {
+      "epoch": 0.44056460059030755,
+      "grad_norm": 5.048880749973647,
+      "learning_rate": 2e-05,
+      "loss": 0.9011,
+      "step": 18509
+    },
+    {
+      "epoch": 0.44058840331333904,
+      "grad_norm": 1.632070206601632,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 18510
+    },
+    {
+      "epoch": 0.4406122060363706,
+      "grad_norm": 4.635829967759478,
+      "learning_rate": 2e-05,
+      "loss": 0.8952,
+      "step": 18511
+    },
+    {
+      "epoch": 0.44063600875940206,
+      "grad_norm": 3.980538900272251,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 18512
+    },
+    {
+      "epoch": 0.4406598114824336,
+      "grad_norm": 0.6633984152079345,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 18513
+    },
+    {
+      "epoch": 0.4406836142054651,
+      "grad_norm": 2.695213002726524,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 18514
+    },
+    {
+      "epoch": 0.44070741692849663,
+      "grad_norm": 0.6161562214333907,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 18515
+    },
+    {
+      "epoch": 0.4407312196515281,
+      "grad_norm": 0.31065530827513205,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 18516
+    },
+    {
+      "epoch": 0.44075502237455966,
+      "grad_norm": 1.60289677695565,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 18517
+    },
+    {
+      "epoch": 0.44077882509759114,
+      "grad_norm": 0.47472474089257466,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 18518
+    },
+    {
+      "epoch": 0.4408026278206227,
+      "grad_norm": 4.266268525643878,
+      "learning_rate": 2e-05,
+      "loss": 0.5011,
+      "step": 18519
+    },
+    {
+      "epoch": 0.44082643054365417,
+      "grad_norm": 10.479735439241365,
+      "learning_rate": 2e-05,
+      "loss": 1.0353,
+      "step": 18520
+    },
+    {
+      "epoch": 0.4408502332666857,
+      "grad_norm": 2.0065245780996026,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 18521
+    },
+    {
+      "epoch": 0.44087403598971725,
+      "grad_norm": 3.8845132906727926,
+      "learning_rate": 2e-05,
+      "loss": 0.3932,
+      "step": 18522
+    },
+    {
+      "epoch": 0.44089783871274874,
+      "grad_norm": 3.662100098791207,
+      "learning_rate": 2e-05,
+      "loss": 0.7818,
+      "step": 18523
+    },
+    {
+      "epoch": 0.4409216414357803,
+      "grad_norm": 4.212080006184783,
+      "learning_rate": 2e-05,
+      "loss": 0.3495,
+      "step": 18524
+    },
+    {
+      "epoch": 0.44094544415881176,
+      "grad_norm": 0.9264376690006034,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 18525
+    },
+    {
+      "epoch": 0.4409692468818433,
+      "grad_norm": 16.65260095049803,
+      "learning_rate": 2e-05,
+      "loss": 0.1428,
+      "step": 18526
+    },
+    {
+      "epoch": 0.4409930496048748,
+      "grad_norm": 4.8235049979084845,
+      "learning_rate": 2e-05,
+      "loss": 0.7692,
+      "step": 18527
+    },
+    {
+      "epoch": 0.44101685232790633,
+      "grad_norm": 0.8315928867839673,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 18528
+    },
+    {
+      "epoch": 0.4410406550509378,
+      "grad_norm": 6.10808943932743,
+      "learning_rate": 2e-05,
+      "loss": 0.8241,
+      "step": 18529
+    },
+    {
+      "epoch": 0.44106445777396935,
+      "grad_norm": 2.0039906769433067,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 18530
+    },
+    {
+      "epoch": 0.44108826049700084,
+      "grad_norm": 11.966416521959818,
+      "learning_rate": 2e-05,
+      "loss": 1.0954,
+      "step": 18531
+    },
+    {
+      "epoch": 0.4411120632200324,
+      "grad_norm": 7.626180533738857,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 18532
+    },
+    {
+      "epoch": 0.44113586594306387,
+      "grad_norm": 10.33512540661021,
+      "learning_rate": 2e-05,
+      "loss": 1.0806,
+      "step": 18533
+    },
+    {
+      "epoch": 0.4411596686660954,
+      "grad_norm": 26.557145897466462,
+      "learning_rate": 2e-05,
+      "loss": 0.5848,
+      "step": 18534
+    },
+    {
+      "epoch": 0.4411834713891269,
+      "grad_norm": 5.978782464005346,
+      "learning_rate": 2e-05,
+      "loss": 0.9277,
+      "step": 18535
+    },
+    {
+      "epoch": 0.44120727411215843,
+      "grad_norm": 3.085293846100224,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 18536
+    },
+    {
+      "epoch": 0.44123107683519,
+      "grad_norm": 181.0087121528523,
+      "learning_rate": 2e-05,
+      "loss": 1.2355,
+      "step": 18537
+    },
+    {
+      "epoch": 0.44125487955822146,
+      "grad_norm": 5.6874901361825225,
+      "learning_rate": 2e-05,
+      "loss": 0.2518,
+      "step": 18538
+    },
+    {
+      "epoch": 0.441278682281253,
+      "grad_norm": 0.5313951102573382,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 18539
+    },
+    {
+      "epoch": 0.4413024850042845,
+      "grad_norm": 4.278307704331171,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 18540
+    },
+    {
+      "epoch": 0.441326287727316,
+      "grad_norm": 0.48967945582798467,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 18541
+    },
+    {
+      "epoch": 0.4413500904503475,
+      "grad_norm": 9.942393978072557,
+      "learning_rate": 2e-05,
+      "loss": 1.2395,
+      "step": 18542
+    },
+    {
+      "epoch": 0.44137389317337905,
+      "grad_norm": 1.8214152973134967,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 18543
+    },
+    {
+      "epoch": 0.44139769589641054,
+      "grad_norm": 6.948867135719761,
+      "learning_rate": 2e-05,
+      "loss": 1.3077,
+      "step": 18544
+    },
+    {
+      "epoch": 0.4414214986194421,
+      "grad_norm": 5.738364640935094,
+      "learning_rate": 2e-05,
+      "loss": 1.4419,
+      "step": 18545
+    },
+    {
+      "epoch": 0.44144530134247356,
+      "grad_norm": 3.613218560123088,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 18546
+    },
+    {
+      "epoch": 0.4414691040655051,
+      "grad_norm": 2.863869648023071,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 18547
+    },
+    {
+      "epoch": 0.4414929067885366,
+      "grad_norm": 7.447626852232986,
+      "learning_rate": 2e-05,
+      "loss": 0.8202,
+      "step": 18548
+    },
+    {
+      "epoch": 0.44151670951156813,
+      "grad_norm": 9.038272902335402,
+      "learning_rate": 2e-05,
+      "loss": 0.9122,
+      "step": 18549
+    },
+    {
+      "epoch": 0.4415405122345996,
+      "grad_norm": 0.27143057900394163,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 18550
+    },
+    {
+      "epoch": 0.44156431495763115,
+      "grad_norm": 0.8563684563209304,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 18551
+    },
+    {
+      "epoch": 0.4415881176806627,
+      "grad_norm": 2.69005007043409,
+      "learning_rate": 2e-05,
+      "loss": 0.4864,
+      "step": 18552
+    },
+    {
+      "epoch": 0.4416119204036942,
+      "grad_norm": 1.5551898393208536,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 18553
+    },
+    {
+      "epoch": 0.4416357231267257,
+      "grad_norm": 2.8791230130514918,
+      "learning_rate": 2e-05,
+      "loss": 0.6124,
+      "step": 18554
+    },
+    {
+      "epoch": 0.4416595258497572,
+      "grad_norm": 4.918927502532116,
+      "learning_rate": 2e-05,
+      "loss": 0.2853,
+      "step": 18555
+    },
+    {
+      "epoch": 0.44168332857278875,
+      "grad_norm": 7.123339780746511,
+      "learning_rate": 2e-05,
+      "loss": 0.8374,
+      "step": 18556
+    },
+    {
+      "epoch": 0.44170713129582023,
+      "grad_norm": 2.1784134528607924,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 18557
+    },
+    {
+      "epoch": 0.4417309340188518,
+      "grad_norm": 2.2266499156935984,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 18558
+    },
+    {
+      "epoch": 0.44175473674188326,
+      "grad_norm": 4.586127894946763,
+      "learning_rate": 2e-05,
+      "loss": 0.2414,
+      "step": 18559
+    },
+    {
+      "epoch": 0.4417785394649148,
+      "grad_norm": 2.43836529146207,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 18560
+    },
+    {
+      "epoch": 0.4418023421879463,
+      "grad_norm": 0.18586668304956122,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18561
+    },
+    {
+      "epoch": 0.4418261449109778,
+      "grad_norm": 5.47255528117818,
+      "learning_rate": 2e-05,
+      "loss": 1.2198,
+      "step": 18562
+    },
+    {
+      "epoch": 0.4418499476340093,
+      "grad_norm": 11.297433110419204,
+      "learning_rate": 2e-05,
+      "loss": 0.2957,
+      "step": 18563
+    },
+    {
+      "epoch": 0.44187375035704085,
+      "grad_norm": 12.237227306451889,
+      "learning_rate": 2e-05,
+      "loss": 0.4449,
+      "step": 18564
+    },
+    {
+      "epoch": 0.44189755308007234,
+      "grad_norm": 2.6121815258097305,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 18565
+    },
+    {
+      "epoch": 0.4419213558031039,
+      "grad_norm": 2.5962426153539364,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 18566
+    },
+    {
+      "epoch": 0.4419451585261354,
+      "grad_norm": 2.6902368723906984,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 18567
+    },
+    {
+      "epoch": 0.4419689612491669,
+      "grad_norm": 0.36411172080154675,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18568
+    },
+    {
+      "epoch": 0.44199276397219844,
+      "grad_norm": 1.7606924552695158,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 18569
+    },
+    {
+      "epoch": 0.44201656669522993,
+      "grad_norm": 9.328375680990328,
+      "learning_rate": 2e-05,
+      "loss": 0.7153,
+      "step": 18570
+    },
+    {
+      "epoch": 0.44204036941826147,
+      "grad_norm": 4.745029465710754,
+      "learning_rate": 2e-05,
+      "loss": 0.5135,
+      "step": 18571
+    },
+    {
+      "epoch": 0.44206417214129295,
+      "grad_norm": 3.0089665315502954,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 18572
+    },
+    {
+      "epoch": 0.4420879748643245,
+      "grad_norm": 0.9642542712660835,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 18573
+    },
+    {
+      "epoch": 0.442111777587356,
+      "grad_norm": 10.077581931154354,
+      "learning_rate": 2e-05,
+      "loss": 1.5667,
+      "step": 18574
+    },
+    {
+      "epoch": 0.4421355803103875,
+      "grad_norm": 5.563385801992994,
+      "learning_rate": 2e-05,
+      "loss": 0.2093,
+      "step": 18575
+    },
+    {
+      "epoch": 0.442159383033419,
+      "grad_norm": 4.586576616542036,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 18576
+    },
+    {
+      "epoch": 0.44218318575645055,
+      "grad_norm": 1.61555478924234,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 18577
+    },
+    {
+      "epoch": 0.44220698847948203,
+      "grad_norm": 7.053582552515965,
+      "learning_rate": 2e-05,
+      "loss": 0.8783,
+      "step": 18578
+    },
+    {
+      "epoch": 0.4422307912025136,
+      "grad_norm": 6.262427251002567,
+      "learning_rate": 2e-05,
+      "loss": 1.0613,
+      "step": 18579
+    },
+    {
+      "epoch": 0.44225459392554506,
+      "grad_norm": 6.067156836891241,
+      "learning_rate": 2e-05,
+      "loss": 0.3174,
+      "step": 18580
+    },
+    {
+      "epoch": 0.4422783966485766,
+      "grad_norm": 3.573955439046618,
+      "learning_rate": 2e-05,
+      "loss": 0.3133,
+      "step": 18581
+    },
+    {
+      "epoch": 0.4423021993716081,
+      "grad_norm": 0.6653125927743018,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 18582
+    },
+    {
+      "epoch": 0.4423260020946396,
+      "grad_norm": 8.059583308402226,
+      "learning_rate": 2e-05,
+      "loss": 0.8235,
+      "step": 18583
+    },
+    {
+      "epoch": 0.44234980481767117,
+      "grad_norm": 2.296144783033111,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 18584
+    },
+    {
+      "epoch": 0.44237360754070265,
+      "grad_norm": 8.826106621111732,
+      "learning_rate": 2e-05,
+      "loss": 0.3021,
+      "step": 18585
+    },
+    {
+      "epoch": 0.4423974102637342,
+      "grad_norm": 3.7787611973894673,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 18586
+    },
+    {
+      "epoch": 0.4424212129867657,
+      "grad_norm": 7.059812071225241,
+      "learning_rate": 2e-05,
+      "loss": 0.3111,
+      "step": 18587
+    },
+    {
+      "epoch": 0.4424450157097972,
+      "grad_norm": 4.249039027106071,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 18588
+    },
+    {
+      "epoch": 0.4424688184328287,
+      "grad_norm": 2.272366054860626,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 18589
+    },
+    {
+      "epoch": 0.44249262115586024,
+      "grad_norm": 1.5575773899493535,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 18590
+    },
+    {
+      "epoch": 0.44251642387889173,
+      "grad_norm": 10.719712806806637,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 18591
+    },
+    {
+      "epoch": 0.44254022660192327,
+      "grad_norm": 4.428946422273874,
+      "learning_rate": 2e-05,
+      "loss": 0.8845,
+      "step": 18592
+    },
+    {
+      "epoch": 0.44256402932495476,
+      "grad_norm": 2.215493554895279,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 18593
+    },
+    {
+      "epoch": 0.4425878320479863,
+      "grad_norm": 2.47273216011301,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 18594
+    },
+    {
+      "epoch": 0.4426116347710178,
+      "grad_norm": 7.409913225349595,
+      "learning_rate": 2e-05,
+      "loss": 0.3602,
+      "step": 18595
+    },
+    {
+      "epoch": 0.4426354374940493,
+      "grad_norm": 3.72952260668377,
+      "learning_rate": 2e-05,
+      "loss": 0.6762,
+      "step": 18596
+    },
+    {
+      "epoch": 0.4426592402170808,
+      "grad_norm": 4.658671595332052,
+      "learning_rate": 2e-05,
+      "loss": 1.2274,
+      "step": 18597
+    },
+    {
+      "epoch": 0.44268304294011235,
+      "grad_norm": 2.8622181905120083,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 18598
+    },
+    {
+      "epoch": 0.4427068456631439,
+      "grad_norm": 1.1863327457197985,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 18599
+    },
+    {
+      "epoch": 0.4427306483861754,
+      "grad_norm": 7.090789692256607,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 18600
+    },
+    {
+      "epoch": 0.4427544511092069,
+      "grad_norm": 4.76109118194435,
+      "learning_rate": 2e-05,
+      "loss": 0.8314,
+      "step": 18601
+    },
+    {
+      "epoch": 0.4427782538322384,
+      "grad_norm": 5.520240540744596,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 18602
+    },
+    {
+      "epoch": 0.44280205655526994,
+      "grad_norm": 1.5694435530511641,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 18603
+    },
+    {
+      "epoch": 0.4428258592783014,
+      "grad_norm": 40.02785079539482,
+      "learning_rate": 2e-05,
+      "loss": 0.2551,
+      "step": 18604
+    },
+    {
+      "epoch": 0.44284966200133297,
+      "grad_norm": 4.233063146453813,
+      "learning_rate": 2e-05,
+      "loss": 0.5831,
+      "step": 18605
+    },
+    {
+      "epoch": 0.44287346472436445,
+      "grad_norm": 3.3021576895904823,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 18606
+    },
+    {
+      "epoch": 0.442897267447396,
+      "grad_norm": 2.455387605767527,
+      "learning_rate": 2e-05,
+      "loss": 0.3289,
+      "step": 18607
+    },
+    {
+      "epoch": 0.4429210701704275,
+      "grad_norm": 1.4988794422894034,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 18608
+    },
+    {
+      "epoch": 0.442944872893459,
+      "grad_norm": 3.4847548946655293,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 18609
+    },
+    {
+      "epoch": 0.4429686756164905,
+      "grad_norm": 4.846928747824662,
+      "learning_rate": 2e-05,
+      "loss": 0.728,
+      "step": 18610
+    },
+    {
+      "epoch": 0.44299247833952204,
+      "grad_norm": 1.955980574785703,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 18611
+    },
+    {
+      "epoch": 0.44301628106255353,
+      "grad_norm": 3.7855877595852974,
+      "learning_rate": 2e-05,
+      "loss": 0.1317,
+      "step": 18612
+    },
+    {
+      "epoch": 0.44304008378558507,
+      "grad_norm": 6.404051731817114,
+      "learning_rate": 2e-05,
+      "loss": 0.3966,
+      "step": 18613
+    },
+    {
+      "epoch": 0.4430638865086166,
+      "grad_norm": 3.1932871924156343,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 18614
+    },
+    {
+      "epoch": 0.4430876892316481,
+      "grad_norm": 2.767117396502253,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 18615
+    },
+    {
+      "epoch": 0.44311149195467964,
+      "grad_norm": 1.3945988580316817,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 18616
+    },
+    {
+      "epoch": 0.4431352946777111,
+      "grad_norm": 1.0341834028425474,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 18617
+    },
+    {
+      "epoch": 0.44315909740074266,
+      "grad_norm": 3.0936491511174453,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 18618
+    },
+    {
+      "epoch": 0.44318290012377415,
+      "grad_norm": 2.0493712630576773,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 18619
+    },
+    {
+      "epoch": 0.4432067028468057,
+      "grad_norm": 12.603769461772885,
+      "learning_rate": 2e-05,
+      "loss": 0.6666,
+      "step": 18620
+    },
+    {
+      "epoch": 0.4432305055698372,
+      "grad_norm": 4.573536364354547,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 18621
+    },
+    {
+      "epoch": 0.4432543082928687,
+      "grad_norm": 1.6531883376501109,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 18622
+    },
+    {
+      "epoch": 0.4432781110159002,
+      "grad_norm": 1.270169083376387,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 18623
+    },
+    {
+      "epoch": 0.44330191373893174,
+      "grad_norm": 4.58100556696618,
+      "learning_rate": 2e-05,
+      "loss": 0.2416,
+      "step": 18624
+    },
+    {
+      "epoch": 0.4433257164619632,
+      "grad_norm": 2.2853082757110754,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 18625
+    },
+    {
+      "epoch": 0.44334951918499477,
+      "grad_norm": 5.85488106499534,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 18626
+    },
+    {
+      "epoch": 0.44337332190802625,
+      "grad_norm": 2.7570386168082455,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 18627
+    },
+    {
+      "epoch": 0.4433971246310578,
+      "grad_norm": 8.327920555713478,
+      "learning_rate": 2e-05,
+      "loss": 2.4053,
+      "step": 18628
+    },
+    {
+      "epoch": 0.44342092735408933,
+      "grad_norm": 1.4058234102076128,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 18629
+    },
+    {
+      "epoch": 0.4434447300771208,
+      "grad_norm": 1.120800186923522,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 18630
+    },
+    {
+      "epoch": 0.44346853280015236,
+      "grad_norm": 4.5304503747302896,
+      "learning_rate": 2e-05,
+      "loss": 0.6985,
+      "step": 18631
+    },
+    {
+      "epoch": 0.44349233552318384,
+      "grad_norm": 34.24198417929472,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 18632
+    },
+    {
+      "epoch": 0.4435161382462154,
+      "grad_norm": 9.240605875249791,
+      "learning_rate": 2e-05,
+      "loss": 0.7978,
+      "step": 18633
+    },
+    {
+      "epoch": 0.44353994096924687,
+      "grad_norm": 5.731161597542123,
+      "learning_rate": 2e-05,
+      "loss": 0.8647,
+      "step": 18634
+    },
+    {
+      "epoch": 0.4435637436922784,
+      "grad_norm": 0.7911865005413687,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 18635
+    },
+    {
+      "epoch": 0.4435875464153099,
+      "grad_norm": 1.7303339491080605,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 18636
+    },
+    {
+      "epoch": 0.44361134913834144,
+      "grad_norm": 2.0179101823102332,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 18637
+    },
+    {
+      "epoch": 0.4436351518613729,
+      "grad_norm": 4.741203341776375,
+      "learning_rate": 2e-05,
+      "loss": 0.3343,
+      "step": 18638
+    },
+    {
+      "epoch": 0.44365895458440446,
+      "grad_norm": 1.4861636006073957,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 18639
+    },
+    {
+      "epoch": 0.44368275730743595,
+      "grad_norm": 1.4817956751223438,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 18640
+    },
+    {
+      "epoch": 0.4437065600304675,
+      "grad_norm": 4.7562634956016545,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 18641
+    },
+    {
+      "epoch": 0.443730362753499,
+      "grad_norm": 0.24356182483027228,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 18642
+    },
+    {
+      "epoch": 0.4437541654765305,
+      "grad_norm": 5.635669310019523,
+      "learning_rate": 2e-05,
+      "loss": 1.0307,
+      "step": 18643
+    },
+    {
+      "epoch": 0.44377796819956206,
+      "grad_norm": 6.529762916134835,
+      "learning_rate": 2e-05,
+      "loss": 0.7206,
+      "step": 18644
+    },
+    {
+      "epoch": 0.44380177092259354,
+      "grad_norm": 0.16415569543142172,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18645
+    },
+    {
+      "epoch": 0.4438255736456251,
+      "grad_norm": 5.987327113112228,
+      "learning_rate": 2e-05,
+      "loss": 1.1812,
+      "step": 18646
+    },
+    {
+      "epoch": 0.44384937636865657,
+      "grad_norm": 4.385491485092918,
+      "learning_rate": 2e-05,
+      "loss": 0.6342,
+      "step": 18647
+    },
+    {
+      "epoch": 0.4438731790916881,
+      "grad_norm": 4.181689648099443,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 18648
+    },
+    {
+      "epoch": 0.4438969818147196,
+      "grad_norm": 1.2830220818255076,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 18649
+    },
+    {
+      "epoch": 0.44392078453775113,
+      "grad_norm": 2.580142144860509,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 18650
+    },
+    {
+      "epoch": 0.4439445872607826,
+      "grad_norm": 4.9027064471021875,
+      "learning_rate": 2e-05,
+      "loss": 0.6487,
+      "step": 18651
+    },
+    {
+      "epoch": 0.44396838998381416,
+      "grad_norm": 4.028186720488581,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 18652
+    },
+    {
+      "epoch": 0.44399219270684565,
+      "grad_norm": 4.229943365499638,
+      "learning_rate": 2e-05,
+      "loss": 0.1206,
+      "step": 18653
+    },
+    {
+      "epoch": 0.4440159954298772,
+      "grad_norm": 1.6568395453426141,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 18654
+    },
+    {
+      "epoch": 0.44403979815290867,
+      "grad_norm": 1.104269905627174,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 18655
+    },
+    {
+      "epoch": 0.4440636008759402,
+      "grad_norm": 2.210095032591558,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 18656
+    },
+    {
+      "epoch": 0.4440874035989717,
+      "grad_norm": 6.7641315863317715,
+      "learning_rate": 2e-05,
+      "loss": 1.3169,
+      "step": 18657
+    },
+    {
+      "epoch": 0.44411120632200324,
+      "grad_norm": 0.1353242508905266,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18658
+    },
+    {
+      "epoch": 0.4441350090450348,
+      "grad_norm": 1.6040820896125036,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 18659
+    },
+    {
+      "epoch": 0.44415881176806626,
+      "grad_norm": 4.365395714722129,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 18660
+    },
+    {
+      "epoch": 0.4441826144910978,
+      "grad_norm": 1.7234383481218356,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 18661
+    },
+    {
+      "epoch": 0.4442064172141293,
+      "grad_norm": 0.24245528139340952,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 18662
+    },
+    {
+      "epoch": 0.44423021993716083,
+      "grad_norm": 11.187812592309381,
+      "learning_rate": 2e-05,
+      "loss": 1.0968,
+      "step": 18663
+    },
+    {
+      "epoch": 0.4442540226601923,
+      "grad_norm": 3.858700689802551,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 18664
+    },
+    {
+      "epoch": 0.44427782538322386,
+      "grad_norm": 20.92227806911634,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 18665
+    },
+    {
+      "epoch": 0.44430162810625534,
+      "grad_norm": 4.078226645573171,
+      "learning_rate": 2e-05,
+      "loss": 0.2012,
+      "step": 18666
+    },
+    {
+      "epoch": 0.4443254308292869,
+      "grad_norm": 11.03793161810996,
+      "learning_rate": 2e-05,
+      "loss": 0.8774,
+      "step": 18667
+    },
+    {
+      "epoch": 0.44434923355231837,
+      "grad_norm": 0.8217591473649062,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 18668
+    },
+    {
+      "epoch": 0.4443730362753499,
+      "grad_norm": 3.3095385249187417,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 18669
+    },
+    {
+      "epoch": 0.4443968389983814,
+      "grad_norm": 5.363198480698844,
+      "learning_rate": 2e-05,
+      "loss": 0.5165,
+      "step": 18670
+    },
+    {
+      "epoch": 0.44442064172141293,
+      "grad_norm": 0.5637441615731622,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 18671
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 4.1018361829835595,
+      "learning_rate": 2e-05,
+      "loss": 0.6632,
+      "step": 18672
+    },
+    {
+      "epoch": 0.44446824716747596,
+      "grad_norm": 5.62780537812293,
+      "learning_rate": 2e-05,
+      "loss": 1.4577,
+      "step": 18673
+    },
+    {
+      "epoch": 0.4444920498905075,
+      "grad_norm": 3.4069715552229125,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 18674
+    },
+    {
+      "epoch": 0.444515852613539,
+      "grad_norm": 1.843633219432497,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 18675
+    },
+    {
+      "epoch": 0.4445396553365705,
+      "grad_norm": 0.7868758698823758,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 18676
+    },
+    {
+      "epoch": 0.444563458059602,
+      "grad_norm": 3.8301128198484995,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 18677
+    },
+    {
+      "epoch": 0.44458726078263355,
+      "grad_norm": 7.6476257975584625,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 18678
+    },
+    {
+      "epoch": 0.44461106350566504,
+      "grad_norm": 3.9877876514399784,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 18679
+    },
+    {
+      "epoch": 0.4446348662286966,
+      "grad_norm": 5.680802911218393,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 18680
+    },
+    {
+      "epoch": 0.44465866895172806,
+      "grad_norm": 3.467885272621997,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 18681
+    },
+    {
+      "epoch": 0.4446824716747596,
+      "grad_norm": 5.881520409790807,
+      "learning_rate": 2e-05,
+      "loss": 0.9524,
+      "step": 18682
+    },
+    {
+      "epoch": 0.4447062743977911,
+      "grad_norm": 3.8654688732157716,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 18683
+    },
+    {
+      "epoch": 0.44473007712082263,
+      "grad_norm": 14.788641076427238,
+      "learning_rate": 2e-05,
+      "loss": 0.9028,
+      "step": 18684
+    },
+    {
+      "epoch": 0.4447538798438541,
+      "grad_norm": 0.06655722684684208,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 18685
+    },
+    {
+      "epoch": 0.44477768256688566,
+      "grad_norm": 2.7307276649424654,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 18686
+    },
+    {
+      "epoch": 0.44480148528991714,
+      "grad_norm": 2.2684891986107423,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 18687
+    },
+    {
+      "epoch": 0.4448252880129487,
+      "grad_norm": 0.36486286320818534,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 18688
+    },
+    {
+      "epoch": 0.4448490907359802,
+      "grad_norm": 3.2234275314714975,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 18689
+    },
+    {
+      "epoch": 0.4448728934590117,
+      "grad_norm": 0.5340134055446576,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 18690
+    },
+    {
+      "epoch": 0.44489669618204325,
+      "grad_norm": 2.9889262150188443,
+      "learning_rate": 2e-05,
+      "loss": 0.3696,
+      "step": 18691
+    },
+    {
+      "epoch": 0.44492049890507473,
+      "grad_norm": 5.6753638600436505,
+      "learning_rate": 2e-05,
+      "loss": 1.1432,
+      "step": 18692
+    },
+    {
+      "epoch": 0.4449443016281063,
+      "grad_norm": 1.16898753498992,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 18693
+    },
+    {
+      "epoch": 0.44496810435113776,
+      "grad_norm": 2.8050573734099022,
+      "learning_rate": 2e-05,
+      "loss": 0.2924,
+      "step": 18694
+    },
+    {
+      "epoch": 0.4449919070741693,
+      "grad_norm": 3.198215439328567,
+      "learning_rate": 2e-05,
+      "loss": 0.1389,
+      "step": 18695
+    },
+    {
+      "epoch": 0.4450157097972008,
+      "grad_norm": 0.4425736008907504,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 18696
+    },
+    {
+      "epoch": 0.4450395125202323,
+      "grad_norm": 3.1480182111260913,
+      "learning_rate": 2e-05,
+      "loss": 0.34,
+      "step": 18697
+    },
+    {
+      "epoch": 0.4450633152432638,
+      "grad_norm": 0.6209418775889758,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 18698
+    },
+    {
+      "epoch": 0.44508711796629535,
+      "grad_norm": 0.23602712841550022,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 18699
+    },
+    {
+      "epoch": 0.44511092068932684,
+      "grad_norm": 0.9098633308792569,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 18700
+    },
+    {
+      "epoch": 0.4451347234123584,
+      "grad_norm": 0.7759954351191736,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 18701
+    },
+    {
+      "epoch": 0.44515852613538986,
+      "grad_norm": 18.923874388969985,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 18702
+    },
+    {
+      "epoch": 0.4451823288584214,
+      "grad_norm": 4.108014832547818,
+      "learning_rate": 2e-05,
+      "loss": 0.8417,
+      "step": 18703
+    },
+    {
+      "epoch": 0.4452061315814529,
+      "grad_norm": 3.318740402087038,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 18704
+    },
+    {
+      "epoch": 0.44522993430448443,
+      "grad_norm": 19.740036448790832,
+      "learning_rate": 2e-05,
+      "loss": 0.522,
+      "step": 18705
+    },
+    {
+      "epoch": 0.445253737027516,
+      "grad_norm": 12.683295516163437,
+      "learning_rate": 2e-05,
+      "loss": 0.9998,
+      "step": 18706
+    },
+    {
+      "epoch": 0.44527753975054746,
+      "grad_norm": 12.179385076146032,
+      "learning_rate": 2e-05,
+      "loss": 1.0099,
+      "step": 18707
+    },
+    {
+      "epoch": 0.445301342473579,
+      "grad_norm": 5.985091997620796,
+      "learning_rate": 2e-05,
+      "loss": 1.0913,
+      "step": 18708
+    },
+    {
+      "epoch": 0.4453251451966105,
+      "grad_norm": 10.487910083592643,
+      "learning_rate": 2e-05,
+      "loss": 0.2782,
+      "step": 18709
+    },
+    {
+      "epoch": 0.445348947919642,
+      "grad_norm": 7.181867592146705,
+      "learning_rate": 2e-05,
+      "loss": 1.2007,
+      "step": 18710
+    },
+    {
+      "epoch": 0.4453727506426735,
+      "grad_norm": 7.9634282666972664,
+      "learning_rate": 2e-05,
+      "loss": 0.7238,
+      "step": 18711
+    },
+    {
+      "epoch": 0.44539655336570505,
+      "grad_norm": 2.895344849989944,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 18712
+    },
+    {
+      "epoch": 0.44542035608873654,
+      "grad_norm": 0.45374784645054794,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18713
+    },
+    {
+      "epoch": 0.4454441588117681,
+      "grad_norm": 4.063514470509442,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 18714
+    },
+    {
+      "epoch": 0.44546796153479956,
+      "grad_norm": 1.9607541115407663,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 18715
+    },
+    {
+      "epoch": 0.4454917642578311,
+      "grad_norm": 5.393471008553247,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 18716
+    },
+    {
+      "epoch": 0.4455155669808626,
+      "grad_norm": 10.126531169381675,
+      "learning_rate": 2e-05,
+      "loss": 1.3518,
+      "step": 18717
+    },
+    {
+      "epoch": 0.44553936970389413,
+      "grad_norm": 3.206353641926005,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 18718
+    },
+    {
+      "epoch": 0.4455631724269256,
+      "grad_norm": 3.379053948517329,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 18719
+    },
+    {
+      "epoch": 0.44558697514995715,
+      "grad_norm": 3.1273829546415746,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 18720
+    },
+    {
+      "epoch": 0.4456107778729887,
+      "grad_norm": 6.7476586953165585,
+      "learning_rate": 2e-05,
+      "loss": 1.1247,
+      "step": 18721
+    },
+    {
+      "epoch": 0.4456345805960202,
+      "grad_norm": 9.01541076786346,
+      "learning_rate": 2e-05,
+      "loss": 0.7295,
+      "step": 18722
+    },
+    {
+      "epoch": 0.4456583833190517,
+      "grad_norm": 2.2775698924439283,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 18723
+    },
+    {
+      "epoch": 0.4456821860420832,
+      "grad_norm": 2.2179557075363214,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 18724
+    },
+    {
+      "epoch": 0.44570598876511475,
+      "grad_norm": 3.486347966483959,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 18725
+    },
+    {
+      "epoch": 0.44572979148814623,
+      "grad_norm": 3.470479338307053,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 18726
+    },
+    {
+      "epoch": 0.4457535942111778,
+      "grad_norm": 2.7595352895965704,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 18727
+    },
+    {
+      "epoch": 0.44577739693420926,
+      "grad_norm": 1.274354134073564,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 18728
+    },
+    {
+      "epoch": 0.4458011996572408,
+      "grad_norm": 3.500134110605197,
+      "learning_rate": 2e-05,
+      "loss": 0.3833,
+      "step": 18729
+    },
+    {
+      "epoch": 0.4458250023802723,
+      "grad_norm": 1.3679039703698115,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 18730
+    },
+    {
+      "epoch": 0.4458488051033038,
+      "grad_norm": 5.595591475452355,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 18731
+    },
+    {
+      "epoch": 0.4458726078263353,
+      "grad_norm": 4.108772466081032,
+      "learning_rate": 2e-05,
+      "loss": 0.7382,
+      "step": 18732
+    },
+    {
+      "epoch": 0.44589641054936685,
+      "grad_norm": 6.400119065917238,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 18733
+    },
+    {
+      "epoch": 0.44592021327239834,
+      "grad_norm": 5.795413646808938,
+      "learning_rate": 2e-05,
+      "loss": 0.8521,
+      "step": 18734
+    },
+    {
+      "epoch": 0.4459440159954299,
+      "grad_norm": 1.485091242347045,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 18735
+    },
+    {
+      "epoch": 0.4459678187184614,
+      "grad_norm": 4.189593698650517,
+      "learning_rate": 2e-05,
+      "loss": 0.6329,
+      "step": 18736
+    },
+    {
+      "epoch": 0.4459916214414929,
+      "grad_norm": 0.35877023844380324,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 18737
+    },
+    {
+      "epoch": 0.44601542416452444,
+      "grad_norm": 1.9013473495772486,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 18738
+    },
+    {
+      "epoch": 0.44603922688755593,
+      "grad_norm": 14.91611118615792,
+      "learning_rate": 2e-05,
+      "loss": 0.9475,
+      "step": 18739
+    },
+    {
+      "epoch": 0.44606302961058747,
+      "grad_norm": 1.8038970058797765,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 18740
+    },
+    {
+      "epoch": 0.44608683233361895,
+      "grad_norm": 6.017881844290871,
+      "learning_rate": 2e-05,
+      "loss": 0.8122,
+      "step": 18741
+    },
+    {
+      "epoch": 0.4461106350566505,
+      "grad_norm": 1.113834824244355,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 18742
+    },
+    {
+      "epoch": 0.446134437779682,
+      "grad_norm": 0.2468886869713082,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 18743
+    },
+    {
+      "epoch": 0.4461582405027135,
+      "grad_norm": 5.761741914546964,
+      "learning_rate": 2e-05,
+      "loss": 0.855,
+      "step": 18744
+    },
+    {
+      "epoch": 0.446182043225745,
+      "grad_norm": 0.6773946143059171,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 18745
+    },
+    {
+      "epoch": 0.44620584594877655,
+      "grad_norm": 0.6087850226420981,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 18746
+    },
+    {
+      "epoch": 0.44622964867180803,
+      "grad_norm": 1.2694719573773887,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 18747
+    },
+    {
+      "epoch": 0.4462534513948396,
+      "grad_norm": 1.9844696209875474,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 18748
+    },
+    {
+      "epoch": 0.44627725411787106,
+      "grad_norm": 0.6523311330363881,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 18749
+    },
+    {
+      "epoch": 0.4463010568409026,
+      "grad_norm": 8.381680170927169,
+      "learning_rate": 2e-05,
+      "loss": 1.339,
+      "step": 18750
+    },
+    {
+      "epoch": 0.44632485956393414,
+      "grad_norm": 2.915937963875025,
+      "learning_rate": 2e-05,
+      "loss": 0.2482,
+      "step": 18751
+    },
+    {
+      "epoch": 0.4463486622869656,
+      "grad_norm": 0.8359824422523449,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 18752
+    },
+    {
+      "epoch": 0.44637246500999717,
+      "grad_norm": 9.099013895119102,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 18753
+    },
+    {
+      "epoch": 0.44639626773302865,
+      "grad_norm": 13.361729446092632,
+      "learning_rate": 2e-05,
+      "loss": 0.8758,
+      "step": 18754
+    },
+    {
+      "epoch": 0.4464200704560602,
+      "grad_norm": 5.562772202770246,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 18755
+    },
+    {
+      "epoch": 0.4464438731790917,
+      "grad_norm": 7.5786800806014645,
+      "learning_rate": 2e-05,
+      "loss": 1.4588,
+      "step": 18756
+    },
+    {
+      "epoch": 0.4464676759021232,
+      "grad_norm": 6.6522425396017875,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 18757
+    },
+    {
+      "epoch": 0.4464914786251547,
+      "grad_norm": 4.41954327761672,
+      "learning_rate": 2e-05,
+      "loss": 1.0128,
+      "step": 18758
+    },
+    {
+      "epoch": 0.44651528134818624,
+      "grad_norm": 150.2419177264473,
+      "learning_rate": 2e-05,
+      "loss": 0.7121,
+      "step": 18759
+    },
+    {
+      "epoch": 0.44653908407121773,
+      "grad_norm": 2.6672537688055233,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 18760
+    },
+    {
+      "epoch": 0.44656288679424927,
+      "grad_norm": 6.881814872769558,
+      "learning_rate": 2e-05,
+      "loss": 1.332,
+      "step": 18761
+    },
+    {
+      "epoch": 0.44658668951728075,
+      "grad_norm": 5.52693583703303,
+      "learning_rate": 2e-05,
+      "loss": 0.3069,
+      "step": 18762
+    },
+    {
+      "epoch": 0.4466104922403123,
+      "grad_norm": 29.458731329819727,
+      "learning_rate": 2e-05,
+      "loss": 1.2779,
+      "step": 18763
+    },
+    {
+      "epoch": 0.4466342949633438,
+      "grad_norm": 2.548466744854634,
+      "learning_rate": 2e-05,
+      "loss": 0.135,
+      "step": 18764
+    },
+    {
+      "epoch": 0.4466580976863753,
+      "grad_norm": 3.733386903632583,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 18765
+    },
+    {
+      "epoch": 0.44668190040940686,
+      "grad_norm": 0.13325538122390923,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 18766
+    },
+    {
+      "epoch": 0.44670570313243835,
+      "grad_norm": 0.17010943885290875,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 18767
+    },
+    {
+      "epoch": 0.4467295058554699,
+      "grad_norm": 2.631904861218798,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 18768
+    },
+    {
+      "epoch": 0.4467533085785014,
+      "grad_norm": 2.772620006175233,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 18769
+    },
+    {
+      "epoch": 0.4467771113015329,
+      "grad_norm": 2.094455183342796,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 18770
+    },
+    {
+      "epoch": 0.4468009140245644,
+      "grad_norm": 2.646486225041595,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 18771
+    },
+    {
+      "epoch": 0.44682471674759594,
+      "grad_norm": 4.917680450740152,
+      "learning_rate": 2e-05,
+      "loss": 0.5883,
+      "step": 18772
+    },
+    {
+      "epoch": 0.4468485194706274,
+      "grad_norm": 1.5714116132414961,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 18773
+    },
+    {
+      "epoch": 0.44687232219365897,
+      "grad_norm": 7.920845658720652,
+      "learning_rate": 2e-05,
+      "loss": 0.643,
+      "step": 18774
+    },
+    {
+      "epoch": 0.44689612491669045,
+      "grad_norm": 0.4806196454423521,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 18775
+    },
+    {
+      "epoch": 0.446919927639722,
+      "grad_norm": 0.46119430060432426,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 18776
+    },
+    {
+      "epoch": 0.4469437303627535,
+      "grad_norm": 2.2559022234808976,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 18777
+    },
+    {
+      "epoch": 0.446967533085785,
+      "grad_norm": 6.741338892807987,
+      "learning_rate": 2e-05,
+      "loss": 1.4646,
+      "step": 18778
+    },
+    {
+      "epoch": 0.4469913358088165,
+      "grad_norm": 1.051492884029132,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 18779
+    },
+    {
+      "epoch": 0.44701513853184804,
+      "grad_norm": 5.7070262677829975,
+      "learning_rate": 2e-05,
+      "loss": 0.245,
+      "step": 18780
+    },
+    {
+      "epoch": 0.4470389412548796,
+      "grad_norm": 2.71758828282661,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 18781
+    },
+    {
+      "epoch": 0.44706274397791107,
+      "grad_norm": 5.103822120858014,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 18782
+    },
+    {
+      "epoch": 0.4470865467009426,
+      "grad_norm": 3.8661950649634123,
+      "learning_rate": 2e-05,
+      "loss": 0.3389,
+      "step": 18783
+    },
+    {
+      "epoch": 0.4471103494239741,
+      "grad_norm": 1.4041077895039455,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 18784
+    },
+    {
+      "epoch": 0.44713415214700564,
+      "grad_norm": 5.054281609929631,
+      "learning_rate": 2e-05,
+      "loss": 0.4802,
+      "step": 18785
+    },
+    {
+      "epoch": 0.4471579548700371,
+      "grad_norm": 4.2739179001810195,
+      "learning_rate": 2e-05,
+      "loss": 0.6581,
+      "step": 18786
+    },
+    {
+      "epoch": 0.44718175759306866,
+      "grad_norm": 3.0886599017091814,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 18787
+    },
+    {
+      "epoch": 0.44720556031610015,
+      "grad_norm": 3.133829209678493,
+      "learning_rate": 2e-05,
+      "loss": 0.7966,
+      "step": 18788
+    },
+    {
+      "epoch": 0.4472293630391317,
+      "grad_norm": 2.2349882593169097,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 18789
+    },
+    {
+      "epoch": 0.4472531657621632,
+      "grad_norm": 4.914010333510087,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 18790
+    },
+    {
+      "epoch": 0.4472769684851947,
+      "grad_norm": 5.229548475914962,
+      "learning_rate": 2e-05,
+      "loss": 0.5562,
+      "step": 18791
+    },
+    {
+      "epoch": 0.4473007712082262,
+      "grad_norm": 9.163843084559822,
+      "learning_rate": 2e-05,
+      "loss": 0.9576,
+      "step": 18792
+    },
+    {
+      "epoch": 0.44732457393125774,
+      "grad_norm": 10.984148594844939,
+      "learning_rate": 2e-05,
+      "loss": 0.558,
+      "step": 18793
+    },
+    {
+      "epoch": 0.4473483766542892,
+      "grad_norm": 4.220274952308726,
+      "learning_rate": 2e-05,
+      "loss": 0.6358,
+      "step": 18794
+    },
+    {
+      "epoch": 0.44737217937732077,
+      "grad_norm": 11.277053632827977,
+      "learning_rate": 2e-05,
+      "loss": 0.3352,
+      "step": 18795
+    },
+    {
+      "epoch": 0.4473959821003523,
+      "grad_norm": 4.556998119318707,
+      "learning_rate": 2e-05,
+      "loss": 0.7397,
+      "step": 18796
+    },
+    {
+      "epoch": 0.4474197848233838,
+      "grad_norm": 11.108005795596435,
+      "learning_rate": 2e-05,
+      "loss": 0.3798,
+      "step": 18797
+    },
+    {
+      "epoch": 0.44744358754641533,
+      "grad_norm": 3.49343070351695,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 18798
+    },
+    {
+      "epoch": 0.4474673902694468,
+      "grad_norm": 5.4129092965371335,
+      "learning_rate": 2e-05,
+      "loss": 0.9662,
+      "step": 18799
+    },
+    {
+      "epoch": 0.44749119299247836,
+      "grad_norm": 1.9000852431172819,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 18800
+    },
+    {
+      "epoch": 0.44751499571550984,
+      "grad_norm": 1.3241241523482077,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 18801
+    },
+    {
+      "epoch": 0.4475387984385414,
+      "grad_norm": 3.3880489151859656,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 18802
+    },
+    {
+      "epoch": 0.44756260116157287,
+      "grad_norm": 4.617352233581598,
+      "learning_rate": 2e-05,
+      "loss": 0.7323,
+      "step": 18803
+    },
+    {
+      "epoch": 0.4475864038846044,
+      "grad_norm": 3.7248305362354497,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 18804
+    },
+    {
+      "epoch": 0.4476102066076359,
+      "grad_norm": 10.834888962593585,
+      "learning_rate": 2e-05,
+      "loss": 0.4224,
+      "step": 18805
+    },
+    {
+      "epoch": 0.44763400933066744,
+      "grad_norm": 5.7736622944375675,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 18806
+    },
+    {
+      "epoch": 0.4476578120536989,
+      "grad_norm": 5.3215308000792,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 18807
+    },
+    {
+      "epoch": 0.44768161477673046,
+      "grad_norm": 3.687169022784363,
+      "learning_rate": 2e-05,
+      "loss": 0.5902,
+      "step": 18808
+    },
+    {
+      "epoch": 0.44770541749976195,
+      "grad_norm": 2.061403355206586,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 18809
+    },
+    {
+      "epoch": 0.4477292202227935,
+      "grad_norm": 2.165637567542724,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 18810
+    },
+    {
+      "epoch": 0.44775302294582503,
+      "grad_norm": 1.2625469751788059,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 18811
+    },
+    {
+      "epoch": 0.4477768256688565,
+      "grad_norm": 2.508808027927869,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 18812
+    },
+    {
+      "epoch": 0.44780062839188806,
+      "grad_norm": 2.3148106508430057,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 18813
+    },
+    {
+      "epoch": 0.44782443111491954,
+      "grad_norm": 1.093808210706335,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 18814
+    },
+    {
+      "epoch": 0.4478482338379511,
+      "grad_norm": 6.19637362775579,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 18815
+    },
+    {
+      "epoch": 0.44787203656098257,
+      "grad_norm": 6.107746927095236,
+      "learning_rate": 2e-05,
+      "loss": 1.0626,
+      "step": 18816
+    },
+    {
+      "epoch": 0.4478958392840141,
+      "grad_norm": 3.7864758476789424,
+      "learning_rate": 2e-05,
+      "loss": 0.8413,
+      "step": 18817
+    },
+    {
+      "epoch": 0.4479196420070456,
+      "grad_norm": 1.9777394478868549,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 18818
+    },
+    {
+      "epoch": 0.44794344473007713,
+      "grad_norm": 2.3725779586627738,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 18819
+    },
+    {
+      "epoch": 0.4479672474531086,
+      "grad_norm": 3.342135562612363,
+      "learning_rate": 2e-05,
+      "loss": 0.1507,
+      "step": 18820
+    },
+    {
+      "epoch": 0.44799105017614016,
+      "grad_norm": 10.590311238635861,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 18821
+    },
+    {
+      "epoch": 0.44801485289917165,
+      "grad_norm": 4.189914924374015,
+      "learning_rate": 2e-05,
+      "loss": 0.7812,
+      "step": 18822
+    },
+    {
+      "epoch": 0.4480386556222032,
+      "grad_norm": 2.6755622465692444,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 18823
+    },
+    {
+      "epoch": 0.44806245834523467,
+      "grad_norm": 4.963666068846019,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 18824
+    },
+    {
+      "epoch": 0.4480862610682662,
+      "grad_norm": 1.3416125520211863,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 18825
+    },
+    {
+      "epoch": 0.4481100637912977,
+      "grad_norm": 2.9973811365068737,
+      "learning_rate": 2e-05,
+      "loss": 0.5244,
+      "step": 18826
+    },
+    {
+      "epoch": 0.44813386651432924,
+      "grad_norm": 3.071330168601088,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 18827
+    },
+    {
+      "epoch": 0.4481576692373608,
+      "grad_norm": 0.22656220796012275,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18828
+    },
+    {
+      "epoch": 0.44818147196039226,
+      "grad_norm": 4.881211513948796,
+      "learning_rate": 2e-05,
+      "loss": 0.5866,
+      "step": 18829
+    },
+    {
+      "epoch": 0.4482052746834238,
+      "grad_norm": 6.0793266911568145,
+      "learning_rate": 2e-05,
+      "loss": 1.2664,
+      "step": 18830
+    },
+    {
+      "epoch": 0.4482290774064553,
+      "grad_norm": 2.4192547218824068,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 18831
+    },
+    {
+      "epoch": 0.44825288012948683,
+      "grad_norm": 5.008233029511169,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 18832
+    },
+    {
+      "epoch": 0.4482766828525183,
+      "grad_norm": 5.526489505117666,
+      "learning_rate": 2e-05,
+      "loss": 0.264,
+      "step": 18833
+    },
+    {
+      "epoch": 0.44830048557554986,
+      "grad_norm": 4.38620180850499,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 18834
+    },
+    {
+      "epoch": 0.44832428829858134,
+      "grad_norm": 0.6420141665019568,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 18835
+    },
+    {
+      "epoch": 0.4483480910216129,
+      "grad_norm": 5.0511155301225,
+      "learning_rate": 2e-05,
+      "loss": 0.6097,
+      "step": 18836
+    },
+    {
+      "epoch": 0.44837189374464437,
+      "grad_norm": 1.6215876896511117,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 18837
+    },
+    {
+      "epoch": 0.4483956964676759,
+      "grad_norm": 2.3200947348705596,
+      "learning_rate": 2e-05,
+      "loss": 0.4237,
+      "step": 18838
+    },
+    {
+      "epoch": 0.4484194991907074,
+      "grad_norm": 3.853447380740311,
+      "learning_rate": 2e-05,
+      "loss": 0.5503,
+      "step": 18839
+    },
+    {
+      "epoch": 0.44844330191373893,
+      "grad_norm": 4.157020169215033,
+      "learning_rate": 2e-05,
+      "loss": 0.426,
+      "step": 18840
+    },
+    {
+      "epoch": 0.4484671046367704,
+      "grad_norm": 43.21477901336078,
+      "learning_rate": 2e-05,
+      "loss": 0.2175,
+      "step": 18841
+    },
+    {
+      "epoch": 0.44849090735980196,
+      "grad_norm": 4.062377575016849,
+      "learning_rate": 2e-05,
+      "loss": 0.1592,
+      "step": 18842
+    },
+    {
+      "epoch": 0.4485147100828335,
+      "grad_norm": 20.442202684784846,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 18843
+    },
+    {
+      "epoch": 0.448538512805865,
+      "grad_norm": 0.705315799969463,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 18844
+    },
+    {
+      "epoch": 0.4485623155288965,
+      "grad_norm": 4.547887223983268,
+      "learning_rate": 2e-05,
+      "loss": 0.8117,
+      "step": 18845
+    },
+    {
+      "epoch": 0.448586118251928,
+      "grad_norm": 1.4715636302458943,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 18846
+    },
+    {
+      "epoch": 0.44860992097495955,
+      "grad_norm": 4.3026320582351625,
+      "learning_rate": 2e-05,
+      "loss": 0.3948,
+      "step": 18847
+    },
+    {
+      "epoch": 0.44863372369799104,
+      "grad_norm": 3.1230454627672266,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 18848
+    },
+    {
+      "epoch": 0.4486575264210226,
+      "grad_norm": 2.4717439900736244,
+      "learning_rate": 2e-05,
+      "loss": 0.326,
+      "step": 18849
+    },
+    {
+      "epoch": 0.44868132914405406,
+      "grad_norm": 3.7267068414904405,
+      "learning_rate": 2e-05,
+      "loss": 0.2597,
+      "step": 18850
+    },
+    {
+      "epoch": 0.4487051318670856,
+      "grad_norm": 3.783848629313774,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 18851
+    },
+    {
+      "epoch": 0.4487289345901171,
+      "grad_norm": 1.5379165720283858,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 18852
+    },
+    {
+      "epoch": 0.44875273731314863,
+      "grad_norm": 14.843810767263856,
+      "learning_rate": 2e-05,
+      "loss": 0.2884,
+      "step": 18853
+    },
+    {
+      "epoch": 0.4487765400361801,
+      "grad_norm": 2.1573961330800184,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 18854
+    },
+    {
+      "epoch": 0.44880034275921166,
+      "grad_norm": 5.430564029978322,
+      "learning_rate": 2e-05,
+      "loss": 1.1021,
+      "step": 18855
+    },
+    {
+      "epoch": 0.44882414548224314,
+      "grad_norm": 5.108307136329895,
+      "learning_rate": 2e-05,
+      "loss": 0.1645,
+      "step": 18856
+    },
+    {
+      "epoch": 0.4488479482052747,
+      "grad_norm": 4.6358350699651485,
+      "learning_rate": 2e-05,
+      "loss": 0.8008,
+      "step": 18857
+    },
+    {
+      "epoch": 0.4488717509283062,
+      "grad_norm": 26.92159340038611,
+      "learning_rate": 2e-05,
+      "loss": 0.651,
+      "step": 18858
+    },
+    {
+      "epoch": 0.4488955536513377,
+      "grad_norm": 1.109210995687992,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 18859
+    },
+    {
+      "epoch": 0.44891935637436925,
+      "grad_norm": 4.7271018492637635,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 18860
+    },
+    {
+      "epoch": 0.44894315909740073,
+      "grad_norm": 2.8044117717504977,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 18861
+    },
+    {
+      "epoch": 0.4489669618204323,
+      "grad_norm": 4.825045389420072,
+      "learning_rate": 2e-05,
+      "loss": 0.3891,
+      "step": 18862
+    },
+    {
+      "epoch": 0.44899076454346376,
+      "grad_norm": 5.69288763770842,
+      "learning_rate": 2e-05,
+      "loss": 1.5215,
+      "step": 18863
+    },
+    {
+      "epoch": 0.4490145672664953,
+      "grad_norm": 0.13322422080909063,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 18864
+    },
+    {
+      "epoch": 0.4490383699895268,
+      "grad_norm": 10.762915040429103,
+      "learning_rate": 2e-05,
+      "loss": 0.774,
+      "step": 18865
+    },
+    {
+      "epoch": 0.4490621727125583,
+      "grad_norm": 4.4859458056263675,
+      "learning_rate": 2e-05,
+      "loss": 0.4353,
+      "step": 18866
+    },
+    {
+      "epoch": 0.4490859754355898,
+      "grad_norm": 4.0415090919057794,
+      "learning_rate": 2e-05,
+      "loss": 0.7779,
+      "step": 18867
+    },
+    {
+      "epoch": 0.44910977815862135,
+      "grad_norm": 3.1419994398636595,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 18868
+    },
+    {
+      "epoch": 0.44913358088165284,
+      "grad_norm": 0.34236200107360437,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 18869
+    },
+    {
+      "epoch": 0.4491573836046844,
+      "grad_norm": 0.382891415379538,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 18870
+    },
+    {
+      "epoch": 0.44918118632771586,
+      "grad_norm": 11.463552822022033,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 18871
+    },
+    {
+      "epoch": 0.4492049890507474,
+      "grad_norm": 2.422664475946427,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 18872
+    },
+    {
+      "epoch": 0.44922879177377895,
+      "grad_norm": 1.576506402822614,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 18873
+    },
+    {
+      "epoch": 0.44925259449681043,
+      "grad_norm": 1.6978454190454972,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 18874
+    },
+    {
+      "epoch": 0.44927639721984197,
+      "grad_norm": 1.386613262378602,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 18875
+    },
+    {
+      "epoch": 0.44930019994287346,
+      "grad_norm": 5.538296496695811,
+      "learning_rate": 2e-05,
+      "loss": 1.1534,
+      "step": 18876
+    },
+    {
+      "epoch": 0.449324002665905,
+      "grad_norm": 0.7155566001469023,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 18877
+    },
+    {
+      "epoch": 0.4493478053889365,
+      "grad_norm": 4.483085330636148,
+      "learning_rate": 2e-05,
+      "loss": 0.934,
+      "step": 18878
+    },
+    {
+      "epoch": 0.449371608111968,
+      "grad_norm": 6.320329516058422,
+      "learning_rate": 2e-05,
+      "loss": 0.8319,
+      "step": 18879
+    },
+    {
+      "epoch": 0.4493954108349995,
+      "grad_norm": 75.57552858968077,
+      "learning_rate": 2e-05,
+      "loss": 0.3377,
+      "step": 18880
+    },
+    {
+      "epoch": 0.44941921355803105,
+      "grad_norm": 1.4012519229053562,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 18881
+    },
+    {
+      "epoch": 0.44944301628106254,
+      "grad_norm": 7.090025974210654,
+      "learning_rate": 2e-05,
+      "loss": 0.9315,
+      "step": 18882
+    },
+    {
+      "epoch": 0.4494668190040941,
+      "grad_norm": 4.85840405557364,
+      "learning_rate": 2e-05,
+      "loss": 0.5425,
+      "step": 18883
+    },
+    {
+      "epoch": 0.44949062172712556,
+      "grad_norm": 3.1283375146219163,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 18884
+    },
+    {
+      "epoch": 0.4495144244501571,
+      "grad_norm": 3.816447011213083,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 18885
+    },
+    {
+      "epoch": 0.4495382271731886,
+      "grad_norm": 0.5225973610580772,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 18886
+    },
+    {
+      "epoch": 0.44956202989622013,
+      "grad_norm": 2.9094426885686655,
+      "learning_rate": 2e-05,
+      "loss": 0.5297,
+      "step": 18887
+    },
+    {
+      "epoch": 0.44958583261925167,
+      "grad_norm": 8.243006388152628,
+      "learning_rate": 2e-05,
+      "loss": 1.0394,
+      "step": 18888
+    },
+    {
+      "epoch": 0.44960963534228315,
+      "grad_norm": 9.98802607580292,
+      "learning_rate": 2e-05,
+      "loss": 0.2421,
+      "step": 18889
+    },
+    {
+      "epoch": 0.4496334380653147,
+      "grad_norm": 2.3636022604537197,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 18890
+    },
+    {
+      "epoch": 0.4496572407883462,
+      "grad_norm": 1.5683097996977746,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 18891
+    },
+    {
+      "epoch": 0.4496810435113777,
+      "grad_norm": 6.945922188002407,
+      "learning_rate": 2e-05,
+      "loss": 0.172,
+      "step": 18892
+    },
+    {
+      "epoch": 0.4497048462344092,
+      "grad_norm": 1.6961068037182077,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 18893
+    },
+    {
+      "epoch": 0.44972864895744075,
+      "grad_norm": 0.5299015822339663,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 18894
+    },
+    {
+      "epoch": 0.44975245168047223,
+      "grad_norm": 7.172962889533618,
+      "learning_rate": 2e-05,
+      "loss": 0.27,
+      "step": 18895
+    },
+    {
+      "epoch": 0.4497762544035038,
+      "grad_norm": 4.194578935247779,
+      "learning_rate": 2e-05,
+      "loss": 0.4918,
+      "step": 18896
+    },
+    {
+      "epoch": 0.44980005712653526,
+      "grad_norm": 2.0133569724756346,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 18897
+    },
+    {
+      "epoch": 0.4498238598495668,
+      "grad_norm": 3.8179883634123177,
+      "learning_rate": 2e-05,
+      "loss": 0.4756,
+      "step": 18898
+    },
+    {
+      "epoch": 0.4498476625725983,
+      "grad_norm": 2.589557294773411,
+      "learning_rate": 2e-05,
+      "loss": 0.397,
+      "step": 18899
+    },
+    {
+      "epoch": 0.4498714652956298,
+      "grad_norm": 3.9696522214231824,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 18900
+    },
+    {
+      "epoch": 0.4498952680186613,
+      "grad_norm": 0.2789136802219699,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 18901
+    },
+    {
+      "epoch": 0.44991907074169285,
+      "grad_norm": 0.41975352683243033,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 18902
+    },
+    {
+      "epoch": 0.4499428734647244,
+      "grad_norm": 1.6821275158055453,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 18903
+    },
+    {
+      "epoch": 0.4499666761877559,
+      "grad_norm": 1.5902364382170893,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 18904
+    },
+    {
+      "epoch": 0.4499904789107874,
+      "grad_norm": 8.452595539094766,
+      "learning_rate": 2e-05,
+      "loss": 0.6211,
+      "step": 18905
+    },
+    {
+      "epoch": 0.4500142816338189,
+      "grad_norm": 0.2743974147580579,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 18906
+    },
+    {
+      "epoch": 0.45003808435685044,
+      "grad_norm": 4.6372728820208975,
+      "learning_rate": 2e-05,
+      "loss": 0.6089,
+      "step": 18907
+    },
+    {
+      "epoch": 0.45006188707988193,
+      "grad_norm": 0.35565874002368486,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 18908
+    },
+    {
+      "epoch": 0.45008568980291347,
+      "grad_norm": 8.912594439948325,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 18909
+    },
+    {
+      "epoch": 0.45010949252594495,
+      "grad_norm": 1.8953722880342392,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 18910
+    },
+    {
+      "epoch": 0.4501332952489765,
+      "grad_norm": 0.15700301595736166,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 18911
+    },
+    {
+      "epoch": 0.450157097972008,
+      "grad_norm": 0.2090574694068531,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 18912
+    },
+    {
+      "epoch": 0.4501809006950395,
+      "grad_norm": 3.031011779384881,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 18913
+    },
+    {
+      "epoch": 0.450204703418071,
+      "grad_norm": 1.1309192602299774,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 18914
+    },
+    {
+      "epoch": 0.45022850614110255,
+      "grad_norm": 1.5837144395472236,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 18915
+    },
+    {
+      "epoch": 0.45025230886413403,
+      "grad_norm": 0.2441895773332798,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 18916
+    },
+    {
+      "epoch": 0.4502761115871656,
+      "grad_norm": 0.46599646129635436,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 18917
+    },
+    {
+      "epoch": 0.4502999143101971,
+      "grad_norm": 12.907407753352812,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 18918
+    },
+    {
+      "epoch": 0.4503237170332286,
+      "grad_norm": 4.52308640207805,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 18919
+    },
+    {
+      "epoch": 0.45034751975626014,
+      "grad_norm": 0.18099326476258762,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 18920
+    },
+    {
+      "epoch": 0.4503713224792916,
+      "grad_norm": 5.479686533176453,
+      "learning_rate": 2e-05,
+      "loss": 0.1934,
+      "step": 18921
+    },
+    {
+      "epoch": 0.45039512520232317,
+      "grad_norm": 5.3733925830636,
+      "learning_rate": 2e-05,
+      "loss": 1.2186,
+      "step": 18922
+    },
+    {
+      "epoch": 0.45041892792535465,
+      "grad_norm": 5.654687319633135,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 18923
+    },
+    {
+      "epoch": 0.4504427306483862,
+      "grad_norm": 1.028256738826058,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 18924
+    },
+    {
+      "epoch": 0.4504665333714177,
+      "grad_norm": 4.168981263955037,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 18925
+    },
+    {
+      "epoch": 0.4504903360944492,
+      "grad_norm": 5.16114116922653,
+      "learning_rate": 2e-05,
+      "loss": 0.3192,
+      "step": 18926
+    },
+    {
+      "epoch": 0.4505141388174807,
+      "grad_norm": 6.41434425674654,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 18927
+    },
+    {
+      "epoch": 0.45053794154051224,
+      "grad_norm": 8.232478065470621,
+      "learning_rate": 2e-05,
+      "loss": 0.9657,
+      "step": 18928
+    },
+    {
+      "epoch": 0.45056174426354373,
+      "grad_norm": 3.0802719521919215,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 18929
+    },
+    {
+      "epoch": 0.45058554698657527,
+      "grad_norm": 15.666188633030892,
+      "learning_rate": 2e-05,
+      "loss": 0.4952,
+      "step": 18930
+    },
+    {
+      "epoch": 0.45060934970960675,
+      "grad_norm": 9.82230664470522,
+      "learning_rate": 2e-05,
+      "loss": 0.4569,
+      "step": 18931
+    },
+    {
+      "epoch": 0.4506331524326383,
+      "grad_norm": 0.1491980400916366,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 18932
+    },
+    {
+      "epoch": 0.45065695515566984,
+      "grad_norm": 10.951320894364992,
+      "learning_rate": 2e-05,
+      "loss": 1.0176,
+      "step": 18933
+    },
+    {
+      "epoch": 0.4506807578787013,
+      "grad_norm": 3.5405346608244472,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 18934
+    },
+    {
+      "epoch": 0.45070456060173286,
+      "grad_norm": 0.6239698731046229,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 18935
+    },
+    {
+      "epoch": 0.45072836332476435,
+      "grad_norm": 5.269070638096991,
+      "learning_rate": 2e-05,
+      "loss": 0.6799,
+      "step": 18936
+    },
+    {
+      "epoch": 0.4507521660477959,
+      "grad_norm": 0.7537804774814285,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 18937
+    },
+    {
+      "epoch": 0.4507759687708274,
+      "grad_norm": 12.181356363480562,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 18938
+    },
+    {
+      "epoch": 0.4507997714938589,
+      "grad_norm": 16.511659177764972,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 18939
+    },
+    {
+      "epoch": 0.4508235742168904,
+      "grad_norm": 5.945050607437809,
+      "learning_rate": 2e-05,
+      "loss": 1.1468,
+      "step": 18940
+    },
+    {
+      "epoch": 0.45084737693992194,
+      "grad_norm": 5.811375807411044,
+      "learning_rate": 2e-05,
+      "loss": 0.2486,
+      "step": 18941
+    },
+    {
+      "epoch": 0.4508711796629534,
+      "grad_norm": 22.36711696372503,
+      "learning_rate": 2e-05,
+      "loss": 3.9311,
+      "step": 18942
+    },
+    {
+      "epoch": 0.45089498238598497,
+      "grad_norm": 0.30057170966274976,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 18943
+    },
+    {
+      "epoch": 0.45091878510901645,
+      "grad_norm": 0.3436897166415997,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 18944
+    },
+    {
+      "epoch": 0.450942587832048,
+      "grad_norm": 2.865765847048671,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 18945
+    },
+    {
+      "epoch": 0.4509663905550795,
+      "grad_norm": 7.354040083837534,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 18946
+    },
+    {
+      "epoch": 0.450990193278111,
+      "grad_norm": 22.23715859024679,
+      "learning_rate": 2e-05,
+      "loss": 0.9483,
+      "step": 18947
+    },
+    {
+      "epoch": 0.4510139960011425,
+      "grad_norm": 6.060539817436351,
+      "learning_rate": 2e-05,
+      "loss": 0.3332,
+      "step": 18948
+    },
+    {
+      "epoch": 0.45103779872417404,
+      "grad_norm": 3.0409947852793153,
+      "learning_rate": 2e-05,
+      "loss": 0.5536,
+      "step": 18949
+    },
+    {
+      "epoch": 0.4510616014472056,
+      "grad_norm": 9.286185845433925,
+      "learning_rate": 2e-05,
+      "loss": 2.3034,
+      "step": 18950
+    },
+    {
+      "epoch": 0.45108540417023707,
+      "grad_norm": 3.9973699371299216,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 18951
+    },
+    {
+      "epoch": 0.4511092068932686,
+      "grad_norm": 0.08884913124824882,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 18952
+    },
+    {
+      "epoch": 0.4511330096163001,
+      "grad_norm": 6.271806112987026,
+      "learning_rate": 2e-05,
+      "loss": 1.0335,
+      "step": 18953
+    },
+    {
+      "epoch": 0.45115681233933164,
+      "grad_norm": 0.4101188877300521,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 18954
+    },
+    {
+      "epoch": 0.4511806150623631,
+      "grad_norm": 11.232372225651288,
+      "learning_rate": 2e-05,
+      "loss": 0.8954,
+      "step": 18955
+    },
+    {
+      "epoch": 0.45120441778539466,
+      "grad_norm": 1.3723280650886127,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 18956
+    },
+    {
+      "epoch": 0.45122822050842615,
+      "grad_norm": 3.722399490952373,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 18957
+    },
+    {
+      "epoch": 0.4512520232314577,
+      "grad_norm": 15.235071669969557,
+      "learning_rate": 2e-05,
+      "loss": 2.1539,
+      "step": 18958
+    },
+    {
+      "epoch": 0.4512758259544892,
+      "grad_norm": 4.411698703474628,
+      "learning_rate": 2e-05,
+      "loss": 0.7528,
+      "step": 18959
+    },
+    {
+      "epoch": 0.4512996286775207,
+      "grad_norm": 3.673930105037477,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 18960
+    },
+    {
+      "epoch": 0.4513234314005522,
+      "grad_norm": 2.273101427743269,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 18961
+    },
+    {
+      "epoch": 0.45134723412358374,
+      "grad_norm": 1.1661651627808527,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 18962
+    },
+    {
+      "epoch": 0.4513710368466152,
+      "grad_norm": 0.8749948834016008,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 18963
+    },
+    {
+      "epoch": 0.45139483956964677,
+      "grad_norm": 4.543120636263661,
+      "learning_rate": 2e-05,
+      "loss": 0.3689,
+      "step": 18964
+    },
+    {
+      "epoch": 0.4514186422926783,
+      "grad_norm": 4.21354250237445,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 18965
+    },
+    {
+      "epoch": 0.4514424450157098,
+      "grad_norm": 0.924140511049289,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 18966
+    },
+    {
+      "epoch": 0.45146624773874133,
+      "grad_norm": 3.1118243444717577,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 18967
+    },
+    {
+      "epoch": 0.4514900504617728,
+      "grad_norm": 4.835999468663794,
+      "learning_rate": 2e-05,
+      "loss": 0.4536,
+      "step": 18968
+    },
+    {
+      "epoch": 0.45151385318480436,
+      "grad_norm": 7.350152909398542,
+      "learning_rate": 2e-05,
+      "loss": 0.7123,
+      "step": 18969
+    },
+    {
+      "epoch": 0.45153765590783584,
+      "grad_norm": 2.510779028194318,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 18970
+    },
+    {
+      "epoch": 0.4515614586308674,
+      "grad_norm": 14.81193737098163,
+      "learning_rate": 2e-05,
+      "loss": 0.712,
+      "step": 18971
+    },
+    {
+      "epoch": 0.45158526135389887,
+      "grad_norm": 4.804940144680358,
+      "learning_rate": 2e-05,
+      "loss": 0.954,
+      "step": 18972
+    },
+    {
+      "epoch": 0.4516090640769304,
+      "grad_norm": 6.528496109561059,
+      "learning_rate": 2e-05,
+      "loss": 0.4928,
+      "step": 18973
+    },
+    {
+      "epoch": 0.4516328667999619,
+      "grad_norm": 4.918590518127466,
+      "learning_rate": 2e-05,
+      "loss": 0.6897,
+      "step": 18974
+    },
+    {
+      "epoch": 0.45165666952299344,
+      "grad_norm": 9.69622054747419,
+      "learning_rate": 2e-05,
+      "loss": 1.0509,
+      "step": 18975
+    },
+    {
+      "epoch": 0.4516804722460249,
+      "grad_norm": 8.23955701129645,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 18976
+    },
+    {
+      "epoch": 0.45170427496905646,
+      "grad_norm": 5.76932387344094,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 18977
+    },
+    {
+      "epoch": 0.45172807769208795,
+      "grad_norm": 5.694561905791498,
+      "learning_rate": 2e-05,
+      "loss": 0.5911,
+      "step": 18978
+    },
+    {
+      "epoch": 0.4517518804151195,
+      "grad_norm": 5.200308971228957,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 18979
+    },
+    {
+      "epoch": 0.45177568313815103,
+      "grad_norm": 2.088174122877213,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 18980
+    },
+    {
+      "epoch": 0.4517994858611825,
+      "grad_norm": 3.8551987335099915,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 18981
+    },
+    {
+      "epoch": 0.45182328858421406,
+      "grad_norm": 12.413726662787946,
+      "learning_rate": 2e-05,
+      "loss": 0.8878,
+      "step": 18982
+    },
+    {
+      "epoch": 0.45184709130724554,
+      "grad_norm": 2.7419409763360894,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 18983
+    },
+    {
+      "epoch": 0.4518708940302771,
+      "grad_norm": 7.531876156318188,
+      "learning_rate": 2e-05,
+      "loss": 0.8139,
+      "step": 18984
+    },
+    {
+      "epoch": 0.45189469675330857,
+      "grad_norm": 13.209602457598269,
+      "learning_rate": 2e-05,
+      "loss": 0.4543,
+      "step": 18985
+    },
+    {
+      "epoch": 0.4519184994763401,
+      "grad_norm": 14.458825893434664,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 18986
+    },
+    {
+      "epoch": 0.4519423021993716,
+      "grad_norm": 7.08907422730419,
+      "learning_rate": 2e-05,
+      "loss": 1.2317,
+      "step": 18987
+    },
+    {
+      "epoch": 0.45196610492240313,
+      "grad_norm": 1.7457843425678445,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 18988
+    },
+    {
+      "epoch": 0.4519899076454346,
+      "grad_norm": 4.091517615859624,
+      "learning_rate": 2e-05,
+      "loss": 0.4458,
+      "step": 18989
+    },
+    {
+      "epoch": 0.45201371036846616,
+      "grad_norm": 13.168578726581313,
+      "learning_rate": 2e-05,
+      "loss": 1.3061,
+      "step": 18990
+    },
+    {
+      "epoch": 0.45203751309149764,
+      "grad_norm": 8.579389334198563,
+      "learning_rate": 2e-05,
+      "loss": 1.5656,
+      "step": 18991
+    },
+    {
+      "epoch": 0.4520613158145292,
+      "grad_norm": 7.241488204457021,
+      "learning_rate": 2e-05,
+      "loss": 0.3709,
+      "step": 18992
+    },
+    {
+      "epoch": 0.45208511853756067,
+      "grad_norm": 15.091348695895471,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 18993
+    },
+    {
+      "epoch": 0.4521089212605922,
+      "grad_norm": 6.27186048677719,
+      "learning_rate": 2e-05,
+      "loss": 1.3459,
+      "step": 18994
+    },
+    {
+      "epoch": 0.45213272398362375,
+      "grad_norm": 35.823568419743445,
+      "learning_rate": 2e-05,
+      "loss": 0.4878,
+      "step": 18995
+    },
+    {
+      "epoch": 0.45215652670665524,
+      "grad_norm": 2.8971850097970826,
+      "learning_rate": 2e-05,
+      "loss": 0.5475,
+      "step": 18996
+    },
+    {
+      "epoch": 0.4521803294296868,
+      "grad_norm": 7.120116197702664,
+      "learning_rate": 2e-05,
+      "loss": 1.4505,
+      "step": 18997
+    },
+    {
+      "epoch": 0.45220413215271826,
+      "grad_norm": 3.8597552296662188,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 18998
+    },
+    {
+      "epoch": 0.4522279348757498,
+      "grad_norm": 0.4887972798933926,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 18999
+    },
+    {
+      "epoch": 0.4522517375987813,
+      "grad_norm": 3.321340533809357,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45227554032181283,
+      "grad_norm": 2.9957144994914815,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 19001
+    },
+    {
+      "epoch": 0.4522993430448443,
+      "grad_norm": 3.7975493757680208,
+      "learning_rate": 2e-05,
+      "loss": 0.5564,
+      "step": 19002
+    },
+    {
+      "epoch": 0.45232314576787586,
+      "grad_norm": 1.5631327868824385,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 19003
+    },
+    {
+      "epoch": 0.45234694849090734,
+      "grad_norm": 2.2509625364262926,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 19004
+    },
+    {
+      "epoch": 0.4523707512139389,
+      "grad_norm": 4.942421333314378,
+      "learning_rate": 2e-05,
+      "loss": 0.8483,
+      "step": 19005
+    },
+    {
+      "epoch": 0.45239455393697037,
+      "grad_norm": 8.695692760346306,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 19006
+    },
+    {
+      "epoch": 0.4524183566600019,
+      "grad_norm": 7.173599097045726,
+      "learning_rate": 2e-05,
+      "loss": 0.9171,
+      "step": 19007
+    },
+    {
+      "epoch": 0.4524421593830334,
+      "grad_norm": 3.356442451802622,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 19008
+    },
+    {
+      "epoch": 0.45246596210606493,
+      "grad_norm": 3.9293710587432953,
+      "learning_rate": 2e-05,
+      "loss": 0.7172,
+      "step": 19009
+    },
+    {
+      "epoch": 0.4524897648290965,
+      "grad_norm": 4.76593047264912,
+      "learning_rate": 2e-05,
+      "loss": 0.9407,
+      "step": 19010
+    },
+    {
+      "epoch": 0.45251356755212796,
+      "grad_norm": 4.330976328918477,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 19011
+    },
+    {
+      "epoch": 0.4525373702751595,
+      "grad_norm": 7.112415030483512,
+      "learning_rate": 2e-05,
+      "loss": 0.9053,
+      "step": 19012
+    },
+    {
+      "epoch": 0.452561172998191,
+      "grad_norm": 1.731060703527297,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 19013
+    },
+    {
+      "epoch": 0.4525849757212225,
+      "grad_norm": 3.875106145951676,
+      "learning_rate": 2e-05,
+      "loss": 0.1033,
+      "step": 19014
+    },
+    {
+      "epoch": 0.452608778444254,
+      "grad_norm": 15.1741576869329,
+      "learning_rate": 2e-05,
+      "loss": 0.3693,
+      "step": 19015
+    },
+    {
+      "epoch": 0.45263258116728555,
+      "grad_norm": 5.542089041953287,
+      "learning_rate": 2e-05,
+      "loss": 0.2053,
+      "step": 19016
+    },
+    {
+      "epoch": 0.45265638389031704,
+      "grad_norm": 3.835280034005505,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 19017
+    },
+    {
+      "epoch": 0.4526801866133486,
+      "grad_norm": 1.8697012991154736,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 19018
+    },
+    {
+      "epoch": 0.45270398933638006,
+      "grad_norm": 1.5021383955497876,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 19019
+    },
+    {
+      "epoch": 0.4527277920594116,
+      "grad_norm": 4.80645549855366,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 19020
+    },
+    {
+      "epoch": 0.4527515947824431,
+      "grad_norm": 7.5905467995212925,
+      "learning_rate": 2e-05,
+      "loss": 1.1229,
+      "step": 19021
+    },
+    {
+      "epoch": 0.45277539750547463,
+      "grad_norm": 2.014347284586435,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 19022
+    },
+    {
+      "epoch": 0.4527992002285061,
+      "grad_norm": 5.243184691215395,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 19023
+    },
+    {
+      "epoch": 0.45282300295153766,
+      "grad_norm": 3.3818937814683663,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 19024
+    },
+    {
+      "epoch": 0.4528468056745692,
+      "grad_norm": 3.694739380060793,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 19025
+    },
+    {
+      "epoch": 0.4528706083976007,
+      "grad_norm": 1.5562775027134756,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 19026
+    },
+    {
+      "epoch": 0.4528944111206322,
+      "grad_norm": 0.36067046332533687,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19027
+    },
+    {
+      "epoch": 0.4529182138436637,
+      "grad_norm": 3.890267692882309,
+      "learning_rate": 2e-05,
+      "loss": 0.4137,
+      "step": 19028
+    },
+    {
+      "epoch": 0.45294201656669525,
+      "grad_norm": 0.7115889808074677,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 19029
+    },
+    {
+      "epoch": 0.45296581928972673,
+      "grad_norm": 7.531873729382958,
+      "learning_rate": 2e-05,
+      "loss": 0.4555,
+      "step": 19030
+    },
+    {
+      "epoch": 0.4529896220127583,
+      "grad_norm": 6.344192639619759,
+      "learning_rate": 2e-05,
+      "loss": 0.2416,
+      "step": 19031
+    },
+    {
+      "epoch": 0.45301342473578976,
+      "grad_norm": 1.2506048421178608,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 19032
+    },
+    {
+      "epoch": 0.4530372274588213,
+      "grad_norm": 1.8233354603127911,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 19033
+    },
+    {
+      "epoch": 0.4530610301818528,
+      "grad_norm": 4.073499533740541,
+      "learning_rate": 2e-05,
+      "loss": 0.4033,
+      "step": 19034
+    },
+    {
+      "epoch": 0.4530848329048843,
+      "grad_norm": 5.777110918945606,
+      "learning_rate": 2e-05,
+      "loss": 0.5618,
+      "step": 19035
+    },
+    {
+      "epoch": 0.4531086356279158,
+      "grad_norm": 3.7877846933687027,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 19036
+    },
+    {
+      "epoch": 0.45313243835094735,
+      "grad_norm": 3.3809223560725945,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 19037
+    },
+    {
+      "epoch": 0.45315624107397884,
+      "grad_norm": 3.372440766988963,
+      "learning_rate": 2e-05,
+      "loss": 0.432,
+      "step": 19038
+    },
+    {
+      "epoch": 0.4531800437970104,
+      "grad_norm": 0.6510205385740891,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 19039
+    },
+    {
+      "epoch": 0.4532038465200419,
+      "grad_norm": 2.06818023329209,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 19040
+    },
+    {
+      "epoch": 0.4532276492430734,
+      "grad_norm": 3.9260586095581664,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 19041
+    },
+    {
+      "epoch": 0.45325145196610495,
+      "grad_norm": 0.40787977118227714,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 19042
+    },
+    {
+      "epoch": 0.45327525468913643,
+      "grad_norm": 9.865448008034933,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 19043
+    },
+    {
+      "epoch": 0.45329905741216797,
+      "grad_norm": 6.316972632933296,
+      "learning_rate": 2e-05,
+      "loss": 1.3843,
+      "step": 19044
+    },
+    {
+      "epoch": 0.45332286013519946,
+      "grad_norm": 6.243599099504732,
+      "learning_rate": 2e-05,
+      "loss": 1.3708,
+      "step": 19045
+    },
+    {
+      "epoch": 0.453346662858231,
+      "grad_norm": 4.454252939783416,
+      "learning_rate": 2e-05,
+      "loss": 0.4395,
+      "step": 19046
+    },
+    {
+      "epoch": 0.4533704655812625,
+      "grad_norm": 6.513659405247337,
+      "learning_rate": 2e-05,
+      "loss": 0.2954,
+      "step": 19047
+    },
+    {
+      "epoch": 0.453394268304294,
+      "grad_norm": 2.025327795525215,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 19048
+    },
+    {
+      "epoch": 0.4534180710273255,
+      "grad_norm": 11.473038375706789,
+      "learning_rate": 2e-05,
+      "loss": 0.5495,
+      "step": 19049
+    },
+    {
+      "epoch": 0.45344187375035705,
+      "grad_norm": 4.679984416643397,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 19050
+    },
+    {
+      "epoch": 0.45346567647338853,
+      "grad_norm": 2.4234144314918105,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 19051
+    },
+    {
+      "epoch": 0.4534894791964201,
+      "grad_norm": 0.7265223736954411,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 19052
+    },
+    {
+      "epoch": 0.45351328191945156,
+      "grad_norm": 3.391372185469865,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 19053
+    },
+    {
+      "epoch": 0.4535370846424831,
+      "grad_norm": 3.606004429313506,
+      "learning_rate": 2e-05,
+      "loss": 0.4451,
+      "step": 19054
+    },
+    {
+      "epoch": 0.45356088736551464,
+      "grad_norm": 5.0041291532255805,
+      "learning_rate": 2e-05,
+      "loss": 0.5084,
+      "step": 19055
+    },
+    {
+      "epoch": 0.4535846900885461,
+      "grad_norm": 4.782990590438338,
+      "learning_rate": 2e-05,
+      "loss": 0.6975,
+      "step": 19056
+    },
+    {
+      "epoch": 0.45360849281157767,
+      "grad_norm": 5.436772814282644,
+      "learning_rate": 2e-05,
+      "loss": 0.8537,
+      "step": 19057
+    },
+    {
+      "epoch": 0.45363229553460915,
+      "grad_norm": 6.565452012290949,
+      "learning_rate": 2e-05,
+      "loss": 0.5458,
+      "step": 19058
+    },
+    {
+      "epoch": 0.4536560982576407,
+      "grad_norm": 3.1675434113796284,
+      "learning_rate": 2e-05,
+      "loss": 0.5471,
+      "step": 19059
+    },
+    {
+      "epoch": 0.4536799009806722,
+      "grad_norm": 1.4912120828072286,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 19060
+    },
+    {
+      "epoch": 0.4537037037037037,
+      "grad_norm": 3.7626843792366116,
+      "learning_rate": 2e-05,
+      "loss": 0.5401,
+      "step": 19061
+    },
+    {
+      "epoch": 0.4537275064267352,
+      "grad_norm": 2.648787366607522,
+      "learning_rate": 2e-05,
+      "loss": 0.1649,
+      "step": 19062
+    },
+    {
+      "epoch": 0.45375130914976675,
+      "grad_norm": 3.6277455136566883,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 19063
+    },
+    {
+      "epoch": 0.45377511187279823,
+      "grad_norm": 8.7255097108306,
+      "learning_rate": 2e-05,
+      "loss": 1.095,
+      "step": 19064
+    },
+    {
+      "epoch": 0.45379891459582977,
+      "grad_norm": 1.992318352908365,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 19065
+    },
+    {
+      "epoch": 0.45382271731886126,
+      "grad_norm": 3.9427234432397613,
+      "learning_rate": 2e-05,
+      "loss": 0.7706,
+      "step": 19066
+    },
+    {
+      "epoch": 0.4538465200418928,
+      "grad_norm": 4.777819495963123,
+      "learning_rate": 2e-05,
+      "loss": 0.3149,
+      "step": 19067
+    },
+    {
+      "epoch": 0.4538703227649243,
+      "grad_norm": 3.7841198085481795,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 19068
+    },
+    {
+      "epoch": 0.4538941254879558,
+      "grad_norm": 4.7008395559908305,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 19069
+    },
+    {
+      "epoch": 0.4539179282109873,
+      "grad_norm": 7.653349093183988,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 19070
+    },
+    {
+      "epoch": 0.45394173093401885,
+      "grad_norm": 5.193137960502688,
+      "learning_rate": 2e-05,
+      "loss": 0.7109,
+      "step": 19071
+    },
+    {
+      "epoch": 0.4539655336570504,
+      "grad_norm": 1.9757172399937557,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 19072
+    },
+    {
+      "epoch": 0.4539893363800819,
+      "grad_norm": 0.4388852802159408,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 19073
+    },
+    {
+      "epoch": 0.4540131391031134,
+      "grad_norm": 5.650098732978031,
+      "learning_rate": 2e-05,
+      "loss": 0.6184,
+      "step": 19074
+    },
+    {
+      "epoch": 0.4540369418261449,
+      "grad_norm": 7.091168514522697,
+      "learning_rate": 2e-05,
+      "loss": 0.7781,
+      "step": 19075
+    },
+    {
+      "epoch": 0.45406074454917644,
+      "grad_norm": 9.766831510606139,
+      "learning_rate": 2e-05,
+      "loss": 1.364,
+      "step": 19076
+    },
+    {
+      "epoch": 0.45408454727220793,
+      "grad_norm": 6.405205726923943,
+      "learning_rate": 2e-05,
+      "loss": 0.8696,
+      "step": 19077
+    },
+    {
+      "epoch": 0.45410834999523947,
+      "grad_norm": 3.355322639067984,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 19078
+    },
+    {
+      "epoch": 0.45413215271827095,
+      "grad_norm": 5.510445047964436,
+      "learning_rate": 2e-05,
+      "loss": 1.0448,
+      "step": 19079
+    },
+    {
+      "epoch": 0.4541559554413025,
+      "grad_norm": 9.825166908273143,
+      "learning_rate": 2e-05,
+      "loss": 1.0271,
+      "step": 19080
+    },
+    {
+      "epoch": 0.454179758164334,
+      "grad_norm": 108.96346431132018,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 19081
+    },
+    {
+      "epoch": 0.4542035608873655,
+      "grad_norm": 4.978943083816442,
+      "learning_rate": 2e-05,
+      "loss": 0.751,
+      "step": 19082
+    },
+    {
+      "epoch": 0.454227363610397,
+      "grad_norm": 7.456629891944291,
+      "learning_rate": 2e-05,
+      "loss": 0.1679,
+      "step": 19083
+    },
+    {
+      "epoch": 0.45425116633342855,
+      "grad_norm": 2.301335229484654,
+      "learning_rate": 2e-05,
+      "loss": 0.3679,
+      "step": 19084
+    },
+    {
+      "epoch": 0.45427496905646003,
+      "grad_norm": 6.861106092158652,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 19085
+    },
+    {
+      "epoch": 0.4542987717794916,
+      "grad_norm": 0.46882178408295616,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 19086
+    },
+    {
+      "epoch": 0.4543225745025231,
+      "grad_norm": 8.679076809689398,
+      "learning_rate": 2e-05,
+      "loss": 1.4934,
+      "step": 19087
+    },
+    {
+      "epoch": 0.4543463772255546,
+      "grad_norm": 8.090848938746573,
+      "learning_rate": 2e-05,
+      "loss": 1.6558,
+      "step": 19088
+    },
+    {
+      "epoch": 0.45437017994858614,
+      "grad_norm": 3.9831189918590244,
+      "learning_rate": 2e-05,
+      "loss": 0.6733,
+      "step": 19089
+    },
+    {
+      "epoch": 0.4543939826716176,
+      "grad_norm": 4.9719720805524235,
+      "learning_rate": 2e-05,
+      "loss": 0.8711,
+      "step": 19090
+    },
+    {
+      "epoch": 0.45441778539464917,
+      "grad_norm": 1.0183138272056405,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 19091
+    },
+    {
+      "epoch": 0.45444158811768065,
+      "grad_norm": 1.4368277516924024,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 19092
+    },
+    {
+      "epoch": 0.4544653908407122,
+      "grad_norm": 2.1281057224751287,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 19093
+    },
+    {
+      "epoch": 0.4544891935637437,
+      "grad_norm": 14.346898202605221,
+      "learning_rate": 2e-05,
+      "loss": 1.0817,
+      "step": 19094
+    },
+    {
+      "epoch": 0.4545129962867752,
+      "grad_norm": 0.08577283240957873,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 19095
+    },
+    {
+      "epoch": 0.4545367990098067,
+      "grad_norm": 1.4531435629254554,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 19096
+    },
+    {
+      "epoch": 0.45456060173283824,
+      "grad_norm": 4.2350173130980195,
+      "learning_rate": 2e-05,
+      "loss": 0.7192,
+      "step": 19097
+    },
+    {
+      "epoch": 0.45458440445586973,
+      "grad_norm": 0.15706460320131568,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 19098
+    },
+    {
+      "epoch": 0.45460820717890127,
+      "grad_norm": 2.643575059665277,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 19099
+    },
+    {
+      "epoch": 0.45463200990193275,
+      "grad_norm": 2.974809106327619,
+      "learning_rate": 2e-05,
+      "loss": 0.1401,
+      "step": 19100
+    },
+    {
+      "epoch": 0.4546558126249643,
+      "grad_norm": 1.9253292204036252,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 19101
+    },
+    {
+      "epoch": 0.45467961534799584,
+      "grad_norm": 1.1754095254245005,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19102
+    },
+    {
+      "epoch": 0.4547034180710273,
+      "grad_norm": 1.3792209764294319,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 19103
+    },
+    {
+      "epoch": 0.45472722079405886,
+      "grad_norm": 4.267162038518813,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 19104
+    },
+    {
+      "epoch": 0.45475102351709035,
+      "grad_norm": 3.178546984951293,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 19105
+    },
+    {
+      "epoch": 0.4547748262401219,
+      "grad_norm": 0.7809262234901816,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 19106
+    },
+    {
+      "epoch": 0.4547986289631534,
+      "grad_norm": 1.714499783333208,
+      "learning_rate": 2e-05,
+      "loss": 0.1007,
+      "step": 19107
+    },
+    {
+      "epoch": 0.4548224316861849,
+      "grad_norm": 2.6918748680288993,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 19108
+    },
+    {
+      "epoch": 0.4548462344092164,
+      "grad_norm": 3.033161237324584,
+      "learning_rate": 2e-05,
+      "loss": 0.2988,
+      "step": 19109
+    },
+    {
+      "epoch": 0.45487003713224794,
+      "grad_norm": 3.781934009345476,
+      "learning_rate": 2e-05,
+      "loss": 0.6712,
+      "step": 19110
+    },
+    {
+      "epoch": 0.4548938398552794,
+      "grad_norm": 0.8339896417213296,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 19111
+    },
+    {
+      "epoch": 0.45491764257831097,
+      "grad_norm": 0.27145387863496573,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 19112
+    },
+    {
+      "epoch": 0.45494144530134245,
+      "grad_norm": 0.5767409151067886,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19113
+    },
+    {
+      "epoch": 0.454965248024374,
+      "grad_norm": 0.2938751479955502,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19114
+    },
+    {
+      "epoch": 0.4549890507474055,
+      "grad_norm": 2.2112458126207057,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 19115
+    },
+    {
+      "epoch": 0.455012853470437,
+      "grad_norm": 2.285858886066715,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 19116
+    },
+    {
+      "epoch": 0.45503665619346856,
+      "grad_norm": 0.41123087755170307,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 19117
+    },
+    {
+      "epoch": 0.45506045891650004,
+      "grad_norm": 4.804103671406162,
+      "learning_rate": 2e-05,
+      "loss": 0.4406,
+      "step": 19118
+    },
+    {
+      "epoch": 0.4550842616395316,
+      "grad_norm": 3.081224582128397,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 19119
+    },
+    {
+      "epoch": 0.45510806436256307,
+      "grad_norm": 5.774651500306999,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 19120
+    },
+    {
+      "epoch": 0.4551318670855946,
+      "grad_norm": 2.1553590421441893,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 19121
+    },
+    {
+      "epoch": 0.4551556698086261,
+      "grad_norm": 3.5438785667559016,
+      "learning_rate": 2e-05,
+      "loss": 0.5521,
+      "step": 19122
+    },
+    {
+      "epoch": 0.45517947253165764,
+      "grad_norm": 0.4627087653967108,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19123
+    },
+    {
+      "epoch": 0.4552032752546891,
+      "grad_norm": 4.2853673603769105,
+      "learning_rate": 2e-05,
+      "loss": 0.9925,
+      "step": 19124
+    },
+    {
+      "epoch": 0.45522707797772066,
+      "grad_norm": 3.531971923216091,
+      "learning_rate": 2e-05,
+      "loss": 0.1632,
+      "step": 19125
+    },
+    {
+      "epoch": 0.45525088070075215,
+      "grad_norm": 1.5289450835881528,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 19126
+    },
+    {
+      "epoch": 0.4552746834237837,
+      "grad_norm": 4.490808511316397,
+      "learning_rate": 2e-05,
+      "loss": 0.3982,
+      "step": 19127
+    },
+    {
+      "epoch": 0.4552984861468152,
+      "grad_norm": 3.4148405772696053,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 19128
+    },
+    {
+      "epoch": 0.4553222888698467,
+      "grad_norm": 7.066257924884904,
+      "learning_rate": 2e-05,
+      "loss": 1.1244,
+      "step": 19129
+    },
+    {
+      "epoch": 0.4553460915928782,
+      "grad_norm": 2.5050219694624607,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 19130
+    },
+    {
+      "epoch": 0.45536989431590974,
+      "grad_norm": 3.2256246208292967,
+      "learning_rate": 2e-05,
+      "loss": 0.4581,
+      "step": 19131
+    },
+    {
+      "epoch": 0.4553936970389413,
+      "grad_norm": 8.563187003355141,
+      "learning_rate": 2e-05,
+      "loss": 0.587,
+      "step": 19132
+    },
+    {
+      "epoch": 0.45541749976197277,
+      "grad_norm": 3.79671799657955,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 19133
+    },
+    {
+      "epoch": 0.4554413024850043,
+      "grad_norm": 3.6890734412689694,
+      "learning_rate": 2e-05,
+      "loss": 0.5722,
+      "step": 19134
+    },
+    {
+      "epoch": 0.4554651052080358,
+      "grad_norm": 5.486879251885363,
+      "learning_rate": 2e-05,
+      "loss": 1.1492,
+      "step": 19135
+    },
+    {
+      "epoch": 0.45548890793106733,
+      "grad_norm": 3.813133947322814,
+      "learning_rate": 2e-05,
+      "loss": 0.3983,
+      "step": 19136
+    },
+    {
+      "epoch": 0.4555127106540988,
+      "grad_norm": 1.8382872178203828,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 19137
+    },
+    {
+      "epoch": 0.45553651337713036,
+      "grad_norm": 0.8730743272334276,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 19138
+    },
+    {
+      "epoch": 0.45556031610016184,
+      "grad_norm": 1.3696952136272593,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 19139
+    },
+    {
+      "epoch": 0.4555841188231934,
+      "grad_norm": 28.97327097038264,
+      "learning_rate": 2e-05,
+      "loss": 0.8078,
+      "step": 19140
+    },
+    {
+      "epoch": 0.45560792154622487,
+      "grad_norm": 3.1586859817994997,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 19141
+    },
+    {
+      "epoch": 0.4556317242692564,
+      "grad_norm": 15.147506607123733,
+      "learning_rate": 2e-05,
+      "loss": 0.7664,
+      "step": 19142
+    },
+    {
+      "epoch": 0.4556555269922879,
+      "grad_norm": 4.432765386506377,
+      "learning_rate": 2e-05,
+      "loss": 0.3474,
+      "step": 19143
+    },
+    {
+      "epoch": 0.45567932971531944,
+      "grad_norm": 4.974348802745308,
+      "learning_rate": 2e-05,
+      "loss": 0.9394,
+      "step": 19144
+    },
+    {
+      "epoch": 0.4557031324383509,
+      "grad_norm": 1.9015636984672257,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 19145
+    },
+    {
+      "epoch": 0.45572693516138246,
+      "grad_norm": 3.28029614831034,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 19146
+    },
+    {
+      "epoch": 0.455750737884414,
+      "grad_norm": 3.9338020102575424,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 19147
+    },
+    {
+      "epoch": 0.4557745406074455,
+      "grad_norm": 15.45328722694176,
+      "learning_rate": 2e-05,
+      "loss": 0.9889,
+      "step": 19148
+    },
+    {
+      "epoch": 0.45579834333047703,
+      "grad_norm": 4.551466362081141,
+      "learning_rate": 2e-05,
+      "loss": 0.6061,
+      "step": 19149
+    },
+    {
+      "epoch": 0.4558221460535085,
+      "grad_norm": 4.675421021627045,
+      "learning_rate": 2e-05,
+      "loss": 0.1751,
+      "step": 19150
+    },
+    {
+      "epoch": 0.45584594877654006,
+      "grad_norm": 7.346818808400616,
+      "learning_rate": 2e-05,
+      "loss": 0.7046,
+      "step": 19151
+    },
+    {
+      "epoch": 0.45586975149957154,
+      "grad_norm": 2.4587240708241227,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 19152
+    },
+    {
+      "epoch": 0.4558935542226031,
+      "grad_norm": 13.174260893367604,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 19153
+    },
+    {
+      "epoch": 0.45591735694563457,
+      "grad_norm": 4.535491114365023,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 19154
+    },
+    {
+      "epoch": 0.4559411596686661,
+      "grad_norm": 5.818310870672264,
+      "learning_rate": 2e-05,
+      "loss": 1.2832,
+      "step": 19155
+    },
+    {
+      "epoch": 0.4559649623916976,
+      "grad_norm": 7.5698496591051505,
+      "learning_rate": 2e-05,
+      "loss": 0.7303,
+      "step": 19156
+    },
+    {
+      "epoch": 0.45598876511472913,
+      "grad_norm": 1.8329228009309912,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 19157
+    },
+    {
+      "epoch": 0.4560125678377606,
+      "grad_norm": 13.582532518911234,
+      "learning_rate": 2e-05,
+      "loss": 1.2522,
+      "step": 19158
+    },
+    {
+      "epoch": 0.45603637056079216,
+      "grad_norm": 6.9125590709613824,
+      "learning_rate": 2e-05,
+      "loss": 1.164,
+      "step": 19159
+    },
+    {
+      "epoch": 0.45606017328382364,
+      "grad_norm": 11.402161223916528,
+      "learning_rate": 2e-05,
+      "loss": 0.482,
+      "step": 19160
+    },
+    {
+      "epoch": 0.4560839760068552,
+      "grad_norm": 2.291980207637415,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 19161
+    },
+    {
+      "epoch": 0.4561077787298867,
+      "grad_norm": 4.264923171325235,
+      "learning_rate": 2e-05,
+      "loss": 0.8601,
+      "step": 19162
+    },
+    {
+      "epoch": 0.4561315814529182,
+      "grad_norm": 2.2527145801264084,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 19163
+    },
+    {
+      "epoch": 0.45615538417594975,
+      "grad_norm": 3.781206396071458,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 19164
+    },
+    {
+      "epoch": 0.45617918689898124,
+      "grad_norm": 0.8455821387780762,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 19165
+    },
+    {
+      "epoch": 0.4562029896220128,
+      "grad_norm": 4.698368920619803,
+      "learning_rate": 2e-05,
+      "loss": 0.943,
+      "step": 19166
+    },
+    {
+      "epoch": 0.45622679234504426,
+      "grad_norm": 12.325256439455885,
+      "learning_rate": 2e-05,
+      "loss": 0.7252,
+      "step": 19167
+    },
+    {
+      "epoch": 0.4562505950680758,
+      "grad_norm": 1368.1121660425886,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 19168
+    },
+    {
+      "epoch": 0.4562743977911073,
+      "grad_norm": 1.482078388658063,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 19169
+    },
+    {
+      "epoch": 0.45629820051413883,
+      "grad_norm": 0.5783509915959276,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 19170
+    },
+    {
+      "epoch": 0.4563220032371703,
+      "grad_norm": 2.592993999130397,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 19171
+    },
+    {
+      "epoch": 0.45634580596020186,
+      "grad_norm": 0.34544264075711395,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19172
+    },
+    {
+      "epoch": 0.45636960868323334,
+      "grad_norm": 3.749837329156239,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 19173
+    },
+    {
+      "epoch": 0.4563934114062649,
+      "grad_norm": 1.1492133638745246,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 19174
+    },
+    {
+      "epoch": 0.45641721412929637,
+      "grad_norm": 4.928600402585489,
+      "learning_rate": 2e-05,
+      "loss": 0.1108,
+      "step": 19175
+    },
+    {
+      "epoch": 0.4564410168523279,
+      "grad_norm": 2.879408885179686,
+      "learning_rate": 2e-05,
+      "loss": 0.5123,
+      "step": 19176
+    },
+    {
+      "epoch": 0.45646481957535945,
+      "grad_norm": 1.9488230151342905,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 19177
+    },
+    {
+      "epoch": 0.45648862229839093,
+      "grad_norm": 0.49032135488346335,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 19178
+    },
+    {
+      "epoch": 0.4565124250214225,
+      "grad_norm": 1.560825782550216,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 19179
+    },
+    {
+      "epoch": 0.45653622774445396,
+      "grad_norm": 2.5046536828238755,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 19180
+    },
+    {
+      "epoch": 0.4565600304674855,
+      "grad_norm": 4.250105202354774,
+      "learning_rate": 2e-05,
+      "loss": 0.9138,
+      "step": 19181
+    },
+    {
+      "epoch": 0.456583833190517,
+      "grad_norm": 6.1159213355173385,
+      "learning_rate": 2e-05,
+      "loss": 0.9925,
+      "step": 19182
+    },
+    {
+      "epoch": 0.4566076359135485,
+      "grad_norm": 2.791939419258226,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 19183
+    },
+    {
+      "epoch": 0.45663143863658,
+      "grad_norm": 26.734287002554744,
+      "learning_rate": 2e-05,
+      "loss": 1.2118,
+      "step": 19184
+    },
+    {
+      "epoch": 0.45665524135961155,
+      "grad_norm": 10.652755368642266,
+      "learning_rate": 2e-05,
+      "loss": 0.3346,
+      "step": 19185
+    },
+    {
+      "epoch": 0.45667904408264304,
+      "grad_norm": 4.109653987869424,
+      "learning_rate": 2e-05,
+      "loss": 0.5798,
+      "step": 19186
+    },
+    {
+      "epoch": 0.4567028468056746,
+      "grad_norm": 3.807748382014669,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 19187
+    },
+    {
+      "epoch": 0.45672664952870606,
+      "grad_norm": 2.22976897095869,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 19188
+    },
+    {
+      "epoch": 0.4567504522517376,
+      "grad_norm": 3.391504225997845,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 19189
+    },
+    {
+      "epoch": 0.4567742549747691,
+      "grad_norm": 1.7635774961948547,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 19190
+    },
+    {
+      "epoch": 0.45679805769780063,
+      "grad_norm": 3.7555397677028677,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 19191
+    },
+    {
+      "epoch": 0.45682186042083217,
+      "grad_norm": 2.8728009359486206,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 19192
+    },
+    {
+      "epoch": 0.45684566314386366,
+      "grad_norm": 15.971979054337355,
+      "learning_rate": 2e-05,
+      "loss": 2.2246,
+      "step": 19193
+    },
+    {
+      "epoch": 0.4568694658668952,
+      "grad_norm": 3.864318491682208,
+      "learning_rate": 2e-05,
+      "loss": 0.6505,
+      "step": 19194
+    },
+    {
+      "epoch": 0.4568932685899267,
+      "grad_norm": 0.11827283480034807,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19195
+    },
+    {
+      "epoch": 0.4569170713129582,
+      "grad_norm": 2.1479030510194383,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 19196
+    },
+    {
+      "epoch": 0.4569408740359897,
+      "grad_norm": 4.018819852239716,
+      "learning_rate": 2e-05,
+      "loss": 0.5113,
+      "step": 19197
+    },
+    {
+      "epoch": 0.45696467675902125,
+      "grad_norm": 4.5016879426619445,
+      "learning_rate": 2e-05,
+      "loss": 0.3259,
+      "step": 19198
+    },
+    {
+      "epoch": 0.45698847948205273,
+      "grad_norm": 7.269951783444779,
+      "learning_rate": 2e-05,
+      "loss": 0.1917,
+      "step": 19199
+    },
+    {
+      "epoch": 0.4570122822050843,
+      "grad_norm": 4.558477042224153,
+      "learning_rate": 2e-05,
+      "loss": 0.6927,
+      "step": 19200
+    },
+    {
+      "epoch": 0.45703608492811576,
+      "grad_norm": 0.509704490213658,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 19201
+    },
+    {
+      "epoch": 0.4570598876511473,
+      "grad_norm": 2.015130856558757,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 19202
+    },
+    {
+      "epoch": 0.4570836903741788,
+      "grad_norm": 3.7131522693287655,
+      "learning_rate": 2e-05,
+      "loss": 0.7855,
+      "step": 19203
+    },
+    {
+      "epoch": 0.4571074930972103,
+      "grad_norm": 0.8350939523579596,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 19204
+    },
+    {
+      "epoch": 0.4571312958202418,
+      "grad_norm": 3.4594307508479916,
+      "learning_rate": 2e-05,
+      "loss": 0.6507,
+      "step": 19205
+    },
+    {
+      "epoch": 0.45715509854327335,
+      "grad_norm": 15.287264284089993,
+      "learning_rate": 2e-05,
+      "loss": 0.3227,
+      "step": 19206
+    },
+    {
+      "epoch": 0.45717890126630484,
+      "grad_norm": 0.8474792536361191,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 19207
+    },
+    {
+      "epoch": 0.4572027039893364,
+      "grad_norm": 38.48944104269886,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 19208
+    },
+    {
+      "epoch": 0.4572265067123679,
+      "grad_norm": 2.0221585386770204,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 19209
+    },
+    {
+      "epoch": 0.4572503094353994,
+      "grad_norm": 65.74078896813235,
+      "learning_rate": 2e-05,
+      "loss": 0.414,
+      "step": 19210
+    },
+    {
+      "epoch": 0.45727411215843095,
+      "grad_norm": 0.8347851354235456,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 19211
+    },
+    {
+      "epoch": 0.45729791488146243,
+      "grad_norm": 1.117571382106713,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 19212
+    },
+    {
+      "epoch": 0.45732171760449397,
+      "grad_norm": 2.855998396182623,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 19213
+    },
+    {
+      "epoch": 0.45734552032752546,
+      "grad_norm": 4.256679823746669,
+      "learning_rate": 2e-05,
+      "loss": 0.5942,
+      "step": 19214
+    },
+    {
+      "epoch": 0.457369323050557,
+      "grad_norm": 0.6263186748608142,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 19215
+    },
+    {
+      "epoch": 0.4573931257735885,
+      "grad_norm": 3.5905433140167538,
+      "learning_rate": 2e-05,
+      "loss": 0.3738,
+      "step": 19216
+    },
+    {
+      "epoch": 0.45741692849662,
+      "grad_norm": 3.7343610263020914,
+      "learning_rate": 2e-05,
+      "loss": 0.759,
+      "step": 19217
+    },
+    {
+      "epoch": 0.4574407312196515,
+      "grad_norm": 8.095010528316395,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 19218
+    },
+    {
+      "epoch": 0.45746453394268305,
+      "grad_norm": 2.1107985770511313,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 19219
+    },
+    {
+      "epoch": 0.45748833666571453,
+      "grad_norm": 3.5308928384305767,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 19220
+    },
+    {
+      "epoch": 0.4575121393887461,
+      "grad_norm": 0.19570578697323626,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 19221
+    },
+    {
+      "epoch": 0.45753594211177756,
+      "grad_norm": 2.81360476058125,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 19222
+    },
+    {
+      "epoch": 0.4575597448348091,
+      "grad_norm": 4.316845130186272,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 19223
+    },
+    {
+      "epoch": 0.45758354755784064,
+      "grad_norm": 5.11002153104835,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 19224
+    },
+    {
+      "epoch": 0.4576073502808721,
+      "grad_norm": 2.290179969171967,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 19225
+    },
+    {
+      "epoch": 0.45763115300390367,
+      "grad_norm": 4.841659701685561,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 19226
+    },
+    {
+      "epoch": 0.45765495572693515,
+      "grad_norm": 1.5610041948394875,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 19227
+    },
+    {
+      "epoch": 0.4576787584499667,
+      "grad_norm": 4.4878728130675265,
+      "learning_rate": 2e-05,
+      "loss": 0.6178,
+      "step": 19228
+    },
+    {
+      "epoch": 0.4577025611729982,
+      "grad_norm": 4.099679533691404,
+      "learning_rate": 2e-05,
+      "loss": 0.4346,
+      "step": 19229
+    },
+    {
+      "epoch": 0.4577263638960297,
+      "grad_norm": 10.588558987751208,
+      "learning_rate": 2e-05,
+      "loss": 0.8966,
+      "step": 19230
+    },
+    {
+      "epoch": 0.4577501666190612,
+      "grad_norm": 5.000661308997513,
+      "learning_rate": 2e-05,
+      "loss": 1.1053,
+      "step": 19231
+    },
+    {
+      "epoch": 0.45777396934209275,
+      "grad_norm": 4.377175497909097,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 19232
+    },
+    {
+      "epoch": 0.45779777206512423,
+      "grad_norm": 0.761989701286902,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 19233
+    },
+    {
+      "epoch": 0.45782157478815577,
+      "grad_norm": 9.230753667516142,
+      "learning_rate": 2e-05,
+      "loss": 0.83,
+      "step": 19234
+    },
+    {
+      "epoch": 0.45784537751118726,
+      "grad_norm": 13.532125558584857,
+      "learning_rate": 2e-05,
+      "loss": 0.8315,
+      "step": 19235
+    },
+    {
+      "epoch": 0.4578691802342188,
+      "grad_norm": 6.001069652816254,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 19236
+    },
+    {
+      "epoch": 0.4578929829572503,
+      "grad_norm": 3.785449749004027,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 19237
+    },
+    {
+      "epoch": 0.4579167856802818,
+      "grad_norm": 7.9804534216726,
+      "learning_rate": 2e-05,
+      "loss": 0.571,
+      "step": 19238
+    },
+    {
+      "epoch": 0.45794058840331336,
+      "grad_norm": 3.2050993727777812,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 19239
+    },
+    {
+      "epoch": 0.45796439112634485,
+      "grad_norm": 11.77906434603141,
+      "learning_rate": 2e-05,
+      "loss": 0.5195,
+      "step": 19240
+    },
+    {
+      "epoch": 0.4579881938493764,
+      "grad_norm": 5.84378211801065,
+      "learning_rate": 2e-05,
+      "loss": 1.1068,
+      "step": 19241
+    },
+    {
+      "epoch": 0.4580119965724079,
+      "grad_norm": 2.7648255696232376,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 19242
+    },
+    {
+      "epoch": 0.4580357992954394,
+      "grad_norm": 5.703125866462604,
+      "learning_rate": 2e-05,
+      "loss": 0.9079,
+      "step": 19243
+    },
+    {
+      "epoch": 0.4580596020184709,
+      "grad_norm": 0.4892121344837166,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 19244
+    },
+    {
+      "epoch": 0.45808340474150244,
+      "grad_norm": 2.176688962460467,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 19245
+    },
+    {
+      "epoch": 0.4581072074645339,
+      "grad_norm": 5.196175247076558,
+      "learning_rate": 2e-05,
+      "loss": 0.6331,
+      "step": 19246
+    },
+    {
+      "epoch": 0.45813101018756547,
+      "grad_norm": 3.147497976734388,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 19247
+    },
+    {
+      "epoch": 0.45815481291059695,
+      "grad_norm": 2.4978238034166074,
+      "learning_rate": 2e-05,
+      "loss": 0.3668,
+      "step": 19248
+    },
+    {
+      "epoch": 0.4581786156336285,
+      "grad_norm": 27.035516287492364,
+      "learning_rate": 2e-05,
+      "loss": 0.9056,
+      "step": 19249
+    },
+    {
+      "epoch": 0.45820241835666,
+      "grad_norm": 6.001217829409684,
+      "learning_rate": 2e-05,
+      "loss": 1.2412,
+      "step": 19250
+    },
+    {
+      "epoch": 0.4582262210796915,
+      "grad_norm": 1.7036902295370273,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 19251
+    },
+    {
+      "epoch": 0.458250023802723,
+      "grad_norm": 4.135384138170621,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 19252
+    },
+    {
+      "epoch": 0.45827382652575455,
+      "grad_norm": 4.0553170667473095,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 19253
+    },
+    {
+      "epoch": 0.4582976292487861,
+      "grad_norm": 0.526053267748399,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 19254
+    },
+    {
+      "epoch": 0.45832143197181757,
+      "grad_norm": 7.113800557660938,
+      "learning_rate": 2e-05,
+      "loss": 0.7932,
+      "step": 19255
+    },
+    {
+      "epoch": 0.4583452346948491,
+      "grad_norm": 2.8688018282206458,
+      "learning_rate": 2e-05,
+      "loss": 0.1916,
+      "step": 19256
+    },
+    {
+      "epoch": 0.4583690374178806,
+      "grad_norm": 3.597434390059288,
+      "learning_rate": 2e-05,
+      "loss": 0.4812,
+      "step": 19257
+    },
+    {
+      "epoch": 0.45839284014091214,
+      "grad_norm": 19.206007077757537,
+      "learning_rate": 2e-05,
+      "loss": 1.6153,
+      "step": 19258
+    },
+    {
+      "epoch": 0.4584166428639436,
+      "grad_norm": 3.694137723302445,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 19259
+    },
+    {
+      "epoch": 0.45844044558697516,
+      "grad_norm": 3.2472964648069986,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 19260
+    },
+    {
+      "epoch": 0.45846424831000665,
+      "grad_norm": 2.926795169416115,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 19261
+    },
+    {
+      "epoch": 0.4584880510330382,
+      "grad_norm": 6.3087020684521455,
+      "learning_rate": 2e-05,
+      "loss": 1.1188,
+      "step": 19262
+    },
+    {
+      "epoch": 0.4585118537560697,
+      "grad_norm": 4.272183082473278,
+      "learning_rate": 2e-05,
+      "loss": 0.2767,
+      "step": 19263
+    },
+    {
+      "epoch": 0.4585356564791012,
+      "grad_norm": 1.0181247309178905,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 19264
+    },
+    {
+      "epoch": 0.4585594592021327,
+      "grad_norm": 3.496935049544023,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 19265
+    },
+    {
+      "epoch": 0.45858326192516424,
+      "grad_norm": 3.947959816814374,
+      "learning_rate": 2e-05,
+      "loss": 0.3076,
+      "step": 19266
+    },
+    {
+      "epoch": 0.45860706464819573,
+      "grad_norm": 2.6865911509546305,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 19267
+    },
+    {
+      "epoch": 0.45863086737122727,
+      "grad_norm": 0.9432495241812133,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 19268
+    },
+    {
+      "epoch": 0.4586546700942588,
+      "grad_norm": 16.019043611197134,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 19269
+    },
+    {
+      "epoch": 0.4586784728172903,
+      "grad_norm": 5.877619397686674,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 19270
+    },
+    {
+      "epoch": 0.45870227554032184,
+      "grad_norm": 5.002643003707763,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 19271
+    },
+    {
+      "epoch": 0.4587260782633533,
+      "grad_norm": 5.9834070436812405,
+      "learning_rate": 2e-05,
+      "loss": 0.6735,
+      "step": 19272
+    },
+    {
+      "epoch": 0.45874988098638486,
+      "grad_norm": 5.559384497153562,
+      "learning_rate": 2e-05,
+      "loss": 0.1009,
+      "step": 19273
+    },
+    {
+      "epoch": 0.45877368370941635,
+      "grad_norm": 6.619532879499513,
+      "learning_rate": 2e-05,
+      "loss": 1.3668,
+      "step": 19274
+    },
+    {
+      "epoch": 0.4587974864324479,
+      "grad_norm": 6.880872680582411,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 19275
+    },
+    {
+      "epoch": 0.4588212891554794,
+      "grad_norm": 2.8422418527642934,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 19276
+    },
+    {
+      "epoch": 0.4588450918785109,
+      "grad_norm": 4.402204381157772,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 19277
+    },
+    {
+      "epoch": 0.4588688946015424,
+      "grad_norm": 4.447560977166241,
+      "learning_rate": 2e-05,
+      "loss": 0.6888,
+      "step": 19278
+    },
+    {
+      "epoch": 0.45889269732457394,
+      "grad_norm": 3.855526775618117,
+      "learning_rate": 2e-05,
+      "loss": 0.6466,
+      "step": 19279
+    },
+    {
+      "epoch": 0.4589165000476054,
+      "grad_norm": 4.1082943638223925,
+      "learning_rate": 2e-05,
+      "loss": 0.7977,
+      "step": 19280
+    },
+    {
+      "epoch": 0.45894030277063697,
+      "grad_norm": 0.4720754889969515,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 19281
+    },
+    {
+      "epoch": 0.45896410549366845,
+      "grad_norm": 1.7627951524386896,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 19282
+    },
+    {
+      "epoch": 0.4589879082167,
+      "grad_norm": 0.28521976694396395,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 19283
+    },
+    {
+      "epoch": 0.45901171093973153,
+      "grad_norm": 1.8369379619863166,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 19284
+    },
+    {
+      "epoch": 0.459035513662763,
+      "grad_norm": 3.7607138273138636,
+      "learning_rate": 2e-05,
+      "loss": 0.5343,
+      "step": 19285
+    },
+    {
+      "epoch": 0.45905931638579456,
+      "grad_norm": 11.825762355579597,
+      "learning_rate": 2e-05,
+      "loss": 1.0971,
+      "step": 19286
+    },
+    {
+      "epoch": 0.45908311910882604,
+      "grad_norm": 3.3559340361622563,
+      "learning_rate": 2e-05,
+      "loss": 0.645,
+      "step": 19287
+    },
+    {
+      "epoch": 0.4591069218318576,
+      "grad_norm": 1.8744338885902123,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 19288
+    },
+    {
+      "epoch": 0.45913072455488907,
+      "grad_norm": 2.660380674573085,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 19289
+    },
+    {
+      "epoch": 0.4591545272779206,
+      "grad_norm": 4.573936034074254,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 19290
+    },
+    {
+      "epoch": 0.4591783300009521,
+      "grad_norm": 1.856524026850013,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 19291
+    },
+    {
+      "epoch": 0.45920213272398364,
+      "grad_norm": 4.889324167942775,
+      "learning_rate": 2e-05,
+      "loss": 0.2054,
+      "step": 19292
+    },
+    {
+      "epoch": 0.4592259354470151,
+      "grad_norm": 0.24846303864428368,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 19293
+    },
+    {
+      "epoch": 0.45924973817004666,
+      "grad_norm": 9.200051123421026,
+      "learning_rate": 2e-05,
+      "loss": 0.6133,
+      "step": 19294
+    },
+    {
+      "epoch": 0.45927354089307815,
+      "grad_norm": 8.562283764083794,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 19295
+    },
+    {
+      "epoch": 0.4592973436161097,
+      "grad_norm": 5.106581140753367,
+      "learning_rate": 2e-05,
+      "loss": 0.5557,
+      "step": 19296
+    },
+    {
+      "epoch": 0.4593211463391412,
+      "grad_norm": 4.712729045428455,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 19297
+    },
+    {
+      "epoch": 0.4593449490621727,
+      "grad_norm": 4.9798647112356536,
+      "learning_rate": 2e-05,
+      "loss": 0.1861,
+      "step": 19298
+    },
+    {
+      "epoch": 0.45936875178520425,
+      "grad_norm": 12.426305318081615,
+      "learning_rate": 2e-05,
+      "loss": 0.528,
+      "step": 19299
+    },
+    {
+      "epoch": 0.45939255450823574,
+      "grad_norm": 1.3954528031005642,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 19300
+    },
+    {
+      "epoch": 0.4594163572312673,
+      "grad_norm": 0.326753650335874,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19301
+    },
+    {
+      "epoch": 0.45944015995429877,
+      "grad_norm": 3.1165538573403406,
+      "learning_rate": 2e-05,
+      "loss": 0.4468,
+      "step": 19302
+    },
+    {
+      "epoch": 0.4594639626773303,
+      "grad_norm": 3.5580145621034824,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 19303
+    },
+    {
+      "epoch": 0.4594877654003618,
+      "grad_norm": 0.378568759431016,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 19304
+    },
+    {
+      "epoch": 0.45951156812339333,
+      "grad_norm": 2.699198660958236,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 19305
+    },
+    {
+      "epoch": 0.4595353708464248,
+      "grad_norm": 2.0622602963985193,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 19306
+    },
+    {
+      "epoch": 0.45955917356945636,
+      "grad_norm": 6.793866608428655,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 19307
+    },
+    {
+      "epoch": 0.45958297629248784,
+      "grad_norm": 4.037947758321038,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 19308
+    },
+    {
+      "epoch": 0.4596067790155194,
+      "grad_norm": 3.143507168533292,
+      "learning_rate": 2e-05,
+      "loss": 0.3553,
+      "step": 19309
+    },
+    {
+      "epoch": 0.45963058173855087,
+      "grad_norm": 0.28120220972152504,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 19310
+    },
+    {
+      "epoch": 0.4596543844615824,
+      "grad_norm": 4.356901090243843,
+      "learning_rate": 2e-05,
+      "loss": 0.6042,
+      "step": 19311
+    },
+    {
+      "epoch": 0.4596781871846139,
+      "grad_norm": 2.8486103205103963,
+      "learning_rate": 2e-05,
+      "loss": 0.4044,
+      "step": 19312
+    },
+    {
+      "epoch": 0.45970198990764544,
+      "grad_norm": 2.6038050362291183,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 19313
+    },
+    {
+      "epoch": 0.459725792630677,
+      "grad_norm": 3.95430703219021,
+      "learning_rate": 2e-05,
+      "loss": 0.6865,
+      "step": 19314
+    },
+    {
+      "epoch": 0.45974959535370846,
+      "grad_norm": 1.36464939581934,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 19315
+    },
+    {
+      "epoch": 0.45977339807674,
+      "grad_norm": 5.056489917329373,
+      "learning_rate": 2e-05,
+      "loss": 0.8386,
+      "step": 19316
+    },
+    {
+      "epoch": 0.4597972007997715,
+      "grad_norm": 3.9792352808101636,
+      "learning_rate": 2e-05,
+      "loss": 0.4866,
+      "step": 19317
+    },
+    {
+      "epoch": 0.45982100352280303,
+      "grad_norm": 2.758543509316833,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 19318
+    },
+    {
+      "epoch": 0.4598448062458345,
+      "grad_norm": 3.2562580881336407,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 19319
+    },
+    {
+      "epoch": 0.45986860896886605,
+      "grad_norm": 6.020434623929092,
+      "learning_rate": 2e-05,
+      "loss": 0.5688,
+      "step": 19320
+    },
+    {
+      "epoch": 0.45989241169189754,
+      "grad_norm": 6.4606022017593885,
+      "learning_rate": 2e-05,
+      "loss": 0.4604,
+      "step": 19321
+    },
+    {
+      "epoch": 0.4599162144149291,
+      "grad_norm": 10.83536432036688,
+      "learning_rate": 2e-05,
+      "loss": 1.4338,
+      "step": 19322
+    },
+    {
+      "epoch": 0.45994001713796057,
+      "grad_norm": 6.407304291394594,
+      "learning_rate": 2e-05,
+      "loss": 0.7922,
+      "step": 19323
+    },
+    {
+      "epoch": 0.4599638198609921,
+      "grad_norm": 5.614359006007905,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 19324
+    },
+    {
+      "epoch": 0.4599876225840236,
+      "grad_norm": 5.464056052300969,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 19325
+    },
+    {
+      "epoch": 0.46001142530705513,
+      "grad_norm": 2.1770727227458786,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 19326
+    },
+    {
+      "epoch": 0.4600352280300866,
+      "grad_norm": 4.648374551606183,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 19327
+    },
+    {
+      "epoch": 0.46005903075311816,
+      "grad_norm": 15.323201832826355,
+      "learning_rate": 2e-05,
+      "loss": 0.417,
+      "step": 19328
+    },
+    {
+      "epoch": 0.46008283347614964,
+      "grad_norm": 3.686515669510431,
+      "learning_rate": 2e-05,
+      "loss": 0.3898,
+      "step": 19329
+    },
+    {
+      "epoch": 0.4601066361991812,
+      "grad_norm": 4.1419018932817515,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 19330
+    },
+    {
+      "epoch": 0.4601304389222127,
+      "grad_norm": 5.431923754335157,
+      "learning_rate": 2e-05,
+      "loss": 0.8763,
+      "step": 19331
+    },
+    {
+      "epoch": 0.4601542416452442,
+      "grad_norm": 4.817915701110084,
+      "learning_rate": 2e-05,
+      "loss": 0.9139,
+      "step": 19332
+    },
+    {
+      "epoch": 0.46017804436827575,
+      "grad_norm": 4.489725491981951,
+      "learning_rate": 2e-05,
+      "loss": 0.984,
+      "step": 19333
+    },
+    {
+      "epoch": 0.46020184709130724,
+      "grad_norm": 10.092636438438054,
+      "learning_rate": 2e-05,
+      "loss": 1.6107,
+      "step": 19334
+    },
+    {
+      "epoch": 0.4602256498143388,
+      "grad_norm": 2.681782619752717,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 19335
+    },
+    {
+      "epoch": 0.46024945253737026,
+      "grad_norm": 15.279231166141484,
+      "learning_rate": 2e-05,
+      "loss": 0.5181,
+      "step": 19336
+    },
+    {
+      "epoch": 0.4602732552604018,
+      "grad_norm": 5.747615498660707,
+      "learning_rate": 2e-05,
+      "loss": 1.446,
+      "step": 19337
+    },
+    {
+      "epoch": 0.4602970579834333,
+      "grad_norm": 12.252570251706915,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 19338
+    },
+    {
+      "epoch": 0.46032086070646483,
+      "grad_norm": 4.788952529493631,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 19339
+    },
+    {
+      "epoch": 0.4603446634294963,
+      "grad_norm": 3.0385259803000633,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 19340
+    },
+    {
+      "epoch": 0.46036846615252786,
+      "grad_norm": 4.172403906350737,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 19341
+    },
+    {
+      "epoch": 0.46039226887555934,
+      "grad_norm": 5.36547960690171,
+      "learning_rate": 2e-05,
+      "loss": 0.7712,
+      "step": 19342
+    },
+    {
+      "epoch": 0.4604160715985909,
+      "grad_norm": 4.7446307072037825,
+      "learning_rate": 2e-05,
+      "loss": 0.6657,
+      "step": 19343
+    },
+    {
+      "epoch": 0.46043987432162237,
+      "grad_norm": 0.7047988910630052,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 19344
+    },
+    {
+      "epoch": 0.4604636770446539,
+      "grad_norm": 2.201285175098482,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 19345
+    },
+    {
+      "epoch": 0.46048747976768545,
+      "grad_norm": 0.9404173439002591,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 19346
+    },
+    {
+      "epoch": 0.46051128249071693,
+      "grad_norm": 2.941962701193925,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 19347
+    },
+    {
+      "epoch": 0.4605350852137485,
+      "grad_norm": 1.8599552451122328,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 19348
+    },
+    {
+      "epoch": 0.46055888793677996,
+      "grad_norm": 3.578656284709498,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 19349
+    },
+    {
+      "epoch": 0.4605826906598115,
+      "grad_norm": 0.15392592212531248,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 19350
+    },
+    {
+      "epoch": 0.460606493382843,
+      "grad_norm": 5.307472002210028,
+      "learning_rate": 2e-05,
+      "loss": 1.1766,
+      "step": 19351
+    },
+    {
+      "epoch": 0.4606302961058745,
+      "grad_norm": 4.5423540233220985,
+      "learning_rate": 2e-05,
+      "loss": 0.7757,
+      "step": 19352
+    },
+    {
+      "epoch": 0.460654098828906,
+      "grad_norm": 12.87887035852578,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 19353
+    },
+    {
+      "epoch": 0.46067790155193755,
+      "grad_norm": 8.567193940498193,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 19354
+    },
+    {
+      "epoch": 0.46070170427496904,
+      "grad_norm": 4.2092138345818535,
+      "learning_rate": 2e-05,
+      "loss": 0.5114,
+      "step": 19355
+    },
+    {
+      "epoch": 0.4607255069980006,
+      "grad_norm": 1.4253442953824609,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 19356
+    },
+    {
+      "epoch": 0.46074930972103206,
+      "grad_norm": 5.499087558782987,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 19357
+    },
+    {
+      "epoch": 0.4607731124440636,
+      "grad_norm": 6.926663287768703,
+      "learning_rate": 2e-05,
+      "loss": 0.326,
+      "step": 19358
+    },
+    {
+      "epoch": 0.4607969151670951,
+      "grad_norm": 0.29661608165738057,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 19359
+    },
+    {
+      "epoch": 0.46082071789012663,
+      "grad_norm": 6.5131639492748645,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 19360
+    },
+    {
+      "epoch": 0.46084452061315817,
+      "grad_norm": 4.058920362795948,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 19361
+    },
+    {
+      "epoch": 0.46086832333618966,
+      "grad_norm": 9.955570940246467,
+      "learning_rate": 2e-05,
+      "loss": 2.0081,
+      "step": 19362
+    },
+    {
+      "epoch": 0.4608921260592212,
+      "grad_norm": 4.581824928861511,
+      "learning_rate": 2e-05,
+      "loss": 0.3656,
+      "step": 19363
+    },
+    {
+      "epoch": 0.4609159287822527,
+      "grad_norm": 0.8745920307943172,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 19364
+    },
+    {
+      "epoch": 0.4609397315052842,
+      "grad_norm": 4.250364392872236,
+      "learning_rate": 2e-05,
+      "loss": 0.2902,
+      "step": 19365
+    },
+    {
+      "epoch": 0.4609635342283157,
+      "grad_norm": 2.7149902668613555,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 19366
+    },
+    {
+      "epoch": 0.46098733695134725,
+      "grad_norm": 4.575581781256204,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 19367
+    },
+    {
+      "epoch": 0.46101113967437873,
+      "grad_norm": 23.368759790067397,
+      "learning_rate": 2e-05,
+      "loss": 1.3694,
+      "step": 19368
+    },
+    {
+      "epoch": 0.4610349423974103,
+      "grad_norm": 0.45380607881325813,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19369
+    },
+    {
+      "epoch": 0.46105874512044176,
+      "grad_norm": 4.141461987712288,
+      "learning_rate": 2e-05,
+      "loss": 0.4063,
+      "step": 19370
+    },
+    {
+      "epoch": 0.4610825478434733,
+      "grad_norm": 4.1209914875089035,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 19371
+    },
+    {
+      "epoch": 0.4611063505665048,
+      "grad_norm": 3.280891680103675,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 19372
+    },
+    {
+      "epoch": 0.4611301532895363,
+      "grad_norm": 11.252107581543097,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 19373
+    },
+    {
+      "epoch": 0.4611539560125678,
+      "grad_norm": 3.166426128869451,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 19374
+    },
+    {
+      "epoch": 0.46117775873559935,
+      "grad_norm": 3.0786580321151846,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 19375
+    },
+    {
+      "epoch": 0.4612015614586309,
+      "grad_norm": 0.10548756076975661,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 19376
+    },
+    {
+      "epoch": 0.4612253641816624,
+      "grad_norm": 5.8269701892163255,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 19377
+    },
+    {
+      "epoch": 0.4612491669046939,
+      "grad_norm": 0.6327976259498104,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19378
+    },
+    {
+      "epoch": 0.4612729696277254,
+      "grad_norm": 5.80735506927091,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 19379
+    },
+    {
+      "epoch": 0.46129677235075695,
+      "grad_norm": 0.36133305609824934,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 19380
+    },
+    {
+      "epoch": 0.46132057507378843,
+      "grad_norm": 2.7191831063900618,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 19381
+    },
+    {
+      "epoch": 0.46134437779681997,
+      "grad_norm": 3.8998287493648474,
+      "learning_rate": 2e-05,
+      "loss": 0.6785,
+      "step": 19382
+    },
+    {
+      "epoch": 0.46136818051985146,
+      "grad_norm": 1.0500497281005774,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 19383
+    },
+    {
+      "epoch": 0.461391983242883,
+      "grad_norm": 3.7643222533785226,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 19384
+    },
+    {
+      "epoch": 0.4614157859659145,
+      "grad_norm": 4.956960524096148,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 19385
+    },
+    {
+      "epoch": 0.461439588688946,
+      "grad_norm": 5.718695038207914,
+      "learning_rate": 2e-05,
+      "loss": 1.4018,
+      "step": 19386
+    },
+    {
+      "epoch": 0.4614633914119775,
+      "grad_norm": 3.3650427788823296,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 19387
+    },
+    {
+      "epoch": 0.46148719413500905,
+      "grad_norm": 6.938101158593612,
+      "learning_rate": 2e-05,
+      "loss": 0.9691,
+      "step": 19388
+    },
+    {
+      "epoch": 0.46151099685804053,
+      "grad_norm": 1.1919973422804309,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 19389
+    },
+    {
+      "epoch": 0.4615347995810721,
+      "grad_norm": 2.7505298990388156,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 19390
+    },
+    {
+      "epoch": 0.4615586023041036,
+      "grad_norm": 3.1585586174115723,
+      "learning_rate": 2e-05,
+      "loss": 0.4419,
+      "step": 19391
+    },
+    {
+      "epoch": 0.4615824050271351,
+      "grad_norm": 3.5826763982874006,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 19392
+    },
+    {
+      "epoch": 0.46160620775016664,
+      "grad_norm": 18.676456144332594,
+      "learning_rate": 2e-05,
+      "loss": 0.1214,
+      "step": 19393
+    },
+    {
+      "epoch": 0.4616300104731981,
+      "grad_norm": 21.910186655077503,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 19394
+    },
+    {
+      "epoch": 0.46165381319622967,
+      "grad_norm": 12.854385383323157,
+      "learning_rate": 2e-05,
+      "loss": 0.5537,
+      "step": 19395
+    },
+    {
+      "epoch": 0.46167761591926115,
+      "grad_norm": 5.350124221968116,
+      "learning_rate": 2e-05,
+      "loss": 0.7256,
+      "step": 19396
+    },
+    {
+      "epoch": 0.4617014186422927,
+      "grad_norm": 4.856118903846876,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 19397
+    },
+    {
+      "epoch": 0.4617252213653242,
+      "grad_norm": 4.598968493802673,
+      "learning_rate": 2e-05,
+      "loss": 0.2704,
+      "step": 19398
+    },
+    {
+      "epoch": 0.4617490240883557,
+      "grad_norm": 16.57655446242542,
+      "learning_rate": 2e-05,
+      "loss": 0.955,
+      "step": 19399
+    },
+    {
+      "epoch": 0.4617728268113872,
+      "grad_norm": 4.548446783236381,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46179662953441875,
+      "grad_norm": 3.0257420942491073,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 19401
+    },
+    {
+      "epoch": 0.46182043225745023,
+      "grad_norm": 3.4149480357805273,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 19402
+    },
+    {
+      "epoch": 0.46184423498048177,
+      "grad_norm": 1.990968041671735,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 19403
+    },
+    {
+      "epoch": 0.46186803770351326,
+      "grad_norm": 7.937401477222945,
+      "learning_rate": 2e-05,
+      "loss": 0.4282,
+      "step": 19404
+    },
+    {
+      "epoch": 0.4618918404265448,
+      "grad_norm": 5.005989504489861,
+      "learning_rate": 2e-05,
+      "loss": 0.7524,
+      "step": 19405
+    },
+    {
+      "epoch": 0.46191564314957634,
+      "grad_norm": 4.245756526452451,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 19406
+    },
+    {
+      "epoch": 0.4619394458726078,
+      "grad_norm": 4.80033323437061,
+      "learning_rate": 2e-05,
+      "loss": 0.6107,
+      "step": 19407
+    },
+    {
+      "epoch": 0.46196324859563936,
+      "grad_norm": 3.1351770511471515,
+      "learning_rate": 2e-05,
+      "loss": 0.1412,
+      "step": 19408
+    },
+    {
+      "epoch": 0.46198705131867085,
+      "grad_norm": 0.9303854960734543,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19409
+    },
+    {
+      "epoch": 0.4620108540417024,
+      "grad_norm": 13.218337504859244,
+      "learning_rate": 2e-05,
+      "loss": 0.6117,
+      "step": 19410
+    },
+    {
+      "epoch": 0.4620346567647339,
+      "grad_norm": 4.0314505361296415,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 19411
+    },
+    {
+      "epoch": 0.4620584594877654,
+      "grad_norm": 2.9226737543511603,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 19412
+    },
+    {
+      "epoch": 0.4620822622107969,
+      "grad_norm": 7.7561209374361875,
+      "learning_rate": 2e-05,
+      "loss": 0.2976,
+      "step": 19413
+    },
+    {
+      "epoch": 0.46210606493382844,
+      "grad_norm": 0.8907387346115758,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 19414
+    },
+    {
+      "epoch": 0.4621298676568599,
+      "grad_norm": 0.40306760385136786,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 19415
+    },
+    {
+      "epoch": 0.46215367037989147,
+      "grad_norm": 1.469619855722965,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 19416
+    },
+    {
+      "epoch": 0.46217747310292295,
+      "grad_norm": 8.605044907666352,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 19417
+    },
+    {
+      "epoch": 0.4622012758259545,
+      "grad_norm": 1.362785734094455,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 19418
+    },
+    {
+      "epoch": 0.462225078548986,
+      "grad_norm": 10.046416748667738,
+      "learning_rate": 2e-05,
+      "loss": 0.8438,
+      "step": 19419
+    },
+    {
+      "epoch": 0.4622488812720175,
+      "grad_norm": 4.807419495276365,
+      "learning_rate": 2e-05,
+      "loss": 0.6456,
+      "step": 19420
+    },
+    {
+      "epoch": 0.46227268399504906,
+      "grad_norm": 0.525594530341359,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 19421
+    },
+    {
+      "epoch": 0.46229648671808055,
+      "grad_norm": 3.9419884664231915,
+      "learning_rate": 2e-05,
+      "loss": 0.4013,
+      "step": 19422
+    },
+    {
+      "epoch": 0.4623202894411121,
+      "grad_norm": 1.6212949815231594,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 19423
+    },
+    {
+      "epoch": 0.46234409216414357,
+      "grad_norm": 6.058008264385979,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 19424
+    },
+    {
+      "epoch": 0.4623678948871751,
+      "grad_norm": 3.4476322354522715,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 19425
+    },
+    {
+      "epoch": 0.4623916976102066,
+      "grad_norm": 2.271705023927668,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 19426
+    },
+    {
+      "epoch": 0.46241550033323814,
+      "grad_norm": 3.386174759378437,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 19427
+    },
+    {
+      "epoch": 0.4624393030562696,
+      "grad_norm": 2.0440123718207195,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 19428
+    },
+    {
+      "epoch": 0.46246310577930116,
+      "grad_norm": 6.13905463849512,
+      "learning_rate": 2e-05,
+      "loss": 0.9581,
+      "step": 19429
+    },
+    {
+      "epoch": 0.46248690850233265,
+      "grad_norm": 11.048515868573084,
+      "learning_rate": 2e-05,
+      "loss": 1.3076,
+      "step": 19430
+    },
+    {
+      "epoch": 0.4625107112253642,
+      "grad_norm": 23.394921789111745,
+      "learning_rate": 2e-05,
+      "loss": 0.6191,
+      "step": 19431
+    },
+    {
+      "epoch": 0.4625345139483957,
+      "grad_norm": 2.450423512173623,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 19432
+    },
+    {
+      "epoch": 0.4625583166714272,
+      "grad_norm": 5.938385724364788,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 19433
+    },
+    {
+      "epoch": 0.4625821193944587,
+      "grad_norm": 7.268091800281544,
+      "learning_rate": 2e-05,
+      "loss": 0.5115,
+      "step": 19434
+    },
+    {
+      "epoch": 0.46260592211749024,
+      "grad_norm": 5.8272370228453125,
+      "learning_rate": 2e-05,
+      "loss": 0.1989,
+      "step": 19435
+    },
+    {
+      "epoch": 0.4626297248405218,
+      "grad_norm": 1.3028882129136172,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 19436
+    },
+    {
+      "epoch": 0.46265352756355327,
+      "grad_norm": 1.6582307865367927,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 19437
+    },
+    {
+      "epoch": 0.4626773302865848,
+      "grad_norm": 18.566696854738428,
+      "learning_rate": 2e-05,
+      "loss": 0.3673,
+      "step": 19438
+    },
+    {
+      "epoch": 0.4627011330096163,
+      "grad_norm": 5.0517415629010936,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 19439
+    },
+    {
+      "epoch": 0.46272493573264784,
+      "grad_norm": 3.5405982306466224,
+      "learning_rate": 2e-05,
+      "loss": 0.2479,
+      "step": 19440
+    },
+    {
+      "epoch": 0.4627487384556793,
+      "grad_norm": 3.5916977217504686,
+      "learning_rate": 2e-05,
+      "loss": 0.3306,
+      "step": 19441
+    },
+    {
+      "epoch": 0.46277254117871086,
+      "grad_norm": 0.5585875040307084,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19442
+    },
+    {
+      "epoch": 0.46279634390174235,
+      "grad_norm": 3.2078577117521845,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 19443
+    },
+    {
+      "epoch": 0.4628201466247739,
+      "grad_norm": 12.939456980395727,
+      "learning_rate": 2e-05,
+      "loss": 0.9718,
+      "step": 19444
+    },
+    {
+      "epoch": 0.4628439493478054,
+      "grad_norm": 4.830426078880732,
+      "learning_rate": 2e-05,
+      "loss": 0.6681,
+      "step": 19445
+    },
+    {
+      "epoch": 0.4628677520708369,
+      "grad_norm": 2.2832014268209155,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 19446
+    },
+    {
+      "epoch": 0.4628915547938684,
+      "grad_norm": 3.0926991061596922,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 19447
+    },
+    {
+      "epoch": 0.46291535751689994,
+      "grad_norm": 8.410701669142862,
+      "learning_rate": 2e-05,
+      "loss": 1.7379,
+      "step": 19448
+    },
+    {
+      "epoch": 0.4629391602399314,
+      "grad_norm": 2.4761622404092747,
+      "learning_rate": 2e-05,
+      "loss": 0.1532,
+      "step": 19449
+    },
+    {
+      "epoch": 0.46296296296296297,
+      "grad_norm": 8.430155213234594,
+      "learning_rate": 2e-05,
+      "loss": 0.3572,
+      "step": 19450
+    },
+    {
+      "epoch": 0.46298676568599445,
+      "grad_norm": 7.187452577379239,
+      "learning_rate": 2e-05,
+      "loss": 2.3973,
+      "step": 19451
+    },
+    {
+      "epoch": 0.463010568409026,
+      "grad_norm": 4.026717287434214,
+      "learning_rate": 2e-05,
+      "loss": 0.4322,
+      "step": 19452
+    },
+    {
+      "epoch": 0.46303437113205753,
+      "grad_norm": 4.3847769641686165,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 19453
+    },
+    {
+      "epoch": 0.463058173855089,
+      "grad_norm": 7.723124741819983,
+      "learning_rate": 2e-05,
+      "loss": 1.5186,
+      "step": 19454
+    },
+    {
+      "epoch": 0.46308197657812056,
+      "grad_norm": 3.6307951125642854,
+      "learning_rate": 2e-05,
+      "loss": 0.4258,
+      "step": 19455
+    },
+    {
+      "epoch": 0.46310577930115204,
+      "grad_norm": 7.104723690897888,
+      "learning_rate": 2e-05,
+      "loss": 1.2544,
+      "step": 19456
+    },
+    {
+      "epoch": 0.4631295820241836,
+      "grad_norm": 6.189117165503862,
+      "learning_rate": 2e-05,
+      "loss": 1.2293,
+      "step": 19457
+    },
+    {
+      "epoch": 0.46315338474721507,
+      "grad_norm": 4.4866336715419,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 19458
+    },
+    {
+      "epoch": 0.4631771874702466,
+      "grad_norm": 34.05922913875402,
+      "learning_rate": 2e-05,
+      "loss": 0.6612,
+      "step": 19459
+    },
+    {
+      "epoch": 0.4632009901932781,
+      "grad_norm": 5.949800945491866,
+      "learning_rate": 2e-05,
+      "loss": 0.9215,
+      "step": 19460
+    },
+    {
+      "epoch": 0.46322479291630964,
+      "grad_norm": 6.149807857594835,
+      "learning_rate": 2e-05,
+      "loss": 0.206,
+      "step": 19461
+    },
+    {
+      "epoch": 0.4632485956393411,
+      "grad_norm": 14.182070013776327,
+      "learning_rate": 2e-05,
+      "loss": 0.4383,
+      "step": 19462
+    },
+    {
+      "epoch": 0.46327239836237266,
+      "grad_norm": 5.773943642785398,
+      "learning_rate": 2e-05,
+      "loss": 0.8961,
+      "step": 19463
+    },
+    {
+      "epoch": 0.46329620108540415,
+      "grad_norm": 9.097087958283455,
+      "learning_rate": 2e-05,
+      "loss": 0.3069,
+      "step": 19464
+    },
+    {
+      "epoch": 0.4633200038084357,
+      "grad_norm": 0.6294736458998028,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 19465
+    },
+    {
+      "epoch": 0.4633438065314672,
+      "grad_norm": 1.364597819602528,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 19466
+    },
+    {
+      "epoch": 0.4633676092544987,
+      "grad_norm": 3.867582451225257,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 19467
+    },
+    {
+      "epoch": 0.46339141197753025,
+      "grad_norm": 12.73535847059007,
+      "learning_rate": 2e-05,
+      "loss": 0.6374,
+      "step": 19468
+    },
+    {
+      "epoch": 0.46341521470056174,
+      "grad_norm": 6.83351767597025,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 19469
+    },
+    {
+      "epoch": 0.4634390174235933,
+      "grad_norm": 2.2824066454020278,
+      "learning_rate": 2e-05,
+      "loss": 0.1305,
+      "step": 19470
+    },
+    {
+      "epoch": 0.46346282014662477,
+      "grad_norm": 14.422594761860656,
+      "learning_rate": 2e-05,
+      "loss": 0.8121,
+      "step": 19471
+    },
+    {
+      "epoch": 0.4634866228696563,
+      "grad_norm": 1.3307308154999447,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 19472
+    },
+    {
+      "epoch": 0.4635104255926878,
+      "grad_norm": 4.004147941336268,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 19473
+    },
+    {
+      "epoch": 0.46353422831571933,
+      "grad_norm": 3.016363280944889,
+      "learning_rate": 2e-05,
+      "loss": 0.3727,
+      "step": 19474
+    },
+    {
+      "epoch": 0.4635580310387508,
+      "grad_norm": 0.1604755073816976,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19475
+    },
+    {
+      "epoch": 0.46358183376178236,
+      "grad_norm": 4.901127939248915,
+      "learning_rate": 2e-05,
+      "loss": 0.7913,
+      "step": 19476
+    },
+    {
+      "epoch": 0.46360563648481384,
+      "grad_norm": 0.12855310248108254,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19477
+    },
+    {
+      "epoch": 0.4636294392078454,
+      "grad_norm": 36.82938977580494,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 19478
+    },
+    {
+      "epoch": 0.46365324193087687,
+      "grad_norm": 7.964028998558464,
+      "learning_rate": 2e-05,
+      "loss": 1.0888,
+      "step": 19479
+    },
+    {
+      "epoch": 0.4636770446539084,
+      "grad_norm": 4.294907622597051,
+      "learning_rate": 2e-05,
+      "loss": 0.4642,
+      "step": 19480
+    },
+    {
+      "epoch": 0.4637008473769399,
+      "grad_norm": 2.693408788824951,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 19481
+    },
+    {
+      "epoch": 0.46372465009997144,
+      "grad_norm": 1.9622228367416148,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 19482
+    },
+    {
+      "epoch": 0.463748452823003,
+      "grad_norm": 2.9789562705417025,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 19483
+    },
+    {
+      "epoch": 0.46377225554603446,
+      "grad_norm": 5.315826852098384,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 19484
+    },
+    {
+      "epoch": 0.463796058269066,
+      "grad_norm": 4.191518662710218,
+      "learning_rate": 2e-05,
+      "loss": 0.6577,
+      "step": 19485
+    },
+    {
+      "epoch": 0.4638198609920975,
+      "grad_norm": 2.7100283073936637,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 19486
+    },
+    {
+      "epoch": 0.46384366371512903,
+      "grad_norm": 3.75134588745713,
+      "learning_rate": 2e-05,
+      "loss": 0.3782,
+      "step": 19487
+    },
+    {
+      "epoch": 0.4638674664381605,
+      "grad_norm": 5.258018374209451,
+      "learning_rate": 2e-05,
+      "loss": 0.1864,
+      "step": 19488
+    },
+    {
+      "epoch": 0.46389126916119205,
+      "grad_norm": 1.7038925134298495,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 19489
+    },
+    {
+      "epoch": 0.46391507188422354,
+      "grad_norm": 6.097744054871693,
+      "learning_rate": 2e-05,
+      "loss": 0.2977,
+      "step": 19490
+    },
+    {
+      "epoch": 0.4639388746072551,
+      "grad_norm": 7.270712541753676,
+      "learning_rate": 2e-05,
+      "loss": 0.3404,
+      "step": 19491
+    },
+    {
+      "epoch": 0.46396267733028657,
+      "grad_norm": 5.634337531600809,
+      "learning_rate": 2e-05,
+      "loss": 0.7055,
+      "step": 19492
+    },
+    {
+      "epoch": 0.4639864800533181,
+      "grad_norm": 0.7716216385989537,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 19493
+    },
+    {
+      "epoch": 0.4640102827763496,
+      "grad_norm": 4.130846187934423,
+      "learning_rate": 2e-05,
+      "loss": 0.1518,
+      "step": 19494
+    },
+    {
+      "epoch": 0.46403408549938113,
+      "grad_norm": 6.71548180409733,
+      "learning_rate": 2e-05,
+      "loss": 0.8667,
+      "step": 19495
+    },
+    {
+      "epoch": 0.4640578882224126,
+      "grad_norm": 5.755553537483818,
+      "learning_rate": 2e-05,
+      "loss": 0.6071,
+      "step": 19496
+    },
+    {
+      "epoch": 0.46408169094544416,
+      "grad_norm": 0.32968271608300825,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 19497
+    },
+    {
+      "epoch": 0.4641054936684757,
+      "grad_norm": 7.383526005563148,
+      "learning_rate": 2e-05,
+      "loss": 0.2963,
+      "step": 19498
+    },
+    {
+      "epoch": 0.4641292963915072,
+      "grad_norm": 5.861978192878158,
+      "learning_rate": 2e-05,
+      "loss": 1.1747,
+      "step": 19499
+    },
+    {
+      "epoch": 0.4641530991145387,
+      "grad_norm": 3.2967061900620265,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 19500
+    },
+    {
+      "epoch": 0.4641769018375702,
+      "grad_norm": 7.697107139478526,
+      "learning_rate": 2e-05,
+      "loss": 0.2278,
+      "step": 19501
+    },
+    {
+      "epoch": 0.46420070456060175,
+      "grad_norm": 0.1856098848458953,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19502
+    },
+    {
+      "epoch": 0.46422450728363324,
+      "grad_norm": 4.186022504113195,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 19503
+    },
+    {
+      "epoch": 0.4642483100066648,
+      "grad_norm": 0.8320474065076484,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 19504
+    },
+    {
+      "epoch": 0.46427211272969626,
+      "grad_norm": 3.234225072617983,
+      "learning_rate": 2e-05,
+      "loss": 0.5643,
+      "step": 19505
+    },
+    {
+      "epoch": 0.4642959154527278,
+      "grad_norm": 1.1115274087059663,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 19506
+    },
+    {
+      "epoch": 0.4643197181757593,
+      "grad_norm": 3.6581998884926454,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 19507
+    },
+    {
+      "epoch": 0.46434352089879083,
+      "grad_norm": 6.716288487233478,
+      "learning_rate": 2e-05,
+      "loss": 0.9166,
+      "step": 19508
+    },
+    {
+      "epoch": 0.4643673236218223,
+      "grad_norm": 2.7648440939666483,
+      "learning_rate": 2e-05,
+      "loss": 0.4098,
+      "step": 19509
+    },
+    {
+      "epoch": 0.46439112634485386,
+      "grad_norm": 6.638481814297872,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 19510
+    },
+    {
+      "epoch": 0.46441492906788534,
+      "grad_norm": 8.664006415177132,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 19511
+    },
+    {
+      "epoch": 0.4644387317909169,
+      "grad_norm": 9.134480084631605,
+      "learning_rate": 2e-05,
+      "loss": 1.2792,
+      "step": 19512
+    },
+    {
+      "epoch": 0.4644625345139484,
+      "grad_norm": 5.269368997189073,
+      "learning_rate": 2e-05,
+      "loss": 0.9929,
+      "step": 19513
+    },
+    {
+      "epoch": 0.4644863372369799,
+      "grad_norm": 7.602817198306064,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 19514
+    },
+    {
+      "epoch": 0.46451013996001145,
+      "grad_norm": 0.597761262128072,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 19515
+    },
+    {
+      "epoch": 0.46453394268304293,
+      "grad_norm": 8.159935539439477,
+      "learning_rate": 2e-05,
+      "loss": 1.2677,
+      "step": 19516
+    },
+    {
+      "epoch": 0.4645577454060745,
+      "grad_norm": 0.10337519366450514,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 19517
+    },
+    {
+      "epoch": 0.46458154812910596,
+      "grad_norm": 1.0586941632236138,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 19518
+    },
+    {
+      "epoch": 0.4646053508521375,
+      "grad_norm": 2.552497447109216,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 19519
+    },
+    {
+      "epoch": 0.464629153575169,
+      "grad_norm": 6.945206743147412,
+      "learning_rate": 2e-05,
+      "loss": 0.4053,
+      "step": 19520
+    },
+    {
+      "epoch": 0.4646529562982005,
+      "grad_norm": 1.0015099462430175,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 19521
+    },
+    {
+      "epoch": 0.464676759021232,
+      "grad_norm": 3.325543303383561,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 19522
+    },
+    {
+      "epoch": 0.46470056174426355,
+      "grad_norm": 2.640785474424445,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 19523
+    },
+    {
+      "epoch": 0.46472436446729504,
+      "grad_norm": 2.574028964307752,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 19524
+    },
+    {
+      "epoch": 0.4647481671903266,
+      "grad_norm": 5.77840838113377,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 19525
+    },
+    {
+      "epoch": 0.46477196991335806,
+      "grad_norm": 1.700678994133326,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 19526
+    },
+    {
+      "epoch": 0.4647957726363896,
+      "grad_norm": 7.13221690428107,
+      "learning_rate": 2e-05,
+      "loss": 1.7947,
+      "step": 19527
+    },
+    {
+      "epoch": 0.46481957535942114,
+      "grad_norm": 5.246072932371724,
+      "learning_rate": 2e-05,
+      "loss": 1.0824,
+      "step": 19528
+    },
+    {
+      "epoch": 0.46484337808245263,
+      "grad_norm": 3.272926899616775,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 19529
+    },
+    {
+      "epoch": 0.46486718080548417,
+      "grad_norm": 3.4593887424206997,
+      "learning_rate": 2e-05,
+      "loss": 0.3221,
+      "step": 19530
+    },
+    {
+      "epoch": 0.46489098352851566,
+      "grad_norm": 2.6833958473053747,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 19531
+    },
+    {
+      "epoch": 0.4649147862515472,
+      "grad_norm": 1.5718342696302747,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 19532
+    },
+    {
+      "epoch": 0.4649385889745787,
+      "grad_norm": 8.286702029777947,
+      "learning_rate": 2e-05,
+      "loss": 2.5388,
+      "step": 19533
+    },
+    {
+      "epoch": 0.4649623916976102,
+      "grad_norm": 2.0428391217748225,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 19534
+    },
+    {
+      "epoch": 0.4649861944206417,
+      "grad_norm": 4.062435701153617,
+      "learning_rate": 2e-05,
+      "loss": 0.4329,
+      "step": 19535
+    },
+    {
+      "epoch": 0.46500999714367325,
+      "grad_norm": 3.42778294403389,
+      "learning_rate": 2e-05,
+      "loss": 0.0926,
+      "step": 19536
+    },
+    {
+      "epoch": 0.46503379986670473,
+      "grad_norm": 7.281991667589919,
+      "learning_rate": 2e-05,
+      "loss": 0.2503,
+      "step": 19537
+    },
+    {
+      "epoch": 0.4650576025897363,
+      "grad_norm": 1.2002152295461295,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 19538
+    },
+    {
+      "epoch": 0.46508140531276776,
+      "grad_norm": 5.2280424715671385,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 19539
+    },
+    {
+      "epoch": 0.4651052080357993,
+      "grad_norm": 29.429515387020555,
+      "learning_rate": 2e-05,
+      "loss": 1.8113,
+      "step": 19540
+    },
+    {
+      "epoch": 0.4651290107588308,
+      "grad_norm": 7.841771275949797,
+      "learning_rate": 2e-05,
+      "loss": 0.2433,
+      "step": 19541
+    },
+    {
+      "epoch": 0.4651528134818623,
+      "grad_norm": 5.871227754894645,
+      "learning_rate": 2e-05,
+      "loss": 0.9921,
+      "step": 19542
+    },
+    {
+      "epoch": 0.46517661620489387,
+      "grad_norm": 6.521907909846726,
+      "learning_rate": 2e-05,
+      "loss": 0.8591,
+      "step": 19543
+    },
+    {
+      "epoch": 0.46520041892792535,
+      "grad_norm": 10.830585552689705,
+      "learning_rate": 2e-05,
+      "loss": 1.0773,
+      "step": 19544
+    },
+    {
+      "epoch": 0.4652242216509569,
+      "grad_norm": 4.985442805157104,
+      "learning_rate": 2e-05,
+      "loss": 0.7644,
+      "step": 19545
+    },
+    {
+      "epoch": 0.4652480243739884,
+      "grad_norm": 6.2744454531314755,
+      "learning_rate": 2e-05,
+      "loss": 1.1983,
+      "step": 19546
+    },
+    {
+      "epoch": 0.4652718270970199,
+      "grad_norm": 11.968705871357999,
+      "learning_rate": 2e-05,
+      "loss": 0.5307,
+      "step": 19547
+    },
+    {
+      "epoch": 0.4652956298200514,
+      "grad_norm": 0.07102369945891758,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 19548
+    },
+    {
+      "epoch": 0.46531943254308294,
+      "grad_norm": 5.751732807613154,
+      "learning_rate": 2e-05,
+      "loss": 0.71,
+      "step": 19549
+    },
+    {
+      "epoch": 0.46534323526611443,
+      "grad_norm": 12.00096768486898,
+      "learning_rate": 2e-05,
+      "loss": 1.7491,
+      "step": 19550
+    },
+    {
+      "epoch": 0.46536703798914597,
+      "grad_norm": 4.016612600018623,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 19551
+    },
+    {
+      "epoch": 0.46539084071217746,
+      "grad_norm": 6.080988242894101,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 19552
+    },
+    {
+      "epoch": 0.465414643435209,
+      "grad_norm": 1.9062471573998787,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 19553
+    },
+    {
+      "epoch": 0.4654384461582405,
+      "grad_norm": 6.376050513850709,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 19554
+    },
+    {
+      "epoch": 0.465462248881272,
+      "grad_norm": 3.815029298793449,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 19555
+    },
+    {
+      "epoch": 0.4654860516043035,
+      "grad_norm": 4.934621425935576,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 19556
+    },
+    {
+      "epoch": 0.46550985432733505,
+      "grad_norm": 2.1527289325498327,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 19557
+    },
+    {
+      "epoch": 0.4655336570503666,
+      "grad_norm": 6.010330324550447,
+      "learning_rate": 2e-05,
+      "loss": 0.5839,
+      "step": 19558
+    },
+    {
+      "epoch": 0.4655574597733981,
+      "grad_norm": 11.709713981899935,
+      "learning_rate": 2e-05,
+      "loss": 1.1425,
+      "step": 19559
+    },
+    {
+      "epoch": 0.4655812624964296,
+      "grad_norm": 6.133119565640077,
+      "learning_rate": 2e-05,
+      "loss": 0.2868,
+      "step": 19560
+    },
+    {
+      "epoch": 0.4656050652194611,
+      "grad_norm": 17.159110124018692,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 19561
+    },
+    {
+      "epoch": 0.46562886794249264,
+      "grad_norm": 0.7288910706470138,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 19562
+    },
+    {
+      "epoch": 0.4656526706655241,
+      "grad_norm": 8.557241884555893,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 19563
+    },
+    {
+      "epoch": 0.46567647338855567,
+      "grad_norm": 9.98667250778765,
+      "learning_rate": 2e-05,
+      "loss": 0.9459,
+      "step": 19564
+    },
+    {
+      "epoch": 0.46570027611158715,
+      "grad_norm": 10.33391420300665,
+      "learning_rate": 2e-05,
+      "loss": 1.1105,
+      "step": 19565
+    },
+    {
+      "epoch": 0.4657240788346187,
+      "grad_norm": 0.5094579870095682,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 19566
+    },
+    {
+      "epoch": 0.4657478815576502,
+      "grad_norm": 0.4293499577533931,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 19567
+    },
+    {
+      "epoch": 0.4657716842806817,
+      "grad_norm": 2.5808785523415922,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 19568
+    },
+    {
+      "epoch": 0.4657954870037132,
+      "grad_norm": 2.361639346828194,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 19569
+    },
+    {
+      "epoch": 0.46581928972674475,
+      "grad_norm": 0.7752778982921161,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 19570
+    },
+    {
+      "epoch": 0.46584309244977623,
+      "grad_norm": 2.6327920698383154,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 19571
+    },
+    {
+      "epoch": 0.46586689517280777,
+      "grad_norm": 0.9226770126372993,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 19572
+    },
+    {
+      "epoch": 0.46589069789583926,
+      "grad_norm": 6.110455319594044,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 19573
+    },
+    {
+      "epoch": 0.4659145006188708,
+      "grad_norm": 4.965024862871417,
+      "learning_rate": 2e-05,
+      "loss": 0.7746,
+      "step": 19574
+    },
+    {
+      "epoch": 0.46593830334190234,
+      "grad_norm": 6.822026258340525,
+      "learning_rate": 2e-05,
+      "loss": 1.7281,
+      "step": 19575
+    },
+    {
+      "epoch": 0.4659621060649338,
+      "grad_norm": 3.0730584742301366,
+      "learning_rate": 2e-05,
+      "loss": 0.4537,
+      "step": 19576
+    },
+    {
+      "epoch": 0.46598590878796536,
+      "grad_norm": 10.07802466993802,
+      "learning_rate": 2e-05,
+      "loss": 1.4924,
+      "step": 19577
+    },
+    {
+      "epoch": 0.46600971151099685,
+      "grad_norm": 1.2545322415558016,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 19578
+    },
+    {
+      "epoch": 0.4660335142340284,
+      "grad_norm": 5.07676739152127,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 19579
+    },
+    {
+      "epoch": 0.4660573169570599,
+      "grad_norm": 5.189237237360762,
+      "learning_rate": 2e-05,
+      "loss": 0.2825,
+      "step": 19580
+    },
+    {
+      "epoch": 0.4660811196800914,
+      "grad_norm": 5.547032819741592,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 19581
+    },
+    {
+      "epoch": 0.4661049224031229,
+      "grad_norm": 2.4836912640918394,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 19582
+    },
+    {
+      "epoch": 0.46612872512615444,
+      "grad_norm": 3.8799720839286036,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 19583
+    },
+    {
+      "epoch": 0.4661525278491859,
+      "grad_norm": 8.746264389147422,
+      "learning_rate": 2e-05,
+      "loss": 0.8677,
+      "step": 19584
+    },
+    {
+      "epoch": 0.46617633057221747,
+      "grad_norm": 7.062208541802481,
+      "learning_rate": 2e-05,
+      "loss": 0.8838,
+      "step": 19585
+    },
+    {
+      "epoch": 0.46620013329524895,
+      "grad_norm": 6.312834817806167,
+      "learning_rate": 2e-05,
+      "loss": 0.834,
+      "step": 19586
+    },
+    {
+      "epoch": 0.4662239360182805,
+      "grad_norm": 1.183062759960415,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 19587
+    },
+    {
+      "epoch": 0.466247738741312,
+      "grad_norm": 0.7406353275683153,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 19588
+    },
+    {
+      "epoch": 0.4662715414643435,
+      "grad_norm": 3.850366793204947,
+      "learning_rate": 2e-05,
+      "loss": 0.8573,
+      "step": 19589
+    },
+    {
+      "epoch": 0.46629534418737506,
+      "grad_norm": 4.984783259268622,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 19590
+    },
+    {
+      "epoch": 0.46631914691040655,
+      "grad_norm": 0.11548580910060693,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 19591
+    },
+    {
+      "epoch": 0.4663429496334381,
+      "grad_norm": 0.4753531255472157,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 19592
+    },
+    {
+      "epoch": 0.46636675235646957,
+      "grad_norm": 6.020904445163524,
+      "learning_rate": 2e-05,
+      "loss": 1.0016,
+      "step": 19593
+    },
+    {
+      "epoch": 0.4663905550795011,
+      "grad_norm": 4.186779010457535,
+      "learning_rate": 2e-05,
+      "loss": 0.6342,
+      "step": 19594
+    },
+    {
+      "epoch": 0.4664143578025326,
+      "grad_norm": 2.101032882432897,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 19595
+    },
+    {
+      "epoch": 0.46643816052556414,
+      "grad_norm": 4.523908918865261,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 19596
+    },
+    {
+      "epoch": 0.4664619632485956,
+      "grad_norm": 16.355110427671942,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 19597
+    },
+    {
+      "epoch": 0.46648576597162716,
+      "grad_norm": 1.9524334119828646,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 19598
+    },
+    {
+      "epoch": 0.46650956869465865,
+      "grad_norm": 0.24253113100129034,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19599
+    },
+    {
+      "epoch": 0.4665333714176902,
+      "grad_norm": 4.848336627164898,
+      "learning_rate": 2e-05,
+      "loss": 0.4099,
+      "step": 19600
+    },
+    {
+      "epoch": 0.4665571741407217,
+      "grad_norm": 6.5705232876514446,
+      "learning_rate": 2e-05,
+      "loss": 1.5155,
+      "step": 19601
+    },
+    {
+      "epoch": 0.4665809768637532,
+      "grad_norm": 10.60875347295474,
+      "learning_rate": 2e-05,
+      "loss": 0.6761,
+      "step": 19602
+    },
+    {
+      "epoch": 0.4666047795867847,
+      "grad_norm": 3.3739364625202413,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 19603
+    },
+    {
+      "epoch": 0.46662858230981624,
+      "grad_norm": 0.46379496313382673,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 19604
+    },
+    {
+      "epoch": 0.4666523850328478,
+      "grad_norm": 3.0043710999352005,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 19605
+    },
+    {
+      "epoch": 0.46667618775587927,
+      "grad_norm": 1.2433255641061893,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 19606
+    },
+    {
+      "epoch": 0.4666999904789108,
+      "grad_norm": 2.8006863885503837,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 19607
+    },
+    {
+      "epoch": 0.4667237932019423,
+      "grad_norm": 7.1850023683875355,
+      "learning_rate": 2e-05,
+      "loss": 0.3382,
+      "step": 19608
+    },
+    {
+      "epoch": 0.46674759592497383,
+      "grad_norm": 6.318854878231715,
+      "learning_rate": 2e-05,
+      "loss": 0.2555,
+      "step": 19609
+    },
+    {
+      "epoch": 0.4667713986480053,
+      "grad_norm": 0.2152161044690487,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 19610
+    },
+    {
+      "epoch": 0.46679520137103686,
+      "grad_norm": 2.8934548047400486,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 19611
+    },
+    {
+      "epoch": 0.46681900409406835,
+      "grad_norm": 4.59860101363547,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 19612
+    },
+    {
+      "epoch": 0.4668428068170999,
+      "grad_norm": 6.88950991322544,
+      "learning_rate": 2e-05,
+      "loss": 1.1342,
+      "step": 19613
+    },
+    {
+      "epoch": 0.46686660954013137,
+      "grad_norm": 1.8014826322305075,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 19614
+    },
+    {
+      "epoch": 0.4668904122631629,
+      "grad_norm": 3.6090763697293022,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 19615
+    },
+    {
+      "epoch": 0.4669142149861944,
+      "grad_norm": 6.72250860086663,
+      "learning_rate": 2e-05,
+      "loss": 0.2212,
+      "step": 19616
+    },
+    {
+      "epoch": 0.46693801770922594,
+      "grad_norm": 5.597852671707399,
+      "learning_rate": 2e-05,
+      "loss": 0.9057,
+      "step": 19617
+    },
+    {
+      "epoch": 0.4669618204322574,
+      "grad_norm": 2.450123611611835,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 19618
+    },
+    {
+      "epoch": 0.46698562315528896,
+      "grad_norm": 6.619507206294688,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 19619
+    },
+    {
+      "epoch": 0.4670094258783205,
+      "grad_norm": 4.5537117573081956,
+      "learning_rate": 2e-05,
+      "loss": 0.8518,
+      "step": 19620
+    },
+    {
+      "epoch": 0.467033228601352,
+      "grad_norm": 4.788043962252985,
+      "learning_rate": 2e-05,
+      "loss": 0.7682,
+      "step": 19621
+    },
+    {
+      "epoch": 0.46705703132438353,
+      "grad_norm": 3.623554794759022,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 19622
+    },
+    {
+      "epoch": 0.467080834047415,
+      "grad_norm": 6.151440900550127,
+      "learning_rate": 2e-05,
+      "loss": 0.6094,
+      "step": 19623
+    },
+    {
+      "epoch": 0.46710463677044656,
+      "grad_norm": 8.40671976718453,
+      "learning_rate": 2e-05,
+      "loss": 0.8909,
+      "step": 19624
+    },
+    {
+      "epoch": 0.46712843949347804,
+      "grad_norm": 4.700941117919306,
+      "learning_rate": 2e-05,
+      "loss": 0.6259,
+      "step": 19625
+    },
+    {
+      "epoch": 0.4671522422165096,
+      "grad_norm": 2.6525396062983506,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 19626
+    },
+    {
+      "epoch": 0.46717604493954107,
+      "grad_norm": 3.6596533090417633,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 19627
+    },
+    {
+      "epoch": 0.4671998476625726,
+      "grad_norm": 4.800685503890731,
+      "learning_rate": 2e-05,
+      "loss": 0.71,
+      "step": 19628
+    },
+    {
+      "epoch": 0.4672236503856041,
+      "grad_norm": 13.020671027685509,
+      "learning_rate": 2e-05,
+      "loss": 0.5783,
+      "step": 19629
+    },
+    {
+      "epoch": 0.46724745310863564,
+      "grad_norm": 8.296931153829288,
+      "learning_rate": 2e-05,
+      "loss": 0.355,
+      "step": 19630
+    },
+    {
+      "epoch": 0.4672712558316671,
+      "grad_norm": 0.5302289257735397,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 19631
+    },
+    {
+      "epoch": 0.46729505855469866,
+      "grad_norm": 3.8236895012011876,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 19632
+    },
+    {
+      "epoch": 0.46731886127773015,
+      "grad_norm": 9.98889638032462,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 19633
+    },
+    {
+      "epoch": 0.4673426640007617,
+      "grad_norm": 9.258019785229992,
+      "learning_rate": 2e-05,
+      "loss": 0.5433,
+      "step": 19634
+    },
+    {
+      "epoch": 0.46736646672379323,
+      "grad_norm": 7.757617222735708,
+      "learning_rate": 2e-05,
+      "loss": 0.4451,
+      "step": 19635
+    },
+    {
+      "epoch": 0.4673902694468247,
+      "grad_norm": 10.634635965578159,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 19636
+    },
+    {
+      "epoch": 0.46741407216985625,
+      "grad_norm": 3.667668611615725,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 19637
+    },
+    {
+      "epoch": 0.46743787489288774,
+      "grad_norm": 0.05447038942240793,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 19638
+    },
+    {
+      "epoch": 0.4674616776159193,
+      "grad_norm": 5.674063546520491,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 19639
+    },
+    {
+      "epoch": 0.46748548033895077,
+      "grad_norm": 1.2154907864584938,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 19640
+    },
+    {
+      "epoch": 0.4675092830619823,
+      "grad_norm": 4.2235080087414785,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 19641
+    },
+    {
+      "epoch": 0.4675330857850138,
+      "grad_norm": 6.138715708423809,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 19642
+    },
+    {
+      "epoch": 0.46755688850804533,
+      "grad_norm": 4.8056574343627165,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 19643
+    },
+    {
+      "epoch": 0.4675806912310768,
+      "grad_norm": 5.246547859417851,
+      "learning_rate": 2e-05,
+      "loss": 0.6201,
+      "step": 19644
+    },
+    {
+      "epoch": 0.46760449395410836,
+      "grad_norm": 1.8918474588343883,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 19645
+    },
+    {
+      "epoch": 0.46762829667713984,
+      "grad_norm": 2.003341631766806,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 19646
+    },
+    {
+      "epoch": 0.4676520994001714,
+      "grad_norm": 1.1529219201249281,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 19647
+    },
+    {
+      "epoch": 0.46767590212320287,
+      "grad_norm": 5.899818031210471,
+      "learning_rate": 2e-05,
+      "loss": 0.7516,
+      "step": 19648
+    },
+    {
+      "epoch": 0.4676997048462344,
+      "grad_norm": 3.822879951895017,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 19649
+    },
+    {
+      "epoch": 0.46772350756926595,
+      "grad_norm": 1.0790892632178533,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 19650
+    },
+    {
+      "epoch": 0.46774731029229744,
+      "grad_norm": 6.510498481279976,
+      "learning_rate": 2e-05,
+      "loss": 0.219,
+      "step": 19651
+    },
+    {
+      "epoch": 0.467771113015329,
+      "grad_norm": 1.0767874827348438,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 19652
+    },
+    {
+      "epoch": 0.46779491573836046,
+      "grad_norm": 0.46868980970165164,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 19653
+    },
+    {
+      "epoch": 0.467818718461392,
+      "grad_norm": 0.26432028707713945,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 19654
+    },
+    {
+      "epoch": 0.4678425211844235,
+      "grad_norm": 7.87164039714678,
+      "learning_rate": 2e-05,
+      "loss": 0.2845,
+      "step": 19655
+    },
+    {
+      "epoch": 0.46786632390745503,
+      "grad_norm": 3.251540777823137,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 19656
+    },
+    {
+      "epoch": 0.4678901266304865,
+      "grad_norm": 9.114034452636512,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 19657
+    },
+    {
+      "epoch": 0.46791392935351805,
+      "grad_norm": 3.670030088757807,
+      "learning_rate": 2e-05,
+      "loss": 0.4329,
+      "step": 19658
+    },
+    {
+      "epoch": 0.46793773207654954,
+      "grad_norm": 5.861912452757241,
+      "learning_rate": 2e-05,
+      "loss": 1.0301,
+      "step": 19659
+    },
+    {
+      "epoch": 0.4679615347995811,
+      "grad_norm": 1.6746824275237482,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 19660
+    },
+    {
+      "epoch": 0.46798533752261257,
+      "grad_norm": 2.203796203919226,
+      "learning_rate": 2e-05,
+      "loss": 0.2107,
+      "step": 19661
+    },
+    {
+      "epoch": 0.4680091402456441,
+      "grad_norm": 2.705922520361847,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 19662
+    },
+    {
+      "epoch": 0.4680329429686756,
+      "grad_norm": 4.617348260355858,
+      "learning_rate": 2e-05,
+      "loss": 0.1727,
+      "step": 19663
+    },
+    {
+      "epoch": 0.46805674569170713,
+      "grad_norm": 6.4181152164077275,
+      "learning_rate": 2e-05,
+      "loss": 1.402,
+      "step": 19664
+    },
+    {
+      "epoch": 0.4680805484147387,
+      "grad_norm": 20.236045379981093,
+      "learning_rate": 2e-05,
+      "loss": 0.547,
+      "step": 19665
+    },
+    {
+      "epoch": 0.46810435113777016,
+      "grad_norm": 7.729575738910128,
+      "learning_rate": 2e-05,
+      "loss": 1.1537,
+      "step": 19666
+    },
+    {
+      "epoch": 0.4681281538608017,
+      "grad_norm": 5.5907299024557675,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 19667
+    },
+    {
+      "epoch": 0.4681519565838332,
+      "grad_norm": 10.045817750006735,
+      "learning_rate": 2e-05,
+      "loss": 0.4218,
+      "step": 19668
+    },
+    {
+      "epoch": 0.4681757593068647,
+      "grad_norm": 2.551380895117395,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 19669
+    },
+    {
+      "epoch": 0.4681995620298962,
+      "grad_norm": 12.932112242844578,
+      "learning_rate": 2e-05,
+      "loss": 0.9088,
+      "step": 19670
+    },
+    {
+      "epoch": 0.46822336475292775,
+      "grad_norm": 4.300373394457786,
+      "learning_rate": 2e-05,
+      "loss": 0.0713,
+      "step": 19671
+    },
+    {
+      "epoch": 0.46824716747595924,
+      "grad_norm": 2.7722814883999165,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 19672
+    },
+    {
+      "epoch": 0.4682709701989908,
+      "grad_norm": 1.831128331226862,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 19673
+    },
+    {
+      "epoch": 0.46829477292202226,
+      "grad_norm": 2.9847274880318984,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 19674
+    },
+    {
+      "epoch": 0.4683185756450538,
+      "grad_norm": 5.3694535787263575,
+      "learning_rate": 2e-05,
+      "loss": 0.4062,
+      "step": 19675
+    },
+    {
+      "epoch": 0.4683423783680853,
+      "grad_norm": 10.172319659706913,
+      "learning_rate": 2e-05,
+      "loss": 0.3977,
+      "step": 19676
+    },
+    {
+      "epoch": 0.46836618109111683,
+      "grad_norm": 8.262508710615307,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 19677
+    },
+    {
+      "epoch": 0.4683899838141483,
+      "grad_norm": 6.706852876278227,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 19678
+    },
+    {
+      "epoch": 0.46841378653717985,
+      "grad_norm": 16.814762784015784,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 19679
+    },
+    {
+      "epoch": 0.4684375892602114,
+      "grad_norm": 9.981812028106436,
+      "learning_rate": 2e-05,
+      "loss": 0.4225,
+      "step": 19680
+    },
+    {
+      "epoch": 0.4684613919832429,
+      "grad_norm": 4.976291760438961,
+      "learning_rate": 2e-05,
+      "loss": 0.2207,
+      "step": 19681
+    },
+    {
+      "epoch": 0.4684851947062744,
+      "grad_norm": 2.030026308155832,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 19682
+    },
+    {
+      "epoch": 0.4685089974293059,
+      "grad_norm": 0.7188053055229185,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 19683
+    },
+    {
+      "epoch": 0.46853280015233745,
+      "grad_norm": 3.967888738902335,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 19684
+    },
+    {
+      "epoch": 0.46855660287536893,
+      "grad_norm": 1.330365834449569,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 19685
+    },
+    {
+      "epoch": 0.4685804055984005,
+      "grad_norm": 2.661826080609403,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 19686
+    },
+    {
+      "epoch": 0.46860420832143196,
+      "grad_norm": 1.9513653795746977,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 19687
+    },
+    {
+      "epoch": 0.4686280110444635,
+      "grad_norm": 4.438288744548106,
+      "learning_rate": 2e-05,
+      "loss": 0.7967,
+      "step": 19688
+    },
+    {
+      "epoch": 0.468651813767495,
+      "grad_norm": 7.4391594069122835,
+      "learning_rate": 2e-05,
+      "loss": 0.9758,
+      "step": 19689
+    },
+    {
+      "epoch": 0.4686756164905265,
+      "grad_norm": 3.945479901397376,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 19690
+    },
+    {
+      "epoch": 0.468699419213558,
+      "grad_norm": 5.97556804990972,
+      "learning_rate": 2e-05,
+      "loss": 1.1167,
+      "step": 19691
+    },
+    {
+      "epoch": 0.46872322193658955,
+      "grad_norm": 6.267609171869162,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 19692
+    },
+    {
+      "epoch": 0.46874702465962104,
+      "grad_norm": 1.7498644498862417,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 19693
+    },
+    {
+      "epoch": 0.4687708273826526,
+      "grad_norm": 0.2918081502493311,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 19694
+    },
+    {
+      "epoch": 0.46879463010568406,
+      "grad_norm": 4.383083052944278,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 19695
+    },
+    {
+      "epoch": 0.4688184328287156,
+      "grad_norm": 0.3281941084511699,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 19696
+    },
+    {
+      "epoch": 0.46884223555174714,
+      "grad_norm": 6.9481897871389915,
+      "learning_rate": 2e-05,
+      "loss": 0.8074,
+      "step": 19697
+    },
+    {
+      "epoch": 0.46886603827477863,
+      "grad_norm": 2.311563718146943,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 19698
+    },
+    {
+      "epoch": 0.46888984099781017,
+      "grad_norm": 3.210710485176616,
+      "learning_rate": 2e-05,
+      "loss": 0.2985,
+      "step": 19699
+    },
+    {
+      "epoch": 0.46891364372084166,
+      "grad_norm": 4.200534260611009,
+      "learning_rate": 2e-05,
+      "loss": 0.3465,
+      "step": 19700
+    },
+    {
+      "epoch": 0.4689374464438732,
+      "grad_norm": 10.806983860963857,
+      "learning_rate": 2e-05,
+      "loss": 0.7006,
+      "step": 19701
+    },
+    {
+      "epoch": 0.4689612491669047,
+      "grad_norm": 5.099956947131068,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 19702
+    },
+    {
+      "epoch": 0.4689850518899362,
+      "grad_norm": 3.7222695021666885,
+      "learning_rate": 2e-05,
+      "loss": 0.7203,
+      "step": 19703
+    },
+    {
+      "epoch": 0.4690088546129677,
+      "grad_norm": 7.461236618940072,
+      "learning_rate": 2e-05,
+      "loss": 0.6955,
+      "step": 19704
+    },
+    {
+      "epoch": 0.46903265733599925,
+      "grad_norm": 3.465799504041324,
+      "learning_rate": 2e-05,
+      "loss": 0.2145,
+      "step": 19705
+    },
+    {
+      "epoch": 0.46905646005903073,
+      "grad_norm": 10.4587732432383,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 19706
+    },
+    {
+      "epoch": 0.4690802627820623,
+      "grad_norm": 3.2115789935208237,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 19707
+    },
+    {
+      "epoch": 0.46910406550509376,
+      "grad_norm": 3.3766667485803863,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 19708
+    },
+    {
+      "epoch": 0.4691278682281253,
+      "grad_norm": 6.571596408099376,
+      "learning_rate": 2e-05,
+      "loss": 1.2936,
+      "step": 19709
+    },
+    {
+      "epoch": 0.4691516709511568,
+      "grad_norm": 5.143267095422717,
+      "learning_rate": 2e-05,
+      "loss": 0.6977,
+      "step": 19710
+    },
+    {
+      "epoch": 0.4691754736741883,
+      "grad_norm": 5.444612934446556,
+      "learning_rate": 2e-05,
+      "loss": 1.055,
+      "step": 19711
+    },
+    {
+      "epoch": 0.46919927639721987,
+      "grad_norm": 3.3864006935550557,
+      "learning_rate": 2e-05,
+      "loss": 0.5708,
+      "step": 19712
+    },
+    {
+      "epoch": 0.46922307912025135,
+      "grad_norm": 2.4481565953180047,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 19713
+    },
+    {
+      "epoch": 0.4692468818432829,
+      "grad_norm": 7.594362304986012,
+      "learning_rate": 2e-05,
+      "loss": 1.2241,
+      "step": 19714
+    },
+    {
+      "epoch": 0.4692706845663144,
+      "grad_norm": 4.755752268568675,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 19715
+    },
+    {
+      "epoch": 0.4692944872893459,
+      "grad_norm": 6.6138291018371955,
+      "learning_rate": 2e-05,
+      "loss": 1.6822,
+      "step": 19716
+    },
+    {
+      "epoch": 0.4693182900123774,
+      "grad_norm": 7.270579935608031,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 19717
+    },
+    {
+      "epoch": 0.46934209273540894,
+      "grad_norm": 5.191178364114117,
+      "learning_rate": 2e-05,
+      "loss": 0.9453,
+      "step": 19718
+    },
+    {
+      "epoch": 0.46936589545844043,
+      "grad_norm": 3.5169988353537973,
+      "learning_rate": 2e-05,
+      "loss": 0.7548,
+      "step": 19719
+    },
+    {
+      "epoch": 0.46938969818147197,
+      "grad_norm": 4.457386968149327,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 19720
+    },
+    {
+      "epoch": 0.46941350090450346,
+      "grad_norm": 1.4852218641845638,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 19721
+    },
+    {
+      "epoch": 0.469437303627535,
+      "grad_norm": 17.388331794242568,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 19722
+    },
+    {
+      "epoch": 0.4694611063505665,
+      "grad_norm": 5.117353540497618,
+      "learning_rate": 2e-05,
+      "loss": 0.8852,
+      "step": 19723
+    },
+    {
+      "epoch": 0.469484909073598,
+      "grad_norm": 0.5855451938980963,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 19724
+    },
+    {
+      "epoch": 0.4695087117966295,
+      "grad_norm": 5.027211501898034,
+      "learning_rate": 2e-05,
+      "loss": 1.1925,
+      "step": 19725
+    },
+    {
+      "epoch": 0.46953251451966105,
+      "grad_norm": 4.200732236885367,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 19726
+    },
+    {
+      "epoch": 0.4695563172426926,
+      "grad_norm": 6.001149836418022,
+      "learning_rate": 2e-05,
+      "loss": 0.5257,
+      "step": 19727
+    },
+    {
+      "epoch": 0.4695801199657241,
+      "grad_norm": 1.7304421953910218,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 19728
+    },
+    {
+      "epoch": 0.4696039226887556,
+      "grad_norm": 4.011775135211699,
+      "learning_rate": 2e-05,
+      "loss": 0.5696,
+      "step": 19729
+    },
+    {
+      "epoch": 0.4696277254117871,
+      "grad_norm": 1.7866080729980138,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 19730
+    },
+    {
+      "epoch": 0.46965152813481864,
+      "grad_norm": 0.23637508476685362,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 19731
+    },
+    {
+      "epoch": 0.4696753308578501,
+      "grad_norm": 0.15798069385353128,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 19732
+    },
+    {
+      "epoch": 0.46969913358088167,
+      "grad_norm": 3.3501605520931137,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 19733
+    },
+    {
+      "epoch": 0.46972293630391315,
+      "grad_norm": 3.0489800310318538,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 19734
+    },
+    {
+      "epoch": 0.4697467390269447,
+      "grad_norm": 9.376322870141777,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 19735
+    },
+    {
+      "epoch": 0.4697705417499762,
+      "grad_norm": 1.6608400704609425,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 19736
+    },
+    {
+      "epoch": 0.4697943444730077,
+      "grad_norm": 4.720892798438442,
+      "learning_rate": 2e-05,
+      "loss": 0.6449,
+      "step": 19737
+    },
+    {
+      "epoch": 0.4698181471960392,
+      "grad_norm": 2.1553860665522713,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 19738
+    },
+    {
+      "epoch": 0.46984194991907074,
+      "grad_norm": 13.614861502427878,
+      "learning_rate": 2e-05,
+      "loss": 0.4331,
+      "step": 19739
+    },
+    {
+      "epoch": 0.46986575264210223,
+      "grad_norm": 9.062381832312742,
+      "learning_rate": 2e-05,
+      "loss": 0.4256,
+      "step": 19740
+    },
+    {
+      "epoch": 0.46988955536513377,
+      "grad_norm": 3.941999878477377,
+      "learning_rate": 2e-05,
+      "loss": 0.5653,
+      "step": 19741
+    },
+    {
+      "epoch": 0.4699133580881653,
+      "grad_norm": 5.290070578616243,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 19742
+    },
+    {
+      "epoch": 0.4699371608111968,
+      "grad_norm": 3.567523245299773,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 19743
+    },
+    {
+      "epoch": 0.46996096353422834,
+      "grad_norm": 6.512477821805977,
+      "learning_rate": 2e-05,
+      "loss": 0.8976,
+      "step": 19744
+    },
+    {
+      "epoch": 0.4699847662572598,
+      "grad_norm": 4.608838338205829,
+      "learning_rate": 2e-05,
+      "loss": 0.4762,
+      "step": 19745
+    },
+    {
+      "epoch": 0.47000856898029136,
+      "grad_norm": 0.10147121156466699,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 19746
+    },
+    {
+      "epoch": 0.47003237170332285,
+      "grad_norm": 0.12593043593599823,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19747
+    },
+    {
+      "epoch": 0.4700561744263544,
+      "grad_norm": 1.1101139497491768,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 19748
+    },
+    {
+      "epoch": 0.4700799771493859,
+      "grad_norm": 7.208008944120932,
+      "learning_rate": 2e-05,
+      "loss": 1.4582,
+      "step": 19749
+    },
+    {
+      "epoch": 0.4701037798724174,
+      "grad_norm": 11.875053971717376,
+      "learning_rate": 2e-05,
+      "loss": 1.2273,
+      "step": 19750
+    },
+    {
+      "epoch": 0.4701275825954489,
+      "grad_norm": 0.12447769817420407,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19751
+    },
+    {
+      "epoch": 0.47015138531848044,
+      "grad_norm": 2.7145925350110622,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 19752
+    },
+    {
+      "epoch": 0.4701751880415119,
+      "grad_norm": 0.744866692806158,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 19753
+    },
+    {
+      "epoch": 0.47019899076454347,
+      "grad_norm": 5.249964305718558,
+      "learning_rate": 2e-05,
+      "loss": 0.6946,
+      "step": 19754
+    },
+    {
+      "epoch": 0.47022279348757495,
+      "grad_norm": 0.03434300568625897,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 19755
+    },
+    {
+      "epoch": 0.4702465962106065,
+      "grad_norm": 1.566303162749759,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 19756
+    },
+    {
+      "epoch": 0.47027039893363803,
+      "grad_norm": 6.68557708073478,
+      "learning_rate": 2e-05,
+      "loss": 0.4321,
+      "step": 19757
+    },
+    {
+      "epoch": 0.4702942016566695,
+      "grad_norm": 3.2073937737353067,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 19758
+    },
+    {
+      "epoch": 0.47031800437970106,
+      "grad_norm": 1.5798948830386463,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 19759
+    },
+    {
+      "epoch": 0.47034180710273255,
+      "grad_norm": 5.2631257145143255,
+      "learning_rate": 2e-05,
+      "loss": 0.143,
+      "step": 19760
+    },
+    {
+      "epoch": 0.4703656098257641,
+      "grad_norm": 5.725051957499965,
+      "learning_rate": 2e-05,
+      "loss": 0.2529,
+      "step": 19761
+    },
+    {
+      "epoch": 0.47038941254879557,
+      "grad_norm": 21.27058119690488,
+      "learning_rate": 2e-05,
+      "loss": 0.5916,
+      "step": 19762
+    },
+    {
+      "epoch": 0.4704132152718271,
+      "grad_norm": 3.51869414364508,
+      "learning_rate": 2e-05,
+      "loss": 0.3302,
+      "step": 19763
+    },
+    {
+      "epoch": 0.4704370179948586,
+      "grad_norm": 0.03038611101879734,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 19764
+    },
+    {
+      "epoch": 0.47046082071789014,
+      "grad_norm": 0.9523558677032357,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 19765
+    },
+    {
+      "epoch": 0.4704846234409216,
+      "grad_norm": 11.219029174801237,
+      "learning_rate": 2e-05,
+      "loss": 0.6676,
+      "step": 19766
+    },
+    {
+      "epoch": 0.47050842616395316,
+      "grad_norm": 4.001893778843292,
+      "learning_rate": 2e-05,
+      "loss": 0.271,
+      "step": 19767
+    },
+    {
+      "epoch": 0.47053222888698465,
+      "grad_norm": 0.42525967958477384,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 19768
+    },
+    {
+      "epoch": 0.4705560316100162,
+      "grad_norm": 2.0712457276421903,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 19769
+    },
+    {
+      "epoch": 0.4705798343330477,
+      "grad_norm": 4.828457850417348,
+      "learning_rate": 2e-05,
+      "loss": 0.3178,
+      "step": 19770
+    },
+    {
+      "epoch": 0.4706036370560792,
+      "grad_norm": 0.21056179168891737,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 19771
+    },
+    {
+      "epoch": 0.47062743977911076,
+      "grad_norm": 4.570113436031768,
+      "learning_rate": 2e-05,
+      "loss": 0.4837,
+      "step": 19772
+    },
+    {
+      "epoch": 0.47065124250214224,
+      "grad_norm": 0.994502139951063,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 19773
+    },
+    {
+      "epoch": 0.4706750452251738,
+      "grad_norm": 2.63858860856504,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 19774
+    },
+    {
+      "epoch": 0.47069884794820527,
+      "grad_norm": 2.5621346822591553,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 19775
+    },
+    {
+      "epoch": 0.4707226506712368,
+      "grad_norm": 0.21958364871870925,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 19776
+    },
+    {
+      "epoch": 0.4707464533942683,
+      "grad_norm": 3.9358267568443948,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 19777
+    },
+    {
+      "epoch": 0.47077025611729983,
+      "grad_norm": 3.633009135135664,
+      "learning_rate": 2e-05,
+      "loss": 0.3474,
+      "step": 19778
+    },
+    {
+      "epoch": 0.4707940588403313,
+      "grad_norm": 2.1170453434299263,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 19779
+    },
+    {
+      "epoch": 0.47081786156336286,
+      "grad_norm": 4.853299604857652,
+      "learning_rate": 2e-05,
+      "loss": 0.7113,
+      "step": 19780
+    },
+    {
+      "epoch": 0.47084166428639435,
+      "grad_norm": 8.808425158030015,
+      "learning_rate": 2e-05,
+      "loss": 0.4022,
+      "step": 19781
+    },
+    {
+      "epoch": 0.4708654670094259,
+      "grad_norm": 1.5296775906393263,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 19782
+    },
+    {
+      "epoch": 0.47088926973245737,
+      "grad_norm": 13.4550361945934,
+      "learning_rate": 2e-05,
+      "loss": 1.9285,
+      "step": 19783
+    },
+    {
+      "epoch": 0.4709130724554889,
+      "grad_norm": 4.228344705553531,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 19784
+    },
+    {
+      "epoch": 0.4709368751785204,
+      "grad_norm": 4.421714098938115,
+      "learning_rate": 2e-05,
+      "loss": 0.3645,
+      "step": 19785
+    },
+    {
+      "epoch": 0.47096067790155194,
+      "grad_norm": 4.366783778655673,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 19786
+    },
+    {
+      "epoch": 0.4709844806245835,
+      "grad_norm": 0.18327713407304977,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 19787
+    },
+    {
+      "epoch": 0.47100828334761496,
+      "grad_norm": 1.198907398403771,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 19788
+    },
+    {
+      "epoch": 0.4710320860706465,
+      "grad_norm": 4.361468640185955,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 19789
+    },
+    {
+      "epoch": 0.471055888793678,
+      "grad_norm": 2.413208872870156,
+      "learning_rate": 2e-05,
+      "loss": 0.2861,
+      "step": 19790
+    },
+    {
+      "epoch": 0.47107969151670953,
+      "grad_norm": 9.55436220600487,
+      "learning_rate": 2e-05,
+      "loss": 1.164,
+      "step": 19791
+    },
+    {
+      "epoch": 0.471103494239741,
+      "grad_norm": 3.548410669099031,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 19792
+    },
+    {
+      "epoch": 0.47112729696277256,
+      "grad_norm": 2.1608158494536926,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 19793
+    },
+    {
+      "epoch": 0.47115109968580404,
+      "grad_norm": 1.3387665402364377,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 19794
+    },
+    {
+      "epoch": 0.4711749024088356,
+      "grad_norm": 4.81033293122027,
+      "learning_rate": 2e-05,
+      "loss": 0.7301,
+      "step": 19795
+    },
+    {
+      "epoch": 0.47119870513186707,
+      "grad_norm": 8.8317653772574,
+      "learning_rate": 2e-05,
+      "loss": 0.4185,
+      "step": 19796
+    },
+    {
+      "epoch": 0.4712225078548986,
+      "grad_norm": 0.37530939342121294,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 19797
+    },
+    {
+      "epoch": 0.4712463105779301,
+      "grad_norm": 4.179483686676184,
+      "learning_rate": 2e-05,
+      "loss": 0.1356,
+      "step": 19798
+    },
+    {
+      "epoch": 0.47127011330096163,
+      "grad_norm": 4.288338815110377,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 19799
+    },
+    {
+      "epoch": 0.4712939160239931,
+      "grad_norm": 5.639377944483175,
+      "learning_rate": 2e-05,
+      "loss": 1.3064,
+      "step": 19800
+    },
+    {
+      "epoch": 0.47131771874702466,
+      "grad_norm": 2.7682691641438866,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 19801
+    },
+    {
+      "epoch": 0.4713415214700562,
+      "grad_norm": 2.8558476947291584,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 19802
+    },
+    {
+      "epoch": 0.4713653241930877,
+      "grad_norm": 8.702958436189034,
+      "learning_rate": 2e-05,
+      "loss": 0.9863,
+      "step": 19803
+    },
+    {
+      "epoch": 0.4713891269161192,
+      "grad_norm": 9.047487339600554,
+      "learning_rate": 2e-05,
+      "loss": 1.2978,
+      "step": 19804
+    },
+    {
+      "epoch": 0.4714129296391507,
+      "grad_norm": 5.701438020022591,
+      "learning_rate": 2e-05,
+      "loss": 0.5346,
+      "step": 19805
+    },
+    {
+      "epoch": 0.47143673236218225,
+      "grad_norm": 4.202630886390224,
+      "learning_rate": 2e-05,
+      "loss": 0.1402,
+      "step": 19806
+    },
+    {
+      "epoch": 0.47146053508521374,
+      "grad_norm": 2.7042821374164685,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 19807
+    },
+    {
+      "epoch": 0.4714843378082453,
+      "grad_norm": 2.3412860171235,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 19808
+    },
+    {
+      "epoch": 0.47150814053127676,
+      "grad_norm": 0.9657337263787017,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 19809
+    },
+    {
+      "epoch": 0.4715319432543083,
+      "grad_norm": 4.528260931260897,
+      "learning_rate": 2e-05,
+      "loss": 0.7673,
+      "step": 19810
+    },
+    {
+      "epoch": 0.4715557459773398,
+      "grad_norm": 0.15377798094428907,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19811
+    },
+    {
+      "epoch": 0.47157954870037133,
+      "grad_norm": 5.316449786478509,
+      "learning_rate": 2e-05,
+      "loss": 0.888,
+      "step": 19812
+    },
+    {
+      "epoch": 0.4716033514234028,
+      "grad_norm": 2.5354022911438934,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 19813
+    },
+    {
+      "epoch": 0.47162715414643436,
+      "grad_norm": 16.314911594559785,
+      "learning_rate": 2e-05,
+      "loss": 0.4517,
+      "step": 19814
+    },
+    {
+      "epoch": 0.47165095686946584,
+      "grad_norm": 2.061195713177454,
+      "learning_rate": 2e-05,
+      "loss": 0.1837,
+      "step": 19815
+    },
+    {
+      "epoch": 0.4716747595924974,
+      "grad_norm": 0.4122450847318264,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19816
+    },
+    {
+      "epoch": 0.4716985623155289,
+      "grad_norm": 3.3648197805465125,
+      "learning_rate": 2e-05,
+      "loss": 0.2413,
+      "step": 19817
+    },
+    {
+      "epoch": 0.4717223650385604,
+      "grad_norm": 9.46021185606151,
+      "learning_rate": 2e-05,
+      "loss": 0.45,
+      "step": 19818
+    },
+    {
+      "epoch": 0.47174616776159195,
+      "grad_norm": 0.08148950974443904,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 19819
+    },
+    {
+      "epoch": 0.47176997048462344,
+      "grad_norm": 6.146152724105377,
+      "learning_rate": 2e-05,
+      "loss": 0.8404,
+      "step": 19820
+    },
+    {
+      "epoch": 0.471793773207655,
+      "grad_norm": 0.5449303870206395,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 19821
+    },
+    {
+      "epoch": 0.47181757593068646,
+      "grad_norm": 2.5163061248030947,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 19822
+    },
+    {
+      "epoch": 0.471841378653718,
+      "grad_norm": 0.9203588350748845,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 19823
+    },
+    {
+      "epoch": 0.4718651813767495,
+      "grad_norm": 15.106717351283951,
+      "learning_rate": 2e-05,
+      "loss": 1.0327,
+      "step": 19824
+    },
+    {
+      "epoch": 0.47188898409978103,
+      "grad_norm": 3.151273521132567,
+      "learning_rate": 2e-05,
+      "loss": 0.4529,
+      "step": 19825
+    },
+    {
+      "epoch": 0.4719127868228125,
+      "grad_norm": 1.8029483432322895,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 19826
+    },
+    {
+      "epoch": 0.47193658954584405,
+      "grad_norm": 6.37214465709616,
+      "learning_rate": 2e-05,
+      "loss": 0.9066,
+      "step": 19827
+    },
+    {
+      "epoch": 0.47196039226887554,
+      "grad_norm": 3.1272092756443173,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 19828
+    },
+    {
+      "epoch": 0.4719841949919071,
+      "grad_norm": 2.480869437245178,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 19829
+    },
+    {
+      "epoch": 0.47200799771493857,
+      "grad_norm": 1.9711186849353983,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 19830
+    },
+    {
+      "epoch": 0.4720318004379701,
+      "grad_norm": 2.9375933224711197,
+      "learning_rate": 2e-05,
+      "loss": 0.118,
+      "step": 19831
+    },
+    {
+      "epoch": 0.4720556031610016,
+      "grad_norm": 5.7724198216608205,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 19832
+    },
+    {
+      "epoch": 0.47207940588403313,
+      "grad_norm": 8.305726365934031,
+      "learning_rate": 2e-05,
+      "loss": 0.61,
+      "step": 19833
+    },
+    {
+      "epoch": 0.4721032086070647,
+      "grad_norm": 2.0092501943058956,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 19834
+    },
+    {
+      "epoch": 0.47212701133009616,
+      "grad_norm": 3.7895118931764173,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 19835
+    },
+    {
+      "epoch": 0.4721508140531277,
+      "grad_norm": 3.9010428499810694,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 19836
+    },
+    {
+      "epoch": 0.4721746167761592,
+      "grad_norm": 0.6868213773284654,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 19837
+    },
+    {
+      "epoch": 0.4721984194991907,
+      "grad_norm": 8.807920763780627,
+      "learning_rate": 2e-05,
+      "loss": 1.3206,
+      "step": 19838
+    },
+    {
+      "epoch": 0.4722222222222222,
+      "grad_norm": 3.275798676366379,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 19839
+    },
+    {
+      "epoch": 0.47224602494525375,
+      "grad_norm": 1.7602123366573263,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 19840
+    },
+    {
+      "epoch": 0.47226982766828524,
+      "grad_norm": 1.796882053902796,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 19841
+    },
+    {
+      "epoch": 0.4722936303913168,
+      "grad_norm": 7.836648897476362,
+      "learning_rate": 2e-05,
+      "loss": 1.8185,
+      "step": 19842
+    },
+    {
+      "epoch": 0.47231743311434826,
+      "grad_norm": 3.6536997657917576,
+      "learning_rate": 2e-05,
+      "loss": 0.6961,
+      "step": 19843
+    },
+    {
+      "epoch": 0.4723412358373798,
+      "grad_norm": 1.0420820130314035,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 19844
+    },
+    {
+      "epoch": 0.4723650385604113,
+      "grad_norm": 2.4867366993914337,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 19845
+    },
+    {
+      "epoch": 0.47238884128344283,
+      "grad_norm": 4.164848723730894,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 19846
+    },
+    {
+      "epoch": 0.4724126440064743,
+      "grad_norm": 3.0014718585846545,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 19847
+    },
+    {
+      "epoch": 0.47243644672950585,
+      "grad_norm": 0.586506750548459,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 19848
+    },
+    {
+      "epoch": 0.4724602494525374,
+      "grad_norm": 1.7048648359943768,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 19849
+    },
+    {
+      "epoch": 0.4724840521755689,
+      "grad_norm": 2.52409248715933,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 19850
+    },
+    {
+      "epoch": 0.4725078548986004,
+      "grad_norm": 3.5443774878202277,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 19851
+    },
+    {
+      "epoch": 0.4725316576216319,
+      "grad_norm": 6.342108744200648,
+      "learning_rate": 2e-05,
+      "loss": 0.9637,
+      "step": 19852
+    },
+    {
+      "epoch": 0.47255546034466345,
+      "grad_norm": 2.9544892750315626,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 19853
+    },
+    {
+      "epoch": 0.47257926306769493,
+      "grad_norm": 4.465318797719687,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 19854
+    },
+    {
+      "epoch": 0.4726030657907265,
+      "grad_norm": 4.61273250250659,
+      "learning_rate": 2e-05,
+      "loss": 0.6878,
+      "step": 19855
+    },
+    {
+      "epoch": 0.47262686851375796,
+      "grad_norm": 5.0882681318613985,
+      "learning_rate": 2e-05,
+      "loss": 0.3691,
+      "step": 19856
+    },
+    {
+      "epoch": 0.4726506712367895,
+      "grad_norm": 3.4207343482898187,
+      "learning_rate": 2e-05,
+      "loss": 0.404,
+      "step": 19857
+    },
+    {
+      "epoch": 0.472674473959821,
+      "grad_norm": 3.3400682457146895,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 19858
+    },
+    {
+      "epoch": 0.4726982766828525,
+      "grad_norm": 4.038069929493143,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 19859
+    },
+    {
+      "epoch": 0.472722079405884,
+      "grad_norm": 4.589703968697744,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 19860
+    },
+    {
+      "epoch": 0.47274588212891555,
+      "grad_norm": 5.626752180786341,
+      "learning_rate": 2e-05,
+      "loss": 0.3239,
+      "step": 19861
+    },
+    {
+      "epoch": 0.47276968485194704,
+      "grad_norm": 4.791339046282965,
+      "learning_rate": 2e-05,
+      "loss": 0.7373,
+      "step": 19862
+    },
+    {
+      "epoch": 0.4727934875749786,
+      "grad_norm": 1.3898262962139134,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 19863
+    },
+    {
+      "epoch": 0.4728172902980101,
+      "grad_norm": 2.9044011264791885,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 19864
+    },
+    {
+      "epoch": 0.4728410930210416,
+      "grad_norm": 5.151011514156409,
+      "learning_rate": 2e-05,
+      "loss": 0.2341,
+      "step": 19865
+    },
+    {
+      "epoch": 0.47286489574407314,
+      "grad_norm": 1.7596895044371401,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 19866
+    },
+    {
+      "epoch": 0.47288869846710463,
+      "grad_norm": 1.7638590009235626,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 19867
+    },
+    {
+      "epoch": 0.47291250119013617,
+      "grad_norm": 5.728224120055212,
+      "learning_rate": 2e-05,
+      "loss": 0.8742,
+      "step": 19868
+    },
+    {
+      "epoch": 0.47293630391316765,
+      "grad_norm": 33.831464655329235,
+      "learning_rate": 2e-05,
+      "loss": 0.7627,
+      "step": 19869
+    },
+    {
+      "epoch": 0.4729601066361992,
+      "grad_norm": 1.003990303116868,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 19870
+    },
+    {
+      "epoch": 0.4729839093592307,
+      "grad_norm": 3.4193396839885084,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 19871
+    },
+    {
+      "epoch": 0.4730077120822622,
+      "grad_norm": 6.060179057974837,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 19872
+    },
+    {
+      "epoch": 0.4730315148052937,
+      "grad_norm": 7.412788348398746,
+      "learning_rate": 2e-05,
+      "loss": 1.7358,
+      "step": 19873
+    },
+    {
+      "epoch": 0.47305531752832525,
+      "grad_norm": 3.8145544813260965,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 19874
+    },
+    {
+      "epoch": 0.47307912025135673,
+      "grad_norm": 1.5302504092334264,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 19875
+    },
+    {
+      "epoch": 0.4731029229743883,
+      "grad_norm": 4.707519140965901,
+      "learning_rate": 2e-05,
+      "loss": 0.6607,
+      "step": 19876
+    },
+    {
+      "epoch": 0.47312672569741976,
+      "grad_norm": 5.354125541538396,
+      "learning_rate": 2e-05,
+      "loss": 1.239,
+      "step": 19877
+    },
+    {
+      "epoch": 0.4731505284204513,
+      "grad_norm": 6.066764549722816,
+      "learning_rate": 2e-05,
+      "loss": 1.1611,
+      "step": 19878
+    },
+    {
+      "epoch": 0.47317433114348284,
+      "grad_norm": 0.2325056192558604,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 19879
+    },
+    {
+      "epoch": 0.4731981338665143,
+      "grad_norm": 13.092507920218681,
+      "learning_rate": 2e-05,
+      "loss": 0.3591,
+      "step": 19880
+    },
+    {
+      "epoch": 0.47322193658954587,
+      "grad_norm": 1.5536011096862263,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 19881
+    },
+    {
+      "epoch": 0.47324573931257735,
+      "grad_norm": 7.1038771544805694,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 19882
+    },
+    {
+      "epoch": 0.4732695420356089,
+      "grad_norm": 2.4797742070211,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 19883
+    },
+    {
+      "epoch": 0.4732933447586404,
+      "grad_norm": 4.164023444555587,
+      "learning_rate": 2e-05,
+      "loss": 0.7033,
+      "step": 19884
+    },
+    {
+      "epoch": 0.4733171474816719,
+      "grad_norm": 1.9798340805171162,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 19885
+    },
+    {
+      "epoch": 0.4733409502047034,
+      "grad_norm": 3.051905403776324,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 19886
+    },
+    {
+      "epoch": 0.47336475292773494,
+      "grad_norm": 3.5018043182385656,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 19887
+    },
+    {
+      "epoch": 0.47338855565076643,
+      "grad_norm": 7.701474644694099,
+      "learning_rate": 2e-05,
+      "loss": 1.4595,
+      "step": 19888
+    },
+    {
+      "epoch": 0.47341235837379797,
+      "grad_norm": 5.086128659112723,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 19889
+    },
+    {
+      "epoch": 0.47343616109682946,
+      "grad_norm": 0.28549195488306617,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 19890
+    },
+    {
+      "epoch": 0.473459963819861,
+      "grad_norm": 1.0495438370586843,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 19891
+    },
+    {
+      "epoch": 0.4734837665428925,
+      "grad_norm": 1.9143684698671593,
+      "learning_rate": 2e-05,
+      "loss": 0.0958,
+      "step": 19892
+    },
+    {
+      "epoch": 0.473507569265924,
+      "grad_norm": 5.665996951981312,
+      "learning_rate": 2e-05,
+      "loss": 0.459,
+      "step": 19893
+    },
+    {
+      "epoch": 0.47353137198895556,
+      "grad_norm": 3.3389204107989303,
+      "learning_rate": 2e-05,
+      "loss": 0.5045,
+      "step": 19894
+    },
+    {
+      "epoch": 0.47355517471198705,
+      "grad_norm": 1.7675544125544842,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 19895
+    },
+    {
+      "epoch": 0.4735789774350186,
+      "grad_norm": 6.94194075737046,
+      "learning_rate": 2e-05,
+      "loss": 0.2245,
+      "step": 19896
+    },
+    {
+      "epoch": 0.4736027801580501,
+      "grad_norm": 1.217851573673004,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 19897
+    },
+    {
+      "epoch": 0.4736265828810816,
+      "grad_norm": 6.42637927045135,
+      "learning_rate": 2e-05,
+      "loss": 1.1554,
+      "step": 19898
+    },
+    {
+      "epoch": 0.4736503856041131,
+      "grad_norm": 6.279953819482252,
+      "learning_rate": 2e-05,
+      "loss": 1.0332,
+      "step": 19899
+    },
+    {
+      "epoch": 0.47367418832714464,
+      "grad_norm": 8.428666069440206,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 19900
+    },
+    {
+      "epoch": 0.4736979910501761,
+      "grad_norm": 3.41057859660067,
+      "learning_rate": 2e-05,
+      "loss": 0.5253,
+      "step": 19901
+    },
+    {
+      "epoch": 0.47372179377320767,
+      "grad_norm": 5.270520774893024,
+      "learning_rate": 2e-05,
+      "loss": 0.8226,
+      "step": 19902
+    },
+    {
+      "epoch": 0.47374559649623915,
+      "grad_norm": 3.7220063603164952,
+      "learning_rate": 2e-05,
+      "loss": 0.3463,
+      "step": 19903
+    },
+    {
+      "epoch": 0.4737693992192707,
+      "grad_norm": 0.07220333524666141,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 19904
+    },
+    {
+      "epoch": 0.4737932019423022,
+      "grad_norm": 5.08165882163981,
+      "learning_rate": 2e-05,
+      "loss": 0.9738,
+      "step": 19905
+    },
+    {
+      "epoch": 0.4738170046653337,
+      "grad_norm": 0.5896415137028665,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 19906
+    },
+    {
+      "epoch": 0.4738408073883652,
+      "grad_norm": 2.5355372728598025,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 19907
+    },
+    {
+      "epoch": 0.47386461011139674,
+      "grad_norm": 5.459950500270269,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 19908
+    },
+    {
+      "epoch": 0.4738884128344283,
+      "grad_norm": 4.1326228397897555,
+      "learning_rate": 2e-05,
+      "loss": 0.5942,
+      "step": 19909
+    },
+    {
+      "epoch": 0.47391221555745977,
+      "grad_norm": 1.8791053326478697,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 19910
+    },
+    {
+      "epoch": 0.4739360182804913,
+      "grad_norm": 2.20298881871995,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 19911
+    },
+    {
+      "epoch": 0.4739598210035228,
+      "grad_norm": 2.967443430455582,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 19912
+    },
+    {
+      "epoch": 0.47398362372655434,
+      "grad_norm": 6.213820317209227,
+      "learning_rate": 2e-05,
+      "loss": 1.3038,
+      "step": 19913
+    },
+    {
+      "epoch": 0.4740074264495858,
+      "grad_norm": 0.5286003109578752,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 19914
+    },
+    {
+      "epoch": 0.47403122917261736,
+      "grad_norm": 7.715916171759212,
+      "learning_rate": 2e-05,
+      "loss": 0.2321,
+      "step": 19915
+    },
+    {
+      "epoch": 0.47405503189564885,
+      "grad_norm": 2.149495808969188,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 19916
+    },
+    {
+      "epoch": 0.4740788346186804,
+      "grad_norm": 2.922578829448267,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 19917
+    },
+    {
+      "epoch": 0.4741026373417119,
+      "grad_norm": 3.6999050275683567,
+      "learning_rate": 2e-05,
+      "loss": 0.6625,
+      "step": 19918
+    },
+    {
+      "epoch": 0.4741264400647434,
+      "grad_norm": 2.1605720456625574,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 19919
+    },
+    {
+      "epoch": 0.4741502427877749,
+      "grad_norm": 2.661906898563942,
+      "learning_rate": 2e-05,
+      "loss": 0.5573,
+      "step": 19920
+    },
+    {
+      "epoch": 0.47417404551080644,
+      "grad_norm": 1.945884176161864,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 19921
+    },
+    {
+      "epoch": 0.4741978482338379,
+      "grad_norm": 10.64489693132926,
+      "learning_rate": 2e-05,
+      "loss": 1.4103,
+      "step": 19922
+    },
+    {
+      "epoch": 0.47422165095686947,
+      "grad_norm": 3.9333195411200403,
+      "learning_rate": 2e-05,
+      "loss": 0.4998,
+      "step": 19923
+    },
+    {
+      "epoch": 0.474245453679901,
+      "grad_norm": 8.578982904720368,
+      "learning_rate": 2e-05,
+      "loss": 0.5136,
+      "step": 19924
+    },
+    {
+      "epoch": 0.4742692564029325,
+      "grad_norm": 0.59706912922663,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 19925
+    },
+    {
+      "epoch": 0.47429305912596403,
+      "grad_norm": 3.19836881041025,
+      "learning_rate": 2e-05,
+      "loss": 0.3624,
+      "step": 19926
+    },
+    {
+      "epoch": 0.4743168618489955,
+      "grad_norm": 11.911075166493264,
+      "learning_rate": 2e-05,
+      "loss": 0.9134,
+      "step": 19927
+    },
+    {
+      "epoch": 0.47434066457202706,
+      "grad_norm": 4.165930003715304,
+      "learning_rate": 2e-05,
+      "loss": 0.3566,
+      "step": 19928
+    },
+    {
+      "epoch": 0.47436446729505855,
+      "grad_norm": 2.609711386647109,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 19929
+    },
+    {
+      "epoch": 0.4743882700180901,
+      "grad_norm": 1.152983336343569,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 19930
+    },
+    {
+      "epoch": 0.47441207274112157,
+      "grad_norm": 5.077881568621417,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 19931
+    },
+    {
+      "epoch": 0.4744358754641531,
+      "grad_norm": 3.5350001430275273,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 19932
+    },
+    {
+      "epoch": 0.4744596781871846,
+      "grad_norm": 0.9069344428474638,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 19933
+    },
+    {
+      "epoch": 0.47448348091021614,
+      "grad_norm": 9.275593188574442,
+      "learning_rate": 2e-05,
+      "loss": 0.7291,
+      "step": 19934
+    },
+    {
+      "epoch": 0.4745072836332476,
+      "grad_norm": 14.080163650273555,
+      "learning_rate": 2e-05,
+      "loss": 0.313,
+      "step": 19935
+    },
+    {
+      "epoch": 0.47453108635627916,
+      "grad_norm": 4.000785830778482,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 19936
+    },
+    {
+      "epoch": 0.47455488907931065,
+      "grad_norm": 6.789793365400881,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 19937
+    },
+    {
+      "epoch": 0.4745786918023422,
+      "grad_norm": 3.283043179135277,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 19938
+    },
+    {
+      "epoch": 0.47460249452537373,
+      "grad_norm": 0.18225230500629663,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19939
+    },
+    {
+      "epoch": 0.4746262972484052,
+      "grad_norm": 2.805155060443047,
+      "learning_rate": 2e-05,
+      "loss": 0.4181,
+      "step": 19940
+    },
+    {
+      "epoch": 0.47465009997143676,
+      "grad_norm": 6.82687151097372,
+      "learning_rate": 2e-05,
+      "loss": 1.065,
+      "step": 19941
+    },
+    {
+      "epoch": 0.47467390269446824,
+      "grad_norm": 2.601482774517911,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 19942
+    },
+    {
+      "epoch": 0.4746977054174998,
+      "grad_norm": 5.437362556435532,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 19943
+    },
+    {
+      "epoch": 0.47472150814053127,
+      "grad_norm": 9.51594161696273,
+      "learning_rate": 2e-05,
+      "loss": 0.7806,
+      "step": 19944
+    },
+    {
+      "epoch": 0.4747453108635628,
+      "grad_norm": 10.65946133196706,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 19945
+    },
+    {
+      "epoch": 0.4747691135865943,
+      "grad_norm": 7.562266240618706,
+      "learning_rate": 2e-05,
+      "loss": 0.5828,
+      "step": 19946
+    },
+    {
+      "epoch": 0.47479291630962583,
+      "grad_norm": 5.202232701493566,
+      "learning_rate": 2e-05,
+      "loss": 0.7605,
+      "step": 19947
+    },
+    {
+      "epoch": 0.4748167190326573,
+      "grad_norm": 1.6000190557871734,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 19948
+    },
+    {
+      "epoch": 0.47484052175568886,
+      "grad_norm": 9.650535852463177,
+      "learning_rate": 2e-05,
+      "loss": 0.356,
+      "step": 19949
+    },
+    {
+      "epoch": 0.47486432447872035,
+      "grad_norm": 10.58771534206688,
+      "learning_rate": 2e-05,
+      "loss": 0.2467,
+      "step": 19950
+    },
+    {
+      "epoch": 0.4748881272017519,
+      "grad_norm": 5.1133901029028825,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 19951
+    },
+    {
+      "epoch": 0.47491192992478337,
+      "grad_norm": 4.364274387449112,
+      "learning_rate": 2e-05,
+      "loss": 0.6255,
+      "step": 19952
+    },
+    {
+      "epoch": 0.4749357326478149,
+      "grad_norm": 2.843548679480986,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 19953
+    },
+    {
+      "epoch": 0.4749595353708464,
+      "grad_norm": 5.659883699165644,
+      "learning_rate": 2e-05,
+      "loss": 0.9029,
+      "step": 19954
+    },
+    {
+      "epoch": 0.47498333809387794,
+      "grad_norm": 0.25165808545937735,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 19955
+    },
+    {
+      "epoch": 0.4750071408169095,
+      "grad_norm": 9.316737147205085,
+      "learning_rate": 2e-05,
+      "loss": 0.2398,
+      "step": 19956
+    },
+    {
+      "epoch": 0.47503094353994096,
+      "grad_norm": 4.029278890964658,
+      "learning_rate": 2e-05,
+      "loss": 0.2442,
+      "step": 19957
+    },
+    {
+      "epoch": 0.4750547462629725,
+      "grad_norm": 2.4184664895258767,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 19958
+    },
+    {
+      "epoch": 0.475078548986004,
+      "grad_norm": 6.187837584360045,
+      "learning_rate": 2e-05,
+      "loss": 1.4415,
+      "step": 19959
+    },
+    {
+      "epoch": 0.47510235170903553,
+      "grad_norm": 8.077917083032524,
+      "learning_rate": 2e-05,
+      "loss": 0.3478,
+      "step": 19960
+    },
+    {
+      "epoch": 0.475126154432067,
+      "grad_norm": 1.2683778179823701,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 19961
+    },
+    {
+      "epoch": 0.47514995715509856,
+      "grad_norm": 0.40680119990602925,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 19962
+    },
+    {
+      "epoch": 0.47517375987813004,
+      "grad_norm": 4.6030918726520875,
+      "learning_rate": 2e-05,
+      "loss": 0.3232,
+      "step": 19963
+    },
+    {
+      "epoch": 0.4751975626011616,
+      "grad_norm": 2.9983761389787986,
+      "learning_rate": 2e-05,
+      "loss": 0.201,
+      "step": 19964
+    },
+    {
+      "epoch": 0.47522136532419307,
+      "grad_norm": 2.9842723996311062,
+      "learning_rate": 2e-05,
+      "loss": 0.5491,
+      "step": 19965
+    },
+    {
+      "epoch": 0.4752451680472246,
+      "grad_norm": 2.6367627085850227,
+      "learning_rate": 2e-05,
+      "loss": 0.3422,
+      "step": 19966
+    },
+    {
+      "epoch": 0.4752689707702561,
+      "grad_norm": 2.2988108735364534,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 19967
+    },
+    {
+      "epoch": 0.47529277349328763,
+      "grad_norm": 1.9488592957766817,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 19968
+    },
+    {
+      "epoch": 0.4753165762163191,
+      "grad_norm": 0.2970506247200081,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 19969
+    },
+    {
+      "epoch": 0.47534037893935066,
+      "grad_norm": 4.147622694371092,
+      "learning_rate": 2e-05,
+      "loss": 0.605,
+      "step": 19970
+    },
+    {
+      "epoch": 0.4753641816623822,
+      "grad_norm": 3.105556171999947,
+      "learning_rate": 2e-05,
+      "loss": 0.6445,
+      "step": 19971
+    },
+    {
+      "epoch": 0.4753879843854137,
+      "grad_norm": 5.4827679064142485,
+      "learning_rate": 2e-05,
+      "loss": 0.6061,
+      "step": 19972
+    },
+    {
+      "epoch": 0.4754117871084452,
+      "grad_norm": 4.299333252086572,
+      "learning_rate": 2e-05,
+      "loss": 0.7366,
+      "step": 19973
+    },
+    {
+      "epoch": 0.4754355898314767,
+      "grad_norm": 0.6261662413245321,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 19974
+    },
+    {
+      "epoch": 0.47545939255450825,
+      "grad_norm": 2.954763806161853,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 19975
+    },
+    {
+      "epoch": 0.47548319527753974,
+      "grad_norm": 5.231012560124916,
+      "learning_rate": 2e-05,
+      "loss": 0.8027,
+      "step": 19976
+    },
+    {
+      "epoch": 0.4755069980005713,
+      "grad_norm": 7.982970188811801,
+      "learning_rate": 2e-05,
+      "loss": 0.6423,
+      "step": 19977
+    },
+    {
+      "epoch": 0.47553080072360276,
+      "grad_norm": 7.619862182853913,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 19978
+    },
+    {
+      "epoch": 0.4755546034466343,
+      "grad_norm": 1.0392901393033351,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 19979
+    },
+    {
+      "epoch": 0.4755784061696658,
+      "grad_norm": 0.1652781490252101,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 19980
+    },
+    {
+      "epoch": 0.47560220889269733,
+      "grad_norm": 2.32412222478423,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 19981
+    },
+    {
+      "epoch": 0.4756260116157288,
+      "grad_norm": 3.0582962130688816,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 19982
+    },
+    {
+      "epoch": 0.47564981433876036,
+      "grad_norm": 4.028147522195281,
+      "learning_rate": 2e-05,
+      "loss": 0.6327,
+      "step": 19983
+    },
+    {
+      "epoch": 0.47567361706179184,
+      "grad_norm": 0.07892590695276679,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 19984
+    },
+    {
+      "epoch": 0.4756974197848234,
+      "grad_norm": 2.3696338113076467,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 19985
+    },
+    {
+      "epoch": 0.4757212225078549,
+      "grad_norm": 5.151625330520126,
+      "learning_rate": 2e-05,
+      "loss": 0.2475,
+      "step": 19986
+    },
+    {
+      "epoch": 0.4757450252308864,
+      "grad_norm": 13.163123598067255,
+      "learning_rate": 2e-05,
+      "loss": 1.2295,
+      "step": 19987
+    },
+    {
+      "epoch": 0.47576882795391795,
+      "grad_norm": 1.261027863070216,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 19988
+    },
+    {
+      "epoch": 0.47579263067694944,
+      "grad_norm": 10.60815675475375,
+      "learning_rate": 2e-05,
+      "loss": 0.3065,
+      "step": 19989
+    },
+    {
+      "epoch": 0.475816433399981,
+      "grad_norm": 1.1716759412038802,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 19990
+    },
+    {
+      "epoch": 0.47584023612301246,
+      "grad_norm": 5.035505244878776,
+      "learning_rate": 2e-05,
+      "loss": 0.7668,
+      "step": 19991
+    },
+    {
+      "epoch": 0.475864038846044,
+      "grad_norm": 3.1566098309859854,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 19992
+    },
+    {
+      "epoch": 0.4758878415690755,
+      "grad_norm": 4.24045708495751,
+      "learning_rate": 2e-05,
+      "loss": 0.6012,
+      "step": 19993
+    },
+    {
+      "epoch": 0.47591164429210703,
+      "grad_norm": 5.2536826073109015,
+      "learning_rate": 2e-05,
+      "loss": 0.765,
+      "step": 19994
+    },
+    {
+      "epoch": 0.4759354470151385,
+      "grad_norm": 4.6216624182590875,
+      "learning_rate": 2e-05,
+      "loss": 1.0453,
+      "step": 19995
+    },
+    {
+      "epoch": 0.47595924973817005,
+      "grad_norm": 0.4040710134576949,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 19996
+    },
+    {
+      "epoch": 0.47598305246120154,
+      "grad_norm": 0.2513242030605147,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 19997
+    },
+    {
+      "epoch": 0.4760068551842331,
+      "grad_norm": 1.936044275080516,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 19998
+    },
+    {
+      "epoch": 0.47603065790726456,
+      "grad_norm": 5.826941263257681,
+      "learning_rate": 2e-05,
+      "loss": 0.9034,
+      "step": 19999
+    },
+    {
+      "epoch": 0.4760544606302961,
+      "grad_norm": 5.742189716168904,
+      "learning_rate": 2e-05,
+      "loss": 0.3477,
+      "step": 20000
+    },
+    {
+      "epoch": 0.47607826335332765,
+      "grad_norm": 0.884798645709164,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 20001
+    },
+    {
+      "epoch": 0.47610206607635913,
+      "grad_norm": 2.0802893002940355,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 20002
+    },
+    {
+      "epoch": 0.4761258687993907,
+      "grad_norm": 10.277462826624767,
+      "learning_rate": 2e-05,
+      "loss": 0.8168,
+      "step": 20003
+    },
+    {
+      "epoch": 0.47614967152242216,
+      "grad_norm": 1.9654080777927108,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 20004
+    },
+    {
+      "epoch": 0.4761734742454537,
+      "grad_norm": 1.28515187554159,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 20005
+    },
+    {
+      "epoch": 0.4761972769684852,
+      "grad_norm": 1.3998092102691517,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 20006
+    },
+    {
+      "epoch": 0.4762210796915167,
+      "grad_norm": 2.393507404924447,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 20007
+    },
+    {
+      "epoch": 0.4762448824145482,
+      "grad_norm": 0.27119106007602406,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20008
+    },
+    {
+      "epoch": 0.47626868513757975,
+      "grad_norm": 7.681421050744365,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 20009
+    },
+    {
+      "epoch": 0.47629248786061124,
+      "grad_norm": 1.6140616967592332,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20010
+    },
+    {
+      "epoch": 0.4763162905836428,
+      "grad_norm": 0.7234573461226675,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 20011
+    },
+    {
+      "epoch": 0.47634009330667426,
+      "grad_norm": 6.799558705623523,
+      "learning_rate": 2e-05,
+      "loss": 1.2099,
+      "step": 20012
+    },
+    {
+      "epoch": 0.4763638960297058,
+      "grad_norm": 4.276524045275011,
+      "learning_rate": 2e-05,
+      "loss": 0.6635,
+      "step": 20013
+    },
+    {
+      "epoch": 0.4763876987527373,
+      "grad_norm": 6.2763534349812815,
+      "learning_rate": 2e-05,
+      "loss": 0.6714,
+      "step": 20014
+    },
+    {
+      "epoch": 0.47641150147576883,
+      "grad_norm": 8.3904816637496,
+      "learning_rate": 2e-05,
+      "loss": 0.2685,
+      "step": 20015
+    },
+    {
+      "epoch": 0.47643530419880037,
+      "grad_norm": 3.8087797633898455,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 20016
+    },
+    {
+      "epoch": 0.47645910692183185,
+      "grad_norm": 3.659120233313428,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 20017
+    },
+    {
+      "epoch": 0.4764829096448634,
+      "grad_norm": 3.279032306778255,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 20018
+    },
+    {
+      "epoch": 0.4765067123678949,
+      "grad_norm": 8.133367175418924,
+      "learning_rate": 2e-05,
+      "loss": 0.8414,
+      "step": 20019
+    },
+    {
+      "epoch": 0.4765305150909264,
+      "grad_norm": 0.043424472348905896,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4765543178139579,
+      "grad_norm": 6.813551727041119,
+      "learning_rate": 2e-05,
+      "loss": 0.8095,
+      "step": 20021
+    },
+    {
+      "epoch": 0.47657812053698945,
+      "grad_norm": 2.3142444277366923,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 20022
+    },
+    {
+      "epoch": 0.47660192326002093,
+      "grad_norm": 1.2961693989849428,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 20023
+    },
+    {
+      "epoch": 0.4766257259830525,
+      "grad_norm": 3.8867726305663286,
+      "learning_rate": 2e-05,
+      "loss": 0.2588,
+      "step": 20024
+    },
+    {
+      "epoch": 0.47664952870608396,
+      "grad_norm": 3.4834461192678643,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 20025
+    },
+    {
+      "epoch": 0.4766733314291155,
+      "grad_norm": 6.794443947668037,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 20026
+    },
+    {
+      "epoch": 0.476697134152147,
+      "grad_norm": 4.327921485960153,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 20027
+    },
+    {
+      "epoch": 0.4767209368751785,
+      "grad_norm": 6.336935780180965,
+      "learning_rate": 2e-05,
+      "loss": 0.3722,
+      "step": 20028
+    },
+    {
+      "epoch": 0.47674473959821,
+      "grad_norm": 8.617743615264192,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 20029
+    },
+    {
+      "epoch": 0.47676854232124155,
+      "grad_norm": 7.090567473310391,
+      "learning_rate": 2e-05,
+      "loss": 0.1987,
+      "step": 20030
+    },
+    {
+      "epoch": 0.4767923450442731,
+      "grad_norm": 1.7476226983196534,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 20031
+    },
+    {
+      "epoch": 0.4768161477673046,
+      "grad_norm": 0.24552994322540345,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20032
+    },
+    {
+      "epoch": 0.4768399504903361,
+      "grad_norm": 4.979009459576432,
+      "learning_rate": 2e-05,
+      "loss": 0.5511,
+      "step": 20033
+    },
+    {
+      "epoch": 0.4768637532133676,
+      "grad_norm": 3.149969730109911,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 20034
+    },
+    {
+      "epoch": 0.47688755593639914,
+      "grad_norm": 3.2737053723145877,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 20035
+    },
+    {
+      "epoch": 0.47691135865943063,
+      "grad_norm": 2.771763349048004,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 20036
+    },
+    {
+      "epoch": 0.47693516138246217,
+      "grad_norm": 7.276986737276241,
+      "learning_rate": 2e-05,
+      "loss": 0.7686,
+      "step": 20037
+    },
+    {
+      "epoch": 0.47695896410549365,
+      "grad_norm": 7.160634069254026,
+      "learning_rate": 2e-05,
+      "loss": 0.6887,
+      "step": 20038
+    },
+    {
+      "epoch": 0.4769827668285252,
+      "grad_norm": 3.6282570582519402,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 20039
+    },
+    {
+      "epoch": 0.4770065695515567,
+      "grad_norm": 8.287037929506626,
+      "learning_rate": 2e-05,
+      "loss": 0.9006,
+      "step": 20040
+    },
+    {
+      "epoch": 0.4770303722745882,
+      "grad_norm": 3.734394949893449,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 20041
+    },
+    {
+      "epoch": 0.4770541749976197,
+      "grad_norm": 4.490829410824968,
+      "learning_rate": 2e-05,
+      "loss": 0.994,
+      "step": 20042
+    },
+    {
+      "epoch": 0.47707797772065125,
+      "grad_norm": 4.3819758792348145,
+      "learning_rate": 2e-05,
+      "loss": 0.238,
+      "step": 20043
+    },
+    {
+      "epoch": 0.47710178044368273,
+      "grad_norm": 8.640940982926985,
+      "learning_rate": 2e-05,
+      "loss": 1.1969,
+      "step": 20044
+    },
+    {
+      "epoch": 0.4771255831667143,
+      "grad_norm": 6.89579572730401,
+      "learning_rate": 2e-05,
+      "loss": 0.2358,
+      "step": 20045
+    },
+    {
+      "epoch": 0.4771493858897458,
+      "grad_norm": 0.5310914119816328,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 20046
+    },
+    {
+      "epoch": 0.4771731886127773,
+      "grad_norm": 3.6464382919160445,
+      "learning_rate": 2e-05,
+      "loss": 0.4995,
+      "step": 20047
+    },
+    {
+      "epoch": 0.47719699133580884,
+      "grad_norm": 1.3812025253502993,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 20048
+    },
+    {
+      "epoch": 0.4772207940588403,
+      "grad_norm": 2.9499408125450723,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 20049
+    },
+    {
+      "epoch": 0.47724459678187187,
+      "grad_norm": 4.00249392310407,
+      "learning_rate": 2e-05,
+      "loss": 0.5416,
+      "step": 20050
+    },
+    {
+      "epoch": 0.47726839950490335,
+      "grad_norm": 12.292304620523845,
+      "learning_rate": 2e-05,
+      "loss": 0.814,
+      "step": 20051
+    },
+    {
+      "epoch": 0.4772922022279349,
+      "grad_norm": 2.3408481731183475,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 20052
+    },
+    {
+      "epoch": 0.4773160049509664,
+      "grad_norm": 3.4518901992934343,
+      "learning_rate": 2e-05,
+      "loss": 0.26,
+      "step": 20053
+    },
+    {
+      "epoch": 0.4773398076739979,
+      "grad_norm": 3.4574420318795633,
+      "learning_rate": 2e-05,
+      "loss": 0.128,
+      "step": 20054
+    },
+    {
+      "epoch": 0.4773636103970294,
+      "grad_norm": 4.1025295864678055,
+      "learning_rate": 2e-05,
+      "loss": 0.6575,
+      "step": 20055
+    },
+    {
+      "epoch": 0.47738741312006094,
+      "grad_norm": 3.005295159652865,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 20056
+    },
+    {
+      "epoch": 0.47741121584309243,
+      "grad_norm": 2.055392659612792,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 20057
+    },
+    {
+      "epoch": 0.47743501856612397,
+      "grad_norm": 1.0391874461504895,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 20058
+    },
+    {
+      "epoch": 0.47745882128915546,
+      "grad_norm": 6.105465174252952,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 20059
+    },
+    {
+      "epoch": 0.477482624012187,
+      "grad_norm": 4.124390716293405,
+      "learning_rate": 2e-05,
+      "loss": 0.5535,
+      "step": 20060
+    },
+    {
+      "epoch": 0.47750642673521854,
+      "grad_norm": 5.752034991029428,
+      "learning_rate": 2e-05,
+      "loss": 0.9906,
+      "step": 20061
+    },
+    {
+      "epoch": 0.47753022945825,
+      "grad_norm": 7.176183736603284,
+      "learning_rate": 2e-05,
+      "loss": 0.2451,
+      "step": 20062
+    },
+    {
+      "epoch": 0.47755403218128156,
+      "grad_norm": 10.97435183411099,
+      "learning_rate": 2e-05,
+      "loss": 1.8069,
+      "step": 20063
+    },
+    {
+      "epoch": 0.47757783490431305,
+      "grad_norm": 13.534371588178656,
+      "learning_rate": 2e-05,
+      "loss": 0.8664,
+      "step": 20064
+    },
+    {
+      "epoch": 0.4776016376273446,
+      "grad_norm": 3.237301082782703,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 20065
+    },
+    {
+      "epoch": 0.4776254403503761,
+      "grad_norm": 5.234119883551838,
+      "learning_rate": 2e-05,
+      "loss": 0.8081,
+      "step": 20066
+    },
+    {
+      "epoch": 0.4776492430734076,
+      "grad_norm": 2.7122648907128326,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 20067
+    },
+    {
+      "epoch": 0.4776730457964391,
+      "grad_norm": 2.988805059406111,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 20068
+    },
+    {
+      "epoch": 0.47769684851947064,
+      "grad_norm": 11.77242372328923,
+      "learning_rate": 2e-05,
+      "loss": 0.4496,
+      "step": 20069
+    },
+    {
+      "epoch": 0.4777206512425021,
+      "grad_norm": 1.30916652879028,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 20070
+    },
+    {
+      "epoch": 0.47774445396553367,
+      "grad_norm": 3.6239463086165276,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 20071
+    },
+    {
+      "epoch": 0.47776825668856515,
+      "grad_norm": 4.408964400780268,
+      "learning_rate": 2e-05,
+      "loss": 0.4962,
+      "step": 20072
+    },
+    {
+      "epoch": 0.4777920594115967,
+      "grad_norm": 0.4307322953134017,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20073
+    },
+    {
+      "epoch": 0.4778158621346282,
+      "grad_norm": 1.3555209064793303,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20074
+    },
+    {
+      "epoch": 0.4778396648576597,
+      "grad_norm": 1.430415378128169,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 20075
+    },
+    {
+      "epoch": 0.4778634675806912,
+      "grad_norm": 4.882808380922065,
+      "learning_rate": 2e-05,
+      "loss": 0.1254,
+      "step": 20076
+    },
+    {
+      "epoch": 0.47788727030372274,
+      "grad_norm": 5.270273121771219,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 20077
+    },
+    {
+      "epoch": 0.4779110730267543,
+      "grad_norm": 7.267543615092319,
+      "learning_rate": 2e-05,
+      "loss": 0.3925,
+      "step": 20078
+    },
+    {
+      "epoch": 0.47793487574978577,
+      "grad_norm": 1.161737229666483,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 20079
+    },
+    {
+      "epoch": 0.4779586784728173,
+      "grad_norm": 6.581715471088353,
+      "learning_rate": 2e-05,
+      "loss": 0.8584,
+      "step": 20080
+    },
+    {
+      "epoch": 0.4779824811958488,
+      "grad_norm": 2.6170871286273507,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 20081
+    },
+    {
+      "epoch": 0.47800628391888034,
+      "grad_norm": 2.1774269730096982,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 20082
+    },
+    {
+      "epoch": 0.4780300866419118,
+      "grad_norm": 10.592123171688407,
+      "learning_rate": 2e-05,
+      "loss": 0.6004,
+      "step": 20083
+    },
+    {
+      "epoch": 0.47805388936494336,
+      "grad_norm": 2.1414236009503056,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 20084
+    },
+    {
+      "epoch": 0.47807769208797485,
+      "grad_norm": 4.135669204734058,
+      "learning_rate": 2e-05,
+      "loss": 0.5931,
+      "step": 20085
+    },
+    {
+      "epoch": 0.4781014948110064,
+      "grad_norm": 0.7863646817382658,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 20086
+    },
+    {
+      "epoch": 0.4781252975340379,
+      "grad_norm": 7.41030194290792,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 20087
+    },
+    {
+      "epoch": 0.4781491002570694,
+      "grad_norm": 2.371214502091966,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 20088
+    },
+    {
+      "epoch": 0.4781729029801009,
+      "grad_norm": 8.062002262234287,
+      "learning_rate": 2e-05,
+      "loss": 0.2738,
+      "step": 20089
+    },
+    {
+      "epoch": 0.47819670570313244,
+      "grad_norm": 2.0701733887200855,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 20090
+    },
+    {
+      "epoch": 0.4782205084261639,
+      "grad_norm": 8.96966868748225,
+      "learning_rate": 2e-05,
+      "loss": 0.2786,
+      "step": 20091
+    },
+    {
+      "epoch": 0.47824431114919547,
+      "grad_norm": 5.527891965753428,
+      "learning_rate": 2e-05,
+      "loss": 0.4608,
+      "step": 20092
+    },
+    {
+      "epoch": 0.478268113872227,
+      "grad_norm": 0.47256925342612827,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 20093
+    },
+    {
+      "epoch": 0.4782919165952585,
+      "grad_norm": 10.447650343364678,
+      "learning_rate": 2e-05,
+      "loss": 0.3684,
+      "step": 20094
+    },
+    {
+      "epoch": 0.47831571931829003,
+      "grad_norm": 6.58480768024222,
+      "learning_rate": 2e-05,
+      "loss": 0.6631,
+      "step": 20095
+    },
+    {
+      "epoch": 0.4783395220413215,
+      "grad_norm": 2.276010830005037,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 20096
+    },
+    {
+      "epoch": 0.47836332476435306,
+      "grad_norm": 7.366772250333075,
+      "learning_rate": 2e-05,
+      "loss": 0.225,
+      "step": 20097
+    },
+    {
+      "epoch": 0.47838712748738454,
+      "grad_norm": 4.314048662781834,
+      "learning_rate": 2e-05,
+      "loss": 0.5968,
+      "step": 20098
+    },
+    {
+      "epoch": 0.4784109302104161,
+      "grad_norm": 3.726148840527642,
+      "learning_rate": 2e-05,
+      "loss": 0.3456,
+      "step": 20099
+    },
+    {
+      "epoch": 0.47843473293344757,
+      "grad_norm": 1.1422015300900794,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4784585356564791,
+      "grad_norm": 0.9652492541978058,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 20101
+    },
+    {
+      "epoch": 0.4784823383795106,
+      "grad_norm": 0.23635721178037986,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 20102
+    },
+    {
+      "epoch": 0.47850614110254214,
+      "grad_norm": 2.8799507250705996,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 20103
+    },
+    {
+      "epoch": 0.4785299438255736,
+      "grad_norm": 4.038424927205481,
+      "learning_rate": 2e-05,
+      "loss": 0.5564,
+      "step": 20104
+    },
+    {
+      "epoch": 0.47855374654860516,
+      "grad_norm": 2.8738895465631304,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 20105
+    },
+    {
+      "epoch": 0.47857754927163665,
+      "grad_norm": 0.860700561328193,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20106
+    },
+    {
+      "epoch": 0.4786013519946682,
+      "grad_norm": 0.4191225002407129,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 20107
+    },
+    {
+      "epoch": 0.47862515471769973,
+      "grad_norm": 21.472002293798855,
+      "learning_rate": 2e-05,
+      "loss": 1.1264,
+      "step": 20108
+    },
+    {
+      "epoch": 0.4786489574407312,
+      "grad_norm": 3.910331375990446,
+      "learning_rate": 2e-05,
+      "loss": 0.5283,
+      "step": 20109
+    },
+    {
+      "epoch": 0.47867276016376276,
+      "grad_norm": 3.907807715567825,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 20110
+    },
+    {
+      "epoch": 0.47869656288679424,
+      "grad_norm": 4.089265605057935,
+      "learning_rate": 2e-05,
+      "loss": 0.6137,
+      "step": 20111
+    },
+    {
+      "epoch": 0.4787203656098258,
+      "grad_norm": 2.8365085667534435,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 20112
+    },
+    {
+      "epoch": 0.47874416833285727,
+      "grad_norm": 8.264704888013513,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 20113
+    },
+    {
+      "epoch": 0.4787679710558888,
+      "grad_norm": 8.640113486428405,
+      "learning_rate": 2e-05,
+      "loss": 1.0478,
+      "step": 20114
+    },
+    {
+      "epoch": 0.4787917737789203,
+      "grad_norm": 9.66570112341849,
+      "learning_rate": 2e-05,
+      "loss": 0.5005,
+      "step": 20115
+    },
+    {
+      "epoch": 0.47881557650195183,
+      "grad_norm": 4.742327550256369,
+      "learning_rate": 2e-05,
+      "loss": 0.6643,
+      "step": 20116
+    },
+    {
+      "epoch": 0.4788393792249833,
+      "grad_norm": 14.907561018301312,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 20117
+    },
+    {
+      "epoch": 0.47886318194801486,
+      "grad_norm": 2.5237987611045822,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 20118
+    },
+    {
+      "epoch": 0.47888698467104635,
+      "grad_norm": 1.9686048521200796,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 20119
+    },
+    {
+      "epoch": 0.4789107873940779,
+      "grad_norm": 3.689313159403729,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 20120
+    },
+    {
+      "epoch": 0.47893459011710937,
+      "grad_norm": 4.748549612791475,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 20121
+    },
+    {
+      "epoch": 0.4789583928401409,
+      "grad_norm": 6.7937479753172605,
+      "learning_rate": 2e-05,
+      "loss": 0.414,
+      "step": 20122
+    },
+    {
+      "epoch": 0.47898219556317245,
+      "grad_norm": 0.9410125884203869,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 20123
+    },
+    {
+      "epoch": 0.47900599828620394,
+      "grad_norm": 4.363495202810665,
+      "learning_rate": 2e-05,
+      "loss": 0.5829,
+      "step": 20124
+    },
+    {
+      "epoch": 0.4790298010092355,
+      "grad_norm": 1.5766959731698629,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 20125
+    },
+    {
+      "epoch": 0.47905360373226696,
+      "grad_norm": 10.26618812425039,
+      "learning_rate": 2e-05,
+      "loss": 0.6134,
+      "step": 20126
+    },
+    {
+      "epoch": 0.4790774064552985,
+      "grad_norm": 5.258327721556569,
+      "learning_rate": 2e-05,
+      "loss": 0.4399,
+      "step": 20127
+    },
+    {
+      "epoch": 0.47910120917833,
+      "grad_norm": 4.818303225022734,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 20128
+    },
+    {
+      "epoch": 0.47912501190136153,
+      "grad_norm": 6.79848351580155,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 20129
+    },
+    {
+      "epoch": 0.479148814624393,
+      "grad_norm": 6.678855264993347,
+      "learning_rate": 2e-05,
+      "loss": 0.5937,
+      "step": 20130
+    },
+    {
+      "epoch": 0.47917261734742456,
+      "grad_norm": 2.094598436201103,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20131
+    },
+    {
+      "epoch": 0.47919642007045604,
+      "grad_norm": 1.091060588979553,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 20132
+    },
+    {
+      "epoch": 0.4792202227934876,
+      "grad_norm": 3.670273926237129,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 20133
+    },
+    {
+      "epoch": 0.47924402551651907,
+      "grad_norm": 6.3666053167650825,
+      "learning_rate": 2e-05,
+      "loss": 0.7155,
+      "step": 20134
+    },
+    {
+      "epoch": 0.4792678282395506,
+      "grad_norm": 3.360975970517437,
+      "learning_rate": 2e-05,
+      "loss": 0.4774,
+      "step": 20135
+    },
+    {
+      "epoch": 0.4792916309625821,
+      "grad_norm": 5.837606768285757,
+      "learning_rate": 2e-05,
+      "loss": 1.255,
+      "step": 20136
+    },
+    {
+      "epoch": 0.47931543368561363,
+      "grad_norm": 1.7899541328706712,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 20137
+    },
+    {
+      "epoch": 0.4793392364086452,
+      "grad_norm": 4.544024396262405,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 20138
+    },
+    {
+      "epoch": 0.47936303913167666,
+      "grad_norm": 2.572413527380259,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 20139
+    },
+    {
+      "epoch": 0.4793868418547082,
+      "grad_norm": 5.4824049862740605,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 20140
+    },
+    {
+      "epoch": 0.4794106445777397,
+      "grad_norm": 2.2165027990921424,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 20141
+    },
+    {
+      "epoch": 0.4794344473007712,
+      "grad_norm": 0.2258858394973946,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20142
+    },
+    {
+      "epoch": 0.4794582500238027,
+      "grad_norm": 2.2738827328783873,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 20143
+    },
+    {
+      "epoch": 0.47948205274683425,
+      "grad_norm": 8.828376644200652,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 20144
+    },
+    {
+      "epoch": 0.47950585546986574,
+      "grad_norm": 3.841446691468914,
+      "learning_rate": 2e-05,
+      "loss": 0.5495,
+      "step": 20145
+    },
+    {
+      "epoch": 0.4795296581928973,
+      "grad_norm": 3.6899491511270277,
+      "learning_rate": 2e-05,
+      "loss": 0.3342,
+      "step": 20146
+    },
+    {
+      "epoch": 0.47955346091592876,
+      "grad_norm": 1.6517365597902394,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 20147
+    },
+    {
+      "epoch": 0.4795772636389603,
+      "grad_norm": 1.2849965903140785,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 20148
+    },
+    {
+      "epoch": 0.4796010663619918,
+      "grad_norm": 1.8405386873983651,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 20149
+    },
+    {
+      "epoch": 0.47962486908502333,
+      "grad_norm": 7.479618372245589,
+      "learning_rate": 2e-05,
+      "loss": 0.3272,
+      "step": 20150
+    },
+    {
+      "epoch": 0.4796486718080548,
+      "grad_norm": 9.079152447731579,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 20151
+    },
+    {
+      "epoch": 0.47967247453108636,
+      "grad_norm": 3.9488355618448656,
+      "learning_rate": 2e-05,
+      "loss": 0.6208,
+      "step": 20152
+    },
+    {
+      "epoch": 0.4796962772541179,
+      "grad_norm": 1.4659906220104246,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 20153
+    },
+    {
+      "epoch": 0.4797200799771494,
+      "grad_norm": 1.5213273009317798,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 20154
+    },
+    {
+      "epoch": 0.4797438827001809,
+      "grad_norm": 1.1515847711355942,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 20155
+    },
+    {
+      "epoch": 0.4797676854232124,
+      "grad_norm": 7.450162252529556,
+      "learning_rate": 2e-05,
+      "loss": 0.273,
+      "step": 20156
+    },
+    {
+      "epoch": 0.47979148814624395,
+      "grad_norm": 4.485309143980619,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 20157
+    },
+    {
+      "epoch": 0.47981529086927543,
+      "grad_norm": 0.436084919501149,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 20158
+    },
+    {
+      "epoch": 0.479839093592307,
+      "grad_norm": 2.680539289320633,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 20159
+    },
+    {
+      "epoch": 0.47986289631533846,
+      "grad_norm": 4.820886649893626,
+      "learning_rate": 2e-05,
+      "loss": 0.736,
+      "step": 20160
+    },
+    {
+      "epoch": 0.47988669903837,
+      "grad_norm": 3.03504037203815,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 20161
+    },
+    {
+      "epoch": 0.4799105017614015,
+      "grad_norm": 4.750890470700288,
+      "learning_rate": 2e-05,
+      "loss": 0.5806,
+      "step": 20162
+    },
+    {
+      "epoch": 0.479934304484433,
+      "grad_norm": 3.243837623517925,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 20163
+    },
+    {
+      "epoch": 0.4799581072074645,
+      "grad_norm": 5.095714342706117,
+      "learning_rate": 2e-05,
+      "loss": 1.2312,
+      "step": 20164
+    },
+    {
+      "epoch": 0.47998190993049605,
+      "grad_norm": 2.6415682566715115,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 20165
+    },
+    {
+      "epoch": 0.48000571265352754,
+      "grad_norm": 2.3950540372841767,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 20166
+    },
+    {
+      "epoch": 0.4800295153765591,
+      "grad_norm": 3.349956604525295,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 20167
+    },
+    {
+      "epoch": 0.4800533180995906,
+      "grad_norm": 11.189621509469262,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 20168
+    },
+    {
+      "epoch": 0.4800771208226221,
+      "grad_norm": 1.5572918498059596,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 20169
+    },
+    {
+      "epoch": 0.48010092354565365,
+      "grad_norm": 3.1654369550440635,
+      "learning_rate": 2e-05,
+      "loss": 0.4504,
+      "step": 20170
+    },
+    {
+      "epoch": 0.48012472626868513,
+      "grad_norm": 0.9234735363178768,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 20171
+    },
+    {
+      "epoch": 0.48014852899171667,
+      "grad_norm": 0.5247470449457055,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20172
+    },
+    {
+      "epoch": 0.48017233171474816,
+      "grad_norm": 2.0569233959039943,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 20173
+    },
+    {
+      "epoch": 0.4801961344377797,
+      "grad_norm": 3.660599018031259,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 20174
+    },
+    {
+      "epoch": 0.4802199371608112,
+      "grad_norm": 6.069237440391558,
+      "learning_rate": 2e-05,
+      "loss": 1.7436,
+      "step": 20175
+    },
+    {
+      "epoch": 0.4802437398838427,
+      "grad_norm": 1.4027709862547215,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 20176
+    },
+    {
+      "epoch": 0.4802675426068742,
+      "grad_norm": 2.1155752806664725,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 20177
+    },
+    {
+      "epoch": 0.48029134532990575,
+      "grad_norm": 3.1896996451078548,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 20178
+    },
+    {
+      "epoch": 0.48031514805293724,
+      "grad_norm": 2.034540454547772,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 20179
+    },
+    {
+      "epoch": 0.4803389507759688,
+      "grad_norm": 4.95429904424744,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 20180
+    },
+    {
+      "epoch": 0.48036275349900026,
+      "grad_norm": 3.840725335213618,
+      "learning_rate": 2e-05,
+      "loss": 0.3755,
+      "step": 20181
+    },
+    {
+      "epoch": 0.4803865562220318,
+      "grad_norm": 3.5652844284016503,
+      "learning_rate": 2e-05,
+      "loss": 0.3126,
+      "step": 20182
+    },
+    {
+      "epoch": 0.48041035894506334,
+      "grad_norm": 0.33214770580813313,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 20183
+    },
+    {
+      "epoch": 0.48043416166809483,
+      "grad_norm": 2.425215963975801,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 20184
+    },
+    {
+      "epoch": 0.48045796439112637,
+      "grad_norm": 1.8182319558372892,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 20185
+    },
+    {
+      "epoch": 0.48048176711415785,
+      "grad_norm": 1.9478386239071783,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 20186
+    },
+    {
+      "epoch": 0.4805055698371894,
+      "grad_norm": 5.23178277536855,
+      "learning_rate": 2e-05,
+      "loss": 0.1573,
+      "step": 20187
+    },
+    {
+      "epoch": 0.4805293725602209,
+      "grad_norm": 2.936564503129745,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 20188
+    },
+    {
+      "epoch": 0.4805531752832524,
+      "grad_norm": 2.815161239722616,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 20189
+    },
+    {
+      "epoch": 0.4805769780062839,
+      "grad_norm": 8.599009941064175,
+      "learning_rate": 2e-05,
+      "loss": 0.5294,
+      "step": 20190
+    },
+    {
+      "epoch": 0.48060078072931545,
+      "grad_norm": 7.00551488383431,
+      "learning_rate": 2e-05,
+      "loss": 0.9136,
+      "step": 20191
+    },
+    {
+      "epoch": 0.48062458345234693,
+      "grad_norm": 1.2935497671819731,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 20192
+    },
+    {
+      "epoch": 0.4806483861753785,
+      "grad_norm": 3.5873161360832553,
+      "learning_rate": 2e-05,
+      "loss": 0.5938,
+      "step": 20193
+    },
+    {
+      "epoch": 0.48067218889840996,
+      "grad_norm": 0.7822139231399305,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 20194
+    },
+    {
+      "epoch": 0.4806959916214415,
+      "grad_norm": 2.062295980631832,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 20195
+    },
+    {
+      "epoch": 0.480719794344473,
+      "grad_norm": 4.154191831423966,
+      "learning_rate": 2e-05,
+      "loss": 0.5552,
+      "step": 20196
+    },
+    {
+      "epoch": 0.4807435970675045,
+      "grad_norm": 5.205511533904436,
+      "learning_rate": 2e-05,
+      "loss": 0.5409,
+      "step": 20197
+    },
+    {
+      "epoch": 0.480767399790536,
+      "grad_norm": 1.223041499567729,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 20198
+    },
+    {
+      "epoch": 0.48079120251356755,
+      "grad_norm": 4.0628942434859,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 20199
+    },
+    {
+      "epoch": 0.4808150052365991,
+      "grad_norm": 2.2735970112044077,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 20200
+    },
+    {
+      "epoch": 0.4808388079596306,
+      "grad_norm": 3.1790944124397793,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 20201
+    },
+    {
+      "epoch": 0.4808626106826621,
+      "grad_norm": 8.031343300907356,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 20202
+    },
+    {
+      "epoch": 0.4808864134056936,
+      "grad_norm": 15.954911717896858,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 20203
+    },
+    {
+      "epoch": 0.48091021612872514,
+      "grad_norm": 4.190302891839739,
+      "learning_rate": 2e-05,
+      "loss": 0.8662,
+      "step": 20204
+    },
+    {
+      "epoch": 0.48093401885175663,
+      "grad_norm": 2.5300187334421045,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 20205
+    },
+    {
+      "epoch": 0.48095782157478817,
+      "grad_norm": 1.9804061889048374,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 20206
+    },
+    {
+      "epoch": 0.48098162429781965,
+      "grad_norm": 4.653048211806477,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 20207
+    },
+    {
+      "epoch": 0.4810054270208512,
+      "grad_norm": 0.4011978181155888,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 20208
+    },
+    {
+      "epoch": 0.4810292297438827,
+      "grad_norm": 3.9380215439866615,
+      "learning_rate": 2e-05,
+      "loss": 0.411,
+      "step": 20209
+    },
+    {
+      "epoch": 0.4810530324669142,
+      "grad_norm": 5.125891645684667,
+      "learning_rate": 2e-05,
+      "loss": 1.1047,
+      "step": 20210
+    },
+    {
+      "epoch": 0.4810768351899457,
+      "grad_norm": 16.482903766427224,
+      "learning_rate": 2e-05,
+      "loss": 1.4909,
+      "step": 20211
+    },
+    {
+      "epoch": 0.48110063791297725,
+      "grad_norm": 6.151722263223408,
+      "learning_rate": 2e-05,
+      "loss": 0.6451,
+      "step": 20212
+    },
+    {
+      "epoch": 0.48112444063600873,
+      "grad_norm": 1.5441127218511703,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 20213
+    },
+    {
+      "epoch": 0.4811482433590403,
+      "grad_norm": 4.218413762148581,
+      "learning_rate": 2e-05,
+      "loss": 0.5004,
+      "step": 20214
+    },
+    {
+      "epoch": 0.4811720460820718,
+      "grad_norm": 3.2366680968159405,
+      "learning_rate": 2e-05,
+      "loss": 0.5606,
+      "step": 20215
+    },
+    {
+      "epoch": 0.4811958488051033,
+      "grad_norm": 1.1179814509227122,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 20216
+    },
+    {
+      "epoch": 0.48121965152813484,
+      "grad_norm": 9.697579685874611,
+      "learning_rate": 2e-05,
+      "loss": 1.371,
+      "step": 20217
+    },
+    {
+      "epoch": 0.4812434542511663,
+      "grad_norm": 4.017169624002868,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 20218
+    },
+    {
+      "epoch": 0.48126725697419787,
+      "grad_norm": 0.8363266680711234,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 20219
+    },
+    {
+      "epoch": 0.48129105969722935,
+      "grad_norm": 3.6401552739978027,
+      "learning_rate": 2e-05,
+      "loss": 0.3561,
+      "step": 20220
+    },
+    {
+      "epoch": 0.4813148624202609,
+      "grad_norm": 3.2863188690251683,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 20221
+    },
+    {
+      "epoch": 0.4813386651432924,
+      "grad_norm": 4.6761546551005715,
+      "learning_rate": 2e-05,
+      "loss": 0.7076,
+      "step": 20222
+    },
+    {
+      "epoch": 0.4813624678663239,
+      "grad_norm": 3.39918849619383,
+      "learning_rate": 2e-05,
+      "loss": 0.2074,
+      "step": 20223
+    },
+    {
+      "epoch": 0.4813862705893554,
+      "grad_norm": 3.674570637854179,
+      "learning_rate": 2e-05,
+      "loss": 0.3622,
+      "step": 20224
+    },
+    {
+      "epoch": 0.48141007331238694,
+      "grad_norm": 23.948296602179774,
+      "learning_rate": 2e-05,
+      "loss": 0.5229,
+      "step": 20225
+    },
+    {
+      "epoch": 0.48143387603541843,
+      "grad_norm": 6.687710183552269,
+      "learning_rate": 2e-05,
+      "loss": 0.8616,
+      "step": 20226
+    },
+    {
+      "epoch": 0.48145767875844997,
+      "grad_norm": 9.292439946122089,
+      "learning_rate": 2e-05,
+      "loss": 0.6339,
+      "step": 20227
+    },
+    {
+      "epoch": 0.48148148148148145,
+      "grad_norm": 1.2185819822057513,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 20228
+    },
+    {
+      "epoch": 0.481505284204513,
+      "grad_norm": 5.880451018655017,
+      "learning_rate": 2e-05,
+      "loss": 0.8448,
+      "step": 20229
+    },
+    {
+      "epoch": 0.48152908692754454,
+      "grad_norm": 2.744976242662451,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 20230
+    },
+    {
+      "epoch": 0.481552889650576,
+      "grad_norm": 1.7626322838518842,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 20231
+    },
+    {
+      "epoch": 0.48157669237360756,
+      "grad_norm": 4.980139387622053,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 20232
+    },
+    {
+      "epoch": 0.48160049509663905,
+      "grad_norm": 7.168346078343194,
+      "learning_rate": 2e-05,
+      "loss": 1.4954,
+      "step": 20233
+    },
+    {
+      "epoch": 0.4816242978196706,
+      "grad_norm": 3.028172414281157,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 20234
+    },
+    {
+      "epoch": 0.4816481005427021,
+      "grad_norm": 2.47122772243686,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 20235
+    },
+    {
+      "epoch": 0.4816719032657336,
+      "grad_norm": 3.990389176287324,
+      "learning_rate": 2e-05,
+      "loss": 0.3087,
+      "step": 20236
+    },
+    {
+      "epoch": 0.4816957059887651,
+      "grad_norm": 3.4049885674765665,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 20237
+    },
+    {
+      "epoch": 0.48171950871179664,
+      "grad_norm": 3.4488489910602684,
+      "learning_rate": 2e-05,
+      "loss": 0.5018,
+      "step": 20238
+    },
+    {
+      "epoch": 0.4817433114348281,
+      "grad_norm": 3.6316801817012783,
+      "learning_rate": 2e-05,
+      "loss": 0.6967,
+      "step": 20239
+    },
+    {
+      "epoch": 0.48176711415785967,
+      "grad_norm": 3.227548153254887,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 20240
+    },
+    {
+      "epoch": 0.48179091688089115,
+      "grad_norm": 5.077042524686074,
+      "learning_rate": 2e-05,
+      "loss": 1.1057,
+      "step": 20241
+    },
+    {
+      "epoch": 0.4818147196039227,
+      "grad_norm": 1.6624167805585097,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 20242
+    },
+    {
+      "epoch": 0.4818385223269542,
+      "grad_norm": 3.5380779102544797,
+      "learning_rate": 2e-05,
+      "loss": 0.2016,
+      "step": 20243
+    },
+    {
+      "epoch": 0.4818623250499857,
+      "grad_norm": 1.62025192641644,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 20244
+    },
+    {
+      "epoch": 0.48188612777301726,
+      "grad_norm": 2.948008830506455,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 20245
+    },
+    {
+      "epoch": 0.48190993049604874,
+      "grad_norm": 3.383643263816915,
+      "learning_rate": 2e-05,
+      "loss": 0.4758,
+      "step": 20246
+    },
+    {
+      "epoch": 0.4819337332190803,
+      "grad_norm": 2.431597223378558,
+      "learning_rate": 2e-05,
+      "loss": 0.3651,
+      "step": 20247
+    },
+    {
+      "epoch": 0.48195753594211177,
+      "grad_norm": 1.3891815780503596,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 20248
+    },
+    {
+      "epoch": 0.4819813386651433,
+      "grad_norm": 5.10713023369388,
+      "learning_rate": 2e-05,
+      "loss": 0.77,
+      "step": 20249
+    },
+    {
+      "epoch": 0.4820051413881748,
+      "grad_norm": 1.2499568555143052,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 20250
+    },
+    {
+      "epoch": 0.48202894411120634,
+      "grad_norm": 33.361097649354306,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 20251
+    },
+    {
+      "epoch": 0.4820527468342378,
+      "grad_norm": 1.7222951380035996,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 20252
+    },
+    {
+      "epoch": 0.48207654955726936,
+      "grad_norm": 6.443343573334499,
+      "learning_rate": 2e-05,
+      "loss": 0.6236,
+      "step": 20253
+    },
+    {
+      "epoch": 0.48210035228030085,
+      "grad_norm": 5.69704117677564,
+      "learning_rate": 2e-05,
+      "loss": 0.8665,
+      "step": 20254
+    },
+    {
+      "epoch": 0.4821241550033324,
+      "grad_norm": 1.2121010570753612,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 20255
+    },
+    {
+      "epoch": 0.4821479577263639,
+      "grad_norm": 3.04392369497135,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 20256
+    },
+    {
+      "epoch": 0.4821717604493954,
+      "grad_norm": 2.5040925332038504,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 20257
+    },
+    {
+      "epoch": 0.4821955631724269,
+      "grad_norm": 3.520705071165592,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 20258
+    },
+    {
+      "epoch": 0.48221936589545844,
+      "grad_norm": 2.993733300822782,
+      "learning_rate": 2e-05,
+      "loss": 0.5941,
+      "step": 20259
+    },
+    {
+      "epoch": 0.48224316861849,
+      "grad_norm": 2.9614346170659105,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 20260
+    },
+    {
+      "epoch": 0.48226697134152147,
+      "grad_norm": 5.248028942142031,
+      "learning_rate": 2e-05,
+      "loss": 1.0646,
+      "step": 20261
+    },
+    {
+      "epoch": 0.482290774064553,
+      "grad_norm": 1.8076119004629732,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 20262
+    },
+    {
+      "epoch": 0.4823145767875845,
+      "grad_norm": 3.475123193630021,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 20263
+    },
+    {
+      "epoch": 0.48233837951061603,
+      "grad_norm": 0.40555989249299096,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20264
+    },
+    {
+      "epoch": 0.4823621822336475,
+      "grad_norm": 19.92538998587568,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 20265
+    },
+    {
+      "epoch": 0.48238598495667906,
+      "grad_norm": 2.177062614490569,
+      "learning_rate": 2e-05,
+      "loss": 0.2489,
+      "step": 20266
+    },
+    {
+      "epoch": 0.48240978767971054,
+      "grad_norm": 1.8252621857976352,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 20267
+    },
+    {
+      "epoch": 0.4824335904027421,
+      "grad_norm": 11.998187877187156,
+      "learning_rate": 2e-05,
+      "loss": 0.9104,
+      "step": 20268
+    },
+    {
+      "epoch": 0.48245739312577357,
+      "grad_norm": 0.8670410837779153,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 20269
+    },
+    {
+      "epoch": 0.4824811958488051,
+      "grad_norm": 4.062424805005671,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 20270
+    },
+    {
+      "epoch": 0.4825049985718366,
+      "grad_norm": 0.6574706688257194,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20271
+    },
+    {
+      "epoch": 0.48252880129486814,
+      "grad_norm": 0.39984281309296377,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20272
+    },
+    {
+      "epoch": 0.4825526040178996,
+      "grad_norm": 18.188955983578424,
+      "learning_rate": 2e-05,
+      "loss": 1.0054,
+      "step": 20273
+    },
+    {
+      "epoch": 0.48257640674093116,
+      "grad_norm": 0.5048769316458106,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20274
+    },
+    {
+      "epoch": 0.4826002094639627,
+      "grad_norm": 10.461780878356825,
+      "learning_rate": 2e-05,
+      "loss": 0.7308,
+      "step": 20275
+    },
+    {
+      "epoch": 0.4826240121869942,
+      "grad_norm": 7.240093458448704,
+      "learning_rate": 2e-05,
+      "loss": 0.0889,
+      "step": 20276
+    },
+    {
+      "epoch": 0.48264781491002573,
+      "grad_norm": 4.803531347034059,
+      "learning_rate": 2e-05,
+      "loss": 0.8133,
+      "step": 20277
+    },
+    {
+      "epoch": 0.4826716176330572,
+      "grad_norm": 3.3665025099106027,
+      "learning_rate": 2e-05,
+      "loss": 0.2091,
+      "step": 20278
+    },
+    {
+      "epoch": 0.48269542035608876,
+      "grad_norm": 0.7069634248463176,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 20279
+    },
+    {
+      "epoch": 0.48271922307912024,
+      "grad_norm": 12.79597219903896,
+      "learning_rate": 2e-05,
+      "loss": 1.1182,
+      "step": 20280
+    },
+    {
+      "epoch": 0.4827430258021518,
+      "grad_norm": 4.689879900730554,
+      "learning_rate": 2e-05,
+      "loss": 0.7344,
+      "step": 20281
+    },
+    {
+      "epoch": 0.48276682852518327,
+      "grad_norm": 0.6367496018811987,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 20282
+    },
+    {
+      "epoch": 0.4827906312482148,
+      "grad_norm": 2.7508301825832557,
+      "learning_rate": 2e-05,
+      "loss": 0.2519,
+      "step": 20283
+    },
+    {
+      "epoch": 0.4828144339712463,
+      "grad_norm": 5.882125653930788,
+      "learning_rate": 2e-05,
+      "loss": 1.0231,
+      "step": 20284
+    },
+    {
+      "epoch": 0.48283823669427783,
+      "grad_norm": 1.9401640099888346,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 20285
+    },
+    {
+      "epoch": 0.4828620394173093,
+      "grad_norm": 1.307393741830523,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 20286
+    },
+    {
+      "epoch": 0.48288584214034086,
+      "grad_norm": 5.631037309165193,
+      "learning_rate": 2e-05,
+      "loss": 0.6478,
+      "step": 20287
+    },
+    {
+      "epoch": 0.48290964486337234,
+      "grad_norm": 3.7304573780121966,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 20288
+    },
+    {
+      "epoch": 0.4829334475864039,
+      "grad_norm": 2.412414754168615,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 20289
+    },
+    {
+      "epoch": 0.4829572503094354,
+      "grad_norm": 8.964145563196011,
+      "learning_rate": 2e-05,
+      "loss": 0.2865,
+      "step": 20290
+    },
+    {
+      "epoch": 0.4829810530324669,
+      "grad_norm": 6.225848745947705,
+      "learning_rate": 2e-05,
+      "loss": 0.3899,
+      "step": 20291
+    },
+    {
+      "epoch": 0.48300485575549845,
+      "grad_norm": 1.2135727264929443,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 20292
+    },
+    {
+      "epoch": 0.48302865847852994,
+      "grad_norm": 9.495758253546946,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 20293
+    },
+    {
+      "epoch": 0.4830524612015615,
+      "grad_norm": 12.04833009104115,
+      "learning_rate": 2e-05,
+      "loss": 2.2445,
+      "step": 20294
+    },
+    {
+      "epoch": 0.48307626392459296,
+      "grad_norm": 2.8600837634886176,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 20295
+    },
+    {
+      "epoch": 0.4831000666476245,
+      "grad_norm": 4.909978015451072,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 20296
+    },
+    {
+      "epoch": 0.483123869370656,
+      "grad_norm": 3.6403189506519276,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 20297
+    },
+    {
+      "epoch": 0.48314767209368753,
+      "grad_norm": 5.640898610799377,
+      "learning_rate": 2e-05,
+      "loss": 1.1028,
+      "step": 20298
+    },
+    {
+      "epoch": 0.483171474816719,
+      "grad_norm": 2.750498399641161,
+      "learning_rate": 2e-05,
+      "loss": 0.3595,
+      "step": 20299
+    },
+    {
+      "epoch": 0.48319527753975056,
+      "grad_norm": 21.493534964405526,
+      "learning_rate": 2e-05,
+      "loss": 1.8612,
+      "step": 20300
+    },
+    {
+      "epoch": 0.48321908026278204,
+      "grad_norm": 3.624524706995795,
+      "learning_rate": 2e-05,
+      "loss": 0.685,
+      "step": 20301
+    },
+    {
+      "epoch": 0.4832428829858136,
+      "grad_norm": 10.713103299630719,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 20302
+    },
+    {
+      "epoch": 0.48326668570884507,
+      "grad_norm": 4.1449220750017854,
+      "learning_rate": 2e-05,
+      "loss": 0.7767,
+      "step": 20303
+    },
+    {
+      "epoch": 0.4832904884318766,
+      "grad_norm": 2.2003188612747917,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 20304
+    },
+    {
+      "epoch": 0.48331429115490815,
+      "grad_norm": 12.137246437317927,
+      "learning_rate": 2e-05,
+      "loss": 0.7289,
+      "step": 20305
+    },
+    {
+      "epoch": 0.48333809387793963,
+      "grad_norm": 5.27599311503151,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 20306
+    },
+    {
+      "epoch": 0.4833618966009712,
+      "grad_norm": 8.046392880613448,
+      "learning_rate": 2e-05,
+      "loss": 0.9602,
+      "step": 20307
+    },
+    {
+      "epoch": 0.48338569932400266,
+      "grad_norm": 0.30704686677912124,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 20308
+    },
+    {
+      "epoch": 0.4834095020470342,
+      "grad_norm": 20.85233770950088,
+      "learning_rate": 2e-05,
+      "loss": 0.5516,
+      "step": 20309
+    },
+    {
+      "epoch": 0.4834333047700657,
+      "grad_norm": 2.575476621294459,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 20310
+    },
+    {
+      "epoch": 0.4834571074930972,
+      "grad_norm": 2.576169109375729,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 20311
+    },
+    {
+      "epoch": 0.4834809102161287,
+      "grad_norm": 7.721562116494622,
+      "learning_rate": 2e-05,
+      "loss": 1.1737,
+      "step": 20312
+    },
+    {
+      "epoch": 0.48350471293916025,
+      "grad_norm": 5.910416242363598,
+      "learning_rate": 2e-05,
+      "loss": 0.8477,
+      "step": 20313
+    },
+    {
+      "epoch": 0.48352851566219174,
+      "grad_norm": 6.328260442053972,
+      "learning_rate": 2e-05,
+      "loss": 0.2372,
+      "step": 20314
+    },
+    {
+      "epoch": 0.4835523183852233,
+      "grad_norm": 2.5023659180592106,
+      "learning_rate": 2e-05,
+      "loss": 0.0924,
+      "step": 20315
+    },
+    {
+      "epoch": 0.48357612110825476,
+      "grad_norm": 0.8451768362650015,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 20316
+    },
+    {
+      "epoch": 0.4835999238312863,
+      "grad_norm": 8.566683042784511,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 20317
+    },
+    {
+      "epoch": 0.4836237265543178,
+      "grad_norm": 1.9901299133490147,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 20318
+    },
+    {
+      "epoch": 0.48364752927734933,
+      "grad_norm": 6.4441482089073086,
+      "learning_rate": 2e-05,
+      "loss": 1.0498,
+      "step": 20319
+    },
+    {
+      "epoch": 0.4836713320003808,
+      "grad_norm": 2.7433351815148854,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 20320
+    },
+    {
+      "epoch": 0.48369513472341236,
+      "grad_norm": 3.49642166394889,
+      "learning_rate": 2e-05,
+      "loss": 0.6892,
+      "step": 20321
+    },
+    {
+      "epoch": 0.4837189374464439,
+      "grad_norm": 7.692859572428748,
+      "learning_rate": 2e-05,
+      "loss": 0.1313,
+      "step": 20322
+    },
+    {
+      "epoch": 0.4837427401694754,
+      "grad_norm": 7.215379671286534,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 20323
+    },
+    {
+      "epoch": 0.4837665428925069,
+      "grad_norm": 5.405830768213168,
+      "learning_rate": 2e-05,
+      "loss": 0.17,
+      "step": 20324
+    },
+    {
+      "epoch": 0.4837903456155384,
+      "grad_norm": 9.527321201920868,
+      "learning_rate": 2e-05,
+      "loss": 0.5221,
+      "step": 20325
+    },
+    {
+      "epoch": 0.48381414833856995,
+      "grad_norm": 3.6468419586552248,
+      "learning_rate": 2e-05,
+      "loss": 0.8943,
+      "step": 20326
+    },
+    {
+      "epoch": 0.48383795106160143,
+      "grad_norm": 2.3804063443074095,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 20327
+    },
+    {
+      "epoch": 0.483861753784633,
+      "grad_norm": 2.896469968249497,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 20328
+    },
+    {
+      "epoch": 0.48388555650766446,
+      "grad_norm": 10.395946158053661,
+      "learning_rate": 2e-05,
+      "loss": 0.125,
+      "step": 20329
+    },
+    {
+      "epoch": 0.483909359230696,
+      "grad_norm": 3.865255297853299,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 20330
+    },
+    {
+      "epoch": 0.4839331619537275,
+      "grad_norm": 2.7762007526029815,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 20331
+    },
+    {
+      "epoch": 0.483956964676759,
+      "grad_norm": 3.3638624965666777,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 20332
+    },
+    {
+      "epoch": 0.4839807673997905,
+      "grad_norm": 5.102918857586542,
+      "learning_rate": 2e-05,
+      "loss": 0.9362,
+      "step": 20333
+    },
+    {
+      "epoch": 0.48400457012282205,
+      "grad_norm": 5.823734954217729,
+      "learning_rate": 2e-05,
+      "loss": 1.0054,
+      "step": 20334
+    },
+    {
+      "epoch": 0.48402837284585354,
+      "grad_norm": 9.806346324822213,
+      "learning_rate": 2e-05,
+      "loss": 0.4016,
+      "step": 20335
+    },
+    {
+      "epoch": 0.4840521755688851,
+      "grad_norm": 2.969543422665651,
+      "learning_rate": 2e-05,
+      "loss": 0.445,
+      "step": 20336
+    },
+    {
+      "epoch": 0.4840759782919166,
+      "grad_norm": 1.022990090047342,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 20337
+    },
+    {
+      "epoch": 0.4840997810149481,
+      "grad_norm": 10.99272975883852,
+      "learning_rate": 2e-05,
+      "loss": 0.5324,
+      "step": 20338
+    },
+    {
+      "epoch": 0.48412358373797965,
+      "grad_norm": 0.5378868158357997,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 20339
+    },
+    {
+      "epoch": 0.48414738646101113,
+      "grad_norm": 5.263384920179268,
+      "learning_rate": 2e-05,
+      "loss": 0.9556,
+      "step": 20340
+    },
+    {
+      "epoch": 0.48417118918404267,
+      "grad_norm": 3.525809537952937,
+      "learning_rate": 2e-05,
+      "loss": 0.5593,
+      "step": 20341
+    },
+    {
+      "epoch": 0.48419499190707416,
+      "grad_norm": 2.5701159222337435,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 20342
+    },
+    {
+      "epoch": 0.4842187946301057,
+      "grad_norm": 12.506934659558517,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 20343
+    },
+    {
+      "epoch": 0.4842425973531372,
+      "grad_norm": 0.7197730214181743,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20344
+    },
+    {
+      "epoch": 0.4842664000761687,
+      "grad_norm": 5.641270407244566,
+      "learning_rate": 2e-05,
+      "loss": 0.2309,
+      "step": 20345
+    },
+    {
+      "epoch": 0.4842902027992002,
+      "grad_norm": 4.2968826874665025,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 20346
+    },
+    {
+      "epoch": 0.48431400552223175,
+      "grad_norm": 1.3944410793595372,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 20347
+    },
+    {
+      "epoch": 0.48433780824526323,
+      "grad_norm": 2.088126218953423,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 20348
+    },
+    {
+      "epoch": 0.4843616109682948,
+      "grad_norm": 6.210268346732456,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 20349
+    },
+    {
+      "epoch": 0.48438541369132626,
+      "grad_norm": 3.040224062074155,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 20350
+    },
+    {
+      "epoch": 0.4844092164143578,
+      "grad_norm": 2.474896285836701,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 20351
+    },
+    {
+      "epoch": 0.48443301913738934,
+      "grad_norm": 4.289614142958782,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 20352
+    },
+    {
+      "epoch": 0.4844568218604208,
+      "grad_norm": 3.3384285894110928,
+      "learning_rate": 2e-05,
+      "loss": 0.3653,
+      "step": 20353
+    },
+    {
+      "epoch": 0.48448062458345237,
+      "grad_norm": 3.247252522367537,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 20354
+    },
+    {
+      "epoch": 0.48450442730648385,
+      "grad_norm": 3.3357396675077844,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 20355
+    },
+    {
+      "epoch": 0.4845282300295154,
+      "grad_norm": 0.06814588473465338,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 20356
+    },
+    {
+      "epoch": 0.4845520327525469,
+      "grad_norm": 2.8340177359753356,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 20357
+    },
+    {
+      "epoch": 0.4845758354755784,
+      "grad_norm": 19.043205571948082,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 20358
+    },
+    {
+      "epoch": 0.4845996381986099,
+      "grad_norm": 2.8958872826810507,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 20359
+    },
+    {
+      "epoch": 0.48462344092164145,
+      "grad_norm": 1.164720821307999,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 20360
+    },
+    {
+      "epoch": 0.48464724364467293,
+      "grad_norm": 0.04272594944579568,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 20361
+    },
+    {
+      "epoch": 0.48467104636770447,
+      "grad_norm": 3.0879482906292637,
+      "learning_rate": 2e-05,
+      "loss": 0.4874,
+      "step": 20362
+    },
+    {
+      "epoch": 0.48469484909073596,
+      "grad_norm": 0.1541437728452644,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 20363
+    },
+    {
+      "epoch": 0.4847186518137675,
+      "grad_norm": 26.60419987939274,
+      "learning_rate": 2e-05,
+      "loss": 0.2263,
+      "step": 20364
+    },
+    {
+      "epoch": 0.484742454536799,
+      "grad_norm": 4.551102492217322,
+      "learning_rate": 2e-05,
+      "loss": 0.6864,
+      "step": 20365
+    },
+    {
+      "epoch": 0.4847662572598305,
+      "grad_norm": 7.838642184448135,
+      "learning_rate": 2e-05,
+      "loss": 0.7691,
+      "step": 20366
+    },
+    {
+      "epoch": 0.48479005998286206,
+      "grad_norm": 1.0455144702749102,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 20367
+    },
+    {
+      "epoch": 0.48481386270589355,
+      "grad_norm": 1.2824225545404946,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 20368
+    },
+    {
+      "epoch": 0.4848376654289251,
+      "grad_norm": 5.198400179496742,
+      "learning_rate": 2e-05,
+      "loss": 0.9851,
+      "step": 20369
+    },
+    {
+      "epoch": 0.4848614681519566,
+      "grad_norm": 3.4378549938635796,
+      "learning_rate": 2e-05,
+      "loss": 0.5199,
+      "step": 20370
+    },
+    {
+      "epoch": 0.4848852708749881,
+      "grad_norm": 6.100150688308553,
+      "learning_rate": 2e-05,
+      "loss": 0.9976,
+      "step": 20371
+    },
+    {
+      "epoch": 0.4849090735980196,
+      "grad_norm": 1.237403227225708,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 20372
+    },
+    {
+      "epoch": 0.48493287632105114,
+      "grad_norm": 7.94005639831708,
+      "learning_rate": 2e-05,
+      "loss": 0.2789,
+      "step": 20373
+    },
+    {
+      "epoch": 0.48495667904408263,
+      "grad_norm": 10.087027000870165,
+      "learning_rate": 2e-05,
+      "loss": 0.5805,
+      "step": 20374
+    },
+    {
+      "epoch": 0.48498048176711417,
+      "grad_norm": 3.8511909203953034,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 20375
+    },
+    {
+      "epoch": 0.48500428449014565,
+      "grad_norm": 5.390182408976522,
+      "learning_rate": 2e-05,
+      "loss": 0.3787,
+      "step": 20376
+    },
+    {
+      "epoch": 0.4850280872131772,
+      "grad_norm": 3.177844477408967,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 20377
+    },
+    {
+      "epoch": 0.4850518899362087,
+      "grad_norm": 7.788257369692608,
+      "learning_rate": 2e-05,
+      "loss": 1.2628,
+      "step": 20378
+    },
+    {
+      "epoch": 0.4850756926592402,
+      "grad_norm": 3.3867075913738645,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 20379
+    },
+    {
+      "epoch": 0.4850994953822717,
+      "grad_norm": 6.4952283898624,
+      "learning_rate": 2e-05,
+      "loss": 0.3227,
+      "step": 20380
+    },
+    {
+      "epoch": 0.48512329810530325,
+      "grad_norm": 7.015020374542453,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 20381
+    },
+    {
+      "epoch": 0.4851471008283348,
+      "grad_norm": 0.5125684036086605,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20382
+    },
+    {
+      "epoch": 0.4851709035513663,
+      "grad_norm": 9.823546556364414,
+      "learning_rate": 2e-05,
+      "loss": 1.7696,
+      "step": 20383
+    },
+    {
+      "epoch": 0.4851947062743978,
+      "grad_norm": 9.027348069590577,
+      "learning_rate": 2e-05,
+      "loss": 0.5492,
+      "step": 20384
+    },
+    {
+      "epoch": 0.4852185089974293,
+      "grad_norm": 4.385082814310063,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 20385
+    },
+    {
+      "epoch": 0.48524231172046084,
+      "grad_norm": 0.20338065417226522,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 20386
+    },
+    {
+      "epoch": 0.4852661144434923,
+      "grad_norm": 3.621849789383277,
+      "learning_rate": 2e-05,
+      "loss": 0.6534,
+      "step": 20387
+    },
+    {
+      "epoch": 0.48528991716652387,
+      "grad_norm": 1.1317659431667129,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 20388
+    },
+    {
+      "epoch": 0.48531371988955535,
+      "grad_norm": 12.408180264104669,
+      "learning_rate": 2e-05,
+      "loss": 0.9253,
+      "step": 20389
+    },
+    {
+      "epoch": 0.4853375226125869,
+      "grad_norm": 3.04063953729802,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 20390
+    },
+    {
+      "epoch": 0.4853613253356184,
+      "grad_norm": 3.0364093732233344,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 20391
+    },
+    {
+      "epoch": 0.4853851280586499,
+      "grad_norm": 2.987452487248253,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 20392
+    },
+    {
+      "epoch": 0.4854089307816814,
+      "grad_norm": 2.3892167231057124,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 20393
+    },
+    {
+      "epoch": 0.48543273350471294,
+      "grad_norm": 1.6020155562491145,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 20394
+    },
+    {
+      "epoch": 0.48545653622774443,
+      "grad_norm": 3.3362687665784887,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 20395
+    },
+    {
+      "epoch": 0.48548033895077597,
+      "grad_norm": 2.120818909427691,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 20396
+    },
+    {
+      "epoch": 0.4855041416738075,
+      "grad_norm": 2.820624377888082,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 20397
+    },
+    {
+      "epoch": 0.485527944396839,
+      "grad_norm": 3.7274612997247005,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 20398
+    },
+    {
+      "epoch": 0.48555174711987054,
+      "grad_norm": 12.228948214222397,
+      "learning_rate": 2e-05,
+      "loss": 0.2568,
+      "step": 20399
+    },
+    {
+      "epoch": 0.485575549842902,
+      "grad_norm": 7.551773926752626,
+      "learning_rate": 2e-05,
+      "loss": 0.7203,
+      "step": 20400
+    },
+    {
+      "epoch": 0.48559935256593356,
+      "grad_norm": 1.7506741327651554,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 20401
+    },
+    {
+      "epoch": 0.48562315528896505,
+      "grad_norm": 7.603884882714389,
+      "learning_rate": 2e-05,
+      "loss": 1.0688,
+      "step": 20402
+    },
+    {
+      "epoch": 0.4856469580119966,
+      "grad_norm": 1.8579644563971522,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 20403
+    },
+    {
+      "epoch": 0.4856707607350281,
+      "grad_norm": 4.888125768533028,
+      "learning_rate": 2e-05,
+      "loss": 0.5923,
+      "step": 20404
+    },
+    {
+      "epoch": 0.4856945634580596,
+      "grad_norm": 7.774282393142998,
+      "learning_rate": 2e-05,
+      "loss": 0.927,
+      "step": 20405
+    },
+    {
+      "epoch": 0.4857183661810911,
+      "grad_norm": 11.551421762593174,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 20406
+    },
+    {
+      "epoch": 0.48574216890412264,
+      "grad_norm": 0.9995231576111496,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20407
+    },
+    {
+      "epoch": 0.4857659716271541,
+      "grad_norm": 3.9257531538609935,
+      "learning_rate": 2e-05,
+      "loss": 0.6553,
+      "step": 20408
+    },
+    {
+      "epoch": 0.48578977435018567,
+      "grad_norm": 2.5392532636875043,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 20409
+    },
+    {
+      "epoch": 0.48581357707321715,
+      "grad_norm": 6.299210178559034,
+      "learning_rate": 2e-05,
+      "loss": 1.2823,
+      "step": 20410
+    },
+    {
+      "epoch": 0.4858373797962487,
+      "grad_norm": 1.2856307994301364,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 20411
+    },
+    {
+      "epoch": 0.48586118251928023,
+      "grad_norm": 4.1590663270364905,
+      "learning_rate": 2e-05,
+      "loss": 0.745,
+      "step": 20412
+    },
+    {
+      "epoch": 0.4858849852423117,
+      "grad_norm": 5.183600566848296,
+      "learning_rate": 2e-05,
+      "loss": 1.0432,
+      "step": 20413
+    },
+    {
+      "epoch": 0.48590878796534326,
+      "grad_norm": 1.6221285253375028,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 20414
+    },
+    {
+      "epoch": 0.48593259068837474,
+      "grad_norm": 1.8425042817535489,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 20415
+    },
+    {
+      "epoch": 0.4859563934114063,
+      "grad_norm": 2.982699290799917,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 20416
+    },
+    {
+      "epoch": 0.48598019613443777,
+      "grad_norm": 4.07163702906187,
+      "learning_rate": 2e-05,
+      "loss": 0.5325,
+      "step": 20417
+    },
+    {
+      "epoch": 0.4860039988574693,
+      "grad_norm": 0.7839474801342468,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 20418
+    },
+    {
+      "epoch": 0.4860278015805008,
+      "grad_norm": 4.973840901061062,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 20419
+    },
+    {
+      "epoch": 0.48605160430353234,
+      "grad_norm": 3.5918719474095404,
+      "learning_rate": 2e-05,
+      "loss": 0.1618,
+      "step": 20420
+    },
+    {
+      "epoch": 0.4860754070265638,
+      "grad_norm": 0.6338304975271912,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 20421
+    },
+    {
+      "epoch": 0.48609920974959536,
+      "grad_norm": 4.36105596432662,
+      "learning_rate": 2e-05,
+      "loss": 1.0774,
+      "step": 20422
+    },
+    {
+      "epoch": 0.48612301247262685,
+      "grad_norm": 1.9450938031954612,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 20423
+    },
+    {
+      "epoch": 0.4861468151956584,
+      "grad_norm": 0.06086151760298467,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 20424
+    },
+    {
+      "epoch": 0.4861706179186899,
+      "grad_norm": 4.174322433812344,
+      "learning_rate": 2e-05,
+      "loss": 0.4922,
+      "step": 20425
+    },
+    {
+      "epoch": 0.4861944206417214,
+      "grad_norm": 0.7634465801400568,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 20426
+    },
+    {
+      "epoch": 0.48621822336475296,
+      "grad_norm": 1.6392093083730923,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 20427
+    },
+    {
+      "epoch": 0.48624202608778444,
+      "grad_norm": 2.876475117785885,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 20428
+    },
+    {
+      "epoch": 0.486265828810816,
+      "grad_norm": 4.324849549426,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 20429
+    },
+    {
+      "epoch": 0.48628963153384747,
+      "grad_norm": 1.4953581535830434,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 20430
+    },
+    {
+      "epoch": 0.486313434256879,
+      "grad_norm": 39.95836995109323,
+      "learning_rate": 2e-05,
+      "loss": 1.7109,
+      "step": 20431
+    },
+    {
+      "epoch": 0.4863372369799105,
+      "grad_norm": 7.190049993108418,
+      "learning_rate": 2e-05,
+      "loss": 0.2069,
+      "step": 20432
+    },
+    {
+      "epoch": 0.48636103970294203,
+      "grad_norm": 0.33152529057526636,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 20433
+    },
+    {
+      "epoch": 0.4863848424259735,
+      "grad_norm": 2.762976787696941,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 20434
+    },
+    {
+      "epoch": 0.48640864514900506,
+      "grad_norm": 4.294427826010393,
+      "learning_rate": 2e-05,
+      "loss": 0.362,
+      "step": 20435
+    },
+    {
+      "epoch": 0.48643244787203654,
+      "grad_norm": 0.761270663623895,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 20436
+    },
+    {
+      "epoch": 0.4864562505950681,
+      "grad_norm": 6.878613441220527,
+      "learning_rate": 2e-05,
+      "loss": 0.5741,
+      "step": 20437
+    },
+    {
+      "epoch": 0.48648005331809957,
+      "grad_norm": 0.36022126420289086,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 20438
+    },
+    {
+      "epoch": 0.4865038560411311,
+      "grad_norm": 1.4071767638930326,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 20439
+    },
+    {
+      "epoch": 0.4865276587641626,
+      "grad_norm": 5.675384734192947,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 20440
+    },
+    {
+      "epoch": 0.48655146148719414,
+      "grad_norm": 2.5029603094554553,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 20441
+    },
+    {
+      "epoch": 0.4865752642102257,
+      "grad_norm": 8.89618253158588,
+      "learning_rate": 2e-05,
+      "loss": 0.9063,
+      "step": 20442
+    },
+    {
+      "epoch": 0.48659906693325716,
+      "grad_norm": 1.7583465440147379,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 20443
+    },
+    {
+      "epoch": 0.4866228696562887,
+      "grad_norm": 6.951225584591293,
+      "learning_rate": 2e-05,
+      "loss": 0.1773,
+      "step": 20444
+    },
+    {
+      "epoch": 0.4866466723793202,
+      "grad_norm": 5.808610739016269,
+      "learning_rate": 2e-05,
+      "loss": 0.3267,
+      "step": 20445
+    },
+    {
+      "epoch": 0.48667047510235173,
+      "grad_norm": 10.286703356199451,
+      "learning_rate": 2e-05,
+      "loss": 0.6424,
+      "step": 20446
+    },
+    {
+      "epoch": 0.4866942778253832,
+      "grad_norm": 5.000960774777115,
+      "learning_rate": 2e-05,
+      "loss": 0.5518,
+      "step": 20447
+    },
+    {
+      "epoch": 0.48671808054841476,
+      "grad_norm": 9.041190499105754,
+      "learning_rate": 2e-05,
+      "loss": 0.6556,
+      "step": 20448
+    },
+    {
+      "epoch": 0.48674188327144624,
+      "grad_norm": 2.1475719338393255,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 20449
+    },
+    {
+      "epoch": 0.4867656859944778,
+      "grad_norm": 5.054346892348217,
+      "learning_rate": 2e-05,
+      "loss": 0.7174,
+      "step": 20450
+    },
+    {
+      "epoch": 0.48678948871750927,
+      "grad_norm": 6.38694887727404,
+      "learning_rate": 2e-05,
+      "loss": 0.8977,
+      "step": 20451
+    },
+    {
+      "epoch": 0.4868132914405408,
+      "grad_norm": 16.505141110155122,
+      "learning_rate": 2e-05,
+      "loss": 0.8015,
+      "step": 20452
+    },
+    {
+      "epoch": 0.4868370941635723,
+      "grad_norm": 1.921756944247693,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 20453
+    },
+    {
+      "epoch": 0.48686089688660383,
+      "grad_norm": 1.0998792554861232,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 20454
+    },
+    {
+      "epoch": 0.4868846996096353,
+      "grad_norm": 4.978676156894617,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 20455
+    },
+    {
+      "epoch": 0.48690850233266686,
+      "grad_norm": 0.7090776997583237,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 20456
+    },
+    {
+      "epoch": 0.48693230505569834,
+      "grad_norm": 1.698571508997432,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 20457
+    },
+    {
+      "epoch": 0.4869561077787299,
+      "grad_norm": 5.992711657927545,
+      "learning_rate": 2e-05,
+      "loss": 0.4556,
+      "step": 20458
+    },
+    {
+      "epoch": 0.4869799105017614,
+      "grad_norm": 9.570982923225664,
+      "learning_rate": 2e-05,
+      "loss": 0.3255,
+      "step": 20459
+    },
+    {
+      "epoch": 0.4870037132247929,
+      "grad_norm": 6.003542455528802,
+      "learning_rate": 2e-05,
+      "loss": 0.9662,
+      "step": 20460
+    },
+    {
+      "epoch": 0.48702751594782445,
+      "grad_norm": 3.708848293888448,
+      "learning_rate": 2e-05,
+      "loss": 0.4554,
+      "step": 20461
+    },
+    {
+      "epoch": 0.48705131867085594,
+      "grad_norm": 1.5320429282045018,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 20462
+    },
+    {
+      "epoch": 0.4870751213938875,
+      "grad_norm": 6.683844807466999,
+      "learning_rate": 2e-05,
+      "loss": 1.6136,
+      "step": 20463
+    },
+    {
+      "epoch": 0.48709892411691896,
+      "grad_norm": 3.4898989201818438,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 20464
+    },
+    {
+      "epoch": 0.4871227268399505,
+      "grad_norm": 0.48702984306208924,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 20465
+    },
+    {
+      "epoch": 0.487146529562982,
+      "grad_norm": 4.749359270219276,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 20466
+    },
+    {
+      "epoch": 0.48717033228601353,
+      "grad_norm": 1.1109354950739088,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 20467
+    },
+    {
+      "epoch": 0.487194135009045,
+      "grad_norm": 11.451870077690351,
+      "learning_rate": 2e-05,
+      "loss": 0.6538,
+      "step": 20468
+    },
+    {
+      "epoch": 0.48721793773207656,
+      "grad_norm": 7.883243529131536,
+      "learning_rate": 2e-05,
+      "loss": 0.3008,
+      "step": 20469
+    },
+    {
+      "epoch": 0.48724174045510804,
+      "grad_norm": 2.1708416452913215,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 20470
+    },
+    {
+      "epoch": 0.4872655431781396,
+      "grad_norm": 3.8509952838821326,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 20471
+    },
+    {
+      "epoch": 0.48728934590117107,
+      "grad_norm": 3.4322050719465227,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 20472
+    },
+    {
+      "epoch": 0.4873131486242026,
+      "grad_norm": 0.7829525519914103,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 20473
+    },
+    {
+      "epoch": 0.48733695134723415,
+      "grad_norm": 4.684772174770511,
+      "learning_rate": 2e-05,
+      "loss": 0.4284,
+      "step": 20474
+    },
+    {
+      "epoch": 0.48736075407026563,
+      "grad_norm": 5.064755796151427,
+      "learning_rate": 2e-05,
+      "loss": 0.3591,
+      "step": 20475
+    },
+    {
+      "epoch": 0.4873845567932972,
+      "grad_norm": 5.085004850378514,
+      "learning_rate": 2e-05,
+      "loss": 0.2078,
+      "step": 20476
+    },
+    {
+      "epoch": 0.48740835951632866,
+      "grad_norm": 3.840987795341644,
+      "learning_rate": 2e-05,
+      "loss": 0.39,
+      "step": 20477
+    },
+    {
+      "epoch": 0.4874321622393602,
+      "grad_norm": 6.875090448248435,
+      "learning_rate": 2e-05,
+      "loss": 0.5733,
+      "step": 20478
+    },
+    {
+      "epoch": 0.4874559649623917,
+      "grad_norm": 1.083835114682471,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 20479
+    },
+    {
+      "epoch": 0.4874797676854232,
+      "grad_norm": 3.667924682362391,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 20480
+    },
+    {
+      "epoch": 0.4875035704084547,
+      "grad_norm": 1.9720105490589415,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 20481
+    },
+    {
+      "epoch": 0.48752737313148625,
+      "grad_norm": 9.510535977994616,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 20482
+    },
+    {
+      "epoch": 0.48755117585451774,
+      "grad_norm": 5.354326356395056,
+      "learning_rate": 2e-05,
+      "loss": 0.1576,
+      "step": 20483
+    },
+    {
+      "epoch": 0.4875749785775493,
+      "grad_norm": 2.7837512027643507,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 20484
+    },
+    {
+      "epoch": 0.48759878130058076,
+      "grad_norm": 5.01713777007011,
+      "learning_rate": 2e-05,
+      "loss": 0.8377,
+      "step": 20485
+    },
+    {
+      "epoch": 0.4876225840236123,
+      "grad_norm": 18.958137649065574,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 20486
+    },
+    {
+      "epoch": 0.4876463867466438,
+      "grad_norm": 7.03921995945305,
+      "learning_rate": 2e-05,
+      "loss": 0.4182,
+      "step": 20487
+    },
+    {
+      "epoch": 0.48767018946967533,
+      "grad_norm": 3.7466509324806387,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 20488
+    },
+    {
+      "epoch": 0.48769399219270687,
+      "grad_norm": 1.4794235531948317,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 20489
+    },
+    {
+      "epoch": 0.48771779491573836,
+      "grad_norm": 13.84145691539087,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 20490
+    },
+    {
+      "epoch": 0.4877415976387699,
+      "grad_norm": 10.065806662000472,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 20491
+    },
+    {
+      "epoch": 0.4877654003618014,
+      "grad_norm": 0.489553414326755,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20492
+    },
+    {
+      "epoch": 0.4877892030848329,
+      "grad_norm": 1.8643365207387472,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 20493
+    },
+    {
+      "epoch": 0.4878130058078644,
+      "grad_norm": 2.992855279489787,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 20494
+    },
+    {
+      "epoch": 0.48783680853089595,
+      "grad_norm": 8.207917038052374,
+      "learning_rate": 2e-05,
+      "loss": 1.1029,
+      "step": 20495
+    },
+    {
+      "epoch": 0.48786061125392743,
+      "grad_norm": 2.947944209620595,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 20496
+    },
+    {
+      "epoch": 0.487884413976959,
+      "grad_norm": 2.475443116797002,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 20497
+    },
+    {
+      "epoch": 0.48790821669999046,
+      "grad_norm": 1.181894372460112,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20498
+    },
+    {
+      "epoch": 0.487932019423022,
+      "grad_norm": 4.598491506050759,
+      "learning_rate": 2e-05,
+      "loss": 0.6492,
+      "step": 20499
+    },
+    {
+      "epoch": 0.4879558221460535,
+      "grad_norm": 4.851287241687821,
+      "learning_rate": 2e-05,
+      "loss": 0.8599,
+      "step": 20500
+    },
+    {
+      "epoch": 0.487979624869085,
+      "grad_norm": 0.7734693057539226,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 20501
+    },
+    {
+      "epoch": 0.4880034275921165,
+      "grad_norm": 3.7968717671312207,
+      "learning_rate": 2e-05,
+      "loss": 0.6425,
+      "step": 20502
+    },
+    {
+      "epoch": 0.48802723031514805,
+      "grad_norm": 3.595735467522146,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 20503
+    },
+    {
+      "epoch": 0.4880510330381796,
+      "grad_norm": 1.5148657601335362,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 20504
+    },
+    {
+      "epoch": 0.4880748357612111,
+      "grad_norm": 4.7386749399578125,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 20505
+    },
+    {
+      "epoch": 0.4880986384842426,
+      "grad_norm": 4.588034704008961,
+      "learning_rate": 2e-05,
+      "loss": 0.2599,
+      "step": 20506
+    },
+    {
+      "epoch": 0.4881224412072741,
+      "grad_norm": 3.3979027733508365,
+      "learning_rate": 2e-05,
+      "loss": 0.2061,
+      "step": 20507
+    },
+    {
+      "epoch": 0.48814624393030565,
+      "grad_norm": 4.586009324254765,
+      "learning_rate": 2e-05,
+      "loss": 0.7576,
+      "step": 20508
+    },
+    {
+      "epoch": 0.48817004665333713,
+      "grad_norm": 3.360265415132383,
+      "learning_rate": 2e-05,
+      "loss": 0.5558,
+      "step": 20509
+    },
+    {
+      "epoch": 0.48819384937636867,
+      "grad_norm": 4.531283419426335,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 20510
+    },
+    {
+      "epoch": 0.48821765209940016,
+      "grad_norm": 11.120474634013187,
+      "learning_rate": 2e-05,
+      "loss": 0.1508,
+      "step": 20511
+    },
+    {
+      "epoch": 0.4882414548224317,
+      "grad_norm": 1.9190000248541192,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 20512
+    },
+    {
+      "epoch": 0.4882652575454632,
+      "grad_norm": 5.530099005762724,
+      "learning_rate": 2e-05,
+      "loss": 0.6664,
+      "step": 20513
+    },
+    {
+      "epoch": 0.4882890602684947,
+      "grad_norm": 3.945713262845486,
+      "learning_rate": 2e-05,
+      "loss": 0.4822,
+      "step": 20514
+    },
+    {
+      "epoch": 0.4883128629915262,
+      "grad_norm": 0.2195235168356012,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 20515
+    },
+    {
+      "epoch": 0.48833666571455775,
+      "grad_norm": 11.453951333621397,
+      "learning_rate": 2e-05,
+      "loss": 1.8867,
+      "step": 20516
+    },
+    {
+      "epoch": 0.48836046843758923,
+      "grad_norm": 5.497710218857442,
+      "learning_rate": 2e-05,
+      "loss": 0.9103,
+      "step": 20517
+    },
+    {
+      "epoch": 0.4883842711606208,
+      "grad_norm": 2.1356831060703465,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 20518
+    },
+    {
+      "epoch": 0.4884080738836523,
+      "grad_norm": 8.337941366145387,
+      "learning_rate": 2e-05,
+      "loss": 0.6393,
+      "step": 20519
+    },
+    {
+      "epoch": 0.4884318766066838,
+      "grad_norm": 2.2718634733211602,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 20520
+    },
+    {
+      "epoch": 0.48845567932971534,
+      "grad_norm": 4.101216261509238,
+      "learning_rate": 2e-05,
+      "loss": 0.6901,
+      "step": 20521
+    },
+    {
+      "epoch": 0.4884794820527468,
+      "grad_norm": 0.20015017521202658,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 20522
+    },
+    {
+      "epoch": 0.48850328477577837,
+      "grad_norm": 0.13276768093612687,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 20523
+    },
+    {
+      "epoch": 0.48852708749880985,
+      "grad_norm": 5.383039064266453,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 20524
+    },
+    {
+      "epoch": 0.4885508902218414,
+      "grad_norm": 6.589313237273348,
+      "learning_rate": 2e-05,
+      "loss": 0.9015,
+      "step": 20525
+    },
+    {
+      "epoch": 0.4885746929448729,
+      "grad_norm": 3.281968941587757,
+      "learning_rate": 2e-05,
+      "loss": 0.1924,
+      "step": 20526
+    },
+    {
+      "epoch": 0.4885984956679044,
+      "grad_norm": 9.796163374944122,
+      "learning_rate": 2e-05,
+      "loss": 1.1373,
+      "step": 20527
+    },
+    {
+      "epoch": 0.4886222983909359,
+      "grad_norm": 1.7463994467592672,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20528
+    },
+    {
+      "epoch": 0.48864610111396745,
+      "grad_norm": 2.5933751163508574,
+      "learning_rate": 2e-05,
+      "loss": 0.4512,
+      "step": 20529
+    },
+    {
+      "epoch": 0.48866990383699893,
+      "grad_norm": 1.4366787147387055,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 20530
+    },
+    {
+      "epoch": 0.48869370656003047,
+      "grad_norm": 2.243676251506687,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 20531
+    },
+    {
+      "epoch": 0.48871750928306196,
+      "grad_norm": 6.945094236777608,
+      "learning_rate": 2e-05,
+      "loss": 1.2838,
+      "step": 20532
+    },
+    {
+      "epoch": 0.4887413120060935,
+      "grad_norm": 5.86726324001306,
+      "learning_rate": 2e-05,
+      "loss": 1.1263,
+      "step": 20533
+    },
+    {
+      "epoch": 0.48876511472912504,
+      "grad_norm": 1.4239566402732486,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 20534
+    },
+    {
+      "epoch": 0.4887889174521565,
+      "grad_norm": 8.210344566006993,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 20535
+    },
+    {
+      "epoch": 0.48881272017518806,
+      "grad_norm": 2.8401340689378887,
+      "learning_rate": 2e-05,
+      "loss": 0.4562,
+      "step": 20536
+    },
+    {
+      "epoch": 0.48883652289821955,
+      "grad_norm": 0.19890242696134355,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20537
+    },
+    {
+      "epoch": 0.4888603256212511,
+      "grad_norm": 0.6424855349318445,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 20538
+    },
+    {
+      "epoch": 0.4888841283442826,
+      "grad_norm": 4.32040538969667,
+      "learning_rate": 2e-05,
+      "loss": 0.7611,
+      "step": 20539
+    },
+    {
+      "epoch": 0.4889079310673141,
+      "grad_norm": 5.279395816086882,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 20540
+    },
+    {
+      "epoch": 0.4889317337903456,
+      "grad_norm": 4.142929530925183,
+      "learning_rate": 2e-05,
+      "loss": 0.8148,
+      "step": 20541
+    },
+    {
+      "epoch": 0.48895553651337714,
+      "grad_norm": 1.6244192551493586,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 20542
+    },
+    {
+      "epoch": 0.4889793392364086,
+      "grad_norm": 11.01270323112948,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 20543
+    },
+    {
+      "epoch": 0.48900314195944017,
+      "grad_norm": 2.782563767827707,
+      "learning_rate": 2e-05,
+      "loss": 0.3917,
+      "step": 20544
+    },
+    {
+      "epoch": 0.48902694468247165,
+      "grad_norm": 6.689848539363341,
+      "learning_rate": 2e-05,
+      "loss": 0.2949,
+      "step": 20545
+    },
+    {
+      "epoch": 0.4890507474055032,
+      "grad_norm": 2.917130702355985,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 20546
+    },
+    {
+      "epoch": 0.4890745501285347,
+      "grad_norm": 0.3328404012194345,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 20547
+    },
+    {
+      "epoch": 0.4890983528515662,
+      "grad_norm": 3.6381627189597023,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 20548
+    },
+    {
+      "epoch": 0.48912215557459776,
+      "grad_norm": 5.346230148736489,
+      "learning_rate": 2e-05,
+      "loss": 0.7333,
+      "step": 20549
+    },
+    {
+      "epoch": 0.48914595829762925,
+      "grad_norm": 7.214490398418479,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 20550
+    },
+    {
+      "epoch": 0.4891697610206608,
+      "grad_norm": 6.657405022549431,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 20551
+    },
+    {
+      "epoch": 0.4891935637436923,
+      "grad_norm": 1.0590972380029726,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 20552
+    },
+    {
+      "epoch": 0.4892173664667238,
+      "grad_norm": 4.762408999843687,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 20553
+    },
+    {
+      "epoch": 0.4892411691897553,
+      "grad_norm": 4.01785814770953,
+      "learning_rate": 2e-05,
+      "loss": 0.2405,
+      "step": 20554
+    },
+    {
+      "epoch": 0.48926497191278684,
+      "grad_norm": 5.624569190711149,
+      "learning_rate": 2e-05,
+      "loss": 0.4366,
+      "step": 20555
+    },
+    {
+      "epoch": 0.4892887746358183,
+      "grad_norm": 3.9146766222194413,
+      "learning_rate": 2e-05,
+      "loss": 0.4136,
+      "step": 20556
+    },
+    {
+      "epoch": 0.48931257735884987,
+      "grad_norm": 2.300848524921438,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 20557
+    },
+    {
+      "epoch": 0.48933638008188135,
+      "grad_norm": 6.997506506096784,
+      "learning_rate": 2e-05,
+      "loss": 0.287,
+      "step": 20558
+    },
+    {
+      "epoch": 0.4893601828049129,
+      "grad_norm": 3.205921027461103,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 20559
+    },
+    {
+      "epoch": 0.4893839855279444,
+      "grad_norm": 0.7074604640345616,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 20560
+    },
+    {
+      "epoch": 0.4894077882509759,
+      "grad_norm": 1.5205646458738449,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 20561
+    },
+    {
+      "epoch": 0.4894315909740074,
+      "grad_norm": 13.102160503404992,
+      "learning_rate": 2e-05,
+      "loss": 0.7781,
+      "step": 20562
+    },
+    {
+      "epoch": 0.48945539369703894,
+      "grad_norm": 3.0291412610753157,
+      "learning_rate": 2e-05,
+      "loss": 0.4633,
+      "step": 20563
+    },
+    {
+      "epoch": 0.4894791964200705,
+      "grad_norm": 0.5024341742476442,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 20564
+    },
+    {
+      "epoch": 0.48950299914310197,
+      "grad_norm": 2.3180044624531124,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 20565
+    },
+    {
+      "epoch": 0.4895268018661335,
+      "grad_norm": 5.783602254917254,
+      "learning_rate": 2e-05,
+      "loss": 1.1352,
+      "step": 20566
+    },
+    {
+      "epoch": 0.489550604589165,
+      "grad_norm": 1.7364999491067536,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 20567
+    },
+    {
+      "epoch": 0.48957440731219654,
+      "grad_norm": 5.509503943450474,
+      "learning_rate": 2e-05,
+      "loss": 0.9636,
+      "step": 20568
+    },
+    {
+      "epoch": 0.489598210035228,
+      "grad_norm": 4.326682663409989,
+      "learning_rate": 2e-05,
+      "loss": 0.207,
+      "step": 20569
+    },
+    {
+      "epoch": 0.48962201275825956,
+      "grad_norm": 16.498478556235458,
+      "learning_rate": 2e-05,
+      "loss": 0.1626,
+      "step": 20570
+    },
+    {
+      "epoch": 0.48964581548129105,
+      "grad_norm": 6.348910906981853,
+      "learning_rate": 2e-05,
+      "loss": 1.4013,
+      "step": 20571
+    },
+    {
+      "epoch": 0.4896696182043226,
+      "grad_norm": 1.4933539706207735,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 20572
+    },
+    {
+      "epoch": 0.4896934209273541,
+      "grad_norm": 10.417971480623772,
+      "learning_rate": 2e-05,
+      "loss": 1.3085,
+      "step": 20573
+    },
+    {
+      "epoch": 0.4897172236503856,
+      "grad_norm": 2.788363609587,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 20574
+    },
+    {
+      "epoch": 0.4897410263734171,
+      "grad_norm": 10.399618082945057,
+      "learning_rate": 2e-05,
+      "loss": 1.0904,
+      "step": 20575
+    },
+    {
+      "epoch": 0.48976482909644864,
+      "grad_norm": 1.3021789634965313,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 20576
+    },
+    {
+      "epoch": 0.4897886318194801,
+      "grad_norm": 8.509501733710303,
+      "learning_rate": 2e-05,
+      "loss": 0.3462,
+      "step": 20577
+    },
+    {
+      "epoch": 0.48981243454251167,
+      "grad_norm": 1.2473495457279407,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 20578
+    },
+    {
+      "epoch": 0.48983623726554315,
+      "grad_norm": 2.582864740787608,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 20579
+    },
+    {
+      "epoch": 0.4898600399885747,
+      "grad_norm": 2.0877887976128635,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 20580
+    },
+    {
+      "epoch": 0.48988384271160623,
+      "grad_norm": 2.5168939952510314,
+      "learning_rate": 2e-05,
+      "loss": 0.3237,
+      "step": 20581
+    },
+    {
+      "epoch": 0.4899076454346377,
+      "grad_norm": 1.9306172812847697,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 20582
+    },
+    {
+      "epoch": 0.48993144815766926,
+      "grad_norm": 1.7591765681189995,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 20583
+    },
+    {
+      "epoch": 0.48995525088070074,
+      "grad_norm": 2.031917767842319,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 20584
+    },
+    {
+      "epoch": 0.4899790536037323,
+      "grad_norm": 1.3479822446684948,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 20585
+    },
+    {
+      "epoch": 0.49000285632676377,
+      "grad_norm": 7.59150805907205,
+      "learning_rate": 2e-05,
+      "loss": 1.0034,
+      "step": 20586
+    },
+    {
+      "epoch": 0.4900266590497953,
+      "grad_norm": 0.42468958444342025,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 20587
+    },
+    {
+      "epoch": 0.4900504617728268,
+      "grad_norm": 1.3512755868225597,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 20588
+    },
+    {
+      "epoch": 0.49007426449585834,
+      "grad_norm": 3.8745614628988676,
+      "learning_rate": 2e-05,
+      "loss": 0.42,
+      "step": 20589
+    },
+    {
+      "epoch": 0.4900980672188898,
+      "grad_norm": 1.1970936243289392,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 20590
+    },
+    {
+      "epoch": 0.49012186994192136,
+      "grad_norm": 6.204201012369708,
+      "learning_rate": 2e-05,
+      "loss": 1.2674,
+      "step": 20591
+    },
+    {
+      "epoch": 0.49014567266495285,
+      "grad_norm": 0.22709640383319574,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 20592
+    },
+    {
+      "epoch": 0.4901694753879844,
+      "grad_norm": 2.333289300949297,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 20593
+    },
+    {
+      "epoch": 0.4901932781110159,
+      "grad_norm": 0.7174271455160887,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20594
+    },
+    {
+      "epoch": 0.4902170808340474,
+      "grad_norm": 4.059293487288984,
+      "learning_rate": 2e-05,
+      "loss": 0.5136,
+      "step": 20595
+    },
+    {
+      "epoch": 0.49024088355707895,
+      "grad_norm": 4.5581409579464545,
+      "learning_rate": 2e-05,
+      "loss": 0.1902,
+      "step": 20596
+    },
+    {
+      "epoch": 0.49026468628011044,
+      "grad_norm": 6.852625140031555,
+      "learning_rate": 2e-05,
+      "loss": 1.0031,
+      "step": 20597
+    },
+    {
+      "epoch": 0.490288489003142,
+      "grad_norm": 2.735031688989322,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 20598
+    },
+    {
+      "epoch": 0.49031229172617347,
+      "grad_norm": 0.6379711331404091,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 20599
+    },
+    {
+      "epoch": 0.490336094449205,
+      "grad_norm": 0.24682936229482005,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 20600
+    },
+    {
+      "epoch": 0.4903598971722365,
+      "grad_norm": 6.312413573442344,
+      "learning_rate": 2e-05,
+      "loss": 0.885,
+      "step": 20601
+    },
+    {
+      "epoch": 0.49038369989526803,
+      "grad_norm": 0.07401543701337951,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 20602
+    },
+    {
+      "epoch": 0.4904075026182995,
+      "grad_norm": 9.400022239055392,
+      "learning_rate": 2e-05,
+      "loss": 0.4729,
+      "step": 20603
+    },
+    {
+      "epoch": 0.49043130534133106,
+      "grad_norm": 3.743447391079495,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 20604
+    },
+    {
+      "epoch": 0.49045510806436254,
+      "grad_norm": 2.2212428323149536,
+      "learning_rate": 2e-05,
+      "loss": 0.0941,
+      "step": 20605
+    },
+    {
+      "epoch": 0.4904789107873941,
+      "grad_norm": 10.150296874941278,
+      "learning_rate": 2e-05,
+      "loss": 1.0802,
+      "step": 20606
+    },
+    {
+      "epoch": 0.49050271351042557,
+      "grad_norm": 4.63882000128209,
+      "learning_rate": 2e-05,
+      "loss": 0.9876,
+      "step": 20607
+    },
+    {
+      "epoch": 0.4905265162334571,
+      "grad_norm": 4.265193090811557,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 20608
+    },
+    {
+      "epoch": 0.4905503189564886,
+      "grad_norm": 4.870079890840539,
+      "learning_rate": 2e-05,
+      "loss": 0.5211,
+      "step": 20609
+    },
+    {
+      "epoch": 0.49057412167952014,
+      "grad_norm": 5.080005207499443,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 20610
+    },
+    {
+      "epoch": 0.4905979244025517,
+      "grad_norm": 3.3563499287394167,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 20611
+    },
+    {
+      "epoch": 0.49062172712558316,
+      "grad_norm": 3.642955778690492,
+      "learning_rate": 2e-05,
+      "loss": 0.7301,
+      "step": 20612
+    },
+    {
+      "epoch": 0.4906455298486147,
+      "grad_norm": 7.536828623003294,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 20613
+    },
+    {
+      "epoch": 0.4906693325716462,
+      "grad_norm": 2.0317616434795727,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20614
+    },
+    {
+      "epoch": 0.49069313529467773,
+      "grad_norm": 0.7022439419368217,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 20615
+    },
+    {
+      "epoch": 0.4907169380177092,
+      "grad_norm": 8.494651551555991,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 20616
+    },
+    {
+      "epoch": 0.49074074074074076,
+      "grad_norm": 6.173462039512166,
+      "learning_rate": 2e-05,
+      "loss": 0.9945,
+      "step": 20617
+    },
+    {
+      "epoch": 0.49076454346377224,
+      "grad_norm": 2.6445136393275983,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 20618
+    },
+    {
+      "epoch": 0.4907883461868038,
+      "grad_norm": 2.193222924082701,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 20619
+    },
+    {
+      "epoch": 0.49081214890983527,
+      "grad_norm": 1.4788324061275149,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 20620
+    },
+    {
+      "epoch": 0.4908359516328668,
+      "grad_norm": 1.356092704512587,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 20621
+    },
+    {
+      "epoch": 0.4908597543558983,
+      "grad_norm": 1.939358261142143,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 20622
+    },
+    {
+      "epoch": 0.49088355707892983,
+      "grad_norm": 1.4501260372357263,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 20623
+    },
+    {
+      "epoch": 0.4909073598019613,
+      "grad_norm": 3.6671845565077943,
+      "learning_rate": 2e-05,
+      "loss": 0.4312,
+      "step": 20624
+    },
+    {
+      "epoch": 0.49093116252499286,
+      "grad_norm": 2.941896169969228,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 20625
+    },
+    {
+      "epoch": 0.4909549652480244,
+      "grad_norm": 6.67209213517744,
+      "learning_rate": 2e-05,
+      "loss": 0.3892,
+      "step": 20626
+    },
+    {
+      "epoch": 0.4909787679710559,
+      "grad_norm": 7.123726000013672,
+      "learning_rate": 2e-05,
+      "loss": 0.2569,
+      "step": 20627
+    },
+    {
+      "epoch": 0.4910025706940874,
+      "grad_norm": 8.128251665453266,
+      "learning_rate": 2e-05,
+      "loss": 0.3164,
+      "step": 20628
+    },
+    {
+      "epoch": 0.4910263734171189,
+      "grad_norm": 3.850212675981821,
+      "learning_rate": 2e-05,
+      "loss": 0.4581,
+      "step": 20629
+    },
+    {
+      "epoch": 0.49105017614015045,
+      "grad_norm": 4.722330939674953,
+      "learning_rate": 2e-05,
+      "loss": 0.6843,
+      "step": 20630
+    },
+    {
+      "epoch": 0.49107397886318194,
+      "grad_norm": 4.937077327707088,
+      "learning_rate": 2e-05,
+      "loss": 0.6127,
+      "step": 20631
+    },
+    {
+      "epoch": 0.4910977815862135,
+      "grad_norm": 1.1893619543179235,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 20632
+    },
+    {
+      "epoch": 0.49112158430924496,
+      "grad_norm": 4.4509824315429505,
+      "learning_rate": 2e-05,
+      "loss": 1.074,
+      "step": 20633
+    },
+    {
+      "epoch": 0.4911453870322765,
+      "grad_norm": 1.0578415211094987,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 20634
+    },
+    {
+      "epoch": 0.491169189755308,
+      "grad_norm": 4.4969867201796605,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 20635
+    },
+    {
+      "epoch": 0.49119299247833953,
+      "grad_norm": 1.9882141685745345,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 20636
+    },
+    {
+      "epoch": 0.491216795201371,
+      "grad_norm": 5.799779307268035,
+      "learning_rate": 2e-05,
+      "loss": 0.8302,
+      "step": 20637
+    },
+    {
+      "epoch": 0.49124059792440256,
+      "grad_norm": 6.249906184435885,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 20638
+    },
+    {
+      "epoch": 0.49126440064743404,
+      "grad_norm": 0.660563016343457,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 20639
+    },
+    {
+      "epoch": 0.4912882033704656,
+      "grad_norm": 2.5085357566441604,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 20640
+    },
+    {
+      "epoch": 0.4913120060934971,
+      "grad_norm": 0.11005978955735427,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 20641
+    },
+    {
+      "epoch": 0.4913358088165286,
+      "grad_norm": 5.158665679766436,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 20642
+    },
+    {
+      "epoch": 0.49135961153956015,
+      "grad_norm": 6.663261976421233,
+      "learning_rate": 2e-05,
+      "loss": 0.2255,
+      "step": 20643
+    },
+    {
+      "epoch": 0.49138341426259163,
+      "grad_norm": 9.802750820327056,
+      "learning_rate": 2e-05,
+      "loss": 0.6697,
+      "step": 20644
+    },
+    {
+      "epoch": 0.4914072169856232,
+      "grad_norm": 6.400437175393217,
+      "learning_rate": 2e-05,
+      "loss": 0.7573,
+      "step": 20645
+    },
+    {
+      "epoch": 0.49143101970865466,
+      "grad_norm": 0.12428522644589501,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 20646
+    },
+    {
+      "epoch": 0.4914548224316862,
+      "grad_norm": 2.6410834484170387,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 20647
+    },
+    {
+      "epoch": 0.4914786251547177,
+      "grad_norm": 2.7140779882635786,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 20648
+    },
+    {
+      "epoch": 0.4915024278777492,
+      "grad_norm": 10.985234963678366,
+      "learning_rate": 2e-05,
+      "loss": 0.6568,
+      "step": 20649
+    },
+    {
+      "epoch": 0.4915262306007807,
+      "grad_norm": 6.165712655143977,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 20650
+    },
+    {
+      "epoch": 0.49155003332381225,
+      "grad_norm": 4.352508954124622,
+      "learning_rate": 2e-05,
+      "loss": 0.3112,
+      "step": 20651
+    },
+    {
+      "epoch": 0.49157383604684374,
+      "grad_norm": 2.1541189321675454,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 20652
+    },
+    {
+      "epoch": 0.4915976387698753,
+      "grad_norm": 5.106635865074501,
+      "learning_rate": 2e-05,
+      "loss": 0.188,
+      "step": 20653
+    },
+    {
+      "epoch": 0.49162144149290676,
+      "grad_norm": 2.303426570609813,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 20654
+    },
+    {
+      "epoch": 0.4916452442159383,
+      "grad_norm": 0.08411127564120915,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 20655
+    },
+    {
+      "epoch": 0.49166904693896984,
+      "grad_norm": 7.405524303509432,
+      "learning_rate": 2e-05,
+      "loss": 1.2212,
+      "step": 20656
+    },
+    {
+      "epoch": 0.49169284966200133,
+      "grad_norm": 4.651738912905814,
+      "learning_rate": 2e-05,
+      "loss": 0.5462,
+      "step": 20657
+    },
+    {
+      "epoch": 0.49171665238503287,
+      "grad_norm": 7.547669987989415,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 20658
+    },
+    {
+      "epoch": 0.49174045510806436,
+      "grad_norm": 1.3992033581715897,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20659
+    },
+    {
+      "epoch": 0.4917642578310959,
+      "grad_norm": 0.9963110852637279,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20660
+    },
+    {
+      "epoch": 0.4917880605541274,
+      "grad_norm": 5.145691089784844,
+      "learning_rate": 2e-05,
+      "loss": 0.1847,
+      "step": 20661
+    },
+    {
+      "epoch": 0.4918118632771589,
+      "grad_norm": 8.699719886763418,
+      "learning_rate": 2e-05,
+      "loss": 0.7124,
+      "step": 20662
+    },
+    {
+      "epoch": 0.4918356660001904,
+      "grad_norm": 9.819628780742402,
+      "learning_rate": 2e-05,
+      "loss": 0.4965,
+      "step": 20663
+    },
+    {
+      "epoch": 0.49185946872322195,
+      "grad_norm": 4.1485584662078185,
+      "learning_rate": 2e-05,
+      "loss": 0.9554,
+      "step": 20664
+    },
+    {
+      "epoch": 0.49188327144625343,
+      "grad_norm": 3.792969055603718,
+      "learning_rate": 2e-05,
+      "loss": 0.3282,
+      "step": 20665
+    },
+    {
+      "epoch": 0.491907074169285,
+      "grad_norm": 13.109232308455413,
+      "learning_rate": 2e-05,
+      "loss": 0.3368,
+      "step": 20666
+    },
+    {
+      "epoch": 0.49193087689231646,
+      "grad_norm": 0.7630612424010561,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 20667
+    },
+    {
+      "epoch": 0.491954679615348,
+      "grad_norm": 4.764137971527624,
+      "learning_rate": 2e-05,
+      "loss": 0.5657,
+      "step": 20668
+    },
+    {
+      "epoch": 0.4919784823383795,
+      "grad_norm": 67.80217947367525,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 20669
+    },
+    {
+      "epoch": 0.492002285061411,
+      "grad_norm": 2.413982923371177,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 20670
+    },
+    {
+      "epoch": 0.49202608778444257,
+      "grad_norm": 5.007001337121891,
+      "learning_rate": 2e-05,
+      "loss": 0.1953,
+      "step": 20671
+    },
+    {
+      "epoch": 0.49204989050747405,
+      "grad_norm": 0.19004874426217386,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 20672
+    },
+    {
+      "epoch": 0.4920736932305056,
+      "grad_norm": 4.840002578129956,
+      "learning_rate": 2e-05,
+      "loss": 0.4095,
+      "step": 20673
+    },
+    {
+      "epoch": 0.4920974959535371,
+      "grad_norm": 1.730945241411015,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 20674
+    },
+    {
+      "epoch": 0.4921212986765686,
+      "grad_norm": 14.010013206568066,
+      "learning_rate": 2e-05,
+      "loss": 0.2598,
+      "step": 20675
+    },
+    {
+      "epoch": 0.4921451013996001,
+      "grad_norm": 5.092838275040772,
+      "learning_rate": 2e-05,
+      "loss": 0.6033,
+      "step": 20676
+    },
+    {
+      "epoch": 0.49216890412263165,
+      "grad_norm": 0.8054367292054501,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 20677
+    },
+    {
+      "epoch": 0.49219270684566313,
+      "grad_norm": 0.8750805407865695,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 20678
+    },
+    {
+      "epoch": 0.49221650956869467,
+      "grad_norm": 2.3400256667163037,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 20679
+    },
+    {
+      "epoch": 0.49224031229172616,
+      "grad_norm": 13.491306920567796,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 20680
+    },
+    {
+      "epoch": 0.4922641150147577,
+      "grad_norm": 1.616045808076974,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 20681
+    },
+    {
+      "epoch": 0.4922879177377892,
+      "grad_norm": 4.877731473437501,
+      "learning_rate": 2e-05,
+      "loss": 0.8142,
+      "step": 20682
+    },
+    {
+      "epoch": 0.4923117204608207,
+      "grad_norm": 2.379838573578563,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 20683
+    },
+    {
+      "epoch": 0.4923355231838522,
+      "grad_norm": 5.6382143455552045,
+      "learning_rate": 2e-05,
+      "loss": 1.0701,
+      "step": 20684
+    },
+    {
+      "epoch": 0.49235932590688375,
+      "grad_norm": 2.664061089833277,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 20685
+    },
+    {
+      "epoch": 0.4923831286299153,
+      "grad_norm": 6.19014893428133,
+      "learning_rate": 2e-05,
+      "loss": 0.8754,
+      "step": 20686
+    },
+    {
+      "epoch": 0.4924069313529468,
+      "grad_norm": 4.416625501167007,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 20687
+    },
+    {
+      "epoch": 0.4924307340759783,
+      "grad_norm": 0.18600692421809079,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20688
+    },
+    {
+      "epoch": 0.4924545367990098,
+      "grad_norm": 0.8255058718909148,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 20689
+    },
+    {
+      "epoch": 0.49247833952204134,
+      "grad_norm": 5.8033154856934335,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 20690
+    },
+    {
+      "epoch": 0.4925021422450728,
+      "grad_norm": 16.992398237480874,
+      "learning_rate": 2e-05,
+      "loss": 0.1758,
+      "step": 20691
+    },
+    {
+      "epoch": 0.49252594496810437,
+      "grad_norm": 2.083344565437772,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 20692
+    },
+    {
+      "epoch": 0.49254974769113585,
+      "grad_norm": 3.6788182404269647,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 20693
+    },
+    {
+      "epoch": 0.4925735504141674,
+      "grad_norm": 9.98528394123587,
+      "learning_rate": 2e-05,
+      "loss": 0.3815,
+      "step": 20694
+    },
+    {
+      "epoch": 0.4925973531371989,
+      "grad_norm": 5.154316418452788,
+      "learning_rate": 2e-05,
+      "loss": 1.2615,
+      "step": 20695
+    },
+    {
+      "epoch": 0.4926211558602304,
+      "grad_norm": 0.3212549938134577,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 20696
+    },
+    {
+      "epoch": 0.4926449585832619,
+      "grad_norm": 6.493064993555714,
+      "learning_rate": 2e-05,
+      "loss": 0.3544,
+      "step": 20697
+    },
+    {
+      "epoch": 0.49266876130629345,
+      "grad_norm": 8.580770392728283,
+      "learning_rate": 2e-05,
+      "loss": 1.762,
+      "step": 20698
+    },
+    {
+      "epoch": 0.49269256402932493,
+      "grad_norm": 4.815538344932967,
+      "learning_rate": 2e-05,
+      "loss": 0.8191,
+      "step": 20699
+    },
+    {
+      "epoch": 0.49271636675235647,
+      "grad_norm": 3.020663779284953,
+      "learning_rate": 2e-05,
+      "loss": 0.2549,
+      "step": 20700
+    },
+    {
+      "epoch": 0.49274016947538796,
+      "grad_norm": 7.196417444406721,
+      "learning_rate": 2e-05,
+      "loss": 0.4365,
+      "step": 20701
+    },
+    {
+      "epoch": 0.4927639721984195,
+      "grad_norm": 1.1552815776399457,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 20702
+    },
+    {
+      "epoch": 0.49278777492145104,
+      "grad_norm": 1.6395962766487036,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 20703
+    },
+    {
+      "epoch": 0.4928115776444825,
+      "grad_norm": 9.26086261388833,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 20704
+    },
+    {
+      "epoch": 0.49283538036751406,
+      "grad_norm": 3.8968908889169778,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 20705
+    },
+    {
+      "epoch": 0.49285918309054555,
+      "grad_norm": 1.6346985390683957,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 20706
+    },
+    {
+      "epoch": 0.4928829858135771,
+      "grad_norm": 5.326994033814088,
+      "learning_rate": 2e-05,
+      "loss": 1.148,
+      "step": 20707
+    },
+    {
+      "epoch": 0.4929067885366086,
+      "grad_norm": 5.029671005514113,
+      "learning_rate": 2e-05,
+      "loss": 0.2929,
+      "step": 20708
+    },
+    {
+      "epoch": 0.4929305912596401,
+      "grad_norm": 0.22887688208840123,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 20709
+    },
+    {
+      "epoch": 0.4929543939826716,
+      "grad_norm": 0.1426009699561468,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 20710
+    },
+    {
+      "epoch": 0.49297819670570314,
+      "grad_norm": 9.35873175258132,
+      "learning_rate": 2e-05,
+      "loss": 0.3319,
+      "step": 20711
+    },
+    {
+      "epoch": 0.4930019994287346,
+      "grad_norm": 52.697164873061446,
+      "learning_rate": 2e-05,
+      "loss": 0.8658,
+      "step": 20712
+    },
+    {
+      "epoch": 0.49302580215176617,
+      "grad_norm": 6.335338572578975,
+      "learning_rate": 2e-05,
+      "loss": 1.0818,
+      "step": 20713
+    },
+    {
+      "epoch": 0.49304960487479765,
+      "grad_norm": 8.514598080021642,
+      "learning_rate": 2e-05,
+      "loss": 1.1619,
+      "step": 20714
+    },
+    {
+      "epoch": 0.4930734075978292,
+      "grad_norm": 0.423822949104966,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 20715
+    },
+    {
+      "epoch": 0.4930972103208607,
+      "grad_norm": 3.3541449223712823,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 20716
+    },
+    {
+      "epoch": 0.4931210130438922,
+      "grad_norm": 2.151497224425466,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 20717
+    },
+    {
+      "epoch": 0.49314481576692376,
+      "grad_norm": 8.407267066649569,
+      "learning_rate": 2e-05,
+      "loss": 1.0832,
+      "step": 20718
+    },
+    {
+      "epoch": 0.49316861848995525,
+      "grad_norm": 9.523078205674539,
+      "learning_rate": 2e-05,
+      "loss": 0.5191,
+      "step": 20719
+    },
+    {
+      "epoch": 0.4931924212129868,
+      "grad_norm": 4.211249725100175,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 20720
+    },
+    {
+      "epoch": 0.49321622393601827,
+      "grad_norm": 5.793323399506329,
+      "learning_rate": 2e-05,
+      "loss": 0.2536,
+      "step": 20721
+    },
+    {
+      "epoch": 0.4932400266590498,
+      "grad_norm": 5.872190546005043,
+      "learning_rate": 2e-05,
+      "loss": 0.236,
+      "step": 20722
+    },
+    {
+      "epoch": 0.4932638293820813,
+      "grad_norm": 0.44522671236169703,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 20723
+    },
+    {
+      "epoch": 0.49328763210511284,
+      "grad_norm": 0.28749333732381904,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 20724
+    },
+    {
+      "epoch": 0.4933114348281443,
+      "grad_norm": 2.4220567330092275,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 20725
+    },
+    {
+      "epoch": 0.49333523755117586,
+      "grad_norm": 5.716865412732357,
+      "learning_rate": 2e-05,
+      "loss": 0.3483,
+      "step": 20726
+    },
+    {
+      "epoch": 0.49335904027420735,
+      "grad_norm": 6.015288362671383,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 20727
+    },
+    {
+      "epoch": 0.4933828429972389,
+      "grad_norm": 3.0773544822532286,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 20728
+    },
+    {
+      "epoch": 0.4934066457202704,
+      "grad_norm": 12.933161190246286,
+      "learning_rate": 2e-05,
+      "loss": 0.8499,
+      "step": 20729
+    },
+    {
+      "epoch": 0.4934304484433019,
+      "grad_norm": 6.770028307259957,
+      "learning_rate": 2e-05,
+      "loss": 0.1755,
+      "step": 20730
+    },
+    {
+      "epoch": 0.4934542511663334,
+      "grad_norm": 1.320590288202918,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 20731
+    },
+    {
+      "epoch": 0.49347805388936494,
+      "grad_norm": 7.155319492702635,
+      "learning_rate": 2e-05,
+      "loss": 0.7333,
+      "step": 20732
+    },
+    {
+      "epoch": 0.4935018566123965,
+      "grad_norm": 6.589569886877759,
+      "learning_rate": 2e-05,
+      "loss": 0.6961,
+      "step": 20733
+    },
+    {
+      "epoch": 0.49352565933542797,
+      "grad_norm": 6.2923625698472465,
+      "learning_rate": 2e-05,
+      "loss": 1.2069,
+      "step": 20734
+    },
+    {
+      "epoch": 0.4935494620584595,
+      "grad_norm": 5.427129183903542,
+      "learning_rate": 2e-05,
+      "loss": 0.7242,
+      "step": 20735
+    },
+    {
+      "epoch": 0.493573264781491,
+      "grad_norm": 1.7903969431780509,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 20736
+    },
+    {
+      "epoch": 0.49359706750452254,
+      "grad_norm": 0.5905375334279961,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 20737
+    },
+    {
+      "epoch": 0.493620870227554,
+      "grad_norm": 3.819256739832579,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 20738
+    },
+    {
+      "epoch": 0.49364467295058556,
+      "grad_norm": 14.424723563275434,
+      "learning_rate": 2e-05,
+      "loss": 0.9356,
+      "step": 20739
+    },
+    {
+      "epoch": 0.49366847567361705,
+      "grad_norm": 2.356476469170137,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 20740
+    },
+    {
+      "epoch": 0.4936922783966486,
+      "grad_norm": 3.920225038057924,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 20741
+    },
+    {
+      "epoch": 0.4937160811196801,
+      "grad_norm": 7.635600874757694,
+      "learning_rate": 2e-05,
+      "loss": 0.1701,
+      "step": 20742
+    },
+    {
+      "epoch": 0.4937398838427116,
+      "grad_norm": 3.7770812546597985,
+      "learning_rate": 2e-05,
+      "loss": 0.7802,
+      "step": 20743
+    },
+    {
+      "epoch": 0.4937636865657431,
+      "grad_norm": 8.424142077672382,
+      "learning_rate": 2e-05,
+      "loss": 0.608,
+      "step": 20744
+    },
+    {
+      "epoch": 0.49378748928877464,
+      "grad_norm": 5.374771883058883,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 20745
+    },
+    {
+      "epoch": 0.4938112920118061,
+      "grad_norm": 6.79110420724366,
+      "learning_rate": 2e-05,
+      "loss": 0.9613,
+      "step": 20746
+    },
+    {
+      "epoch": 0.49383509473483767,
+      "grad_norm": 4.322754974843141,
+      "learning_rate": 2e-05,
+      "loss": 0.3919,
+      "step": 20747
+    },
+    {
+      "epoch": 0.4938588974578692,
+      "grad_norm": 6.345749192653524,
+      "learning_rate": 2e-05,
+      "loss": 1.3931,
+      "step": 20748
+    },
+    {
+      "epoch": 0.4938827001809007,
+      "grad_norm": 4.247692778755299,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20749
+    },
+    {
+      "epoch": 0.49390650290393223,
+      "grad_norm": 9.684078928857195,
+      "learning_rate": 2e-05,
+      "loss": 1.5603,
+      "step": 20750
+    },
+    {
+      "epoch": 0.4939303056269637,
+      "grad_norm": 4.076506719209243,
+      "learning_rate": 2e-05,
+      "loss": 0.5152,
+      "step": 20751
+    },
+    {
+      "epoch": 0.49395410834999526,
+      "grad_norm": 5.365104028572692,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 20752
+    },
+    {
+      "epoch": 0.49397791107302674,
+      "grad_norm": 4.3087208183823975,
+      "learning_rate": 2e-05,
+      "loss": 0.7838,
+      "step": 20753
+    },
+    {
+      "epoch": 0.4940017137960583,
+      "grad_norm": 6.143276304496619,
+      "learning_rate": 2e-05,
+      "loss": 0.246,
+      "step": 20754
+    },
+    {
+      "epoch": 0.49402551651908977,
+      "grad_norm": 2.916386655886814,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 20755
+    },
+    {
+      "epoch": 0.4940493192421213,
+      "grad_norm": 2.005736288791427,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 20756
+    },
+    {
+      "epoch": 0.4940731219651528,
+      "grad_norm": 2.0692923365785467,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 20757
+    },
+    {
+      "epoch": 0.49409692468818434,
+      "grad_norm": 0.22742168058331141,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20758
+    },
+    {
+      "epoch": 0.4941207274112158,
+      "grad_norm": 3.4444488156425686,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 20759
+    },
+    {
+      "epoch": 0.49414453013424736,
+      "grad_norm": 0.09373305957024003,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 20760
+    },
+    {
+      "epoch": 0.49416833285727885,
+      "grad_norm": 7.685690779210229,
+      "learning_rate": 2e-05,
+      "loss": 0.5165,
+      "step": 20761
+    },
+    {
+      "epoch": 0.4941921355803104,
+      "grad_norm": 2.492446789014186,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 20762
+    },
+    {
+      "epoch": 0.49421593830334193,
+      "grad_norm": 3.282594235728466,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 20763
+    },
+    {
+      "epoch": 0.4942397410263734,
+      "grad_norm": 4.824354795264007,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 20764
+    },
+    {
+      "epoch": 0.49426354374940495,
+      "grad_norm": 4.458705205562984,
+      "learning_rate": 2e-05,
+      "loss": 0.3913,
+      "step": 20765
+    },
+    {
+      "epoch": 0.49428734647243644,
+      "grad_norm": 2.503297859095163,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 20766
+    },
+    {
+      "epoch": 0.494311149195468,
+      "grad_norm": 6.180419840488231,
+      "learning_rate": 2e-05,
+      "loss": 0.8276,
+      "step": 20767
+    },
+    {
+      "epoch": 0.49433495191849947,
+      "grad_norm": 2.216148918774762,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 20768
+    },
+    {
+      "epoch": 0.494358754641531,
+      "grad_norm": 3.159958324662718,
+      "learning_rate": 2e-05,
+      "loss": 0.1265,
+      "step": 20769
+    },
+    {
+      "epoch": 0.4943825573645625,
+      "grad_norm": 2.624932725958409,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 20770
+    },
+    {
+      "epoch": 0.49440636008759403,
+      "grad_norm": 9.05278059633656,
+      "learning_rate": 2e-05,
+      "loss": 1.4379,
+      "step": 20771
+    },
+    {
+      "epoch": 0.4944301628106255,
+      "grad_norm": 6.341476417541861,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 20772
+    },
+    {
+      "epoch": 0.49445396553365706,
+      "grad_norm": 0.7945189672345296,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 20773
+    },
+    {
+      "epoch": 0.49447776825668854,
+      "grad_norm": 9.486910003650765,
+      "learning_rate": 2e-05,
+      "loss": 1.1341,
+      "step": 20774
+    },
+    {
+      "epoch": 0.4945015709797201,
+      "grad_norm": 5.107290313770045,
+      "learning_rate": 2e-05,
+      "loss": 0.8459,
+      "step": 20775
+    },
+    {
+      "epoch": 0.49452537370275157,
+      "grad_norm": 3.1215988354556012,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 20776
+    },
+    {
+      "epoch": 0.4945491764257831,
+      "grad_norm": 7.386701790650735,
+      "learning_rate": 2e-05,
+      "loss": 0.6535,
+      "step": 20777
+    },
+    {
+      "epoch": 0.49457297914881465,
+      "grad_norm": 1.4253680357916199,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 20778
+    },
+    {
+      "epoch": 0.49459678187184614,
+      "grad_norm": 4.18962990946331,
+      "learning_rate": 2e-05,
+      "loss": 0.9648,
+      "step": 20779
+    },
+    {
+      "epoch": 0.4946205845948777,
+      "grad_norm": 10.985994008429083,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 20780
+    },
+    {
+      "epoch": 0.49464438731790916,
+      "grad_norm": 5.085438540842573,
+      "learning_rate": 2e-05,
+      "loss": 0.2512,
+      "step": 20781
+    },
+    {
+      "epoch": 0.4946681900409407,
+      "grad_norm": 3.9423572647762914,
+      "learning_rate": 2e-05,
+      "loss": 0.4311,
+      "step": 20782
+    },
+    {
+      "epoch": 0.4946919927639722,
+      "grad_norm": 2.8853092737567194,
+      "learning_rate": 2e-05,
+      "loss": 0.4534,
+      "step": 20783
+    },
+    {
+      "epoch": 0.49471579548700373,
+      "grad_norm": 0.9930595883851427,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 20784
+    },
+    {
+      "epoch": 0.4947395982100352,
+      "grad_norm": 9.099330223710133,
+      "learning_rate": 2e-05,
+      "loss": 0.4879,
+      "step": 20785
+    },
+    {
+      "epoch": 0.49476340093306675,
+      "grad_norm": 4.118555040882856,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 20786
+    },
+    {
+      "epoch": 0.49478720365609824,
+      "grad_norm": 6.1633414036468075,
+      "learning_rate": 2e-05,
+      "loss": 1.0482,
+      "step": 20787
+    },
+    {
+      "epoch": 0.4948110063791298,
+      "grad_norm": 2.236925831005588,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 20788
+    },
+    {
+      "epoch": 0.49483480910216127,
+      "grad_norm": 1.6878362423049729,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 20789
+    },
+    {
+      "epoch": 0.4948586118251928,
+      "grad_norm": 1.2923177332201974,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 20790
+    },
+    {
+      "epoch": 0.4948824145482243,
+      "grad_norm": 19.176871264051048,
+      "learning_rate": 2e-05,
+      "loss": 0.3781,
+      "step": 20791
+    },
+    {
+      "epoch": 0.49490621727125583,
+      "grad_norm": 2.624504420246937,
+      "learning_rate": 2e-05,
+      "loss": 0.1913,
+      "step": 20792
+    },
+    {
+      "epoch": 0.4949300199942874,
+      "grad_norm": 6.318176808086073,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 20793
+    },
+    {
+      "epoch": 0.49495382271731886,
+      "grad_norm": 25.834066632310197,
+      "learning_rate": 2e-05,
+      "loss": 1.0071,
+      "step": 20794
+    },
+    {
+      "epoch": 0.4949776254403504,
+      "grad_norm": 3.8957585193208226,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 20795
+    },
+    {
+      "epoch": 0.4950014281633819,
+      "grad_norm": 6.798764435021792,
+      "learning_rate": 2e-05,
+      "loss": 0.8865,
+      "step": 20796
+    },
+    {
+      "epoch": 0.4950252308864134,
+      "grad_norm": 14.335783284621314,
+      "learning_rate": 2e-05,
+      "loss": 0.4103,
+      "step": 20797
+    },
+    {
+      "epoch": 0.4950490336094449,
+      "grad_norm": 6.330828652703612,
+      "learning_rate": 2e-05,
+      "loss": 0.6141,
+      "step": 20798
+    },
+    {
+      "epoch": 0.49507283633247645,
+      "grad_norm": 3.0221454585885494,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 20799
+    },
+    {
+      "epoch": 0.49509663905550794,
+      "grad_norm": 5.310476624779123,
+      "learning_rate": 2e-05,
+      "loss": 1.1891,
+      "step": 20800
+    },
+    {
+      "epoch": 0.4951204417785395,
+      "grad_norm": 6.15936015095532,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 20801
+    },
+    {
+      "epoch": 0.49514424450157096,
+      "grad_norm": 1.6214492918827408,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 20802
+    },
+    {
+      "epoch": 0.4951680472246025,
+      "grad_norm": 3.3025489458786237,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 20803
+    },
+    {
+      "epoch": 0.495191849947634,
+      "grad_norm": 2.9323059117818207,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 20804
+    },
+    {
+      "epoch": 0.49521565267066553,
+      "grad_norm": 1.9879394818704885,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 20805
+    },
+    {
+      "epoch": 0.495239455393697,
+      "grad_norm": 0.8630413158385924,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 20806
+    },
+    {
+      "epoch": 0.49526325811672856,
+      "grad_norm": 19.47954300919552,
+      "learning_rate": 2e-05,
+      "loss": 0.4219,
+      "step": 20807
+    },
+    {
+      "epoch": 0.4952870608397601,
+      "grad_norm": 3.6649647904472453,
+      "learning_rate": 2e-05,
+      "loss": 0.0733,
+      "step": 20808
+    },
+    {
+      "epoch": 0.4953108635627916,
+      "grad_norm": 1.884246903312779,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 20809
+    },
+    {
+      "epoch": 0.4953346662858231,
+      "grad_norm": 5.499876271034974,
+      "learning_rate": 2e-05,
+      "loss": 1.1814,
+      "step": 20810
+    },
+    {
+      "epoch": 0.4953584690088546,
+      "grad_norm": 0.35134593484800203,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 20811
+    },
+    {
+      "epoch": 0.49538227173188615,
+      "grad_norm": 0.806947960798785,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 20812
+    },
+    {
+      "epoch": 0.49540607445491763,
+      "grad_norm": 12.399219535090094,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 20813
+    },
+    {
+      "epoch": 0.4954298771779492,
+      "grad_norm": 0.1910028522891111,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 20814
+    },
+    {
+      "epoch": 0.49545367990098066,
+      "grad_norm": 0.08388881787563632,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 20815
+    },
+    {
+      "epoch": 0.4954774826240122,
+      "grad_norm": 6.092356157725373,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 20816
+    },
+    {
+      "epoch": 0.4955012853470437,
+      "grad_norm": 2.3198861170244727,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 20817
+    },
+    {
+      "epoch": 0.4955250880700752,
+      "grad_norm": 6.317027823825802,
+      "learning_rate": 2e-05,
+      "loss": 0.2302,
+      "step": 20818
+    },
+    {
+      "epoch": 0.4955488907931067,
+      "grad_norm": 4.305533951391801,
+      "learning_rate": 2e-05,
+      "loss": 0.9009,
+      "step": 20819
+    },
+    {
+      "epoch": 0.49557269351613825,
+      "grad_norm": 12.511459573039705,
+      "learning_rate": 2e-05,
+      "loss": 0.2037,
+      "step": 20820
+    },
+    {
+      "epoch": 0.49559649623916974,
+      "grad_norm": 0.7239088700688995,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 20821
+    },
+    {
+      "epoch": 0.4956202989622013,
+      "grad_norm": 6.688681835173619,
+      "learning_rate": 2e-05,
+      "loss": 0.693,
+      "step": 20822
+    },
+    {
+      "epoch": 0.49564410168523276,
+      "grad_norm": 4.904237627828628,
+      "learning_rate": 2e-05,
+      "loss": 1.3094,
+      "step": 20823
+    },
+    {
+      "epoch": 0.4956679044082643,
+      "grad_norm": 7.453041648532033,
+      "learning_rate": 2e-05,
+      "loss": 0.7396,
+      "step": 20824
+    },
+    {
+      "epoch": 0.49569170713129584,
+      "grad_norm": 4.738649985917784,
+      "learning_rate": 2e-05,
+      "loss": 0.7163,
+      "step": 20825
+    },
+    {
+      "epoch": 0.49571550985432733,
+      "grad_norm": 3.6932299245078006,
+      "learning_rate": 2e-05,
+      "loss": 0.6335,
+      "step": 20826
+    },
+    {
+      "epoch": 0.49573931257735887,
+      "grad_norm": 7.497467256569579,
+      "learning_rate": 2e-05,
+      "loss": 0.8867,
+      "step": 20827
+    },
+    {
+      "epoch": 0.49576311530039036,
+      "grad_norm": 4.054587884571741,
+      "learning_rate": 2e-05,
+      "loss": 0.5518,
+      "step": 20828
+    },
+    {
+      "epoch": 0.4957869180234219,
+      "grad_norm": 5.790955185417464,
+      "learning_rate": 2e-05,
+      "loss": 0.7717,
+      "step": 20829
+    },
+    {
+      "epoch": 0.4958107207464534,
+      "grad_norm": 8.007981257517262,
+      "learning_rate": 2e-05,
+      "loss": 0.4229,
+      "step": 20830
+    },
+    {
+      "epoch": 0.4958345234694849,
+      "grad_norm": 4.380174072656437,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 20831
+    },
+    {
+      "epoch": 0.4958583261925164,
+      "grad_norm": 5.771566173988488,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 20832
+    },
+    {
+      "epoch": 0.49588212891554795,
+      "grad_norm": 4.349484705742967,
+      "learning_rate": 2e-05,
+      "loss": 0.5732,
+      "step": 20833
+    },
+    {
+      "epoch": 0.49590593163857943,
+      "grad_norm": 4.5230534760778385,
+      "learning_rate": 2e-05,
+      "loss": 0.5582,
+      "step": 20834
+    },
+    {
+      "epoch": 0.495929734361611,
+      "grad_norm": 3.0682183590035734,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 20835
+    },
+    {
+      "epoch": 0.49595353708464246,
+      "grad_norm": 10.076444551409217,
+      "learning_rate": 2e-05,
+      "loss": 0.5881,
+      "step": 20836
+    },
+    {
+      "epoch": 0.495977339807674,
+      "grad_norm": 1.3411927574988136,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 20837
+    },
+    {
+      "epoch": 0.4960011425307055,
+      "grad_norm": 3.573033337851234,
+      "learning_rate": 2e-05,
+      "loss": 0.1407,
+      "step": 20838
+    },
+    {
+      "epoch": 0.496024945253737,
+      "grad_norm": 4.5802187875801454,
+      "learning_rate": 2e-05,
+      "loss": 0.7665,
+      "step": 20839
+    },
+    {
+      "epoch": 0.49604874797676857,
+      "grad_norm": 2.564100684945986,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 20840
+    },
+    {
+      "epoch": 0.49607255069980005,
+      "grad_norm": 9.55209604667522,
+      "learning_rate": 2e-05,
+      "loss": 0.7969,
+      "step": 20841
+    },
+    {
+      "epoch": 0.4960963534228316,
+      "grad_norm": 1.2080108100867033,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 20842
+    },
+    {
+      "epoch": 0.4961201561458631,
+      "grad_norm": 3.607883710270262,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 20843
+    },
+    {
+      "epoch": 0.4961439588688946,
+      "grad_norm": 5.065309316948443,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 20844
+    },
+    {
+      "epoch": 0.4961677615919261,
+      "grad_norm": 2.8515957668559775,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 20845
+    },
+    {
+      "epoch": 0.49619156431495764,
+      "grad_norm": 5.031339290238476,
+      "learning_rate": 2e-05,
+      "loss": 0.2485,
+      "step": 20846
+    },
+    {
+      "epoch": 0.49621536703798913,
+      "grad_norm": 1.4680084430587514,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 20847
+    },
+    {
+      "epoch": 0.49623916976102067,
+      "grad_norm": 7.375770060226776,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 20848
+    },
+    {
+      "epoch": 0.49626297248405216,
+      "grad_norm": 2.631667380921108,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 20849
+    },
+    {
+      "epoch": 0.4962867752070837,
+      "grad_norm": 0.19109173397342533,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 20850
+    },
+    {
+      "epoch": 0.4963105779301152,
+      "grad_norm": 4.213731711730749,
+      "learning_rate": 2e-05,
+      "loss": 0.217,
+      "step": 20851
+    },
+    {
+      "epoch": 0.4963343806531467,
+      "grad_norm": 4.282630874124342,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 20852
+    },
+    {
+      "epoch": 0.4963581833761782,
+      "grad_norm": 17.78240644742879,
+      "learning_rate": 2e-05,
+      "loss": 1.5255,
+      "step": 20853
+    },
+    {
+      "epoch": 0.49638198609920975,
+      "grad_norm": 0.6998098748339582,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 20854
+    },
+    {
+      "epoch": 0.4964057888222413,
+      "grad_norm": 4.711010491303827,
+      "learning_rate": 2e-05,
+      "loss": 0.5848,
+      "step": 20855
+    },
+    {
+      "epoch": 0.4964295915452728,
+      "grad_norm": 10.76370846361777,
+      "learning_rate": 2e-05,
+      "loss": 1.1595,
+      "step": 20856
+    },
+    {
+      "epoch": 0.4964533942683043,
+      "grad_norm": 13.592886615622596,
+      "learning_rate": 2e-05,
+      "loss": 0.8101,
+      "step": 20857
+    },
+    {
+      "epoch": 0.4964771969913358,
+      "grad_norm": 9.978127551696174,
+      "learning_rate": 2e-05,
+      "loss": 0.4837,
+      "step": 20858
+    },
+    {
+      "epoch": 0.49650099971436734,
+      "grad_norm": 3.8485150942783006,
+      "learning_rate": 2e-05,
+      "loss": 0.4663,
+      "step": 20859
+    },
+    {
+      "epoch": 0.4965248024373988,
+      "grad_norm": 3.7096327006875116,
+      "learning_rate": 2e-05,
+      "loss": 0.3543,
+      "step": 20860
+    },
+    {
+      "epoch": 0.49654860516043037,
+      "grad_norm": 2.156719749999469,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 20861
+    },
+    {
+      "epoch": 0.49657240788346185,
+      "grad_norm": 4.630468922966174,
+      "learning_rate": 2e-05,
+      "loss": 0.378,
+      "step": 20862
+    },
+    {
+      "epoch": 0.4965962106064934,
+      "grad_norm": 3.303902702860861,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 20863
+    },
+    {
+      "epoch": 0.4966200133295249,
+      "grad_norm": 2.8222697380764408,
+      "learning_rate": 2e-05,
+      "loss": 0.3957,
+      "step": 20864
+    },
+    {
+      "epoch": 0.4966438160525564,
+      "grad_norm": 3.3435010062487716,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 20865
+    },
+    {
+      "epoch": 0.4966676187755879,
+      "grad_norm": 3.2607779031814776,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 20866
+    },
+    {
+      "epoch": 0.49669142149861945,
+      "grad_norm": 6.278928955935623,
+      "learning_rate": 2e-05,
+      "loss": 1.5234,
+      "step": 20867
+    },
+    {
+      "epoch": 0.49671522422165093,
+      "grad_norm": 6.042212926884526,
+      "learning_rate": 2e-05,
+      "loss": 0.5256,
+      "step": 20868
+    },
+    {
+      "epoch": 0.49673902694468247,
+      "grad_norm": 4.261553341894988,
+      "learning_rate": 2e-05,
+      "loss": 0.6254,
+      "step": 20869
+    },
+    {
+      "epoch": 0.496762829667714,
+      "grad_norm": 4.645250524073398,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 20870
+    },
+    {
+      "epoch": 0.4967866323907455,
+      "grad_norm": 6.2571468739968905,
+      "learning_rate": 2e-05,
+      "loss": 0.7934,
+      "step": 20871
+    },
+    {
+      "epoch": 0.49681043511377704,
+      "grad_norm": 8.624644915286853,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 20872
+    },
+    {
+      "epoch": 0.4968342378368085,
+      "grad_norm": 2.9830839550146386,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 20873
+    },
+    {
+      "epoch": 0.49685804055984006,
+      "grad_norm": 2.1575711178422363,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 20874
+    },
+    {
+      "epoch": 0.49688184328287155,
+      "grad_norm": 1.8933495260640265,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 20875
+    },
+    {
+      "epoch": 0.4969056460059031,
+      "grad_norm": 5.599694824927434,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 20876
+    },
+    {
+      "epoch": 0.4969294487289346,
+      "grad_norm": 3.594021523688249,
+      "learning_rate": 2e-05,
+      "loss": 0.6011,
+      "step": 20877
+    },
+    {
+      "epoch": 0.4969532514519661,
+      "grad_norm": 5.785326169000753,
+      "learning_rate": 2e-05,
+      "loss": 0.3075,
+      "step": 20878
+    },
+    {
+      "epoch": 0.4969770541749976,
+      "grad_norm": 0.19586133914069925,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 20879
+    },
+    {
+      "epoch": 0.49700085689802914,
+      "grad_norm": 5.445729408202985,
+      "learning_rate": 2e-05,
+      "loss": 1.0661,
+      "step": 20880
+    },
+    {
+      "epoch": 0.4970246596210606,
+      "grad_norm": 2.1371623876623387,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 20881
+    },
+    {
+      "epoch": 0.49704846234409217,
+      "grad_norm": 4.442361132926266,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 20882
+    },
+    {
+      "epoch": 0.49707226506712365,
+      "grad_norm": 5.585744516875477,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 20883
+    },
+    {
+      "epoch": 0.4970960677901552,
+      "grad_norm": 3.1743566027990173,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 20884
+    },
+    {
+      "epoch": 0.49711987051318673,
+      "grad_norm": 0.18556925664838148,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 20885
+    },
+    {
+      "epoch": 0.4971436732362182,
+      "grad_norm": 2.15946395000327,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 20886
+    },
+    {
+      "epoch": 0.49716747595924976,
+      "grad_norm": 0.824300888544324,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 20887
+    },
+    {
+      "epoch": 0.49719127868228125,
+      "grad_norm": 12.047502987099627,
+      "learning_rate": 2e-05,
+      "loss": 0.5369,
+      "step": 20888
+    },
+    {
+      "epoch": 0.4972150814053128,
+      "grad_norm": 2.9188030077324965,
+      "learning_rate": 2e-05,
+      "loss": 0.336,
+      "step": 20889
+    },
+    {
+      "epoch": 0.49723888412834427,
+      "grad_norm": 0.6100629361972821,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 20890
+    },
+    {
+      "epoch": 0.4972626868513758,
+      "grad_norm": 8.77016891893822,
+      "learning_rate": 2e-05,
+      "loss": 0.7264,
+      "step": 20891
+    },
+    {
+      "epoch": 0.4972864895744073,
+      "grad_norm": 1.1832400598873642,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 20892
+    },
+    {
+      "epoch": 0.49731029229743884,
+      "grad_norm": 10.731705864564812,
+      "learning_rate": 2e-05,
+      "loss": 0.79,
+      "step": 20893
+    },
+    {
+      "epoch": 0.4973340950204703,
+      "grad_norm": 0.7520477025895904,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 20894
+    },
+    {
+      "epoch": 0.49735789774350186,
+      "grad_norm": 2.4836577837358136,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 20895
+    },
+    {
+      "epoch": 0.49738170046653335,
+      "grad_norm": 2.1734543544843055,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 20896
+    },
+    {
+      "epoch": 0.4974055031895649,
+      "grad_norm": 0.32591871624536106,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 20897
+    },
+    {
+      "epoch": 0.4974293059125964,
+      "grad_norm": 2.736379130410066,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 20898
+    },
+    {
+      "epoch": 0.4974531086356279,
+      "grad_norm": 2.56876660381654,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 20899
+    },
+    {
+      "epoch": 0.49747691135865946,
+      "grad_norm": 5.414248322675842,
+      "learning_rate": 2e-05,
+      "loss": 0.809,
+      "step": 20900
+    },
+    {
+      "epoch": 0.49750071408169094,
+      "grad_norm": 2.0739411860394084,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 20901
+    },
+    {
+      "epoch": 0.4975245168047225,
+      "grad_norm": 1.6105256475545842,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 20902
+    },
+    {
+      "epoch": 0.49754831952775397,
+      "grad_norm": 13.634654449489329,
+      "learning_rate": 2e-05,
+      "loss": 1.3446,
+      "step": 20903
+    },
+    {
+      "epoch": 0.4975721222507855,
+      "grad_norm": 1.6919841490730838,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 20904
+    },
+    {
+      "epoch": 0.497595924973817,
+      "grad_norm": 3.5552809445702827,
+      "learning_rate": 2e-05,
+      "loss": 0.6194,
+      "step": 20905
+    },
+    {
+      "epoch": 0.49761972769684854,
+      "grad_norm": 1.044108078229505,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 20906
+    },
+    {
+      "epoch": 0.49764353041988,
+      "grad_norm": 6.94544684942619,
+      "learning_rate": 2e-05,
+      "loss": 0.5375,
+      "step": 20907
+    },
+    {
+      "epoch": 0.49766733314291156,
+      "grad_norm": 4.76382045055172,
+      "learning_rate": 2e-05,
+      "loss": 0.5048,
+      "step": 20908
+    },
+    {
+      "epoch": 0.49769113586594305,
+      "grad_norm": 4.340063095553605,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 20909
+    },
+    {
+      "epoch": 0.4977149385889746,
+      "grad_norm": 6.979743842093878,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 20910
+    },
+    {
+      "epoch": 0.49773874131200607,
+      "grad_norm": 4.525354964965514,
+      "learning_rate": 2e-05,
+      "loss": 0.1249,
+      "step": 20911
+    },
+    {
+      "epoch": 0.4977625440350376,
+      "grad_norm": 9.789759726040273,
+      "learning_rate": 2e-05,
+      "loss": 0.4777,
+      "step": 20912
+    },
+    {
+      "epoch": 0.4977863467580691,
+      "grad_norm": 5.639091899261441,
+      "learning_rate": 2e-05,
+      "loss": 0.719,
+      "step": 20913
+    },
+    {
+      "epoch": 0.49781014948110064,
+      "grad_norm": 4.627662162418414,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 20914
+    },
+    {
+      "epoch": 0.4978339522041322,
+      "grad_norm": 6.2288543658394655,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 20915
+    },
+    {
+      "epoch": 0.49785775492716366,
+      "grad_norm": 3.987863964463856,
+      "learning_rate": 2e-05,
+      "loss": 0.6228,
+      "step": 20916
+    },
+    {
+      "epoch": 0.4978815576501952,
+      "grad_norm": 3.992174840596593,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 20917
+    },
+    {
+      "epoch": 0.4979053603732267,
+      "grad_norm": 18.0574803800391,
+      "learning_rate": 2e-05,
+      "loss": 1.4499,
+      "step": 20918
+    },
+    {
+      "epoch": 0.49792916309625823,
+      "grad_norm": 6.89722675992103,
+      "learning_rate": 2e-05,
+      "loss": 1.1915,
+      "step": 20919
+    },
+    {
+      "epoch": 0.4979529658192897,
+      "grad_norm": 4.648581913342508,
+      "learning_rate": 2e-05,
+      "loss": 0.7347,
+      "step": 20920
+    },
+    {
+      "epoch": 0.49797676854232126,
+      "grad_norm": 1.011268903002127,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 20921
+    },
+    {
+      "epoch": 0.49800057126535274,
+      "grad_norm": 4.076156324761163,
+      "learning_rate": 2e-05,
+      "loss": 0.6006,
+      "step": 20922
+    },
+    {
+      "epoch": 0.4980243739883843,
+      "grad_norm": 5.115334475141837,
+      "learning_rate": 2e-05,
+      "loss": 0.3957,
+      "step": 20923
+    },
+    {
+      "epoch": 0.49804817671141577,
+      "grad_norm": 1.7777610581640866,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 20924
+    },
+    {
+      "epoch": 0.4980719794344473,
+      "grad_norm": 0.21031151537127482,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 20925
+    },
+    {
+      "epoch": 0.4980957821574788,
+      "grad_norm": 11.48912330036589,
+      "learning_rate": 2e-05,
+      "loss": 0.7389,
+      "step": 20926
+    },
+    {
+      "epoch": 0.49811958488051034,
+      "grad_norm": 5.497032315969468,
+      "learning_rate": 2e-05,
+      "loss": 0.5993,
+      "step": 20927
+    },
+    {
+      "epoch": 0.4981433876035418,
+      "grad_norm": 5.520956916730911,
+      "learning_rate": 2e-05,
+      "loss": 0.8863,
+      "step": 20928
+    },
+    {
+      "epoch": 0.49816719032657336,
+      "grad_norm": 4.077605381767219,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 20929
+    },
+    {
+      "epoch": 0.4981909930496049,
+      "grad_norm": 0.42191448951534555,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 20930
+    },
+    {
+      "epoch": 0.4982147957726364,
+      "grad_norm": 2.8788674108076124,
+      "learning_rate": 2e-05,
+      "loss": 0.5343,
+      "step": 20931
+    },
+    {
+      "epoch": 0.49823859849566793,
+      "grad_norm": 3.7152418576033766,
+      "learning_rate": 2e-05,
+      "loss": 0.7615,
+      "step": 20932
+    },
+    {
+      "epoch": 0.4982624012186994,
+      "grad_norm": 5.447341423194241,
+      "learning_rate": 2e-05,
+      "loss": 0.7433,
+      "step": 20933
+    },
+    {
+      "epoch": 0.49828620394173095,
+      "grad_norm": 0.9207244046963435,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 20934
+    },
+    {
+      "epoch": 0.49831000666476244,
+      "grad_norm": 0.735433910821134,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 20935
+    },
+    {
+      "epoch": 0.498333809387794,
+      "grad_norm": 8.729215480877135,
+      "learning_rate": 2e-05,
+      "loss": 0.3915,
+      "step": 20936
+    },
+    {
+      "epoch": 0.49835761211082547,
+      "grad_norm": 1.5199596319194533,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 20937
+    },
+    {
+      "epoch": 0.498381414833857,
+      "grad_norm": 0.22266840491684642,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 20938
+    },
+    {
+      "epoch": 0.4984052175568885,
+      "grad_norm": 3.8531888367644367,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 20939
+    },
+    {
+      "epoch": 0.49842902027992003,
+      "grad_norm": 11.911479194397511,
+      "learning_rate": 2e-05,
+      "loss": 0.8504,
+      "step": 20940
+    },
+    {
+      "epoch": 0.4984528230029515,
+      "grad_norm": 4.095856000647027,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 20941
+    },
+    {
+      "epoch": 0.49847662572598306,
+      "grad_norm": 0.12728037415679283,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 20942
+    },
+    {
+      "epoch": 0.49850042844901454,
+      "grad_norm": 4.370953863119481,
+      "learning_rate": 2e-05,
+      "loss": 0.6683,
+      "step": 20943
+    },
+    {
+      "epoch": 0.4985242311720461,
+      "grad_norm": 5.607013039523733,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 20944
+    },
+    {
+      "epoch": 0.49854803389507757,
+      "grad_norm": 11.075864382777706,
+      "learning_rate": 2e-05,
+      "loss": 0.6848,
+      "step": 20945
+    },
+    {
+      "epoch": 0.4985718366181091,
+      "grad_norm": 19.562621422258033,
+      "learning_rate": 2e-05,
+      "loss": 0.9438,
+      "step": 20946
+    },
+    {
+      "epoch": 0.49859563934114065,
+      "grad_norm": 9.652226829364654,
+      "learning_rate": 2e-05,
+      "loss": 0.2263,
+      "step": 20947
+    },
+    {
+      "epoch": 0.49861944206417214,
+      "grad_norm": 1.9617142745673555,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 20948
+    },
+    {
+      "epoch": 0.4986432447872037,
+      "grad_norm": 2.0807250460008424,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 20949
+    },
+    {
+      "epoch": 0.49866704751023516,
+      "grad_norm": 0.767047820220528,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 20950
+    },
+    {
+      "epoch": 0.4986908502332667,
+      "grad_norm": 7.357757729429034,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 20951
+    },
+    {
+      "epoch": 0.4987146529562982,
+      "grad_norm": 11.388114725225797,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 20952
+    },
+    {
+      "epoch": 0.49873845567932973,
+      "grad_norm": 3.2784554899961154,
+      "learning_rate": 2e-05,
+      "loss": 0.6205,
+      "step": 20953
+    },
+    {
+      "epoch": 0.4987622584023612,
+      "grad_norm": 0.33828248356318275,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 20954
+    },
+    {
+      "epoch": 0.49878606112539275,
+      "grad_norm": 0.7492651144408294,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 20955
+    },
+    {
+      "epoch": 0.49880986384842424,
+      "grad_norm": 8.902296274012084,
+      "learning_rate": 2e-05,
+      "loss": 0.3328,
+      "step": 20956
+    },
+    {
+      "epoch": 0.4988336665714558,
+      "grad_norm": 1.7776029449789144,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 20957
+    },
+    {
+      "epoch": 0.49885746929448727,
+      "grad_norm": 69.19718171939937,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 20958
+    },
+    {
+      "epoch": 0.4988812720175188,
+      "grad_norm": 5.615948640897586,
+      "learning_rate": 2e-05,
+      "loss": 0.5325,
+      "step": 20959
+    },
+    {
+      "epoch": 0.4989050747405503,
+      "grad_norm": 0.8203732177104285,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 20960
+    },
+    {
+      "epoch": 0.49892887746358183,
+      "grad_norm": 11.990722276649233,
+      "learning_rate": 2e-05,
+      "loss": 0.4499,
+      "step": 20961
+    },
+    {
+      "epoch": 0.4989526801866134,
+      "grad_norm": 2.665949166465104,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 20962
+    },
+    {
+      "epoch": 0.49897648290964486,
+      "grad_norm": 7.052721517050669,
+      "learning_rate": 2e-05,
+      "loss": 0.2836,
+      "step": 20963
+    },
+    {
+      "epoch": 0.4990002856326764,
+      "grad_norm": 0.20014596915061433,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 20964
+    },
+    {
+      "epoch": 0.4990240883557079,
+      "grad_norm": 0.20835539170806094,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 20965
+    },
+    {
+      "epoch": 0.4990478910787394,
+      "grad_norm": 0.963367308835847,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 20966
+    },
+    {
+      "epoch": 0.4990716938017709,
+      "grad_norm": 3.5046174264952774,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 20967
+    },
+    {
+      "epoch": 0.49909549652480245,
+      "grad_norm": 0.9988193549958843,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 20968
+    },
+    {
+      "epoch": 0.49911929924783394,
+      "grad_norm": 3.240594940261999,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 20969
+    },
+    {
+      "epoch": 0.4991431019708655,
+      "grad_norm": 4.376273638773076,
+      "learning_rate": 2e-05,
+      "loss": 0.5427,
+      "step": 20970
+    },
+    {
+      "epoch": 0.49916690469389696,
+      "grad_norm": 14.611242159863046,
+      "learning_rate": 2e-05,
+      "loss": 0.6237,
+      "step": 20971
+    },
+    {
+      "epoch": 0.4991907074169285,
+      "grad_norm": 3.328535039846946,
+      "learning_rate": 2e-05,
+      "loss": 0.1911,
+      "step": 20972
+    },
+    {
+      "epoch": 0.49921451013996,
+      "grad_norm": 0.8774038431472501,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 20973
+    },
+    {
+      "epoch": 0.49923831286299153,
+      "grad_norm": 7.4621091011263845,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 20974
+    },
+    {
+      "epoch": 0.499262115586023,
+      "grad_norm": 8.413251478049146,
+      "learning_rate": 2e-05,
+      "loss": 1.5177,
+      "step": 20975
+    },
+    {
+      "epoch": 0.49928591830905455,
+      "grad_norm": 2.4509394625605756,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 20976
+    },
+    {
+      "epoch": 0.4993097210320861,
+      "grad_norm": 7.04771267902825,
+      "learning_rate": 2e-05,
+      "loss": 0.9061,
+      "step": 20977
+    },
+    {
+      "epoch": 0.4993335237551176,
+      "grad_norm": 10.00580778216239,
+      "learning_rate": 2e-05,
+      "loss": 1.0729,
+      "step": 20978
+    },
+    {
+      "epoch": 0.4993573264781491,
+      "grad_norm": 7.000037759078767,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 20979
+    },
+    {
+      "epoch": 0.4993811292011806,
+      "grad_norm": 0.6234649624319706,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 20980
+    },
+    {
+      "epoch": 0.49940493192421215,
+      "grad_norm": 3.502567718138646,
+      "learning_rate": 2e-05,
+      "loss": 0.8564,
+      "step": 20981
+    },
+    {
+      "epoch": 0.49942873464724363,
+      "grad_norm": 5.823780304117821,
+      "learning_rate": 2e-05,
+      "loss": 1.2222,
+      "step": 20982
+    },
+    {
+      "epoch": 0.4994525373702752,
+      "grad_norm": 4.981460213821847,
+      "learning_rate": 2e-05,
+      "loss": 0.8362,
+      "step": 20983
+    },
+    {
+      "epoch": 0.49947634009330666,
+      "grad_norm": 5.1710264874669285,
+      "learning_rate": 2e-05,
+      "loss": 1.0458,
+      "step": 20984
+    },
+    {
+      "epoch": 0.4995001428163382,
+      "grad_norm": 3.951748278926731,
+      "learning_rate": 2e-05,
+      "loss": 0.6373,
+      "step": 20985
+    },
+    {
+      "epoch": 0.4995239455393697,
+      "grad_norm": 4.285697898658497,
+      "learning_rate": 2e-05,
+      "loss": 0.2453,
+      "step": 20986
+    },
+    {
+      "epoch": 0.4995477482624012,
+      "grad_norm": 5.489262866884939,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 20987
+    },
+    {
+      "epoch": 0.4995715509854327,
+      "grad_norm": 4.735259263831132,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 20988
+    },
+    {
+      "epoch": 0.49959535370846425,
+      "grad_norm": 2.6110052227672296,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 20989
+    },
+    {
+      "epoch": 0.49961915643149574,
+      "grad_norm": 7.863697195719308,
+      "learning_rate": 2e-05,
+      "loss": 1.2684,
+      "step": 20990
+    },
+    {
+      "epoch": 0.4996429591545273,
+      "grad_norm": 6.672470481793723,
+      "learning_rate": 2e-05,
+      "loss": 1.0392,
+      "step": 20991
+    },
+    {
+      "epoch": 0.4996667618775588,
+      "grad_norm": 4.704101126045975,
+      "learning_rate": 2e-05,
+      "loss": 0.52,
+      "step": 20992
+    },
+    {
+      "epoch": 0.4996905646005903,
+      "grad_norm": 2.101965157905767,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 20993
+    },
+    {
+      "epoch": 0.49971436732362184,
+      "grad_norm": 2.3289031183238103,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 20994
+    },
+    {
+      "epoch": 0.49973817004665333,
+      "grad_norm": 2.665275326830597,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 20995
+    },
+    {
+      "epoch": 0.49976197276968487,
+      "grad_norm": 5.049641279251228,
+      "learning_rate": 2e-05,
+      "loss": 0.8826,
+      "step": 20996
+    },
+    {
+      "epoch": 0.49978577549271636,
+      "grad_norm": 6.563781108043384,
+      "learning_rate": 2e-05,
+      "loss": 0.1494,
+      "step": 20997
+    },
+    {
+      "epoch": 0.4998095782157479,
+      "grad_norm": 4.998568978570067,
+      "learning_rate": 2e-05,
+      "loss": 0.4539,
+      "step": 20998
+    },
+    {
+      "epoch": 0.4998333809387794,
+      "grad_norm": 4.978760331033603,
+      "learning_rate": 2e-05,
+      "loss": 0.153,
+      "step": 20999
+    },
+    {
+      "epoch": 0.4998571836618109,
+      "grad_norm": 4.328102467286432,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 21000
+    },
+    {
+      "epoch": 0.4998809863848424,
+      "grad_norm": 6.357926487349912,
+      "learning_rate": 2e-05,
+      "loss": 0.235,
+      "step": 21001
+    },
+    {
+      "epoch": 0.49990478910787395,
+      "grad_norm": 2.2281994892678725,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 21002
+    },
+    {
+      "epoch": 0.49992859183090543,
+      "grad_norm": 4.219154583792186,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 21003
+    },
+    {
+      "epoch": 0.499952394553937,
+      "grad_norm": 2.474946377393474,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 21004
+    },
+    {
+      "epoch": 0.49997619727696846,
+      "grad_norm": 7.595050323426101,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 21005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8007312073920754,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 21006
+    },
+    {
+      "epoch": 0.5000238027230315,
+      "grad_norm": 0.2806913370741585,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21007
+    },
+    {
+      "epoch": 0.5000476054460631,
+      "grad_norm": 12.82006714673021,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 21008
+    },
+    {
+      "epoch": 0.5000714081690946,
+      "grad_norm": 2.254018819742113,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 21009
+    },
+    {
+      "epoch": 0.500095210892126,
+      "grad_norm": 7.415342302044189,
+      "learning_rate": 2e-05,
+      "loss": 0.9856,
+      "step": 21010
+    },
+    {
+      "epoch": 0.5001190136151575,
+      "grad_norm": 2.511258996752917,
+      "learning_rate": 2e-05,
+      "loss": 0.1416,
+      "step": 21011
+    },
+    {
+      "epoch": 0.5001428163381891,
+      "grad_norm": 3.4492422998023047,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 21012
+    },
+    {
+      "epoch": 0.5001666190612206,
+      "grad_norm": 3.7511446225562657,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 21013
+    },
+    {
+      "epoch": 0.5001904217842521,
+      "grad_norm": 0.6611128441360331,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 21014
+    },
+    {
+      "epoch": 0.5002142245072836,
+      "grad_norm": 5.150630520974564,
+      "learning_rate": 2e-05,
+      "loss": 0.7188,
+      "step": 21015
+    },
+    {
+      "epoch": 0.5002380272303152,
+      "grad_norm": 4.7164280437376425,
+      "learning_rate": 2e-05,
+      "loss": 0.4419,
+      "step": 21016
+    },
+    {
+      "epoch": 0.5002618299533467,
+      "grad_norm": 4.287875026927348,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 21017
+    },
+    {
+      "epoch": 0.5002856326763782,
+      "grad_norm": 5.817066767400245,
+      "learning_rate": 2e-05,
+      "loss": 1.3279,
+      "step": 21018
+    },
+    {
+      "epoch": 0.5003094353994096,
+      "grad_norm": 0.02810761042022525,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 21019
+    },
+    {
+      "epoch": 0.5003332381224412,
+      "grad_norm": 4.808442867697928,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 21020
+    },
+    {
+      "epoch": 0.5003570408454727,
+      "grad_norm": 0.09053066646415973,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21021
+    },
+    {
+      "epoch": 0.5003808435685042,
+      "grad_norm": 3.663974196663123,
+      "learning_rate": 2e-05,
+      "loss": 0.2634,
+      "step": 21022
+    },
+    {
+      "epoch": 0.5004046462915358,
+      "grad_norm": 6.527985730286874,
+      "learning_rate": 2e-05,
+      "loss": 0.9888,
+      "step": 21023
+    },
+    {
+      "epoch": 0.5004284490145673,
+      "grad_norm": 0.3005447145670848,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 21024
+    },
+    {
+      "epoch": 0.5004522517375988,
+      "grad_norm": 8.506194395334022,
+      "learning_rate": 2e-05,
+      "loss": 0.1217,
+      "step": 21025
+    },
+    {
+      "epoch": 0.5004760544606303,
+      "grad_norm": 0.18943902251283615,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 21026
+    },
+    {
+      "epoch": 0.5004998571836619,
+      "grad_norm": 0.6228634881748573,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 21027
+    },
+    {
+      "epoch": 0.5005236599066933,
+      "grad_norm": 7.5665357739147945,
+      "learning_rate": 2e-05,
+      "loss": 0.7192,
+      "step": 21028
+    },
+    {
+      "epoch": 0.5005474626297248,
+      "grad_norm": 3.074835910998617,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 21029
+    },
+    {
+      "epoch": 0.5005712653527563,
+      "grad_norm": 22.421051940305908,
+      "learning_rate": 2e-05,
+      "loss": 1.7951,
+      "step": 21030
+    },
+    {
+      "epoch": 0.5005950680757879,
+      "grad_norm": 1.8713911589434498,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 21031
+    },
+    {
+      "epoch": 0.5006188707988194,
+      "grad_norm": 0.501164167859051,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 21032
+    },
+    {
+      "epoch": 0.5006426735218509,
+      "grad_norm": 1.693075000998029,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 21033
+    },
+    {
+      "epoch": 0.5006664762448824,
+      "grad_norm": 0.9674936276127726,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 21034
+    },
+    {
+      "epoch": 0.500690278967914,
+      "grad_norm": 0.49580315466638125,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 21035
+    },
+    {
+      "epoch": 0.5007140816909454,
+      "grad_norm": 1.2377977347551974,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 21036
+    },
+    {
+      "epoch": 0.5007378844139769,
+      "grad_norm": 1.0410355208854232,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 21037
+    },
+    {
+      "epoch": 0.5007616871370085,
+      "grad_norm": 30.114934541161393,
+      "learning_rate": 2e-05,
+      "loss": 0.946,
+      "step": 21038
+    },
+    {
+      "epoch": 0.50078548986004,
+      "grad_norm": 1.8766699462953278,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 21039
+    },
+    {
+      "epoch": 0.5008092925830715,
+      "grad_norm": 0.8965504297074435,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 21040
+    },
+    {
+      "epoch": 0.500833095306103,
+      "grad_norm": 6.480229522634908,
+      "learning_rate": 2e-05,
+      "loss": 1.6983,
+      "step": 21041
+    },
+    {
+      "epoch": 0.5008568980291346,
+      "grad_norm": 5.877382261178783,
+      "learning_rate": 2e-05,
+      "loss": 0.4183,
+      "step": 21042
+    },
+    {
+      "epoch": 0.5008807007521661,
+      "grad_norm": 1.4872000136229204,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 21043
+    },
+    {
+      "epoch": 0.5009045034751975,
+      "grad_norm": 2.8672991044666163,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 21044
+    },
+    {
+      "epoch": 0.500928306198229,
+      "grad_norm": 4.630057133326639,
+      "learning_rate": 2e-05,
+      "loss": 0.1911,
+      "step": 21045
+    },
+    {
+      "epoch": 0.5009521089212606,
+      "grad_norm": 4.8139424625648655,
+      "learning_rate": 2e-05,
+      "loss": 0.3012,
+      "step": 21046
+    },
+    {
+      "epoch": 0.5009759116442921,
+      "grad_norm": 7.39131820907126,
+      "learning_rate": 2e-05,
+      "loss": 0.5414,
+      "step": 21047
+    },
+    {
+      "epoch": 0.5009997143673236,
+      "grad_norm": 0.127824133215924,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 21048
+    },
+    {
+      "epoch": 0.5010235170903551,
+      "grad_norm": 0.3038684357896319,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 21049
+    },
+    {
+      "epoch": 0.5010473198133867,
+      "grad_norm": 3.6262674813514844,
+      "learning_rate": 2e-05,
+      "loss": 0.3818,
+      "step": 21050
+    },
+    {
+      "epoch": 0.5010711225364182,
+      "grad_norm": 4.0352272871709935,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 21051
+    },
+    {
+      "epoch": 0.5010949252594497,
+      "grad_norm": 4.517403124861056,
+      "learning_rate": 2e-05,
+      "loss": 0.3852,
+      "step": 21052
+    },
+    {
+      "epoch": 0.5011187279824812,
+      "grad_norm": 5.215538674086584,
+      "learning_rate": 2e-05,
+      "loss": 0.8231,
+      "step": 21053
+    },
+    {
+      "epoch": 0.5011425307055127,
+      "grad_norm": 13.680143977378966,
+      "learning_rate": 2e-05,
+      "loss": 0.8116,
+      "step": 21054
+    },
+    {
+      "epoch": 0.5011663334285442,
+      "grad_norm": 3.93202349091481,
+      "learning_rate": 2e-05,
+      "loss": 0.9258,
+      "step": 21055
+    },
+    {
+      "epoch": 0.5011901361515757,
+      "grad_norm": 1.806700760442087,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 21056
+    },
+    {
+      "epoch": 0.5012139388746073,
+      "grad_norm": 6.02919624284531,
+      "learning_rate": 2e-05,
+      "loss": 0.6637,
+      "step": 21057
+    },
+    {
+      "epoch": 0.5012377415976388,
+      "grad_norm": 3.0965125500999675,
+      "learning_rate": 2e-05,
+      "loss": 0.4434,
+      "step": 21058
+    },
+    {
+      "epoch": 0.5012615443206703,
+      "grad_norm": 5.148606048475708,
+      "learning_rate": 2e-05,
+      "loss": 0.6189,
+      "step": 21059
+    },
+    {
+      "epoch": 0.5012853470437018,
+      "grad_norm": 5.950434564604569,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 21060
+    },
+    {
+      "epoch": 0.5013091497667334,
+      "grad_norm": 5.5248010184751335,
+      "learning_rate": 2e-05,
+      "loss": 0.8925,
+      "step": 21061
+    },
+    {
+      "epoch": 0.5013329524897648,
+      "grad_norm": 4.8666550222334015,
+      "learning_rate": 2e-05,
+      "loss": 0.1755,
+      "step": 21062
+    },
+    {
+      "epoch": 0.5013567552127963,
+      "grad_norm": 7.094134133674103,
+      "learning_rate": 2e-05,
+      "loss": 0.8371,
+      "step": 21063
+    },
+    {
+      "epoch": 0.5013805579358278,
+      "grad_norm": 2.6895339414383668,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 21064
+    },
+    {
+      "epoch": 0.5014043606588594,
+      "grad_norm": 5.280226955764804,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 21065
+    },
+    {
+      "epoch": 0.5014281633818909,
+      "grad_norm": 1.4808157398681923,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 21066
+    },
+    {
+      "epoch": 0.5014519661049224,
+      "grad_norm": 0.567632445165832,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 21067
+    },
+    {
+      "epoch": 0.501475768827954,
+      "grad_norm": 4.202674471105267,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 21068
+    },
+    {
+      "epoch": 0.5014995715509855,
+      "grad_norm": 0.1714180807339947,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 21069
+    },
+    {
+      "epoch": 0.5015233742740169,
+      "grad_norm": 5.284925495787734,
+      "learning_rate": 2e-05,
+      "loss": 1.1025,
+      "step": 21070
+    },
+    {
+      "epoch": 0.5015471769970484,
+      "grad_norm": 3.6614906545540324,
+      "learning_rate": 2e-05,
+      "loss": 0.3944,
+      "step": 21071
+    },
+    {
+      "epoch": 0.50157097972008,
+      "grad_norm": 3.26033257202956,
+      "learning_rate": 2e-05,
+      "loss": 0.0892,
+      "step": 21072
+    },
+    {
+      "epoch": 0.5015947824431115,
+      "grad_norm": 8.481340573905438,
+      "learning_rate": 2e-05,
+      "loss": 0.6377,
+      "step": 21073
+    },
+    {
+      "epoch": 0.501618585166143,
+      "grad_norm": 0.5840610496465014,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 21074
+    },
+    {
+      "epoch": 0.5016423878891745,
+      "grad_norm": 6.798176521048199,
+      "learning_rate": 2e-05,
+      "loss": 1.1425,
+      "step": 21075
+    },
+    {
+      "epoch": 0.5016661906122061,
+      "grad_norm": 4.429903587756735,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 21076
+    },
+    {
+      "epoch": 0.5016899933352376,
+      "grad_norm": 4.140544740615496,
+      "learning_rate": 2e-05,
+      "loss": 0.4326,
+      "step": 21077
+    },
+    {
+      "epoch": 0.501713796058269,
+      "grad_norm": 3.5626308177704433,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 21078
+    },
+    {
+      "epoch": 0.5017375987813005,
+      "grad_norm": 2.8569760530902846,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 21079
+    },
+    {
+      "epoch": 0.5017614015043321,
+      "grad_norm": 1.690042313769108,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 21080
+    },
+    {
+      "epoch": 0.5017852042273636,
+      "grad_norm": 1.676739006573341,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 21081
+    },
+    {
+      "epoch": 0.5018090069503951,
+      "grad_norm": 22.896924218175766,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 21082
+    },
+    {
+      "epoch": 0.5018328096734267,
+      "grad_norm": 2.3428956883761822,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 21083
+    },
+    {
+      "epoch": 0.5018566123964582,
+      "grad_norm": 5.819781466894758,
+      "learning_rate": 2e-05,
+      "loss": 0.6549,
+      "step": 21084
+    },
+    {
+      "epoch": 0.5018804151194897,
+      "grad_norm": 10.638912260132017,
+      "learning_rate": 2e-05,
+      "loss": 1.2141,
+      "step": 21085
+    },
+    {
+      "epoch": 0.5019042178425211,
+      "grad_norm": 3.054001218097439,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 21086
+    },
+    {
+      "epoch": 0.5019280205655527,
+      "grad_norm": 23.354030676695015,
+      "learning_rate": 2e-05,
+      "loss": 1.1641,
+      "step": 21087
+    },
+    {
+      "epoch": 0.5019518232885842,
+      "grad_norm": 1.0003000561967188,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 21088
+    },
+    {
+      "epoch": 0.5019756260116157,
+      "grad_norm": 7.369551379031301,
+      "learning_rate": 2e-05,
+      "loss": 0.9057,
+      "step": 21089
+    },
+    {
+      "epoch": 0.5019994287346472,
+      "grad_norm": 3.870030284940371,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 21090
+    },
+    {
+      "epoch": 0.5020232314576788,
+      "grad_norm": 2.533131686155609,
+      "learning_rate": 2e-05,
+      "loss": 0.2569,
+      "step": 21091
+    },
+    {
+      "epoch": 0.5020470341807103,
+      "grad_norm": 3.5408841343994726,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 21092
+    },
+    {
+      "epoch": 0.5020708369037418,
+      "grad_norm": 5.2369420713196595,
+      "learning_rate": 2e-05,
+      "loss": 0.7063,
+      "step": 21093
+    },
+    {
+      "epoch": 0.5020946396267733,
+      "grad_norm": 0.34177214933701744,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21094
+    },
+    {
+      "epoch": 0.5021184423498048,
+      "grad_norm": 6.339518003339641,
+      "learning_rate": 2e-05,
+      "loss": 1.466,
+      "step": 21095
+    },
+    {
+      "epoch": 0.5021422450728363,
+      "grad_norm": 3.509607039730339,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 21096
+    },
+    {
+      "epoch": 0.5021660477958678,
+      "grad_norm": 6.417310901063608,
+      "learning_rate": 2e-05,
+      "loss": 1.0258,
+      "step": 21097
+    },
+    {
+      "epoch": 0.5021898505188994,
+      "grad_norm": 10.272085503117792,
+      "learning_rate": 2e-05,
+      "loss": 0.7621,
+      "step": 21098
+    },
+    {
+      "epoch": 0.5022136532419309,
+      "grad_norm": 3.110656867718516,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 21099
+    },
+    {
+      "epoch": 0.5022374559649624,
+      "grad_norm": 0.7634561125960561,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5022612586879939,
+      "grad_norm": 0.0714304189623468,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 21101
+    },
+    {
+      "epoch": 0.5022850614110255,
+      "grad_norm": 5.572236384744683,
+      "learning_rate": 2e-05,
+      "loss": 1.1845,
+      "step": 21102
+    },
+    {
+      "epoch": 0.502308864134057,
+      "grad_norm": 2.8361782451235333,
+      "learning_rate": 2e-05,
+      "loss": 0.3788,
+      "step": 21103
+    },
+    {
+      "epoch": 0.5023326668570884,
+      "grad_norm": 2.379714196294901,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 21104
+    },
+    {
+      "epoch": 0.5023564695801199,
+      "grad_norm": 6.558224984927397,
+      "learning_rate": 2e-05,
+      "loss": 0.7557,
+      "step": 21105
+    },
+    {
+      "epoch": 0.5023802723031515,
+      "grad_norm": 5.537991292377249,
+      "learning_rate": 2e-05,
+      "loss": 0.7414,
+      "step": 21106
+    },
+    {
+      "epoch": 0.502404075026183,
+      "grad_norm": 1.7375878866831174,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 21107
+    },
+    {
+      "epoch": 0.5024278777492145,
+      "grad_norm": 7.088054494195011,
+      "learning_rate": 2e-05,
+      "loss": 0.6873,
+      "step": 21108
+    },
+    {
+      "epoch": 0.502451680472246,
+      "grad_norm": 1.495598806327716,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 21109
+    },
+    {
+      "epoch": 0.5024754831952776,
+      "grad_norm": 5.979237789143563,
+      "learning_rate": 2e-05,
+      "loss": 1.7931,
+      "step": 21110
+    },
+    {
+      "epoch": 0.5024992859183091,
+      "grad_norm": 3.344684535987929,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 21111
+    },
+    {
+      "epoch": 0.5025230886413405,
+      "grad_norm": 2.894712652382604,
+      "learning_rate": 2e-05,
+      "loss": 0.4316,
+      "step": 21112
+    },
+    {
+      "epoch": 0.5025468913643721,
+      "grad_norm": 5.369268750864929,
+      "learning_rate": 2e-05,
+      "loss": 1.3413,
+      "step": 21113
+    },
+    {
+      "epoch": 0.5025706940874036,
+      "grad_norm": 1.039844411841975,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 21114
+    },
+    {
+      "epoch": 0.5025944968104351,
+      "grad_norm": 2.357174983052094,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 21115
+    },
+    {
+      "epoch": 0.5026182995334666,
+      "grad_norm": 5.477459646490748,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 21116
+    },
+    {
+      "epoch": 0.5026421022564982,
+      "grad_norm": 3.489993390027929,
+      "learning_rate": 2e-05,
+      "loss": 0.5018,
+      "step": 21117
+    },
+    {
+      "epoch": 0.5026659049795297,
+      "grad_norm": 2.540072208321598,
+      "learning_rate": 2e-05,
+      "loss": 0.2911,
+      "step": 21118
+    },
+    {
+      "epoch": 0.5026897077025612,
+      "grad_norm": 0.8185818423847092,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 21119
+    },
+    {
+      "epoch": 0.5027135104255926,
+      "grad_norm": 4.56563873022537,
+      "learning_rate": 2e-05,
+      "loss": 0.9078,
+      "step": 21120
+    },
+    {
+      "epoch": 0.5027373131486242,
+      "grad_norm": 0.38965143731764706,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 21121
+    },
+    {
+      "epoch": 0.5027611158716557,
+      "grad_norm": 2.180091030072913,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 21122
+    },
+    {
+      "epoch": 0.5027849185946872,
+      "grad_norm": 10.86429566138169,
+      "learning_rate": 2e-05,
+      "loss": 0.4028,
+      "step": 21123
+    },
+    {
+      "epoch": 0.5028087213177187,
+      "grad_norm": 1.031358419253745,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 21124
+    },
+    {
+      "epoch": 0.5028325240407503,
+      "grad_norm": 16.429783775382322,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 21125
+    },
+    {
+      "epoch": 0.5028563267637818,
+      "grad_norm": 0.487358222445409,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 21126
+    },
+    {
+      "epoch": 0.5028801294868133,
+      "grad_norm": 2.529120995757905,
+      "learning_rate": 2e-05,
+      "loss": 0.1508,
+      "step": 21127
+    },
+    {
+      "epoch": 0.5029039322098449,
+      "grad_norm": 4.994437331367599,
+      "learning_rate": 2e-05,
+      "loss": 0.8176,
+      "step": 21128
+    },
+    {
+      "epoch": 0.5029277349328763,
+      "grad_norm": 2.528398144932718,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 21129
+    },
+    {
+      "epoch": 0.5029515376559078,
+      "grad_norm": 4.394419546470504,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 21130
+    },
+    {
+      "epoch": 0.5029753403789393,
+      "grad_norm": 14.313006549064063,
+      "learning_rate": 2e-05,
+      "loss": 0.6479,
+      "step": 21131
+    },
+    {
+      "epoch": 0.5029991431019709,
+      "grad_norm": 19.364993734168554,
+      "learning_rate": 2e-05,
+      "loss": 0.7089,
+      "step": 21132
+    },
+    {
+      "epoch": 0.5030229458250024,
+      "grad_norm": 10.08595993634695,
+      "learning_rate": 2e-05,
+      "loss": 0.3392,
+      "step": 21133
+    },
+    {
+      "epoch": 0.5030467485480339,
+      "grad_norm": 3.0255050212321493,
+      "learning_rate": 2e-05,
+      "loss": 0.6757,
+      "step": 21134
+    },
+    {
+      "epoch": 0.5030705512710654,
+      "grad_norm": 0.6445334425204394,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 21135
+    },
+    {
+      "epoch": 0.503094353994097,
+      "grad_norm": 5.617591837564135,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 21136
+    },
+    {
+      "epoch": 0.5031181567171285,
+      "grad_norm": 3.5654987078501494,
+      "learning_rate": 2e-05,
+      "loss": 0.5461,
+      "step": 21137
+    },
+    {
+      "epoch": 0.5031419594401599,
+      "grad_norm": 1.8683186893327606,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 21138
+    },
+    {
+      "epoch": 0.5031657621631914,
+      "grad_norm": 3.8306096069593534,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 21139
+    },
+    {
+      "epoch": 0.503189564886223,
+      "grad_norm": 5.5498859166306715,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 21140
+    },
+    {
+      "epoch": 0.5032133676092545,
+      "grad_norm": 1.5050759184049858,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 21141
+    },
+    {
+      "epoch": 0.503237170332286,
+      "grad_norm": 5.847065733232555,
+      "learning_rate": 2e-05,
+      "loss": 0.8435,
+      "step": 21142
+    },
+    {
+      "epoch": 0.5032609730553175,
+      "grad_norm": 8.192957238223512,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 21143
+    },
+    {
+      "epoch": 0.5032847757783491,
+      "grad_norm": 3.7343834459515834,
+      "learning_rate": 2e-05,
+      "loss": 0.6208,
+      "step": 21144
+    },
+    {
+      "epoch": 0.5033085785013806,
+      "grad_norm": 0.18767765730583033,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21145
+    },
+    {
+      "epoch": 0.503332381224412,
+      "grad_norm": 6.685247194697754,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 21146
+    },
+    {
+      "epoch": 0.5033561839474436,
+      "grad_norm": 0.2701337203385631,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 21147
+    },
+    {
+      "epoch": 0.5033799866704751,
+      "grad_norm": 6.152829433872454,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 21148
+    },
+    {
+      "epoch": 0.5034037893935066,
+      "grad_norm": 5.31566850549824,
+      "learning_rate": 2e-05,
+      "loss": 0.8714,
+      "step": 21149
+    },
+    {
+      "epoch": 0.5034275921165381,
+      "grad_norm": 8.040687630857697,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 21150
+    },
+    {
+      "epoch": 0.5034513948395697,
+      "grad_norm": 4.646205073768194,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 21151
+    },
+    {
+      "epoch": 0.5034751975626012,
+      "grad_norm": 4.839078942190402,
+      "learning_rate": 2e-05,
+      "loss": 0.3703,
+      "step": 21152
+    },
+    {
+      "epoch": 0.5034990002856327,
+      "grad_norm": 6.477367432576535,
+      "learning_rate": 2e-05,
+      "loss": 0.9067,
+      "step": 21153
+    },
+    {
+      "epoch": 0.5035228030086641,
+      "grad_norm": 7.9880060317130335,
+      "learning_rate": 2e-05,
+      "loss": 0.1772,
+      "step": 21154
+    },
+    {
+      "epoch": 0.5035466057316957,
+      "grad_norm": 1.6983327911758939,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 21155
+    },
+    {
+      "epoch": 0.5035704084547272,
+      "grad_norm": 5.404907875528928,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 21156
+    },
+    {
+      "epoch": 0.5035942111777587,
+      "grad_norm": 0.11538808649537603,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 21157
+    },
+    {
+      "epoch": 0.5036180139007902,
+      "grad_norm": 2.932355200865457,
+      "learning_rate": 2e-05,
+      "loss": 0.4987,
+      "step": 21158
+    },
+    {
+      "epoch": 0.5036418166238218,
+      "grad_norm": 12.580757980469086,
+      "learning_rate": 2e-05,
+      "loss": 0.882,
+      "step": 21159
+    },
+    {
+      "epoch": 0.5036656193468533,
+      "grad_norm": 4.794383615542266,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 21160
+    },
+    {
+      "epoch": 0.5036894220698848,
+      "grad_norm": 52.62840586095447,
+      "learning_rate": 2e-05,
+      "loss": 0.8147,
+      "step": 21161
+    },
+    {
+      "epoch": 0.5037132247929164,
+      "grad_norm": 0.4904725982438889,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 21162
+    },
+    {
+      "epoch": 0.5037370275159478,
+      "grad_norm": 10.162490888843452,
+      "learning_rate": 2e-05,
+      "loss": 0.6376,
+      "step": 21163
+    },
+    {
+      "epoch": 0.5037608302389793,
+      "grad_norm": 0.20972807196962398,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 21164
+    },
+    {
+      "epoch": 0.5037846329620108,
+      "grad_norm": 2.303786231852558,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 21165
+    },
+    {
+      "epoch": 0.5038084356850424,
+      "grad_norm": 5.364184584286977,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 21166
+    },
+    {
+      "epoch": 0.5038322384080739,
+      "grad_norm": 4.801944409444456,
+      "learning_rate": 2e-05,
+      "loss": 0.7607,
+      "step": 21167
+    },
+    {
+      "epoch": 0.5038560411311054,
+      "grad_norm": 3.510374422240023,
+      "learning_rate": 2e-05,
+      "loss": 0.5838,
+      "step": 21168
+    },
+    {
+      "epoch": 0.5038798438541369,
+      "grad_norm": 4.709566297085439,
+      "learning_rate": 2e-05,
+      "loss": 0.2321,
+      "step": 21169
+    },
+    {
+      "epoch": 0.5039036465771685,
+      "grad_norm": 3.9287960271939264,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 21170
+    },
+    {
+      "epoch": 0.5039274493002,
+      "grad_norm": 64.47314517097182,
+      "learning_rate": 2e-05,
+      "loss": 0.434,
+      "step": 21171
+    },
+    {
+      "epoch": 0.5039512520232314,
+      "grad_norm": 7.829508601800549,
+      "learning_rate": 2e-05,
+      "loss": 0.4017,
+      "step": 21172
+    },
+    {
+      "epoch": 0.5039750547462629,
+      "grad_norm": 6.013113784224,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 21173
+    },
+    {
+      "epoch": 0.5039988574692945,
+      "grad_norm": 5.596138565153769,
+      "learning_rate": 2e-05,
+      "loss": 0.2129,
+      "step": 21174
+    },
+    {
+      "epoch": 0.504022660192326,
+      "grad_norm": 0.4474216608816669,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21175
+    },
+    {
+      "epoch": 0.5040464629153575,
+      "grad_norm": 1.648186166311534,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 21176
+    },
+    {
+      "epoch": 0.5040702656383891,
+      "grad_norm": 4.419804177139538,
+      "learning_rate": 2e-05,
+      "loss": 0.6608,
+      "step": 21177
+    },
+    {
+      "epoch": 0.5040940683614206,
+      "grad_norm": 1.0316144470158009,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 21178
+    },
+    {
+      "epoch": 0.504117871084452,
+      "grad_norm": 7.497174339572278,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 21179
+    },
+    {
+      "epoch": 0.5041416738074835,
+      "grad_norm": 7.990791287746866,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 21180
+    },
+    {
+      "epoch": 0.5041654765305151,
+      "grad_norm": 1.357813467664756,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 21181
+    },
+    {
+      "epoch": 0.5041892792535466,
+      "grad_norm": 2.421720650435278,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 21182
+    },
+    {
+      "epoch": 0.5042130819765781,
+      "grad_norm": 5.517708090482054,
+      "learning_rate": 2e-05,
+      "loss": 0.2929,
+      "step": 21183
+    },
+    {
+      "epoch": 0.5042368846996096,
+      "grad_norm": 3.0670017905384954,
+      "learning_rate": 2e-05,
+      "loss": 0.2891,
+      "step": 21184
+    },
+    {
+      "epoch": 0.5042606874226412,
+      "grad_norm": 4.845943793848045,
+      "learning_rate": 2e-05,
+      "loss": 0.2048,
+      "step": 21185
+    },
+    {
+      "epoch": 0.5042844901456727,
+      "grad_norm": 0.5917470689639626,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 21186
+    },
+    {
+      "epoch": 0.5043082928687042,
+      "grad_norm": 3.7201377116238317,
+      "learning_rate": 2e-05,
+      "loss": 0.5835,
+      "step": 21187
+    },
+    {
+      "epoch": 0.5043320955917356,
+      "grad_norm": 0.5656358790465212,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 21188
+    },
+    {
+      "epoch": 0.5043558983147672,
+      "grad_norm": 5.0917914778052396,
+      "learning_rate": 2e-05,
+      "loss": 1.0817,
+      "step": 21189
+    },
+    {
+      "epoch": 0.5043797010377987,
+      "grad_norm": 4.435317783703029,
+      "learning_rate": 2e-05,
+      "loss": 0.8424,
+      "step": 21190
+    },
+    {
+      "epoch": 0.5044035037608302,
+      "grad_norm": 0.39260413456684135,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 21191
+    },
+    {
+      "epoch": 0.5044273064838618,
+      "grad_norm": 25.36292857519183,
+      "learning_rate": 2e-05,
+      "loss": 0.8615,
+      "step": 21192
+    },
+    {
+      "epoch": 0.5044511092068933,
+      "grad_norm": 17.937026030629152,
+      "learning_rate": 2e-05,
+      "loss": 0.2391,
+      "step": 21193
+    },
+    {
+      "epoch": 0.5044749119299248,
+      "grad_norm": 5.438098405393168,
+      "learning_rate": 2e-05,
+      "loss": 0.4725,
+      "step": 21194
+    },
+    {
+      "epoch": 0.5044987146529563,
+      "grad_norm": 1.139394790163032,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 21195
+    },
+    {
+      "epoch": 0.5045225173759879,
+      "grad_norm": 3.8851465206046423,
+      "learning_rate": 2e-05,
+      "loss": 0.7582,
+      "step": 21196
+    },
+    {
+      "epoch": 0.5045463200990193,
+      "grad_norm": 18.94773657613513,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 21197
+    },
+    {
+      "epoch": 0.5045701228220508,
+      "grad_norm": 1.9108174916157987,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 21198
+    },
+    {
+      "epoch": 0.5045939255450823,
+      "grad_norm": 3.704532927554986,
+      "learning_rate": 2e-05,
+      "loss": 0.5811,
+      "step": 21199
+    },
+    {
+      "epoch": 0.5046177282681139,
+      "grad_norm": 1.530599544226184,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 21200
+    },
+    {
+      "epoch": 0.5046415309911454,
+      "grad_norm": 5.568898615399413,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 21201
+    },
+    {
+      "epoch": 0.5046653337141769,
+      "grad_norm": 4.6780809669820185,
+      "learning_rate": 2e-05,
+      "loss": 0.762,
+      "step": 21202
+    },
+    {
+      "epoch": 0.5046891364372084,
+      "grad_norm": 7.046034868427438,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 21203
+    },
+    {
+      "epoch": 0.50471293916024,
+      "grad_norm": 8.912075505666266,
+      "learning_rate": 2e-05,
+      "loss": 1.2273,
+      "step": 21204
+    },
+    {
+      "epoch": 0.5047367418832714,
+      "grad_norm": 2.329365678066342,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 21205
+    },
+    {
+      "epoch": 0.5047605446063029,
+      "grad_norm": 0.6530891099304427,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 21206
+    },
+    {
+      "epoch": 0.5047843473293345,
+      "grad_norm": 0.30577970298401963,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 21207
+    },
+    {
+      "epoch": 0.504808150052366,
+      "grad_norm": 3.1144568257601173,
+      "learning_rate": 2e-05,
+      "loss": 0.3677,
+      "step": 21208
+    },
+    {
+      "epoch": 0.5048319527753975,
+      "grad_norm": 2.6364596139743393,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 21209
+    },
+    {
+      "epoch": 0.504855755498429,
+      "grad_norm": 1.2583104359279245,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 21210
+    },
+    {
+      "epoch": 0.5048795582214606,
+      "grad_norm": 3.5396045429738487,
+      "learning_rate": 2e-05,
+      "loss": 0.5777,
+      "step": 21211
+    },
+    {
+      "epoch": 0.5049033609444921,
+      "grad_norm": 4.051798800507296,
+      "learning_rate": 2e-05,
+      "loss": 0.5416,
+      "step": 21212
+    },
+    {
+      "epoch": 0.5049271636675235,
+      "grad_norm": 17.214572565529433,
+      "learning_rate": 2e-05,
+      "loss": 1.1368,
+      "step": 21213
+    },
+    {
+      "epoch": 0.504950966390555,
+      "grad_norm": 3.471804038100921,
+      "learning_rate": 2e-05,
+      "loss": 0.6165,
+      "step": 21214
+    },
+    {
+      "epoch": 0.5049747691135866,
+      "grad_norm": 4.049505352620938,
+      "learning_rate": 2e-05,
+      "loss": 0.6852,
+      "step": 21215
+    },
+    {
+      "epoch": 0.5049985718366181,
+      "grad_norm": 6.262774341705785,
+      "learning_rate": 2e-05,
+      "loss": 0.6931,
+      "step": 21216
+    },
+    {
+      "epoch": 0.5050223745596496,
+      "grad_norm": 10.958675313597194,
+      "learning_rate": 2e-05,
+      "loss": 0.5633,
+      "step": 21217
+    },
+    {
+      "epoch": 0.5050461772826811,
+      "grad_norm": 6.837957775552293,
+      "learning_rate": 2e-05,
+      "loss": 0.8729,
+      "step": 21218
+    },
+    {
+      "epoch": 0.5050699800057127,
+      "grad_norm": 3.5925762037132074,
+      "learning_rate": 2e-05,
+      "loss": 0.4959,
+      "step": 21219
+    },
+    {
+      "epoch": 0.5050937827287442,
+      "grad_norm": 5.839270530974775,
+      "learning_rate": 2e-05,
+      "loss": 0.2415,
+      "step": 21220
+    },
+    {
+      "epoch": 0.5051175854517757,
+      "grad_norm": 2.225325453754094,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 21221
+    },
+    {
+      "epoch": 0.5051413881748072,
+      "grad_norm": 5.132648906866464,
+      "learning_rate": 2e-05,
+      "loss": 0.1646,
+      "step": 21222
+    },
+    {
+      "epoch": 0.5051651908978387,
+      "grad_norm": 5.147814653275179,
+      "learning_rate": 2e-05,
+      "loss": 0.3263,
+      "step": 21223
+    },
+    {
+      "epoch": 0.5051889936208702,
+      "grad_norm": 74.81398205124256,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 21224
+    },
+    {
+      "epoch": 0.5052127963439017,
+      "grad_norm": 8.171695448598655,
+      "learning_rate": 2e-05,
+      "loss": 0.2278,
+      "step": 21225
+    },
+    {
+      "epoch": 0.5052365990669333,
+      "grad_norm": 0.8731138879401276,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 21226
+    },
+    {
+      "epoch": 0.5052604017899648,
+      "grad_norm": 17.84372941260452,
+      "learning_rate": 2e-05,
+      "loss": 0.7715,
+      "step": 21227
+    },
+    {
+      "epoch": 0.5052842045129963,
+      "grad_norm": 4.04209232001426,
+      "learning_rate": 2e-05,
+      "loss": 0.8753,
+      "step": 21228
+    },
+    {
+      "epoch": 0.5053080072360278,
+      "grad_norm": 2.752443363393734,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 21229
+    },
+    {
+      "epoch": 0.5053318099590594,
+      "grad_norm": 3.1047476289543683,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5053556126820908,
+      "grad_norm": 3.1772494066056596,
+      "learning_rate": 2e-05,
+      "loss": 0.5747,
+      "step": 21231
+    },
+    {
+      "epoch": 0.5053794154051223,
+      "grad_norm": 6.448680468046477,
+      "learning_rate": 2e-05,
+      "loss": 0.6549,
+      "step": 21232
+    },
+    {
+      "epoch": 0.5054032181281538,
+      "grad_norm": 0.32455775516378055,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21233
+    },
+    {
+      "epoch": 0.5054270208511854,
+      "grad_norm": 1.4832379323213685,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 21234
+    },
+    {
+      "epoch": 0.5054508235742169,
+      "grad_norm": 4.764836590232409,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 21235
+    },
+    {
+      "epoch": 0.5054746262972484,
+      "grad_norm": 5.073012558285287,
+      "learning_rate": 2e-05,
+      "loss": 0.4997,
+      "step": 21236
+    },
+    {
+      "epoch": 0.50549842902028,
+      "grad_norm": 3.0872883365262833,
+      "learning_rate": 2e-05,
+      "loss": 0.4584,
+      "step": 21237
+    },
+    {
+      "epoch": 0.5055222317433115,
+      "grad_norm": 3.109471934179287,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 21238
+    },
+    {
+      "epoch": 0.5055460344663429,
+      "grad_norm": 3.987303709967376,
+      "learning_rate": 2e-05,
+      "loss": 0.9109,
+      "step": 21239
+    },
+    {
+      "epoch": 0.5055698371893744,
+      "grad_norm": 4.585956820512051,
+      "learning_rate": 2e-05,
+      "loss": 0.3436,
+      "step": 21240
+    },
+    {
+      "epoch": 0.505593639912406,
+      "grad_norm": 5.689347028251708,
+      "learning_rate": 2e-05,
+      "loss": 0.1755,
+      "step": 21241
+    },
+    {
+      "epoch": 0.5056174426354375,
+      "grad_norm": 3.322188777813483,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 21242
+    },
+    {
+      "epoch": 0.505641245358469,
+      "grad_norm": 12.472555057337674,
+      "learning_rate": 2e-05,
+      "loss": 0.9337,
+      "step": 21243
+    },
+    {
+      "epoch": 0.5056650480815005,
+      "grad_norm": 2.530816260666159,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 21244
+    },
+    {
+      "epoch": 0.5056888508045321,
+      "grad_norm": 0.04490283334962292,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 21245
+    },
+    {
+      "epoch": 0.5057126535275636,
+      "grad_norm": 1.0283105561161352,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 21246
+    },
+    {
+      "epoch": 0.505736456250595,
+      "grad_norm": 15.826248793871391,
+      "learning_rate": 2e-05,
+      "loss": 0.836,
+      "step": 21247
+    },
+    {
+      "epoch": 0.5057602589736265,
+      "grad_norm": 14.413322052097682,
+      "learning_rate": 2e-05,
+      "loss": 1.3847,
+      "step": 21248
+    },
+    {
+      "epoch": 0.5057840616966581,
+      "grad_norm": 4.051568102478036,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 21249
+    },
+    {
+      "epoch": 0.5058078644196896,
+      "grad_norm": 0.5932915138223259,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5058316671427211,
+      "grad_norm": 7.480233329819192,
+      "learning_rate": 2e-05,
+      "loss": 1.0549,
+      "step": 21251
+    },
+    {
+      "epoch": 0.5058554698657527,
+      "grad_norm": 1.0882870975114307,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 21252
+    },
+    {
+      "epoch": 0.5058792725887842,
+      "grad_norm": 0.14372549194563441,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 21253
+    },
+    {
+      "epoch": 0.5059030753118157,
+      "grad_norm": 4.071925787116423,
+      "learning_rate": 2e-05,
+      "loss": 0.7009,
+      "step": 21254
+    },
+    {
+      "epoch": 0.5059268780348471,
+      "grad_norm": 2.577782464480181,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 21255
+    },
+    {
+      "epoch": 0.5059506807578787,
+      "grad_norm": 4.436434484703696,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 21256
+    },
+    {
+      "epoch": 0.5059744834809102,
+      "grad_norm": 6.460632714883209,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 21257
+    },
+    {
+      "epoch": 0.5059982862039417,
+      "grad_norm": 0.034695463591243626,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 21258
+    },
+    {
+      "epoch": 0.5060220889269732,
+      "grad_norm": 6.714207747219091,
+      "learning_rate": 2e-05,
+      "loss": 1.3723,
+      "step": 21259
+    },
+    {
+      "epoch": 0.5060458916500048,
+      "grad_norm": 17.138329845917102,
+      "learning_rate": 2e-05,
+      "loss": 0.8373,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5060696943730363,
+      "grad_norm": 4.142801185980391,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 21261
+    },
+    {
+      "epoch": 0.5060934970960678,
+      "grad_norm": 4.703552255794362,
+      "learning_rate": 2e-05,
+      "loss": 0.6704,
+      "step": 21262
+    },
+    {
+      "epoch": 0.5061172998190993,
+      "grad_norm": 7.049863308895901,
+      "learning_rate": 2e-05,
+      "loss": 0.6556,
+      "step": 21263
+    },
+    {
+      "epoch": 0.5061411025421308,
+      "grad_norm": 8.814758319150016,
+      "learning_rate": 2e-05,
+      "loss": 0.454,
+      "step": 21264
+    },
+    {
+      "epoch": 0.5061649052651623,
+      "grad_norm": 2.765908638975233,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 21265
+    },
+    {
+      "epoch": 0.5061887079881938,
+      "grad_norm": 13.213731353416538,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 21266
+    },
+    {
+      "epoch": 0.5062125107112254,
+      "grad_norm": 6.993405378854368,
+      "learning_rate": 2e-05,
+      "loss": 0.2585,
+      "step": 21267
+    },
+    {
+      "epoch": 0.5062363134342569,
+      "grad_norm": 1.414384073586502,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 21268
+    },
+    {
+      "epoch": 0.5062601161572884,
+      "grad_norm": 3.911940813770942,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 21269
+    },
+    {
+      "epoch": 0.5062839188803199,
+      "grad_norm": 5.988063361871999,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 21270
+    },
+    {
+      "epoch": 0.5063077216033515,
+      "grad_norm": 3.166818335014649,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 21271
+    },
+    {
+      "epoch": 0.506331524326383,
+      "grad_norm": 4.408005030663078,
+      "learning_rate": 2e-05,
+      "loss": 0.4566,
+      "step": 21272
+    },
+    {
+      "epoch": 0.5063553270494144,
+      "grad_norm": 3.488520223817308,
+      "learning_rate": 2e-05,
+      "loss": 0.5346,
+      "step": 21273
+    },
+    {
+      "epoch": 0.5063791297724459,
+      "grad_norm": 2.4882478017568697,
+      "learning_rate": 2e-05,
+      "loss": 0.3039,
+      "step": 21274
+    },
+    {
+      "epoch": 0.5064029324954775,
+      "grad_norm": 4.269465543198624,
+      "learning_rate": 2e-05,
+      "loss": 1.505,
+      "step": 21275
+    },
+    {
+      "epoch": 0.506426735218509,
+      "grad_norm": 4.970992705982167,
+      "learning_rate": 2e-05,
+      "loss": 0.5761,
+      "step": 21276
+    },
+    {
+      "epoch": 0.5064505379415405,
+      "grad_norm": 3.006538337337995,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 21277
+    },
+    {
+      "epoch": 0.506474340664572,
+      "grad_norm": 16.065007215082556,
+      "learning_rate": 2e-05,
+      "loss": 0.7456,
+      "step": 21278
+    },
+    {
+      "epoch": 0.5064981433876036,
+      "grad_norm": 2.704249689021655,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 21279
+    },
+    {
+      "epoch": 0.5065219461106351,
+      "grad_norm": 2.0955157384940213,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5065457488336665,
+      "grad_norm": 4.642590649071549,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 21281
+    },
+    {
+      "epoch": 0.5065695515566981,
+      "grad_norm": 1.3503912968755218,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 21282
+    },
+    {
+      "epoch": 0.5065933542797296,
+      "grad_norm": 4.804884743991234,
+      "learning_rate": 2e-05,
+      "loss": 0.4316,
+      "step": 21283
+    },
+    {
+      "epoch": 0.5066171570027611,
+      "grad_norm": 5.777186990818342,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 21284
+    },
+    {
+      "epoch": 0.5066409597257926,
+      "grad_norm": 4.904712015900245,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 21285
+    },
+    {
+      "epoch": 0.5066647624488242,
+      "grad_norm": 7.754920896617759,
+      "learning_rate": 2e-05,
+      "loss": 0.7246,
+      "step": 21286
+    },
+    {
+      "epoch": 0.5066885651718557,
+      "grad_norm": 6.357184691371293,
+      "learning_rate": 2e-05,
+      "loss": 0.243,
+      "step": 21287
+    },
+    {
+      "epoch": 0.5067123678948872,
+      "grad_norm": 4.5486675655346325,
+      "learning_rate": 2e-05,
+      "loss": 0.8222,
+      "step": 21288
+    },
+    {
+      "epoch": 0.5067361706179186,
+      "grad_norm": 5.102453216400652,
+      "learning_rate": 2e-05,
+      "loss": 0.7024,
+      "step": 21289
+    },
+    {
+      "epoch": 0.5067599733409502,
+      "grad_norm": 4.453847532615111,
+      "learning_rate": 2e-05,
+      "loss": 0.7501,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5067837760639817,
+      "grad_norm": 0.33727510059200627,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 21291
+    },
+    {
+      "epoch": 0.5068075787870132,
+      "grad_norm": 0.21589312044129477,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21292
+    },
+    {
+      "epoch": 0.5068313815100447,
+      "grad_norm": 3.229571984674851,
+      "learning_rate": 2e-05,
+      "loss": 0.4038,
+      "step": 21293
+    },
+    {
+      "epoch": 0.5068551842330763,
+      "grad_norm": 0.5124025580024782,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 21294
+    },
+    {
+      "epoch": 0.5068789869561078,
+      "grad_norm": 3.321235069788577,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 21295
+    },
+    {
+      "epoch": 0.5069027896791393,
+      "grad_norm": 0.30093286825279336,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21296
+    },
+    {
+      "epoch": 0.5069265924021709,
+      "grad_norm": 10.766856559541107,
+      "learning_rate": 2e-05,
+      "loss": 1.2884,
+      "step": 21297
+    },
+    {
+      "epoch": 0.5069503951252023,
+      "grad_norm": 3.435924511067249,
+      "learning_rate": 2e-05,
+      "loss": 0.5747,
+      "step": 21298
+    },
+    {
+      "epoch": 0.5069741978482338,
+      "grad_norm": 2.936495627170838,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 21299
+    },
+    {
+      "epoch": 0.5069980005712653,
+      "grad_norm": 3.218046177231349,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5070218032942969,
+      "grad_norm": 0.20713642551137373,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 21301
+    },
+    {
+      "epoch": 0.5070456060173284,
+      "grad_norm": 6.270320764556285,
+      "learning_rate": 2e-05,
+      "loss": 0.3763,
+      "step": 21302
+    },
+    {
+      "epoch": 0.5070694087403599,
+      "grad_norm": 4.097139848238552,
+      "learning_rate": 2e-05,
+      "loss": 0.6582,
+      "step": 21303
+    },
+    {
+      "epoch": 0.5070932114633914,
+      "grad_norm": 5.49449837124989,
+      "learning_rate": 2e-05,
+      "loss": 0.4857,
+      "step": 21304
+    },
+    {
+      "epoch": 0.507117014186423,
+      "grad_norm": 3.0500358029638592,
+      "learning_rate": 2e-05,
+      "loss": 0.384,
+      "step": 21305
+    },
+    {
+      "epoch": 0.5071408169094545,
+      "grad_norm": 8.081992171585243,
+      "learning_rate": 2e-05,
+      "loss": 0.285,
+      "step": 21306
+    },
+    {
+      "epoch": 0.5071646196324859,
+      "grad_norm": 3.242533289116694,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 21307
+    },
+    {
+      "epoch": 0.5071884223555174,
+      "grad_norm": 3.178380915086551,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 21308
+    },
+    {
+      "epoch": 0.507212225078549,
+      "grad_norm": 1.294723471486939,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 21309
+    },
+    {
+      "epoch": 0.5072360278015805,
+      "grad_norm": 1.547321432622339,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 21310
+    },
+    {
+      "epoch": 0.507259830524612,
+      "grad_norm": 3.1752155432950104,
+      "learning_rate": 2e-05,
+      "loss": 0.4537,
+      "step": 21311
+    },
+    {
+      "epoch": 0.5072836332476436,
+      "grad_norm": 5.029051567927068,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 21312
+    },
+    {
+      "epoch": 0.5073074359706751,
+      "grad_norm": 4.011973099716972,
+      "learning_rate": 2e-05,
+      "loss": 0.7994,
+      "step": 21313
+    },
+    {
+      "epoch": 0.5073312386937066,
+      "grad_norm": 4.184234906900047,
+      "learning_rate": 2e-05,
+      "loss": 0.6749,
+      "step": 21314
+    },
+    {
+      "epoch": 0.507355041416738,
+      "grad_norm": 1.374240385840991,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 21315
+    },
+    {
+      "epoch": 0.5073788441397696,
+      "grad_norm": 2.175675637443023,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 21316
+    },
+    {
+      "epoch": 0.5074026468628011,
+      "grad_norm": 6.420165867512541,
+      "learning_rate": 2e-05,
+      "loss": 1.5266,
+      "step": 21317
+    },
+    {
+      "epoch": 0.5074264495858326,
+      "grad_norm": 4.37779786361824,
+      "learning_rate": 2e-05,
+      "loss": 0.6962,
+      "step": 21318
+    },
+    {
+      "epoch": 0.5074502523088641,
+      "grad_norm": 4.158391253687289,
+      "learning_rate": 2e-05,
+      "loss": 0.7349,
+      "step": 21319
+    },
+    {
+      "epoch": 0.5074740550318957,
+      "grad_norm": 1.7512580862778724,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5074978577549272,
+      "grad_norm": 3.996731474424277,
+      "learning_rate": 2e-05,
+      "loss": 0.5768,
+      "step": 21321
+    },
+    {
+      "epoch": 0.5075216604779587,
+      "grad_norm": 8.07880077484685,
+      "learning_rate": 2e-05,
+      "loss": 1.1211,
+      "step": 21322
+    },
+    {
+      "epoch": 0.5075454632009901,
+      "grad_norm": 3.3763291960395656,
+      "learning_rate": 2e-05,
+      "loss": 0.4371,
+      "step": 21323
+    },
+    {
+      "epoch": 0.5075692659240217,
+      "grad_norm": 4.1120320394740135,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 21324
+    },
+    {
+      "epoch": 0.5075930686470532,
+      "grad_norm": 13.429528074325741,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 21325
+    },
+    {
+      "epoch": 0.5076168713700847,
+      "grad_norm": 1.6001435901716374,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 21326
+    },
+    {
+      "epoch": 0.5076406740931163,
+      "grad_norm": 2.9026876060229005,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 21327
+    },
+    {
+      "epoch": 0.5076644768161478,
+      "grad_norm": 4.56288633119939,
+      "learning_rate": 2e-05,
+      "loss": 0.8553,
+      "step": 21328
+    },
+    {
+      "epoch": 0.5076882795391793,
+      "grad_norm": 4.374093413823598,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 21329
+    },
+    {
+      "epoch": 0.5077120822622108,
+      "grad_norm": 19.854290950995196,
+      "learning_rate": 2e-05,
+      "loss": 1.1989,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5077358849852424,
+      "grad_norm": 5.6536500126870886,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 21331
+    },
+    {
+      "epoch": 0.5077596877082738,
+      "grad_norm": 9.332869693749835,
+      "learning_rate": 2e-05,
+      "loss": 0.7011,
+      "step": 21332
+    },
+    {
+      "epoch": 0.5077834904313053,
+      "grad_norm": 2.7124026759161,
+      "learning_rate": 2e-05,
+      "loss": 0.2873,
+      "step": 21333
+    },
+    {
+      "epoch": 0.5078072931543368,
+      "grad_norm": 2.090844238601335,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 21334
+    },
+    {
+      "epoch": 0.5078310958773684,
+      "grad_norm": 0.9234830086270961,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 21335
+    },
+    {
+      "epoch": 0.5078548986003999,
+      "grad_norm": 2.754986328847308,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 21336
+    },
+    {
+      "epoch": 0.5078787013234314,
+      "grad_norm": 0.6924603258105706,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 21337
+    },
+    {
+      "epoch": 0.5079025040464629,
+      "grad_norm": 5.8466222990718375,
+      "learning_rate": 2e-05,
+      "loss": 1.2093,
+      "step": 21338
+    },
+    {
+      "epoch": 0.5079263067694945,
+      "grad_norm": 3.6756732880036282,
+      "learning_rate": 2e-05,
+      "loss": 0.558,
+      "step": 21339
+    },
+    {
+      "epoch": 0.507950109492526,
+      "grad_norm": 8.80339838940777,
+      "learning_rate": 2e-05,
+      "loss": 0.196,
+      "step": 21340
+    },
+    {
+      "epoch": 0.5079739122155574,
+      "grad_norm": 3.847023511828029,
+      "learning_rate": 2e-05,
+      "loss": 0.478,
+      "step": 21341
+    },
+    {
+      "epoch": 0.507997714938589,
+      "grad_norm": 1.2885841543038754,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21342
+    },
+    {
+      "epoch": 0.5080215176616205,
+      "grad_norm": 0.16868945977967145,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21343
+    },
+    {
+      "epoch": 0.508045320384652,
+      "grad_norm": 1.6862117075419811,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 21344
+    },
+    {
+      "epoch": 0.5080691231076835,
+      "grad_norm": 4.680267908691585,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 21345
+    },
+    {
+      "epoch": 0.5080929258307151,
+      "grad_norm": 9.474813410943291,
+      "learning_rate": 2e-05,
+      "loss": 0.1847,
+      "step": 21346
+    },
+    {
+      "epoch": 0.5081167285537466,
+      "grad_norm": 2.3236601635475407,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 21347
+    },
+    {
+      "epoch": 0.508140531276778,
+      "grad_norm": 3.3022567903230304,
+      "learning_rate": 2e-05,
+      "loss": 0.7832,
+      "step": 21348
+    },
+    {
+      "epoch": 0.5081643339998095,
+      "grad_norm": 6.192298649754222,
+      "learning_rate": 2e-05,
+      "loss": 1.0927,
+      "step": 21349
+    },
+    {
+      "epoch": 0.5081881367228411,
+      "grad_norm": 2.921016499930005,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5082119394458726,
+      "grad_norm": 5.463758497889853,
+      "learning_rate": 2e-05,
+      "loss": 0.6657,
+      "step": 21351
+    },
+    {
+      "epoch": 0.5082357421689041,
+      "grad_norm": 0.33572176987637165,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 21352
+    },
+    {
+      "epoch": 0.5082595448919356,
+      "grad_norm": 1.77067138303975,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 21353
+    },
+    {
+      "epoch": 0.5082833476149672,
+      "grad_norm": 0.1938038155546515,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 21354
+    },
+    {
+      "epoch": 0.5083071503379987,
+      "grad_norm": 6.373342201823837,
+      "learning_rate": 2e-05,
+      "loss": 1.3062,
+      "step": 21355
+    },
+    {
+      "epoch": 0.5083309530610302,
+      "grad_norm": 0.191204187680033,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21356
+    },
+    {
+      "epoch": 0.5083547557840618,
+      "grad_norm": 13.114546906746417,
+      "learning_rate": 2e-05,
+      "loss": 0.8661,
+      "step": 21357
+    },
+    {
+      "epoch": 0.5083785585070932,
+      "grad_norm": 2.4791424729809655,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 21358
+    },
+    {
+      "epoch": 0.5084023612301247,
+      "grad_norm": 4.04787366482098,
+      "learning_rate": 2e-05,
+      "loss": 0.2556,
+      "step": 21359
+    },
+    {
+      "epoch": 0.5084261639531562,
+      "grad_norm": 4.225012512495777,
+      "learning_rate": 2e-05,
+      "loss": 0.5975,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5084499666761878,
+      "grad_norm": 5.287920568517625,
+      "learning_rate": 2e-05,
+      "loss": 0.997,
+      "step": 21361
+    },
+    {
+      "epoch": 0.5084737693992193,
+      "grad_norm": 9.930001711423948,
+      "learning_rate": 2e-05,
+      "loss": 0.5876,
+      "step": 21362
+    },
+    {
+      "epoch": 0.5084975721222508,
+      "grad_norm": 0.33586266820800986,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 21363
+    },
+    {
+      "epoch": 0.5085213748452823,
+      "grad_norm": 3.4535179200179478,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 21364
+    },
+    {
+      "epoch": 0.5085451775683139,
+      "grad_norm": 5.758918553206285,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 21365
+    },
+    {
+      "epoch": 0.5085689802913453,
+      "grad_norm": 6.149672097760236,
+      "learning_rate": 2e-05,
+      "loss": 1.1459,
+      "step": 21366
+    },
+    {
+      "epoch": 0.5085927830143768,
+      "grad_norm": 4.799329352024787,
+      "learning_rate": 2e-05,
+      "loss": 0.4065,
+      "step": 21367
+    },
+    {
+      "epoch": 0.5086165857374083,
+      "grad_norm": 7.44782152296692,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 21368
+    },
+    {
+      "epoch": 0.5086403884604399,
+      "grad_norm": 0.6578878976589183,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21369
+    },
+    {
+      "epoch": 0.5086641911834714,
+      "grad_norm": 11.484763592062208,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5086879939065029,
+      "grad_norm": 2.0854130847861483,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 21371
+    },
+    {
+      "epoch": 0.5087117966295345,
+      "grad_norm": 4.6101487991522765,
+      "learning_rate": 2e-05,
+      "loss": 0.4802,
+      "step": 21372
+    },
+    {
+      "epoch": 0.508735599352566,
+      "grad_norm": 1.5187555869623297,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 21373
+    },
+    {
+      "epoch": 0.5087594020755974,
+      "grad_norm": 5.326060903727114,
+      "learning_rate": 2e-05,
+      "loss": 0.6606,
+      "step": 21374
+    },
+    {
+      "epoch": 0.5087832047986289,
+      "grad_norm": 0.2917404673763112,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 21375
+    },
+    {
+      "epoch": 0.5088070075216605,
+      "grad_norm": 3.2141626684419986,
+      "learning_rate": 2e-05,
+      "loss": 0.4019,
+      "step": 21376
+    },
+    {
+      "epoch": 0.508830810244692,
+      "grad_norm": 15.964366271545238,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 21377
+    },
+    {
+      "epoch": 0.5088546129677235,
+      "grad_norm": 3.382587056536366,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 21378
+    },
+    {
+      "epoch": 0.508878415690755,
+      "grad_norm": 3.9539043195351695,
+      "learning_rate": 2e-05,
+      "loss": 0.6763,
+      "step": 21379
+    },
+    {
+      "epoch": 0.5089022184137866,
+      "grad_norm": 5.220316360126783,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5089260211368181,
+      "grad_norm": 3.6404120071018866,
+      "learning_rate": 2e-05,
+      "loss": 0.7067,
+      "step": 21381
+    },
+    {
+      "epoch": 0.5089498238598495,
+      "grad_norm": 3.6232854988048,
+      "learning_rate": 2e-05,
+      "loss": 0.5971,
+      "step": 21382
+    },
+    {
+      "epoch": 0.508973626582881,
+      "grad_norm": 88.85030747740458,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21383
+    },
+    {
+      "epoch": 0.5089974293059126,
+      "grad_norm": 3.278670878966288,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 21384
+    },
+    {
+      "epoch": 0.5090212320289441,
+      "grad_norm": 2.9143011081511014,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 21385
+    },
+    {
+      "epoch": 0.5090450347519756,
+      "grad_norm": 4.110181324048474,
+      "learning_rate": 2e-05,
+      "loss": 0.4179,
+      "step": 21386
+    },
+    {
+      "epoch": 0.5090688374750071,
+      "grad_norm": 6.816620000498153,
+      "learning_rate": 2e-05,
+      "loss": 0.2269,
+      "step": 21387
+    },
+    {
+      "epoch": 0.5090926401980387,
+      "grad_norm": 3.389957950030047,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 21388
+    },
+    {
+      "epoch": 0.5091164429210702,
+      "grad_norm": 0.22431191690513128,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 21389
+    },
+    {
+      "epoch": 0.5091402456441017,
+      "grad_norm": 2.3469905839706797,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5091640483671332,
+      "grad_norm": 4.530394349316471,
+      "learning_rate": 2e-05,
+      "loss": 0.1683,
+      "step": 21391
+    },
+    {
+      "epoch": 0.5091878510901647,
+      "grad_norm": 6.981744141762049,
+      "learning_rate": 2e-05,
+      "loss": 0.458,
+      "step": 21392
+    },
+    {
+      "epoch": 0.5092116538131962,
+      "grad_norm": 3.5644473463036928,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 21393
+    },
+    {
+      "epoch": 0.5092354565362277,
+      "grad_norm": 1.769436857385495,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 21394
+    },
+    {
+      "epoch": 0.5092592592592593,
+      "grad_norm": 1.2513010368057247,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 21395
+    },
+    {
+      "epoch": 0.5092830619822908,
+      "grad_norm": 28.12917179516184,
+      "learning_rate": 2e-05,
+      "loss": 0.3024,
+      "step": 21396
+    },
+    {
+      "epoch": 0.5093068647053223,
+      "grad_norm": 3.48995744352533,
+      "learning_rate": 2e-05,
+      "loss": 0.357,
+      "step": 21397
+    },
+    {
+      "epoch": 0.5093306674283538,
+      "grad_norm": 4.0247259712820185,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 21398
+    },
+    {
+      "epoch": 0.5093544701513854,
+      "grad_norm": 4.112929433516194,
+      "learning_rate": 2e-05,
+      "loss": 0.7355,
+      "step": 21399
+    },
+    {
+      "epoch": 0.5093782728744168,
+      "grad_norm": 8.883046910920365,
+      "learning_rate": 2e-05,
+      "loss": 1.1932,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5094020755974483,
+      "grad_norm": 3.852794488110489,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 21401
+    },
+    {
+      "epoch": 0.5094258783204798,
+      "grad_norm": 0.052945074301189714,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21402
+    },
+    {
+      "epoch": 0.5094496810435114,
+      "grad_norm": 15.58299977414292,
+      "learning_rate": 2e-05,
+      "loss": 0.9398,
+      "step": 21403
+    },
+    {
+      "epoch": 0.5094734837665429,
+      "grad_norm": 2.840256445323788,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 21404
+    },
+    {
+      "epoch": 0.5094972864895744,
+      "grad_norm": 0.8429655086527572,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 21405
+    },
+    {
+      "epoch": 0.509521089212606,
+      "grad_norm": 2.268461534839542,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 21406
+    },
+    {
+      "epoch": 0.5095448919356375,
+      "grad_norm": 4.338901773099596,
+      "learning_rate": 2e-05,
+      "loss": 0.7667,
+      "step": 21407
+    },
+    {
+      "epoch": 0.5095686946586689,
+      "grad_norm": 8.634451223134214,
+      "learning_rate": 2e-05,
+      "loss": 1.474,
+      "step": 21408
+    },
+    {
+      "epoch": 0.5095924973817004,
+      "grad_norm": 3.855544853629631,
+      "learning_rate": 2e-05,
+      "loss": 0.4145,
+      "step": 21409
+    },
+    {
+      "epoch": 0.509616300104732,
+      "grad_norm": 0.9576706425748508,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 21410
+    },
+    {
+      "epoch": 0.5096401028277635,
+      "grad_norm": 0.03532779541546907,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 21411
+    },
+    {
+      "epoch": 0.509663905550795,
+      "grad_norm": 0.4736541490803341,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 21412
+    },
+    {
+      "epoch": 0.5096877082738265,
+      "grad_norm": 2.9273458397264136,
+      "learning_rate": 2e-05,
+      "loss": 0.2607,
+      "step": 21413
+    },
+    {
+      "epoch": 0.5097115109968581,
+      "grad_norm": 2.978254757570927,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 21414
+    },
+    {
+      "epoch": 0.5097353137198896,
+      "grad_norm": 2.8943612996017376,
+      "learning_rate": 2e-05,
+      "loss": 0.3143,
+      "step": 21415
+    },
+    {
+      "epoch": 0.509759116442921,
+      "grad_norm": 0.09487685449647883,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 21416
+    },
+    {
+      "epoch": 0.5097829191659525,
+      "grad_norm": 3.9199416767085666,
+      "learning_rate": 2e-05,
+      "loss": 0.418,
+      "step": 21417
+    },
+    {
+      "epoch": 0.5098067218889841,
+      "grad_norm": 0.2259249244266216,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 21418
+    },
+    {
+      "epoch": 0.5098305246120156,
+      "grad_norm": 1.7206474164344931,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 21419
+    },
+    {
+      "epoch": 0.5098543273350471,
+      "grad_norm": 0.9041761491030856,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 21420
+    },
+    {
+      "epoch": 0.5098781300580787,
+      "grad_norm": 2.185233088471328,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 21421
+    },
+    {
+      "epoch": 0.5099019327811102,
+      "grad_norm": 4.555659419687407,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 21422
+    },
+    {
+      "epoch": 0.5099257355041417,
+      "grad_norm": 5.256253195018325,
+      "learning_rate": 2e-05,
+      "loss": 0.5983,
+      "step": 21423
+    },
+    {
+      "epoch": 0.5099495382271731,
+      "grad_norm": 1.9009439893679707,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 21424
+    },
+    {
+      "epoch": 0.5099733409502047,
+      "grad_norm": 1.1328299737618834,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 21425
+    },
+    {
+      "epoch": 0.5099971436732362,
+      "grad_norm": 4.300811962018068,
+      "learning_rate": 2e-05,
+      "loss": 0.4381,
+      "step": 21426
+    },
+    {
+      "epoch": 0.5100209463962677,
+      "grad_norm": 1.972107240063683,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 21427
+    },
+    {
+      "epoch": 0.5100447491192992,
+      "grad_norm": 0.978898659838648,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 21428
+    },
+    {
+      "epoch": 0.5100685518423308,
+      "grad_norm": 0.08056562291357684,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21429
+    },
+    {
+      "epoch": 0.5100923545653623,
+      "grad_norm": 2.342810790322622,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5101161572883938,
+      "grad_norm": 9.259717662901204,
+      "learning_rate": 2e-05,
+      "loss": 0.8621,
+      "step": 21431
+    },
+    {
+      "epoch": 0.5101399600114253,
+      "grad_norm": 2.979522060727547,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 21432
+    },
+    {
+      "epoch": 0.5101637627344568,
+      "grad_norm": 0.12428885905528382,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 21433
+    },
+    {
+      "epoch": 0.5101875654574883,
+      "grad_norm": 3.991153304831888,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 21434
+    },
+    {
+      "epoch": 0.5102113681805198,
+      "grad_norm": 4.295859510932197,
+      "learning_rate": 2e-05,
+      "loss": 0.6086,
+      "step": 21435
+    },
+    {
+      "epoch": 0.5102351709035514,
+      "grad_norm": 3.125064921514609,
+      "learning_rate": 2e-05,
+      "loss": 0.124,
+      "step": 21436
+    },
+    {
+      "epoch": 0.5102589736265829,
+      "grad_norm": 6.985946735397518,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 21437
+    },
+    {
+      "epoch": 0.5102827763496144,
+      "grad_norm": 4.350364117473112,
+      "learning_rate": 2e-05,
+      "loss": 0.7066,
+      "step": 21438
+    },
+    {
+      "epoch": 0.5103065790726459,
+      "grad_norm": 2.134669286936372,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 21439
+    },
+    {
+      "epoch": 0.5103303817956775,
+      "grad_norm": 5.311102426925881,
+      "learning_rate": 2e-05,
+      "loss": 0.2963,
+      "step": 21440
+    },
+    {
+      "epoch": 0.510354184518709,
+      "grad_norm": 0.35351654045028663,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21441
+    },
+    {
+      "epoch": 0.5103779872417404,
+      "grad_norm": 8.126099447374354,
+      "learning_rate": 2e-05,
+      "loss": 0.4662,
+      "step": 21442
+    },
+    {
+      "epoch": 0.5104017899647719,
+      "grad_norm": 13.773503518895138,
+      "learning_rate": 2e-05,
+      "loss": 1.0416,
+      "step": 21443
+    },
+    {
+      "epoch": 0.5104255926878035,
+      "grad_norm": 11.959485063077702,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 21444
+    },
+    {
+      "epoch": 0.510449395410835,
+      "grad_norm": 11.950781038510755,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 21445
+    },
+    {
+      "epoch": 0.5104731981338665,
+      "grad_norm": 7.377662111348026,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 21446
+    },
+    {
+      "epoch": 0.510497000856898,
+      "grad_norm": 3.986780616439357,
+      "learning_rate": 2e-05,
+      "loss": 0.636,
+      "step": 21447
+    },
+    {
+      "epoch": 0.5105208035799296,
+      "grad_norm": 3.5120410060661342,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 21448
+    },
+    {
+      "epoch": 0.5105446063029611,
+      "grad_norm": 1.245836343972201,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 21449
+    },
+    {
+      "epoch": 0.5105684090259925,
+      "grad_norm": 1.4679817246615183,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 21450
+    },
+    {
+      "epoch": 0.5105922117490241,
+      "grad_norm": 3.8902246766911683,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 21451
+    },
+    {
+      "epoch": 0.5106160144720556,
+      "grad_norm": 2.8207756624184603,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 21452
+    },
+    {
+      "epoch": 0.5106398171950871,
+      "grad_norm": 4.0762340333462355,
+      "learning_rate": 2e-05,
+      "loss": 0.982,
+      "step": 21453
+    },
+    {
+      "epoch": 0.5106636199181186,
+      "grad_norm": 4.63131450581304,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 21454
+    },
+    {
+      "epoch": 0.5106874226411502,
+      "grad_norm": 3.7880739777866212,
+      "learning_rate": 2e-05,
+      "loss": 0.1811,
+      "step": 21455
+    },
+    {
+      "epoch": 0.5107112253641817,
+      "grad_norm": 3.283971795224282,
+      "learning_rate": 2e-05,
+      "loss": 0.3456,
+      "step": 21456
+    },
+    {
+      "epoch": 0.5107350280872132,
+      "grad_norm": 0.24917841287398879,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 21457
+    },
+    {
+      "epoch": 0.5107588308102446,
+      "grad_norm": 6.121329585339064,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 21458
+    },
+    {
+      "epoch": 0.5107826335332762,
+      "grad_norm": 7.055372974688015,
+      "learning_rate": 2e-05,
+      "loss": 0.371,
+      "step": 21459
+    },
+    {
+      "epoch": 0.5108064362563077,
+      "grad_norm": 6.452952841604987,
+      "learning_rate": 2e-05,
+      "loss": 0.2004,
+      "step": 21460
+    },
+    {
+      "epoch": 0.5108302389793392,
+      "grad_norm": 0.013077958093483309,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 21461
+    },
+    {
+      "epoch": 0.5108540417023707,
+      "grad_norm": 3.911310702403965,
+      "learning_rate": 2e-05,
+      "loss": 0.2107,
+      "step": 21462
+    },
+    {
+      "epoch": 0.5108778444254023,
+      "grad_norm": 0.21375578514846072,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21463
+    },
+    {
+      "epoch": 0.5109016471484338,
+      "grad_norm": 3.6481267710506415,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 21464
+    },
+    {
+      "epoch": 0.5109254498714653,
+      "grad_norm": 0.1937987932016869,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21465
+    },
+    {
+      "epoch": 0.5109492525944969,
+      "grad_norm": 3.327899630311618,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 21466
+    },
+    {
+      "epoch": 0.5109730553175283,
+      "grad_norm": 4.23847087690574,
+      "learning_rate": 2e-05,
+      "loss": 0.1906,
+      "step": 21467
+    },
+    {
+      "epoch": 0.5109968580405598,
+      "grad_norm": 4.82992569103149,
+      "learning_rate": 2e-05,
+      "loss": 0.4244,
+      "step": 21468
+    },
+    {
+      "epoch": 0.5110206607635913,
+      "grad_norm": 3034.080003082163,
+      "learning_rate": 2e-05,
+      "loss": 0.8802,
+      "step": 21469
+    },
+    {
+      "epoch": 0.5110444634866229,
+      "grad_norm": 0.5473376618737317,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 21470
+    },
+    {
+      "epoch": 0.5110682662096544,
+      "grad_norm": 2.415965450208656,
+      "learning_rate": 2e-05,
+      "loss": 0.1378,
+      "step": 21471
+    },
+    {
+      "epoch": 0.5110920689326859,
+      "grad_norm": 3.669818579900911,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 21472
+    },
+    {
+      "epoch": 0.5111158716557174,
+      "grad_norm": 0.8052738327141236,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 21473
+    },
+    {
+      "epoch": 0.511139674378749,
+      "grad_norm": 2.845928127354446,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 21474
+    },
+    {
+      "epoch": 0.5111634771017805,
+      "grad_norm": 5.38777974946595,
+      "learning_rate": 2e-05,
+      "loss": 0.2192,
+      "step": 21475
+    },
+    {
+      "epoch": 0.5111872798248119,
+      "grad_norm": 8.191783621845335,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 21476
+    },
+    {
+      "epoch": 0.5112110825478434,
+      "grad_norm": 0.7173061632463732,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 21477
+    },
+    {
+      "epoch": 0.511234885270875,
+      "grad_norm": 4.505037324684714,
+      "learning_rate": 2e-05,
+      "loss": 0.5852,
+      "step": 21478
+    },
+    {
+      "epoch": 0.5112586879939065,
+      "grad_norm": 1.4088514987586072,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 21479
+    },
+    {
+      "epoch": 0.511282490716938,
+      "grad_norm": 2.579871347425341,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 21480
+    },
+    {
+      "epoch": 0.5113062934399696,
+      "grad_norm": 53.269727066846556,
+      "learning_rate": 2e-05,
+      "loss": 1.4976,
+      "step": 21481
+    },
+    {
+      "epoch": 0.5113300961630011,
+      "grad_norm": 4.827408701119217,
+      "learning_rate": 2e-05,
+      "loss": 0.7465,
+      "step": 21482
+    },
+    {
+      "epoch": 0.5113538988860326,
+      "grad_norm": 6.84178236891671,
+      "learning_rate": 2e-05,
+      "loss": 0.6889,
+      "step": 21483
+    },
+    {
+      "epoch": 0.511377701609064,
+      "grad_norm": 18.71160491219182,
+      "learning_rate": 2e-05,
+      "loss": 0.8588,
+      "step": 21484
+    },
+    {
+      "epoch": 0.5114015043320956,
+      "grad_norm": 27.9251217828198,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 21485
+    },
+    {
+      "epoch": 0.5114253070551271,
+      "grad_norm": 3.0980164980668743,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 21486
+    },
+    {
+      "epoch": 0.5114491097781586,
+      "grad_norm": 2.8912066846810154,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 21487
+    },
+    {
+      "epoch": 0.5114729125011901,
+      "grad_norm": 3.046789264036465,
+      "learning_rate": 2e-05,
+      "loss": 0.4694,
+      "step": 21488
+    },
+    {
+      "epoch": 0.5114967152242217,
+      "grad_norm": 0.5256305678327291,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 21489
+    },
+    {
+      "epoch": 0.5115205179472532,
+      "grad_norm": 21.080510371091755,
+      "learning_rate": 2e-05,
+      "loss": 0.35,
+      "step": 21490
+    },
+    {
+      "epoch": 0.5115443206702847,
+      "grad_norm": 2.6160601564216406,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 21491
+    },
+    {
+      "epoch": 0.5115681233933161,
+      "grad_norm": 11.909296755707539,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 21492
+    },
+    {
+      "epoch": 0.5115919261163477,
+      "grad_norm": 0.48965411956914756,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 21493
+    },
+    {
+      "epoch": 0.5116157288393792,
+      "grad_norm": 2.9206134022874974,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 21494
+    },
+    {
+      "epoch": 0.5116395315624107,
+      "grad_norm": 0.4738076690803154,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 21495
+    },
+    {
+      "epoch": 0.5116633342854423,
+      "grad_norm": 2.7547916567496937,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 21496
+    },
+    {
+      "epoch": 0.5116871370084738,
+      "grad_norm": 0.8213775008760543,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 21497
+    },
+    {
+      "epoch": 0.5117109397315053,
+      "grad_norm": 5.3765614650507505,
+      "learning_rate": 2e-05,
+      "loss": 0.5409,
+      "step": 21498
+    },
+    {
+      "epoch": 0.5117347424545368,
+      "grad_norm": 6.287788277236454,
+      "learning_rate": 2e-05,
+      "loss": 0.8292,
+      "step": 21499
+    },
+    {
+      "epoch": 0.5117585451775684,
+      "grad_norm": 3.1384229182372176,
+      "learning_rate": 2e-05,
+      "loss": 0.468,
+      "step": 21500
+    },
+    {
+      "epoch": 0.5117823479005998,
+      "grad_norm": 3.9754847204333457,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 21501
+    },
+    {
+      "epoch": 0.5118061506236313,
+      "grad_norm": 61.00415629845939,
+      "learning_rate": 2e-05,
+      "loss": 2.0775,
+      "step": 21502
+    },
+    {
+      "epoch": 0.5118299533466628,
+      "grad_norm": 12.364706966855346,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 21503
+    },
+    {
+      "epoch": 0.5118537560696944,
+      "grad_norm": 49.03989163844711,
+      "learning_rate": 2e-05,
+      "loss": 0.9677,
+      "step": 21504
+    },
+    {
+      "epoch": 0.5118775587927259,
+      "grad_norm": 17.243362981258343,
+      "learning_rate": 2e-05,
+      "loss": 0.6863,
+      "step": 21505
+    },
+    {
+      "epoch": 0.5119013615157574,
+      "grad_norm": 8.220643218841284,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 21506
+    },
+    {
+      "epoch": 0.5119251642387889,
+      "grad_norm": 10.866609637157557,
+      "learning_rate": 2e-05,
+      "loss": 1.1299,
+      "step": 21507
+    },
+    {
+      "epoch": 0.5119489669618205,
+      "grad_norm": 12.088252831108083,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 21508
+    },
+    {
+      "epoch": 0.511972769684852,
+      "grad_norm": 4.699262320778538,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 21509
+    },
+    {
+      "epoch": 0.5119965724078834,
+      "grad_norm": 5.63712715033628,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 21510
+    },
+    {
+      "epoch": 0.512020375130915,
+      "grad_norm": 3.1832745909296793,
+      "learning_rate": 2e-05,
+      "loss": 0.5255,
+      "step": 21511
+    },
+    {
+      "epoch": 0.5120441778539465,
+      "grad_norm": 7.3272320560769115,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 21512
+    },
+    {
+      "epoch": 0.512067980576978,
+      "grad_norm": 3.2398682384651227,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 21513
+    },
+    {
+      "epoch": 0.5120917833000095,
+      "grad_norm": 2.4765093472779585,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 21514
+    },
+    {
+      "epoch": 0.5121155860230411,
+      "grad_norm": 348.62103059568824,
+      "learning_rate": 2e-05,
+      "loss": 0.764,
+      "step": 21515
+    },
+    {
+      "epoch": 0.5121393887460726,
+      "grad_norm": 9.37737277849738,
+      "learning_rate": 2e-05,
+      "loss": 0.9176,
+      "step": 21516
+    },
+    {
+      "epoch": 0.512163191469104,
+      "grad_norm": 7.049844960930883,
+      "learning_rate": 2e-05,
+      "loss": 0.3192,
+      "step": 21517
+    },
+    {
+      "epoch": 0.5121869941921355,
+      "grad_norm": 2.704108297601047,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 21518
+    },
+    {
+      "epoch": 0.5122107969151671,
+      "grad_norm": 2.5107885829155663,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 21519
+    },
+    {
+      "epoch": 0.5122345996381986,
+      "grad_norm": 3.1565407039028104,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 21520
+    },
+    {
+      "epoch": 0.5122584023612301,
+      "grad_norm": 5.781165250166239,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 21521
+    },
+    {
+      "epoch": 0.5122822050842616,
+      "grad_norm": 3.740857815039618,
+      "learning_rate": 2e-05,
+      "loss": 0.4536,
+      "step": 21522
+    },
+    {
+      "epoch": 0.5123060078072932,
+      "grad_norm": 4.889455939374436,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 21523
+    },
+    {
+      "epoch": 0.5123298105303247,
+      "grad_norm": 10.15256766161682,
+      "learning_rate": 2e-05,
+      "loss": 1.2627,
+      "step": 21524
+    },
+    {
+      "epoch": 0.5123536132533562,
+      "grad_norm": 6.814329341525046,
+      "learning_rate": 2e-05,
+      "loss": 0.8195,
+      "step": 21525
+    },
+    {
+      "epoch": 0.5123774159763878,
+      "grad_norm": 0.0711599836122045,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 21526
+    },
+    {
+      "epoch": 0.5124012186994192,
+      "grad_norm": 2.2993956301467975,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 21527
+    },
+    {
+      "epoch": 0.5124250214224507,
+      "grad_norm": 6.7071246753811105,
+      "learning_rate": 2e-05,
+      "loss": 1.0636,
+      "step": 21528
+    },
+    {
+      "epoch": 0.5124488241454822,
+      "grad_norm": 5.338241943279653,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 21529
+    },
+    {
+      "epoch": 0.5124726268685138,
+      "grad_norm": 3.2179483009838017,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 21530
+    },
+    {
+      "epoch": 0.5124964295915453,
+      "grad_norm": 0.9740395259880715,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 21531
+    },
+    {
+      "epoch": 0.5125202323145768,
+      "grad_norm": 5.641641687312586,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 21532
+    },
+    {
+      "epoch": 0.5125440350376083,
+      "grad_norm": 1.649627437475571,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 21533
+    },
+    {
+      "epoch": 0.5125678377606399,
+      "grad_norm": 5.612213435690779,
+      "learning_rate": 2e-05,
+      "loss": 1.1865,
+      "step": 21534
+    },
+    {
+      "epoch": 0.5125916404836713,
+      "grad_norm": 2.6855542200196396,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 21535
+    },
+    {
+      "epoch": 0.5126154432067028,
+      "grad_norm": 1.9932036283385708,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 21536
+    },
+    {
+      "epoch": 0.5126392459297343,
+      "grad_norm": 0.3883799390961276,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 21537
+    },
+    {
+      "epoch": 0.5126630486527659,
+      "grad_norm": 7.0913477112929995,
+      "learning_rate": 2e-05,
+      "loss": 0.5193,
+      "step": 21538
+    },
+    {
+      "epoch": 0.5126868513757974,
+      "grad_norm": 3.5929288862712645,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 21539
+    },
+    {
+      "epoch": 0.5127106540988289,
+      "grad_norm": 1.2028187512394748,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 21540
+    },
+    {
+      "epoch": 0.5127344568218605,
+      "grad_norm": 1.945350850414797,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 21541
+    },
+    {
+      "epoch": 0.512758259544892,
+      "grad_norm": 0.8844238326149413,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 21542
+    },
+    {
+      "epoch": 0.5127820622679234,
+      "grad_norm": 6.1688703050241065,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 21543
+    },
+    {
+      "epoch": 0.5128058649909549,
+      "grad_norm": 6.172662125579248,
+      "learning_rate": 2e-05,
+      "loss": 0.4341,
+      "step": 21544
+    },
+    {
+      "epoch": 0.5128296677139865,
+      "grad_norm": 7.432334333089767,
+      "learning_rate": 2e-05,
+      "loss": 1.1644,
+      "step": 21545
+    },
+    {
+      "epoch": 0.512853470437018,
+      "grad_norm": 3.996028544600571,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 21546
+    },
+    {
+      "epoch": 0.5128772731600495,
+      "grad_norm": 4.186685357287518,
+      "learning_rate": 2e-05,
+      "loss": 0.5638,
+      "step": 21547
+    },
+    {
+      "epoch": 0.512901075883081,
+      "grad_norm": 0.15164829293084897,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 21548
+    },
+    {
+      "epoch": 0.5129248786061126,
+      "grad_norm": 4.181439349516296,
+      "learning_rate": 2e-05,
+      "loss": 0.1053,
+      "step": 21549
+    },
+    {
+      "epoch": 0.5129486813291441,
+      "grad_norm": 13.180924178834957,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 21550
+    },
+    {
+      "epoch": 0.5129724840521755,
+      "grad_norm": 2.923732235324443,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 21551
+    },
+    {
+      "epoch": 0.512996286775207,
+      "grad_norm": 5.125529491812458,
+      "learning_rate": 2e-05,
+      "loss": 0.8665,
+      "step": 21552
+    },
+    {
+      "epoch": 0.5130200894982386,
+      "grad_norm": 3.7344951472209416,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 21553
+    },
+    {
+      "epoch": 0.5130438922212701,
+      "grad_norm": 2.563513739816819,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 21554
+    },
+    {
+      "epoch": 0.5130676949443016,
+      "grad_norm": 3.067193007572645,
+      "learning_rate": 2e-05,
+      "loss": 0.5426,
+      "step": 21555
+    },
+    {
+      "epoch": 0.5130914976673332,
+      "grad_norm": 3.205677922688305,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 21556
+    },
+    {
+      "epoch": 0.5131153003903647,
+      "grad_norm": 2.250535827992069,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 21557
+    },
+    {
+      "epoch": 0.5131391031133962,
+      "grad_norm": 2.276283180848532,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 21558
+    },
+    {
+      "epoch": 0.5131629058364277,
+      "grad_norm": 0.2916274184047028,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 21559
+    },
+    {
+      "epoch": 0.5131867085594592,
+      "grad_norm": 11.01669007909931,
+      "learning_rate": 2e-05,
+      "loss": 1.4334,
+      "step": 21560
+    },
+    {
+      "epoch": 0.5132105112824907,
+      "grad_norm": 3.5242117855036144,
+      "learning_rate": 2e-05,
+      "loss": 0.3812,
+      "step": 21561
+    },
+    {
+      "epoch": 0.5132343140055222,
+      "grad_norm": 3.8496133099760383,
+      "learning_rate": 2e-05,
+      "loss": 0.1422,
+      "step": 21562
+    },
+    {
+      "epoch": 0.5132581167285537,
+      "grad_norm": 12.82430579779803,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 21563
+    },
+    {
+      "epoch": 0.5132819194515853,
+      "grad_norm": 2.59693001108304,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 21564
+    },
+    {
+      "epoch": 0.5133057221746168,
+      "grad_norm": 1.693187446958841,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 21565
+    },
+    {
+      "epoch": 0.5133295248976483,
+      "grad_norm": 6.525348844871291,
+      "learning_rate": 2e-05,
+      "loss": 0.3843,
+      "step": 21566
+    },
+    {
+      "epoch": 0.5133533276206798,
+      "grad_norm": 6.342722174066006,
+      "learning_rate": 2e-05,
+      "loss": 1.3984,
+      "step": 21567
+    },
+    {
+      "epoch": 0.5133771303437114,
+      "grad_norm": 1.3368121862221916,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 21568
+    },
+    {
+      "epoch": 0.5134009330667428,
+      "grad_norm": 1.1170231311088996,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 21569
+    },
+    {
+      "epoch": 0.5134247357897743,
+      "grad_norm": 0.4808212065946725,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 21570
+    },
+    {
+      "epoch": 0.5134485385128059,
+      "grad_norm": 1.023141377570121,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 21571
+    },
+    {
+      "epoch": 0.5134723412358374,
+      "grad_norm": 5.936488615834083,
+      "learning_rate": 2e-05,
+      "loss": 1.3381,
+      "step": 21572
+    },
+    {
+      "epoch": 0.5134961439588689,
+      "grad_norm": 1.9700975208876161,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 21573
+    },
+    {
+      "epoch": 0.5135199466819004,
+      "grad_norm": 0.712143268575965,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21574
+    },
+    {
+      "epoch": 0.513543749404932,
+      "grad_norm": 5.049123438981985,
+      "learning_rate": 2e-05,
+      "loss": 0.269,
+      "step": 21575
+    },
+    {
+      "epoch": 0.5135675521279635,
+      "grad_norm": 5.006957399517753,
+      "learning_rate": 2e-05,
+      "loss": 0.6439,
+      "step": 21576
+    },
+    {
+      "epoch": 0.5135913548509949,
+      "grad_norm": 4.357051176475256,
+      "learning_rate": 2e-05,
+      "loss": 0.7718,
+      "step": 21577
+    },
+    {
+      "epoch": 0.5136151575740264,
+      "grad_norm": 5.09337378592189,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 21578
+    },
+    {
+      "epoch": 0.513638960297058,
+      "grad_norm": 6.387385563176448,
+      "learning_rate": 2e-05,
+      "loss": 0.5279,
+      "step": 21579
+    },
+    {
+      "epoch": 0.5136627630200895,
+      "grad_norm": 0.6411797334578205,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 21580
+    },
+    {
+      "epoch": 0.513686565743121,
+      "grad_norm": 4.438930913573792,
+      "learning_rate": 2e-05,
+      "loss": 0.1436,
+      "step": 21581
+    },
+    {
+      "epoch": 0.5137103684661525,
+      "grad_norm": 2.495465488110383,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 21582
+    },
+    {
+      "epoch": 0.5137341711891841,
+      "grad_norm": 1.7341743661736446,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 21583
+    },
+    {
+      "epoch": 0.5137579739122156,
+      "grad_norm": 4.524072953775762,
+      "learning_rate": 2e-05,
+      "loss": 0.4843,
+      "step": 21584
+    },
+    {
+      "epoch": 0.513781776635247,
+      "grad_norm": 0.9230297813813816,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 21585
+    },
+    {
+      "epoch": 0.5138055793582786,
+      "grad_norm": 0.17310029329650786,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 21586
+    },
+    {
+      "epoch": 0.5138293820813101,
+      "grad_norm": 0.2861216515816443,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21587
+    },
+    {
+      "epoch": 0.5138531848043416,
+      "grad_norm": 6.326462451922838,
+      "learning_rate": 2e-05,
+      "loss": 1.3047,
+      "step": 21588
+    },
+    {
+      "epoch": 0.5138769875273731,
+      "grad_norm": 2.6859006673433736,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 21589
+    },
+    {
+      "epoch": 0.5139007902504047,
+      "grad_norm": 8.95369895853364,
+      "learning_rate": 2e-05,
+      "loss": 0.6194,
+      "step": 21590
+    },
+    {
+      "epoch": 0.5139245929734362,
+      "grad_norm": 0.2372777715649399,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 21591
+    },
+    {
+      "epoch": 0.5139483956964677,
+      "grad_norm": 2.6978400584426474,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 21592
+    },
+    {
+      "epoch": 0.5139721984194991,
+      "grad_norm": 0.13289464974174606,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 21593
+    },
+    {
+      "epoch": 0.5139960011425307,
+      "grad_norm": 4.281784047589075,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 21594
+    },
+    {
+      "epoch": 0.5140198038655622,
+      "grad_norm": 5.1855588996531825,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 21595
+    },
+    {
+      "epoch": 0.5140436065885937,
+      "grad_norm": 9.950828255208487,
+      "learning_rate": 2e-05,
+      "loss": 3.1216,
+      "step": 21596
+    },
+    {
+      "epoch": 0.5140674093116252,
+      "grad_norm": 4.520138709646828,
+      "learning_rate": 2e-05,
+      "loss": 0.7061,
+      "step": 21597
+    },
+    {
+      "epoch": 0.5140912120346568,
+      "grad_norm": 0.5370034109712903,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21598
+    },
+    {
+      "epoch": 0.5141150147576883,
+      "grad_norm": 5.019657596123988,
+      "learning_rate": 2e-05,
+      "loss": 0.9617,
+      "step": 21599
+    },
+    {
+      "epoch": 0.5141388174807198,
+      "grad_norm": 0.9331233661731819,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 21600
+    },
+    {
+      "epoch": 0.5141626202037514,
+      "grad_norm": 5.350295044766193,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 21601
+    },
+    {
+      "epoch": 0.5141864229267828,
+      "grad_norm": 3.214816550140302,
+      "learning_rate": 2e-05,
+      "loss": 0.5639,
+      "step": 21602
+    },
+    {
+      "epoch": 0.5142102256498143,
+      "grad_norm": 21.801079482590644,
+      "learning_rate": 2e-05,
+      "loss": 0.3674,
+      "step": 21603
+    },
+    {
+      "epoch": 0.5142340283728458,
+      "grad_norm": 0.1664416079493497,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 21604
+    },
+    {
+      "epoch": 0.5142578310958774,
+      "grad_norm": 0.4029628960350291,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 21605
+    },
+    {
+      "epoch": 0.5142816338189089,
+      "grad_norm": 5.376187968094118,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 21606
+    },
+    {
+      "epoch": 0.5143054365419404,
+      "grad_norm": 1.6459819274942467,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 21607
+    },
+    {
+      "epoch": 0.5143292392649719,
+      "grad_norm": 1.8253107534421797,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 21608
+    },
+    {
+      "epoch": 0.5143530419880035,
+      "grad_norm": 1.4084865682514531,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 21609
+    },
+    {
+      "epoch": 0.514376844711035,
+      "grad_norm": 7.197453500342068,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 21610
+    },
+    {
+      "epoch": 0.5144006474340664,
+      "grad_norm": 7.442535909012211,
+      "learning_rate": 2e-05,
+      "loss": 0.1915,
+      "step": 21611
+    },
+    {
+      "epoch": 0.5144244501570979,
+      "grad_norm": 0.7366620125679457,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 21612
+    },
+    {
+      "epoch": 0.5144482528801295,
+      "grad_norm": 4.279010441281429,
+      "learning_rate": 2e-05,
+      "loss": 0.37,
+      "step": 21613
+    },
+    {
+      "epoch": 0.514472055603161,
+      "grad_norm": 0.040372466562130914,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 21614
+    },
+    {
+      "epoch": 0.5144958583261925,
+      "grad_norm": 13.543150983598915,
+      "learning_rate": 2e-05,
+      "loss": 0.4633,
+      "step": 21615
+    },
+    {
+      "epoch": 0.5145196610492241,
+      "grad_norm": 3.2562802224451275,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 21616
+    },
+    {
+      "epoch": 0.5145434637722556,
+      "grad_norm": 9.632442065171723,
+      "learning_rate": 2e-05,
+      "loss": 0.763,
+      "step": 21617
+    },
+    {
+      "epoch": 0.5145672664952871,
+      "grad_norm": 0.8487870952512698,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 21618
+    },
+    {
+      "epoch": 0.5145910692183185,
+      "grad_norm": 2.419565502843091,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 21619
+    },
+    {
+      "epoch": 0.5146148719413501,
+      "grad_norm": 0.460352964952727,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 21620
+    },
+    {
+      "epoch": 0.5146386746643816,
+      "grad_norm": 1.6784153944703317,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 21621
+    },
+    {
+      "epoch": 0.5146624773874131,
+      "grad_norm": 9.714342553403863,
+      "learning_rate": 2e-05,
+      "loss": 0.4677,
+      "step": 21622
+    },
+    {
+      "epoch": 0.5146862801104446,
+      "grad_norm": 0.9378478269679141,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 21623
+    },
+    {
+      "epoch": 0.5147100828334762,
+      "grad_norm": 2.718631597618821,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 21624
+    },
+    {
+      "epoch": 0.5147338855565077,
+      "grad_norm": 4.50206483457503,
+      "learning_rate": 2e-05,
+      "loss": 0.738,
+      "step": 21625
+    },
+    {
+      "epoch": 0.5147576882795392,
+      "grad_norm": 2.149159725925803,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 21626
+    },
+    {
+      "epoch": 0.5147814910025706,
+      "grad_norm": 11.565769749815493,
+      "learning_rate": 2e-05,
+      "loss": 1.8947,
+      "step": 21627
+    },
+    {
+      "epoch": 0.5148052937256022,
+      "grad_norm": 0.7670449616673338,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 21628
+    },
+    {
+      "epoch": 0.5148290964486337,
+      "grad_norm": 4.083739150840218,
+      "learning_rate": 2e-05,
+      "loss": 0.7889,
+      "step": 21629
+    },
+    {
+      "epoch": 0.5148528991716652,
+      "grad_norm": 1.8473927477694945,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 21630
+    },
+    {
+      "epoch": 0.5148767018946968,
+      "grad_norm": 2.3867032363343115,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 21631
+    },
+    {
+      "epoch": 0.5149005046177283,
+      "grad_norm": 0.8197485406091307,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 21632
+    },
+    {
+      "epoch": 0.5149243073407598,
+      "grad_norm": 4.702377773095548,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 21633
+    },
+    {
+      "epoch": 0.5149481100637913,
+      "grad_norm": 3.630072502272982,
+      "learning_rate": 2e-05,
+      "loss": 0.4674,
+      "step": 21634
+    },
+    {
+      "epoch": 0.5149719127868229,
+      "grad_norm": 12.567309301620956,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 21635
+    },
+    {
+      "epoch": 0.5149957155098543,
+      "grad_norm": 9.211984168765415,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 21636
+    },
+    {
+      "epoch": 0.5150195182328858,
+      "grad_norm": 2.8345759822768786,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 21637
+    },
+    {
+      "epoch": 0.5150433209559173,
+      "grad_norm": 4.815964071230896,
+      "learning_rate": 2e-05,
+      "loss": 0.8568,
+      "step": 21638
+    },
+    {
+      "epoch": 0.5150671236789489,
+      "grad_norm": 3.8451060126701364,
+      "learning_rate": 2e-05,
+      "loss": 0.4459,
+      "step": 21639
+    },
+    {
+      "epoch": 0.5150909264019804,
+      "grad_norm": 1.9802595642698686,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 21640
+    },
+    {
+      "epoch": 0.5151147291250119,
+      "grad_norm": 2.246519941004949,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 21641
+    },
+    {
+      "epoch": 0.5151385318480434,
+      "grad_norm": 4.412886709985743,
+      "learning_rate": 2e-05,
+      "loss": 0.5234,
+      "step": 21642
+    },
+    {
+      "epoch": 0.515162334571075,
+      "grad_norm": 60.18736383567584,
+      "learning_rate": 2e-05,
+      "loss": 1.1609,
+      "step": 21643
+    },
+    {
+      "epoch": 0.5151861372941064,
+      "grad_norm": 5.112935873357863,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 21644
+    },
+    {
+      "epoch": 0.5152099400171379,
+      "grad_norm": 2.4651008676619113,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 21645
+    },
+    {
+      "epoch": 0.5152337427401694,
+      "grad_norm": 14.35845804045557,
+      "learning_rate": 2e-05,
+      "loss": 0.4058,
+      "step": 21646
+    },
+    {
+      "epoch": 0.515257545463201,
+      "grad_norm": 3.991222475877194,
+      "learning_rate": 2e-05,
+      "loss": 0.4148,
+      "step": 21647
+    },
+    {
+      "epoch": 0.5152813481862325,
+      "grad_norm": 2.018917710314479,
+      "learning_rate": 2e-05,
+      "loss": 0.1815,
+      "step": 21648
+    },
+    {
+      "epoch": 0.515305150909264,
+      "grad_norm": 8.040484929806448,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 21649
+    },
+    {
+      "epoch": 0.5153289536322956,
+      "grad_norm": 6.444977967328703,
+      "learning_rate": 2e-05,
+      "loss": 0.2347,
+      "step": 21650
+    },
+    {
+      "epoch": 0.5153527563553271,
+      "grad_norm": 1.1979181915221626,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 21651
+    },
+    {
+      "epoch": 0.5153765590783586,
+      "grad_norm": 0.43536941089646436,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21652
+    },
+    {
+      "epoch": 0.51540036180139,
+      "grad_norm": 2.4837440231092613,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 21653
+    },
+    {
+      "epoch": 0.5154241645244216,
+      "grad_norm": 4.06421758937881,
+      "learning_rate": 2e-05,
+      "loss": 0.8542,
+      "step": 21654
+    },
+    {
+      "epoch": 0.5154479672474531,
+      "grad_norm": 7.799020721501754,
+      "learning_rate": 2e-05,
+      "loss": 1.5923,
+      "step": 21655
+    },
+    {
+      "epoch": 0.5154717699704846,
+      "grad_norm": 2.9158793574662503,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 21656
+    },
+    {
+      "epoch": 0.5154955726935161,
+      "grad_norm": 0.051653221041309705,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21657
+    },
+    {
+      "epoch": 0.5155193754165477,
+      "grad_norm": 7.7076434033679515,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 21658
+    },
+    {
+      "epoch": 0.5155431781395792,
+      "grad_norm": 1.4848414654852022,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 21659
+    },
+    {
+      "epoch": 0.5155669808626107,
+      "grad_norm": 0.5441925029604979,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 21660
+    },
+    {
+      "epoch": 0.5155907835856421,
+      "grad_norm": 5.516769961389619,
+      "learning_rate": 2e-05,
+      "loss": 0.8829,
+      "step": 21661
+    },
+    {
+      "epoch": 0.5156145863086737,
+      "grad_norm": 0.22116618860108578,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 21662
+    },
+    {
+      "epoch": 0.5156383890317052,
+      "grad_norm": 0.6704410060973354,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 21663
+    },
+    {
+      "epoch": 0.5156621917547367,
+      "grad_norm": 3.836368723282657,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 21664
+    },
+    {
+      "epoch": 0.5156859944777683,
+      "grad_norm": 1.3081208354078784,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 21665
+    },
+    {
+      "epoch": 0.5157097972007998,
+      "grad_norm": 4.3616006172344415,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 21666
+    },
+    {
+      "epoch": 0.5157335999238313,
+      "grad_norm": 2.9128619987435673,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 21667
+    },
+    {
+      "epoch": 0.5157574026468628,
+      "grad_norm": 3.3361036765965015,
+      "learning_rate": 2e-05,
+      "loss": 0.5636,
+      "step": 21668
+    },
+    {
+      "epoch": 0.5157812053698944,
+      "grad_norm": 0.10180348309804127,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 21669
+    },
+    {
+      "epoch": 0.5158050080929258,
+      "grad_norm": 4.506691163960679,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 21670
+    },
+    {
+      "epoch": 0.5158288108159573,
+      "grad_norm": 6.376789920422654,
+      "learning_rate": 2e-05,
+      "loss": 0.3159,
+      "step": 21671
+    },
+    {
+      "epoch": 0.5158526135389888,
+      "grad_norm": 9.156353020951435,
+      "learning_rate": 2e-05,
+      "loss": 0.7534,
+      "step": 21672
+    },
+    {
+      "epoch": 0.5158764162620204,
+      "grad_norm": 8.16204041727873,
+      "learning_rate": 2e-05,
+      "loss": 0.9585,
+      "step": 21673
+    },
+    {
+      "epoch": 0.5159002189850519,
+      "grad_norm": 0.3648101327426456,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 21674
+    },
+    {
+      "epoch": 0.5159240217080834,
+      "grad_norm": 2.1257540478547745,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 21675
+    },
+    {
+      "epoch": 0.5159478244311149,
+      "grad_norm": 4.61616431392873,
+      "learning_rate": 2e-05,
+      "loss": 0.4759,
+      "step": 21676
+    },
+    {
+      "epoch": 0.5159716271541465,
+      "grad_norm": 4.145565369231464,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 21677
+    },
+    {
+      "epoch": 0.515995429877178,
+      "grad_norm": 4.0279987296070905,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 21678
+    },
+    {
+      "epoch": 0.5160192326002094,
+      "grad_norm": 1.8512258105437116,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 21679
+    },
+    {
+      "epoch": 0.516043035323241,
+      "grad_norm": 2.0322056585211854,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 21680
+    },
+    {
+      "epoch": 0.5160668380462725,
+      "grad_norm": 4.724336187147526,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 21681
+    },
+    {
+      "epoch": 0.516090640769304,
+      "grad_norm": 5.518222957929009,
+      "learning_rate": 2e-05,
+      "loss": 0.844,
+      "step": 21682
+    },
+    {
+      "epoch": 0.5161144434923355,
+      "grad_norm": 6.829523109590383,
+      "learning_rate": 2e-05,
+      "loss": 1.2827,
+      "step": 21683
+    },
+    {
+      "epoch": 0.5161382462153671,
+      "grad_norm": 1.7340927755849018,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 21684
+    },
+    {
+      "epoch": 0.5161620489383986,
+      "grad_norm": 11.800826635252934,
+      "learning_rate": 2e-05,
+      "loss": 0.3951,
+      "step": 21685
+    },
+    {
+      "epoch": 0.51618585166143,
+      "grad_norm": 7.033742777873087,
+      "learning_rate": 2e-05,
+      "loss": 0.4458,
+      "step": 21686
+    },
+    {
+      "epoch": 0.5162096543844615,
+      "grad_norm": 2.7671534958926483,
+      "learning_rate": 2e-05,
+      "loss": 0.6968,
+      "step": 21687
+    },
+    {
+      "epoch": 0.5162334571074931,
+      "grad_norm": 8.382210731641575,
+      "learning_rate": 2e-05,
+      "loss": 1.2157,
+      "step": 21688
+    },
+    {
+      "epoch": 0.5162572598305246,
+      "grad_norm": 115.6595151224126,
+      "learning_rate": 2e-05,
+      "loss": 0.8643,
+      "step": 21689
+    },
+    {
+      "epoch": 0.5162810625535561,
+      "grad_norm": 2.162112546318018,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 21690
+    },
+    {
+      "epoch": 0.5163048652765876,
+      "grad_norm": 6.71241363157529,
+      "learning_rate": 2e-05,
+      "loss": 0.2737,
+      "step": 21691
+    },
+    {
+      "epoch": 0.5163286679996192,
+      "grad_norm": 8.958013081804406,
+      "learning_rate": 2e-05,
+      "loss": 0.9511,
+      "step": 21692
+    },
+    {
+      "epoch": 0.5163524707226507,
+      "grad_norm": 0.6107237293580889,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 21693
+    },
+    {
+      "epoch": 0.5163762734456822,
+      "grad_norm": 1.7082571665588326,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 21694
+    },
+    {
+      "epoch": 0.5164000761687138,
+      "grad_norm": 0.24958631329591727,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 21695
+    },
+    {
+      "epoch": 0.5164238788917452,
+      "grad_norm": 2.951767252552388,
+      "learning_rate": 2e-05,
+      "loss": 0.3914,
+      "step": 21696
+    },
+    {
+      "epoch": 0.5164476816147767,
+      "grad_norm": 0.39970822512146553,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 21697
+    },
+    {
+      "epoch": 0.5164714843378082,
+      "grad_norm": 13.533962070861843,
+      "learning_rate": 2e-05,
+      "loss": 0.5251,
+      "step": 21698
+    },
+    {
+      "epoch": 0.5164952870608398,
+      "grad_norm": 1.927185189601026,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 21699
+    },
+    {
+      "epoch": 0.5165190897838713,
+      "grad_norm": 4.2320239677403135,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 21700
+    },
+    {
+      "epoch": 0.5165428925069028,
+      "grad_norm": 3.9539479541157796,
+      "learning_rate": 2e-05,
+      "loss": 0.5047,
+      "step": 21701
+    },
+    {
+      "epoch": 0.5165666952299343,
+      "grad_norm": 4.914531308523853,
+      "learning_rate": 2e-05,
+      "loss": 0.5202,
+      "step": 21702
+    },
+    {
+      "epoch": 0.5165904979529659,
+      "grad_norm": 5.849943554390457,
+      "learning_rate": 2e-05,
+      "loss": 0.3181,
+      "step": 21703
+    },
+    {
+      "epoch": 0.5166143006759973,
+      "grad_norm": 2.379047977899081,
+      "learning_rate": 2e-05,
+      "loss": 0.441,
+      "step": 21704
+    },
+    {
+      "epoch": 0.5166381033990288,
+      "grad_norm": 2.134164550111441,
+      "learning_rate": 2e-05,
+      "loss": 0.2069,
+      "step": 21705
+    },
+    {
+      "epoch": 0.5166619061220603,
+      "grad_norm": 4.421604140548027,
+      "learning_rate": 2e-05,
+      "loss": 0.9033,
+      "step": 21706
+    },
+    {
+      "epoch": 0.5166857088450919,
+      "grad_norm": 6.663489628523278,
+      "learning_rate": 2e-05,
+      "loss": 0.9471,
+      "step": 21707
+    },
+    {
+      "epoch": 0.5167095115681234,
+      "grad_norm": 1.5358614440094267,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 21708
+    },
+    {
+      "epoch": 0.5167333142911549,
+      "grad_norm": 5.6587611948567105,
+      "learning_rate": 2e-05,
+      "loss": 1.4821,
+      "step": 21709
+    },
+    {
+      "epoch": 0.5167571170141865,
+      "grad_norm": 4.635968655238487,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 21710
+    },
+    {
+      "epoch": 0.516780919737218,
+      "grad_norm": 1.1755314979191913,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 21711
+    },
+    {
+      "epoch": 0.5168047224602494,
+      "grad_norm": 9.774774942271623,
+      "learning_rate": 2e-05,
+      "loss": 1.3419,
+      "step": 21712
+    },
+    {
+      "epoch": 0.5168285251832809,
+      "grad_norm": 12.599774218098123,
+      "learning_rate": 2e-05,
+      "loss": 0.4491,
+      "step": 21713
+    },
+    {
+      "epoch": 0.5168523279063125,
+      "grad_norm": 7.7894307775588025,
+      "learning_rate": 2e-05,
+      "loss": 0.7948,
+      "step": 21714
+    },
+    {
+      "epoch": 0.516876130629344,
+      "grad_norm": 1.9643156778539368,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 21715
+    },
+    {
+      "epoch": 0.5168999333523755,
+      "grad_norm": 6.014405763310109,
+      "learning_rate": 2e-05,
+      "loss": 1.5286,
+      "step": 21716
+    },
+    {
+      "epoch": 0.516923736075407,
+      "grad_norm": 1.752073416443757,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 21717
+    },
+    {
+      "epoch": 0.5169475387984386,
+      "grad_norm": 17.62445151357789,
+      "learning_rate": 2e-05,
+      "loss": 0.4162,
+      "step": 21718
+    },
+    {
+      "epoch": 0.5169713415214701,
+      "grad_norm": 0.841136856483478,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 21719
+    },
+    {
+      "epoch": 0.5169951442445015,
+      "grad_norm": 5.3177284351762175,
+      "learning_rate": 2e-05,
+      "loss": 0.1612,
+      "step": 21720
+    },
+    {
+      "epoch": 0.517018946967533,
+      "grad_norm": 1.4681648230194486,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 21721
+    },
+    {
+      "epoch": 0.5170427496905646,
+      "grad_norm": 1.2847369086769542,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 21722
+    },
+    {
+      "epoch": 0.5170665524135961,
+      "grad_norm": 4.071498127567032,
+      "learning_rate": 2e-05,
+      "loss": 0.3788,
+      "step": 21723
+    },
+    {
+      "epoch": 0.5170903551366276,
+      "grad_norm": 0.4049918814660162,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 21724
+    },
+    {
+      "epoch": 0.5171141578596592,
+      "grad_norm": 1.299208692597322,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 21725
+    },
+    {
+      "epoch": 0.5171379605826907,
+      "grad_norm": 0.34903094609051377,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 21726
+    },
+    {
+      "epoch": 0.5171617633057222,
+      "grad_norm": 3.44491584345125,
+      "learning_rate": 2e-05,
+      "loss": 0.6078,
+      "step": 21727
+    },
+    {
+      "epoch": 0.5171855660287537,
+      "grad_norm": 11.229048388192542,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 21728
+    },
+    {
+      "epoch": 0.5172093687517852,
+      "grad_norm": 1.0925367776358998,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 21729
+    },
+    {
+      "epoch": 0.5172331714748167,
+      "grad_norm": 3.121013683503434,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 21730
+    },
+    {
+      "epoch": 0.5172569741978482,
+      "grad_norm": 5.6723695413130155,
+      "learning_rate": 2e-05,
+      "loss": 0.9664,
+      "step": 21731
+    },
+    {
+      "epoch": 0.5172807769208797,
+      "grad_norm": 2.992280820921326,
+      "learning_rate": 2e-05,
+      "loss": 0.161,
+      "step": 21732
+    },
+    {
+      "epoch": 0.5173045796439113,
+      "grad_norm": 2.368158704885155,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 21733
+    },
+    {
+      "epoch": 0.5173283823669428,
+      "grad_norm": 4.047467586960031,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 21734
+    },
+    {
+      "epoch": 0.5173521850899743,
+      "grad_norm": 5.617932532469556,
+      "learning_rate": 2e-05,
+      "loss": 0.4766,
+      "step": 21735
+    },
+    {
+      "epoch": 0.5173759878130058,
+      "grad_norm": 5.238972310367828,
+      "learning_rate": 2e-05,
+      "loss": 0.6417,
+      "step": 21736
+    },
+    {
+      "epoch": 0.5173997905360374,
+      "grad_norm": 2.501800039386019,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 21737
+    },
+    {
+      "epoch": 0.5174235932590688,
+      "grad_norm": 5.546713063556363,
+      "learning_rate": 2e-05,
+      "loss": 1.0584,
+      "step": 21738
+    },
+    {
+      "epoch": 0.5174473959821003,
+      "grad_norm": 5.703813499449048,
+      "learning_rate": 2e-05,
+      "loss": 0.872,
+      "step": 21739
+    },
+    {
+      "epoch": 0.5174711987051319,
+      "grad_norm": 5.708397803036026,
+      "learning_rate": 2e-05,
+      "loss": 0.5159,
+      "step": 21740
+    },
+    {
+      "epoch": 0.5174950014281634,
+      "grad_norm": 2.147037979100435,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 21741
+    },
+    {
+      "epoch": 0.5175188041511949,
+      "grad_norm": 4.157200374290113,
+      "learning_rate": 2e-05,
+      "loss": 0.4071,
+      "step": 21742
+    },
+    {
+      "epoch": 0.5175426068742264,
+      "grad_norm": 3.800128512613878,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 21743
+    },
+    {
+      "epoch": 0.517566409597258,
+      "grad_norm": 0.1370095095604455,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 21744
+    },
+    {
+      "epoch": 0.5175902123202895,
+      "grad_norm": 3.8288232959035597,
+      "learning_rate": 2e-05,
+      "loss": 0.5664,
+      "step": 21745
+    },
+    {
+      "epoch": 0.5176140150433209,
+      "grad_norm": 1.2806831479971825,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 21746
+    },
+    {
+      "epoch": 0.5176378177663524,
+      "grad_norm": 5.677116616324804,
+      "learning_rate": 2e-05,
+      "loss": 0.6412,
+      "step": 21747
+    },
+    {
+      "epoch": 0.517661620489384,
+      "grad_norm": 4.6393896960125725,
+      "learning_rate": 2e-05,
+      "loss": 0.6603,
+      "step": 21748
+    },
+    {
+      "epoch": 0.5176854232124155,
+      "grad_norm": 9.941645536440904,
+      "learning_rate": 2e-05,
+      "loss": 0.6475,
+      "step": 21749
+    },
+    {
+      "epoch": 0.517709225935447,
+      "grad_norm": 4.082315483203818,
+      "learning_rate": 2e-05,
+      "loss": 0.2437,
+      "step": 21750
+    },
+    {
+      "epoch": 0.5177330286584785,
+      "grad_norm": 9.578409907202772,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 21751
+    },
+    {
+      "epoch": 0.5177568313815101,
+      "grad_norm": 3.096126292468072,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 21752
+    },
+    {
+      "epoch": 0.5177806341045416,
+      "grad_norm": 7.896219265838867,
+      "learning_rate": 2e-05,
+      "loss": 0.4958,
+      "step": 21753
+    },
+    {
+      "epoch": 0.517804436827573,
+      "grad_norm": 5.577165470860975,
+      "learning_rate": 2e-05,
+      "loss": 0.4345,
+      "step": 21754
+    },
+    {
+      "epoch": 0.5178282395506046,
+      "grad_norm": 4.218092497911558,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 21755
+    },
+    {
+      "epoch": 0.5178520422736361,
+      "grad_norm": 4.642076110162351,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 21756
+    },
+    {
+      "epoch": 0.5178758449966676,
+      "grad_norm": 3.9019125041598657,
+      "learning_rate": 2e-05,
+      "loss": 0.8435,
+      "step": 21757
+    },
+    {
+      "epoch": 0.5178996477196991,
+      "grad_norm": 0.6721296130866786,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 21758
+    },
+    {
+      "epoch": 0.5179234504427307,
+      "grad_norm": 2.3612371078034333,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 21759
+    },
+    {
+      "epoch": 0.5179472531657622,
+      "grad_norm": 1.2353948636239724,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 21760
+    },
+    {
+      "epoch": 0.5179710558887937,
+      "grad_norm": 1.9870852778680184,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 21761
+    },
+    {
+      "epoch": 0.5179948586118251,
+      "grad_norm": 2.4521498333521117,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 21762
+    },
+    {
+      "epoch": 0.5180186613348567,
+      "grad_norm": 8.300298119737782,
+      "learning_rate": 2e-05,
+      "loss": 0.5264,
+      "step": 21763
+    },
+    {
+      "epoch": 0.5180424640578882,
+      "grad_norm": 0.7754948132819997,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 21764
+    },
+    {
+      "epoch": 0.5180662667809197,
+      "grad_norm": 19.377548087273183,
+      "learning_rate": 2e-05,
+      "loss": 0.3601,
+      "step": 21765
+    },
+    {
+      "epoch": 0.5180900695039512,
+      "grad_norm": 16.18273464702955,
+      "learning_rate": 2e-05,
+      "loss": 0.8833,
+      "step": 21766
+    },
+    {
+      "epoch": 0.5181138722269828,
+      "grad_norm": 5.061733120401273,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 21767
+    },
+    {
+      "epoch": 0.5181376749500143,
+      "grad_norm": 1.5778262045333322,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 21768
+    },
+    {
+      "epoch": 0.5181614776730458,
+      "grad_norm": 4.63016940106159,
+      "learning_rate": 2e-05,
+      "loss": 0.7429,
+      "step": 21769
+    },
+    {
+      "epoch": 0.5181852803960774,
+      "grad_norm": 0.16579807233793595,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 21770
+    },
+    {
+      "epoch": 0.5182090831191088,
+      "grad_norm": 4.226581640778308,
+      "learning_rate": 2e-05,
+      "loss": 0.9254,
+      "step": 21771
+    },
+    {
+      "epoch": 0.5182328858421403,
+      "grad_norm": 6.336764542498763,
+      "learning_rate": 2e-05,
+      "loss": 0.2326,
+      "step": 21772
+    },
+    {
+      "epoch": 0.5182566885651718,
+      "grad_norm": 6.089218894806958,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 21773
+    },
+    {
+      "epoch": 0.5182804912882034,
+      "grad_norm": 6.535483724104878,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 21774
+    },
+    {
+      "epoch": 0.5183042940112349,
+      "grad_norm": 0.7263690092897989,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 21775
+    },
+    {
+      "epoch": 0.5183280967342664,
+      "grad_norm": 0.10103009341888716,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 21776
+    },
+    {
+      "epoch": 0.5183518994572979,
+      "grad_norm": 0.9882726155463786,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 21777
+    },
+    {
+      "epoch": 0.5183757021803295,
+      "grad_norm": 2.0977626935363936,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 21778
+    },
+    {
+      "epoch": 0.518399504903361,
+      "grad_norm": 4.779773672790036,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 21779
+    },
+    {
+      "epoch": 0.5184233076263924,
+      "grad_norm": 2.5278779016191324,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 21780
+    },
+    {
+      "epoch": 0.5184471103494239,
+      "grad_norm": 5.169572508711934,
+      "learning_rate": 2e-05,
+      "loss": 0.819,
+      "step": 21781
+    },
+    {
+      "epoch": 0.5184709130724555,
+      "grad_norm": 17.461743569465682,
+      "learning_rate": 2e-05,
+      "loss": 0.8031,
+      "step": 21782
+    },
+    {
+      "epoch": 0.518494715795487,
+      "grad_norm": 3.7368856166093765,
+      "learning_rate": 2e-05,
+      "loss": 0.7308,
+      "step": 21783
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 1.380617665160497,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 21784
+    },
+    {
+      "epoch": 0.5185423212415501,
+      "grad_norm": 4.3802459261032425,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 21785
+    },
+    {
+      "epoch": 0.5185661239645816,
+      "grad_norm": 12.09527013367144,
+      "learning_rate": 2e-05,
+      "loss": 1.6211,
+      "step": 21786
+    },
+    {
+      "epoch": 0.5185899266876131,
+      "grad_norm": 1.5152586385928493,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 21787
+    },
+    {
+      "epoch": 0.5186137294106445,
+      "grad_norm": 10.901917996147676,
+      "learning_rate": 2e-05,
+      "loss": 0.6234,
+      "step": 21788
+    },
+    {
+      "epoch": 0.5186375321336761,
+      "grad_norm": 6.758005597553074,
+      "learning_rate": 2e-05,
+      "loss": 0.6502,
+      "step": 21789
+    },
+    {
+      "epoch": 0.5186613348567076,
+      "grad_norm": 1.744160488642934,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 21790
+    },
+    {
+      "epoch": 0.5186851375797391,
+      "grad_norm": 4.790091483822041,
+      "learning_rate": 2e-05,
+      "loss": 1.0538,
+      "step": 21791
+    },
+    {
+      "epoch": 0.5187089403027706,
+      "grad_norm": 0.02876825545663424,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 21792
+    },
+    {
+      "epoch": 0.5187327430258022,
+      "grad_norm": 1.1350877220519724,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 21793
+    },
+    {
+      "epoch": 0.5187565457488337,
+      "grad_norm": 4.496636640860572,
+      "learning_rate": 2e-05,
+      "loss": 0.6689,
+      "step": 21794
+    },
+    {
+      "epoch": 0.5187803484718652,
+      "grad_norm": 2.3322595574869807,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 21795
+    },
+    {
+      "epoch": 0.5188041511948966,
+      "grad_norm": 3.4495962338602184,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 21796
+    },
+    {
+      "epoch": 0.5188279539179282,
+      "grad_norm": 4.26692770132852,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 21797
+    },
+    {
+      "epoch": 0.5188517566409597,
+      "grad_norm": 2.52403779586865,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 21798
+    },
+    {
+      "epoch": 0.5188755593639912,
+      "grad_norm": 8.508673582892207,
+      "learning_rate": 2e-05,
+      "loss": 0.2437,
+      "step": 21799
+    },
+    {
+      "epoch": 0.5188993620870228,
+      "grad_norm": 6.997421126524512,
+      "learning_rate": 2e-05,
+      "loss": 1.3985,
+      "step": 21800
+    },
+    {
+      "epoch": 0.5189231648100543,
+      "grad_norm": 6.649028467367687,
+      "learning_rate": 2e-05,
+      "loss": 1.1114,
+      "step": 21801
+    },
+    {
+      "epoch": 0.5189469675330858,
+      "grad_norm": 6.651578802486793,
+      "learning_rate": 2e-05,
+      "loss": 0.1279,
+      "step": 21802
+    },
+    {
+      "epoch": 0.5189707702561173,
+      "grad_norm": 1.4333702559458255,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 21803
+    },
+    {
+      "epoch": 0.5189945729791489,
+      "grad_norm": 0.052626244631074716,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 21804
+    },
+    {
+      "epoch": 0.5190183757021803,
+      "grad_norm": 4.497214845456934,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 21805
+    },
+    {
+      "epoch": 0.5190421784252118,
+      "grad_norm": 4.165395597055074,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 21806
+    },
+    {
+      "epoch": 0.5190659811482433,
+      "grad_norm": 5.737547110469571,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 21807
+    },
+    {
+      "epoch": 0.5190897838712749,
+      "grad_norm": 8.693310886842433,
+      "learning_rate": 2e-05,
+      "loss": 0.3857,
+      "step": 21808
+    },
+    {
+      "epoch": 0.5191135865943064,
+      "grad_norm": 12.30013116703409,
+      "learning_rate": 2e-05,
+      "loss": 1.3348,
+      "step": 21809
+    },
+    {
+      "epoch": 0.5191373893173379,
+      "grad_norm": 3.299764138543734,
+      "learning_rate": 2e-05,
+      "loss": 0.447,
+      "step": 21810
+    },
+    {
+      "epoch": 0.5191611920403694,
+      "grad_norm": 5.287742959727389,
+      "learning_rate": 2e-05,
+      "loss": 1.2488,
+      "step": 21811
+    },
+    {
+      "epoch": 0.519184994763401,
+      "grad_norm": 0.09081508366010825,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 21812
+    },
+    {
+      "epoch": 0.5192087974864324,
+      "grad_norm": 5.902868091008257,
+      "learning_rate": 2e-05,
+      "loss": 1.3784,
+      "step": 21813
+    },
+    {
+      "epoch": 0.5192326002094639,
+      "grad_norm": 3.8649599997820556,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 21814
+    },
+    {
+      "epoch": 0.5192564029324955,
+      "grad_norm": 0.10544929404960617,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 21815
+    },
+    {
+      "epoch": 0.519280205655527,
+      "grad_norm": 7.916107469343052,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 21816
+    },
+    {
+      "epoch": 0.5193040083785585,
+      "grad_norm": 0.3920974748810095,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 21817
+    },
+    {
+      "epoch": 0.51932781110159,
+      "grad_norm": 2.4372820612289514,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 21818
+    },
+    {
+      "epoch": 0.5193516138246216,
+      "grad_norm": 2.778204017282119,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 21819
+    },
+    {
+      "epoch": 0.5193754165476531,
+      "grad_norm": 0.02574085573779035,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 21820
+    },
+    {
+      "epoch": 0.5193992192706846,
+      "grad_norm": 0.7514554654551937,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 21821
+    },
+    {
+      "epoch": 0.519423021993716,
+      "grad_norm": 2.811779182582531,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 21822
+    },
+    {
+      "epoch": 0.5194468247167476,
+      "grad_norm": 4.028632204650981,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 21823
+    },
+    {
+      "epoch": 0.5194706274397791,
+      "grad_norm": 7.861781649473051,
+      "learning_rate": 2e-05,
+      "loss": 0.812,
+      "step": 21824
+    },
+    {
+      "epoch": 0.5194944301628106,
+      "grad_norm": 0.05914279473100038,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 21825
+    },
+    {
+      "epoch": 0.5195182328858421,
+      "grad_norm": 930.8716273332752,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 21826
+    },
+    {
+      "epoch": 0.5195420356088737,
+      "grad_norm": 7.814758658231858,
+      "learning_rate": 2e-05,
+      "loss": 0.7926,
+      "step": 21827
+    },
+    {
+      "epoch": 0.5195658383319052,
+      "grad_norm": 9.159313305185483,
+      "learning_rate": 2e-05,
+      "loss": 0.6439,
+      "step": 21828
+    },
+    {
+      "epoch": 0.5195896410549367,
+      "grad_norm": 1.161891253296832,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 21829
+    },
+    {
+      "epoch": 0.5196134437779683,
+      "grad_norm": 2.6489214213388816,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 21830
+    },
+    {
+      "epoch": 0.5196372465009997,
+      "grad_norm": 1.5115897640492149,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 21831
+    },
+    {
+      "epoch": 0.5196610492240312,
+      "grad_norm": 4.03843837361355,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 21832
+    },
+    {
+      "epoch": 0.5196848519470627,
+      "grad_norm": 12.877058763048051,
+      "learning_rate": 2e-05,
+      "loss": 0.722,
+      "step": 21833
+    },
+    {
+      "epoch": 0.5197086546700943,
+      "grad_norm": 3.8190752621116197,
+      "learning_rate": 2e-05,
+      "loss": 0.4203,
+      "step": 21834
+    },
+    {
+      "epoch": 0.5197324573931258,
+      "grad_norm": 3.570794044880344,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 21835
+    },
+    {
+      "epoch": 0.5197562601161573,
+      "grad_norm": 3.74367687652261,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 21836
+    },
+    {
+      "epoch": 0.5197800628391888,
+      "grad_norm": 1.005856266917702,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 21837
+    },
+    {
+      "epoch": 0.5198038655622204,
+      "grad_norm": 3.476704964544226,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 21838
+    },
+    {
+      "epoch": 0.5198276682852518,
+      "grad_norm": 3.404598798134898,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 21839
+    },
+    {
+      "epoch": 0.5198514710082833,
+      "grad_norm": 1.2789820352482437,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 21840
+    },
+    {
+      "epoch": 0.5198752737313148,
+      "grad_norm": 2.246514116680799,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 21841
+    },
+    {
+      "epoch": 0.5198990764543464,
+      "grad_norm": 0.9171577696780798,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 21842
+    },
+    {
+      "epoch": 0.5199228791773779,
+      "grad_norm": 3.6161305678760116,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 21843
+    },
+    {
+      "epoch": 0.5199466819004094,
+      "grad_norm": 0.09389076598426516,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 21844
+    },
+    {
+      "epoch": 0.519970484623441,
+      "grad_norm": 2.5816770369159965,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 21845
+    },
+    {
+      "epoch": 0.5199942873464725,
+      "grad_norm": 6.478729432283219,
+      "learning_rate": 2e-05,
+      "loss": 1.1212,
+      "step": 21846
+    },
+    {
+      "epoch": 0.520018090069504,
+      "grad_norm": 3.720699188660147,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 21847
+    },
+    {
+      "epoch": 0.5200418927925354,
+      "grad_norm": 9.773326145491868,
+      "learning_rate": 2e-05,
+      "loss": 0.4793,
+      "step": 21848
+    },
+    {
+      "epoch": 0.520065695515567,
+      "grad_norm": 4.010134110047824,
+      "learning_rate": 2e-05,
+      "loss": 0.7381,
+      "step": 21849
+    },
+    {
+      "epoch": 0.5200894982385985,
+      "grad_norm": 0.6018426895189642,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 21850
+    },
+    {
+      "epoch": 0.52011330096163,
+      "grad_norm": 4.37296960245558,
+      "learning_rate": 2e-05,
+      "loss": 0.7692,
+      "step": 21851
+    },
+    {
+      "epoch": 0.5201371036846615,
+      "grad_norm": 6.597769575558894,
+      "learning_rate": 2e-05,
+      "loss": 0.5869,
+      "step": 21852
+    },
+    {
+      "epoch": 0.5201609064076931,
+      "grad_norm": 8.117105794302054,
+      "learning_rate": 2e-05,
+      "loss": 0.8512,
+      "step": 21853
+    },
+    {
+      "epoch": 0.5201847091307246,
+      "grad_norm": 0.18721122142606347,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 21854
+    },
+    {
+      "epoch": 0.520208511853756,
+      "grad_norm": 9.51763015168342,
+      "learning_rate": 2e-05,
+      "loss": 1.6013,
+      "step": 21855
+    },
+    {
+      "epoch": 0.5202323145767875,
+      "grad_norm": 6.155061954226813,
+      "learning_rate": 2e-05,
+      "loss": 0.3938,
+      "step": 21856
+    },
+    {
+      "epoch": 0.5202561172998191,
+      "grad_norm": 3.710204789075531,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 21857
+    },
+    {
+      "epoch": 0.5202799200228506,
+      "grad_norm": 2.9830175034171234,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 21858
+    },
+    {
+      "epoch": 0.5203037227458821,
+      "grad_norm": 6.740918056105867,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 21859
+    },
+    {
+      "epoch": 0.5203275254689137,
+      "grad_norm": 10.397745478945255,
+      "learning_rate": 2e-05,
+      "loss": 0.8651,
+      "step": 21860
+    },
+    {
+      "epoch": 0.5203513281919452,
+      "grad_norm": 4.937678666539197,
+      "learning_rate": 2e-05,
+      "loss": 1.2752,
+      "step": 21861
+    },
+    {
+      "epoch": 0.5203751309149767,
+      "grad_norm": 0.6144243609316024,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 21862
+    },
+    {
+      "epoch": 0.5203989336380082,
+      "grad_norm": 11.428322220810726,
+      "learning_rate": 2e-05,
+      "loss": 0.5915,
+      "step": 21863
+    },
+    {
+      "epoch": 0.5204227363610398,
+      "grad_norm": 10.886080067957002,
+      "learning_rate": 2e-05,
+      "loss": 0.2123,
+      "step": 21864
+    },
+    {
+      "epoch": 0.5204465390840712,
+      "grad_norm": 2.4784215529301075,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 21865
+    },
+    {
+      "epoch": 0.5204703418071027,
+      "grad_norm": 5.2773913313761085,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 21866
+    },
+    {
+      "epoch": 0.5204941445301342,
+      "grad_norm": 2.987510928516345,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 21867
+    },
+    {
+      "epoch": 0.5205179472531658,
+      "grad_norm": 5.418559359461325,
+      "learning_rate": 2e-05,
+      "loss": 0.6953,
+      "step": 21868
+    },
+    {
+      "epoch": 0.5205417499761973,
+      "grad_norm": 5.707384472993989,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 21869
+    },
+    {
+      "epoch": 0.5205655526992288,
+      "grad_norm": 7.737881027480864,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 21870
+    },
+    {
+      "epoch": 0.5205893554222603,
+      "grad_norm": 10.181168573413903,
+      "learning_rate": 2e-05,
+      "loss": 0.6401,
+      "step": 21871
+    },
+    {
+      "epoch": 0.5206131581452919,
+      "grad_norm": 7.750609626854952,
+      "learning_rate": 2e-05,
+      "loss": 0.8563,
+      "step": 21872
+    },
+    {
+      "epoch": 0.5206369608683233,
+      "grad_norm": 6.151236461217421,
+      "learning_rate": 2e-05,
+      "loss": 0.258,
+      "step": 21873
+    },
+    {
+      "epoch": 0.5206607635913548,
+      "grad_norm": 5.033767081710829,
+      "learning_rate": 2e-05,
+      "loss": 0.1756,
+      "step": 21874
+    },
+    {
+      "epoch": 0.5206845663143864,
+      "grad_norm": 0.5807538680568696,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 21875
+    },
+    {
+      "epoch": 0.5207083690374179,
+      "grad_norm": 5.094860238442237,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 21876
+    },
+    {
+      "epoch": 0.5207321717604494,
+      "grad_norm": 2.7248665054146834,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 21877
+    },
+    {
+      "epoch": 0.5207559744834809,
+      "grad_norm": 0.21483625454799785,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 21878
+    },
+    {
+      "epoch": 0.5207797772065125,
+      "grad_norm": 5.189550825518024,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 21879
+    },
+    {
+      "epoch": 0.520803579929544,
+      "grad_norm": 4.196197488115417,
+      "learning_rate": 2e-05,
+      "loss": 0.7599,
+      "step": 21880
+    },
+    {
+      "epoch": 0.5208273826525754,
+      "grad_norm": 2.1000858180431012,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 21881
+    },
+    {
+      "epoch": 0.5208511853756069,
+      "grad_norm": 2.7047443972694913,
+      "learning_rate": 2e-05,
+      "loss": 0.4113,
+      "step": 21882
+    },
+    {
+      "epoch": 0.5208749880986385,
+      "grad_norm": 4.014572325615689,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 21883
+    },
+    {
+      "epoch": 0.52089879082167,
+      "grad_norm": 3.5466411647172333,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 21884
+    },
+    {
+      "epoch": 0.5209225935447015,
+      "grad_norm": 6.388337466893629,
+      "learning_rate": 2e-05,
+      "loss": 1.5647,
+      "step": 21885
+    },
+    {
+      "epoch": 0.520946396267733,
+      "grad_norm": 0.9853491617077145,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 21886
+    },
+    {
+      "epoch": 0.5209701989907646,
+      "grad_norm": 3.053018695103096,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 21887
+    },
+    {
+      "epoch": 0.5209940017137961,
+      "grad_norm": 6.969284001674363,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 21888
+    },
+    {
+      "epoch": 0.5210178044368275,
+      "grad_norm": 3.4233718512609728,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 21889
+    },
+    {
+      "epoch": 0.521041607159859,
+      "grad_norm": 22.796326440686748,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 21890
+    },
+    {
+      "epoch": 0.5210654098828906,
+      "grad_norm": 1.4275869868156632,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 21891
+    },
+    {
+      "epoch": 0.5210892126059221,
+      "grad_norm": 0.8180502749384515,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 21892
+    },
+    {
+      "epoch": 0.5211130153289536,
+      "grad_norm": 7.024833532506996,
+      "learning_rate": 2e-05,
+      "loss": 1.0472,
+      "step": 21893
+    },
+    {
+      "epoch": 0.5211368180519852,
+      "grad_norm": 7.626240825693202,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 21894
+    },
+    {
+      "epoch": 0.5211606207750167,
+      "grad_norm": 5.845443845596322,
+      "learning_rate": 2e-05,
+      "loss": 0.2028,
+      "step": 21895
+    },
+    {
+      "epoch": 0.5211844234980482,
+      "grad_norm": 4.251001685131433,
+      "learning_rate": 2e-05,
+      "loss": 0.7176,
+      "step": 21896
+    },
+    {
+      "epoch": 0.5212082262210797,
+      "grad_norm": 3.6932788681655606,
+      "learning_rate": 2e-05,
+      "loss": 0.3484,
+      "step": 21897
+    },
+    {
+      "epoch": 0.5212320289441112,
+      "grad_norm": 2.45894323793241,
+      "learning_rate": 2e-05,
+      "loss": 0.3867,
+      "step": 21898
+    },
+    {
+      "epoch": 0.5212558316671427,
+      "grad_norm": 0.27565434669686034,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21899
+    },
+    {
+      "epoch": 0.5212796343901742,
+      "grad_norm": 6.129579348949535,
+      "learning_rate": 2e-05,
+      "loss": 1.046,
+      "step": 21900
+    },
+    {
+      "epoch": 0.5213034371132057,
+      "grad_norm": 4.413114501535922,
+      "learning_rate": 2e-05,
+      "loss": 0.4357,
+      "step": 21901
+    },
+    {
+      "epoch": 0.5213272398362373,
+      "grad_norm": 9.406166841433835,
+      "learning_rate": 2e-05,
+      "loss": 0.7246,
+      "step": 21902
+    },
+    {
+      "epoch": 0.5213510425592688,
+      "grad_norm": 2.761330079837116,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 21903
+    },
+    {
+      "epoch": 0.5213748452823003,
+      "grad_norm": 4.107985343645178,
+      "learning_rate": 2e-05,
+      "loss": 0.5897,
+      "step": 21904
+    },
+    {
+      "epoch": 0.5213986480053318,
+      "grad_norm": 9.153063576407195,
+      "learning_rate": 2e-05,
+      "loss": 0.2357,
+      "step": 21905
+    },
+    {
+      "epoch": 0.5214224507283634,
+      "grad_norm": 2.3661301527711434,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 21906
+    },
+    {
+      "epoch": 0.5214462534513948,
+      "grad_norm": 2.5459080881083995,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 21907
+    },
+    {
+      "epoch": 0.5214700561744263,
+      "grad_norm": 3.3799781654527514,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 21908
+    },
+    {
+      "epoch": 0.5214938588974579,
+      "grad_norm": 5.150649089355488,
+      "learning_rate": 2e-05,
+      "loss": 0.1922,
+      "step": 21909
+    },
+    {
+      "epoch": 0.5215176616204894,
+      "grad_norm": 3.4706837119210863,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 21910
+    },
+    {
+      "epoch": 0.5215414643435209,
+      "grad_norm": 3.73190405133485,
+      "learning_rate": 2e-05,
+      "loss": 0.5469,
+      "step": 21911
+    },
+    {
+      "epoch": 0.5215652670665524,
+      "grad_norm": 0.7695367336939358,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 21912
+    },
+    {
+      "epoch": 0.521589069789584,
+      "grad_norm": 2.506621162739235,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 21913
+    },
+    {
+      "epoch": 0.5216128725126155,
+      "grad_norm": 5.375699917347266,
+      "learning_rate": 2e-05,
+      "loss": 0.792,
+      "step": 21914
+    },
+    {
+      "epoch": 0.5216366752356469,
+      "grad_norm": 10.169342614813583,
+      "learning_rate": 2e-05,
+      "loss": 0.5823,
+      "step": 21915
+    },
+    {
+      "epoch": 0.5216604779586784,
+      "grad_norm": 3.007619844238875,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 21916
+    },
+    {
+      "epoch": 0.52168428068171,
+      "grad_norm": 1.9702817487422026,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 21917
+    },
+    {
+      "epoch": 0.5217080834047415,
+      "grad_norm": 4.408260706775931,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 21918
+    },
+    {
+      "epoch": 0.521731886127773,
+      "grad_norm": 6.29010149124135,
+      "learning_rate": 2e-05,
+      "loss": 0.977,
+      "step": 21919
+    },
+    {
+      "epoch": 0.5217556888508045,
+      "grad_norm": 2.771243185323216,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 21920
+    },
+    {
+      "epoch": 0.5217794915738361,
+      "grad_norm": 6.718469757235966,
+      "learning_rate": 2e-05,
+      "loss": 0.1473,
+      "step": 21921
+    },
+    {
+      "epoch": 0.5218032942968676,
+      "grad_norm": 7.913832408307575,
+      "learning_rate": 2e-05,
+      "loss": 1.2967,
+      "step": 21922
+    },
+    {
+      "epoch": 0.521827097019899,
+      "grad_norm": 6.401151973513115,
+      "learning_rate": 2e-05,
+      "loss": 1.6688,
+      "step": 21923
+    },
+    {
+      "epoch": 0.5218508997429306,
+      "grad_norm": 3.4663033845893034,
+      "learning_rate": 2e-05,
+      "loss": 0.2453,
+      "step": 21924
+    },
+    {
+      "epoch": 0.5218747024659621,
+      "grad_norm": 3.691428413423962,
+      "learning_rate": 2e-05,
+      "loss": 0.274,
+      "step": 21925
+    },
+    {
+      "epoch": 0.5218985051889936,
+      "grad_norm": 3.4935971332610447,
+      "learning_rate": 2e-05,
+      "loss": 0.5027,
+      "step": 21926
+    },
+    {
+      "epoch": 0.5219223079120251,
+      "grad_norm": 2.2141040582444282,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 21927
+    },
+    {
+      "epoch": 0.5219461106350567,
+      "grad_norm": 12.025569128048303,
+      "learning_rate": 2e-05,
+      "loss": 0.4626,
+      "step": 21928
+    },
+    {
+      "epoch": 0.5219699133580882,
+      "grad_norm": 0.20112769041374273,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 21929
+    },
+    {
+      "epoch": 0.5219937160811197,
+      "grad_norm": 1.6958218340352564,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 21930
+    },
+    {
+      "epoch": 0.5220175188041511,
+      "grad_norm": 4.038678245287899,
+      "learning_rate": 2e-05,
+      "loss": 0.7767,
+      "step": 21931
+    },
+    {
+      "epoch": 0.5220413215271827,
+      "grad_norm": 5.111283681761954,
+      "learning_rate": 2e-05,
+      "loss": 0.677,
+      "step": 21932
+    },
+    {
+      "epoch": 0.5220651242502142,
+      "grad_norm": 3.7258544712665826,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 21933
+    },
+    {
+      "epoch": 0.5220889269732457,
+      "grad_norm": 7.719518963408786,
+      "learning_rate": 2e-05,
+      "loss": 0.8517,
+      "step": 21934
+    },
+    {
+      "epoch": 0.5221127296962772,
+      "grad_norm": 3.9792541812921294,
+      "learning_rate": 2e-05,
+      "loss": 0.6989,
+      "step": 21935
+    },
+    {
+      "epoch": 0.5221365324193088,
+      "grad_norm": 2.0829436097687455,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 21936
+    },
+    {
+      "epoch": 0.5221603351423403,
+      "grad_norm": 3.854583912111803,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 21937
+    },
+    {
+      "epoch": 0.5221841378653718,
+      "grad_norm": 6.219603610425894,
+      "learning_rate": 2e-05,
+      "loss": 0.7255,
+      "step": 21938
+    },
+    {
+      "epoch": 0.5222079405884034,
+      "grad_norm": 6.818610961975346,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 21939
+    },
+    {
+      "epoch": 0.5222317433114348,
+      "grad_norm": 5.605214386331679,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 21940
+    },
+    {
+      "epoch": 0.5222555460344663,
+      "grad_norm": 6.981951404808238,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 21941
+    },
+    {
+      "epoch": 0.5222793487574978,
+      "grad_norm": 0.3309352407678855,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 21942
+    },
+    {
+      "epoch": 0.5223031514805294,
+      "grad_norm": 3.5043377512320357,
+      "learning_rate": 2e-05,
+      "loss": 0.1289,
+      "step": 21943
+    },
+    {
+      "epoch": 0.5223269542035609,
+      "grad_norm": 4.769940947265545,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 21944
+    },
+    {
+      "epoch": 0.5223507569265924,
+      "grad_norm": 5.165920942013939,
+      "learning_rate": 2e-05,
+      "loss": 0.6625,
+      "step": 21945
+    },
+    {
+      "epoch": 0.5223745596496239,
+      "grad_norm": 3.621722574329807,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 21946
+    },
+    {
+      "epoch": 0.5223983623726555,
+      "grad_norm": 4.583409889810742,
+      "learning_rate": 2e-05,
+      "loss": 0.641,
+      "step": 21947
+    },
+    {
+      "epoch": 0.522422165095687,
+      "grad_norm": 5.761567324966004,
+      "learning_rate": 2e-05,
+      "loss": 0.9147,
+      "step": 21948
+    },
+    {
+      "epoch": 0.5224459678187184,
+      "grad_norm": 0.95473458705868,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 21949
+    },
+    {
+      "epoch": 0.5224697705417499,
+      "grad_norm": 7.743527509647886,
+      "learning_rate": 2e-05,
+      "loss": 0.5865,
+      "step": 21950
+    },
+    {
+      "epoch": 0.5224935732647815,
+      "grad_norm": 0.5970994462828951,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 21951
+    },
+    {
+      "epoch": 0.522517375987813,
+      "grad_norm": 0.30218756678653275,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 21952
+    },
+    {
+      "epoch": 0.5225411787108445,
+      "grad_norm": 3.4845137129089196,
+      "learning_rate": 2e-05,
+      "loss": 0.5308,
+      "step": 21953
+    },
+    {
+      "epoch": 0.5225649814338761,
+      "grad_norm": 7.42050103790806,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 21954
+    },
+    {
+      "epoch": 0.5225887841569076,
+      "grad_norm": 4.747180136273152,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 21955
+    },
+    {
+      "epoch": 0.5226125868799391,
+      "grad_norm": 6.152015584351308,
+      "learning_rate": 2e-05,
+      "loss": 0.5054,
+      "step": 21956
+    },
+    {
+      "epoch": 0.5226363896029705,
+      "grad_norm": 2.5263027022875804,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 21957
+    },
+    {
+      "epoch": 0.5226601923260021,
+      "grad_norm": 1.2715405037063647,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 21958
+    },
+    {
+      "epoch": 0.5226839950490336,
+      "grad_norm": 12.09422836144831,
+      "learning_rate": 2e-05,
+      "loss": 2.1995,
+      "step": 21959
+    },
+    {
+      "epoch": 0.5227077977720651,
+      "grad_norm": 4.952167955967561,
+      "learning_rate": 2e-05,
+      "loss": 0.7517,
+      "step": 21960
+    },
+    {
+      "epoch": 0.5227316004950966,
+      "grad_norm": 0.0505781646898084,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 21961
+    },
+    {
+      "epoch": 0.5227554032181282,
+      "grad_norm": 2.563428516564539,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 21962
+    },
+    {
+      "epoch": 0.5227792059411597,
+      "grad_norm": 4.869395324512624,
+      "learning_rate": 2e-05,
+      "loss": 0.6356,
+      "step": 21963
+    },
+    {
+      "epoch": 0.5228030086641912,
+      "grad_norm": 1.3220930838855258,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 21964
+    },
+    {
+      "epoch": 0.5228268113872226,
+      "grad_norm": 4.572689378764214,
+      "learning_rate": 2e-05,
+      "loss": 0.3682,
+      "step": 21965
+    },
+    {
+      "epoch": 0.5228506141102542,
+      "grad_norm": 1.3238648944227298,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 21966
+    },
+    {
+      "epoch": 0.5228744168332857,
+      "grad_norm": 3.9055209752335185,
+      "learning_rate": 2e-05,
+      "loss": 0.3544,
+      "step": 21967
+    },
+    {
+      "epoch": 0.5228982195563172,
+      "grad_norm": 2.0000077227240656,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 21968
+    },
+    {
+      "epoch": 0.5229220222793488,
+      "grad_norm": 4.451312326840984,
+      "learning_rate": 2e-05,
+      "loss": 0.7401,
+      "step": 21969
+    },
+    {
+      "epoch": 0.5229458250023803,
+      "grad_norm": 5.3128112670584855,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 21970
+    },
+    {
+      "epoch": 0.5229696277254118,
+      "grad_norm": 4.241563969629129,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 21971
+    },
+    {
+      "epoch": 0.5229934304484433,
+      "grad_norm": 0.5407599060213719,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 21972
+    },
+    {
+      "epoch": 0.5230172331714749,
+      "grad_norm": 5.530029156328728,
+      "learning_rate": 2e-05,
+      "loss": 0.7024,
+      "step": 21973
+    },
+    {
+      "epoch": 0.5230410358945063,
+      "grad_norm": 4.009052709481623,
+      "learning_rate": 2e-05,
+      "loss": 0.5168,
+      "step": 21974
+    },
+    {
+      "epoch": 0.5230648386175378,
+      "grad_norm": 3.503749528307265,
+      "learning_rate": 2e-05,
+      "loss": 0.8722,
+      "step": 21975
+    },
+    {
+      "epoch": 0.5230886413405693,
+      "grad_norm": 4.307447850860776,
+      "learning_rate": 2e-05,
+      "loss": 0.615,
+      "step": 21976
+    },
+    {
+      "epoch": 0.5231124440636009,
+      "grad_norm": 0.10699571360872921,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 21977
+    },
+    {
+      "epoch": 0.5231362467866324,
+      "grad_norm": 1.4924064577839826,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 21978
+    },
+    {
+      "epoch": 0.5231600495096639,
+      "grad_norm": 5.52203650221948,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 21979
+    },
+    {
+      "epoch": 0.5231838522326954,
+      "grad_norm": 2.3168948223002874,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 21980
+    },
+    {
+      "epoch": 0.523207654955727,
+      "grad_norm": 5.255346368510996,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 21981
+    },
+    {
+      "epoch": 0.5232314576787584,
+      "grad_norm": 5.239016364130656,
+      "learning_rate": 2e-05,
+      "loss": 0.7002,
+      "step": 21982
+    },
+    {
+      "epoch": 0.5232552604017899,
+      "grad_norm": 8.31247378554699,
+      "learning_rate": 2e-05,
+      "loss": 0.3142,
+      "step": 21983
+    },
+    {
+      "epoch": 0.5232790631248215,
+      "grad_norm": 5.773602251472783,
+      "learning_rate": 2e-05,
+      "loss": 0.5362,
+      "step": 21984
+    },
+    {
+      "epoch": 0.523302865847853,
+      "grad_norm": 3.0607353188229784,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 21985
+    },
+    {
+      "epoch": 0.5233266685708845,
+      "grad_norm": 12.90603462916201,
+      "learning_rate": 2e-05,
+      "loss": 0.622,
+      "step": 21986
+    },
+    {
+      "epoch": 0.523350471293916,
+      "grad_norm": 2.383941612210342,
+      "learning_rate": 2e-05,
+      "loss": 0.2121,
+      "step": 21987
+    },
+    {
+      "epoch": 0.5233742740169476,
+      "grad_norm": 11.036339051242408,
+      "learning_rate": 2e-05,
+      "loss": 0.6687,
+      "step": 21988
+    },
+    {
+      "epoch": 0.5233980767399791,
+      "grad_norm": 2.693338135844133,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 21989
+    },
+    {
+      "epoch": 0.5234218794630106,
+      "grad_norm": 4.95414380864877,
+      "learning_rate": 2e-05,
+      "loss": 0.7551,
+      "step": 21990
+    },
+    {
+      "epoch": 0.523445682186042,
+      "grad_norm": 6.384840221995735,
+      "learning_rate": 2e-05,
+      "loss": 0.2461,
+      "step": 21991
+    },
+    {
+      "epoch": 0.5234694849090736,
+      "grad_norm": 5.511621845950299,
+      "learning_rate": 2e-05,
+      "loss": 0.3105,
+      "step": 21992
+    },
+    {
+      "epoch": 0.5234932876321051,
+      "grad_norm": 1.7699385830785506,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 21993
+    },
+    {
+      "epoch": 0.5235170903551366,
+      "grad_norm": 0.5664694603718456,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 21994
+    },
+    {
+      "epoch": 0.5235408930781681,
+      "grad_norm": 0.3612601101662088,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 21995
+    },
+    {
+      "epoch": 0.5235646958011997,
+      "grad_norm": 5.70373127144617,
+      "learning_rate": 2e-05,
+      "loss": 0.8944,
+      "step": 21996
+    },
+    {
+      "epoch": 0.5235884985242312,
+      "grad_norm": 8.778372290112344,
+      "learning_rate": 2e-05,
+      "loss": 0.2155,
+      "step": 21997
+    },
+    {
+      "epoch": 0.5236123012472627,
+      "grad_norm": 2.727595856264933,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 21998
+    },
+    {
+      "epoch": 0.5236361039702943,
+      "grad_norm": 0.6206211854011909,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 21999
+    },
+    {
+      "epoch": 0.5236599066933257,
+      "grad_norm": 9.299685422533585,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 22000
+    },
+    {
+      "epoch": 0.5236837094163572,
+      "grad_norm": 3.3802262180743314,
+      "learning_rate": 2e-05,
+      "loss": 0.399,
+      "step": 22001
+    },
+    {
+      "epoch": 0.5237075121393887,
+      "grad_norm": 2.5819884083040905,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 22002
+    },
+    {
+      "epoch": 0.5237313148624203,
+      "grad_norm": 3.043788588000054,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 22003
+    },
+    {
+      "epoch": 0.5237551175854518,
+      "grad_norm": 4.833154384156527,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 22004
+    },
+    {
+      "epoch": 0.5237789203084833,
+      "grad_norm": 8.959033800951682,
+      "learning_rate": 2e-05,
+      "loss": 0.2907,
+      "step": 22005
+    },
+    {
+      "epoch": 0.5238027230315148,
+      "grad_norm": 2.0960180920937783,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 22006
+    },
+    {
+      "epoch": 0.5238265257545464,
+      "grad_norm": 2.212858986279327,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 22007
+    },
+    {
+      "epoch": 0.5238503284775778,
+      "grad_norm": 3.661809726294741,
+      "learning_rate": 2e-05,
+      "loss": 0.6238,
+      "step": 22008
+    },
+    {
+      "epoch": 0.5238741312006093,
+      "grad_norm": 12.742349344851197,
+      "learning_rate": 2e-05,
+      "loss": 0.8132,
+      "step": 22009
+    },
+    {
+      "epoch": 0.5238979339236408,
+      "grad_norm": 0.05940744152697778,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 22010
+    },
+    {
+      "epoch": 0.5239217366466724,
+      "grad_norm": 10.11903072512291,
+      "learning_rate": 2e-05,
+      "loss": 0.4769,
+      "step": 22011
+    },
+    {
+      "epoch": 0.5239455393697039,
+      "grad_norm": 5.834470289843441,
+      "learning_rate": 2e-05,
+      "loss": 1.2186,
+      "step": 22012
+    },
+    {
+      "epoch": 0.5239693420927354,
+      "grad_norm": 1.737877369561946,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 22013
+    },
+    {
+      "epoch": 0.523993144815767,
+      "grad_norm": 1.3821515991804911,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 22014
+    },
+    {
+      "epoch": 0.5240169475387985,
+      "grad_norm": 7.992298592055347,
+      "learning_rate": 2e-05,
+      "loss": 0.3117,
+      "step": 22015
+    },
+    {
+      "epoch": 0.52404075026183,
+      "grad_norm": 1.7052361880423283,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 22016
+    },
+    {
+      "epoch": 0.5240645529848614,
+      "grad_norm": 1.9725299090498905,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 22017
+    },
+    {
+      "epoch": 0.524088355707893,
+      "grad_norm": 1.855416520782796,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 22018
+    },
+    {
+      "epoch": 0.5241121584309245,
+      "grad_norm": 7.186245907679274,
+      "learning_rate": 2e-05,
+      "loss": 1.7169,
+      "step": 22019
+    },
+    {
+      "epoch": 0.524135961153956,
+      "grad_norm": 3.0800767904841666,
+      "learning_rate": 2e-05,
+      "loss": 0.0903,
+      "step": 22020
+    },
+    {
+      "epoch": 0.5241597638769875,
+      "grad_norm": 3.2821665340380544,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 22021
+    },
+    {
+      "epoch": 0.5241835666000191,
+      "grad_norm": 0.19976176785765512,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22022
+    },
+    {
+      "epoch": 0.5242073693230506,
+      "grad_norm": 2.6257754590138953,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 22023
+    },
+    {
+      "epoch": 0.524231172046082,
+      "grad_norm": 1.8003935114076541,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 22024
+    },
+    {
+      "epoch": 0.5242549747691135,
+      "grad_norm": 2.0534814007278777,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 22025
+    },
+    {
+      "epoch": 0.5242787774921451,
+      "grad_norm": 3.0114963865885374,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 22026
+    },
+    {
+      "epoch": 0.5243025802151766,
+      "grad_norm": 2.152246968098403,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 22027
+    },
+    {
+      "epoch": 0.5243263829382081,
+      "grad_norm": 1.102649530234642,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 22028
+    },
+    {
+      "epoch": 0.5243501856612397,
+      "grad_norm": 10.082511645783978,
+      "learning_rate": 2e-05,
+      "loss": 1.5854,
+      "step": 22029
+    },
+    {
+      "epoch": 0.5243739883842712,
+      "grad_norm": 2.4956297851183913,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 22030
+    },
+    {
+      "epoch": 0.5243977911073027,
+      "grad_norm": 2.9656070789695863,
+      "learning_rate": 2e-05,
+      "loss": 0.5121,
+      "step": 22031
+    },
+    {
+      "epoch": 0.5244215938303342,
+      "grad_norm": 8.15966852094258,
+      "learning_rate": 2e-05,
+      "loss": 0.6005,
+      "step": 22032
+    },
+    {
+      "epoch": 0.5244453965533658,
+      "grad_norm": 2.087297492743814,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 22033
+    },
+    {
+      "epoch": 0.5244691992763972,
+      "grad_norm": 3.109422789851922,
+      "learning_rate": 2e-05,
+      "loss": 0.4021,
+      "step": 22034
+    },
+    {
+      "epoch": 0.5244930019994287,
+      "grad_norm": 3.9617973094290586,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 22035
+    },
+    {
+      "epoch": 0.5245168047224602,
+      "grad_norm": 5.484372274635376,
+      "learning_rate": 2e-05,
+      "loss": 0.1214,
+      "step": 22036
+    },
+    {
+      "epoch": 0.5245406074454918,
+      "grad_norm": 1.1413901172350465,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 22037
+    },
+    {
+      "epoch": 0.5245644101685233,
+      "grad_norm": 3.496268693180529,
+      "learning_rate": 2e-05,
+      "loss": 0.3731,
+      "step": 22038
+    },
+    {
+      "epoch": 0.5245882128915548,
+      "grad_norm": 0.3908225972275485,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22039
+    },
+    {
+      "epoch": 0.5246120156145863,
+      "grad_norm": 5.420604705666778,
+      "learning_rate": 2e-05,
+      "loss": 0.9018,
+      "step": 22040
+    },
+    {
+      "epoch": 0.5246358183376179,
+      "grad_norm": 0.34780569913238313,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 22041
+    },
+    {
+      "epoch": 0.5246596210606493,
+      "grad_norm": 3.7884067596656603,
+      "learning_rate": 2e-05,
+      "loss": 0.6275,
+      "step": 22042
+    },
+    {
+      "epoch": 0.5246834237836808,
+      "grad_norm": 10.218893533102268,
+      "learning_rate": 2e-05,
+      "loss": 0.9323,
+      "step": 22043
+    },
+    {
+      "epoch": 0.5247072265067124,
+      "grad_norm": 12.997385879104408,
+      "learning_rate": 2e-05,
+      "loss": 0.9366,
+      "step": 22044
+    },
+    {
+      "epoch": 0.5247310292297439,
+      "grad_norm": 1.0927940481225136,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 22045
+    },
+    {
+      "epoch": 0.5247548319527754,
+      "grad_norm": 4.142855252465713,
+      "learning_rate": 2e-05,
+      "loss": 0.7543,
+      "step": 22046
+    },
+    {
+      "epoch": 0.5247786346758069,
+      "grad_norm": 16.901154468298685,
+      "learning_rate": 2e-05,
+      "loss": 0.5314,
+      "step": 22047
+    },
+    {
+      "epoch": 0.5248024373988385,
+      "grad_norm": 5.0406048269428485,
+      "learning_rate": 2e-05,
+      "loss": 1.0335,
+      "step": 22048
+    },
+    {
+      "epoch": 0.52482624012187,
+      "grad_norm": 1.7685868460247711,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 22049
+    },
+    {
+      "epoch": 0.5248500428449014,
+      "grad_norm": 6.9276749940969164,
+      "learning_rate": 2e-05,
+      "loss": 1.2202,
+      "step": 22050
+    },
+    {
+      "epoch": 0.5248738455679329,
+      "grad_norm": 4.659203472345252,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 22051
+    },
+    {
+      "epoch": 0.5248976482909645,
+      "grad_norm": 7.130304264586141,
+      "learning_rate": 2e-05,
+      "loss": 0.9906,
+      "step": 22052
+    },
+    {
+      "epoch": 0.524921451013996,
+      "grad_norm": 5.419252979086506,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 22053
+    },
+    {
+      "epoch": 0.5249452537370275,
+      "grad_norm": 3.19895903904919,
+      "learning_rate": 2e-05,
+      "loss": 0.8142,
+      "step": 22054
+    },
+    {
+      "epoch": 0.524969056460059,
+      "grad_norm": 4.24686392245471,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 22055
+    },
+    {
+      "epoch": 0.5249928591830906,
+      "grad_norm": 18.038646578902796,
+      "learning_rate": 2e-05,
+      "loss": 0.214,
+      "step": 22056
+    },
+    {
+      "epoch": 0.5250166619061221,
+      "grad_norm": 6.2849068300173085,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 22057
+    },
+    {
+      "epoch": 0.5250404646291535,
+      "grad_norm": 0.557191725592947,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 22058
+    },
+    {
+      "epoch": 0.5250642673521851,
+      "grad_norm": 4.919084874107567,
+      "learning_rate": 2e-05,
+      "loss": 1.1563,
+      "step": 22059
+    },
+    {
+      "epoch": 0.5250880700752166,
+      "grad_norm": 3.2743256243689887,
+      "learning_rate": 2e-05,
+      "loss": 0.3586,
+      "step": 22060
+    },
+    {
+      "epoch": 0.5251118727982481,
+      "grad_norm": 6.99185784145576,
+      "learning_rate": 2e-05,
+      "loss": 1.5017,
+      "step": 22061
+    },
+    {
+      "epoch": 0.5251356755212796,
+      "grad_norm": 4.814342992108507,
+      "learning_rate": 2e-05,
+      "loss": 0.4862,
+      "step": 22062
+    },
+    {
+      "epoch": 0.5251594782443112,
+      "grad_norm": 4.836621208148421,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 22063
+    },
+    {
+      "epoch": 0.5251832809673427,
+      "grad_norm": 2.3240130448467737,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 22064
+    },
+    {
+      "epoch": 0.5252070836903742,
+      "grad_norm": 3.5189345013354654,
+      "learning_rate": 2e-05,
+      "loss": 0.4632,
+      "step": 22065
+    },
+    {
+      "epoch": 0.5252308864134057,
+      "grad_norm": 1.758453507715487,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 22066
+    },
+    {
+      "epoch": 0.5252546891364372,
+      "grad_norm": 6.472173737119299,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 22067
+    },
+    {
+      "epoch": 0.5252784918594687,
+      "grad_norm": 4.007845107452679,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 22068
+    },
+    {
+      "epoch": 0.5253022945825002,
+      "grad_norm": 8.322666698701918,
+      "learning_rate": 2e-05,
+      "loss": 0.888,
+      "step": 22069
+    },
+    {
+      "epoch": 0.5253260973055317,
+      "grad_norm": 8.18805094787955,
+      "learning_rate": 2e-05,
+      "loss": 0.4836,
+      "step": 22070
+    },
+    {
+      "epoch": 0.5253499000285633,
+      "grad_norm": 0.5622301043866996,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 22071
+    },
+    {
+      "epoch": 0.5253737027515948,
+      "grad_norm": 5.600307017244978,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 22072
+    },
+    {
+      "epoch": 0.5253975054746263,
+      "grad_norm": 2.5232859553566973,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 22073
+    },
+    {
+      "epoch": 0.5254213081976579,
+      "grad_norm": 4.497368789286723,
+      "learning_rate": 2e-05,
+      "loss": 0.818,
+      "step": 22074
+    },
+    {
+      "epoch": 0.5254451109206894,
+      "grad_norm": 2.3635338510528685,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 22075
+    },
+    {
+      "epoch": 0.5254689136437208,
+      "grad_norm": 0.1598729171190295,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22076
+    },
+    {
+      "epoch": 0.5254927163667523,
+      "grad_norm": 5.814698275732247,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 22077
+    },
+    {
+      "epoch": 0.5255165190897839,
+      "grad_norm": 1.9307008898585627,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 22078
+    },
+    {
+      "epoch": 0.5255403218128154,
+      "grad_norm": 3.9980619632822902,
+      "learning_rate": 2e-05,
+      "loss": 0.2247,
+      "step": 22079
+    },
+    {
+      "epoch": 0.5255641245358469,
+      "grad_norm": 3.5163654824827315,
+      "learning_rate": 2e-05,
+      "loss": 0.276,
+      "step": 22080
+    },
+    {
+      "epoch": 0.5255879272588784,
+      "grad_norm": 3.199130932751495,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 22081
+    },
+    {
+      "epoch": 0.52561172998191,
+      "grad_norm": 1.6673888349576562,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 22082
+    },
+    {
+      "epoch": 0.5256355327049415,
+      "grad_norm": 1.0956451535971252,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 22083
+    },
+    {
+      "epoch": 0.5256593354279729,
+      "grad_norm": 5.375238947928932,
+      "learning_rate": 2e-05,
+      "loss": 0.6667,
+      "step": 22084
+    },
+    {
+      "epoch": 0.5256831381510044,
+      "grad_norm": 3.246243709271314,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 22085
+    },
+    {
+      "epoch": 0.525706940874036,
+      "grad_norm": 0.18634879611226787,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 22086
+    },
+    {
+      "epoch": 0.5257307435970675,
+      "grad_norm": 5.8927734811719015,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 22087
+    },
+    {
+      "epoch": 0.525754546320099,
+      "grad_norm": 4.312311039597536,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 22088
+    },
+    {
+      "epoch": 0.5257783490431306,
+      "grad_norm": 1.9475417775845154,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 22089
+    },
+    {
+      "epoch": 0.5258021517661621,
+      "grad_norm": 1.4139007429478196,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 22090
+    },
+    {
+      "epoch": 0.5258259544891936,
+      "grad_norm": 6.153475426726369,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 22091
+    },
+    {
+      "epoch": 0.525849757212225,
+      "grad_norm": 4.8638058354797575,
+      "learning_rate": 2e-05,
+      "loss": 0.1797,
+      "step": 22092
+    },
+    {
+      "epoch": 0.5258735599352566,
+      "grad_norm": 3.368605953865852,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 22093
+    },
+    {
+      "epoch": 0.5258973626582881,
+      "grad_norm": 5.29045721491506,
+      "learning_rate": 2e-05,
+      "loss": 0.4334,
+      "step": 22094
+    },
+    {
+      "epoch": 0.5259211653813196,
+      "grad_norm": 3.3528229621834105,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 22095
+    },
+    {
+      "epoch": 0.5259449681043511,
+      "grad_norm": 7.036662397257929,
+      "learning_rate": 2e-05,
+      "loss": 0.4496,
+      "step": 22096
+    },
+    {
+      "epoch": 0.5259687708273827,
+      "grad_norm": 85.08104086584824,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 22097
+    },
+    {
+      "epoch": 0.5259925735504142,
+      "grad_norm": 1.8305798503360358,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 22098
+    },
+    {
+      "epoch": 0.5260163762734457,
+      "grad_norm": 24.720346613188752,
+      "learning_rate": 2e-05,
+      "loss": 0.3533,
+      "step": 22099
+    },
+    {
+      "epoch": 0.5260401789964771,
+      "grad_norm": 0.5536033226058408,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 22100
+    },
+    {
+      "epoch": 0.5260639817195087,
+      "grad_norm": 8.79236262697502,
+      "learning_rate": 2e-05,
+      "loss": 1.3235,
+      "step": 22101
+    },
+    {
+      "epoch": 0.5260877844425402,
+      "grad_norm": 0.8921969803600145,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 22102
+    },
+    {
+      "epoch": 0.5261115871655717,
+      "grad_norm": 4.2214875792311854,
+      "learning_rate": 2e-05,
+      "loss": 0.7999,
+      "step": 22103
+    },
+    {
+      "epoch": 0.5261353898886033,
+      "grad_norm": 0.8539752782902735,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 22104
+    },
+    {
+      "epoch": 0.5261591926116348,
+      "grad_norm": 0.2001081687488431,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 22105
+    },
+    {
+      "epoch": 0.5261829953346663,
+      "grad_norm": 3.9426624804521597,
+      "learning_rate": 2e-05,
+      "loss": 0.6144,
+      "step": 22106
+    },
+    {
+      "epoch": 0.5262067980576978,
+      "grad_norm": 1.8374216622939785,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 22107
+    },
+    {
+      "epoch": 0.5262306007807294,
+      "grad_norm": 0.9529817289374057,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 22108
+    },
+    {
+      "epoch": 0.5262544035037608,
+      "grad_norm": 5.8288868209229285,
+      "learning_rate": 2e-05,
+      "loss": 1.1709,
+      "step": 22109
+    },
+    {
+      "epoch": 0.5262782062267923,
+      "grad_norm": 5.0226808989463905,
+      "learning_rate": 2e-05,
+      "loss": 0.8323,
+      "step": 22110
+    },
+    {
+      "epoch": 0.5263020089498238,
+      "grad_norm": 4.3962859677503445,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 22111
+    },
+    {
+      "epoch": 0.5263258116728554,
+      "grad_norm": 6.035331968930307,
+      "learning_rate": 2e-05,
+      "loss": 0.8505,
+      "step": 22112
+    },
+    {
+      "epoch": 0.5263496143958869,
+      "grad_norm": 3.173601747535838,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 22113
+    },
+    {
+      "epoch": 0.5263734171189184,
+      "grad_norm": 2.9290459098399304,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 22114
+    },
+    {
+      "epoch": 0.5263972198419499,
+      "grad_norm": 2.356258195641808,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 22115
+    },
+    {
+      "epoch": 0.5264210225649815,
+      "grad_norm": 4.584493212750292,
+      "learning_rate": 2e-05,
+      "loss": 0.7794,
+      "step": 22116
+    },
+    {
+      "epoch": 0.526444825288013,
+      "grad_norm": 5.246454427648672,
+      "learning_rate": 2e-05,
+      "loss": 0.1768,
+      "step": 22117
+    },
+    {
+      "epoch": 0.5264686280110444,
+      "grad_norm": 22.207075035898786,
+      "learning_rate": 2e-05,
+      "loss": 0.833,
+      "step": 22118
+    },
+    {
+      "epoch": 0.526492430734076,
+      "grad_norm": 0.04341024288995434,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22119
+    },
+    {
+      "epoch": 0.5265162334571075,
+      "grad_norm": 0.06197426240114726,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22120
+    },
+    {
+      "epoch": 0.526540036180139,
+      "grad_norm": 3.5961931636182842,
+      "learning_rate": 2e-05,
+      "loss": 0.7327,
+      "step": 22121
+    },
+    {
+      "epoch": 0.5265638389031705,
+      "grad_norm": 4.315911498655481,
+      "learning_rate": 2e-05,
+      "loss": 0.693,
+      "step": 22122
+    },
+    {
+      "epoch": 0.5265876416262021,
+      "grad_norm": 7.238591387804731,
+      "learning_rate": 2e-05,
+      "loss": 1.2322,
+      "step": 22123
+    },
+    {
+      "epoch": 0.5266114443492336,
+      "grad_norm": 7.692035469529745,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 22124
+    },
+    {
+      "epoch": 0.5266352470722651,
+      "grad_norm": 5.143071895620533,
+      "learning_rate": 2e-05,
+      "loss": 0.3501,
+      "step": 22125
+    },
+    {
+      "epoch": 0.5266590497952965,
+      "grad_norm": 2.8861635584965697,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 22126
+    },
+    {
+      "epoch": 0.5266828525183281,
+      "grad_norm": 2.888043784567731,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 22127
+    },
+    {
+      "epoch": 0.5267066552413596,
+      "grad_norm": 5.297766082772918,
+      "learning_rate": 2e-05,
+      "loss": 0.9211,
+      "step": 22128
+    },
+    {
+      "epoch": 0.5267304579643911,
+      "grad_norm": 4.753595435707717,
+      "learning_rate": 2e-05,
+      "loss": 0.2681,
+      "step": 22129
+    },
+    {
+      "epoch": 0.5267542606874226,
+      "grad_norm": 0.47359503436947054,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 22130
+    },
+    {
+      "epoch": 0.5267780634104542,
+      "grad_norm": 122.61315055058095,
+      "learning_rate": 2e-05,
+      "loss": 0.5042,
+      "step": 22131
+    },
+    {
+      "epoch": 0.5268018661334857,
+      "grad_norm": 1.953475167552891,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 22132
+    },
+    {
+      "epoch": 0.5268256688565172,
+      "grad_norm": 2.1598830064468473,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 22133
+    },
+    {
+      "epoch": 0.5268494715795486,
+      "grad_norm": 0.833379618353338,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 22134
+    },
+    {
+      "epoch": 0.5268732743025802,
+      "grad_norm": 2.0851728913332015,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 22135
+    },
+    {
+      "epoch": 0.5268970770256117,
+      "grad_norm": 0.35823636366208456,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 22136
+    },
+    {
+      "epoch": 0.5269208797486432,
+      "grad_norm": 4.550678857650661,
+      "learning_rate": 2e-05,
+      "loss": 0.5304,
+      "step": 22137
+    },
+    {
+      "epoch": 0.5269446824716748,
+      "grad_norm": 3.39101543227882,
+      "learning_rate": 2e-05,
+      "loss": 0.1412,
+      "step": 22138
+    },
+    {
+      "epoch": 0.5269684851947063,
+      "grad_norm": 8.908723491985373,
+      "learning_rate": 2e-05,
+      "loss": 1.3196,
+      "step": 22139
+    },
+    {
+      "epoch": 0.5269922879177378,
+      "grad_norm": 2.5258840368294,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 22140
+    },
+    {
+      "epoch": 0.5270160906407693,
+      "grad_norm": 4.7325123573309815,
+      "learning_rate": 2e-05,
+      "loss": 0.5647,
+      "step": 22141
+    },
+    {
+      "epoch": 0.5270398933638009,
+      "grad_norm": 4.265041391805668,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 22142
+    },
+    {
+      "epoch": 0.5270636960868323,
+      "grad_norm": 2.538502865056734,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 22143
+    },
+    {
+      "epoch": 0.5270874988098638,
+      "grad_norm": 5.67442406929864,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 22144
+    },
+    {
+      "epoch": 0.5271113015328953,
+      "grad_norm": 4.226821545558776,
+      "learning_rate": 2e-05,
+      "loss": 0.9545,
+      "step": 22145
+    },
+    {
+      "epoch": 0.5271351042559269,
+      "grad_norm": 6.023413451183048,
+      "learning_rate": 2e-05,
+      "loss": 0.7755,
+      "step": 22146
+    },
+    {
+      "epoch": 0.5271589069789584,
+      "grad_norm": 2.561854893837606,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 22147
+    },
+    {
+      "epoch": 0.5271827097019899,
+      "grad_norm": 1.7983983335223328,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 22148
+    },
+    {
+      "epoch": 0.5272065124250214,
+      "grad_norm": 6.7659420698138355,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 22149
+    },
+    {
+      "epoch": 0.527230315148053,
+      "grad_norm": 11.581117621970979,
+      "learning_rate": 2e-05,
+      "loss": 0.7277,
+      "step": 22150
+    },
+    {
+      "epoch": 0.5272541178710844,
+      "grad_norm": 2.53424896377172,
+      "learning_rate": 2e-05,
+      "loss": 0.2683,
+      "step": 22151
+    },
+    {
+      "epoch": 0.5272779205941159,
+      "grad_norm": 2.06418839879276,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 22152
+    },
+    {
+      "epoch": 0.5273017233171475,
+      "grad_norm": 0.39571145025003,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 22153
+    },
+    {
+      "epoch": 0.527325526040179,
+      "grad_norm": 4.16822773241982,
+      "learning_rate": 2e-05,
+      "loss": 0.6914,
+      "step": 22154
+    },
+    {
+      "epoch": 0.5273493287632105,
+      "grad_norm": 3.7869035201721117,
+      "learning_rate": 2e-05,
+      "loss": 0.538,
+      "step": 22155
+    },
+    {
+      "epoch": 0.527373131486242,
+      "grad_norm": 1.6298607775738547,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 22156
+    },
+    {
+      "epoch": 0.5273969342092736,
+      "grad_norm": 2.992940789251542,
+      "learning_rate": 2e-05,
+      "loss": 0.5746,
+      "step": 22157
+    },
+    {
+      "epoch": 0.5274207369323051,
+      "grad_norm": 8.974443188357467,
+      "learning_rate": 2e-05,
+      "loss": 0.3305,
+      "step": 22158
+    },
+    {
+      "epoch": 0.5274445396553366,
+      "grad_norm": 7.523510044581184,
+      "learning_rate": 2e-05,
+      "loss": 0.7993,
+      "step": 22159
+    },
+    {
+      "epoch": 0.527468342378368,
+      "grad_norm": 5.863227655930616,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 22160
+    },
+    {
+      "epoch": 0.5274921451013996,
+      "grad_norm": 9.955626302712995,
+      "learning_rate": 2e-05,
+      "loss": 0.7923,
+      "step": 22161
+    },
+    {
+      "epoch": 0.5275159478244311,
+      "grad_norm": 6.050667069273741,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 22162
+    },
+    {
+      "epoch": 0.5275397505474626,
+      "grad_norm": 3.4468854888287788,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 22163
+    },
+    {
+      "epoch": 0.5275635532704941,
+      "grad_norm": 3.800498803369027,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 22164
+    },
+    {
+      "epoch": 0.5275873559935257,
+      "grad_norm": 4.679377755371728,
+      "learning_rate": 2e-05,
+      "loss": 0.4665,
+      "step": 22165
+    },
+    {
+      "epoch": 0.5276111587165572,
+      "grad_norm": 10.85409814332365,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 22166
+    },
+    {
+      "epoch": 0.5276349614395887,
+      "grad_norm": 11.211861850915936,
+      "learning_rate": 2e-05,
+      "loss": 1.0904,
+      "step": 22167
+    },
+    {
+      "epoch": 0.5276587641626203,
+      "grad_norm": 2.3880700434483444,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 22168
+    },
+    {
+      "epoch": 0.5276825668856517,
+      "grad_norm": 2.5518707492833292,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 22169
+    },
+    {
+      "epoch": 0.5277063696086832,
+      "grad_norm": 6.796529182619959,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 22170
+    },
+    {
+      "epoch": 0.5277301723317147,
+      "grad_norm": 1.7034111534149237,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 22171
+    },
+    {
+      "epoch": 0.5277539750547463,
+      "grad_norm": 3.9709173794705683,
+      "learning_rate": 2e-05,
+      "loss": 0.6413,
+      "step": 22172
+    },
+    {
+      "epoch": 0.5277777777777778,
+      "grad_norm": 0.3507015002365672,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22173
+    },
+    {
+      "epoch": 0.5278015805008093,
+      "grad_norm": 2.462603813073834,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 22174
+    },
+    {
+      "epoch": 0.5278253832238408,
+      "grad_norm": 2.3730641817014666,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 22175
+    },
+    {
+      "epoch": 0.5278491859468724,
+      "grad_norm": 0.49917804732166954,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 22176
+    },
+    {
+      "epoch": 0.5278729886699038,
+      "grad_norm": 1.444427379824584,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 22177
+    },
+    {
+      "epoch": 0.5278967913929353,
+      "grad_norm": 4.112035808032628,
+      "learning_rate": 2e-05,
+      "loss": 0.7166,
+      "step": 22178
+    },
+    {
+      "epoch": 0.5279205941159668,
+      "grad_norm": 4.798834523072874,
+      "learning_rate": 2e-05,
+      "loss": 0.9402,
+      "step": 22179
+    },
+    {
+      "epoch": 0.5279443968389984,
+      "grad_norm": 4.863684825314535,
+      "learning_rate": 2e-05,
+      "loss": 0.91,
+      "step": 22180
+    },
+    {
+      "epoch": 0.5279681995620299,
+      "grad_norm": 3.4751770335726424,
+      "learning_rate": 2e-05,
+      "loss": 0.7492,
+      "step": 22181
+    },
+    {
+      "epoch": 0.5279920022850614,
+      "grad_norm": 3.6661027462329616,
+      "learning_rate": 2e-05,
+      "loss": 0.3328,
+      "step": 22182
+    },
+    {
+      "epoch": 0.528015805008093,
+      "grad_norm": 2.081216287679573,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 22183
+    },
+    {
+      "epoch": 0.5280396077311245,
+      "grad_norm": 2.4439859856398014,
+      "learning_rate": 2e-05,
+      "loss": 0.214,
+      "step": 22184
+    },
+    {
+      "epoch": 0.528063410454156,
+      "grad_norm": 4.94043759111286,
+      "learning_rate": 2e-05,
+      "loss": 0.8673,
+      "step": 22185
+    },
+    {
+      "epoch": 0.5280872131771874,
+      "grad_norm": 0.5619241112542811,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 22186
+    },
+    {
+      "epoch": 0.528111015900219,
+      "grad_norm": 7.526661978736272,
+      "learning_rate": 2e-05,
+      "loss": 1.0498,
+      "step": 22187
+    },
+    {
+      "epoch": 0.5281348186232505,
+      "grad_norm": 3.9133644206987745,
+      "learning_rate": 2e-05,
+      "loss": 0.6949,
+      "step": 22188
+    },
+    {
+      "epoch": 0.528158621346282,
+      "grad_norm": 1.441153151659376,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 22189
+    },
+    {
+      "epoch": 0.5281824240693135,
+      "grad_norm": 2.9639852540003293,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 22190
+    },
+    {
+      "epoch": 0.5282062267923451,
+      "grad_norm": 4.936777318251511,
+      "learning_rate": 2e-05,
+      "loss": 0.9586,
+      "step": 22191
+    },
+    {
+      "epoch": 0.5282300295153766,
+      "grad_norm": 2.223949586921863,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 22192
+    },
+    {
+      "epoch": 0.528253832238408,
+      "grad_norm": 1.7338710824635541,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 22193
+    },
+    {
+      "epoch": 0.5282776349614395,
+      "grad_norm": 3.6117582198628018,
+      "learning_rate": 2e-05,
+      "loss": 0.5457,
+      "step": 22194
+    },
+    {
+      "epoch": 0.5283014376844711,
+      "grad_norm": 3.845664305276789,
+      "learning_rate": 2e-05,
+      "loss": 0.5068,
+      "step": 22195
+    },
+    {
+      "epoch": 0.5283252404075026,
+      "grad_norm": 4.226641087511424,
+      "learning_rate": 2e-05,
+      "loss": 0.4741,
+      "step": 22196
+    },
+    {
+      "epoch": 0.5283490431305341,
+      "grad_norm": 2.564272279759631,
+      "learning_rate": 2e-05,
+      "loss": 0.4357,
+      "step": 22197
+    },
+    {
+      "epoch": 0.5283728458535657,
+      "grad_norm": 6.2054445132307325,
+      "learning_rate": 2e-05,
+      "loss": 0.801,
+      "step": 22198
+    },
+    {
+      "epoch": 0.5283966485765972,
+      "grad_norm": 4.667079371130809,
+      "learning_rate": 2e-05,
+      "loss": 0.2591,
+      "step": 22199
+    },
+    {
+      "epoch": 0.5284204512996287,
+      "grad_norm": 4.5058659112889705,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 22200
+    },
+    {
+      "epoch": 0.5284442540226602,
+      "grad_norm": 3.371501197085644,
+      "learning_rate": 2e-05,
+      "loss": 0.5039,
+      "step": 22201
+    },
+    {
+      "epoch": 0.5284680567456917,
+      "grad_norm": 0.22379248057899806,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 22202
+    },
+    {
+      "epoch": 0.5284918594687232,
+      "grad_norm": 0.03925598666258832,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22203
+    },
+    {
+      "epoch": 0.5285156621917547,
+      "grad_norm": 1.9189963965130719,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 22204
+    },
+    {
+      "epoch": 0.5285394649147862,
+      "grad_norm": 2.6260192804039493,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 22205
+    },
+    {
+      "epoch": 0.5285632676378178,
+      "grad_norm": 6.839028914079991,
+      "learning_rate": 2e-05,
+      "loss": 0.3789,
+      "step": 22206
+    },
+    {
+      "epoch": 0.5285870703608493,
+      "grad_norm": 10.91577786161321,
+      "learning_rate": 2e-05,
+      "loss": 0.2144,
+      "step": 22207
+    },
+    {
+      "epoch": 0.5286108730838808,
+      "grad_norm": 11.859907782307557,
+      "learning_rate": 2e-05,
+      "loss": 0.8247,
+      "step": 22208
+    },
+    {
+      "epoch": 0.5286346758069123,
+      "grad_norm": 1.1909694753853408,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 22209
+    },
+    {
+      "epoch": 0.5286584785299439,
+      "grad_norm": 3.8190563138178395,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 22210
+    },
+    {
+      "epoch": 0.5286822812529753,
+      "grad_norm": 4.344114462673495,
+      "learning_rate": 2e-05,
+      "loss": 0.4514,
+      "step": 22211
+    },
+    {
+      "epoch": 0.5287060839760068,
+      "grad_norm": 1.1274275710815498,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 22212
+    },
+    {
+      "epoch": 0.5287298866990384,
+      "grad_norm": 0.6449782582867237,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 22213
+    },
+    {
+      "epoch": 0.5287536894220699,
+      "grad_norm": 1.8027972257240221,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 22214
+    },
+    {
+      "epoch": 0.5287774921451014,
+      "grad_norm": 5.203975933526999,
+      "learning_rate": 2e-05,
+      "loss": 0.9187,
+      "step": 22215
+    },
+    {
+      "epoch": 0.5288012948681329,
+      "grad_norm": 0.12074929413263527,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22216
+    },
+    {
+      "epoch": 0.5288250975911645,
+      "grad_norm": 1.6261875921854116,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 22217
+    },
+    {
+      "epoch": 0.528848900314196,
+      "grad_norm": 12.704124806905423,
+      "learning_rate": 2e-05,
+      "loss": 0.6399,
+      "step": 22218
+    },
+    {
+      "epoch": 0.5288727030372274,
+      "grad_norm": 5.409975922957654,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 22219
+    },
+    {
+      "epoch": 0.5288965057602589,
+      "grad_norm": 0.5077168692929664,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 22220
+    },
+    {
+      "epoch": 0.5289203084832905,
+      "grad_norm": 4.467732000481689,
+      "learning_rate": 2e-05,
+      "loss": 0.235,
+      "step": 22221
+    },
+    {
+      "epoch": 0.528944111206322,
+      "grad_norm": 1.240100234402432,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 22222
+    },
+    {
+      "epoch": 0.5289679139293535,
+      "grad_norm": 1.6914743677739854,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 22223
+    },
+    {
+      "epoch": 0.528991716652385,
+      "grad_norm": 5.903442136213406,
+      "learning_rate": 2e-05,
+      "loss": 0.6252,
+      "step": 22224
+    },
+    {
+      "epoch": 0.5290155193754166,
+      "grad_norm": 3.074454470463018,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 22225
+    },
+    {
+      "epoch": 0.5290393220984481,
+      "grad_norm": 18.45950158550479,
+      "learning_rate": 2e-05,
+      "loss": 0.9822,
+      "step": 22226
+    },
+    {
+      "epoch": 0.5290631248214795,
+      "grad_norm": 1.3240205796528375,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 22227
+    },
+    {
+      "epoch": 0.5290869275445111,
+      "grad_norm": 3.868853428356964,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 22228
+    },
+    {
+      "epoch": 0.5291107302675426,
+      "grad_norm": 0.6821247023649817,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 22229
+    },
+    {
+      "epoch": 0.5291345329905741,
+      "grad_norm": 1.542608829152176,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 22230
+    },
+    {
+      "epoch": 0.5291583357136056,
+      "grad_norm": 2.732293791654915,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 22231
+    },
+    {
+      "epoch": 0.5291821384366372,
+      "grad_norm": 0.2565846380944649,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 22232
+    },
+    {
+      "epoch": 0.5292059411596687,
+      "grad_norm": 15.460333986117998,
+      "learning_rate": 2e-05,
+      "loss": 0.6117,
+      "step": 22233
+    },
+    {
+      "epoch": 0.5292297438827002,
+      "grad_norm": 6.468780004375133,
+      "learning_rate": 2e-05,
+      "loss": 1.2285,
+      "step": 22234
+    },
+    {
+      "epoch": 0.5292535466057317,
+      "grad_norm": 0.965870586812276,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 22235
+    },
+    {
+      "epoch": 0.5292773493287632,
+      "grad_norm": 2.6657537231497934,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 22236
+    },
+    {
+      "epoch": 0.5293011520517947,
+      "grad_norm": 1.1863607881082894,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 22237
+    },
+    {
+      "epoch": 0.5293249547748262,
+      "grad_norm": 3.0863147405620994,
+      "learning_rate": 2e-05,
+      "loss": 0.6235,
+      "step": 22238
+    },
+    {
+      "epoch": 0.5293487574978577,
+      "grad_norm": 6.700373037058873,
+      "learning_rate": 2e-05,
+      "loss": 1.0615,
+      "step": 22239
+    },
+    {
+      "epoch": 0.5293725602208893,
+      "grad_norm": 1.4509068538960714,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 22240
+    },
+    {
+      "epoch": 0.5293963629439208,
+      "grad_norm": 4.761749959904872,
+      "learning_rate": 2e-05,
+      "loss": 0.6503,
+      "step": 22241
+    },
+    {
+      "epoch": 0.5294201656669523,
+      "grad_norm": 5.332494320028288,
+      "learning_rate": 2e-05,
+      "loss": 0.846,
+      "step": 22242
+    },
+    {
+      "epoch": 0.5294439683899839,
+      "grad_norm": 3.5787410342627983,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 22243
+    },
+    {
+      "epoch": 0.5294677711130154,
+      "grad_norm": 2.9161185724607774,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 22244
+    },
+    {
+      "epoch": 0.5294915738360468,
+      "grad_norm": 3.531879926988475,
+      "learning_rate": 2e-05,
+      "loss": 0.5455,
+      "step": 22245
+    },
+    {
+      "epoch": 0.5295153765590783,
+      "grad_norm": 3.2203222782919743,
+      "learning_rate": 2e-05,
+      "loss": 0.3836,
+      "step": 22246
+    },
+    {
+      "epoch": 0.5295391792821099,
+      "grad_norm": 9.390497402766814,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 22247
+    },
+    {
+      "epoch": 0.5295629820051414,
+      "grad_norm": 4.954227308097684,
+      "learning_rate": 2e-05,
+      "loss": 0.6925,
+      "step": 22248
+    },
+    {
+      "epoch": 0.5295867847281729,
+      "grad_norm": 4.215290076621734,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 22249
+    },
+    {
+      "epoch": 0.5296105874512044,
+      "grad_norm": 4.988409083562582,
+      "learning_rate": 2e-05,
+      "loss": 0.2614,
+      "step": 22250
+    },
+    {
+      "epoch": 0.529634390174236,
+      "grad_norm": 2.3714687413994935,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 22251
+    },
+    {
+      "epoch": 0.5296581928972675,
+      "grad_norm": 10.592907986171173,
+      "learning_rate": 2e-05,
+      "loss": 0.3098,
+      "step": 22252
+    },
+    {
+      "epoch": 0.5296819956202989,
+      "grad_norm": 3.3478745666626235,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 22253
+    },
+    {
+      "epoch": 0.5297057983433304,
+      "grad_norm": 4.558591305497323,
+      "learning_rate": 2e-05,
+      "loss": 0.4578,
+      "step": 22254
+    },
+    {
+      "epoch": 0.529729601066362,
+      "grad_norm": 2.2690023500941785,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 22255
+    },
+    {
+      "epoch": 0.5297534037893935,
+      "grad_norm": 5.717716051359755,
+      "learning_rate": 2e-05,
+      "loss": 0.833,
+      "step": 22256
+    },
+    {
+      "epoch": 0.529777206512425,
+      "grad_norm": 9.160029357088185,
+      "learning_rate": 2e-05,
+      "loss": 0.7856,
+      "step": 22257
+    },
+    {
+      "epoch": 0.5298010092354566,
+      "grad_norm": 0.8384718208396928,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 22258
+    },
+    {
+      "epoch": 0.5298248119584881,
+      "grad_norm": 0.3983655440955478,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22259
+    },
+    {
+      "epoch": 0.5298486146815196,
+      "grad_norm": 4.669521741913824,
+      "learning_rate": 2e-05,
+      "loss": 0.3213,
+      "step": 22260
+    },
+    {
+      "epoch": 0.529872417404551,
+      "grad_norm": 13.218473279125645,
+      "learning_rate": 2e-05,
+      "loss": 0.5439,
+      "step": 22261
+    },
+    {
+      "epoch": 0.5298962201275826,
+      "grad_norm": 4.757796076974051,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 22262
+    },
+    {
+      "epoch": 0.5299200228506141,
+      "grad_norm": 0.5051558657101212,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 22263
+    },
+    {
+      "epoch": 0.5299438255736456,
+      "grad_norm": 9.191723781266958,
+      "learning_rate": 2e-05,
+      "loss": 0.2563,
+      "step": 22264
+    },
+    {
+      "epoch": 0.5299676282966771,
+      "grad_norm": 2.8995929942991507,
+      "learning_rate": 2e-05,
+      "loss": 0.2973,
+      "step": 22265
+    },
+    {
+      "epoch": 0.5299914310197087,
+      "grad_norm": 6.9343823493527434,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 22266
+    },
+    {
+      "epoch": 0.5300152337427402,
+      "grad_norm": 5.4542282822231645,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 22267
+    },
+    {
+      "epoch": 0.5300390364657717,
+      "grad_norm": 9.964736215934703,
+      "learning_rate": 2e-05,
+      "loss": 0.6995,
+      "step": 22268
+    },
+    {
+      "epoch": 0.5300628391888031,
+      "grad_norm": 8.109073344052721,
+      "learning_rate": 2e-05,
+      "loss": 0.3232,
+      "step": 22269
+    },
+    {
+      "epoch": 0.5300866419118347,
+      "grad_norm": 3.3099626729839065,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 22270
+    },
+    {
+      "epoch": 0.5301104446348662,
+      "grad_norm": 4.983788262706087,
+      "learning_rate": 2e-05,
+      "loss": 0.7076,
+      "step": 22271
+    },
+    {
+      "epoch": 0.5301342473578977,
+      "grad_norm": 1.2108556386662526,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 22272
+    },
+    {
+      "epoch": 0.5301580500809293,
+      "grad_norm": 0.812502079964457,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 22273
+    },
+    {
+      "epoch": 0.5301818528039608,
+      "grad_norm": 9.330416784502148,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 22274
+    },
+    {
+      "epoch": 0.5302056555269923,
+      "grad_norm": 4.773914975913964,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 22275
+    },
+    {
+      "epoch": 0.5302294582500238,
+      "grad_norm": 1.8379679637333692,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 22276
+    },
+    {
+      "epoch": 0.5302532609730554,
+      "grad_norm": 2.5197947978167385,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 22277
+    },
+    {
+      "epoch": 0.5302770636960868,
+      "grad_norm": 4.461915990824536,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 22278
+    },
+    {
+      "epoch": 0.5303008664191183,
+      "grad_norm": 4.237426240539204,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 22279
+    },
+    {
+      "epoch": 0.5303246691421498,
+      "grad_norm": 1.1374197333699807,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 22280
+    },
+    {
+      "epoch": 0.5303484718651814,
+      "grad_norm": 5.37733705978069,
+      "learning_rate": 2e-05,
+      "loss": 0.5369,
+      "step": 22281
+    },
+    {
+      "epoch": 0.5303722745882129,
+      "grad_norm": 4.735908398450506,
+      "learning_rate": 2e-05,
+      "loss": 0.5833,
+      "step": 22282
+    },
+    {
+      "epoch": 0.5303960773112444,
+      "grad_norm": 21.36386414083597,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 22283
+    },
+    {
+      "epoch": 0.5304198800342759,
+      "grad_norm": 4.033932349501519,
+      "learning_rate": 2e-05,
+      "loss": 0.7842,
+      "step": 22284
+    },
+    {
+      "epoch": 0.5304436827573075,
+      "grad_norm": 3.524759461126312,
+      "learning_rate": 2e-05,
+      "loss": 0.454,
+      "step": 22285
+    },
+    {
+      "epoch": 0.530467485480339,
+      "grad_norm": 4.690128687025587,
+      "learning_rate": 2e-05,
+      "loss": 0.4466,
+      "step": 22286
+    },
+    {
+      "epoch": 0.5304912882033704,
+      "grad_norm": 7.122998100476535,
+      "learning_rate": 2e-05,
+      "loss": 1.11,
+      "step": 22287
+    },
+    {
+      "epoch": 0.530515090926402,
+      "grad_norm": 0.8148539046595797,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 22288
+    },
+    {
+      "epoch": 0.5305388936494335,
+      "grad_norm": 4.381708206810407,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 22289
+    },
+    {
+      "epoch": 0.530562696372465,
+      "grad_norm": 3.8497843968339436,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 22290
+    },
+    {
+      "epoch": 0.5305864990954965,
+      "grad_norm": 6.259292975158339,
+      "learning_rate": 2e-05,
+      "loss": 0.7313,
+      "step": 22291
+    },
+    {
+      "epoch": 0.5306103018185281,
+      "grad_norm": 6.9580859519986324,
+      "learning_rate": 2e-05,
+      "loss": 0.6978,
+      "step": 22292
+    },
+    {
+      "epoch": 0.5306341045415596,
+      "grad_norm": 2.6770769262750655,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 22293
+    },
+    {
+      "epoch": 0.5306579072645911,
+      "grad_norm": 3.5898854498755584,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 22294
+    },
+    {
+      "epoch": 0.5306817099876225,
+      "grad_norm": 5.2932898222136116,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 22295
+    },
+    {
+      "epoch": 0.5307055127106541,
+      "grad_norm": 0.10903838153210173,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 22296
+    },
+    {
+      "epoch": 0.5307293154336856,
+      "grad_norm": 14.075479195840034,
+      "learning_rate": 2e-05,
+      "loss": 0.8958,
+      "step": 22297
+    },
+    {
+      "epoch": 0.5307531181567171,
+      "grad_norm": 6.022481427747966,
+      "learning_rate": 2e-05,
+      "loss": 0.3223,
+      "step": 22298
+    },
+    {
+      "epoch": 0.5307769208797486,
+      "grad_norm": 7.715991593132554,
+      "learning_rate": 2e-05,
+      "loss": 0.3985,
+      "step": 22299
+    },
+    {
+      "epoch": 0.5308007236027802,
+      "grad_norm": 0.205377104922574,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 22300
+    },
+    {
+      "epoch": 0.5308245263258117,
+      "grad_norm": 4.336108842644914,
+      "learning_rate": 2e-05,
+      "loss": 0.2418,
+      "step": 22301
+    },
+    {
+      "epoch": 0.5308483290488432,
+      "grad_norm": 3.9814130860881485,
+      "learning_rate": 2e-05,
+      "loss": 0.6078,
+      "step": 22302
+    },
+    {
+      "epoch": 0.5308721317718748,
+      "grad_norm": 0.27658766064593154,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 22303
+    },
+    {
+      "epoch": 0.5308959344949062,
+      "grad_norm": 1.9230012149619096,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 22304
+    },
+    {
+      "epoch": 0.5309197372179377,
+      "grad_norm": 2.815956755089797,
+      "learning_rate": 2e-05,
+      "loss": 0.2084,
+      "step": 22305
+    },
+    {
+      "epoch": 0.5309435399409692,
+      "grad_norm": 2.7468429258796068,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 22306
+    },
+    {
+      "epoch": 0.5309673426640008,
+      "grad_norm": 0.1303658034559078,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 22307
+    },
+    {
+      "epoch": 0.5309911453870323,
+      "grad_norm": 2.487316914501038,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 22308
+    },
+    {
+      "epoch": 0.5310149481100638,
+      "grad_norm": 4.1092632130317455,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 22309
+    },
+    {
+      "epoch": 0.5310387508330953,
+      "grad_norm": 1.555220884425503,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 22310
+    },
+    {
+      "epoch": 0.5310625535561269,
+      "grad_norm": 3.6581424913809992,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 22311
+    },
+    {
+      "epoch": 0.5310863562791583,
+      "grad_norm": 1.6198450455755977,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 22312
+    },
+    {
+      "epoch": 0.5311101590021898,
+      "grad_norm": 0.4178967107257123,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22313
+    },
+    {
+      "epoch": 0.5311339617252213,
+      "grad_norm": 0.1744663166750311,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22314
+    },
+    {
+      "epoch": 0.5311577644482529,
+      "grad_norm": 0.2202776053773964,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22315
+    },
+    {
+      "epoch": 0.5311815671712844,
+      "grad_norm": 0.4329827399000289,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 22316
+    },
+    {
+      "epoch": 0.5312053698943159,
+      "grad_norm": 7.51289898931363,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 22317
+    },
+    {
+      "epoch": 0.5312291726173475,
+      "grad_norm": 0.7178319308188359,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 22318
+    },
+    {
+      "epoch": 0.531252975340379,
+      "grad_norm": 3.808421643415294,
+      "learning_rate": 2e-05,
+      "loss": 0.4933,
+      "step": 22319
+    },
+    {
+      "epoch": 0.5312767780634104,
+      "grad_norm": 9.067752568197479,
+      "learning_rate": 2e-05,
+      "loss": 0.3803,
+      "step": 22320
+    },
+    {
+      "epoch": 0.5313005807864419,
+      "grad_norm": 11.443926622403911,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 22321
+    },
+    {
+      "epoch": 0.5313243835094735,
+      "grad_norm": 4.050193591767971,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 22322
+    },
+    {
+      "epoch": 0.531348186232505,
+      "grad_norm": 4.037634084331205,
+      "learning_rate": 2e-05,
+      "loss": 0.6614,
+      "step": 22323
+    },
+    {
+      "epoch": 0.5313719889555365,
+      "grad_norm": 3.5036119661593372,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 22324
+    },
+    {
+      "epoch": 0.531395791678568,
+      "grad_norm": 4.807602014138022,
+      "learning_rate": 2e-05,
+      "loss": 0.1813,
+      "step": 22325
+    },
+    {
+      "epoch": 0.5314195944015996,
+      "grad_norm": 11.843859618775038,
+      "learning_rate": 2e-05,
+      "loss": 0.9523,
+      "step": 22326
+    },
+    {
+      "epoch": 0.5314433971246311,
+      "grad_norm": 2.7380617892146635,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 22327
+    },
+    {
+      "epoch": 0.5314671998476626,
+      "grad_norm": 5.278396172858978,
+      "learning_rate": 2e-05,
+      "loss": 0.8885,
+      "step": 22328
+    },
+    {
+      "epoch": 0.531491002570694,
+      "grad_norm": 9.078025122227261,
+      "learning_rate": 2e-05,
+      "loss": 1.4137,
+      "step": 22329
+    },
+    {
+      "epoch": 0.5315148052937256,
+      "grad_norm": 4.63186693021318,
+      "learning_rate": 2e-05,
+      "loss": 0.7466,
+      "step": 22330
+    },
+    {
+      "epoch": 0.5315386080167571,
+      "grad_norm": 0.09394227920379458,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 22331
+    },
+    {
+      "epoch": 0.5315624107397886,
+      "grad_norm": 12.76771816301567,
+      "learning_rate": 2e-05,
+      "loss": 0.3657,
+      "step": 22332
+    },
+    {
+      "epoch": 0.5315862134628202,
+      "grad_norm": 5.3929334276810375,
+      "learning_rate": 2e-05,
+      "loss": 0.536,
+      "step": 22333
+    },
+    {
+      "epoch": 0.5316100161858517,
+      "grad_norm": 3.625007711665783,
+      "learning_rate": 2e-05,
+      "loss": 0.5819,
+      "step": 22334
+    },
+    {
+      "epoch": 0.5316338189088832,
+      "grad_norm": 3.307786800126182,
+      "learning_rate": 2e-05,
+      "loss": 0.4603,
+      "step": 22335
+    },
+    {
+      "epoch": 0.5316576216319147,
+      "grad_norm": 2.430937081197339,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 22336
+    },
+    {
+      "epoch": 0.5316814243549463,
+      "grad_norm": 3.0177099260666007,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 22337
+    },
+    {
+      "epoch": 0.5317052270779777,
+      "grad_norm": 2.319593937067713,
+      "learning_rate": 2e-05,
+      "loss": 0.3342,
+      "step": 22338
+    },
+    {
+      "epoch": 0.5317290298010092,
+      "grad_norm": 22.706922679790186,
+      "learning_rate": 2e-05,
+      "loss": 0.9858,
+      "step": 22339
+    },
+    {
+      "epoch": 0.5317528325240407,
+      "grad_norm": 3.4850593015718854,
+      "learning_rate": 2e-05,
+      "loss": 0.3792,
+      "step": 22340
+    },
+    {
+      "epoch": 0.5317766352470723,
+      "grad_norm": 1.0555317972174576,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 22341
+    },
+    {
+      "epoch": 0.5318004379701038,
+      "grad_norm": 3.6163726974539236,
+      "learning_rate": 2e-05,
+      "loss": 0.6358,
+      "step": 22342
+    },
+    {
+      "epoch": 0.5318242406931353,
+      "grad_norm": 0.5076464441434372,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 22343
+    },
+    {
+      "epoch": 0.5318480434161668,
+      "grad_norm": 7.270879876145573,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 22344
+    },
+    {
+      "epoch": 0.5318718461391984,
+      "grad_norm": 5.5442872321097,
+      "learning_rate": 2e-05,
+      "loss": 0.795,
+      "step": 22345
+    },
+    {
+      "epoch": 0.5318956488622298,
+      "grad_norm": 2.031365738970871,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 22346
+    },
+    {
+      "epoch": 0.5319194515852613,
+      "grad_norm": 6.988951429187119,
+      "learning_rate": 2e-05,
+      "loss": 1.4736,
+      "step": 22347
+    },
+    {
+      "epoch": 0.5319432543082929,
+      "grad_norm": 10.195056822863256,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 22348
+    },
+    {
+      "epoch": 0.5319670570313244,
+      "grad_norm": 0.6091630335828002,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 22349
+    },
+    {
+      "epoch": 0.5319908597543559,
+      "grad_norm": 4.253631136112538,
+      "learning_rate": 2e-05,
+      "loss": 0.4615,
+      "step": 22350
+    },
+    {
+      "epoch": 0.5320146624773874,
+      "grad_norm": 4.156792429503766,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 22351
+    },
+    {
+      "epoch": 0.532038465200419,
+      "grad_norm": 3.7547190706080578,
+      "learning_rate": 2e-05,
+      "loss": 0.3292,
+      "step": 22352
+    },
+    {
+      "epoch": 0.5320622679234505,
+      "grad_norm": 2.9367731167304387,
+      "learning_rate": 2e-05,
+      "loss": 0.4241,
+      "step": 22353
+    },
+    {
+      "epoch": 0.532086070646482,
+      "grad_norm": 3.9075162122053446,
+      "learning_rate": 2e-05,
+      "loss": 0.4881,
+      "step": 22354
+    },
+    {
+      "epoch": 0.5321098733695134,
+      "grad_norm": 4.562623744096495,
+      "learning_rate": 2e-05,
+      "loss": 0.8137,
+      "step": 22355
+    },
+    {
+      "epoch": 0.532133676092545,
+      "grad_norm": 3.1471694922090983,
+      "learning_rate": 2e-05,
+      "loss": 0.1687,
+      "step": 22356
+    },
+    {
+      "epoch": 0.5321574788155765,
+      "grad_norm": 0.4391361702605823,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22357
+    },
+    {
+      "epoch": 0.532181281538608,
+      "grad_norm": 1.8249967094281687,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 22358
+    },
+    {
+      "epoch": 0.5322050842616395,
+      "grad_norm": 12.171739404155531,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 22359
+    },
+    {
+      "epoch": 0.5322288869846711,
+      "grad_norm": 0.311268143805006,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 22360
+    },
+    {
+      "epoch": 0.5322526897077026,
+      "grad_norm": 11.75095908912998,
+      "learning_rate": 2e-05,
+      "loss": 0.5484,
+      "step": 22361
+    },
+    {
+      "epoch": 0.532276492430734,
+      "grad_norm": 0.5528480083860733,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 22362
+    },
+    {
+      "epoch": 0.5323002951537656,
+      "grad_norm": 2.1839186320000645,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 22363
+    },
+    {
+      "epoch": 0.5323240978767971,
+      "grad_norm": 6.712435719593774,
+      "learning_rate": 2e-05,
+      "loss": 1.2796,
+      "step": 22364
+    },
+    {
+      "epoch": 0.5323479005998286,
+      "grad_norm": 3.2328080346566743,
+      "learning_rate": 2e-05,
+      "loss": 0.3702,
+      "step": 22365
+    },
+    {
+      "epoch": 0.5323717033228601,
+      "grad_norm": 3.551953171573482,
+      "learning_rate": 2e-05,
+      "loss": 0.6263,
+      "step": 22366
+    },
+    {
+      "epoch": 0.5323955060458917,
+      "grad_norm": 2.682793421318363,
+      "learning_rate": 2e-05,
+      "loss": 0.5669,
+      "step": 22367
+    },
+    {
+      "epoch": 0.5324193087689232,
+      "grad_norm": 13.081713103984919,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 22368
+    },
+    {
+      "epoch": 0.5324431114919547,
+      "grad_norm": 13.503641098347314,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 22369
+    },
+    {
+      "epoch": 0.5324669142149862,
+      "grad_norm": 5.9620838495533945,
+      "learning_rate": 2e-05,
+      "loss": 1.0558,
+      "step": 22370
+    },
+    {
+      "epoch": 0.5324907169380177,
+      "grad_norm": 9.092316574493621,
+      "learning_rate": 2e-05,
+      "loss": 0.2125,
+      "step": 22371
+    },
+    {
+      "epoch": 0.5325145196610492,
+      "grad_norm": 1.5686970063183554,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 22372
+    },
+    {
+      "epoch": 0.5325383223840807,
+      "grad_norm": 5.36502256578651,
+      "learning_rate": 2e-05,
+      "loss": 1.0585,
+      "step": 22373
+    },
+    {
+      "epoch": 0.5325621251071122,
+      "grad_norm": 18.38995129261934,
+      "learning_rate": 2e-05,
+      "loss": 0.4836,
+      "step": 22374
+    },
+    {
+      "epoch": 0.5325859278301438,
+      "grad_norm": 8.662540832459515,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 22375
+    },
+    {
+      "epoch": 0.5326097305531753,
+      "grad_norm": 5.956614070733502,
+      "learning_rate": 2e-05,
+      "loss": 0.992,
+      "step": 22376
+    },
+    {
+      "epoch": 0.5326335332762068,
+      "grad_norm": 0.2398781639083214,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 22377
+    },
+    {
+      "epoch": 0.5326573359992384,
+      "grad_norm": 5.57345254467942,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 22378
+    },
+    {
+      "epoch": 0.5326811387222699,
+      "grad_norm": 1.9802520678531095,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 22379
+    },
+    {
+      "epoch": 0.5327049414453013,
+      "grad_norm": 5.367046961619925,
+      "learning_rate": 2e-05,
+      "loss": 1.1429,
+      "step": 22380
+    },
+    {
+      "epoch": 0.5327287441683328,
+      "grad_norm": 3.528696559677901,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 22381
+    },
+    {
+      "epoch": 0.5327525468913644,
+      "grad_norm": 3.461133980900575,
+      "learning_rate": 2e-05,
+      "loss": 0.5628,
+      "step": 22382
+    },
+    {
+      "epoch": 0.5327763496143959,
+      "grad_norm": 7.3553177648030275,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 22383
+    },
+    {
+      "epoch": 0.5328001523374274,
+      "grad_norm": 1.9560508725043173,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 22384
+    },
+    {
+      "epoch": 0.5328239550604589,
+      "grad_norm": 7.507543006700542,
+      "learning_rate": 2e-05,
+      "loss": 1.2194,
+      "step": 22385
+    },
+    {
+      "epoch": 0.5328477577834905,
+      "grad_norm": 4.221607690797368,
+      "learning_rate": 2e-05,
+      "loss": 0.1718,
+      "step": 22386
+    },
+    {
+      "epoch": 0.532871560506522,
+      "grad_norm": 3.9059512436108306,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 22387
+    },
+    {
+      "epoch": 0.5328953632295534,
+      "grad_norm": 3.2548147141677997,
+      "learning_rate": 2e-05,
+      "loss": 0.2656,
+      "step": 22388
+    },
+    {
+      "epoch": 0.5329191659525849,
+      "grad_norm": 9.177027632831026,
+      "learning_rate": 2e-05,
+      "loss": 0.7622,
+      "step": 22389
+    },
+    {
+      "epoch": 0.5329429686756165,
+      "grad_norm": 2.8166085836235077,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 22390
+    },
+    {
+      "epoch": 0.532966771398648,
+      "grad_norm": 3.6370746196479855,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 22391
+    },
+    {
+      "epoch": 0.5329905741216795,
+      "grad_norm": 4.709074833768834,
+      "learning_rate": 2e-05,
+      "loss": 0.26,
+      "step": 22392
+    },
+    {
+      "epoch": 0.533014376844711,
+      "grad_norm": 4.505647302402259,
+      "learning_rate": 2e-05,
+      "loss": 1.1824,
+      "step": 22393
+    },
+    {
+      "epoch": 0.5330381795677426,
+      "grad_norm": 3.9097521267324673,
+      "learning_rate": 2e-05,
+      "loss": 0.3181,
+      "step": 22394
+    },
+    {
+      "epoch": 0.5330619822907741,
+      "grad_norm": 10.794238645994138,
+      "learning_rate": 2e-05,
+      "loss": 0.333,
+      "step": 22395
+    },
+    {
+      "epoch": 0.5330857850138055,
+      "grad_norm": 9.36044256390228,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 22396
+    },
+    {
+      "epoch": 0.5331095877368371,
+      "grad_norm": 6.93815323581556,
+      "learning_rate": 2e-05,
+      "loss": 0.2659,
+      "step": 22397
+    },
+    {
+      "epoch": 0.5331333904598686,
+      "grad_norm": 0.597601846089627,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 22398
+    },
+    {
+      "epoch": 0.5331571931829001,
+      "grad_norm": 4.5298163350878875,
+      "learning_rate": 2e-05,
+      "loss": 0.5271,
+      "step": 22399
+    },
+    {
+      "epoch": 0.5331809959059316,
+      "grad_norm": 0.08130535591166112,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 22400
+    },
+    {
+      "epoch": 0.5332047986289632,
+      "grad_norm": 5.9879992180327815,
+      "learning_rate": 2e-05,
+      "loss": 0.71,
+      "step": 22401
+    },
+    {
+      "epoch": 0.5332286013519947,
+      "grad_norm": 8.767528191097668,
+      "learning_rate": 2e-05,
+      "loss": 0.3644,
+      "step": 22402
+    },
+    {
+      "epoch": 0.5332524040750262,
+      "grad_norm": 6.578985545425384,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 22403
+    },
+    {
+      "epoch": 0.5332762067980577,
+      "grad_norm": 6.299858984160144,
+      "learning_rate": 2e-05,
+      "loss": 0.2358,
+      "step": 22404
+    },
+    {
+      "epoch": 0.5333000095210892,
+      "grad_norm": 0.6346223856811436,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 22405
+    },
+    {
+      "epoch": 0.5333238122441207,
+      "grad_norm": 7.079254205520416,
+      "learning_rate": 2e-05,
+      "loss": 0.8778,
+      "step": 22406
+    },
+    {
+      "epoch": 0.5333476149671522,
+      "grad_norm": 13.768451595587974,
+      "learning_rate": 2e-05,
+      "loss": 0.9226,
+      "step": 22407
+    },
+    {
+      "epoch": 0.5333714176901837,
+      "grad_norm": 3.442097533379313,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 22408
+    },
+    {
+      "epoch": 0.5333952204132153,
+      "grad_norm": 4.653620684160519,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 22409
+    },
+    {
+      "epoch": 0.5334190231362468,
+      "grad_norm": 6.441811277461805,
+      "learning_rate": 2e-05,
+      "loss": 1.3099,
+      "step": 22410
+    },
+    {
+      "epoch": 0.5334428258592783,
+      "grad_norm": 3.7692982308030922,
+      "learning_rate": 2e-05,
+      "loss": 0.5224,
+      "step": 22411
+    },
+    {
+      "epoch": 0.5334666285823099,
+      "grad_norm": 10.918659091971254,
+      "learning_rate": 2e-05,
+      "loss": 0.8658,
+      "step": 22412
+    },
+    {
+      "epoch": 0.5334904313053414,
+      "grad_norm": 6.284692130397333,
+      "learning_rate": 2e-05,
+      "loss": 0.6004,
+      "step": 22413
+    },
+    {
+      "epoch": 0.5335142340283728,
+      "grad_norm": 4.926116680497035,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 22414
+    },
+    {
+      "epoch": 0.5335380367514043,
+      "grad_norm": 5.33416821042099,
+      "learning_rate": 2e-05,
+      "loss": 0.8037,
+      "step": 22415
+    },
+    {
+      "epoch": 0.5335618394744359,
+      "grad_norm": 11.130302546789599,
+      "learning_rate": 2e-05,
+      "loss": 0.4387,
+      "step": 22416
+    },
+    {
+      "epoch": 0.5335856421974674,
+      "grad_norm": 0.7901507209533494,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 22417
+    },
+    {
+      "epoch": 0.5336094449204989,
+      "grad_norm": 3.9008777097563625,
+      "learning_rate": 2e-05,
+      "loss": 0.1876,
+      "step": 22418
+    },
+    {
+      "epoch": 0.5336332476435304,
+      "grad_norm": 5.093924109456032,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 22419
+    },
+    {
+      "epoch": 0.533657050366562,
+      "grad_norm": 6.318502439747921,
+      "learning_rate": 2e-05,
+      "loss": 0.2061,
+      "step": 22420
+    },
+    {
+      "epoch": 0.5336808530895935,
+      "grad_norm": 3.317252420377297,
+      "learning_rate": 2e-05,
+      "loss": 0.308,
+      "step": 22421
+    },
+    {
+      "epoch": 0.5337046558126249,
+      "grad_norm": 4.932617312563521,
+      "learning_rate": 2e-05,
+      "loss": 0.7566,
+      "step": 22422
+    },
+    {
+      "epoch": 0.5337284585356564,
+      "grad_norm": 1.0140849875631346,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 22423
+    },
+    {
+      "epoch": 0.533752261258688,
+      "grad_norm": 0.1346429825989065,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22424
+    },
+    {
+      "epoch": 0.5337760639817195,
+      "grad_norm": 0.15419401773891092,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 22425
+    },
+    {
+      "epoch": 0.533799866704751,
+      "grad_norm": 5.115849672892339,
+      "learning_rate": 2e-05,
+      "loss": 1.0538,
+      "step": 22426
+    },
+    {
+      "epoch": 0.5338236694277826,
+      "grad_norm": 14.781266698100364,
+      "learning_rate": 2e-05,
+      "loss": 0.6255,
+      "step": 22427
+    },
+    {
+      "epoch": 0.5338474721508141,
+      "grad_norm": 1.9613705095712792,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 22428
+    },
+    {
+      "epoch": 0.5338712748738456,
+      "grad_norm": 1.89989670329931,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 22429
+    },
+    {
+      "epoch": 0.533895077596877,
+      "grad_norm": 0.9601367829787245,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 22430
+    },
+    {
+      "epoch": 0.5339188803199086,
+      "grad_norm": 1.3857298460115575,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 22431
+    },
+    {
+      "epoch": 0.5339426830429401,
+      "grad_norm": 3.055421363761799,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 22432
+    },
+    {
+      "epoch": 0.5339664857659716,
+      "grad_norm": 4.288399309902736,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 22433
+    },
+    {
+      "epoch": 0.5339902884890031,
+      "grad_norm": 2.9136790452891312,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 22434
+    },
+    {
+      "epoch": 0.5340140912120347,
+      "grad_norm": 2.2946310510278027,
+      "learning_rate": 2e-05,
+      "loss": 0.1422,
+      "step": 22435
+    },
+    {
+      "epoch": 0.5340378939350662,
+      "grad_norm": 2.777092116391339,
+      "learning_rate": 2e-05,
+      "loss": 0.3291,
+      "step": 22436
+    },
+    {
+      "epoch": 0.5340616966580977,
+      "grad_norm": 4.368789697011771,
+      "learning_rate": 2e-05,
+      "loss": 0.458,
+      "step": 22437
+    },
+    {
+      "epoch": 0.5340854993811291,
+      "grad_norm": 5.321491433439038,
+      "learning_rate": 2e-05,
+      "loss": 1.2739,
+      "step": 22438
+    },
+    {
+      "epoch": 0.5341093021041607,
+      "grad_norm": 4.138179828849002,
+      "learning_rate": 2e-05,
+      "loss": 0.5345,
+      "step": 22439
+    },
+    {
+      "epoch": 0.5341331048271922,
+      "grad_norm": 2.197009757669626,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 22440
+    },
+    {
+      "epoch": 0.5341569075502237,
+      "grad_norm": 8.875919029317586,
+      "learning_rate": 2e-05,
+      "loss": 0.4521,
+      "step": 22441
+    },
+    {
+      "epoch": 0.5341807102732553,
+      "grad_norm": 0.8270733891243366,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 22442
+    },
+    {
+      "epoch": 0.5342045129962868,
+      "grad_norm": 5.89136478768556,
+      "learning_rate": 2e-05,
+      "loss": 1.1505,
+      "step": 22443
+    },
+    {
+      "epoch": 0.5342283157193183,
+      "grad_norm": 2.990041601598314,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 22444
+    },
+    {
+      "epoch": 0.5342521184423498,
+      "grad_norm": 3.674547058021427,
+      "learning_rate": 2e-05,
+      "loss": 0.3976,
+      "step": 22445
+    },
+    {
+      "epoch": 0.5342759211653814,
+      "grad_norm": 0.7046189164390078,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 22446
+    },
+    {
+      "epoch": 0.5342997238884128,
+      "grad_norm": 0.08817507181684509,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 22447
+    },
+    {
+      "epoch": 0.5343235266114443,
+      "grad_norm": 7.783405090373927,
+      "learning_rate": 2e-05,
+      "loss": 0.1506,
+      "step": 22448
+    },
+    {
+      "epoch": 0.5343473293344758,
+      "grad_norm": 6.663157427560832,
+      "learning_rate": 2e-05,
+      "loss": 0.2976,
+      "step": 22449
+    },
+    {
+      "epoch": 0.5343711320575074,
+      "grad_norm": 8.107886972116251,
+      "learning_rate": 2e-05,
+      "loss": 0.3,
+      "step": 22450
+    },
+    {
+      "epoch": 0.5343949347805389,
+      "grad_norm": 0.2087936117805154,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 22451
+    },
+    {
+      "epoch": 0.5344187375035704,
+      "grad_norm": 3.1923300718760355,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 22452
+    },
+    {
+      "epoch": 0.5344425402266019,
+      "grad_norm": 8.846828860408726,
+      "learning_rate": 2e-05,
+      "loss": 0.4436,
+      "step": 22453
+    },
+    {
+      "epoch": 0.5344663429496335,
+      "grad_norm": 7.715388949122688,
+      "learning_rate": 2e-05,
+      "loss": 0.6677,
+      "step": 22454
+    },
+    {
+      "epoch": 0.534490145672665,
+      "grad_norm": 0.8399655085297534,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 22455
+    },
+    {
+      "epoch": 0.5345139483956964,
+      "grad_norm": 2.735059678150086,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 22456
+    },
+    {
+      "epoch": 0.534537751118728,
+      "grad_norm": 5.269019548937775,
+      "learning_rate": 2e-05,
+      "loss": 0.3173,
+      "step": 22457
+    },
+    {
+      "epoch": 0.5345615538417595,
+      "grad_norm": 0.05132068307088514,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 22458
+    },
+    {
+      "epoch": 0.534585356564791,
+      "grad_norm": 0.951007584373992,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 22459
+    },
+    {
+      "epoch": 0.5346091592878225,
+      "grad_norm": 5.482914590521554,
+      "learning_rate": 2e-05,
+      "loss": 1.048,
+      "step": 22460
+    },
+    {
+      "epoch": 0.5346329620108541,
+      "grad_norm": 0.0495303757880436,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 22461
+    },
+    {
+      "epoch": 0.5346567647338856,
+      "grad_norm": 7.680986389822649,
+      "learning_rate": 2e-05,
+      "loss": 0.6579,
+      "step": 22462
+    },
+    {
+      "epoch": 0.534680567456917,
+      "grad_norm": 0.2319818643082406,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22463
+    },
+    {
+      "epoch": 0.5347043701799485,
+      "grad_norm": 2.898860716202639,
+      "learning_rate": 2e-05,
+      "loss": 0.529,
+      "step": 22464
+    },
+    {
+      "epoch": 0.5347281729029801,
+      "grad_norm": 9.875000990418922,
+      "learning_rate": 2e-05,
+      "loss": 0.881,
+      "step": 22465
+    },
+    {
+      "epoch": 0.5347519756260116,
+      "grad_norm": 6.221954576962553,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 22466
+    },
+    {
+      "epoch": 0.5347757783490431,
+      "grad_norm": 2.800961562428422,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 22467
+    },
+    {
+      "epoch": 0.5347995810720746,
+      "grad_norm": 3.5915671398159685,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 22468
+    },
+    {
+      "epoch": 0.5348233837951062,
+      "grad_norm": 5.278119410053902,
+      "learning_rate": 2e-05,
+      "loss": 0.5936,
+      "step": 22469
+    },
+    {
+      "epoch": 0.5348471865181377,
+      "grad_norm": 14.536800183677814,
+      "learning_rate": 2e-05,
+      "loss": 1.8072,
+      "step": 22470
+    },
+    {
+      "epoch": 0.5348709892411692,
+      "grad_norm": 4.790736698632483,
+      "learning_rate": 2e-05,
+      "loss": 0.7666,
+      "step": 22471
+    },
+    {
+      "epoch": 0.5348947919642008,
+      "grad_norm": 5.491980653085978,
+      "learning_rate": 2e-05,
+      "loss": 0.4181,
+      "step": 22472
+    },
+    {
+      "epoch": 0.5349185946872322,
+      "grad_norm": 0.6072956655314664,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 22473
+    },
+    {
+      "epoch": 0.5349423974102637,
+      "grad_norm": 22.532646703982277,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 22474
+    },
+    {
+      "epoch": 0.5349662001332952,
+      "grad_norm": 1.241062149037698,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 22475
+    },
+    {
+      "epoch": 0.5349900028563268,
+      "grad_norm": 7.2913070417715655,
+      "learning_rate": 2e-05,
+      "loss": 1.4031,
+      "step": 22476
+    },
+    {
+      "epoch": 0.5350138055793583,
+      "grad_norm": 4.54916190755301,
+      "learning_rate": 2e-05,
+      "loss": 0.9337,
+      "step": 22477
+    },
+    {
+      "epoch": 0.5350376083023898,
+      "grad_norm": 4.54573933188399,
+      "learning_rate": 2e-05,
+      "loss": 0.8182,
+      "step": 22478
+    },
+    {
+      "epoch": 0.5350614110254213,
+      "grad_norm": 4.2624571603683865,
+      "learning_rate": 2e-05,
+      "loss": 0.4702,
+      "step": 22479
+    },
+    {
+      "epoch": 0.5350852137484529,
+      "grad_norm": 1.045464870687549,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 22480
+    },
+    {
+      "epoch": 0.5351090164714843,
+      "grad_norm": 0.8530265962249242,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 22481
+    },
+    {
+      "epoch": 0.5351328191945158,
+      "grad_norm": 0.6437998947050597,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 22482
+    },
+    {
+      "epoch": 0.5351566219175473,
+      "grad_norm": 1.4832675867536664,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 22483
+    },
+    {
+      "epoch": 0.5351804246405789,
+      "grad_norm": 3.011601775433518,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 22484
+    },
+    {
+      "epoch": 0.5352042273636104,
+      "grad_norm": 0.7585421744589506,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 22485
+    },
+    {
+      "epoch": 0.5352280300866419,
+      "grad_norm": 0.8578847976723438,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 22486
+    },
+    {
+      "epoch": 0.5352518328096735,
+      "grad_norm": 0.17427964209615576,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 22487
+    },
+    {
+      "epoch": 0.535275635532705,
+      "grad_norm": 7.39697908550297,
+      "learning_rate": 2e-05,
+      "loss": 0.6271,
+      "step": 22488
+    },
+    {
+      "epoch": 0.5352994382557364,
+      "grad_norm": 6.096048852718831,
+      "learning_rate": 2e-05,
+      "loss": 1.236,
+      "step": 22489
+    },
+    {
+      "epoch": 0.5353232409787679,
+      "grad_norm": 5.5097093488832565,
+      "learning_rate": 2e-05,
+      "loss": 0.5167,
+      "step": 22490
+    },
+    {
+      "epoch": 0.5353470437017995,
+      "grad_norm": 3.814745475102466,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 22491
+    },
+    {
+      "epoch": 0.535370846424831,
+      "grad_norm": 2.548804947666897,
+      "learning_rate": 2e-05,
+      "loss": 0.2485,
+      "step": 22492
+    },
+    {
+      "epoch": 0.5353946491478625,
+      "grad_norm": 6.786593362310863,
+      "learning_rate": 2e-05,
+      "loss": 0.2701,
+      "step": 22493
+    },
+    {
+      "epoch": 0.535418451870894,
+      "grad_norm": 12.91607612757235,
+      "learning_rate": 2e-05,
+      "loss": 1.0046,
+      "step": 22494
+    },
+    {
+      "epoch": 0.5354422545939256,
+      "grad_norm": 9.133747895662125,
+      "learning_rate": 2e-05,
+      "loss": 1.4226,
+      "step": 22495
+    },
+    {
+      "epoch": 0.5354660573169571,
+      "grad_norm": 2.8971841374791296,
+      "learning_rate": 2e-05,
+      "loss": 0.1964,
+      "step": 22496
+    },
+    {
+      "epoch": 0.5354898600399886,
+      "grad_norm": 4.3365312000661005,
+      "learning_rate": 2e-05,
+      "loss": 0.4591,
+      "step": 22497
+    },
+    {
+      "epoch": 0.53551366276302,
+      "grad_norm": 6.0522949975557525,
+      "learning_rate": 2e-05,
+      "loss": 0.5679,
+      "step": 22498
+    },
+    {
+      "epoch": 0.5355374654860516,
+      "grad_norm": 2.7367818647737083,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 22499
+    },
+    {
+      "epoch": 0.5355612682090831,
+      "grad_norm": 2.2122128568761013,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 22500
+    },
+    {
+      "epoch": 0.5355850709321146,
+      "grad_norm": 6.1052133375701505,
+      "learning_rate": 2e-05,
+      "loss": 1.6373,
+      "step": 22501
+    },
+    {
+      "epoch": 0.5356088736551462,
+      "grad_norm": 3.105505843019017,
+      "learning_rate": 2e-05,
+      "loss": 0.2806,
+      "step": 22502
+    },
+    {
+      "epoch": 0.5356326763781777,
+      "grad_norm": 12.224645120834957,
+      "learning_rate": 2e-05,
+      "loss": 0.9571,
+      "step": 22503
+    },
+    {
+      "epoch": 0.5356564791012092,
+      "grad_norm": 0.25585712277412653,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22504
+    },
+    {
+      "epoch": 0.5356802818242407,
+      "grad_norm": 4.719729434700084,
+      "learning_rate": 2e-05,
+      "loss": 1.0629,
+      "step": 22505
+    },
+    {
+      "epoch": 0.5357040845472723,
+      "grad_norm": 3.8384181265266863,
+      "learning_rate": 2e-05,
+      "loss": 0.1203,
+      "step": 22506
+    },
+    {
+      "epoch": 0.5357278872703037,
+      "grad_norm": 13.740131158595288,
+      "learning_rate": 2e-05,
+      "loss": 0.5252,
+      "step": 22507
+    },
+    {
+      "epoch": 0.5357516899933352,
+      "grad_norm": 7.51478948017939,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 22508
+    },
+    {
+      "epoch": 0.5357754927163667,
+      "grad_norm": 4.8805814940255425,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 22509
+    },
+    {
+      "epoch": 0.5357992954393983,
+      "grad_norm": 0.9075700713551905,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 22510
+    },
+    {
+      "epoch": 0.5358230981624298,
+      "grad_norm": 4.0966081182123375,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 22511
+    },
+    {
+      "epoch": 0.5358469008854613,
+      "grad_norm": 0.29946671733875824,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 22512
+    },
+    {
+      "epoch": 0.5358707036084928,
+      "grad_norm": 3.395379318140403,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 22513
+    },
+    {
+      "epoch": 0.5358945063315244,
+      "grad_norm": 4.504699784372231,
+      "learning_rate": 2e-05,
+      "loss": 0.7522,
+      "step": 22514
+    },
+    {
+      "epoch": 0.5359183090545558,
+      "grad_norm": 2.6599656068960975,
+      "learning_rate": 2e-05,
+      "loss": 0.2061,
+      "step": 22515
+    },
+    {
+      "epoch": 0.5359421117775873,
+      "grad_norm": 0.9027727706409827,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 22516
+    },
+    {
+      "epoch": 0.5359659145006189,
+      "grad_norm": 2.2441685725443343,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 22517
+    },
+    {
+      "epoch": 0.5359897172236504,
+      "grad_norm": 6.731772251275802,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 22518
+    },
+    {
+      "epoch": 0.5360135199466819,
+      "grad_norm": 5.058014990685243,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 22519
+    },
+    {
+      "epoch": 0.5360373226697134,
+      "grad_norm": 6.251395515820375,
+      "learning_rate": 2e-05,
+      "loss": 0.553,
+      "step": 22520
+    },
+    {
+      "epoch": 0.536061125392745,
+      "grad_norm": 3.760028507329616,
+      "learning_rate": 2e-05,
+      "loss": 0.824,
+      "step": 22521
+    },
+    {
+      "epoch": 0.5360849281157765,
+      "grad_norm": 0.058346367440601635,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 22522
+    },
+    {
+      "epoch": 0.536108730838808,
+      "grad_norm": 0.16530756415893497,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 22523
+    },
+    {
+      "epoch": 0.5361325335618394,
+      "grad_norm": 5.687767320137851,
+      "learning_rate": 2e-05,
+      "loss": 0.2871,
+      "step": 22524
+    },
+    {
+      "epoch": 0.536156336284871,
+      "grad_norm": 3.250853915162945,
+      "learning_rate": 2e-05,
+      "loss": 0.1618,
+      "step": 22525
+    },
+    {
+      "epoch": 0.5361801390079025,
+      "grad_norm": 3.6018526324752793,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 22526
+    },
+    {
+      "epoch": 0.536203941730934,
+      "grad_norm": 0.15076129224145313,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 22527
+    },
+    {
+      "epoch": 0.5362277444539655,
+      "grad_norm": 2.937044781999315,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 22528
+    },
+    {
+      "epoch": 0.5362515471769971,
+      "grad_norm": 2.639375334190947,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 22529
+    },
+    {
+      "epoch": 0.5362753499000286,
+      "grad_norm": 3.5068931231778464,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 22530
+    },
+    {
+      "epoch": 0.53629915262306,
+      "grad_norm": 4.527417029046222,
+      "learning_rate": 2e-05,
+      "loss": 0.4553,
+      "step": 22531
+    },
+    {
+      "epoch": 0.5363229553460916,
+      "grad_norm": 3.6524717951935632,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 22532
+    },
+    {
+      "epoch": 0.5363467580691231,
+      "grad_norm": 7.4515806149555495,
+      "learning_rate": 2e-05,
+      "loss": 0.2871,
+      "step": 22533
+    },
+    {
+      "epoch": 0.5363705607921546,
+      "grad_norm": 3.992102466145443,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 22534
+    },
+    {
+      "epoch": 0.5363943635151861,
+      "grad_norm": 6.442982880510362,
+      "learning_rate": 2e-05,
+      "loss": 1.1256,
+      "step": 22535
+    },
+    {
+      "epoch": 0.5364181662382177,
+      "grad_norm": 0.1036380232714177,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 22536
+    },
+    {
+      "epoch": 0.5364419689612492,
+      "grad_norm": 7.012799816410213,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 22537
+    },
+    {
+      "epoch": 0.5364657716842807,
+      "grad_norm": 2.3239745051919,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 22538
+    },
+    {
+      "epoch": 0.5364895744073122,
+      "grad_norm": 6.89540506836432,
+      "learning_rate": 2e-05,
+      "loss": 0.1387,
+      "step": 22539
+    },
+    {
+      "epoch": 0.5365133771303437,
+      "grad_norm": 2.1339782542416375,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 22540
+    },
+    {
+      "epoch": 0.5365371798533752,
+      "grad_norm": 3.727929023647857,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 22541
+    },
+    {
+      "epoch": 0.5365609825764067,
+      "grad_norm": 6.036194829433152,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 22542
+    },
+    {
+      "epoch": 0.5365847852994382,
+      "grad_norm": 3.435744325237494,
+      "learning_rate": 2e-05,
+      "loss": 0.3506,
+      "step": 22543
+    },
+    {
+      "epoch": 0.5366085880224698,
+      "grad_norm": 5.4032360811274724,
+      "learning_rate": 2e-05,
+      "loss": 0.728,
+      "step": 22544
+    },
+    {
+      "epoch": 0.5366323907455013,
+      "grad_norm": 3.359031758923435,
+      "learning_rate": 2e-05,
+      "loss": 0.5522,
+      "step": 22545
+    },
+    {
+      "epoch": 0.5366561934685328,
+      "grad_norm": 4.03733254263775,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 22546
+    },
+    {
+      "epoch": 0.5366799961915644,
+      "grad_norm": 4.728657857161233,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 22547
+    },
+    {
+      "epoch": 0.5367037989145959,
+      "grad_norm": 26.79484177953822,
+      "learning_rate": 2e-05,
+      "loss": 0.4479,
+      "step": 22548
+    },
+    {
+      "epoch": 0.5367276016376273,
+      "grad_norm": 3.423427670888298,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 22549
+    },
+    {
+      "epoch": 0.5367514043606588,
+      "grad_norm": 2.5875028221116523,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 22550
+    },
+    {
+      "epoch": 0.5367752070836904,
+      "grad_norm": 0.6336502025488135,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 22551
+    },
+    {
+      "epoch": 0.5367990098067219,
+      "grad_norm": 3.287429362890807,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 22552
+    },
+    {
+      "epoch": 0.5368228125297534,
+      "grad_norm": 2.671852504740798,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 22553
+    },
+    {
+      "epoch": 0.5368466152527849,
+      "grad_norm": 3.9971074900476893,
+      "learning_rate": 2e-05,
+      "loss": 0.4179,
+      "step": 22554
+    },
+    {
+      "epoch": 0.5368704179758165,
+      "grad_norm": 14.625123223474063,
+      "learning_rate": 2e-05,
+      "loss": 0.3881,
+      "step": 22555
+    },
+    {
+      "epoch": 0.536894220698848,
+      "grad_norm": 4.0377095173267605,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 22556
+    },
+    {
+      "epoch": 0.5369180234218794,
+      "grad_norm": 4.932023851843362,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 22557
+    },
+    {
+      "epoch": 0.5369418261449109,
+      "grad_norm": 2.932469223283281,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 22558
+    },
+    {
+      "epoch": 0.5369656288679425,
+      "grad_norm": 6.403280342678653,
+      "learning_rate": 2e-05,
+      "loss": 0.1708,
+      "step": 22559
+    },
+    {
+      "epoch": 0.536989431590974,
+      "grad_norm": 1.0455086876524815,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 22560
+    },
+    {
+      "epoch": 0.5370132343140055,
+      "grad_norm": 6.137408333355624,
+      "learning_rate": 2e-05,
+      "loss": 1.5316,
+      "step": 22561
+    },
+    {
+      "epoch": 0.5370370370370371,
+      "grad_norm": 3.771603723768494,
+      "learning_rate": 2e-05,
+      "loss": 0.546,
+      "step": 22562
+    },
+    {
+      "epoch": 0.5370608397600686,
+      "grad_norm": 9.411994870517926,
+      "learning_rate": 2e-05,
+      "loss": 1.15,
+      "step": 22563
+    },
+    {
+      "epoch": 0.5370846424831001,
+      "grad_norm": 0.18296720422726193,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 22564
+    },
+    {
+      "epoch": 0.5371084452061315,
+      "grad_norm": 4.3737851849029115,
+      "learning_rate": 2e-05,
+      "loss": 0.7003,
+      "step": 22565
+    },
+    {
+      "epoch": 0.5371322479291631,
+      "grad_norm": 9.258024608464972,
+      "learning_rate": 2e-05,
+      "loss": 0.6124,
+      "step": 22566
+    },
+    {
+      "epoch": 0.5371560506521946,
+      "grad_norm": 4.148040391750722,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 22567
+    },
+    {
+      "epoch": 0.5371798533752261,
+      "grad_norm": 1.8579713541236598,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 22568
+    },
+    {
+      "epoch": 0.5372036560982576,
+      "grad_norm": 6.446574879860835,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 22569
+    },
+    {
+      "epoch": 0.5372274588212892,
+      "grad_norm": 3.7016233409746757,
+      "learning_rate": 2e-05,
+      "loss": 0.1358,
+      "step": 22570
+    },
+    {
+      "epoch": 0.5372512615443207,
+      "grad_norm": 2.77912348844356,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 22571
+    },
+    {
+      "epoch": 0.5372750642673522,
+      "grad_norm": 5.0156970795047595,
+      "learning_rate": 2e-05,
+      "loss": 0.8561,
+      "step": 22572
+    },
+    {
+      "epoch": 0.5372988669903837,
+      "grad_norm": 5.903163572905825,
+      "learning_rate": 2e-05,
+      "loss": 0.6822,
+      "step": 22573
+    },
+    {
+      "epoch": 0.5373226697134152,
+      "grad_norm": 2.0908903228358424,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 22574
+    },
+    {
+      "epoch": 0.5373464724364467,
+      "grad_norm": 2.4421890385760157,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 22575
+    },
+    {
+      "epoch": 0.5373702751594782,
+      "grad_norm": 3.0291134664511388,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 22576
+    },
+    {
+      "epoch": 0.5373940778825098,
+      "grad_norm": 0.8177884684389336,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 22577
+    },
+    {
+      "epoch": 0.5374178806055413,
+      "grad_norm": 7.749843357788521,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 22578
+    },
+    {
+      "epoch": 0.5374416833285728,
+      "grad_norm": 0.5646350694887361,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 22579
+    },
+    {
+      "epoch": 0.5374654860516043,
+      "grad_norm": 0.10945492276445978,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 22580
+    },
+    {
+      "epoch": 0.5374892887746359,
+      "grad_norm": 3.7302467596862185,
+      "learning_rate": 2e-05,
+      "loss": 0.8496,
+      "step": 22581
+    },
+    {
+      "epoch": 0.5375130914976674,
+      "grad_norm": 18.216723354835242,
+      "learning_rate": 2e-05,
+      "loss": 1.046,
+      "step": 22582
+    },
+    {
+      "epoch": 0.5375368942206988,
+      "grad_norm": 3.292703808388696,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 22583
+    },
+    {
+      "epoch": 0.5375606969437303,
+      "grad_norm": 4.17584620727804,
+      "learning_rate": 2e-05,
+      "loss": 0.7572,
+      "step": 22584
+    },
+    {
+      "epoch": 0.5375844996667619,
+      "grad_norm": 3.4066177466759986,
+      "learning_rate": 2e-05,
+      "loss": 0.7852,
+      "step": 22585
+    },
+    {
+      "epoch": 0.5376083023897934,
+      "grad_norm": 1.5273126015457261,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22586
+    },
+    {
+      "epoch": 0.5376321051128249,
+      "grad_norm": 0.6841422891228487,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 22587
+    },
+    {
+      "epoch": 0.5376559078358564,
+      "grad_norm": 2.2660336303574287,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 22588
+    },
+    {
+      "epoch": 0.537679710558888,
+      "grad_norm": 6.52109197949213,
+      "learning_rate": 2e-05,
+      "loss": 0.9648,
+      "step": 22589
+    },
+    {
+      "epoch": 0.5377035132819195,
+      "grad_norm": 1.5278900133074307,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 22590
+    },
+    {
+      "epoch": 0.5377273160049509,
+      "grad_norm": 1.5028174923155262,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 22591
+    },
+    {
+      "epoch": 0.5377511187279825,
+      "grad_norm": 4.514425734717972,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 22592
+    },
+    {
+      "epoch": 0.537774921451014,
+      "grad_norm": 0.2836145446511831,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 22593
+    },
+    {
+      "epoch": 0.5377987241740455,
+      "grad_norm": 0.2669972978296844,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22594
+    },
+    {
+      "epoch": 0.537822526897077,
+      "grad_norm": 1.0952252633921393,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 22595
+    },
+    {
+      "epoch": 0.5378463296201086,
+      "grad_norm": 2.9629603676168776,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 22596
+    },
+    {
+      "epoch": 0.5378701323431401,
+      "grad_norm": 9.999037997594105,
+      "learning_rate": 2e-05,
+      "loss": 0.7781,
+      "step": 22597
+    },
+    {
+      "epoch": 0.5378939350661716,
+      "grad_norm": 9.314535594954723,
+      "learning_rate": 2e-05,
+      "loss": 0.895,
+      "step": 22598
+    },
+    {
+      "epoch": 0.537917737789203,
+      "grad_norm": 5.3973906600401635,
+      "learning_rate": 2e-05,
+      "loss": 0.9162,
+      "step": 22599
+    },
+    {
+      "epoch": 0.5379415405122346,
+      "grad_norm": 12.408725406713785,
+      "learning_rate": 2e-05,
+      "loss": 0.5994,
+      "step": 22600
+    },
+    {
+      "epoch": 0.5379653432352661,
+      "grad_norm": 6.5554139230471,
+      "learning_rate": 2e-05,
+      "loss": 1.2876,
+      "step": 22601
+    },
+    {
+      "epoch": 0.5379891459582976,
+      "grad_norm": 1.3085854645525268,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 22602
+    },
+    {
+      "epoch": 0.5380129486813291,
+      "grad_norm": 2.8250325745856224,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 22603
+    },
+    {
+      "epoch": 0.5380367514043607,
+      "grad_norm": 4.299469181922022,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 22604
+    },
+    {
+      "epoch": 0.5380605541273922,
+      "grad_norm": 14.17102041839011,
+      "learning_rate": 2e-05,
+      "loss": 1.0004,
+      "step": 22605
+    },
+    {
+      "epoch": 0.5380843568504237,
+      "grad_norm": 3.9238097807889516,
+      "learning_rate": 2e-05,
+      "loss": 0.9317,
+      "step": 22606
+    },
+    {
+      "epoch": 0.5381081595734553,
+      "grad_norm": 4.557359733785327,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 22607
+    },
+    {
+      "epoch": 0.5381319622964867,
+      "grad_norm": 2.4948462658167254,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 22608
+    },
+    {
+      "epoch": 0.5381557650195182,
+      "grad_norm": 5.801484436601733,
+      "learning_rate": 2e-05,
+      "loss": 0.7493,
+      "step": 22609
+    },
+    {
+      "epoch": 0.5381795677425497,
+      "grad_norm": 1.006399080353193,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 22610
+    },
+    {
+      "epoch": 0.5382033704655813,
+      "grad_norm": 2.3842832910789196,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 22611
+    },
+    {
+      "epoch": 0.5382271731886128,
+      "grad_norm": 7.021387171645567,
+      "learning_rate": 2e-05,
+      "loss": 1.4523,
+      "step": 22612
+    },
+    {
+      "epoch": 0.5382509759116443,
+      "grad_norm": 5.588595281545871,
+      "learning_rate": 2e-05,
+      "loss": 0.94,
+      "step": 22613
+    },
+    {
+      "epoch": 0.5382747786346758,
+      "grad_norm": 0.9124374857741817,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 22614
+    },
+    {
+      "epoch": 0.5382985813577074,
+      "grad_norm": 3.1399490553970293,
+      "learning_rate": 2e-05,
+      "loss": 0.6333,
+      "step": 22615
+    },
+    {
+      "epoch": 0.5383223840807388,
+      "grad_norm": 5.10527415337884,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 22616
+    },
+    {
+      "epoch": 0.5383461868037703,
+      "grad_norm": 9.963985806176181,
+      "learning_rate": 2e-05,
+      "loss": 0.574,
+      "step": 22617
+    },
+    {
+      "epoch": 0.5383699895268018,
+      "grad_norm": 2.4381263932115846,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 22618
+    },
+    {
+      "epoch": 0.5383937922498334,
+      "grad_norm": 13.177165514091138,
+      "learning_rate": 2e-05,
+      "loss": 0.1914,
+      "step": 22619
+    },
+    {
+      "epoch": 0.5384175949728649,
+      "grad_norm": 16.682866611399405,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 22620
+    },
+    {
+      "epoch": 0.5384413976958964,
+      "grad_norm": 2.963887420350883,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 22621
+    },
+    {
+      "epoch": 0.538465200418928,
+      "grad_norm": 17.894629726068192,
+      "learning_rate": 2e-05,
+      "loss": 0.6668,
+      "step": 22622
+    },
+    {
+      "epoch": 0.5384890031419595,
+      "grad_norm": 3.824800299350626,
+      "learning_rate": 2e-05,
+      "loss": 0.5318,
+      "step": 22623
+    },
+    {
+      "epoch": 0.538512805864991,
+      "grad_norm": 4.767787446477638,
+      "learning_rate": 2e-05,
+      "loss": 0.5834,
+      "step": 22624
+    },
+    {
+      "epoch": 0.5385366085880224,
+      "grad_norm": 0.17845262273024595,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22625
+    },
+    {
+      "epoch": 0.538560411311054,
+      "grad_norm": 8.137657873931465,
+      "learning_rate": 2e-05,
+      "loss": 1.2086,
+      "step": 22626
+    },
+    {
+      "epoch": 0.5385842140340855,
+      "grad_norm": 3.1189516215766093,
+      "learning_rate": 2e-05,
+      "loss": 0.1574,
+      "step": 22627
+    },
+    {
+      "epoch": 0.538608016757117,
+      "grad_norm": 4.354600194911905,
+      "learning_rate": 2e-05,
+      "loss": 0.6469,
+      "step": 22628
+    },
+    {
+      "epoch": 0.5386318194801485,
+      "grad_norm": 3.9909127412557233,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 22629
+    },
+    {
+      "epoch": 0.5386556222031801,
+      "grad_norm": 1.0024191424596554,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 22630
+    },
+    {
+      "epoch": 0.5386794249262116,
+      "grad_norm": 3.250578746367818,
+      "learning_rate": 2e-05,
+      "loss": 0.5568,
+      "step": 22631
+    },
+    {
+      "epoch": 0.538703227649243,
+      "grad_norm": 0.9666227071520814,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 22632
+    },
+    {
+      "epoch": 0.5387270303722745,
+      "grad_norm": 2.9534383119897463,
+      "learning_rate": 2e-05,
+      "loss": 0.5896,
+      "step": 22633
+    },
+    {
+      "epoch": 0.5387508330953061,
+      "grad_norm": 4.679538119145584,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 22634
+    },
+    {
+      "epoch": 0.5387746358183376,
+      "grad_norm": 3.1216388928054215,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 22635
+    },
+    {
+      "epoch": 0.5387984385413691,
+      "grad_norm": 3.6600227206475,
+      "learning_rate": 2e-05,
+      "loss": 0.4921,
+      "step": 22636
+    },
+    {
+      "epoch": 0.5388222412644006,
+      "grad_norm": 3.5357337850620425,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 22637
+    },
+    {
+      "epoch": 0.5388460439874322,
+      "grad_norm": 10.259440099217047,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 22638
+    },
+    {
+      "epoch": 0.5388698467104637,
+      "grad_norm": 10.360517127860408,
+      "learning_rate": 2e-05,
+      "loss": 0.2676,
+      "step": 22639
+    },
+    {
+      "epoch": 0.5388936494334952,
+      "grad_norm": 5.813192384789505,
+      "learning_rate": 2e-05,
+      "loss": 0.3911,
+      "step": 22640
+    },
+    {
+      "epoch": 0.5389174521565268,
+      "grad_norm": 4.668820977227506,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 22641
+    },
+    {
+      "epoch": 0.5389412548795582,
+      "grad_norm": 1.7277476867874044,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 22642
+    },
+    {
+      "epoch": 0.5389650576025897,
+      "grad_norm": 2.741667498234083,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 22643
+    },
+    {
+      "epoch": 0.5389888603256212,
+      "grad_norm": 1.4497906675542815,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 22644
+    },
+    {
+      "epoch": 0.5390126630486528,
+      "grad_norm": 1.8572260962654499,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 22645
+    },
+    {
+      "epoch": 0.5390364657716843,
+      "grad_norm": 2.4161176832877618,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 22646
+    },
+    {
+      "epoch": 0.5390602684947158,
+      "grad_norm": 13.734096044585014,
+      "learning_rate": 2e-05,
+      "loss": 1.0679,
+      "step": 22647
+    },
+    {
+      "epoch": 0.5390840712177473,
+      "grad_norm": 6.1695051622103145,
+      "learning_rate": 2e-05,
+      "loss": 0.3135,
+      "step": 22648
+    },
+    {
+      "epoch": 0.5391078739407789,
+      "grad_norm": 6.329680063874259,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 22649
+    },
+    {
+      "epoch": 0.5391316766638103,
+      "grad_norm": 0.7672525634653631,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 22650
+    },
+    {
+      "epoch": 0.5391554793868418,
+      "grad_norm": 1.1001152544913786,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 22651
+    },
+    {
+      "epoch": 0.5391792821098733,
+      "grad_norm": 6.566340104084711,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 22652
+    },
+    {
+      "epoch": 0.5392030848329049,
+      "grad_norm": 2.8153592504194735,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 22653
+    },
+    {
+      "epoch": 0.5392268875559364,
+      "grad_norm": 4.51298798878318,
+      "learning_rate": 2e-05,
+      "loss": 0.9921,
+      "step": 22654
+    },
+    {
+      "epoch": 0.5392506902789679,
+      "grad_norm": 0.13575814673656875,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22655
+    },
+    {
+      "epoch": 0.5392744930019995,
+      "grad_norm": 0.7136944449957958,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 22656
+    },
+    {
+      "epoch": 0.539298295725031,
+      "grad_norm": 3.6862070400945157,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 22657
+    },
+    {
+      "epoch": 0.5393220984480624,
+      "grad_norm": 2.0684251035661347,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 22658
+    },
+    {
+      "epoch": 0.5393459011710939,
+      "grad_norm": 5.781936630490822,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 22659
+    },
+    {
+      "epoch": 0.5393697038941255,
+      "grad_norm": 0.23019150301261532,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 22660
+    },
+    {
+      "epoch": 0.539393506617157,
+      "grad_norm": 3.5586809536567743,
+      "learning_rate": 2e-05,
+      "loss": 0.5396,
+      "step": 22661
+    },
+    {
+      "epoch": 0.5394173093401885,
+      "grad_norm": 0.4339214493472701,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 22662
+    },
+    {
+      "epoch": 0.53944111206322,
+      "grad_norm": 6.261383446072862,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 22663
+    },
+    {
+      "epoch": 0.5394649147862516,
+      "grad_norm": 10.185666082199875,
+      "learning_rate": 2e-05,
+      "loss": 0.878,
+      "step": 22664
+    },
+    {
+      "epoch": 0.5394887175092831,
+      "grad_norm": 0.6708752665617307,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 22665
+    },
+    {
+      "epoch": 0.5395125202323146,
+      "grad_norm": 0.5851577411300042,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 22666
+    },
+    {
+      "epoch": 0.539536322955346,
+      "grad_norm": 10.990635907606455,
+      "learning_rate": 2e-05,
+      "loss": 0.8484,
+      "step": 22667
+    },
+    {
+      "epoch": 0.5395601256783776,
+      "grad_norm": 0.2202223005209738,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 22668
+    },
+    {
+      "epoch": 0.5395839284014091,
+      "grad_norm": 8.421019177844592,
+      "learning_rate": 2e-05,
+      "loss": 1.5575,
+      "step": 22669
+    },
+    {
+      "epoch": 0.5396077311244406,
+      "grad_norm": 17.841227980893695,
+      "learning_rate": 2e-05,
+      "loss": 0.5314,
+      "step": 22670
+    },
+    {
+      "epoch": 0.5396315338474722,
+      "grad_norm": 4.657058631610325,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 22671
+    },
+    {
+      "epoch": 0.5396553365705037,
+      "grad_norm": 7.954115263147116,
+      "learning_rate": 2e-05,
+      "loss": 0.1499,
+      "step": 22672
+    },
+    {
+      "epoch": 0.5396791392935352,
+      "grad_norm": 5.898271141576208,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 22673
+    },
+    {
+      "epoch": 0.5397029420165667,
+      "grad_norm": 4.377137916803772,
+      "learning_rate": 2e-05,
+      "loss": 0.5036,
+      "step": 22674
+    },
+    {
+      "epoch": 0.5397267447395983,
+      "grad_norm": 2.6905381598274705,
+      "learning_rate": 2e-05,
+      "loss": 0.1184,
+      "step": 22675
+    },
+    {
+      "epoch": 0.5397505474626297,
+      "grad_norm": 4.066636772259197,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 22676
+    },
+    {
+      "epoch": 0.5397743501856612,
+      "grad_norm": 0.12300246030975984,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22677
+    },
+    {
+      "epoch": 0.5397981529086927,
+      "grad_norm": 6.543898735909248,
+      "learning_rate": 2e-05,
+      "loss": 0.1794,
+      "step": 22678
+    },
+    {
+      "epoch": 0.5398219556317243,
+      "grad_norm": 6.316523513963928,
+      "learning_rate": 2e-05,
+      "loss": 1.2098,
+      "step": 22679
+    },
+    {
+      "epoch": 0.5398457583547558,
+      "grad_norm": 13.069769117521378,
+      "learning_rate": 2e-05,
+      "loss": 1.5247,
+      "step": 22680
+    },
+    {
+      "epoch": 0.5398695610777873,
+      "grad_norm": 5.792659732401025,
+      "learning_rate": 2e-05,
+      "loss": 0.6259,
+      "step": 22681
+    },
+    {
+      "epoch": 0.5398933638008188,
+      "grad_norm": 5.401077909628978,
+      "learning_rate": 2e-05,
+      "loss": 0.6881,
+      "step": 22682
+    },
+    {
+      "epoch": 0.5399171665238504,
+      "grad_norm": 4.0542074956450245,
+      "learning_rate": 2e-05,
+      "loss": 0.6432,
+      "step": 22683
+    },
+    {
+      "epoch": 0.5399409692468818,
+      "grad_norm": 4.541653828038664,
+      "learning_rate": 2e-05,
+      "loss": 1.0888,
+      "step": 22684
+    },
+    {
+      "epoch": 0.5399647719699133,
+      "grad_norm": 7.134705280246062,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 22685
+    },
+    {
+      "epoch": 0.5399885746929449,
+      "grad_norm": 5.290401356453049,
+      "learning_rate": 2e-05,
+      "loss": 0.901,
+      "step": 22686
+    },
+    {
+      "epoch": 0.5400123774159764,
+      "grad_norm": 6.405712586241902,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 22687
+    },
+    {
+      "epoch": 0.5400361801390079,
+      "grad_norm": 8.84671234789834,
+      "learning_rate": 2e-05,
+      "loss": 0.8911,
+      "step": 22688
+    },
+    {
+      "epoch": 0.5400599828620394,
+      "grad_norm": 0.09315686591514126,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 22689
+    },
+    {
+      "epoch": 0.540083785585071,
+      "grad_norm": 4.596512203537107,
+      "learning_rate": 2e-05,
+      "loss": 0.2169,
+      "step": 22690
+    },
+    {
+      "epoch": 0.5401075883081025,
+      "grad_norm": 0.11671030004583995,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 22691
+    },
+    {
+      "epoch": 0.540131391031134,
+      "grad_norm": 0.4469496485435637,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22692
+    },
+    {
+      "epoch": 0.5401551937541654,
+      "grad_norm": 4.740953895022817,
+      "learning_rate": 2e-05,
+      "loss": 1.13,
+      "step": 22693
+    },
+    {
+      "epoch": 0.540178996477197,
+      "grad_norm": 5.730740754422056,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 22694
+    },
+    {
+      "epoch": 0.5402027992002285,
+      "grad_norm": 6.515852489919605,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 22695
+    },
+    {
+      "epoch": 0.54022660192326,
+      "grad_norm": 3.0613223255269775,
+      "learning_rate": 2e-05,
+      "loss": 0.2765,
+      "step": 22696
+    },
+    {
+      "epoch": 0.5402504046462915,
+      "grad_norm": 4.263312411836493,
+      "learning_rate": 2e-05,
+      "loss": 0.4391,
+      "step": 22697
+    },
+    {
+      "epoch": 0.5402742073693231,
+      "grad_norm": 3.2817568819145744,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 22698
+    },
+    {
+      "epoch": 0.5402980100923546,
+      "grad_norm": 4.817995341067815,
+      "learning_rate": 2e-05,
+      "loss": 0.7492,
+      "step": 22699
+    },
+    {
+      "epoch": 0.540321812815386,
+      "grad_norm": 5.103523828214267,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 22700
+    },
+    {
+      "epoch": 0.5403456155384176,
+      "grad_norm": 0.2511655428706966,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 22701
+    },
+    {
+      "epoch": 0.5403694182614491,
+      "grad_norm": 1.604303081098427,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 22702
+    },
+    {
+      "epoch": 0.5403932209844806,
+      "grad_norm": 6.649537665065331,
+      "learning_rate": 2e-05,
+      "loss": 0.8125,
+      "step": 22703
+    },
+    {
+      "epoch": 0.5404170237075121,
+      "grad_norm": 3.345352398132426,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 22704
+    },
+    {
+      "epoch": 0.5404408264305437,
+      "grad_norm": 4.68183400471538,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 22705
+    },
+    {
+      "epoch": 0.5404646291535752,
+      "grad_norm": 3.837019565402171,
+      "learning_rate": 2e-05,
+      "loss": 0.5381,
+      "step": 22706
+    },
+    {
+      "epoch": 0.5404884318766067,
+      "grad_norm": 0.3795736966735705,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 22707
+    },
+    {
+      "epoch": 0.5405122345996382,
+      "grad_norm": 5.266392258460304,
+      "learning_rate": 2e-05,
+      "loss": 0.0937,
+      "step": 22708
+    },
+    {
+      "epoch": 0.5405360373226697,
+      "grad_norm": 4.879970277102584,
+      "learning_rate": 2e-05,
+      "loss": 0.5165,
+      "step": 22709
+    },
+    {
+      "epoch": 0.5405598400457012,
+      "grad_norm": 4.09678380684013,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 22710
+    },
+    {
+      "epoch": 0.5405836427687327,
+      "grad_norm": 12.241809855960812,
+      "learning_rate": 2e-05,
+      "loss": 0.9607,
+      "step": 22711
+    },
+    {
+      "epoch": 0.5406074454917642,
+      "grad_norm": 2.3001223942649687,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 22712
+    },
+    {
+      "epoch": 0.5406312482147958,
+      "grad_norm": 24.469330495059218,
+      "learning_rate": 2e-05,
+      "loss": 0.6679,
+      "step": 22713
+    },
+    {
+      "epoch": 0.5406550509378273,
+      "grad_norm": 0.9257940026846577,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 22714
+    },
+    {
+      "epoch": 0.5406788536608588,
+      "grad_norm": 6.009208098266857,
+      "learning_rate": 2e-05,
+      "loss": 0.5699,
+      "step": 22715
+    },
+    {
+      "epoch": 0.5407026563838904,
+      "grad_norm": 1.2021383550756057,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 22716
+    },
+    {
+      "epoch": 0.5407264591069219,
+      "grad_norm": 4.658659330427487,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 22717
+    },
+    {
+      "epoch": 0.5407502618299533,
+      "grad_norm": 42.410955124287476,
+      "learning_rate": 2e-05,
+      "loss": 0.6828,
+      "step": 22718
+    },
+    {
+      "epoch": 0.5407740645529848,
+      "grad_norm": 5.089874398170024,
+      "learning_rate": 2e-05,
+      "loss": 0.7755,
+      "step": 22719
+    },
+    {
+      "epoch": 0.5407978672760164,
+      "grad_norm": 0.20585628268018466,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 22720
+    },
+    {
+      "epoch": 0.5408216699990479,
+      "grad_norm": 11.486492819114842,
+      "learning_rate": 2e-05,
+      "loss": 1.4197,
+      "step": 22721
+    },
+    {
+      "epoch": 0.5408454727220794,
+      "grad_norm": 9.988295589331146,
+      "learning_rate": 2e-05,
+      "loss": 1.1128,
+      "step": 22722
+    },
+    {
+      "epoch": 0.5408692754451109,
+      "grad_norm": 3.104778344118601,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 22723
+    },
+    {
+      "epoch": 0.5408930781681425,
+      "grad_norm": 3.1560689520069753,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 22724
+    },
+    {
+      "epoch": 0.540916880891174,
+      "grad_norm": 5.36157268571923,
+      "learning_rate": 2e-05,
+      "loss": 0.9496,
+      "step": 22725
+    },
+    {
+      "epoch": 0.5409406836142054,
+      "grad_norm": 4.7950233628225645,
+      "learning_rate": 2e-05,
+      "loss": 0.7662,
+      "step": 22726
+    },
+    {
+      "epoch": 0.5409644863372369,
+      "grad_norm": 6.649502378071193,
+      "learning_rate": 2e-05,
+      "loss": 0.977,
+      "step": 22727
+    },
+    {
+      "epoch": 0.5409882890602685,
+      "grad_norm": 1.546208672600596,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 22728
+    },
+    {
+      "epoch": 0.5410120917833,
+      "grad_norm": 0.841450870969653,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 22729
+    },
+    {
+      "epoch": 0.5410358945063315,
+      "grad_norm": 8.765747719750745,
+      "learning_rate": 2e-05,
+      "loss": 0.3889,
+      "step": 22730
+    },
+    {
+      "epoch": 0.5410596972293631,
+      "grad_norm": 3.760216428034427,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 22731
+    },
+    {
+      "epoch": 0.5410834999523946,
+      "grad_norm": 0.5014660699654875,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 22732
+    },
+    {
+      "epoch": 0.5411073026754261,
+      "grad_norm": 5.590667647963874,
+      "learning_rate": 2e-05,
+      "loss": 0.6467,
+      "step": 22733
+    },
+    {
+      "epoch": 0.5411311053984575,
+      "grad_norm": 1.0831773727442162,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 22734
+    },
+    {
+      "epoch": 0.5411549081214891,
+      "grad_norm": 3.6952357385413306,
+      "learning_rate": 2e-05,
+      "loss": 0.7097,
+      "step": 22735
+    },
+    {
+      "epoch": 0.5411787108445206,
+      "grad_norm": 4.9325153595246825,
+      "learning_rate": 2e-05,
+      "loss": 0.5551,
+      "step": 22736
+    },
+    {
+      "epoch": 0.5412025135675521,
+      "grad_norm": 3.7553799731073,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 22737
+    },
+    {
+      "epoch": 0.5412263162905836,
+      "grad_norm": 0.03317037875232765,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 22738
+    },
+    {
+      "epoch": 0.5412501190136152,
+      "grad_norm": 0.5984768956806954,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 22739
+    },
+    {
+      "epoch": 0.5412739217366467,
+      "grad_norm": 3.1795250595734315,
+      "learning_rate": 2e-05,
+      "loss": 0.2747,
+      "step": 22740
+    },
+    {
+      "epoch": 0.5412977244596782,
+      "grad_norm": 13.933650092826078,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 22741
+    },
+    {
+      "epoch": 0.5413215271827096,
+      "grad_norm": 0.5195153793124765,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 22742
+    },
+    {
+      "epoch": 0.5413453299057412,
+      "grad_norm": 4.503237538191284,
+      "learning_rate": 2e-05,
+      "loss": 1.0594,
+      "step": 22743
+    },
+    {
+      "epoch": 0.5413691326287727,
+      "grad_norm": 4.54288687991474,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 22744
+    },
+    {
+      "epoch": 0.5413929353518042,
+      "grad_norm": 4.498953613398091,
+      "learning_rate": 2e-05,
+      "loss": 0.8005,
+      "step": 22745
+    },
+    {
+      "epoch": 0.5414167380748358,
+      "grad_norm": 1.8913783884089197,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 22746
+    },
+    {
+      "epoch": 0.5414405407978673,
+      "grad_norm": 1.0619481158948636,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 22747
+    },
+    {
+      "epoch": 0.5414643435208988,
+      "grad_norm": 2.3528680902614276,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 22748
+    },
+    {
+      "epoch": 0.5414881462439303,
+      "grad_norm": 4.702705545730052,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 22749
+    },
+    {
+      "epoch": 0.5415119489669619,
+      "grad_norm": 9.467233947682788,
+      "learning_rate": 2e-05,
+      "loss": 0.657,
+      "step": 22750
+    },
+    {
+      "epoch": 0.5415357516899933,
+      "grad_norm": 0.9202737919686199,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 22751
+    },
+    {
+      "epoch": 0.5415595544130248,
+      "grad_norm": 7.463752703797694,
+      "learning_rate": 2e-05,
+      "loss": 1.3948,
+      "step": 22752
+    },
+    {
+      "epoch": 0.5415833571360563,
+      "grad_norm": 0.8382353538774121,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 22753
+    },
+    {
+      "epoch": 0.5416071598590879,
+      "grad_norm": 12.646640985408295,
+      "learning_rate": 2e-05,
+      "loss": 0.4243,
+      "step": 22754
+    },
+    {
+      "epoch": 0.5416309625821194,
+      "grad_norm": 2.9568676360031616,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 22755
+    },
+    {
+      "epoch": 0.5416547653051509,
+      "grad_norm": 0.42040031338734823,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 22756
+    },
+    {
+      "epoch": 0.5416785680281824,
+      "grad_norm": 2.773918126167004,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 22757
+    },
+    {
+      "epoch": 0.541702370751214,
+      "grad_norm": 0.37652062141034104,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 22758
+    },
+    {
+      "epoch": 0.5417261734742455,
+      "grad_norm": 7.49798053130873,
+      "learning_rate": 2e-05,
+      "loss": 0.305,
+      "step": 22759
+    },
+    {
+      "epoch": 0.5417499761972769,
+      "grad_norm": 1.7394281325899756,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 22760
+    },
+    {
+      "epoch": 0.5417737789203085,
+      "grad_norm": 3.299126348554114,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 22761
+    },
+    {
+      "epoch": 0.54179758164334,
+      "grad_norm": 3.1007877185028736,
+      "learning_rate": 2e-05,
+      "loss": 0.3715,
+      "step": 22762
+    },
+    {
+      "epoch": 0.5418213843663715,
+      "grad_norm": 0.24225542784909337,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 22763
+    },
+    {
+      "epoch": 0.541845187089403,
+      "grad_norm": 4.111796943341481,
+      "learning_rate": 2e-05,
+      "loss": 0.5746,
+      "step": 22764
+    },
+    {
+      "epoch": 0.5418689898124346,
+      "grad_norm": 3.1381808682739014,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 22765
+    },
+    {
+      "epoch": 0.5418927925354661,
+      "grad_norm": 3.9280691435300743,
+      "learning_rate": 2e-05,
+      "loss": 0.5035,
+      "step": 22766
+    },
+    {
+      "epoch": 0.5419165952584976,
+      "grad_norm": 5.781253931033523,
+      "learning_rate": 2e-05,
+      "loss": 0.3834,
+      "step": 22767
+    },
+    {
+      "epoch": 0.541940397981529,
+      "grad_norm": 4.7333187991658585,
+      "learning_rate": 2e-05,
+      "loss": 0.8982,
+      "step": 22768
+    },
+    {
+      "epoch": 0.5419642007045606,
+      "grad_norm": 0.11011699821282535,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 22769
+    },
+    {
+      "epoch": 0.5419880034275921,
+      "grad_norm": 8.868498732876407,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 22770
+    },
+    {
+      "epoch": 0.5420118061506236,
+      "grad_norm": 0.23198032859394735,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 22771
+    },
+    {
+      "epoch": 0.5420356088736551,
+      "grad_norm": 6.734835046797357,
+      "learning_rate": 2e-05,
+      "loss": 0.8529,
+      "step": 22772
+    },
+    {
+      "epoch": 0.5420594115966867,
+      "grad_norm": 1.312770612858699,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 22773
+    },
+    {
+      "epoch": 0.5420832143197182,
+      "grad_norm": 1.3467444397613613,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 22774
+    },
+    {
+      "epoch": 0.5421070170427497,
+      "grad_norm": 0.5791898839569225,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 22775
+    },
+    {
+      "epoch": 0.5421308197657813,
+      "grad_norm": 2.096935967406542,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 22776
+    },
+    {
+      "epoch": 0.5421546224888127,
+      "grad_norm": 8.266443630899335,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 22777
+    },
+    {
+      "epoch": 0.5421784252118442,
+      "grad_norm": 6.882278870291539,
+      "learning_rate": 2e-05,
+      "loss": 0.3485,
+      "step": 22778
+    },
+    {
+      "epoch": 0.5422022279348757,
+      "grad_norm": 0.4616542076923037,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22779
+    },
+    {
+      "epoch": 0.5422260306579073,
+      "grad_norm": 3.039204221585457,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 22780
+    },
+    {
+      "epoch": 0.5422498333809388,
+      "grad_norm": 0.5588137074957449,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 22781
+    },
+    {
+      "epoch": 0.5422736361039703,
+      "grad_norm": 3.816028097982074,
+      "learning_rate": 2e-05,
+      "loss": 0.5219,
+      "step": 22782
+    },
+    {
+      "epoch": 0.5422974388270018,
+      "grad_norm": 15.378756135316122,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 22783
+    },
+    {
+      "epoch": 0.5423212415500334,
+      "grad_norm": 0.9763481318234418,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 22784
+    },
+    {
+      "epoch": 0.5423450442730648,
+      "grad_norm": 5.064403235562604,
+      "learning_rate": 2e-05,
+      "loss": 0.9507,
+      "step": 22785
+    },
+    {
+      "epoch": 0.5423688469960963,
+      "grad_norm": 1.7708392639343007,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 22786
+    },
+    {
+      "epoch": 0.5423926497191278,
+      "grad_norm": 6.025422963906845,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 22787
+    },
+    {
+      "epoch": 0.5424164524421594,
+      "grad_norm": 2.8549167752646234,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 22788
+    },
+    {
+      "epoch": 0.5424402551651909,
+      "grad_norm": 4.86396562089407,
+      "learning_rate": 2e-05,
+      "loss": 0.5548,
+      "step": 22789
+    },
+    {
+      "epoch": 0.5424640578882224,
+      "grad_norm": 4.432138884965005,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 22790
+    },
+    {
+      "epoch": 0.542487860611254,
+      "grad_norm": 7.300724945625989,
+      "learning_rate": 2e-05,
+      "loss": 1.5561,
+      "step": 22791
+    },
+    {
+      "epoch": 0.5425116633342855,
+      "grad_norm": 0.7318777743666433,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 22792
+    },
+    {
+      "epoch": 0.542535466057317,
+      "grad_norm": 0.024324586134342567,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 22793
+    },
+    {
+      "epoch": 0.5425592687803484,
+      "grad_norm": 11.202083541388086,
+      "learning_rate": 2e-05,
+      "loss": 0.5689,
+      "step": 22794
+    },
+    {
+      "epoch": 0.54258307150338,
+      "grad_norm": 3.778464050718965,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 22795
+    },
+    {
+      "epoch": 0.5426068742264115,
+      "grad_norm": 1.9624736873974262,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 22796
+    },
+    {
+      "epoch": 0.542630676949443,
+      "grad_norm": 0.9618316422734129,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 22797
+    },
+    {
+      "epoch": 0.5426544796724745,
+      "grad_norm": 6.40119597918536,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 22798
+    },
+    {
+      "epoch": 0.5426782823955061,
+      "grad_norm": 0.18245845396541252,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22799
+    },
+    {
+      "epoch": 0.5427020851185376,
+      "grad_norm": 3.4366088875840983,
+      "learning_rate": 2e-05,
+      "loss": 0.3866,
+      "step": 22800
+    },
+    {
+      "epoch": 0.542725887841569,
+      "grad_norm": 2.191881070019364,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 22801
+    },
+    {
+      "epoch": 0.5427496905646005,
+      "grad_norm": 9.685253922309007,
+      "learning_rate": 2e-05,
+      "loss": 0.6271,
+      "step": 22802
+    },
+    {
+      "epoch": 0.5427734932876321,
+      "grad_norm": 4.446202447696132,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 22803
+    },
+    {
+      "epoch": 0.5427972960106636,
+      "grad_norm": 1.159263871109425,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 22804
+    },
+    {
+      "epoch": 0.5428210987336951,
+      "grad_norm": 3.0334427059381817,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 22805
+    },
+    {
+      "epoch": 0.5428449014567267,
+      "grad_norm": 0.3799768374319729,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 22806
+    },
+    {
+      "epoch": 0.5428687041797582,
+      "grad_norm": 0.5411949264417681,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 22807
+    },
+    {
+      "epoch": 0.5428925069027897,
+      "grad_norm": 0.06772736450912573,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 22808
+    },
+    {
+      "epoch": 0.5429163096258212,
+      "grad_norm": 7.4254026459868605,
+      "learning_rate": 2e-05,
+      "loss": 0.3883,
+      "step": 22809
+    },
+    {
+      "epoch": 0.5429401123488528,
+      "grad_norm": 3.027701299496457,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 22810
+    },
+    {
+      "epoch": 0.5429639150718842,
+      "grad_norm": 3.7824570844137217,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 22811
+    },
+    {
+      "epoch": 0.5429877177949157,
+      "grad_norm": 7.292902742582227,
+      "learning_rate": 2e-05,
+      "loss": 1.1198,
+      "step": 22812
+    },
+    {
+      "epoch": 0.5430115205179472,
+      "grad_norm": 4.583535178103057,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 22813
+    },
+    {
+      "epoch": 0.5430353232409788,
+      "grad_norm": 7.602211830333275,
+      "learning_rate": 2e-05,
+      "loss": 0.2261,
+      "step": 22814
+    },
+    {
+      "epoch": 0.5430591259640103,
+      "grad_norm": 0.5189625534143237,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 22815
+    },
+    {
+      "epoch": 0.5430829286870418,
+      "grad_norm": 3.3056932663426375,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 22816
+    },
+    {
+      "epoch": 0.5431067314100733,
+      "grad_norm": 3.1581450787703793,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 22817
+    },
+    {
+      "epoch": 0.5431305341331049,
+      "grad_norm": 0.11975072576875835,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 22818
+    },
+    {
+      "epoch": 0.5431543368561363,
+      "grad_norm": 4.117597413475315,
+      "learning_rate": 2e-05,
+      "loss": 0.1705,
+      "step": 22819
+    },
+    {
+      "epoch": 0.5431781395791678,
+      "grad_norm": 0.9389908815865875,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 22820
+    },
+    {
+      "epoch": 0.5432019423021994,
+      "grad_norm": 5.065225983130392,
+      "learning_rate": 2e-05,
+      "loss": 0.6184,
+      "step": 22821
+    },
+    {
+      "epoch": 0.5432257450252309,
+      "grad_norm": 3.5753845452623545,
+      "learning_rate": 2e-05,
+      "loss": 0.6934,
+      "step": 22822
+    },
+    {
+      "epoch": 0.5432495477482624,
+      "grad_norm": 3.694151155681388,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 22823
+    },
+    {
+      "epoch": 0.5432733504712939,
+      "grad_norm": 3.8647264779114994,
+      "learning_rate": 2e-05,
+      "loss": 0.7076,
+      "step": 22824
+    },
+    {
+      "epoch": 0.5432971531943255,
+      "grad_norm": 5.73462272957713,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 22825
+    },
+    {
+      "epoch": 0.543320955917357,
+      "grad_norm": 3.081665458317681,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 22826
+    },
+    {
+      "epoch": 0.5433447586403884,
+      "grad_norm": 3.566403683016914,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 22827
+    },
+    {
+      "epoch": 0.5433685613634199,
+      "grad_norm": 11.637692877671661,
+      "learning_rate": 2e-05,
+      "loss": 0.8327,
+      "step": 22828
+    },
+    {
+      "epoch": 0.5433923640864515,
+      "grad_norm": 5.555000733291288,
+      "learning_rate": 2e-05,
+      "loss": 0.8609,
+      "step": 22829
+    },
+    {
+      "epoch": 0.543416166809483,
+      "grad_norm": 0.7133178003570821,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 22830
+    },
+    {
+      "epoch": 0.5434399695325145,
+      "grad_norm": 4.1506050412743,
+      "learning_rate": 2e-05,
+      "loss": 0.5232,
+      "step": 22831
+    },
+    {
+      "epoch": 0.543463772255546,
+      "grad_norm": 7.181422334907572,
+      "learning_rate": 2e-05,
+      "loss": 1.506,
+      "step": 22832
+    },
+    {
+      "epoch": 0.5434875749785776,
+      "grad_norm": 0.5238920215243729,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 22833
+    },
+    {
+      "epoch": 0.5435113777016091,
+      "grad_norm": 6.325836607844904,
+      "learning_rate": 2e-05,
+      "loss": 1.0018,
+      "step": 22834
+    },
+    {
+      "epoch": 0.5435351804246406,
+      "grad_norm": 4.6225487215808885,
+      "learning_rate": 2e-05,
+      "loss": 0.5203,
+      "step": 22835
+    },
+    {
+      "epoch": 0.5435589831476721,
+      "grad_norm": 5.944456248927425,
+      "learning_rate": 2e-05,
+      "loss": 1.0872,
+      "step": 22836
+    },
+    {
+      "epoch": 0.5435827858707036,
+      "grad_norm": 13.146949140686315,
+      "learning_rate": 2e-05,
+      "loss": 0.6052,
+      "step": 22837
+    },
+    {
+      "epoch": 0.5436065885937351,
+      "grad_norm": 4.565138266419746,
+      "learning_rate": 2e-05,
+      "loss": 0.6014,
+      "step": 22838
+    },
+    {
+      "epoch": 0.5436303913167666,
+      "grad_norm": 5.744545087570509,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 22839
+    },
+    {
+      "epoch": 0.5436541940397982,
+      "grad_norm": 2.803983955852645,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 22840
+    },
+    {
+      "epoch": 0.5436779967628297,
+      "grad_norm": 4.104896758640713,
+      "learning_rate": 2e-05,
+      "loss": 0.8664,
+      "step": 22841
+    },
+    {
+      "epoch": 0.5437017994858612,
+      "grad_norm": 3.6411360686715626,
+      "learning_rate": 2e-05,
+      "loss": 0.6871,
+      "step": 22842
+    },
+    {
+      "epoch": 0.5437256022088927,
+      "grad_norm": 0.6749599748464141,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 22843
+    },
+    {
+      "epoch": 0.5437494049319243,
+      "grad_norm": 4.6764624004610145,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 22844
+    },
+    {
+      "epoch": 0.5437732076549557,
+      "grad_norm": 7.268462610611024,
+      "learning_rate": 2e-05,
+      "loss": 0.3259,
+      "step": 22845
+    },
+    {
+      "epoch": 0.5437970103779872,
+      "grad_norm": 1.6976734412035053,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 22846
+    },
+    {
+      "epoch": 0.5438208131010187,
+      "grad_norm": 4.252367167345574,
+      "learning_rate": 2e-05,
+      "loss": 0.4714,
+      "step": 22847
+    },
+    {
+      "epoch": 0.5438446158240503,
+      "grad_norm": 0.4466326088939834,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 22848
+    },
+    {
+      "epoch": 0.5438684185470818,
+      "grad_norm": 3.04226390674473,
+      "learning_rate": 2e-05,
+      "loss": 0.2616,
+      "step": 22849
+    },
+    {
+      "epoch": 0.5438922212701133,
+      "grad_norm": 6.411218441053391,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 22850
+    },
+    {
+      "epoch": 0.5439160239931449,
+      "grad_norm": 6.438819396404457,
+      "learning_rate": 2e-05,
+      "loss": 1.192,
+      "step": 22851
+    },
+    {
+      "epoch": 0.5439398267161764,
+      "grad_norm": 0.31950755750695736,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 22852
+    },
+    {
+      "epoch": 0.5439636294392078,
+      "grad_norm": 3.4462390239195697,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 22853
+    },
+    {
+      "epoch": 0.5439874321622393,
+      "grad_norm": 6.845059111051776,
+      "learning_rate": 2e-05,
+      "loss": 0.6656,
+      "step": 22854
+    },
+    {
+      "epoch": 0.5440112348852709,
+      "grad_norm": 0.6933714811701668,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 22855
+    },
+    {
+      "epoch": 0.5440350376083024,
+      "grad_norm": 4.71250424757685,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 22856
+    },
+    {
+      "epoch": 0.5440588403313339,
+      "grad_norm": 4.623522248439793,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 22857
+    },
+    {
+      "epoch": 0.5440826430543654,
+      "grad_norm": 120.22906754079094,
+      "learning_rate": 2e-05,
+      "loss": 0.8463,
+      "step": 22858
+    },
+    {
+      "epoch": 0.544106445777397,
+      "grad_norm": 2.0743865250762767,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 22859
+    },
+    {
+      "epoch": 0.5441302485004285,
+      "grad_norm": 2.6685153375406814,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 22860
+    },
+    {
+      "epoch": 0.54415405122346,
+      "grad_norm": 7.152166048155829,
+      "learning_rate": 2e-05,
+      "loss": 0.7348,
+      "step": 22861
+    },
+    {
+      "epoch": 0.5441778539464914,
+      "grad_norm": 8.94767610103589,
+      "learning_rate": 2e-05,
+      "loss": 0.4535,
+      "step": 22862
+    },
+    {
+      "epoch": 0.544201656669523,
+      "grad_norm": 0.4243747880834284,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 22863
+    },
+    {
+      "epoch": 0.5442254593925545,
+      "grad_norm": 3.7661850785444226,
+      "learning_rate": 2e-05,
+      "loss": 0.5021,
+      "step": 22864
+    },
+    {
+      "epoch": 0.544249262115586,
+      "grad_norm": 9.538057621589013,
+      "learning_rate": 2e-05,
+      "loss": 0.556,
+      "step": 22865
+    },
+    {
+      "epoch": 0.5442730648386176,
+      "grad_norm": 4.908618714241924,
+      "learning_rate": 2e-05,
+      "loss": 0.5927,
+      "step": 22866
+    },
+    {
+      "epoch": 0.5442968675616491,
+      "grad_norm": 3.274949723927061,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 22867
+    },
+    {
+      "epoch": 0.5443206702846806,
+      "grad_norm": 9.40855540049351,
+      "learning_rate": 2e-05,
+      "loss": 1.5466,
+      "step": 22868
+    },
+    {
+      "epoch": 0.544344473007712,
+      "grad_norm": 6.696475785289463,
+      "learning_rate": 2e-05,
+      "loss": 0.937,
+      "step": 22869
+    },
+    {
+      "epoch": 0.5443682757307436,
+      "grad_norm": 4.119789090276273,
+      "learning_rate": 2e-05,
+      "loss": 0.6101,
+      "step": 22870
+    },
+    {
+      "epoch": 0.5443920784537751,
+      "grad_norm": 4.952123530046138,
+      "learning_rate": 2e-05,
+      "loss": 0.7399,
+      "step": 22871
+    },
+    {
+      "epoch": 0.5444158811768066,
+      "grad_norm": 5.702744771402113,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 22872
+    },
+    {
+      "epoch": 0.5444396838998381,
+      "grad_norm": 5.117972560806496,
+      "learning_rate": 2e-05,
+      "loss": 0.9636,
+      "step": 22873
+    },
+    {
+      "epoch": 0.5444634866228697,
+      "grad_norm": 3.9032640149069735,
+      "learning_rate": 2e-05,
+      "loss": 0.771,
+      "step": 22874
+    },
+    {
+      "epoch": 0.5444872893459012,
+      "grad_norm": 0.07515198908195544,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 22875
+    },
+    {
+      "epoch": 0.5445110920689327,
+      "grad_norm": 0.4995809218890666,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 22876
+    },
+    {
+      "epoch": 0.5445348947919642,
+      "grad_norm": 3.67728162669214,
+      "learning_rate": 2e-05,
+      "loss": 0.7713,
+      "step": 22877
+    },
+    {
+      "epoch": 0.5445586975149957,
+      "grad_norm": 13.594059039026835,
+      "learning_rate": 2e-05,
+      "loss": 0.1631,
+      "step": 22878
+    },
+    {
+      "epoch": 0.5445825002380272,
+      "grad_norm": 3.416429693641402,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 22879
+    },
+    {
+      "epoch": 0.5446063029610587,
+      "grad_norm": 2.9635990576263347,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 22880
+    },
+    {
+      "epoch": 0.5446301056840903,
+      "grad_norm": 2.880319956991559,
+      "learning_rate": 2e-05,
+      "loss": 0.4306,
+      "step": 22881
+    },
+    {
+      "epoch": 0.5446539084071218,
+      "grad_norm": 1.469850136206285,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 22882
+    },
+    {
+      "epoch": 0.5446777111301533,
+      "grad_norm": 6.902076091789557,
+      "learning_rate": 2e-05,
+      "loss": 0.1243,
+      "step": 22883
+    },
+    {
+      "epoch": 0.5447015138531848,
+      "grad_norm": 3.860384462304338,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 22884
+    },
+    {
+      "epoch": 0.5447253165762164,
+      "grad_norm": 5.9375539865019515,
+      "learning_rate": 2e-05,
+      "loss": 0.2165,
+      "step": 22885
+    },
+    {
+      "epoch": 0.5447491192992479,
+      "grad_norm": 0.9930915717472045,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 22886
+    },
+    {
+      "epoch": 0.5447729220222793,
+      "grad_norm": 5.2255508794914425,
+      "learning_rate": 2e-05,
+      "loss": 0.4797,
+      "step": 22887
+    },
+    {
+      "epoch": 0.5447967247453108,
+      "grad_norm": 2.2749847849698197,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 22888
+    },
+    {
+      "epoch": 0.5448205274683424,
+      "grad_norm": 1.6887378559223298,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 22889
+    },
+    {
+      "epoch": 0.5448443301913739,
+      "grad_norm": 0.40981922251310615,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 22890
+    },
+    {
+      "epoch": 0.5448681329144054,
+      "grad_norm": 0.11050135469911034,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 22891
+    },
+    {
+      "epoch": 0.5448919356374369,
+      "grad_norm": 6.341928070549442,
+      "learning_rate": 2e-05,
+      "loss": 0.4757,
+      "step": 22892
+    },
+    {
+      "epoch": 0.5449157383604685,
+      "grad_norm": 4.331706026832419,
+      "learning_rate": 2e-05,
+      "loss": 0.8319,
+      "step": 22893
+    },
+    {
+      "epoch": 0.5449395410835,
+      "grad_norm": 2.617876625042407,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 22894
+    },
+    {
+      "epoch": 0.5449633438065314,
+      "grad_norm": 3.63363144667206,
+      "learning_rate": 2e-05,
+      "loss": 0.2284,
+      "step": 22895
+    },
+    {
+      "epoch": 0.5449871465295629,
+      "grad_norm": 2.240599817587148,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 22896
+    },
+    {
+      "epoch": 0.5450109492525945,
+      "grad_norm": 0.7751871901682813,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 22897
+    },
+    {
+      "epoch": 0.545034751975626,
+      "grad_norm": 1.0540042988330536,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 22898
+    },
+    {
+      "epoch": 0.5450585546986575,
+      "grad_norm": 6.864383122985481,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 22899
+    },
+    {
+      "epoch": 0.5450823574216891,
+      "grad_norm": 0.255529209763541,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 22900
+    },
+    {
+      "epoch": 0.5451061601447206,
+      "grad_norm": 7.264649228630086,
+      "learning_rate": 2e-05,
+      "loss": 1.3923,
+      "step": 22901
+    },
+    {
+      "epoch": 0.5451299628677521,
+      "grad_norm": 3.0590449228698966,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 22902
+    },
+    {
+      "epoch": 0.5451537655907835,
+      "grad_norm": 1.1205291292300126,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 22903
+    },
+    {
+      "epoch": 0.5451775683138151,
+      "grad_norm": 4.6541509584526874,
+      "learning_rate": 2e-05,
+      "loss": 0.6585,
+      "step": 22904
+    },
+    {
+      "epoch": 0.5452013710368466,
+      "grad_norm": 6.239117199368821,
+      "learning_rate": 2e-05,
+      "loss": 0.2832,
+      "step": 22905
+    },
+    {
+      "epoch": 0.5452251737598781,
+      "grad_norm": 0.5780497892088745,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 22906
+    },
+    {
+      "epoch": 0.5452489764829096,
+      "grad_norm": 10.129675645703397,
+      "learning_rate": 2e-05,
+      "loss": 1.0737,
+      "step": 22907
+    },
+    {
+      "epoch": 0.5452727792059412,
+      "grad_norm": 1.5109687808542906,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 22908
+    },
+    {
+      "epoch": 0.5452965819289727,
+      "grad_norm": 1.8813057143773566,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 22909
+    },
+    {
+      "epoch": 0.5453203846520042,
+      "grad_norm": 3.8262958993641694,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 22910
+    },
+    {
+      "epoch": 0.5453441873750356,
+      "grad_norm": 6.112690101738325,
+      "learning_rate": 2e-05,
+      "loss": 0.3435,
+      "step": 22911
+    },
+    {
+      "epoch": 0.5453679900980672,
+      "grad_norm": 0.8797610105719624,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 22912
+    },
+    {
+      "epoch": 0.5453917928210987,
+      "grad_norm": 6.420699370145425,
+      "learning_rate": 2e-05,
+      "loss": 1.6707,
+      "step": 22913
+    },
+    {
+      "epoch": 0.5454155955441302,
+      "grad_norm": 0.41966389600513004,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 22914
+    },
+    {
+      "epoch": 0.5454393982671618,
+      "grad_norm": 4.2886732904668055,
+      "learning_rate": 2e-05,
+      "loss": 0.5275,
+      "step": 22915
+    },
+    {
+      "epoch": 0.5454632009901933,
+      "grad_norm": 3.2794898354965816,
+      "learning_rate": 2e-05,
+      "loss": 0.4627,
+      "step": 22916
+    },
+    {
+      "epoch": 0.5454870037132248,
+      "grad_norm": 2.3888518846714146,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 22917
+    },
+    {
+      "epoch": 0.5455108064362563,
+      "grad_norm": 4.156197457724742,
+      "learning_rate": 2e-05,
+      "loss": 0.4831,
+      "step": 22918
+    },
+    {
+      "epoch": 0.5455346091592879,
+      "grad_norm": 6.020784776816079,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 22919
+    },
+    {
+      "epoch": 0.5455584118823193,
+      "grad_norm": 4.138151116556441,
+      "learning_rate": 2e-05,
+      "loss": 0.2212,
+      "step": 22920
+    },
+    {
+      "epoch": 0.5455822146053508,
+      "grad_norm": 4.318717982925561,
+      "learning_rate": 2e-05,
+      "loss": 1.0032,
+      "step": 22921
+    },
+    {
+      "epoch": 0.5456060173283823,
+      "grad_norm": 3.364960359026367,
+      "learning_rate": 2e-05,
+      "loss": 0.5972,
+      "step": 22922
+    },
+    {
+      "epoch": 0.5456298200514139,
+      "grad_norm": 1.1413408955012874,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 22923
+    },
+    {
+      "epoch": 0.5456536227744454,
+      "grad_norm": 2.4606165493466965,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 22924
+    },
+    {
+      "epoch": 0.5456774254974769,
+      "grad_norm": 4.128132661265879,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 22925
+    },
+    {
+      "epoch": 0.5457012282205084,
+      "grad_norm": 8.011277586382459,
+      "learning_rate": 2e-05,
+      "loss": 1.1337,
+      "step": 22926
+    },
+    {
+      "epoch": 0.54572503094354,
+      "grad_norm": 1.8706969924474939,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 22927
+    },
+    {
+      "epoch": 0.5457488336665715,
+      "grad_norm": 3.280383953561626,
+      "learning_rate": 2e-05,
+      "loss": 0.4754,
+      "step": 22928
+    },
+    {
+      "epoch": 0.5457726363896029,
+      "grad_norm": 2.85817004779204,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 22929
+    },
+    {
+      "epoch": 0.5457964391126345,
+      "grad_norm": 6.498014360015154,
+      "learning_rate": 2e-05,
+      "loss": 0.9423,
+      "step": 22930
+    },
+    {
+      "epoch": 0.545820241835666,
+      "grad_norm": 3.1533774088596633,
+      "learning_rate": 2e-05,
+      "loss": 0.1431,
+      "step": 22931
+    },
+    {
+      "epoch": 0.5458440445586975,
+      "grad_norm": 4.858137648410712,
+      "learning_rate": 2e-05,
+      "loss": 0.7398,
+      "step": 22932
+    },
+    {
+      "epoch": 0.545867847281729,
+      "grad_norm": 4.276262522005873,
+      "learning_rate": 2e-05,
+      "loss": 0.468,
+      "step": 22933
+    },
+    {
+      "epoch": 0.5458916500047606,
+      "grad_norm": 0.1721377739428227,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 22934
+    },
+    {
+      "epoch": 0.5459154527277921,
+      "grad_norm": 5.597314242601323,
+      "learning_rate": 2e-05,
+      "loss": 0.7779,
+      "step": 22935
+    },
+    {
+      "epoch": 0.5459392554508236,
+      "grad_norm": 11.885909938484781,
+      "learning_rate": 2e-05,
+      "loss": 0.283,
+      "step": 22936
+    },
+    {
+      "epoch": 0.545963058173855,
+      "grad_norm": 1.0173980008887074,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 22937
+    },
+    {
+      "epoch": 0.5459868608968866,
+      "grad_norm": 2.8397027747796857,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 22938
+    },
+    {
+      "epoch": 0.5460106636199181,
+      "grad_norm": 1.3320821915760297,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 22939
+    },
+    {
+      "epoch": 0.5460344663429496,
+      "grad_norm": 1.5657788330798281,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 22940
+    },
+    {
+      "epoch": 0.5460582690659811,
+      "grad_norm": 3.876151095764906,
+      "learning_rate": 2e-05,
+      "loss": 0.5075,
+      "step": 22941
+    },
+    {
+      "epoch": 0.5460820717890127,
+      "grad_norm": 16.074148023333123,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 22942
+    },
+    {
+      "epoch": 0.5461058745120442,
+      "grad_norm": 2.82091179099325,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 22943
+    },
+    {
+      "epoch": 0.5461296772350757,
+      "grad_norm": 3.219243402644358,
+      "learning_rate": 2e-05,
+      "loss": 0.5627,
+      "step": 22944
+    },
+    {
+      "epoch": 0.5461534799581073,
+      "grad_norm": 5.272400493525886,
+      "learning_rate": 2e-05,
+      "loss": 0.6358,
+      "step": 22945
+    },
+    {
+      "epoch": 0.5461772826811387,
+      "grad_norm": 17.00395607721769,
+      "learning_rate": 2e-05,
+      "loss": 0.3407,
+      "step": 22946
+    },
+    {
+      "epoch": 0.5462010854041702,
+      "grad_norm": 3.0342834926668174,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 22947
+    },
+    {
+      "epoch": 0.5462248881272017,
+      "grad_norm": 7.504788199335373,
+      "learning_rate": 2e-05,
+      "loss": 1.1293,
+      "step": 22948
+    },
+    {
+      "epoch": 0.5462486908502333,
+      "grad_norm": 5.013449562687734,
+      "learning_rate": 2e-05,
+      "loss": 1.0712,
+      "step": 22949
+    },
+    {
+      "epoch": 0.5462724935732648,
+      "grad_norm": 4.348531917167669,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 22950
+    },
+    {
+      "epoch": 0.5462962962962963,
+      "grad_norm": 0.13097652429204884,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 22951
+    },
+    {
+      "epoch": 0.5463200990193278,
+      "grad_norm": 0.4158714560783258,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 22952
+    },
+    {
+      "epoch": 0.5463439017423594,
+      "grad_norm": 0.5073709425226396,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 22953
+    },
+    {
+      "epoch": 0.5463677044653908,
+      "grad_norm": 0.6500115369542894,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 22954
+    },
+    {
+      "epoch": 0.5463915071884223,
+      "grad_norm": 0.05599270409950952,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 22955
+    },
+    {
+      "epoch": 0.5464153099114538,
+      "grad_norm": 0.3704026531946515,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 22956
+    },
+    {
+      "epoch": 0.5464391126344854,
+      "grad_norm": 0.5360651286641452,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 22957
+    },
+    {
+      "epoch": 0.5464629153575169,
+      "grad_norm": 7.415179841460442,
+      "learning_rate": 2e-05,
+      "loss": 0.9329,
+      "step": 22958
+    },
+    {
+      "epoch": 0.5464867180805484,
+      "grad_norm": 3.699389305464514,
+      "learning_rate": 2e-05,
+      "loss": 0.4603,
+      "step": 22959
+    },
+    {
+      "epoch": 0.54651052080358,
+      "grad_norm": 0.6875652517715302,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 22960
+    },
+    {
+      "epoch": 0.5465343235266115,
+      "grad_norm": 4.559235620978137,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 22961
+    },
+    {
+      "epoch": 0.546558126249643,
+      "grad_norm": 0.6699179094102575,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 22962
+    },
+    {
+      "epoch": 0.5465819289726744,
+      "grad_norm": 2.0356834120596328,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 22963
+    },
+    {
+      "epoch": 0.546605731695706,
+      "grad_norm": 17.54058390465191,
+      "learning_rate": 2e-05,
+      "loss": 0.5826,
+      "step": 22964
+    },
+    {
+      "epoch": 0.5466295344187375,
+      "grad_norm": 19.958030031539227,
+      "learning_rate": 2e-05,
+      "loss": 0.7409,
+      "step": 22965
+    },
+    {
+      "epoch": 0.546653337141769,
+      "grad_norm": 0.2575101157043211,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 22966
+    },
+    {
+      "epoch": 0.5466771398648005,
+      "grad_norm": 6.3970626507974435,
+      "learning_rate": 2e-05,
+      "loss": 0.4448,
+      "step": 22967
+    },
+    {
+      "epoch": 0.5467009425878321,
+      "grad_norm": 1.323559996211847,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 22968
+    },
+    {
+      "epoch": 0.5467247453108636,
+      "grad_norm": 3.5653541083381812,
+      "learning_rate": 2e-05,
+      "loss": 0.5091,
+      "step": 22969
+    },
+    {
+      "epoch": 0.546748548033895,
+      "grad_norm": 5.791481197709162,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 22970
+    },
+    {
+      "epoch": 0.5467723507569265,
+      "grad_norm": 5.82946597963059,
+      "learning_rate": 2e-05,
+      "loss": 0.585,
+      "step": 22971
+    },
+    {
+      "epoch": 0.5467961534799581,
+      "grad_norm": 1.5798715820941347,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 22972
+    },
+    {
+      "epoch": 0.5468199562029896,
+      "grad_norm": 3.9626582183016814,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 22973
+    },
+    {
+      "epoch": 0.5468437589260211,
+      "grad_norm": 10.897061294919917,
+      "learning_rate": 2e-05,
+      "loss": 0.3689,
+      "step": 22974
+    },
+    {
+      "epoch": 0.5468675616490527,
+      "grad_norm": 2.441385776283207,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 22975
+    },
+    {
+      "epoch": 0.5468913643720842,
+      "grad_norm": 6.371429935568212,
+      "learning_rate": 2e-05,
+      "loss": 1.1489,
+      "step": 22976
+    },
+    {
+      "epoch": 0.5469151670951157,
+      "grad_norm": 0.8123946105095355,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 22977
+    },
+    {
+      "epoch": 0.5469389698181472,
+      "grad_norm": 0.30019589156561144,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 22978
+    },
+    {
+      "epoch": 0.5469627725411788,
+      "grad_norm": 3.9032532947893785,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 22979
+    },
+    {
+      "epoch": 0.5469865752642102,
+      "grad_norm": 4.589803265648294,
+      "learning_rate": 2e-05,
+      "loss": 0.6007,
+      "step": 22980
+    },
+    {
+      "epoch": 0.5470103779872417,
+      "grad_norm": 1.2594778226660264,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 22981
+    },
+    {
+      "epoch": 0.5470341807102732,
+      "grad_norm": 4.764107879220777,
+      "learning_rate": 2e-05,
+      "loss": 0.7023,
+      "step": 22982
+    },
+    {
+      "epoch": 0.5470579834333048,
+      "grad_norm": 3.295099056922548,
+      "learning_rate": 2e-05,
+      "loss": 0.7517,
+      "step": 22983
+    },
+    {
+      "epoch": 0.5470817861563363,
+      "grad_norm": 2.902418811461765,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 22984
+    },
+    {
+      "epoch": 0.5471055888793678,
+      "grad_norm": 0.25040355868035136,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 22985
+    },
+    {
+      "epoch": 0.5471293916023993,
+      "grad_norm": 4.812188755713862,
+      "learning_rate": 2e-05,
+      "loss": 1.1189,
+      "step": 22986
+    },
+    {
+      "epoch": 0.5471531943254309,
+      "grad_norm": 5.219328824612514,
+      "learning_rate": 2e-05,
+      "loss": 0.9115,
+      "step": 22987
+    },
+    {
+      "epoch": 0.5471769970484623,
+      "grad_norm": 0.1651976781695388,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 22988
+    },
+    {
+      "epoch": 0.5472007997714938,
+      "grad_norm": 1.6065173782665398,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 22989
+    },
+    {
+      "epoch": 0.5472246024945254,
+      "grad_norm": 3.490599198955777,
+      "learning_rate": 2e-05,
+      "loss": 0.5263,
+      "step": 22990
+    },
+    {
+      "epoch": 0.5472484052175569,
+      "grad_norm": 2.9362772173794895,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 22991
+    },
+    {
+      "epoch": 0.5472722079405884,
+      "grad_norm": 4.273147679128034,
+      "learning_rate": 2e-05,
+      "loss": 0.6776,
+      "step": 22992
+    },
+    {
+      "epoch": 0.5472960106636199,
+      "grad_norm": 2.620375947429775,
+      "learning_rate": 2e-05,
+      "loss": 0.6468,
+      "step": 22993
+    },
+    {
+      "epoch": 0.5473198133866515,
+      "grad_norm": 3.2997012694157926,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 22994
+    },
+    {
+      "epoch": 0.547343616109683,
+      "grad_norm": 7.238571984894038,
+      "learning_rate": 2e-05,
+      "loss": 1.6014,
+      "step": 22995
+    },
+    {
+      "epoch": 0.5473674188327144,
+      "grad_norm": 10.813034353525198,
+      "learning_rate": 2e-05,
+      "loss": 0.2851,
+      "step": 22996
+    },
+    {
+      "epoch": 0.5473912215557459,
+      "grad_norm": 3.0634085598689023,
+      "learning_rate": 2e-05,
+      "loss": 0.3976,
+      "step": 22997
+    },
+    {
+      "epoch": 0.5474150242787775,
+      "grad_norm": 7.801602490920998,
+      "learning_rate": 2e-05,
+      "loss": 1.187,
+      "step": 22998
+    },
+    {
+      "epoch": 0.547438827001809,
+      "grad_norm": 2.065496076524958,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 22999
+    },
+    {
+      "epoch": 0.5474626297248405,
+      "grad_norm": 7.4982363589337755,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 23000
+    },
+    {
+      "epoch": 0.547486432447872,
+      "grad_norm": 16.851418184871825,
+      "learning_rate": 2e-05,
+      "loss": 0.8548,
+      "step": 23001
+    },
+    {
+      "epoch": 0.5475102351709036,
+      "grad_norm": 2.1531391834717004,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 23002
+    },
+    {
+      "epoch": 0.5475340378939351,
+      "grad_norm": 4.514072793116891,
+      "learning_rate": 2e-05,
+      "loss": 0.9388,
+      "step": 23003
+    },
+    {
+      "epoch": 0.5475578406169666,
+      "grad_norm": 4.345525223559913,
+      "learning_rate": 2e-05,
+      "loss": 0.294,
+      "step": 23004
+    },
+    {
+      "epoch": 0.5475816433399981,
+      "grad_norm": 3.7342265837467608,
+      "learning_rate": 2e-05,
+      "loss": 0.4421,
+      "step": 23005
+    },
+    {
+      "epoch": 0.5476054460630296,
+      "grad_norm": 4.794259946691225,
+      "learning_rate": 2e-05,
+      "loss": 0.6873,
+      "step": 23006
+    },
+    {
+      "epoch": 0.5476292487860611,
+      "grad_norm": 9.550690979533504,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 23007
+    },
+    {
+      "epoch": 0.5476530515090926,
+      "grad_norm": 2.056683170748236,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 23008
+    },
+    {
+      "epoch": 0.5476768542321242,
+      "grad_norm": 1.8467888315458598,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 23009
+    },
+    {
+      "epoch": 0.5477006569551557,
+      "grad_norm": 5.378153983867599,
+      "learning_rate": 2e-05,
+      "loss": 0.7164,
+      "step": 23010
+    },
+    {
+      "epoch": 0.5477244596781872,
+      "grad_norm": 0.3418565610690071,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 23011
+    },
+    {
+      "epoch": 0.5477482624012187,
+      "grad_norm": 3.5396471477296276,
+      "learning_rate": 2e-05,
+      "loss": 0.3296,
+      "step": 23012
+    },
+    {
+      "epoch": 0.5477720651242503,
+      "grad_norm": 2.8389223022658463,
+      "learning_rate": 2e-05,
+      "loss": 0.4658,
+      "step": 23013
+    },
+    {
+      "epoch": 0.5477958678472817,
+      "grad_norm": 0.5496157543259937,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 23014
+    },
+    {
+      "epoch": 0.5478196705703132,
+      "grad_norm": 0.3437534426813805,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 23015
+    },
+    {
+      "epoch": 0.5478434732933447,
+      "grad_norm": 3.660385938050139,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 23016
+    },
+    {
+      "epoch": 0.5478672760163763,
+      "grad_norm": 4.309344665943391,
+      "learning_rate": 2e-05,
+      "loss": 0.4718,
+      "step": 23017
+    },
+    {
+      "epoch": 0.5478910787394078,
+      "grad_norm": 3.4434635590054663,
+      "learning_rate": 2e-05,
+      "loss": 0.663,
+      "step": 23018
+    },
+    {
+      "epoch": 0.5479148814624393,
+      "grad_norm": 3.0437707946455332,
+      "learning_rate": 2e-05,
+      "loss": 0.3431,
+      "step": 23019
+    },
+    {
+      "epoch": 0.5479386841854709,
+      "grad_norm": 1.4345491970631756,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 23020
+    },
+    {
+      "epoch": 0.5479624869085024,
+      "grad_norm": 1.9535945253889753,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 23021
+    },
+    {
+      "epoch": 0.5479862896315338,
+      "grad_norm": 6.407960241302218,
+      "learning_rate": 2e-05,
+      "loss": 0.4148,
+      "step": 23022
+    },
+    {
+      "epoch": 0.5480100923545653,
+      "grad_norm": 2.667031141393428,
+      "learning_rate": 2e-05,
+      "loss": 0.6141,
+      "step": 23023
+    },
+    {
+      "epoch": 0.5480338950775969,
+      "grad_norm": 7.788008007027778,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 23024
+    },
+    {
+      "epoch": 0.5480576978006284,
+      "grad_norm": 0.7693488443807288,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 23025
+    },
+    {
+      "epoch": 0.5480815005236599,
+      "grad_norm": 3.3084946406450886,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 23026
+    },
+    {
+      "epoch": 0.5481053032466914,
+      "grad_norm": 7.58050071845236,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 23027
+    },
+    {
+      "epoch": 0.548129105969723,
+      "grad_norm": 5.230518656912376,
+      "learning_rate": 2e-05,
+      "loss": 0.6804,
+      "step": 23028
+    },
+    {
+      "epoch": 0.5481529086927545,
+      "grad_norm": 4.8734762993483205,
+      "learning_rate": 2e-05,
+      "loss": 0.789,
+      "step": 23029
+    },
+    {
+      "epoch": 0.5481767114157859,
+      "grad_norm": 4.737931370430098,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 23030
+    },
+    {
+      "epoch": 0.5482005141388174,
+      "grad_norm": 0.39117585242903585,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 23031
+    },
+    {
+      "epoch": 0.548224316861849,
+      "grad_norm": 0.2809761571432911,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 23032
+    },
+    {
+      "epoch": 0.5482481195848805,
+      "grad_norm": 1.2547420743506383,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 23033
+    },
+    {
+      "epoch": 0.548271922307912,
+      "grad_norm": 5.706618665323359,
+      "learning_rate": 2e-05,
+      "loss": 0.6316,
+      "step": 23034
+    },
+    {
+      "epoch": 0.5482957250309436,
+      "grad_norm": 3.890085184442586,
+      "learning_rate": 2e-05,
+      "loss": 0.6888,
+      "step": 23035
+    },
+    {
+      "epoch": 0.5483195277539751,
+      "grad_norm": 2.753436792055915,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 23036
+    },
+    {
+      "epoch": 0.5483433304770066,
+      "grad_norm": 16.278119363406578,
+      "learning_rate": 2e-05,
+      "loss": 1.1018,
+      "step": 23037
+    },
+    {
+      "epoch": 0.548367133200038,
+      "grad_norm": 3.6290216869634264,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 23038
+    },
+    {
+      "epoch": 0.5483909359230696,
+      "grad_norm": 3.3989714415618835,
+      "learning_rate": 2e-05,
+      "loss": 0.4099,
+      "step": 23039
+    },
+    {
+      "epoch": 0.5484147386461011,
+      "grad_norm": 12.32550136588663,
+      "learning_rate": 2e-05,
+      "loss": 0.289,
+      "step": 23040
+    },
+    {
+      "epoch": 0.5484385413691326,
+      "grad_norm": 4.965033238065733,
+      "learning_rate": 2e-05,
+      "loss": 0.2145,
+      "step": 23041
+    },
+    {
+      "epoch": 0.5484623440921641,
+      "grad_norm": 4.76465225091649,
+      "learning_rate": 2e-05,
+      "loss": 0.5029,
+      "step": 23042
+    },
+    {
+      "epoch": 0.5484861468151957,
+      "grad_norm": 0.6662120797452684,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 23043
+    },
+    {
+      "epoch": 0.5485099495382272,
+      "grad_norm": 3.969122348322818,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 23044
+    },
+    {
+      "epoch": 0.5485337522612587,
+      "grad_norm": 4.69183388695638,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 23045
+    },
+    {
+      "epoch": 0.5485575549842902,
+      "grad_norm": 0.44674636040805693,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 23046
+    },
+    {
+      "epoch": 0.5485813577073217,
+      "grad_norm": 27.777670123835104,
+      "learning_rate": 2e-05,
+      "loss": 0.9304,
+      "step": 23047
+    },
+    {
+      "epoch": 0.5486051604303532,
+      "grad_norm": 2.3388001249307093,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 23048
+    },
+    {
+      "epoch": 0.5486289631533847,
+      "grad_norm": 11.435005404789054,
+      "learning_rate": 2e-05,
+      "loss": 0.8674,
+      "step": 23049
+    },
+    {
+      "epoch": 0.5486527658764163,
+      "grad_norm": 144.98397881373955,
+      "learning_rate": 2e-05,
+      "loss": 0.1775,
+      "step": 23050
+    },
+    {
+      "epoch": 0.5486765685994478,
+      "grad_norm": 0.05191805296354437,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 23051
+    },
+    {
+      "epoch": 0.5487003713224793,
+      "grad_norm": 5.70597119596175,
+      "learning_rate": 2e-05,
+      "loss": 0.5889,
+      "step": 23052
+    },
+    {
+      "epoch": 0.5487241740455108,
+      "grad_norm": 1.042868720549086,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 23053
+    },
+    {
+      "epoch": 0.5487479767685424,
+      "grad_norm": 2.6753871523616266,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 23054
+    },
+    {
+      "epoch": 0.5487717794915739,
+      "grad_norm": 3.7098184400707326,
+      "learning_rate": 2e-05,
+      "loss": 0.402,
+      "step": 23055
+    },
+    {
+      "epoch": 0.5487955822146053,
+      "grad_norm": 3.7366138612216497,
+      "learning_rate": 2e-05,
+      "loss": 0.7764,
+      "step": 23056
+    },
+    {
+      "epoch": 0.5488193849376368,
+      "grad_norm": 1.7370405321350844,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 23057
+    },
+    {
+      "epoch": 0.5488431876606684,
+      "grad_norm": 0.4765567393705987,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 23058
+    },
+    {
+      "epoch": 0.5488669903836999,
+      "grad_norm": 5.992768563727671,
+      "learning_rate": 2e-05,
+      "loss": 0.5272,
+      "step": 23059
+    },
+    {
+      "epoch": 0.5488907931067314,
+      "grad_norm": 3.3057595347412096,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 23060
+    },
+    {
+      "epoch": 0.5489145958297629,
+      "grad_norm": 6.106512672787414,
+      "learning_rate": 2e-05,
+      "loss": 0.8769,
+      "step": 23061
+    },
+    {
+      "epoch": 0.5489383985527945,
+      "grad_norm": 3.3782708166577518,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 23062
+    },
+    {
+      "epoch": 0.548962201275826,
+      "grad_norm": 3.1950760471676203,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 23063
+    },
+    {
+      "epoch": 0.5489860039988574,
+      "grad_norm": 6.213139272173055,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 23064
+    },
+    {
+      "epoch": 0.549009806721889,
+      "grad_norm": 5.945756041363739,
+      "learning_rate": 2e-05,
+      "loss": 1.0735,
+      "step": 23065
+    },
+    {
+      "epoch": 0.5490336094449205,
+      "grad_norm": 17.836251754237672,
+      "learning_rate": 2e-05,
+      "loss": 0.3032,
+      "step": 23066
+    },
+    {
+      "epoch": 0.549057412167952,
+      "grad_norm": 6.348211732257609,
+      "learning_rate": 2e-05,
+      "loss": 1.3013,
+      "step": 23067
+    },
+    {
+      "epoch": 0.5490812148909835,
+      "grad_norm": 4.89908699688917,
+      "learning_rate": 2e-05,
+      "loss": 0.5956,
+      "step": 23068
+    },
+    {
+      "epoch": 0.5491050176140151,
+      "grad_norm": 2.6136859392865053,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 23069
+    },
+    {
+      "epoch": 0.5491288203370466,
+      "grad_norm": 1.7079140152528374,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 23070
+    },
+    {
+      "epoch": 0.5491526230600781,
+      "grad_norm": 2.410537560179095,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 23071
+    },
+    {
+      "epoch": 0.5491764257831095,
+      "grad_norm": 3.784192353263621,
+      "learning_rate": 2e-05,
+      "loss": 0.2684,
+      "step": 23072
+    },
+    {
+      "epoch": 0.5492002285061411,
+      "grad_norm": 4.398306026061334,
+      "learning_rate": 2e-05,
+      "loss": 0.8989,
+      "step": 23073
+    },
+    {
+      "epoch": 0.5492240312291726,
+      "grad_norm": 0.7204096053662018,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 23074
+    },
+    {
+      "epoch": 0.5492478339522041,
+      "grad_norm": 16.901995321074082,
+      "learning_rate": 2e-05,
+      "loss": 1.1287,
+      "step": 23075
+    },
+    {
+      "epoch": 0.5492716366752356,
+      "grad_norm": 3.7826285214569255,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 23076
+    },
+    {
+      "epoch": 0.5492954393982672,
+      "grad_norm": 4.740242105513097,
+      "learning_rate": 2e-05,
+      "loss": 0.3973,
+      "step": 23077
+    },
+    {
+      "epoch": 0.5493192421212987,
+      "grad_norm": 13.19098694093717,
+      "learning_rate": 2e-05,
+      "loss": 1.0411,
+      "step": 23078
+    },
+    {
+      "epoch": 0.5493430448443302,
+      "grad_norm": 2.67051375923769,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 23079
+    },
+    {
+      "epoch": 0.5493668475673618,
+      "grad_norm": 4.055037728629029,
+      "learning_rate": 2e-05,
+      "loss": 0.8928,
+      "step": 23080
+    },
+    {
+      "epoch": 0.5493906502903932,
+      "grad_norm": 2.8309265669713635,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 23081
+    },
+    {
+      "epoch": 0.5494144530134247,
+      "grad_norm": 0.6365686170332703,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 23082
+    },
+    {
+      "epoch": 0.5494382557364562,
+      "grad_norm": 7.372455308026145,
+      "learning_rate": 2e-05,
+      "loss": 0.494,
+      "step": 23083
+    },
+    {
+      "epoch": 0.5494620584594878,
+      "grad_norm": 7.047866229028463,
+      "learning_rate": 2e-05,
+      "loss": 0.2567,
+      "step": 23084
+    },
+    {
+      "epoch": 0.5494858611825193,
+      "grad_norm": 13.552398859500684,
+      "learning_rate": 2e-05,
+      "loss": 0.5053,
+      "step": 23085
+    },
+    {
+      "epoch": 0.5495096639055508,
+      "grad_norm": 3.2430827768318053,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 23086
+    },
+    {
+      "epoch": 0.5495334666285823,
+      "grad_norm": 21.882943928289354,
+      "learning_rate": 2e-05,
+      "loss": 0.379,
+      "step": 23087
+    },
+    {
+      "epoch": 0.5495572693516139,
+      "grad_norm": 2.1529806417868182,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 23088
+    },
+    {
+      "epoch": 0.5495810720746453,
+      "grad_norm": 6.9406467868110715,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 23089
+    },
+    {
+      "epoch": 0.5496048747976768,
+      "grad_norm": 4.652564622747084,
+      "learning_rate": 2e-05,
+      "loss": 0.5674,
+      "step": 23090
+    },
+    {
+      "epoch": 0.5496286775207083,
+      "grad_norm": 1.9974547303186616,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 23091
+    },
+    {
+      "epoch": 0.5496524802437399,
+      "grad_norm": 0.31117841891894243,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 23092
+    },
+    {
+      "epoch": 0.5496762829667714,
+      "grad_norm": 1.2869130902072357,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 23093
+    },
+    {
+      "epoch": 0.5497000856898029,
+      "grad_norm": 11.127944557492407,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 23094
+    },
+    {
+      "epoch": 0.5497238884128345,
+      "grad_norm": 4.789126379885191,
+      "learning_rate": 2e-05,
+      "loss": 1.2758,
+      "step": 23095
+    },
+    {
+      "epoch": 0.549747691135866,
+      "grad_norm": 4.971332760746302,
+      "learning_rate": 2e-05,
+      "loss": 0.6793,
+      "step": 23096
+    },
+    {
+      "epoch": 0.5497714938588975,
+      "grad_norm": 3.894929865753609,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 23097
+    },
+    {
+      "epoch": 0.5497952965819289,
+      "grad_norm": 4.109738233906066,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 23098
+    },
+    {
+      "epoch": 0.5498190993049605,
+      "grad_norm": 5.335265780244381,
+      "learning_rate": 2e-05,
+      "loss": 0.7951,
+      "step": 23099
+    },
+    {
+      "epoch": 0.549842902027992,
+      "grad_norm": 2.2146907564216414,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 23100
+    },
+    {
+      "epoch": 0.5498667047510235,
+      "grad_norm": 2.649840327220498,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 23101
+    },
+    {
+      "epoch": 0.549890507474055,
+      "grad_norm": 2.5680400893472273,
+      "learning_rate": 2e-05,
+      "loss": 0.5583,
+      "step": 23102
+    },
+    {
+      "epoch": 0.5499143101970866,
+      "grad_norm": 12.549780434447742,
+      "learning_rate": 2e-05,
+      "loss": 0.8933,
+      "step": 23103
+    },
+    {
+      "epoch": 0.5499381129201181,
+      "grad_norm": 0.024240141255818342,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 23104
+    },
+    {
+      "epoch": 0.5499619156431496,
+      "grad_norm": 0.7985285315905025,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 23105
+    },
+    {
+      "epoch": 0.549985718366181,
+      "grad_norm": 8.385021740851593,
+      "learning_rate": 2e-05,
+      "loss": 0.9652,
+      "step": 23106
+    },
+    {
+      "epoch": 0.5500095210892126,
+      "grad_norm": 3.3198863290492726,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 23107
+    },
+    {
+      "epoch": 0.5500333238122441,
+      "grad_norm": 7.100753695413379,
+      "learning_rate": 2e-05,
+      "loss": 0.2801,
+      "step": 23108
+    },
+    {
+      "epoch": 0.5500571265352756,
+      "grad_norm": 12.174921856098775,
+      "learning_rate": 2e-05,
+      "loss": 0.608,
+      "step": 23109
+    },
+    {
+      "epoch": 0.5500809292583072,
+      "grad_norm": 0.03751129201195874,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 23110
+    },
+    {
+      "epoch": 0.5501047319813387,
+      "grad_norm": 4.223860433395269,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 23111
+    },
+    {
+      "epoch": 0.5501285347043702,
+      "grad_norm": 4.683873884417246,
+      "learning_rate": 2e-05,
+      "loss": 0.6849,
+      "step": 23112
+    },
+    {
+      "epoch": 0.5501523374274017,
+      "grad_norm": 1.4564027366005772,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 23113
+    },
+    {
+      "epoch": 0.5501761401504333,
+      "grad_norm": 0.07556276007446815,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 23114
+    },
+    {
+      "epoch": 0.5501999428734647,
+      "grad_norm": 9.46962280710497,
+      "learning_rate": 2e-05,
+      "loss": 0.5982,
+      "step": 23115
+    },
+    {
+      "epoch": 0.5502237455964962,
+      "grad_norm": 10.59644686204311,
+      "learning_rate": 2e-05,
+      "loss": 0.1931,
+      "step": 23116
+    },
+    {
+      "epoch": 0.5502475483195277,
+      "grad_norm": 2.9294318198182663,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 23117
+    },
+    {
+      "epoch": 0.5502713510425593,
+      "grad_norm": 9.854568671450586,
+      "learning_rate": 2e-05,
+      "loss": 0.5492,
+      "step": 23118
+    },
+    {
+      "epoch": 0.5502951537655908,
+      "grad_norm": 3.431705357023055,
+      "learning_rate": 2e-05,
+      "loss": 0.5409,
+      "step": 23119
+    },
+    {
+      "epoch": 0.5503189564886223,
+      "grad_norm": 4.917698011883207,
+      "learning_rate": 2e-05,
+      "loss": 1.2306,
+      "step": 23120
+    },
+    {
+      "epoch": 0.5503427592116538,
+      "grad_norm": 3.006151400459248,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 23121
+    },
+    {
+      "epoch": 0.5503665619346854,
+      "grad_norm": 5.232467900711973,
+      "learning_rate": 2e-05,
+      "loss": 0.8655,
+      "step": 23122
+    },
+    {
+      "epoch": 0.5503903646577168,
+      "grad_norm": 0.9949707544621322,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 23123
+    },
+    {
+      "epoch": 0.5504141673807483,
+      "grad_norm": 0.7394740284220175,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 23124
+    },
+    {
+      "epoch": 0.5504379701037799,
+      "grad_norm": 2.242632586124548,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 23125
+    },
+    {
+      "epoch": 0.5504617728268114,
+      "grad_norm": 4.755639875018195,
+      "learning_rate": 2e-05,
+      "loss": 0.2364,
+      "step": 23126
+    },
+    {
+      "epoch": 0.5504855755498429,
+      "grad_norm": 4.361659078531236,
+      "learning_rate": 2e-05,
+      "loss": 0.934,
+      "step": 23127
+    },
+    {
+      "epoch": 0.5505093782728744,
+      "grad_norm": 2.6178677583844476,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 23128
+    },
+    {
+      "epoch": 0.550533180995906,
+      "grad_norm": 9.320802458463502,
+      "learning_rate": 2e-05,
+      "loss": 0.3214,
+      "step": 23129
+    },
+    {
+      "epoch": 0.5505569837189375,
+      "grad_norm": 4.738386920031698,
+      "learning_rate": 2e-05,
+      "loss": 0.6989,
+      "step": 23130
+    },
+    {
+      "epoch": 0.550580786441969,
+      "grad_norm": 3.8419239603694035,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 23131
+    },
+    {
+      "epoch": 0.5506045891650004,
+      "grad_norm": 2.6387301348767687,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 23132
+    },
+    {
+      "epoch": 0.550628391888032,
+      "grad_norm": 0.18913595136566,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 23133
+    },
+    {
+      "epoch": 0.5506521946110635,
+      "grad_norm": 3.9024278955395446,
+      "learning_rate": 2e-05,
+      "loss": 0.8718,
+      "step": 23134
+    },
+    {
+      "epoch": 0.550675997334095,
+      "grad_norm": 6.224828933803619,
+      "learning_rate": 2e-05,
+      "loss": 1.6243,
+      "step": 23135
+    },
+    {
+      "epoch": 0.5506998000571265,
+      "grad_norm": 0.10446734307987633,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 23136
+    },
+    {
+      "epoch": 0.5507236027801581,
+      "grad_norm": 2.025253022716895,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 23137
+    },
+    {
+      "epoch": 0.5507474055031896,
+      "grad_norm": 2.857552703658718,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 23138
+    },
+    {
+      "epoch": 0.550771208226221,
+      "grad_norm": 3.2587942564397143,
+      "learning_rate": 2e-05,
+      "loss": 0.5009,
+      "step": 23139
+    },
+    {
+      "epoch": 0.5507950109492525,
+      "grad_norm": 1.314184333070164,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 23140
+    },
+    {
+      "epoch": 0.5508188136722841,
+      "grad_norm": 4.917235110757618,
+      "learning_rate": 2e-05,
+      "loss": 0.5962,
+      "step": 23141
+    },
+    {
+      "epoch": 0.5508426163953156,
+      "grad_norm": 1.1366315303161223,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 23142
+    },
+    {
+      "epoch": 0.5508664191183471,
+      "grad_norm": 1.6148484581851064,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 23143
+    },
+    {
+      "epoch": 0.5508902218413787,
+      "grad_norm": 4.257563113000133,
+      "learning_rate": 2e-05,
+      "loss": 0.5622,
+      "step": 23144
+    },
+    {
+      "epoch": 0.5509140245644102,
+      "grad_norm": 6.355709236234751,
+      "learning_rate": 2e-05,
+      "loss": 0.7074,
+      "step": 23145
+    },
+    {
+      "epoch": 0.5509378272874417,
+      "grad_norm": 3.431084873389278,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 23146
+    },
+    {
+      "epoch": 0.5509616300104732,
+      "grad_norm": 13.028385192470889,
+      "learning_rate": 2e-05,
+      "loss": 0.4695,
+      "step": 23147
+    },
+    {
+      "epoch": 0.5509854327335048,
+      "grad_norm": 0.5289601952539329,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 23148
+    },
+    {
+      "epoch": 0.5510092354565362,
+      "grad_norm": 2.962420789255025,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 23149
+    },
+    {
+      "epoch": 0.5510330381795677,
+      "grad_norm": 2.095742067697537,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 23150
+    },
+    {
+      "epoch": 0.5510568409025992,
+      "grad_norm": 5.117148736788069,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 23151
+    },
+    {
+      "epoch": 0.5510806436256308,
+      "grad_norm": 7.707870993063044,
+      "learning_rate": 2e-05,
+      "loss": 1.1756,
+      "step": 23152
+    },
+    {
+      "epoch": 0.5511044463486623,
+      "grad_norm": 2.7901674824513685,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 23153
+    },
+    {
+      "epoch": 0.5511282490716938,
+      "grad_norm": 3.9111700823773634,
+      "learning_rate": 2e-05,
+      "loss": 0.5696,
+      "step": 23154
+    },
+    {
+      "epoch": 0.5511520517947253,
+      "grad_norm": 0.08639315226794397,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 23155
+    },
+    {
+      "epoch": 0.5511758545177569,
+      "grad_norm": 7.910617170306165,
+      "learning_rate": 2e-05,
+      "loss": 1.195,
+      "step": 23156
+    },
+    {
+      "epoch": 0.5511996572407883,
+      "grad_norm": 6.1268009987072505,
+      "learning_rate": 2e-05,
+      "loss": 0.8354,
+      "step": 23157
+    },
+    {
+      "epoch": 0.5512234599638198,
+      "grad_norm": 0.30889039948804065,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 23158
+    },
+    {
+      "epoch": 0.5512472626868514,
+      "grad_norm": 5.099887768078176,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 23159
+    },
+    {
+      "epoch": 0.5512710654098829,
+      "grad_norm": 5.948603469295366,
+      "learning_rate": 2e-05,
+      "loss": 1.0681,
+      "step": 23160
+    },
+    {
+      "epoch": 0.5512948681329144,
+      "grad_norm": 3.9538817651119893,
+      "learning_rate": 2e-05,
+      "loss": 0.6905,
+      "step": 23161
+    },
+    {
+      "epoch": 0.5513186708559459,
+      "grad_norm": 23.845795234270927,
+      "learning_rate": 2e-05,
+      "loss": 0.2161,
+      "step": 23162
+    },
+    {
+      "epoch": 0.5513424735789775,
+      "grad_norm": 1.9952022866001031,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 23163
+    },
+    {
+      "epoch": 0.551366276302009,
+      "grad_norm": 5.391807758006394,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 23164
+    },
+    {
+      "epoch": 0.5513900790250404,
+      "grad_norm": 1.701318867010743,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 23165
+    },
+    {
+      "epoch": 0.5514138817480719,
+      "grad_norm": 6.238203283907894,
+      "learning_rate": 2e-05,
+      "loss": 0.5996,
+      "step": 23166
+    },
+    {
+      "epoch": 0.5514376844711035,
+      "grad_norm": 3.4683863462132516,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 23167
+    },
+    {
+      "epoch": 0.551461487194135,
+      "grad_norm": 5.277931893343518,
+      "learning_rate": 2e-05,
+      "loss": 0.6221,
+      "step": 23168
+    },
+    {
+      "epoch": 0.5514852899171665,
+      "grad_norm": 4.808199302649424,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 23169
+    },
+    {
+      "epoch": 0.551509092640198,
+      "grad_norm": 14.39322543467824,
+      "learning_rate": 2e-05,
+      "loss": 0.5757,
+      "step": 23170
+    },
+    {
+      "epoch": 0.5515328953632296,
+      "grad_norm": 4.96838289033419,
+      "learning_rate": 2e-05,
+      "loss": 1.1599,
+      "step": 23171
+    },
+    {
+      "epoch": 0.5515566980862611,
+      "grad_norm": 0.3337651878569357,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 23172
+    },
+    {
+      "epoch": 0.5515805008092926,
+      "grad_norm": 5.456964490013834,
+      "learning_rate": 2e-05,
+      "loss": 0.5138,
+      "step": 23173
+    },
+    {
+      "epoch": 0.5516043035323241,
+      "grad_norm": 0.31739584500377344,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 23174
+    },
+    {
+      "epoch": 0.5516281062553556,
+      "grad_norm": 11.936580712577202,
+      "learning_rate": 2e-05,
+      "loss": 1.0951,
+      "step": 23175
+    },
+    {
+      "epoch": 0.5516519089783871,
+      "grad_norm": 18.97879054648904,
+      "learning_rate": 2e-05,
+      "loss": 1.947,
+      "step": 23176
+    },
+    {
+      "epoch": 0.5516757117014186,
+      "grad_norm": 0.8144591941219508,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 23177
+    },
+    {
+      "epoch": 0.5516995144244502,
+      "grad_norm": 4.2108691042973,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 23178
+    },
+    {
+      "epoch": 0.5517233171474817,
+      "grad_norm": 0.9724098101252161,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 23179
+    },
+    {
+      "epoch": 0.5517471198705132,
+      "grad_norm": 4.144265828361098,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 23180
+    },
+    {
+      "epoch": 0.5517709225935447,
+      "grad_norm": 1.4964636552158908,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 23181
+    },
+    {
+      "epoch": 0.5517947253165763,
+      "grad_norm": 0.3183458615493456,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 23182
+    },
+    {
+      "epoch": 0.5518185280396077,
+      "grad_norm": 1.7804271577883959,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 23183
+    },
+    {
+      "epoch": 0.5518423307626392,
+      "grad_norm": 139.14746690235089,
+      "learning_rate": 2e-05,
+      "loss": 0.2612,
+      "step": 23184
+    },
+    {
+      "epoch": 0.5518661334856707,
+      "grad_norm": 1.5500277018741833,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 23185
+    },
+    {
+      "epoch": 0.5518899362087023,
+      "grad_norm": 4.303261614865555,
+      "learning_rate": 2e-05,
+      "loss": 0.7906,
+      "step": 23186
+    },
+    {
+      "epoch": 0.5519137389317338,
+      "grad_norm": 7.191757856602899,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 23187
+    },
+    {
+      "epoch": 0.5519375416547653,
+      "grad_norm": 0.303530123342312,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 23188
+    },
+    {
+      "epoch": 0.5519613443777969,
+      "grad_norm": 12.72756828289748,
+      "learning_rate": 2e-05,
+      "loss": 1.2158,
+      "step": 23189
+    },
+    {
+      "epoch": 0.5519851471008284,
+      "grad_norm": 102.40101969578195,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 23190
+    },
+    {
+      "epoch": 0.5520089498238598,
+      "grad_norm": 5.151186840394756,
+      "learning_rate": 2e-05,
+      "loss": 1.087,
+      "step": 23191
+    },
+    {
+      "epoch": 0.5520327525468913,
+      "grad_norm": 5.430521078750158,
+      "learning_rate": 2e-05,
+      "loss": 0.3143,
+      "step": 23192
+    },
+    {
+      "epoch": 0.5520565552699229,
+      "grad_norm": 2.5535059804849487,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 23193
+    },
+    {
+      "epoch": 0.5520803579929544,
+      "grad_norm": 4.3520435532859665,
+      "learning_rate": 2e-05,
+      "loss": 0.6118,
+      "step": 23194
+    },
+    {
+      "epoch": 0.5521041607159859,
+      "grad_norm": 14.4693114961231,
+      "learning_rate": 2e-05,
+      "loss": 0.7912,
+      "step": 23195
+    },
+    {
+      "epoch": 0.5521279634390174,
+      "grad_norm": 4.298353196389415,
+      "learning_rate": 2e-05,
+      "loss": 0.2801,
+      "step": 23196
+    },
+    {
+      "epoch": 0.552151766162049,
+      "grad_norm": 3.4400030370218295,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 23197
+    },
+    {
+      "epoch": 0.5521755688850805,
+      "grad_norm": 2.5324070332825785,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 23198
+    },
+    {
+      "epoch": 0.5521993716081119,
+      "grad_norm": 6.105732062043285,
+      "learning_rate": 2e-05,
+      "loss": 0.5381,
+      "step": 23199
+    },
+    {
+      "epoch": 0.5522231743311434,
+      "grad_norm": 4.402454628647647,
+      "learning_rate": 2e-05,
+      "loss": 0.8241,
+      "step": 23200
+    },
+    {
+      "epoch": 0.552246977054175,
+      "grad_norm": 4.023853823260271,
+      "learning_rate": 2e-05,
+      "loss": 0.4838,
+      "step": 23201
+    },
+    {
+      "epoch": 0.5522707797772065,
+      "grad_norm": 3.0900615515579557,
+      "learning_rate": 2e-05,
+      "loss": 0.5327,
+      "step": 23202
+    },
+    {
+      "epoch": 0.552294582500238,
+      "grad_norm": 1.3146007594897995,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 23203
+    },
+    {
+      "epoch": 0.5523183852232696,
+      "grad_norm": 0.19300988295953903,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 23204
+    },
+    {
+      "epoch": 0.5523421879463011,
+      "grad_norm": 0.08577990341365094,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 23205
+    },
+    {
+      "epoch": 0.5523659906693326,
+      "grad_norm": 7.643520835650165,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 23206
+    },
+    {
+      "epoch": 0.552389793392364,
+      "grad_norm": 3.9886622052987164,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 23207
+    },
+    {
+      "epoch": 0.5524135961153956,
+      "grad_norm": 1287.950402545258,
+      "learning_rate": 2e-05,
+      "loss": 1.1761,
+      "step": 23208
+    },
+    {
+      "epoch": 0.5524373988384271,
+      "grad_norm": 11.02328905401024,
+      "learning_rate": 2e-05,
+      "loss": 0.3159,
+      "step": 23209
+    },
+    {
+      "epoch": 0.5524612015614586,
+      "grad_norm": 0.3699728228040003,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 23210
+    },
+    {
+      "epoch": 0.5524850042844901,
+      "grad_norm": 8.306797379828236,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 23211
+    },
+    {
+      "epoch": 0.5525088070075217,
+      "grad_norm": 3.4076622255824205,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 23212
+    },
+    {
+      "epoch": 0.5525326097305532,
+      "grad_norm": 3.4047928569253307,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 23213
+    },
+    {
+      "epoch": 0.5525564124535847,
+      "grad_norm": 2.210684851939396,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 23214
+    },
+    {
+      "epoch": 0.5525802151766162,
+      "grad_norm": 1.861674880755389,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 23215
+    },
+    {
+      "epoch": 0.5526040178996477,
+      "grad_norm": 0.8143709479645984,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 23216
+    },
+    {
+      "epoch": 0.5526278206226792,
+      "grad_norm": 3.3415026029723878,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 23217
+    },
+    {
+      "epoch": 0.5526516233457107,
+      "grad_norm": 5.011624308392792,
+      "learning_rate": 2e-05,
+      "loss": 0.5105,
+      "step": 23218
+    },
+    {
+      "epoch": 0.5526754260687423,
+      "grad_norm": 19.969604100073113,
+      "learning_rate": 2e-05,
+      "loss": 0.6918,
+      "step": 23219
+    },
+    {
+      "epoch": 0.5526992287917738,
+      "grad_norm": 3.56896363380352,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 23220
+    },
+    {
+      "epoch": 0.5527230315148053,
+      "grad_norm": 6.12418021248365,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 23221
+    },
+    {
+      "epoch": 0.5527468342378368,
+      "grad_norm": 3.9297820822850427,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 23222
+    },
+    {
+      "epoch": 0.5527706369608684,
+      "grad_norm": 5.536728247008528,
+      "learning_rate": 2e-05,
+      "loss": 1.0592,
+      "step": 23223
+    },
+    {
+      "epoch": 0.5527944396838999,
+      "grad_norm": 16.88952152513957,
+      "learning_rate": 2e-05,
+      "loss": 1.5301,
+      "step": 23224
+    },
+    {
+      "epoch": 0.5528182424069313,
+      "grad_norm": 2.1700049707993463,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 23225
+    },
+    {
+      "epoch": 0.5528420451299628,
+      "grad_norm": 7.537400289596203,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 23226
+    },
+    {
+      "epoch": 0.5528658478529944,
+      "grad_norm": 0.21394474230211988,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 23227
+    },
+    {
+      "epoch": 0.5528896505760259,
+      "grad_norm": 1.6003772792181983,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 23228
+    },
+    {
+      "epoch": 0.5529134532990574,
+      "grad_norm": 3.6573297049954565,
+      "learning_rate": 2e-05,
+      "loss": 0.2813,
+      "step": 23229
+    },
+    {
+      "epoch": 0.5529372560220889,
+      "grad_norm": 4.178458041456353,
+      "learning_rate": 2e-05,
+      "loss": 0.5339,
+      "step": 23230
+    },
+    {
+      "epoch": 0.5529610587451205,
+      "grad_norm": 0.9026129778345849,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 23231
+    },
+    {
+      "epoch": 0.552984861468152,
+      "grad_norm": 2.4877659916849324,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 23232
+    },
+    {
+      "epoch": 0.5530086641911834,
+      "grad_norm": 4.551984011536978,
+      "learning_rate": 2e-05,
+      "loss": 1.094,
+      "step": 23233
+    },
+    {
+      "epoch": 0.553032466914215,
+      "grad_norm": 6.004509342227998,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 23234
+    },
+    {
+      "epoch": 0.5530562696372465,
+      "grad_norm": 0.6294938169769786,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 23235
+    },
+    {
+      "epoch": 0.553080072360278,
+      "grad_norm": 4.179441824109359,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 23236
+    },
+    {
+      "epoch": 0.5531038750833095,
+      "grad_norm": 4.171444824437364,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 23237
+    },
+    {
+      "epoch": 0.5531276778063411,
+      "grad_norm": 9.51033450216539,
+      "learning_rate": 2e-05,
+      "loss": 0.9804,
+      "step": 23238
+    },
+    {
+      "epoch": 0.5531514805293726,
+      "grad_norm": 5.822491158205117,
+      "learning_rate": 2e-05,
+      "loss": 1.5257,
+      "step": 23239
+    },
+    {
+      "epoch": 0.5531752832524041,
+      "grad_norm": 6.9055164993974465,
+      "learning_rate": 2e-05,
+      "loss": 1.1544,
+      "step": 23240
+    },
+    {
+      "epoch": 0.5531990859754355,
+      "grad_norm": 20.315540958776374,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 23241
+    },
+    {
+      "epoch": 0.5532228886984671,
+      "grad_norm": 2.231964879733817,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 23242
+    },
+    {
+      "epoch": 0.5532466914214986,
+      "grad_norm": 0.5499069024443086,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 23243
+    },
+    {
+      "epoch": 0.5532704941445301,
+      "grad_norm": 3.805354235617049,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 23244
+    },
+    {
+      "epoch": 0.5532942968675616,
+      "grad_norm": 1.3041223024714825,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 23245
+    },
+    {
+      "epoch": 0.5533180995905932,
+      "grad_norm": 1.555897072701581,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 23246
+    },
+    {
+      "epoch": 0.5533419023136247,
+      "grad_norm": 3.430100022301388,
+      "learning_rate": 2e-05,
+      "loss": 0.5584,
+      "step": 23247
+    },
+    {
+      "epoch": 0.5533657050366562,
+      "grad_norm": 1.251335146213964,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 23248
+    },
+    {
+      "epoch": 0.5533895077596878,
+      "grad_norm": 0.2222417046226638,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 23249
+    },
+    {
+      "epoch": 0.5534133104827192,
+      "grad_norm": 1.5367149114443137,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 23250
+    },
+    {
+      "epoch": 0.5534371132057507,
+      "grad_norm": 0.2315043813475682,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 23251
+    },
+    {
+      "epoch": 0.5534609159287822,
+      "grad_norm": 1.4809216865801718,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 23252
+    },
+    {
+      "epoch": 0.5534847186518138,
+      "grad_norm": 2.365529085742261,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 23253
+    },
+    {
+      "epoch": 0.5535085213748453,
+      "grad_norm": 2.9746207857244134,
+      "learning_rate": 2e-05,
+      "loss": 0.5435,
+      "step": 23254
+    },
+    {
+      "epoch": 0.5535323240978768,
+      "grad_norm": 4.90185353172304,
+      "learning_rate": 2e-05,
+      "loss": 1.1167,
+      "step": 23255
+    },
+    {
+      "epoch": 0.5535561268209083,
+      "grad_norm": 0.3065790132819926,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 23256
+    },
+    {
+      "epoch": 0.5535799295439399,
+      "grad_norm": 11.427441671114188,
+      "learning_rate": 2e-05,
+      "loss": 0.4088,
+      "step": 23257
+    },
+    {
+      "epoch": 0.5536037322669713,
+      "grad_norm": 9.441834418769043,
+      "learning_rate": 2e-05,
+      "loss": 0.3926,
+      "step": 23258
+    },
+    {
+      "epoch": 0.5536275349900028,
+      "grad_norm": 1.010981383635415,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 23259
+    },
+    {
+      "epoch": 0.5536513377130343,
+      "grad_norm": 2.26322773020599,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 23260
+    },
+    {
+      "epoch": 0.5536751404360659,
+      "grad_norm": 2.5779341256905943,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 23261
+    },
+    {
+      "epoch": 0.5536989431590974,
+      "grad_norm": 2.4509617762884046,
+      "learning_rate": 2e-05,
+      "loss": 0.282,
+      "step": 23262
+    },
+    {
+      "epoch": 0.5537227458821289,
+      "grad_norm": 0.5181542470156568,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 23263
+    },
+    {
+      "epoch": 0.5537465486051605,
+      "grad_norm": 1.0215461900036873,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 23264
+    },
+    {
+      "epoch": 0.553770351328192,
+      "grad_norm": 0.16485525533965395,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 23265
+    },
+    {
+      "epoch": 0.5537941540512235,
+      "grad_norm": 20.168592110613766,
+      "learning_rate": 2e-05,
+      "loss": 1.1203,
+      "step": 23266
+    },
+    {
+      "epoch": 0.5538179567742549,
+      "grad_norm": 2.2883561549043794,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 23267
+    },
+    {
+      "epoch": 0.5538417594972865,
+      "grad_norm": 11.405761191348462,
+      "learning_rate": 2e-05,
+      "loss": 0.6908,
+      "step": 23268
+    },
+    {
+      "epoch": 0.553865562220318,
+      "grad_norm": 9.041084496703819,
+      "learning_rate": 2e-05,
+      "loss": 0.1785,
+      "step": 23269
+    },
+    {
+      "epoch": 0.5538893649433495,
+      "grad_norm": 0.11440836254218908,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 23270
+    },
+    {
+      "epoch": 0.553913167666381,
+      "grad_norm": 0.10482262199032619,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 23271
+    },
+    {
+      "epoch": 0.5539369703894126,
+      "grad_norm": 1.0919224503893967,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 23272
+    },
+    {
+      "epoch": 0.5539607731124441,
+      "grad_norm": 0.5713822303842584,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 23273
+    },
+    {
+      "epoch": 0.5539845758354756,
+      "grad_norm": 0.04669944326754494,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 23274
+    },
+    {
+      "epoch": 0.554008378558507,
+      "grad_norm": 0.07848953144534385,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 23275
+    },
+    {
+      "epoch": 0.5540321812815386,
+      "grad_norm": 1.1325507875518106,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 23276
+    },
+    {
+      "epoch": 0.5540559840045701,
+      "grad_norm": 3.964375020982426,
+      "learning_rate": 2e-05,
+      "loss": 1.0666,
+      "step": 23277
+    },
+    {
+      "epoch": 0.5540797867276016,
+      "grad_norm": 1.57592396920974,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 23278
+    },
+    {
+      "epoch": 0.5541035894506332,
+      "grad_norm": 1.538245948639811,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 23279
+    },
+    {
+      "epoch": 0.5541273921736647,
+      "grad_norm": 3.14504148517495,
+      "learning_rate": 2e-05,
+      "loss": 0.329,
+      "step": 23280
+    },
+    {
+      "epoch": 0.5541511948966962,
+      "grad_norm": 6.85507408968312,
+      "learning_rate": 2e-05,
+      "loss": 1.38,
+      "step": 23281
+    },
+    {
+      "epoch": 0.5541749976197277,
+      "grad_norm": 0.6350121617301823,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 23282
+    },
+    {
+      "epoch": 0.5541988003427593,
+      "grad_norm": 1.4637538556233154,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 23283
+    },
+    {
+      "epoch": 0.5542226030657907,
+      "grad_norm": 2.3548608124782664,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 23284
+    },
+    {
+      "epoch": 0.5542464057888222,
+      "grad_norm": 11.232651923206408,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 23285
+    },
+    {
+      "epoch": 0.5542702085118537,
+      "grad_norm": 2.6886453646570705,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 23286
+    },
+    {
+      "epoch": 0.5542940112348853,
+      "grad_norm": 3.046217516126098,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 23287
+    },
+    {
+      "epoch": 0.5543178139579168,
+      "grad_norm": 18.147109275896163,
+      "learning_rate": 2e-05,
+      "loss": 0.8346,
+      "step": 23288
+    },
+    {
+      "epoch": 0.5543416166809483,
+      "grad_norm": 3.2281067279709226,
+      "learning_rate": 2e-05,
+      "loss": 0.4266,
+      "step": 23289
+    },
+    {
+      "epoch": 0.5543654194039798,
+      "grad_norm": 4.015308926318881,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 23290
+    },
+    {
+      "epoch": 0.5543892221270114,
+      "grad_norm": 2.611309775569212,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 23291
+    },
+    {
+      "epoch": 0.5544130248500428,
+      "grad_norm": 3.2202398420265395,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 23292
+    },
+    {
+      "epoch": 0.5544368275730743,
+      "grad_norm": 4.015675939087868,
+      "learning_rate": 2e-05,
+      "loss": 0.6406,
+      "step": 23293
+    },
+    {
+      "epoch": 0.5544606302961059,
+      "grad_norm": 2.7641435329641397,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 23294
+    },
+    {
+      "epoch": 0.5544844330191374,
+      "grad_norm": 2.9159633777439264,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 23295
+    },
+    {
+      "epoch": 0.5545082357421689,
+      "grad_norm": 0.07140639948935903,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 23296
+    },
+    {
+      "epoch": 0.5545320384652004,
+      "grad_norm": 4.467017680542117,
+      "learning_rate": 2e-05,
+      "loss": 1.0523,
+      "step": 23297
+    },
+    {
+      "epoch": 0.554555841188232,
+      "grad_norm": 9.739440201143331,
+      "learning_rate": 2e-05,
+      "loss": 0.6563,
+      "step": 23298
+    },
+    {
+      "epoch": 0.5545796439112635,
+      "grad_norm": 6.1696553285922295,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 23299
+    },
+    {
+      "epoch": 0.554603446634295,
+      "grad_norm": 4.901508706570263,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 23300
+    },
+    {
+      "epoch": 0.5546272493573264,
+      "grad_norm": 25.959216703917647,
+      "learning_rate": 2e-05,
+      "loss": 0.4137,
+      "step": 23301
+    },
+    {
+      "epoch": 0.554651052080358,
+      "grad_norm": 0.18062753392445147,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 23302
+    },
+    {
+      "epoch": 0.5546748548033895,
+      "grad_norm": 2.5913591002501244,
+      "learning_rate": 2e-05,
+      "loss": 0.4493,
+      "step": 23303
+    },
+    {
+      "epoch": 0.554698657526421,
+      "grad_norm": 4.173801479194133,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 23304
+    },
+    {
+      "epoch": 0.5547224602494525,
+      "grad_norm": 7.784746287286256,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 23305
+    },
+    {
+      "epoch": 0.5547462629724841,
+      "grad_norm": 6.441428869796042,
+      "learning_rate": 2e-05,
+      "loss": 0.4174,
+      "step": 23306
+    },
+    {
+      "epoch": 0.5547700656955156,
+      "grad_norm": 1.147864197095536,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 23307
+    },
+    {
+      "epoch": 0.554793868418547,
+      "grad_norm": 5.605755589961388,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 23308
+    },
+    {
+      "epoch": 0.5548176711415786,
+      "grad_norm": 2.5094556716832237,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 23309
+    },
+    {
+      "epoch": 0.5548414738646101,
+      "grad_norm": 3.5061120851548746,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 23310
+    },
+    {
+      "epoch": 0.5548652765876416,
+      "grad_norm": 5.397661390658169,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 23311
+    },
+    {
+      "epoch": 0.5548890793106731,
+      "grad_norm": 2.7779011711484727,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 23312
+    },
+    {
+      "epoch": 0.5549128820337047,
+      "grad_norm": 0.17549336311969438,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 23313
+    },
+    {
+      "epoch": 0.5549366847567362,
+      "grad_norm": 12.787158677410126,
+      "learning_rate": 2e-05,
+      "loss": 0.2605,
+      "step": 23314
+    },
+    {
+      "epoch": 0.5549604874797677,
+      "grad_norm": 3.905044161703645,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 23315
+    },
+    {
+      "epoch": 0.5549842902027992,
+      "grad_norm": 5.270560146429104,
+      "learning_rate": 2e-05,
+      "loss": 0.8908,
+      "step": 23316
+    },
+    {
+      "epoch": 0.5550080929258308,
+      "grad_norm": 2.5106888894023545,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 23317
+    },
+    {
+      "epoch": 0.5550318956488622,
+      "grad_norm": 1.455209682175026,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 23318
+    },
+    {
+      "epoch": 0.5550556983718937,
+      "grad_norm": 0.32203406149718294,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 23319
+    },
+    {
+      "epoch": 0.5550795010949252,
+      "grad_norm": 2.9055464498332504,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 23320
+    },
+    {
+      "epoch": 0.5551033038179568,
+      "grad_norm": 7.681315665978013,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 23321
+    },
+    {
+      "epoch": 0.5551271065409883,
+      "grad_norm": 4.352002162059325,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 23322
+    },
+    {
+      "epoch": 0.5551509092640198,
+      "grad_norm": 3.1927336853819126,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 23323
+    },
+    {
+      "epoch": 0.5551747119870514,
+      "grad_norm": 4.543248058482253,
+      "learning_rate": 2e-05,
+      "loss": 0.468,
+      "step": 23324
+    },
+    {
+      "epoch": 0.5551985147100829,
+      "grad_norm": 6.575340247766345,
+      "learning_rate": 2e-05,
+      "loss": 0.791,
+      "step": 23325
+    },
+    {
+      "epoch": 0.5552223174331143,
+      "grad_norm": 4.2219617578659685,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 23326
+    },
+    {
+      "epoch": 0.5552461201561458,
+      "grad_norm": 8.756216353491542,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 23327
+    },
+    {
+      "epoch": 0.5552699228791774,
+      "grad_norm": 2.8112846103324927,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 23328
+    },
+    {
+      "epoch": 0.5552937256022089,
+      "grad_norm": 1.4954381490242434,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 23329
+    },
+    {
+      "epoch": 0.5553175283252404,
+      "grad_norm": 2.555600815605428,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 23330
+    },
+    {
+      "epoch": 0.5553413310482719,
+      "grad_norm": 6.264375121359739,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 23331
+    },
+    {
+      "epoch": 0.5553651337713035,
+      "grad_norm": 3.7790502828492807,
+      "learning_rate": 2e-05,
+      "loss": 0.666,
+      "step": 23332
+    },
+    {
+      "epoch": 0.555388936494335,
+      "grad_norm": 3.2258828770323196,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 23333
+    },
+    {
+      "epoch": 0.5554127392173664,
+      "grad_norm": 6.009887308084463,
+      "learning_rate": 2e-05,
+      "loss": 1.0835,
+      "step": 23334
+    },
+    {
+      "epoch": 0.5554365419403979,
+      "grad_norm": 0.5048503957522336,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 23335
+    },
+    {
+      "epoch": 0.5554603446634295,
+      "grad_norm": 1.6828524227579296,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 23336
+    },
+    {
+      "epoch": 0.555484147386461,
+      "grad_norm": 6.275038903518894,
+      "learning_rate": 2e-05,
+      "loss": 1.7775,
+      "step": 23337
+    },
+    {
+      "epoch": 0.5555079501094925,
+      "grad_norm": 0.343715625770767,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 23338
+    },
+    {
+      "epoch": 0.5555317528325241,
+      "grad_norm": 8.898625322003271,
+      "learning_rate": 2e-05,
+      "loss": 0.7352,
+      "step": 23339
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.5037419903922824,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 23340
+    },
+    {
+      "epoch": 0.5555793582785871,
+      "grad_norm": 7.12786430084178,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 23341
+    },
+    {
+      "epoch": 0.5556031610016186,
+      "grad_norm": 1.3158502431320969,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 23342
+    },
+    {
+      "epoch": 0.5556269637246501,
+      "grad_norm": 4.691202341331959,
+      "learning_rate": 2e-05,
+      "loss": 0.8097,
+      "step": 23343
+    },
+    {
+      "epoch": 0.5556507664476816,
+      "grad_norm": 6.612025327804663,
+      "learning_rate": 2e-05,
+      "loss": 0.3298,
+      "step": 23344
+    },
+    {
+      "epoch": 0.5556745691707131,
+      "grad_norm": 10.273576371103644,
+      "learning_rate": 2e-05,
+      "loss": 1.4004,
+      "step": 23345
+    },
+    {
+      "epoch": 0.5556983718937446,
+      "grad_norm": 8.056746511395536,
+      "learning_rate": 2e-05,
+      "loss": 1.0093,
+      "step": 23346
+    },
+    {
+      "epoch": 0.5557221746167762,
+      "grad_norm": 3.065270193897997,
+      "learning_rate": 2e-05,
+      "loss": 0.1221,
+      "step": 23347
+    },
+    {
+      "epoch": 0.5557459773398077,
+      "grad_norm": 3.471132261500945,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 23348
+    },
+    {
+      "epoch": 0.5557697800628392,
+      "grad_norm": 0.44818027194683374,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 23349
+    },
+    {
+      "epoch": 0.5557935827858707,
+      "grad_norm": 2.8423579044437273,
+      "learning_rate": 2e-05,
+      "loss": 0.2611,
+      "step": 23350
+    },
+    {
+      "epoch": 0.5558173855089023,
+      "grad_norm": 3.8439839293417273,
+      "learning_rate": 2e-05,
+      "loss": 0.7506,
+      "step": 23351
+    },
+    {
+      "epoch": 0.5558411882319337,
+      "grad_norm": 3.1702514209510735,
+      "learning_rate": 2e-05,
+      "loss": 0.3831,
+      "step": 23352
+    },
+    {
+      "epoch": 0.5558649909549652,
+      "grad_norm": 0.9094836434594586,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 23353
+    },
+    {
+      "epoch": 0.5558887936779968,
+      "grad_norm": 0.23801761520159653,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 23354
+    },
+    {
+      "epoch": 0.5559125964010283,
+      "grad_norm": 5.489396646722751,
+      "learning_rate": 2e-05,
+      "loss": 0.6123,
+      "step": 23355
+    },
+    {
+      "epoch": 0.5559363991240598,
+      "grad_norm": 0.24677460284377728,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 23356
+    },
+    {
+      "epoch": 0.5559602018470913,
+      "grad_norm": 1.5483207080028814,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 23357
+    },
+    {
+      "epoch": 0.5559840045701229,
+      "grad_norm": 9.564799128799578,
+      "learning_rate": 2e-05,
+      "loss": 0.3342,
+      "step": 23358
+    },
+    {
+      "epoch": 0.5560078072931544,
+      "grad_norm": 1.5084231820590208,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 23359
+    },
+    {
+      "epoch": 0.5560316100161858,
+      "grad_norm": 4.121924387357893,
+      "learning_rate": 2e-05,
+      "loss": 0.5892,
+      "step": 23360
+    },
+    {
+      "epoch": 0.5560554127392173,
+      "grad_norm": 0.13500475443353877,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 23361
+    },
+    {
+      "epoch": 0.5560792154622489,
+      "grad_norm": 0.7124953351600152,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 23362
+    },
+    {
+      "epoch": 0.5561030181852804,
+      "grad_norm": 2.601884852050293,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 23363
+    },
+    {
+      "epoch": 0.5561268209083119,
+      "grad_norm": 0.3196440995561727,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 23364
+    },
+    {
+      "epoch": 0.5561506236313434,
+      "grad_norm": 30.89848216053975,
+      "learning_rate": 2e-05,
+      "loss": 1.535,
+      "step": 23365
+    },
+    {
+      "epoch": 0.556174426354375,
+      "grad_norm": 4.162568747990345,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 23366
+    },
+    {
+      "epoch": 0.5561982290774065,
+      "grad_norm": 4.186983863464344,
+      "learning_rate": 2e-05,
+      "loss": 0.4336,
+      "step": 23367
+    },
+    {
+      "epoch": 0.5562220318004379,
+      "grad_norm": 0.17635592898834707,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 23368
+    },
+    {
+      "epoch": 0.5562458345234695,
+      "grad_norm": 2.7741620900773425,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 23369
+    },
+    {
+      "epoch": 0.556269637246501,
+      "grad_norm": 4.108303398701169,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 23370
+    },
+    {
+      "epoch": 0.5562934399695325,
+      "grad_norm": 4.944528701987703,
+      "learning_rate": 2e-05,
+      "loss": 0.4288,
+      "step": 23371
+    },
+    {
+      "epoch": 0.556317242692564,
+      "grad_norm": 0.7165677925635685,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 23372
+    },
+    {
+      "epoch": 0.5563410454155956,
+      "grad_norm": 2.8701696623118247,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 23373
+    },
+    {
+      "epoch": 0.5563648481386271,
+      "grad_norm": 3.8939691547963684,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 23374
+    },
+    {
+      "epoch": 0.5563886508616586,
+      "grad_norm": 4.5679134234922225,
+      "learning_rate": 2e-05,
+      "loss": 0.6128,
+      "step": 23375
+    },
+    {
+      "epoch": 0.55641245358469,
+      "grad_norm": 36.010155621375134,
+      "learning_rate": 2e-05,
+      "loss": 0.2941,
+      "step": 23376
+    },
+    {
+      "epoch": 0.5564362563077216,
+      "grad_norm": 3.896663441955378,
+      "learning_rate": 2e-05,
+      "loss": 0.6746,
+      "step": 23377
+    },
+    {
+      "epoch": 0.5564600590307531,
+      "grad_norm": 1.8572735980278663,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 23378
+    },
+    {
+      "epoch": 0.5564838617537846,
+      "grad_norm": 0.21652462931855265,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 23379
+    },
+    {
+      "epoch": 0.5565076644768161,
+      "grad_norm": 1.3399812879472568,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 23380
+    },
+    {
+      "epoch": 0.5565314671998477,
+      "grad_norm": 3.771994467580612,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 23381
+    },
+    {
+      "epoch": 0.5565552699228792,
+      "grad_norm": 0.47575656932334387,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 23382
+    },
+    {
+      "epoch": 0.5565790726459107,
+      "grad_norm": 0.12305905883759576,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 23383
+    },
+    {
+      "epoch": 0.5566028753689422,
+      "grad_norm": 5.296628484832712,
+      "learning_rate": 2e-05,
+      "loss": 0.3271,
+      "step": 23384
+    },
+    {
+      "epoch": 0.5566266780919737,
+      "grad_norm": 4.704441905826207,
+      "learning_rate": 2e-05,
+      "loss": 0.8347,
+      "step": 23385
+    },
+    {
+      "epoch": 0.5566504808150052,
+      "grad_norm": 0.9569379614650895,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 23386
+    },
+    {
+      "epoch": 0.5566742835380367,
+      "grad_norm": 3.891719430687156,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 23387
+    },
+    {
+      "epoch": 0.5566980862610683,
+      "grad_norm": 2.2000814532892323,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 23388
+    },
+    {
+      "epoch": 0.5567218889840998,
+      "grad_norm": 1.5302930134688246,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 23389
+    },
+    {
+      "epoch": 0.5567456917071313,
+      "grad_norm": 6.416083726429619,
+      "learning_rate": 2e-05,
+      "loss": 0.6499,
+      "step": 23390
+    },
+    {
+      "epoch": 0.5567694944301628,
+      "grad_norm": 3.019109999263509,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 23391
+    },
+    {
+      "epoch": 0.5567932971531944,
+      "grad_norm": 3.2917563140672224,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 23392
+    },
+    {
+      "epoch": 0.5568170998762259,
+      "grad_norm": 4.074225277172952,
+      "learning_rate": 2e-05,
+      "loss": 0.5859,
+      "step": 23393
+    },
+    {
+      "epoch": 0.5568409025992573,
+      "grad_norm": 7.981848500791449,
+      "learning_rate": 2e-05,
+      "loss": 0.5378,
+      "step": 23394
+    },
+    {
+      "epoch": 0.5568647053222888,
+      "grad_norm": 7.079523021280061,
+      "learning_rate": 2e-05,
+      "loss": 0.9917,
+      "step": 23395
+    },
+    {
+      "epoch": 0.5568885080453204,
+      "grad_norm": 2.961470477906496,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 23396
+    },
+    {
+      "epoch": 0.5569123107683519,
+      "grad_norm": 0.6563197412872281,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 23397
+    },
+    {
+      "epoch": 0.5569361134913834,
+      "grad_norm": 4.022563454388163,
+      "learning_rate": 2e-05,
+      "loss": 0.4508,
+      "step": 23398
+    },
+    {
+      "epoch": 0.5569599162144149,
+      "grad_norm": 3.8152354709762575,
+      "learning_rate": 2e-05,
+      "loss": 0.6284,
+      "step": 23399
+    },
+    {
+      "epoch": 0.5569837189374465,
+      "grad_norm": 4.292702448492391,
+      "learning_rate": 2e-05,
+      "loss": 0.7647,
+      "step": 23400
+    },
+    {
+      "epoch": 0.557007521660478,
+      "grad_norm": 3.1618450212826943,
+      "learning_rate": 2e-05,
+      "loss": 0.2398,
+      "step": 23401
+    },
+    {
+      "epoch": 0.5570313243835094,
+      "grad_norm": 4.316406038905545,
+      "learning_rate": 2e-05,
+      "loss": 0.6372,
+      "step": 23402
+    },
+    {
+      "epoch": 0.557055127106541,
+      "grad_norm": 3.9490081596533497,
+      "learning_rate": 2e-05,
+      "loss": 0.4893,
+      "step": 23403
+    },
+    {
+      "epoch": 0.5570789298295725,
+      "grad_norm": 1.9287879037295466,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 23404
+    },
+    {
+      "epoch": 0.557102732552604,
+      "grad_norm": 4.361495126864441,
+      "learning_rate": 2e-05,
+      "loss": 0.3862,
+      "step": 23405
+    },
+    {
+      "epoch": 0.5571265352756355,
+      "grad_norm": 4.509637805481789,
+      "learning_rate": 2e-05,
+      "loss": 0.4541,
+      "step": 23406
+    },
+    {
+      "epoch": 0.5571503379986671,
+      "grad_norm": 0.15748546439195213,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 23407
+    },
+    {
+      "epoch": 0.5571741407216986,
+      "grad_norm": 3.857469118672336,
+      "learning_rate": 2e-05,
+      "loss": 0.5682,
+      "step": 23408
+    },
+    {
+      "epoch": 0.5571979434447301,
+      "grad_norm": 0.049747247938154236,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 23409
+    },
+    {
+      "epoch": 0.5572217461677615,
+      "grad_norm": 3.6389670357546935,
+      "learning_rate": 2e-05,
+      "loss": 0.4745,
+      "step": 23410
+    },
+    {
+      "epoch": 0.5572455488907931,
+      "grad_norm": 6.338759594276821,
+      "learning_rate": 2e-05,
+      "loss": 1.5299,
+      "step": 23411
+    },
+    {
+      "epoch": 0.5572693516138246,
+      "grad_norm": 3.283226032125727,
+      "learning_rate": 2e-05,
+      "loss": 0.3478,
+      "step": 23412
+    },
+    {
+      "epoch": 0.5572931543368561,
+      "grad_norm": 10.35485259478312,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 23413
+    },
+    {
+      "epoch": 0.5573169570598876,
+      "grad_norm": 9.160117332460212,
+      "learning_rate": 2e-05,
+      "loss": 0.9717,
+      "step": 23414
+    },
+    {
+      "epoch": 0.5573407597829192,
+      "grad_norm": 5.368912164368905,
+      "learning_rate": 2e-05,
+      "loss": 0.937,
+      "step": 23415
+    },
+    {
+      "epoch": 0.5573645625059507,
+      "grad_norm": 6.823819188023444,
+      "learning_rate": 2e-05,
+      "loss": 0.244,
+      "step": 23416
+    },
+    {
+      "epoch": 0.5573883652289822,
+      "grad_norm": 0.30470437607791667,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 23417
+    },
+    {
+      "epoch": 0.5574121679520138,
+      "grad_norm": 27.374901319642472,
+      "learning_rate": 2e-05,
+      "loss": 0.331,
+      "step": 23418
+    },
+    {
+      "epoch": 0.5574359706750452,
+      "grad_norm": 0.18875707496697444,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 23419
+    },
+    {
+      "epoch": 0.5574597733980767,
+      "grad_norm": 1.6037400757922422,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 23420
+    },
+    {
+      "epoch": 0.5574835761211082,
+      "grad_norm": 12.776934090119571,
+      "learning_rate": 2e-05,
+      "loss": 0.571,
+      "step": 23421
+    },
+    {
+      "epoch": 0.5575073788441398,
+      "grad_norm": 2.382070597665928,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 23422
+    },
+    {
+      "epoch": 0.5575311815671713,
+      "grad_norm": 3.7467723777352635,
+      "learning_rate": 2e-05,
+      "loss": 0.6897,
+      "step": 23423
+    },
+    {
+      "epoch": 0.5575549842902028,
+      "grad_norm": 0.32884112395973353,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 23424
+    },
+    {
+      "epoch": 0.5575787870132343,
+      "grad_norm": 2.4297382705960713,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 23425
+    },
+    {
+      "epoch": 0.5576025897362659,
+      "grad_norm": 4.5752958887721284,
+      "learning_rate": 2e-05,
+      "loss": 0.7896,
+      "step": 23426
+    },
+    {
+      "epoch": 0.5576263924592973,
+      "grad_norm": 2.4267263174774816,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 23427
+    },
+    {
+      "epoch": 0.5576501951823288,
+      "grad_norm": 0.15033962156633104,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 23428
+    },
+    {
+      "epoch": 0.5576739979053603,
+      "grad_norm": 2.0998779979367534,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 23429
+    },
+    {
+      "epoch": 0.5576978006283919,
+      "grad_norm": 6.935221306645527,
+      "learning_rate": 2e-05,
+      "loss": 0.2344,
+      "step": 23430
+    },
+    {
+      "epoch": 0.5577216033514234,
+      "grad_norm": 3.0578235084849017,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 23431
+    },
+    {
+      "epoch": 0.5577454060744549,
+      "grad_norm": 2.278080894731413,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 23432
+    },
+    {
+      "epoch": 0.5577692087974865,
+      "grad_norm": 5.684009799621356,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 23433
+    },
+    {
+      "epoch": 0.557793011520518,
+      "grad_norm": 9.483493509946161,
+      "learning_rate": 2e-05,
+      "loss": 1.1602,
+      "step": 23434
+    },
+    {
+      "epoch": 0.5578168142435495,
+      "grad_norm": 1.0386441955830696,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 23435
+    },
+    {
+      "epoch": 0.5578406169665809,
+      "grad_norm": 2.70820737882403,
+      "learning_rate": 2e-05,
+      "loss": 0.4198,
+      "step": 23436
+    },
+    {
+      "epoch": 0.5578644196896125,
+      "grad_norm": 0.36382828013020135,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 23437
+    },
+    {
+      "epoch": 0.557888222412644,
+      "grad_norm": 0.011346051413367662,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 23438
+    },
+    {
+      "epoch": 0.5579120251356755,
+      "grad_norm": 4.723444315928409,
+      "learning_rate": 2e-05,
+      "loss": 0.7285,
+      "step": 23439
+    },
+    {
+      "epoch": 0.557935827858707,
+      "grad_norm": 14.269768044261975,
+      "learning_rate": 2e-05,
+      "loss": 1.2234,
+      "step": 23440
+    },
+    {
+      "epoch": 0.5579596305817386,
+      "grad_norm": 3.0390285342399888,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 23441
+    },
+    {
+      "epoch": 0.5579834333047701,
+      "grad_norm": 10.147474562059703,
+      "learning_rate": 2e-05,
+      "loss": 0.7096,
+      "step": 23442
+    },
+    {
+      "epoch": 0.5580072360278016,
+      "grad_norm": 3.9224217806686346,
+      "learning_rate": 2e-05,
+      "loss": 0.8879,
+      "step": 23443
+    },
+    {
+      "epoch": 0.558031038750833,
+      "grad_norm": 4.19545387743886,
+      "learning_rate": 2e-05,
+      "loss": 0.6351,
+      "step": 23444
+    },
+    {
+      "epoch": 0.5580548414738646,
+      "grad_norm": 3.546343686595775,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 23445
+    },
+    {
+      "epoch": 0.5580786441968961,
+      "grad_norm": 0.41066004724250527,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 23446
+    },
+    {
+      "epoch": 0.5581024469199276,
+      "grad_norm": 3.268321625736475,
+      "learning_rate": 2e-05,
+      "loss": 0.4934,
+      "step": 23447
+    },
+    {
+      "epoch": 0.5581262496429592,
+      "grad_norm": 5.587889713703546,
+      "learning_rate": 2e-05,
+      "loss": 1.0632,
+      "step": 23448
+    },
+    {
+      "epoch": 0.5581500523659907,
+      "grad_norm": 5.444501802715968,
+      "learning_rate": 2e-05,
+      "loss": 0.757,
+      "step": 23449
+    },
+    {
+      "epoch": 0.5581738550890222,
+      "grad_norm": 2.82996808786796,
+      "learning_rate": 2e-05,
+      "loss": 0.4454,
+      "step": 23450
+    },
+    {
+      "epoch": 0.5581976578120537,
+      "grad_norm": 4.573615588575373,
+      "learning_rate": 2e-05,
+      "loss": 0.176,
+      "step": 23451
+    },
+    {
+      "epoch": 0.5582214605350853,
+      "grad_norm": 0.030200018930950964,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 23452
+    },
+    {
+      "epoch": 0.5582452632581167,
+      "grad_norm": 4.734107485892027,
+      "learning_rate": 2e-05,
+      "loss": 0.5501,
+      "step": 23453
+    },
+    {
+      "epoch": 0.5582690659811482,
+      "grad_norm": 2.6803221069057948,
+      "learning_rate": 2e-05,
+      "loss": 0.2678,
+      "step": 23454
+    },
+    {
+      "epoch": 0.5582928687041797,
+      "grad_norm": 4.875533042887185,
+      "learning_rate": 2e-05,
+      "loss": 0.7355,
+      "step": 23455
+    },
+    {
+      "epoch": 0.5583166714272113,
+      "grad_norm": 6.780498621221761,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 23456
+    },
+    {
+      "epoch": 0.5583404741502428,
+      "grad_norm": 7.372034938011253,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 23457
+    },
+    {
+      "epoch": 0.5583642768732743,
+      "grad_norm": 24.775362251894983,
+      "learning_rate": 2e-05,
+      "loss": 1.2616,
+      "step": 23458
+    },
+    {
+      "epoch": 0.5583880795963058,
+      "grad_norm": 1.408928005449738,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 23459
+    },
+    {
+      "epoch": 0.5584118823193374,
+      "grad_norm": 2.5784891700463577,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 23460
+    },
+    {
+      "epoch": 0.5584356850423688,
+      "grad_norm": 13.619396894874125,
+      "learning_rate": 2e-05,
+      "loss": 1.6844,
+      "step": 23461
+    },
+    {
+      "epoch": 0.5584594877654003,
+      "grad_norm": 0.7994753898349363,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 23462
+    },
+    {
+      "epoch": 0.5584832904884319,
+      "grad_norm": 5.454713614617456,
+      "learning_rate": 2e-05,
+      "loss": 0.9657,
+      "step": 23463
+    },
+    {
+      "epoch": 0.5585070932114634,
+      "grad_norm": 0.5347588334881862,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 23464
+    },
+    {
+      "epoch": 0.5585308959344949,
+      "grad_norm": 2.5321596480443582,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 23465
+    },
+    {
+      "epoch": 0.5585546986575264,
+      "grad_norm": 2.3891702868717433,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 23466
+    },
+    {
+      "epoch": 0.558578501380558,
+      "grad_norm": 4.021221274491137,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 23467
+    },
+    {
+      "epoch": 0.5586023041035895,
+      "grad_norm": 2.8982473794400345,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 23468
+    },
+    {
+      "epoch": 0.558626106826621,
+      "grad_norm": 0.8223965286428792,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 23469
+    },
+    {
+      "epoch": 0.5586499095496524,
+      "grad_norm": 4.579210488399868,
+      "learning_rate": 2e-05,
+      "loss": 0.654,
+      "step": 23470
+    },
+    {
+      "epoch": 0.558673712272684,
+      "grad_norm": 3.167595622875873,
+      "learning_rate": 2e-05,
+      "loss": 0.5414,
+      "step": 23471
+    },
+    {
+      "epoch": 0.5586975149957155,
+      "grad_norm": 5.731376430403825,
+      "learning_rate": 2e-05,
+      "loss": 0.3858,
+      "step": 23472
+    },
+    {
+      "epoch": 0.558721317718747,
+      "grad_norm": 7.653246722252139,
+      "learning_rate": 2e-05,
+      "loss": 0.4435,
+      "step": 23473
+    },
+    {
+      "epoch": 0.5587451204417785,
+      "grad_norm": 4.398541558785164,
+      "learning_rate": 2e-05,
+      "loss": 1.0107,
+      "step": 23474
+    },
+    {
+      "epoch": 0.5587689231648101,
+      "grad_norm": 3.6171857205579845,
+      "learning_rate": 2e-05,
+      "loss": 0.6546,
+      "step": 23475
+    },
+    {
+      "epoch": 0.5587927258878416,
+      "grad_norm": 1.489812941677011,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 23476
+    },
+    {
+      "epoch": 0.558816528610873,
+      "grad_norm": 1.3471738420402235,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 23477
+    },
+    {
+      "epoch": 0.5588403313339046,
+      "grad_norm": 0.4290328914269138,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 23478
+    },
+    {
+      "epoch": 0.5588641340569361,
+      "grad_norm": 9.144450242703103,
+      "learning_rate": 2e-05,
+      "loss": 1.1205,
+      "step": 23479
+    },
+    {
+      "epoch": 0.5588879367799676,
+      "grad_norm": 6.422702126678873,
+      "learning_rate": 2e-05,
+      "loss": 0.8966,
+      "step": 23480
+    },
+    {
+      "epoch": 0.5589117395029991,
+      "grad_norm": 0.3644053151174368,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 23481
+    },
+    {
+      "epoch": 0.5589355422260307,
+      "grad_norm": 1.7028124238154754,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 23482
+    },
+    {
+      "epoch": 0.5589593449490622,
+      "grad_norm": 9.128417704021158,
+      "learning_rate": 2e-05,
+      "loss": 1.4207,
+      "step": 23483
+    },
+    {
+      "epoch": 0.5589831476720937,
+      "grad_norm": 9.525554144004843,
+      "learning_rate": 2e-05,
+      "loss": 0.8041,
+      "step": 23484
+    },
+    {
+      "epoch": 0.5590069503951252,
+      "grad_norm": 6.434432176876926,
+      "learning_rate": 2e-05,
+      "loss": 0.5264,
+      "step": 23485
+    },
+    {
+      "epoch": 0.5590307531181568,
+      "grad_norm": 2.8733787358090406,
+      "learning_rate": 2e-05,
+      "loss": 0.6201,
+      "step": 23486
+    },
+    {
+      "epoch": 0.5590545558411882,
+      "grad_norm": 12.394847592252166,
+      "learning_rate": 2e-05,
+      "loss": 0.3449,
+      "step": 23487
+    },
+    {
+      "epoch": 0.5590783585642197,
+      "grad_norm": 0.5640065359138304,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 23488
+    },
+    {
+      "epoch": 0.5591021612872512,
+      "grad_norm": 5.14881413853883,
+      "learning_rate": 2e-05,
+      "loss": 0.5583,
+      "step": 23489
+    },
+    {
+      "epoch": 0.5591259640102828,
+      "grad_norm": 1.4823891633836852,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 23490
+    },
+    {
+      "epoch": 0.5591497667333143,
+      "grad_norm": 4.710798971554679,
+      "learning_rate": 2e-05,
+      "loss": 0.5618,
+      "step": 23491
+    },
+    {
+      "epoch": 0.5591735694563458,
+      "grad_norm": 2.303170517165596,
+      "learning_rate": 2e-05,
+      "loss": 0.1164,
+      "step": 23492
+    },
+    {
+      "epoch": 0.5591973721793774,
+      "grad_norm": 6.378442886213223,
+      "learning_rate": 2e-05,
+      "loss": 1.0726,
+      "step": 23493
+    },
+    {
+      "epoch": 0.5592211749024089,
+      "grad_norm": 0.11809118868777693,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 23494
+    },
+    {
+      "epoch": 0.5592449776254403,
+      "grad_norm": 5.814478623747566,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 23495
+    },
+    {
+      "epoch": 0.5592687803484718,
+      "grad_norm": 4.109894908798255,
+      "learning_rate": 2e-05,
+      "loss": 0.6141,
+      "step": 23496
+    },
+    {
+      "epoch": 0.5592925830715034,
+      "grad_norm": 25.390638884169423,
+      "learning_rate": 2e-05,
+      "loss": 1.2364,
+      "step": 23497
+    },
+    {
+      "epoch": 0.5593163857945349,
+      "grad_norm": 0.3461973602148789,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 23498
+    },
+    {
+      "epoch": 0.5593401885175664,
+      "grad_norm": 5.250383354196089,
+      "learning_rate": 2e-05,
+      "loss": 0.6961,
+      "step": 23499
+    },
+    {
+      "epoch": 0.5593639912405979,
+      "grad_norm": 8.600299440451126,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 23500
+    },
+    {
+      "epoch": 0.5593877939636295,
+      "grad_norm": 2.080328329899595,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 23501
+    },
+    {
+      "epoch": 0.559411596686661,
+      "grad_norm": 2.6576682895425807,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 23502
+    },
+    {
+      "epoch": 0.5594353994096924,
+      "grad_norm": 9.667633234118457,
+      "learning_rate": 2e-05,
+      "loss": 0.7629,
+      "step": 23503
+    },
+    {
+      "epoch": 0.5594592021327239,
+      "grad_norm": 11.928414003233598,
+      "learning_rate": 2e-05,
+      "loss": 0.3941,
+      "step": 23504
+    },
+    {
+      "epoch": 0.5594830048557555,
+      "grad_norm": 0.505882236274612,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 23505
+    },
+    {
+      "epoch": 0.559506807578787,
+      "grad_norm": 4.478827104075852,
+      "learning_rate": 2e-05,
+      "loss": 0.3309,
+      "step": 23506
+    },
+    {
+      "epoch": 0.5595306103018185,
+      "grad_norm": 4.461510977888284,
+      "learning_rate": 2e-05,
+      "loss": 0.4303,
+      "step": 23507
+    },
+    {
+      "epoch": 0.5595544130248501,
+      "grad_norm": 0.028819987051887087,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 23508
+    },
+    {
+      "epoch": 0.5595782157478816,
+      "grad_norm": 7.814757896672268,
+      "learning_rate": 2e-05,
+      "loss": 1.2953,
+      "step": 23509
+    },
+    {
+      "epoch": 0.5596020184709131,
+      "grad_norm": 0.34349634122160627,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 23510
+    },
+    {
+      "epoch": 0.5596258211939446,
+      "grad_norm": 6.642145109484174,
+      "learning_rate": 2e-05,
+      "loss": 0.3426,
+      "step": 23511
+    },
+    {
+      "epoch": 0.5596496239169761,
+      "grad_norm": 5.295140152314077,
+      "learning_rate": 2e-05,
+      "loss": 0.6257,
+      "step": 23512
+    },
+    {
+      "epoch": 0.5596734266400076,
+      "grad_norm": 9.49343860119522,
+      "learning_rate": 2e-05,
+      "loss": 0.3698,
+      "step": 23513
+    },
+    {
+      "epoch": 0.5596972293630391,
+      "grad_norm": 0.01890024183886041,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 23514
+    },
+    {
+      "epoch": 0.5597210320860706,
+      "grad_norm": 0.06263298840329223,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 23515
+    },
+    {
+      "epoch": 0.5597448348091022,
+      "grad_norm": 3.1880895284467363,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 23516
+    },
+    {
+      "epoch": 0.5597686375321337,
+      "grad_norm": 0.26600337668337587,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 23517
+    },
+    {
+      "epoch": 0.5597924402551652,
+      "grad_norm": 4.68255412930361,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 23518
+    },
+    {
+      "epoch": 0.5598162429781967,
+      "grad_norm": 9.850137303642262,
+      "learning_rate": 2e-05,
+      "loss": 0.9039,
+      "step": 23519
+    },
+    {
+      "epoch": 0.5598400457012283,
+      "grad_norm": 0.5695446476067283,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 23520
+    },
+    {
+      "epoch": 0.5598638484242597,
+      "grad_norm": 2.971189449613201,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 23521
+    },
+    {
+      "epoch": 0.5598876511472912,
+      "grad_norm": 8.994239421659122,
+      "learning_rate": 2e-05,
+      "loss": 1.1379,
+      "step": 23522
+    },
+    {
+      "epoch": 0.5599114538703228,
+      "grad_norm": 3.467827876849123,
+      "learning_rate": 2e-05,
+      "loss": 0.1473,
+      "step": 23523
+    },
+    {
+      "epoch": 0.5599352565933543,
+      "grad_norm": 1.285412878390327,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 23524
+    },
+    {
+      "epoch": 0.5599590593163858,
+      "grad_norm": 0.8885176977367392,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 23525
+    },
+    {
+      "epoch": 0.5599828620394173,
+      "grad_norm": 2.464829032280658,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 23526
+    },
+    {
+      "epoch": 0.5600066647624489,
+      "grad_norm": 4.767625943962837,
+      "learning_rate": 2e-05,
+      "loss": 0.456,
+      "step": 23527
+    },
+    {
+      "epoch": 0.5600304674854804,
+      "grad_norm": 0.5280553913604545,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 23528
+    },
+    {
+      "epoch": 0.5600542702085118,
+      "grad_norm": 7.677820712648908,
+      "learning_rate": 2e-05,
+      "loss": 1.1706,
+      "step": 23529
+    },
+    {
+      "epoch": 0.5600780729315433,
+      "grad_norm": 10.175937007637991,
+      "learning_rate": 2e-05,
+      "loss": 1.1157,
+      "step": 23530
+    },
+    {
+      "epoch": 0.5601018756545749,
+      "grad_norm": 5.413538815791706,
+      "learning_rate": 2e-05,
+      "loss": 0.8288,
+      "step": 23531
+    },
+    {
+      "epoch": 0.5601256783776064,
+      "grad_norm": 4.823645084503221,
+      "learning_rate": 2e-05,
+      "loss": 0.5744,
+      "step": 23532
+    },
+    {
+      "epoch": 0.5601494811006379,
+      "grad_norm": 2.0895199460882807,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 23533
+    },
+    {
+      "epoch": 0.5601732838236694,
+      "grad_norm": 2.4633647688824136,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 23534
+    },
+    {
+      "epoch": 0.560197086546701,
+      "grad_norm": 4.444847478719571,
+      "learning_rate": 2e-05,
+      "loss": 0.5016,
+      "step": 23535
+    },
+    {
+      "epoch": 0.5602208892697325,
+      "grad_norm": 4.099195722126599,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 23536
+    },
+    {
+      "epoch": 0.5602446919927639,
+      "grad_norm": 2.343850809051752,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 23537
+    },
+    {
+      "epoch": 0.5602684947157955,
+      "grad_norm": 6.040786151692373,
+      "learning_rate": 2e-05,
+      "loss": 1.0317,
+      "step": 23538
+    },
+    {
+      "epoch": 0.560292297438827,
+      "grad_norm": 5.696496263867322,
+      "learning_rate": 2e-05,
+      "loss": 0.4538,
+      "step": 23539
+    },
+    {
+      "epoch": 0.5603161001618585,
+      "grad_norm": 4.143461217516811,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 23540
+    },
+    {
+      "epoch": 0.56033990288489,
+      "grad_norm": 4.58076445677589,
+      "learning_rate": 2e-05,
+      "loss": 0.389,
+      "step": 23541
+    },
+    {
+      "epoch": 0.5603637056079216,
+      "grad_norm": 0.20851334763471788,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 23542
+    },
+    {
+      "epoch": 0.5603875083309531,
+      "grad_norm": 0.9202245111367172,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 23543
+    },
+    {
+      "epoch": 0.5604113110539846,
+      "grad_norm": 2.9575435552042313,
+      "learning_rate": 2e-05,
+      "loss": 0.1426,
+      "step": 23544
+    },
+    {
+      "epoch": 0.560435113777016,
+      "grad_norm": 30.20465946039165,
+      "learning_rate": 2e-05,
+      "loss": 1.5792,
+      "step": 23545
+    },
+    {
+      "epoch": 0.5604589165000476,
+      "grad_norm": 4.385976977172414,
+      "learning_rate": 2e-05,
+      "loss": 0.4681,
+      "step": 23546
+    },
+    {
+      "epoch": 0.5604827192230791,
+      "grad_norm": 3.789717357038268,
+      "learning_rate": 2e-05,
+      "loss": 0.3423,
+      "step": 23547
+    },
+    {
+      "epoch": 0.5605065219461106,
+      "grad_norm": 6.033577064488573,
+      "learning_rate": 2e-05,
+      "loss": 1.1796,
+      "step": 23548
+    },
+    {
+      "epoch": 0.5605303246691421,
+      "grad_norm": 0.4047022644432887,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 23549
+    },
+    {
+      "epoch": 0.5605541273921737,
+      "grad_norm": 3.6093285468992176,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 23550
+    },
+    {
+      "epoch": 0.5605779301152052,
+      "grad_norm": 4.824019871532466,
+      "learning_rate": 2e-05,
+      "loss": 0.866,
+      "step": 23551
+    },
+    {
+      "epoch": 0.5606017328382367,
+      "grad_norm": 1.3725861491929991,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 23552
+    },
+    {
+      "epoch": 0.5606255355612683,
+      "grad_norm": 6.125913533797798,
+      "learning_rate": 2e-05,
+      "loss": 1.2529,
+      "step": 23553
+    },
+    {
+      "epoch": 0.5606493382842997,
+      "grad_norm": 2.6883149342289574,
+      "learning_rate": 2e-05,
+      "loss": 0.1184,
+      "step": 23554
+    },
+    {
+      "epoch": 0.5606731410073312,
+      "grad_norm": 5.241215092614703,
+      "learning_rate": 2e-05,
+      "loss": 0.2328,
+      "step": 23555
+    },
+    {
+      "epoch": 0.5606969437303627,
+      "grad_norm": 1.819089021383161,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 23556
+    },
+    {
+      "epoch": 0.5607207464533943,
+      "grad_norm": 4.8589067862326845,
+      "learning_rate": 2e-05,
+      "loss": 0.7377,
+      "step": 23557
+    },
+    {
+      "epoch": 0.5607445491764258,
+      "grad_norm": 5.534318913229761,
+      "learning_rate": 2e-05,
+      "loss": 0.2126,
+      "step": 23558
+    },
+    {
+      "epoch": 0.5607683518994573,
+      "grad_norm": 4.415631375111588,
+      "learning_rate": 2e-05,
+      "loss": 0.5961,
+      "step": 23559
+    },
+    {
+      "epoch": 0.5607921546224888,
+      "grad_norm": 1.2807531257162172,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 23560
+    },
+    {
+      "epoch": 0.5608159573455204,
+      "grad_norm": 2.927431551741815,
+      "learning_rate": 2e-05,
+      "loss": 0.3814,
+      "step": 23561
+    },
+    {
+      "epoch": 0.5608397600685519,
+      "grad_norm": 6.40148724272092,
+      "learning_rate": 2e-05,
+      "loss": 0.0642,
+      "step": 23562
+    },
+    {
+      "epoch": 0.5608635627915833,
+      "grad_norm": 2.688304979228216,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 23563
+    },
+    {
+      "epoch": 0.5608873655146148,
+      "grad_norm": 4.545268177334064,
+      "learning_rate": 2e-05,
+      "loss": 0.8386,
+      "step": 23564
+    },
+    {
+      "epoch": 0.5609111682376464,
+      "grad_norm": 3.194897472488533,
+      "learning_rate": 2e-05,
+      "loss": 0.1235,
+      "step": 23565
+    },
+    {
+      "epoch": 0.5609349709606779,
+      "grad_norm": 0.6280029444633755,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 23566
+    },
+    {
+      "epoch": 0.5609587736837094,
+      "grad_norm": 0.06966053116002752,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 23567
+    },
+    {
+      "epoch": 0.560982576406741,
+      "grad_norm": 4.60038272915934,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 23568
+    },
+    {
+      "epoch": 0.5610063791297725,
+      "grad_norm": 4.364600108411,
+      "learning_rate": 2e-05,
+      "loss": 0.2886,
+      "step": 23569
+    },
+    {
+      "epoch": 0.561030181852804,
+      "grad_norm": 1.3774005594489929,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 23570
+    },
+    {
+      "epoch": 0.5610539845758354,
+      "grad_norm": 2.3079034019623172,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 23571
+    },
+    {
+      "epoch": 0.561077787298867,
+      "grad_norm": 1.103993635933805,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 23572
+    },
+    {
+      "epoch": 0.5611015900218985,
+      "grad_norm": 25.468043108241456,
+      "learning_rate": 2e-05,
+      "loss": 0.2785,
+      "step": 23573
+    },
+    {
+      "epoch": 0.56112539274493,
+      "grad_norm": 1.7857835022440935,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 23574
+    },
+    {
+      "epoch": 0.5611491954679615,
+      "grad_norm": 3.4452836176566612,
+      "learning_rate": 2e-05,
+      "loss": 0.464,
+      "step": 23575
+    },
+    {
+      "epoch": 0.5611729981909931,
+      "grad_norm": 9.021301609095781,
+      "learning_rate": 2e-05,
+      "loss": 0.9832,
+      "step": 23576
+    },
+    {
+      "epoch": 0.5611968009140246,
+      "grad_norm": 4.644816663655037,
+      "learning_rate": 2e-05,
+      "loss": 0.4372,
+      "step": 23577
+    },
+    {
+      "epoch": 0.5612206036370561,
+      "grad_norm": 1.701324514048088,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 23578
+    },
+    {
+      "epoch": 0.5612444063600875,
+      "grad_norm": 3.234753868544703,
+      "learning_rate": 2e-05,
+      "loss": 0.3665,
+      "step": 23579
+    },
+    {
+      "epoch": 0.5612682090831191,
+      "grad_norm": 4.6250636809192205,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 23580
+    },
+    {
+      "epoch": 0.5612920118061506,
+      "grad_norm": 3.609376317904949,
+      "learning_rate": 2e-05,
+      "loss": 0.3984,
+      "step": 23581
+    },
+    {
+      "epoch": 0.5613158145291821,
+      "grad_norm": 5.81621848780761,
+      "learning_rate": 2e-05,
+      "loss": 0.9689,
+      "step": 23582
+    },
+    {
+      "epoch": 0.5613396172522137,
+      "grad_norm": 2.4693695943677088,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 23583
+    },
+    {
+      "epoch": 0.5613634199752452,
+      "grad_norm": 0.7536319804621413,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 23584
+    },
+    {
+      "epoch": 0.5613872226982767,
+      "grad_norm": 0.08178568743782894,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 23585
+    },
+    {
+      "epoch": 0.5614110254213082,
+      "grad_norm": 0.15174805258156432,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 23586
+    },
+    {
+      "epoch": 0.5614348281443398,
+      "grad_norm": 5.976037208946654,
+      "learning_rate": 2e-05,
+      "loss": 0.6808,
+      "step": 23587
+    },
+    {
+      "epoch": 0.5614586308673712,
+      "grad_norm": 0.1001506861175383,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 23588
+    },
+    {
+      "epoch": 0.5614824335904027,
+      "grad_norm": 8.271462620226446,
+      "learning_rate": 2e-05,
+      "loss": 0.4902,
+      "step": 23589
+    },
+    {
+      "epoch": 0.5615062363134342,
+      "grad_norm": 0.5227402017032396,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 23590
+    },
+    {
+      "epoch": 0.5615300390364658,
+      "grad_norm": 2.3225936312620323,
+      "learning_rate": 2e-05,
+      "loss": 0.1212,
+      "step": 23591
+    },
+    {
+      "epoch": 0.5615538417594973,
+      "grad_norm": 15.575494333731411,
+      "learning_rate": 2e-05,
+      "loss": 0.6216,
+      "step": 23592
+    },
+    {
+      "epoch": 0.5615776444825288,
+      "grad_norm": 2.3172259268934505,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 23593
+    },
+    {
+      "epoch": 0.5616014472055603,
+      "grad_norm": 5.574953976021031,
+      "learning_rate": 2e-05,
+      "loss": 0.806,
+      "step": 23594
+    },
+    {
+      "epoch": 0.5616252499285919,
+      "grad_norm": 5.904072284185753,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 23595
+    },
+    {
+      "epoch": 0.5616490526516233,
+      "grad_norm": 4.572104250678089,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 23596
+    },
+    {
+      "epoch": 0.5616728553746548,
+      "grad_norm": 11.891798963463176,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 23597
+    },
+    {
+      "epoch": 0.5616966580976864,
+      "grad_norm": 1.0759980383294703,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 23598
+    },
+    {
+      "epoch": 0.5617204608207179,
+      "grad_norm": 2.1986957897698676,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 23599
+    },
+    {
+      "epoch": 0.5617442635437494,
+      "grad_norm": 2.649378143739667,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 23600
+    },
+    {
+      "epoch": 0.5617680662667809,
+      "grad_norm": 0.9763234575422267,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 23601
+    },
+    {
+      "epoch": 0.5617918689898125,
+      "grad_norm": 4.127225938390272,
+      "learning_rate": 2e-05,
+      "loss": 0.5589,
+      "step": 23602
+    },
+    {
+      "epoch": 0.561815671712844,
+      "grad_norm": 0.4569179373807233,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 23603
+    },
+    {
+      "epoch": 0.5618394744358755,
+      "grad_norm": 0.16185250626914047,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 23604
+    },
+    {
+      "epoch": 0.5618632771589069,
+      "grad_norm": 5.388486228406459,
+      "learning_rate": 2e-05,
+      "loss": 0.947,
+      "step": 23605
+    },
+    {
+      "epoch": 0.5618870798819385,
+      "grad_norm": 0.40499023226243,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 23606
+    },
+    {
+      "epoch": 0.56191088260497,
+      "grad_norm": 0.7041505415344668,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 23607
+    },
+    {
+      "epoch": 0.5619346853280015,
+      "grad_norm": 2.4184406418758613,
+      "learning_rate": 2e-05,
+      "loss": 0.1878,
+      "step": 23608
+    },
+    {
+      "epoch": 0.561958488051033,
+      "grad_norm": 8.08838114430617,
+      "learning_rate": 2e-05,
+      "loss": 2.3881,
+      "step": 23609
+    },
+    {
+      "epoch": 0.5619822907740646,
+      "grad_norm": 4.851966249155535,
+      "learning_rate": 2e-05,
+      "loss": 0.6244,
+      "step": 23610
+    },
+    {
+      "epoch": 0.5620060934970961,
+      "grad_norm": 1.441834067836161,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 23611
+    },
+    {
+      "epoch": 0.5620298962201276,
+      "grad_norm": 5.252410287299622,
+      "learning_rate": 2e-05,
+      "loss": 0.6868,
+      "step": 23612
+    },
+    {
+      "epoch": 0.5620536989431592,
+      "grad_norm": 1.7427003346032484,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 23613
+    },
+    {
+      "epoch": 0.5620775016661906,
+      "grad_norm": 6.576228453339574,
+      "learning_rate": 2e-05,
+      "loss": 1.5819,
+      "step": 23614
+    },
+    {
+      "epoch": 0.5621013043892221,
+      "grad_norm": 4.961232681631522,
+      "learning_rate": 2e-05,
+      "loss": 0.5429,
+      "step": 23615
+    },
+    {
+      "epoch": 0.5621251071122536,
+      "grad_norm": 8.695964792581913,
+      "learning_rate": 2e-05,
+      "loss": 0.8815,
+      "step": 23616
+    },
+    {
+      "epoch": 0.5621489098352852,
+      "grad_norm": 2.251121359632438,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 23617
+    },
+    {
+      "epoch": 0.5621727125583167,
+      "grad_norm": 2.4828378995758453,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 23618
+    },
+    {
+      "epoch": 0.5621965152813482,
+      "grad_norm": 4.381691797703248,
+      "learning_rate": 2e-05,
+      "loss": 0.7299,
+      "step": 23619
+    },
+    {
+      "epoch": 0.5622203180043797,
+      "grad_norm": 6.615396481690908,
+      "learning_rate": 2e-05,
+      "loss": 0.9224,
+      "step": 23620
+    },
+    {
+      "epoch": 0.5622441207274113,
+      "grad_norm": 5.075644769210227,
+      "learning_rate": 2e-05,
+      "loss": 1.1651,
+      "step": 23621
+    },
+    {
+      "epoch": 0.5622679234504427,
+      "grad_norm": 1.0784847023037298,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 23622
+    },
+    {
+      "epoch": 0.5622917261734742,
+      "grad_norm": 0.06586278566949173,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 23623
+    },
+    {
+      "epoch": 0.5623155288965057,
+      "grad_norm": 8.283831564330992,
+      "learning_rate": 2e-05,
+      "loss": 0.311,
+      "step": 23624
+    },
+    {
+      "epoch": 0.5623393316195373,
+      "grad_norm": 4.294814463603779,
+      "learning_rate": 2e-05,
+      "loss": 0.5313,
+      "step": 23625
+    },
+    {
+      "epoch": 0.5623631343425688,
+      "grad_norm": 5.054322947836179,
+      "learning_rate": 2e-05,
+      "loss": 0.825,
+      "step": 23626
+    },
+    {
+      "epoch": 0.5623869370656003,
+      "grad_norm": 1.0744552877116234,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 23627
+    },
+    {
+      "epoch": 0.5624107397886319,
+      "grad_norm": 4.191042213687313,
+      "learning_rate": 2e-05,
+      "loss": 0.8062,
+      "step": 23628
+    },
+    {
+      "epoch": 0.5624345425116634,
+      "grad_norm": 8.73841308059886,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 23629
+    },
+    {
+      "epoch": 0.5624583452346948,
+      "grad_norm": 7.27429484006945,
+      "learning_rate": 2e-05,
+      "loss": 0.165,
+      "step": 23630
+    },
+    {
+      "epoch": 0.5624821479577263,
+      "grad_norm": 0.5138194528244742,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 23631
+    },
+    {
+      "epoch": 0.5625059506807579,
+      "grad_norm": 4.7331740003547935,
+      "learning_rate": 2e-05,
+      "loss": 0.632,
+      "step": 23632
+    },
+    {
+      "epoch": 0.5625297534037894,
+      "grad_norm": 0.987573912866852,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 23633
+    },
+    {
+      "epoch": 0.5625535561268209,
+      "grad_norm": 217.91869805199116,
+      "learning_rate": 2e-05,
+      "loss": 0.2702,
+      "step": 23634
+    },
+    {
+      "epoch": 0.5625773588498524,
+      "grad_norm": 1.8887292819869679,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 23635
+    },
+    {
+      "epoch": 0.562601161572884,
+      "grad_norm": 5.58637837143299,
+      "learning_rate": 2e-05,
+      "loss": 0.0823,
+      "step": 23636
+    },
+    {
+      "epoch": 0.5626249642959155,
+      "grad_norm": 1.3696469931814519,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 23637
+    },
+    {
+      "epoch": 0.562648767018947,
+      "grad_norm": 5.957885931337166,
+      "learning_rate": 2e-05,
+      "loss": 0.25,
+      "step": 23638
+    },
+    {
+      "epoch": 0.5626725697419784,
+      "grad_norm": 1.1500854532013203,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 23639
+    },
+    {
+      "epoch": 0.56269637246501,
+      "grad_norm": 10.66187476455134,
+      "learning_rate": 2e-05,
+      "loss": 1.9277,
+      "step": 23640
+    },
+    {
+      "epoch": 0.5627201751880415,
+      "grad_norm": 10.740951782898417,
+      "learning_rate": 2e-05,
+      "loss": 1.0872,
+      "step": 23641
+    },
+    {
+      "epoch": 0.562743977911073,
+      "grad_norm": 5.724738768599851,
+      "learning_rate": 2e-05,
+      "loss": 1.3944,
+      "step": 23642
+    },
+    {
+      "epoch": 0.5627677806341045,
+      "grad_norm": 0.29127669212272944,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 23643
+    },
+    {
+      "epoch": 0.5627915833571361,
+      "grad_norm": 0.2520688659811904,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 23644
+    },
+    {
+      "epoch": 0.5628153860801676,
+      "grad_norm": 4.960754444146888,
+      "learning_rate": 2e-05,
+      "loss": 1.1398,
+      "step": 23645
+    },
+    {
+      "epoch": 0.562839188803199,
+      "grad_norm": 10.943530224179005,
+      "learning_rate": 2e-05,
+      "loss": 1.7225,
+      "step": 23646
+    },
+    {
+      "epoch": 0.5628629915262306,
+      "grad_norm": 7.1721847641176675,
+      "learning_rate": 2e-05,
+      "loss": 0.6097,
+      "step": 23647
+    },
+    {
+      "epoch": 0.5628867942492621,
+      "grad_norm": 9.968685634568818,
+      "learning_rate": 2e-05,
+      "loss": 0.4837,
+      "step": 23648
+    },
+    {
+      "epoch": 0.5629105969722936,
+      "grad_norm": 0.7813779410482029,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 23649
+    },
+    {
+      "epoch": 0.5629343996953251,
+      "grad_norm": 4.951169107026309,
+      "learning_rate": 2e-05,
+      "loss": 0.5803,
+      "step": 23650
+    },
+    {
+      "epoch": 0.5629582024183567,
+      "grad_norm": 4.43216236399537,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 23651
+    },
+    {
+      "epoch": 0.5629820051413882,
+      "grad_norm": 4.119815928620837,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 23652
+    },
+    {
+      "epoch": 0.5630058078644197,
+      "grad_norm": 0.03163056882310566,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 23653
+    },
+    {
+      "epoch": 0.5630296105874512,
+      "grad_norm": 4.202200262313232,
+      "learning_rate": 2e-05,
+      "loss": 0.5996,
+      "step": 23654
+    },
+    {
+      "epoch": 0.5630534133104828,
+      "grad_norm": 4.771181422583791,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 23655
+    },
+    {
+      "epoch": 0.5630772160335142,
+      "grad_norm": 4.108241309499987,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 23656
+    },
+    {
+      "epoch": 0.5631010187565457,
+      "grad_norm": 4.6062609074710705,
+      "learning_rate": 2e-05,
+      "loss": 0.5279,
+      "step": 23657
+    },
+    {
+      "epoch": 0.5631248214795772,
+      "grad_norm": 20.697158069912973,
+      "learning_rate": 2e-05,
+      "loss": 0.4607,
+      "step": 23658
+    },
+    {
+      "epoch": 0.5631486242026088,
+      "grad_norm": 3.7990025766827826,
+      "learning_rate": 2e-05,
+      "loss": 0.6273,
+      "step": 23659
+    },
+    {
+      "epoch": 0.5631724269256403,
+      "grad_norm": 0.631168807728204,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 23660
+    },
+    {
+      "epoch": 0.5631962296486718,
+      "grad_norm": 0.6693881810412082,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 23661
+    },
+    {
+      "epoch": 0.5632200323717034,
+      "grad_norm": 2.8812133095548913,
+      "learning_rate": 2e-05,
+      "loss": 0.4658,
+      "step": 23662
+    },
+    {
+      "epoch": 0.5632438350947349,
+      "grad_norm": 0.3626416819951946,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 23663
+    },
+    {
+      "epoch": 0.5632676378177663,
+      "grad_norm": 3.311065886805636,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 23664
+    },
+    {
+      "epoch": 0.5632914405407978,
+      "grad_norm": 4.92105409622274,
+      "learning_rate": 2e-05,
+      "loss": 0.6816,
+      "step": 23665
+    },
+    {
+      "epoch": 0.5633152432638294,
+      "grad_norm": 2.595830922325741,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 23666
+    },
+    {
+      "epoch": 0.5633390459868609,
+      "grad_norm": 0.2090205740258525,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 23667
+    },
+    {
+      "epoch": 0.5633628487098924,
+      "grad_norm": 5.0142230909348084,
+      "learning_rate": 2e-05,
+      "loss": 0.7005,
+      "step": 23668
+    },
+    {
+      "epoch": 0.5633866514329239,
+      "grad_norm": 1.2987437509173239,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 23669
+    },
+    {
+      "epoch": 0.5634104541559555,
+      "grad_norm": 0.266630311092668,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 23670
+    },
+    {
+      "epoch": 0.563434256878987,
+      "grad_norm": 0.3580298128779957,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 23671
+    },
+    {
+      "epoch": 0.5634580596020184,
+      "grad_norm": 2.42172824693454,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 23672
+    },
+    {
+      "epoch": 0.5634818623250499,
+      "grad_norm": 4.605596596046979,
+      "learning_rate": 2e-05,
+      "loss": 0.3916,
+      "step": 23673
+    },
+    {
+      "epoch": 0.5635056650480815,
+      "grad_norm": 3.7245889348094483,
+      "learning_rate": 2e-05,
+      "loss": 0.1194,
+      "step": 23674
+    },
+    {
+      "epoch": 0.563529467771113,
+      "grad_norm": 2.5856346634067373,
+      "learning_rate": 2e-05,
+      "loss": 0.4324,
+      "step": 23675
+    },
+    {
+      "epoch": 0.5635532704941445,
+      "grad_norm": 1.3197596867107984,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 23676
+    },
+    {
+      "epoch": 0.5635770732171761,
+      "grad_norm": 4.082589420782624,
+      "learning_rate": 2e-05,
+      "loss": 0.5333,
+      "step": 23677
+    },
+    {
+      "epoch": 0.5636008759402076,
+      "grad_norm": 5.580259403255721,
+      "learning_rate": 2e-05,
+      "loss": 0.151,
+      "step": 23678
+    },
+    {
+      "epoch": 0.5636246786632391,
+      "grad_norm": 10.328643286302267,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 23679
+    },
+    {
+      "epoch": 0.5636484813862706,
+      "grad_norm": 2.8292979890741794,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 23680
+    },
+    {
+      "epoch": 0.5636722841093021,
+      "grad_norm": 4.478075742503016,
+      "learning_rate": 2e-05,
+      "loss": 0.9596,
+      "step": 23681
+    },
+    {
+      "epoch": 0.5636960868323336,
+      "grad_norm": 0.3146048332425113,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 23682
+    },
+    {
+      "epoch": 0.5637198895553651,
+      "grad_norm": 4.7701543329950615,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 23683
+    },
+    {
+      "epoch": 0.5637436922783966,
+      "grad_norm": 5.908090152673516,
+      "learning_rate": 2e-05,
+      "loss": 1.3507,
+      "step": 23684
+    },
+    {
+      "epoch": 0.5637674950014282,
+      "grad_norm": 4.271704474238724,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 23685
+    },
+    {
+      "epoch": 0.5637912977244597,
+      "grad_norm": 5.693441372359023,
+      "learning_rate": 2e-05,
+      "loss": 0.5833,
+      "step": 23686
+    },
+    {
+      "epoch": 0.5638151004474912,
+      "grad_norm": 0.09442071561914223,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 23687
+    },
+    {
+      "epoch": 0.5638389031705227,
+      "grad_norm": 8.958337027890224,
+      "learning_rate": 2e-05,
+      "loss": 0.5469,
+      "step": 23688
+    },
+    {
+      "epoch": 0.5638627058935543,
+      "grad_norm": 8.94815412340302,
+      "learning_rate": 2e-05,
+      "loss": 1.0838,
+      "step": 23689
+    },
+    {
+      "epoch": 0.5638865086165857,
+      "grad_norm": 6.581188338231058,
+      "learning_rate": 2e-05,
+      "loss": 0.9949,
+      "step": 23690
+    },
+    {
+      "epoch": 0.5639103113396172,
+      "grad_norm": 7.614091247581253,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 23691
+    },
+    {
+      "epoch": 0.5639341140626488,
+      "grad_norm": 5.721536253857568,
+      "learning_rate": 2e-05,
+      "loss": 0.1742,
+      "step": 23692
+    },
+    {
+      "epoch": 0.5639579167856803,
+      "grad_norm": 5.028555976995796,
+      "learning_rate": 2e-05,
+      "loss": 1.2254,
+      "step": 23693
+    },
+    {
+      "epoch": 0.5639817195087118,
+      "grad_norm": 3.3805253730507054,
+      "learning_rate": 2e-05,
+      "loss": 0.3045,
+      "step": 23694
+    },
+    {
+      "epoch": 0.5640055222317433,
+      "grad_norm": 9.077752211022213,
+      "learning_rate": 2e-05,
+      "loss": 0.819,
+      "step": 23695
+    },
+    {
+      "epoch": 0.5640293249547749,
+      "grad_norm": 8.210713376228538,
+      "learning_rate": 2e-05,
+      "loss": 1.3193,
+      "step": 23696
+    },
+    {
+      "epoch": 0.5640531276778064,
+      "grad_norm": 5.550729084674367,
+      "learning_rate": 2e-05,
+      "loss": 0.1774,
+      "step": 23697
+    },
+    {
+      "epoch": 0.5640769304008378,
+      "grad_norm": 3.0981358536448655,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 23698
+    },
+    {
+      "epoch": 0.5641007331238693,
+      "grad_norm": 7.217007322833095,
+      "learning_rate": 2e-05,
+      "loss": 1.2777,
+      "step": 23699
+    },
+    {
+      "epoch": 0.5641245358469009,
+      "grad_norm": 1.5166174240071124,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 23700
+    },
+    {
+      "epoch": 0.5641483385699324,
+      "grad_norm": 4.9028687856244595,
+      "learning_rate": 2e-05,
+      "loss": 0.683,
+      "step": 23701
+    },
+    {
+      "epoch": 0.5641721412929639,
+      "grad_norm": 2.4546687956262434,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 23702
+    },
+    {
+      "epoch": 0.5641959440159954,
+      "grad_norm": 3.6992186770834796,
+      "learning_rate": 2e-05,
+      "loss": 0.4074,
+      "step": 23703
+    },
+    {
+      "epoch": 0.564219746739027,
+      "grad_norm": 4.830878806163074,
+      "learning_rate": 2e-05,
+      "loss": 0.8467,
+      "step": 23704
+    },
+    {
+      "epoch": 0.5642435494620585,
+      "grad_norm": 3.8480577081715883,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 23705
+    },
+    {
+      "epoch": 0.5642673521850899,
+      "grad_norm": 15.418303203943712,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 23706
+    },
+    {
+      "epoch": 0.5642911549081215,
+      "grad_norm": 4.963082140376213,
+      "learning_rate": 2e-05,
+      "loss": 0.5209,
+      "step": 23707
+    },
+    {
+      "epoch": 0.564314957631153,
+      "grad_norm": 4.190719217517215,
+      "learning_rate": 2e-05,
+      "loss": 0.207,
+      "step": 23708
+    },
+    {
+      "epoch": 0.5643387603541845,
+      "grad_norm": 2.790839391216834,
+      "learning_rate": 2e-05,
+      "loss": 0.5052,
+      "step": 23709
+    },
+    {
+      "epoch": 0.564362563077216,
+      "grad_norm": 8.214801571681539,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 23710
+    },
+    {
+      "epoch": 0.5643863658002476,
+      "grad_norm": 0.27627393735157774,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 23711
+    },
+    {
+      "epoch": 0.5644101685232791,
+      "grad_norm": 0.06231549399340174,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 23712
+    },
+    {
+      "epoch": 0.5644339712463106,
+      "grad_norm": 5.03338090388845,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 23713
+    },
+    {
+      "epoch": 0.564457773969342,
+      "grad_norm": 3.9894266901556983,
+      "learning_rate": 2e-05,
+      "loss": 0.3329,
+      "step": 23714
+    },
+    {
+      "epoch": 0.5644815766923736,
+      "grad_norm": 4.028483045392067,
+      "learning_rate": 2e-05,
+      "loss": 0.8038,
+      "step": 23715
+    },
+    {
+      "epoch": 0.5645053794154051,
+      "grad_norm": 0.7400531393027479,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 23716
+    },
+    {
+      "epoch": 0.5645291821384366,
+      "grad_norm": 5.881369843175737,
+      "learning_rate": 2e-05,
+      "loss": 0.7564,
+      "step": 23717
+    },
+    {
+      "epoch": 0.5645529848614681,
+      "grad_norm": 0.2844590252765408,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 23718
+    },
+    {
+      "epoch": 0.5645767875844997,
+      "grad_norm": 4.21150033001145,
+      "learning_rate": 2e-05,
+      "loss": 0.1829,
+      "step": 23719
+    },
+    {
+      "epoch": 0.5646005903075312,
+      "grad_norm": 3.49419808011144,
+      "learning_rate": 2e-05,
+      "loss": 0.614,
+      "step": 23720
+    },
+    {
+      "epoch": 0.5646243930305627,
+      "grad_norm": 11.52531262564269,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 23721
+    },
+    {
+      "epoch": 0.5646481957535943,
+      "grad_norm": 0.7133174337908,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 23722
+    },
+    {
+      "epoch": 0.5646719984766257,
+      "grad_norm": 3.3173702413089408,
+      "learning_rate": 2e-05,
+      "loss": 0.507,
+      "step": 23723
+    },
+    {
+      "epoch": 0.5646958011996572,
+      "grad_norm": 0.7200782479923717,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 23724
+    },
+    {
+      "epoch": 0.5647196039226887,
+      "grad_norm": 11.212582156969061,
+      "learning_rate": 2e-05,
+      "loss": 0.4267,
+      "step": 23725
+    },
+    {
+      "epoch": 0.5647434066457203,
+      "grad_norm": 1.8033363282069008,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 23726
+    },
+    {
+      "epoch": 0.5647672093687518,
+      "grad_norm": 3.77015216267358,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 23727
+    },
+    {
+      "epoch": 0.5647910120917833,
+      "grad_norm": 2.532569623628008,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 23728
+    },
+    {
+      "epoch": 0.5648148148148148,
+      "grad_norm": 11.006601329072607,
+      "learning_rate": 2e-05,
+      "loss": 0.55,
+      "step": 23729
+    },
+    {
+      "epoch": 0.5648386175378464,
+      "grad_norm": 0.38496754419381535,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 23730
+    },
+    {
+      "epoch": 0.5648624202608779,
+      "grad_norm": 3.5187663815700967,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 23731
+    },
+    {
+      "epoch": 0.5648862229839093,
+      "grad_norm": 2.590136202832611,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 23732
+    },
+    {
+      "epoch": 0.5649100257069408,
+      "grad_norm": 0.7806832688625386,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 23733
+    },
+    {
+      "epoch": 0.5649338284299724,
+      "grad_norm": 6.826701917485738,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 23734
+    },
+    {
+      "epoch": 0.5649576311530039,
+      "grad_norm": 3.453760527336063,
+      "learning_rate": 2e-05,
+      "loss": 0.0759,
+      "step": 23735
+    },
+    {
+      "epoch": 0.5649814338760354,
+      "grad_norm": 6.842841379285109,
+      "learning_rate": 2e-05,
+      "loss": 0.1989,
+      "step": 23736
+    },
+    {
+      "epoch": 0.565005236599067,
+      "grad_norm": 4.222546374837373,
+      "learning_rate": 2e-05,
+      "loss": 0.7057,
+      "step": 23737
+    },
+    {
+      "epoch": 0.5650290393220985,
+      "grad_norm": 5.709920354713132,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 23738
+    },
+    {
+      "epoch": 0.56505284204513,
+      "grad_norm": 3.7988120111547947,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 23739
+    },
+    {
+      "epoch": 0.5650766447681614,
+      "grad_norm": 4.43086565984021,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 23740
+    },
+    {
+      "epoch": 0.565100447491193,
+      "grad_norm": 2.6285981211967755,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 23741
+    },
+    {
+      "epoch": 0.5651242502142245,
+      "grad_norm": 3.109886376809572,
+      "learning_rate": 2e-05,
+      "loss": 0.1632,
+      "step": 23742
+    },
+    {
+      "epoch": 0.565148052937256,
+      "grad_norm": 1.9942513713537817,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 23743
+    },
+    {
+      "epoch": 0.5651718556602875,
+      "grad_norm": 4.646954075205923,
+      "learning_rate": 2e-05,
+      "loss": 0.1632,
+      "step": 23744
+    },
+    {
+      "epoch": 0.5651956583833191,
+      "grad_norm": 16.05840058772676,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 23745
+    },
+    {
+      "epoch": 0.5652194611063506,
+      "grad_norm": 6.0786688240181235,
+      "learning_rate": 2e-05,
+      "loss": 1.3623,
+      "step": 23746
+    },
+    {
+      "epoch": 0.5652432638293821,
+      "grad_norm": 4.314508721655243,
+      "learning_rate": 2e-05,
+      "loss": 0.8612,
+      "step": 23747
+    },
+    {
+      "epoch": 0.5652670665524135,
+      "grad_norm": 5.14825388823043,
+      "learning_rate": 2e-05,
+      "loss": 0.7523,
+      "step": 23748
+    },
+    {
+      "epoch": 0.5652908692754451,
+      "grad_norm": 1.8512274343865405,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 23749
+    },
+    {
+      "epoch": 0.5653146719984766,
+      "grad_norm": 9.156550462714833,
+      "learning_rate": 2e-05,
+      "loss": 0.2022,
+      "step": 23750
+    },
+    {
+      "epoch": 0.5653384747215081,
+      "grad_norm": 2.618451620183181,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 23751
+    },
+    {
+      "epoch": 0.5653622774445397,
+      "grad_norm": 0.5072691358974803,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 23752
+    },
+    {
+      "epoch": 0.5653860801675712,
+      "grad_norm": 5.718277624395915,
+      "learning_rate": 2e-05,
+      "loss": 0.1327,
+      "step": 23753
+    },
+    {
+      "epoch": 0.5654098828906027,
+      "grad_norm": 4.35534769515806,
+      "learning_rate": 2e-05,
+      "loss": 0.6049,
+      "step": 23754
+    },
+    {
+      "epoch": 0.5654336856136342,
+      "grad_norm": 7.136739940139264,
+      "learning_rate": 2e-05,
+      "loss": 1.1351,
+      "step": 23755
+    },
+    {
+      "epoch": 0.5654574883366658,
+      "grad_norm": 1.9122711667336922,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 23756
+    },
+    {
+      "epoch": 0.5654812910596972,
+      "grad_norm": 0.1224205029085919,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 23757
+    },
+    {
+      "epoch": 0.5655050937827287,
+      "grad_norm": 5.358915304802375,
+      "learning_rate": 2e-05,
+      "loss": 0.2984,
+      "step": 23758
+    },
+    {
+      "epoch": 0.5655288965057602,
+      "grad_norm": 2.2758349556479653,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 23759
+    },
+    {
+      "epoch": 0.5655526992287918,
+      "grad_norm": 3.5751532084800006,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 23760
+    },
+    {
+      "epoch": 0.5655765019518233,
+      "grad_norm": 3.6748653933733966,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 23761
+    },
+    {
+      "epoch": 0.5656003046748548,
+      "grad_norm": 0.4502640066047885,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 23762
+    },
+    {
+      "epoch": 0.5656241073978863,
+      "grad_norm": 4.921355661619365,
+      "learning_rate": 2e-05,
+      "loss": 0.2071,
+      "step": 23763
+    },
+    {
+      "epoch": 0.5656479101209179,
+      "grad_norm": 4.370937288243485,
+      "learning_rate": 2e-05,
+      "loss": 0.524,
+      "step": 23764
+    },
+    {
+      "epoch": 0.5656717128439493,
+      "grad_norm": 11.837568995110958,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 23765
+    },
+    {
+      "epoch": 0.5656955155669808,
+      "grad_norm": 1.2557826935729692,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 23766
+    },
+    {
+      "epoch": 0.5657193182900124,
+      "grad_norm": 0.9191046311564565,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 23767
+    },
+    {
+      "epoch": 0.5657431210130439,
+      "grad_norm": 4.961066898256205,
+      "learning_rate": 2e-05,
+      "loss": 0.1977,
+      "step": 23768
+    },
+    {
+      "epoch": 0.5657669237360754,
+      "grad_norm": 11.965791185354306,
+      "learning_rate": 2e-05,
+      "loss": 0.5937,
+      "step": 23769
+    },
+    {
+      "epoch": 0.5657907264591069,
+      "grad_norm": 4.110402354848559,
+      "learning_rate": 2e-05,
+      "loss": 0.1091,
+      "step": 23770
+    },
+    {
+      "epoch": 0.5658145291821385,
+      "grad_norm": 0.18195231406913043,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 23771
+    },
+    {
+      "epoch": 0.56583833190517,
+      "grad_norm": 2.81210582920885,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 23772
+    },
+    {
+      "epoch": 0.5658621346282015,
+      "grad_norm": 0.7538367162877594,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 23773
+    },
+    {
+      "epoch": 0.5658859373512329,
+      "grad_norm": 3.1660220218017385,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 23774
+    },
+    {
+      "epoch": 0.5659097400742645,
+      "grad_norm": 5.0083095065938785,
+      "learning_rate": 2e-05,
+      "loss": 0.6174,
+      "step": 23775
+    },
+    {
+      "epoch": 0.565933542797296,
+      "grad_norm": 2.1108854998010247,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 23776
+    },
+    {
+      "epoch": 0.5659573455203275,
+      "grad_norm": 1.6581655653863838,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 23777
+    },
+    {
+      "epoch": 0.565981148243359,
+      "grad_norm": 2.9308077591953663,
+      "learning_rate": 2e-05,
+      "loss": 0.1182,
+      "step": 23778
+    },
+    {
+      "epoch": 0.5660049509663906,
+      "grad_norm": 4.110443226414913,
+      "learning_rate": 2e-05,
+      "loss": 0.7985,
+      "step": 23779
+    },
+    {
+      "epoch": 0.5660287536894221,
+      "grad_norm": 6.692949076860461,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 23780
+    },
+    {
+      "epoch": 0.5660525564124536,
+      "grad_norm": 2.059932156281285,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 23781
+    },
+    {
+      "epoch": 0.5660763591354852,
+      "grad_norm": 0.3092160500671583,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 23782
+    },
+    {
+      "epoch": 0.5661001618585166,
+      "grad_norm": 1.5094455681647436,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 23783
+    },
+    {
+      "epoch": 0.5661239645815481,
+      "grad_norm": 4.780032429822848,
+      "learning_rate": 2e-05,
+      "loss": 0.4448,
+      "step": 23784
+    },
+    {
+      "epoch": 0.5661477673045796,
+      "grad_norm": 1.7641839455176183,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 23785
+    },
+    {
+      "epoch": 0.5661715700276112,
+      "grad_norm": 3.7175375042720753,
+      "learning_rate": 2e-05,
+      "loss": 0.6226,
+      "step": 23786
+    },
+    {
+      "epoch": 0.5661953727506427,
+      "grad_norm": 5.204087292554018,
+      "learning_rate": 2e-05,
+      "loss": 1.16,
+      "step": 23787
+    },
+    {
+      "epoch": 0.5662191754736742,
+      "grad_norm": 0.5473221133598624,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 23788
+    },
+    {
+      "epoch": 0.5662429781967057,
+      "grad_norm": 0.04903550659893782,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 23789
+    },
+    {
+      "epoch": 0.5662667809197373,
+      "grad_norm": 2.432471074564722,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 23790
+    },
+    {
+      "epoch": 0.5662905836427687,
+      "grad_norm": 1.9496882020612114,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 23791
+    },
+    {
+      "epoch": 0.5663143863658002,
+      "grad_norm": 0.9866437442763385,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 23792
+    },
+    {
+      "epoch": 0.5663381890888317,
+      "grad_norm": 2.4337288633500656,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 23793
+    },
+    {
+      "epoch": 0.5663619918118633,
+      "grad_norm": 6.417131952508503,
+      "learning_rate": 2e-05,
+      "loss": 0.4999,
+      "step": 23794
+    },
+    {
+      "epoch": 0.5663857945348948,
+      "grad_norm": 2.0014389970662596,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 23795
+    },
+    {
+      "epoch": 0.5664095972579263,
+      "grad_norm": 7.584353342879464,
+      "learning_rate": 2e-05,
+      "loss": 0.9124,
+      "step": 23796
+    },
+    {
+      "epoch": 0.5664333999809579,
+      "grad_norm": 0.22021474695088256,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 23797
+    },
+    {
+      "epoch": 0.5664572027039894,
+      "grad_norm": 3.9804726809949917,
+      "learning_rate": 2e-05,
+      "loss": 0.5423,
+      "step": 23798
+    },
+    {
+      "epoch": 0.5664810054270208,
+      "grad_norm": 5.126836584318755,
+      "learning_rate": 2e-05,
+      "loss": 0.7938,
+      "step": 23799
+    },
+    {
+      "epoch": 0.5665048081500523,
+      "grad_norm": 2.8584587394643894,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 23800
+    },
+    {
+      "epoch": 0.5665286108730839,
+      "grad_norm": 5.157767323442694,
+      "learning_rate": 2e-05,
+      "loss": 1.4609,
+      "step": 23801
+    },
+    {
+      "epoch": 0.5665524135961154,
+      "grad_norm": 0.13418987689870218,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 23802
+    },
+    {
+      "epoch": 0.5665762163191469,
+      "grad_norm": 5.347777178120294,
+      "learning_rate": 2e-05,
+      "loss": 1.0668,
+      "step": 23803
+    },
+    {
+      "epoch": 0.5666000190421784,
+      "grad_norm": 4.146901282192191,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 23804
+    },
+    {
+      "epoch": 0.56662382176521,
+      "grad_norm": 0.7291816519431065,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 23805
+    },
+    {
+      "epoch": 0.5666476244882415,
+      "grad_norm": 10.762723030211118,
+      "learning_rate": 2e-05,
+      "loss": 1.087,
+      "step": 23806
+    },
+    {
+      "epoch": 0.566671427211273,
+      "grad_norm": 3.3305565336174845,
+      "learning_rate": 2e-05,
+      "loss": 0.4854,
+      "step": 23807
+    },
+    {
+      "epoch": 0.5666952299343044,
+      "grad_norm": 6.777408043743418,
+      "learning_rate": 2e-05,
+      "loss": 0.7506,
+      "step": 23808
+    },
+    {
+      "epoch": 0.566719032657336,
+      "grad_norm": 1.0307321802452867,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 23809
+    },
+    {
+      "epoch": 0.5667428353803675,
+      "grad_norm": 3.00667065993181,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 23810
+    },
+    {
+      "epoch": 0.566766638103399,
+      "grad_norm": 1.1466557086448261,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 23811
+    },
+    {
+      "epoch": 0.5667904408264306,
+      "grad_norm": 0.18298042313469423,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 23812
+    },
+    {
+      "epoch": 0.5668142435494621,
+      "grad_norm": 2.891513894125871,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 23813
+    },
+    {
+      "epoch": 0.5668380462724936,
+      "grad_norm": 1.1311194055306582,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 23814
+    },
+    {
+      "epoch": 0.566861848995525,
+      "grad_norm": 0.36629094153855535,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 23815
+    },
+    {
+      "epoch": 0.5668856517185566,
+      "grad_norm": 0.8797448025968693,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 23816
+    },
+    {
+      "epoch": 0.5669094544415881,
+      "grad_norm": 0.2711926878868571,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 23817
+    },
+    {
+      "epoch": 0.5669332571646196,
+      "grad_norm": 4.632645283385919,
+      "learning_rate": 2e-05,
+      "loss": 0.6968,
+      "step": 23818
+    },
+    {
+      "epoch": 0.5669570598876511,
+      "grad_norm": 3.911051668603635,
+      "learning_rate": 2e-05,
+      "loss": 0.4821,
+      "step": 23819
+    },
+    {
+      "epoch": 0.5669808626106827,
+      "grad_norm": 5.802578247244645,
+      "learning_rate": 2e-05,
+      "loss": 0.9822,
+      "step": 23820
+    },
+    {
+      "epoch": 0.5670046653337142,
+      "grad_norm": 0.12097129038101989,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 23821
+    },
+    {
+      "epoch": 0.5670284680567457,
+      "grad_norm": 5.286157818272356,
+      "learning_rate": 2e-05,
+      "loss": 0.2084,
+      "step": 23822
+    },
+    {
+      "epoch": 0.5670522707797772,
+      "grad_norm": 2.885642124891662,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 23823
+    },
+    {
+      "epoch": 0.5670760735028088,
+      "grad_norm": 10.213225151167467,
+      "learning_rate": 2e-05,
+      "loss": 0.7651,
+      "step": 23824
+    },
+    {
+      "epoch": 0.5670998762258402,
+      "grad_norm": 5.742339641566172,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 23825
+    },
+    {
+      "epoch": 0.5671236789488717,
+      "grad_norm": 3.1413528599810285,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 23826
+    },
+    {
+      "epoch": 0.5671474816719033,
+      "grad_norm": 11.021121302487373,
+      "learning_rate": 2e-05,
+      "loss": 0.5959,
+      "step": 23827
+    },
+    {
+      "epoch": 0.5671712843949348,
+      "grad_norm": 0.07897002452897278,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 23828
+    },
+    {
+      "epoch": 0.5671950871179663,
+      "grad_norm": 3.61951872392117,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 23829
+    },
+    {
+      "epoch": 0.5672188898409978,
+      "grad_norm": 5.122319260456953,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 23830
+    },
+    {
+      "epoch": 0.5672426925640294,
+      "grad_norm": 1.9359963720388504,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 23831
+    },
+    {
+      "epoch": 0.5672664952870609,
+      "grad_norm": 0.11160369411464138,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 23832
+    },
+    {
+      "epoch": 0.5672902980100923,
+      "grad_norm": 3.0364298614736662,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 23833
+    },
+    {
+      "epoch": 0.5673141007331238,
+      "grad_norm": 11.130834033017111,
+      "learning_rate": 2e-05,
+      "loss": 0.5023,
+      "step": 23834
+    },
+    {
+      "epoch": 0.5673379034561554,
+      "grad_norm": 4.804741945918681,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 23835
+    },
+    {
+      "epoch": 0.5673617061791869,
+      "grad_norm": 2.865581131065412,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 23836
+    },
+    {
+      "epoch": 0.5673855089022184,
+      "grad_norm": 5.776784178092409,
+      "learning_rate": 2e-05,
+      "loss": 1.184,
+      "step": 23837
+    },
+    {
+      "epoch": 0.5674093116252499,
+      "grad_norm": 4.376074865570502,
+      "learning_rate": 2e-05,
+      "loss": 0.373,
+      "step": 23838
+    },
+    {
+      "epoch": 0.5674331143482815,
+      "grad_norm": 5.207199821837875,
+      "learning_rate": 2e-05,
+      "loss": 0.5252,
+      "step": 23839
+    },
+    {
+      "epoch": 0.567456917071313,
+      "grad_norm": 5.505307154352507,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 23840
+    },
+    {
+      "epoch": 0.5674807197943444,
+      "grad_norm": 10.878421646838481,
+      "learning_rate": 2e-05,
+      "loss": 0.4809,
+      "step": 23841
+    },
+    {
+      "epoch": 0.567504522517376,
+      "grad_norm": 4.835497584734523,
+      "learning_rate": 2e-05,
+      "loss": 0.7129,
+      "step": 23842
+    },
+    {
+      "epoch": 0.5675283252404075,
+      "grad_norm": 8.144683982129155,
+      "learning_rate": 2e-05,
+      "loss": 1.0722,
+      "step": 23843
+    },
+    {
+      "epoch": 0.567552127963439,
+      "grad_norm": 0.755344431020373,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 23844
+    },
+    {
+      "epoch": 0.5675759306864705,
+      "grad_norm": 4.243369981802588,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 23845
+    },
+    {
+      "epoch": 0.5675997334095021,
+      "grad_norm": 5.839987747258296,
+      "learning_rate": 2e-05,
+      "loss": 0.9049,
+      "step": 23846
+    },
+    {
+      "epoch": 0.5676235361325336,
+      "grad_norm": 5.707794829848002,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 23847
+    },
+    {
+      "epoch": 0.5676473388555651,
+      "grad_norm": 4.242525705216814,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 23848
+    },
+    {
+      "epoch": 0.5676711415785965,
+      "grad_norm": 5.120692801908223,
+      "learning_rate": 2e-05,
+      "loss": 0.1833,
+      "step": 23849
+    },
+    {
+      "epoch": 0.5676949443016281,
+      "grad_norm": 1.0872146476798945,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 23850
+    },
+    {
+      "epoch": 0.5677187470246596,
+      "grad_norm": 4.245199802546142,
+      "learning_rate": 2e-05,
+      "loss": 0.4357,
+      "step": 23851
+    },
+    {
+      "epoch": 0.5677425497476911,
+      "grad_norm": 3.0080906536959158,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 23852
+    },
+    {
+      "epoch": 0.5677663524707226,
+      "grad_norm": 4.854220402325814,
+      "learning_rate": 2e-05,
+      "loss": 0.8633,
+      "step": 23853
+    },
+    {
+      "epoch": 0.5677901551937542,
+      "grad_norm": 2.940380647615301,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 23854
+    },
+    {
+      "epoch": 0.5678139579167857,
+      "grad_norm": 0.11659471029544373,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 23855
+    },
+    {
+      "epoch": 0.5678377606398172,
+      "grad_norm": 1.7287003023666612,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 23856
+    },
+    {
+      "epoch": 0.5678615633628488,
+      "grad_norm": 8.745064207223175,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 23857
+    },
+    {
+      "epoch": 0.5678853660858802,
+      "grad_norm": 4.480468447753106,
+      "learning_rate": 2e-05,
+      "loss": 0.9777,
+      "step": 23858
+    },
+    {
+      "epoch": 0.5679091688089117,
+      "grad_norm": 1.0830527071309128,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 23859
+    },
+    {
+      "epoch": 0.5679329715319432,
+      "grad_norm": 1.4373043073606464,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 23860
+    },
+    {
+      "epoch": 0.5679567742549748,
+      "grad_norm": 3.745116212844519,
+      "learning_rate": 2e-05,
+      "loss": 0.9738,
+      "step": 23861
+    },
+    {
+      "epoch": 0.5679805769780063,
+      "grad_norm": 17.417713768159476,
+      "learning_rate": 2e-05,
+      "loss": 1.0909,
+      "step": 23862
+    },
+    {
+      "epoch": 0.5680043797010378,
+      "grad_norm": 10.405744496414531,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 23863
+    },
+    {
+      "epoch": 0.5680281824240693,
+      "grad_norm": 0.17209400469373473,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 23864
+    },
+    {
+      "epoch": 0.5680519851471009,
+      "grad_norm": 1.7436972645291193,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 23865
+    },
+    {
+      "epoch": 0.5680757878701324,
+      "grad_norm": 5.881862726097575,
+      "learning_rate": 2e-05,
+      "loss": 0.7211,
+      "step": 23866
+    },
+    {
+      "epoch": 0.5680995905931638,
+      "grad_norm": 10.419828542841843,
+      "learning_rate": 2e-05,
+      "loss": 1.3676,
+      "step": 23867
+    },
+    {
+      "epoch": 0.5681233933161953,
+      "grad_norm": 1.6944555981306288,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 23868
+    },
+    {
+      "epoch": 0.5681471960392269,
+      "grad_norm": 3.993081332501109,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 23869
+    },
+    {
+      "epoch": 0.5681709987622584,
+      "grad_norm": 4.375890758846013,
+      "learning_rate": 2e-05,
+      "loss": 0.9095,
+      "step": 23870
+    },
+    {
+      "epoch": 0.5681948014852899,
+      "grad_norm": 1.3596267978210466,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 23871
+    },
+    {
+      "epoch": 0.5682186042083215,
+      "grad_norm": 7.969006997202396,
+      "learning_rate": 2e-05,
+      "loss": 1.157,
+      "step": 23872
+    },
+    {
+      "epoch": 0.568242406931353,
+      "grad_norm": 3.4126930810396874,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 23873
+    },
+    {
+      "epoch": 0.5682662096543845,
+      "grad_norm": 0.36497889480347484,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 23874
+    },
+    {
+      "epoch": 0.5682900123774159,
+      "grad_norm": 2.612470695081606,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 23875
+    },
+    {
+      "epoch": 0.5683138151004475,
+      "grad_norm": 9.741798391484537,
+      "learning_rate": 2e-05,
+      "loss": 0.566,
+      "step": 23876
+    },
+    {
+      "epoch": 0.568337617823479,
+      "grad_norm": 3.519963036164852,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 23877
+    },
+    {
+      "epoch": 0.5683614205465105,
+      "grad_norm": 8.251226793868543,
+      "learning_rate": 2e-05,
+      "loss": 0.507,
+      "step": 23878
+    },
+    {
+      "epoch": 0.568385223269542,
+      "grad_norm": 8.165924791377137,
+      "learning_rate": 2e-05,
+      "loss": 1.1895,
+      "step": 23879
+    },
+    {
+      "epoch": 0.5684090259925736,
+      "grad_norm": 2.4604493473044373,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 23880
+    },
+    {
+      "epoch": 0.5684328287156051,
+      "grad_norm": 0.8194269272213284,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 23881
+    },
+    {
+      "epoch": 0.5684566314386366,
+      "grad_norm": 4.7417855182054875,
+      "learning_rate": 2e-05,
+      "loss": 1.3263,
+      "step": 23882
+    },
+    {
+      "epoch": 0.568480434161668,
+      "grad_norm": 6.078817310665593,
+      "learning_rate": 2e-05,
+      "loss": 0.3051,
+      "step": 23883
+    },
+    {
+      "epoch": 0.5685042368846996,
+      "grad_norm": 0.3563692178397229,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 23884
+    },
+    {
+      "epoch": 0.5685280396077311,
+      "grad_norm": 3.3337467788032056,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 23885
+    },
+    {
+      "epoch": 0.5685518423307626,
+      "grad_norm": 2.2423532988446646,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 23886
+    },
+    {
+      "epoch": 0.5685756450537941,
+      "grad_norm": 9.836634222867842,
+      "learning_rate": 2e-05,
+      "loss": 0.3989,
+      "step": 23887
+    },
+    {
+      "epoch": 0.5685994477768257,
+      "grad_norm": 1.8962239805300283,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 23888
+    },
+    {
+      "epoch": 0.5686232504998572,
+      "grad_norm": 6.536640951395031,
+      "learning_rate": 2e-05,
+      "loss": 1.4782,
+      "step": 23889
+    },
+    {
+      "epoch": 0.5686470532228887,
+      "grad_norm": 0.9940584621637025,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 23890
+    },
+    {
+      "epoch": 0.5686708559459203,
+      "grad_norm": 9.934328303167039,
+      "learning_rate": 2e-05,
+      "loss": 0.4643,
+      "step": 23891
+    },
+    {
+      "epoch": 0.5686946586689517,
+      "grad_norm": 1.8112972753270398,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 23892
+    },
+    {
+      "epoch": 0.5687184613919832,
+      "grad_norm": 5.789353136626013,
+      "learning_rate": 2e-05,
+      "loss": 1.3367,
+      "step": 23893
+    },
+    {
+      "epoch": 0.5687422641150147,
+      "grad_norm": 2.934613661218787,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 23894
+    },
+    {
+      "epoch": 0.5687660668380463,
+      "grad_norm": 1.3144191264579015,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 23895
+    },
+    {
+      "epoch": 0.5687898695610778,
+      "grad_norm": 2.726230846958437,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 23896
+    },
+    {
+      "epoch": 0.5688136722841093,
+      "grad_norm": 4.255575194121049,
+      "learning_rate": 2e-05,
+      "loss": 0.514,
+      "step": 23897
+    },
+    {
+      "epoch": 0.5688374750071408,
+      "grad_norm": 0.05377767056543172,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 23898
+    },
+    {
+      "epoch": 0.5688612777301724,
+      "grad_norm": 0.7915781996263327,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 23899
+    },
+    {
+      "epoch": 0.5688850804532039,
+      "grad_norm": 4.560483644935045,
+      "learning_rate": 2e-05,
+      "loss": 0.8865,
+      "step": 23900
+    },
+    {
+      "epoch": 0.5689088831762353,
+      "grad_norm": 3.1892615409963843,
+      "learning_rate": 2e-05,
+      "loss": 0.2573,
+      "step": 23901
+    },
+    {
+      "epoch": 0.5689326858992668,
+      "grad_norm": 0.02385030092885196,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 23902
+    },
+    {
+      "epoch": 0.5689564886222984,
+      "grad_norm": 4.8843153087341,
+      "learning_rate": 2e-05,
+      "loss": 0.4912,
+      "step": 23903
+    },
+    {
+      "epoch": 0.5689802913453299,
+      "grad_norm": 9.883581964791846,
+      "learning_rate": 2e-05,
+      "loss": 0.7693,
+      "step": 23904
+    },
+    {
+      "epoch": 0.5690040940683614,
+      "grad_norm": 4.892457963283395,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 23905
+    },
+    {
+      "epoch": 0.569027896791393,
+      "grad_norm": 2.4355468217679883,
+      "learning_rate": 2e-05,
+      "loss": 0.4289,
+      "step": 23906
+    },
+    {
+      "epoch": 0.5690516995144245,
+      "grad_norm": 5.735355972142253,
+      "learning_rate": 2e-05,
+      "loss": 1.1729,
+      "step": 23907
+    },
+    {
+      "epoch": 0.569075502237456,
+      "grad_norm": 2.1548817734639423,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 23908
+    },
+    {
+      "epoch": 0.5690993049604874,
+      "grad_norm": 3.752742838128631,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 23909
+    },
+    {
+      "epoch": 0.569123107683519,
+      "grad_norm": 3.6929223880597446,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 23910
+    },
+    {
+      "epoch": 0.5691469104065505,
+      "grad_norm": 1.0656827661243407,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 23911
+    },
+    {
+      "epoch": 0.569170713129582,
+      "grad_norm": 3.1845826210097594,
+      "learning_rate": 2e-05,
+      "loss": 0.5559,
+      "step": 23912
+    },
+    {
+      "epoch": 0.5691945158526135,
+      "grad_norm": 8.037123203006557,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 23913
+    },
+    {
+      "epoch": 0.5692183185756451,
+      "grad_norm": 3.3984598406799016,
+      "learning_rate": 2e-05,
+      "loss": 0.4096,
+      "step": 23914
+    },
+    {
+      "epoch": 0.5692421212986766,
+      "grad_norm": 6.434675418670305,
+      "learning_rate": 2e-05,
+      "loss": 1.1429,
+      "step": 23915
+    },
+    {
+      "epoch": 0.5692659240217081,
+      "grad_norm": 9.064462386249854,
+      "learning_rate": 2e-05,
+      "loss": 1.2716,
+      "step": 23916
+    },
+    {
+      "epoch": 0.5692897267447395,
+      "grad_norm": 4.353736297489082,
+      "learning_rate": 2e-05,
+      "loss": 0.7742,
+      "step": 23917
+    },
+    {
+      "epoch": 0.5693135294677711,
+      "grad_norm": 5.0794830752400335,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 23918
+    },
+    {
+      "epoch": 0.5693373321908026,
+      "grad_norm": 2.6301582439483786,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 23919
+    },
+    {
+      "epoch": 0.5693611349138341,
+      "grad_norm": 0.11223489070964228,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 23920
+    },
+    {
+      "epoch": 0.5693849376368657,
+      "grad_norm": 4.139995164045902,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 23921
+    },
+    {
+      "epoch": 0.5694087403598972,
+      "grad_norm": 6.833516679472486,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 23922
+    },
+    {
+      "epoch": 0.5694325430829287,
+      "grad_norm": 4.197094522438935,
+      "learning_rate": 2e-05,
+      "loss": 0.0936,
+      "step": 23923
+    },
+    {
+      "epoch": 0.5694563458059602,
+      "grad_norm": 10.454675745697308,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 23924
+    },
+    {
+      "epoch": 0.5694801485289918,
+      "grad_norm": 2.541687992666981,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 23925
+    },
+    {
+      "epoch": 0.5695039512520232,
+      "grad_norm": 4.286611289071804,
+      "learning_rate": 2e-05,
+      "loss": 0.242,
+      "step": 23926
+    },
+    {
+      "epoch": 0.5695277539750547,
+      "grad_norm": 4.924603805926758,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 23927
+    },
+    {
+      "epoch": 0.5695515566980862,
+      "grad_norm": 2.3819715671024086,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 23928
+    },
+    {
+      "epoch": 0.5695753594211178,
+      "grad_norm": 1.2954106498137075,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 23929
+    },
+    {
+      "epoch": 0.5695991621441493,
+      "grad_norm": 7.026117002724451,
+      "learning_rate": 2e-05,
+      "loss": 0.895,
+      "step": 23930
+    },
+    {
+      "epoch": 0.5696229648671808,
+      "grad_norm": 1.5008785952328605,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 23931
+    },
+    {
+      "epoch": 0.5696467675902123,
+      "grad_norm": 8.610441557328857,
+      "learning_rate": 2e-05,
+      "loss": 0.3394,
+      "step": 23932
+    },
+    {
+      "epoch": 0.5696705703132439,
+      "grad_norm": 39.533761942733,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 23933
+    },
+    {
+      "epoch": 0.5696943730362753,
+      "grad_norm": 3.703086150404602,
+      "learning_rate": 2e-05,
+      "loss": 0.5719,
+      "step": 23934
+    },
+    {
+      "epoch": 0.5697181757593068,
+      "grad_norm": 96.83893595114685,
+      "learning_rate": 2e-05,
+      "loss": 1.0479,
+      "step": 23935
+    },
+    {
+      "epoch": 0.5697419784823384,
+      "grad_norm": 0.025047145167007474,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 23936
+    },
+    {
+      "epoch": 0.5697657812053699,
+      "grad_norm": 14.567465038509251,
+      "learning_rate": 2e-05,
+      "loss": 0.8131,
+      "step": 23937
+    },
+    {
+      "epoch": 0.5697895839284014,
+      "grad_norm": 2.4234719305926142,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 23938
+    },
+    {
+      "epoch": 0.5698133866514329,
+      "grad_norm": 0.7880925609601619,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 23939
+    },
+    {
+      "epoch": 0.5698371893744645,
+      "grad_norm": 4.302745006322182,
+      "learning_rate": 2e-05,
+      "loss": 0.6535,
+      "step": 23940
+    },
+    {
+      "epoch": 0.569860992097496,
+      "grad_norm": 3.465034072995439,
+      "learning_rate": 2e-05,
+      "loss": 0.448,
+      "step": 23941
+    },
+    {
+      "epoch": 0.5698847948205275,
+      "grad_norm": 0.16797202297570205,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 23942
+    },
+    {
+      "epoch": 0.5699085975435589,
+      "grad_norm": 1.6296992406155102,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 23943
+    },
+    {
+      "epoch": 0.5699324002665905,
+      "grad_norm": 0.12520360271349695,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 23944
+    },
+    {
+      "epoch": 0.569956202989622,
+      "grad_norm": 4.1361579043954,
+      "learning_rate": 2e-05,
+      "loss": 0.1562,
+      "step": 23945
+    },
+    {
+      "epoch": 0.5699800057126535,
+      "grad_norm": 3.520722742141676,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 23946
+    },
+    {
+      "epoch": 0.570003808435685,
+      "grad_norm": 4.343991656986132,
+      "learning_rate": 2e-05,
+      "loss": 0.5339,
+      "step": 23947
+    },
+    {
+      "epoch": 0.5700276111587166,
+      "grad_norm": 3.75745762912396,
+      "learning_rate": 2e-05,
+      "loss": 0.8367,
+      "step": 23948
+    },
+    {
+      "epoch": 0.5700514138817481,
+      "grad_norm": 3.780666614857049,
+      "learning_rate": 2e-05,
+      "loss": 0.6764,
+      "step": 23949
+    },
+    {
+      "epoch": 0.5700752166047796,
+      "grad_norm": 4.464949652268186,
+      "learning_rate": 2e-05,
+      "loss": 0.9001,
+      "step": 23950
+    },
+    {
+      "epoch": 0.5700990193278112,
+      "grad_norm": 0.5347055496015635,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 23951
+    },
+    {
+      "epoch": 0.5701228220508426,
+      "grad_norm": 3.0740223784777303,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 23952
+    },
+    {
+      "epoch": 0.5701466247738741,
+      "grad_norm": 1.402563928489668,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 23953
+    },
+    {
+      "epoch": 0.5701704274969056,
+      "grad_norm": 3.9913611973153813,
+      "learning_rate": 2e-05,
+      "loss": 0.4655,
+      "step": 23954
+    },
+    {
+      "epoch": 0.5701942302199372,
+      "grad_norm": 1.965335686205779,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 23955
+    },
+    {
+      "epoch": 0.5702180329429687,
+      "grad_norm": 5.339113091992387,
+      "learning_rate": 2e-05,
+      "loss": 1.1056,
+      "step": 23956
+    },
+    {
+      "epoch": 0.5702418356660002,
+      "grad_norm": 0.10225731156588609,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 23957
+    },
+    {
+      "epoch": 0.5702656383890317,
+      "grad_norm": 4.501422020064661,
+      "learning_rate": 2e-05,
+      "loss": 0.1874,
+      "step": 23958
+    },
+    {
+      "epoch": 0.5702894411120633,
+      "grad_norm": 4.5059388509735205,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 23959
+    },
+    {
+      "epoch": 0.5703132438350947,
+      "grad_norm": 10.59042319420478,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 23960
+    },
+    {
+      "epoch": 0.5703370465581262,
+      "grad_norm": 5.8675357263761985,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 23961
+    },
+    {
+      "epoch": 0.5703608492811577,
+      "grad_norm": 0.9658378962434206,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 23962
+    },
+    {
+      "epoch": 0.5703846520041893,
+      "grad_norm": 3.094408385706755,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 23963
+    },
+    {
+      "epoch": 0.5704084547272208,
+      "grad_norm": 0.6382334508665412,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 23964
+    },
+    {
+      "epoch": 0.5704322574502523,
+      "grad_norm": 6.914832079804019,
+      "learning_rate": 2e-05,
+      "loss": 0.2067,
+      "step": 23965
+    },
+    {
+      "epoch": 0.5704560601732839,
+      "grad_norm": 3.914604083940625,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 23966
+    },
+    {
+      "epoch": 0.5704798628963154,
+      "grad_norm": 1.2568564305243328,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 23967
+    },
+    {
+      "epoch": 0.5705036656193468,
+      "grad_norm": 3.9848595327813046,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 23968
+    },
+    {
+      "epoch": 0.5705274683423783,
+      "grad_norm": 9.452471048580662,
+      "learning_rate": 2e-05,
+      "loss": 0.6349,
+      "step": 23969
+    },
+    {
+      "epoch": 0.5705512710654099,
+      "grad_norm": 5.569897893271892,
+      "learning_rate": 2e-05,
+      "loss": 0.9905,
+      "step": 23970
+    },
+    {
+      "epoch": 0.5705750737884414,
+      "grad_norm": 2.908004869994059,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 23971
+    },
+    {
+      "epoch": 0.5705988765114729,
+      "grad_norm": 3.478804104739504,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 23972
+    },
+    {
+      "epoch": 0.5706226792345044,
+      "grad_norm": 0.6153537057571272,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 23973
+    },
+    {
+      "epoch": 0.570646481957536,
+      "grad_norm": 5.137981902827846,
+      "learning_rate": 2e-05,
+      "loss": 0.9127,
+      "step": 23974
+    },
+    {
+      "epoch": 0.5706702846805675,
+      "grad_norm": 9.324581623797735,
+      "learning_rate": 2e-05,
+      "loss": 0.4342,
+      "step": 23975
+    },
+    {
+      "epoch": 0.570694087403599,
+      "grad_norm": 2.586736352522972,
+      "learning_rate": 2e-05,
+      "loss": 0.3413,
+      "step": 23976
+    },
+    {
+      "epoch": 0.5707178901266304,
+      "grad_norm": 5.748817947000745,
+      "learning_rate": 2e-05,
+      "loss": 0.0969,
+      "step": 23977
+    },
+    {
+      "epoch": 0.570741692849662,
+      "grad_norm": 3.912549553678283,
+      "learning_rate": 2e-05,
+      "loss": 0.5632,
+      "step": 23978
+    },
+    {
+      "epoch": 0.5707654955726935,
+      "grad_norm": 3.4331580179351953,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 23979
+    },
+    {
+      "epoch": 0.570789298295725,
+      "grad_norm": 7.3627759343773205,
+      "learning_rate": 2e-05,
+      "loss": 0.1384,
+      "step": 23980
+    },
+    {
+      "epoch": 0.5708131010187566,
+      "grad_norm": 3.2765140881162913,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 23981
+    },
+    {
+      "epoch": 0.5708369037417881,
+      "grad_norm": 3.1984622355268892,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 23982
+    },
+    {
+      "epoch": 0.5708607064648196,
+      "grad_norm": 0.46742113505984534,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 23983
+    },
+    {
+      "epoch": 0.570884509187851,
+      "grad_norm": 0.14180641956521275,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 23984
+    },
+    {
+      "epoch": 0.5709083119108826,
+      "grad_norm": 13.423703963365877,
+      "learning_rate": 2e-05,
+      "loss": 1.4394,
+      "step": 23985
+    },
+    {
+      "epoch": 0.5709321146339141,
+      "grad_norm": 13.049269684366868,
+      "learning_rate": 2e-05,
+      "loss": 0.4439,
+      "step": 23986
+    },
+    {
+      "epoch": 0.5709559173569456,
+      "grad_norm": 1.6169793039643852,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 23987
+    },
+    {
+      "epoch": 0.5709797200799771,
+      "grad_norm": 2.9845019719808237,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 23988
+    },
+    {
+      "epoch": 0.5710035228030087,
+      "grad_norm": 5.991037590140435,
+      "learning_rate": 2e-05,
+      "loss": 0.8124,
+      "step": 23989
+    },
+    {
+      "epoch": 0.5710273255260402,
+      "grad_norm": 3.606971691573531,
+      "learning_rate": 2e-05,
+      "loss": 0.1205,
+      "step": 23990
+    },
+    {
+      "epoch": 0.5710511282490717,
+      "grad_norm": 1.9822910822938404,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 23991
+    },
+    {
+      "epoch": 0.5710749309721032,
+      "grad_norm": 9.678236517757043,
+      "learning_rate": 2e-05,
+      "loss": 0.2396,
+      "step": 23992
+    },
+    {
+      "epoch": 0.5710987336951348,
+      "grad_norm": 1.1106829729319825,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 23993
+    },
+    {
+      "epoch": 0.5711225364181662,
+      "grad_norm": 25.98046249812792,
+      "learning_rate": 2e-05,
+      "loss": 0.6188,
+      "step": 23994
+    },
+    {
+      "epoch": 0.5711463391411977,
+      "grad_norm": 1.2506013295352267,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 23995
+    },
+    {
+      "epoch": 0.5711701418642293,
+      "grad_norm": 1.2800797835989863,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 23996
+    },
+    {
+      "epoch": 0.5711939445872608,
+      "grad_norm": 8.826494408590218,
+      "learning_rate": 2e-05,
+      "loss": 0.3658,
+      "step": 23997
+    },
+    {
+      "epoch": 0.5712177473102923,
+      "grad_norm": 10.07615440064539,
+      "learning_rate": 2e-05,
+      "loss": 1.1778,
+      "step": 23998
+    },
+    {
+      "epoch": 0.5712415500333238,
+      "grad_norm": 0.6855363416966315,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 23999
+    },
+    {
+      "epoch": 0.5712653527563554,
+      "grad_norm": 3.2648233274834086,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 24000
+    },
+    {
+      "epoch": 0.5712891554793869,
+      "grad_norm": 6.720670499915337,
+      "learning_rate": 2e-05,
+      "loss": 0.6215,
+      "step": 24001
+    },
+    {
+      "epoch": 0.5713129582024183,
+      "grad_norm": 0.985732993022781,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 24002
+    },
+    {
+      "epoch": 0.5713367609254498,
+      "grad_norm": 4.700155324091536,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 24003
+    },
+    {
+      "epoch": 0.5713605636484814,
+      "grad_norm": 3.8843131359241863,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 24004
+    },
+    {
+      "epoch": 0.5713843663715129,
+      "grad_norm": 0.8515727945297653,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 24005
+    },
+    {
+      "epoch": 0.5714081690945444,
+      "grad_norm": 5.20951085438382,
+      "learning_rate": 2e-05,
+      "loss": 1.0723,
+      "step": 24006
+    },
+    {
+      "epoch": 0.5714319718175759,
+      "grad_norm": 5.737467427709999,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 24007
+    },
+    {
+      "epoch": 0.5714557745406075,
+      "grad_norm": 3.0653822844928973,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 24008
+    },
+    {
+      "epoch": 0.571479577263639,
+      "grad_norm": 8.200082250136209,
+      "learning_rate": 2e-05,
+      "loss": 0.9161,
+      "step": 24009
+    },
+    {
+      "epoch": 0.5715033799866704,
+      "grad_norm": 1.3417856570772733,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 24010
+    },
+    {
+      "epoch": 0.571527182709702,
+      "grad_norm": 7.881307032502403,
+      "learning_rate": 2e-05,
+      "loss": 0.2745,
+      "step": 24011
+    },
+    {
+      "epoch": 0.5715509854327335,
+      "grad_norm": 9.697149015990933,
+      "learning_rate": 2e-05,
+      "loss": 0.5638,
+      "step": 24012
+    },
+    {
+      "epoch": 0.571574788155765,
+      "grad_norm": 2.8589096978095845,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 24013
+    },
+    {
+      "epoch": 0.5715985908787965,
+      "grad_norm": 3.7032027595461785,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 24014
+    },
+    {
+      "epoch": 0.5716223936018281,
+      "grad_norm": 6.823927283741459,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 24015
+    },
+    {
+      "epoch": 0.5716461963248596,
+      "grad_norm": 5.06280608551771,
+      "learning_rate": 2e-05,
+      "loss": 1.0642,
+      "step": 24016
+    },
+    {
+      "epoch": 0.5716699990478911,
+      "grad_norm": 2.8021477204028407,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 24017
+    },
+    {
+      "epoch": 0.5716938017709225,
+      "grad_norm": 4.5203440756165305,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 24018
+    },
+    {
+      "epoch": 0.5717176044939541,
+      "grad_norm": 1.394939727785483,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 24019
+    },
+    {
+      "epoch": 0.5717414072169856,
+      "grad_norm": 4.595322736743299,
+      "learning_rate": 2e-05,
+      "loss": 0.9464,
+      "step": 24020
+    },
+    {
+      "epoch": 0.5717652099400171,
+      "grad_norm": 2.217599813996039,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 24021
+    },
+    {
+      "epoch": 0.5717890126630486,
+      "grad_norm": 3.6154180919525123,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 24022
+    },
+    {
+      "epoch": 0.5718128153860802,
+      "grad_norm": 12.154980090693147,
+      "learning_rate": 2e-05,
+      "loss": 0.9625,
+      "step": 24023
+    },
+    {
+      "epoch": 0.5718366181091117,
+      "grad_norm": 2.455947692988215,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 24024
+    },
+    {
+      "epoch": 0.5718604208321432,
+      "grad_norm": 9.33552262196622,
+      "learning_rate": 2e-05,
+      "loss": 0.3574,
+      "step": 24025
+    },
+    {
+      "epoch": 0.5718842235551748,
+      "grad_norm": 1.1640502186573407,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 24026
+    },
+    {
+      "epoch": 0.5719080262782062,
+      "grad_norm": 0.6387508513606,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 24027
+    },
+    {
+      "epoch": 0.5719318290012377,
+      "grad_norm": 0.4287032854508466,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 24028
+    },
+    {
+      "epoch": 0.5719556317242692,
+      "grad_norm": 3.781353194830193,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 24029
+    },
+    {
+      "epoch": 0.5719794344473008,
+      "grad_norm": 3.129141883963844,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 24030
+    },
+    {
+      "epoch": 0.5720032371703323,
+      "grad_norm": 1.6863993995610262,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 24031
+    },
+    {
+      "epoch": 0.5720270398933638,
+      "grad_norm": 4.263071725995157,
+      "learning_rate": 2e-05,
+      "loss": 0.4794,
+      "step": 24032
+    },
+    {
+      "epoch": 0.5720508426163953,
+      "grad_norm": 0.021812231193266682,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 24033
+    },
+    {
+      "epoch": 0.5720746453394269,
+      "grad_norm": 4.16551571803075,
+      "learning_rate": 2e-05,
+      "loss": 0.661,
+      "step": 24034
+    },
+    {
+      "epoch": 0.5720984480624584,
+      "grad_norm": 3.2747163434835564,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 24035
+    },
+    {
+      "epoch": 0.5721222507854898,
+      "grad_norm": 22.40699107738367,
+      "learning_rate": 2e-05,
+      "loss": 1.8607,
+      "step": 24036
+    },
+    {
+      "epoch": 0.5721460535085213,
+      "grad_norm": 2.6330914011734263,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 24037
+    },
+    {
+      "epoch": 0.5721698562315529,
+      "grad_norm": 4.427532082638582,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 24038
+    },
+    {
+      "epoch": 0.5721936589545844,
+      "grad_norm": 5.741234268275697,
+      "learning_rate": 2e-05,
+      "loss": 1.2945,
+      "step": 24039
+    },
+    {
+      "epoch": 0.5722174616776159,
+      "grad_norm": 3.2927895697862457,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 24040
+    },
+    {
+      "epoch": 0.5722412644006475,
+      "grad_norm": 0.9698937808761148,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 24041
+    },
+    {
+      "epoch": 0.572265067123679,
+      "grad_norm": 0.22950772066908753,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 24042
+    },
+    {
+      "epoch": 0.5722888698467105,
+      "grad_norm": 3.4079375669611967,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 24043
+    },
+    {
+      "epoch": 0.5723126725697419,
+      "grad_norm": 5.124787189353555,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 24044
+    },
+    {
+      "epoch": 0.5723364752927735,
+      "grad_norm": 2.151178057158224,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 24045
+    },
+    {
+      "epoch": 0.572360278015805,
+      "grad_norm": 1.3169836534155526,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 24046
+    },
+    {
+      "epoch": 0.5723840807388365,
+      "grad_norm": 0.9025939886142182,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 24047
+    },
+    {
+      "epoch": 0.572407883461868,
+      "grad_norm": 1.2914171862218415,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 24048
+    },
+    {
+      "epoch": 0.5724316861848996,
+      "grad_norm": 4.231997517749036,
+      "learning_rate": 2e-05,
+      "loss": 0.5236,
+      "step": 24049
+    },
+    {
+      "epoch": 0.5724554889079311,
+      "grad_norm": 1.777125391122027,
+      "learning_rate": 2e-05,
+      "loss": 0.2411,
+      "step": 24050
+    },
+    {
+      "epoch": 0.5724792916309626,
+      "grad_norm": 2.0084401061658346,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 24051
+    },
+    {
+      "epoch": 0.572503094353994,
+      "grad_norm": 9.237797751413739,
+      "learning_rate": 2e-05,
+      "loss": 1.7247,
+      "step": 24052
+    },
+    {
+      "epoch": 0.5725268970770256,
+      "grad_norm": 6.869662208514839,
+      "learning_rate": 2e-05,
+      "loss": 0.5108,
+      "step": 24053
+    },
+    {
+      "epoch": 0.5725506998000571,
+      "grad_norm": 8.460981363077451,
+      "learning_rate": 2e-05,
+      "loss": 0.1036,
+      "step": 24054
+    },
+    {
+      "epoch": 0.5725745025230886,
+      "grad_norm": 15.297609606676842,
+      "learning_rate": 2e-05,
+      "loss": 1.5173,
+      "step": 24055
+    },
+    {
+      "epoch": 0.5725983052461202,
+      "grad_norm": 0.08389637580523912,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 24056
+    },
+    {
+      "epoch": 0.5726221079691517,
+      "grad_norm": 0.3926985096977133,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 24057
+    },
+    {
+      "epoch": 0.5726459106921832,
+      "grad_norm": 3.161194085562658,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 24058
+    },
+    {
+      "epoch": 0.5726697134152147,
+      "grad_norm": 4.200362966370056,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 24059
+    },
+    {
+      "epoch": 0.5726935161382463,
+      "grad_norm": 4.950675183982472,
+      "learning_rate": 2e-05,
+      "loss": 0.7997,
+      "step": 24060
+    },
+    {
+      "epoch": 0.5727173188612777,
+      "grad_norm": 7.1822769904801875,
+      "learning_rate": 2e-05,
+      "loss": 1.1596,
+      "step": 24061
+    },
+    {
+      "epoch": 0.5727411215843092,
+      "grad_norm": 2.9010204093460588,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 24062
+    },
+    {
+      "epoch": 0.5727649243073407,
+      "grad_norm": 2.946167311935691,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 24063
+    },
+    {
+      "epoch": 0.5727887270303723,
+      "grad_norm": 4.358721723878023,
+      "learning_rate": 2e-05,
+      "loss": 0.8686,
+      "step": 24064
+    },
+    {
+      "epoch": 0.5728125297534038,
+      "grad_norm": 2.183873899203516,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 24065
+    },
+    {
+      "epoch": 0.5728363324764353,
+      "grad_norm": 5.587803469566814,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 24066
+    },
+    {
+      "epoch": 0.5728601351994668,
+      "grad_norm": 2.003831246867892,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 24067
+    },
+    {
+      "epoch": 0.5728839379224984,
+      "grad_norm": 9.988881242445748,
+      "learning_rate": 2e-05,
+      "loss": 0.5205,
+      "step": 24068
+    },
+    {
+      "epoch": 0.5729077406455299,
+      "grad_norm": 1.694411871992916,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 24069
+    },
+    {
+      "epoch": 0.5729315433685613,
+      "grad_norm": 6.762795854006584,
+      "learning_rate": 2e-05,
+      "loss": 2.0332,
+      "step": 24070
+    },
+    {
+      "epoch": 0.5729553460915929,
+      "grad_norm": 1.5546731123732014,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 24071
+    },
+    {
+      "epoch": 0.5729791488146244,
+      "grad_norm": 6.629992890470628,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 24072
+    },
+    {
+      "epoch": 0.5730029515376559,
+      "grad_norm": 0.37008738793234947,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 24073
+    },
+    {
+      "epoch": 0.5730267542606874,
+      "grad_norm": 5.641886569355234,
+      "learning_rate": 2e-05,
+      "loss": 0.72,
+      "step": 24074
+    },
+    {
+      "epoch": 0.573050556983719,
+      "grad_norm": 5.607315742562505,
+      "learning_rate": 2e-05,
+      "loss": 0.7412,
+      "step": 24075
+    },
+    {
+      "epoch": 0.5730743597067505,
+      "grad_norm": 4.79645722754368,
+      "learning_rate": 2e-05,
+      "loss": 0.6539,
+      "step": 24076
+    },
+    {
+      "epoch": 0.573098162429782,
+      "grad_norm": 4.125355036792001,
+      "learning_rate": 2e-05,
+      "loss": 0.8125,
+      "step": 24077
+    },
+    {
+      "epoch": 0.5731219651528134,
+      "grad_norm": 3.1076843978959183,
+      "learning_rate": 2e-05,
+      "loss": 0.6116,
+      "step": 24078
+    },
+    {
+      "epoch": 0.573145767875845,
+      "grad_norm": 6.951656758731519,
+      "learning_rate": 2e-05,
+      "loss": 1.5918,
+      "step": 24079
+    },
+    {
+      "epoch": 0.5731695705988765,
+      "grad_norm": 2.7514896643766114,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 24080
+    },
+    {
+      "epoch": 0.573193373321908,
+      "grad_norm": 4.700400349447967,
+      "learning_rate": 2e-05,
+      "loss": 0.5407,
+      "step": 24081
+    },
+    {
+      "epoch": 0.5732171760449395,
+      "grad_norm": 1.144333023092071,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 24082
+    },
+    {
+      "epoch": 0.5732409787679711,
+      "grad_norm": 4.511818184431574,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 24083
+    },
+    {
+      "epoch": 0.5732647814910026,
+      "grad_norm": 4.536047666368324,
+      "learning_rate": 2e-05,
+      "loss": 1.2008,
+      "step": 24084
+    },
+    {
+      "epoch": 0.5732885842140341,
+      "grad_norm": 0.41215483786893187,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 24085
+    },
+    {
+      "epoch": 0.5733123869370657,
+      "grad_norm": 0.4678340630935239,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 24086
+    },
+    {
+      "epoch": 0.5733361896600971,
+      "grad_norm": 3.923447284006332,
+      "learning_rate": 2e-05,
+      "loss": 0.0952,
+      "step": 24087
+    },
+    {
+      "epoch": 0.5733599923831286,
+      "grad_norm": 0.41100144135823835,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 24088
+    },
+    {
+      "epoch": 0.5733837951061601,
+      "grad_norm": 5.346698908310024,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 24089
+    },
+    {
+      "epoch": 0.5734075978291917,
+      "grad_norm": 0.08827089949987722,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 24090
+    },
+    {
+      "epoch": 0.5734314005522232,
+      "grad_norm": 3.9468228477178253,
+      "learning_rate": 2e-05,
+      "loss": 0.6237,
+      "step": 24091
+    },
+    {
+      "epoch": 0.5734552032752547,
+      "grad_norm": 0.29091908763302315,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 24092
+    },
+    {
+      "epoch": 0.5734790059982862,
+      "grad_norm": 5.189332965150831,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 24093
+    },
+    {
+      "epoch": 0.5735028087213178,
+      "grad_norm": 12.11024828135778,
+      "learning_rate": 2e-05,
+      "loss": 1.1495,
+      "step": 24094
+    },
+    {
+      "epoch": 0.5735266114443492,
+      "grad_norm": 4.490162585870365,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 24095
+    },
+    {
+      "epoch": 0.5735504141673807,
+      "grad_norm": 9.794893723143442,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 24096
+    },
+    {
+      "epoch": 0.5735742168904122,
+      "grad_norm": 1.368429842232597,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 24097
+    },
+    {
+      "epoch": 0.5735980196134438,
+      "grad_norm": 4.5363772842736,
+      "learning_rate": 2e-05,
+      "loss": 0.1931,
+      "step": 24098
+    },
+    {
+      "epoch": 0.5736218223364753,
+      "grad_norm": 0.4128138249341741,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 24099
+    },
+    {
+      "epoch": 0.5736456250595068,
+      "grad_norm": 0.6738429289532912,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 24100
+    },
+    {
+      "epoch": 0.5736694277825384,
+      "grad_norm": 3.7915605497616425,
+      "learning_rate": 2e-05,
+      "loss": 0.2697,
+      "step": 24101
+    },
+    {
+      "epoch": 0.5736932305055699,
+      "grad_norm": 22.808550145265972,
+      "learning_rate": 2e-05,
+      "loss": 1.0665,
+      "step": 24102
+    },
+    {
+      "epoch": 0.5737170332286013,
+      "grad_norm": 6.917995901807409,
+      "learning_rate": 2e-05,
+      "loss": 0.7648,
+      "step": 24103
+    },
+    {
+      "epoch": 0.5737408359516328,
+      "grad_norm": 0.33779135576999564,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 24104
+    },
+    {
+      "epoch": 0.5737646386746644,
+      "grad_norm": 1.38376482634959,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 24105
+    },
+    {
+      "epoch": 0.5737884413976959,
+      "grad_norm": 3.136819088033883,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 24106
+    },
+    {
+      "epoch": 0.5738122441207274,
+      "grad_norm": 1.7034220131429023,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 24107
+    },
+    {
+      "epoch": 0.5738360468437589,
+      "grad_norm": 2.331721187902384,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 24108
+    },
+    {
+      "epoch": 0.5738598495667905,
+      "grad_norm": 0.4667434918705581,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 24109
+    },
+    {
+      "epoch": 0.573883652289822,
+      "grad_norm": 4.995852379205873,
+      "learning_rate": 2e-05,
+      "loss": 0.1665,
+      "step": 24110
+    },
+    {
+      "epoch": 0.5739074550128535,
+      "grad_norm": 4.435523383259425,
+      "learning_rate": 2e-05,
+      "loss": 0.6108,
+      "step": 24111
+    },
+    {
+      "epoch": 0.5739312577358849,
+      "grad_norm": 3.348778355693933,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 24112
+    },
+    {
+      "epoch": 0.5739550604589165,
+      "grad_norm": 2.2193550161637323,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 24113
+    },
+    {
+      "epoch": 0.573978863181948,
+      "grad_norm": 0.22971623354561352,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 24114
+    },
+    {
+      "epoch": 0.5740026659049795,
+      "grad_norm": 9.5911479703435,
+      "learning_rate": 2e-05,
+      "loss": 0.7378,
+      "step": 24115
+    },
+    {
+      "epoch": 0.5740264686280111,
+      "grad_norm": 1.3055823330643197,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 24116
+    },
+    {
+      "epoch": 0.5740502713510426,
+      "grad_norm": 3.5331942483305157,
+      "learning_rate": 2e-05,
+      "loss": 0.7725,
+      "step": 24117
+    },
+    {
+      "epoch": 0.5740740740740741,
+      "grad_norm": 14.243248068279211,
+      "learning_rate": 2e-05,
+      "loss": 1.2323,
+      "step": 24118
+    },
+    {
+      "epoch": 0.5740978767971056,
+      "grad_norm": 0.8414592329229249,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 24119
+    },
+    {
+      "epoch": 0.5741216795201372,
+      "grad_norm": 3.131904639993521,
+      "learning_rate": 2e-05,
+      "loss": 0.322,
+      "step": 24120
+    },
+    {
+      "epoch": 0.5741454822431686,
+      "grad_norm": 2.7176654533787823,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 24121
+    },
+    {
+      "epoch": 0.5741692849662001,
+      "grad_norm": 4.33724062353247,
+      "learning_rate": 2e-05,
+      "loss": 0.555,
+      "step": 24122
+    },
+    {
+      "epoch": 0.5741930876892316,
+      "grad_norm": 5.346952258974706,
+      "learning_rate": 2e-05,
+      "loss": 0.9203,
+      "step": 24123
+    },
+    {
+      "epoch": 0.5742168904122632,
+      "grad_norm": 2.9272226011041718,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 24124
+    },
+    {
+      "epoch": 0.5742406931352947,
+      "grad_norm": 0.09446769793835441,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 24125
+    },
+    {
+      "epoch": 0.5742644958583262,
+      "grad_norm": 9.721528754204842,
+      "learning_rate": 2e-05,
+      "loss": 0.4716,
+      "step": 24126
+    },
+    {
+      "epoch": 0.5742882985813577,
+      "grad_norm": 2.153823533980214,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 24127
+    },
+    {
+      "epoch": 0.5743121013043893,
+      "grad_norm": 10.296655311353089,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 24128
+    },
+    {
+      "epoch": 0.5743359040274207,
+      "grad_norm": 4.131226656177349,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 24129
+    },
+    {
+      "epoch": 0.5743597067504522,
+      "grad_norm": 0.15054385536744752,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 24130
+    },
+    {
+      "epoch": 0.5743835094734838,
+      "grad_norm": 0.2554593271985833,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 24131
+    },
+    {
+      "epoch": 0.5744073121965153,
+      "grad_norm": 4.813580657377216,
+      "learning_rate": 2e-05,
+      "loss": 0.2058,
+      "step": 24132
+    },
+    {
+      "epoch": 0.5744311149195468,
+      "grad_norm": 1.2339265467215903,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 24133
+    },
+    {
+      "epoch": 0.5744549176425783,
+      "grad_norm": 0.7072420808236083,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 24134
+    },
+    {
+      "epoch": 0.5744787203656099,
+      "grad_norm": 5.107519355718142,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 24135
+    },
+    {
+      "epoch": 0.5745025230886414,
+      "grad_norm": 3.406801247054666,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 24136
+    },
+    {
+      "epoch": 0.5745263258116728,
+      "grad_norm": 3.7538839390969705,
+      "learning_rate": 2e-05,
+      "loss": 0.6378,
+      "step": 24137
+    },
+    {
+      "epoch": 0.5745501285347043,
+      "grad_norm": 5.406099708633586,
+      "learning_rate": 2e-05,
+      "loss": 0.8555,
+      "step": 24138
+    },
+    {
+      "epoch": 0.5745739312577359,
+      "grad_norm": 4.921290031197621,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 24139
+    },
+    {
+      "epoch": 0.5745977339807674,
+      "grad_norm": 2.0059949227741223,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 24140
+    },
+    {
+      "epoch": 0.5746215367037989,
+      "grad_norm": 6.534564754255715,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 24141
+    },
+    {
+      "epoch": 0.5746453394268304,
+      "grad_norm": 3.009958462463181,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 24142
+    },
+    {
+      "epoch": 0.574669142149862,
+      "grad_norm": 0.7322639194518817,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 24143
+    },
+    {
+      "epoch": 0.5746929448728935,
+      "grad_norm": 3.9170326970477847,
+      "learning_rate": 2e-05,
+      "loss": 0.6852,
+      "step": 24144
+    },
+    {
+      "epoch": 0.574716747595925,
+      "grad_norm": 7.246558471667951,
+      "learning_rate": 2e-05,
+      "loss": 0.5984,
+      "step": 24145
+    },
+    {
+      "epoch": 0.5747405503189564,
+      "grad_norm": 6.458918958232033,
+      "learning_rate": 2e-05,
+      "loss": 1.6097,
+      "step": 24146
+    },
+    {
+      "epoch": 0.574764353041988,
+      "grad_norm": 3.934815700638894,
+      "learning_rate": 2e-05,
+      "loss": 0.7796,
+      "step": 24147
+    },
+    {
+      "epoch": 0.5747881557650195,
+      "grad_norm": 4.951015961958984,
+      "learning_rate": 2e-05,
+      "loss": 0.7478,
+      "step": 24148
+    },
+    {
+      "epoch": 0.574811958488051,
+      "grad_norm": 7.518710812737936,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 24149
+    },
+    {
+      "epoch": 0.5748357612110826,
+      "grad_norm": 2.3930712802497345,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 24150
+    },
+    {
+      "epoch": 0.5748595639341141,
+      "grad_norm": 2.604582692921063,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 24151
+    },
+    {
+      "epoch": 0.5748833666571456,
+      "grad_norm": 5.620791736167438,
+      "learning_rate": 2e-05,
+      "loss": 0.7517,
+      "step": 24152
+    },
+    {
+      "epoch": 0.574907169380177,
+      "grad_norm": 10.225097206088495,
+      "learning_rate": 2e-05,
+      "loss": 0.9849,
+      "step": 24153
+    },
+    {
+      "epoch": 0.5749309721032086,
+      "grad_norm": 11.140441160623938,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 24154
+    },
+    {
+      "epoch": 0.5749547748262401,
+      "grad_norm": 6.519674266767532,
+      "learning_rate": 2e-05,
+      "loss": 1.2291,
+      "step": 24155
+    },
+    {
+      "epoch": 0.5749785775492716,
+      "grad_norm": 5.9795771915826625,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 24156
+    },
+    {
+      "epoch": 0.5750023802723031,
+      "grad_norm": 3.6833791681393264,
+      "learning_rate": 2e-05,
+      "loss": 0.9492,
+      "step": 24157
+    },
+    {
+      "epoch": 0.5750261829953347,
+      "grad_norm": 3.5570823570803722,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 24158
+    },
+    {
+      "epoch": 0.5750499857183662,
+      "grad_norm": 3.7454268045519887,
+      "learning_rate": 2e-05,
+      "loss": 0.5011,
+      "step": 24159
+    },
+    {
+      "epoch": 0.5750737884413977,
+      "grad_norm": 9.191172799286214,
+      "learning_rate": 2e-05,
+      "loss": 0.2064,
+      "step": 24160
+    },
+    {
+      "epoch": 0.5750975911644292,
+      "grad_norm": 5.22218417092784,
+      "learning_rate": 2e-05,
+      "loss": 0.8451,
+      "step": 24161
+    },
+    {
+      "epoch": 0.5751213938874608,
+      "grad_norm": 2.1143758143100833,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 24162
+    },
+    {
+      "epoch": 0.5751451966104922,
+      "grad_norm": 3.676349241815845,
+      "learning_rate": 2e-05,
+      "loss": 0.4532,
+      "step": 24163
+    },
+    {
+      "epoch": 0.5751689993335237,
+      "grad_norm": 2.9394536127124917,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 24164
+    },
+    {
+      "epoch": 0.5751928020565553,
+      "grad_norm": 8.028559579174981,
+      "learning_rate": 2e-05,
+      "loss": 0.2072,
+      "step": 24165
+    },
+    {
+      "epoch": 0.5752166047795868,
+      "grad_norm": 1.9765276141640211,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 24166
+    },
+    {
+      "epoch": 0.5752404075026183,
+      "grad_norm": 3.875272468585239,
+      "learning_rate": 2e-05,
+      "loss": 0.296,
+      "step": 24167
+    },
+    {
+      "epoch": 0.5752642102256498,
+      "grad_norm": 4.449133572681322,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 24168
+    },
+    {
+      "epoch": 0.5752880129486814,
+      "grad_norm": 2.7273731838008564,
+      "learning_rate": 2e-05,
+      "loss": 0.2121,
+      "step": 24169
+    },
+    {
+      "epoch": 0.5753118156717129,
+      "grad_norm": 6.354366501280842,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 24170
+    },
+    {
+      "epoch": 0.5753356183947443,
+      "grad_norm": 8.82202492557301,
+      "learning_rate": 2e-05,
+      "loss": 1.1127,
+      "step": 24171
+    },
+    {
+      "epoch": 0.5753594211177758,
+      "grad_norm": 0.48741259445804525,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 24172
+    },
+    {
+      "epoch": 0.5753832238408074,
+      "grad_norm": 3.886028267709417,
+      "learning_rate": 2e-05,
+      "loss": 0.2491,
+      "step": 24173
+    },
+    {
+      "epoch": 0.5754070265638389,
+      "grad_norm": 3.8069278054115974,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 24174
+    },
+    {
+      "epoch": 0.5754308292868704,
+      "grad_norm": 6.910736941030205,
+      "learning_rate": 2e-05,
+      "loss": 0.885,
+      "step": 24175
+    },
+    {
+      "epoch": 0.5754546320099019,
+      "grad_norm": 6.140331644467802,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 24176
+    },
+    {
+      "epoch": 0.5754784347329335,
+      "grad_norm": 7.730658708823468,
+      "learning_rate": 2e-05,
+      "loss": 0.6901,
+      "step": 24177
+    },
+    {
+      "epoch": 0.575502237455965,
+      "grad_norm": 9.335567975380892,
+      "learning_rate": 2e-05,
+      "loss": 1.48,
+      "step": 24178
+    },
+    {
+      "epoch": 0.5755260401789964,
+      "grad_norm": 2.5905033072665495,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 24179
+    },
+    {
+      "epoch": 0.575549842902028,
+      "grad_norm": 2.071218324151261,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 24180
+    },
+    {
+      "epoch": 0.5755736456250595,
+      "grad_norm": 0.09740803425190295,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 24181
+    },
+    {
+      "epoch": 0.575597448348091,
+      "grad_norm": 0.29092791843472554,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 24182
+    },
+    {
+      "epoch": 0.5756212510711225,
+      "grad_norm": 5.933065349119793,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 24183
+    },
+    {
+      "epoch": 0.5756450537941541,
+      "grad_norm": 3.4361743069637716,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 24184
+    },
+    {
+      "epoch": 0.5756688565171856,
+      "grad_norm": 3.8524892050664272,
+      "learning_rate": 2e-05,
+      "loss": 0.8128,
+      "step": 24185
+    },
+    {
+      "epoch": 0.5756926592402171,
+      "grad_norm": 3.9403376035651347,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 24186
+    },
+    {
+      "epoch": 0.5757164619632485,
+      "grad_norm": 0.8083014354557704,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 24187
+    },
+    {
+      "epoch": 0.5757402646862801,
+      "grad_norm": 5.492358863884067,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 24188
+    },
+    {
+      "epoch": 0.5757640674093116,
+      "grad_norm": 4.451655949922705,
+      "learning_rate": 2e-05,
+      "loss": 0.8317,
+      "step": 24189
+    },
+    {
+      "epoch": 0.5757878701323431,
+      "grad_norm": 0.9601821534988126,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 24190
+    },
+    {
+      "epoch": 0.5758116728553746,
+      "grad_norm": 4.090543820676832,
+      "learning_rate": 2e-05,
+      "loss": 0.5891,
+      "step": 24191
+    },
+    {
+      "epoch": 0.5758354755784062,
+      "grad_norm": 2.3434138642281863,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 24192
+    },
+    {
+      "epoch": 0.5758592783014377,
+      "grad_norm": 3.8123031767337374,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 24193
+    },
+    {
+      "epoch": 0.5758830810244692,
+      "grad_norm": 4.236796156871937,
+      "learning_rate": 2e-05,
+      "loss": 0.6565,
+      "step": 24194
+    },
+    {
+      "epoch": 0.5759068837475008,
+      "grad_norm": 0.5885965546202708,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 24195
+    },
+    {
+      "epoch": 0.5759306864705322,
+      "grad_norm": 4.918827439927988,
+      "learning_rate": 2e-05,
+      "loss": 0.763,
+      "step": 24196
+    },
+    {
+      "epoch": 0.5759544891935637,
+      "grad_norm": 2.9527608107325505,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 24197
+    },
+    {
+      "epoch": 0.5759782919165952,
+      "grad_norm": 1.6027390780171051,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 24198
+    },
+    {
+      "epoch": 0.5760020946396268,
+      "grad_norm": 5.051906661416418,
+      "learning_rate": 2e-05,
+      "loss": 0.2884,
+      "step": 24199
+    },
+    {
+      "epoch": 0.5760258973626583,
+      "grad_norm": 2.712840910253578,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 24200
+    },
+    {
+      "epoch": 0.5760497000856898,
+      "grad_norm": 0.07171431018306303,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 24201
+    },
+    {
+      "epoch": 0.5760735028087213,
+      "grad_norm": 1.5116380503482831,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 24202
+    },
+    {
+      "epoch": 0.5760973055317529,
+      "grad_norm": 3.2579490030515514,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 24203
+    },
+    {
+      "epoch": 0.5761211082547844,
+      "grad_norm": 3.1347946760864076,
+      "learning_rate": 2e-05,
+      "loss": 0.059,
+      "step": 24204
+    },
+    {
+      "epoch": 0.5761449109778158,
+      "grad_norm": 2.3984526809564177,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 24205
+    },
+    {
+      "epoch": 0.5761687137008473,
+      "grad_norm": 7.376353110682292,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 24206
+    },
+    {
+      "epoch": 0.5761925164238789,
+      "grad_norm": 11.922407550677464,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 24207
+    },
+    {
+      "epoch": 0.5762163191469104,
+      "grad_norm": 3.391333294262488,
+      "learning_rate": 2e-05,
+      "loss": 0.6049,
+      "step": 24208
+    },
+    {
+      "epoch": 0.5762401218699419,
+      "grad_norm": 2.5230701336878987,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 24209
+    },
+    {
+      "epoch": 0.5762639245929735,
+      "grad_norm": 1.5637180769088606,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 24210
+    },
+    {
+      "epoch": 0.576287727316005,
+      "grad_norm": 0.04535959863851917,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 24211
+    },
+    {
+      "epoch": 0.5763115300390365,
+      "grad_norm": 9.617039406544523,
+      "learning_rate": 2e-05,
+      "loss": 0.3133,
+      "step": 24212
+    },
+    {
+      "epoch": 0.5763353327620679,
+      "grad_norm": 10.892178282964649,
+      "learning_rate": 2e-05,
+      "loss": 0.4943,
+      "step": 24213
+    },
+    {
+      "epoch": 0.5763591354850995,
+      "grad_norm": 6.067774800125769,
+      "learning_rate": 2e-05,
+      "loss": 0.7087,
+      "step": 24214
+    },
+    {
+      "epoch": 0.576382938208131,
+      "grad_norm": 0.44171867550987315,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 24215
+    },
+    {
+      "epoch": 0.5764067409311625,
+      "grad_norm": 3.1861450927016275,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 24216
+    },
+    {
+      "epoch": 0.576430543654194,
+      "grad_norm": 1.867216891307957,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 24217
+    },
+    {
+      "epoch": 0.5764543463772256,
+      "grad_norm": 6.29143601439922,
+      "learning_rate": 2e-05,
+      "loss": 0.8507,
+      "step": 24218
+    },
+    {
+      "epoch": 0.5764781491002571,
+      "grad_norm": 4.599211837442465,
+      "learning_rate": 2e-05,
+      "loss": 1.1912,
+      "step": 24219
+    },
+    {
+      "epoch": 0.5765019518232886,
+      "grad_norm": 13.891922990771763,
+      "learning_rate": 2e-05,
+      "loss": 0.3162,
+      "step": 24220
+    },
+    {
+      "epoch": 0.57652575454632,
+      "grad_norm": 4.946502540931124,
+      "learning_rate": 2e-05,
+      "loss": 0.4747,
+      "step": 24221
+    },
+    {
+      "epoch": 0.5765495572693516,
+      "grad_norm": 3.6876843121388627,
+      "learning_rate": 2e-05,
+      "loss": 0.7558,
+      "step": 24222
+    },
+    {
+      "epoch": 0.5765733599923831,
+      "grad_norm": 5.896052157601052,
+      "learning_rate": 2e-05,
+      "loss": 0.7022,
+      "step": 24223
+    },
+    {
+      "epoch": 0.5765971627154146,
+      "grad_norm": 1.975471192211508,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 24224
+    },
+    {
+      "epoch": 0.5766209654384462,
+      "grad_norm": 4.377921857643979,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 24225
+    },
+    {
+      "epoch": 0.5766447681614777,
+      "grad_norm": 1.9978198896980002,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 24226
+    },
+    {
+      "epoch": 0.5766685708845092,
+      "grad_norm": 0.19982672037367355,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 24227
+    },
+    {
+      "epoch": 0.5766923736075407,
+      "grad_norm": 9.261707250978619,
+      "learning_rate": 2e-05,
+      "loss": 0.9872,
+      "step": 24228
+    },
+    {
+      "epoch": 0.5767161763305723,
+      "grad_norm": 4.697127590323912,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 24229
+    },
+    {
+      "epoch": 0.5767399790536037,
+      "grad_norm": 7.269178529621125,
+      "learning_rate": 2e-05,
+      "loss": 0.143,
+      "step": 24230
+    },
+    {
+      "epoch": 0.5767637817766352,
+      "grad_norm": 0.4819720447079146,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 24231
+    },
+    {
+      "epoch": 0.5767875844996667,
+      "grad_norm": 3.071509866687543,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 24232
+    },
+    {
+      "epoch": 0.5768113872226983,
+      "grad_norm": 4.018012760830349,
+      "learning_rate": 2e-05,
+      "loss": 0.5955,
+      "step": 24233
+    },
+    {
+      "epoch": 0.5768351899457298,
+      "grad_norm": 9.795684902060488,
+      "learning_rate": 2e-05,
+      "loss": 0.279,
+      "step": 24234
+    },
+    {
+      "epoch": 0.5768589926687613,
+      "grad_norm": 8.054161630831251,
+      "learning_rate": 2e-05,
+      "loss": 0.6878,
+      "step": 24235
+    },
+    {
+      "epoch": 0.5768827953917928,
+      "grad_norm": 5.351293142667724,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 24236
+    },
+    {
+      "epoch": 0.5769065981148244,
+      "grad_norm": 7.8675300323816035,
+      "learning_rate": 2e-05,
+      "loss": 1.2182,
+      "step": 24237
+    },
+    {
+      "epoch": 0.5769304008378559,
+      "grad_norm": 8.08248509656359,
+      "learning_rate": 2e-05,
+      "loss": 1.234,
+      "step": 24238
+    },
+    {
+      "epoch": 0.5769542035608873,
+      "grad_norm": 8.340723249591727,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 24239
+    },
+    {
+      "epoch": 0.5769780062839189,
+      "grad_norm": 3.0429226484595953,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 24240
+    },
+    {
+      "epoch": 0.5770018090069504,
+      "grad_norm": 4.613229471704415,
+      "learning_rate": 2e-05,
+      "loss": 0.6196,
+      "step": 24241
+    },
+    {
+      "epoch": 0.5770256117299819,
+      "grad_norm": 3.0036341616908295,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 24242
+    },
+    {
+      "epoch": 0.5770494144530134,
+      "grad_norm": 12.49410442178385,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 24243
+    },
+    {
+      "epoch": 0.577073217176045,
+      "grad_norm": 0.24166530755782714,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 24244
+    },
+    {
+      "epoch": 0.5770970198990765,
+      "grad_norm": 0.16870566035493534,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 24245
+    },
+    {
+      "epoch": 0.577120822622108,
+      "grad_norm": 4.182617569281896,
+      "learning_rate": 2e-05,
+      "loss": 0.4262,
+      "step": 24246
+    },
+    {
+      "epoch": 0.5771446253451394,
+      "grad_norm": 0.5805302586730227,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 24247
+    },
+    {
+      "epoch": 0.577168428068171,
+      "grad_norm": 2.451682579360751,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 24248
+    },
+    {
+      "epoch": 0.5771922307912025,
+      "grad_norm": 2.456121984264212,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 24249
+    },
+    {
+      "epoch": 0.577216033514234,
+      "grad_norm": 17.96454877092327,
+      "learning_rate": 2e-05,
+      "loss": 0.8076,
+      "step": 24250
+    },
+    {
+      "epoch": 0.5772398362372655,
+      "grad_norm": 5.336984994118518,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 24251
+    },
+    {
+      "epoch": 0.5772636389602971,
+      "grad_norm": 0.9778618605079162,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 24252
+    },
+    {
+      "epoch": 0.5772874416833286,
+      "grad_norm": 3.8485549417567717,
+      "learning_rate": 2e-05,
+      "loss": 0.5763,
+      "step": 24253
+    },
+    {
+      "epoch": 0.5773112444063601,
+      "grad_norm": 6.53365105681013,
+      "learning_rate": 2e-05,
+      "loss": 1.3524,
+      "step": 24254
+    },
+    {
+      "epoch": 0.5773350471293917,
+      "grad_norm": 2.1439101573707755,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 24255
+    },
+    {
+      "epoch": 0.5773588498524231,
+      "grad_norm": 7.694698137860993,
+      "learning_rate": 2e-05,
+      "loss": 0.9846,
+      "step": 24256
+    },
+    {
+      "epoch": 0.5773826525754546,
+      "grad_norm": 2.4368728429264657,
+      "learning_rate": 2e-05,
+      "loss": 0.2022,
+      "step": 24257
+    },
+    {
+      "epoch": 0.5774064552984861,
+      "grad_norm": 4.032326780444892,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 24258
+    },
+    {
+      "epoch": 0.5774302580215177,
+      "grad_norm": 7.3068492307842785,
+      "learning_rate": 2e-05,
+      "loss": 0.3668,
+      "step": 24259
+    },
+    {
+      "epoch": 0.5774540607445492,
+      "grad_norm": 4.387535904695798,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 24260
+    },
+    {
+      "epoch": 0.5774778634675807,
+      "grad_norm": 2.6354718573542795,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 24261
+    },
+    {
+      "epoch": 0.5775016661906122,
+      "grad_norm": 0.9375652092836313,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 24262
+    },
+    {
+      "epoch": 0.5775254689136438,
+      "grad_norm": 4.594735894510954,
+      "learning_rate": 2e-05,
+      "loss": 0.5571,
+      "step": 24263
+    },
+    {
+      "epoch": 0.5775492716366752,
+      "grad_norm": 8.760454782203215,
+      "learning_rate": 2e-05,
+      "loss": 1.0637,
+      "step": 24264
+    },
+    {
+      "epoch": 0.5775730743597067,
+      "grad_norm": 2.673703328327233,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 24265
+    },
+    {
+      "epoch": 0.5775968770827382,
+      "grad_norm": 3.275884576305105,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 24266
+    },
+    {
+      "epoch": 0.5776206798057698,
+      "grad_norm": 6.835004432398868,
+      "learning_rate": 2e-05,
+      "loss": 0.7421,
+      "step": 24267
+    },
+    {
+      "epoch": 0.5776444825288013,
+      "grad_norm": 3.7211084204469214,
+      "learning_rate": 2e-05,
+      "loss": 0.5358,
+      "step": 24268
+    },
+    {
+      "epoch": 0.5776682852518328,
+      "grad_norm": 4.4462628278386624,
+      "learning_rate": 2e-05,
+      "loss": 0.6061,
+      "step": 24269
+    },
+    {
+      "epoch": 0.5776920879748644,
+      "grad_norm": 5.055379323013722,
+      "learning_rate": 2e-05,
+      "loss": 0.4514,
+      "step": 24270
+    },
+    {
+      "epoch": 0.5777158906978959,
+      "grad_norm": 3.6623577152081936,
+      "learning_rate": 2e-05,
+      "loss": 0.5162,
+      "step": 24271
+    },
+    {
+      "epoch": 0.5777396934209273,
+      "grad_norm": 2.8840504304455163,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 24272
+    },
+    {
+      "epoch": 0.5777634961439588,
+      "grad_norm": 0.5344664272335862,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24273
+    },
+    {
+      "epoch": 0.5777872988669904,
+      "grad_norm": 8.069368065583191,
+      "learning_rate": 2e-05,
+      "loss": 1.5372,
+      "step": 24274
+    },
+    {
+      "epoch": 0.5778111015900219,
+      "grad_norm": 6.34941722253266,
+      "learning_rate": 2e-05,
+      "loss": 0.792,
+      "step": 24275
+    },
+    {
+      "epoch": 0.5778349043130534,
+      "grad_norm": 3.6557755815821924,
+      "learning_rate": 2e-05,
+      "loss": 0.6418,
+      "step": 24276
+    },
+    {
+      "epoch": 0.5778587070360849,
+      "grad_norm": 2.8471468758206866,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 24277
+    },
+    {
+      "epoch": 0.5778825097591165,
+      "grad_norm": 11.493705827886242,
+      "learning_rate": 2e-05,
+      "loss": 0.3805,
+      "step": 24278
+    },
+    {
+      "epoch": 0.577906312482148,
+      "grad_norm": 3.1795792953955013,
+      "learning_rate": 2e-05,
+      "loss": 0.5182,
+      "step": 24279
+    },
+    {
+      "epoch": 0.5779301152051795,
+      "grad_norm": 5.0403256570087525,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 24280
+    },
+    {
+      "epoch": 0.5779539179282109,
+      "grad_norm": 0.12469435514075008,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 24281
+    },
+    {
+      "epoch": 0.5779777206512425,
+      "grad_norm": 1.9801279946039103,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 24282
+    },
+    {
+      "epoch": 0.578001523374274,
+      "grad_norm": 3.899250352034136,
+      "learning_rate": 2e-05,
+      "loss": 0.7469,
+      "step": 24283
+    },
+    {
+      "epoch": 0.5780253260973055,
+      "grad_norm": 0.6495341003965491,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 24284
+    },
+    {
+      "epoch": 0.5780491288203371,
+      "grad_norm": 1.8794874348299162,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 24285
+    },
+    {
+      "epoch": 0.5780729315433686,
+      "grad_norm": 5.336443824275639,
+      "learning_rate": 2e-05,
+      "loss": 0.4673,
+      "step": 24286
+    },
+    {
+      "epoch": 0.5780967342664001,
+      "grad_norm": 10.730157554303023,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 24287
+    },
+    {
+      "epoch": 0.5781205369894316,
+      "grad_norm": 1.0483344058682391,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 24288
+    },
+    {
+      "epoch": 0.5781443397124632,
+      "grad_norm": 1.8881444511287846,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 24289
+    },
+    {
+      "epoch": 0.5781681424354946,
+      "grad_norm": 0.2980376712719835,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 24290
+    },
+    {
+      "epoch": 0.5781919451585261,
+      "grad_norm": 6.680589788360175,
+      "learning_rate": 2e-05,
+      "loss": 0.1959,
+      "step": 24291
+    },
+    {
+      "epoch": 0.5782157478815576,
+      "grad_norm": 2.921936839543204,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 24292
+    },
+    {
+      "epoch": 0.5782395506045892,
+      "grad_norm": 2.4801583198608457,
+      "learning_rate": 2e-05,
+      "loss": 0.2412,
+      "step": 24293
+    },
+    {
+      "epoch": 0.5782633533276207,
+      "grad_norm": 2.751125198842747,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 24294
+    },
+    {
+      "epoch": 0.5782871560506522,
+      "grad_norm": 1.1150684906163804,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 24295
+    },
+    {
+      "epoch": 0.5783109587736837,
+      "grad_norm": 3.6139544686048084,
+      "learning_rate": 2e-05,
+      "loss": 0.4796,
+      "step": 24296
+    },
+    {
+      "epoch": 0.5783347614967153,
+      "grad_norm": 0.29458451892805937,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 24297
+    },
+    {
+      "epoch": 0.5783585642197467,
+      "grad_norm": 5.965855627107345,
+      "learning_rate": 2e-05,
+      "loss": 0.9702,
+      "step": 24298
+    },
+    {
+      "epoch": 0.5783823669427782,
+      "grad_norm": 0.5397262694638246,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 24299
+    },
+    {
+      "epoch": 0.5784061696658098,
+      "grad_norm": 4.661752704802577,
+      "learning_rate": 2e-05,
+      "loss": 0.7537,
+      "step": 24300
+    },
+    {
+      "epoch": 0.5784299723888413,
+      "grad_norm": 0.1128617656405075,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 24301
+    },
+    {
+      "epoch": 0.5784537751118728,
+      "grad_norm": 3.443823903196845,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 24302
+    },
+    {
+      "epoch": 0.5784775778349043,
+      "grad_norm": 4.4187525712777855,
+      "learning_rate": 2e-05,
+      "loss": 0.8078,
+      "step": 24303
+    },
+    {
+      "epoch": 0.5785013805579359,
+      "grad_norm": 4.52229138110038,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 24304
+    },
+    {
+      "epoch": 0.5785251832809674,
+      "grad_norm": 4.58289612897369,
+      "learning_rate": 2e-05,
+      "loss": 0.737,
+      "step": 24305
+    },
+    {
+      "epoch": 0.5785489860039988,
+      "grad_norm": 0.3505236475604572,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 24306
+    },
+    {
+      "epoch": 0.5785727887270303,
+      "grad_norm": 0.10739370100504804,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 24307
+    },
+    {
+      "epoch": 0.5785965914500619,
+      "grad_norm": 0.2611029268656813,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 24308
+    },
+    {
+      "epoch": 0.5786203941730934,
+      "grad_norm": 0.4191956871152642,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 24309
+    },
+    {
+      "epoch": 0.5786441968961249,
+      "grad_norm": 0.13227000538987782,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 24310
+    },
+    {
+      "epoch": 0.5786679996191564,
+      "grad_norm": 4.243012094567168,
+      "learning_rate": 2e-05,
+      "loss": 0.5639,
+      "step": 24311
+    },
+    {
+      "epoch": 0.578691802342188,
+      "grad_norm": 1.6231285092465615,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 24312
+    },
+    {
+      "epoch": 0.5787156050652195,
+      "grad_norm": 0.08825291042913115,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 24313
+    },
+    {
+      "epoch": 0.578739407788251,
+      "grad_norm": 6.559168639015097,
+      "learning_rate": 2e-05,
+      "loss": 1.1561,
+      "step": 24314
+    },
+    {
+      "epoch": 0.5787632105112825,
+      "grad_norm": 0.08868018126315545,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 24315
+    },
+    {
+      "epoch": 0.578787013234314,
+      "grad_norm": 5.4545538471634565,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 24316
+    },
+    {
+      "epoch": 0.5788108159573455,
+      "grad_norm": 2.9729767534596117,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 24317
+    },
+    {
+      "epoch": 0.578834618680377,
+      "grad_norm": 2.670136873686655,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 24318
+    },
+    {
+      "epoch": 0.5788584214034086,
+      "grad_norm": 1.404144985585169,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 24319
+    },
+    {
+      "epoch": 0.5788822241264401,
+      "grad_norm": 1.3575807917646223,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 24320
+    },
+    {
+      "epoch": 0.5789060268494716,
+      "grad_norm": 16.37835645102066,
+      "learning_rate": 2e-05,
+      "loss": 0.7198,
+      "step": 24321
+    },
+    {
+      "epoch": 0.578929829572503,
+      "grad_norm": 3.876476996045419,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 24322
+    },
+    {
+      "epoch": 0.5789536322955346,
+      "grad_norm": 4.163538456847771,
+      "learning_rate": 2e-05,
+      "loss": 0.4687,
+      "step": 24323
+    },
+    {
+      "epoch": 0.5789774350185661,
+      "grad_norm": 4.014554376988193,
+      "learning_rate": 2e-05,
+      "loss": 0.5097,
+      "step": 24324
+    },
+    {
+      "epoch": 0.5790012377415976,
+      "grad_norm": 3.130625466618195,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 24325
+    },
+    {
+      "epoch": 0.5790250404646291,
+      "grad_norm": 6.10326666698118,
+      "learning_rate": 2e-05,
+      "loss": 1.0769,
+      "step": 24326
+    },
+    {
+      "epoch": 0.5790488431876607,
+      "grad_norm": 13.485397680926708,
+      "learning_rate": 2e-05,
+      "loss": 0.2561,
+      "step": 24327
+    },
+    {
+      "epoch": 0.5790726459106922,
+      "grad_norm": 0.9503818457853076,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 24328
+    },
+    {
+      "epoch": 0.5790964486337237,
+      "grad_norm": 14.97893165753902,
+      "learning_rate": 2e-05,
+      "loss": 0.4845,
+      "step": 24329
+    },
+    {
+      "epoch": 0.5791202513567553,
+      "grad_norm": 3.7568977881795855,
+      "learning_rate": 2e-05,
+      "loss": 0.5881,
+      "step": 24330
+    },
+    {
+      "epoch": 0.5791440540797868,
+      "grad_norm": 5.86900890135291,
+      "learning_rate": 2e-05,
+      "loss": 0.7362,
+      "step": 24331
+    },
+    {
+      "epoch": 0.5791678568028182,
+      "grad_norm": 3.344589437029491,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 24332
+    },
+    {
+      "epoch": 0.5791916595258497,
+      "grad_norm": 6.3124833084934595,
+      "learning_rate": 2e-05,
+      "loss": 1.2954,
+      "step": 24333
+    },
+    {
+      "epoch": 0.5792154622488813,
+      "grad_norm": 2.349610764255483,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 24334
+    },
+    {
+      "epoch": 0.5792392649719128,
+      "grad_norm": 10.904708355635563,
+      "learning_rate": 2e-05,
+      "loss": 0.292,
+      "step": 24335
+    },
+    {
+      "epoch": 0.5792630676949443,
+      "grad_norm": 2.3740106705624684,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 24336
+    },
+    {
+      "epoch": 0.5792868704179758,
+      "grad_norm": 4.659024832976563,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 24337
+    },
+    {
+      "epoch": 0.5793106731410074,
+      "grad_norm": 3.2032365878220523,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 24338
+    },
+    {
+      "epoch": 0.5793344758640389,
+      "grad_norm": 6.3558881270376055,
+      "learning_rate": 2e-05,
+      "loss": 0.3835,
+      "step": 24339
+    },
+    {
+      "epoch": 0.5793582785870703,
+      "grad_norm": 3.8717040149263355,
+      "learning_rate": 2e-05,
+      "loss": 0.7775,
+      "step": 24340
+    },
+    {
+      "epoch": 0.5793820813101018,
+      "grad_norm": 9.435734175597895,
+      "learning_rate": 2e-05,
+      "loss": 0.3283,
+      "step": 24341
+    },
+    {
+      "epoch": 0.5794058840331334,
+      "grad_norm": 17.219085641080223,
+      "learning_rate": 2e-05,
+      "loss": 0.9313,
+      "step": 24342
+    },
+    {
+      "epoch": 0.5794296867561649,
+      "grad_norm": 3.936680686942663,
+      "learning_rate": 2e-05,
+      "loss": 0.4332,
+      "step": 24343
+    },
+    {
+      "epoch": 0.5794534894791964,
+      "grad_norm": 0.26005720047187236,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 24344
+    },
+    {
+      "epoch": 0.579477292202228,
+      "grad_norm": 6.690796121779738,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 24345
+    },
+    {
+      "epoch": 0.5795010949252595,
+      "grad_norm": 2.0909528412324434,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 24346
+    },
+    {
+      "epoch": 0.579524897648291,
+      "grad_norm": 6.260200983227926,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 24347
+    },
+    {
+      "epoch": 0.5795487003713224,
+      "grad_norm": 2.705399128874584,
+      "learning_rate": 2e-05,
+      "loss": 0.4305,
+      "step": 24348
+    },
+    {
+      "epoch": 0.579572503094354,
+      "grad_norm": 0.23727955892950822,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 24349
+    },
+    {
+      "epoch": 0.5795963058173855,
+      "grad_norm": 11.839866072404902,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 24350
+    },
+    {
+      "epoch": 0.579620108540417,
+      "grad_norm": 5.769470790017027,
+      "learning_rate": 2e-05,
+      "loss": 0.9084,
+      "step": 24351
+    },
+    {
+      "epoch": 0.5796439112634485,
+      "grad_norm": 8.318765690371759,
+      "learning_rate": 2e-05,
+      "loss": 1.3144,
+      "step": 24352
+    },
+    {
+      "epoch": 0.5796677139864801,
+      "grad_norm": 0.4628592779649262,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 24353
+    },
+    {
+      "epoch": 0.5796915167095116,
+      "grad_norm": 5.90682907153533,
+      "learning_rate": 2e-05,
+      "loss": 1.2878,
+      "step": 24354
+    },
+    {
+      "epoch": 0.5797153194325431,
+      "grad_norm": 25.571451512894686,
+      "learning_rate": 2e-05,
+      "loss": 0.7648,
+      "step": 24355
+    },
+    {
+      "epoch": 0.5797391221555745,
+      "grad_norm": 6.439375644154172,
+      "learning_rate": 2e-05,
+      "loss": 2.0775,
+      "step": 24356
+    },
+    {
+      "epoch": 0.5797629248786061,
+      "grad_norm": 80.82935827507914,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 24357
+    },
+    {
+      "epoch": 0.5797867276016376,
+      "grad_norm": 3.014179540539834,
+      "learning_rate": 2e-05,
+      "loss": 0.1999,
+      "step": 24358
+    },
+    {
+      "epoch": 0.5798105303246691,
+      "grad_norm": 1.8025425458455595,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 24359
+    },
+    {
+      "epoch": 0.5798343330477007,
+      "grad_norm": 6.43554128814147,
+      "learning_rate": 2e-05,
+      "loss": 0.2022,
+      "step": 24360
+    },
+    {
+      "epoch": 0.5798581357707322,
+      "grad_norm": 4.36722070099916,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 24361
+    },
+    {
+      "epoch": 0.5798819384937637,
+      "grad_norm": 6.393737510171421,
+      "learning_rate": 2e-05,
+      "loss": 0.1937,
+      "step": 24362
+    },
+    {
+      "epoch": 0.5799057412167952,
+      "grad_norm": 3.61579455683486,
+      "learning_rate": 2e-05,
+      "loss": 0.5512,
+      "step": 24363
+    },
+    {
+      "epoch": 0.5799295439398268,
+      "grad_norm": 1.8361784324984767,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 24364
+    },
+    {
+      "epoch": 0.5799533466628582,
+      "grad_norm": 4.642641448916021,
+      "learning_rate": 2e-05,
+      "loss": 0.9724,
+      "step": 24365
+    },
+    {
+      "epoch": 0.5799771493858897,
+      "grad_norm": 3.61956833534707,
+      "learning_rate": 2e-05,
+      "loss": 1.0655,
+      "step": 24366
+    },
+    {
+      "epoch": 0.5800009521089212,
+      "grad_norm": 3.4573043100637464,
+      "learning_rate": 2e-05,
+      "loss": 0.4005,
+      "step": 24367
+    },
+    {
+      "epoch": 0.5800247548319528,
+      "grad_norm": 3.574880462725357,
+      "learning_rate": 2e-05,
+      "loss": 0.3915,
+      "step": 24368
+    },
+    {
+      "epoch": 0.5800485575549843,
+      "grad_norm": 1.9816170316355044,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 24369
+    },
+    {
+      "epoch": 0.5800723602780158,
+      "grad_norm": 0.07246176963893605,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 24370
+    },
+    {
+      "epoch": 0.5800961630010473,
+      "grad_norm": 1.5046216275866042,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 24371
+    },
+    {
+      "epoch": 0.5801199657240789,
+      "grad_norm": 1.9635626129160202,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 24372
+    },
+    {
+      "epoch": 0.5801437684471104,
+      "grad_norm": 6.31583062607428,
+      "learning_rate": 2e-05,
+      "loss": 0.7931,
+      "step": 24373
+    },
+    {
+      "epoch": 0.5801675711701418,
+      "grad_norm": 4.508347341423188,
+      "learning_rate": 2e-05,
+      "loss": 0.4802,
+      "step": 24374
+    },
+    {
+      "epoch": 0.5801913738931734,
+      "grad_norm": 4.65771827598482,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 24375
+    },
+    {
+      "epoch": 0.5802151766162049,
+      "grad_norm": 26.257615883404668,
+      "learning_rate": 2e-05,
+      "loss": 0.7604,
+      "step": 24376
+    },
+    {
+      "epoch": 0.5802389793392364,
+      "grad_norm": 7.101380776687558,
+      "learning_rate": 2e-05,
+      "loss": 0.8074,
+      "step": 24377
+    },
+    {
+      "epoch": 0.5802627820622679,
+      "grad_norm": 0.20659648028950472,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 24378
+    },
+    {
+      "epoch": 0.5802865847852995,
+      "grad_norm": 6.675281593281909,
+      "learning_rate": 2e-05,
+      "loss": 1.3474,
+      "step": 24379
+    },
+    {
+      "epoch": 0.580310387508331,
+      "grad_norm": 2.945154303098446,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 24380
+    },
+    {
+      "epoch": 0.5803341902313625,
+      "grad_norm": 4.57084288461793,
+      "learning_rate": 2e-05,
+      "loss": 0.2877,
+      "step": 24381
+    },
+    {
+      "epoch": 0.5803579929543939,
+      "grad_norm": 8.19771886584853,
+      "learning_rate": 2e-05,
+      "loss": 0.3978,
+      "step": 24382
+    },
+    {
+      "epoch": 0.5803817956774255,
+      "grad_norm": 7.6420881043545705,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 24383
+    },
+    {
+      "epoch": 0.580405598400457,
+      "grad_norm": 4.007566851323191,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 24384
+    },
+    {
+      "epoch": 0.5804294011234885,
+      "grad_norm": 0.20733011267703427,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 24385
+    },
+    {
+      "epoch": 0.58045320384652,
+      "grad_norm": 59.22099725314987,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 24386
+    },
+    {
+      "epoch": 0.5804770065695516,
+      "grad_norm": 4.170328435366736,
+      "learning_rate": 2e-05,
+      "loss": 0.2123,
+      "step": 24387
+    },
+    {
+      "epoch": 0.5805008092925831,
+      "grad_norm": 6.585952411993424,
+      "learning_rate": 2e-05,
+      "loss": 0.8066,
+      "step": 24388
+    },
+    {
+      "epoch": 0.5805246120156146,
+      "grad_norm": 11.353820333063524,
+      "learning_rate": 2e-05,
+      "loss": 0.8488,
+      "step": 24389
+    },
+    {
+      "epoch": 0.580548414738646,
+      "grad_norm": 4.202880673496109,
+      "learning_rate": 2e-05,
+      "loss": 0.5812,
+      "step": 24390
+    },
+    {
+      "epoch": 0.5805722174616776,
+      "grad_norm": 2.5411769427612363,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 24391
+    },
+    {
+      "epoch": 0.5805960201847091,
+      "grad_norm": 7.106173947400272,
+      "learning_rate": 2e-05,
+      "loss": 0.4022,
+      "step": 24392
+    },
+    {
+      "epoch": 0.5806198229077406,
+      "grad_norm": 10.308010922817518,
+      "learning_rate": 2e-05,
+      "loss": 0.6112,
+      "step": 24393
+    },
+    {
+      "epoch": 0.5806436256307722,
+      "grad_norm": 0.11760702701097935,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 24394
+    },
+    {
+      "epoch": 0.5806674283538037,
+      "grad_norm": 2.153969294581833,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 24395
+    },
+    {
+      "epoch": 0.5806912310768352,
+      "grad_norm": 5.311524222298807,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 24396
+    },
+    {
+      "epoch": 0.5807150337998667,
+      "grad_norm": 8.134606952356359,
+      "learning_rate": 2e-05,
+      "loss": 1.1431,
+      "step": 24397
+    },
+    {
+      "epoch": 0.5807388365228983,
+      "grad_norm": 7.840636163544527,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 24398
+    },
+    {
+      "epoch": 0.5807626392459297,
+      "grad_norm": 1.4164678999020361,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 24399
+    },
+    {
+      "epoch": 0.5807864419689612,
+      "grad_norm": 6.227632390289594,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 24400
+    },
+    {
+      "epoch": 0.5808102446919927,
+      "grad_norm": 3.131064299063699,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 24401
+    },
+    {
+      "epoch": 0.5808340474150243,
+      "grad_norm": 1.6165314174984604,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 24402
+    },
+    {
+      "epoch": 0.5808578501380558,
+      "grad_norm": 11.880200915796324,
+      "learning_rate": 2e-05,
+      "loss": 0.6654,
+      "step": 24403
+    },
+    {
+      "epoch": 0.5808816528610873,
+      "grad_norm": 2.4863022142979747,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 24404
+    },
+    {
+      "epoch": 0.5809054555841188,
+      "grad_norm": 1.670880793358151,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 24405
+    },
+    {
+      "epoch": 0.5809292583071504,
+      "grad_norm": 0.24910599175892015,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24406
+    },
+    {
+      "epoch": 0.5809530610301818,
+      "grad_norm": 3.360255825698005,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 24407
+    },
+    {
+      "epoch": 0.5809768637532133,
+      "grad_norm": 6.086020005273145,
+      "learning_rate": 2e-05,
+      "loss": 0.5293,
+      "step": 24408
+    },
+    {
+      "epoch": 0.5810006664762449,
+      "grad_norm": 0.1211183670708245,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 24409
+    },
+    {
+      "epoch": 0.5810244691992764,
+      "grad_norm": 7.945695116279144,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 24410
+    },
+    {
+      "epoch": 0.5810482719223079,
+      "grad_norm": 0.4350765852222099,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 24411
+    },
+    {
+      "epoch": 0.5810720746453394,
+      "grad_norm": 10.700051335079515,
+      "learning_rate": 2e-05,
+      "loss": 1.1732,
+      "step": 24412
+    },
+    {
+      "epoch": 0.581095877368371,
+      "grad_norm": 3.5987420068332816,
+      "learning_rate": 2e-05,
+      "loss": 0.1541,
+      "step": 24413
+    },
+    {
+      "epoch": 0.5811196800914025,
+      "grad_norm": 3.7583915740834906,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 24414
+    },
+    {
+      "epoch": 0.581143482814434,
+      "grad_norm": 1.6896285465316998,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 24415
+    },
+    {
+      "epoch": 0.5811672855374654,
+      "grad_norm": 18.897749987103023,
+      "learning_rate": 2e-05,
+      "loss": 0.8044,
+      "step": 24416
+    },
+    {
+      "epoch": 0.581191088260497,
+      "grad_norm": 8.62850610678058,
+      "learning_rate": 2e-05,
+      "loss": 0.3724,
+      "step": 24417
+    },
+    {
+      "epoch": 0.5812148909835285,
+      "grad_norm": 4.904951834233735,
+      "learning_rate": 2e-05,
+      "loss": 0.9665,
+      "step": 24418
+    },
+    {
+      "epoch": 0.58123869370656,
+      "grad_norm": 5.992967234894199,
+      "learning_rate": 2e-05,
+      "loss": 0.3417,
+      "step": 24419
+    },
+    {
+      "epoch": 0.5812624964295915,
+      "grad_norm": 3.401993558364331,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 24420
+    },
+    {
+      "epoch": 0.5812862991526231,
+      "grad_norm": 10.698254666413813,
+      "learning_rate": 2e-05,
+      "loss": 0.5679,
+      "step": 24421
+    },
+    {
+      "epoch": 0.5813101018756546,
+      "grad_norm": 1.3709776259484676,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 24422
+    },
+    {
+      "epoch": 0.5813339045986861,
+      "grad_norm": 19.2551077083272,
+      "learning_rate": 2e-05,
+      "loss": 1.1439,
+      "step": 24423
+    },
+    {
+      "epoch": 0.5813577073217177,
+      "grad_norm": 0.23428378576075673,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 24424
+    },
+    {
+      "epoch": 0.5813815100447491,
+      "grad_norm": 4.272522726858533,
+      "learning_rate": 2e-05,
+      "loss": 0.663,
+      "step": 24425
+    },
+    {
+      "epoch": 0.5814053127677806,
+      "grad_norm": 2.213673132043368,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 24426
+    },
+    {
+      "epoch": 0.5814291154908121,
+      "grad_norm": 7.011507908693762,
+      "learning_rate": 2e-05,
+      "loss": 1.4603,
+      "step": 24427
+    },
+    {
+      "epoch": 0.5814529182138437,
+      "grad_norm": 1.2815863444084084,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 24428
+    },
+    {
+      "epoch": 0.5814767209368752,
+      "grad_norm": 4.961868090090668,
+      "learning_rate": 2e-05,
+      "loss": 1.1035,
+      "step": 24429
+    },
+    {
+      "epoch": 0.5815005236599067,
+      "grad_norm": 3.6448620446902913,
+      "learning_rate": 2e-05,
+      "loss": 0.1812,
+      "step": 24430
+    },
+    {
+      "epoch": 0.5815243263829382,
+      "grad_norm": 15.631499641497122,
+      "learning_rate": 2e-05,
+      "loss": 0.8092,
+      "step": 24431
+    },
+    {
+      "epoch": 0.5815481291059698,
+      "grad_norm": 3.259615483039956,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 24432
+    },
+    {
+      "epoch": 0.5815719318290012,
+      "grad_norm": 0.3570537938974978,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 24433
+    },
+    {
+      "epoch": 0.5815957345520327,
+      "grad_norm": 6.297115088955463,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 24434
+    },
+    {
+      "epoch": 0.5816195372750642,
+      "grad_norm": 1.3671676353715823,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 24435
+    },
+    {
+      "epoch": 0.5816433399980958,
+      "grad_norm": 5.533886813336823,
+      "learning_rate": 2e-05,
+      "loss": 1.0799,
+      "step": 24436
+    },
+    {
+      "epoch": 0.5816671427211273,
+      "grad_norm": 5.8251995224911495,
+      "learning_rate": 2e-05,
+      "loss": 0.8125,
+      "step": 24437
+    },
+    {
+      "epoch": 0.5816909454441588,
+      "grad_norm": 1.7002437807435464,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 24438
+    },
+    {
+      "epoch": 0.5817147481671904,
+      "grad_norm": 0.1408545520432042,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 24439
+    },
+    {
+      "epoch": 0.5817385508902219,
+      "grad_norm": 2.232040826956706,
+      "learning_rate": 2e-05,
+      "loss": 0.1827,
+      "step": 24440
+    },
+    {
+      "epoch": 0.5817623536132533,
+      "grad_norm": 1.8045634099890062,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 24441
+    },
+    {
+      "epoch": 0.5817861563362848,
+      "grad_norm": 5.075840853094465,
+      "learning_rate": 2e-05,
+      "loss": 0.1157,
+      "step": 24442
+    },
+    {
+      "epoch": 0.5818099590593164,
+      "grad_norm": 5.619926590506345,
+      "learning_rate": 2e-05,
+      "loss": 0.9379,
+      "step": 24443
+    },
+    {
+      "epoch": 0.5818337617823479,
+      "grad_norm": 0.4680928001579108,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 24444
+    },
+    {
+      "epoch": 0.5818575645053794,
+      "grad_norm": 2.1534824656659763,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 24445
+    },
+    {
+      "epoch": 0.5818813672284109,
+      "grad_norm": 5.479851413433836,
+      "learning_rate": 2e-05,
+      "loss": 1.0999,
+      "step": 24446
+    },
+    {
+      "epoch": 0.5819051699514425,
+      "grad_norm": 14.386360083494907,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 24447
+    },
+    {
+      "epoch": 0.581928972674474,
+      "grad_norm": 1.5034706936558857,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 24448
+    },
+    {
+      "epoch": 0.5819527753975055,
+      "grad_norm": 0.4013592732884586,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 24449
+    },
+    {
+      "epoch": 0.5819765781205369,
+      "grad_norm": 2.0429323276341824,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 24450
+    },
+    {
+      "epoch": 0.5820003808435685,
+      "grad_norm": 7.677028155521396,
+      "learning_rate": 2e-05,
+      "loss": 1.0623,
+      "step": 24451
+    },
+    {
+      "epoch": 0.5820241835666,
+      "grad_norm": 0.19617773614884934,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24452
+    },
+    {
+      "epoch": 0.5820479862896315,
+      "grad_norm": 7.9475440952618905,
+      "learning_rate": 2e-05,
+      "loss": 1.3715,
+      "step": 24453
+    },
+    {
+      "epoch": 0.5820717890126631,
+      "grad_norm": 1.2405055385897934,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 24454
+    },
+    {
+      "epoch": 0.5820955917356946,
+      "grad_norm": 4.979504010711476,
+      "learning_rate": 2e-05,
+      "loss": 1.7968,
+      "step": 24455
+    },
+    {
+      "epoch": 0.5821193944587261,
+      "grad_norm": 5.465163936956726,
+      "learning_rate": 2e-05,
+      "loss": 0.7239,
+      "step": 24456
+    },
+    {
+      "epoch": 0.5821431971817576,
+      "grad_norm": 35.5267407551473,
+      "learning_rate": 2e-05,
+      "loss": 0.2191,
+      "step": 24457
+    },
+    {
+      "epoch": 0.5821669999047892,
+      "grad_norm": 2.842848049201232,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 24458
+    },
+    {
+      "epoch": 0.5821908026278206,
+      "grad_norm": 2.529726025423406,
+      "learning_rate": 2e-05,
+      "loss": 0.0696,
+      "step": 24459
+    },
+    {
+      "epoch": 0.5822146053508521,
+      "grad_norm": 2.336226455017303,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 24460
+    },
+    {
+      "epoch": 0.5822384080738836,
+      "grad_norm": 3.7254689761704465,
+      "learning_rate": 2e-05,
+      "loss": 0.6212,
+      "step": 24461
+    },
+    {
+      "epoch": 0.5822622107969152,
+      "grad_norm": 0.39406606098589375,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 24462
+    },
+    {
+      "epoch": 0.5822860135199467,
+      "grad_norm": 4.02429800829172,
+      "learning_rate": 2e-05,
+      "loss": 0.4228,
+      "step": 24463
+    },
+    {
+      "epoch": 0.5823098162429782,
+      "grad_norm": 1.4979615105281154,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 24464
+    },
+    {
+      "epoch": 0.5823336189660097,
+      "grad_norm": 14.57316932917539,
+      "learning_rate": 2e-05,
+      "loss": 0.723,
+      "step": 24465
+    },
+    {
+      "epoch": 0.5823574216890413,
+      "grad_norm": 0.6353383273979033,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 24466
+    },
+    {
+      "epoch": 0.5823812244120727,
+      "grad_norm": 9.124897208780759,
+      "learning_rate": 2e-05,
+      "loss": 0.7054,
+      "step": 24467
+    },
+    {
+      "epoch": 0.5824050271351042,
+      "grad_norm": 0.02437351518985048,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 24468
+    },
+    {
+      "epoch": 0.5824288298581358,
+      "grad_norm": 1.2798469798174827,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 24469
+    },
+    {
+      "epoch": 0.5824526325811673,
+      "grad_norm": 0.10599067243283422,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 24470
+    },
+    {
+      "epoch": 0.5824764353041988,
+      "grad_norm": 10.136372767050794,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 24471
+    },
+    {
+      "epoch": 0.5825002380272303,
+      "grad_norm": 1.8366684021471178,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 24472
+    },
+    {
+      "epoch": 0.5825240407502619,
+      "grad_norm": 5.6972207239428005,
+      "learning_rate": 2e-05,
+      "loss": 0.298,
+      "step": 24473
+    },
+    {
+      "epoch": 0.5825478434732934,
+      "grad_norm": 2.3629452225253678,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 24474
+    },
+    {
+      "epoch": 0.5825716461963248,
+      "grad_norm": 2.96807936151508,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 24475
+    },
+    {
+      "epoch": 0.5825954489193563,
+      "grad_norm": 1.0433128652105192,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 24476
+    },
+    {
+      "epoch": 0.5826192516423879,
+      "grad_norm": 2.7438397867468143,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 24477
+    },
+    {
+      "epoch": 0.5826430543654194,
+      "grad_norm": 7.915503838563957,
+      "learning_rate": 2e-05,
+      "loss": 0.2186,
+      "step": 24478
+    },
+    {
+      "epoch": 0.5826668570884509,
+      "grad_norm": 3.65506907754928,
+      "learning_rate": 2e-05,
+      "loss": 0.4135,
+      "step": 24479
+    },
+    {
+      "epoch": 0.5826906598114824,
+      "grad_norm": 7.876169880846921,
+      "learning_rate": 2e-05,
+      "loss": 0.2442,
+      "step": 24480
+    },
+    {
+      "epoch": 0.582714462534514,
+      "grad_norm": 7.570162362505544,
+      "learning_rate": 2e-05,
+      "loss": 1.3006,
+      "step": 24481
+    },
+    {
+      "epoch": 0.5827382652575455,
+      "grad_norm": 5.76258486388554,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 24482
+    },
+    {
+      "epoch": 0.582762067980577,
+      "grad_norm": 4.865902563785465,
+      "learning_rate": 2e-05,
+      "loss": 0.4768,
+      "step": 24483
+    },
+    {
+      "epoch": 0.5827858707036085,
+      "grad_norm": 3.1913583457535255,
+      "learning_rate": 2e-05,
+      "loss": 0.4256,
+      "step": 24484
+    },
+    {
+      "epoch": 0.58280967342664,
+      "grad_norm": 6.477417033870055,
+      "learning_rate": 2e-05,
+      "loss": 1.1758,
+      "step": 24485
+    },
+    {
+      "epoch": 0.5828334761496715,
+      "grad_norm": 4.823597283676364,
+      "learning_rate": 2e-05,
+      "loss": 0.7905,
+      "step": 24486
+    },
+    {
+      "epoch": 0.582857278872703,
+      "grad_norm": 0.3822240699712329,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 24487
+    },
+    {
+      "epoch": 0.5828810815957346,
+      "grad_norm": 14.222914838884277,
+      "learning_rate": 2e-05,
+      "loss": 1.0437,
+      "step": 24488
+    },
+    {
+      "epoch": 0.5829048843187661,
+      "grad_norm": 15.039483846570011,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 24489
+    },
+    {
+      "epoch": 0.5829286870417976,
+      "grad_norm": 0.8316500369517336,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 24490
+    },
+    {
+      "epoch": 0.582952489764829,
+      "grad_norm": 0.4066841063695728,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 24491
+    },
+    {
+      "epoch": 0.5829762924878606,
+      "grad_norm": 12.146777292209174,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 24492
+    },
+    {
+      "epoch": 0.5830000952108921,
+      "grad_norm": 19.941417937467886,
+      "learning_rate": 2e-05,
+      "loss": 0.5247,
+      "step": 24493
+    },
+    {
+      "epoch": 0.5830238979339236,
+      "grad_norm": 1.6108624298427072,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 24494
+    },
+    {
+      "epoch": 0.5830477006569551,
+      "grad_norm": 4.528834847171486,
+      "learning_rate": 2e-05,
+      "loss": 0.2391,
+      "step": 24495
+    },
+    {
+      "epoch": 0.5830715033799867,
+      "grad_norm": 1.128384653590322,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 24496
+    },
+    {
+      "epoch": 0.5830953061030182,
+      "grad_norm": 4.115977345424548,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 24497
+    },
+    {
+      "epoch": 0.5831191088260497,
+      "grad_norm": 8.320686036628885,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 24498
+    },
+    {
+      "epoch": 0.5831429115490813,
+      "grad_norm": 6.749670697083734,
+      "learning_rate": 2e-05,
+      "loss": 0.8464,
+      "step": 24499
+    },
+    {
+      "epoch": 0.5831667142721128,
+      "grad_norm": 2.593367488007719,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 24500
+    },
+    {
+      "epoch": 0.5831905169951442,
+      "grad_norm": 3.3818438753067688,
+      "learning_rate": 2e-05,
+      "loss": 0.7236,
+      "step": 24501
+    },
+    {
+      "epoch": 0.5832143197181757,
+      "grad_norm": 2.885151707895594,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 24502
+    },
+    {
+      "epoch": 0.5832381224412073,
+      "grad_norm": 4.322938629331024,
+      "learning_rate": 2e-05,
+      "loss": 0.6026,
+      "step": 24503
+    },
+    {
+      "epoch": 0.5832619251642388,
+      "grad_norm": 0.8703960785408036,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 24504
+    },
+    {
+      "epoch": 0.5832857278872703,
+      "grad_norm": 0.8162543391361216,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 24505
+    },
+    {
+      "epoch": 0.5833095306103018,
+      "grad_norm": 1.4923390531427105,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 24506
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 4.900942495885964,
+      "learning_rate": 2e-05,
+      "loss": 1.0593,
+      "step": 24507
+    },
+    {
+      "epoch": 0.5833571360563649,
+      "grad_norm": 0.6093092625111916,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 24508
+    },
+    {
+      "epoch": 0.5833809387793963,
+      "grad_norm": 3.9543329872910458,
+      "learning_rate": 2e-05,
+      "loss": 0.6138,
+      "step": 24509
+    },
+    {
+      "epoch": 0.5834047415024278,
+      "grad_norm": 3.328319535428819,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 24510
+    },
+    {
+      "epoch": 0.5834285442254594,
+      "grad_norm": 5.39112821388119,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 24511
+    },
+    {
+      "epoch": 0.5834523469484909,
+      "grad_norm": 10.920327869341886,
+      "learning_rate": 2e-05,
+      "loss": 0.1846,
+      "step": 24512
+    },
+    {
+      "epoch": 0.5834761496715224,
+      "grad_norm": 3.5716447106561597,
+      "learning_rate": 2e-05,
+      "loss": 0.429,
+      "step": 24513
+    },
+    {
+      "epoch": 0.583499952394554,
+      "grad_norm": 1.1554957144137508,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 24514
+    },
+    {
+      "epoch": 0.5835237551175855,
+      "grad_norm": 3.020513206353941,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 24515
+    },
+    {
+      "epoch": 0.583547557840617,
+      "grad_norm": 5.819750491729969,
+      "learning_rate": 2e-05,
+      "loss": 0.9201,
+      "step": 24516
+    },
+    {
+      "epoch": 0.5835713605636484,
+      "grad_norm": 0.1703273108003442,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 24517
+    },
+    {
+      "epoch": 0.58359516328668,
+      "grad_norm": 4.429329769126117,
+      "learning_rate": 2e-05,
+      "loss": 0.5775,
+      "step": 24518
+    },
+    {
+      "epoch": 0.5836189660097115,
+      "grad_norm": 5.7244460150694065,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 24519
+    },
+    {
+      "epoch": 0.583642768732743,
+      "grad_norm": 0.42756450842561816,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 24520
+    },
+    {
+      "epoch": 0.5836665714557745,
+      "grad_norm": 1.8154092502007404,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 24521
+    },
+    {
+      "epoch": 0.5836903741788061,
+      "grad_norm": 6.735811804099878,
+      "learning_rate": 2e-05,
+      "loss": 1.6581,
+      "step": 24522
+    },
+    {
+      "epoch": 0.5837141769018376,
+      "grad_norm": 2.937582762760096,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 24523
+    },
+    {
+      "epoch": 0.5837379796248691,
+      "grad_norm": 10.81601710477427,
+      "learning_rate": 2e-05,
+      "loss": 0.9645,
+      "step": 24524
+    },
+    {
+      "epoch": 0.5837617823479005,
+      "grad_norm": 4.180239838855143,
+      "learning_rate": 2e-05,
+      "loss": 0.5615,
+      "step": 24525
+    },
+    {
+      "epoch": 0.5837855850709321,
+      "grad_norm": 0.27934447904465326,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 24526
+    },
+    {
+      "epoch": 0.5838093877939636,
+      "grad_norm": 10.031444598039522,
+      "learning_rate": 2e-05,
+      "loss": 1.6135,
+      "step": 24527
+    },
+    {
+      "epoch": 0.5838331905169951,
+      "grad_norm": 5.485962626473386,
+      "learning_rate": 2e-05,
+      "loss": 0.615,
+      "step": 24528
+    },
+    {
+      "epoch": 0.5838569932400267,
+      "grad_norm": 1.3221925818283038,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 24529
+    },
+    {
+      "epoch": 0.5838807959630582,
+      "grad_norm": 10.721181754292354,
+      "learning_rate": 2e-05,
+      "loss": 0.7846,
+      "step": 24530
+    },
+    {
+      "epoch": 0.5839045986860897,
+      "grad_norm": 5.983601500141608,
+      "learning_rate": 2e-05,
+      "loss": 0.1603,
+      "step": 24531
+    },
+    {
+      "epoch": 0.5839284014091212,
+      "grad_norm": 8.108473696407293,
+      "learning_rate": 2e-05,
+      "loss": 0.6666,
+      "step": 24532
+    },
+    {
+      "epoch": 0.5839522041321528,
+      "grad_norm": 3.861898766059231,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 24533
+    },
+    {
+      "epoch": 0.5839760068551842,
+      "grad_norm": 3.019436585595809,
+      "learning_rate": 2e-05,
+      "loss": 0.0488,
+      "step": 24534
+    },
+    {
+      "epoch": 0.5839998095782157,
+      "grad_norm": 5.985539526583513,
+      "learning_rate": 2e-05,
+      "loss": 0.6841,
+      "step": 24535
+    },
+    {
+      "epoch": 0.5840236123012472,
+      "grad_norm": 0.30240366824249815,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 24536
+    },
+    {
+      "epoch": 0.5840474150242788,
+      "grad_norm": 0.6036721656935723,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 24537
+    },
+    {
+      "epoch": 0.5840712177473103,
+      "grad_norm": 1.0476258036639345,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 24538
+    },
+    {
+      "epoch": 0.5840950204703418,
+      "grad_norm": 4.745507022311813,
+      "learning_rate": 2e-05,
+      "loss": 0.7048,
+      "step": 24539
+    },
+    {
+      "epoch": 0.5841188231933733,
+      "grad_norm": 5.824738723472507,
+      "learning_rate": 2e-05,
+      "loss": 0.3698,
+      "step": 24540
+    },
+    {
+      "epoch": 0.5841426259164049,
+      "grad_norm": 2.473249972857874,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 24541
+    },
+    {
+      "epoch": 0.5841664286394364,
+      "grad_norm": 4.960667883638005,
+      "learning_rate": 2e-05,
+      "loss": 0.3868,
+      "step": 24542
+    },
+    {
+      "epoch": 0.5841902313624678,
+      "grad_norm": 0.5759328351181145,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 24543
+    },
+    {
+      "epoch": 0.5842140340854994,
+      "grad_norm": 2.001326743791429,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 24544
+    },
+    {
+      "epoch": 0.5842378368085309,
+      "grad_norm": 4.898734329091255,
+      "learning_rate": 2e-05,
+      "loss": 0.7857,
+      "step": 24545
+    },
+    {
+      "epoch": 0.5842616395315624,
+      "grad_norm": 5.125174916970732,
+      "learning_rate": 2e-05,
+      "loss": 0.4964,
+      "step": 24546
+    },
+    {
+      "epoch": 0.5842854422545939,
+      "grad_norm": 1.7370226563340259,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 24547
+    },
+    {
+      "epoch": 0.5843092449776255,
+      "grad_norm": 1.858501309275725,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 24548
+    },
+    {
+      "epoch": 0.584333047700657,
+      "grad_norm": 6.445021280195922,
+      "learning_rate": 2e-05,
+      "loss": 0.1645,
+      "step": 24549
+    },
+    {
+      "epoch": 0.5843568504236885,
+      "grad_norm": 3.4447069393348992,
+      "learning_rate": 2e-05,
+      "loss": 0.4375,
+      "step": 24550
+    },
+    {
+      "epoch": 0.5843806531467199,
+      "grad_norm": 3.713995329878754,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 24551
+    },
+    {
+      "epoch": 0.5844044558697515,
+      "grad_norm": 0.11307377247427985,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 24552
+    },
+    {
+      "epoch": 0.584428258592783,
+      "grad_norm": 5.1757211429093575,
+      "learning_rate": 2e-05,
+      "loss": 0.2535,
+      "step": 24553
+    },
+    {
+      "epoch": 0.5844520613158145,
+      "grad_norm": 7.098518165777812,
+      "learning_rate": 2e-05,
+      "loss": 0.416,
+      "step": 24554
+    },
+    {
+      "epoch": 0.584475864038846,
+      "grad_norm": 0.06772770297231069,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 24555
+    },
+    {
+      "epoch": 0.5844996667618776,
+      "grad_norm": 3.1785109878036595,
+      "learning_rate": 2e-05,
+      "loss": 0.4519,
+      "step": 24556
+    },
+    {
+      "epoch": 0.5845234694849091,
+      "grad_norm": 3.097054777750799,
+      "learning_rate": 2e-05,
+      "loss": 0.349,
+      "step": 24557
+    },
+    {
+      "epoch": 0.5845472722079406,
+      "grad_norm": 3.1837527348548416,
+      "learning_rate": 2e-05,
+      "loss": 0.7807,
+      "step": 24558
+    },
+    {
+      "epoch": 0.5845710749309722,
+      "grad_norm": 4.797865049421982,
+      "learning_rate": 2e-05,
+      "loss": 0.5765,
+      "step": 24559
+    },
+    {
+      "epoch": 0.5845948776540036,
+      "grad_norm": 1.3027297174446557,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 24560
+    },
+    {
+      "epoch": 0.5846186803770351,
+      "grad_norm": 6.296902972422782,
+      "learning_rate": 2e-05,
+      "loss": 1.2903,
+      "step": 24561
+    },
+    {
+      "epoch": 0.5846424831000666,
+      "grad_norm": 3.7528506113479136,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 24562
+    },
+    {
+      "epoch": 0.5846662858230982,
+      "grad_norm": 5.038990331484763,
+      "learning_rate": 2e-05,
+      "loss": 0.7868,
+      "step": 24563
+    },
+    {
+      "epoch": 0.5846900885461297,
+      "grad_norm": 0.8124729491506453,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 24564
+    },
+    {
+      "epoch": 0.5847138912691612,
+      "grad_norm": 0.19021151370861353,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24565
+    },
+    {
+      "epoch": 0.5847376939921927,
+      "grad_norm": 3.817049293239293,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 24566
+    },
+    {
+      "epoch": 0.5847614967152243,
+      "grad_norm": 0.6062429805529921,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 24567
+    },
+    {
+      "epoch": 0.5847852994382557,
+      "grad_norm": 0.893731500548019,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 24568
+    },
+    {
+      "epoch": 0.5848091021612872,
+      "grad_norm": 3.116152499473973,
+      "learning_rate": 2e-05,
+      "loss": 0.3721,
+      "step": 24569
+    },
+    {
+      "epoch": 0.5848329048843187,
+      "grad_norm": 11.470424858719396,
+      "learning_rate": 2e-05,
+      "loss": 0.4425,
+      "step": 24570
+    },
+    {
+      "epoch": 0.5848567076073503,
+      "grad_norm": 4.76202383106129,
+      "learning_rate": 2e-05,
+      "loss": 0.8432,
+      "step": 24571
+    },
+    {
+      "epoch": 0.5848805103303818,
+      "grad_norm": 1.3222363139345,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 24572
+    },
+    {
+      "epoch": 0.5849043130534133,
+      "grad_norm": 3.5519532433064116,
+      "learning_rate": 2e-05,
+      "loss": 0.149,
+      "step": 24573
+    },
+    {
+      "epoch": 0.5849281157764449,
+      "grad_norm": 5.312158291809599,
+      "learning_rate": 2e-05,
+      "loss": 0.9363,
+      "step": 24574
+    },
+    {
+      "epoch": 0.5849519184994764,
+      "grad_norm": 9.299512742377548,
+      "learning_rate": 2e-05,
+      "loss": 0.4513,
+      "step": 24575
+    },
+    {
+      "epoch": 0.5849757212225078,
+      "grad_norm": 3.527710296545616,
+      "learning_rate": 2e-05,
+      "loss": 0.671,
+      "step": 24576
+    },
+    {
+      "epoch": 0.5849995239455393,
+      "grad_norm": 5.46595653208528,
+      "learning_rate": 2e-05,
+      "loss": 0.6527,
+      "step": 24577
+    },
+    {
+      "epoch": 0.5850233266685709,
+      "grad_norm": 0.9281168310464608,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 24578
+    },
+    {
+      "epoch": 0.5850471293916024,
+      "grad_norm": 6.4729170069560755,
+      "learning_rate": 2e-05,
+      "loss": 1.1652,
+      "step": 24579
+    },
+    {
+      "epoch": 0.5850709321146339,
+      "grad_norm": 2.933985493783792,
+      "learning_rate": 2e-05,
+      "loss": 0.4468,
+      "step": 24580
+    },
+    {
+      "epoch": 0.5850947348376654,
+      "grad_norm": 3.1127051836376682,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 24581
+    },
+    {
+      "epoch": 0.585118537560697,
+      "grad_norm": 4.60045712554793,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 24582
+    },
+    {
+      "epoch": 0.5851423402837285,
+      "grad_norm": 4.520423119532849,
+      "learning_rate": 2e-05,
+      "loss": 0.9441,
+      "step": 24583
+    },
+    {
+      "epoch": 0.58516614300676,
+      "grad_norm": 6.576919778738933,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 24584
+    },
+    {
+      "epoch": 0.5851899457297914,
+      "grad_norm": 7.634993447254914,
+      "learning_rate": 2e-05,
+      "loss": 0.2574,
+      "step": 24585
+    },
+    {
+      "epoch": 0.585213748452823,
+      "grad_norm": 0.32970545737289525,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 24586
+    },
+    {
+      "epoch": 0.5852375511758545,
+      "grad_norm": 4.078716811912898,
+      "learning_rate": 2e-05,
+      "loss": 0.907,
+      "step": 24587
+    },
+    {
+      "epoch": 0.585261353898886,
+      "grad_norm": 5.483197340989405,
+      "learning_rate": 2e-05,
+      "loss": 0.6863,
+      "step": 24588
+    },
+    {
+      "epoch": 0.5852851566219176,
+      "grad_norm": 9.35189178737774,
+      "learning_rate": 2e-05,
+      "loss": 1.8922,
+      "step": 24589
+    },
+    {
+      "epoch": 0.5853089593449491,
+      "grad_norm": 2.68765144610037,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 24590
+    },
+    {
+      "epoch": 0.5853327620679806,
+      "grad_norm": 0.7338115588957397,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 24591
+    },
+    {
+      "epoch": 0.5853565647910121,
+      "grad_norm": 0.3708529411041155,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 24592
+    },
+    {
+      "epoch": 0.5853803675140437,
+      "grad_norm": 7.6106425346367335,
+      "learning_rate": 2e-05,
+      "loss": 1.1207,
+      "step": 24593
+    },
+    {
+      "epoch": 0.5854041702370751,
+      "grad_norm": 0.42094475298482664,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 24594
+    },
+    {
+      "epoch": 0.5854279729601066,
+      "grad_norm": 4.8618194462632625,
+      "learning_rate": 2e-05,
+      "loss": 0.232,
+      "step": 24595
+    },
+    {
+      "epoch": 0.5854517756831381,
+      "grad_norm": 1.0941808605853782,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 24596
+    },
+    {
+      "epoch": 0.5854755784061697,
+      "grad_norm": 2.042551280227327,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 24597
+    },
+    {
+      "epoch": 0.5854993811292012,
+      "grad_norm": 2.592756708394954,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 24598
+    },
+    {
+      "epoch": 0.5855231838522327,
+      "grad_norm": 3.3117615979239314,
+      "learning_rate": 2e-05,
+      "loss": 0.1293,
+      "step": 24599
+    },
+    {
+      "epoch": 0.5855469865752642,
+      "grad_norm": 8.790260385829653,
+      "learning_rate": 2e-05,
+      "loss": 0.9519,
+      "step": 24600
+    },
+    {
+      "epoch": 0.5855707892982958,
+      "grad_norm": 5.691991161625046,
+      "learning_rate": 2e-05,
+      "loss": 1.1655,
+      "step": 24601
+    },
+    {
+      "epoch": 0.5855945920213272,
+      "grad_norm": 2.795406612115244,
+      "learning_rate": 2e-05,
+      "loss": 0.3972,
+      "step": 24602
+    },
+    {
+      "epoch": 0.5856183947443587,
+      "grad_norm": 1.0994288857027743,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 24603
+    },
+    {
+      "epoch": 0.5856421974673903,
+      "grad_norm": 1.062284543012495,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 24604
+    },
+    {
+      "epoch": 0.5856660001904218,
+      "grad_norm": 2.2592739600045793,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 24605
+    },
+    {
+      "epoch": 0.5856898029134533,
+      "grad_norm": 6.376746436139472,
+      "learning_rate": 2e-05,
+      "loss": 1.1684,
+      "step": 24606
+    },
+    {
+      "epoch": 0.5857136056364848,
+      "grad_norm": 4.991419061492042,
+      "learning_rate": 2e-05,
+      "loss": 0.6853,
+      "step": 24607
+    },
+    {
+      "epoch": 0.5857374083595164,
+      "grad_norm": 1.04141957139149,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 24608
+    },
+    {
+      "epoch": 0.5857612110825479,
+      "grad_norm": 3.134094807514415,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 24609
+    },
+    {
+      "epoch": 0.5857850138055793,
+      "grad_norm": 3.2831467616404386,
+      "learning_rate": 2e-05,
+      "loss": 0.4552,
+      "step": 24610
+    },
+    {
+      "epoch": 0.5858088165286108,
+      "grad_norm": 2.6897160644729228,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 24611
+    },
+    {
+      "epoch": 0.5858326192516424,
+      "grad_norm": 5.064723008809446,
+      "learning_rate": 2e-05,
+      "loss": 0.6746,
+      "step": 24612
+    },
+    {
+      "epoch": 0.5858564219746739,
+      "grad_norm": 3.3326979252585858,
+      "learning_rate": 2e-05,
+      "loss": 0.3657,
+      "step": 24613
+    },
+    {
+      "epoch": 0.5858802246977054,
+      "grad_norm": 4.712120326456533,
+      "learning_rate": 2e-05,
+      "loss": 0.2513,
+      "step": 24614
+    },
+    {
+      "epoch": 0.5859040274207369,
+      "grad_norm": 0.34631157141981084,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 24615
+    },
+    {
+      "epoch": 0.5859278301437685,
+      "grad_norm": 0.8523303798120881,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 24616
+    },
+    {
+      "epoch": 0.5859516328668,
+      "grad_norm": 2.2889679248740227,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 24617
+    },
+    {
+      "epoch": 0.5859754355898315,
+      "grad_norm": 4.938518259673556,
+      "learning_rate": 2e-05,
+      "loss": 0.6324,
+      "step": 24618
+    },
+    {
+      "epoch": 0.585999238312863,
+      "grad_norm": 5.859932879430394,
+      "learning_rate": 2e-05,
+      "loss": 0.5239,
+      "step": 24619
+    },
+    {
+      "epoch": 0.5860230410358945,
+      "grad_norm": 6.118773218972969,
+      "learning_rate": 2e-05,
+      "loss": 0.5053,
+      "step": 24620
+    },
+    {
+      "epoch": 0.586046843758926,
+      "grad_norm": 0.29888156680940725,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 24621
+    },
+    {
+      "epoch": 0.5860706464819575,
+      "grad_norm": 8.896308526641672,
+      "learning_rate": 2e-05,
+      "loss": 0.7348,
+      "step": 24622
+    },
+    {
+      "epoch": 0.5860944492049891,
+      "grad_norm": 4.251274719036778,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 24623
+    },
+    {
+      "epoch": 0.5861182519280206,
+      "grad_norm": 0.18690417299903647,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24624
+    },
+    {
+      "epoch": 0.5861420546510521,
+      "grad_norm": 2.3474485724797396,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 24625
+    },
+    {
+      "epoch": 0.5861658573740836,
+      "grad_norm": 7.982381976928224,
+      "learning_rate": 2e-05,
+      "loss": 0.4482,
+      "step": 24626
+    },
+    {
+      "epoch": 0.5861896600971152,
+      "grad_norm": 6.5627907660551505,
+      "learning_rate": 2e-05,
+      "loss": 0.7469,
+      "step": 24627
+    },
+    {
+      "epoch": 0.5862134628201466,
+      "grad_norm": 4.431317285810692,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 24628
+    },
+    {
+      "epoch": 0.5862372655431781,
+      "grad_norm": 3.557051374223546,
+      "learning_rate": 2e-05,
+      "loss": 0.4206,
+      "step": 24629
+    },
+    {
+      "epoch": 0.5862610682662096,
+      "grad_norm": 2.491957987304139,
+      "learning_rate": 2e-05,
+      "loss": 0.1351,
+      "step": 24630
+    },
+    {
+      "epoch": 0.5862848709892412,
+      "grad_norm": 3.2822261091782443,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 24631
+    },
+    {
+      "epoch": 0.5863086737122727,
+      "grad_norm": 6.160262792459885,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 24632
+    },
+    {
+      "epoch": 0.5863324764353042,
+      "grad_norm": 3.268268902813584,
+      "learning_rate": 2e-05,
+      "loss": 0.5127,
+      "step": 24633
+    },
+    {
+      "epoch": 0.5863562791583357,
+      "grad_norm": 4.645080953110462,
+      "learning_rate": 2e-05,
+      "loss": 0.4398,
+      "step": 24634
+    },
+    {
+      "epoch": 0.5863800818813673,
+      "grad_norm": 4.359520372565657,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 24635
+    },
+    {
+      "epoch": 0.5864038846043987,
+      "grad_norm": 2.4427008483632466,
+      "learning_rate": 2e-05,
+      "loss": 0.3728,
+      "step": 24636
+    },
+    {
+      "epoch": 0.5864276873274302,
+      "grad_norm": 6.845617710312143,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 24637
+    },
+    {
+      "epoch": 0.5864514900504618,
+      "grad_norm": 3.3061613140177655,
+      "learning_rate": 2e-05,
+      "loss": 0.8227,
+      "step": 24638
+    },
+    {
+      "epoch": 0.5864752927734933,
+      "grad_norm": 29.982481057368087,
+      "learning_rate": 2e-05,
+      "loss": 0.2423,
+      "step": 24639
+    },
+    {
+      "epoch": 0.5864990954965248,
+      "grad_norm": 1.2599641144144251,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 24640
+    },
+    {
+      "epoch": 0.5865228982195563,
+      "grad_norm": 2.085153921938262,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 24641
+    },
+    {
+      "epoch": 0.5865467009425879,
+      "grad_norm": 3.891958323471796,
+      "learning_rate": 2e-05,
+      "loss": 0.4678,
+      "step": 24642
+    },
+    {
+      "epoch": 0.5865705036656194,
+      "grad_norm": 7.552140615621259,
+      "learning_rate": 2e-05,
+      "loss": 0.5133,
+      "step": 24643
+    },
+    {
+      "epoch": 0.5865943063886508,
+      "grad_norm": 10.69100137768507,
+      "learning_rate": 2e-05,
+      "loss": 0.8311,
+      "step": 24644
+    },
+    {
+      "epoch": 0.5866181091116823,
+      "grad_norm": 1.9942197479888633,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 24645
+    },
+    {
+      "epoch": 0.5866419118347139,
+      "grad_norm": 0.5317212577483946,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 24646
+    },
+    {
+      "epoch": 0.5866657145577454,
+      "grad_norm": 0.7107226645582027,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 24647
+    },
+    {
+      "epoch": 0.5866895172807769,
+      "grad_norm": 3.5623227202832664,
+      "learning_rate": 2e-05,
+      "loss": 0.5685,
+      "step": 24648
+    },
+    {
+      "epoch": 0.5867133200038084,
+      "grad_norm": 1.318179255994893,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 24649
+    },
+    {
+      "epoch": 0.58673712272684,
+      "grad_norm": 2.403419857561929,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 24650
+    },
+    {
+      "epoch": 0.5867609254498715,
+      "grad_norm": 2.5999574586541843,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 24651
+    },
+    {
+      "epoch": 0.586784728172903,
+      "grad_norm": 0.7099535810989287,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 24652
+    },
+    {
+      "epoch": 0.5868085308959345,
+      "grad_norm": 0.47483857723583023,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 24653
+    },
+    {
+      "epoch": 0.586832333618966,
+      "grad_norm": 0.48431847590359056,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 24654
+    },
+    {
+      "epoch": 0.5868561363419975,
+      "grad_norm": 4.6275194716952655,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 24655
+    },
+    {
+      "epoch": 0.586879939065029,
+      "grad_norm": 5.387027354807388,
+      "learning_rate": 2e-05,
+      "loss": 0.4224,
+      "step": 24656
+    },
+    {
+      "epoch": 0.5869037417880606,
+      "grad_norm": 10.304949063583422,
+      "learning_rate": 2e-05,
+      "loss": 0.4511,
+      "step": 24657
+    },
+    {
+      "epoch": 0.5869275445110921,
+      "grad_norm": 7.84396003467697,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 24658
+    },
+    {
+      "epoch": 0.5869513472341236,
+      "grad_norm": 2.6800259885209585,
+      "learning_rate": 2e-05,
+      "loss": 0.4344,
+      "step": 24659
+    },
+    {
+      "epoch": 0.586975149957155,
+      "grad_norm": 7.071122555396654,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 24660
+    },
+    {
+      "epoch": 0.5869989526801866,
+      "grad_norm": 0.7461000016160116,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 24661
+    },
+    {
+      "epoch": 0.5870227554032181,
+      "grad_norm": 3.8160275090232183,
+      "learning_rate": 2e-05,
+      "loss": 0.4352,
+      "step": 24662
+    },
+    {
+      "epoch": 0.5870465581262496,
+      "grad_norm": 4.605582859962298,
+      "learning_rate": 2e-05,
+      "loss": 0.5613,
+      "step": 24663
+    },
+    {
+      "epoch": 0.5870703608492811,
+      "grad_norm": 1.0749791450567407,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 24664
+    },
+    {
+      "epoch": 0.5870941635723127,
+      "grad_norm": 21.780088727612178,
+      "learning_rate": 2e-05,
+      "loss": 0.9726,
+      "step": 24665
+    },
+    {
+      "epoch": 0.5871179662953442,
+      "grad_norm": 4.500662179685736,
+      "learning_rate": 2e-05,
+      "loss": 0.1583,
+      "step": 24666
+    },
+    {
+      "epoch": 0.5871417690183757,
+      "grad_norm": 3.5075663691131753,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 24667
+    },
+    {
+      "epoch": 0.5871655717414073,
+      "grad_norm": 3.4712433872510595,
+      "learning_rate": 2e-05,
+      "loss": 0.3403,
+      "step": 24668
+    },
+    {
+      "epoch": 0.5871893744644388,
+      "grad_norm": 1.0691499347007896,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 24669
+    },
+    {
+      "epoch": 0.5872131771874702,
+      "grad_norm": 12.94555091345416,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 24670
+    },
+    {
+      "epoch": 0.5872369799105017,
+      "grad_norm": 5.090132456458726,
+      "learning_rate": 2e-05,
+      "loss": 0.9972,
+      "step": 24671
+    },
+    {
+      "epoch": 0.5872607826335333,
+      "grad_norm": 4.4758725105358135,
+      "learning_rate": 2e-05,
+      "loss": 0.7038,
+      "step": 24672
+    },
+    {
+      "epoch": 0.5872845853565648,
+      "grad_norm": 1.8149440122063198,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 24673
+    },
+    {
+      "epoch": 0.5873083880795963,
+      "grad_norm": 1.5647187369124678,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 24674
+    },
+    {
+      "epoch": 0.5873321908026278,
+      "grad_norm": 5.761323146325854,
+      "learning_rate": 2e-05,
+      "loss": 0.2335,
+      "step": 24675
+    },
+    {
+      "epoch": 0.5873559935256594,
+      "grad_norm": 8.370970251989505,
+      "learning_rate": 2e-05,
+      "loss": 1.2012,
+      "step": 24676
+    },
+    {
+      "epoch": 0.5873797962486909,
+      "grad_norm": 0.8650588212826684,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 24677
+    },
+    {
+      "epoch": 0.5874035989717223,
+      "grad_norm": 6.669929369953088,
+      "learning_rate": 2e-05,
+      "loss": 0.812,
+      "step": 24678
+    },
+    {
+      "epoch": 0.5874274016947538,
+      "grad_norm": 2.2436336644820933,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 24679
+    },
+    {
+      "epoch": 0.5874512044177854,
+      "grad_norm": 3.887142383048761,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 24680
+    },
+    {
+      "epoch": 0.5874750071408169,
+      "grad_norm": 0.699130673394739,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 24681
+    },
+    {
+      "epoch": 0.5874988098638484,
+      "grad_norm": 4.235207443228934,
+      "learning_rate": 2e-05,
+      "loss": 0.9021,
+      "step": 24682
+    },
+    {
+      "epoch": 0.58752261258688,
+      "grad_norm": 3.5684324267769982,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 24683
+    },
+    {
+      "epoch": 0.5875464153099115,
+      "grad_norm": 4.5883849635134615,
+      "learning_rate": 2e-05,
+      "loss": 1.0555,
+      "step": 24684
+    },
+    {
+      "epoch": 0.587570218032943,
+      "grad_norm": 0.8604323418286751,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 24685
+    },
+    {
+      "epoch": 0.5875940207559744,
+      "grad_norm": 3.739672663779674,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 24686
+    },
+    {
+      "epoch": 0.587617823479006,
+      "grad_norm": 1.2848261423943677,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 24687
+    },
+    {
+      "epoch": 0.5876416262020375,
+      "grad_norm": 7.8275913016144605,
+      "learning_rate": 2e-05,
+      "loss": 0.7943,
+      "step": 24688
+    },
+    {
+      "epoch": 0.587665428925069,
+      "grad_norm": 3.4006790517209575,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 24689
+    },
+    {
+      "epoch": 0.5876892316481005,
+      "grad_norm": 5.1915750528059625,
+      "learning_rate": 2e-05,
+      "loss": 0.6056,
+      "step": 24690
+    },
+    {
+      "epoch": 0.5877130343711321,
+      "grad_norm": 2.5759020138923985,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 24691
+    },
+    {
+      "epoch": 0.5877368370941636,
+      "grad_norm": 0.3685221853021211,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 24692
+    },
+    {
+      "epoch": 0.5877606398171951,
+      "grad_norm": 1.0180576222753484,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 24693
+    },
+    {
+      "epoch": 0.5877844425402265,
+      "grad_norm": 5.953650009241104,
+      "learning_rate": 2e-05,
+      "loss": 1.347,
+      "step": 24694
+    },
+    {
+      "epoch": 0.5878082452632581,
+      "grad_norm": 7.657847287713764,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 24695
+    },
+    {
+      "epoch": 0.5878320479862896,
+      "grad_norm": 7.562104374480845,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 24696
+    },
+    {
+      "epoch": 0.5878558507093211,
+      "grad_norm": 3.308981549224036,
+      "learning_rate": 2e-05,
+      "loss": 0.3807,
+      "step": 24697
+    },
+    {
+      "epoch": 0.5878796534323527,
+      "grad_norm": 0.6336885102141704,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 24698
+    },
+    {
+      "epoch": 0.5879034561553842,
+      "grad_norm": 1.9332228945971168,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 24699
+    },
+    {
+      "epoch": 0.5879272588784157,
+      "grad_norm": 0.7399725762604901,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 24700
+    },
+    {
+      "epoch": 0.5879510616014472,
+      "grad_norm": 5.94872150447727,
+      "learning_rate": 2e-05,
+      "loss": 0.5603,
+      "step": 24701
+    },
+    {
+      "epoch": 0.5879748643244788,
+      "grad_norm": 6.602743953619212,
+      "learning_rate": 2e-05,
+      "loss": 0.9785,
+      "step": 24702
+    },
+    {
+      "epoch": 0.5879986670475102,
+      "grad_norm": 2.390116875319352,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 24703
+    },
+    {
+      "epoch": 0.5880224697705417,
+      "grad_norm": 5.310289711898084,
+      "learning_rate": 2e-05,
+      "loss": 0.2405,
+      "step": 24704
+    },
+    {
+      "epoch": 0.5880462724935732,
+      "grad_norm": 4.096731743194578,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 24705
+    },
+    {
+      "epoch": 0.5880700752166048,
+      "grad_norm": 2.92294657554564,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 24706
+    },
+    {
+      "epoch": 0.5880938779396363,
+      "grad_norm": 7.4319479411514004,
+      "learning_rate": 2e-05,
+      "loss": 1.8389,
+      "step": 24707
+    },
+    {
+      "epoch": 0.5881176806626678,
+      "grad_norm": 1.469780939168535,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 24708
+    },
+    {
+      "epoch": 0.5881414833856993,
+      "grad_norm": 4.542223921844646,
+      "learning_rate": 2e-05,
+      "loss": 0.3656,
+      "step": 24709
+    },
+    {
+      "epoch": 0.5881652861087309,
+      "grad_norm": 5.8637424855366405,
+      "learning_rate": 2e-05,
+      "loss": 0.7626,
+      "step": 24710
+    },
+    {
+      "epoch": 0.5881890888317624,
+      "grad_norm": 7.512879991347153,
+      "learning_rate": 2e-05,
+      "loss": 0.9693,
+      "step": 24711
+    },
+    {
+      "epoch": 0.5882128915547938,
+      "grad_norm": 3.4056362214570566,
+      "learning_rate": 2e-05,
+      "loss": 0.6949,
+      "step": 24712
+    },
+    {
+      "epoch": 0.5882366942778254,
+      "grad_norm": 8.76128718910579,
+      "learning_rate": 2e-05,
+      "loss": 1.4495,
+      "step": 24713
+    },
+    {
+      "epoch": 0.5882604970008569,
+      "grad_norm": 0.0954824761729651,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 24714
+    },
+    {
+      "epoch": 0.5882842997238884,
+      "grad_norm": 0.5661240182994012,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 24715
+    },
+    {
+      "epoch": 0.5883081024469199,
+      "grad_norm": 0.876254160337495,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 24716
+    },
+    {
+      "epoch": 0.5883319051699515,
+      "grad_norm": 16.338876419984604,
+      "learning_rate": 2e-05,
+      "loss": 0.4469,
+      "step": 24717
+    },
+    {
+      "epoch": 0.588355707892983,
+      "grad_norm": 1.3115179347035393,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 24718
+    },
+    {
+      "epoch": 0.5883795106160145,
+      "grad_norm": 6.4344732601694705,
+      "learning_rate": 2e-05,
+      "loss": 0.4972,
+      "step": 24719
+    },
+    {
+      "epoch": 0.5884033133390459,
+      "grad_norm": 1.9508567823755298,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 24720
+    },
+    {
+      "epoch": 0.5884271160620775,
+      "grad_norm": 0.5186169993789956,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 24721
+    },
+    {
+      "epoch": 0.588450918785109,
+      "grad_norm": 3.5554431129558193,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 24722
+    },
+    {
+      "epoch": 0.5884747215081405,
+      "grad_norm": 1.7147209873528895,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 24723
+    },
+    {
+      "epoch": 0.588498524231172,
+      "grad_norm": 1.9072767387464884,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 24724
+    },
+    {
+      "epoch": 0.5885223269542036,
+      "grad_norm": 1.4835991316939103,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 24725
+    },
+    {
+      "epoch": 0.5885461296772351,
+      "grad_norm": 5.912137877656929,
+      "learning_rate": 2e-05,
+      "loss": 1.3991,
+      "step": 24726
+    },
+    {
+      "epoch": 0.5885699324002666,
+      "grad_norm": 4.690449059776985,
+      "learning_rate": 2e-05,
+      "loss": 0.9792,
+      "step": 24727
+    },
+    {
+      "epoch": 0.5885937351232982,
+      "grad_norm": 1.3319003007319947,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 24728
+    },
+    {
+      "epoch": 0.5886175378463296,
+      "grad_norm": 2.348723942250063,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 24729
+    },
+    {
+      "epoch": 0.5886413405693611,
+      "grad_norm": 1.3436592999171693,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 24730
+    },
+    {
+      "epoch": 0.5886651432923926,
+      "grad_norm": 2.95151418695883,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 24731
+    },
+    {
+      "epoch": 0.5886889460154242,
+      "grad_norm": 6.157900564603109,
+      "learning_rate": 2e-05,
+      "loss": 0.7571,
+      "step": 24732
+    },
+    {
+      "epoch": 0.5887127487384557,
+      "grad_norm": 2.631804616657992,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 24733
+    },
+    {
+      "epoch": 0.5887365514614872,
+      "grad_norm": 5.821672283445376,
+      "learning_rate": 2e-05,
+      "loss": 0.8773,
+      "step": 24734
+    },
+    {
+      "epoch": 0.5887603541845187,
+      "grad_norm": 3.3552142377909786,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 24735
+    },
+    {
+      "epoch": 0.5887841569075503,
+      "grad_norm": 9.784473139445087,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 24736
+    },
+    {
+      "epoch": 0.5888079596305817,
+      "grad_norm": 1.7843377385638302,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 24737
+    },
+    {
+      "epoch": 0.5888317623536132,
+      "grad_norm": 13.362509753742469,
+      "learning_rate": 2e-05,
+      "loss": 0.659,
+      "step": 24738
+    },
+    {
+      "epoch": 0.5888555650766447,
+      "grad_norm": 3.2723638612393775,
+      "learning_rate": 2e-05,
+      "loss": 0.716,
+      "step": 24739
+    },
+    {
+      "epoch": 0.5888793677996763,
+      "grad_norm": 0.8952512300629304,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 24740
+    },
+    {
+      "epoch": 0.5889031705227078,
+      "grad_norm": 1.4687768547191704,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 24741
+    },
+    {
+      "epoch": 0.5889269732457393,
+      "grad_norm": 0.06037183855030235,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 24742
+    },
+    {
+      "epoch": 0.5889507759687709,
+      "grad_norm": 5.460505752869424,
+      "learning_rate": 2e-05,
+      "loss": 0.194,
+      "step": 24743
+    },
+    {
+      "epoch": 0.5889745786918024,
+      "grad_norm": 3.895079030736122,
+      "learning_rate": 2e-05,
+      "loss": 0.7902,
+      "step": 24744
+    },
+    {
+      "epoch": 0.5889983814148338,
+      "grad_norm": 5.616750951315629,
+      "learning_rate": 2e-05,
+      "loss": 0.5027,
+      "step": 24745
+    },
+    {
+      "epoch": 0.5890221841378653,
+      "grad_norm": 9.516578927320491,
+      "learning_rate": 2e-05,
+      "loss": 0.4561,
+      "step": 24746
+    },
+    {
+      "epoch": 0.5890459868608969,
+      "grad_norm": 11.179858836195756,
+      "learning_rate": 2e-05,
+      "loss": 0.3664,
+      "step": 24747
+    },
+    {
+      "epoch": 0.5890697895839284,
+      "grad_norm": 3.555587239970357,
+      "learning_rate": 2e-05,
+      "loss": 0.1331,
+      "step": 24748
+    },
+    {
+      "epoch": 0.5890935923069599,
+      "grad_norm": 5.5358386900417065,
+      "learning_rate": 2e-05,
+      "loss": 0.8915,
+      "step": 24749
+    },
+    {
+      "epoch": 0.5891173950299914,
+      "grad_norm": 3.995934048209209,
+      "learning_rate": 2e-05,
+      "loss": 1.0675,
+      "step": 24750
+    },
+    {
+      "epoch": 0.589141197753023,
+      "grad_norm": 2.669730983734982,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 24751
+    },
+    {
+      "epoch": 0.5891650004760545,
+      "grad_norm": 4.523595045371588,
+      "learning_rate": 2e-05,
+      "loss": 0.6237,
+      "step": 24752
+    },
+    {
+      "epoch": 0.589188803199086,
+      "grad_norm": 6.349838726380846,
+      "learning_rate": 2e-05,
+      "loss": 1.2577,
+      "step": 24753
+    },
+    {
+      "epoch": 0.5892126059221174,
+      "grad_norm": 1.7172377922526298,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 24754
+    },
+    {
+      "epoch": 0.589236408645149,
+      "grad_norm": 4.341542784304366,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 24755
+    },
+    {
+      "epoch": 0.5892602113681805,
+      "grad_norm": 0.6346295854015115,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 24756
+    },
+    {
+      "epoch": 0.589284014091212,
+      "grad_norm": 6.140391284540455,
+      "learning_rate": 2e-05,
+      "loss": 0.4423,
+      "step": 24757
+    },
+    {
+      "epoch": 0.5893078168142436,
+      "grad_norm": 1.2519634669717146,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 24758
+    },
+    {
+      "epoch": 0.5893316195372751,
+      "grad_norm": 7.643819891485452,
+      "learning_rate": 2e-05,
+      "loss": 0.7872,
+      "step": 24759
+    },
+    {
+      "epoch": 0.5893554222603066,
+      "grad_norm": 6.290145004667926,
+      "learning_rate": 2e-05,
+      "loss": 1.087,
+      "step": 24760
+    },
+    {
+      "epoch": 0.5893792249833381,
+      "grad_norm": 2.693220461396117,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 24761
+    },
+    {
+      "epoch": 0.5894030277063697,
+      "grad_norm": 0.9946189086050962,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 24762
+    },
+    {
+      "epoch": 0.5894268304294011,
+      "grad_norm": 1.912650799828834,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 24763
+    },
+    {
+      "epoch": 0.5894506331524326,
+      "grad_norm": 4.9651912493639125,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 24764
+    },
+    {
+      "epoch": 0.5894744358754641,
+      "grad_norm": 9.618958229704546,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 24765
+    },
+    {
+      "epoch": 0.5894982385984957,
+      "grad_norm": 4.928215179367613,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 24766
+    },
+    {
+      "epoch": 0.5895220413215272,
+      "grad_norm": 3.5861486981326096,
+      "learning_rate": 2e-05,
+      "loss": 0.8544,
+      "step": 24767
+    },
+    {
+      "epoch": 0.5895458440445587,
+      "grad_norm": 5.182766075822359,
+      "learning_rate": 2e-05,
+      "loss": 1.0836,
+      "step": 24768
+    },
+    {
+      "epoch": 0.5895696467675902,
+      "grad_norm": 3.0287432759993895,
+      "learning_rate": 2e-05,
+      "loss": 0.0901,
+      "step": 24769
+    },
+    {
+      "epoch": 0.5895934494906218,
+      "grad_norm": 0.4456114188026091,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 24770
+    },
+    {
+      "epoch": 0.5896172522136532,
+      "grad_norm": 0.9365938563734862,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 24771
+    },
+    {
+      "epoch": 0.5896410549366847,
+      "grad_norm": 4.172943003319888,
+      "learning_rate": 2e-05,
+      "loss": 0.2827,
+      "step": 24772
+    },
+    {
+      "epoch": 0.5896648576597163,
+      "grad_norm": 7.623618141709704,
+      "learning_rate": 2e-05,
+      "loss": 0.3415,
+      "step": 24773
+    },
+    {
+      "epoch": 0.5896886603827478,
+      "grad_norm": 2.6663917444261878,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 24774
+    },
+    {
+      "epoch": 0.5897124631057793,
+      "grad_norm": 3.2822758663879696,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 24775
+    },
+    {
+      "epoch": 0.5897362658288108,
+      "grad_norm": 12.612970813835473,
+      "learning_rate": 2e-05,
+      "loss": 0.6203,
+      "step": 24776
+    },
+    {
+      "epoch": 0.5897600685518424,
+      "grad_norm": 7.384662865630145,
+      "learning_rate": 2e-05,
+      "loss": 0.2531,
+      "step": 24777
+    },
+    {
+      "epoch": 0.5897838712748739,
+      "grad_norm": 11.437800782716927,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 24778
+    },
+    {
+      "epoch": 0.5898076739979053,
+      "grad_norm": 3.2224226115752526,
+      "learning_rate": 2e-05,
+      "loss": 0.3058,
+      "step": 24779
+    },
+    {
+      "epoch": 0.5898314767209368,
+      "grad_norm": 1.6256043875041557,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 24780
+    },
+    {
+      "epoch": 0.5898552794439684,
+      "grad_norm": 0.8281047300617753,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 24781
+    },
+    {
+      "epoch": 0.5898790821669999,
+      "grad_norm": 6.941712745683554,
+      "learning_rate": 2e-05,
+      "loss": 0.8884,
+      "step": 24782
+    },
+    {
+      "epoch": 0.5899028848900314,
+      "grad_norm": 4.23713072590734,
+      "learning_rate": 2e-05,
+      "loss": 0.7526,
+      "step": 24783
+    },
+    {
+      "epoch": 0.5899266876130629,
+      "grad_norm": 0.176470530524258,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 24784
+    },
+    {
+      "epoch": 0.5899504903360945,
+      "grad_norm": 4.945733179891642,
+      "learning_rate": 2e-05,
+      "loss": 0.1719,
+      "step": 24785
+    },
+    {
+      "epoch": 0.589974293059126,
+      "grad_norm": 1.6032450194548529,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 24786
+    },
+    {
+      "epoch": 0.5899980957821575,
+      "grad_norm": 2.295585064919542,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 24787
+    },
+    {
+      "epoch": 0.590021898505189,
+      "grad_norm": 5.430362287372454,
+      "learning_rate": 2e-05,
+      "loss": 0.9093,
+      "step": 24788
+    },
+    {
+      "epoch": 0.5900457012282205,
+      "grad_norm": 3.63006516314978,
+      "learning_rate": 2e-05,
+      "loss": 0.7688,
+      "step": 24789
+    },
+    {
+      "epoch": 0.590069503951252,
+      "grad_norm": 5.498144991496145,
+      "learning_rate": 2e-05,
+      "loss": 0.4693,
+      "step": 24790
+    },
+    {
+      "epoch": 0.5900933066742835,
+      "grad_norm": 12.113833065357909,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 24791
+    },
+    {
+      "epoch": 0.5901171093973151,
+      "grad_norm": 3.2193385542797084,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 24792
+    },
+    {
+      "epoch": 0.5901409121203466,
+      "grad_norm": 5.480210376474584,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 24793
+    },
+    {
+      "epoch": 0.5901647148433781,
+      "grad_norm": 5.794591936951132,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 24794
+    },
+    {
+      "epoch": 0.5901885175664096,
+      "grad_norm": 1.3971351491184631,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 24795
+    },
+    {
+      "epoch": 0.5902123202894412,
+      "grad_norm": 2.855469871832998,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 24796
+    },
+    {
+      "epoch": 0.5902361230124726,
+      "grad_norm": 0.19859511608380284,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 24797
+    },
+    {
+      "epoch": 0.5902599257355041,
+      "grad_norm": 2.7079008987665363,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 24798
+    },
+    {
+      "epoch": 0.5902837284585356,
+      "grad_norm": 1.2259562443258147,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 24799
+    },
+    {
+      "epoch": 0.5903075311815672,
+      "grad_norm": 4.083656232944231,
+      "learning_rate": 2e-05,
+      "loss": 0.2634,
+      "step": 24800
+    },
+    {
+      "epoch": 0.5903313339045987,
+      "grad_norm": 5.589817680034049,
+      "learning_rate": 2e-05,
+      "loss": 0.5336,
+      "step": 24801
+    },
+    {
+      "epoch": 0.5903551366276302,
+      "grad_norm": 3.6676445561643263,
+      "learning_rate": 2e-05,
+      "loss": 0.4066,
+      "step": 24802
+    },
+    {
+      "epoch": 0.5903789393506618,
+      "grad_norm": 3.729382637582646,
+      "learning_rate": 2e-05,
+      "loss": 0.5019,
+      "step": 24803
+    },
+    {
+      "epoch": 0.5904027420736933,
+      "grad_norm": 2.643378712757461,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 24804
+    },
+    {
+      "epoch": 0.5904265447967247,
+      "grad_norm": 13.147716409582406,
+      "learning_rate": 2e-05,
+      "loss": 1.3206,
+      "step": 24805
+    },
+    {
+      "epoch": 0.5904503475197562,
+      "grad_norm": 9.523984986926857,
+      "learning_rate": 2e-05,
+      "loss": 0.8223,
+      "step": 24806
+    },
+    {
+      "epoch": 0.5904741502427878,
+      "grad_norm": 4.805768830526769,
+      "learning_rate": 2e-05,
+      "loss": 1.41,
+      "step": 24807
+    },
+    {
+      "epoch": 0.5904979529658193,
+      "grad_norm": 1.0765094665357389,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 24808
+    },
+    {
+      "epoch": 0.5905217556888508,
+      "grad_norm": 0.43025085016739983,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 24809
+    },
+    {
+      "epoch": 0.5905455584118823,
+      "grad_norm": 3.640063633720369,
+      "learning_rate": 2e-05,
+      "loss": 0.57,
+      "step": 24810
+    },
+    {
+      "epoch": 0.5905693611349139,
+      "grad_norm": 7.590183719134678,
+      "learning_rate": 2e-05,
+      "loss": 1.5834,
+      "step": 24811
+    },
+    {
+      "epoch": 0.5905931638579454,
+      "grad_norm": 3.482477515749085,
+      "learning_rate": 2e-05,
+      "loss": 0.6146,
+      "step": 24812
+    },
+    {
+      "epoch": 0.5906169665809768,
+      "grad_norm": 7.54351872851844,
+      "learning_rate": 2e-05,
+      "loss": 0.3172,
+      "step": 24813
+    },
+    {
+      "epoch": 0.5906407693040083,
+      "grad_norm": 6.472618633488857,
+      "learning_rate": 2e-05,
+      "loss": 0.1965,
+      "step": 24814
+    },
+    {
+      "epoch": 0.5906645720270399,
+      "grad_norm": 3.3410982856826856,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 24815
+    },
+    {
+      "epoch": 0.5906883747500714,
+      "grad_norm": 0.3653091599060993,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 24816
+    },
+    {
+      "epoch": 0.5907121774731029,
+      "grad_norm": 5.250547142961259,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 24817
+    },
+    {
+      "epoch": 0.5907359801961345,
+      "grad_norm": 4.789681560607492,
+      "learning_rate": 2e-05,
+      "loss": 0.6062,
+      "step": 24818
+    },
+    {
+      "epoch": 0.590759782919166,
+      "grad_norm": 1.919021011748996,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 24819
+    },
+    {
+      "epoch": 0.5907835856421975,
+      "grad_norm": 5.789128838716567,
+      "learning_rate": 2e-05,
+      "loss": 0.2178,
+      "step": 24820
+    },
+    {
+      "epoch": 0.590807388365229,
+      "grad_norm": 6.062916808747137,
+      "learning_rate": 2e-05,
+      "loss": 1.4322,
+      "step": 24821
+    },
+    {
+      "epoch": 0.5908311910882605,
+      "grad_norm": 3.9881060592621944,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 24822
+    },
+    {
+      "epoch": 0.590854993811292,
+      "grad_norm": 4.177220582105078,
+      "learning_rate": 2e-05,
+      "loss": 0.1212,
+      "step": 24823
+    },
+    {
+      "epoch": 0.5908787965343235,
+      "grad_norm": 1.5590451162740444,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 24824
+    },
+    {
+      "epoch": 0.590902599257355,
+      "grad_norm": 3.683669567340053,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 24825
+    },
+    {
+      "epoch": 0.5909264019803866,
+      "grad_norm": 0.559653345068644,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 24826
+    },
+    {
+      "epoch": 0.5909502047034181,
+      "grad_norm": 0.648503145866037,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 24827
+    },
+    {
+      "epoch": 0.5909740074264496,
+      "grad_norm": 30.00688975178336,
+      "learning_rate": 2e-05,
+      "loss": 0.4154,
+      "step": 24828
+    },
+    {
+      "epoch": 0.590997810149481,
+      "grad_norm": 1.486699018638596,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 24829
+    },
+    {
+      "epoch": 0.5910216128725126,
+      "grad_norm": 4.0417131582583865,
+      "learning_rate": 2e-05,
+      "loss": 0.2759,
+      "step": 24830
+    },
+    {
+      "epoch": 0.5910454155955441,
+      "grad_norm": 11.616476043130811,
+      "learning_rate": 2e-05,
+      "loss": 0.2762,
+      "step": 24831
+    },
+    {
+      "epoch": 0.5910692183185756,
+      "grad_norm": 6.883366057853321,
+      "learning_rate": 2e-05,
+      "loss": 0.8858,
+      "step": 24832
+    },
+    {
+      "epoch": 0.5910930210416072,
+      "grad_norm": 4.295214964018016,
+      "learning_rate": 2e-05,
+      "loss": 0.623,
+      "step": 24833
+    },
+    {
+      "epoch": 0.5911168237646387,
+      "grad_norm": 5.38799182212865,
+      "learning_rate": 2e-05,
+      "loss": 0.6683,
+      "step": 24834
+    },
+    {
+      "epoch": 0.5911406264876702,
+      "grad_norm": 2.3168425106117736,
+      "learning_rate": 2e-05,
+      "loss": 0.3482,
+      "step": 24835
+    },
+    {
+      "epoch": 0.5911644292107017,
+      "grad_norm": 0.4796359724578691,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 24836
+    },
+    {
+      "epoch": 0.5911882319337333,
+      "grad_norm": 3.8211321643443465,
+      "learning_rate": 2e-05,
+      "loss": 0.5897,
+      "step": 24837
+    },
+    {
+      "epoch": 0.5912120346567648,
+      "grad_norm": 0.7812327089064439,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 24838
+    },
+    {
+      "epoch": 0.5912358373797962,
+      "grad_norm": 2.1056424897361046,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 24839
+    },
+    {
+      "epoch": 0.5912596401028277,
+      "grad_norm": 6.96848400834548,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 24840
+    },
+    {
+      "epoch": 0.5912834428258593,
+      "grad_norm": 3.3266297738481128,
+      "learning_rate": 2e-05,
+      "loss": 0.4295,
+      "step": 24841
+    },
+    {
+      "epoch": 0.5913072455488908,
+      "grad_norm": 8.193465468320708,
+      "learning_rate": 2e-05,
+      "loss": 0.83,
+      "step": 24842
+    },
+    {
+      "epoch": 0.5913310482719223,
+      "grad_norm": 1.9703495747101916,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 24843
+    },
+    {
+      "epoch": 0.5913548509949538,
+      "grad_norm": 5.719370137923309,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 24844
+    },
+    {
+      "epoch": 0.5913786537179854,
+      "grad_norm": 10.92109452052904,
+      "learning_rate": 2e-05,
+      "loss": 1.0616,
+      "step": 24845
+    },
+    {
+      "epoch": 0.5914024564410169,
+      "grad_norm": 1.7512280943789689,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 24846
+    },
+    {
+      "epoch": 0.5914262591640483,
+      "grad_norm": 5.270749401485093,
+      "learning_rate": 2e-05,
+      "loss": 0.7176,
+      "step": 24847
+    },
+    {
+      "epoch": 0.5914500618870799,
+      "grad_norm": 3.05018373151736,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 24848
+    },
+    {
+      "epoch": 0.5914738646101114,
+      "grad_norm": 3.212201720226276,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 24849
+    },
+    {
+      "epoch": 0.5914976673331429,
+      "grad_norm": 3.8725461150834954,
+      "learning_rate": 2e-05,
+      "loss": 0.2906,
+      "step": 24850
+    },
+    {
+      "epoch": 0.5915214700561744,
+      "grad_norm": 4.601126920792071,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 24851
+    },
+    {
+      "epoch": 0.591545272779206,
+      "grad_norm": 4.149005836873699,
+      "learning_rate": 2e-05,
+      "loss": 0.4873,
+      "step": 24852
+    },
+    {
+      "epoch": 0.5915690755022375,
+      "grad_norm": 8.746884259260323,
+      "learning_rate": 2e-05,
+      "loss": 0.4242,
+      "step": 24853
+    },
+    {
+      "epoch": 0.591592878225269,
+      "grad_norm": 1.7628603518934784,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 24854
+    },
+    {
+      "epoch": 0.5916166809483004,
+      "grad_norm": 2.124305789202147,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 24855
+    },
+    {
+      "epoch": 0.591640483671332,
+      "grad_norm": 0.21354582591406415,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 24856
+    },
+    {
+      "epoch": 0.5916642863943635,
+      "grad_norm": 6.108824415296087,
+      "learning_rate": 2e-05,
+      "loss": 1.122,
+      "step": 24857
+    },
+    {
+      "epoch": 0.591688089117395,
+      "grad_norm": 0.26452889663029483,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 24858
+    },
+    {
+      "epoch": 0.5917118918404265,
+      "grad_norm": 1.840036975323818,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 24859
+    },
+    {
+      "epoch": 0.5917356945634581,
+      "grad_norm": 3.910351428356484,
+      "learning_rate": 2e-05,
+      "loss": 0.5868,
+      "step": 24860
+    },
+    {
+      "epoch": 0.5917594972864896,
+      "grad_norm": 0.7860119933194833,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 24861
+    },
+    {
+      "epoch": 0.5917833000095211,
+      "grad_norm": 2.469616444170719,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 24862
+    },
+    {
+      "epoch": 0.5918071027325527,
+      "grad_norm": 0.1686018324657104,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 24863
+    },
+    {
+      "epoch": 0.5918309054555841,
+      "grad_norm": 0.9840006624920544,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 24864
+    },
+    {
+      "epoch": 0.5918547081786156,
+      "grad_norm": 4.784397630237411,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 24865
+    },
+    {
+      "epoch": 0.5918785109016471,
+      "grad_norm": 3.385438877360252,
+      "learning_rate": 2e-05,
+      "loss": 0.4144,
+      "step": 24866
+    },
+    {
+      "epoch": 0.5919023136246787,
+      "grad_norm": 1.4880349190964481,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 24867
+    },
+    {
+      "epoch": 0.5919261163477102,
+      "grad_norm": 0.6766673775723804,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 24868
+    },
+    {
+      "epoch": 0.5919499190707417,
+      "grad_norm": 2.5887242273041124,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 24869
+    },
+    {
+      "epoch": 0.5919737217937732,
+      "grad_norm": 5.080992858027415,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 24870
+    },
+    {
+      "epoch": 0.5919975245168048,
+      "grad_norm": 6.166065787535249,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 24871
+    },
+    {
+      "epoch": 0.5920213272398362,
+      "grad_norm": 3.860653372499577,
+      "learning_rate": 2e-05,
+      "loss": 0.452,
+      "step": 24872
+    },
+    {
+      "epoch": 0.5920451299628677,
+      "grad_norm": 3.7520215241321555,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 24873
+    },
+    {
+      "epoch": 0.5920689326858992,
+      "grad_norm": 4.411676618677668,
+      "learning_rate": 2e-05,
+      "loss": 0.8381,
+      "step": 24874
+    },
+    {
+      "epoch": 0.5920927354089308,
+      "grad_norm": 0.2689405066525606,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 24875
+    },
+    {
+      "epoch": 0.5921165381319623,
+      "grad_norm": 14.941541564371903,
+      "learning_rate": 2e-05,
+      "loss": 0.5355,
+      "step": 24876
+    },
+    {
+      "epoch": 0.5921403408549938,
+      "grad_norm": 0.21553737968298434,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 24877
+    },
+    {
+      "epoch": 0.5921641435780254,
+      "grad_norm": 0.577192109552926,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 24878
+    },
+    {
+      "epoch": 0.5921879463010569,
+      "grad_norm": 1.4711111416832359,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 24879
+    },
+    {
+      "epoch": 0.5922117490240884,
+      "grad_norm": 2.4137975750158285,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 24880
+    },
+    {
+      "epoch": 0.5922355517471198,
+      "grad_norm": 0.2554345587248256,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 24881
+    },
+    {
+      "epoch": 0.5922593544701514,
+      "grad_norm": 2.995815774389973,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 24882
+    },
+    {
+      "epoch": 0.5922831571931829,
+      "grad_norm": 3.6998006806953416,
+      "learning_rate": 2e-05,
+      "loss": 0.422,
+      "step": 24883
+    },
+    {
+      "epoch": 0.5923069599162144,
+      "grad_norm": 0.3212186211725342,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 24884
+    },
+    {
+      "epoch": 0.5923307626392459,
+      "grad_norm": 0.41138498170072346,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 24885
+    },
+    {
+      "epoch": 0.5923545653622775,
+      "grad_norm": 0.1840255515181052,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 24886
+    },
+    {
+      "epoch": 0.592378368085309,
+      "grad_norm": 3.715549553036681,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 24887
+    },
+    {
+      "epoch": 0.5924021708083405,
+      "grad_norm": 1.3060291539957463,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 24888
+    },
+    {
+      "epoch": 0.5924259735313719,
+      "grad_norm": 1.4577762696294343,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 24889
+    },
+    {
+      "epoch": 0.5924497762544035,
+      "grad_norm": 0.32453258275454494,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 24890
+    },
+    {
+      "epoch": 0.592473578977435,
+      "grad_norm": 2.294681048909177,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 24891
+    },
+    {
+      "epoch": 0.5924973817004665,
+      "grad_norm": 3.1897474965729606,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 24892
+    },
+    {
+      "epoch": 0.592521184423498,
+      "grad_norm": 0.16916800524292328,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 24893
+    },
+    {
+      "epoch": 0.5925449871465296,
+      "grad_norm": 2.62574849859796,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 24894
+    },
+    {
+      "epoch": 0.5925687898695611,
+      "grad_norm": 3.182680215659436,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 24895
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 2.8062967508000187,
+      "learning_rate": 2e-05,
+      "loss": 0.2925,
+      "step": 24896
+    },
+    {
+      "epoch": 0.5926163953156242,
+      "grad_norm": 5.625415846105703,
+      "learning_rate": 2e-05,
+      "loss": 1.4939,
+      "step": 24897
+    },
+    {
+      "epoch": 0.5926401980386556,
+      "grad_norm": 5.458026873652017,
+      "learning_rate": 2e-05,
+      "loss": 0.7291,
+      "step": 24898
+    },
+    {
+      "epoch": 0.5926640007616871,
+      "grad_norm": 2.6423336027422355,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 24899
+    },
+    {
+      "epoch": 0.5926878034847186,
+      "grad_norm": 0.5434015735338472,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 24900
+    },
+    {
+      "epoch": 0.5927116062077502,
+      "grad_norm": 0.1699686815562835,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 24901
+    },
+    {
+      "epoch": 0.5927354089307817,
+      "grad_norm": 6.918574966841418,
+      "learning_rate": 2e-05,
+      "loss": 0.5815,
+      "step": 24902
+    },
+    {
+      "epoch": 0.5927592116538132,
+      "grad_norm": 1.7874598571537996,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 24903
+    },
+    {
+      "epoch": 0.5927830143768447,
+      "grad_norm": 3.810206869856067,
+      "learning_rate": 2e-05,
+      "loss": 0.6524,
+      "step": 24904
+    },
+    {
+      "epoch": 0.5928068170998763,
+      "grad_norm": 2.48275618956733,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 24905
+    },
+    {
+      "epoch": 0.5928306198229077,
+      "grad_norm": 2.3038889480058753,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 24906
+    },
+    {
+      "epoch": 0.5928544225459392,
+      "grad_norm": 1.7926373734406307,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 24907
+    },
+    {
+      "epoch": 0.5928782252689707,
+      "grad_norm": 0.5330841561418742,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 24908
+    },
+    {
+      "epoch": 0.5929020279920023,
+      "grad_norm": 9.721744280301408,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 24909
+    },
+    {
+      "epoch": 0.5929258307150338,
+      "grad_norm": 2.5981185663567485,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 24910
+    },
+    {
+      "epoch": 0.5929496334380653,
+      "grad_norm": 1.3227454712108773,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 24911
+    },
+    {
+      "epoch": 0.5929734361610969,
+      "grad_norm": 1.6240087786610555,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 24912
+    },
+    {
+      "epoch": 0.5929972388841284,
+      "grad_norm": 4.0897213938428,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 24913
+    },
+    {
+      "epoch": 0.5930210416071598,
+      "grad_norm": 4.714220977432513,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 24914
+    },
+    {
+      "epoch": 0.5930448443301913,
+      "grad_norm": 4.47188051573092,
+      "learning_rate": 2e-05,
+      "loss": 0.5955,
+      "step": 24915
+    },
+    {
+      "epoch": 0.5930686470532229,
+      "grad_norm": 1.05536787472654,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 24916
+    },
+    {
+      "epoch": 0.5930924497762544,
+      "grad_norm": 7.291241197264146,
+      "learning_rate": 2e-05,
+      "loss": 1.7771,
+      "step": 24917
+    },
+    {
+      "epoch": 0.5931162524992859,
+      "grad_norm": 4.237158054716839,
+      "learning_rate": 2e-05,
+      "loss": 0.4703,
+      "step": 24918
+    },
+    {
+      "epoch": 0.5931400552223174,
+      "grad_norm": 5.949926978537658,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 24919
+    },
+    {
+      "epoch": 0.593163857945349,
+      "grad_norm": 0.15759498765271446,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 24920
+    },
+    {
+      "epoch": 0.5931876606683805,
+      "grad_norm": 5.727655401376692,
+      "learning_rate": 2e-05,
+      "loss": 0.1598,
+      "step": 24921
+    },
+    {
+      "epoch": 0.593211463391412,
+      "grad_norm": 2.3665817186430607,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 24922
+    },
+    {
+      "epoch": 0.5932352661144434,
+      "grad_norm": 8.897808822063086,
+      "learning_rate": 2e-05,
+      "loss": 0.2977,
+      "step": 24923
+    },
+    {
+      "epoch": 0.593259068837475,
+      "grad_norm": 0.3551523169289032,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 24924
+    },
+    {
+      "epoch": 0.5932828715605065,
+      "grad_norm": 2.2436487123853928,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 24925
+    },
+    {
+      "epoch": 0.593306674283538,
+      "grad_norm": 6.532766774105083,
+      "learning_rate": 2e-05,
+      "loss": 0.2003,
+      "step": 24926
+    },
+    {
+      "epoch": 0.5933304770065696,
+      "grad_norm": 3.1724987988697637,
+      "learning_rate": 2e-05,
+      "loss": 0.1194,
+      "step": 24927
+    },
+    {
+      "epoch": 0.5933542797296011,
+      "grad_norm": 4.759136797927663,
+      "learning_rate": 2e-05,
+      "loss": 0.5806,
+      "step": 24928
+    },
+    {
+      "epoch": 0.5933780824526326,
+      "grad_norm": 8.74950613227086,
+      "learning_rate": 2e-05,
+      "loss": 1.1113,
+      "step": 24929
+    },
+    {
+      "epoch": 0.5934018851756641,
+      "grad_norm": 13.027602600409631,
+      "learning_rate": 2e-05,
+      "loss": 0.8447,
+      "step": 24930
+    },
+    {
+      "epoch": 0.5934256878986957,
+      "grad_norm": 5.373365453390267,
+      "learning_rate": 2e-05,
+      "loss": 0.6227,
+      "step": 24931
+    },
+    {
+      "epoch": 0.5934494906217271,
+      "grad_norm": 10.099544924865969,
+      "learning_rate": 2e-05,
+      "loss": 0.2886,
+      "step": 24932
+    },
+    {
+      "epoch": 0.5934732933447586,
+      "grad_norm": 0.011436006450619545,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 24933
+    },
+    {
+      "epoch": 0.5934970960677901,
+      "grad_norm": 3.1941793917639605,
+      "learning_rate": 2e-05,
+      "loss": 0.525,
+      "step": 24934
+    },
+    {
+      "epoch": 0.5935208987908217,
+      "grad_norm": 19.468572031457764,
+      "learning_rate": 2e-05,
+      "loss": 0.3104,
+      "step": 24935
+    },
+    {
+      "epoch": 0.5935447015138532,
+      "grad_norm": 9.354179438426712,
+      "learning_rate": 2e-05,
+      "loss": 1.7845,
+      "step": 24936
+    },
+    {
+      "epoch": 0.5935685042368847,
+      "grad_norm": 6.6078992467856805,
+      "learning_rate": 2e-05,
+      "loss": 1.2961,
+      "step": 24937
+    },
+    {
+      "epoch": 0.5935923069599162,
+      "grad_norm": 2.9639316013143056,
+      "learning_rate": 2e-05,
+      "loss": 0.3701,
+      "step": 24938
+    },
+    {
+      "epoch": 0.5936161096829478,
+      "grad_norm": 5.07486771188625,
+      "learning_rate": 2e-05,
+      "loss": 0.5077,
+      "step": 24939
+    },
+    {
+      "epoch": 0.5936399124059792,
+      "grad_norm": 4.498032748366772,
+      "learning_rate": 2e-05,
+      "loss": 0.4911,
+      "step": 24940
+    },
+    {
+      "epoch": 0.5936637151290107,
+      "grad_norm": 0.27910526206852393,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 24941
+    },
+    {
+      "epoch": 0.5936875178520423,
+      "grad_norm": 3.4728985935245675,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 24942
+    },
+    {
+      "epoch": 0.5937113205750738,
+      "grad_norm": 1.255806965284665,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 24943
+    },
+    {
+      "epoch": 0.5937351232981053,
+      "grad_norm": 3.675820542177693,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 24944
+    },
+    {
+      "epoch": 0.5937589260211368,
+      "grad_norm": 4.727365674193828,
+      "learning_rate": 2e-05,
+      "loss": 0.6582,
+      "step": 24945
+    },
+    {
+      "epoch": 0.5937827287441684,
+      "grad_norm": 5.359361023235015,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 24946
+    },
+    {
+      "epoch": 0.5938065314671999,
+      "grad_norm": 3.3528395042771644,
+      "learning_rate": 2e-05,
+      "loss": 0.2414,
+      "step": 24947
+    },
+    {
+      "epoch": 0.5938303341902313,
+      "grad_norm": 8.926574864455436,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 24948
+    },
+    {
+      "epoch": 0.5938541369132628,
+      "grad_norm": 10.095773099918054,
+      "learning_rate": 2e-05,
+      "loss": 0.328,
+      "step": 24949
+    },
+    {
+      "epoch": 0.5938779396362944,
+      "grad_norm": 0.22503715308186328,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 24950
+    },
+    {
+      "epoch": 0.5939017423593259,
+      "grad_norm": 2.322909428493751,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 24951
+    },
+    {
+      "epoch": 0.5939255450823574,
+      "grad_norm": 3.644730632457269,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 24952
+    },
+    {
+      "epoch": 0.5939493478053889,
+      "grad_norm": 6.396488523359271,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 24953
+    },
+    {
+      "epoch": 0.5939731505284205,
+      "grad_norm": 2.589258245933825,
+      "learning_rate": 2e-05,
+      "loss": 0.3726,
+      "step": 24954
+    },
+    {
+      "epoch": 0.593996953251452,
+      "grad_norm": 3.4151589251792527,
+      "learning_rate": 2e-05,
+      "loss": 0.2787,
+      "step": 24955
+    },
+    {
+      "epoch": 0.5940207559744834,
+      "grad_norm": 3.168424367965,
+      "learning_rate": 2e-05,
+      "loss": 0.3464,
+      "step": 24956
+    },
+    {
+      "epoch": 0.594044558697515,
+      "grad_norm": 3.13516448107002,
+      "learning_rate": 2e-05,
+      "loss": 0.1926,
+      "step": 24957
+    },
+    {
+      "epoch": 0.5940683614205465,
+      "grad_norm": 5.629602349372463,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 24958
+    },
+    {
+      "epoch": 0.594092164143578,
+      "grad_norm": 6.307731725552011,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 24959
+    },
+    {
+      "epoch": 0.5941159668666095,
+      "grad_norm": 5.660691141924647,
+      "learning_rate": 2e-05,
+      "loss": 1.0951,
+      "step": 24960
+    },
+    {
+      "epoch": 0.5941397695896411,
+      "grad_norm": 2.5563002726192106,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 24961
+    },
+    {
+      "epoch": 0.5941635723126726,
+      "grad_norm": 1.3344185689340544,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 24962
+    },
+    {
+      "epoch": 0.5941873750357041,
+      "grad_norm": 2.840376120648084,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 24963
+    },
+    {
+      "epoch": 0.5942111777587356,
+      "grad_norm": 3.3954633935169536,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 24964
+    },
+    {
+      "epoch": 0.5942349804817671,
+      "grad_norm": 2.298738019179549,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 24965
+    },
+    {
+      "epoch": 0.5942587832047986,
+      "grad_norm": 2.7557832476780653,
+      "learning_rate": 2e-05,
+      "loss": 0.3215,
+      "step": 24966
+    },
+    {
+      "epoch": 0.5942825859278301,
+      "grad_norm": 10.15760850839251,
+      "learning_rate": 2e-05,
+      "loss": 0.634,
+      "step": 24967
+    },
+    {
+      "epoch": 0.5943063886508616,
+      "grad_norm": 4.86677776226929,
+      "learning_rate": 2e-05,
+      "loss": 0.6679,
+      "step": 24968
+    },
+    {
+      "epoch": 0.5943301913738932,
+      "grad_norm": 5.245270389880481,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 24969
+    },
+    {
+      "epoch": 0.5943539940969247,
+      "grad_norm": 3.4745182592342427,
+      "learning_rate": 2e-05,
+      "loss": 0.1528,
+      "step": 24970
+    },
+    {
+      "epoch": 0.5943777968199562,
+      "grad_norm": 3.0895492044089727,
+      "learning_rate": 2e-05,
+      "loss": 0.4675,
+      "step": 24971
+    },
+    {
+      "epoch": 0.5944015995429878,
+      "grad_norm": 4.260187630241845,
+      "learning_rate": 2e-05,
+      "loss": 0.8291,
+      "step": 24972
+    },
+    {
+      "epoch": 0.5944254022660193,
+      "grad_norm": 7.197480637782391,
+      "learning_rate": 2e-05,
+      "loss": 1.0389,
+      "step": 24973
+    },
+    {
+      "epoch": 0.5944492049890507,
+      "grad_norm": 0.918100912684654,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 24974
+    },
+    {
+      "epoch": 0.5944730077120822,
+      "grad_norm": 3.639801492558577,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 24975
+    },
+    {
+      "epoch": 0.5944968104351138,
+      "grad_norm": 2.9910523536036306,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 24976
+    },
+    {
+      "epoch": 0.5945206131581453,
+      "grad_norm": 4.235620139895125,
+      "learning_rate": 2e-05,
+      "loss": 0.674,
+      "step": 24977
+    },
+    {
+      "epoch": 0.5945444158811768,
+      "grad_norm": 6.804859872871307,
+      "learning_rate": 2e-05,
+      "loss": 1.2058,
+      "step": 24978
+    },
+    {
+      "epoch": 0.5945682186042083,
+      "grad_norm": 2.9647573851023745,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 24979
+    },
+    {
+      "epoch": 0.5945920213272399,
+      "grad_norm": 4.327658723801095,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 24980
+    },
+    {
+      "epoch": 0.5946158240502714,
+      "grad_norm": 12.535689823355947,
+      "learning_rate": 2e-05,
+      "loss": 0.5516,
+      "step": 24981
+    },
+    {
+      "epoch": 0.5946396267733028,
+      "grad_norm": 1.434440399395225,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 24982
+    },
+    {
+      "epoch": 0.5946634294963343,
+      "grad_norm": 3.8013562901249376,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 24983
+    },
+    {
+      "epoch": 0.5946872322193659,
+      "grad_norm": 1.8251989965325437,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 24984
+    },
+    {
+      "epoch": 0.5947110349423974,
+      "grad_norm": 1.6462156124886311,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 24985
+    },
+    {
+      "epoch": 0.5947348376654289,
+      "grad_norm": 4.617343265765482,
+      "learning_rate": 2e-05,
+      "loss": 0.7214,
+      "step": 24986
+    },
+    {
+      "epoch": 0.5947586403884605,
+      "grad_norm": 3.8134479230243783,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 24987
+    },
+    {
+      "epoch": 0.594782443111492,
+      "grad_norm": 0.6763748184464946,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 24988
+    },
+    {
+      "epoch": 0.5948062458345235,
+      "grad_norm": 4.9555469886247705,
+      "learning_rate": 2e-05,
+      "loss": 0.8473,
+      "step": 24989
+    },
+    {
+      "epoch": 0.594830048557555,
+      "grad_norm": 3.0643551752460443,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 24990
+    },
+    {
+      "epoch": 0.5948538512805865,
+      "grad_norm": 10.11585096033115,
+      "learning_rate": 2e-05,
+      "loss": 0.6301,
+      "step": 24991
+    },
+    {
+      "epoch": 0.594877654003618,
+      "grad_norm": 2.877538765412247,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 24992
+    },
+    {
+      "epoch": 0.5949014567266495,
+      "grad_norm": 1.1715966564772,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 24993
+    },
+    {
+      "epoch": 0.594925259449681,
+      "grad_norm": 3.235112895200311,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 24994
+    },
+    {
+      "epoch": 0.5949490621727126,
+      "grad_norm": 4.72545953547504,
+      "learning_rate": 2e-05,
+      "loss": 0.9374,
+      "step": 24995
+    },
+    {
+      "epoch": 0.5949728648957441,
+      "grad_norm": 1.3278751841983523,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 24996
+    },
+    {
+      "epoch": 0.5949966676187756,
+      "grad_norm": 3.826277600294871,
+      "learning_rate": 2e-05,
+      "loss": 0.5823,
+      "step": 24997
+    },
+    {
+      "epoch": 0.595020470341807,
+      "grad_norm": 5.54018557884993,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 24998
+    },
+    {
+      "epoch": 0.5950442730648386,
+      "grad_norm": 5.812403080737135,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 24999
+    },
+    {
+      "epoch": 0.5950680757878701,
+      "grad_norm": 2.891190388156234,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 25000
+    },
+    {
+      "epoch": 0.5950918785109016,
+      "grad_norm": 0.7589584062445999,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 25001
+    },
+    {
+      "epoch": 0.5951156812339332,
+      "grad_norm": 3.1937433969804903,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 25002
+    },
+    {
+      "epoch": 0.5951394839569647,
+      "grad_norm": 0.10626784705862767,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 25003
+    },
+    {
+      "epoch": 0.5951632866799962,
+      "grad_norm": 3.753646003474193,
+      "learning_rate": 2e-05,
+      "loss": 0.579,
+      "step": 25004
+    },
+    {
+      "epoch": 0.5951870894030277,
+      "grad_norm": 14.087409314043322,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 25005
+    },
+    {
+      "epoch": 0.5952108921260593,
+      "grad_norm": 1.3601187899579725,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 25006
+    },
+    {
+      "epoch": 0.5952346948490908,
+      "grad_norm": 8.218087749358194,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 25007
+    },
+    {
+      "epoch": 0.5952584975721222,
+      "grad_norm": 15.754343385627381,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 25008
+    },
+    {
+      "epoch": 0.5952823002951537,
+      "grad_norm": 3.470806842179444,
+      "learning_rate": 2e-05,
+      "loss": 0.473,
+      "step": 25009
+    },
+    {
+      "epoch": 0.5953061030181853,
+      "grad_norm": 3.7058256950093225,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 25010
+    },
+    {
+      "epoch": 0.5953299057412168,
+      "grad_norm": 9.386476016692429,
+      "learning_rate": 2e-05,
+      "loss": 1.0442,
+      "step": 25011
+    },
+    {
+      "epoch": 0.5953537084642483,
+      "grad_norm": 3.3943430953902456,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 25012
+    },
+    {
+      "epoch": 0.5953775111872798,
+      "grad_norm": 5.841001884314594,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 25013
+    },
+    {
+      "epoch": 0.5954013139103114,
+      "grad_norm": 1.9676333438022862,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 25014
+    },
+    {
+      "epoch": 0.5954251166333429,
+      "grad_norm": 0.11866382304448951,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 25015
+    },
+    {
+      "epoch": 0.5954489193563743,
+      "grad_norm": 2.881837871923179,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 25016
+    },
+    {
+      "epoch": 0.5954727220794059,
+      "grad_norm": 5.483080060102995,
+      "learning_rate": 2e-05,
+      "loss": 0.1566,
+      "step": 25017
+    },
+    {
+      "epoch": 0.5954965248024374,
+      "grad_norm": 9.00317206539869,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 25018
+    },
+    {
+      "epoch": 0.5955203275254689,
+      "grad_norm": 6.402210969958146,
+      "learning_rate": 2e-05,
+      "loss": 0.9102,
+      "step": 25019
+    },
+    {
+      "epoch": 0.5955441302485004,
+      "grad_norm": 2.0448102686267586,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 25020
+    },
+    {
+      "epoch": 0.595567932971532,
+      "grad_norm": 0.5874388875417618,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 25021
+    },
+    {
+      "epoch": 0.5955917356945635,
+      "grad_norm": 2.864898874745414,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 25022
+    },
+    {
+      "epoch": 0.595615538417595,
+      "grad_norm": 10.322014296600909,
+      "learning_rate": 2e-05,
+      "loss": 0.754,
+      "step": 25023
+    },
+    {
+      "epoch": 0.5956393411406264,
+      "grad_norm": 0.5843920292058049,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 25024
+    },
+    {
+      "epoch": 0.595663143863658,
+      "grad_norm": 4.902259088970759,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 25025
+    },
+    {
+      "epoch": 0.5956869465866895,
+      "grad_norm": 2.7868894430316593,
+      "learning_rate": 2e-05,
+      "loss": 0.3167,
+      "step": 25026
+    },
+    {
+      "epoch": 0.595710749309721,
+      "grad_norm": 11.31108117940124,
+      "learning_rate": 2e-05,
+      "loss": 0.8611,
+      "step": 25027
+    },
+    {
+      "epoch": 0.5957345520327525,
+      "grad_norm": 2.4831684455787375,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 25028
+    },
+    {
+      "epoch": 0.5957583547557841,
+      "grad_norm": 0.6200742766312214,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 25029
+    },
+    {
+      "epoch": 0.5957821574788156,
+      "grad_norm": 6.393411860862047,
+      "learning_rate": 2e-05,
+      "loss": 0.7692,
+      "step": 25030
+    },
+    {
+      "epoch": 0.5958059602018471,
+      "grad_norm": 1.7026335626793387,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 25031
+    },
+    {
+      "epoch": 0.5958297629248787,
+      "grad_norm": 5.877134575257255,
+      "learning_rate": 2e-05,
+      "loss": 0.4363,
+      "step": 25032
+    },
+    {
+      "epoch": 0.5958535656479101,
+      "grad_norm": 7.617046822197021,
+      "learning_rate": 2e-05,
+      "loss": 0.3848,
+      "step": 25033
+    },
+    {
+      "epoch": 0.5958773683709416,
+      "grad_norm": 1.786792083380449,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 25034
+    },
+    {
+      "epoch": 0.5959011710939731,
+      "grad_norm": 9.190774835384428,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 25035
+    },
+    {
+      "epoch": 0.5959249738170047,
+      "grad_norm": 4.961747232472004,
+      "learning_rate": 2e-05,
+      "loss": 0.848,
+      "step": 25036
+    },
+    {
+      "epoch": 0.5959487765400362,
+      "grad_norm": 2.7766920465664064,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 25037
+    },
+    {
+      "epoch": 0.5959725792630677,
+      "grad_norm": 3.730479852841462,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 25038
+    },
+    {
+      "epoch": 0.5959963819860992,
+      "grad_norm": 3.291458399391335,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 25039
+    },
+    {
+      "epoch": 0.5960201847091308,
+      "grad_norm": 0.9824928774915637,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 25040
+    },
+    {
+      "epoch": 0.5960439874321622,
+      "grad_norm": 4.22080907344044,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 25041
+    },
+    {
+      "epoch": 0.5960677901551937,
+      "grad_norm": 1.1899358132442748,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 25042
+    },
+    {
+      "epoch": 0.5960915928782252,
+      "grad_norm": 4.15003288206486,
+      "learning_rate": 2e-05,
+      "loss": 0.5124,
+      "step": 25043
+    },
+    {
+      "epoch": 0.5961153956012568,
+      "grad_norm": 5.466362363001525,
+      "learning_rate": 2e-05,
+      "loss": 0.6179,
+      "step": 25044
+    },
+    {
+      "epoch": 0.5961391983242883,
+      "grad_norm": 8.911320874858923,
+      "learning_rate": 2e-05,
+      "loss": 0.939,
+      "step": 25045
+    },
+    {
+      "epoch": 0.5961630010473198,
+      "grad_norm": 4.066991831289418,
+      "learning_rate": 2e-05,
+      "loss": 0.7437,
+      "step": 25046
+    },
+    {
+      "epoch": 0.5961868037703514,
+      "grad_norm": 3.9092262940965763,
+      "learning_rate": 2e-05,
+      "loss": 0.7419,
+      "step": 25047
+    },
+    {
+      "epoch": 0.5962106064933829,
+      "grad_norm": 4.518699976051538,
+      "learning_rate": 2e-05,
+      "loss": 0.5729,
+      "step": 25048
+    },
+    {
+      "epoch": 0.5962344092164144,
+      "grad_norm": 105.47943830676185,
+      "learning_rate": 2e-05,
+      "loss": 0.9578,
+      "step": 25049
+    },
+    {
+      "epoch": 0.5962582119394458,
+      "grad_norm": 3.6486588250034755,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 25050
+    },
+    {
+      "epoch": 0.5962820146624774,
+      "grad_norm": 6.799304606265724,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 25051
+    },
+    {
+      "epoch": 0.5963058173855089,
+      "grad_norm": 7.082606663299622,
+      "learning_rate": 2e-05,
+      "loss": 0.8067,
+      "step": 25052
+    },
+    {
+      "epoch": 0.5963296201085404,
+      "grad_norm": 3.7979035444790097,
+      "learning_rate": 2e-05,
+      "loss": 0.6652,
+      "step": 25053
+    },
+    {
+      "epoch": 0.5963534228315719,
+      "grad_norm": 5.045600781230008,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 25054
+    },
+    {
+      "epoch": 0.5963772255546035,
+      "grad_norm": 4.690449328216861,
+      "learning_rate": 2e-05,
+      "loss": 0.3665,
+      "step": 25055
+    },
+    {
+      "epoch": 0.596401028277635,
+      "grad_norm": 7.73234749002998,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 25056
+    },
+    {
+      "epoch": 0.5964248310006665,
+      "grad_norm": 0.2784949886285792,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 25057
+    },
+    {
+      "epoch": 0.5964486337236979,
+      "grad_norm": 9.529002093496006,
+      "learning_rate": 2e-05,
+      "loss": 0.5179,
+      "step": 25058
+    },
+    {
+      "epoch": 0.5964724364467295,
+      "grad_norm": 2.344294352899395,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 25059
+    },
+    {
+      "epoch": 0.596496239169761,
+      "grad_norm": 18.345688446792483,
+      "learning_rate": 2e-05,
+      "loss": 1.0338,
+      "step": 25060
+    },
+    {
+      "epoch": 0.5965200418927925,
+      "grad_norm": 0.9282749194956363,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 25061
+    },
+    {
+      "epoch": 0.5965438446158241,
+      "grad_norm": 4.753654644941424,
+      "learning_rate": 2e-05,
+      "loss": 0.8568,
+      "step": 25062
+    },
+    {
+      "epoch": 0.5965676473388556,
+      "grad_norm": 2.8427110704092167,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 25063
+    },
+    {
+      "epoch": 0.5965914500618871,
+      "grad_norm": 4.184400661946967,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 25064
+    },
+    {
+      "epoch": 0.5966152527849186,
+      "grad_norm": 29.77373371836774,
+      "learning_rate": 2e-05,
+      "loss": 1.2712,
+      "step": 25065
+    },
+    {
+      "epoch": 0.5966390555079502,
+      "grad_norm": 6.960183633041376,
+      "learning_rate": 2e-05,
+      "loss": 0.9627,
+      "step": 25066
+    },
+    {
+      "epoch": 0.5966628582309816,
+      "grad_norm": 7.536153037887215,
+      "learning_rate": 2e-05,
+      "loss": 0.6097,
+      "step": 25067
+    },
+    {
+      "epoch": 0.5966866609540131,
+      "grad_norm": 2.586615565675688,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 25068
+    },
+    {
+      "epoch": 0.5967104636770446,
+      "grad_norm": 18.620793914039748,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 25069
+    },
+    {
+      "epoch": 0.5967342664000762,
+      "grad_norm": 5.305055599290965,
+      "learning_rate": 2e-05,
+      "loss": 0.7817,
+      "step": 25070
+    },
+    {
+      "epoch": 0.5967580691231077,
+      "grad_norm": 4.9665204932684786,
+      "learning_rate": 2e-05,
+      "loss": 0.2,
+      "step": 25071
+    },
+    {
+      "epoch": 0.5967818718461392,
+      "grad_norm": 2.73936028294177,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 25072
+    },
+    {
+      "epoch": 0.5968056745691707,
+      "grad_norm": 4.335029624326565,
+      "learning_rate": 2e-05,
+      "loss": 1.0025,
+      "step": 25073
+    },
+    {
+      "epoch": 0.5968294772922023,
+      "grad_norm": 5.113140539057639,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 25074
+    },
+    {
+      "epoch": 0.5968532800152337,
+      "grad_norm": 0.31346560931107625,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 25075
+    },
+    {
+      "epoch": 0.5968770827382652,
+      "grad_norm": 6.012012825242318,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 25076
+    },
+    {
+      "epoch": 0.5969008854612968,
+      "grad_norm": 5.72785919939222,
+      "learning_rate": 2e-05,
+      "loss": 1.0147,
+      "step": 25077
+    },
+    {
+      "epoch": 0.5969246881843283,
+      "grad_norm": 2.538700938344379,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 25078
+    },
+    {
+      "epoch": 0.5969484909073598,
+      "grad_norm": 0.3186001397148045,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 25079
+    },
+    {
+      "epoch": 0.5969722936303913,
+      "grad_norm": 5.460321328228489,
+      "learning_rate": 2e-05,
+      "loss": 0.7609,
+      "step": 25080
+    },
+    {
+      "epoch": 0.5969960963534229,
+      "grad_norm": 4.0278970958669955,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 25081
+    },
+    {
+      "epoch": 0.5970198990764544,
+      "grad_norm": 0.35932918645446604,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 25082
+    },
+    {
+      "epoch": 0.5970437017994858,
+      "grad_norm": 4.776497255292836,
+      "learning_rate": 2e-05,
+      "loss": 0.7603,
+      "step": 25083
+    },
+    {
+      "epoch": 0.5970675045225173,
+      "grad_norm": 1.8337245913037543,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 25084
+    },
+    {
+      "epoch": 0.5970913072455489,
+      "grad_norm": 8.51214607296249,
+      "learning_rate": 2e-05,
+      "loss": 0.8064,
+      "step": 25085
+    },
+    {
+      "epoch": 0.5971151099685804,
+      "grad_norm": 1.2551042447896252,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 25086
+    },
+    {
+      "epoch": 0.5971389126916119,
+      "grad_norm": 3.5133797619887903,
+      "learning_rate": 2e-05,
+      "loss": 0.3216,
+      "step": 25087
+    },
+    {
+      "epoch": 0.5971627154146434,
+      "grad_norm": 0.14096970741818096,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 25088
+    },
+    {
+      "epoch": 0.597186518137675,
+      "grad_norm": 0.5837544732569232,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 25089
+    },
+    {
+      "epoch": 0.5972103208607065,
+      "grad_norm": 0.828210871015518,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 25090
+    },
+    {
+      "epoch": 0.597234123583738,
+      "grad_norm": 2.887388757877111,
+      "learning_rate": 2e-05,
+      "loss": 0.53,
+      "step": 25091
+    },
+    {
+      "epoch": 0.5972579263067695,
+      "grad_norm": 3.571408939091419,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 25092
+    },
+    {
+      "epoch": 0.597281729029801,
+      "grad_norm": 7.764245157855251,
+      "learning_rate": 2e-05,
+      "loss": 0.1492,
+      "step": 25093
+    },
+    {
+      "epoch": 0.5973055317528325,
+      "grad_norm": 0.559170556970674,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 25094
+    },
+    {
+      "epoch": 0.597329334475864,
+      "grad_norm": 5.733774914916336,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 25095
+    },
+    {
+      "epoch": 0.5973531371988956,
+      "grad_norm": 8.816092302813656,
+      "learning_rate": 2e-05,
+      "loss": 1.5996,
+      "step": 25096
+    },
+    {
+      "epoch": 0.5973769399219271,
+      "grad_norm": 5.080076662694892,
+      "learning_rate": 2e-05,
+      "loss": 0.8215,
+      "step": 25097
+    },
+    {
+      "epoch": 0.5974007426449586,
+      "grad_norm": 1.2731187153933978,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 25098
+    },
+    {
+      "epoch": 0.5974245453679901,
+      "grad_norm": 0.3911840231919746,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 25099
+    },
+    {
+      "epoch": 0.5974483480910217,
+      "grad_norm": 0.08517023588568522,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 25100
+    },
+    {
+      "epoch": 0.5974721508140531,
+      "grad_norm": 0.8256683925690085,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 25101
+    },
+    {
+      "epoch": 0.5974959535370846,
+      "grad_norm": 8.677816291113096,
+      "learning_rate": 2e-05,
+      "loss": 1.0135,
+      "step": 25102
+    },
+    {
+      "epoch": 0.5975197562601161,
+      "grad_norm": 1.0504253326979989,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 25103
+    },
+    {
+      "epoch": 0.5975435589831477,
+      "grad_norm": 4.489236218048831,
+      "learning_rate": 2e-05,
+      "loss": 1.0672,
+      "step": 25104
+    },
+    {
+      "epoch": 0.5975673617061792,
+      "grad_norm": 6.5709434509991,
+      "learning_rate": 2e-05,
+      "loss": 1.0683,
+      "step": 25105
+    },
+    {
+      "epoch": 0.5975911644292107,
+      "grad_norm": 3.5014774950453122,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 25106
+    },
+    {
+      "epoch": 0.5976149671522423,
+      "grad_norm": 6.300775165891865,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 25107
+    },
+    {
+      "epoch": 0.5976387698752738,
+      "grad_norm": 6.130295093628522,
+      "learning_rate": 2e-05,
+      "loss": 0.5528,
+      "step": 25108
+    },
+    {
+      "epoch": 0.5976625725983052,
+      "grad_norm": 25.446627579023623,
+      "learning_rate": 2e-05,
+      "loss": 0.7497,
+      "step": 25109
+    },
+    {
+      "epoch": 0.5976863753213367,
+      "grad_norm": 8.558503137850133,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 25110
+    },
+    {
+      "epoch": 0.5977101780443683,
+      "grad_norm": 6.303234088561841,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 25111
+    },
+    {
+      "epoch": 0.5977339807673998,
+      "grad_norm": 0.04229797934317799,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 25112
+    },
+    {
+      "epoch": 0.5977577834904313,
+      "grad_norm": 5.481441145891635,
+      "learning_rate": 2e-05,
+      "loss": 0.7314,
+      "step": 25113
+    },
+    {
+      "epoch": 0.5977815862134628,
+      "grad_norm": 0.5401044017025319,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 25114
+    },
+    {
+      "epoch": 0.5978053889364944,
+      "grad_norm": 5.396978084940304,
+      "learning_rate": 2e-05,
+      "loss": 0.5641,
+      "step": 25115
+    },
+    {
+      "epoch": 0.5978291916595259,
+      "grad_norm": 3.9394928583534647,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 25116
+    },
+    {
+      "epoch": 0.5978529943825573,
+      "grad_norm": 6.6822270643814665,
+      "learning_rate": 2e-05,
+      "loss": 0.524,
+      "step": 25117
+    },
+    {
+      "epoch": 0.5978767971055888,
+      "grad_norm": 3.5433168399049104,
+      "learning_rate": 2e-05,
+      "loss": 0.6173,
+      "step": 25118
+    },
+    {
+      "epoch": 0.5979005998286204,
+      "grad_norm": 2.3112540982747336,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 25119
+    },
+    {
+      "epoch": 0.5979244025516519,
+      "grad_norm": 15.040007152080923,
+      "learning_rate": 2e-05,
+      "loss": 1.0735,
+      "step": 25120
+    },
+    {
+      "epoch": 0.5979482052746834,
+      "grad_norm": 0.34116560372676724,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 25121
+    },
+    {
+      "epoch": 0.597972007997715,
+      "grad_norm": 3.669733281064269,
+      "learning_rate": 2e-05,
+      "loss": 0.0951,
+      "step": 25122
+    },
+    {
+      "epoch": 0.5979958107207465,
+      "grad_norm": 4.538881203209865,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 25123
+    },
+    {
+      "epoch": 0.598019613443778,
+      "grad_norm": 2.239135574533712,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 25124
+    },
+    {
+      "epoch": 0.5980434161668094,
+      "grad_norm": 0.24480150917719004,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 25125
+    },
+    {
+      "epoch": 0.598067218889841,
+      "grad_norm": 4.411603412454688,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 25126
+    },
+    {
+      "epoch": 0.5980910216128725,
+      "grad_norm": 2.875638022007935,
+      "learning_rate": 2e-05,
+      "loss": 0.2212,
+      "step": 25127
+    },
+    {
+      "epoch": 0.598114824335904,
+      "grad_norm": 0.5874671507835955,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 25128
+    },
+    {
+      "epoch": 0.5981386270589355,
+      "grad_norm": 5.462662647517258,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 25129
+    },
+    {
+      "epoch": 0.5981624297819671,
+      "grad_norm": 2.4487373202301397,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 25130
+    },
+    {
+      "epoch": 0.5981862325049986,
+      "grad_norm": 2.7238955879023585,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 25131
+    },
+    {
+      "epoch": 0.5982100352280301,
+      "grad_norm": 2.485475997948369,
+      "learning_rate": 2e-05,
+      "loss": 0.2485,
+      "step": 25132
+    },
+    {
+      "epoch": 0.5982338379510616,
+      "grad_norm": 4.202646283770302,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 25133
+    },
+    {
+      "epoch": 0.5982576406740931,
+      "grad_norm": 8.820315516860203,
+      "learning_rate": 2e-05,
+      "loss": 1.1867,
+      "step": 25134
+    },
+    {
+      "epoch": 0.5982814433971246,
+      "grad_norm": 10.135253300649172,
+      "learning_rate": 2e-05,
+      "loss": 0.5209,
+      "step": 25135
+    },
+    {
+      "epoch": 0.5983052461201561,
+      "grad_norm": 5.574282251222302,
+      "learning_rate": 2e-05,
+      "loss": 0.7787,
+      "step": 25136
+    },
+    {
+      "epoch": 0.5983290488431876,
+      "grad_norm": 0.47129784188580165,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 25137
+    },
+    {
+      "epoch": 0.5983528515662192,
+      "grad_norm": 2.468406216457024,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 25138
+    },
+    {
+      "epoch": 0.5983766542892507,
+      "grad_norm": 3.7297129975051666,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 25139
+    },
+    {
+      "epoch": 0.5984004570122822,
+      "grad_norm": 6.001019034129829,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 25140
+    },
+    {
+      "epoch": 0.5984242597353138,
+      "grad_norm": 0.15724863724205898,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 25141
+    },
+    {
+      "epoch": 0.5984480624583453,
+      "grad_norm": 5.501692128719405,
+      "learning_rate": 2e-05,
+      "loss": 0.2792,
+      "step": 25142
+    },
+    {
+      "epoch": 0.5984718651813767,
+      "grad_norm": 3.122021905118327,
+      "learning_rate": 2e-05,
+      "loss": 0.373,
+      "step": 25143
+    },
+    {
+      "epoch": 0.5984956679044082,
+      "grad_norm": 11.777833727616201,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 25144
+    },
+    {
+      "epoch": 0.5985194706274398,
+      "grad_norm": 2.4121218365924526,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 25145
+    },
+    {
+      "epoch": 0.5985432733504713,
+      "grad_norm": 3.537775520196091,
+      "learning_rate": 2e-05,
+      "loss": 0.5828,
+      "step": 25146
+    },
+    {
+      "epoch": 0.5985670760735028,
+      "grad_norm": 7.699206379764931,
+      "learning_rate": 2e-05,
+      "loss": 0.9351,
+      "step": 25147
+    },
+    {
+      "epoch": 0.5985908787965343,
+      "grad_norm": 5.133001008535065,
+      "learning_rate": 2e-05,
+      "loss": 1.1719,
+      "step": 25148
+    },
+    {
+      "epoch": 0.5986146815195659,
+      "grad_norm": 3.0208020356902323,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 25149
+    },
+    {
+      "epoch": 0.5986384842425974,
+      "grad_norm": 0.577625398454665,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 25150
+    },
+    {
+      "epoch": 0.5986622869656288,
+      "grad_norm": 7.641828567023865,
+      "learning_rate": 2e-05,
+      "loss": 1.7442,
+      "step": 25151
+    },
+    {
+      "epoch": 0.5986860896886603,
+      "grad_norm": 0.5496736980872347,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 25152
+    },
+    {
+      "epoch": 0.5987098924116919,
+      "grad_norm": 8.196006788373031,
+      "learning_rate": 2e-05,
+      "loss": 0.8381,
+      "step": 25153
+    },
+    {
+      "epoch": 0.5987336951347234,
+      "grad_norm": 1.214032350020886,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 25154
+    },
+    {
+      "epoch": 0.5987574978577549,
+      "grad_norm": 6.035738304108162,
+      "learning_rate": 2e-05,
+      "loss": 0.9431,
+      "step": 25155
+    },
+    {
+      "epoch": 0.5987813005807865,
+      "grad_norm": 0.4528715074383555,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 25156
+    },
+    {
+      "epoch": 0.598805103303818,
+      "grad_norm": 5.165731169595326,
+      "learning_rate": 2e-05,
+      "loss": 0.8322,
+      "step": 25157
+    },
+    {
+      "epoch": 0.5988289060268495,
+      "grad_norm": 11.774842723358661,
+      "learning_rate": 2e-05,
+      "loss": 0.5151,
+      "step": 25158
+    },
+    {
+      "epoch": 0.598852708749881,
+      "grad_norm": 4.556154398986505,
+      "learning_rate": 2e-05,
+      "loss": 0.4559,
+      "step": 25159
+    },
+    {
+      "epoch": 0.5988765114729125,
+      "grad_norm": 9.021871686778528,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 25160
+    },
+    {
+      "epoch": 0.598900314195944,
+      "grad_norm": 5.126294231480786,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 25161
+    },
+    {
+      "epoch": 0.5989241169189755,
+      "grad_norm": 8.718428742682878,
+      "learning_rate": 2e-05,
+      "loss": 0.6054,
+      "step": 25162
+    },
+    {
+      "epoch": 0.598947919642007,
+      "grad_norm": 1.9763519102708802,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 25163
+    },
+    {
+      "epoch": 0.5989717223650386,
+      "grad_norm": 0.07129464909559254,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 25164
+    },
+    {
+      "epoch": 0.5989955250880701,
+      "grad_norm": 0.1167127458163919,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 25165
+    },
+    {
+      "epoch": 0.5990193278111016,
+      "grad_norm": 4.760571505527506,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 25166
+    },
+    {
+      "epoch": 0.599043130534133,
+      "grad_norm": 0.34809168955035574,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 25167
+    },
+    {
+      "epoch": 0.5990669332571646,
+      "grad_norm": 3.94532327202391,
+      "learning_rate": 2e-05,
+      "loss": 0.6088,
+      "step": 25168
+    },
+    {
+      "epoch": 0.5990907359801961,
+      "grad_norm": 1.5522472501866267,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 25169
+    },
+    {
+      "epoch": 0.5991145387032276,
+      "grad_norm": 4.491442811044793,
+      "learning_rate": 2e-05,
+      "loss": 0.6866,
+      "step": 25170
+    },
+    {
+      "epoch": 0.5991383414262592,
+      "grad_norm": 1.4681236597509584,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 25171
+    },
+    {
+      "epoch": 0.5991621441492907,
+      "grad_norm": 3.693152602045834,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 25172
+    },
+    {
+      "epoch": 0.5991859468723222,
+      "grad_norm": 3.2425832411791364,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 25173
+    },
+    {
+      "epoch": 0.5992097495953537,
+      "grad_norm": 4.952826794102687,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 25174
+    },
+    {
+      "epoch": 0.5992335523183853,
+      "grad_norm": 4.02449371513538,
+      "learning_rate": 2e-05,
+      "loss": 0.4602,
+      "step": 25175
+    },
+    {
+      "epoch": 0.5992573550414168,
+      "grad_norm": 5.873506590330041,
+      "learning_rate": 2e-05,
+      "loss": 0.5641,
+      "step": 25176
+    },
+    {
+      "epoch": 0.5992811577644482,
+      "grad_norm": 5.956339913771013,
+      "learning_rate": 2e-05,
+      "loss": 0.3502,
+      "step": 25177
+    },
+    {
+      "epoch": 0.5993049604874797,
+      "grad_norm": 2.1819479000727022,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 25178
+    },
+    {
+      "epoch": 0.5993287632105113,
+      "grad_norm": 1.9642375222819273,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 25179
+    },
+    {
+      "epoch": 0.5993525659335428,
+      "grad_norm": 3.026288256045382,
+      "learning_rate": 2e-05,
+      "loss": 0.2947,
+      "step": 25180
+    },
+    {
+      "epoch": 0.5993763686565743,
+      "grad_norm": 11.769165515801536,
+      "learning_rate": 2e-05,
+      "loss": 0.3618,
+      "step": 25181
+    },
+    {
+      "epoch": 0.5994001713796058,
+      "grad_norm": 5.4529326044120285,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 25182
+    },
+    {
+      "epoch": 0.5994239741026374,
+      "grad_norm": 4.606399969024386,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 25183
+    },
+    {
+      "epoch": 0.5994477768256689,
+      "grad_norm": 1.398375850466891,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 25184
+    },
+    {
+      "epoch": 0.5994715795487003,
+      "grad_norm": 5.34771522392193,
+      "learning_rate": 2e-05,
+      "loss": 1.1921,
+      "step": 25185
+    },
+    {
+      "epoch": 0.5994953822717319,
+      "grad_norm": 0.10820356855716949,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 25186
+    },
+    {
+      "epoch": 0.5995191849947634,
+      "grad_norm": 13.75096974842043,
+      "learning_rate": 2e-05,
+      "loss": 0.5046,
+      "step": 25187
+    },
+    {
+      "epoch": 0.5995429877177949,
+      "grad_norm": 10.565995265746237,
+      "learning_rate": 2e-05,
+      "loss": 0.2261,
+      "step": 25188
+    },
+    {
+      "epoch": 0.5995667904408264,
+      "grad_norm": 2.6464241772131367,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 25189
+    },
+    {
+      "epoch": 0.599590593163858,
+      "grad_norm": 5.861613994395924,
+      "learning_rate": 2e-05,
+      "loss": 0.8889,
+      "step": 25190
+    },
+    {
+      "epoch": 0.5996143958868895,
+      "grad_norm": 1.9485206447343597,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 25191
+    },
+    {
+      "epoch": 0.599638198609921,
+      "grad_norm": 2.9812262051638783,
+      "learning_rate": 2e-05,
+      "loss": 0.5163,
+      "step": 25192
+    },
+    {
+      "epoch": 0.5996620013329524,
+      "grad_norm": 4.103403180303391,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 25193
+    },
+    {
+      "epoch": 0.599685804055984,
+      "grad_norm": 5.182132834729627,
+      "learning_rate": 2e-05,
+      "loss": 0.623,
+      "step": 25194
+    },
+    {
+      "epoch": 0.5997096067790155,
+      "grad_norm": 7.370643686154989,
+      "learning_rate": 2e-05,
+      "loss": 0.9647,
+      "step": 25195
+    },
+    {
+      "epoch": 0.599733409502047,
+      "grad_norm": 1.473202486889066,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 25196
+    },
+    {
+      "epoch": 0.5997572122250785,
+      "grad_norm": 5.027632857385216,
+      "learning_rate": 2e-05,
+      "loss": 0.7035,
+      "step": 25197
+    },
+    {
+      "epoch": 0.5997810149481101,
+      "grad_norm": 5.824997159048007,
+      "learning_rate": 2e-05,
+      "loss": 1.1358,
+      "step": 25198
+    },
+    {
+      "epoch": 0.5998048176711416,
+      "grad_norm": 0.18712158998599815,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 25199
+    },
+    {
+      "epoch": 0.5998286203941731,
+      "grad_norm": 2.79013939595675,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 25200
+    },
+    {
+      "epoch": 0.5998524231172047,
+      "grad_norm": 7.84830163726241,
+      "learning_rate": 2e-05,
+      "loss": 1.2566,
+      "step": 25201
+    },
+    {
+      "epoch": 0.5998762258402361,
+      "grad_norm": 0.2734598672232521,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 25202
+    },
+    {
+      "epoch": 0.5999000285632676,
+      "grad_norm": 4.591739772914951,
+      "learning_rate": 2e-05,
+      "loss": 0.9801,
+      "step": 25203
+    },
+    {
+      "epoch": 0.5999238312862991,
+      "grad_norm": 11.314002232636973,
+      "learning_rate": 2e-05,
+      "loss": 0.4733,
+      "step": 25204
+    },
+    {
+      "epoch": 0.5999476340093307,
+      "grad_norm": 0.33065558202013384,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 25205
+    },
+    {
+      "epoch": 0.5999714367323622,
+      "grad_norm": 0.1240832804090045,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 25206
+    },
+    {
+      "epoch": 0.5999952394553937,
+      "grad_norm": 6.519431292328625,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 25207
+    },
+    {
+      "epoch": 0.6000190421784252,
+      "grad_norm": 1.5486717701043646,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 25208
+    },
+    {
+      "epoch": 0.6000428449014568,
+      "grad_norm": 4.119331328114241,
+      "learning_rate": 2e-05,
+      "loss": 0.5915,
+      "step": 25209
+    },
+    {
+      "epoch": 0.6000666476244882,
+      "grad_norm": 10.831182107941604,
+      "learning_rate": 2e-05,
+      "loss": 1.1627,
+      "step": 25210
+    },
+    {
+      "epoch": 0.6000904503475197,
+      "grad_norm": 3.452473634933774,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 25211
+    },
+    {
+      "epoch": 0.6001142530705512,
+      "grad_norm": 3.7884495798392934,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 25212
+    },
+    {
+      "epoch": 0.6001380557935828,
+      "grad_norm": 14.453908819179405,
+      "learning_rate": 2e-05,
+      "loss": 0.6461,
+      "step": 25213
+    },
+    {
+      "epoch": 0.6001618585166143,
+      "grad_norm": 4.465020405634774,
+      "learning_rate": 2e-05,
+      "loss": 0.4013,
+      "step": 25214
+    },
+    {
+      "epoch": 0.6001856612396458,
+      "grad_norm": 5.618496683016591,
+      "learning_rate": 2e-05,
+      "loss": 0.2656,
+      "step": 25215
+    },
+    {
+      "epoch": 0.6002094639626774,
+      "grad_norm": 2.657998132566291,
+      "learning_rate": 2e-05,
+      "loss": 0.2926,
+      "step": 25216
+    },
+    {
+      "epoch": 0.6002332666857089,
+      "grad_norm": 3.7701417902775893,
+      "learning_rate": 2e-05,
+      "loss": 0.6411,
+      "step": 25217
+    },
+    {
+      "epoch": 0.6002570694087404,
+      "grad_norm": 4.768375439613747,
+      "learning_rate": 2e-05,
+      "loss": 0.9024,
+      "step": 25218
+    },
+    {
+      "epoch": 0.6002808721317718,
+      "grad_norm": 5.024965305366855,
+      "learning_rate": 2e-05,
+      "loss": 0.5173,
+      "step": 25219
+    },
+    {
+      "epoch": 0.6003046748548034,
+      "grad_norm": 4.789111421164335,
+      "learning_rate": 2e-05,
+      "loss": 0.708,
+      "step": 25220
+    },
+    {
+      "epoch": 0.6003284775778349,
+      "grad_norm": 2.8840002734804404,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 25221
+    },
+    {
+      "epoch": 0.6003522803008664,
+      "grad_norm": 1.8782236271415675,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 25222
+    },
+    {
+      "epoch": 0.6003760830238979,
+      "grad_norm": 3.3419151396378024,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 25223
+    },
+    {
+      "epoch": 0.6003998857469295,
+      "grad_norm": 4.758481546968891,
+      "learning_rate": 2e-05,
+      "loss": 1.4076,
+      "step": 25224
+    },
+    {
+      "epoch": 0.600423688469961,
+      "grad_norm": 0.7082538496282546,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 25225
+    },
+    {
+      "epoch": 0.6004474911929925,
+      "grad_norm": 10.226419827094709,
+      "learning_rate": 2e-05,
+      "loss": 0.5371,
+      "step": 25226
+    },
+    {
+      "epoch": 0.6004712939160239,
+      "grad_norm": 2.7622390630634595,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 25227
+    },
+    {
+      "epoch": 0.6004950966390555,
+      "grad_norm": 1.9468007636111506,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 25228
+    },
+    {
+      "epoch": 0.600518899362087,
+      "grad_norm": 11.393275622769135,
+      "learning_rate": 2e-05,
+      "loss": 0.4438,
+      "step": 25229
+    },
+    {
+      "epoch": 0.6005427020851185,
+      "grad_norm": 1.0588750347236195,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 25230
+    },
+    {
+      "epoch": 0.6005665048081501,
+      "grad_norm": 0.7768712483820234,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 25231
+    },
+    {
+      "epoch": 0.6005903075311816,
+      "grad_norm": 7.1925881697394365,
+      "learning_rate": 2e-05,
+      "loss": 1.0836,
+      "step": 25232
+    },
+    {
+      "epoch": 0.6006141102542131,
+      "grad_norm": 1.919757679145656,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 25233
+    },
+    {
+      "epoch": 0.6006379129772446,
+      "grad_norm": 5.103071981435438,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 25234
+    },
+    {
+      "epoch": 0.6006617157002762,
+      "grad_norm": 3.17684282901303,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 25235
+    },
+    {
+      "epoch": 0.6006855184233076,
+      "grad_norm": 7.687786652509737,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 25236
+    },
+    {
+      "epoch": 0.6007093211463391,
+      "grad_norm": 2.409449620830289,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 25237
+    },
+    {
+      "epoch": 0.6007331238693706,
+      "grad_norm": 3.4788072873491362,
+      "learning_rate": 2e-05,
+      "loss": 0.8475,
+      "step": 25238
+    },
+    {
+      "epoch": 0.6007569265924022,
+      "grad_norm": 1.900395276321345,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 25239
+    },
+    {
+      "epoch": 0.6007807293154337,
+      "grad_norm": 0.24856174935363237,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 25240
+    },
+    {
+      "epoch": 0.6008045320384652,
+      "grad_norm": 4.076490748316194,
+      "learning_rate": 2e-05,
+      "loss": 0.4256,
+      "step": 25241
+    },
+    {
+      "epoch": 0.6008283347614967,
+      "grad_norm": 4.987460403152488,
+      "learning_rate": 2e-05,
+      "loss": 1.026,
+      "step": 25242
+    },
+    {
+      "epoch": 0.6008521374845283,
+      "grad_norm": 10.39014329995339,
+      "learning_rate": 2e-05,
+      "loss": 0.3718,
+      "step": 25243
+    },
+    {
+      "epoch": 0.6008759402075597,
+      "grad_norm": 1.5668443091203406,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 25244
+    },
+    {
+      "epoch": 0.6008997429305912,
+      "grad_norm": 1.2415483299164378,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 25245
+    },
+    {
+      "epoch": 0.6009235456536228,
+      "grad_norm": 0.9570148369878778,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 25246
+    },
+    {
+      "epoch": 0.6009473483766543,
+      "grad_norm": 1.40069984366026,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 25247
+    },
+    {
+      "epoch": 0.6009711510996858,
+      "grad_norm": 1.755174290947518,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 25248
+    },
+    {
+      "epoch": 0.6009949538227173,
+      "grad_norm": 0.06640820972148526,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 25249
+    },
+    {
+      "epoch": 0.6010187565457489,
+      "grad_norm": 5.339979443195326,
+      "learning_rate": 2e-05,
+      "loss": 0.6821,
+      "step": 25250
+    },
+    {
+      "epoch": 0.6010425592687804,
+      "grad_norm": 5.659823183817648,
+      "learning_rate": 2e-05,
+      "loss": 0.12,
+      "step": 25251
+    },
+    {
+      "epoch": 0.6010663619918118,
+      "grad_norm": 6.5364198220275735,
+      "learning_rate": 2e-05,
+      "loss": 1.4704,
+      "step": 25252
+    },
+    {
+      "epoch": 0.6010901647148433,
+      "grad_norm": 3.8091352055251875,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 25253
+    },
+    {
+      "epoch": 0.6011139674378749,
+      "grad_norm": 3.2984232245684186,
+      "learning_rate": 2e-05,
+      "loss": 0.3286,
+      "step": 25254
+    },
+    {
+      "epoch": 0.6011377701609064,
+      "grad_norm": 6.0247113985898695,
+      "learning_rate": 2e-05,
+      "loss": 1.0968,
+      "step": 25255
+    },
+    {
+      "epoch": 0.6011615728839379,
+      "grad_norm": 5.242145221422084,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 25256
+    },
+    {
+      "epoch": 0.6011853756069694,
+      "grad_norm": 2.1980644638041196,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 25257
+    },
+    {
+      "epoch": 0.601209178330001,
+      "grad_norm": 5.4393298620522135,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 25258
+    },
+    {
+      "epoch": 0.6012329810530325,
+      "grad_norm": 2.9299164748116366,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 25259
+    },
+    {
+      "epoch": 0.601256783776064,
+      "grad_norm": 6.326649944553296,
+      "learning_rate": 2e-05,
+      "loss": 1.2694,
+      "step": 25260
+    },
+    {
+      "epoch": 0.6012805864990955,
+      "grad_norm": 0.3174905019151831,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 25261
+    },
+    {
+      "epoch": 0.601304389222127,
+      "grad_norm": 2.93222285007576,
+      "learning_rate": 2e-05,
+      "loss": 0.6719,
+      "step": 25262
+    },
+    {
+      "epoch": 0.6013281919451585,
+      "grad_norm": 2.3463391203772126,
+      "learning_rate": 2e-05,
+      "loss": 0.1179,
+      "step": 25263
+    },
+    {
+      "epoch": 0.60135199466819,
+      "grad_norm": 2.6807079192232117,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 25264
+    },
+    {
+      "epoch": 0.6013757973912216,
+      "grad_norm": 3.761130003614957,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 25265
+    },
+    {
+      "epoch": 0.6013996001142531,
+      "grad_norm": 3.046173345325671,
+      "learning_rate": 2e-05,
+      "loss": 0.0926,
+      "step": 25266
+    },
+    {
+      "epoch": 0.6014234028372846,
+      "grad_norm": 5.757159130396885,
+      "learning_rate": 2e-05,
+      "loss": 0.8666,
+      "step": 25267
+    },
+    {
+      "epoch": 0.6014472055603161,
+      "grad_norm": 0.27107655626300703,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 25268
+    },
+    {
+      "epoch": 0.6014710082833477,
+      "grad_norm": 3.1256907806938936,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 25269
+    },
+    {
+      "epoch": 0.6014948110063791,
+      "grad_norm": 0.4343544977189168,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 25270
+    },
+    {
+      "epoch": 0.6015186137294106,
+      "grad_norm": 0.47064532266453385,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 25271
+    },
+    {
+      "epoch": 0.6015424164524421,
+      "grad_norm": 3.7773192687165644,
+      "learning_rate": 2e-05,
+      "loss": 0.5475,
+      "step": 25272
+    },
+    {
+      "epoch": 0.6015662191754737,
+      "grad_norm": 4.215954989934737,
+      "learning_rate": 2e-05,
+      "loss": 0.7035,
+      "step": 25273
+    },
+    {
+      "epoch": 0.6015900218985052,
+      "grad_norm": 6.3017276494715615,
+      "learning_rate": 2e-05,
+      "loss": 0.7821,
+      "step": 25274
+    },
+    {
+      "epoch": 0.6016138246215367,
+      "grad_norm": 1.539882635531769,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 25275
+    },
+    {
+      "epoch": 0.6016376273445683,
+      "grad_norm": 4.625798755462831,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 25276
+    },
+    {
+      "epoch": 0.6016614300675998,
+      "grad_norm": 0.805133493452093,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 25277
+    },
+    {
+      "epoch": 0.6016852327906312,
+      "grad_norm": 11.332997109067648,
+      "learning_rate": 2e-05,
+      "loss": 0.4918,
+      "step": 25278
+    },
+    {
+      "epoch": 0.6017090355136627,
+      "grad_norm": 0.25894004145778615,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 25279
+    },
+    {
+      "epoch": 0.6017328382366943,
+      "grad_norm": 2.0044108774842733,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 25280
+    },
+    {
+      "epoch": 0.6017566409597258,
+      "grad_norm": 2.370858103624745,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 25281
+    },
+    {
+      "epoch": 0.6017804436827573,
+      "grad_norm": 4.001550579985253,
+      "learning_rate": 2e-05,
+      "loss": 0.544,
+      "step": 25282
+    },
+    {
+      "epoch": 0.6018042464057888,
+      "grad_norm": 6.316272619818039,
+      "learning_rate": 2e-05,
+      "loss": 1.1534,
+      "step": 25283
+    },
+    {
+      "epoch": 0.6018280491288204,
+      "grad_norm": 10.2291423185097,
+      "learning_rate": 2e-05,
+      "loss": 0.872,
+      "step": 25284
+    },
+    {
+      "epoch": 0.6018518518518519,
+      "grad_norm": 0.638403517107157,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 25285
+    },
+    {
+      "epoch": 0.6018756545748833,
+      "grad_norm": 3.2815507872163265,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 25286
+    },
+    {
+      "epoch": 0.6018994572979148,
+      "grad_norm": 3.97218908789467,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 25287
+    },
+    {
+      "epoch": 0.6019232600209464,
+      "grad_norm": 3.977179123351731,
+      "learning_rate": 2e-05,
+      "loss": 0.453,
+      "step": 25288
+    },
+    {
+      "epoch": 0.6019470627439779,
+      "grad_norm": 6.035353208473055,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 25289
+    },
+    {
+      "epoch": 0.6019708654670094,
+      "grad_norm": 0.9564657631769286,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 25290
+    },
+    {
+      "epoch": 0.601994668190041,
+      "grad_norm": 3.9818328657602526,
+      "learning_rate": 2e-05,
+      "loss": 0.6845,
+      "step": 25291
+    },
+    {
+      "epoch": 0.6020184709130725,
+      "grad_norm": 5.024898366642485,
+      "learning_rate": 2e-05,
+      "loss": 0.7494,
+      "step": 25292
+    },
+    {
+      "epoch": 0.602042273636104,
+      "grad_norm": 2.5663496272114448,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 25293
+    },
+    {
+      "epoch": 0.6020660763591354,
+      "grad_norm": 1.7190902482211599,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 25294
+    },
+    {
+      "epoch": 0.602089879082167,
+      "grad_norm": 7.6149233852127,
+      "learning_rate": 2e-05,
+      "loss": 0.2321,
+      "step": 25295
+    },
+    {
+      "epoch": 0.6021136818051985,
+      "grad_norm": 3.272653915151136,
+      "learning_rate": 2e-05,
+      "loss": 0.3999,
+      "step": 25296
+    },
+    {
+      "epoch": 0.60213748452823,
+      "grad_norm": 2.2179357404785893,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 25297
+    },
+    {
+      "epoch": 0.6021612872512615,
+      "grad_norm": 3.248540497708035,
+      "learning_rate": 2e-05,
+      "loss": 0.3016,
+      "step": 25298
+    },
+    {
+      "epoch": 0.6021850899742931,
+      "grad_norm": 3.143296333629364,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 25299
+    },
+    {
+      "epoch": 0.6022088926973246,
+      "grad_norm": 15.554851927944673,
+      "learning_rate": 2e-05,
+      "loss": 0.6868,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6022326954203561,
+      "grad_norm": 6.4167959767204525,
+      "learning_rate": 2e-05,
+      "loss": 0.2547,
+      "step": 25301
+    },
+    {
+      "epoch": 0.6022564981433876,
+      "grad_norm": 1.7944103579176074,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 25302
+    },
+    {
+      "epoch": 0.6022803008664191,
+      "grad_norm": 8.290149146920907,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 25303
+    },
+    {
+      "epoch": 0.6023041035894506,
+      "grad_norm": 19.71263804587585,
+      "learning_rate": 2e-05,
+      "loss": 1.1646,
+      "step": 25304
+    },
+    {
+      "epoch": 0.6023279063124821,
+      "grad_norm": 2.0052086608195565,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 25305
+    },
+    {
+      "epoch": 0.6023517090355137,
+      "grad_norm": 10.610255246388238,
+      "learning_rate": 2e-05,
+      "loss": 0.6802,
+      "step": 25306
+    },
+    {
+      "epoch": 0.6023755117585452,
+      "grad_norm": 6.012849329593004,
+      "learning_rate": 2e-05,
+      "loss": 0.814,
+      "step": 25307
+    },
+    {
+      "epoch": 0.6023993144815767,
+      "grad_norm": 12.630548327575942,
+      "learning_rate": 2e-05,
+      "loss": 1.3184,
+      "step": 25308
+    },
+    {
+      "epoch": 0.6024231172046082,
+      "grad_norm": 0.13856026348085593,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 25309
+    },
+    {
+      "epoch": 0.6024469199276398,
+      "grad_norm": 14.400639983660339,
+      "learning_rate": 2e-05,
+      "loss": 0.3334,
+      "step": 25310
+    },
+    {
+      "epoch": 0.6024707226506713,
+      "grad_norm": 2.296600932703432,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 25311
+    },
+    {
+      "epoch": 0.6024945253737027,
+      "grad_norm": 1.5076226521140195,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 25312
+    },
+    {
+      "epoch": 0.6025183280967342,
+      "grad_norm": 3.5117737854285695,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 25313
+    },
+    {
+      "epoch": 0.6025421308197658,
+      "grad_norm": 10.761892407220914,
+      "learning_rate": 2e-05,
+      "loss": 0.3267,
+      "step": 25314
+    },
+    {
+      "epoch": 0.6025659335427973,
+      "grad_norm": 10.584919350338074,
+      "learning_rate": 2e-05,
+      "loss": 0.4331,
+      "step": 25315
+    },
+    {
+      "epoch": 0.6025897362658288,
+      "grad_norm": 1.1211666857687188,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 25316
+    },
+    {
+      "epoch": 0.6026135389888603,
+      "grad_norm": 7.872141258324494,
+      "learning_rate": 2e-05,
+      "loss": 0.3968,
+      "step": 25317
+    },
+    {
+      "epoch": 0.6026373417118919,
+      "grad_norm": 0.4339937901862886,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 25318
+    },
+    {
+      "epoch": 0.6026611444349234,
+      "grad_norm": 1.436721037662947,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 25319
+    },
+    {
+      "epoch": 0.6026849471579548,
+      "grad_norm": 2.2610753024436234,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 25320
+    },
+    {
+      "epoch": 0.6027087498809864,
+      "grad_norm": 9.393749408806388,
+      "learning_rate": 2e-05,
+      "loss": 0.2662,
+      "step": 25321
+    },
+    {
+      "epoch": 0.6027325526040179,
+      "grad_norm": 6.960100150415714,
+      "learning_rate": 2e-05,
+      "loss": 0.7275,
+      "step": 25322
+    },
+    {
+      "epoch": 0.6027563553270494,
+      "grad_norm": 5.643895224142743,
+      "learning_rate": 2e-05,
+      "loss": 0.6303,
+      "step": 25323
+    },
+    {
+      "epoch": 0.6027801580500809,
+      "grad_norm": 3.7966401072807208,
+      "learning_rate": 2e-05,
+      "loss": 0.4683,
+      "step": 25324
+    },
+    {
+      "epoch": 0.6028039607731125,
+      "grad_norm": 0.3550414859103916,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 25325
+    },
+    {
+      "epoch": 0.602827763496144,
+      "grad_norm": 0.9704451066558167,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 25326
+    },
+    {
+      "epoch": 0.6028515662191755,
+      "grad_norm": 0.8447108732870101,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 25327
+    },
+    {
+      "epoch": 0.602875368942207,
+      "grad_norm": 3.899736839702106,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 25328
+    },
+    {
+      "epoch": 0.6028991716652385,
+      "grad_norm": 1.9842877958963838,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 25329
+    },
+    {
+      "epoch": 0.60292297438827,
+      "grad_norm": 2.711169063327124,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 25330
+    },
+    {
+      "epoch": 0.6029467771113015,
+      "grad_norm": 1.5463836647336946,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 25331
+    },
+    {
+      "epoch": 0.602970579834333,
+      "grad_norm": 3.8761740070753334,
+      "learning_rate": 2e-05,
+      "loss": 0.4318,
+      "step": 25332
+    },
+    {
+      "epoch": 0.6029943825573646,
+      "grad_norm": 5.209980877066025,
+      "learning_rate": 2e-05,
+      "loss": 1.1345,
+      "step": 25333
+    },
+    {
+      "epoch": 0.6030181852803961,
+      "grad_norm": 5.40745461822554,
+      "learning_rate": 2e-05,
+      "loss": 1.4556,
+      "step": 25334
+    },
+    {
+      "epoch": 0.6030419880034276,
+      "grad_norm": 0.935697985546533,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 25335
+    },
+    {
+      "epoch": 0.6030657907264592,
+      "grad_norm": 4.326682647894468,
+      "learning_rate": 2e-05,
+      "loss": 0.6349,
+      "step": 25336
+    },
+    {
+      "epoch": 0.6030895934494906,
+      "grad_norm": 0.518372143130606,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 25337
+    },
+    {
+      "epoch": 0.6031133961725221,
+      "grad_norm": 9.959643879727272,
+      "learning_rate": 2e-05,
+      "loss": 1.0435,
+      "step": 25338
+    },
+    {
+      "epoch": 0.6031371988955536,
+      "grad_norm": 5.212236930833666,
+      "learning_rate": 2e-05,
+      "loss": 0.7316,
+      "step": 25339
+    },
+    {
+      "epoch": 0.6031610016185852,
+      "grad_norm": 2.9835082654509293,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 25340
+    },
+    {
+      "epoch": 0.6031848043416167,
+      "grad_norm": 5.2973562969677355,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 25341
+    },
+    {
+      "epoch": 0.6032086070646482,
+      "grad_norm": 4.674777000294071,
+      "learning_rate": 2e-05,
+      "loss": 0.7976,
+      "step": 25342
+    },
+    {
+      "epoch": 0.6032324097876797,
+      "grad_norm": 2.5345581326954147,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 25343
+    },
+    {
+      "epoch": 0.6032562125107113,
+      "grad_norm": 0.6843581162659639,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 25344
+    },
+    {
+      "epoch": 0.6032800152337428,
+      "grad_norm": 4.591125115374396,
+      "learning_rate": 2e-05,
+      "loss": 0.7122,
+      "step": 25345
+    },
+    {
+      "epoch": 0.6033038179567742,
+      "grad_norm": 4.161762000309369,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 25346
+    },
+    {
+      "epoch": 0.6033276206798057,
+      "grad_norm": 4.8256614827965505,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 25347
+    },
+    {
+      "epoch": 0.6033514234028373,
+      "grad_norm": 2.949671927406647,
+      "learning_rate": 2e-05,
+      "loss": 0.5864,
+      "step": 25348
+    },
+    {
+      "epoch": 0.6033752261258688,
+      "grad_norm": 0.17050900553899342,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 25349
+    },
+    {
+      "epoch": 0.6033990288489003,
+      "grad_norm": 30.751816420351005,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 25350
+    },
+    {
+      "epoch": 0.6034228315719319,
+      "grad_norm": 1.8211814373015476,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 25351
+    },
+    {
+      "epoch": 0.6034466342949634,
+      "grad_norm": 0.4584235424420414,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 25352
+    },
+    {
+      "epoch": 0.6034704370179949,
+      "grad_norm": 5.3145774042792855,
+      "learning_rate": 2e-05,
+      "loss": 0.7693,
+      "step": 25353
+    },
+    {
+      "epoch": 0.6034942397410263,
+      "grad_norm": 3.5731540594109816,
+      "learning_rate": 2e-05,
+      "loss": 0.5452,
+      "step": 25354
+    },
+    {
+      "epoch": 0.6035180424640579,
+      "grad_norm": 4.086597607888022,
+      "learning_rate": 2e-05,
+      "loss": 0.5588,
+      "step": 25355
+    },
+    {
+      "epoch": 0.6035418451870894,
+      "grad_norm": 3.869735142599339,
+      "learning_rate": 2e-05,
+      "loss": 0.3082,
+      "step": 25356
+    },
+    {
+      "epoch": 0.6035656479101209,
+      "grad_norm": 2.797297172993724,
+      "learning_rate": 2e-05,
+      "loss": 0.3079,
+      "step": 25357
+    },
+    {
+      "epoch": 0.6035894506331524,
+      "grad_norm": 3.325273139062655,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 25358
+    },
+    {
+      "epoch": 0.603613253356184,
+      "grad_norm": 4.026682196414398,
+      "learning_rate": 2e-05,
+      "loss": 0.7423,
+      "step": 25359
+    },
+    {
+      "epoch": 0.6036370560792155,
+      "grad_norm": 3.980847851032189,
+      "learning_rate": 2e-05,
+      "loss": 0.1792,
+      "step": 25360
+    },
+    {
+      "epoch": 0.603660858802247,
+      "grad_norm": 1.6432160994429073,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 25361
+    },
+    {
+      "epoch": 0.6036846615252784,
+      "grad_norm": 3.949899258162204,
+      "learning_rate": 2e-05,
+      "loss": 0.7552,
+      "step": 25362
+    },
+    {
+      "epoch": 0.60370846424831,
+      "grad_norm": 9.475411651371367,
+      "learning_rate": 2e-05,
+      "loss": 0.7652,
+      "step": 25363
+    },
+    {
+      "epoch": 0.6037322669713415,
+      "grad_norm": 0.09510392783923065,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 25364
+    },
+    {
+      "epoch": 0.603756069694373,
+      "grad_norm": 31.225960011821968,
+      "learning_rate": 2e-05,
+      "loss": 2.0485,
+      "step": 25365
+    },
+    {
+      "epoch": 0.6037798724174046,
+      "grad_norm": 3.0932874845371674,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 25366
+    },
+    {
+      "epoch": 0.6038036751404361,
+      "grad_norm": 3.5164116195660347,
+      "learning_rate": 2e-05,
+      "loss": 0.2692,
+      "step": 25367
+    },
+    {
+      "epoch": 0.6038274778634676,
+      "grad_norm": 1.892536976829595,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 25368
+    },
+    {
+      "epoch": 0.6038512805864991,
+      "grad_norm": 2.191502824453653,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 25369
+    },
+    {
+      "epoch": 0.6038750833095307,
+      "grad_norm": 10.429581677265531,
+      "learning_rate": 2e-05,
+      "loss": 1.0018,
+      "step": 25370
+    },
+    {
+      "epoch": 0.6038988860325621,
+      "grad_norm": 5.295525338383363,
+      "learning_rate": 2e-05,
+      "loss": 0.5862,
+      "step": 25371
+    },
+    {
+      "epoch": 0.6039226887555936,
+      "grad_norm": 0.21535344008528665,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 25372
+    },
+    {
+      "epoch": 0.6039464914786251,
+      "grad_norm": 5.400901210670193,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 25373
+    },
+    {
+      "epoch": 0.6039702942016567,
+      "grad_norm": 0.30511289872687614,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 25374
+    },
+    {
+      "epoch": 0.6039940969246882,
+      "grad_norm": 11.29835682940019,
+      "learning_rate": 2e-05,
+      "loss": 0.2427,
+      "step": 25375
+    },
+    {
+      "epoch": 0.6040178996477197,
+      "grad_norm": 4.692196508867799,
+      "learning_rate": 2e-05,
+      "loss": 0.7075,
+      "step": 25376
+    },
+    {
+      "epoch": 0.6040417023707512,
+      "grad_norm": 9.757260443711807,
+      "learning_rate": 2e-05,
+      "loss": 0.4988,
+      "step": 25377
+    },
+    {
+      "epoch": 0.6040655050937828,
+      "grad_norm": 0.1655656528184171,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 25378
+    },
+    {
+      "epoch": 0.6040893078168142,
+      "grad_norm": 5.614293153336377,
+      "learning_rate": 2e-05,
+      "loss": 1.7431,
+      "step": 25379
+    },
+    {
+      "epoch": 0.6041131105398457,
+      "grad_norm": 1.1481086338533093,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 25380
+    },
+    {
+      "epoch": 0.6041369132628773,
+      "grad_norm": 8.587082812590602,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 25381
+    },
+    {
+      "epoch": 0.6041607159859088,
+      "grad_norm": 4.111041331588705,
+      "learning_rate": 2e-05,
+      "loss": 0.8997,
+      "step": 25382
+    },
+    {
+      "epoch": 0.6041845187089403,
+      "grad_norm": 3.8633447116897184,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 25383
+    },
+    {
+      "epoch": 0.6042083214319718,
+      "grad_norm": 4.265396759482424,
+      "learning_rate": 2e-05,
+      "loss": 0.6121,
+      "step": 25384
+    },
+    {
+      "epoch": 0.6042321241550034,
+      "grad_norm": 4.738126452960772,
+      "learning_rate": 2e-05,
+      "loss": 0.4414,
+      "step": 25385
+    },
+    {
+      "epoch": 0.6042559268780349,
+      "grad_norm": 4.9013079303723925,
+      "learning_rate": 2e-05,
+      "loss": 0.4739,
+      "step": 25386
+    },
+    {
+      "epoch": 0.6042797296010664,
+      "grad_norm": 1.714661848334266,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 25387
+    },
+    {
+      "epoch": 0.6043035323240978,
+      "grad_norm": 2.4788424814499974,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 25388
+    },
+    {
+      "epoch": 0.6043273350471294,
+      "grad_norm": 4.825479004226064,
+      "learning_rate": 2e-05,
+      "loss": 0.7714,
+      "step": 25389
+    },
+    {
+      "epoch": 0.6043511377701609,
+      "grad_norm": 1.4777057305653485,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 25390
+    },
+    {
+      "epoch": 0.6043749404931924,
+      "grad_norm": 14.40305920160303,
+      "learning_rate": 2e-05,
+      "loss": 0.6858,
+      "step": 25391
+    },
+    {
+      "epoch": 0.6043987432162239,
+      "grad_norm": 0.9783888430598011,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 25392
+    },
+    {
+      "epoch": 0.6044225459392555,
+      "grad_norm": 1.3155498352376864,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 25393
+    },
+    {
+      "epoch": 0.604446348662287,
+      "grad_norm": 0.41148206314165386,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 25394
+    },
+    {
+      "epoch": 0.6044701513853185,
+      "grad_norm": 8.540100588870683,
+      "learning_rate": 2e-05,
+      "loss": 1.3626,
+      "step": 25395
+    },
+    {
+      "epoch": 0.6044939541083499,
+      "grad_norm": 1.5783739741794884,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 25396
+    },
+    {
+      "epoch": 0.6045177568313815,
+      "grad_norm": 0.9157513682799263,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 25397
+    },
+    {
+      "epoch": 0.604541559554413,
+      "grad_norm": 1.4112960892917121,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 25398
+    },
+    {
+      "epoch": 0.6045653622774445,
+      "grad_norm": 2.9687825938734047,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 25399
+    },
+    {
+      "epoch": 0.6045891650004761,
+      "grad_norm": 0.5913993191333632,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 25400
+    },
+    {
+      "epoch": 0.6046129677235076,
+      "grad_norm": 1.039922440763504,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 25401
+    },
+    {
+      "epoch": 0.6046367704465391,
+      "grad_norm": 2.3643742233337477,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 25402
+    },
+    {
+      "epoch": 0.6046605731695706,
+      "grad_norm": 7.947753668607995,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 25403
+    },
+    {
+      "epoch": 0.6046843758926022,
+      "grad_norm": 2.6419906583779453,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 25404
+    },
+    {
+      "epoch": 0.6047081786156336,
+      "grad_norm": 7.370374832609304,
+      "learning_rate": 2e-05,
+      "loss": 0.3902,
+      "step": 25405
+    },
+    {
+      "epoch": 0.6047319813386651,
+      "grad_norm": 10.201005187925137,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 25406
+    },
+    {
+      "epoch": 0.6047557840616966,
+      "grad_norm": 3.5149761782934075,
+      "learning_rate": 2e-05,
+      "loss": 0.6851,
+      "step": 25407
+    },
+    {
+      "epoch": 0.6047795867847282,
+      "grad_norm": 3.124616275911299,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 25408
+    },
+    {
+      "epoch": 0.6048033895077597,
+      "grad_norm": 6.339209002889008,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 25409
+    },
+    {
+      "epoch": 0.6048271922307912,
+      "grad_norm": 15.102517148647431,
+      "learning_rate": 2e-05,
+      "loss": 0.4937,
+      "step": 25410
+    },
+    {
+      "epoch": 0.6048509949538227,
+      "grad_norm": 3.1943545682491297,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 25411
+    },
+    {
+      "epoch": 0.6048747976768543,
+      "grad_norm": 3.1800337111221286,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 25412
+    },
+    {
+      "epoch": 0.6048986003998857,
+      "grad_norm": 2.239633521615876,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 25413
+    },
+    {
+      "epoch": 0.6049224031229172,
+      "grad_norm": 7.191483878148468,
+      "learning_rate": 2e-05,
+      "loss": 0.6175,
+      "step": 25414
+    },
+    {
+      "epoch": 0.6049462058459488,
+      "grad_norm": 2.477628055141829,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 25415
+    },
+    {
+      "epoch": 0.6049700085689803,
+      "grad_norm": 1.7798971013284626,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 25416
+    },
+    {
+      "epoch": 0.6049938112920118,
+      "grad_norm": 2.2194266416525994,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 25417
+    },
+    {
+      "epoch": 0.6050176140150433,
+      "grad_norm": 6.826736968771346,
+      "learning_rate": 2e-05,
+      "loss": 0.6218,
+      "step": 25418
+    },
+    {
+      "epoch": 0.6050414167380749,
+      "grad_norm": 0.23788028315577595,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 25419
+    },
+    {
+      "epoch": 0.6050652194611064,
+      "grad_norm": 2.3996330066853635,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 25420
+    },
+    {
+      "epoch": 0.6050890221841378,
+      "grad_norm": 3.5508969479705264,
+      "learning_rate": 2e-05,
+      "loss": 0.4921,
+      "step": 25421
+    },
+    {
+      "epoch": 0.6051128249071693,
+      "grad_norm": 2.4004019165417394,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 25422
+    },
+    {
+      "epoch": 0.6051366276302009,
+      "grad_norm": 6.837997528057722,
+      "learning_rate": 2e-05,
+      "loss": 0.7303,
+      "step": 25423
+    },
+    {
+      "epoch": 0.6051604303532324,
+      "grad_norm": 5.2785064709828875,
+      "learning_rate": 2e-05,
+      "loss": 1.0457,
+      "step": 25424
+    },
+    {
+      "epoch": 0.6051842330762639,
+      "grad_norm": 0.4396083254015557,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 25425
+    },
+    {
+      "epoch": 0.6052080357992954,
+      "grad_norm": 2.4929370046249124,
+      "learning_rate": 2e-05,
+      "loss": 0.3675,
+      "step": 25426
+    },
+    {
+      "epoch": 0.605231838522327,
+      "grad_norm": 13.486906039682616,
+      "learning_rate": 2e-05,
+      "loss": 0.7163,
+      "step": 25427
+    },
+    {
+      "epoch": 0.6052556412453585,
+      "grad_norm": 0.9268222436184871,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 25428
+    },
+    {
+      "epoch": 0.60527944396839,
+      "grad_norm": 6.275954965955439,
+      "learning_rate": 2e-05,
+      "loss": 0.147,
+      "step": 25429
+    },
+    {
+      "epoch": 0.6053032466914215,
+      "grad_norm": 2.528814027702569,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 25430
+    },
+    {
+      "epoch": 0.605327049414453,
+      "grad_norm": 3.6542567565759976,
+      "learning_rate": 2e-05,
+      "loss": 0.7213,
+      "step": 25431
+    },
+    {
+      "epoch": 0.6053508521374845,
+      "grad_norm": 2.2315298177225324,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 25432
+    },
+    {
+      "epoch": 0.605374654860516,
+      "grad_norm": 0.1800830469792579,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 25433
+    },
+    {
+      "epoch": 0.6053984575835476,
+      "grad_norm": 3.534820517715324,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 25434
+    },
+    {
+      "epoch": 0.6054222603065791,
+      "grad_norm": 0.6056884869150332,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 25435
+    },
+    {
+      "epoch": 0.6054460630296106,
+      "grad_norm": 3.1483768641826786,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 25436
+    },
+    {
+      "epoch": 0.6054698657526421,
+      "grad_norm": 5.832949958016267,
+      "learning_rate": 2e-05,
+      "loss": 1.014,
+      "step": 25437
+    },
+    {
+      "epoch": 0.6054936684756737,
+      "grad_norm": 1.5926861430962604,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 25438
+    },
+    {
+      "epoch": 0.6055174711987051,
+      "grad_norm": 1.8419326604051414,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 25439
+    },
+    {
+      "epoch": 0.6055412739217366,
+      "grad_norm": 0.17359055071820706,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 25440
+    },
+    {
+      "epoch": 0.6055650766447681,
+      "grad_norm": 5.9297742338267,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 25441
+    },
+    {
+      "epoch": 0.6055888793677997,
+      "grad_norm": 3.0608397476947697,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 25442
+    },
+    {
+      "epoch": 0.6056126820908312,
+      "grad_norm": 8.308813589914044,
+      "learning_rate": 2e-05,
+      "loss": 1.7971,
+      "step": 25443
+    },
+    {
+      "epoch": 0.6056364848138627,
+      "grad_norm": 3.4277468258526036,
+      "learning_rate": 2e-05,
+      "loss": 0.7073,
+      "step": 25444
+    },
+    {
+      "epoch": 0.6056602875368943,
+      "grad_norm": 11.175930319063424,
+      "learning_rate": 2e-05,
+      "loss": 0.6018,
+      "step": 25445
+    },
+    {
+      "epoch": 0.6056840902599258,
+      "grad_norm": 11.779624331068208,
+      "learning_rate": 2e-05,
+      "loss": 0.696,
+      "step": 25446
+    },
+    {
+      "epoch": 0.6057078929829572,
+      "grad_norm": 3.349870677336115,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 25447
+    },
+    {
+      "epoch": 0.6057316957059887,
+      "grad_norm": 7.0300009236058525,
+      "learning_rate": 2e-05,
+      "loss": 0.3118,
+      "step": 25448
+    },
+    {
+      "epoch": 0.6057554984290203,
+      "grad_norm": 0.36679066852737274,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 25449
+    },
+    {
+      "epoch": 0.6057793011520518,
+      "grad_norm": 3.305780460946015,
+      "learning_rate": 2e-05,
+      "loss": 0.4387,
+      "step": 25450
+    },
+    {
+      "epoch": 0.6058031038750833,
+      "grad_norm": 6.833082984841447,
+      "learning_rate": 2e-05,
+      "loss": 0.8422,
+      "step": 25451
+    },
+    {
+      "epoch": 0.6058269065981148,
+      "grad_norm": 8.200187336928956,
+      "learning_rate": 2e-05,
+      "loss": 0.5777,
+      "step": 25452
+    },
+    {
+      "epoch": 0.6058507093211464,
+      "grad_norm": 2.000193944738499,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 25453
+    },
+    {
+      "epoch": 0.6058745120441779,
+      "grad_norm": 11.406672405106207,
+      "learning_rate": 2e-05,
+      "loss": 1.0676,
+      "step": 25454
+    },
+    {
+      "epoch": 0.6058983147672093,
+      "grad_norm": 7.565095605473225,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 25455
+    },
+    {
+      "epoch": 0.6059221174902408,
+      "grad_norm": 2.2040085208991864,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 25456
+    },
+    {
+      "epoch": 0.6059459202132724,
+      "grad_norm": 0.5834426432101044,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 25457
+    },
+    {
+      "epoch": 0.6059697229363039,
+      "grad_norm": 7.594179871212307,
+      "learning_rate": 2e-05,
+      "loss": 0.3747,
+      "step": 25458
+    },
+    {
+      "epoch": 0.6059935256593354,
+      "grad_norm": 1.5723389747392764,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 25459
+    },
+    {
+      "epoch": 0.606017328382367,
+      "grad_norm": 0.22525109564759235,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 25460
+    },
+    {
+      "epoch": 0.6060411311053985,
+      "grad_norm": 0.5044177480238383,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 25461
+    },
+    {
+      "epoch": 0.60606493382843,
+      "grad_norm": 1.146351984749938,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 25462
+    },
+    {
+      "epoch": 0.6060887365514614,
+      "grad_norm": 3.894036185102023,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 25463
+    },
+    {
+      "epoch": 0.606112539274493,
+      "grad_norm": 1.3610374202126647,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 25464
+    },
+    {
+      "epoch": 0.6061363419975245,
+      "grad_norm": 6.001747584436692,
+      "learning_rate": 2e-05,
+      "loss": 0.9337,
+      "step": 25465
+    },
+    {
+      "epoch": 0.606160144720556,
+      "grad_norm": 4.200452970197764,
+      "learning_rate": 2e-05,
+      "loss": 0.7111,
+      "step": 25466
+    },
+    {
+      "epoch": 0.6061839474435875,
+      "grad_norm": 7.064579638672341,
+      "learning_rate": 2e-05,
+      "loss": 1.0416,
+      "step": 25467
+    },
+    {
+      "epoch": 0.6062077501666191,
+      "grad_norm": 4.456887872219047,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 25468
+    },
+    {
+      "epoch": 0.6062315528896506,
+      "grad_norm": 1.3606273647925269,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 25469
+    },
+    {
+      "epoch": 0.6062553556126821,
+      "grad_norm": 4.303710255285197,
+      "learning_rate": 2e-05,
+      "loss": 0.1007,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6062791583357136,
+      "grad_norm": 7.149647154404213,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 25471
+    },
+    {
+      "epoch": 0.6063029610587451,
+      "grad_norm": 8.376689863802556,
+      "learning_rate": 2e-05,
+      "loss": 1.0877,
+      "step": 25472
+    },
+    {
+      "epoch": 0.6063267637817766,
+      "grad_norm": 0.3097017647806506,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 25473
+    },
+    {
+      "epoch": 0.6063505665048081,
+      "grad_norm": 5.178672482369449,
+      "learning_rate": 2e-05,
+      "loss": 0.9624,
+      "step": 25474
+    },
+    {
+      "epoch": 0.6063743692278397,
+      "grad_norm": 4.517221360133363,
+      "learning_rate": 2e-05,
+      "loss": 0.9342,
+      "step": 25475
+    },
+    {
+      "epoch": 0.6063981719508712,
+      "grad_norm": 4.0164266730110825,
+      "learning_rate": 2e-05,
+      "loss": 0.5973,
+      "step": 25476
+    },
+    {
+      "epoch": 0.6064219746739027,
+      "grad_norm": 3.146769449403833,
+      "learning_rate": 2e-05,
+      "loss": 0.6061,
+      "step": 25477
+    },
+    {
+      "epoch": 0.6064457773969342,
+      "grad_norm": 2.8461766179032444,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 25478
+    },
+    {
+      "epoch": 0.6064695801199658,
+      "grad_norm": 2.344402954163021,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 25479
+    },
+    {
+      "epoch": 0.6064933828429973,
+      "grad_norm": 18.77942461089197,
+      "learning_rate": 2e-05,
+      "loss": 0.4396,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6065171855660287,
+      "grad_norm": 9.246295252358609,
+      "learning_rate": 2e-05,
+      "loss": 0.4475,
+      "step": 25481
+    },
+    {
+      "epoch": 0.6065409882890602,
+      "grad_norm": 5.012357115802832,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 25482
+    },
+    {
+      "epoch": 0.6065647910120918,
+      "grad_norm": 2.527030812059068,
+      "learning_rate": 2e-05,
+      "loss": 0.0651,
+      "step": 25483
+    },
+    {
+      "epoch": 0.6065885937351233,
+      "grad_norm": 7.9989756619420405,
+      "learning_rate": 2e-05,
+      "loss": 0.7326,
+      "step": 25484
+    },
+    {
+      "epoch": 0.6066123964581548,
+      "grad_norm": 5.31848724572029,
+      "learning_rate": 2e-05,
+      "loss": 1.0756,
+      "step": 25485
+    },
+    {
+      "epoch": 0.6066361991811863,
+      "grad_norm": 5.290582890519231,
+      "learning_rate": 2e-05,
+      "loss": 0.9632,
+      "step": 25486
+    },
+    {
+      "epoch": 0.6066600019042179,
+      "grad_norm": 0.07965127700344082,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 25487
+    },
+    {
+      "epoch": 0.6066838046272494,
+      "grad_norm": 5.962905073625781,
+      "learning_rate": 2e-05,
+      "loss": 0.378,
+      "step": 25488
+    },
+    {
+      "epoch": 0.6067076073502808,
+      "grad_norm": 3.4736440745300325,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 25489
+    },
+    {
+      "epoch": 0.6067314100733124,
+      "grad_norm": 5.217598024714228,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6067552127963439,
+      "grad_norm": 7.561409620178241,
+      "learning_rate": 2e-05,
+      "loss": 0.3951,
+      "step": 25491
+    },
+    {
+      "epoch": 0.6067790155193754,
+      "grad_norm": 15.055911521800825,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 25492
+    },
+    {
+      "epoch": 0.6068028182424069,
+      "grad_norm": 3.9909039622550595,
+      "learning_rate": 2e-05,
+      "loss": 0.455,
+      "step": 25493
+    },
+    {
+      "epoch": 0.6068266209654385,
+      "grad_norm": 5.224368059852479,
+      "learning_rate": 2e-05,
+      "loss": 1.3483,
+      "step": 25494
+    },
+    {
+      "epoch": 0.60685042368847,
+      "grad_norm": 5.073907361209476,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 25495
+    },
+    {
+      "epoch": 0.6068742264115015,
+      "grad_norm": 8.716309788118753,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 25496
+    },
+    {
+      "epoch": 0.606898029134533,
+      "grad_norm": 4.146871165511368,
+      "learning_rate": 2e-05,
+      "loss": 0.7337,
+      "step": 25497
+    },
+    {
+      "epoch": 0.6069218318575645,
+      "grad_norm": 0.3364350638225875,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 25498
+    },
+    {
+      "epoch": 0.606945634580596,
+      "grad_norm": 5.978189378391872,
+      "learning_rate": 2e-05,
+      "loss": 1.2839,
+      "step": 25499
+    },
+    {
+      "epoch": 0.6069694373036275,
+      "grad_norm": 1.6466931758992036,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 25500
+    },
+    {
+      "epoch": 0.606993240026659,
+      "grad_norm": 3.168809269313975,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 25501
+    },
+    {
+      "epoch": 0.6070170427496906,
+      "grad_norm": 2.1468374644822714,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 25502
+    },
+    {
+      "epoch": 0.6070408454727221,
+      "grad_norm": 1.1793350342735471,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 25503
+    },
+    {
+      "epoch": 0.6070646481957536,
+      "grad_norm": 6.153424835552244,
+      "learning_rate": 2e-05,
+      "loss": 0.7878,
+      "step": 25504
+    },
+    {
+      "epoch": 0.6070884509187852,
+      "grad_norm": 2.0916325706558614,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 25505
+    },
+    {
+      "epoch": 0.6071122536418166,
+      "grad_norm": 2.116799350900668,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 25506
+    },
+    {
+      "epoch": 0.6071360563648481,
+      "grad_norm": 5.632825935008332,
+      "learning_rate": 2e-05,
+      "loss": 1.6837,
+      "step": 25507
+    },
+    {
+      "epoch": 0.6071598590878796,
+      "grad_norm": 3.0942999668988938,
+      "learning_rate": 2e-05,
+      "loss": 0.3401,
+      "step": 25508
+    },
+    {
+      "epoch": 0.6071836618109112,
+      "grad_norm": 5.387010688579265,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 25509
+    },
+    {
+      "epoch": 0.6072074645339427,
+      "grad_norm": 5.563443713923094,
+      "learning_rate": 2e-05,
+      "loss": 0.3276,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6072312672569742,
+      "grad_norm": 9.194554573665975,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 25511
+    },
+    {
+      "epoch": 0.6072550699800057,
+      "grad_norm": 7.750241933621526,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 25512
+    },
+    {
+      "epoch": 0.6072788727030373,
+      "grad_norm": 12.301817853857038,
+      "learning_rate": 2e-05,
+      "loss": 1.4612,
+      "step": 25513
+    },
+    {
+      "epoch": 0.6073026754260687,
+      "grad_norm": 4.972117286861102,
+      "learning_rate": 2e-05,
+      "loss": 0.4432,
+      "step": 25514
+    },
+    {
+      "epoch": 0.6073264781491002,
+      "grad_norm": 40.586812747732495,
+      "learning_rate": 2e-05,
+      "loss": 0.7248,
+      "step": 25515
+    },
+    {
+      "epoch": 0.6073502808721317,
+      "grad_norm": 0.1277955132670518,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 25516
+    },
+    {
+      "epoch": 0.6073740835951633,
+      "grad_norm": 1.5934323918885724,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 25517
+    },
+    {
+      "epoch": 0.6073978863181948,
+      "grad_norm": 3.3246550534083315,
+      "learning_rate": 2e-05,
+      "loss": 0.3406,
+      "step": 25518
+    },
+    {
+      "epoch": 0.6074216890412263,
+      "grad_norm": 4.05486013624042,
+      "learning_rate": 2e-05,
+      "loss": 0.0882,
+      "step": 25519
+    },
+    {
+      "epoch": 0.6074454917642579,
+      "grad_norm": 3.6560672086685426,
+      "learning_rate": 2e-05,
+      "loss": 0.7247,
+      "step": 25520
+    },
+    {
+      "epoch": 0.6074692944872894,
+      "grad_norm": 6.974582006204706,
+      "learning_rate": 2e-05,
+      "loss": 1.4074,
+      "step": 25521
+    },
+    {
+      "epoch": 0.6074930972103209,
+      "grad_norm": 4.97092176624545,
+      "learning_rate": 2e-05,
+      "loss": 0.7989,
+      "step": 25522
+    },
+    {
+      "epoch": 0.6075168999333523,
+      "grad_norm": 3.408096189539585,
+      "learning_rate": 2e-05,
+      "loss": 0.4014,
+      "step": 25523
+    },
+    {
+      "epoch": 0.6075407026563839,
+      "grad_norm": 3.238632973317287,
+      "learning_rate": 2e-05,
+      "loss": 0.4763,
+      "step": 25524
+    },
+    {
+      "epoch": 0.6075645053794154,
+      "grad_norm": 13.779308752430575,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 25525
+    },
+    {
+      "epoch": 0.6075883081024469,
+      "grad_norm": 4.879727684303085,
+      "learning_rate": 2e-05,
+      "loss": 0.7202,
+      "step": 25526
+    },
+    {
+      "epoch": 0.6076121108254784,
+      "grad_norm": 5.413360490056231,
+      "learning_rate": 2e-05,
+      "loss": 1.6944,
+      "step": 25527
+    },
+    {
+      "epoch": 0.60763591354851,
+      "grad_norm": 6.992022394178764,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 25528
+    },
+    {
+      "epoch": 0.6076597162715415,
+      "grad_norm": 4.079197690391602,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 25529
+    },
+    {
+      "epoch": 0.607683518994573,
+      "grad_norm": 5.784208485231336,
+      "learning_rate": 2e-05,
+      "loss": 0.2048,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6077073217176044,
+      "grad_norm": 2.068213107694429,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 25531
+    },
+    {
+      "epoch": 0.607731124440636,
+      "grad_norm": 5.661551590805324,
+      "learning_rate": 2e-05,
+      "loss": 0.6013,
+      "step": 25532
+    },
+    {
+      "epoch": 0.6077549271636675,
+      "grad_norm": 2.3384485829659787,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 25533
+    },
+    {
+      "epoch": 0.607778729886699,
+      "grad_norm": 1.598758176814469,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 25534
+    },
+    {
+      "epoch": 0.6078025326097306,
+      "grad_norm": 6.827946843180191,
+      "learning_rate": 2e-05,
+      "loss": 0.5892,
+      "step": 25535
+    },
+    {
+      "epoch": 0.6078263353327621,
+      "grad_norm": 5.870963135353007,
+      "learning_rate": 2e-05,
+      "loss": 0.9582,
+      "step": 25536
+    },
+    {
+      "epoch": 0.6078501380557936,
+      "grad_norm": 9.773114972734687,
+      "learning_rate": 2e-05,
+      "loss": 0.2577,
+      "step": 25537
+    },
+    {
+      "epoch": 0.6078739407788251,
+      "grad_norm": 4.058142624823558,
+      "learning_rate": 2e-05,
+      "loss": 0.125,
+      "step": 25538
+    },
+    {
+      "epoch": 0.6078977435018567,
+      "grad_norm": 5.4908009937218525,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 25539
+    },
+    {
+      "epoch": 0.6079215462248881,
+      "grad_norm": 3.1079755346335087,
+      "learning_rate": 2e-05,
+      "loss": 0.4334,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6079453489479196,
+      "grad_norm": 17.848030731362357,
+      "learning_rate": 2e-05,
+      "loss": 0.2047,
+      "step": 25541
+    },
+    {
+      "epoch": 0.6079691516709511,
+      "grad_norm": 3.8317095062806263,
+      "learning_rate": 2e-05,
+      "loss": 0.6465,
+      "step": 25542
+    },
+    {
+      "epoch": 0.6079929543939827,
+      "grad_norm": 4.4129026316851725,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 25543
+    },
+    {
+      "epoch": 0.6080167571170142,
+      "grad_norm": 4.969691825742841,
+      "learning_rate": 2e-05,
+      "loss": 0.7832,
+      "step": 25544
+    },
+    {
+      "epoch": 0.6080405598400457,
+      "grad_norm": 3.4478511810552503,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 25545
+    },
+    {
+      "epoch": 0.6080643625630772,
+      "grad_norm": 2.830214810896429,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 25546
+    },
+    {
+      "epoch": 0.6080881652861088,
+      "grad_norm": 0.884048476140793,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 25547
+    },
+    {
+      "epoch": 0.6081119680091402,
+      "grad_norm": 11.827030346100704,
+      "learning_rate": 2e-05,
+      "loss": 1.0359,
+      "step": 25548
+    },
+    {
+      "epoch": 0.6081357707321717,
+      "grad_norm": 10.223358943183953,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 25549
+    },
+    {
+      "epoch": 0.6081595734552033,
+      "grad_norm": 3.3192121580087486,
+      "learning_rate": 2e-05,
+      "loss": 0.529,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6081833761782348,
+      "grad_norm": 4.306638974788404,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 25551
+    },
+    {
+      "epoch": 0.6082071789012663,
+      "grad_norm": 3.883688808860007,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 25552
+    },
+    {
+      "epoch": 0.6082309816242978,
+      "grad_norm": 5.0366673186702355,
+      "learning_rate": 2e-05,
+      "loss": 0.6338,
+      "step": 25553
+    },
+    {
+      "epoch": 0.6082547843473294,
+      "grad_norm": 0.3079402443635425,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 25554
+    },
+    {
+      "epoch": 0.6082785870703609,
+      "grad_norm": 8.646550050152205,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 25555
+    },
+    {
+      "epoch": 0.6083023897933924,
+      "grad_norm": 0.4914487237439726,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 25556
+    },
+    {
+      "epoch": 0.6083261925164238,
+      "grad_norm": 0.09657002110652291,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 25557
+    },
+    {
+      "epoch": 0.6083499952394554,
+      "grad_norm": 4.470945220638604,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 25558
+    },
+    {
+      "epoch": 0.6083737979624869,
+      "grad_norm": 4.285184893435913,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 25559
+    },
+    {
+      "epoch": 0.6083976006855184,
+      "grad_norm": 0.6746298662096982,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6084214034085499,
+      "grad_norm": 2.770631321216264,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 25561
+    },
+    {
+      "epoch": 0.6084452061315815,
+      "grad_norm": 4.6164878987440305,
+      "learning_rate": 2e-05,
+      "loss": 0.8507,
+      "step": 25562
+    },
+    {
+      "epoch": 0.608469008854613,
+      "grad_norm": 3.4687407738651066,
+      "learning_rate": 2e-05,
+      "loss": 0.3837,
+      "step": 25563
+    },
+    {
+      "epoch": 0.6084928115776445,
+      "grad_norm": 1.7561379918214899,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 25564
+    },
+    {
+      "epoch": 0.608516614300676,
+      "grad_norm": 1.9947675133299856,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 25565
+    },
+    {
+      "epoch": 0.6085404170237075,
+      "grad_norm": 5.020596272639757,
+      "learning_rate": 2e-05,
+      "loss": 1.1023,
+      "step": 25566
+    },
+    {
+      "epoch": 0.608564219746739,
+      "grad_norm": 4.523365934187167,
+      "learning_rate": 2e-05,
+      "loss": 0.6808,
+      "step": 25567
+    },
+    {
+      "epoch": 0.6085880224697705,
+      "grad_norm": 8.348918128145257,
+      "learning_rate": 2e-05,
+      "loss": 0.5057,
+      "step": 25568
+    },
+    {
+      "epoch": 0.6086118251928021,
+      "grad_norm": 6.568912973724922,
+      "learning_rate": 2e-05,
+      "loss": 1.5466,
+      "step": 25569
+    },
+    {
+      "epoch": 0.6086356279158336,
+      "grad_norm": 15.378993751452708,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6086594306388651,
+      "grad_norm": 2.1001853397465653,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 25571
+    },
+    {
+      "epoch": 0.6086832333618966,
+      "grad_norm": 0.7810175288664861,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 25572
+    },
+    {
+      "epoch": 0.6087070360849282,
+      "grad_norm": 3.7958573872076107,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 25573
+    },
+    {
+      "epoch": 0.6087308388079596,
+      "grad_norm": 1.5917137867195472,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 25574
+    },
+    {
+      "epoch": 0.6087546415309911,
+      "grad_norm": 0.7616839089840144,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 25575
+    },
+    {
+      "epoch": 0.6087784442540226,
+      "grad_norm": 0.2570722296524814,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 25576
+    },
+    {
+      "epoch": 0.6088022469770542,
+      "grad_norm": 4.64110341762463,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 25577
+    },
+    {
+      "epoch": 0.6088260497000857,
+      "grad_norm": 0.3252828886630209,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 25578
+    },
+    {
+      "epoch": 0.6088498524231172,
+      "grad_norm": 5.349891178418903,
+      "learning_rate": 2e-05,
+      "loss": 0.6052,
+      "step": 25579
+    },
+    {
+      "epoch": 0.6088736551461488,
+      "grad_norm": 4.104575719486432,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 25580
+    },
+    {
+      "epoch": 0.6088974578691803,
+      "grad_norm": 0.5514950710083716,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 25581
+    },
+    {
+      "epoch": 0.6089212605922117,
+      "grad_norm": 2.4090404890023804,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 25582
+    },
+    {
+      "epoch": 0.6089450633152432,
+      "grad_norm": 2.1878910129303955,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 25583
+    },
+    {
+      "epoch": 0.6089688660382748,
+      "grad_norm": 5.099181337077615,
+      "learning_rate": 2e-05,
+      "loss": 0.7058,
+      "step": 25584
+    },
+    {
+      "epoch": 0.6089926687613063,
+      "grad_norm": 8.034496573798336,
+      "learning_rate": 2e-05,
+      "loss": 0.9834,
+      "step": 25585
+    },
+    {
+      "epoch": 0.6090164714843378,
+      "grad_norm": 5.589984260221777,
+      "learning_rate": 2e-05,
+      "loss": 0.38,
+      "step": 25586
+    },
+    {
+      "epoch": 0.6090402742073693,
+      "grad_norm": 8.89739162636933,
+      "learning_rate": 2e-05,
+      "loss": 0.4728,
+      "step": 25587
+    },
+    {
+      "epoch": 0.6090640769304009,
+      "grad_norm": 3.9238591125934357,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 25588
+    },
+    {
+      "epoch": 0.6090878796534324,
+      "grad_norm": 0.22487359396512835,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 25589
+    },
+    {
+      "epoch": 0.6091116823764638,
+      "grad_norm": 5.031192742115005,
+      "learning_rate": 2e-05,
+      "loss": 0.4375,
+      "step": 25590
+    },
+    {
+      "epoch": 0.6091354850994953,
+      "grad_norm": 4.815653289903236,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 25591
+    },
+    {
+      "epoch": 0.6091592878225269,
+      "grad_norm": 4.865744492759039,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 25592
+    },
+    {
+      "epoch": 0.6091830905455584,
+      "grad_norm": 0.9704108219363524,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 25593
+    },
+    {
+      "epoch": 0.6092068932685899,
+      "grad_norm": 4.610668394502281,
+      "learning_rate": 2e-05,
+      "loss": 0.6246,
+      "step": 25594
+    },
+    {
+      "epoch": 0.6092306959916215,
+      "grad_norm": 5.868518400059496,
+      "learning_rate": 2e-05,
+      "loss": 0.1726,
+      "step": 25595
+    },
+    {
+      "epoch": 0.609254498714653,
+      "grad_norm": 1.7453439374804922,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 25596
+    },
+    {
+      "epoch": 0.6092783014376845,
+      "grad_norm": 17.64093299184972,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 25597
+    },
+    {
+      "epoch": 0.609302104160716,
+      "grad_norm": 6.435194901896358,
+      "learning_rate": 2e-05,
+      "loss": 1.3964,
+      "step": 25598
+    },
+    {
+      "epoch": 0.6093259068837475,
+      "grad_norm": 3.99382710101041,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 25599
+    },
+    {
+      "epoch": 0.609349709606779,
+      "grad_norm": 7.90022883914692,
+      "learning_rate": 2e-05,
+      "loss": 0.3544,
+      "step": 25600
+    },
+    {
+      "epoch": 0.6093735123298105,
+      "grad_norm": 5.152761523089955,
+      "learning_rate": 2e-05,
+      "loss": 0.7677,
+      "step": 25601
+    },
+    {
+      "epoch": 0.609397315052842,
+      "grad_norm": 5.993587894294966,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 25602
+    },
+    {
+      "epoch": 0.6094211177758736,
+      "grad_norm": 6.698447599528696,
+      "learning_rate": 2e-05,
+      "loss": 0.1565,
+      "step": 25603
+    },
+    {
+      "epoch": 0.6094449204989051,
+      "grad_norm": 1.5274319661566624,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 25604
+    },
+    {
+      "epoch": 0.6094687232219366,
+      "grad_norm": 0.13594587576262493,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 25605
+    },
+    {
+      "epoch": 0.609492525944968,
+      "grad_norm": 0.17533601711202015,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 25606
+    },
+    {
+      "epoch": 0.6095163286679997,
+      "grad_norm": 2.902580813933787,
+      "learning_rate": 2e-05,
+      "loss": 0.3959,
+      "step": 25607
+    },
+    {
+      "epoch": 0.6095401313910311,
+      "grad_norm": 4.898432991805166,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 25608
+    },
+    {
+      "epoch": 0.6095639341140626,
+      "grad_norm": 4.7519911415448055,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 25609
+    },
+    {
+      "epoch": 0.6095877368370942,
+      "grad_norm": 4.677149329626105,
+      "learning_rate": 2e-05,
+      "loss": 0.9859,
+      "step": 25610
+    },
+    {
+      "epoch": 0.6096115395601257,
+      "grad_norm": 2.8531573709823723,
+      "learning_rate": 2e-05,
+      "loss": 0.1354,
+      "step": 25611
+    },
+    {
+      "epoch": 0.6096353422831572,
+      "grad_norm": 2.1552572558050036,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 25612
+    },
+    {
+      "epoch": 0.6096591450061887,
+      "grad_norm": 10.912745972730496,
+      "learning_rate": 2e-05,
+      "loss": 0.5931,
+      "step": 25613
+    },
+    {
+      "epoch": 0.6096829477292203,
+      "grad_norm": 1.0972315189978812,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 25614
+    },
+    {
+      "epoch": 0.6097067504522518,
+      "grad_norm": 1.2926093675127457,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 25615
+    },
+    {
+      "epoch": 0.6097305531752832,
+      "grad_norm": 0.21728840714903963,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 25616
+    },
+    {
+      "epoch": 0.6097543558983147,
+      "grad_norm": 0.5100635199101283,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 25617
+    },
+    {
+      "epoch": 0.6097781586213463,
+      "grad_norm": 2.2023864601374155,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 25618
+    },
+    {
+      "epoch": 0.6098019613443778,
+      "grad_norm": 5.379542412555774,
+      "learning_rate": 2e-05,
+      "loss": 1.2367,
+      "step": 25619
+    },
+    {
+      "epoch": 0.6098257640674093,
+      "grad_norm": 6.33870098935973,
+      "learning_rate": 2e-05,
+      "loss": 1.5317,
+      "step": 25620
+    },
+    {
+      "epoch": 0.6098495667904408,
+      "grad_norm": 4.364560401434893,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 25621
+    },
+    {
+      "epoch": 0.6098733695134724,
+      "grad_norm": 13.512486241971741,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 25622
+    },
+    {
+      "epoch": 0.6098971722365039,
+      "grad_norm": 4.095685665133035,
+      "learning_rate": 2e-05,
+      "loss": 0.3085,
+      "step": 25623
+    },
+    {
+      "epoch": 0.6099209749595353,
+      "grad_norm": 6.935781704286362,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 25624
+    },
+    {
+      "epoch": 0.6099447776825669,
+      "grad_norm": 4.420288908626014,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 25625
+    },
+    {
+      "epoch": 0.6099685804055984,
+      "grad_norm": 0.6814434797518881,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 25626
+    },
+    {
+      "epoch": 0.6099923831286299,
+      "grad_norm": 0.36243129692478643,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 25627
+    },
+    {
+      "epoch": 0.6100161858516614,
+      "grad_norm": 9.110400589177274,
+      "learning_rate": 2e-05,
+      "loss": 2.3621,
+      "step": 25628
+    },
+    {
+      "epoch": 0.610039988574693,
+      "grad_norm": 2.7612054399591397,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 25629
+    },
+    {
+      "epoch": 0.6100637912977245,
+      "grad_norm": 2.4955719008882284,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 25630
+    },
+    {
+      "epoch": 0.610087594020756,
+      "grad_norm": 6.633381975515761,
+      "learning_rate": 2e-05,
+      "loss": 1.2711,
+      "step": 25631
+    },
+    {
+      "epoch": 0.6101113967437874,
+      "grad_norm": 0.7911184292636925,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 25632
+    },
+    {
+      "epoch": 0.610135199466819,
+      "grad_norm": 0.20294848981777924,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 25633
+    },
+    {
+      "epoch": 0.6101590021898505,
+      "grad_norm": 0.28760179114535733,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 25634
+    },
+    {
+      "epoch": 0.610182804912882,
+      "grad_norm": 0.07571230388607128,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 25635
+    },
+    {
+      "epoch": 0.6102066076359135,
+      "grad_norm": 7.480438243927955,
+      "learning_rate": 2e-05,
+      "loss": 1.4052,
+      "step": 25636
+    },
+    {
+      "epoch": 0.6102304103589451,
+      "grad_norm": 2.928671575480269,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 25637
+    },
+    {
+      "epoch": 0.6102542130819766,
+      "grad_norm": 6.392073063695924,
+      "learning_rate": 2e-05,
+      "loss": 0.3019,
+      "step": 25638
+    },
+    {
+      "epoch": 0.6102780158050081,
+      "grad_norm": 2.5128441981230427,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 25639
+    },
+    {
+      "epoch": 0.6103018185280396,
+      "grad_norm": 6.856310169198241,
+      "learning_rate": 2e-05,
+      "loss": 0.3788,
+      "step": 25640
+    },
+    {
+      "epoch": 0.6103256212510711,
+      "grad_norm": 0.1635170198262198,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 25641
+    },
+    {
+      "epoch": 0.6103494239741026,
+      "grad_norm": 3.0186848721701893,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 25642
+    },
+    {
+      "epoch": 0.6103732266971341,
+      "grad_norm": 1.8888442952117876,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 25643
+    },
+    {
+      "epoch": 0.6103970294201657,
+      "grad_norm": 11.8164476178464,
+      "learning_rate": 2e-05,
+      "loss": 0.3365,
+      "step": 25644
+    },
+    {
+      "epoch": 0.6104208321431972,
+      "grad_norm": 2.7476739627510236,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 25645
+    },
+    {
+      "epoch": 0.6104446348662287,
+      "grad_norm": 3.665878221154515,
+      "learning_rate": 2e-05,
+      "loss": 0.3909,
+      "step": 25646
+    },
+    {
+      "epoch": 0.6104684375892602,
+      "grad_norm": 3.5532156040310006,
+      "learning_rate": 2e-05,
+      "loss": 0.4314,
+      "step": 25647
+    },
+    {
+      "epoch": 0.6104922403122918,
+      "grad_norm": 11.931764693079753,
+      "learning_rate": 2e-05,
+      "loss": 1.1289,
+      "step": 25648
+    },
+    {
+      "epoch": 0.6105160430353233,
+      "grad_norm": 10.351668195970895,
+      "learning_rate": 2e-05,
+      "loss": 0.9258,
+      "step": 25649
+    },
+    {
+      "epoch": 0.6105398457583547,
+      "grad_norm": 5.026256509936718,
+      "learning_rate": 2e-05,
+      "loss": 0.6921,
+      "step": 25650
+    },
+    {
+      "epoch": 0.6105636484813862,
+      "grad_norm": 2.449773066630797,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 25651
+    },
+    {
+      "epoch": 0.6105874512044178,
+      "grad_norm": 0.8631939662895112,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 25652
+    },
+    {
+      "epoch": 0.6106112539274493,
+      "grad_norm": 1.6904050894700298,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 25653
+    },
+    {
+      "epoch": 0.6106350566504808,
+      "grad_norm": 3.9920474066170164,
+      "learning_rate": 2e-05,
+      "loss": 0.715,
+      "step": 25654
+    },
+    {
+      "epoch": 0.6106588593735123,
+      "grad_norm": 0.5144377886179112,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 25655
+    },
+    {
+      "epoch": 0.6106826620965439,
+      "grad_norm": 6.408746529416061,
+      "learning_rate": 2e-05,
+      "loss": 1.0136,
+      "step": 25656
+    },
+    {
+      "epoch": 0.6107064648195754,
+      "grad_norm": 3.7656321641433834,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 25657
+    },
+    {
+      "epoch": 0.6107302675426068,
+      "grad_norm": 2.8984517698735135,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 25658
+    },
+    {
+      "epoch": 0.6107540702656384,
+      "grad_norm": 3.9314655671347785,
+      "learning_rate": 2e-05,
+      "loss": 0.5215,
+      "step": 25659
+    },
+    {
+      "epoch": 0.6107778729886699,
+      "grad_norm": 7.818061951813476,
+      "learning_rate": 2e-05,
+      "loss": 1.032,
+      "step": 25660
+    },
+    {
+      "epoch": 0.6108016757117014,
+      "grad_norm": 4.36117829337288,
+      "learning_rate": 2e-05,
+      "loss": 0.2404,
+      "step": 25661
+    },
+    {
+      "epoch": 0.6108254784347329,
+      "grad_norm": 18.73696561343519,
+      "learning_rate": 2e-05,
+      "loss": 0.2979,
+      "step": 25662
+    },
+    {
+      "epoch": 0.6108492811577645,
+      "grad_norm": 3.5386397357120227,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 25663
+    },
+    {
+      "epoch": 0.610873083880796,
+      "grad_norm": 2.0006675696237948,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 25664
+    },
+    {
+      "epoch": 0.6108968866038275,
+      "grad_norm": 3.125820238644683,
+      "learning_rate": 2e-05,
+      "loss": 0.4756,
+      "step": 25665
+    },
+    {
+      "epoch": 0.610920689326859,
+      "grad_norm": 5.723268891166976,
+      "learning_rate": 2e-05,
+      "loss": 1.4938,
+      "step": 25666
+    },
+    {
+      "epoch": 0.6109444920498905,
+      "grad_norm": 4.139036319089499,
+      "learning_rate": 2e-05,
+      "loss": 0.4618,
+      "step": 25667
+    },
+    {
+      "epoch": 0.610968294772922,
+      "grad_norm": 2.0069383328137786,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 25668
+    },
+    {
+      "epoch": 0.6109920974959535,
+      "grad_norm": 2.698139854781896,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 25669
+    },
+    {
+      "epoch": 0.611015900218985,
+      "grad_norm": 7.097903849833028,
+      "learning_rate": 2e-05,
+      "loss": 0.4966,
+      "step": 25670
+    },
+    {
+      "epoch": 0.6110397029420166,
+      "grad_norm": 0.9279316379548764,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 25671
+    },
+    {
+      "epoch": 0.6110635056650481,
+      "grad_norm": 8.478190667570262,
+      "learning_rate": 2e-05,
+      "loss": 0.1706,
+      "step": 25672
+    },
+    {
+      "epoch": 0.6110873083880796,
+      "grad_norm": 9.165369221610344,
+      "learning_rate": 2e-05,
+      "loss": 2.1119,
+      "step": 25673
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 6.584050724152354,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 25674
+    },
+    {
+      "epoch": 0.6111349138341426,
+      "grad_norm": 4.0816406074987786,
+      "learning_rate": 2e-05,
+      "loss": 0.943,
+      "step": 25675
+    },
+    {
+      "epoch": 0.6111587165571741,
+      "grad_norm": 9.888948206370614,
+      "learning_rate": 2e-05,
+      "loss": 0.3823,
+      "step": 25676
+    },
+    {
+      "epoch": 0.6111825192802056,
+      "grad_norm": 2.1053496877741122,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 25677
+    },
+    {
+      "epoch": 0.6112063220032372,
+      "grad_norm": 3.1069568265383167,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 25678
+    },
+    {
+      "epoch": 0.6112301247262687,
+      "grad_norm": 0.6331337020641171,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 25679
+    },
+    {
+      "epoch": 0.6112539274493002,
+      "grad_norm": 3.5316073554024734,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 25680
+    },
+    {
+      "epoch": 0.6112777301723317,
+      "grad_norm": 8.612104000530069,
+      "learning_rate": 2e-05,
+      "loss": 0.2513,
+      "step": 25681
+    },
+    {
+      "epoch": 0.6113015328953633,
+      "grad_norm": 3.811322964326963,
+      "learning_rate": 2e-05,
+      "loss": 0.5265,
+      "step": 25682
+    },
+    {
+      "epoch": 0.6113253356183947,
+      "grad_norm": 0.9892813454483567,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 25683
+    },
+    {
+      "epoch": 0.6113491383414262,
+      "grad_norm": 0.19839164175065313,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 25684
+    },
+    {
+      "epoch": 0.6113729410644577,
+      "grad_norm": 1.6428829487659746,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 25685
+    },
+    {
+      "epoch": 0.6113967437874893,
+      "grad_norm": 4.701518673869111,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 25686
+    },
+    {
+      "epoch": 0.6114205465105208,
+      "grad_norm": 5.062936065792587,
+      "learning_rate": 2e-05,
+      "loss": 0.2787,
+      "step": 25687
+    },
+    {
+      "epoch": 0.6114443492335523,
+      "grad_norm": 4.590483463873298,
+      "learning_rate": 2e-05,
+      "loss": 0.6598,
+      "step": 25688
+    },
+    {
+      "epoch": 0.6114681519565839,
+      "grad_norm": 4.362184997706813,
+      "learning_rate": 2e-05,
+      "loss": 1.0064,
+      "step": 25689
+    },
+    {
+      "epoch": 0.6114919546796154,
+      "grad_norm": 11.679481864741152,
+      "learning_rate": 2e-05,
+      "loss": 0.6331,
+      "step": 25690
+    },
+    {
+      "epoch": 0.6115157574026469,
+      "grad_norm": 0.598738306291321,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 25691
+    },
+    {
+      "epoch": 0.6115395601256783,
+      "grad_norm": 20.702241211066088,
+      "learning_rate": 2e-05,
+      "loss": 0.4295,
+      "step": 25692
+    },
+    {
+      "epoch": 0.6115633628487099,
+      "grad_norm": 6.643604216909172,
+      "learning_rate": 2e-05,
+      "loss": 0.8662,
+      "step": 25693
+    },
+    {
+      "epoch": 0.6115871655717414,
+      "grad_norm": 3.552251165184338,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 25694
+    },
+    {
+      "epoch": 0.6116109682947729,
+      "grad_norm": 2.626172988333296,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 25695
+    },
+    {
+      "epoch": 0.6116347710178044,
+      "grad_norm": 4.304094717911553,
+      "learning_rate": 2e-05,
+      "loss": 0.1121,
+      "step": 25696
+    },
+    {
+      "epoch": 0.611658573740836,
+      "grad_norm": 2.791824029139113,
+      "learning_rate": 2e-05,
+      "loss": 0.456,
+      "step": 25697
+    },
+    {
+      "epoch": 0.6116823764638675,
+      "grad_norm": 9.42480394404986,
+      "learning_rate": 2e-05,
+      "loss": 0.6375,
+      "step": 25698
+    },
+    {
+      "epoch": 0.611706179186899,
+      "grad_norm": 41.053849868819235,
+      "learning_rate": 2e-05,
+      "loss": 0.223,
+      "step": 25699
+    },
+    {
+      "epoch": 0.6117299819099304,
+      "grad_norm": 2.5893868432331755,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 25700
+    },
+    {
+      "epoch": 0.611753784632962,
+      "grad_norm": 4.913455131572179,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 25701
+    },
+    {
+      "epoch": 0.6117775873559935,
+      "grad_norm": 4.759491949793523,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 25702
+    },
+    {
+      "epoch": 0.611801390079025,
+      "grad_norm": 7.41459349544518,
+      "learning_rate": 2e-05,
+      "loss": 1.17,
+      "step": 25703
+    },
+    {
+      "epoch": 0.6118251928020566,
+      "grad_norm": 4.382670992182473,
+      "learning_rate": 2e-05,
+      "loss": 0.7722,
+      "step": 25704
+    },
+    {
+      "epoch": 0.6118489955250881,
+      "grad_norm": 0.7415985088437858,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 25705
+    },
+    {
+      "epoch": 0.6118727982481196,
+      "grad_norm": 5.632594190882694,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 25706
+    },
+    {
+      "epoch": 0.6118966009711511,
+      "grad_norm": 6.04424977682525,
+      "learning_rate": 2e-05,
+      "loss": 0.7997,
+      "step": 25707
+    },
+    {
+      "epoch": 0.6119204036941827,
+      "grad_norm": 3.456700385465801,
+      "learning_rate": 2e-05,
+      "loss": 0.2022,
+      "step": 25708
+    },
+    {
+      "epoch": 0.6119442064172141,
+      "grad_norm": 1.7133012130963854,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 25709
+    },
+    {
+      "epoch": 0.6119680091402456,
+      "grad_norm": 10.151491910989577,
+      "learning_rate": 2e-05,
+      "loss": 2.3483,
+      "step": 25710
+    },
+    {
+      "epoch": 0.6119918118632771,
+      "grad_norm": 12.384193355769192,
+      "learning_rate": 2e-05,
+      "loss": 0.9271,
+      "step": 25711
+    },
+    {
+      "epoch": 0.6120156145863087,
+      "grad_norm": 0.3957803550897277,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 25712
+    },
+    {
+      "epoch": 0.6120394173093402,
+      "grad_norm": 1.4305087137465184,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 25713
+    },
+    {
+      "epoch": 0.6120632200323717,
+      "grad_norm": 2.9079512574835737,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 25714
+    },
+    {
+      "epoch": 0.6120870227554032,
+      "grad_norm": 0.10486390370095226,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 25715
+    },
+    {
+      "epoch": 0.6121108254784348,
+      "grad_norm": 2.7883410488894564,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 25716
+    },
+    {
+      "epoch": 0.6121346282014662,
+      "grad_norm": 0.8871188853219966,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 25717
+    },
+    {
+      "epoch": 0.6121584309244977,
+      "grad_norm": 6.353087749060865,
+      "learning_rate": 2e-05,
+      "loss": 1.0589,
+      "step": 25718
+    },
+    {
+      "epoch": 0.6121822336475293,
+      "grad_norm": 2.0612478523009403,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 25719
+    },
+    {
+      "epoch": 0.6122060363705608,
+      "grad_norm": 3.536954976048253,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 25720
+    },
+    {
+      "epoch": 0.6122298390935923,
+      "grad_norm": 6.814616134396159,
+      "learning_rate": 2e-05,
+      "loss": 1.1441,
+      "step": 25721
+    },
+    {
+      "epoch": 0.6122536418166238,
+      "grad_norm": 4.748824936857627,
+      "learning_rate": 2e-05,
+      "loss": 0.8681,
+      "step": 25722
+    },
+    {
+      "epoch": 0.6122774445396554,
+      "grad_norm": 1.5204926999738655,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 25723
+    },
+    {
+      "epoch": 0.6123012472626869,
+      "grad_norm": 3.0950725883392156,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 25724
+    },
+    {
+      "epoch": 0.6123250499857184,
+      "grad_norm": 14.315298460405886,
+      "learning_rate": 2e-05,
+      "loss": 1.0816,
+      "step": 25725
+    },
+    {
+      "epoch": 0.6123488527087498,
+      "grad_norm": 6.052241115411245,
+      "learning_rate": 2e-05,
+      "loss": 0.7521,
+      "step": 25726
+    },
+    {
+      "epoch": 0.6123726554317814,
+      "grad_norm": 0.10029203530155757,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 25727
+    },
+    {
+      "epoch": 0.6123964581548129,
+      "grad_norm": 5.610894362525535,
+      "learning_rate": 2e-05,
+      "loss": 0.6066,
+      "step": 25728
+    },
+    {
+      "epoch": 0.6124202608778444,
+      "grad_norm": 3.0636761663094045,
+      "learning_rate": 2e-05,
+      "loss": 0.646,
+      "step": 25729
+    },
+    {
+      "epoch": 0.6124440636008759,
+      "grad_norm": 2.736005575526522,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 25730
+    },
+    {
+      "epoch": 0.6124678663239075,
+      "grad_norm": 2.4206814159680627,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 25731
+    },
+    {
+      "epoch": 0.612491669046939,
+      "grad_norm": 0.9814126484860967,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 25732
+    },
+    {
+      "epoch": 0.6125154717699705,
+      "grad_norm": 2.5591591321799063,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 25733
+    },
+    {
+      "epoch": 0.612539274493002,
+      "grad_norm": 0.3672371442786537,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 25734
+    },
+    {
+      "epoch": 0.6125630772160335,
+      "grad_norm": 6.030057074583793,
+      "learning_rate": 2e-05,
+      "loss": 0.5048,
+      "step": 25735
+    },
+    {
+      "epoch": 0.612586879939065,
+      "grad_norm": 28.148268886089937,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 25736
+    },
+    {
+      "epoch": 0.6126106826620965,
+      "grad_norm": 5.139955126534517,
+      "learning_rate": 2e-05,
+      "loss": 0.9139,
+      "step": 25737
+    },
+    {
+      "epoch": 0.6126344853851281,
+      "grad_norm": 25.736556000337185,
+      "learning_rate": 2e-05,
+      "loss": 0.5187,
+      "step": 25738
+    },
+    {
+      "epoch": 0.6126582881081596,
+      "grad_norm": 11.57998084575571,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 25739
+    },
+    {
+      "epoch": 0.6126820908311911,
+      "grad_norm": 0.3241201956410768,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 25740
+    },
+    {
+      "epoch": 0.6127058935542226,
+      "grad_norm": 1.4207678811602944,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 25741
+    },
+    {
+      "epoch": 0.6127296962772542,
+      "grad_norm": 3.378882699710949,
+      "learning_rate": 2e-05,
+      "loss": 0.3291,
+      "step": 25742
+    },
+    {
+      "epoch": 0.6127534990002856,
+      "grad_norm": 6.276207582133666,
+      "learning_rate": 2e-05,
+      "loss": 0.1635,
+      "step": 25743
+    },
+    {
+      "epoch": 0.6127773017233171,
+      "grad_norm": 2.41347741482174,
+      "learning_rate": 2e-05,
+      "loss": 0.1374,
+      "step": 25744
+    },
+    {
+      "epoch": 0.6128011044463486,
+      "grad_norm": 25.44393008791512,
+      "learning_rate": 2e-05,
+      "loss": 0.6898,
+      "step": 25745
+    },
+    {
+      "epoch": 0.6128249071693802,
+      "grad_norm": 16.08707117022789,
+      "learning_rate": 2e-05,
+      "loss": 1.0335,
+      "step": 25746
+    },
+    {
+      "epoch": 0.6128487098924117,
+      "grad_norm": 4.679730831469177,
+      "learning_rate": 2e-05,
+      "loss": 0.8027,
+      "step": 25747
+    },
+    {
+      "epoch": 0.6128725126154432,
+      "grad_norm": 6.6311046860012866,
+      "learning_rate": 2e-05,
+      "loss": 0.6462,
+      "step": 25748
+    },
+    {
+      "epoch": 0.6128963153384748,
+      "grad_norm": 7.079006972554565,
+      "learning_rate": 2e-05,
+      "loss": 0.7874,
+      "step": 25749
+    },
+    {
+      "epoch": 0.6129201180615063,
+      "grad_norm": 1.8547251869258483,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 25750
+    },
+    {
+      "epoch": 0.6129439207845377,
+      "grad_norm": 3.7548659252538026,
+      "learning_rate": 2e-05,
+      "loss": 0.3576,
+      "step": 25751
+    },
+    {
+      "epoch": 0.6129677235075692,
+      "grad_norm": 0.2076871356747671,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 25752
+    },
+    {
+      "epoch": 0.6129915262306008,
+      "grad_norm": 3.5215822390380875,
+      "learning_rate": 2e-05,
+      "loss": 0.4441,
+      "step": 25753
+    },
+    {
+      "epoch": 0.6130153289536323,
+      "grad_norm": 3.753665816598073,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 25754
+    },
+    {
+      "epoch": 0.6130391316766638,
+      "grad_norm": 19.918400057186993,
+      "learning_rate": 2e-05,
+      "loss": 1.3037,
+      "step": 25755
+    },
+    {
+      "epoch": 0.6130629343996953,
+      "grad_norm": 3.6212573252627824,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 25756
+    },
+    {
+      "epoch": 0.6130867371227269,
+      "grad_norm": 10.816901729011633,
+      "learning_rate": 2e-05,
+      "loss": 1.1018,
+      "step": 25757
+    },
+    {
+      "epoch": 0.6131105398457584,
+      "grad_norm": 2.507422258924249,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 25758
+    },
+    {
+      "epoch": 0.6131343425687898,
+      "grad_norm": 0.8752955540443443,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 25759
+    },
+    {
+      "epoch": 0.6131581452918213,
+      "grad_norm": 5.6986925403966,
+      "learning_rate": 2e-05,
+      "loss": 0.2609,
+      "step": 25760
+    },
+    {
+      "epoch": 0.6131819480148529,
+      "grad_norm": 6.134454109533581,
+      "learning_rate": 2e-05,
+      "loss": 1.23,
+      "step": 25761
+    },
+    {
+      "epoch": 0.6132057507378844,
+      "grad_norm": 9.55541553431177,
+      "learning_rate": 2e-05,
+      "loss": 0.3183,
+      "step": 25762
+    },
+    {
+      "epoch": 0.6132295534609159,
+      "grad_norm": 2.445345428886483,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 25763
+    },
+    {
+      "epoch": 0.6132533561839475,
+      "grad_norm": 4.395768129444448,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 25764
+    },
+    {
+      "epoch": 0.613277158906979,
+      "grad_norm": 32.927638897202655,
+      "learning_rate": 2e-05,
+      "loss": 0.816,
+      "step": 25765
+    },
+    {
+      "epoch": 0.6133009616300105,
+      "grad_norm": 3.127035260269401,
+      "learning_rate": 2e-05,
+      "loss": 0.4615,
+      "step": 25766
+    },
+    {
+      "epoch": 0.613324764353042,
+      "grad_norm": 0.681759351537123,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 25767
+    },
+    {
+      "epoch": 0.6133485670760735,
+      "grad_norm": 1.1954247914700862,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 25768
+    },
+    {
+      "epoch": 0.613372369799105,
+      "grad_norm": 0.4384372504452821,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 25769
+    },
+    {
+      "epoch": 0.6133961725221365,
+      "grad_norm": 6.457088904703959,
+      "learning_rate": 2e-05,
+      "loss": 0.3065,
+      "step": 25770
+    },
+    {
+      "epoch": 0.613419975245168,
+      "grad_norm": 4.14371426533077,
+      "learning_rate": 2e-05,
+      "loss": 0.5525,
+      "step": 25771
+    },
+    {
+      "epoch": 0.6134437779681996,
+      "grad_norm": 1.2080144734168976,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 25772
+    },
+    {
+      "epoch": 0.6134675806912311,
+      "grad_norm": 9.19798064943787,
+      "learning_rate": 2e-05,
+      "loss": 1.3276,
+      "step": 25773
+    },
+    {
+      "epoch": 0.6134913834142626,
+      "grad_norm": 0.07954740339568898,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 25774
+    },
+    {
+      "epoch": 0.613515186137294,
+      "grad_norm": 0.1196886484042218,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 25775
+    },
+    {
+      "epoch": 0.6135389888603257,
+      "grad_norm": 2.1301892465900147,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 25776
+    },
+    {
+      "epoch": 0.6135627915833571,
+      "grad_norm": 1.7812487198455134,
+      "learning_rate": 2e-05,
+      "loss": 0.1821,
+      "step": 25777
+    },
+    {
+      "epoch": 0.6135865943063886,
+      "grad_norm": 1.5894869535966458,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 25778
+    },
+    {
+      "epoch": 0.6136103970294202,
+      "grad_norm": 5.886533341155983,
+      "learning_rate": 2e-05,
+      "loss": 0.7388,
+      "step": 25779
+    },
+    {
+      "epoch": 0.6136341997524517,
+      "grad_norm": 3.6074998683177717,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 25780
+    },
+    {
+      "epoch": 0.6136580024754832,
+      "grad_norm": 4.9245353328185,
+      "learning_rate": 2e-05,
+      "loss": 0.2626,
+      "step": 25781
+    },
+    {
+      "epoch": 0.6136818051985147,
+      "grad_norm": 11.610318862393168,
+      "learning_rate": 2e-05,
+      "loss": 1.0015,
+      "step": 25782
+    },
+    {
+      "epoch": 0.6137056079215463,
+      "grad_norm": 0.4548954603798845,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 25783
+    },
+    {
+      "epoch": 0.6137294106445778,
+      "grad_norm": 10.003054018665388,
+      "learning_rate": 2e-05,
+      "loss": 0.4857,
+      "step": 25784
+    },
+    {
+      "epoch": 0.6137532133676092,
+      "grad_norm": 0.27403289836895983,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 25785
+    },
+    {
+      "epoch": 0.6137770160906407,
+      "grad_norm": 16.23978532434877,
+      "learning_rate": 2e-05,
+      "loss": 0.7627,
+      "step": 25786
+    },
+    {
+      "epoch": 0.6138008188136723,
+      "grad_norm": 0.40394684336300973,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 25787
+    },
+    {
+      "epoch": 0.6138246215367038,
+      "grad_norm": 0.6342963234230544,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 25788
+    },
+    {
+      "epoch": 0.6138484242597353,
+      "grad_norm": 6.032771122598648,
+      "learning_rate": 2e-05,
+      "loss": 0.9375,
+      "step": 25789
+    },
+    {
+      "epoch": 0.6138722269827668,
+      "grad_norm": 11.226804932570102,
+      "learning_rate": 2e-05,
+      "loss": 0.4341,
+      "step": 25790
+    },
+    {
+      "epoch": 0.6138960297057984,
+      "grad_norm": 6.151216603023836,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 25791
+    },
+    {
+      "epoch": 0.6139198324288299,
+      "grad_norm": 5.8335380236703,
+      "learning_rate": 2e-05,
+      "loss": 0.2676,
+      "step": 25792
+    },
+    {
+      "epoch": 0.6139436351518613,
+      "grad_norm": 6.080386866436215,
+      "learning_rate": 2e-05,
+      "loss": 1.0854,
+      "step": 25793
+    },
+    {
+      "epoch": 0.6139674378748929,
+      "grad_norm": 5.266375414525462,
+      "learning_rate": 2e-05,
+      "loss": 0.8671,
+      "step": 25794
+    },
+    {
+      "epoch": 0.6139912405979244,
+      "grad_norm": 2.8568584853592256,
+      "learning_rate": 2e-05,
+      "loss": 0.4473,
+      "step": 25795
+    },
+    {
+      "epoch": 0.6140150433209559,
+      "grad_norm": 2.5927480420832785,
+      "learning_rate": 2e-05,
+      "loss": 0.1267,
+      "step": 25796
+    },
+    {
+      "epoch": 0.6140388460439874,
+      "grad_norm": 5.80467710094319,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 25797
+    },
+    {
+      "epoch": 0.614062648767019,
+      "grad_norm": 9.06076360637996,
+      "learning_rate": 2e-05,
+      "loss": 0.5091,
+      "step": 25798
+    },
+    {
+      "epoch": 0.6140864514900505,
+      "grad_norm": 3.59579550792276,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 25799
+    },
+    {
+      "epoch": 0.614110254213082,
+      "grad_norm": 0.4302584543764997,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 25800
+    },
+    {
+      "epoch": 0.6141340569361134,
+      "grad_norm": 27.614707263467917,
+      "learning_rate": 2e-05,
+      "loss": 0.2354,
+      "step": 25801
+    },
+    {
+      "epoch": 0.614157859659145,
+      "grad_norm": 5.134333273928601,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 25802
+    },
+    {
+      "epoch": 0.6141816623821765,
+      "grad_norm": 0.5886878534471115,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 25803
+    },
+    {
+      "epoch": 0.614205465105208,
+      "grad_norm": 0.12897963553814898,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 25804
+    },
+    {
+      "epoch": 0.6142292678282395,
+      "grad_norm": 6.284071332185316,
+      "learning_rate": 2e-05,
+      "loss": 1.4489,
+      "step": 25805
+    },
+    {
+      "epoch": 0.6142530705512711,
+      "grad_norm": 6.989962571326413,
+      "learning_rate": 2e-05,
+      "loss": 1.0598,
+      "step": 25806
+    },
+    {
+      "epoch": 0.6142768732743026,
+      "grad_norm": 5.058198807271293,
+      "learning_rate": 2e-05,
+      "loss": 0.33,
+      "step": 25807
+    },
+    {
+      "epoch": 0.6143006759973341,
+      "grad_norm": 2.15380838958213,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 25808
+    },
+    {
+      "epoch": 0.6143244787203657,
+      "grad_norm": 4.578758968105481,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 25809
+    },
+    {
+      "epoch": 0.6143482814433971,
+      "grad_norm": 5.153102036610845,
+      "learning_rate": 2e-05,
+      "loss": 1.342,
+      "step": 25810
+    },
+    {
+      "epoch": 0.6143720841664286,
+      "grad_norm": 3.205430044902039,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 25811
+    },
+    {
+      "epoch": 0.6143958868894601,
+      "grad_norm": 6.489689883661059,
+      "learning_rate": 2e-05,
+      "loss": 1.6651,
+      "step": 25812
+    },
+    {
+      "epoch": 0.6144196896124917,
+      "grad_norm": 2.152369055904245,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 25813
+    },
+    {
+      "epoch": 0.6144434923355232,
+      "grad_norm": 1.7305866875025695,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 25814
+    },
+    {
+      "epoch": 0.6144672950585547,
+      "grad_norm": 25.5667252856397,
+      "learning_rate": 2e-05,
+      "loss": 0.4487,
+      "step": 25815
+    },
+    {
+      "epoch": 0.6144910977815862,
+      "grad_norm": 3.3637825490658924,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 25816
+    },
+    {
+      "epoch": 0.6145149005046178,
+      "grad_norm": 4.795241317726662,
+      "learning_rate": 2e-05,
+      "loss": 0.5837,
+      "step": 25817
+    },
+    {
+      "epoch": 0.6145387032276493,
+      "grad_norm": 7.661553696709542,
+      "learning_rate": 2e-05,
+      "loss": 1.1424,
+      "step": 25818
+    },
+    {
+      "epoch": 0.6145625059506807,
+      "grad_norm": 4.649696419216187,
+      "learning_rate": 2e-05,
+      "loss": 1.1193,
+      "step": 25819
+    },
+    {
+      "epoch": 0.6145863086737122,
+      "grad_norm": 4.021397258315325,
+      "learning_rate": 2e-05,
+      "loss": 0.7915,
+      "step": 25820
+    },
+    {
+      "epoch": 0.6146101113967438,
+      "grad_norm": 2.1030429707595557,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 25821
+    },
+    {
+      "epoch": 0.6146339141197753,
+      "grad_norm": 4.52229178961088,
+      "learning_rate": 2e-05,
+      "loss": 1.003,
+      "step": 25822
+    },
+    {
+      "epoch": 0.6146577168428068,
+      "grad_norm": 3.403688505294174,
+      "learning_rate": 2e-05,
+      "loss": 0.4826,
+      "step": 25823
+    },
+    {
+      "epoch": 0.6146815195658384,
+      "grad_norm": 1.5907161979075894,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 25824
+    },
+    {
+      "epoch": 0.6147053222888699,
+      "grad_norm": 12.42748467477004,
+      "learning_rate": 2e-05,
+      "loss": 0.5027,
+      "step": 25825
+    },
+    {
+      "epoch": 0.6147291250119014,
+      "grad_norm": 3.3365388461782106,
+      "learning_rate": 2e-05,
+      "loss": 0.4218,
+      "step": 25826
+    },
+    {
+      "epoch": 0.6147529277349328,
+      "grad_norm": 4.773921664887103,
+      "learning_rate": 2e-05,
+      "loss": 0.8435,
+      "step": 25827
+    },
+    {
+      "epoch": 0.6147767304579644,
+      "grad_norm": 2.0798955889829656,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 25828
+    },
+    {
+      "epoch": 0.6148005331809959,
+      "grad_norm": 2.1941640614252496,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 25829
+    },
+    {
+      "epoch": 0.6148243359040274,
+      "grad_norm": 6.328550625965886,
+      "learning_rate": 2e-05,
+      "loss": 1.1867,
+      "step": 25830
+    },
+    {
+      "epoch": 0.6148481386270589,
+      "grad_norm": 9.330459221562442,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 25831
+    },
+    {
+      "epoch": 0.6148719413500905,
+      "grad_norm": 3.0977719933515067,
+      "learning_rate": 2e-05,
+      "loss": 0.4988,
+      "step": 25832
+    },
+    {
+      "epoch": 0.614895744073122,
+      "grad_norm": 4.992683373695947,
+      "learning_rate": 2e-05,
+      "loss": 1.0951,
+      "step": 25833
+    },
+    {
+      "epoch": 0.6149195467961535,
+      "grad_norm": 2.5572263385931833,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 25834
+    },
+    {
+      "epoch": 0.614943349519185,
+      "grad_norm": 3.1924187673541913,
+      "learning_rate": 2e-05,
+      "loss": 0.3553,
+      "step": 25835
+    },
+    {
+      "epoch": 0.6149671522422165,
+      "grad_norm": 3.436933761894755,
+      "learning_rate": 2e-05,
+      "loss": 0.7096,
+      "step": 25836
+    },
+    {
+      "epoch": 0.614990954965248,
+      "grad_norm": 1.5121249550985714,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 25837
+    },
+    {
+      "epoch": 0.6150147576882795,
+      "grad_norm": 0.32576504185978367,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 25838
+    },
+    {
+      "epoch": 0.6150385604113111,
+      "grad_norm": 3.540500513359347,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 25839
+    },
+    {
+      "epoch": 0.6150623631343426,
+      "grad_norm": 3.7858731815476316,
+      "learning_rate": 2e-05,
+      "loss": 0.7841,
+      "step": 25840
+    },
+    {
+      "epoch": 0.6150861658573741,
+      "grad_norm": 0.18803272968587414,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 25841
+    },
+    {
+      "epoch": 0.6151099685804056,
+      "grad_norm": 4.387555616161562,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 25842
+    },
+    {
+      "epoch": 0.6151337713034372,
+      "grad_norm": 0.8443617732821791,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 25843
+    },
+    {
+      "epoch": 0.6151575740264686,
+      "grad_norm": 0.3600536434468711,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 25844
+    },
+    {
+      "epoch": 0.6151813767495001,
+      "grad_norm": 7.246760457121364,
+      "learning_rate": 2e-05,
+      "loss": 0.9603,
+      "step": 25845
+    },
+    {
+      "epoch": 0.6152051794725316,
+      "grad_norm": 4.665970533683552,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 25846
+    },
+    {
+      "epoch": 0.6152289821955632,
+      "grad_norm": 4.196327013634786,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 25847
+    },
+    {
+      "epoch": 0.6152527849185947,
+      "grad_norm": 5.34238111481017,
+      "learning_rate": 2e-05,
+      "loss": 0.2289,
+      "step": 25848
+    },
+    {
+      "epoch": 0.6152765876416262,
+      "grad_norm": 5.368130266688346,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 25849
+    },
+    {
+      "epoch": 0.6153003903646577,
+      "grad_norm": 0.6439781707052842,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 25850
+    },
+    {
+      "epoch": 0.6153241930876893,
+      "grad_norm": 8.619829276864982,
+      "learning_rate": 2e-05,
+      "loss": 0.5305,
+      "step": 25851
+    },
+    {
+      "epoch": 0.6153479958107207,
+      "grad_norm": 1.6471653551484642,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 25852
+    },
+    {
+      "epoch": 0.6153717985337522,
+      "grad_norm": 0.7382125009397079,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 25853
+    },
+    {
+      "epoch": 0.6153956012567838,
+      "grad_norm": 0.23012047419121062,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 25854
+    },
+    {
+      "epoch": 0.6154194039798153,
+      "grad_norm": 5.855677516938742,
+      "learning_rate": 2e-05,
+      "loss": 0.5468,
+      "step": 25855
+    },
+    {
+      "epoch": 0.6154432067028468,
+      "grad_norm": 0.6838726859358862,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 25856
+    },
+    {
+      "epoch": 0.6154670094258783,
+      "grad_norm": 2.6687026864182113,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 25857
+    },
+    {
+      "epoch": 0.6154908121489099,
+      "grad_norm": 0.35741966187339835,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 25858
+    },
+    {
+      "epoch": 0.6155146148719414,
+      "grad_norm": 2.231449600289437,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 25859
+    },
+    {
+      "epoch": 0.6155384175949729,
+      "grad_norm": 3.0401402462181197,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 25860
+    },
+    {
+      "epoch": 0.6155622203180043,
+      "grad_norm": 2.3118338159741976,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 25861
+    },
+    {
+      "epoch": 0.6155860230410359,
+      "grad_norm": 4.548223400466826,
+      "learning_rate": 2e-05,
+      "loss": 0.4013,
+      "step": 25862
+    },
+    {
+      "epoch": 0.6156098257640674,
+      "grad_norm": 14.639256871829522,
+      "learning_rate": 2e-05,
+      "loss": 0.4885,
+      "step": 25863
+    },
+    {
+      "epoch": 0.6156336284870989,
+      "grad_norm": 9.491882750670982,
+      "learning_rate": 2e-05,
+      "loss": 0.3999,
+      "step": 25864
+    },
+    {
+      "epoch": 0.6156574312101304,
+      "grad_norm": 12.020954699228595,
+      "learning_rate": 2e-05,
+      "loss": 0.3663,
+      "step": 25865
+    },
+    {
+      "epoch": 0.615681233933162,
+      "grad_norm": 8.403798133903475,
+      "learning_rate": 2e-05,
+      "loss": 0.863,
+      "step": 25866
+    },
+    {
+      "epoch": 0.6157050366561935,
+      "grad_norm": 5.679865382207297,
+      "learning_rate": 2e-05,
+      "loss": 0.7389,
+      "step": 25867
+    },
+    {
+      "epoch": 0.615728839379225,
+      "grad_norm": 3.694702094129135,
+      "learning_rate": 2e-05,
+      "loss": 0.5025,
+      "step": 25868
+    },
+    {
+      "epoch": 0.6157526421022566,
+      "grad_norm": 7.761425926976443,
+      "learning_rate": 2e-05,
+      "loss": 1.0252,
+      "step": 25869
+    },
+    {
+      "epoch": 0.615776444825288,
+      "grad_norm": 3.4016499383742866,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 25870
+    },
+    {
+      "epoch": 0.6158002475483195,
+      "grad_norm": 12.749769828087892,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 25871
+    },
+    {
+      "epoch": 0.615824050271351,
+      "grad_norm": 3.3486972915769018,
+      "learning_rate": 2e-05,
+      "loss": 0.3796,
+      "step": 25872
+    },
+    {
+      "epoch": 0.6158478529943826,
+      "grad_norm": 0.6919173391464591,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 25873
+    },
+    {
+      "epoch": 0.6158716557174141,
+      "grad_norm": 4.359559806613751,
+      "learning_rate": 2e-05,
+      "loss": 0.421,
+      "step": 25874
+    },
+    {
+      "epoch": 0.6158954584404456,
+      "grad_norm": 3.986556497374916,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 25875
+    },
+    {
+      "epoch": 0.6159192611634771,
+      "grad_norm": 8.233202508174754,
+      "learning_rate": 2e-05,
+      "loss": 1.1515,
+      "step": 25876
+    },
+    {
+      "epoch": 0.6159430638865087,
+      "grad_norm": 1.2433932958801563,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 25877
+    },
+    {
+      "epoch": 0.6159668666095401,
+      "grad_norm": 1.9037517310252678,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 25878
+    },
+    {
+      "epoch": 0.6159906693325716,
+      "grad_norm": 0.12422933252668886,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 25879
+    },
+    {
+      "epoch": 0.6160144720556031,
+      "grad_norm": 0.5080650453891983,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 25880
+    },
+    {
+      "epoch": 0.6160382747786347,
+      "grad_norm": 4.475437197584475,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 25881
+    },
+    {
+      "epoch": 0.6160620775016662,
+      "grad_norm": 4.105685199469414,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 25882
+    },
+    {
+      "epoch": 0.6160858802246977,
+      "grad_norm": 1.3832340089324908,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 25883
+    },
+    {
+      "epoch": 0.6161096829477292,
+      "grad_norm": 3.6175422616973205,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 25884
+    },
+    {
+      "epoch": 0.6161334856707608,
+      "grad_norm": 4.164700460001898,
+      "learning_rate": 2e-05,
+      "loss": 0.5365,
+      "step": 25885
+    },
+    {
+      "epoch": 0.6161572883937922,
+      "grad_norm": 6.989205227053679,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 25886
+    },
+    {
+      "epoch": 0.6161810911168237,
+      "grad_norm": 41.09125141818919,
+      "learning_rate": 2e-05,
+      "loss": 0.5354,
+      "step": 25887
+    },
+    {
+      "epoch": 0.6162048938398553,
+      "grad_norm": 3.931562776130839,
+      "learning_rate": 2e-05,
+      "loss": 0.368,
+      "step": 25888
+    },
+    {
+      "epoch": 0.6162286965628868,
+      "grad_norm": 23.86795974739112,
+      "learning_rate": 2e-05,
+      "loss": 0.7257,
+      "step": 25889
+    },
+    {
+      "epoch": 0.6162524992859183,
+      "grad_norm": 9.364494943262999,
+      "learning_rate": 2e-05,
+      "loss": 0.3633,
+      "step": 25890
+    },
+    {
+      "epoch": 0.6162763020089498,
+      "grad_norm": 10.510181260570562,
+      "learning_rate": 2e-05,
+      "loss": 0.4893,
+      "step": 25891
+    },
+    {
+      "epoch": 0.6163001047319814,
+      "grad_norm": 10.688919102907924,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 25892
+    },
+    {
+      "epoch": 0.6163239074550129,
+      "grad_norm": 8.556241805169751,
+      "learning_rate": 2e-05,
+      "loss": 1.5402,
+      "step": 25893
+    },
+    {
+      "epoch": 0.6163477101780444,
+      "grad_norm": 0.16437849721303194,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 25894
+    },
+    {
+      "epoch": 0.6163715129010758,
+      "grad_norm": 5.853768556412706,
+      "learning_rate": 2e-05,
+      "loss": 0.503,
+      "step": 25895
+    },
+    {
+      "epoch": 0.6163953156241074,
+      "grad_norm": 3.554798036427388,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 25896
+    },
+    {
+      "epoch": 0.6164191183471389,
+      "grad_norm": 2.743948056184417,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 25897
+    },
+    {
+      "epoch": 0.6164429210701704,
+      "grad_norm": 2.450415220325102,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 25898
+    },
+    {
+      "epoch": 0.6164667237932019,
+      "grad_norm": 2.223048679421151,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 25899
+    },
+    {
+      "epoch": 0.6164905265162335,
+      "grad_norm": 1.0632917852245853,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 25900
+    },
+    {
+      "epoch": 0.616514329239265,
+      "grad_norm": 6.486953884058258,
+      "learning_rate": 2e-05,
+      "loss": 0.1008,
+      "step": 25901
+    },
+    {
+      "epoch": 0.6165381319622965,
+      "grad_norm": 5.815457264226449,
+      "learning_rate": 2e-05,
+      "loss": 0.691,
+      "step": 25902
+    },
+    {
+      "epoch": 0.616561934685328,
+      "grad_norm": 4.639384348232056,
+      "learning_rate": 2e-05,
+      "loss": 0.6094,
+      "step": 25903
+    },
+    {
+      "epoch": 0.6165857374083595,
+      "grad_norm": 6.221454962450513,
+      "learning_rate": 2e-05,
+      "loss": 1.0065,
+      "step": 25904
+    },
+    {
+      "epoch": 0.616609540131391,
+      "grad_norm": 2.3239582916905612,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 25905
+    },
+    {
+      "epoch": 0.6166333428544225,
+      "grad_norm": 0.6179749665261746,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 25906
+    },
+    {
+      "epoch": 0.6166571455774541,
+      "grad_norm": 18.46878878156861,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 25907
+    },
+    {
+      "epoch": 0.6166809483004856,
+      "grad_norm": 7.062502517827503,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 25908
+    },
+    {
+      "epoch": 0.6167047510235171,
+      "grad_norm": 3.679268239039882,
+      "learning_rate": 2e-05,
+      "loss": 0.2335,
+      "step": 25909
+    },
+    {
+      "epoch": 0.6167285537465486,
+      "grad_norm": 4.375014937436517,
+      "learning_rate": 2e-05,
+      "loss": 0.852,
+      "step": 25910
+    },
+    {
+      "epoch": 0.6167523564695802,
+      "grad_norm": 3.1972915918184253,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 25911
+    },
+    {
+      "epoch": 0.6167761591926116,
+      "grad_norm": 8.061101977334161,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 25912
+    },
+    {
+      "epoch": 0.6167999619156431,
+      "grad_norm": 3.204069586228611,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 25913
+    },
+    {
+      "epoch": 0.6168237646386746,
+      "grad_norm": 1.769948516413749,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 25914
+    },
+    {
+      "epoch": 0.6168475673617062,
+      "grad_norm": 5.398498498751964,
+      "learning_rate": 2e-05,
+      "loss": 1.0082,
+      "step": 25915
+    },
+    {
+      "epoch": 0.6168713700847377,
+      "grad_norm": 5.183072408632717,
+      "learning_rate": 2e-05,
+      "loss": 1.3927,
+      "step": 25916
+    },
+    {
+      "epoch": 0.6168951728077692,
+      "grad_norm": 7.703096026145013,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 25917
+    },
+    {
+      "epoch": 0.6169189755308008,
+      "grad_norm": 2.6460148834822115,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 25918
+    },
+    {
+      "epoch": 0.6169427782538323,
+      "grad_norm": 3.2186580558800895,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 25919
+    },
+    {
+      "epoch": 0.6169665809768637,
+      "grad_norm": 1.6112405323763375,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 25920
+    },
+    {
+      "epoch": 0.6169903836998952,
+      "grad_norm": 7.455120054148624,
+      "learning_rate": 2e-05,
+      "loss": 1.4405,
+      "step": 25921
+    },
+    {
+      "epoch": 0.6170141864229268,
+      "grad_norm": 2.8621828186079186,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 25922
+    },
+    {
+      "epoch": 0.6170379891459583,
+      "grad_norm": 1.7562652463232389,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 25923
+    },
+    {
+      "epoch": 0.6170617918689898,
+      "grad_norm": 0.01698670254099758,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 25924
+    },
+    {
+      "epoch": 0.6170855945920213,
+      "grad_norm": 9.854632068911286,
+      "learning_rate": 2e-05,
+      "loss": 0.2903,
+      "step": 25925
+    },
+    {
+      "epoch": 0.6171093973150529,
+      "grad_norm": 5.837439595183355,
+      "learning_rate": 2e-05,
+      "loss": 0.7112,
+      "step": 25926
+    },
+    {
+      "epoch": 0.6171332000380844,
+      "grad_norm": 0.03169097687758116,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 25927
+    },
+    {
+      "epoch": 0.6171570027611158,
+      "grad_norm": 7.917321589829206,
+      "learning_rate": 2e-05,
+      "loss": 0.1875,
+      "step": 25928
+    },
+    {
+      "epoch": 0.6171808054841473,
+      "grad_norm": 0.10135670579547512,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 25929
+    },
+    {
+      "epoch": 0.6172046082071789,
+      "grad_norm": 3.8253459941473187,
+      "learning_rate": 2e-05,
+      "loss": 0.5564,
+      "step": 25930
+    },
+    {
+      "epoch": 0.6172284109302104,
+      "grad_norm": 3.9539791794515264,
+      "learning_rate": 2e-05,
+      "loss": 0.5111,
+      "step": 25931
+    },
+    {
+      "epoch": 0.6172522136532419,
+      "grad_norm": 0.3260935083275329,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 25932
+    },
+    {
+      "epoch": 0.6172760163762735,
+      "grad_norm": 4.988099166492537,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 25933
+    },
+    {
+      "epoch": 0.617299819099305,
+      "grad_norm": 3.1139579802322737,
+      "learning_rate": 2e-05,
+      "loss": 0.4659,
+      "step": 25934
+    },
+    {
+      "epoch": 0.6173236218223365,
+      "grad_norm": 0.2418283473538727,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 25935
+    },
+    {
+      "epoch": 0.617347424545368,
+      "grad_norm": 5.200351251717395,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 25936
+    },
+    {
+      "epoch": 0.6173712272683995,
+      "grad_norm": 5.2248211578051595,
+      "learning_rate": 2e-05,
+      "loss": 0.7891,
+      "step": 25937
+    },
+    {
+      "epoch": 0.617395029991431,
+      "grad_norm": 11.343176879802009,
+      "learning_rate": 2e-05,
+      "loss": 0.469,
+      "step": 25938
+    },
+    {
+      "epoch": 0.6174188327144625,
+      "grad_norm": 10.432866212349742,
+      "learning_rate": 2e-05,
+      "loss": 1.6217,
+      "step": 25939
+    },
+    {
+      "epoch": 0.617442635437494,
+      "grad_norm": 1.5988448408624103,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 25940
+    },
+    {
+      "epoch": 0.6174664381605256,
+      "grad_norm": 2.5616361599243693,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 25941
+    },
+    {
+      "epoch": 0.6174902408835571,
+      "grad_norm": 6.84700156892804,
+      "learning_rate": 2e-05,
+      "loss": 1.0133,
+      "step": 25942
+    },
+    {
+      "epoch": 0.6175140436065886,
+      "grad_norm": 6.550382540730409,
+      "learning_rate": 2e-05,
+      "loss": 0.1907,
+      "step": 25943
+    },
+    {
+      "epoch": 0.61753784632962,
+      "grad_norm": 6.409187819210833,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 25944
+    },
+    {
+      "epoch": 0.6175616490526517,
+      "grad_norm": 1.4037389967673772,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 25945
+    },
+    {
+      "epoch": 0.6175854517756831,
+      "grad_norm": 0.02226871890914437,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 25946
+    },
+    {
+      "epoch": 0.6176092544987146,
+      "grad_norm": 0.17622406734131751,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 25947
+    },
+    {
+      "epoch": 0.6176330572217462,
+      "grad_norm": 3.4453357631772312,
+      "learning_rate": 2e-05,
+      "loss": 0.6484,
+      "step": 25948
+    },
+    {
+      "epoch": 0.6176568599447777,
+      "grad_norm": 4.198711182640632,
+      "learning_rate": 2e-05,
+      "loss": 0.4488,
+      "step": 25949
+    },
+    {
+      "epoch": 0.6176806626678092,
+      "grad_norm": 4.697960648014932,
+      "learning_rate": 2e-05,
+      "loss": 0.4454,
+      "step": 25950
+    },
+    {
+      "epoch": 0.6177044653908407,
+      "grad_norm": 16.63913027385899,
+      "learning_rate": 2e-05,
+      "loss": 0.2422,
+      "step": 25951
+    },
+    {
+      "epoch": 0.6177282681138723,
+      "grad_norm": 2.2823872399177607,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 25952
+    },
+    {
+      "epoch": 0.6177520708369038,
+      "grad_norm": 1.643577783386523,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 25953
+    },
+    {
+      "epoch": 0.6177758735599352,
+      "grad_norm": 2.032637802615013,
+      "learning_rate": 2e-05,
+      "loss": 0.0926,
+      "step": 25954
+    },
+    {
+      "epoch": 0.6177996762829667,
+      "grad_norm": 0.6120524755210834,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 25955
+    },
+    {
+      "epoch": 0.6178234790059983,
+      "grad_norm": 5.123244090197912,
+      "learning_rate": 2e-05,
+      "loss": 1.1621,
+      "step": 25956
+    },
+    {
+      "epoch": 0.6178472817290298,
+      "grad_norm": 1.713601795728142,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 25957
+    },
+    {
+      "epoch": 0.6178710844520613,
+      "grad_norm": 3.33528204910711,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 25958
+    },
+    {
+      "epoch": 0.6178948871750928,
+      "grad_norm": 5.996988515658253,
+      "learning_rate": 2e-05,
+      "loss": 0.268,
+      "step": 25959
+    },
+    {
+      "epoch": 0.6179186898981244,
+      "grad_norm": 0.6374071179457513,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 25960
+    },
+    {
+      "epoch": 0.6179424926211559,
+      "grad_norm": 0.878328519144814,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 25961
+    },
+    {
+      "epoch": 0.6179662953441873,
+      "grad_norm": 2.3495358035227842,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 25962
+    },
+    {
+      "epoch": 0.6179900980672189,
+      "grad_norm": 2.397748446180589,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 25963
+    },
+    {
+      "epoch": 0.6180139007902504,
+      "grad_norm": 15.247534550012826,
+      "learning_rate": 2e-05,
+      "loss": 0.8274,
+      "step": 25964
+    },
+    {
+      "epoch": 0.6180377035132819,
+      "grad_norm": 8.636702061854917,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 25965
+    },
+    {
+      "epoch": 0.6180615062363134,
+      "grad_norm": 0.7933503090564005,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 25966
+    },
+    {
+      "epoch": 0.618085308959345,
+      "grad_norm": 6.186825517340292,
+      "learning_rate": 2e-05,
+      "loss": 1.2227,
+      "step": 25967
+    },
+    {
+      "epoch": 0.6181091116823765,
+      "grad_norm": 1.694376755994197,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 25968
+    },
+    {
+      "epoch": 0.618132914405408,
+      "grad_norm": 4.562784341765203,
+      "learning_rate": 2e-05,
+      "loss": 0.1807,
+      "step": 25969
+    },
+    {
+      "epoch": 0.6181567171284394,
+      "grad_norm": 7.345497496436415,
+      "learning_rate": 2e-05,
+      "loss": 1.0884,
+      "step": 25970
+    },
+    {
+      "epoch": 0.618180519851471,
+      "grad_norm": 1.98439326693964,
+      "learning_rate": 2e-05,
+      "loss": 0.118,
+      "step": 25971
+    },
+    {
+      "epoch": 0.6182043225745025,
+      "grad_norm": 0.23043026304282835,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 25972
+    },
+    {
+      "epoch": 0.618228125297534,
+      "grad_norm": 2.878239219380623,
+      "learning_rate": 2e-05,
+      "loss": 0.4311,
+      "step": 25973
+    },
+    {
+      "epoch": 0.6182519280205655,
+      "grad_norm": 0.4528609811009444,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 25974
+    },
+    {
+      "epoch": 0.6182757307435971,
+      "grad_norm": 4.848251554528171,
+      "learning_rate": 2e-05,
+      "loss": 0.4862,
+      "step": 25975
+    },
+    {
+      "epoch": 0.6182995334666286,
+      "grad_norm": 3.6459949850149345,
+      "learning_rate": 2e-05,
+      "loss": 0.5656,
+      "step": 25976
+    },
+    {
+      "epoch": 0.6183233361896601,
+      "grad_norm": 1.0794145381096112,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 25977
+    },
+    {
+      "epoch": 0.6183471389126917,
+      "grad_norm": 2.735824402952102,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 25978
+    },
+    {
+      "epoch": 0.6183709416357231,
+      "grad_norm": 3.6057394568398946,
+      "learning_rate": 2e-05,
+      "loss": 0.6586,
+      "step": 25979
+    },
+    {
+      "epoch": 0.6183947443587546,
+      "grad_norm": 6.360429268893766,
+      "learning_rate": 2e-05,
+      "loss": 0.7241,
+      "step": 25980
+    },
+    {
+      "epoch": 0.6184185470817861,
+      "grad_norm": 2.9590127378632674,
+      "learning_rate": 2e-05,
+      "loss": 0.5966,
+      "step": 25981
+    },
+    {
+      "epoch": 0.6184423498048177,
+      "grad_norm": 0.659274366594803,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 25982
+    },
+    {
+      "epoch": 0.6184661525278492,
+      "grad_norm": 2.586358227482968,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 25983
+    },
+    {
+      "epoch": 0.6184899552508807,
+      "grad_norm": 2.5927349805213145,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 25984
+    },
+    {
+      "epoch": 0.6185137579739122,
+      "grad_norm": 12.626412009792718,
+      "learning_rate": 2e-05,
+      "loss": 1.6786,
+      "step": 25985
+    },
+    {
+      "epoch": 0.6185375606969438,
+      "grad_norm": 19.45420827003802,
+      "learning_rate": 2e-05,
+      "loss": 0.4599,
+      "step": 25986
+    },
+    {
+      "epoch": 0.6185613634199753,
+      "grad_norm": 4.331322902814198,
+      "learning_rate": 2e-05,
+      "loss": 0.5001,
+      "step": 25987
+    },
+    {
+      "epoch": 0.6185851661430067,
+      "grad_norm": 7.993814108476431,
+      "learning_rate": 2e-05,
+      "loss": 0.4129,
+      "step": 25988
+    },
+    {
+      "epoch": 0.6186089688660382,
+      "grad_norm": 18.368286714794955,
+      "learning_rate": 2e-05,
+      "loss": 1.0031,
+      "step": 25989
+    },
+    {
+      "epoch": 0.6186327715890698,
+      "grad_norm": 1.105263315019631,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 25990
+    },
+    {
+      "epoch": 0.6186565743121013,
+      "grad_norm": 0.6943181378705073,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 25991
+    },
+    {
+      "epoch": 0.6186803770351328,
+      "grad_norm": 7.381807048261138,
+      "learning_rate": 2e-05,
+      "loss": 1.0816,
+      "step": 25992
+    },
+    {
+      "epoch": 0.6187041797581644,
+      "grad_norm": 3.9300929083166216,
+      "learning_rate": 2e-05,
+      "loss": 0.2137,
+      "step": 25993
+    },
+    {
+      "epoch": 0.6187279824811959,
+      "grad_norm": 3.098439983241175,
+      "learning_rate": 2e-05,
+      "loss": 0.3381,
+      "step": 25994
+    },
+    {
+      "epoch": 0.6187517852042274,
+      "grad_norm": 4.375170145414566,
+      "learning_rate": 2e-05,
+      "loss": 0.3323,
+      "step": 25995
+    },
+    {
+      "epoch": 0.6187755879272588,
+      "grad_norm": 12.42742293204209,
+      "learning_rate": 2e-05,
+      "loss": 0.7917,
+      "step": 25996
+    },
+    {
+      "epoch": 0.6187993906502904,
+      "grad_norm": 0.3414665984504336,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 25997
+    },
+    {
+      "epoch": 0.6188231933733219,
+      "grad_norm": 2.6874125319326136,
+      "learning_rate": 2e-05,
+      "loss": 0.427,
+      "step": 25998
+    },
+    {
+      "epoch": 0.6188469960963534,
+      "grad_norm": 5.8642897066582815,
+      "learning_rate": 2e-05,
+      "loss": 0.9747,
+      "step": 25999
+    },
+    {
+      "epoch": 0.6188707988193849,
+      "grad_norm": 0.9288943752528078,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 26000
+    },
+    {
+      "epoch": 0.6188946015424165,
+      "grad_norm": 1.732692941300765,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 26001
+    },
+    {
+      "epoch": 0.618918404265448,
+      "grad_norm": 3.495363521996492,
+      "learning_rate": 2e-05,
+      "loss": 0.7159,
+      "step": 26002
+    },
+    {
+      "epoch": 0.6189422069884795,
+      "grad_norm": 1.271912040638748,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 26003
+    },
+    {
+      "epoch": 0.618966009711511,
+      "grad_norm": 5.90665325589698,
+      "learning_rate": 2e-05,
+      "loss": 0.8352,
+      "step": 26004
+    },
+    {
+      "epoch": 0.6189898124345425,
+      "grad_norm": 4.783398356385492,
+      "learning_rate": 2e-05,
+      "loss": 0.4314,
+      "step": 26005
+    },
+    {
+      "epoch": 0.619013615157574,
+      "grad_norm": 4.4237769510553075,
+      "learning_rate": 2e-05,
+      "loss": 0.7284,
+      "step": 26006
+    },
+    {
+      "epoch": 0.6190374178806055,
+      "grad_norm": 5.36350429354412,
+      "learning_rate": 2e-05,
+      "loss": 0.9783,
+      "step": 26007
+    },
+    {
+      "epoch": 0.6190612206036371,
+      "grad_norm": 0.035180712692275636,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 26008
+    },
+    {
+      "epoch": 0.6190850233266686,
+      "grad_norm": 1.8423837876741167,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 26009
+    },
+    {
+      "epoch": 0.6191088260497001,
+      "grad_norm": 3.5735512993721192,
+      "learning_rate": 2e-05,
+      "loss": 0.3515,
+      "step": 26010
+    },
+    {
+      "epoch": 0.6191326287727316,
+      "grad_norm": 3.1181476633919583,
+      "learning_rate": 2e-05,
+      "loss": 0.4293,
+      "step": 26011
+    },
+    {
+      "epoch": 0.6191564314957632,
+      "grad_norm": 11.51435729571919,
+      "learning_rate": 2e-05,
+      "loss": 1.5106,
+      "step": 26012
+    },
+    {
+      "epoch": 0.6191802342187946,
+      "grad_norm": 6.0344319045238715,
+      "learning_rate": 2e-05,
+      "loss": 0.3721,
+      "step": 26013
+    },
+    {
+      "epoch": 0.6192040369418261,
+      "grad_norm": 5.087897534054902,
+      "learning_rate": 2e-05,
+      "loss": 0.5081,
+      "step": 26014
+    },
+    {
+      "epoch": 0.6192278396648576,
+      "grad_norm": 5.760053630750354,
+      "learning_rate": 2e-05,
+      "loss": 0.942,
+      "step": 26015
+    },
+    {
+      "epoch": 0.6192516423878892,
+      "grad_norm": 9.512528457622055,
+      "learning_rate": 2e-05,
+      "loss": 0.7108,
+      "step": 26016
+    },
+    {
+      "epoch": 0.6192754451109207,
+      "grad_norm": 2.747326098743956,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 26017
+    },
+    {
+      "epoch": 0.6192992478339522,
+      "grad_norm": 9.653992085476848,
+      "learning_rate": 2e-05,
+      "loss": 1.1891,
+      "step": 26018
+    },
+    {
+      "epoch": 0.6193230505569837,
+      "grad_norm": 4.66819269132451,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 26019
+    },
+    {
+      "epoch": 0.6193468532800153,
+      "grad_norm": 3.460122129053305,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 26020
+    },
+    {
+      "epoch": 0.6193706560030467,
+      "grad_norm": 2.701342700454215,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 26021
+    },
+    {
+      "epoch": 0.6193944587260782,
+      "grad_norm": 0.03954371948374167,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 26022
+    },
+    {
+      "epoch": 0.6194182614491098,
+      "grad_norm": 2.7374800753240423,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 26023
+    },
+    {
+      "epoch": 0.6194420641721413,
+      "grad_norm": 3.047702239088065,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 26024
+    },
+    {
+      "epoch": 0.6194658668951728,
+      "grad_norm": 2.5983875908722402,
+      "learning_rate": 2e-05,
+      "loss": 0.3179,
+      "step": 26025
+    },
+    {
+      "epoch": 0.6194896696182043,
+      "grad_norm": 2.055274010915437,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 26026
+    },
+    {
+      "epoch": 0.6195134723412359,
+      "grad_norm": 35.275523222027054,
+      "learning_rate": 2e-05,
+      "loss": 1.4527,
+      "step": 26027
+    },
+    {
+      "epoch": 0.6195372750642674,
+      "grad_norm": 4.872729170100413,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 26028
+    },
+    {
+      "epoch": 0.6195610777872989,
+      "grad_norm": 7.889780752116,
+      "learning_rate": 2e-05,
+      "loss": 1.4293,
+      "step": 26029
+    },
+    {
+      "epoch": 0.6195848805103303,
+      "grad_norm": 1.3151002381964871,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 26030
+    },
+    {
+      "epoch": 0.6196086832333619,
+      "grad_norm": 0.2934416331976377,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 26031
+    },
+    {
+      "epoch": 0.6196324859563934,
+      "grad_norm": 6.918872884200102,
+      "learning_rate": 2e-05,
+      "loss": 0.2021,
+      "step": 26032
+    },
+    {
+      "epoch": 0.6196562886794249,
+      "grad_norm": 1.3181777021875365,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 26033
+    },
+    {
+      "epoch": 0.6196800914024564,
+      "grad_norm": 1.7183728856283123,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 26034
+    },
+    {
+      "epoch": 0.619703894125488,
+      "grad_norm": 1.5311883029238678,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 26035
+    },
+    {
+      "epoch": 0.6197276968485195,
+      "grad_norm": 0.7484495323585196,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 26036
+    },
+    {
+      "epoch": 0.619751499571551,
+      "grad_norm": 2.8972763176210687,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 26037
+    },
+    {
+      "epoch": 0.6197753022945826,
+      "grad_norm": 10.69738433227811,
+      "learning_rate": 2e-05,
+      "loss": 0.9556,
+      "step": 26038
+    },
+    {
+      "epoch": 0.619799105017614,
+      "grad_norm": 7.091836495851968,
+      "learning_rate": 2e-05,
+      "loss": 1.244,
+      "step": 26039
+    },
+    {
+      "epoch": 0.6198229077406455,
+      "grad_norm": 11.00357421760431,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 26040
+    },
+    {
+      "epoch": 0.619846710463677,
+      "grad_norm": 11.707644344899554,
+      "learning_rate": 2e-05,
+      "loss": 0.4351,
+      "step": 26041
+    },
+    {
+      "epoch": 0.6198705131867086,
+      "grad_norm": 0.33280576285164054,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 26042
+    },
+    {
+      "epoch": 0.6198943159097401,
+      "grad_norm": 4.480219209392228,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 26043
+    },
+    {
+      "epoch": 0.6199181186327716,
+      "grad_norm": 0.17109015393640006,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 26044
+    },
+    {
+      "epoch": 0.6199419213558031,
+      "grad_norm": 6.442031731389072,
+      "learning_rate": 2e-05,
+      "loss": 0.2843,
+      "step": 26045
+    },
+    {
+      "epoch": 0.6199657240788347,
+      "grad_norm": 2.66190544015124,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 26046
+    },
+    {
+      "epoch": 0.6199895268018661,
+      "grad_norm": 0.8085749113556107,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 26047
+    },
+    {
+      "epoch": 0.6200133295248976,
+      "grad_norm": 8.920482133961526,
+      "learning_rate": 2e-05,
+      "loss": 0.4732,
+      "step": 26048
+    },
+    {
+      "epoch": 0.6200371322479291,
+      "grad_norm": 2.7702713542653408,
+      "learning_rate": 2e-05,
+      "loss": 0.3378,
+      "step": 26049
+    },
+    {
+      "epoch": 0.6200609349709607,
+      "grad_norm": 10.49510662487942,
+      "learning_rate": 2e-05,
+      "loss": 0.6699,
+      "step": 26050
+    },
+    {
+      "epoch": 0.6200847376939922,
+      "grad_norm": 5.706262911726033,
+      "learning_rate": 2e-05,
+      "loss": 1.1396,
+      "step": 26051
+    },
+    {
+      "epoch": 0.6201085404170237,
+      "grad_norm": 5.912419414100027,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 26052
+    },
+    {
+      "epoch": 0.6201323431400553,
+      "grad_norm": 0.07598526249690923,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 26053
+    },
+    {
+      "epoch": 0.6201561458630868,
+      "grad_norm": 4.600197002129594,
+      "learning_rate": 2e-05,
+      "loss": 1.106,
+      "step": 26054
+    },
+    {
+      "epoch": 0.6201799485861182,
+      "grad_norm": 0.6077732895474092,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 26055
+    },
+    {
+      "epoch": 0.6202037513091497,
+      "grad_norm": 0.24881299778995414,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 26056
+    },
+    {
+      "epoch": 0.6202275540321813,
+      "grad_norm": 4.635050639131869,
+      "learning_rate": 2e-05,
+      "loss": 0.5501,
+      "step": 26057
+    },
+    {
+      "epoch": 0.6202513567552128,
+      "grad_norm": 1.2012240231197007,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 26058
+    },
+    {
+      "epoch": 0.6202751594782443,
+      "grad_norm": 3.6695181622849398,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 26059
+    },
+    {
+      "epoch": 0.6202989622012758,
+      "grad_norm": 14.146300061986866,
+      "learning_rate": 2e-05,
+      "loss": 0.9226,
+      "step": 26060
+    },
+    {
+      "epoch": 0.6203227649243074,
+      "grad_norm": 0.9556125848556679,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 26061
+    },
+    {
+      "epoch": 0.6203465676473389,
+      "grad_norm": 4.844696159997595,
+      "learning_rate": 2e-05,
+      "loss": 0.5618,
+      "step": 26062
+    },
+    {
+      "epoch": 0.6203703703703703,
+      "grad_norm": 3.4855241080137573,
+      "learning_rate": 2e-05,
+      "loss": 0.4369,
+      "step": 26063
+    },
+    {
+      "epoch": 0.6203941730934018,
+      "grad_norm": 9.92276177982577,
+      "learning_rate": 2e-05,
+      "loss": 0.8954,
+      "step": 26064
+    },
+    {
+      "epoch": 0.6204179758164334,
+      "grad_norm": 1.1943332760996628,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 26065
+    },
+    {
+      "epoch": 0.6204417785394649,
+      "grad_norm": 3.239406342347935,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 26066
+    },
+    {
+      "epoch": 0.6204655812624964,
+      "grad_norm": 2.5753853239919056,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 26067
+    },
+    {
+      "epoch": 0.620489383985528,
+      "grad_norm": 8.40000319344181,
+      "learning_rate": 2e-05,
+      "loss": 0.2029,
+      "step": 26068
+    },
+    {
+      "epoch": 0.6205131867085595,
+      "grad_norm": 10.568187763250476,
+      "learning_rate": 2e-05,
+      "loss": 0.2102,
+      "step": 26069
+    },
+    {
+      "epoch": 0.620536989431591,
+      "grad_norm": 6.882746165665561,
+      "learning_rate": 2e-05,
+      "loss": 0.289,
+      "step": 26070
+    },
+    {
+      "epoch": 0.6205607921546225,
+      "grad_norm": 4.492594257600959,
+      "learning_rate": 2e-05,
+      "loss": 0.6153,
+      "step": 26071
+    },
+    {
+      "epoch": 0.620584594877654,
+      "grad_norm": 3.8801011718972322,
+      "learning_rate": 2e-05,
+      "loss": 0.6512,
+      "step": 26072
+    },
+    {
+      "epoch": 0.6206083976006855,
+      "grad_norm": 7.517650896203862,
+      "learning_rate": 2e-05,
+      "loss": 0.5803,
+      "step": 26073
+    },
+    {
+      "epoch": 0.620632200323717,
+      "grad_norm": 2.9868972288806104,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 26074
+    },
+    {
+      "epoch": 0.6206560030467485,
+      "grad_norm": 6.514475176152182,
+      "learning_rate": 2e-05,
+      "loss": 0.4235,
+      "step": 26075
+    },
+    {
+      "epoch": 0.6206798057697801,
+      "grad_norm": 3.8579201823361164,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 26076
+    },
+    {
+      "epoch": 0.6207036084928116,
+      "grad_norm": 77.31646866959271,
+      "learning_rate": 2e-05,
+      "loss": 0.6048,
+      "step": 26077
+    },
+    {
+      "epoch": 0.6207274112158431,
+      "grad_norm": 5.217297602553964,
+      "learning_rate": 2e-05,
+      "loss": 0.7825,
+      "step": 26078
+    },
+    {
+      "epoch": 0.6207512139388746,
+      "grad_norm": 4.09518811443656,
+      "learning_rate": 2e-05,
+      "loss": 0.5739,
+      "step": 26079
+    },
+    {
+      "epoch": 0.6207750166619062,
+      "grad_norm": 6.20416422889327,
+      "learning_rate": 2e-05,
+      "loss": 0.2587,
+      "step": 26080
+    },
+    {
+      "epoch": 0.6207988193849376,
+      "grad_norm": 6.190601600390762,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 26081
+    },
+    {
+      "epoch": 0.6208226221079691,
+      "grad_norm": 4.957988957229742,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 26082
+    },
+    {
+      "epoch": 0.6208464248310007,
+      "grad_norm": 6.5116467549459,
+      "learning_rate": 2e-05,
+      "loss": 0.2499,
+      "step": 26083
+    },
+    {
+      "epoch": 0.6208702275540322,
+      "grad_norm": 4.07013114417234,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 26084
+    },
+    {
+      "epoch": 0.6208940302770637,
+      "grad_norm": 2.180354533467832,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 26085
+    },
+    {
+      "epoch": 0.6209178330000952,
+      "grad_norm": 3.105659279162431,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 26086
+    },
+    {
+      "epoch": 0.6209416357231268,
+      "grad_norm": 1.7965179782931302,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 26087
+    },
+    {
+      "epoch": 0.6209654384461583,
+      "grad_norm": 4.0188575709382715,
+      "learning_rate": 2e-05,
+      "loss": 0.5477,
+      "step": 26088
+    },
+    {
+      "epoch": 0.6209892411691897,
+      "grad_norm": 3.5529729232074905,
+      "learning_rate": 2e-05,
+      "loss": 0.5123,
+      "step": 26089
+    },
+    {
+      "epoch": 0.6210130438922212,
+      "grad_norm": 4.000952083845828,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 26090
+    },
+    {
+      "epoch": 0.6210368466152528,
+      "grad_norm": 0.1309588038288312,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 26091
+    },
+    {
+      "epoch": 0.6210606493382843,
+      "grad_norm": 5.389982106633235,
+      "learning_rate": 2e-05,
+      "loss": 0.6903,
+      "step": 26092
+    },
+    {
+      "epoch": 0.6210844520613158,
+      "grad_norm": 0.4860880208940699,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 26093
+    },
+    {
+      "epoch": 0.6211082547843473,
+      "grad_norm": 3.2415585421777187,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 26094
+    },
+    {
+      "epoch": 0.6211320575073789,
+      "grad_norm": 3.697517330453249,
+      "learning_rate": 2e-05,
+      "loss": 0.5729,
+      "step": 26095
+    },
+    {
+      "epoch": 0.6211558602304104,
+      "grad_norm": 0.6175628600268384,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 26096
+    },
+    {
+      "epoch": 0.6211796629534418,
+      "grad_norm": 0.8854113203695378,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 26097
+    },
+    {
+      "epoch": 0.6212034656764734,
+      "grad_norm": 0.49126966113180515,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 26098
+    },
+    {
+      "epoch": 0.6212272683995049,
+      "grad_norm": 0.1652895997970892,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 26099
+    },
+    {
+      "epoch": 0.6212510711225364,
+      "grad_norm": 3.1707127958438606,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 26100
+    },
+    {
+      "epoch": 0.6212748738455679,
+      "grad_norm": 1.5672453294602204,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 26101
+    },
+    {
+      "epoch": 0.6212986765685995,
+      "grad_norm": 2.477313057560148,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 26102
+    },
+    {
+      "epoch": 0.621322479291631,
+      "grad_norm": 3.318643248368727,
+      "learning_rate": 2e-05,
+      "loss": 0.227,
+      "step": 26103
+    },
+    {
+      "epoch": 0.6213462820146625,
+      "grad_norm": 2.8625675595996016,
+      "learning_rate": 2e-05,
+      "loss": 0.1503,
+      "step": 26104
+    },
+    {
+      "epoch": 0.621370084737694,
+      "grad_norm": 7.291327838804458,
+      "learning_rate": 2e-05,
+      "loss": 0.1972,
+      "step": 26105
+    },
+    {
+      "epoch": 0.6213938874607255,
+      "grad_norm": 7.202150139583064,
+      "learning_rate": 2e-05,
+      "loss": 1.2396,
+      "step": 26106
+    },
+    {
+      "epoch": 0.621417690183757,
+      "grad_norm": 7.053887464790874,
+      "learning_rate": 2e-05,
+      "loss": 0.6524,
+      "step": 26107
+    },
+    {
+      "epoch": 0.6214414929067885,
+      "grad_norm": 3.101345515686218,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 26108
+    },
+    {
+      "epoch": 0.62146529562982,
+      "grad_norm": 4.9893155628849515,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 26109
+    },
+    {
+      "epoch": 0.6214890983528516,
+      "grad_norm": 6.715678609910354,
+      "learning_rate": 2e-05,
+      "loss": 0.6716,
+      "step": 26110
+    },
+    {
+      "epoch": 0.6215129010758831,
+      "grad_norm": 5.863232644741671,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 26111
+    },
+    {
+      "epoch": 0.6215367037989146,
+      "grad_norm": 2.96626746258452,
+      "learning_rate": 2e-05,
+      "loss": 0.1646,
+      "step": 26112
+    },
+    {
+      "epoch": 0.6215605065219462,
+      "grad_norm": 0.1964118691655647,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 26113
+    },
+    {
+      "epoch": 0.6215843092449777,
+      "grad_norm": 2.57686009738097,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 26114
+    },
+    {
+      "epoch": 0.6216081119680091,
+      "grad_norm": 0.47845949984768343,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 26115
+    },
+    {
+      "epoch": 0.6216319146910406,
+      "grad_norm": 0.9731343283866316,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 26116
+    },
+    {
+      "epoch": 0.6216557174140722,
+      "grad_norm": 2.6391960939411905,
+      "learning_rate": 2e-05,
+      "loss": 0.4236,
+      "step": 26117
+    },
+    {
+      "epoch": 0.6216795201371037,
+      "grad_norm": 3.0512493692865785,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 26118
+    },
+    {
+      "epoch": 0.6217033228601352,
+      "grad_norm": 6.686978914484944,
+      "learning_rate": 2e-05,
+      "loss": 0.2088,
+      "step": 26119
+    },
+    {
+      "epoch": 0.6217271255831667,
+      "grad_norm": 1.44003955030918,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 26120
+    },
+    {
+      "epoch": 0.6217509283061983,
+      "grad_norm": 1.041086724683228,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 26121
+    },
+    {
+      "epoch": 0.6217747310292298,
+      "grad_norm": 0.9969208376057245,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 26122
+    },
+    {
+      "epoch": 0.6217985337522612,
+      "grad_norm": 3.955507999141744,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 26123
+    },
+    {
+      "epoch": 0.6218223364752927,
+      "grad_norm": 3.3216739339353283,
+      "learning_rate": 2e-05,
+      "loss": 0.4604,
+      "step": 26124
+    },
+    {
+      "epoch": 0.6218461391983243,
+      "grad_norm": 1.7396610748050905,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 26125
+    },
+    {
+      "epoch": 0.6218699419213558,
+      "grad_norm": 0.1310263750400365,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 26126
+    },
+    {
+      "epoch": 0.6218937446443873,
+      "grad_norm": 0.4291498822000711,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 26127
+    },
+    {
+      "epoch": 0.6219175473674189,
+      "grad_norm": 10.837766149682576,
+      "learning_rate": 2e-05,
+      "loss": 0.5345,
+      "step": 26128
+    },
+    {
+      "epoch": 0.6219413500904504,
+      "grad_norm": 3.117627025588217,
+      "learning_rate": 2e-05,
+      "loss": 0.4966,
+      "step": 26129
+    },
+    {
+      "epoch": 0.6219651528134819,
+      "grad_norm": 3.3469936953371744,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 26130
+    },
+    {
+      "epoch": 0.6219889555365133,
+      "grad_norm": 1.9698514507220921,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 26131
+    },
+    {
+      "epoch": 0.6220127582595449,
+      "grad_norm": 4.008774333030249,
+      "learning_rate": 2e-05,
+      "loss": 0.5477,
+      "step": 26132
+    },
+    {
+      "epoch": 0.6220365609825764,
+      "grad_norm": 3.4169374513009205,
+      "learning_rate": 2e-05,
+      "loss": 0.6605,
+      "step": 26133
+    },
+    {
+      "epoch": 0.6220603637056079,
+      "grad_norm": 0.061129277207468344,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 26134
+    },
+    {
+      "epoch": 0.6220841664286394,
+      "grad_norm": 5.330188220201223,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 26135
+    },
+    {
+      "epoch": 0.622107969151671,
+      "grad_norm": 0.07596495959921223,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 26136
+    },
+    {
+      "epoch": 0.6221317718747025,
+      "grad_norm": 0.6773491212716246,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 26137
+    },
+    {
+      "epoch": 0.622155574597734,
+      "grad_norm": 4.376246877098335,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 26138
+    },
+    {
+      "epoch": 0.6221793773207654,
+      "grad_norm": 9.33866745486223,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 26139
+    },
+    {
+      "epoch": 0.622203180043797,
+      "grad_norm": 3.4397942394017798,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 26140
+    },
+    {
+      "epoch": 0.6222269827668285,
+      "grad_norm": 7.03385654818656,
+      "learning_rate": 2e-05,
+      "loss": 0.2895,
+      "step": 26141
+    },
+    {
+      "epoch": 0.62225078548986,
+      "grad_norm": 0.8112118690887193,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 26142
+    },
+    {
+      "epoch": 0.6222745882128915,
+      "grad_norm": 7.902215698202342,
+      "learning_rate": 2e-05,
+      "loss": 0.9603,
+      "step": 26143
+    },
+    {
+      "epoch": 0.6222983909359231,
+      "grad_norm": 3.4990042865408895,
+      "learning_rate": 2e-05,
+      "loss": 0.216,
+      "step": 26144
+    },
+    {
+      "epoch": 0.6223221936589546,
+      "grad_norm": 3.4837132067229644,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 26145
+    },
+    {
+      "epoch": 0.6223459963819861,
+      "grad_norm": 4.716665316426999,
+      "learning_rate": 2e-05,
+      "loss": 0.5024,
+      "step": 26146
+    },
+    {
+      "epoch": 0.6223697991050177,
+      "grad_norm": 1.7050198002955714,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 26147
+    },
+    {
+      "epoch": 0.6223936018280491,
+      "grad_norm": 0.39906095026047766,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 26148
+    },
+    {
+      "epoch": 0.6224174045510806,
+      "grad_norm": 0.48675974463557314,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 26149
+    },
+    {
+      "epoch": 0.6224412072741121,
+      "grad_norm": 17.779446465896793,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 26150
+    },
+    {
+      "epoch": 0.6224650099971437,
+      "grad_norm": 2.7114575180575478,
+      "learning_rate": 2e-05,
+      "loss": 0.1026,
+      "step": 26151
+    },
+    {
+      "epoch": 0.6224888127201752,
+      "grad_norm": 0.7941904782772593,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 26152
+    },
+    {
+      "epoch": 0.6225126154432067,
+      "grad_norm": 3.856623588811998,
+      "learning_rate": 2e-05,
+      "loss": 0.8567,
+      "step": 26153
+    },
+    {
+      "epoch": 0.6225364181662382,
+      "grad_norm": 5.157245871011717,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 26154
+    },
+    {
+      "epoch": 0.6225602208892698,
+      "grad_norm": 4.906358216721267,
+      "learning_rate": 2e-05,
+      "loss": 1.1171,
+      "step": 26155
+    },
+    {
+      "epoch": 0.6225840236123013,
+      "grad_norm": 0.25558685648692747,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 26156
+    },
+    {
+      "epoch": 0.6226078263353327,
+      "grad_norm": 2.1103313391162786,
+      "learning_rate": 2e-05,
+      "loss": 0.1673,
+      "step": 26157
+    },
+    {
+      "epoch": 0.6226316290583642,
+      "grad_norm": 5.950950692645895,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 26158
+    },
+    {
+      "epoch": 0.6226554317813958,
+      "grad_norm": 3.080157051607003,
+      "learning_rate": 2e-05,
+      "loss": 0.4541,
+      "step": 26159
+    },
+    {
+      "epoch": 0.6226792345044273,
+      "grad_norm": 7.5609820245792205,
+      "learning_rate": 2e-05,
+      "loss": 1.26,
+      "step": 26160
+    },
+    {
+      "epoch": 0.6227030372274588,
+      "grad_norm": 3.7771138067885004,
+      "learning_rate": 2e-05,
+      "loss": 0.0951,
+      "step": 26161
+    },
+    {
+      "epoch": 0.6227268399504904,
+      "grad_norm": 9.461073498800781,
+      "learning_rate": 2e-05,
+      "loss": 0.7294,
+      "step": 26162
+    },
+    {
+      "epoch": 0.6227506426735219,
+      "grad_norm": 5.168107164270649,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 26163
+    },
+    {
+      "epoch": 0.6227744453965534,
+      "grad_norm": 0.22207792341542695,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 26164
+    },
+    {
+      "epoch": 0.6227982481195848,
+      "grad_norm": 3.510493765700318,
+      "learning_rate": 2e-05,
+      "loss": 0.4937,
+      "step": 26165
+    },
+    {
+      "epoch": 0.6228220508426164,
+      "grad_norm": 1.4129975705513365,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 26166
+    },
+    {
+      "epoch": 0.6228458535656479,
+      "grad_norm": 9.647905286735574,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 26167
+    },
+    {
+      "epoch": 0.6228696562886794,
+      "grad_norm": 5.366188849106972,
+      "learning_rate": 2e-05,
+      "loss": 0.5244,
+      "step": 26168
+    },
+    {
+      "epoch": 0.6228934590117109,
+      "grad_norm": 5.489078185376668,
+      "learning_rate": 2e-05,
+      "loss": 1.2107,
+      "step": 26169
+    },
+    {
+      "epoch": 0.6229172617347425,
+      "grad_norm": 7.178545749323181,
+      "learning_rate": 2e-05,
+      "loss": 1.0033,
+      "step": 26170
+    },
+    {
+      "epoch": 0.622941064457774,
+      "grad_norm": 0.946621741403903,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 26171
+    },
+    {
+      "epoch": 0.6229648671808055,
+      "grad_norm": 6.016184966718706,
+      "learning_rate": 2e-05,
+      "loss": 0.3385,
+      "step": 26172
+    },
+    {
+      "epoch": 0.622988669903837,
+      "grad_norm": 1.9615531801072656,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 26173
+    },
+    {
+      "epoch": 0.6230124726268685,
+      "grad_norm": 1.1617579697898657,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 26174
+    },
+    {
+      "epoch": 0.6230362753499,
+      "grad_norm": 20.5211774534268,
+      "learning_rate": 2e-05,
+      "loss": 0.9132,
+      "step": 26175
+    },
+    {
+      "epoch": 0.6230600780729315,
+      "grad_norm": 2.743296103022159,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 26176
+    },
+    {
+      "epoch": 0.6230838807959631,
+      "grad_norm": 8.640186053495988,
+      "learning_rate": 2e-05,
+      "loss": 0.3115,
+      "step": 26177
+    },
+    {
+      "epoch": 0.6231076835189946,
+      "grad_norm": 23.59820694806887,
+      "learning_rate": 2e-05,
+      "loss": 0.6568,
+      "step": 26178
+    },
+    {
+      "epoch": 0.6231314862420261,
+      "grad_norm": 4.0639014703742555,
+      "learning_rate": 2e-05,
+      "loss": 0.6999,
+      "step": 26179
+    },
+    {
+      "epoch": 0.6231552889650576,
+      "grad_norm": 1.306602921621007,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 26180
+    },
+    {
+      "epoch": 0.6231790916880892,
+      "grad_norm": 2.9552519466427167,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 26181
+    },
+    {
+      "epoch": 0.6232028944111206,
+      "grad_norm": 5.103035733954933,
+      "learning_rate": 2e-05,
+      "loss": 1.08,
+      "step": 26182
+    },
+    {
+      "epoch": 0.6232266971341521,
+      "grad_norm": 5.3834824766902525,
+      "learning_rate": 2e-05,
+      "loss": 0.7937,
+      "step": 26183
+    },
+    {
+      "epoch": 0.6232504998571836,
+      "grad_norm": 4.551350430430914,
+      "learning_rate": 2e-05,
+      "loss": 0.9625,
+      "step": 26184
+    },
+    {
+      "epoch": 0.6232743025802152,
+      "grad_norm": 1.3846290642990444,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 26185
+    },
+    {
+      "epoch": 0.6232981053032467,
+      "grad_norm": 8.270502021446308,
+      "learning_rate": 2e-05,
+      "loss": 1.5564,
+      "step": 26186
+    },
+    {
+      "epoch": 0.6233219080262782,
+      "grad_norm": 3.534090906709527,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 26187
+    },
+    {
+      "epoch": 0.6233457107493097,
+      "grad_norm": 0.47679655276474453,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 26188
+    },
+    {
+      "epoch": 0.6233695134723413,
+      "grad_norm": 3.9966067582195137,
+      "learning_rate": 2e-05,
+      "loss": 0.7363,
+      "step": 26189
+    },
+    {
+      "epoch": 0.6233933161953727,
+      "grad_norm": 0.14605001882286198,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 26190
+    },
+    {
+      "epoch": 0.6234171189184042,
+      "grad_norm": 1.542287085104212,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 26191
+    },
+    {
+      "epoch": 0.6234409216414358,
+      "grad_norm": 1.9247102056380438,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 26192
+    },
+    {
+      "epoch": 0.6234647243644673,
+      "grad_norm": 0.07490237129947025,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 26193
+    },
+    {
+      "epoch": 0.6234885270874988,
+      "grad_norm": 0.16226084768845755,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 26194
+    },
+    {
+      "epoch": 0.6235123298105303,
+      "grad_norm": 16.121892995803908,
+      "learning_rate": 2e-05,
+      "loss": 1.524,
+      "step": 26195
+    },
+    {
+      "epoch": 0.6235361325335619,
+      "grad_norm": 3.0798347411667746,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 26196
+    },
+    {
+      "epoch": 0.6235599352565934,
+      "grad_norm": 2.7997796336260947,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 26197
+    },
+    {
+      "epoch": 0.6235837379796249,
+      "grad_norm": 3.5646397420515794,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 26198
+    },
+    {
+      "epoch": 0.6236075407026563,
+      "grad_norm": 0.029305154807161355,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 26199
+    },
+    {
+      "epoch": 0.6236313434256879,
+      "grad_norm": 1.2919412984596204,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 26200
+    },
+    {
+      "epoch": 0.6236551461487194,
+      "grad_norm": 1.6791865048756107,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 26201
+    },
+    {
+      "epoch": 0.6236789488717509,
+      "grad_norm": 8.38196173825338,
+      "learning_rate": 2e-05,
+      "loss": 1.0982,
+      "step": 26202
+    },
+    {
+      "epoch": 0.6237027515947824,
+      "grad_norm": 11.702929893172415,
+      "learning_rate": 2e-05,
+      "loss": 0.4765,
+      "step": 26203
+    },
+    {
+      "epoch": 0.623726554317814,
+      "grad_norm": 4.159948197242484,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 26204
+    },
+    {
+      "epoch": 0.6237503570408455,
+      "grad_norm": 10.948793653204742,
+      "learning_rate": 2e-05,
+      "loss": 1.2212,
+      "step": 26205
+    },
+    {
+      "epoch": 0.623774159763877,
+      "grad_norm": 4.401379826761359,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 26206
+    },
+    {
+      "epoch": 0.6237979624869086,
+      "grad_norm": 4.744469938503239,
+      "learning_rate": 2e-05,
+      "loss": 0.7698,
+      "step": 26207
+    },
+    {
+      "epoch": 0.62382176520994,
+      "grad_norm": 4.404587811933315,
+      "learning_rate": 2e-05,
+      "loss": 0.7663,
+      "step": 26208
+    },
+    {
+      "epoch": 0.6238455679329715,
+      "grad_norm": 4.842677456916545,
+      "learning_rate": 2e-05,
+      "loss": 1.3189,
+      "step": 26209
+    },
+    {
+      "epoch": 0.623869370656003,
+      "grad_norm": 8.011036999554642,
+      "learning_rate": 2e-05,
+      "loss": 0.6854,
+      "step": 26210
+    },
+    {
+      "epoch": 0.6238931733790346,
+      "grad_norm": 14.600167704969639,
+      "learning_rate": 2e-05,
+      "loss": 0.3624,
+      "step": 26211
+    },
+    {
+      "epoch": 0.6239169761020661,
+      "grad_norm": 7.119249913756167,
+      "learning_rate": 2e-05,
+      "loss": 0.3142,
+      "step": 26212
+    },
+    {
+      "epoch": 0.6239407788250976,
+      "grad_norm": 4.668155613699712,
+      "learning_rate": 2e-05,
+      "loss": 0.6353,
+      "step": 26213
+    },
+    {
+      "epoch": 0.6239645815481291,
+      "grad_norm": 4.825129195669745,
+      "learning_rate": 2e-05,
+      "loss": 0.2783,
+      "step": 26214
+    },
+    {
+      "epoch": 0.6239883842711607,
+      "grad_norm": 4.0544105101742325,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 26215
+    },
+    {
+      "epoch": 0.6240121869941921,
+      "grad_norm": 0.23205619887576695,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 26216
+    },
+    {
+      "epoch": 0.6240359897172236,
+      "grad_norm": 2.115090489906461,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 26217
+    },
+    {
+      "epoch": 0.6240597924402551,
+      "grad_norm": 3.9497964652391238,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 26218
+    },
+    {
+      "epoch": 0.6240835951632867,
+      "grad_norm": 3.416644307065319,
+      "learning_rate": 2e-05,
+      "loss": 0.4193,
+      "step": 26219
+    },
+    {
+      "epoch": 0.6241073978863182,
+      "grad_norm": 0.5026227677124852,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 26220
+    },
+    {
+      "epoch": 0.6241312006093497,
+      "grad_norm": 3.0951984896533657,
+      "learning_rate": 2e-05,
+      "loss": 0.6113,
+      "step": 26221
+    },
+    {
+      "epoch": 0.6241550033323813,
+      "grad_norm": 2.766217661513695,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 26222
+    },
+    {
+      "epoch": 0.6241788060554128,
+      "grad_norm": 2.627070733724559,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 26223
+    },
+    {
+      "epoch": 0.6242026087784442,
+      "grad_norm": 7.341511179517153,
+      "learning_rate": 2e-05,
+      "loss": 2.1699,
+      "step": 26224
+    },
+    {
+      "epoch": 0.6242264115014757,
+      "grad_norm": 1.8001725392673409,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 26225
+    },
+    {
+      "epoch": 0.6242502142245073,
+      "grad_norm": 0.5062114751794187,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 26226
+    },
+    {
+      "epoch": 0.6242740169475388,
+      "grad_norm": 11.98441644253313,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 26227
+    },
+    {
+      "epoch": 0.6242978196705703,
+      "grad_norm": 0.16658416489254055,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 26228
+    },
+    {
+      "epoch": 0.6243216223936018,
+      "grad_norm": 0.6953216381339454,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 26229
+    },
+    {
+      "epoch": 0.6243454251166334,
+      "grad_norm": 7.315435066272734,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 26230
+    },
+    {
+      "epoch": 0.6243692278396649,
+      "grad_norm": 0.1236555202404809,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 26231
+    },
+    {
+      "epoch": 0.6243930305626963,
+      "grad_norm": 0.613641713897237,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 26232
+    },
+    {
+      "epoch": 0.6244168332857278,
+      "grad_norm": 6.974303803621742,
+      "learning_rate": 2e-05,
+      "loss": 0.4011,
+      "step": 26233
+    },
+    {
+      "epoch": 0.6244406360087594,
+      "grad_norm": 5.609689014008628,
+      "learning_rate": 2e-05,
+      "loss": 0.9026,
+      "step": 26234
+    },
+    {
+      "epoch": 0.6244644387317909,
+      "grad_norm": 3.9031917753397525,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 26235
+    },
+    {
+      "epoch": 0.6244882414548224,
+      "grad_norm": 5.2976391284322375,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 26236
+    },
+    {
+      "epoch": 0.624512044177854,
+      "grad_norm": 3.0124611294433112,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 26237
+    },
+    {
+      "epoch": 0.6245358469008855,
+      "grad_norm": 7.447988225930389,
+      "learning_rate": 2e-05,
+      "loss": 0.4933,
+      "step": 26238
+    },
+    {
+      "epoch": 0.624559649623917,
+      "grad_norm": 4.921429315637673,
+      "learning_rate": 2e-05,
+      "loss": 0.1392,
+      "step": 26239
+    },
+    {
+      "epoch": 0.6245834523469485,
+      "grad_norm": 0.8014955622336815,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 26240
+    },
+    {
+      "epoch": 0.62460725506998,
+      "grad_norm": 5.111095607106607,
+      "learning_rate": 2e-05,
+      "loss": 0.7612,
+      "step": 26241
+    },
+    {
+      "epoch": 0.6246310577930115,
+      "grad_norm": 17.647427861979885,
+      "learning_rate": 2e-05,
+      "loss": 0.5164,
+      "step": 26242
+    },
+    {
+      "epoch": 0.624654860516043,
+      "grad_norm": 3.1390310292934567,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 26243
+    },
+    {
+      "epoch": 0.6246786632390745,
+      "grad_norm": 14.394299589502602,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 26244
+    },
+    {
+      "epoch": 0.6247024659621061,
+      "grad_norm": 0.5487414278732312,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 26245
+    },
+    {
+      "epoch": 0.6247262686851376,
+      "grad_norm": 6.50674125133827,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 26246
+    },
+    {
+      "epoch": 0.6247500714081691,
+      "grad_norm": 3.220617649932591,
+      "learning_rate": 2e-05,
+      "loss": 0.8404,
+      "step": 26247
+    },
+    {
+      "epoch": 0.6247738741312006,
+      "grad_norm": 3.5278425865822576,
+      "learning_rate": 2e-05,
+      "loss": 0.4593,
+      "step": 26248
+    },
+    {
+      "epoch": 0.6247976768542322,
+      "grad_norm": 3.0382917747935716,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 26249
+    },
+    {
+      "epoch": 0.6248214795772636,
+      "grad_norm": 3.6762936610128003,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 26250
+    },
+    {
+      "epoch": 0.6248452823002951,
+      "grad_norm": 1.0217762821418577,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 26251
+    },
+    {
+      "epoch": 0.6248690850233267,
+      "grad_norm": 5.085593837344327,
+      "learning_rate": 2e-05,
+      "loss": 0.7757,
+      "step": 26252
+    },
+    {
+      "epoch": 0.6248928877463582,
+      "grad_norm": 0.5893944485554959,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 26253
+    },
+    {
+      "epoch": 0.6249166904693897,
+      "grad_norm": 7.856663731475519,
+      "learning_rate": 2e-05,
+      "loss": 0.534,
+      "step": 26254
+    },
+    {
+      "epoch": 0.6249404931924212,
+      "grad_norm": 5.171709618659238,
+      "learning_rate": 2e-05,
+      "loss": 1.0731,
+      "step": 26255
+    },
+    {
+      "epoch": 0.6249642959154528,
+      "grad_norm": 7.861039550707168,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 26256
+    },
+    {
+      "epoch": 0.6249880986384843,
+      "grad_norm": 12.687550400558992,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 26257
+    },
+    {
+      "epoch": 0.6250119013615157,
+      "grad_norm": 4.76011627586083,
+      "learning_rate": 2e-05,
+      "loss": 0.9771,
+      "step": 26258
+    },
+    {
+      "epoch": 0.6250357040845472,
+      "grad_norm": 3.7678784488547072,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 26259
+    },
+    {
+      "epoch": 0.6250595068075788,
+      "grad_norm": 3.810316045845253,
+      "learning_rate": 2e-05,
+      "loss": 0.9204,
+      "step": 26260
+    },
+    {
+      "epoch": 0.6250833095306103,
+      "grad_norm": 4.392255437969336,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 26261
+    },
+    {
+      "epoch": 0.6251071122536418,
+      "grad_norm": 3.401914261386405,
+      "learning_rate": 2e-05,
+      "loss": 0.4505,
+      "step": 26262
+    },
+    {
+      "epoch": 0.6251309149766733,
+      "grad_norm": 5.101837288707115,
+      "learning_rate": 2e-05,
+      "loss": 0.6362,
+      "step": 26263
+    },
+    {
+      "epoch": 0.6251547176997049,
+      "grad_norm": 0.4905399895928336,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 26264
+    },
+    {
+      "epoch": 0.6251785204227364,
+      "grad_norm": 5.775037730280124,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 26265
+    },
+    {
+      "epoch": 0.6252023231457678,
+      "grad_norm": 2.4974663194317217,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 26266
+    },
+    {
+      "epoch": 0.6252261258687994,
+      "grad_norm": 3.959177443742514,
+      "learning_rate": 2e-05,
+      "loss": 0.6867,
+      "step": 26267
+    },
+    {
+      "epoch": 0.6252499285918309,
+      "grad_norm": 4.332190399635712,
+      "learning_rate": 2e-05,
+      "loss": 0.5575,
+      "step": 26268
+    },
+    {
+      "epoch": 0.6252737313148624,
+      "grad_norm": 8.77087314416199,
+      "learning_rate": 2e-05,
+      "loss": 0.9566,
+      "step": 26269
+    },
+    {
+      "epoch": 0.6252975340378939,
+      "grad_norm": 2.8596777050957627,
+      "learning_rate": 2e-05,
+      "loss": 0.5458,
+      "step": 26270
+    },
+    {
+      "epoch": 0.6253213367609255,
+      "grad_norm": 0.13414432475430485,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 26271
+    },
+    {
+      "epoch": 0.625345139483957,
+      "grad_norm": 4.865613491457648,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 26272
+    },
+    {
+      "epoch": 0.6253689422069885,
+      "grad_norm": 1.2024549792822723,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 26273
+    },
+    {
+      "epoch": 0.62539274493002,
+      "grad_norm": 10.937545881786804,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 26274
+    },
+    {
+      "epoch": 0.6254165476530515,
+      "grad_norm": 8.1277700051006,
+      "learning_rate": 2e-05,
+      "loss": 0.3602,
+      "step": 26275
+    },
+    {
+      "epoch": 0.625440350376083,
+      "grad_norm": 0.5585345949704039,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 26276
+    },
+    {
+      "epoch": 0.6254641530991145,
+      "grad_norm": 4.807508621441196,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 26277
+    },
+    {
+      "epoch": 0.625487955822146,
+      "grad_norm": 0.9891801525934188,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 26278
+    },
+    {
+      "epoch": 0.6255117585451776,
+      "grad_norm": 0.5215633741185074,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 26279
+    },
+    {
+      "epoch": 0.6255355612682091,
+      "grad_norm": 2.1213764584351464,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 26280
+    },
+    {
+      "epoch": 0.6255593639912406,
+      "grad_norm": 5.4289363929565395,
+      "learning_rate": 2e-05,
+      "loss": 0.9017,
+      "step": 26281
+    },
+    {
+      "epoch": 0.6255831667142722,
+      "grad_norm": 1.3193555268151287,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 26282
+    },
+    {
+      "epoch": 0.6256069694373037,
+      "grad_norm": 6.832278281104992,
+      "learning_rate": 2e-05,
+      "loss": 0.8564,
+      "step": 26283
+    },
+    {
+      "epoch": 0.6256307721603351,
+      "grad_norm": 8.753310753377635,
+      "learning_rate": 2e-05,
+      "loss": 0.9196,
+      "step": 26284
+    },
+    {
+      "epoch": 0.6256545748833666,
+      "grad_norm": 2.6548243660278823,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 26285
+    },
+    {
+      "epoch": 0.6256783776063982,
+      "grad_norm": 8.108252732062166,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 26286
+    },
+    {
+      "epoch": 0.6257021803294297,
+      "grad_norm": 0.09797706525949858,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 26287
+    },
+    {
+      "epoch": 0.6257259830524612,
+      "grad_norm": 6.01473387429862,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 26288
+    },
+    {
+      "epoch": 0.6257497857754927,
+      "grad_norm": 3.8616260204946844,
+      "learning_rate": 2e-05,
+      "loss": 0.6723,
+      "step": 26289
+    },
+    {
+      "epoch": 0.6257735884985243,
+      "grad_norm": 3.335285638162926,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 26290
+    },
+    {
+      "epoch": 0.6257973912215558,
+      "grad_norm": 0.07010206875737636,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 26291
+    },
+    {
+      "epoch": 0.6258211939445872,
+      "grad_norm": 6.3071096610407,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 26292
+    },
+    {
+      "epoch": 0.6258449966676187,
+      "grad_norm": 3.5666124527424445,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 26293
+    },
+    {
+      "epoch": 0.6258687993906503,
+      "grad_norm": 2.1746954814644863,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 26294
+    },
+    {
+      "epoch": 0.6258926021136818,
+      "grad_norm": 5.818498699925978,
+      "learning_rate": 2e-05,
+      "loss": 0.9121,
+      "step": 26295
+    },
+    {
+      "epoch": 0.6259164048367133,
+      "grad_norm": 3.583977751192455,
+      "learning_rate": 2e-05,
+      "loss": 0.5637,
+      "step": 26296
+    },
+    {
+      "epoch": 0.6259402075597449,
+      "grad_norm": 4.705588940468139,
+      "learning_rate": 2e-05,
+      "loss": 1.0059,
+      "step": 26297
+    },
+    {
+      "epoch": 0.6259640102827764,
+      "grad_norm": 5.470502944805894,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 26298
+    },
+    {
+      "epoch": 0.6259878130058079,
+      "grad_norm": 11.728950903096353,
+      "learning_rate": 2e-05,
+      "loss": 0.0987,
+      "step": 26299
+    },
+    {
+      "epoch": 0.6260116157288393,
+      "grad_norm": 2.3056483170980373,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 26300
+    },
+    {
+      "epoch": 0.6260354184518709,
+      "grad_norm": 7.974020513582926,
+      "learning_rate": 2e-05,
+      "loss": 0.5353,
+      "step": 26301
+    },
+    {
+      "epoch": 0.6260592211749024,
+      "grad_norm": 6.8192430631689795,
+      "learning_rate": 2e-05,
+      "loss": 0.0914,
+      "step": 26302
+    },
+    {
+      "epoch": 0.6260830238979339,
+      "grad_norm": 0.2652212653528604,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 26303
+    },
+    {
+      "epoch": 0.6261068266209654,
+      "grad_norm": 0.9031151279505332,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 26304
+    },
+    {
+      "epoch": 0.626130629343997,
+      "grad_norm": 0.10128886513386136,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26305
+    },
+    {
+      "epoch": 0.6261544320670285,
+      "grad_norm": 2.2798313109519333,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 26306
+    },
+    {
+      "epoch": 0.62617823479006,
+      "grad_norm": 5.0345473190055925,
+      "learning_rate": 2e-05,
+      "loss": 1.6398,
+      "step": 26307
+    },
+    {
+      "epoch": 0.6262020375130914,
+      "grad_norm": 5.634279793466799,
+      "learning_rate": 2e-05,
+      "loss": 0.8283,
+      "step": 26308
+    },
+    {
+      "epoch": 0.626225840236123,
+      "grad_norm": 1.5031514841331028,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 26309
+    },
+    {
+      "epoch": 0.6262496429591545,
+      "grad_norm": 1.174258901207029,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 26310
+    },
+    {
+      "epoch": 0.626273445682186,
+      "grad_norm": 5.634903576607204,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 26311
+    },
+    {
+      "epoch": 0.6262972484052176,
+      "grad_norm": 8.427966845698773,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 26312
+    },
+    {
+      "epoch": 0.6263210511282491,
+      "grad_norm": 5.494836364629946,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 26313
+    },
+    {
+      "epoch": 0.6263448538512806,
+      "grad_norm": 20.579908772055596,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 26314
+    },
+    {
+      "epoch": 0.6263686565743121,
+      "grad_norm": 2.590172037115155,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 26315
+    },
+    {
+      "epoch": 0.6263924592973437,
+      "grad_norm": 3.8351623305491587,
+      "learning_rate": 2e-05,
+      "loss": 0.3317,
+      "step": 26316
+    },
+    {
+      "epoch": 0.6264162620203751,
+      "grad_norm": 4.955915989251054,
+      "learning_rate": 2e-05,
+      "loss": 0.9688,
+      "step": 26317
+    },
+    {
+      "epoch": 0.6264400647434066,
+      "grad_norm": 5.073518023283942,
+      "learning_rate": 2e-05,
+      "loss": 0.2811,
+      "step": 26318
+    },
+    {
+      "epoch": 0.6264638674664381,
+      "grad_norm": 10.377089908545305,
+      "learning_rate": 2e-05,
+      "loss": 0.4799,
+      "step": 26319
+    },
+    {
+      "epoch": 0.6264876701894697,
+      "grad_norm": 8.175904258309034,
+      "learning_rate": 2e-05,
+      "loss": 0.4744,
+      "step": 26320
+    },
+    {
+      "epoch": 0.6265114729125012,
+      "grad_norm": 14.213658473980933,
+      "learning_rate": 2e-05,
+      "loss": 0.5627,
+      "step": 26321
+    },
+    {
+      "epoch": 0.6265352756355327,
+      "grad_norm": 3.509174581511072,
+      "learning_rate": 2e-05,
+      "loss": 0.7441,
+      "step": 26322
+    },
+    {
+      "epoch": 0.6265590783585642,
+      "grad_norm": 2.7857809559359152,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 26323
+    },
+    {
+      "epoch": 0.6265828810815958,
+      "grad_norm": 5.246091926554562,
+      "learning_rate": 2e-05,
+      "loss": 0.5993,
+      "step": 26324
+    },
+    {
+      "epoch": 0.6266066838046273,
+      "grad_norm": 2.0810209579380907,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 26325
+    },
+    {
+      "epoch": 0.6266304865276587,
+      "grad_norm": 5.324960822536248,
+      "learning_rate": 2e-05,
+      "loss": 0.7741,
+      "step": 26326
+    },
+    {
+      "epoch": 0.6266542892506903,
+      "grad_norm": 0.6732775805272263,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 26327
+    },
+    {
+      "epoch": 0.6266780919737218,
+      "grad_norm": 6.988501819897167,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 26328
+    },
+    {
+      "epoch": 0.6267018946967533,
+      "grad_norm": 0.689682089639759,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 26329
+    },
+    {
+      "epoch": 0.6267256974197848,
+      "grad_norm": 2.5206795698795372,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 26330
+    },
+    {
+      "epoch": 0.6267495001428164,
+      "grad_norm": 9.985702929926662,
+      "learning_rate": 2e-05,
+      "loss": 0.7623,
+      "step": 26331
+    },
+    {
+      "epoch": 0.6267733028658479,
+      "grad_norm": 4.750411228608674,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 26332
+    },
+    {
+      "epoch": 0.6267971055888794,
+      "grad_norm": 0.6747306000621047,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 26333
+    },
+    {
+      "epoch": 0.6268209083119108,
+      "grad_norm": 1.0327292294975914,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 26334
+    },
+    {
+      "epoch": 0.6268447110349424,
+      "grad_norm": 3.224958913049826,
+      "learning_rate": 2e-05,
+      "loss": 0.6176,
+      "step": 26335
+    },
+    {
+      "epoch": 0.6268685137579739,
+      "grad_norm": 0.6304107455841171,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 26336
+    },
+    {
+      "epoch": 0.6268923164810054,
+      "grad_norm": 0.7923619797561341,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 26337
+    },
+    {
+      "epoch": 0.6269161192040369,
+      "grad_norm": 9.282920552402675,
+      "learning_rate": 2e-05,
+      "loss": 0.1672,
+      "step": 26338
+    },
+    {
+      "epoch": 0.6269399219270685,
+      "grad_norm": 2.503349251780338,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 26339
+    },
+    {
+      "epoch": 0.6269637246501,
+      "grad_norm": 0.39270865094729923,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 26340
+    },
+    {
+      "epoch": 0.6269875273731315,
+      "grad_norm": 8.379493155074094,
+      "learning_rate": 2e-05,
+      "loss": 0.2002,
+      "step": 26341
+    },
+    {
+      "epoch": 0.627011330096163,
+      "grad_norm": 7.330022721280854,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 26342
+    },
+    {
+      "epoch": 0.6270351328191945,
+      "grad_norm": 14.845688367888876,
+      "learning_rate": 2e-05,
+      "loss": 0.4561,
+      "step": 26343
+    },
+    {
+      "epoch": 0.627058935542226,
+      "grad_norm": 2.162336089838665,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 26344
+    },
+    {
+      "epoch": 0.6270827382652575,
+      "grad_norm": 2.4261939673531803,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 26345
+    },
+    {
+      "epoch": 0.6271065409882891,
+      "grad_norm": 0.534373522283054,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 26346
+    },
+    {
+      "epoch": 0.6271303437113206,
+      "grad_norm": 13.103072324876717,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 26347
+    },
+    {
+      "epoch": 0.6271541464343521,
+      "grad_norm": 1.3675712892322793,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 26348
+    },
+    {
+      "epoch": 0.6271779491573836,
+      "grad_norm": 2.2755048169902934,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 26349
+    },
+    {
+      "epoch": 0.6272017518804152,
+      "grad_norm": 3.2631795073832857,
+      "learning_rate": 2e-05,
+      "loss": 0.867,
+      "step": 26350
+    },
+    {
+      "epoch": 0.6272255546034466,
+      "grad_norm": 4.751426705086184,
+      "learning_rate": 2e-05,
+      "loss": 0.8396,
+      "step": 26351
+    },
+    {
+      "epoch": 0.6272493573264781,
+      "grad_norm": 11.918093691041761,
+      "learning_rate": 2e-05,
+      "loss": 0.5113,
+      "step": 26352
+    },
+    {
+      "epoch": 0.6272731600495096,
+      "grad_norm": 0.11227576843077876,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 26353
+    },
+    {
+      "epoch": 0.6272969627725412,
+      "grad_norm": 3.3147405913288948,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 26354
+    },
+    {
+      "epoch": 0.6273207654955727,
+      "grad_norm": 6.308949302020143,
+      "learning_rate": 2e-05,
+      "loss": 1.1477,
+      "step": 26355
+    },
+    {
+      "epoch": 0.6273445682186042,
+      "grad_norm": 2.6353706994588944,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 26356
+    },
+    {
+      "epoch": 0.6273683709416358,
+      "grad_norm": 3.578609865896185,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 26357
+    },
+    {
+      "epoch": 0.6273921736646673,
+      "grad_norm": 6.749592573297346,
+      "learning_rate": 2e-05,
+      "loss": 0.9921,
+      "step": 26358
+    },
+    {
+      "epoch": 0.6274159763876987,
+      "grad_norm": 1.6381138535413193,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 26359
+    },
+    {
+      "epoch": 0.6274397791107302,
+      "grad_norm": 0.1818054740247237,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 26360
+    },
+    {
+      "epoch": 0.6274635818337618,
+      "grad_norm": 6.693996304584665,
+      "learning_rate": 2e-05,
+      "loss": 0.5457,
+      "step": 26361
+    },
+    {
+      "epoch": 0.6274873845567933,
+      "grad_norm": 36.061389703134104,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 26362
+    },
+    {
+      "epoch": 0.6275111872798248,
+      "grad_norm": 9.254069580190903,
+      "learning_rate": 2e-05,
+      "loss": 0.2903,
+      "step": 26363
+    },
+    {
+      "epoch": 0.6275349900028563,
+      "grad_norm": 7.385216348879928,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 26364
+    },
+    {
+      "epoch": 0.6275587927258879,
+      "grad_norm": 5.270308300006339,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 26365
+    },
+    {
+      "epoch": 0.6275825954489194,
+      "grad_norm": 5.59055960864549,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 26366
+    },
+    {
+      "epoch": 0.6276063981719509,
+      "grad_norm": 16.335088582796025,
+      "learning_rate": 2e-05,
+      "loss": 0.2843,
+      "step": 26367
+    },
+    {
+      "epoch": 0.6276302008949823,
+      "grad_norm": 3.816468966317389,
+      "learning_rate": 2e-05,
+      "loss": 0.5253,
+      "step": 26368
+    },
+    {
+      "epoch": 0.6276540036180139,
+      "grad_norm": 2.812699392872204,
+      "learning_rate": 2e-05,
+      "loss": 0.4814,
+      "step": 26369
+    },
+    {
+      "epoch": 0.6276778063410454,
+      "grad_norm": 4.8666092146535265,
+      "learning_rate": 2e-05,
+      "loss": 0.6992,
+      "step": 26370
+    },
+    {
+      "epoch": 0.6277016090640769,
+      "grad_norm": 10.635868266626296,
+      "learning_rate": 2e-05,
+      "loss": 0.4252,
+      "step": 26371
+    },
+    {
+      "epoch": 0.6277254117871085,
+      "grad_norm": 4.500941824448517,
+      "learning_rate": 2e-05,
+      "loss": 1.0037,
+      "step": 26372
+    },
+    {
+      "epoch": 0.62774921451014,
+      "grad_norm": 0.6086717905714288,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 26373
+    },
+    {
+      "epoch": 0.6277730172331715,
+      "grad_norm": 4.812745676797997,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 26374
+    },
+    {
+      "epoch": 0.627796819956203,
+      "grad_norm": 0.7316534615751559,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 26375
+    },
+    {
+      "epoch": 0.6278206226792346,
+      "grad_norm": 3.3009443384831734,
+      "learning_rate": 2e-05,
+      "loss": 0.5023,
+      "step": 26376
+    },
+    {
+      "epoch": 0.627844425402266,
+      "grad_norm": 6.0890950559761485,
+      "learning_rate": 2e-05,
+      "loss": 1.2118,
+      "step": 26377
+    },
+    {
+      "epoch": 0.6278682281252975,
+      "grad_norm": 13.078696936053419,
+      "learning_rate": 2e-05,
+      "loss": 0.9023,
+      "step": 26378
+    },
+    {
+      "epoch": 0.627892030848329,
+      "grad_norm": 4.080660757223222,
+      "learning_rate": 2e-05,
+      "loss": 0.6995,
+      "step": 26379
+    },
+    {
+      "epoch": 0.6279158335713606,
+      "grad_norm": 2.1176031481468622,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 26380
+    },
+    {
+      "epoch": 0.6279396362943921,
+      "grad_norm": 4.5920717675050655,
+      "learning_rate": 2e-05,
+      "loss": 1.0621,
+      "step": 26381
+    },
+    {
+      "epoch": 0.6279634390174236,
+      "grad_norm": 5.351873724950273,
+      "learning_rate": 2e-05,
+      "loss": 0.1494,
+      "step": 26382
+    },
+    {
+      "epoch": 0.6279872417404551,
+      "grad_norm": 0.8428581268257966,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 26383
+    },
+    {
+      "epoch": 0.6280110444634867,
+      "grad_norm": 2.046432552056603,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 26384
+    },
+    {
+      "epoch": 0.6280348471865181,
+      "grad_norm": 4.415655476376585,
+      "learning_rate": 2e-05,
+      "loss": 0.7371,
+      "step": 26385
+    },
+    {
+      "epoch": 0.6280586499095496,
+      "grad_norm": 1.3545823759008957,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 26386
+    },
+    {
+      "epoch": 0.6280824526325811,
+      "grad_norm": 2.369115341288615,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 26387
+    },
+    {
+      "epoch": 0.6281062553556127,
+      "grad_norm": 3.487501671947495,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 26388
+    },
+    {
+      "epoch": 0.6281300580786442,
+      "grad_norm": 6.503893302779103,
+      "learning_rate": 2e-05,
+      "loss": 1.277,
+      "step": 26389
+    },
+    {
+      "epoch": 0.6281538608016757,
+      "grad_norm": 8.069505156096156,
+      "learning_rate": 2e-05,
+      "loss": 0.9481,
+      "step": 26390
+    },
+    {
+      "epoch": 0.6281776635247073,
+      "grad_norm": 7.012253191896257,
+      "learning_rate": 2e-05,
+      "loss": 0.4609,
+      "step": 26391
+    },
+    {
+      "epoch": 0.6282014662477388,
+      "grad_norm": 5.089565439049884,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 26392
+    },
+    {
+      "epoch": 0.6282252689707702,
+      "grad_norm": 1.585671349983924,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 26393
+    },
+    {
+      "epoch": 0.6282490716938017,
+      "grad_norm": 2.9596662595255094,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 26394
+    },
+    {
+      "epoch": 0.6282728744168333,
+      "grad_norm": 20.265924239334385,
+      "learning_rate": 2e-05,
+      "loss": 0.2699,
+      "step": 26395
+    },
+    {
+      "epoch": 0.6282966771398648,
+      "grad_norm": 2.343776438598563,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 26396
+    },
+    {
+      "epoch": 0.6283204798628963,
+      "grad_norm": 1.6264332446331,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 26397
+    },
+    {
+      "epoch": 0.6283442825859278,
+      "grad_norm": 0.7091050836963825,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 26398
+    },
+    {
+      "epoch": 0.6283680853089594,
+      "grad_norm": 0.0759893908986047,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26399
+    },
+    {
+      "epoch": 0.6283918880319909,
+      "grad_norm": 2.786736919073456,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 26400
+    },
+    {
+      "epoch": 0.6284156907550223,
+      "grad_norm": 3.6131929202750928,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 26401
+    },
+    {
+      "epoch": 0.6284394934780538,
+      "grad_norm": 13.657521621637482,
+      "learning_rate": 2e-05,
+      "loss": 1.2407,
+      "step": 26402
+    },
+    {
+      "epoch": 0.6284632962010854,
+      "grad_norm": 7.003760882052972,
+      "learning_rate": 2e-05,
+      "loss": 0.673,
+      "step": 26403
+    },
+    {
+      "epoch": 0.6284870989241169,
+      "grad_norm": 2.714071047283565,
+      "learning_rate": 2e-05,
+      "loss": 0.3705,
+      "step": 26404
+    },
+    {
+      "epoch": 0.6285109016471484,
+      "grad_norm": 4.858144423530252,
+      "learning_rate": 2e-05,
+      "loss": 0.9653,
+      "step": 26405
+    },
+    {
+      "epoch": 0.62853470437018,
+      "grad_norm": 5.758296216585653,
+      "learning_rate": 2e-05,
+      "loss": 0.7968,
+      "step": 26406
+    },
+    {
+      "epoch": 0.6285585070932115,
+      "grad_norm": 4.114300228542951,
+      "learning_rate": 2e-05,
+      "loss": 0.8882,
+      "step": 26407
+    },
+    {
+      "epoch": 0.628582309816243,
+      "grad_norm": 5.248525539664728,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 26408
+    },
+    {
+      "epoch": 0.6286061125392745,
+      "grad_norm": 2.8492337020029015,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 26409
+    },
+    {
+      "epoch": 0.628629915262306,
+      "grad_norm": 9.35135262118286,
+      "learning_rate": 2e-05,
+      "loss": 0.9343,
+      "step": 26410
+    },
+    {
+      "epoch": 0.6286537179853375,
+      "grad_norm": 32.585073790327144,
+      "learning_rate": 2e-05,
+      "loss": 0.2593,
+      "step": 26411
+    },
+    {
+      "epoch": 0.628677520708369,
+      "grad_norm": 4.226387484881696,
+      "learning_rate": 2e-05,
+      "loss": 0.436,
+      "step": 26412
+    },
+    {
+      "epoch": 0.6287013234314005,
+      "grad_norm": 4.37410858793952,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 26413
+    },
+    {
+      "epoch": 0.6287251261544321,
+      "grad_norm": 0.692773037162781,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 26414
+    },
+    {
+      "epoch": 0.6287489288774636,
+      "grad_norm": 0.8582856573887127,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 26415
+    },
+    {
+      "epoch": 0.6287727316004951,
+      "grad_norm": 1.2094907235652153,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 26416
+    },
+    {
+      "epoch": 0.6287965343235266,
+      "grad_norm": 5.0006191373551045,
+      "learning_rate": 2e-05,
+      "loss": 1.456,
+      "step": 26417
+    },
+    {
+      "epoch": 0.6288203370465582,
+      "grad_norm": 3.6391305915571164,
+      "learning_rate": 2e-05,
+      "loss": 0.4517,
+      "step": 26418
+    },
+    {
+      "epoch": 0.6288441397695896,
+      "grad_norm": 0.48406343002449265,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 26419
+    },
+    {
+      "epoch": 0.6288679424926211,
+      "grad_norm": 0.07523385882408923,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26420
+    },
+    {
+      "epoch": 0.6288917452156527,
+      "grad_norm": 9.450687365710053,
+      "learning_rate": 2e-05,
+      "loss": 0.8777,
+      "step": 26421
+    },
+    {
+      "epoch": 0.6289155479386842,
+      "grad_norm": 2.8704091545457175,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 26422
+    },
+    {
+      "epoch": 0.6289393506617157,
+      "grad_norm": 0.7819962990869211,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 26423
+    },
+    {
+      "epoch": 0.6289631533847472,
+      "grad_norm": 2.7851502898623304,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 26424
+    },
+    {
+      "epoch": 0.6289869561077788,
+      "grad_norm": 0.35895116732062116,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 26425
+    },
+    {
+      "epoch": 0.6290107588308103,
+      "grad_norm": 4.8337661314673905,
+      "learning_rate": 2e-05,
+      "loss": 0.8743,
+      "step": 26426
+    },
+    {
+      "epoch": 0.6290345615538417,
+      "grad_norm": 6.173892024686408,
+      "learning_rate": 2e-05,
+      "loss": 0.2761,
+      "step": 26427
+    },
+    {
+      "epoch": 0.6290583642768732,
+      "grad_norm": 0.48941288381496084,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 26428
+    },
+    {
+      "epoch": 0.6290821669999048,
+      "grad_norm": 3.9441030798322463,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 26429
+    },
+    {
+      "epoch": 0.6291059697229363,
+      "grad_norm": 2.1265692789572586,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 26430
+    },
+    {
+      "epoch": 0.6291297724459678,
+      "grad_norm": 0.09444179338543458,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26431
+    },
+    {
+      "epoch": 0.6291535751689993,
+      "grad_norm": 0.06637716543467242,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26432
+    },
+    {
+      "epoch": 0.6291773778920309,
+      "grad_norm": 3.3597948486205884,
+      "learning_rate": 2e-05,
+      "loss": 0.3785,
+      "step": 26433
+    },
+    {
+      "epoch": 0.6292011806150624,
+      "grad_norm": 5.399477411014538,
+      "learning_rate": 2e-05,
+      "loss": 0.3684,
+      "step": 26434
+    },
+    {
+      "epoch": 0.6292249833380938,
+      "grad_norm": 0.07989683869729035,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 26435
+    },
+    {
+      "epoch": 0.6292487860611254,
+      "grad_norm": 0.2745963097379672,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 26436
+    },
+    {
+      "epoch": 0.6292725887841569,
+      "grad_norm": 3.9804418798762615,
+      "learning_rate": 2e-05,
+      "loss": 0.6631,
+      "step": 26437
+    },
+    {
+      "epoch": 0.6292963915071884,
+      "grad_norm": 5.57296925877602,
+      "learning_rate": 2e-05,
+      "loss": 0.5217,
+      "step": 26438
+    },
+    {
+      "epoch": 0.6293201942302199,
+      "grad_norm": 0.09587348191585951,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 26439
+    },
+    {
+      "epoch": 0.6293439969532515,
+      "grad_norm": 13.40820200919398,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 26440
+    },
+    {
+      "epoch": 0.629367799676283,
+      "grad_norm": 0.29804561985254857,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 26441
+    },
+    {
+      "epoch": 0.6293916023993145,
+      "grad_norm": 3.6373602076047025,
+      "learning_rate": 2e-05,
+      "loss": 0.5398,
+      "step": 26442
+    },
+    {
+      "epoch": 0.629415405122346,
+      "grad_norm": 4.5398459802632,
+      "learning_rate": 2e-05,
+      "loss": 0.4911,
+      "step": 26443
+    },
+    {
+      "epoch": 0.6294392078453775,
+      "grad_norm": 0.13099550665356385,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 26444
+    },
+    {
+      "epoch": 0.629463010568409,
+      "grad_norm": 1.9522169003364656,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 26445
+    },
+    {
+      "epoch": 0.6294868132914405,
+      "grad_norm": 1.1547046982860063,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 26446
+    },
+    {
+      "epoch": 0.629510616014472,
+      "grad_norm": 2.8423448184227413,
+      "learning_rate": 2e-05,
+      "loss": 0.2058,
+      "step": 26447
+    },
+    {
+      "epoch": 0.6295344187375036,
+      "grad_norm": 9.297025983638944,
+      "learning_rate": 2e-05,
+      "loss": 0.2768,
+      "step": 26448
+    },
+    {
+      "epoch": 0.6295582214605351,
+      "grad_norm": 2.60215492856225,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 26449
+    },
+    {
+      "epoch": 0.6295820241835666,
+      "grad_norm": 7.965693981131,
+      "learning_rate": 2e-05,
+      "loss": 0.5443,
+      "step": 26450
+    },
+    {
+      "epoch": 0.6296058269065982,
+      "grad_norm": 5.288787160563436,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 26451
+    },
+    {
+      "epoch": 0.6296296296296297,
+      "grad_norm": 1.4697303285451462,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 26452
+    },
+    {
+      "epoch": 0.6296534323526611,
+      "grad_norm": 0.5144670606524971,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 26453
+    },
+    {
+      "epoch": 0.6296772350756926,
+      "grad_norm": 1.9189579196932736,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 26454
+    },
+    {
+      "epoch": 0.6297010377987242,
+      "grad_norm": 0.06728702338293477,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 26455
+    },
+    {
+      "epoch": 0.6297248405217557,
+      "grad_norm": 10.790357153728714,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 26456
+    },
+    {
+      "epoch": 0.6297486432447872,
+      "grad_norm": 4.007410277960089,
+      "learning_rate": 2e-05,
+      "loss": 0.4389,
+      "step": 26457
+    },
+    {
+      "epoch": 0.6297724459678187,
+      "grad_norm": 0.32389293562939375,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 26458
+    },
+    {
+      "epoch": 0.6297962486908503,
+      "grad_norm": 8.777398291651856,
+      "learning_rate": 2e-05,
+      "loss": 1.6617,
+      "step": 26459
+    },
+    {
+      "epoch": 0.6298200514138818,
+      "grad_norm": 4.847289965903843,
+      "learning_rate": 2e-05,
+      "loss": 0.1479,
+      "step": 26460
+    },
+    {
+      "epoch": 0.6298438541369132,
+      "grad_norm": 0.07729664165429062,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 26461
+    },
+    {
+      "epoch": 0.6298676568599447,
+      "grad_norm": 4.113568668726123,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 26462
+    },
+    {
+      "epoch": 0.6298914595829763,
+      "grad_norm": 0.5611213266166764,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 26463
+    },
+    {
+      "epoch": 0.6299152623060078,
+      "grad_norm": 10.091883718544532,
+      "learning_rate": 2e-05,
+      "loss": 0.4425,
+      "step": 26464
+    },
+    {
+      "epoch": 0.6299390650290393,
+      "grad_norm": 2.04733869022091,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 26465
+    },
+    {
+      "epoch": 0.6299628677520709,
+      "grad_norm": 3.0633743709239294,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 26466
+    },
+    {
+      "epoch": 0.6299866704751024,
+      "grad_norm": 4.401819530331118,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 26467
+    },
+    {
+      "epoch": 0.6300104731981339,
+      "grad_norm": 3.9854010214369358,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 26468
+    },
+    {
+      "epoch": 0.6300342759211653,
+      "grad_norm": 6.7037850070744085,
+      "learning_rate": 2e-05,
+      "loss": 1.4177,
+      "step": 26469
+    },
+    {
+      "epoch": 0.6300580786441969,
+      "grad_norm": 4.67984750348177,
+      "learning_rate": 2e-05,
+      "loss": 0.993,
+      "step": 26470
+    },
+    {
+      "epoch": 0.6300818813672284,
+      "grad_norm": 1.8803501851989797,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 26471
+    },
+    {
+      "epoch": 0.6301056840902599,
+      "grad_norm": 2.2996767936275893,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 26472
+    },
+    {
+      "epoch": 0.6301294868132914,
+      "grad_norm": 5.092595031870385,
+      "learning_rate": 2e-05,
+      "loss": 1.0761,
+      "step": 26473
+    },
+    {
+      "epoch": 0.630153289536323,
+      "grad_norm": 0.3679420444903901,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 26474
+    },
+    {
+      "epoch": 0.6301770922593545,
+      "grad_norm": 0.3362924926453554,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 26475
+    },
+    {
+      "epoch": 0.630200894982386,
+      "grad_norm": 9.371462656408484,
+      "learning_rate": 2e-05,
+      "loss": 0.9705,
+      "step": 26476
+    },
+    {
+      "epoch": 0.6302246977054174,
+      "grad_norm": 7.680310134390926,
+      "learning_rate": 2e-05,
+      "loss": 1.0777,
+      "step": 26477
+    },
+    {
+      "epoch": 0.630248500428449,
+      "grad_norm": 11.266007699900936,
+      "learning_rate": 2e-05,
+      "loss": 0.5239,
+      "step": 26478
+    },
+    {
+      "epoch": 0.6302723031514805,
+      "grad_norm": 3.945660762028501,
+      "learning_rate": 2e-05,
+      "loss": 0.5891,
+      "step": 26479
+    },
+    {
+      "epoch": 0.630296105874512,
+      "grad_norm": 3.8856424164193517,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 26480
+    },
+    {
+      "epoch": 0.6303199085975436,
+      "grad_norm": 3.1149117994013067,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 26481
+    },
+    {
+      "epoch": 0.6303437113205751,
+      "grad_norm": 3.8613692903823997,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 26482
+    },
+    {
+      "epoch": 0.6303675140436066,
+      "grad_norm": 4.360470447083768,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 26483
+    },
+    {
+      "epoch": 0.6303913167666381,
+      "grad_norm": 3.4951952220041487,
+      "learning_rate": 2e-05,
+      "loss": 0.4882,
+      "step": 26484
+    },
+    {
+      "epoch": 0.6304151194896697,
+      "grad_norm": 1.4276739683746336,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 26485
+    },
+    {
+      "epoch": 0.6304389222127011,
+      "grad_norm": 0.22042373594346495,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 26486
+    },
+    {
+      "epoch": 0.6304627249357326,
+      "grad_norm": 1.9218128632944163,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 26487
+    },
+    {
+      "epoch": 0.6304865276587641,
+      "grad_norm": 0.13137737890605747,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 26488
+    },
+    {
+      "epoch": 0.6305103303817957,
+      "grad_norm": 6.379144674159523,
+      "learning_rate": 2e-05,
+      "loss": 0.2189,
+      "step": 26489
+    },
+    {
+      "epoch": 0.6305341331048272,
+      "grad_norm": 5.3053565915166425,
+      "learning_rate": 2e-05,
+      "loss": 0.4096,
+      "step": 26490
+    },
+    {
+      "epoch": 0.6305579358278587,
+      "grad_norm": 0.19114051131788368,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 26491
+    },
+    {
+      "epoch": 0.6305817385508902,
+      "grad_norm": 0.11694904345523893,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 26492
+    },
+    {
+      "epoch": 0.6306055412739218,
+      "grad_norm": 4.849622294335034,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 26493
+    },
+    {
+      "epoch": 0.6306293439969533,
+      "grad_norm": 3.21752969371305,
+      "learning_rate": 2e-05,
+      "loss": 0.1019,
+      "step": 26494
+    },
+    {
+      "epoch": 0.6306531467199847,
+      "grad_norm": 7.5305692850654316,
+      "learning_rate": 2e-05,
+      "loss": 0.4236,
+      "step": 26495
+    },
+    {
+      "epoch": 0.6306769494430163,
+      "grad_norm": 11.685366748439213,
+      "learning_rate": 2e-05,
+      "loss": 0.7015,
+      "step": 26496
+    },
+    {
+      "epoch": 0.6307007521660478,
+      "grad_norm": 12.17206931570935,
+      "learning_rate": 2e-05,
+      "loss": 0.4109,
+      "step": 26497
+    },
+    {
+      "epoch": 0.6307245548890793,
+      "grad_norm": 0.545200933085904,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 26498
+    },
+    {
+      "epoch": 0.6307483576121108,
+      "grad_norm": 2.360082875067702,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 26499
+    },
+    {
+      "epoch": 0.6307721603351424,
+      "grad_norm": 4.017100777935534,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 26500
+    },
+    {
+      "epoch": 0.6307959630581739,
+      "grad_norm": 13.839198163965033,
+      "learning_rate": 2e-05,
+      "loss": 1.5249,
+      "step": 26501
+    },
+    {
+      "epoch": 0.6308197657812054,
+      "grad_norm": 0.07425735429538145,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 26502
+    },
+    {
+      "epoch": 0.6308435685042368,
+      "grad_norm": 5.851681657024843,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 26503
+    },
+    {
+      "epoch": 0.6308673712272684,
+      "grad_norm": 0.23745601579815148,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 26504
+    },
+    {
+      "epoch": 0.6308911739502999,
+      "grad_norm": 1.4188965103403508,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 26505
+    },
+    {
+      "epoch": 0.6309149766733314,
+      "grad_norm": 2.1371822473809696,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 26506
+    },
+    {
+      "epoch": 0.6309387793963629,
+      "grad_norm": 0.16436894826458537,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 26507
+    },
+    {
+      "epoch": 0.6309625821193945,
+      "grad_norm": 4.463548352269671,
+      "learning_rate": 2e-05,
+      "loss": 0.4836,
+      "step": 26508
+    },
+    {
+      "epoch": 0.630986384842426,
+      "grad_norm": 1.621820084656095,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 26509
+    },
+    {
+      "epoch": 0.6310101875654575,
+      "grad_norm": 2.027110883789399,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 26510
+    },
+    {
+      "epoch": 0.631033990288489,
+      "grad_norm": 0.07157816636671234,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 26511
+    },
+    {
+      "epoch": 0.6310577930115205,
+      "grad_norm": 3.5800574902714524,
+      "learning_rate": 2e-05,
+      "loss": 0.6151,
+      "step": 26512
+    },
+    {
+      "epoch": 0.631081595734552,
+      "grad_norm": 2.345763525538606,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 26513
+    },
+    {
+      "epoch": 0.6311053984575835,
+      "grad_norm": 9.414431187300327,
+      "learning_rate": 2e-05,
+      "loss": 1.4544,
+      "step": 26514
+    },
+    {
+      "epoch": 0.6311292011806151,
+      "grad_norm": 8.39891336600363,
+      "learning_rate": 2e-05,
+      "loss": 1.8148,
+      "step": 26515
+    },
+    {
+      "epoch": 0.6311530039036466,
+      "grad_norm": 0.6476207115370605,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 26516
+    },
+    {
+      "epoch": 0.6311768066266781,
+      "grad_norm": 2.9361226558806885,
+      "learning_rate": 2e-05,
+      "loss": 0.4865,
+      "step": 26517
+    },
+    {
+      "epoch": 0.6312006093497096,
+      "grad_norm": 0.8754116675131453,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 26518
+    },
+    {
+      "epoch": 0.6312244120727412,
+      "grad_norm": 5.463935463762045,
+      "learning_rate": 2e-05,
+      "loss": 0.5396,
+      "step": 26519
+    },
+    {
+      "epoch": 0.6312482147957726,
+      "grad_norm": 0.065185098654083,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 26520
+    },
+    {
+      "epoch": 0.6312720175188041,
+      "grad_norm": 1.5393546986811046,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 26521
+    },
+    {
+      "epoch": 0.6312958202418356,
+      "grad_norm": 2.236636894350463,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 26522
+    },
+    {
+      "epoch": 0.6313196229648672,
+      "grad_norm": 4.735890447730038,
+      "learning_rate": 2e-05,
+      "loss": 0.5941,
+      "step": 26523
+    },
+    {
+      "epoch": 0.6313434256878987,
+      "grad_norm": 1.3491741478905623,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 26524
+    },
+    {
+      "epoch": 0.6313672284109302,
+      "grad_norm": 4.184328402789115,
+      "learning_rate": 2e-05,
+      "loss": 0.5218,
+      "step": 26525
+    },
+    {
+      "epoch": 0.6313910311339618,
+      "grad_norm": 9.12475844208885,
+      "learning_rate": 2e-05,
+      "loss": 0.782,
+      "step": 26526
+    },
+    {
+      "epoch": 0.6314148338569933,
+      "grad_norm": 3.039038982835797,
+      "learning_rate": 2e-05,
+      "loss": 0.4972,
+      "step": 26527
+    },
+    {
+      "epoch": 0.6314386365800247,
+      "grad_norm": 7.698138907368726,
+      "learning_rate": 2e-05,
+      "loss": 1.2194,
+      "step": 26528
+    },
+    {
+      "epoch": 0.6314624393030562,
+      "grad_norm": 0.9163057286160013,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 26529
+    },
+    {
+      "epoch": 0.6314862420260878,
+      "grad_norm": 7.093794223097061,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 26530
+    },
+    {
+      "epoch": 0.6315100447491193,
+      "grad_norm": 2.1446041026067917,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 26531
+    },
+    {
+      "epoch": 0.6315338474721508,
+      "grad_norm": 3.2135296462916254,
+      "learning_rate": 2e-05,
+      "loss": 0.6671,
+      "step": 26532
+    },
+    {
+      "epoch": 0.6315576501951823,
+      "grad_norm": 5.064847676966718,
+      "learning_rate": 2e-05,
+      "loss": 0.1904,
+      "step": 26533
+    },
+    {
+      "epoch": 0.6315814529182139,
+      "grad_norm": 55.55176827063841,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 26534
+    },
+    {
+      "epoch": 0.6316052556412454,
+      "grad_norm": 3.951544392047596,
+      "learning_rate": 2e-05,
+      "loss": 0.7313,
+      "step": 26535
+    },
+    {
+      "epoch": 0.6316290583642769,
+      "grad_norm": 3.4485309638710806,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 26536
+    },
+    {
+      "epoch": 0.6316528610873083,
+      "grad_norm": 1.6200408032009685,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 26537
+    },
+    {
+      "epoch": 0.6316766638103399,
+      "grad_norm": 0.05989768782210049,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 26538
+    },
+    {
+      "epoch": 0.6317004665333714,
+      "grad_norm": 0.33381810118664657,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 26539
+    },
+    {
+      "epoch": 0.6317242692564029,
+      "grad_norm": 5.059782157227655,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 26540
+    },
+    {
+      "epoch": 0.6317480719794345,
+      "grad_norm": 1.3273550676135093,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 26541
+    },
+    {
+      "epoch": 0.631771874702466,
+      "grad_norm": 4.741181273571576,
+      "learning_rate": 2e-05,
+      "loss": 0.8942,
+      "step": 26542
+    },
+    {
+      "epoch": 0.6317956774254975,
+      "grad_norm": 0.033323154446945655,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 26543
+    },
+    {
+      "epoch": 0.631819480148529,
+      "grad_norm": 0.4540309591532028,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 26544
+    },
+    {
+      "epoch": 0.6318432828715606,
+      "grad_norm": 0.23989877876002694,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 26545
+    },
+    {
+      "epoch": 0.631867085594592,
+      "grad_norm": 4.634090783214929,
+      "learning_rate": 2e-05,
+      "loss": 0.5645,
+      "step": 26546
+    },
+    {
+      "epoch": 0.6318908883176235,
+      "grad_norm": 0.10900888941625357,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 26547
+    },
+    {
+      "epoch": 0.631914691040655,
+      "grad_norm": 0.21897587775908572,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 26548
+    },
+    {
+      "epoch": 0.6319384937636866,
+      "grad_norm": 2.9873129707816166,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 26549
+    },
+    {
+      "epoch": 0.6319622964867181,
+      "grad_norm": 11.839736325318238,
+      "learning_rate": 2e-05,
+      "loss": 0.854,
+      "step": 26550
+    },
+    {
+      "epoch": 0.6319860992097496,
+      "grad_norm": 4.211699441121639,
+      "learning_rate": 2e-05,
+      "loss": 0.7854,
+      "step": 26551
+    },
+    {
+      "epoch": 0.6320099019327811,
+      "grad_norm": 38.091609721850126,
+      "learning_rate": 2e-05,
+      "loss": 0.5524,
+      "step": 26552
+    },
+    {
+      "epoch": 0.6320337046558127,
+      "grad_norm": 0.14136207358207836,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 26553
+    },
+    {
+      "epoch": 0.6320575073788441,
+      "grad_norm": 1.0804341141604157,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 26554
+    },
+    {
+      "epoch": 0.6320813101018756,
+      "grad_norm": 3.211267464926883,
+      "learning_rate": 2e-05,
+      "loss": 0.0732,
+      "step": 26555
+    },
+    {
+      "epoch": 0.6321051128249072,
+      "grad_norm": 0.7711012488868082,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 26556
+    },
+    {
+      "epoch": 0.6321289155479387,
+      "grad_norm": 7.568187108035825,
+      "learning_rate": 2e-05,
+      "loss": 1.1409,
+      "step": 26557
+    },
+    {
+      "epoch": 0.6321527182709702,
+      "grad_norm": 4.697946626772025,
+      "learning_rate": 2e-05,
+      "loss": 0.6536,
+      "step": 26558
+    },
+    {
+      "epoch": 0.6321765209940017,
+      "grad_norm": 5.269907865961903,
+      "learning_rate": 2e-05,
+      "loss": 0.1463,
+      "step": 26559
+    },
+    {
+      "epoch": 0.6322003237170333,
+      "grad_norm": 1.3981535346250302,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 26560
+    },
+    {
+      "epoch": 0.6322241264400648,
+      "grad_norm": 3.633990294093739,
+      "learning_rate": 2e-05,
+      "loss": 0.7277,
+      "step": 26561
+    },
+    {
+      "epoch": 0.6322479291630962,
+      "grad_norm": 5.197758235112276,
+      "learning_rate": 2e-05,
+      "loss": 1.2679,
+      "step": 26562
+    },
+    {
+      "epoch": 0.6322717318861277,
+      "grad_norm": 2.555784291506209,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 26563
+    },
+    {
+      "epoch": 0.6322955346091593,
+      "grad_norm": 0.46005980461947515,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 26564
+    },
+    {
+      "epoch": 0.6323193373321908,
+      "grad_norm": 13.440979730530602,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 26565
+    },
+    {
+      "epoch": 0.6323431400552223,
+      "grad_norm": 12.24637435885336,
+      "learning_rate": 2e-05,
+      "loss": 1.4828,
+      "step": 26566
+    },
+    {
+      "epoch": 0.6323669427782538,
+      "grad_norm": 5.185266713458156,
+      "learning_rate": 2e-05,
+      "loss": 0.8042,
+      "step": 26567
+    },
+    {
+      "epoch": 0.6323907455012854,
+      "grad_norm": 2.874819266713366,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 26568
+    },
+    {
+      "epoch": 0.6324145482243169,
+      "grad_norm": 21.497246004722435,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 26569
+    },
+    {
+      "epoch": 0.6324383509473483,
+      "grad_norm": 0.1367841673675526,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 26570
+    },
+    {
+      "epoch": 0.63246215367038,
+      "grad_norm": 5.6504301028682,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 26571
+    },
+    {
+      "epoch": 0.6324859563934114,
+      "grad_norm": 2.5881606136533954,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 26572
+    },
+    {
+      "epoch": 0.6325097591164429,
+      "grad_norm": 2.725326195936787,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 26573
+    },
+    {
+      "epoch": 0.6325335618394744,
+      "grad_norm": 18.01797798131373,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 26574
+    },
+    {
+      "epoch": 0.632557364562506,
+      "grad_norm": 0.01883886697490458,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 26575
+    },
+    {
+      "epoch": 0.6325811672855375,
+      "grad_norm": 0.5292929121217714,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 26576
+    },
+    {
+      "epoch": 0.632604970008569,
+      "grad_norm": 5.532454713039422,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 26577
+    },
+    {
+      "epoch": 0.6326287727316005,
+      "grad_norm": 2.733181390296489,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 26578
+    },
+    {
+      "epoch": 0.632652575454632,
+      "grad_norm": 2.7890510543463276,
+      "learning_rate": 2e-05,
+      "loss": 0.2895,
+      "step": 26579
+    },
+    {
+      "epoch": 0.6326763781776635,
+      "grad_norm": 0.41432534780826397,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 26580
+    },
+    {
+      "epoch": 0.632700180900695,
+      "grad_norm": 2.4719689967574476,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 26581
+    },
+    {
+      "epoch": 0.6327239836237265,
+      "grad_norm": 3.6820253880800253,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 26582
+    },
+    {
+      "epoch": 0.6327477863467581,
+      "grad_norm": 84.90072086416242,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 26583
+    },
+    {
+      "epoch": 0.6327715890697896,
+      "grad_norm": 4.404972325898798,
+      "learning_rate": 2e-05,
+      "loss": 0.8974,
+      "step": 26584
+    },
+    {
+      "epoch": 0.6327953917928211,
+      "grad_norm": 0.24551117159599617,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 26585
+    },
+    {
+      "epoch": 0.6328191945158527,
+      "grad_norm": 3.081569458651689,
+      "learning_rate": 2e-05,
+      "loss": 0.4384,
+      "step": 26586
+    },
+    {
+      "epoch": 0.6328429972388842,
+      "grad_norm": 2.065908157064574,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 26587
+    },
+    {
+      "epoch": 0.6328667999619156,
+      "grad_norm": 3.3939240340118135,
+      "learning_rate": 2e-05,
+      "loss": 0.1809,
+      "step": 26588
+    },
+    {
+      "epoch": 0.6328906026849471,
+      "grad_norm": 7.419388965941536,
+      "learning_rate": 2e-05,
+      "loss": 1.4675,
+      "step": 26589
+    },
+    {
+      "epoch": 0.6329144054079787,
+      "grad_norm": 7.084245992669849,
+      "learning_rate": 2e-05,
+      "loss": 0.82,
+      "step": 26590
+    },
+    {
+      "epoch": 0.6329382081310102,
+      "grad_norm": 6.512720419497247,
+      "learning_rate": 2e-05,
+      "loss": 1.1819,
+      "step": 26591
+    },
+    {
+      "epoch": 0.6329620108540417,
+      "grad_norm": 2.885310492082547,
+      "learning_rate": 2e-05,
+      "loss": 0.4276,
+      "step": 26592
+    },
+    {
+      "epoch": 0.6329858135770732,
+      "grad_norm": 5.431239038071458,
+      "learning_rate": 2e-05,
+      "loss": 1.0501,
+      "step": 26593
+    },
+    {
+      "epoch": 0.6330096163001048,
+      "grad_norm": 5.539714189045644,
+      "learning_rate": 2e-05,
+      "loss": 0.9734,
+      "step": 26594
+    },
+    {
+      "epoch": 0.6330334190231363,
+      "grad_norm": 6.487564446589778,
+      "learning_rate": 2e-05,
+      "loss": 0.6566,
+      "step": 26595
+    },
+    {
+      "epoch": 0.6330572217461677,
+      "grad_norm": 3.836704153426463,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 26596
+    },
+    {
+      "epoch": 0.6330810244691992,
+      "grad_norm": 4.350058455340211,
+      "learning_rate": 2e-05,
+      "loss": 0.431,
+      "step": 26597
+    },
+    {
+      "epoch": 0.6331048271922308,
+      "grad_norm": 0.5064253927874641,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 26598
+    },
+    {
+      "epoch": 0.6331286299152623,
+      "grad_norm": 5.55181915463386,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 26599
+    },
+    {
+      "epoch": 0.6331524326382938,
+      "grad_norm": 0.6746528498762084,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 26600
+    },
+    {
+      "epoch": 0.6331762353613254,
+      "grad_norm": 4.164549010377275,
+      "learning_rate": 2e-05,
+      "loss": 0.7633,
+      "step": 26601
+    },
+    {
+      "epoch": 0.6332000380843569,
+      "grad_norm": 4.549070504277083,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 26602
+    },
+    {
+      "epoch": 0.6332238408073884,
+      "grad_norm": 5.19668455626602,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 26603
+    },
+    {
+      "epoch": 0.6332476435304198,
+      "grad_norm": 0.9830960092605414,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 26604
+    },
+    {
+      "epoch": 0.6332714462534514,
+      "grad_norm": 3.8077411902252165,
+      "learning_rate": 2e-05,
+      "loss": 0.5024,
+      "step": 26605
+    },
+    {
+      "epoch": 0.6332952489764829,
+      "grad_norm": 7.0490970339777865,
+      "learning_rate": 2e-05,
+      "loss": 0.2117,
+      "step": 26606
+    },
+    {
+      "epoch": 0.6333190516995144,
+      "grad_norm": 3.0227360637269487,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 26607
+    },
+    {
+      "epoch": 0.6333428544225459,
+      "grad_norm": 5.516919640392798,
+      "learning_rate": 2e-05,
+      "loss": 0.8801,
+      "step": 26608
+    },
+    {
+      "epoch": 0.6333666571455775,
+      "grad_norm": 5.705925293387842,
+      "learning_rate": 2e-05,
+      "loss": 0.3187,
+      "step": 26609
+    },
+    {
+      "epoch": 0.633390459868609,
+      "grad_norm": 0.8931561601404764,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 26610
+    },
+    {
+      "epoch": 0.6334142625916405,
+      "grad_norm": 4.388384537628975,
+      "learning_rate": 2e-05,
+      "loss": 0.1807,
+      "step": 26611
+    },
+    {
+      "epoch": 0.633438065314672,
+      "grad_norm": 9.117437595843402,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 26612
+    },
+    {
+      "epoch": 0.6334618680377035,
+      "grad_norm": 2.002613324707035,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 26613
+    },
+    {
+      "epoch": 0.633485670760735,
+      "grad_norm": 3.947243618998032,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 26614
+    },
+    {
+      "epoch": 0.6335094734837665,
+      "grad_norm": 3.8538997526214636,
+      "learning_rate": 2e-05,
+      "loss": 0.3052,
+      "step": 26615
+    },
+    {
+      "epoch": 0.6335332762067981,
+      "grad_norm": 6.006230551353076,
+      "learning_rate": 2e-05,
+      "loss": 0.6748,
+      "step": 26616
+    },
+    {
+      "epoch": 0.6335570789298296,
+      "grad_norm": 5.707086449392671,
+      "learning_rate": 2e-05,
+      "loss": 1.4437,
+      "step": 26617
+    },
+    {
+      "epoch": 0.6335808816528611,
+      "grad_norm": 2.9696350710606465,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 26618
+    },
+    {
+      "epoch": 0.6336046843758926,
+      "grad_norm": 0.10858589464492263,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 26619
+    },
+    {
+      "epoch": 0.6336284870989242,
+      "grad_norm": 0.6217165967448137,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 26620
+    },
+    {
+      "epoch": 0.6336522898219556,
+      "grad_norm": 5.068779164879493,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 26621
+    },
+    {
+      "epoch": 0.6336760925449871,
+      "grad_norm": 1.845745379022483,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 26622
+    },
+    {
+      "epoch": 0.6336998952680186,
+      "grad_norm": 1.7557529832482772,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 26623
+    },
+    {
+      "epoch": 0.6337236979910502,
+      "grad_norm": 2.9592655495335825,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 26624
+    },
+    {
+      "epoch": 0.6337475007140817,
+      "grad_norm": 2.1753840381059137,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 26625
+    },
+    {
+      "epoch": 0.6337713034371132,
+      "grad_norm": 2.4043807088397156,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 26626
+    },
+    {
+      "epoch": 0.6337951061601447,
+      "grad_norm": 2.88053591538058,
+      "learning_rate": 2e-05,
+      "loss": 0.0858,
+      "step": 26627
+    },
+    {
+      "epoch": 0.6338189088831763,
+      "grad_norm": 3.9328952613290373,
+      "learning_rate": 2e-05,
+      "loss": 0.2562,
+      "step": 26628
+    },
+    {
+      "epoch": 0.6338427116062078,
+      "grad_norm": 4.835885930409889,
+      "learning_rate": 2e-05,
+      "loss": 0.8883,
+      "step": 26629
+    },
+    {
+      "epoch": 0.6338665143292392,
+      "grad_norm": 0.5051852252940126,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 26630
+    },
+    {
+      "epoch": 0.6338903170522708,
+      "grad_norm": 6.588825693094805,
+      "learning_rate": 2e-05,
+      "loss": 0.9014,
+      "step": 26631
+    },
+    {
+      "epoch": 0.6339141197753023,
+      "grad_norm": 4.292309053791485,
+      "learning_rate": 2e-05,
+      "loss": 1.0283,
+      "step": 26632
+    },
+    {
+      "epoch": 0.6339379224983338,
+      "grad_norm": 6.882193948641681,
+      "learning_rate": 2e-05,
+      "loss": 0.337,
+      "step": 26633
+    },
+    {
+      "epoch": 0.6339617252213653,
+      "grad_norm": 0.17965995586631778,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 26634
+    },
+    {
+      "epoch": 0.6339855279443969,
+      "grad_norm": 4.804288284409387,
+      "learning_rate": 2e-05,
+      "loss": 0.5331,
+      "step": 26635
+    },
+    {
+      "epoch": 0.6340093306674284,
+      "grad_norm": 0.8551534467107436,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 26636
+    },
+    {
+      "epoch": 0.6340331333904599,
+      "grad_norm": 3.168669597171925,
+      "learning_rate": 2e-05,
+      "loss": 0.3247,
+      "step": 26637
+    },
+    {
+      "epoch": 0.6340569361134913,
+      "grad_norm": 0.11644460698521926,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 26638
+    },
+    {
+      "epoch": 0.6340807388365229,
+      "grad_norm": 4.9884485794063895,
+      "learning_rate": 2e-05,
+      "loss": 0.9129,
+      "step": 26639
+    },
+    {
+      "epoch": 0.6341045415595544,
+      "grad_norm": 4.24736385492405,
+      "learning_rate": 2e-05,
+      "loss": 0.7794,
+      "step": 26640
+    },
+    {
+      "epoch": 0.6341283442825859,
+      "grad_norm": 26.88383497435312,
+      "learning_rate": 2e-05,
+      "loss": 0.7408,
+      "step": 26641
+    },
+    {
+      "epoch": 0.6341521470056174,
+      "grad_norm": 0.6604764859602742,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 26642
+    },
+    {
+      "epoch": 0.634175949728649,
+      "grad_norm": 3.521659707189647,
+      "learning_rate": 2e-05,
+      "loss": 0.6908,
+      "step": 26643
+    },
+    {
+      "epoch": 0.6341997524516805,
+      "grad_norm": 4.648803791013375,
+      "learning_rate": 2e-05,
+      "loss": 0.7508,
+      "step": 26644
+    },
+    {
+      "epoch": 0.634223555174712,
+      "grad_norm": 0.09352683023103125,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 26645
+    },
+    {
+      "epoch": 0.6342473578977434,
+      "grad_norm": 9.667439575975298,
+      "learning_rate": 2e-05,
+      "loss": 0.7006,
+      "step": 26646
+    },
+    {
+      "epoch": 0.634271160620775,
+      "grad_norm": 6.66653918346483,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 26647
+    },
+    {
+      "epoch": 0.6342949633438065,
+      "grad_norm": 4.068636534845709,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 26648
+    },
+    {
+      "epoch": 0.634318766066838,
+      "grad_norm": 3.8204022522053767,
+      "learning_rate": 2e-05,
+      "loss": 0.2239,
+      "step": 26649
+    },
+    {
+      "epoch": 0.6343425687898696,
+      "grad_norm": 4.087742660886168,
+      "learning_rate": 2e-05,
+      "loss": 0.3161,
+      "step": 26650
+    },
+    {
+      "epoch": 0.6343663715129011,
+      "grad_norm": 7.3938066810822045,
+      "learning_rate": 2e-05,
+      "loss": 0.5376,
+      "step": 26651
+    },
+    {
+      "epoch": 0.6343901742359326,
+      "grad_norm": 3.015121650321991,
+      "learning_rate": 2e-05,
+      "loss": 0.4917,
+      "step": 26652
+    },
+    {
+      "epoch": 0.6344139769589641,
+      "grad_norm": 0.5685860054363233,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 26653
+    },
+    {
+      "epoch": 0.6344377796819957,
+      "grad_norm": 2.7162483020798787,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 26654
+    },
+    {
+      "epoch": 0.6344615824050271,
+      "grad_norm": 0.6957189510531,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 26655
+    },
+    {
+      "epoch": 0.6344853851280586,
+      "grad_norm": 8.147528747917596,
+      "learning_rate": 2e-05,
+      "loss": 0.1526,
+      "step": 26656
+    },
+    {
+      "epoch": 0.6345091878510901,
+      "grad_norm": 0.7960602288687477,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 26657
+    },
+    {
+      "epoch": 0.6345329905741217,
+      "grad_norm": 2.8334378989608484,
+      "learning_rate": 2e-05,
+      "loss": 0.4219,
+      "step": 26658
+    },
+    {
+      "epoch": 0.6345567932971532,
+      "grad_norm": 7.598266782788432,
+      "learning_rate": 2e-05,
+      "loss": 1.6157,
+      "step": 26659
+    },
+    {
+      "epoch": 0.6345805960201847,
+      "grad_norm": 12.687189048412769,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 26660
+    },
+    {
+      "epoch": 0.6346043987432162,
+      "grad_norm": 1.7330741342410645,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 26661
+    },
+    {
+      "epoch": 0.6346282014662478,
+      "grad_norm": 2.172338604717554,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 26662
+    },
+    {
+      "epoch": 0.6346520041892793,
+      "grad_norm": 7.423370435931913,
+      "learning_rate": 2e-05,
+      "loss": 1.4414,
+      "step": 26663
+    },
+    {
+      "epoch": 0.6346758069123107,
+      "grad_norm": 0.162861878804219,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 26664
+    },
+    {
+      "epoch": 0.6346996096353423,
+      "grad_norm": 0.5203432806997059,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 26665
+    },
+    {
+      "epoch": 0.6347234123583738,
+      "grad_norm": 4.124836916582147,
+      "learning_rate": 2e-05,
+      "loss": 0.5476,
+      "step": 26666
+    },
+    {
+      "epoch": 0.6347472150814053,
+      "grad_norm": 3.00203559329545,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 26667
+    },
+    {
+      "epoch": 0.6347710178044368,
+      "grad_norm": 7.682837392384928,
+      "learning_rate": 2e-05,
+      "loss": 0.1474,
+      "step": 26668
+    },
+    {
+      "epoch": 0.6347948205274684,
+      "grad_norm": 4.159317422140092,
+      "learning_rate": 2e-05,
+      "loss": 0.572,
+      "step": 26669
+    },
+    {
+      "epoch": 0.6348186232504999,
+      "grad_norm": 5.87115795252833,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 26670
+    },
+    {
+      "epoch": 0.6348424259735314,
+      "grad_norm": 1.929111126011244,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 26671
+    },
+    {
+      "epoch": 0.6348662286965628,
+      "grad_norm": 5.19652240815812,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 26672
+    },
+    {
+      "epoch": 0.6348900314195944,
+      "grad_norm": 1.8474912916946038,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 26673
+    },
+    {
+      "epoch": 0.6349138341426259,
+      "grad_norm": 3.5053994428543795,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 26674
+    },
+    {
+      "epoch": 0.6349376368656574,
+      "grad_norm": 0.7509386563956622,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 26675
+    },
+    {
+      "epoch": 0.6349614395886889,
+      "grad_norm": 1.2019289503002555,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 26676
+    },
+    {
+      "epoch": 0.6349852423117205,
+      "grad_norm": 6.611128447951044,
+      "learning_rate": 2e-05,
+      "loss": 0.1591,
+      "step": 26677
+    },
+    {
+      "epoch": 0.635009045034752,
+      "grad_norm": 0.5441139305035525,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 26678
+    },
+    {
+      "epoch": 0.6350328477577835,
+      "grad_norm": 5.611077794654775,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 26679
+    },
+    {
+      "epoch": 0.635056650480815,
+      "grad_norm": 2.5176923286229225,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 26680
+    },
+    {
+      "epoch": 0.6350804532038465,
+      "grad_norm": 5.389507274662299,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 26681
+    },
+    {
+      "epoch": 0.635104255926878,
+      "grad_norm": 3.6191852140939162,
+      "learning_rate": 2e-05,
+      "loss": 0.4896,
+      "step": 26682
+    },
+    {
+      "epoch": 0.6351280586499095,
+      "grad_norm": 4.6849736525276064,
+      "learning_rate": 2e-05,
+      "loss": 0.8981,
+      "step": 26683
+    },
+    {
+      "epoch": 0.6351518613729411,
+      "grad_norm": 14.174480173093782,
+      "learning_rate": 2e-05,
+      "loss": 0.8111,
+      "step": 26684
+    },
+    {
+      "epoch": 0.6351756640959726,
+      "grad_norm": 4.231483746191371,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 26685
+    },
+    {
+      "epoch": 0.6351994668190041,
+      "grad_norm": 7.203033228182244,
+      "learning_rate": 2e-05,
+      "loss": 0.7685,
+      "step": 26686
+    },
+    {
+      "epoch": 0.6352232695420356,
+      "grad_norm": 2.271886134602802,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 26687
+    },
+    {
+      "epoch": 0.6352470722650672,
+      "grad_norm": 3.827129489789016,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 26688
+    },
+    {
+      "epoch": 0.6352708749880986,
+      "grad_norm": 0.8241136863919083,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 26689
+    },
+    {
+      "epoch": 0.6352946777111301,
+      "grad_norm": 3.7798661642071925,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 26690
+    },
+    {
+      "epoch": 0.6353184804341616,
+      "grad_norm": 4.440097574195503,
+      "learning_rate": 2e-05,
+      "loss": 0.5423,
+      "step": 26691
+    },
+    {
+      "epoch": 0.6353422831571932,
+      "grad_norm": 1.6305795442456918,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 26692
+    },
+    {
+      "epoch": 0.6353660858802247,
+      "grad_norm": 6.091092467482877,
+      "learning_rate": 2e-05,
+      "loss": 0.8409,
+      "step": 26693
+    },
+    {
+      "epoch": 0.6353898886032562,
+      "grad_norm": 0.9765128941295754,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 26694
+    },
+    {
+      "epoch": 0.6354136913262878,
+      "grad_norm": 0.5750025303272057,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 26695
+    },
+    {
+      "epoch": 0.6354374940493193,
+      "grad_norm": 9.43599149956305,
+      "learning_rate": 2e-05,
+      "loss": 0.174,
+      "step": 26696
+    },
+    {
+      "epoch": 0.6354612967723507,
+      "grad_norm": 5.509667023425315,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 26697
+    },
+    {
+      "epoch": 0.6354850994953822,
+      "grad_norm": 1.0722139602811545,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 26698
+    },
+    {
+      "epoch": 0.6355089022184138,
+      "grad_norm": 6.06991066319864,
+      "learning_rate": 2e-05,
+      "loss": 1.2097,
+      "step": 26699
+    },
+    {
+      "epoch": 0.6355327049414453,
+      "grad_norm": 3.206585108331771,
+      "learning_rate": 2e-05,
+      "loss": 0.6526,
+      "step": 26700
+    },
+    {
+      "epoch": 0.6355565076644768,
+      "grad_norm": 2.3154226023277675,
+      "learning_rate": 2e-05,
+      "loss": 0.3008,
+      "step": 26701
+    },
+    {
+      "epoch": 0.6355803103875083,
+      "grad_norm": 0.3670544443518567,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 26702
+    },
+    {
+      "epoch": 0.6356041131105399,
+      "grad_norm": 3.2801903391913547,
+      "learning_rate": 2e-05,
+      "loss": 0.3416,
+      "step": 26703
+    },
+    {
+      "epoch": 0.6356279158335714,
+      "grad_norm": 1.9770180324969615,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 26704
+    },
+    {
+      "epoch": 0.6356517185566029,
+      "grad_norm": 3.282833988412741,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 26705
+    },
+    {
+      "epoch": 0.6356755212796343,
+      "grad_norm": 0.652711376385682,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 26706
+    },
+    {
+      "epoch": 0.6356993240026659,
+      "grad_norm": 0.8665565220002988,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 26707
+    },
+    {
+      "epoch": 0.6357231267256974,
+      "grad_norm": 1.783507950227776,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 26708
+    },
+    {
+      "epoch": 0.6357469294487289,
+      "grad_norm": 2.8783715903216653,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 26709
+    },
+    {
+      "epoch": 0.6357707321717605,
+      "grad_norm": 2.8475379112832324,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 26710
+    },
+    {
+      "epoch": 0.635794534894792,
+      "grad_norm": 0.10986463687514957,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 26711
+    },
+    {
+      "epoch": 0.6358183376178235,
+      "grad_norm": 1.6915379516785984,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 26712
+    },
+    {
+      "epoch": 0.635842140340855,
+      "grad_norm": 1.1035893680427362,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 26713
+    },
+    {
+      "epoch": 0.6358659430638866,
+      "grad_norm": 1.767880635625884,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 26714
+    },
+    {
+      "epoch": 0.635889745786918,
+      "grad_norm": 5.718523474492837,
+      "learning_rate": 2e-05,
+      "loss": 1.0156,
+      "step": 26715
+    },
+    {
+      "epoch": 0.6359135485099495,
+      "grad_norm": 7.917064579125032,
+      "learning_rate": 2e-05,
+      "loss": 0.2684,
+      "step": 26716
+    },
+    {
+      "epoch": 0.635937351232981,
+      "grad_norm": 2.793630017058929,
+      "learning_rate": 2e-05,
+      "loss": 0.4617,
+      "step": 26717
+    },
+    {
+      "epoch": 0.6359611539560126,
+      "grad_norm": 3.303432148292086,
+      "learning_rate": 2e-05,
+      "loss": 0.7205,
+      "step": 26718
+    },
+    {
+      "epoch": 0.6359849566790441,
+      "grad_norm": 3.871200744176814,
+      "learning_rate": 2e-05,
+      "loss": 0.5475,
+      "step": 26719
+    },
+    {
+      "epoch": 0.6360087594020756,
+      "grad_norm": 4.021792112471368,
+      "learning_rate": 2e-05,
+      "loss": 0.6544,
+      "step": 26720
+    },
+    {
+      "epoch": 0.6360325621251071,
+      "grad_norm": 2.6885957297550296,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 26721
+    },
+    {
+      "epoch": 0.6360563648481387,
+      "grad_norm": 5.457066480261014,
+      "learning_rate": 2e-05,
+      "loss": 0.8619,
+      "step": 26722
+    },
+    {
+      "epoch": 0.6360801675711701,
+      "grad_norm": 0.8824467069194499,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 26723
+    },
+    {
+      "epoch": 0.6361039702942016,
+      "grad_norm": 2.015292222127728,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 26724
+    },
+    {
+      "epoch": 0.6361277730172332,
+      "grad_norm": 4.2454024493440174,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 26725
+    },
+    {
+      "epoch": 0.6361515757402647,
+      "grad_norm": 0.01484860624996755,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 26726
+    },
+    {
+      "epoch": 0.6361753784632962,
+      "grad_norm": 2.0374326268996463,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 26727
+    },
+    {
+      "epoch": 0.6361991811863277,
+      "grad_norm": 2.765100183348495,
+      "learning_rate": 2e-05,
+      "loss": 0.2951,
+      "step": 26728
+    },
+    {
+      "epoch": 0.6362229839093593,
+      "grad_norm": 2.6563698265535662,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 26729
+    },
+    {
+      "epoch": 0.6362467866323908,
+      "grad_norm": 3.467792287214208,
+      "learning_rate": 2e-05,
+      "loss": 0.4748,
+      "step": 26730
+    },
+    {
+      "epoch": 0.6362705893554222,
+      "grad_norm": 1.3742034269089431,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 26731
+    },
+    {
+      "epoch": 0.6362943920784537,
+      "grad_norm": 0.03463378167590768,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 26732
+    },
+    {
+      "epoch": 0.6363181948014853,
+      "grad_norm": 9.513300586621748,
+      "learning_rate": 2e-05,
+      "loss": 0.233,
+      "step": 26733
+    },
+    {
+      "epoch": 0.6363419975245168,
+      "grad_norm": 2.818287927118027,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 26734
+    },
+    {
+      "epoch": 0.6363658002475483,
+      "grad_norm": 0.44750667621951845,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 26735
+    },
+    {
+      "epoch": 0.6363896029705798,
+      "grad_norm": 4.830900576147444,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 26736
+    },
+    {
+      "epoch": 0.6364134056936114,
+      "grad_norm": 4.399829961494469,
+      "learning_rate": 2e-05,
+      "loss": 1.1294,
+      "step": 26737
+    },
+    {
+      "epoch": 0.6364372084166429,
+      "grad_norm": 6.676896604121712,
+      "learning_rate": 2e-05,
+      "loss": 0.8382,
+      "step": 26738
+    },
+    {
+      "epoch": 0.6364610111396743,
+      "grad_norm": 20.848016151376328,
+      "learning_rate": 2e-05,
+      "loss": 1.037,
+      "step": 26739
+    },
+    {
+      "epoch": 0.636484813862706,
+      "grad_norm": 0.43953575377551884,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 26740
+    },
+    {
+      "epoch": 0.6365086165857374,
+      "grad_norm": 3.5580968724620754,
+      "learning_rate": 2e-05,
+      "loss": 0.569,
+      "step": 26741
+    },
+    {
+      "epoch": 0.6365324193087689,
+      "grad_norm": 0.7875751690237188,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 26742
+    },
+    {
+      "epoch": 0.6365562220318004,
+      "grad_norm": 4.956986534223213,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 26743
+    },
+    {
+      "epoch": 0.636580024754832,
+      "grad_norm": 1.1548343129587817,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 26744
+    },
+    {
+      "epoch": 0.6366038274778635,
+      "grad_norm": 4.004546382671789,
+      "learning_rate": 2e-05,
+      "loss": 0.6398,
+      "step": 26745
+    },
+    {
+      "epoch": 0.636627630200895,
+      "grad_norm": 0.46202563626357634,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 26746
+    },
+    {
+      "epoch": 0.6366514329239265,
+      "grad_norm": 3.286977046115916,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 26747
+    },
+    {
+      "epoch": 0.636675235646958,
+      "grad_norm": 3.3507922134908457,
+      "learning_rate": 2e-05,
+      "loss": 0.2718,
+      "step": 26748
+    },
+    {
+      "epoch": 0.6366990383699895,
+      "grad_norm": 12.406747956099444,
+      "learning_rate": 2e-05,
+      "loss": 0.6349,
+      "step": 26749
+    },
+    {
+      "epoch": 0.636722841093021,
+      "grad_norm": 0.7950288733495268,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 26750
+    },
+    {
+      "epoch": 0.6367466438160525,
+      "grad_norm": 5.52722904030024,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 26751
+    },
+    {
+      "epoch": 0.6367704465390841,
+      "grad_norm": 2.925266040949297,
+      "learning_rate": 2e-05,
+      "loss": 0.6652,
+      "step": 26752
+    },
+    {
+      "epoch": 0.6367942492621156,
+      "grad_norm": 8.123403309456862,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 26753
+    },
+    {
+      "epoch": 0.6368180519851471,
+      "grad_norm": 0.1709589498998413,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 26754
+    },
+    {
+      "epoch": 0.6368418547081787,
+      "grad_norm": 3.4312021604511624,
+      "learning_rate": 2e-05,
+      "loss": 0.3589,
+      "step": 26755
+    },
+    {
+      "epoch": 0.6368656574312102,
+      "grad_norm": 4.335975701820246,
+      "learning_rate": 2e-05,
+      "loss": 0.3288,
+      "step": 26756
+    },
+    {
+      "epoch": 0.6368894601542416,
+      "grad_norm": 4.420487513596669,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 26757
+    },
+    {
+      "epoch": 0.6369132628772731,
+      "grad_norm": 4.65573025956196,
+      "learning_rate": 2e-05,
+      "loss": 0.2013,
+      "step": 26758
+    },
+    {
+      "epoch": 0.6369370656003047,
+      "grad_norm": 5.653973311165107,
+      "learning_rate": 2e-05,
+      "loss": 0.6104,
+      "step": 26759
+    },
+    {
+      "epoch": 0.6369608683233362,
+      "grad_norm": 3.4249506542770116,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 26760
+    },
+    {
+      "epoch": 0.6369846710463677,
+      "grad_norm": 0.06940479816196438,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 26761
+    },
+    {
+      "epoch": 0.6370084737693992,
+      "grad_norm": 0.7181829107679842,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 26762
+    },
+    {
+      "epoch": 0.6370322764924308,
+      "grad_norm": 2.7157417312000836,
+      "learning_rate": 2e-05,
+      "loss": 0.3938,
+      "step": 26763
+    },
+    {
+      "epoch": 0.6370560792154623,
+      "grad_norm": 2.7629623749322048,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 26764
+    },
+    {
+      "epoch": 0.6370798819384937,
+      "grad_norm": 2.2597194063825965,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 26765
+    },
+    {
+      "epoch": 0.6371036846615252,
+      "grad_norm": 19.212765491591128,
+      "learning_rate": 2e-05,
+      "loss": 1.0707,
+      "step": 26766
+    },
+    {
+      "epoch": 0.6371274873845568,
+      "grad_norm": 6.1254916615730455,
+      "learning_rate": 2e-05,
+      "loss": 1.1694,
+      "step": 26767
+    },
+    {
+      "epoch": 0.6371512901075883,
+      "grad_norm": 5.603189134972987,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 26768
+    },
+    {
+      "epoch": 0.6371750928306198,
+      "grad_norm": 19.43746707133858,
+      "learning_rate": 2e-05,
+      "loss": 1.3528,
+      "step": 26769
+    },
+    {
+      "epoch": 0.6371988955536514,
+      "grad_norm": 4.371365619037503,
+      "learning_rate": 2e-05,
+      "loss": 0.1193,
+      "step": 26770
+    },
+    {
+      "epoch": 0.6372226982766829,
+      "grad_norm": 3.3558829704271895,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 26771
+    },
+    {
+      "epoch": 0.6372465009997144,
+      "grad_norm": 1.1978359519351933,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 26772
+    },
+    {
+      "epoch": 0.6372703037227458,
+      "grad_norm": 0.041827012842743844,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 26773
+    },
+    {
+      "epoch": 0.6372941064457774,
+      "grad_norm": 2.703779566990195,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 26774
+    },
+    {
+      "epoch": 0.6373179091688089,
+      "grad_norm": 2.9860705251449944,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 26775
+    },
+    {
+      "epoch": 0.6373417118918404,
+      "grad_norm": 4.39324954852135,
+      "learning_rate": 2e-05,
+      "loss": 0.8727,
+      "step": 26776
+    },
+    {
+      "epoch": 0.6373655146148719,
+      "grad_norm": 2.7121549119842245,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 26777
+    },
+    {
+      "epoch": 0.6373893173379035,
+      "grad_norm": 7.347819333083404,
+      "learning_rate": 2e-05,
+      "loss": 0.5588,
+      "step": 26778
+    },
+    {
+      "epoch": 0.637413120060935,
+      "grad_norm": 0.04535811653129841,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 26779
+    },
+    {
+      "epoch": 0.6374369227839665,
+      "grad_norm": 2.2720751991226575,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 26780
+    },
+    {
+      "epoch": 0.637460725506998,
+      "grad_norm": 22.818772679655986,
+      "learning_rate": 2e-05,
+      "loss": 0.3701,
+      "step": 26781
+    },
+    {
+      "epoch": 0.6374845282300295,
+      "grad_norm": 1.8865887255875988,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 26782
+    },
+    {
+      "epoch": 0.637508330953061,
+      "grad_norm": 5.717197245332998,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 26783
+    },
+    {
+      "epoch": 0.6375321336760925,
+      "grad_norm": 2.24461193900595,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 26784
+    },
+    {
+      "epoch": 0.6375559363991241,
+      "grad_norm": 0.4981347359049915,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 26785
+    },
+    {
+      "epoch": 0.6375797391221556,
+      "grad_norm": 0.8081395991341564,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 26786
+    },
+    {
+      "epoch": 0.6376035418451871,
+      "grad_norm": 4.1237686053093405,
+      "learning_rate": 2e-05,
+      "loss": 0.6155,
+      "step": 26787
+    },
+    {
+      "epoch": 0.6376273445682186,
+      "grad_norm": 7.517737652829996,
+      "learning_rate": 2e-05,
+      "loss": 0.8643,
+      "step": 26788
+    },
+    {
+      "epoch": 0.6376511472912502,
+      "grad_norm": 10.600205103812208,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 26789
+    },
+    {
+      "epoch": 0.6376749500142816,
+      "grad_norm": 18.808394677895574,
+      "learning_rate": 2e-05,
+      "loss": 1.4178,
+      "step": 26790
+    },
+    {
+      "epoch": 0.6376987527373131,
+      "grad_norm": 5.241036666472022,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 26791
+    },
+    {
+      "epoch": 0.6377225554603446,
+      "grad_norm": 0.026421081651302754,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 26792
+    },
+    {
+      "epoch": 0.6377463581833762,
+      "grad_norm": 3.7985471935836186,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 26793
+    },
+    {
+      "epoch": 0.6377701609064077,
+      "grad_norm": 3.870711058113271,
+      "learning_rate": 2e-05,
+      "loss": 0.7068,
+      "step": 26794
+    },
+    {
+      "epoch": 0.6377939636294392,
+      "grad_norm": 6.35680363110393,
+      "learning_rate": 2e-05,
+      "loss": 0.586,
+      "step": 26795
+    },
+    {
+      "epoch": 0.6378177663524707,
+      "grad_norm": 51.05556713021598,
+      "learning_rate": 2e-05,
+      "loss": 0.7363,
+      "step": 26796
+    },
+    {
+      "epoch": 0.6378415690755023,
+      "grad_norm": 0.30471896766787465,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 26797
+    },
+    {
+      "epoch": 0.6378653717985338,
+      "grad_norm": 9.809593377186028,
+      "learning_rate": 2e-05,
+      "loss": 0.6417,
+      "step": 26798
+    },
+    {
+      "epoch": 0.6378891745215652,
+      "grad_norm": 5.036371003260238,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 26799
+    },
+    {
+      "epoch": 0.6379129772445968,
+      "grad_norm": 3.599985226200064,
+      "learning_rate": 2e-05,
+      "loss": 0.8296,
+      "step": 26800
+    },
+    {
+      "epoch": 0.6379367799676283,
+      "grad_norm": 3.2520747009846,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 26801
+    },
+    {
+      "epoch": 0.6379605826906598,
+      "grad_norm": 8.198050795425837,
+      "learning_rate": 2e-05,
+      "loss": 1.1661,
+      "step": 26802
+    },
+    {
+      "epoch": 0.6379843854136913,
+      "grad_norm": 2.8575631094162173,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 26803
+    },
+    {
+      "epoch": 0.6380081881367229,
+      "grad_norm": 1.1811364555996708,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 26804
+    },
+    {
+      "epoch": 0.6380319908597544,
+      "grad_norm": 1.3421107516667525,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 26805
+    },
+    {
+      "epoch": 0.6380557935827859,
+      "grad_norm": 1.5935160576561374,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 26806
+    },
+    {
+      "epoch": 0.6380795963058173,
+      "grad_norm": 2.752492699002722,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 26807
+    },
+    {
+      "epoch": 0.6381033990288489,
+      "grad_norm": 5.614842596228289,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 26808
+    },
+    {
+      "epoch": 0.6381272017518804,
+      "grad_norm": 3.580472529706326,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 26809
+    },
+    {
+      "epoch": 0.6381510044749119,
+      "grad_norm": 3.7845821958281136,
+      "learning_rate": 2e-05,
+      "loss": 0.5208,
+      "step": 26810
+    },
+    {
+      "epoch": 0.6381748071979434,
+      "grad_norm": 3.639376469474545,
+      "learning_rate": 2e-05,
+      "loss": 0.2341,
+      "step": 26811
+    },
+    {
+      "epoch": 0.638198609920975,
+      "grad_norm": 15.470212228936205,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 26812
+    },
+    {
+      "epoch": 0.6382224126440065,
+      "grad_norm": 4.399085815530456,
+      "learning_rate": 2e-05,
+      "loss": 0.7243,
+      "step": 26813
+    },
+    {
+      "epoch": 0.638246215367038,
+      "grad_norm": 4.145450542395388,
+      "learning_rate": 2e-05,
+      "loss": 0.8512,
+      "step": 26814
+    },
+    {
+      "epoch": 0.6382700180900696,
+      "grad_norm": 3.281183301664131,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 26815
+    },
+    {
+      "epoch": 0.638293820813101,
+      "grad_norm": 28.450108927975634,
+      "learning_rate": 2e-05,
+      "loss": 0.5326,
+      "step": 26816
+    },
+    {
+      "epoch": 0.6383176235361325,
+      "grad_norm": 8.037346463705196,
+      "learning_rate": 2e-05,
+      "loss": 1.1284,
+      "step": 26817
+    },
+    {
+      "epoch": 0.638341426259164,
+      "grad_norm": 4.048713165770834,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 26818
+    },
+    {
+      "epoch": 0.6383652289821956,
+      "grad_norm": 7.321312011814521,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 26819
+    },
+    {
+      "epoch": 0.6383890317052271,
+      "grad_norm": 2.8791948932025226,
+      "learning_rate": 2e-05,
+      "loss": 0.4518,
+      "step": 26820
+    },
+    {
+      "epoch": 0.6384128344282586,
+      "grad_norm": 0.18785528326397144,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 26821
+    },
+    {
+      "epoch": 0.6384366371512901,
+      "grad_norm": 4.803307616414587,
+      "learning_rate": 2e-05,
+      "loss": 0.3764,
+      "step": 26822
+    },
+    {
+      "epoch": 0.6384604398743217,
+      "grad_norm": 6.602921553772116,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 26823
+    },
+    {
+      "epoch": 0.6384842425973531,
+      "grad_norm": 2.430977109791014,
+      "learning_rate": 2e-05,
+      "loss": 0.1845,
+      "step": 26824
+    },
+    {
+      "epoch": 0.6385080453203846,
+      "grad_norm": 3.1331181411616345,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 26825
+    },
+    {
+      "epoch": 0.6385318480434161,
+      "grad_norm": 4.0498764112202315,
+      "learning_rate": 2e-05,
+      "loss": 0.6304,
+      "step": 26826
+    },
+    {
+      "epoch": 0.6385556507664477,
+      "grad_norm": 3.3672477728548915,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 26827
+    },
+    {
+      "epoch": 0.6385794534894792,
+      "grad_norm": 2.6210977599328915,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 26828
+    },
+    {
+      "epoch": 0.6386032562125107,
+      "grad_norm": 2.784696797037397,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 26829
+    },
+    {
+      "epoch": 0.6386270589355423,
+      "grad_norm": 5.705484257886584,
+      "learning_rate": 2e-05,
+      "loss": 0.6087,
+      "step": 26830
+    },
+    {
+      "epoch": 0.6386508616585738,
+      "grad_norm": 0.8880796329271091,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 26831
+    },
+    {
+      "epoch": 0.6386746643816053,
+      "grad_norm": 8.221814646774117,
+      "learning_rate": 2e-05,
+      "loss": 1.0139,
+      "step": 26832
+    },
+    {
+      "epoch": 0.6386984671046367,
+      "grad_norm": 3.7854075141574803,
+      "learning_rate": 2e-05,
+      "loss": 0.8064,
+      "step": 26833
+    },
+    {
+      "epoch": 0.6387222698276683,
+      "grad_norm": 2.491123830765135,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 26834
+    },
+    {
+      "epoch": 0.6387460725506998,
+      "grad_norm": 3.2115692634952318,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 26835
+    },
+    {
+      "epoch": 0.6387698752737313,
+      "grad_norm": 7.865155935363141,
+      "learning_rate": 2e-05,
+      "loss": 0.2929,
+      "step": 26836
+    },
+    {
+      "epoch": 0.6387936779967628,
+      "grad_norm": 3.9121360588360297,
+      "learning_rate": 2e-05,
+      "loss": 0.6299,
+      "step": 26837
+    },
+    {
+      "epoch": 0.6388174807197944,
+      "grad_norm": 3.9426879619867883,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 26838
+    },
+    {
+      "epoch": 0.6388412834428259,
+      "grad_norm": 0.695826571670096,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 26839
+    },
+    {
+      "epoch": 0.6388650861658574,
+      "grad_norm": 2.4770953879112225,
+      "learning_rate": 2e-05,
+      "loss": 0.0807,
+      "step": 26840
+    },
+    {
+      "epoch": 0.6388888888888888,
+      "grad_norm": 1.6261968170570342,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 26841
+    },
+    {
+      "epoch": 0.6389126916119204,
+      "grad_norm": 3.252755962738955,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 26842
+    },
+    {
+      "epoch": 0.6389364943349519,
+      "grad_norm": 1.057959892386181,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 26843
+    },
+    {
+      "epoch": 0.6389602970579834,
+      "grad_norm": 4.441784417978573,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 26844
+    },
+    {
+      "epoch": 0.638984099781015,
+      "grad_norm": 4.133524084634861,
+      "learning_rate": 2e-05,
+      "loss": 0.4562,
+      "step": 26845
+    },
+    {
+      "epoch": 0.6390079025040465,
+      "grad_norm": 2.742775170780677,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 26846
+    },
+    {
+      "epoch": 0.639031705227078,
+      "grad_norm": 0.38105498056396986,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 26847
+    },
+    {
+      "epoch": 0.6390555079501095,
+      "grad_norm": 7.152616880828371,
+      "learning_rate": 2e-05,
+      "loss": 0.9765,
+      "step": 26848
+    },
+    {
+      "epoch": 0.639079310673141,
+      "grad_norm": 1.9171866667970952,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 26849
+    },
+    {
+      "epoch": 0.6391031133961725,
+      "grad_norm": 8.195904640040526,
+      "learning_rate": 2e-05,
+      "loss": 1.334,
+      "step": 26850
+    },
+    {
+      "epoch": 0.639126916119204,
+      "grad_norm": 5.578501317764381,
+      "learning_rate": 2e-05,
+      "loss": 0.7833,
+      "step": 26851
+    },
+    {
+      "epoch": 0.6391507188422355,
+      "grad_norm": 1.2259740362658365,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 26852
+    },
+    {
+      "epoch": 0.6391745215652671,
+      "grad_norm": 1.1170761491034849,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 26853
+    },
+    {
+      "epoch": 0.6391983242882986,
+      "grad_norm": 0.2513001813338046,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 26854
+    },
+    {
+      "epoch": 0.6392221270113301,
+      "grad_norm": 0.5600207640833826,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 26855
+    },
+    {
+      "epoch": 0.6392459297343616,
+      "grad_norm": 5.981905939143008,
+      "learning_rate": 2e-05,
+      "loss": 0.7281,
+      "step": 26856
+    },
+    {
+      "epoch": 0.6392697324573932,
+      "grad_norm": 4.446011483701543,
+      "learning_rate": 2e-05,
+      "loss": 0.5615,
+      "step": 26857
+    },
+    {
+      "epoch": 0.6392935351804246,
+      "grad_norm": 0.22608571413341205,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 26858
+    },
+    {
+      "epoch": 0.6393173379034561,
+      "grad_norm": 3.3341210521107003,
+      "learning_rate": 2e-05,
+      "loss": 0.1577,
+      "step": 26859
+    },
+    {
+      "epoch": 0.6393411406264877,
+      "grad_norm": 3.6095995239742775,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 26860
+    },
+    {
+      "epoch": 0.6393649433495192,
+      "grad_norm": 1.2176989421569204,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 26861
+    },
+    {
+      "epoch": 0.6393887460725507,
+      "grad_norm": 5.21611220484904,
+      "learning_rate": 2e-05,
+      "loss": 0.5184,
+      "step": 26862
+    },
+    {
+      "epoch": 0.6394125487955822,
+      "grad_norm": 1.0281577172588807,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 26863
+    },
+    {
+      "epoch": 0.6394363515186138,
+      "grad_norm": 3.3303881985895085,
+      "learning_rate": 2e-05,
+      "loss": 0.1228,
+      "step": 26864
+    },
+    {
+      "epoch": 0.6394601542416453,
+      "grad_norm": 4.343455183337602,
+      "learning_rate": 2e-05,
+      "loss": 0.6472,
+      "step": 26865
+    },
+    {
+      "epoch": 0.6394839569646767,
+      "grad_norm": 1.291113039689585,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 26866
+    },
+    {
+      "epoch": 0.6395077596877082,
+      "grad_norm": 5.881071748171174,
+      "learning_rate": 2e-05,
+      "loss": 0.1937,
+      "step": 26867
+    },
+    {
+      "epoch": 0.6395315624107398,
+      "grad_norm": 0.10974334504694586,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 26868
+    },
+    {
+      "epoch": 0.6395553651337713,
+      "grad_norm": 3.3827649262107804,
+      "learning_rate": 2e-05,
+      "loss": 0.7397,
+      "step": 26869
+    },
+    {
+      "epoch": 0.6395791678568028,
+      "grad_norm": 19.967315047041662,
+      "learning_rate": 2e-05,
+      "loss": 0.8141,
+      "step": 26870
+    },
+    {
+      "epoch": 0.6396029705798343,
+      "grad_norm": 16.295511042905847,
+      "learning_rate": 2e-05,
+      "loss": 0.7675,
+      "step": 26871
+    },
+    {
+      "epoch": 0.6396267733028659,
+      "grad_norm": 6.5109057109747255,
+      "learning_rate": 2e-05,
+      "loss": 1.5572,
+      "step": 26872
+    },
+    {
+      "epoch": 0.6396505760258974,
+      "grad_norm": 5.076869829297572,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 26873
+    },
+    {
+      "epoch": 0.6396743787489289,
+      "grad_norm": 4.950534015409855,
+      "learning_rate": 2e-05,
+      "loss": 0.1572,
+      "step": 26874
+    },
+    {
+      "epoch": 0.6396981814719604,
+      "grad_norm": 0.28484790328941884,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 26875
+    },
+    {
+      "epoch": 0.6397219841949919,
+      "grad_norm": 11.83072581946074,
+      "learning_rate": 2e-05,
+      "loss": 0.4896,
+      "step": 26876
+    },
+    {
+      "epoch": 0.6397457869180234,
+      "grad_norm": 9.148952707626265,
+      "learning_rate": 2e-05,
+      "loss": 0.6527,
+      "step": 26877
+    },
+    {
+      "epoch": 0.6397695896410549,
+      "grad_norm": 6.645539846382954,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 26878
+    },
+    {
+      "epoch": 0.6397933923640865,
+      "grad_norm": 2.448661669706491,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 26879
+    },
+    {
+      "epoch": 0.639817195087118,
+      "grad_norm": 12.080747957833628,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 26880
+    },
+    {
+      "epoch": 0.6398409978101495,
+      "grad_norm": 0.2605111042312047,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 26881
+    },
+    {
+      "epoch": 0.639864800533181,
+      "grad_norm": 0.9277386369770052,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 26882
+    },
+    {
+      "epoch": 0.6398886032562126,
+      "grad_norm": 3.1818577448376772,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 26883
+    },
+    {
+      "epoch": 0.639912405979244,
+      "grad_norm": 0.2732988124925353,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 26884
+    },
+    {
+      "epoch": 0.6399362087022755,
+      "grad_norm": 0.8111118439124101,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 26885
+    },
+    {
+      "epoch": 0.639960011425307,
+      "grad_norm": 2.3854999743098393,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 26886
+    },
+    {
+      "epoch": 0.6399838141483386,
+      "grad_norm": 7.585335912611255,
+      "learning_rate": 2e-05,
+      "loss": 1.4964,
+      "step": 26887
+    },
+    {
+      "epoch": 0.6400076168713701,
+      "grad_norm": 0.4765179188189209,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 26888
+    },
+    {
+      "epoch": 0.6400314195944016,
+      "grad_norm": 0.1467201349616304,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 26889
+    },
+    {
+      "epoch": 0.6400552223174331,
+      "grad_norm": 5.104460557450477,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 26890
+    },
+    {
+      "epoch": 0.6400790250404647,
+      "grad_norm": 0.20573651992345102,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 26891
+    },
+    {
+      "epoch": 0.6401028277634961,
+      "grad_norm": 0.023602189297964767,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 26892
+    },
+    {
+      "epoch": 0.6401266304865276,
+      "grad_norm": 0.20355305102483373,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 26893
+    },
+    {
+      "epoch": 0.6401504332095592,
+      "grad_norm": 4.21061950552565,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 26894
+    },
+    {
+      "epoch": 0.6401742359325907,
+      "grad_norm": 3.3725857608842893,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 26895
+    },
+    {
+      "epoch": 0.6401980386556222,
+      "grad_norm": 8.40318673816985,
+      "learning_rate": 2e-05,
+      "loss": 1.9656,
+      "step": 26896
+    },
+    {
+      "epoch": 0.6402218413786537,
+      "grad_norm": 7.600329140646733,
+      "learning_rate": 2e-05,
+      "loss": 0.9837,
+      "step": 26897
+    },
+    {
+      "epoch": 0.6402456441016853,
+      "grad_norm": 1.0098667906837069,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 26898
+    },
+    {
+      "epoch": 0.6402694468247168,
+      "grad_norm": 0.1535818641752382,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 26899
+    },
+    {
+      "epoch": 0.6402932495477482,
+      "grad_norm": 0.6792430073921196,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 26900
+    },
+    {
+      "epoch": 0.6403170522707797,
+      "grad_norm": 2.2001529401113373,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 26901
+    },
+    {
+      "epoch": 0.6403408549938113,
+      "grad_norm": 15.394518045426528,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 26902
+    },
+    {
+      "epoch": 0.6403646577168428,
+      "grad_norm": 3.7910638035592816,
+      "learning_rate": 2e-05,
+      "loss": 0.4676,
+      "step": 26903
+    },
+    {
+      "epoch": 0.6403884604398743,
+      "grad_norm": 0.7332475250278438,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 26904
+    },
+    {
+      "epoch": 0.6404122631629058,
+      "grad_norm": 3.8518085304135217,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 26905
+    },
+    {
+      "epoch": 0.6404360658859374,
+      "grad_norm": 2.2782989802417064,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 26906
+    },
+    {
+      "epoch": 0.6404598686089689,
+      "grad_norm": 0.6921720879068414,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 26907
+    },
+    {
+      "epoch": 0.6404836713320003,
+      "grad_norm": 0.1541668841415164,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 26908
+    },
+    {
+      "epoch": 0.640507474055032,
+      "grad_norm": 1.6521681729877022,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 26909
+    },
+    {
+      "epoch": 0.6405312767780634,
+      "grad_norm": 4.4315997514172745,
+      "learning_rate": 2e-05,
+      "loss": 0.486,
+      "step": 26910
+    },
+    {
+      "epoch": 0.6405550795010949,
+      "grad_norm": 15.435699833043518,
+      "learning_rate": 2e-05,
+      "loss": 1.3357,
+      "step": 26911
+    },
+    {
+      "epoch": 0.6405788822241264,
+      "grad_norm": 10.970850703691614,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 26912
+    },
+    {
+      "epoch": 0.640602684947158,
+      "grad_norm": 2.6358676059772828,
+      "learning_rate": 2e-05,
+      "loss": 0.1291,
+      "step": 26913
+    },
+    {
+      "epoch": 0.6406264876701895,
+      "grad_norm": 8.255455213756596,
+      "learning_rate": 2e-05,
+      "loss": 2.2266,
+      "step": 26914
+    },
+    {
+      "epoch": 0.640650290393221,
+      "grad_norm": 17.97747424300683,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 26915
+    },
+    {
+      "epoch": 0.6406740931162525,
+      "grad_norm": 3.681619403132718,
+      "learning_rate": 2e-05,
+      "loss": 0.6279,
+      "step": 26916
+    },
+    {
+      "epoch": 0.640697895839284,
+      "grad_norm": 3.7697280856920834,
+      "learning_rate": 2e-05,
+      "loss": 0.4521,
+      "step": 26917
+    },
+    {
+      "epoch": 0.6407216985623155,
+      "grad_norm": 0.13642352894604481,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 26918
+    },
+    {
+      "epoch": 0.640745501285347,
+      "grad_norm": 0.040267776350859194,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 26919
+    },
+    {
+      "epoch": 0.6407693040083785,
+      "grad_norm": 0.4133456260359237,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 26920
+    },
+    {
+      "epoch": 0.6407931067314101,
+      "grad_norm": 5.324501960238153,
+      "learning_rate": 2e-05,
+      "loss": 0.6101,
+      "step": 26921
+    },
+    {
+      "epoch": 0.6408169094544416,
+      "grad_norm": 21.60044255937004,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 26922
+    },
+    {
+      "epoch": 0.6408407121774731,
+      "grad_norm": 4.122258227969393,
+      "learning_rate": 2e-05,
+      "loss": 0.4397,
+      "step": 26923
+    },
+    {
+      "epoch": 0.6408645149005047,
+      "grad_norm": 2.084334669275544,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 26924
+    },
+    {
+      "epoch": 0.6408883176235362,
+      "grad_norm": 3.9948254023626175,
+      "learning_rate": 2e-05,
+      "loss": 0.1527,
+      "step": 26925
+    },
+    {
+      "epoch": 0.6409121203465676,
+      "grad_norm": 6.277341354154229,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 26926
+    },
+    {
+      "epoch": 0.6409359230695991,
+      "grad_norm": 11.225205208734884,
+      "learning_rate": 2e-05,
+      "loss": 0.6977,
+      "step": 26927
+    },
+    {
+      "epoch": 0.6409597257926307,
+      "grad_norm": 3.8691625963636196,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 26928
+    },
+    {
+      "epoch": 0.6409835285156622,
+      "grad_norm": 0.26994499443131986,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 26929
+    },
+    {
+      "epoch": 0.6410073312386937,
+      "grad_norm": 2.2362057258919763,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 26930
+    },
+    {
+      "epoch": 0.6410311339617252,
+      "grad_norm": 5.101191411540716,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 26931
+    },
+    {
+      "epoch": 0.6410549366847568,
+      "grad_norm": 15.240134185142693,
+      "learning_rate": 2e-05,
+      "loss": 0.7235,
+      "step": 26932
+    },
+    {
+      "epoch": 0.6410787394077883,
+      "grad_norm": 0.08110154096703967,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 26933
+    },
+    {
+      "epoch": 0.6411025421308197,
+      "grad_norm": 1.4140819364560706,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 26934
+    },
+    {
+      "epoch": 0.6411263448538512,
+      "grad_norm": 7.029610995507823,
+      "learning_rate": 2e-05,
+      "loss": 0.4528,
+      "step": 26935
+    },
+    {
+      "epoch": 0.6411501475768828,
+      "grad_norm": 5.869314482284984,
+      "learning_rate": 2e-05,
+      "loss": 1.3383,
+      "step": 26936
+    },
+    {
+      "epoch": 0.6411739502999143,
+      "grad_norm": 6.065195799592587,
+      "learning_rate": 2e-05,
+      "loss": 0.7965,
+      "step": 26937
+    },
+    {
+      "epoch": 0.6411977530229458,
+      "grad_norm": 0.4714857542287196,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 26938
+    },
+    {
+      "epoch": 0.6412215557459774,
+      "grad_norm": 4.314918585291352,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 26939
+    },
+    {
+      "epoch": 0.6412453584690089,
+      "grad_norm": 2.9158642828233448,
+      "learning_rate": 2e-05,
+      "loss": 0.4435,
+      "step": 26940
+    },
+    {
+      "epoch": 0.6412691611920404,
+      "grad_norm": 3.552278127699082,
+      "learning_rate": 2e-05,
+      "loss": 0.6076,
+      "step": 26941
+    },
+    {
+      "epoch": 0.6412929639150718,
+      "grad_norm": 3.11852900396929,
+      "learning_rate": 2e-05,
+      "loss": 0.0544,
+      "step": 26942
+    },
+    {
+      "epoch": 0.6413167666381034,
+      "grad_norm": 10.23225696637085,
+      "learning_rate": 2e-05,
+      "loss": 1.1121,
+      "step": 26943
+    },
+    {
+      "epoch": 0.6413405693611349,
+      "grad_norm": 3.8137977900323765,
+      "learning_rate": 2e-05,
+      "loss": 0.9168,
+      "step": 26944
+    },
+    {
+      "epoch": 0.6413643720841664,
+      "grad_norm": 0.5513811635666762,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 26945
+    },
+    {
+      "epoch": 0.6413881748071979,
+      "grad_norm": 5.987434910446309,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 26946
+    },
+    {
+      "epoch": 0.6414119775302295,
+      "grad_norm": 4.179226461072424,
+      "learning_rate": 2e-05,
+      "loss": 0.466,
+      "step": 26947
+    },
+    {
+      "epoch": 0.641435780253261,
+      "grad_norm": 4.63267974423662,
+      "learning_rate": 2e-05,
+      "loss": 0.1979,
+      "step": 26948
+    },
+    {
+      "epoch": 0.6414595829762925,
+      "grad_norm": 5.115984561586397,
+      "learning_rate": 2e-05,
+      "loss": 0.9615,
+      "step": 26949
+    },
+    {
+      "epoch": 0.641483385699324,
+      "grad_norm": 0.16229411646455785,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 26950
+    },
+    {
+      "epoch": 0.6415071884223555,
+      "grad_norm": 6.530416756358963,
+      "learning_rate": 2e-05,
+      "loss": 1.4043,
+      "step": 26951
+    },
+    {
+      "epoch": 0.641530991145387,
+      "grad_norm": 11.691614783987083,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 26952
+    },
+    {
+      "epoch": 0.6415547938684185,
+      "grad_norm": 5.347029711925027,
+      "learning_rate": 2e-05,
+      "loss": 0.2174,
+      "step": 26953
+    },
+    {
+      "epoch": 0.6415785965914501,
+      "grad_norm": 5.99478528183841,
+      "learning_rate": 2e-05,
+      "loss": 0.7405,
+      "step": 26954
+    },
+    {
+      "epoch": 0.6416023993144816,
+      "grad_norm": 4.992370322440201,
+      "learning_rate": 2e-05,
+      "loss": 0.482,
+      "step": 26955
+    },
+    {
+      "epoch": 0.6416262020375131,
+      "grad_norm": 2.769409455963765,
+      "learning_rate": 2e-05,
+      "loss": 0.635,
+      "step": 26956
+    },
+    {
+      "epoch": 0.6416500047605446,
+      "grad_norm": 4.162375116103417,
+      "learning_rate": 2e-05,
+      "loss": 1.1416,
+      "step": 26957
+    },
+    {
+      "epoch": 0.6416738074835762,
+      "grad_norm": 0.47905279396084416,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 26958
+    },
+    {
+      "epoch": 0.6416976102066076,
+      "grad_norm": 1.359444699788887,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 26959
+    },
+    {
+      "epoch": 0.6417214129296391,
+      "grad_norm": 1.8498016138187343,
+      "learning_rate": 2e-05,
+      "loss": 0.2441,
+      "step": 26960
+    },
+    {
+      "epoch": 0.6417452156526706,
+      "grad_norm": 5.325009598214866,
+      "learning_rate": 2e-05,
+      "loss": 0.663,
+      "step": 26961
+    },
+    {
+      "epoch": 0.6417690183757022,
+      "grad_norm": 3.38342586322599,
+      "learning_rate": 2e-05,
+      "loss": 0.7038,
+      "step": 26962
+    },
+    {
+      "epoch": 0.6417928210987337,
+      "grad_norm": 0.1303903743749207,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 26963
+    },
+    {
+      "epoch": 0.6418166238217652,
+      "grad_norm": 0.5018494524139806,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 26964
+    },
+    {
+      "epoch": 0.6418404265447967,
+      "grad_norm": 2.599193809409846,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 26965
+    },
+    {
+      "epoch": 0.6418642292678283,
+      "grad_norm": 0.39400064792720085,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 26966
+    },
+    {
+      "epoch": 0.6418880319908598,
+      "grad_norm": 0.3271173734408117,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 26967
+    },
+    {
+      "epoch": 0.6419118347138912,
+      "grad_norm": 3.6995515720946903,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 26968
+    },
+    {
+      "epoch": 0.6419356374369228,
+      "grad_norm": 0.32534060966443934,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 26969
+    },
+    {
+      "epoch": 0.6419594401599543,
+      "grad_norm": 2.2136177023727277,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 26970
+    },
+    {
+      "epoch": 0.6419832428829858,
+      "grad_norm": 0.6765985501824333,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 26971
+    },
+    {
+      "epoch": 0.6420070456060173,
+      "grad_norm": 0.16030724029820215,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 26972
+    },
+    {
+      "epoch": 0.6420308483290489,
+      "grad_norm": 9.711831687380954,
+      "learning_rate": 2e-05,
+      "loss": 0.7273,
+      "step": 26973
+    },
+    {
+      "epoch": 0.6420546510520804,
+      "grad_norm": 3.9352700305922923,
+      "learning_rate": 2e-05,
+      "loss": 0.34,
+      "step": 26974
+    },
+    {
+      "epoch": 0.6420784537751119,
+      "grad_norm": 4.555827329025385,
+      "learning_rate": 2e-05,
+      "loss": 1.0314,
+      "step": 26975
+    },
+    {
+      "epoch": 0.6421022564981433,
+      "grad_norm": 0.042638225088468734,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 26976
+    },
+    {
+      "epoch": 0.6421260592211749,
+      "grad_norm": 0.5699080712861376,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 26977
+    },
+    {
+      "epoch": 0.6421498619442064,
+      "grad_norm": 8.161056764967924,
+      "learning_rate": 2e-05,
+      "loss": 0.749,
+      "step": 26978
+    },
+    {
+      "epoch": 0.6421736646672379,
+      "grad_norm": 0.452253988287011,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 26979
+    },
+    {
+      "epoch": 0.6421974673902694,
+      "grad_norm": 6.600244596444895,
+      "learning_rate": 2e-05,
+      "loss": 1.27,
+      "step": 26980
+    },
+    {
+      "epoch": 0.642221270113301,
+      "grad_norm": 1.8593140494351843,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 26981
+    },
+    {
+      "epoch": 0.6422450728363325,
+      "grad_norm": 0.20487299935933334,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 26982
+    },
+    {
+      "epoch": 0.642268875559364,
+      "grad_norm": 14.234605985383476,
+      "learning_rate": 2e-05,
+      "loss": 0.6902,
+      "step": 26983
+    },
+    {
+      "epoch": 0.6422926782823956,
+      "grad_norm": 3.315819723952038,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 26984
+    },
+    {
+      "epoch": 0.642316481005427,
+      "grad_norm": 3.9613294277169735,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 26985
+    },
+    {
+      "epoch": 0.6423402837284585,
+      "grad_norm": 3.1204189483795615,
+      "learning_rate": 2e-05,
+      "loss": 0.4355,
+      "step": 26986
+    },
+    {
+      "epoch": 0.64236408645149,
+      "grad_norm": 0.029922337645136438,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 26987
+    },
+    {
+      "epoch": 0.6423878891745216,
+      "grad_norm": 0.044463640108702046,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 26988
+    },
+    {
+      "epoch": 0.6424116918975531,
+      "grad_norm": 8.245416639104548,
+      "learning_rate": 2e-05,
+      "loss": 0.7822,
+      "step": 26989
+    },
+    {
+      "epoch": 0.6424354946205846,
+      "grad_norm": 3.7444149918969716,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 26990
+    },
+    {
+      "epoch": 0.6424592973436161,
+      "grad_norm": 0.10469764710634956,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 26991
+    },
+    {
+      "epoch": 0.6424831000666477,
+      "grad_norm": 4.67422479792972,
+      "learning_rate": 2e-05,
+      "loss": 0.7101,
+      "step": 26992
+    },
+    {
+      "epoch": 0.6425069027896791,
+      "grad_norm": 3.020478305876873,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 26993
+    },
+    {
+      "epoch": 0.6425307055127106,
+      "grad_norm": 5.286757541319552,
+      "learning_rate": 2e-05,
+      "loss": 0.52,
+      "step": 26994
+    },
+    {
+      "epoch": 0.6425545082357421,
+      "grad_norm": 0.034681934629244716,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 26995
+    },
+    {
+      "epoch": 0.6425783109587737,
+      "grad_norm": 5.065891706205147,
+      "learning_rate": 2e-05,
+      "loss": 0.4188,
+      "step": 26996
+    },
+    {
+      "epoch": 0.6426021136818052,
+      "grad_norm": 0.016444413174920222,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 26997
+    },
+    {
+      "epoch": 0.6426259164048367,
+      "grad_norm": 11.334295943415114,
+      "learning_rate": 2e-05,
+      "loss": 0.9311,
+      "step": 26998
+    },
+    {
+      "epoch": 0.6426497191278683,
+      "grad_norm": 6.05299160860839,
+      "learning_rate": 2e-05,
+      "loss": 0.607,
+      "step": 26999
+    },
+    {
+      "epoch": 0.6426735218508998,
+      "grad_norm": 6.263390368134137,
+      "learning_rate": 2e-05,
+      "loss": 1.5681,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6426973245739313,
+      "grad_norm": 2.1320913659933467,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 27001
+    },
+    {
+      "epoch": 0.6427211272969627,
+      "grad_norm": 3.5172125596744737,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 27002
+    },
+    {
+      "epoch": 0.6427449300199943,
+      "grad_norm": 4.862497086129022,
+      "learning_rate": 2e-05,
+      "loss": 0.6396,
+      "step": 27003
+    },
+    {
+      "epoch": 0.6427687327430258,
+      "grad_norm": 2.520308859343461,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 27004
+    },
+    {
+      "epoch": 0.6427925354660573,
+      "grad_norm": 3.052361657941514,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 27005
+    },
+    {
+      "epoch": 0.6428163381890888,
+      "grad_norm": 2.0538622057672877,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 27006
+    },
+    {
+      "epoch": 0.6428401409121204,
+      "grad_norm": 9.020836058358212,
+      "learning_rate": 2e-05,
+      "loss": 0.2038,
+      "step": 27007
+    },
+    {
+      "epoch": 0.6428639436351519,
+      "grad_norm": 6.282774038702346,
+      "learning_rate": 2e-05,
+      "loss": 0.1939,
+      "step": 27008
+    },
+    {
+      "epoch": 0.6428877463581834,
+      "grad_norm": 5.416829620166428,
+      "learning_rate": 2e-05,
+      "loss": 0.7249,
+      "step": 27009
+    },
+    {
+      "epoch": 0.6429115490812148,
+      "grad_norm": 0.03705345729489777,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 27010
+    },
+    {
+      "epoch": 0.6429353518042464,
+      "grad_norm": 0.011838909026203843,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 27011
+    },
+    {
+      "epoch": 0.6429591545272779,
+      "grad_norm": 7.091603361234874,
+      "learning_rate": 2e-05,
+      "loss": 0.2694,
+      "step": 27012
+    },
+    {
+      "epoch": 0.6429829572503094,
+      "grad_norm": 3.0935476729222926,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 27013
+    },
+    {
+      "epoch": 0.643006759973341,
+      "grad_norm": 0.7981861263650261,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 27014
+    },
+    {
+      "epoch": 0.6430305626963725,
+      "grad_norm": 3.2242869094237854,
+      "learning_rate": 2e-05,
+      "loss": 0.6634,
+      "step": 27015
+    },
+    {
+      "epoch": 0.643054365419404,
+      "grad_norm": 1.5148465333617351,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 27016
+    },
+    {
+      "epoch": 0.6430781681424355,
+      "grad_norm": 3.4642081038180392,
+      "learning_rate": 2e-05,
+      "loss": 0.3141,
+      "step": 27017
+    },
+    {
+      "epoch": 0.643101970865467,
+      "grad_norm": 3.9104208014883097,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 27018
+    },
+    {
+      "epoch": 0.6431257735884985,
+      "grad_norm": 8.449337586590854,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 27019
+    },
+    {
+      "epoch": 0.64314957631153,
+      "grad_norm": 2.475651567752823,
+      "learning_rate": 2e-05,
+      "loss": 0.4466,
+      "step": 27020
+    },
+    {
+      "epoch": 0.6431733790345615,
+      "grad_norm": 2.645013420416893,
+      "learning_rate": 2e-05,
+      "loss": 0.4177,
+      "step": 27021
+    },
+    {
+      "epoch": 0.6431971817575931,
+      "grad_norm": 0.5118789333304915,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 27022
+    },
+    {
+      "epoch": 0.6432209844806246,
+      "grad_norm": 1.9748251966022798,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 27023
+    },
+    {
+      "epoch": 0.6432447872036561,
+      "grad_norm": 0.5474805293969833,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 27024
+    },
+    {
+      "epoch": 0.6432685899266876,
+      "grad_norm": 6.249979842260215,
+      "learning_rate": 2e-05,
+      "loss": 0.8866,
+      "step": 27025
+    },
+    {
+      "epoch": 0.6432923926497192,
+      "grad_norm": 3.6042139498215784,
+      "learning_rate": 2e-05,
+      "loss": 0.4435,
+      "step": 27026
+    },
+    {
+      "epoch": 0.6433161953727506,
+      "grad_norm": 11.232250397319072,
+      "learning_rate": 2e-05,
+      "loss": 1.6846,
+      "step": 27027
+    },
+    {
+      "epoch": 0.6433399980957821,
+      "grad_norm": 0.14073409250937158,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 27028
+    },
+    {
+      "epoch": 0.6433638008188137,
+      "grad_norm": 0.7463122990510646,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 27029
+    },
+    {
+      "epoch": 0.6433876035418452,
+      "grad_norm": 0.09885522844241372,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 27030
+    },
+    {
+      "epoch": 0.6434114062648767,
+      "grad_norm": 2.3591342258420003,
+      "learning_rate": 2e-05,
+      "loss": 0.304,
+      "step": 27031
+    },
+    {
+      "epoch": 0.6434352089879082,
+      "grad_norm": 3.6179236811211255,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 27032
+    },
+    {
+      "epoch": 0.6434590117109398,
+      "grad_norm": 3.9391700105250886,
+      "learning_rate": 2e-05,
+      "loss": 0.7847,
+      "step": 27033
+    },
+    {
+      "epoch": 0.6434828144339713,
+      "grad_norm": 4.647676026598946,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 27034
+    },
+    {
+      "epoch": 0.6435066171570027,
+      "grad_norm": 7.93526071183375,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 27035
+    },
+    {
+      "epoch": 0.6435304198800342,
+      "grad_norm": 2.7567661908044965,
+      "learning_rate": 2e-05,
+      "loss": 0.5174,
+      "step": 27036
+    },
+    {
+      "epoch": 0.6435542226030658,
+      "grad_norm": 12.131890713244173,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 27037
+    },
+    {
+      "epoch": 0.6435780253260973,
+      "grad_norm": 0.3113354179388048,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 27038
+    },
+    {
+      "epoch": 0.6436018280491288,
+      "grad_norm": 4.541635722762354,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 27039
+    },
+    {
+      "epoch": 0.6436256307721603,
+      "grad_norm": 2.547751943288624,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 27040
+    },
+    {
+      "epoch": 0.6436494334951919,
+      "grad_norm": 0.16110805049500376,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 27041
+    },
+    {
+      "epoch": 0.6436732362182234,
+      "grad_norm": 7.374189881499136,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 27042
+    },
+    {
+      "epoch": 0.6436970389412549,
+      "grad_norm": 0.5639678533927163,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 27043
+    },
+    {
+      "epoch": 0.6437208416642864,
+      "grad_norm": 2.9338959551126513,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 27044
+    },
+    {
+      "epoch": 0.6437446443873179,
+      "grad_norm": 1.7299020481681875,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 27045
+    },
+    {
+      "epoch": 0.6437684471103494,
+      "grad_norm": 8.37686825916186,
+      "learning_rate": 2e-05,
+      "loss": 1.979,
+      "step": 27046
+    },
+    {
+      "epoch": 0.6437922498333809,
+      "grad_norm": 1.2541631460881428,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 27047
+    },
+    {
+      "epoch": 0.6438160525564125,
+      "grad_norm": 6.323208824403519,
+      "learning_rate": 2e-05,
+      "loss": 1.4009,
+      "step": 27048
+    },
+    {
+      "epoch": 0.643839855279444,
+      "grad_norm": 0.441617916941346,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 27049
+    },
+    {
+      "epoch": 0.6438636580024755,
+      "grad_norm": 7.742869616302298,
+      "learning_rate": 2e-05,
+      "loss": 0.283,
+      "step": 27050
+    },
+    {
+      "epoch": 0.643887460725507,
+      "grad_norm": 4.675115924973789,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 27051
+    },
+    {
+      "epoch": 0.6439112634485386,
+      "grad_norm": 2.450825494718547,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 27052
+    },
+    {
+      "epoch": 0.64393506617157,
+      "grad_norm": 3.8590175001610545,
+      "learning_rate": 2e-05,
+      "loss": 0.765,
+      "step": 27053
+    },
+    {
+      "epoch": 0.6439588688946015,
+      "grad_norm": 3.0276104157355626,
+      "learning_rate": 2e-05,
+      "loss": 0.2586,
+      "step": 27054
+    },
+    {
+      "epoch": 0.643982671617633,
+      "grad_norm": 0.11469169110387127,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 27055
+    },
+    {
+      "epoch": 0.6440064743406646,
+      "grad_norm": 3.74574400605242,
+      "learning_rate": 2e-05,
+      "loss": 0.4761,
+      "step": 27056
+    },
+    {
+      "epoch": 0.6440302770636961,
+      "grad_norm": 1.0311120620470007,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 27057
+    },
+    {
+      "epoch": 0.6440540797867276,
+      "grad_norm": 5.067231269868989,
+      "learning_rate": 2e-05,
+      "loss": 1.6345,
+      "step": 27058
+    },
+    {
+      "epoch": 0.6440778825097592,
+      "grad_norm": 6.677067089290437,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 27059
+    },
+    {
+      "epoch": 0.6441016852327907,
+      "grad_norm": 4.73615587387403,
+      "learning_rate": 2e-05,
+      "loss": 0.3287,
+      "step": 27060
+    },
+    {
+      "epoch": 0.6441254879558221,
+      "grad_norm": 1.9322483936463202,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 27061
+    },
+    {
+      "epoch": 0.6441492906788536,
+      "grad_norm": 8.033578426469274,
+      "learning_rate": 2e-05,
+      "loss": 0.8762,
+      "step": 27062
+    },
+    {
+      "epoch": 0.6441730934018852,
+      "grad_norm": 3.8269447807439745,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 27063
+    },
+    {
+      "epoch": 0.6441968961249167,
+      "grad_norm": 2.798381466041365,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 27064
+    },
+    {
+      "epoch": 0.6442206988479482,
+      "grad_norm": 7.014003460083686,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 27065
+    },
+    {
+      "epoch": 0.6442445015709797,
+      "grad_norm": 4.1694478422167025,
+      "learning_rate": 2e-05,
+      "loss": 0.6084,
+      "step": 27066
+    },
+    {
+      "epoch": 0.6442683042940113,
+      "grad_norm": 6.63142372339856,
+      "learning_rate": 2e-05,
+      "loss": 1.0531,
+      "step": 27067
+    },
+    {
+      "epoch": 0.6442921070170428,
+      "grad_norm": 9.564818962972064,
+      "learning_rate": 2e-05,
+      "loss": 0.5426,
+      "step": 27068
+    },
+    {
+      "epoch": 0.6443159097400742,
+      "grad_norm": 2.5022521661096184,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 27069
+    },
+    {
+      "epoch": 0.6443397124631057,
+      "grad_norm": 9.853374134065172,
+      "learning_rate": 2e-05,
+      "loss": 0.1486,
+      "step": 27070
+    },
+    {
+      "epoch": 0.6443635151861373,
+      "grad_norm": 1.462183090143921,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 27071
+    },
+    {
+      "epoch": 0.6443873179091688,
+      "grad_norm": 5.088063870505325,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 27072
+    },
+    {
+      "epoch": 0.6444111206322003,
+      "grad_norm": 0.8639834260823381,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 27073
+    },
+    {
+      "epoch": 0.6444349233552319,
+      "grad_norm": 0.5893857298782762,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 27074
+    },
+    {
+      "epoch": 0.6444587260782634,
+      "grad_norm": 0.986187279213795,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 27075
+    },
+    {
+      "epoch": 0.6444825288012949,
+      "grad_norm": 15.577556995414996,
+      "learning_rate": 2e-05,
+      "loss": 0.7003,
+      "step": 27076
+    },
+    {
+      "epoch": 0.6445063315243263,
+      "grad_norm": 9.793233673912093,
+      "learning_rate": 2e-05,
+      "loss": 0.3214,
+      "step": 27077
+    },
+    {
+      "epoch": 0.6445301342473579,
+      "grad_norm": 3.5441442923905835,
+      "learning_rate": 2e-05,
+      "loss": 0.7186,
+      "step": 27078
+    },
+    {
+      "epoch": 0.6445539369703894,
+      "grad_norm": 2.4648423797346406,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 27079
+    },
+    {
+      "epoch": 0.6445777396934209,
+      "grad_norm": 5.973786528541473,
+      "learning_rate": 2e-05,
+      "loss": 1.3549,
+      "step": 27080
+    },
+    {
+      "epoch": 0.6446015424164524,
+      "grad_norm": 5.2539675786187265,
+      "learning_rate": 2e-05,
+      "loss": 0.9809,
+      "step": 27081
+    },
+    {
+      "epoch": 0.644625345139484,
+      "grad_norm": 8.235705851836777,
+      "learning_rate": 2e-05,
+      "loss": 0.3836,
+      "step": 27082
+    },
+    {
+      "epoch": 0.6446491478625155,
+      "grad_norm": 1.23377222013054,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 27083
+    },
+    {
+      "epoch": 0.644672950585547,
+      "grad_norm": 5.730167047693928,
+      "learning_rate": 2e-05,
+      "loss": 1.1403,
+      "step": 27084
+    },
+    {
+      "epoch": 0.6446967533085785,
+      "grad_norm": 5.244557190206848,
+      "learning_rate": 2e-05,
+      "loss": 0.9421,
+      "step": 27085
+    },
+    {
+      "epoch": 0.64472055603161,
+      "grad_norm": 0.9056886075184729,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 27086
+    },
+    {
+      "epoch": 0.6447443587546415,
+      "grad_norm": 1.967968097665002,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 27087
+    },
+    {
+      "epoch": 0.644768161477673,
+      "grad_norm": 11.848475289213098,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 27088
+    },
+    {
+      "epoch": 0.6447919642007046,
+      "grad_norm": 5.544659257523816,
+      "learning_rate": 2e-05,
+      "loss": 0.6029,
+      "step": 27089
+    },
+    {
+      "epoch": 0.6448157669237361,
+      "grad_norm": 6.25556958662372,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 27090
+    },
+    {
+      "epoch": 0.6448395696467676,
+      "grad_norm": 2.0259447854917054,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 27091
+    },
+    {
+      "epoch": 0.6448633723697991,
+      "grad_norm": 2.972963427841519,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 27092
+    },
+    {
+      "epoch": 0.6448871750928307,
+      "grad_norm": 6.380104365298768,
+      "learning_rate": 2e-05,
+      "loss": 0.6326,
+      "step": 27093
+    },
+    {
+      "epoch": 0.6449109778158622,
+      "grad_norm": 4.676540262240918,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 27094
+    },
+    {
+      "epoch": 0.6449347805388936,
+      "grad_norm": 7.469889217128288,
+      "learning_rate": 2e-05,
+      "loss": 0.7777,
+      "step": 27095
+    },
+    {
+      "epoch": 0.6449585832619251,
+      "grad_norm": 4.8506003066081,
+      "learning_rate": 2e-05,
+      "loss": 0.7525,
+      "step": 27096
+    },
+    {
+      "epoch": 0.6449823859849567,
+      "grad_norm": 4.424908086862171,
+      "learning_rate": 2e-05,
+      "loss": 0.64,
+      "step": 27097
+    },
+    {
+      "epoch": 0.6450061887079882,
+      "grad_norm": 0.2009525992318075,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 27098
+    },
+    {
+      "epoch": 0.6450299914310197,
+      "grad_norm": 6.406644325994725,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 27099
+    },
+    {
+      "epoch": 0.6450537941540512,
+      "grad_norm": 3.414891853303404,
+      "learning_rate": 2e-05,
+      "loss": 0.6555,
+      "step": 27100
+    },
+    {
+      "epoch": 0.6450775968770828,
+      "grad_norm": 1.7351974025274697,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 27101
+    },
+    {
+      "epoch": 0.6451013996001143,
+      "grad_norm": 12.70271565752036,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 27102
+    },
+    {
+      "epoch": 0.6451252023231457,
+      "grad_norm": 4.0584290409915456,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 27103
+    },
+    {
+      "epoch": 0.6451490050461773,
+      "grad_norm": 2.9277756803991144,
+      "learning_rate": 2e-05,
+      "loss": 0.2659,
+      "step": 27104
+    },
+    {
+      "epoch": 0.6451728077692088,
+      "grad_norm": 3.3938980281237248,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 27105
+    },
+    {
+      "epoch": 0.6451966104922403,
+      "grad_norm": 4.782288040514163,
+      "learning_rate": 2e-05,
+      "loss": 0.7766,
+      "step": 27106
+    },
+    {
+      "epoch": 0.6452204132152718,
+      "grad_norm": 0.13164186230129432,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 27107
+    },
+    {
+      "epoch": 0.6452442159383034,
+      "grad_norm": 4.024882173773745,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 27108
+    },
+    {
+      "epoch": 0.6452680186613349,
+      "grad_norm": 10.31875143218883,
+      "learning_rate": 2e-05,
+      "loss": 0.7276,
+      "step": 27109
+    },
+    {
+      "epoch": 0.6452918213843664,
+      "grad_norm": 0.2672082029466578,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 27110
+    },
+    {
+      "epoch": 0.6453156241073978,
+      "grad_norm": 0.10615394186876154,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 27111
+    },
+    {
+      "epoch": 0.6453394268304294,
+      "grad_norm": 3.842151560640956,
+      "learning_rate": 2e-05,
+      "loss": 0.1195,
+      "step": 27112
+    },
+    {
+      "epoch": 0.6453632295534609,
+      "grad_norm": 4.697646950333215,
+      "learning_rate": 2e-05,
+      "loss": 1.0806,
+      "step": 27113
+    },
+    {
+      "epoch": 0.6453870322764924,
+      "grad_norm": 0.12291583135899117,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 27114
+    },
+    {
+      "epoch": 0.6454108349995239,
+      "grad_norm": 4.567699997027311,
+      "learning_rate": 2e-05,
+      "loss": 0.8271,
+      "step": 27115
+    },
+    {
+      "epoch": 0.6454346377225555,
+      "grad_norm": 2.1424280265415385,
+      "learning_rate": 2e-05,
+      "loss": 0.3976,
+      "step": 27116
+    },
+    {
+      "epoch": 0.645458440445587,
+      "grad_norm": 3.2935885967790757,
+      "learning_rate": 2e-05,
+      "loss": 0.6158,
+      "step": 27117
+    },
+    {
+      "epoch": 0.6454822431686185,
+      "grad_norm": 5.860503171792679,
+      "learning_rate": 2e-05,
+      "loss": 0.9324,
+      "step": 27118
+    },
+    {
+      "epoch": 0.6455060458916501,
+      "grad_norm": 0.022368759923008744,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 27119
+    },
+    {
+      "epoch": 0.6455298486146815,
+      "grad_norm": 1.771382194998533,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 27120
+    },
+    {
+      "epoch": 0.645553651337713,
+      "grad_norm": 6.023610930813593,
+      "learning_rate": 2e-05,
+      "loss": 0.4175,
+      "step": 27121
+    },
+    {
+      "epoch": 0.6455774540607445,
+      "grad_norm": 0.037370767180216445,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 27122
+    },
+    {
+      "epoch": 0.6456012567837761,
+      "grad_norm": 2.885628218615805,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 27123
+    },
+    {
+      "epoch": 0.6456250595068076,
+      "grad_norm": 11.3007179270386,
+      "learning_rate": 2e-05,
+      "loss": 0.8366,
+      "step": 27124
+    },
+    {
+      "epoch": 0.6456488622298391,
+      "grad_norm": 0.18871667983462956,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27125
+    },
+    {
+      "epoch": 0.6456726649528706,
+      "grad_norm": 1.1688107222212245,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 27126
+    },
+    {
+      "epoch": 0.6456964676759022,
+      "grad_norm": 0.8409909635605978,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 27127
+    },
+    {
+      "epoch": 0.6457202703989336,
+      "grad_norm": 4.037418727758162,
+      "learning_rate": 2e-05,
+      "loss": 0.2708,
+      "step": 27128
+    },
+    {
+      "epoch": 0.6457440731219651,
+      "grad_norm": 5.766346029759947,
+      "learning_rate": 2e-05,
+      "loss": 1.8826,
+      "step": 27129
+    },
+    {
+      "epoch": 0.6457678758449966,
+      "grad_norm": 7.360705575537137,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 27130
+    },
+    {
+      "epoch": 0.6457916785680282,
+      "grad_norm": 2.2869083629376017,
+      "learning_rate": 2e-05,
+      "loss": 0.4089,
+      "step": 27131
+    },
+    {
+      "epoch": 0.6458154812910597,
+      "grad_norm": 0.6387664209475136,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 27132
+    },
+    {
+      "epoch": 0.6458392840140912,
+      "grad_norm": 0.8434896022091101,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 27133
+    },
+    {
+      "epoch": 0.6458630867371227,
+      "grad_norm": 5.566872213895363,
+      "learning_rate": 2e-05,
+      "loss": 0.1697,
+      "step": 27134
+    },
+    {
+      "epoch": 0.6458868894601543,
+      "grad_norm": 3.8513535983742018,
+      "learning_rate": 2e-05,
+      "loss": 0.5521,
+      "step": 27135
+    },
+    {
+      "epoch": 0.6459106921831858,
+      "grad_norm": 3.2430130765060983,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 27136
+    },
+    {
+      "epoch": 0.6459344949062172,
+      "grad_norm": 16.156548564847522,
+      "learning_rate": 2e-05,
+      "loss": 1.0028,
+      "step": 27137
+    },
+    {
+      "epoch": 0.6459582976292488,
+      "grad_norm": 0.26472632694326315,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 27138
+    },
+    {
+      "epoch": 0.6459821003522803,
+      "grad_norm": 3.179455911498914,
+      "learning_rate": 2e-05,
+      "loss": 0.0671,
+      "step": 27139
+    },
+    {
+      "epoch": 0.6460059030753118,
+      "grad_norm": 2.0915440956549496,
+      "learning_rate": 2e-05,
+      "loss": 0.2746,
+      "step": 27140
+    },
+    {
+      "epoch": 0.6460297057983433,
+      "grad_norm": 3.0802918562269546,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 27141
+    },
+    {
+      "epoch": 0.6460535085213749,
+      "grad_norm": 5.096801000762133,
+      "learning_rate": 2e-05,
+      "loss": 0.8739,
+      "step": 27142
+    },
+    {
+      "epoch": 0.6460773112444064,
+      "grad_norm": 6.309462723863115,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 27143
+    },
+    {
+      "epoch": 0.6461011139674379,
+      "grad_norm": 16.066800054682574,
+      "learning_rate": 2e-05,
+      "loss": 0.7051,
+      "step": 27144
+    },
+    {
+      "epoch": 0.6461249166904693,
+      "grad_norm": 0.5968434818770759,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 27145
+    },
+    {
+      "epoch": 0.6461487194135009,
+      "grad_norm": 11.755793135431563,
+      "learning_rate": 2e-05,
+      "loss": 0.4958,
+      "step": 27146
+    },
+    {
+      "epoch": 0.6461725221365324,
+      "grad_norm": 2.9864152889447184,
+      "learning_rate": 2e-05,
+      "loss": 0.4847,
+      "step": 27147
+    },
+    {
+      "epoch": 0.6461963248595639,
+      "grad_norm": 4.761032597119342,
+      "learning_rate": 2e-05,
+      "loss": 0.1387,
+      "step": 27148
+    },
+    {
+      "epoch": 0.6462201275825954,
+      "grad_norm": 6.333553423224961,
+      "learning_rate": 2e-05,
+      "loss": 1.9256,
+      "step": 27149
+    },
+    {
+      "epoch": 0.646243930305627,
+      "grad_norm": 15.064261374339228,
+      "learning_rate": 2e-05,
+      "loss": 0.4289,
+      "step": 27150
+    },
+    {
+      "epoch": 0.6462677330286585,
+      "grad_norm": 1.327780200693818,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 27151
+    },
+    {
+      "epoch": 0.64629153575169,
+      "grad_norm": 0.1551519153287479,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 27152
+    },
+    {
+      "epoch": 0.6463153384747216,
+      "grad_norm": 0.7319113943705682,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 27153
+    },
+    {
+      "epoch": 0.646339141197753,
+      "grad_norm": 2.4597006646971606,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 27154
+    },
+    {
+      "epoch": 0.6463629439207845,
+      "grad_norm": 3.786802706511079,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 27155
+    },
+    {
+      "epoch": 0.646386746643816,
+      "grad_norm": 0.29227737371572493,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 27156
+    },
+    {
+      "epoch": 0.6464105493668476,
+      "grad_norm": 1.3939818488832814,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 27157
+    },
+    {
+      "epoch": 0.6464343520898791,
+      "grad_norm": 2.758869915300396,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 27158
+    },
+    {
+      "epoch": 0.6464581548129106,
+      "grad_norm": 6.290614086050216,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 27159
+    },
+    {
+      "epoch": 0.6464819575359421,
+      "grad_norm": 8.557297055863106,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 27160
+    },
+    {
+      "epoch": 0.6465057602589737,
+      "grad_norm": 3.9133247683316656,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 27161
+    },
+    {
+      "epoch": 0.6465295629820051,
+      "grad_norm": 1.5084224396648056,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 27162
+    },
+    {
+      "epoch": 0.6465533657050366,
+      "grad_norm": 1.8984552772466896,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 27163
+    },
+    {
+      "epoch": 0.6465771684280681,
+      "grad_norm": 3.5141129332175534,
+      "learning_rate": 2e-05,
+      "loss": 0.5142,
+      "step": 27164
+    },
+    {
+      "epoch": 0.6466009711510997,
+      "grad_norm": 6.981832375001132,
+      "learning_rate": 2e-05,
+      "loss": 0.9995,
+      "step": 27165
+    },
+    {
+      "epoch": 0.6466247738741312,
+      "grad_norm": 4.076475103060824,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 27166
+    },
+    {
+      "epoch": 0.6466485765971627,
+      "grad_norm": 1.9384650818786493,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 27167
+    },
+    {
+      "epoch": 0.6466723793201943,
+      "grad_norm": 0.4968128222491147,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 27168
+    },
+    {
+      "epoch": 0.6466961820432258,
+      "grad_norm": 3.3703012640498167,
+      "learning_rate": 2e-05,
+      "loss": 0.411,
+      "step": 27169
+    },
+    {
+      "epoch": 0.6467199847662572,
+      "grad_norm": 0.4780617188662554,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 27170
+    },
+    {
+      "epoch": 0.6467437874892887,
+      "grad_norm": 13.841529474250471,
+      "learning_rate": 2e-05,
+      "loss": 1.5493,
+      "step": 27171
+    },
+    {
+      "epoch": 0.6467675902123203,
+      "grad_norm": 2.1462986260971553,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 27172
+    },
+    {
+      "epoch": 0.6467913929353518,
+      "grad_norm": 3.694961330157811,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 27173
+    },
+    {
+      "epoch": 0.6468151956583833,
+      "grad_norm": 7.125538918784112,
+      "learning_rate": 2e-05,
+      "loss": 0.9024,
+      "step": 27174
+    },
+    {
+      "epoch": 0.6468389983814148,
+      "grad_norm": 0.03707021185625707,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 27175
+    },
+    {
+      "epoch": 0.6468628011044464,
+      "grad_norm": 0.06116048234841915,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 27176
+    },
+    {
+      "epoch": 0.6468866038274779,
+      "grad_norm": 4.173760819917013,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 27177
+    },
+    {
+      "epoch": 0.6469104065505094,
+      "grad_norm": 0.04939010958046907,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 27178
+    },
+    {
+      "epoch": 0.6469342092735408,
+      "grad_norm": 3.607043812150142,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 27179
+    },
+    {
+      "epoch": 0.6469580119965724,
+      "grad_norm": 13.359052007736476,
+      "learning_rate": 2e-05,
+      "loss": 0.3141,
+      "step": 27180
+    },
+    {
+      "epoch": 0.6469818147196039,
+      "grad_norm": 3.611603388357671,
+      "learning_rate": 2e-05,
+      "loss": 0.1161,
+      "step": 27181
+    },
+    {
+      "epoch": 0.6470056174426354,
+      "grad_norm": 2.1941504130345937,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 27182
+    },
+    {
+      "epoch": 0.647029420165667,
+      "grad_norm": 3.118663214267875,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 27183
+    },
+    {
+      "epoch": 0.6470532228886985,
+      "grad_norm": 5.565405206664078,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 27184
+    },
+    {
+      "epoch": 0.64707702561173,
+      "grad_norm": 5.536487679202605,
+      "learning_rate": 2e-05,
+      "loss": 0.7353,
+      "step": 27185
+    },
+    {
+      "epoch": 0.6471008283347615,
+      "grad_norm": 6.936026827542723,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 27186
+    },
+    {
+      "epoch": 0.647124631057793,
+      "grad_norm": 1.5598365317766798,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 27187
+    },
+    {
+      "epoch": 0.6471484337808245,
+      "grad_norm": 0.4553972487061556,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 27188
+    },
+    {
+      "epoch": 0.647172236503856,
+      "grad_norm": 6.698088390452032,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 27189
+    },
+    {
+      "epoch": 0.6471960392268875,
+      "grad_norm": 4.369693528609373,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 27190
+    },
+    {
+      "epoch": 0.6472198419499191,
+      "grad_norm": 3.8183473090058344,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 27191
+    },
+    {
+      "epoch": 0.6472436446729506,
+      "grad_norm": 0.13355772286645892,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 27192
+    },
+    {
+      "epoch": 0.6472674473959821,
+      "grad_norm": 2.348747684602954,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 27193
+    },
+    {
+      "epoch": 0.6472912501190136,
+      "grad_norm": 3.348958837248685,
+      "learning_rate": 2e-05,
+      "loss": 0.3931,
+      "step": 27194
+    },
+    {
+      "epoch": 0.6473150528420452,
+      "grad_norm": 10.711897838208564,
+      "learning_rate": 2e-05,
+      "loss": 0.4607,
+      "step": 27195
+    },
+    {
+      "epoch": 0.6473388555650766,
+      "grad_norm": 7.207985994948128,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 27196
+    },
+    {
+      "epoch": 0.6473626582881081,
+      "grad_norm": 1.5593409626688133,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 27197
+    },
+    {
+      "epoch": 0.6473864610111397,
+      "grad_norm": 7.149870711148845,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 27198
+    },
+    {
+      "epoch": 0.6474102637341712,
+      "grad_norm": 0.7339778192093785,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 27199
+    },
+    {
+      "epoch": 0.6474340664572027,
+      "grad_norm": 9.262956848043817,
+      "learning_rate": 2e-05,
+      "loss": 0.2181,
+      "step": 27200
+    },
+    {
+      "epoch": 0.6474578691802342,
+      "grad_norm": 3.1971183787828763,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 27201
+    },
+    {
+      "epoch": 0.6474816719032658,
+      "grad_norm": 0.23829849361207253,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 27202
+    },
+    {
+      "epoch": 0.6475054746262973,
+      "grad_norm": 4.478265249701936,
+      "learning_rate": 2e-05,
+      "loss": 0.8265,
+      "step": 27203
+    },
+    {
+      "epoch": 0.6475292773493287,
+      "grad_norm": 0.8812555973702128,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 27204
+    },
+    {
+      "epoch": 0.6475530800723602,
+      "grad_norm": 7.532519415179031,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 27205
+    },
+    {
+      "epoch": 0.6475768827953918,
+      "grad_norm": 11.816340868574065,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 27206
+    },
+    {
+      "epoch": 0.6476006855184233,
+      "grad_norm": 2.64771089406509,
+      "learning_rate": 2e-05,
+      "loss": 0.3888,
+      "step": 27207
+    },
+    {
+      "epoch": 0.6476244882414548,
+      "grad_norm": 0.592120378923369,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 27208
+    },
+    {
+      "epoch": 0.6476482909644863,
+      "grad_norm": 7.043672724346991,
+      "learning_rate": 2e-05,
+      "loss": 0.7792,
+      "step": 27209
+    },
+    {
+      "epoch": 0.6476720936875179,
+      "grad_norm": 0.42421102193206367,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 27210
+    },
+    {
+      "epoch": 0.6476958964105494,
+      "grad_norm": 11.333934079253401,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 27211
+    },
+    {
+      "epoch": 0.6477196991335809,
+      "grad_norm": 6.683880874769116,
+      "learning_rate": 2e-05,
+      "loss": 1.0326,
+      "step": 27212
+    },
+    {
+      "epoch": 0.6477435018566124,
+      "grad_norm": 6.247419345812511,
+      "learning_rate": 2e-05,
+      "loss": 0.2753,
+      "step": 27213
+    },
+    {
+      "epoch": 0.6477673045796439,
+      "grad_norm": 0.5460580463506083,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 27214
+    },
+    {
+      "epoch": 0.6477911073026754,
+      "grad_norm": 2.061434876659338,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 27215
+    },
+    {
+      "epoch": 0.6478149100257069,
+      "grad_norm": 3.5069591128613067,
+      "learning_rate": 2e-05,
+      "loss": 0.7159,
+      "step": 27216
+    },
+    {
+      "epoch": 0.6478387127487385,
+      "grad_norm": 5.187329685883737,
+      "learning_rate": 2e-05,
+      "loss": 0.5669,
+      "step": 27217
+    },
+    {
+      "epoch": 0.64786251547177,
+      "grad_norm": 3.5148206405996802,
+      "learning_rate": 2e-05,
+      "loss": 0.8374,
+      "step": 27218
+    },
+    {
+      "epoch": 0.6478863181948015,
+      "grad_norm": 3.0502810632450785,
+      "learning_rate": 2e-05,
+      "loss": 0.5381,
+      "step": 27219
+    },
+    {
+      "epoch": 0.647910120917833,
+      "grad_norm": 5.853021341159537,
+      "learning_rate": 2e-05,
+      "loss": 0.2233,
+      "step": 27220
+    },
+    {
+      "epoch": 0.6479339236408646,
+      "grad_norm": 5.620583130907373,
+      "learning_rate": 2e-05,
+      "loss": 1.1874,
+      "step": 27221
+    },
+    {
+      "epoch": 0.647957726363896,
+      "grad_norm": 3.1149011670002182,
+      "learning_rate": 2e-05,
+      "loss": 0.2042,
+      "step": 27222
+    },
+    {
+      "epoch": 0.6479815290869275,
+      "grad_norm": 11.174214515052842,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 27223
+    },
+    {
+      "epoch": 0.648005331809959,
+      "grad_norm": 2.013997517693867,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 27224
+    },
+    {
+      "epoch": 0.6480291345329906,
+      "grad_norm": 1.0919895034064229,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 27225
+    },
+    {
+      "epoch": 0.6480529372560221,
+      "grad_norm": 5.345808007691553,
+      "learning_rate": 2e-05,
+      "loss": 1.377,
+      "step": 27226
+    },
+    {
+      "epoch": 0.6480767399790536,
+      "grad_norm": 1.4992937348455642,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 27227
+    },
+    {
+      "epoch": 0.6481005427020852,
+      "grad_norm": 1.653355213132562,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 27228
+    },
+    {
+      "epoch": 0.6481243454251167,
+      "grad_norm": 1.9885594041278416,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 27229
+    },
+    {
+      "epoch": 0.6481481481481481,
+      "grad_norm": 2.3958219835346446,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 27230
+    },
+    {
+      "epoch": 0.6481719508711796,
+      "grad_norm": 4.059859863326623,
+      "learning_rate": 2e-05,
+      "loss": 0.4705,
+      "step": 27231
+    },
+    {
+      "epoch": 0.6481957535942112,
+      "grad_norm": 1.114204552346764,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 27232
+    },
+    {
+      "epoch": 0.6482195563172427,
+      "grad_norm": 3.7888069808821956,
+      "learning_rate": 2e-05,
+      "loss": 0.5536,
+      "step": 27233
+    },
+    {
+      "epoch": 0.6482433590402742,
+      "grad_norm": 1.885401056503809,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 27234
+    },
+    {
+      "epoch": 0.6482671617633057,
+      "grad_norm": 3.643825372512679,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 27235
+    },
+    {
+      "epoch": 0.6482909644863373,
+      "grad_norm": 0.0824239185588841,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 27236
+    },
+    {
+      "epoch": 0.6483147672093688,
+      "grad_norm": 3.606254496871583,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 27237
+    },
+    {
+      "epoch": 0.6483385699324002,
+      "grad_norm": 4.49933740907293,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 27238
+    },
+    {
+      "epoch": 0.6483623726554317,
+      "grad_norm": 0.023744829772802095,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 27239
+    },
+    {
+      "epoch": 0.6483861753784633,
+      "grad_norm": 2.8413792051178333,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 27240
+    },
+    {
+      "epoch": 0.6484099781014948,
+      "grad_norm": 0.12416479269722808,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27241
+    },
+    {
+      "epoch": 0.6484337808245263,
+      "grad_norm": 4.312538976105556,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 27242
+    },
+    {
+      "epoch": 0.6484575835475579,
+      "grad_norm": 0.2561064343204797,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 27243
+    },
+    {
+      "epoch": 0.6484813862705894,
+      "grad_norm": 16.22770746441594,
+      "learning_rate": 2e-05,
+      "loss": 1.1069,
+      "step": 27244
+    },
+    {
+      "epoch": 0.6485051889936209,
+      "grad_norm": 0.435999290468983,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 27245
+    },
+    {
+      "epoch": 0.6485289917166523,
+      "grad_norm": 0.6148984273587137,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 27246
+    },
+    {
+      "epoch": 0.6485527944396839,
+      "grad_norm": 1.9698698906858851,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 27247
+    },
+    {
+      "epoch": 0.6485765971627154,
+      "grad_norm": 0.646789063187286,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 27248
+    },
+    {
+      "epoch": 0.6486003998857469,
+      "grad_norm": 1.8973166775106436,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 27249
+    },
+    {
+      "epoch": 0.6486242026087784,
+      "grad_norm": 1.7373916034888839,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 27250
+    },
+    {
+      "epoch": 0.64864800533181,
+      "grad_norm": 4.171715642873998,
+      "learning_rate": 2e-05,
+      "loss": 0.8842,
+      "step": 27251
+    },
+    {
+      "epoch": 0.6486718080548415,
+      "grad_norm": 0.719386041142822,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 27252
+    },
+    {
+      "epoch": 0.648695610777873,
+      "grad_norm": 0.860135442207758,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 27253
+    },
+    {
+      "epoch": 0.6487194135009045,
+      "grad_norm": 0.05002175629954077,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 27254
+    },
+    {
+      "epoch": 0.648743216223936,
+      "grad_norm": 0.6107969439478832,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 27255
+    },
+    {
+      "epoch": 0.6487670189469675,
+      "grad_norm": 3.34591109427466,
+      "learning_rate": 2e-05,
+      "loss": 0.6243,
+      "step": 27256
+    },
+    {
+      "epoch": 0.648790821669999,
+      "grad_norm": 6.133837213856421,
+      "learning_rate": 2e-05,
+      "loss": 1.2678,
+      "step": 27257
+    },
+    {
+      "epoch": 0.6488146243930306,
+      "grad_norm": 2.6745271363735945,
+      "learning_rate": 2e-05,
+      "loss": 0.2647,
+      "step": 27258
+    },
+    {
+      "epoch": 0.6488384271160621,
+      "grad_norm": 1.0144705558639218,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 27259
+    },
+    {
+      "epoch": 0.6488622298390936,
+      "grad_norm": 0.0768200500439903,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 27260
+    },
+    {
+      "epoch": 0.6488860325621251,
+      "grad_norm": 5.142333470971178,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 27261
+    },
+    {
+      "epoch": 0.6489098352851567,
+      "grad_norm": 4.501336134156783,
+      "learning_rate": 2e-05,
+      "loss": 0.7593,
+      "step": 27262
+    },
+    {
+      "epoch": 0.6489336380081882,
+      "grad_norm": 7.7250407607937,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 27263
+    },
+    {
+      "epoch": 0.6489574407312196,
+      "grad_norm": 0.723235079234848,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 27264
+    },
+    {
+      "epoch": 0.6489812434542511,
+      "grad_norm": 3.0647759469367895,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 27265
+    },
+    {
+      "epoch": 0.6490050461772827,
+      "grad_norm": 0.017994956831127736,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 27266
+    },
+    {
+      "epoch": 0.6490288489003142,
+      "grad_norm": 4.653238829444279,
+      "learning_rate": 2e-05,
+      "loss": 0.6576,
+      "step": 27267
+    },
+    {
+      "epoch": 0.6490526516233457,
+      "grad_norm": 7.645140525915796,
+      "learning_rate": 2e-05,
+      "loss": 1.5833,
+      "step": 27268
+    },
+    {
+      "epoch": 0.6490764543463772,
+      "grad_norm": 1.2163709337316806,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 27269
+    },
+    {
+      "epoch": 0.6491002570694088,
+      "grad_norm": 4.354004622472748,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 27270
+    },
+    {
+      "epoch": 0.6491240597924403,
+      "grad_norm": 3.3650231883823998,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 27271
+    },
+    {
+      "epoch": 0.6491478625154717,
+      "grad_norm": 5.231515101588632,
+      "learning_rate": 2e-05,
+      "loss": 0.2873,
+      "step": 27272
+    },
+    {
+      "epoch": 0.6491716652385033,
+      "grad_norm": 4.60418396829077,
+      "learning_rate": 2e-05,
+      "loss": 0.4959,
+      "step": 27273
+    },
+    {
+      "epoch": 0.6491954679615348,
+      "grad_norm": 0.1477435978906373,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27274
+    },
+    {
+      "epoch": 0.6492192706845663,
+      "grad_norm": 17.87445453165853,
+      "learning_rate": 2e-05,
+      "loss": 0.63,
+      "step": 27275
+    },
+    {
+      "epoch": 0.6492430734075978,
+      "grad_norm": 4.086160848438267,
+      "learning_rate": 2e-05,
+      "loss": 0.7612,
+      "step": 27276
+    },
+    {
+      "epoch": 0.6492668761306294,
+      "grad_norm": 1.050551386439508,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 27277
+    },
+    {
+      "epoch": 0.6492906788536609,
+      "grad_norm": 1.0336948556798022,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 27278
+    },
+    {
+      "epoch": 0.6493144815766924,
+      "grad_norm": 3.57805785451035,
+      "learning_rate": 2e-05,
+      "loss": 0.4223,
+      "step": 27279
+    },
+    {
+      "epoch": 0.6493382842997238,
+      "grad_norm": 2.86265557380415,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 27280
+    },
+    {
+      "epoch": 0.6493620870227554,
+      "grad_norm": 12.844316918972645,
+      "learning_rate": 2e-05,
+      "loss": 0.6968,
+      "step": 27281
+    },
+    {
+      "epoch": 0.6493858897457869,
+      "grad_norm": 3.498444306046423,
+      "learning_rate": 2e-05,
+      "loss": 0.4539,
+      "step": 27282
+    },
+    {
+      "epoch": 0.6494096924688184,
+      "grad_norm": 1.159962594089886,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 27283
+    },
+    {
+      "epoch": 0.6494334951918499,
+      "grad_norm": 12.326037103842276,
+      "learning_rate": 2e-05,
+      "loss": 0.9864,
+      "step": 27284
+    },
+    {
+      "epoch": 0.6494572979148815,
+      "grad_norm": 4.820203488206683,
+      "learning_rate": 2e-05,
+      "loss": 1.184,
+      "step": 27285
+    },
+    {
+      "epoch": 0.649481100637913,
+      "grad_norm": 3.799675097500785,
+      "learning_rate": 2e-05,
+      "loss": 0.4789,
+      "step": 27286
+    },
+    {
+      "epoch": 0.6495049033609445,
+      "grad_norm": 2.9574755508526476,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 27287
+    },
+    {
+      "epoch": 0.6495287060839761,
+      "grad_norm": 4.022254485017231,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 27288
+    },
+    {
+      "epoch": 0.6495525088070075,
+      "grad_norm": 0.1201171105882259,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27289
+    },
+    {
+      "epoch": 0.649576311530039,
+      "grad_norm": 2.688221514199707,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 27290
+    },
+    {
+      "epoch": 0.6496001142530705,
+      "grad_norm": 10.673289176520179,
+      "learning_rate": 2e-05,
+      "loss": 0.4024,
+      "step": 27291
+    },
+    {
+      "epoch": 0.6496239169761021,
+      "grad_norm": 3.0855533933017503,
+      "learning_rate": 2e-05,
+      "loss": 0.4601,
+      "step": 27292
+    },
+    {
+      "epoch": 0.6496477196991336,
+      "grad_norm": 1.1708743074380281,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 27293
+    },
+    {
+      "epoch": 0.6496715224221651,
+      "grad_norm": 7.148545318161947,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 27294
+    },
+    {
+      "epoch": 0.6496953251451966,
+      "grad_norm": 6.461516408675668,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 27295
+    },
+    {
+      "epoch": 0.6497191278682282,
+      "grad_norm": 8.415485299295483,
+      "learning_rate": 2e-05,
+      "loss": 0.5302,
+      "step": 27296
+    },
+    {
+      "epoch": 0.6497429305912596,
+      "grad_norm": 1.664364953973765,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 27297
+    },
+    {
+      "epoch": 0.6497667333142911,
+      "grad_norm": 3.4812506906384604,
+      "learning_rate": 2e-05,
+      "loss": 0.6287,
+      "step": 27298
+    },
+    {
+      "epoch": 0.6497905360373226,
+      "grad_norm": 0.6180420122582947,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 27299
+    },
+    {
+      "epoch": 0.6498143387603542,
+      "grad_norm": 4.316217878337581,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 27300
+    },
+    {
+      "epoch": 0.6498381414833857,
+      "grad_norm": 0.4102886818341943,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 27301
+    },
+    {
+      "epoch": 0.6498619442064172,
+      "grad_norm": 1.308743047417761,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 27302
+    },
+    {
+      "epoch": 0.6498857469294488,
+      "grad_norm": 0.023356028686758155,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 27303
+    },
+    {
+      "epoch": 0.6499095496524803,
+      "grad_norm": 5.076708592845313,
+      "learning_rate": 2e-05,
+      "loss": 0.7238,
+      "step": 27304
+    },
+    {
+      "epoch": 0.6499333523755118,
+      "grad_norm": 0.5546365296364563,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 27305
+    },
+    {
+      "epoch": 0.6499571550985432,
+      "grad_norm": 6.279785443556702,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 27306
+    },
+    {
+      "epoch": 0.6499809578215748,
+      "grad_norm": 36.1200180731795,
+      "learning_rate": 2e-05,
+      "loss": 0.7448,
+      "step": 27307
+    },
+    {
+      "epoch": 0.6500047605446063,
+      "grad_norm": 2.4156672859448585,
+      "learning_rate": 2e-05,
+      "loss": 0.1841,
+      "step": 27308
+    },
+    {
+      "epoch": 0.6500285632676378,
+      "grad_norm": 1.8687805017663113,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 27309
+    },
+    {
+      "epoch": 0.6500523659906693,
+      "grad_norm": 26.914207380615643,
+      "learning_rate": 2e-05,
+      "loss": 2.2808,
+      "step": 27310
+    },
+    {
+      "epoch": 0.6500761687137009,
+      "grad_norm": 2.8043304615591937,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 27311
+    },
+    {
+      "epoch": 0.6500999714367324,
+      "grad_norm": 8.099817950685967,
+      "learning_rate": 2e-05,
+      "loss": 0.0802,
+      "step": 27312
+    },
+    {
+      "epoch": 0.6501237741597639,
+      "grad_norm": 5.993112531762829,
+      "learning_rate": 2e-05,
+      "loss": 0.2425,
+      "step": 27313
+    },
+    {
+      "epoch": 0.6501475768827953,
+      "grad_norm": 4.260050447848591,
+      "learning_rate": 2e-05,
+      "loss": 0.6634,
+      "step": 27314
+    },
+    {
+      "epoch": 0.6501713796058269,
+      "grad_norm": 3.5884163340764066,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 27315
+    },
+    {
+      "epoch": 0.6501951823288584,
+      "grad_norm": 2.2625760855656796,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 27316
+    },
+    {
+      "epoch": 0.6502189850518899,
+      "grad_norm": 9.380986191418488,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 27317
+    },
+    {
+      "epoch": 0.6502427877749215,
+      "grad_norm": 5.621875617009391,
+      "learning_rate": 2e-05,
+      "loss": 1.2426,
+      "step": 27318
+    },
+    {
+      "epoch": 0.650266590497953,
+      "grad_norm": 1.631903111565198,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 27319
+    },
+    {
+      "epoch": 0.6502903932209845,
+      "grad_norm": 4.554576592998911,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 27320
+    },
+    {
+      "epoch": 0.650314195944016,
+      "grad_norm": 4.338930380154819,
+      "learning_rate": 2e-05,
+      "loss": 0.7205,
+      "step": 27321
+    },
+    {
+      "epoch": 0.6503379986670476,
+      "grad_norm": 0.5518010675389228,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 27322
+    },
+    {
+      "epoch": 0.650361801390079,
+      "grad_norm": 0.29841245292638874,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 27323
+    },
+    {
+      "epoch": 0.6503856041131105,
+      "grad_norm": 3.0110689994502917,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 27324
+    },
+    {
+      "epoch": 0.650409406836142,
+      "grad_norm": 2.682081475995961,
+      "learning_rate": 2e-05,
+      "loss": 0.2894,
+      "step": 27325
+    },
+    {
+      "epoch": 0.6504332095591736,
+      "grad_norm": 0.06823682031393165,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 27326
+    },
+    {
+      "epoch": 0.6504570122822051,
+      "grad_norm": 2.769639216783471,
+      "learning_rate": 2e-05,
+      "loss": 0.3171,
+      "step": 27327
+    },
+    {
+      "epoch": 0.6504808150052366,
+      "grad_norm": 0.4361157009884796,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 27328
+    },
+    {
+      "epoch": 0.6505046177282681,
+      "grad_norm": 3.349207568408041,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 27329
+    },
+    {
+      "epoch": 0.6505284204512997,
+      "grad_norm": 32.53398941927649,
+      "learning_rate": 2e-05,
+      "loss": 0.5187,
+      "step": 27330
+    },
+    {
+      "epoch": 0.6505522231743311,
+      "grad_norm": 2.1218004389731755,
+      "learning_rate": 2e-05,
+      "loss": 0.1549,
+      "step": 27331
+    },
+    {
+      "epoch": 0.6505760258973626,
+      "grad_norm": 1.5158485240504447,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 27332
+    },
+    {
+      "epoch": 0.6505998286203942,
+      "grad_norm": 2.6053443032262185,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 27333
+    },
+    {
+      "epoch": 0.6506236313434257,
+      "grad_norm": 5.186243823428245,
+      "learning_rate": 2e-05,
+      "loss": 0.9281,
+      "step": 27334
+    },
+    {
+      "epoch": 0.6506474340664572,
+      "grad_norm": 5.479517864959901,
+      "learning_rate": 2e-05,
+      "loss": 0.0898,
+      "step": 27335
+    },
+    {
+      "epoch": 0.6506712367894887,
+      "grad_norm": 3.1354266115591103,
+      "learning_rate": 2e-05,
+      "loss": 0.4715,
+      "step": 27336
+    },
+    {
+      "epoch": 0.6506950395125203,
+      "grad_norm": 4.668551819090237,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 27337
+    },
+    {
+      "epoch": 0.6507188422355518,
+      "grad_norm": 6.954084184203343,
+      "learning_rate": 2e-05,
+      "loss": 0.9753,
+      "step": 27338
+    },
+    {
+      "epoch": 0.6507426449585832,
+      "grad_norm": 1.6413273909128603,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 27339
+    },
+    {
+      "epoch": 0.6507664476816147,
+      "grad_norm": 1.9773372756153147,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 27340
+    },
+    {
+      "epoch": 0.6507902504046463,
+      "grad_norm": 0.021381366862083005,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 27341
+    },
+    {
+      "epoch": 0.6508140531276778,
+      "grad_norm": 2.260115601991048,
+      "learning_rate": 2e-05,
+      "loss": 0.0925,
+      "step": 27342
+    },
+    {
+      "epoch": 0.6508378558507093,
+      "grad_norm": 11.557238033117176,
+      "learning_rate": 2e-05,
+      "loss": 1.4145,
+      "step": 27343
+    },
+    {
+      "epoch": 0.6508616585737408,
+      "grad_norm": 13.128546692204509,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 27344
+    },
+    {
+      "epoch": 0.6508854612967724,
+      "grad_norm": 0.8517529322655762,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 27345
+    },
+    {
+      "epoch": 0.6509092640198039,
+      "grad_norm": 3.427439296343955,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 27346
+    },
+    {
+      "epoch": 0.6509330667428354,
+      "grad_norm": 3.4691001284481082,
+      "learning_rate": 2e-05,
+      "loss": 0.3417,
+      "step": 27347
+    },
+    {
+      "epoch": 0.650956869465867,
+      "grad_norm": 3.3164266896578365,
+      "learning_rate": 2e-05,
+      "loss": 0.4058,
+      "step": 27348
+    },
+    {
+      "epoch": 0.6509806721888984,
+      "grad_norm": 4.857450431311778,
+      "learning_rate": 2e-05,
+      "loss": 0.832,
+      "step": 27349
+    },
+    {
+      "epoch": 0.6510044749119299,
+      "grad_norm": 0.14108702175812582,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 27350
+    },
+    {
+      "epoch": 0.6510282776349614,
+      "grad_norm": 15.505265423195924,
+      "learning_rate": 2e-05,
+      "loss": 0.8099,
+      "step": 27351
+    },
+    {
+      "epoch": 0.651052080357993,
+      "grad_norm": 2.1516921116889662,
+      "learning_rate": 2e-05,
+      "loss": 0.5242,
+      "step": 27352
+    },
+    {
+      "epoch": 0.6510758830810245,
+      "grad_norm": 9.058490228423281,
+      "learning_rate": 2e-05,
+      "loss": 1.608,
+      "step": 27353
+    },
+    {
+      "epoch": 0.651099685804056,
+      "grad_norm": 6.748143669959056,
+      "learning_rate": 2e-05,
+      "loss": 0.8349,
+      "step": 27354
+    },
+    {
+      "epoch": 0.6511234885270875,
+      "grad_norm": 0.48264052608178293,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 27355
+    },
+    {
+      "epoch": 0.651147291250119,
+      "grad_norm": 9.187950447994718,
+      "learning_rate": 2e-05,
+      "loss": 0.4911,
+      "step": 27356
+    },
+    {
+      "epoch": 0.6511710939731505,
+      "grad_norm": 0.8328715464822907,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 27357
+    },
+    {
+      "epoch": 0.651194896696182,
+      "grad_norm": 1.2619549096417941,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 27358
+    },
+    {
+      "epoch": 0.6512186994192135,
+      "grad_norm": 2.665031202648482,
+      "learning_rate": 2e-05,
+      "loss": 0.3421,
+      "step": 27359
+    },
+    {
+      "epoch": 0.6512425021422451,
+      "grad_norm": 5.080018478768883,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 27360
+    },
+    {
+      "epoch": 0.6512663048652766,
+      "grad_norm": 11.637422839825476,
+      "learning_rate": 2e-05,
+      "loss": 1.2838,
+      "step": 27361
+    },
+    {
+      "epoch": 0.6512901075883081,
+      "grad_norm": 4.673356196572989,
+      "learning_rate": 2e-05,
+      "loss": 0.8898,
+      "step": 27362
+    },
+    {
+      "epoch": 0.6513139103113397,
+      "grad_norm": 4.499811816583962,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 27363
+    },
+    {
+      "epoch": 0.6513377130343712,
+      "grad_norm": 2.588641802238249,
+      "learning_rate": 2e-05,
+      "loss": 0.4733,
+      "step": 27364
+    },
+    {
+      "epoch": 0.6513615157574026,
+      "grad_norm": 3.656767247300546,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 27365
+    },
+    {
+      "epoch": 0.6513853184804341,
+      "grad_norm": 3.496476112501003,
+      "learning_rate": 2e-05,
+      "loss": 0.1812,
+      "step": 27366
+    },
+    {
+      "epoch": 0.6514091212034657,
+      "grad_norm": 5.100016092595639,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 27367
+    },
+    {
+      "epoch": 0.6514329239264972,
+      "grad_norm": 2.79670109343983,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 27368
+    },
+    {
+      "epoch": 0.6514567266495287,
+      "grad_norm": 3.4517527349877293,
+      "learning_rate": 2e-05,
+      "loss": 0.288,
+      "step": 27369
+    },
+    {
+      "epoch": 0.6514805293725602,
+      "grad_norm": 0.6640145743731987,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 27370
+    },
+    {
+      "epoch": 0.6515043320955918,
+      "grad_norm": 6.094467969271445,
+      "learning_rate": 2e-05,
+      "loss": 0.9005,
+      "step": 27371
+    },
+    {
+      "epoch": 0.6515281348186233,
+      "grad_norm": 3.7460557065643814,
+      "learning_rate": 2e-05,
+      "loss": 0.4967,
+      "step": 27372
+    },
+    {
+      "epoch": 0.6515519375416547,
+      "grad_norm": 1.5624044684507754,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 27373
+    },
+    {
+      "epoch": 0.6515757402646862,
+      "grad_norm": 0.2926075209460004,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 27374
+    },
+    {
+      "epoch": 0.6515995429877178,
+      "grad_norm": 0.024938287497746615,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 27375
+    },
+    {
+      "epoch": 0.6516233457107493,
+      "grad_norm": 2.9880791111563028,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 27376
+    },
+    {
+      "epoch": 0.6516471484337808,
+      "grad_norm": 12.863963513106802,
+      "learning_rate": 2e-05,
+      "loss": 0.4153,
+      "step": 27377
+    },
+    {
+      "epoch": 0.6516709511568124,
+      "grad_norm": 4.862699441374417,
+      "learning_rate": 2e-05,
+      "loss": 0.6111,
+      "step": 27378
+    },
+    {
+      "epoch": 0.6516947538798439,
+      "grad_norm": 0.2777841481626373,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 27379
+    },
+    {
+      "epoch": 0.6517185566028754,
+      "grad_norm": 0.6464368230101246,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 27380
+    },
+    {
+      "epoch": 0.6517423593259069,
+      "grad_norm": 3.5793668181611933,
+      "learning_rate": 2e-05,
+      "loss": 0.5548,
+      "step": 27381
+    },
+    {
+      "epoch": 0.6517661620489384,
+      "grad_norm": 2.1262133783145933,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 27382
+    },
+    {
+      "epoch": 0.6517899647719699,
+      "grad_norm": 2.9388010567969474,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 27383
+    },
+    {
+      "epoch": 0.6518137674950014,
+      "grad_norm": 3.4900646727076596,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 27384
+    },
+    {
+      "epoch": 0.6518375702180329,
+      "grad_norm": 8.091073196309376,
+      "learning_rate": 2e-05,
+      "loss": 0.5708,
+      "step": 27385
+    },
+    {
+      "epoch": 0.6518613729410645,
+      "grad_norm": 0.1540509794643469,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 27386
+    },
+    {
+      "epoch": 0.651885175664096,
+      "grad_norm": 0.3133224123698667,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 27387
+    },
+    {
+      "epoch": 0.6519089783871275,
+      "grad_norm": 1.7039909205090535,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 27388
+    },
+    {
+      "epoch": 0.651932781110159,
+      "grad_norm": 23.48745763307274,
+      "learning_rate": 2e-05,
+      "loss": 0.463,
+      "step": 27389
+    },
+    {
+      "epoch": 0.6519565838331906,
+      "grad_norm": 2.841754928043472,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 27390
+    },
+    {
+      "epoch": 0.651980386556222,
+      "grad_norm": 6.379902493950773,
+      "learning_rate": 2e-05,
+      "loss": 0.2903,
+      "step": 27391
+    },
+    {
+      "epoch": 0.6520041892792535,
+      "grad_norm": 2.532705740270565,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 27392
+    },
+    {
+      "epoch": 0.652027992002285,
+      "grad_norm": 4.4124688432977806,
+      "learning_rate": 2e-05,
+      "loss": 0.8049,
+      "step": 27393
+    },
+    {
+      "epoch": 0.6520517947253166,
+      "grad_norm": 5.75907049084075,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 27394
+    },
+    {
+      "epoch": 0.6520755974483481,
+      "grad_norm": 9.943741508116817,
+      "learning_rate": 2e-05,
+      "loss": 0.6173,
+      "step": 27395
+    },
+    {
+      "epoch": 0.6520994001713796,
+      "grad_norm": 6.520814550279615,
+      "learning_rate": 2e-05,
+      "loss": 1.3594,
+      "step": 27396
+    },
+    {
+      "epoch": 0.6521232028944112,
+      "grad_norm": 0.24077949282552288,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 27397
+    },
+    {
+      "epoch": 0.6521470056174427,
+      "grad_norm": 6.380465157735764,
+      "learning_rate": 2e-05,
+      "loss": 1.0348,
+      "step": 27398
+    },
+    {
+      "epoch": 0.6521708083404741,
+      "grad_norm": 1.4752686027754975,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 27399
+    },
+    {
+      "epoch": 0.6521946110635056,
+      "grad_norm": 4.576469843655284,
+      "learning_rate": 2e-05,
+      "loss": 0.7567,
+      "step": 27400
+    },
+    {
+      "epoch": 0.6522184137865372,
+      "grad_norm": 3.3627816516373006,
+      "learning_rate": 2e-05,
+      "loss": 0.5582,
+      "step": 27401
+    },
+    {
+      "epoch": 0.6522422165095687,
+      "grad_norm": 0.15749674987635487,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 27402
+    },
+    {
+      "epoch": 0.6522660192326002,
+      "grad_norm": 2.621997553796633,
+      "learning_rate": 2e-05,
+      "loss": 0.3354,
+      "step": 27403
+    },
+    {
+      "epoch": 0.6522898219556317,
+      "grad_norm": 6.022170434659098,
+      "learning_rate": 2e-05,
+      "loss": 0.1805,
+      "step": 27404
+    },
+    {
+      "epoch": 0.6523136246786633,
+      "grad_norm": 0.3047814308162247,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 27405
+    },
+    {
+      "epoch": 0.6523374274016948,
+      "grad_norm": 2.3251171312522403,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 27406
+    },
+    {
+      "epoch": 0.6523612301247262,
+      "grad_norm": 0.4893732366885568,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 27407
+    },
+    {
+      "epoch": 0.6523850328477577,
+      "grad_norm": 5.696192658505683,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 27408
+    },
+    {
+      "epoch": 0.6524088355707893,
+      "grad_norm": 2.805639266313234,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 27409
+    },
+    {
+      "epoch": 0.6524326382938208,
+      "grad_norm": 7.755714966142582,
+      "learning_rate": 2e-05,
+      "loss": 0.3615,
+      "step": 27410
+    },
+    {
+      "epoch": 0.6524564410168523,
+      "grad_norm": 7.542353702984835,
+      "learning_rate": 2e-05,
+      "loss": 1.0192,
+      "step": 27411
+    },
+    {
+      "epoch": 0.6524802437398839,
+      "grad_norm": 5.067747946424836,
+      "learning_rate": 2e-05,
+      "loss": 1.2114,
+      "step": 27412
+    },
+    {
+      "epoch": 0.6525040464629154,
+      "grad_norm": 1.5167207537376435,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 27413
+    },
+    {
+      "epoch": 0.6525278491859469,
+      "grad_norm": 0.6836349993187024,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 27414
+    },
+    {
+      "epoch": 0.6525516519089783,
+      "grad_norm": 3.2217550192096676,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 27415
+    },
+    {
+      "epoch": 0.6525754546320099,
+      "grad_norm": 10.633760249135701,
+      "learning_rate": 2e-05,
+      "loss": 0.3692,
+      "step": 27416
+    },
+    {
+      "epoch": 0.6525992573550414,
+      "grad_norm": 1.1418960651569994,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 27417
+    },
+    {
+      "epoch": 0.6526230600780729,
+      "grad_norm": 3.452388124898174,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 27418
+    },
+    {
+      "epoch": 0.6526468628011044,
+      "grad_norm": 5.60274518852494,
+      "learning_rate": 2e-05,
+      "loss": 1.2158,
+      "step": 27419
+    },
+    {
+      "epoch": 0.652670665524136,
+      "grad_norm": 6.010516701986727,
+      "learning_rate": 2e-05,
+      "loss": 1.2848,
+      "step": 27420
+    },
+    {
+      "epoch": 0.6526944682471675,
+      "grad_norm": 0.20124907134704922,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 27421
+    },
+    {
+      "epoch": 0.652718270970199,
+      "grad_norm": 2.0832078938089262,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 27422
+    },
+    {
+      "epoch": 0.6527420736932305,
+      "grad_norm": 6.951735753558691,
+      "learning_rate": 2e-05,
+      "loss": 0.6393,
+      "step": 27423
+    },
+    {
+      "epoch": 0.652765876416262,
+      "grad_norm": 1.8414041162228527,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 27424
+    },
+    {
+      "epoch": 0.6527896791392935,
+      "grad_norm": 0.2044611819720536,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 27425
+    },
+    {
+      "epoch": 0.652813481862325,
+      "grad_norm": 0.8839668310317386,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 27426
+    },
+    {
+      "epoch": 0.6528372845853566,
+      "grad_norm": 0.9311373111960286,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 27427
+    },
+    {
+      "epoch": 0.6528610873083881,
+      "grad_norm": 3.8355031649801434,
+      "learning_rate": 2e-05,
+      "loss": 0.6079,
+      "step": 27428
+    },
+    {
+      "epoch": 0.6528848900314196,
+      "grad_norm": 0.15475160756257525,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 27429
+    },
+    {
+      "epoch": 0.6529086927544511,
+      "grad_norm": 6.065431538916436,
+      "learning_rate": 2e-05,
+      "loss": 0.5301,
+      "step": 27430
+    },
+    {
+      "epoch": 0.6529324954774827,
+      "grad_norm": 19.188252968513915,
+      "learning_rate": 2e-05,
+      "loss": 0.5169,
+      "step": 27431
+    },
+    {
+      "epoch": 0.6529562982005142,
+      "grad_norm": 2.5868328884491136,
+      "learning_rate": 2e-05,
+      "loss": 0.4643,
+      "step": 27432
+    },
+    {
+      "epoch": 0.6529801009235456,
+      "grad_norm": 4.829457467414095,
+      "learning_rate": 2e-05,
+      "loss": 0.7142,
+      "step": 27433
+    },
+    {
+      "epoch": 0.6530039036465771,
+      "grad_norm": 0.3016630959121644,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 27434
+    },
+    {
+      "epoch": 0.6530277063696087,
+      "grad_norm": 4.197172418582534,
+      "learning_rate": 2e-05,
+      "loss": 0.735,
+      "step": 27435
+    },
+    {
+      "epoch": 0.6530515090926402,
+      "grad_norm": 2.6787791835702386,
+      "learning_rate": 2e-05,
+      "loss": 0.4584,
+      "step": 27436
+    },
+    {
+      "epoch": 0.6530753118156717,
+      "grad_norm": 9.281326824368884,
+      "learning_rate": 2e-05,
+      "loss": 0.7118,
+      "step": 27437
+    },
+    {
+      "epoch": 0.6530991145387032,
+      "grad_norm": 3.6225986281712808,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 27438
+    },
+    {
+      "epoch": 0.6531229172617348,
+      "grad_norm": 0.1499175017096967,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 27439
+    },
+    {
+      "epoch": 0.6531467199847663,
+      "grad_norm": 5.350672838322839,
+      "learning_rate": 2e-05,
+      "loss": 1.1453,
+      "step": 27440
+    },
+    {
+      "epoch": 0.6531705227077977,
+      "grad_norm": 10.31690847457353,
+      "learning_rate": 2e-05,
+      "loss": 0.3144,
+      "step": 27441
+    },
+    {
+      "epoch": 0.6531943254308293,
+      "grad_norm": 2.1883559979912253,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 27442
+    },
+    {
+      "epoch": 0.6532181281538608,
+      "grad_norm": 1.3470810630410637,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 27443
+    },
+    {
+      "epoch": 0.6532419308768923,
+      "grad_norm": 2.3838802567615334,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 27444
+    },
+    {
+      "epoch": 0.6532657335999238,
+      "grad_norm": 1.959704247916019,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 27445
+    },
+    {
+      "epoch": 0.6532895363229554,
+      "grad_norm": 2.474565078324214,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 27446
+    },
+    {
+      "epoch": 0.6533133390459869,
+      "grad_norm": 6.245876689185832,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 27447
+    },
+    {
+      "epoch": 0.6533371417690184,
+      "grad_norm": 0.11177054822778039,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 27448
+    },
+    {
+      "epoch": 0.6533609444920498,
+      "grad_norm": 1.627521549368172,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 27449
+    },
+    {
+      "epoch": 0.6533847472150814,
+      "grad_norm": 3.7923309268146195,
+      "learning_rate": 2e-05,
+      "loss": 0.7167,
+      "step": 27450
+    },
+    {
+      "epoch": 0.6534085499381129,
+      "grad_norm": 1.561183408450843,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 27451
+    },
+    {
+      "epoch": 0.6534323526611444,
+      "grad_norm": 9.83231195086856,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 27452
+    },
+    {
+      "epoch": 0.6534561553841759,
+      "grad_norm": 0.4040684097255921,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 27453
+    },
+    {
+      "epoch": 0.6534799581072075,
+      "grad_norm": 2.1546647417368248,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 27454
+    },
+    {
+      "epoch": 0.653503760830239,
+      "grad_norm": 5.283343736888313,
+      "learning_rate": 2e-05,
+      "loss": 0.5732,
+      "step": 27455
+    },
+    {
+      "epoch": 0.6535275635532705,
+      "grad_norm": 10.602369863669406,
+      "learning_rate": 2e-05,
+      "loss": 0.7626,
+      "step": 27456
+    },
+    {
+      "epoch": 0.6535513662763021,
+      "grad_norm": 4.5539211969753035,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 27457
+    },
+    {
+      "epoch": 0.6535751689993335,
+      "grad_norm": 4.4856461387292,
+      "learning_rate": 2e-05,
+      "loss": 0.7531,
+      "step": 27458
+    },
+    {
+      "epoch": 0.653598971722365,
+      "grad_norm": 0.04780953302570752,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 27459
+    },
+    {
+      "epoch": 0.6536227744453965,
+      "grad_norm": 0.5766170353421712,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 27460
+    },
+    {
+      "epoch": 0.6536465771684281,
+      "grad_norm": 5.476378364799793,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 27461
+    },
+    {
+      "epoch": 0.6536703798914596,
+      "grad_norm": 2.6704445401768186,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 27462
+    },
+    {
+      "epoch": 0.6536941826144911,
+      "grad_norm": 2.9494103405754926,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 27463
+    },
+    {
+      "epoch": 0.6537179853375226,
+      "grad_norm": 4.242532269782975,
+      "learning_rate": 2e-05,
+      "loss": 0.3768,
+      "step": 27464
+    },
+    {
+      "epoch": 0.6537417880605542,
+      "grad_norm": 4.552601969551183,
+      "learning_rate": 2e-05,
+      "loss": 0.7395,
+      "step": 27465
+    },
+    {
+      "epoch": 0.6537655907835856,
+      "grad_norm": 0.6527636679361702,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 27466
+    },
+    {
+      "epoch": 0.6537893935066171,
+      "grad_norm": 1.2417615539436582,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 27467
+    },
+    {
+      "epoch": 0.6538131962296486,
+      "grad_norm": 4.871431247909882,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 27468
+    },
+    {
+      "epoch": 0.6538369989526802,
+      "grad_norm": 0.2864171384233133,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 27469
+    },
+    {
+      "epoch": 0.6538608016757117,
+      "grad_norm": 1.5083673604145826,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 27470
+    },
+    {
+      "epoch": 0.6538846043987432,
+      "grad_norm": 5.349212733095334,
+      "learning_rate": 2e-05,
+      "loss": 0.8403,
+      "step": 27471
+    },
+    {
+      "epoch": 0.6539084071217748,
+      "grad_norm": 2.1660597516045446,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 27472
+    },
+    {
+      "epoch": 0.6539322098448063,
+      "grad_norm": 8.321525442034426,
+      "learning_rate": 2e-05,
+      "loss": 1.5121,
+      "step": 27473
+    },
+    {
+      "epoch": 0.6539560125678378,
+      "grad_norm": 2.194606980199417,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 27474
+    },
+    {
+      "epoch": 0.6539798152908692,
+      "grad_norm": 4.960355223666695,
+      "learning_rate": 2e-05,
+      "loss": 0.5722,
+      "step": 27475
+    },
+    {
+      "epoch": 0.6540036180139008,
+      "grad_norm": 0.15467885171127171,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 27476
+    },
+    {
+      "epoch": 0.6540274207369323,
+      "grad_norm": 3.770501379593487,
+      "learning_rate": 2e-05,
+      "loss": 0.6251,
+      "step": 27477
+    },
+    {
+      "epoch": 0.6540512234599638,
+      "grad_norm": 0.2803364937536346,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 27478
+    },
+    {
+      "epoch": 0.6540750261829953,
+      "grad_norm": 11.110198776778498,
+      "learning_rate": 2e-05,
+      "loss": 0.6634,
+      "step": 27479
+    },
+    {
+      "epoch": 0.6540988289060269,
+      "grad_norm": 3.835867454663003,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 27480
+    },
+    {
+      "epoch": 0.6541226316290584,
+      "grad_norm": 5.807326733432471,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 27481
+    },
+    {
+      "epoch": 0.6541464343520899,
+      "grad_norm": 5.907747799979188,
+      "learning_rate": 2e-05,
+      "loss": 0.739,
+      "step": 27482
+    },
+    {
+      "epoch": 0.6541702370751213,
+      "grad_norm": 2.9552824222653604,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 27483
+    },
+    {
+      "epoch": 0.6541940397981529,
+      "grad_norm": 4.860985117647164,
+      "learning_rate": 2e-05,
+      "loss": 0.7634,
+      "step": 27484
+    },
+    {
+      "epoch": 0.6542178425211844,
+      "grad_norm": 1.7309292116524282,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 27485
+    },
+    {
+      "epoch": 0.6542416452442159,
+      "grad_norm": 3.525413869505859,
+      "learning_rate": 2e-05,
+      "loss": 0.2935,
+      "step": 27486
+    },
+    {
+      "epoch": 0.6542654479672475,
+      "grad_norm": 3.04625556628553,
+      "learning_rate": 2e-05,
+      "loss": 0.3299,
+      "step": 27487
+    },
+    {
+      "epoch": 0.654289250690279,
+      "grad_norm": 1.7961686301905329,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 27488
+    },
+    {
+      "epoch": 0.6543130534133105,
+      "grad_norm": 5.922682669350876,
+      "learning_rate": 2e-05,
+      "loss": 1.2277,
+      "step": 27489
+    },
+    {
+      "epoch": 0.654336856136342,
+      "grad_norm": 0.07280988716213761,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 27490
+    },
+    {
+      "epoch": 0.6543606588593736,
+      "grad_norm": 8.703770585894684,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 27491
+    },
+    {
+      "epoch": 0.654384461582405,
+      "grad_norm": 2.6115533419164563,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 27492
+    },
+    {
+      "epoch": 0.6544082643054365,
+      "grad_norm": 1.4271391408017657,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 27493
+    },
+    {
+      "epoch": 0.654432067028468,
+      "grad_norm": 4.836810596941107,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 27494
+    },
+    {
+      "epoch": 0.6544558697514996,
+      "grad_norm": 0.39285755114935905,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 27495
+    },
+    {
+      "epoch": 0.6544796724745311,
+      "grad_norm": 13.94980977701032,
+      "learning_rate": 2e-05,
+      "loss": 1.1817,
+      "step": 27496
+    },
+    {
+      "epoch": 0.6545034751975626,
+      "grad_norm": 6.714042046749553,
+      "learning_rate": 2e-05,
+      "loss": 1.0229,
+      "step": 27497
+    },
+    {
+      "epoch": 0.6545272779205941,
+      "grad_norm": 2.273679034343552,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 27498
+    },
+    {
+      "epoch": 0.6545510806436257,
+      "grad_norm": 0.10220460584099055,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 27499
+    },
+    {
+      "epoch": 0.6545748833666571,
+      "grad_norm": 1.7878226161649418,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 27500
+    },
+    {
+      "epoch": 0.6545986860896886,
+      "grad_norm": 7.2309244548330875,
+      "learning_rate": 2e-05,
+      "loss": 0.1574,
+      "step": 27501
+    },
+    {
+      "epoch": 0.6546224888127202,
+      "grad_norm": 6.451639121208145,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 27502
+    },
+    {
+      "epoch": 0.6546462915357517,
+      "grad_norm": 16.62634842642342,
+      "learning_rate": 2e-05,
+      "loss": 0.9233,
+      "step": 27503
+    },
+    {
+      "epoch": 0.6546700942587832,
+      "grad_norm": 5.437632241376955,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 27504
+    },
+    {
+      "epoch": 0.6546938969818147,
+      "grad_norm": 2.0362592313439163,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 27505
+    },
+    {
+      "epoch": 0.6547176997048463,
+      "grad_norm": 5.06269031080406,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 27506
+    },
+    {
+      "epoch": 0.6547415024278778,
+      "grad_norm": 4.1816666368192,
+      "learning_rate": 2e-05,
+      "loss": 0.7745,
+      "step": 27507
+    },
+    {
+      "epoch": 0.6547653051509092,
+      "grad_norm": 2.8316739442067043,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 27508
+    },
+    {
+      "epoch": 0.6547891078739407,
+      "grad_norm": 2.6655501029247293,
+      "learning_rate": 2e-05,
+      "loss": 0.289,
+      "step": 27509
+    },
+    {
+      "epoch": 0.6548129105969723,
+      "grad_norm": 2.7868906544467773,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 27510
+    },
+    {
+      "epoch": 0.6548367133200038,
+      "grad_norm": 2.2187805778260703,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 27511
+    },
+    {
+      "epoch": 0.6548605160430353,
+      "grad_norm": 2.703761554478491,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 27512
+    },
+    {
+      "epoch": 0.6548843187660668,
+      "grad_norm": 3.229643435277584,
+      "learning_rate": 2e-05,
+      "loss": 0.4486,
+      "step": 27513
+    },
+    {
+      "epoch": 0.6549081214890984,
+      "grad_norm": 8.710802913142428,
+      "learning_rate": 2e-05,
+      "loss": 0.3043,
+      "step": 27514
+    },
+    {
+      "epoch": 0.6549319242121299,
+      "grad_norm": 2.9741266414887795,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 27515
+    },
+    {
+      "epoch": 0.6549557269351614,
+      "grad_norm": 4.996770215579647,
+      "learning_rate": 2e-05,
+      "loss": 0.6141,
+      "step": 27516
+    },
+    {
+      "epoch": 0.654979529658193,
+      "grad_norm": 1.6061491443027711,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 27517
+    },
+    {
+      "epoch": 0.6550033323812244,
+      "grad_norm": 0.8926944321777119,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 27518
+    },
+    {
+      "epoch": 0.6550271351042559,
+      "grad_norm": 6.46694237065531,
+      "learning_rate": 2e-05,
+      "loss": 1.771,
+      "step": 27519
+    },
+    {
+      "epoch": 0.6550509378272874,
+      "grad_norm": 8.713878246512762,
+      "learning_rate": 2e-05,
+      "loss": 0.5443,
+      "step": 27520
+    },
+    {
+      "epoch": 0.655074740550319,
+      "grad_norm": 8.485873382380763,
+      "learning_rate": 2e-05,
+      "loss": 1.671,
+      "step": 27521
+    },
+    {
+      "epoch": 0.6550985432733505,
+      "grad_norm": 7.5700105908482955,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 27522
+    },
+    {
+      "epoch": 0.655122345996382,
+      "grad_norm": 0.3983962407661411,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 27523
+    },
+    {
+      "epoch": 0.6551461487194135,
+      "grad_norm": 5.833746535545382,
+      "learning_rate": 2e-05,
+      "loss": 1.0843,
+      "step": 27524
+    },
+    {
+      "epoch": 0.655169951442445,
+      "grad_norm": 0.11346011572609958,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27525
+    },
+    {
+      "epoch": 0.6551937541654765,
+      "grad_norm": 2.0300009127767913,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 27526
+    },
+    {
+      "epoch": 0.655217556888508,
+      "grad_norm": 0.6110417190158736,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 27527
+    },
+    {
+      "epoch": 0.6552413596115395,
+      "grad_norm": 3.234223419712729,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 27528
+    },
+    {
+      "epoch": 0.6552651623345711,
+      "grad_norm": 3.611150684304536,
+      "learning_rate": 2e-05,
+      "loss": 0.3994,
+      "step": 27529
+    },
+    {
+      "epoch": 0.6552889650576026,
+      "grad_norm": 4.871654564428861,
+      "learning_rate": 2e-05,
+      "loss": 0.6341,
+      "step": 27530
+    },
+    {
+      "epoch": 0.6553127677806341,
+      "grad_norm": 5.978045568419295,
+      "learning_rate": 2e-05,
+      "loss": 0.4756,
+      "step": 27531
+    },
+    {
+      "epoch": 0.6553365705036657,
+      "grad_norm": 2.456069592506029,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 27532
+    },
+    {
+      "epoch": 0.6553603732266972,
+      "grad_norm": 3.3155252608480446,
+      "learning_rate": 2e-05,
+      "loss": 0.1787,
+      "step": 27533
+    },
+    {
+      "epoch": 0.6553841759497286,
+      "grad_norm": 5.995180611641101,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 27534
+    },
+    {
+      "epoch": 0.6554079786727601,
+      "grad_norm": 10.012553051150267,
+      "learning_rate": 2e-05,
+      "loss": 0.8034,
+      "step": 27535
+    },
+    {
+      "epoch": 0.6554317813957917,
+      "grad_norm": 0.037113906094739066,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 27536
+    },
+    {
+      "epoch": 0.6554555841188232,
+      "grad_norm": 0.6955708410584236,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 27537
+    },
+    {
+      "epoch": 0.6554793868418547,
+      "grad_norm": 5.009517234416266,
+      "learning_rate": 2e-05,
+      "loss": 0.8061,
+      "step": 27538
+    },
+    {
+      "epoch": 0.6555031895648862,
+      "grad_norm": 3.5793450929864523,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 27539
+    },
+    {
+      "epoch": 0.6555269922879178,
+      "grad_norm": 4.027069341881067,
+      "learning_rate": 2e-05,
+      "loss": 0.5913,
+      "step": 27540
+    },
+    {
+      "epoch": 0.6555507950109493,
+      "grad_norm": 0.6478128650700052,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 27541
+    },
+    {
+      "epoch": 0.6555745977339807,
+      "grad_norm": 1.4425248175778598,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 27542
+    },
+    {
+      "epoch": 0.6555984004570122,
+      "grad_norm": 0.4916976158459222,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 27543
+    },
+    {
+      "epoch": 0.6556222031800438,
+      "grad_norm": 4.977236952496483,
+      "learning_rate": 2e-05,
+      "loss": 0.171,
+      "step": 27544
+    },
+    {
+      "epoch": 0.6556460059030753,
+      "grad_norm": 10.286400606209902,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 27545
+    },
+    {
+      "epoch": 0.6556698086261068,
+      "grad_norm": 0.22756029094057098,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 27546
+    },
+    {
+      "epoch": 0.6556936113491384,
+      "grad_norm": 6.759102410876599,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 27547
+    },
+    {
+      "epoch": 0.6557174140721699,
+      "grad_norm": 3.9238947551066996,
+      "learning_rate": 2e-05,
+      "loss": 0.3845,
+      "step": 27548
+    },
+    {
+      "epoch": 0.6557412167952014,
+      "grad_norm": 2.9168450498081597,
+      "learning_rate": 2e-05,
+      "loss": 0.6031,
+      "step": 27549
+    },
+    {
+      "epoch": 0.6557650195182329,
+      "grad_norm": 1.0730416211458338,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 27550
+    },
+    {
+      "epoch": 0.6557888222412644,
+      "grad_norm": 3.4880751414619087,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 27551
+    },
+    {
+      "epoch": 0.6558126249642959,
+      "grad_norm": 1.6455623580298124,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 27552
+    },
+    {
+      "epoch": 0.6558364276873274,
+      "grad_norm": 9.809983735079879,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 27553
+    },
+    {
+      "epoch": 0.6558602304103589,
+      "grad_norm": 3.1281519500612025,
+      "learning_rate": 2e-05,
+      "loss": 0.403,
+      "step": 27554
+    },
+    {
+      "epoch": 0.6558840331333905,
+      "grad_norm": 3.854319178600332,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 27555
+    },
+    {
+      "epoch": 0.655907835856422,
+      "grad_norm": 5.172881088187883,
+      "learning_rate": 2e-05,
+      "loss": 0.7298,
+      "step": 27556
+    },
+    {
+      "epoch": 0.6559316385794535,
+      "grad_norm": 7.257711559952875,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 27557
+    },
+    {
+      "epoch": 0.655955441302485,
+      "grad_norm": 1.121216625889505,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 27558
+    },
+    {
+      "epoch": 0.6559792440255166,
+      "grad_norm": 1.7198689187831964,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 27559
+    },
+    {
+      "epoch": 0.656003046748548,
+      "grad_norm": 0.36717525905365805,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 27560
+    },
+    {
+      "epoch": 0.6560268494715795,
+      "grad_norm": 0.5735643984630205,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 27561
+    },
+    {
+      "epoch": 0.6560506521946111,
+      "grad_norm": 2.924628176490005,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 27562
+    },
+    {
+      "epoch": 0.6560744549176426,
+      "grad_norm": 4.758627620308323,
+      "learning_rate": 2e-05,
+      "loss": 0.5419,
+      "step": 27563
+    },
+    {
+      "epoch": 0.6560982576406741,
+      "grad_norm": 0.8259705308684082,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 27564
+    },
+    {
+      "epoch": 0.6561220603637056,
+      "grad_norm": 7.184425150639153,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 27565
+    },
+    {
+      "epoch": 0.6561458630867372,
+      "grad_norm": 5.139670494823777,
+      "learning_rate": 2e-05,
+      "loss": 0.2337,
+      "step": 27566
+    },
+    {
+      "epoch": 0.6561696658097687,
+      "grad_norm": 10.326862264962427,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 27567
+    },
+    {
+      "epoch": 0.6561934685328001,
+      "grad_norm": 4.587830400919788,
+      "learning_rate": 2e-05,
+      "loss": 0.1705,
+      "step": 27568
+    },
+    {
+      "epoch": 0.6562172712558316,
+      "grad_norm": 4.215124130900124,
+      "learning_rate": 2e-05,
+      "loss": 0.5207,
+      "step": 27569
+    },
+    {
+      "epoch": 0.6562410739788632,
+      "grad_norm": 3.1032344010749746,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 27570
+    },
+    {
+      "epoch": 0.6562648767018947,
+      "grad_norm": 1.7287799318874277,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 27571
+    },
+    {
+      "epoch": 0.6562886794249262,
+      "grad_norm": 2.459129575865222,
+      "learning_rate": 2e-05,
+      "loss": 0.1713,
+      "step": 27572
+    },
+    {
+      "epoch": 0.6563124821479577,
+      "grad_norm": 0.9814817581384849,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 27573
+    },
+    {
+      "epoch": 0.6563362848709893,
+      "grad_norm": 3.3183043463373734,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 27574
+    },
+    {
+      "epoch": 0.6563600875940208,
+      "grad_norm": 0.08021370604334356,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27575
+    },
+    {
+      "epoch": 0.6563838903170522,
+      "grad_norm": 18.87113013290228,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 27576
+    },
+    {
+      "epoch": 0.6564076930400838,
+      "grad_norm": 5.1063376995842,
+      "learning_rate": 2e-05,
+      "loss": 0.3981,
+      "step": 27577
+    },
+    {
+      "epoch": 0.6564314957631153,
+      "grad_norm": 4.0088302883142015,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 27578
+    },
+    {
+      "epoch": 0.6564552984861468,
+      "grad_norm": 1.1345410591872014,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 27579
+    },
+    {
+      "epoch": 0.6564791012091783,
+      "grad_norm": 0.12370261906941624,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 27580
+    },
+    {
+      "epoch": 0.6565029039322099,
+      "grad_norm": 3.5327177106018284,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 27581
+    },
+    {
+      "epoch": 0.6565267066552414,
+      "grad_norm": 0.10747690745337866,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 27582
+    },
+    {
+      "epoch": 0.6565505093782729,
+      "grad_norm": 9.201473192871749,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 27583
+    },
+    {
+      "epoch": 0.6565743121013043,
+      "grad_norm": 2.9123369104979404,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 27584
+    },
+    {
+      "epoch": 0.6565981148243359,
+      "grad_norm": 0.6772635198838555,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 27585
+    },
+    {
+      "epoch": 0.6566219175473674,
+      "grad_norm": 0.9839162173032034,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 27586
+    },
+    {
+      "epoch": 0.6566457202703989,
+      "grad_norm": 0.19537745490303474,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 27587
+    },
+    {
+      "epoch": 0.6566695229934304,
+      "grad_norm": 1.3272109045637828,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 27588
+    },
+    {
+      "epoch": 0.656693325716462,
+      "grad_norm": 1.5906008326705299,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 27589
+    },
+    {
+      "epoch": 0.6567171284394935,
+      "grad_norm": 1.5582265205383943,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 27590
+    },
+    {
+      "epoch": 0.656740931162525,
+      "grad_norm": 2.1991783158052294,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 27591
+    },
+    {
+      "epoch": 0.6567647338855566,
+      "grad_norm": 2.685309464287122,
+      "learning_rate": 2e-05,
+      "loss": 0.4503,
+      "step": 27592
+    },
+    {
+      "epoch": 0.656788536608588,
+      "grad_norm": 4.459326592479273,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 27593
+    },
+    {
+      "epoch": 0.6568123393316195,
+      "grad_norm": 5.5793433353816555,
+      "learning_rate": 2e-05,
+      "loss": 0.7385,
+      "step": 27594
+    },
+    {
+      "epoch": 0.656836142054651,
+      "grad_norm": 3.0241259103491913,
+      "learning_rate": 2e-05,
+      "loss": 0.2943,
+      "step": 27595
+    },
+    {
+      "epoch": 0.6568599447776826,
+      "grad_norm": 3.3845671860452944,
+      "learning_rate": 2e-05,
+      "loss": 0.3396,
+      "step": 27596
+    },
+    {
+      "epoch": 0.6568837475007141,
+      "grad_norm": 0.682850035641712,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 27597
+    },
+    {
+      "epoch": 0.6569075502237456,
+      "grad_norm": 0.08939612678909303,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 27598
+    },
+    {
+      "epoch": 0.6569313529467771,
+      "grad_norm": 3.794361017477657,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 27599
+    },
+    {
+      "epoch": 0.6569551556698087,
+      "grad_norm": 0.04691439312710744,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 27600
+    },
+    {
+      "epoch": 0.6569789583928402,
+      "grad_norm": 29.835930089401643,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 27601
+    },
+    {
+      "epoch": 0.6570027611158716,
+      "grad_norm": 3.868929124028142,
+      "learning_rate": 2e-05,
+      "loss": 0.4854,
+      "step": 27602
+    },
+    {
+      "epoch": 0.6570265638389031,
+      "grad_norm": 2.0139844381893837,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 27603
+    },
+    {
+      "epoch": 0.6570503665619347,
+      "grad_norm": 1.045286991343366,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 27604
+    },
+    {
+      "epoch": 0.6570741692849662,
+      "grad_norm": 0.046251804099127256,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 27605
+    },
+    {
+      "epoch": 0.6570979720079977,
+      "grad_norm": 0.6959574575805016,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 27606
+    },
+    {
+      "epoch": 0.6571217747310293,
+      "grad_norm": 0.23983809042072085,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 27607
+    },
+    {
+      "epoch": 0.6571455774540608,
+      "grad_norm": 0.9254526682043228,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 27608
+    },
+    {
+      "epoch": 0.6571693801770923,
+      "grad_norm": 0.7486093243495443,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 27609
+    },
+    {
+      "epoch": 0.6571931829001237,
+      "grad_norm": 0.6525960960433388,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 27610
+    },
+    {
+      "epoch": 0.6572169856231553,
+      "grad_norm": 2.791243839890558,
+      "learning_rate": 2e-05,
+      "loss": 0.4921,
+      "step": 27611
+    },
+    {
+      "epoch": 0.6572407883461868,
+      "grad_norm": 0.3639343735153569,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 27612
+    },
+    {
+      "epoch": 0.6572645910692183,
+      "grad_norm": 9.23157183176158,
+      "learning_rate": 2e-05,
+      "loss": 1.0708,
+      "step": 27613
+    },
+    {
+      "epoch": 0.6572883937922498,
+      "grad_norm": 2.5372896803798994,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 27614
+    },
+    {
+      "epoch": 0.6573121965152814,
+      "grad_norm": 0.7218773348849511,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 27615
+    },
+    {
+      "epoch": 0.6573359992383129,
+      "grad_norm": 3.1714806615959215,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 27616
+    },
+    {
+      "epoch": 0.6573598019613444,
+      "grad_norm": 0.23221688966674178,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 27617
+    },
+    {
+      "epoch": 0.6573836046843758,
+      "grad_norm": 5.071022562822755,
+      "learning_rate": 2e-05,
+      "loss": 0.5251,
+      "step": 27618
+    },
+    {
+      "epoch": 0.6574074074074074,
+      "grad_norm": 4.6653142290716945,
+      "learning_rate": 2e-05,
+      "loss": 0.8609,
+      "step": 27619
+    },
+    {
+      "epoch": 0.6574312101304389,
+      "grad_norm": 3.041394803943192,
+      "learning_rate": 2e-05,
+      "loss": 0.4184,
+      "step": 27620
+    },
+    {
+      "epoch": 0.6574550128534704,
+      "grad_norm": 29.111711246658803,
+      "learning_rate": 2e-05,
+      "loss": 0.5822,
+      "step": 27621
+    },
+    {
+      "epoch": 0.657478815576502,
+      "grad_norm": 3.8046387993008555,
+      "learning_rate": 2e-05,
+      "loss": 0.727,
+      "step": 27622
+    },
+    {
+      "epoch": 0.6575026182995335,
+      "grad_norm": 1.877399014593531,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 27623
+    },
+    {
+      "epoch": 0.657526421022565,
+      "grad_norm": 4.754781062857859,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 27624
+    },
+    {
+      "epoch": 0.6575502237455965,
+      "grad_norm": 3.2226785112779575,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 27625
+    },
+    {
+      "epoch": 0.6575740264686281,
+      "grad_norm": 11.6598966880849,
+      "learning_rate": 2e-05,
+      "loss": 0.3857,
+      "step": 27626
+    },
+    {
+      "epoch": 0.6575978291916595,
+      "grad_norm": 8.96423595572194,
+      "learning_rate": 2e-05,
+      "loss": 0.4643,
+      "step": 27627
+    },
+    {
+      "epoch": 0.657621631914691,
+      "grad_norm": 0.8774938700424038,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 27628
+    },
+    {
+      "epoch": 0.6576454346377225,
+      "grad_norm": 5.551964614543091,
+      "learning_rate": 2e-05,
+      "loss": 0.4839,
+      "step": 27629
+    },
+    {
+      "epoch": 0.6576692373607541,
+      "grad_norm": 2.2258012226996944,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 27630
+    },
+    {
+      "epoch": 0.6576930400837856,
+      "grad_norm": 3.421352087199763,
+      "learning_rate": 2e-05,
+      "loss": 0.6761,
+      "step": 27631
+    },
+    {
+      "epoch": 0.6577168428068171,
+      "grad_norm": 0.050200325483517404,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 27632
+    },
+    {
+      "epoch": 0.6577406455298486,
+      "grad_norm": 6.175946240348069,
+      "learning_rate": 2e-05,
+      "loss": 0.8961,
+      "step": 27633
+    },
+    {
+      "epoch": 0.6577644482528802,
+      "grad_norm": 4.912824457911991,
+      "learning_rate": 2e-05,
+      "loss": 0.8237,
+      "step": 27634
+    },
+    {
+      "epoch": 0.6577882509759116,
+      "grad_norm": 2.214750863885467,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 27635
+    },
+    {
+      "epoch": 0.6578120536989431,
+      "grad_norm": 11.175131187729,
+      "learning_rate": 2e-05,
+      "loss": 0.9531,
+      "step": 27636
+    },
+    {
+      "epoch": 0.6578358564219746,
+      "grad_norm": 6.094625714562407,
+      "learning_rate": 2e-05,
+      "loss": 0.8695,
+      "step": 27637
+    },
+    {
+      "epoch": 0.6578596591450062,
+      "grad_norm": 0.3197459954397288,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 27638
+    },
+    {
+      "epoch": 0.6578834618680377,
+      "grad_norm": 10.130276387416588,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 27639
+    },
+    {
+      "epoch": 0.6579072645910692,
+      "grad_norm": 1.4443000378822046,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 27640
+    },
+    {
+      "epoch": 0.6579310673141008,
+      "grad_norm": 2.395351629142685,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 27641
+    },
+    {
+      "epoch": 0.6579548700371323,
+      "grad_norm": 52.78755886300166,
+      "learning_rate": 2e-05,
+      "loss": 2.7198,
+      "step": 27642
+    },
+    {
+      "epoch": 0.6579786727601638,
+      "grad_norm": 6.199904176195358,
+      "learning_rate": 2e-05,
+      "loss": 0.9329,
+      "step": 27643
+    },
+    {
+      "epoch": 0.6580024754831952,
+      "grad_norm": 7.8623384654722965,
+      "learning_rate": 2e-05,
+      "loss": 0.3573,
+      "step": 27644
+    },
+    {
+      "epoch": 0.6580262782062268,
+      "grad_norm": 4.551403695743111,
+      "learning_rate": 2e-05,
+      "loss": 0.9135,
+      "step": 27645
+    },
+    {
+      "epoch": 0.6580500809292583,
+      "grad_norm": 0.11721119494022764,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 27646
+    },
+    {
+      "epoch": 0.6580738836522898,
+      "grad_norm": 2.684610103832439,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 27647
+    },
+    {
+      "epoch": 0.6580976863753213,
+      "grad_norm": 7.316325282479723,
+      "learning_rate": 2e-05,
+      "loss": 1.0171,
+      "step": 27648
+    },
+    {
+      "epoch": 0.6581214890983529,
+      "grad_norm": 3.69862036518618,
+      "learning_rate": 2e-05,
+      "loss": 0.3443,
+      "step": 27649
+    },
+    {
+      "epoch": 0.6581452918213844,
+      "grad_norm": 3.279986481744227,
+      "learning_rate": 2e-05,
+      "loss": 0.0613,
+      "step": 27650
+    },
+    {
+      "epoch": 0.6581690945444159,
+      "grad_norm": 1.7780219490754818,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 27651
+    },
+    {
+      "epoch": 0.6581928972674473,
+      "grad_norm": 0.8445049669454583,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 27652
+    },
+    {
+      "epoch": 0.6582166999904789,
+      "grad_norm": 3.8223171097524435,
+      "learning_rate": 2e-05,
+      "loss": 0.4529,
+      "step": 27653
+    },
+    {
+      "epoch": 0.6582405027135104,
+      "grad_norm": 5.450427247762608,
+      "learning_rate": 2e-05,
+      "loss": 0.6168,
+      "step": 27654
+    },
+    {
+      "epoch": 0.6582643054365419,
+      "grad_norm": 4.736041243580323,
+      "learning_rate": 2e-05,
+      "loss": 1.1379,
+      "step": 27655
+    },
+    {
+      "epoch": 0.6582881081595735,
+      "grad_norm": 3.5689221458725444,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 27656
+    },
+    {
+      "epoch": 0.658311910882605,
+      "grad_norm": 0.1967504969704142,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 27657
+    },
+    {
+      "epoch": 0.6583357136056365,
+      "grad_norm": 3.7754430124488043,
+      "learning_rate": 2e-05,
+      "loss": 0.6641,
+      "step": 27658
+    },
+    {
+      "epoch": 0.658359516328668,
+      "grad_norm": 4.824792135162507,
+      "learning_rate": 2e-05,
+      "loss": 0.9814,
+      "step": 27659
+    },
+    {
+      "epoch": 0.6583833190516996,
+      "grad_norm": 0.7808328345125445,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 27660
+    },
+    {
+      "epoch": 0.658407121774731,
+      "grad_norm": 7.497298476183909,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 27661
+    },
+    {
+      "epoch": 0.6584309244977625,
+      "grad_norm": 19.544730253583207,
+      "learning_rate": 2e-05,
+      "loss": 0.377,
+      "step": 27662
+    },
+    {
+      "epoch": 0.658454727220794,
+      "grad_norm": 5.5986336219049715,
+      "learning_rate": 2e-05,
+      "loss": 0.6283,
+      "step": 27663
+    },
+    {
+      "epoch": 0.6584785299438256,
+      "grad_norm": 9.352332404015065,
+      "learning_rate": 2e-05,
+      "loss": 1.3584,
+      "step": 27664
+    },
+    {
+      "epoch": 0.6585023326668571,
+      "grad_norm": 5.721694874000002,
+      "learning_rate": 2e-05,
+      "loss": 0.1986,
+      "step": 27665
+    },
+    {
+      "epoch": 0.6585261353898886,
+      "grad_norm": 2.0305463346153183,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 27666
+    },
+    {
+      "epoch": 0.6585499381129201,
+      "grad_norm": 3.4734759936302506,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 27667
+    },
+    {
+      "epoch": 0.6585737408359517,
+      "grad_norm": 0.14486148934359544,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 27668
+    },
+    {
+      "epoch": 0.6585975435589831,
+      "grad_norm": 1.863087874950124,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 27669
+    },
+    {
+      "epoch": 0.6586213462820146,
+      "grad_norm": 2.603263983496189,
+      "learning_rate": 2e-05,
+      "loss": 0.3781,
+      "step": 27670
+    },
+    {
+      "epoch": 0.6586451490050462,
+      "grad_norm": 2.578358391856092,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 27671
+    },
+    {
+      "epoch": 0.6586689517280777,
+      "grad_norm": 5.324096922947859,
+      "learning_rate": 2e-05,
+      "loss": 0.8032,
+      "step": 27672
+    },
+    {
+      "epoch": 0.6586927544511092,
+      "grad_norm": 3.1916502997635274,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 27673
+    },
+    {
+      "epoch": 0.6587165571741407,
+      "grad_norm": 4.657947491763029,
+      "learning_rate": 2e-05,
+      "loss": 0.9341,
+      "step": 27674
+    },
+    {
+      "epoch": 0.6587403598971723,
+      "grad_norm": 23.629345548627455,
+      "learning_rate": 2e-05,
+      "loss": 0.4438,
+      "step": 27675
+    },
+    {
+      "epoch": 0.6587641626202038,
+      "grad_norm": 14.551569761747212,
+      "learning_rate": 2e-05,
+      "loss": 1.2941,
+      "step": 27676
+    },
+    {
+      "epoch": 0.6587879653432352,
+      "grad_norm": 3.7559757208679168,
+      "learning_rate": 2e-05,
+      "loss": 0.5295,
+      "step": 27677
+    },
+    {
+      "epoch": 0.6588117680662667,
+      "grad_norm": 2.3818517365301144,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 27678
+    },
+    {
+      "epoch": 0.6588355707892983,
+      "grad_norm": 11.085900292382076,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 27679
+    },
+    {
+      "epoch": 0.6588593735123298,
+      "grad_norm": 3.747395409367522,
+      "learning_rate": 2e-05,
+      "loss": 0.6545,
+      "step": 27680
+    },
+    {
+      "epoch": 0.6588831762353613,
+      "grad_norm": 3.806206282859785,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 27681
+    },
+    {
+      "epoch": 0.6589069789583928,
+      "grad_norm": 0.3393948669796162,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 27682
+    },
+    {
+      "epoch": 0.6589307816814244,
+      "grad_norm": 4.69747968480206,
+      "learning_rate": 2e-05,
+      "loss": 0.9253,
+      "step": 27683
+    },
+    {
+      "epoch": 0.6589545844044559,
+      "grad_norm": 0.08982539057901517,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 27684
+    },
+    {
+      "epoch": 0.6589783871274874,
+      "grad_norm": 3.7262427632049113,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 27685
+    },
+    {
+      "epoch": 0.659002189850519,
+      "grad_norm": 9.965301317264135,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 27686
+    },
+    {
+      "epoch": 0.6590259925735504,
+      "grad_norm": 3.022562521304001,
+      "learning_rate": 2e-05,
+      "loss": 0.0983,
+      "step": 27687
+    },
+    {
+      "epoch": 0.6590497952965819,
+      "grad_norm": 0.6505947218105779,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 27688
+    },
+    {
+      "epoch": 0.6590735980196134,
+      "grad_norm": 4.651062491914582,
+      "learning_rate": 2e-05,
+      "loss": 1.0815,
+      "step": 27689
+    },
+    {
+      "epoch": 0.659097400742645,
+      "grad_norm": 2.2794807565902975,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 27690
+    },
+    {
+      "epoch": 0.6591212034656765,
+      "grad_norm": 6.803311080381933,
+      "learning_rate": 2e-05,
+      "loss": 1.186,
+      "step": 27691
+    },
+    {
+      "epoch": 0.659145006188708,
+      "grad_norm": 0.14762938296956032,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 27692
+    },
+    {
+      "epoch": 0.6591688089117395,
+      "grad_norm": 0.3070443470817588,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 27693
+    },
+    {
+      "epoch": 0.659192611634771,
+      "grad_norm": 4.970203175273325,
+      "learning_rate": 2e-05,
+      "loss": 0.5924,
+      "step": 27694
+    },
+    {
+      "epoch": 0.6592164143578025,
+      "grad_norm": 2.411571270509563,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 27695
+    },
+    {
+      "epoch": 0.659240217080834,
+      "grad_norm": 0.21016580863157308,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 27696
+    },
+    {
+      "epoch": 0.6592640198038655,
+      "grad_norm": 7.37191523139193,
+      "learning_rate": 2e-05,
+      "loss": 0.3676,
+      "step": 27697
+    },
+    {
+      "epoch": 0.6592878225268971,
+      "grad_norm": 1.6724837555608212,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 27698
+    },
+    {
+      "epoch": 0.6593116252499286,
+      "grad_norm": 2.8280062878482903,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 27699
+    },
+    {
+      "epoch": 0.6593354279729601,
+      "grad_norm": 0.5832393968874968,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 27700
+    },
+    {
+      "epoch": 0.6593592306959917,
+      "grad_norm": 0.5547716724009724,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 27701
+    },
+    {
+      "epoch": 0.6593830334190232,
+      "grad_norm": 1.8596753687434455,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 27702
+    },
+    {
+      "epoch": 0.6594068361420546,
+      "grad_norm": 2.703954731157348,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 27703
+    },
+    {
+      "epoch": 0.6594306388650861,
+      "grad_norm": 3.4898830361262854,
+      "learning_rate": 2e-05,
+      "loss": 0.4044,
+      "step": 27704
+    },
+    {
+      "epoch": 0.6594544415881177,
+      "grad_norm": 1.1165238810580598,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 27705
+    },
+    {
+      "epoch": 0.6594782443111492,
+      "grad_norm": 2.0098866403953046,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 27706
+    },
+    {
+      "epoch": 0.6595020470341807,
+      "grad_norm": 0.5835303148063413,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 27707
+    },
+    {
+      "epoch": 0.6595258497572122,
+      "grad_norm": 3.909495683134475,
+      "learning_rate": 2e-05,
+      "loss": 0.8984,
+      "step": 27708
+    },
+    {
+      "epoch": 0.6595496524802438,
+      "grad_norm": 6.295515447790989,
+      "learning_rate": 2e-05,
+      "loss": 1.2685,
+      "step": 27709
+    },
+    {
+      "epoch": 0.6595734552032753,
+      "grad_norm": 0.5534774729819759,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 27710
+    },
+    {
+      "epoch": 0.6595972579263067,
+      "grad_norm": 5.442193313095669,
+      "learning_rate": 2e-05,
+      "loss": 1.5576,
+      "step": 27711
+    },
+    {
+      "epoch": 0.6596210606493382,
+      "grad_norm": 11.233276306354668,
+      "learning_rate": 2e-05,
+      "loss": 0.9262,
+      "step": 27712
+    },
+    {
+      "epoch": 0.6596448633723698,
+      "grad_norm": 6.671556186210734,
+      "learning_rate": 2e-05,
+      "loss": 1.6123,
+      "step": 27713
+    },
+    {
+      "epoch": 0.6596686660954013,
+      "grad_norm": 9.39845988495977,
+      "learning_rate": 2e-05,
+      "loss": 0.2946,
+      "step": 27714
+    },
+    {
+      "epoch": 0.6596924688184328,
+      "grad_norm": 1.346993936601718,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 27715
+    },
+    {
+      "epoch": 0.6597162715414644,
+      "grad_norm": 2.3530221722822096,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 27716
+    },
+    {
+      "epoch": 0.6597400742644959,
+      "grad_norm": 4.0966706117433525,
+      "learning_rate": 2e-05,
+      "loss": 0.3106,
+      "step": 27717
+    },
+    {
+      "epoch": 0.6597638769875274,
+      "grad_norm": 7.494852003071234,
+      "learning_rate": 2e-05,
+      "loss": 0.3815,
+      "step": 27718
+    },
+    {
+      "epoch": 0.6597876797105588,
+      "grad_norm": 1.7470704745105436,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 27719
+    },
+    {
+      "epoch": 0.6598114824335904,
+      "grad_norm": 4.268796970803146,
+      "learning_rate": 2e-05,
+      "loss": 0.7164,
+      "step": 27720
+    },
+    {
+      "epoch": 0.6598352851566219,
+      "grad_norm": 6.226700855465147,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 27721
+    },
+    {
+      "epoch": 0.6598590878796534,
+      "grad_norm": 2.416166558688868,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 27722
+    },
+    {
+      "epoch": 0.6598828906026849,
+      "grad_norm": 5.040694107643927,
+      "learning_rate": 2e-05,
+      "loss": 1.2759,
+      "step": 27723
+    },
+    {
+      "epoch": 0.6599066933257165,
+      "grad_norm": 31.611752730614956,
+      "learning_rate": 2e-05,
+      "loss": 0.9045,
+      "step": 27724
+    },
+    {
+      "epoch": 0.659930496048748,
+      "grad_norm": 5.091572986848162,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 27725
+    },
+    {
+      "epoch": 0.6599542987717795,
+      "grad_norm": 3.124607513752459,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 27726
+    },
+    {
+      "epoch": 0.659978101494811,
+      "grad_norm": 8.68852467956068,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 27727
+    },
+    {
+      "epoch": 0.6600019042178425,
+      "grad_norm": 4.458062434812622,
+      "learning_rate": 2e-05,
+      "loss": 0.5836,
+      "step": 27728
+    },
+    {
+      "epoch": 0.660025706940874,
+      "grad_norm": 5.875819135279031,
+      "learning_rate": 2e-05,
+      "loss": 0.7221,
+      "step": 27729
+    },
+    {
+      "epoch": 0.6600495096639055,
+      "grad_norm": 1.883382117013838,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 27730
+    },
+    {
+      "epoch": 0.6600733123869371,
+      "grad_norm": 3.580730112147396,
+      "learning_rate": 2e-05,
+      "loss": 0.1821,
+      "step": 27731
+    },
+    {
+      "epoch": 0.6600971151099686,
+      "grad_norm": 2.9303934849188766,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 27732
+    },
+    {
+      "epoch": 0.6601209178330001,
+      "grad_norm": 2.5438623092048425,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 27733
+    },
+    {
+      "epoch": 0.6601447205560316,
+      "grad_norm": 0.4159401802099108,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 27734
+    },
+    {
+      "epoch": 0.6601685232790632,
+      "grad_norm": 8.439755454438947,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 27735
+    },
+    {
+      "epoch": 0.6601923260020947,
+      "grad_norm": 0.4827607907041175,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 27736
+    },
+    {
+      "epoch": 0.6602161287251261,
+      "grad_norm": 10.58992596145829,
+      "learning_rate": 2e-05,
+      "loss": 1.0474,
+      "step": 27737
+    },
+    {
+      "epoch": 0.6602399314481576,
+      "grad_norm": 5.017497627948763,
+      "learning_rate": 2e-05,
+      "loss": 1.2905,
+      "step": 27738
+    },
+    {
+      "epoch": 0.6602637341711892,
+      "grad_norm": 0.9441860299843724,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 27739
+    },
+    {
+      "epoch": 0.6602875368942207,
+      "grad_norm": 5.357361548885359,
+      "learning_rate": 2e-05,
+      "loss": 0.8097,
+      "step": 27740
+    },
+    {
+      "epoch": 0.6603113396172522,
+      "grad_norm": 7.1579115937952436,
+      "learning_rate": 2e-05,
+      "loss": 1.2127,
+      "step": 27741
+    },
+    {
+      "epoch": 0.6603351423402837,
+      "grad_norm": 9.20967638731061,
+      "learning_rate": 2e-05,
+      "loss": 0.331,
+      "step": 27742
+    },
+    {
+      "epoch": 0.6603589450633153,
+      "grad_norm": 3.081941116768504,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 27743
+    },
+    {
+      "epoch": 0.6603827477863468,
+      "grad_norm": 9.214723616161606,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 27744
+    },
+    {
+      "epoch": 0.6604065505093782,
+      "grad_norm": 8.265708800669268,
+      "learning_rate": 2e-05,
+      "loss": 0.2372,
+      "step": 27745
+    },
+    {
+      "epoch": 0.6604303532324098,
+      "grad_norm": 22.45835425683574,
+      "learning_rate": 2e-05,
+      "loss": 0.3197,
+      "step": 27746
+    },
+    {
+      "epoch": 0.6604541559554413,
+      "grad_norm": 5.99094231173421,
+      "learning_rate": 2e-05,
+      "loss": 1.6407,
+      "step": 27747
+    },
+    {
+      "epoch": 0.6604779586784728,
+      "grad_norm": 3.187929405732065,
+      "learning_rate": 2e-05,
+      "loss": 0.5281,
+      "step": 27748
+    },
+    {
+      "epoch": 0.6605017614015043,
+      "grad_norm": 4.151747768205897,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 27749
+    },
+    {
+      "epoch": 0.6605255641245359,
+      "grad_norm": 0.5553425111986331,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 27750
+    },
+    {
+      "epoch": 0.6605493668475674,
+      "grad_norm": 5.637128684965677,
+      "learning_rate": 2e-05,
+      "loss": 0.213,
+      "step": 27751
+    },
+    {
+      "epoch": 0.6605731695705989,
+      "grad_norm": 4.57702363952713,
+      "learning_rate": 2e-05,
+      "loss": 0.5952,
+      "step": 27752
+    },
+    {
+      "epoch": 0.6605969722936303,
+      "grad_norm": 0.7483039948154918,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 27753
+    },
+    {
+      "epoch": 0.6606207750166619,
+      "grad_norm": 0.12592700306796517,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 27754
+    },
+    {
+      "epoch": 0.6606445777396934,
+      "grad_norm": 1.7461403807969655,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 27755
+    },
+    {
+      "epoch": 0.6606683804627249,
+      "grad_norm": 6.309230034481342,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 27756
+    },
+    {
+      "epoch": 0.6606921831857564,
+      "grad_norm": 3.8160745336720128,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 27757
+    },
+    {
+      "epoch": 0.660715985908788,
+      "grad_norm": 3.956075113548175,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 27758
+    },
+    {
+      "epoch": 0.6607397886318195,
+      "grad_norm": 0.18504761419458898,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 27759
+    },
+    {
+      "epoch": 0.660763591354851,
+      "grad_norm": 5.1700426130027095,
+      "learning_rate": 2e-05,
+      "loss": 0.7581,
+      "step": 27760
+    },
+    {
+      "epoch": 0.6607873940778826,
+      "grad_norm": 4.891473554242543,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 27761
+    },
+    {
+      "epoch": 0.660811196800914,
+      "grad_norm": 2.77237833755612,
+      "learning_rate": 2e-05,
+      "loss": 0.139,
+      "step": 27762
+    },
+    {
+      "epoch": 0.6608349995239455,
+      "grad_norm": 3.3776698292558165,
+      "learning_rate": 2e-05,
+      "loss": 0.6991,
+      "step": 27763
+    },
+    {
+      "epoch": 0.660858802246977,
+      "grad_norm": 5.221783387416904,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 27764
+    },
+    {
+      "epoch": 0.6608826049700086,
+      "grad_norm": 6.4362083062403,
+      "learning_rate": 2e-05,
+      "loss": 0.5099,
+      "step": 27765
+    },
+    {
+      "epoch": 0.6609064076930401,
+      "grad_norm": 4.471759932221152,
+      "learning_rate": 2e-05,
+      "loss": 0.6865,
+      "step": 27766
+    },
+    {
+      "epoch": 0.6609302104160716,
+      "grad_norm": 0.31819272821410954,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 27767
+    },
+    {
+      "epoch": 0.6609540131391031,
+      "grad_norm": 3.291873385654452,
+      "learning_rate": 2e-05,
+      "loss": 0.5377,
+      "step": 27768
+    },
+    {
+      "epoch": 0.6609778158621347,
+      "grad_norm": 15.891533847916689,
+      "learning_rate": 2e-05,
+      "loss": 0.9554,
+      "step": 27769
+    },
+    {
+      "epoch": 0.6610016185851662,
+      "grad_norm": 4.306388474749144,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 27770
+    },
+    {
+      "epoch": 0.6610254213081976,
+      "grad_norm": 4.52206710256074,
+      "learning_rate": 2e-05,
+      "loss": 0.8608,
+      "step": 27771
+    },
+    {
+      "epoch": 0.6610492240312291,
+      "grad_norm": 10.68032947538742,
+      "learning_rate": 2e-05,
+      "loss": 0.6742,
+      "step": 27772
+    },
+    {
+      "epoch": 0.6610730267542607,
+      "grad_norm": 8.248307543236361,
+      "learning_rate": 2e-05,
+      "loss": 0.4072,
+      "step": 27773
+    },
+    {
+      "epoch": 0.6610968294772922,
+      "grad_norm": 1.1976922087660324,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 27774
+    },
+    {
+      "epoch": 0.6611206322003237,
+      "grad_norm": 2.0321944970250723,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 27775
+    },
+    {
+      "epoch": 0.6611444349233553,
+      "grad_norm": 6.790861007689428,
+      "learning_rate": 2e-05,
+      "loss": 0.1602,
+      "step": 27776
+    },
+    {
+      "epoch": 0.6611682376463868,
+      "grad_norm": 11.989997664867518,
+      "learning_rate": 2e-05,
+      "loss": 1.1123,
+      "step": 27777
+    },
+    {
+      "epoch": 0.6611920403694183,
+      "grad_norm": 4.0939648575108825,
+      "learning_rate": 2e-05,
+      "loss": 0.9636,
+      "step": 27778
+    },
+    {
+      "epoch": 0.6612158430924497,
+      "grad_norm": 3.4143692428241894,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 27779
+    },
+    {
+      "epoch": 0.6612396458154813,
+      "grad_norm": 3.901904656754677,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 27780
+    },
+    {
+      "epoch": 0.6612634485385128,
+      "grad_norm": 3.309272935619292,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 27781
+    },
+    {
+      "epoch": 0.6612872512615443,
+      "grad_norm": 5.795231384436917,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 27782
+    },
+    {
+      "epoch": 0.6613110539845758,
+      "grad_norm": 2.1314658184210713,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 27783
+    },
+    {
+      "epoch": 0.6613348567076074,
+      "grad_norm": 6.264980758670116,
+      "learning_rate": 2e-05,
+      "loss": 0.6987,
+      "step": 27784
+    },
+    {
+      "epoch": 0.6613586594306389,
+      "grad_norm": 4.5299039937349646,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 27785
+    },
+    {
+      "epoch": 0.6613824621536704,
+      "grad_norm": 5.325710621082411,
+      "learning_rate": 2e-05,
+      "loss": 0.2579,
+      "step": 27786
+    },
+    {
+      "epoch": 0.6614062648767018,
+      "grad_norm": 3.1328020856616243,
+      "learning_rate": 2e-05,
+      "loss": 0.8863,
+      "step": 27787
+    },
+    {
+      "epoch": 0.6614300675997334,
+      "grad_norm": 0.5352433638731006,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 27788
+    },
+    {
+      "epoch": 0.6614538703227649,
+      "grad_norm": 0.8361272898835554,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 27789
+    },
+    {
+      "epoch": 0.6614776730457964,
+      "grad_norm": 1.6890283002240951,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 27790
+    },
+    {
+      "epoch": 0.661501475768828,
+      "grad_norm": 0.5177709606438077,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 27791
+    },
+    {
+      "epoch": 0.6615252784918595,
+      "grad_norm": 2.9665788547332013,
+      "learning_rate": 2e-05,
+      "loss": 0.2816,
+      "step": 27792
+    },
+    {
+      "epoch": 0.661549081214891,
+      "grad_norm": 10.377867926280462,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 27793
+    },
+    {
+      "epoch": 0.6615728839379225,
+      "grad_norm": 2.576495844069876,
+      "learning_rate": 2e-05,
+      "loss": 0.1831,
+      "step": 27794
+    },
+    {
+      "epoch": 0.6615966866609541,
+      "grad_norm": 1.0726979923601783,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 27795
+    },
+    {
+      "epoch": 0.6616204893839855,
+      "grad_norm": 0.9470131401956685,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 27796
+    },
+    {
+      "epoch": 0.661644292107017,
+      "grad_norm": 4.835846142716575,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 27797
+    },
+    {
+      "epoch": 0.6616680948300485,
+      "grad_norm": 4.166002592795677,
+      "learning_rate": 2e-05,
+      "loss": 0.5303,
+      "step": 27798
+    },
+    {
+      "epoch": 0.6616918975530801,
+      "grad_norm": 6.315143620045786,
+      "learning_rate": 2e-05,
+      "loss": 0.1248,
+      "step": 27799
+    },
+    {
+      "epoch": 0.6617157002761116,
+      "grad_norm": 0.5112496253329548,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 27800
+    },
+    {
+      "epoch": 0.6617395029991431,
+      "grad_norm": 2.2068882972431116,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 27801
+    },
+    {
+      "epoch": 0.6617633057221746,
+      "grad_norm": 5.0761465001685755,
+      "learning_rate": 2e-05,
+      "loss": 0.4985,
+      "step": 27802
+    },
+    {
+      "epoch": 0.6617871084452062,
+      "grad_norm": 19.72085178295422,
+      "learning_rate": 2e-05,
+      "loss": 0.7938,
+      "step": 27803
+    },
+    {
+      "epoch": 0.6618109111682376,
+      "grad_norm": 0.05108190762589647,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 27804
+    },
+    {
+      "epoch": 0.6618347138912691,
+      "grad_norm": 11.531515844605904,
+      "learning_rate": 2e-05,
+      "loss": 1.0199,
+      "step": 27805
+    },
+    {
+      "epoch": 0.6618585166143007,
+      "grad_norm": 1.3599291882601594,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 27806
+    },
+    {
+      "epoch": 0.6618823193373322,
+      "grad_norm": 2.750976372360631,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 27807
+    },
+    {
+      "epoch": 0.6619061220603637,
+      "grad_norm": 2.0925549541706316,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 27808
+    },
+    {
+      "epoch": 0.6619299247833952,
+      "grad_norm": 5.007816023792032,
+      "learning_rate": 2e-05,
+      "loss": 0.6693,
+      "step": 27809
+    },
+    {
+      "epoch": 0.6619537275064268,
+      "grad_norm": 0.7717652871253985,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 27810
+    },
+    {
+      "epoch": 0.6619775302294583,
+      "grad_norm": 8.804422900657098,
+      "learning_rate": 2e-05,
+      "loss": 0.4384,
+      "step": 27811
+    },
+    {
+      "epoch": 0.6620013329524898,
+      "grad_norm": 0.6754205711215346,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 27812
+    },
+    {
+      "epoch": 0.6620251356755212,
+      "grad_norm": 11.147351014675184,
+      "learning_rate": 2e-05,
+      "loss": 0.6069,
+      "step": 27813
+    },
+    {
+      "epoch": 0.6620489383985528,
+      "grad_norm": 4.508120700616587,
+      "learning_rate": 2e-05,
+      "loss": 0.9549,
+      "step": 27814
+    },
+    {
+      "epoch": 0.6620727411215843,
+      "grad_norm": 5.0591977357792475,
+      "learning_rate": 2e-05,
+      "loss": 0.6337,
+      "step": 27815
+    },
+    {
+      "epoch": 0.6620965438446158,
+      "grad_norm": 9.080014512472486,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 27816
+    },
+    {
+      "epoch": 0.6621203465676473,
+      "grad_norm": 2.7643879016908954,
+      "learning_rate": 2e-05,
+      "loss": 0.3532,
+      "step": 27817
+    },
+    {
+      "epoch": 0.6621441492906789,
+      "grad_norm": 5.338291791955538,
+      "learning_rate": 2e-05,
+      "loss": 1.3458,
+      "step": 27818
+    },
+    {
+      "epoch": 0.6621679520137104,
+      "grad_norm": 10.82725611388854,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 27819
+    },
+    {
+      "epoch": 0.6621917547367419,
+      "grad_norm": 4.53849631054414,
+      "learning_rate": 2e-05,
+      "loss": 0.1296,
+      "step": 27820
+    },
+    {
+      "epoch": 0.6622155574597735,
+      "grad_norm": 6.050627748642158,
+      "learning_rate": 2e-05,
+      "loss": 0.8541,
+      "step": 27821
+    },
+    {
+      "epoch": 0.6622393601828049,
+      "grad_norm": 5.036017333719395,
+      "learning_rate": 2e-05,
+      "loss": 0.5902,
+      "step": 27822
+    },
+    {
+      "epoch": 0.6622631629058364,
+      "grad_norm": 4.939315246914471,
+      "learning_rate": 2e-05,
+      "loss": 0.5951,
+      "step": 27823
+    },
+    {
+      "epoch": 0.6622869656288679,
+      "grad_norm": 3.312234717095135,
+      "learning_rate": 2e-05,
+      "loss": 0.2279,
+      "step": 27824
+    },
+    {
+      "epoch": 0.6623107683518995,
+      "grad_norm": 1.6263202753584323,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 27825
+    },
+    {
+      "epoch": 0.662334571074931,
+      "grad_norm": 6.406271304506208,
+      "learning_rate": 2e-05,
+      "loss": 0.4837,
+      "step": 27826
+    },
+    {
+      "epoch": 0.6623583737979625,
+      "grad_norm": 3.242199279610357,
+      "learning_rate": 2e-05,
+      "loss": 0.4765,
+      "step": 27827
+    },
+    {
+      "epoch": 0.662382176520994,
+      "grad_norm": 3.2352358395624092,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 27828
+    },
+    {
+      "epoch": 0.6624059792440256,
+      "grad_norm": 2.6842187645168463,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 27829
+    },
+    {
+      "epoch": 0.662429781967057,
+      "grad_norm": 9.96357120329478,
+      "learning_rate": 2e-05,
+      "loss": 1.5049,
+      "step": 27830
+    },
+    {
+      "epoch": 0.6624535846900885,
+      "grad_norm": 2.5368284319442607,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 27831
+    },
+    {
+      "epoch": 0.66247738741312,
+      "grad_norm": 14.326160215534767,
+      "learning_rate": 2e-05,
+      "loss": 0.9184,
+      "step": 27832
+    },
+    {
+      "epoch": 0.6625011901361516,
+      "grad_norm": 9.078533144408112,
+      "learning_rate": 2e-05,
+      "loss": 0.4274,
+      "step": 27833
+    },
+    {
+      "epoch": 0.6625249928591831,
+      "grad_norm": 1.4398506692216058,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 27834
+    },
+    {
+      "epoch": 0.6625487955822146,
+      "grad_norm": 1.2329711911548913,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 27835
+    },
+    {
+      "epoch": 0.6625725983052462,
+      "grad_norm": 2.9074802211476345,
+      "learning_rate": 2e-05,
+      "loss": 0.4583,
+      "step": 27836
+    },
+    {
+      "epoch": 0.6625964010282777,
+      "grad_norm": 2.934608143992625,
+      "learning_rate": 2e-05,
+      "loss": 0.3555,
+      "step": 27837
+    },
+    {
+      "epoch": 0.6626202037513091,
+      "grad_norm": 1.2636397112829099,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 27838
+    },
+    {
+      "epoch": 0.6626440064743406,
+      "grad_norm": 3.7961117492766863,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 27839
+    },
+    {
+      "epoch": 0.6626678091973722,
+      "grad_norm": 6.879052069719508,
+      "learning_rate": 2e-05,
+      "loss": 0.996,
+      "step": 27840
+    },
+    {
+      "epoch": 0.6626916119204037,
+      "grad_norm": 0.0337650634847103,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 27841
+    },
+    {
+      "epoch": 0.6627154146434352,
+      "grad_norm": 1.464189223189696,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 27842
+    },
+    {
+      "epoch": 0.6627392173664667,
+      "grad_norm": 0.11812508171311516,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 27843
+    },
+    {
+      "epoch": 0.6627630200894983,
+      "grad_norm": 3.1433453602750228,
+      "learning_rate": 2e-05,
+      "loss": 0.8949,
+      "step": 27844
+    },
+    {
+      "epoch": 0.6627868228125298,
+      "grad_norm": 6.91513437125159,
+      "learning_rate": 2e-05,
+      "loss": 0.9497,
+      "step": 27845
+    },
+    {
+      "epoch": 0.6628106255355612,
+      "grad_norm": 7.30819997261324,
+      "learning_rate": 2e-05,
+      "loss": 1.1471,
+      "step": 27846
+    },
+    {
+      "epoch": 0.6628344282585927,
+      "grad_norm": 1.8477184110448106,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 27847
+    },
+    {
+      "epoch": 0.6628582309816243,
+      "grad_norm": 4.5484239234768316,
+      "learning_rate": 2e-05,
+      "loss": 0.9112,
+      "step": 27848
+    },
+    {
+      "epoch": 0.6628820337046558,
+      "grad_norm": 3.150028352575084,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 27849
+    },
+    {
+      "epoch": 0.6629058364276873,
+      "grad_norm": 1.7416747287058847,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 27850
+    },
+    {
+      "epoch": 0.6629296391507189,
+      "grad_norm": 1.4842989171854282,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 27851
+    },
+    {
+      "epoch": 0.6629534418737504,
+      "grad_norm": 5.243223405254911,
+      "learning_rate": 2e-05,
+      "loss": 0.0717,
+      "step": 27852
+    },
+    {
+      "epoch": 0.6629772445967819,
+      "grad_norm": 2.624309350966783,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 27853
+    },
+    {
+      "epoch": 0.6630010473198134,
+      "grad_norm": 6.424960384785803,
+      "learning_rate": 2e-05,
+      "loss": 0.8248,
+      "step": 27854
+    },
+    {
+      "epoch": 0.663024850042845,
+      "grad_norm": 0.2766485256747039,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 27855
+    },
+    {
+      "epoch": 0.6630486527658764,
+      "grad_norm": 2.6803129999838875,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 27856
+    },
+    {
+      "epoch": 0.6630724554889079,
+      "grad_norm": 4.0295296706542425,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 27857
+    },
+    {
+      "epoch": 0.6630962582119394,
+      "grad_norm": 0.8077661789940748,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 27858
+    },
+    {
+      "epoch": 0.663120060934971,
+      "grad_norm": 4.993648869785565,
+      "learning_rate": 2e-05,
+      "loss": 0.1579,
+      "step": 27859
+    },
+    {
+      "epoch": 0.6631438636580025,
+      "grad_norm": 2.981656925276605,
+      "learning_rate": 2e-05,
+      "loss": 0.3642,
+      "step": 27860
+    },
+    {
+      "epoch": 0.663167666381034,
+      "grad_norm": 0.04089415006446632,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 27861
+    },
+    {
+      "epoch": 0.6631914691040655,
+      "grad_norm": 5.259892978426441,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 27862
+    },
+    {
+      "epoch": 0.663215271827097,
+      "grad_norm": 0.042811851235442774,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 27863
+    },
+    {
+      "epoch": 0.6632390745501285,
+      "grad_norm": 7.263232504264711,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 27864
+    },
+    {
+      "epoch": 0.66326287727316,
+      "grad_norm": 3.3724727756432795,
+      "learning_rate": 2e-05,
+      "loss": 0.5928,
+      "step": 27865
+    },
+    {
+      "epoch": 0.6632866799961916,
+      "grad_norm": 1.8149021837699006,
+      "learning_rate": 2e-05,
+      "loss": 0.1669,
+      "step": 27866
+    },
+    {
+      "epoch": 0.6633104827192231,
+      "grad_norm": 0.6290942266301389,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 27867
+    },
+    {
+      "epoch": 0.6633342854422546,
+      "grad_norm": 0.11748545365986264,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 27868
+    },
+    {
+      "epoch": 0.6633580881652861,
+      "grad_norm": 6.961919278782119,
+      "learning_rate": 2e-05,
+      "loss": 1.3531,
+      "step": 27869
+    },
+    {
+      "epoch": 0.6633818908883177,
+      "grad_norm": 0.3784056396563839,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 27870
+    },
+    {
+      "epoch": 0.6634056936113492,
+      "grad_norm": 2.8663409254091805,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 27871
+    },
+    {
+      "epoch": 0.6634294963343806,
+      "grad_norm": 0.3653968873224659,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 27872
+    },
+    {
+      "epoch": 0.6634532990574121,
+      "grad_norm": 1.6962705357877257,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 27873
+    },
+    {
+      "epoch": 0.6634771017804437,
+      "grad_norm": 6.010401456207814,
+      "learning_rate": 2e-05,
+      "loss": 0.7855,
+      "step": 27874
+    },
+    {
+      "epoch": 0.6635009045034752,
+      "grad_norm": 3.0141795262950937,
+      "learning_rate": 2e-05,
+      "loss": 0.5538,
+      "step": 27875
+    },
+    {
+      "epoch": 0.6635247072265067,
+      "grad_norm": 58.55928087860289,
+      "learning_rate": 2e-05,
+      "loss": 0.484,
+      "step": 27876
+    },
+    {
+      "epoch": 0.6635485099495382,
+      "grad_norm": 2.0310673552967127,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 27877
+    },
+    {
+      "epoch": 0.6635723126725698,
+      "grad_norm": 0.09709019248813798,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 27878
+    },
+    {
+      "epoch": 0.6635961153956013,
+      "grad_norm": 6.212800658447042,
+      "learning_rate": 2e-05,
+      "loss": 1.199,
+      "step": 27879
+    },
+    {
+      "epoch": 0.6636199181186327,
+      "grad_norm": 0.7995377853754444,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 27880
+    },
+    {
+      "epoch": 0.6636437208416643,
+      "grad_norm": 0.23172998101129935,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 27881
+    },
+    {
+      "epoch": 0.6636675235646958,
+      "grad_norm": 1.360084511893107,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 27882
+    },
+    {
+      "epoch": 0.6636913262877273,
+      "grad_norm": 3.2529113627770383,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 27883
+    },
+    {
+      "epoch": 0.6637151290107588,
+      "grad_norm": 0.7806828266422241,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 27884
+    },
+    {
+      "epoch": 0.6637389317337904,
+      "grad_norm": 5.409848798634058,
+      "learning_rate": 2e-05,
+      "loss": 1.0382,
+      "step": 27885
+    },
+    {
+      "epoch": 0.6637627344568219,
+      "grad_norm": 3.516508814198749,
+      "learning_rate": 2e-05,
+      "loss": 0.4292,
+      "step": 27886
+    },
+    {
+      "epoch": 0.6637865371798534,
+      "grad_norm": 2.8481839592012683,
+      "learning_rate": 2e-05,
+      "loss": 0.3838,
+      "step": 27887
+    },
+    {
+      "epoch": 0.6638103399028848,
+      "grad_norm": 3.6917549872262296,
+      "learning_rate": 2e-05,
+      "loss": 0.6303,
+      "step": 27888
+    },
+    {
+      "epoch": 0.6638341426259164,
+      "grad_norm": 9.761059051546926,
+      "learning_rate": 2e-05,
+      "loss": 0.8983,
+      "step": 27889
+    },
+    {
+      "epoch": 0.6638579453489479,
+      "grad_norm": 9.193352647813192,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 27890
+    },
+    {
+      "epoch": 0.6638817480719794,
+      "grad_norm": 4.959296918849719,
+      "learning_rate": 2e-05,
+      "loss": 1.1318,
+      "step": 27891
+    },
+    {
+      "epoch": 0.6639055507950109,
+      "grad_norm": 8.770375920750594,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 27892
+    },
+    {
+      "epoch": 0.6639293535180425,
+      "grad_norm": 5.699945253303394,
+      "learning_rate": 2e-05,
+      "loss": 1.0401,
+      "step": 27893
+    },
+    {
+      "epoch": 0.663953156241074,
+      "grad_norm": 6.1903370428336215,
+      "learning_rate": 2e-05,
+      "loss": 1.3128,
+      "step": 27894
+    },
+    {
+      "epoch": 0.6639769589641055,
+      "grad_norm": 2.6454262018654124,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 27895
+    },
+    {
+      "epoch": 0.664000761687137,
+      "grad_norm": 4.89778487307379,
+      "learning_rate": 2e-05,
+      "loss": 0.8503,
+      "step": 27896
+    },
+    {
+      "epoch": 0.6640245644101685,
+      "grad_norm": 21.32666002241931,
+      "learning_rate": 2e-05,
+      "loss": 1.3811,
+      "step": 27897
+    },
+    {
+      "epoch": 0.6640483671332,
+      "grad_norm": 9.368004916454057,
+      "learning_rate": 2e-05,
+      "loss": 0.4774,
+      "step": 27898
+    },
+    {
+      "epoch": 0.6640721698562315,
+      "grad_norm": 3.950391637061805,
+      "learning_rate": 2e-05,
+      "loss": 0.6816,
+      "step": 27899
+    },
+    {
+      "epoch": 0.6640959725792631,
+      "grad_norm": 3.565882756955104,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 27900
+    },
+    {
+      "epoch": 0.6641197753022946,
+      "grad_norm": 0.3190338383670388,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 27901
+    },
+    {
+      "epoch": 0.6641435780253261,
+      "grad_norm": 5.346344772242125,
+      "learning_rate": 2e-05,
+      "loss": 1.0814,
+      "step": 27902
+    },
+    {
+      "epoch": 0.6641673807483576,
+      "grad_norm": 29.570640631340126,
+      "learning_rate": 2e-05,
+      "loss": 0.548,
+      "step": 27903
+    },
+    {
+      "epoch": 0.6641911834713892,
+      "grad_norm": 2.7199741365309227,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 27904
+    },
+    {
+      "epoch": 0.6642149861944207,
+      "grad_norm": 0.9278729246098357,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 27905
+    },
+    {
+      "epoch": 0.6642387889174521,
+      "grad_norm": 4.207467874934519,
+      "learning_rate": 2e-05,
+      "loss": 0.2566,
+      "step": 27906
+    },
+    {
+      "epoch": 0.6642625916404836,
+      "grad_norm": 8.648604777810489,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 27907
+    },
+    {
+      "epoch": 0.6642863943635152,
+      "grad_norm": 0.28620311222840683,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 27908
+    },
+    {
+      "epoch": 0.6643101970865467,
+      "grad_norm": 2.446303775565399,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 27909
+    },
+    {
+      "epoch": 0.6643339998095782,
+      "grad_norm": 2.691958440400325,
+      "learning_rate": 2e-05,
+      "loss": 0.3954,
+      "step": 27910
+    },
+    {
+      "epoch": 0.6643578025326097,
+      "grad_norm": 1.1093760602447928,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 27911
+    },
+    {
+      "epoch": 0.6643816052556413,
+      "grad_norm": 3.3065197396076287,
+      "learning_rate": 2e-05,
+      "loss": 0.2577,
+      "step": 27912
+    },
+    {
+      "epoch": 0.6644054079786728,
+      "grad_norm": 2.3542706697503197,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 27913
+    },
+    {
+      "epoch": 0.6644292107017042,
+      "grad_norm": 1.5939629478200923,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 27914
+    },
+    {
+      "epoch": 0.6644530134247358,
+      "grad_norm": 5.5125374866386325,
+      "learning_rate": 2e-05,
+      "loss": 0.8121,
+      "step": 27915
+    },
+    {
+      "epoch": 0.6644768161477673,
+      "grad_norm": 0.4188062926234487,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 27916
+    },
+    {
+      "epoch": 0.6645006188707988,
+      "grad_norm": 0.3559122528022509,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 27917
+    },
+    {
+      "epoch": 0.6645244215938303,
+      "grad_norm": 0.4827660352323995,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 27918
+    },
+    {
+      "epoch": 0.6645482243168619,
+      "grad_norm": 5.067481248646937,
+      "learning_rate": 2e-05,
+      "loss": 1.4033,
+      "step": 27919
+    },
+    {
+      "epoch": 0.6645720270398934,
+      "grad_norm": 9.007761342638739,
+      "learning_rate": 2e-05,
+      "loss": 0.7927,
+      "step": 27920
+    },
+    {
+      "epoch": 0.6645958297629249,
+      "grad_norm": 0.10477457136564744,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 27921
+    },
+    {
+      "epoch": 0.6646196324859563,
+      "grad_norm": 3.263585468469026,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 27922
+    },
+    {
+      "epoch": 0.6646434352089879,
+      "grad_norm": 4.069885389874867,
+      "learning_rate": 2e-05,
+      "loss": 0.5949,
+      "step": 27923
+    },
+    {
+      "epoch": 0.6646672379320194,
+      "grad_norm": 0.4634022358913262,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 27924
+    },
+    {
+      "epoch": 0.6646910406550509,
+      "grad_norm": 0.0545145001531997,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 27925
+    },
+    {
+      "epoch": 0.6647148433780824,
+      "grad_norm": 3.2826533284602535,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 27926
+    },
+    {
+      "epoch": 0.664738646101114,
+      "grad_norm": 4.935270698704421,
+      "learning_rate": 2e-05,
+      "loss": 0.8336,
+      "step": 27927
+    },
+    {
+      "epoch": 0.6647624488241455,
+      "grad_norm": 6.253621228629246,
+      "learning_rate": 2e-05,
+      "loss": 0.8224,
+      "step": 27928
+    },
+    {
+      "epoch": 0.664786251547177,
+      "grad_norm": 0.6362581723191235,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 27929
+    },
+    {
+      "epoch": 0.6648100542702086,
+      "grad_norm": 3.726065040518442,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 27930
+    },
+    {
+      "epoch": 0.66483385699324,
+      "grad_norm": 15.55924422756061,
+      "learning_rate": 2e-05,
+      "loss": 0.3071,
+      "step": 27931
+    },
+    {
+      "epoch": 0.6648576597162715,
+      "grad_norm": 1.8612813361247125,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 27932
+    },
+    {
+      "epoch": 0.664881462439303,
+      "grad_norm": 0.2277960220686676,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 27933
+    },
+    {
+      "epoch": 0.6649052651623346,
+      "grad_norm": 2.6087725541807436,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 27934
+    },
+    {
+      "epoch": 0.6649290678853661,
+      "grad_norm": 2.071726558263131,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 27935
+    },
+    {
+      "epoch": 0.6649528706083976,
+      "grad_norm": 0.06475175876582626,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 27936
+    },
+    {
+      "epoch": 0.6649766733314291,
+      "grad_norm": 4.247265713806652,
+      "learning_rate": 2e-05,
+      "loss": 0.9183,
+      "step": 27937
+    },
+    {
+      "epoch": 0.6650004760544607,
+      "grad_norm": 14.816887005564599,
+      "learning_rate": 2e-05,
+      "loss": 0.8366,
+      "step": 27938
+    },
+    {
+      "epoch": 0.6650242787774922,
+      "grad_norm": 5.504047791039688,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 27939
+    },
+    {
+      "epoch": 0.6650480815005236,
+      "grad_norm": 5.208010935100569,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 27940
+    },
+    {
+      "epoch": 0.6650718842235551,
+      "grad_norm": 1.9180580899925759,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 27941
+    },
+    {
+      "epoch": 0.6650956869465867,
+      "grad_norm": 12.472943250859371,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 27942
+    },
+    {
+      "epoch": 0.6651194896696182,
+      "grad_norm": 6.92784655989351,
+      "learning_rate": 2e-05,
+      "loss": 0.2544,
+      "step": 27943
+    },
+    {
+      "epoch": 0.6651432923926497,
+      "grad_norm": 7.506035517229499,
+      "learning_rate": 2e-05,
+      "loss": 0.212,
+      "step": 27944
+    },
+    {
+      "epoch": 0.6651670951156813,
+      "grad_norm": 5.348827254351541,
+      "learning_rate": 2e-05,
+      "loss": 0.9323,
+      "step": 27945
+    },
+    {
+      "epoch": 0.6651908978387128,
+      "grad_norm": 2.8430597339668466,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 27946
+    },
+    {
+      "epoch": 0.6652147005617443,
+      "grad_norm": 9.777880782501597,
+      "learning_rate": 2e-05,
+      "loss": 0.2586,
+      "step": 27947
+    },
+    {
+      "epoch": 0.6652385032847757,
+      "grad_norm": 0.579643542733404,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 27948
+    },
+    {
+      "epoch": 0.6652623060078073,
+      "grad_norm": 9.057583285823233,
+      "learning_rate": 2e-05,
+      "loss": 1.2106,
+      "step": 27949
+    },
+    {
+      "epoch": 0.6652861087308388,
+      "grad_norm": 3.6318688487322976,
+      "learning_rate": 2e-05,
+      "loss": 0.1903,
+      "step": 27950
+    },
+    {
+      "epoch": 0.6653099114538703,
+      "grad_norm": 9.005272141731432,
+      "learning_rate": 2e-05,
+      "loss": 1.7341,
+      "step": 27951
+    },
+    {
+      "epoch": 0.6653337141769018,
+      "grad_norm": 7.913936841747871,
+      "learning_rate": 2e-05,
+      "loss": 1.3633,
+      "step": 27952
+    },
+    {
+      "epoch": 0.6653575168999334,
+      "grad_norm": 1.523725461193295,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 27953
+    },
+    {
+      "epoch": 0.6653813196229649,
+      "grad_norm": 3.0083642145211598,
+      "learning_rate": 2e-05,
+      "loss": 0.3778,
+      "step": 27954
+    },
+    {
+      "epoch": 0.6654051223459964,
+      "grad_norm": 0.06509587713807502,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 27955
+    },
+    {
+      "epoch": 0.6654289250690278,
+      "grad_norm": 2.601369291886714,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 27956
+    },
+    {
+      "epoch": 0.6654527277920594,
+      "grad_norm": 0.14937361492579312,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 27957
+    },
+    {
+      "epoch": 0.6654765305150909,
+      "grad_norm": 2.883073396395244,
+      "learning_rate": 2e-05,
+      "loss": 0.3309,
+      "step": 27958
+    },
+    {
+      "epoch": 0.6655003332381224,
+      "grad_norm": 5.384527440738528,
+      "learning_rate": 2e-05,
+      "loss": 0.283,
+      "step": 27959
+    },
+    {
+      "epoch": 0.665524135961154,
+      "grad_norm": 5.23745748037084,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 27960
+    },
+    {
+      "epoch": 0.6655479386841855,
+      "grad_norm": 0.18093163046563712,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 27961
+    },
+    {
+      "epoch": 0.665571741407217,
+      "grad_norm": 0.2551918336804578,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 27962
+    },
+    {
+      "epoch": 0.6655955441302485,
+      "grad_norm": 4.466021826843465,
+      "learning_rate": 2e-05,
+      "loss": 0.9254,
+      "step": 27963
+    },
+    {
+      "epoch": 0.6656193468532801,
+      "grad_norm": 0.12647821638589676,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 27964
+    },
+    {
+      "epoch": 0.6656431495763115,
+      "grad_norm": 1.2404032597814365,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 27965
+    },
+    {
+      "epoch": 0.665666952299343,
+      "grad_norm": 12.19340659213514,
+      "learning_rate": 2e-05,
+      "loss": 1.014,
+      "step": 27966
+    },
+    {
+      "epoch": 0.6656907550223745,
+      "grad_norm": 5.296351410245173,
+      "learning_rate": 2e-05,
+      "loss": 0.2967,
+      "step": 27967
+    },
+    {
+      "epoch": 0.6657145577454061,
+      "grad_norm": 10.913347758647355,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 27968
+    },
+    {
+      "epoch": 0.6657383604684376,
+      "grad_norm": 0.04173958464278468,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 27969
+    },
+    {
+      "epoch": 0.6657621631914691,
+      "grad_norm": 3.7001206828133757,
+      "learning_rate": 2e-05,
+      "loss": 0.4148,
+      "step": 27970
+    },
+    {
+      "epoch": 0.6657859659145006,
+      "grad_norm": 5.105490083433235,
+      "learning_rate": 2e-05,
+      "loss": 0.2543,
+      "step": 27971
+    },
+    {
+      "epoch": 0.6658097686375322,
+      "grad_norm": 2.1304801424280297,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 27972
+    },
+    {
+      "epoch": 0.6658335713605636,
+      "grad_norm": 6.064524573969816,
+      "learning_rate": 2e-05,
+      "loss": 0.9554,
+      "step": 27973
+    },
+    {
+      "epoch": 0.6658573740835951,
+      "grad_norm": 7.228590068634069,
+      "learning_rate": 2e-05,
+      "loss": 1.1367,
+      "step": 27974
+    },
+    {
+      "epoch": 0.6658811768066267,
+      "grad_norm": 1.1896468685890789,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 27975
+    },
+    {
+      "epoch": 0.6659049795296582,
+      "grad_norm": 1.434545220519441,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 27976
+    },
+    {
+      "epoch": 0.6659287822526897,
+      "grad_norm": 1.4779179655842385,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 27977
+    },
+    {
+      "epoch": 0.6659525849757212,
+      "grad_norm": 0.2349851324240972,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 27978
+    },
+    {
+      "epoch": 0.6659763876987528,
+      "grad_norm": 13.121338579544103,
+      "learning_rate": 2e-05,
+      "loss": 0.6273,
+      "step": 27979
+    },
+    {
+      "epoch": 0.6660001904217843,
+      "grad_norm": 5.528976592231239,
+      "learning_rate": 2e-05,
+      "loss": 0.702,
+      "step": 27980
+    },
+    {
+      "epoch": 0.6660239931448158,
+      "grad_norm": 1.9734722478148299,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 27981
+    },
+    {
+      "epoch": 0.6660477958678472,
+      "grad_norm": 14.574734103606989,
+      "learning_rate": 2e-05,
+      "loss": 1.0351,
+      "step": 27982
+    },
+    {
+      "epoch": 0.6660715985908788,
+      "grad_norm": 2.8142560926748073,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 27983
+    },
+    {
+      "epoch": 0.6660954013139103,
+      "grad_norm": 2.919734367476826,
+      "learning_rate": 2e-05,
+      "loss": 0.8135,
+      "step": 27984
+    },
+    {
+      "epoch": 0.6661192040369418,
+      "grad_norm": 22.522621860323476,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 27985
+    },
+    {
+      "epoch": 0.6661430067599733,
+      "grad_norm": 13.65471949620716,
+      "learning_rate": 2e-05,
+      "loss": 0.7878,
+      "step": 27986
+    },
+    {
+      "epoch": 0.6661668094830049,
+      "grad_norm": 2.6891635969240584,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 27987
+    },
+    {
+      "epoch": 0.6661906122060364,
+      "grad_norm": 4.004764192878528,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 27988
+    },
+    {
+      "epoch": 0.6662144149290679,
+      "grad_norm": 6.718988596339373,
+      "learning_rate": 2e-05,
+      "loss": 0.1958,
+      "step": 27989
+    },
+    {
+      "epoch": 0.6662382176520995,
+      "grad_norm": 17.39091807980725,
+      "learning_rate": 2e-05,
+      "loss": 0.3792,
+      "step": 27990
+    },
+    {
+      "epoch": 0.6662620203751309,
+      "grad_norm": 2.536714417908404,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 27991
+    },
+    {
+      "epoch": 0.6662858230981624,
+      "grad_norm": 4.949635852832244,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 27992
+    },
+    {
+      "epoch": 0.6663096258211939,
+      "grad_norm": 1.4298223275887112,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 27993
+    },
+    {
+      "epoch": 0.6663334285442255,
+      "grad_norm": 1.8801685392485148,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 27994
+    },
+    {
+      "epoch": 0.666357231267257,
+      "grad_norm": 2.052835101846978,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 27995
+    },
+    {
+      "epoch": 0.6663810339902885,
+      "grad_norm": 5.393090888579663,
+      "learning_rate": 2e-05,
+      "loss": 0.8669,
+      "step": 27996
+    },
+    {
+      "epoch": 0.66640483671332,
+      "grad_norm": 3.0373008193997886,
+      "learning_rate": 2e-05,
+      "loss": 0.465,
+      "step": 27997
+    },
+    {
+      "epoch": 0.6664286394363516,
+      "grad_norm": 2.8957509393526264,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 27998
+    },
+    {
+      "epoch": 0.666452442159383,
+      "grad_norm": 0.8820262280962713,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 27999
+    },
+    {
+      "epoch": 0.6664762448824145,
+      "grad_norm": 7.569968123114966,
+      "learning_rate": 2e-05,
+      "loss": 0.4641,
+      "step": 28000
+    },
+    {
+      "epoch": 0.666500047605446,
+      "grad_norm": 3.576173321871921,
+      "learning_rate": 2e-05,
+      "loss": 0.1334,
+      "step": 28001
+    },
+    {
+      "epoch": 0.6665238503284776,
+      "grad_norm": 0.5411541798291902,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 28002
+    },
+    {
+      "epoch": 0.6665476530515091,
+      "grad_norm": 1.3329149968037506,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 28003
+    },
+    {
+      "epoch": 0.6665714557745406,
+      "grad_norm": 1.3382804901424672,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 28004
+    },
+    {
+      "epoch": 0.6665952584975722,
+      "grad_norm": 4.988880500693899,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 28005
+    },
+    {
+      "epoch": 0.6666190612206037,
+      "grad_norm": 0.4038024524036174,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 28006
+    },
+    {
+      "epoch": 0.6666428639436351,
+      "grad_norm": 0.1239396191884428,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 28007
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 9.543169760148748,
+      "learning_rate": 2e-05,
+      "loss": 1.2351,
+      "step": 28008
+    },
+    {
+      "epoch": 0.6666904693896982,
+      "grad_norm": 5.528868775921935,
+      "learning_rate": 2e-05,
+      "loss": 0.7253,
+      "step": 28009
+    },
+    {
+      "epoch": 0.6667142721127297,
+      "grad_norm": 1.4021235403643242,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 28010
+    },
+    {
+      "epoch": 0.6667380748357612,
+      "grad_norm": 0.10210787629335616,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 28011
+    },
+    {
+      "epoch": 0.6667618775587927,
+      "grad_norm": 3.9426998923521435,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 28012
+    },
+    {
+      "epoch": 0.6667856802818243,
+      "grad_norm": 5.488669492710011,
+      "learning_rate": 2e-05,
+      "loss": 0.6379,
+      "step": 28013
+    },
+    {
+      "epoch": 0.6668094830048558,
+      "grad_norm": 4.2546050980852055,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 28014
+    },
+    {
+      "epoch": 0.6668332857278872,
+      "grad_norm": 0.21491919316415484,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 28015
+    },
+    {
+      "epoch": 0.6668570884509187,
+      "grad_norm": 4.551776978403157,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 28016
+    },
+    {
+      "epoch": 0.6668808911739503,
+      "grad_norm": 1.7549854032556556,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 28017
+    },
+    {
+      "epoch": 0.6669046938969818,
+      "grad_norm": 4.323784832313365,
+      "learning_rate": 2e-05,
+      "loss": 0.3046,
+      "step": 28018
+    },
+    {
+      "epoch": 0.6669284966200133,
+      "grad_norm": 5.192275090296618,
+      "learning_rate": 2e-05,
+      "loss": 0.3591,
+      "step": 28019
+    },
+    {
+      "epoch": 0.6669522993430449,
+      "grad_norm": 0.49496595394285037,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 28020
+    },
+    {
+      "epoch": 0.6669761020660764,
+      "grad_norm": 4.665457751488101,
+      "learning_rate": 2e-05,
+      "loss": 1.2085,
+      "step": 28021
+    },
+    {
+      "epoch": 0.6669999047891079,
+      "grad_norm": 3.5397901662820765,
+      "learning_rate": 2e-05,
+      "loss": 0.4989,
+      "step": 28022
+    },
+    {
+      "epoch": 0.6670237075121394,
+      "grad_norm": 1.3877852690530403,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 28023
+    },
+    {
+      "epoch": 0.667047510235171,
+      "grad_norm": 0.06670084206127108,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28024
+    },
+    {
+      "epoch": 0.6670713129582024,
+      "grad_norm": 6.451367120850581,
+      "learning_rate": 2e-05,
+      "loss": 0.4482,
+      "step": 28025
+    },
+    {
+      "epoch": 0.6670951156812339,
+      "grad_norm": 2.1816725587971395,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 28026
+    },
+    {
+      "epoch": 0.6671189184042654,
+      "grad_norm": 2.705712865334189,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 28027
+    },
+    {
+      "epoch": 0.667142721127297,
+      "grad_norm": 11.658235466022402,
+      "learning_rate": 2e-05,
+      "loss": 0.3414,
+      "step": 28028
+    },
+    {
+      "epoch": 0.6671665238503285,
+      "grad_norm": 2.650591970181042,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 28029
+    },
+    {
+      "epoch": 0.66719032657336,
+      "grad_norm": 1.6067558730258866,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 28030
+    },
+    {
+      "epoch": 0.6672141292963915,
+      "grad_norm": 12.170731307925289,
+      "learning_rate": 2e-05,
+      "loss": 0.123,
+      "step": 28031
+    },
+    {
+      "epoch": 0.667237932019423,
+      "grad_norm": 1.6975071134189177,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 28032
+    },
+    {
+      "epoch": 0.6672617347424545,
+      "grad_norm": 4.208716172090204,
+      "learning_rate": 2e-05,
+      "loss": 0.5585,
+      "step": 28033
+    },
+    {
+      "epoch": 0.667285537465486,
+      "grad_norm": 1.9176692788798204,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 28034
+    },
+    {
+      "epoch": 0.6673093401885176,
+      "grad_norm": 3.2025990193728884,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 28035
+    },
+    {
+      "epoch": 0.6673331429115491,
+      "grad_norm": 1.4963154054729526,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 28036
+    },
+    {
+      "epoch": 0.6673569456345806,
+      "grad_norm": 4.228935066980461,
+      "learning_rate": 2e-05,
+      "loss": 0.7308,
+      "step": 28037
+    },
+    {
+      "epoch": 0.6673807483576121,
+      "grad_norm": 3.9739165662025355,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 28038
+    },
+    {
+      "epoch": 0.6674045510806437,
+      "grad_norm": 27.162610993571892,
+      "learning_rate": 2e-05,
+      "loss": 0.7747,
+      "step": 28039
+    },
+    {
+      "epoch": 0.6674283538036752,
+      "grad_norm": 4.260560456804218,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 28040
+    },
+    {
+      "epoch": 0.6674521565267066,
+      "grad_norm": 8.531489804607991,
+      "learning_rate": 2e-05,
+      "loss": 0.1185,
+      "step": 28041
+    },
+    {
+      "epoch": 0.6674759592497381,
+      "grad_norm": 4.003242835418496,
+      "learning_rate": 2e-05,
+      "loss": 0.5019,
+      "step": 28042
+    },
+    {
+      "epoch": 0.6674997619727697,
+      "grad_norm": 10.278629790443915,
+      "learning_rate": 2e-05,
+      "loss": 1.5175,
+      "step": 28043
+    },
+    {
+      "epoch": 0.6675235646958012,
+      "grad_norm": 16.617074962882054,
+      "learning_rate": 2e-05,
+      "loss": 0.3002,
+      "step": 28044
+    },
+    {
+      "epoch": 0.6675473674188327,
+      "grad_norm": 0.08821747017768557,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28045
+    },
+    {
+      "epoch": 0.6675711701418642,
+      "grad_norm": 9.753061144463862,
+      "learning_rate": 2e-05,
+      "loss": 0.8268,
+      "step": 28046
+    },
+    {
+      "epoch": 0.6675949728648958,
+      "grad_norm": 4.843256832351021,
+      "learning_rate": 2e-05,
+      "loss": 0.7567,
+      "step": 28047
+    },
+    {
+      "epoch": 0.6676187755879273,
+      "grad_norm": 1.1458086347731817,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 28048
+    },
+    {
+      "epoch": 0.6676425783109587,
+      "grad_norm": 11.338695290494075,
+      "learning_rate": 2e-05,
+      "loss": 0.6715,
+      "step": 28049
+    },
+    {
+      "epoch": 0.6676663810339903,
+      "grad_norm": 0.7726924399538029,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 28050
+    },
+    {
+      "epoch": 0.6676901837570218,
+      "grad_norm": 4.474962843165757,
+      "learning_rate": 2e-05,
+      "loss": 0.5692,
+      "step": 28051
+    },
+    {
+      "epoch": 0.6677139864800533,
+      "grad_norm": 6.553018112420368,
+      "learning_rate": 2e-05,
+      "loss": 0.3588,
+      "step": 28052
+    },
+    {
+      "epoch": 0.6677377892030848,
+      "grad_norm": 5.317008942304769,
+      "learning_rate": 2e-05,
+      "loss": 0.1263,
+      "step": 28053
+    },
+    {
+      "epoch": 0.6677615919261164,
+      "grad_norm": 1.608864388856001,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 28054
+    },
+    {
+      "epoch": 0.6677853946491479,
+      "grad_norm": 1.1011535811209758,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 28055
+    },
+    {
+      "epoch": 0.6678091973721794,
+      "grad_norm": 0.6837158274929855,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 28056
+    },
+    {
+      "epoch": 0.6678330000952108,
+      "grad_norm": 7.480127205832304,
+      "learning_rate": 2e-05,
+      "loss": 0.7765,
+      "step": 28057
+    },
+    {
+      "epoch": 0.6678568028182424,
+      "grad_norm": 4.15240826030495,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 28058
+    },
+    {
+      "epoch": 0.6678806055412739,
+      "grad_norm": 1.5713078081853162,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 28059
+    },
+    {
+      "epoch": 0.6679044082643054,
+      "grad_norm": 0.5024914175943872,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 28060
+    },
+    {
+      "epoch": 0.6679282109873369,
+      "grad_norm": 3.28024864807256,
+      "learning_rate": 2e-05,
+      "loss": 0.3392,
+      "step": 28061
+    },
+    {
+      "epoch": 0.6679520137103685,
+      "grad_norm": 2.3821455562583123,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 28062
+    },
+    {
+      "epoch": 0.6679758164334,
+      "grad_norm": 2.3013142362885626,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 28063
+    },
+    {
+      "epoch": 0.6679996191564315,
+      "grad_norm": 25.137529256951463,
+      "learning_rate": 2e-05,
+      "loss": 0.7605,
+      "step": 28064
+    },
+    {
+      "epoch": 0.6680234218794631,
+      "grad_norm": 2.5557619910082985,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 28065
+    },
+    {
+      "epoch": 0.6680472246024945,
+      "grad_norm": 3.8913120709440614,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 28066
+    },
+    {
+      "epoch": 0.668071027325526,
+      "grad_norm": 2.4940656099898995,
+      "learning_rate": 2e-05,
+      "loss": 0.349,
+      "step": 28067
+    },
+    {
+      "epoch": 0.6680948300485575,
+      "grad_norm": 5.516777535861509,
+      "learning_rate": 2e-05,
+      "loss": 1.0337,
+      "step": 28068
+    },
+    {
+      "epoch": 0.6681186327715891,
+      "grad_norm": 3.4961966055167677,
+      "learning_rate": 2e-05,
+      "loss": 0.5291,
+      "step": 28069
+    },
+    {
+      "epoch": 0.6681424354946206,
+      "grad_norm": 5.6955961282807355,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 28070
+    },
+    {
+      "epoch": 0.6681662382176521,
+      "grad_norm": 3.4170911051321387,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 28071
+    },
+    {
+      "epoch": 0.6681900409406836,
+      "grad_norm": 7.198550029633188,
+      "learning_rate": 2e-05,
+      "loss": 0.1699,
+      "step": 28072
+    },
+    {
+      "epoch": 0.6682138436637152,
+      "grad_norm": 0.1449259293200756,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 28073
+    },
+    {
+      "epoch": 0.6682376463867467,
+      "grad_norm": 2.359016247635572,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 28074
+    },
+    {
+      "epoch": 0.6682614491097781,
+      "grad_norm": 28.67651227561769,
+      "learning_rate": 2e-05,
+      "loss": 0.4554,
+      "step": 28075
+    },
+    {
+      "epoch": 0.6682852518328096,
+      "grad_norm": 8.302014060821023,
+      "learning_rate": 2e-05,
+      "loss": 0.2872,
+      "step": 28076
+    },
+    {
+      "epoch": 0.6683090545558412,
+      "grad_norm": 0.02115937366119716,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28077
+    },
+    {
+      "epoch": 0.6683328572788727,
+      "grad_norm": 3.252162752292735,
+      "learning_rate": 2e-05,
+      "loss": 0.4575,
+      "step": 28078
+    },
+    {
+      "epoch": 0.6683566600019042,
+      "grad_norm": 0.6332597128831564,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 28079
+    },
+    {
+      "epoch": 0.6683804627249358,
+      "grad_norm": 0.23710914467452768,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 28080
+    },
+    {
+      "epoch": 0.6684042654479673,
+      "grad_norm": 0.26378795087279516,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 28081
+    },
+    {
+      "epoch": 0.6684280681709988,
+      "grad_norm": 5.085685791921075,
+      "learning_rate": 2e-05,
+      "loss": 0.5958,
+      "step": 28082
+    },
+    {
+      "epoch": 0.6684518708940302,
+      "grad_norm": 5.732759219986804,
+      "learning_rate": 2e-05,
+      "loss": 0.1457,
+      "step": 28083
+    },
+    {
+      "epoch": 0.6684756736170618,
+      "grad_norm": 3.1031958578576053,
+      "learning_rate": 2e-05,
+      "loss": 0.3964,
+      "step": 28084
+    },
+    {
+      "epoch": 0.6684994763400933,
+      "grad_norm": 11.7776034599623,
+      "learning_rate": 2e-05,
+      "loss": 1.1225,
+      "step": 28085
+    },
+    {
+      "epoch": 0.6685232790631248,
+      "grad_norm": 1.3313174626415534,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 28086
+    },
+    {
+      "epoch": 0.6685470817861563,
+      "grad_norm": 10.034779631612361,
+      "learning_rate": 2e-05,
+      "loss": 1.1917,
+      "step": 28087
+    },
+    {
+      "epoch": 0.6685708845091879,
+      "grad_norm": 1.9843591237536007,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 28088
+    },
+    {
+      "epoch": 0.6685946872322194,
+      "grad_norm": 3.954939044080802,
+      "learning_rate": 2e-05,
+      "loss": 0.2247,
+      "step": 28089
+    },
+    {
+      "epoch": 0.6686184899552509,
+      "grad_norm": 2.960428028408347,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 28090
+    },
+    {
+      "epoch": 0.6686422926782823,
+      "grad_norm": 11.672492481170611,
+      "learning_rate": 2e-05,
+      "loss": 0.7678,
+      "step": 28091
+    },
+    {
+      "epoch": 0.6686660954013139,
+      "grad_norm": 0.5647410087403026,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 28092
+    },
+    {
+      "epoch": 0.6686898981243454,
+      "grad_norm": 1.64421967160731,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 28093
+    },
+    {
+      "epoch": 0.6687137008473769,
+      "grad_norm": 3.083859497396668,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 28094
+    },
+    {
+      "epoch": 0.6687375035704085,
+      "grad_norm": 5.198605408420382,
+      "learning_rate": 2e-05,
+      "loss": 0.6492,
+      "step": 28095
+    },
+    {
+      "epoch": 0.66876130629344,
+      "grad_norm": 3.2168353712638527,
+      "learning_rate": 2e-05,
+      "loss": 0.3521,
+      "step": 28096
+    },
+    {
+      "epoch": 0.6687851090164715,
+      "grad_norm": 0.2967655580079735,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 28097
+    },
+    {
+      "epoch": 0.668808911739503,
+      "grad_norm": 4.548426137120192,
+      "learning_rate": 2e-05,
+      "loss": 0.1504,
+      "step": 28098
+    },
+    {
+      "epoch": 0.6688327144625346,
+      "grad_norm": 1.9734921445963296,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 28099
+    },
+    {
+      "epoch": 0.668856517185566,
+      "grad_norm": 3.536458939421463,
+      "learning_rate": 2e-05,
+      "loss": 0.532,
+      "step": 28100
+    },
+    {
+      "epoch": 0.6688803199085975,
+      "grad_norm": 0.01928696464948164,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28101
+    },
+    {
+      "epoch": 0.668904122631629,
+      "grad_norm": 5.871867055870772,
+      "learning_rate": 2e-05,
+      "loss": 1.0184,
+      "step": 28102
+    },
+    {
+      "epoch": 0.6689279253546606,
+      "grad_norm": 4.355682356574125,
+      "learning_rate": 2e-05,
+      "loss": 0.946,
+      "step": 28103
+    },
+    {
+      "epoch": 0.6689517280776921,
+      "grad_norm": 1.8493949468793984,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 28104
+    },
+    {
+      "epoch": 0.6689755308007236,
+      "grad_norm": 0.5435683032541333,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 28105
+    },
+    {
+      "epoch": 0.6689993335237551,
+      "grad_norm": 7.066871731364942,
+      "learning_rate": 2e-05,
+      "loss": 1.5161,
+      "step": 28106
+    },
+    {
+      "epoch": 0.6690231362467867,
+      "grad_norm": 10.695118068340156,
+      "learning_rate": 2e-05,
+      "loss": 0.4638,
+      "step": 28107
+    },
+    {
+      "epoch": 0.6690469389698182,
+      "grad_norm": 2.0007311156186853,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 28108
+    },
+    {
+      "epoch": 0.6690707416928496,
+      "grad_norm": 3.6977785047052136,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 28109
+    },
+    {
+      "epoch": 0.6690945444158812,
+      "grad_norm": 4.6619661416083025,
+      "learning_rate": 2e-05,
+      "loss": 0.5991,
+      "step": 28110
+    },
+    {
+      "epoch": 0.6691183471389127,
+      "grad_norm": 7.434362196882398,
+      "learning_rate": 2e-05,
+      "loss": 0.9617,
+      "step": 28111
+    },
+    {
+      "epoch": 0.6691421498619442,
+      "grad_norm": 5.536037535247755,
+      "learning_rate": 2e-05,
+      "loss": 0.7158,
+      "step": 28112
+    },
+    {
+      "epoch": 0.6691659525849757,
+      "grad_norm": 1.234314363532631,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 28113
+    },
+    {
+      "epoch": 0.6691897553080073,
+      "grad_norm": 2.0829774406455366,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 28114
+    },
+    {
+      "epoch": 0.6692135580310388,
+      "grad_norm": 0.07721721431578839,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 28115
+    },
+    {
+      "epoch": 0.6692373607540703,
+      "grad_norm": 3.0561930806974744,
+      "learning_rate": 2e-05,
+      "loss": 0.133,
+      "step": 28116
+    },
+    {
+      "epoch": 0.6692611634771017,
+      "grad_norm": 4.710065092908894,
+      "learning_rate": 2e-05,
+      "loss": 0.77,
+      "step": 28117
+    },
+    {
+      "epoch": 0.6692849662001333,
+      "grad_norm": 1.4839043245245713,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 28118
+    },
+    {
+      "epoch": 0.6693087689231648,
+      "grad_norm": 1.547700281235002,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 28119
+    },
+    {
+      "epoch": 0.6693325716461963,
+      "grad_norm": 8.36050335606743,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 28120
+    },
+    {
+      "epoch": 0.6693563743692278,
+      "grad_norm": 0.4221603886426567,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 28121
+    },
+    {
+      "epoch": 0.6693801770922594,
+      "grad_norm": 1.7261489054009584,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 28122
+    },
+    {
+      "epoch": 0.6694039798152909,
+      "grad_norm": 3.670662257346996,
+      "learning_rate": 2e-05,
+      "loss": 0.4462,
+      "step": 28123
+    },
+    {
+      "epoch": 0.6694277825383224,
+      "grad_norm": 0.5928333348706158,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 28124
+    },
+    {
+      "epoch": 0.669451585261354,
+      "grad_norm": 3.1484767793443473,
+      "learning_rate": 2e-05,
+      "loss": 0.786,
+      "step": 28125
+    },
+    {
+      "epoch": 0.6694753879843854,
+      "grad_norm": 4.071583158065157,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 28126
+    },
+    {
+      "epoch": 0.6694991907074169,
+      "grad_norm": 8.48452143654963,
+      "learning_rate": 2e-05,
+      "loss": 1.1401,
+      "step": 28127
+    },
+    {
+      "epoch": 0.6695229934304484,
+      "grad_norm": 7.490911489712221,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 28128
+    },
+    {
+      "epoch": 0.66954679615348,
+      "grad_norm": 7.806167373205406,
+      "learning_rate": 2e-05,
+      "loss": 1.1182,
+      "step": 28129
+    },
+    {
+      "epoch": 0.6695705988765115,
+      "grad_norm": 0.06302706742664745,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 28130
+    },
+    {
+      "epoch": 0.669594401599543,
+      "grad_norm": 0.5372691489904791,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 28131
+    },
+    {
+      "epoch": 0.6696182043225745,
+      "grad_norm": 0.14557972508117797,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 28132
+    },
+    {
+      "epoch": 0.6696420070456061,
+      "grad_norm": 12.920819866706642,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 28133
+    },
+    {
+      "epoch": 0.6696658097686375,
+      "grad_norm": 0.5384575283908719,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 28134
+    },
+    {
+      "epoch": 0.669689612491669,
+      "grad_norm": 5.278545485408601,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 28135
+    },
+    {
+      "epoch": 0.6697134152147005,
+      "grad_norm": 1.8693913047687063,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 28136
+    },
+    {
+      "epoch": 0.6697372179377321,
+      "grad_norm": 4.8167868652487424,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 28137
+    },
+    {
+      "epoch": 0.6697610206607636,
+      "grad_norm": 4.7473892600030485,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 28138
+    },
+    {
+      "epoch": 0.6697848233837951,
+      "grad_norm": 0.055544646223886024,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 28139
+    },
+    {
+      "epoch": 0.6698086261068266,
+      "grad_norm": 3.4897349932519255,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 28140
+    },
+    {
+      "epoch": 0.6698324288298582,
+      "grad_norm": 11.440063383096087,
+      "learning_rate": 2e-05,
+      "loss": 0.4946,
+      "step": 28141
+    },
+    {
+      "epoch": 0.6698562315528896,
+      "grad_norm": 6.03544198633537,
+      "learning_rate": 2e-05,
+      "loss": 0.4405,
+      "step": 28142
+    },
+    {
+      "epoch": 0.6698800342759211,
+      "grad_norm": 1.020577865211575,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 28143
+    },
+    {
+      "epoch": 0.6699038369989527,
+      "grad_norm": 0.05382805925220367,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 28144
+    },
+    {
+      "epoch": 0.6699276397219842,
+      "grad_norm": 0.1485736083197211,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 28145
+    },
+    {
+      "epoch": 0.6699514424450157,
+      "grad_norm": 8.379748320057628,
+      "learning_rate": 2e-05,
+      "loss": 0.8761,
+      "step": 28146
+    },
+    {
+      "epoch": 0.6699752451680472,
+      "grad_norm": 1.995592456617868,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 28147
+    },
+    {
+      "epoch": 0.6699990478910788,
+      "grad_norm": 11.042758325032446,
+      "learning_rate": 2e-05,
+      "loss": 0.6774,
+      "step": 28148
+    },
+    {
+      "epoch": 0.6700228506141103,
+      "grad_norm": 10.882284056954306,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 28149
+    },
+    {
+      "epoch": 0.6700466533371418,
+      "grad_norm": 1.8572463743208925,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 28150
+    },
+    {
+      "epoch": 0.6700704560601732,
+      "grad_norm": 4.733907533932002,
+      "learning_rate": 2e-05,
+      "loss": 1.1366,
+      "step": 28151
+    },
+    {
+      "epoch": 0.6700942587832048,
+      "grad_norm": 0.3212455895646919,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 28152
+    },
+    {
+      "epoch": 0.6701180615062363,
+      "grad_norm": 0.4449213533936686,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 28153
+    },
+    {
+      "epoch": 0.6701418642292678,
+      "grad_norm": 5.149449055750408,
+      "learning_rate": 2e-05,
+      "loss": 1.3737,
+      "step": 28154
+    },
+    {
+      "epoch": 0.6701656669522993,
+      "grad_norm": 6.68707894996141,
+      "learning_rate": 2e-05,
+      "loss": 1.1627,
+      "step": 28155
+    },
+    {
+      "epoch": 0.6701894696753309,
+      "grad_norm": 1.2939214960479304,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 28156
+    },
+    {
+      "epoch": 0.6702132723983624,
+      "grad_norm": 1.3307547723073263,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 28157
+    },
+    {
+      "epoch": 0.6702370751213939,
+      "grad_norm": 0.5912729581877455,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 28158
+    },
+    {
+      "epoch": 0.6702608778444255,
+      "grad_norm": 3.7523295449277327,
+      "learning_rate": 2e-05,
+      "loss": 0.6722,
+      "step": 28159
+    },
+    {
+      "epoch": 0.6702846805674569,
+      "grad_norm": 5.7531437737136795,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 28160
+    },
+    {
+      "epoch": 0.6703084832904884,
+      "grad_norm": 2.653467750737904,
+      "learning_rate": 2e-05,
+      "loss": 0.4023,
+      "step": 28161
+    },
+    {
+      "epoch": 0.6703322860135199,
+      "grad_norm": 9.97377256660196,
+      "learning_rate": 2e-05,
+      "loss": 0.4412,
+      "step": 28162
+    },
+    {
+      "epoch": 0.6703560887365515,
+      "grad_norm": 0.013738731281511535,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28163
+    },
+    {
+      "epoch": 0.670379891459583,
+      "grad_norm": 0.09760729560979178,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 28164
+    },
+    {
+      "epoch": 0.6704036941826145,
+      "grad_norm": 5.135075485586634,
+      "learning_rate": 2e-05,
+      "loss": 0.6659,
+      "step": 28165
+    },
+    {
+      "epoch": 0.670427496905646,
+      "grad_norm": 9.825557625265665,
+      "learning_rate": 2e-05,
+      "loss": 0.3589,
+      "step": 28166
+    },
+    {
+      "epoch": 0.6704512996286776,
+      "grad_norm": 4.920939196657386,
+      "learning_rate": 2e-05,
+      "loss": 0.7347,
+      "step": 28167
+    },
+    {
+      "epoch": 0.670475102351709,
+      "grad_norm": 11.209225899140153,
+      "learning_rate": 2e-05,
+      "loss": 1.1565,
+      "step": 28168
+    },
+    {
+      "epoch": 0.6704989050747405,
+      "grad_norm": 8.161094757203664,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 28169
+    },
+    {
+      "epoch": 0.670522707797772,
+      "grad_norm": 2.7845243005728113,
+      "learning_rate": 2e-05,
+      "loss": 0.4782,
+      "step": 28170
+    },
+    {
+      "epoch": 0.6705465105208036,
+      "grad_norm": 1.3360373783830863,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 28171
+    },
+    {
+      "epoch": 0.6705703132438351,
+      "grad_norm": 1.6246292103297666,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 28172
+    },
+    {
+      "epoch": 0.6705941159668666,
+      "grad_norm": 2.88150436286269,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 28173
+    },
+    {
+      "epoch": 0.6706179186898982,
+      "grad_norm": 1.3469036382411037,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 28174
+    },
+    {
+      "epoch": 0.6706417214129297,
+      "grad_norm": 6.50196070167098,
+      "learning_rate": 2e-05,
+      "loss": 0.6424,
+      "step": 28175
+    },
+    {
+      "epoch": 0.6706655241359611,
+      "grad_norm": 1.8662171215427135,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 28176
+    },
+    {
+      "epoch": 0.6706893268589926,
+      "grad_norm": 7.471189383066563,
+      "learning_rate": 2e-05,
+      "loss": 2.0286,
+      "step": 28177
+    },
+    {
+      "epoch": 0.6707131295820242,
+      "grad_norm": 6.318341158170281,
+      "learning_rate": 2e-05,
+      "loss": 0.2486,
+      "step": 28178
+    },
+    {
+      "epoch": 0.6707369323050557,
+      "grad_norm": 1.047544047822121,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 28179
+    },
+    {
+      "epoch": 0.6707607350280872,
+      "grad_norm": 3.877153132833163,
+      "learning_rate": 2e-05,
+      "loss": 0.5657,
+      "step": 28180
+    },
+    {
+      "epoch": 0.6707845377511187,
+      "grad_norm": 7.20602824145244,
+      "learning_rate": 2e-05,
+      "loss": 1.0691,
+      "step": 28181
+    },
+    {
+      "epoch": 0.6708083404741503,
+      "grad_norm": 3.2380785651448454,
+      "learning_rate": 2e-05,
+      "loss": 0.4617,
+      "step": 28182
+    },
+    {
+      "epoch": 0.6708321431971818,
+      "grad_norm": 8.428877847082163,
+      "learning_rate": 2e-05,
+      "loss": 2.0814,
+      "step": 28183
+    },
+    {
+      "epoch": 0.6708559459202132,
+      "grad_norm": 2.8009072510978825,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 28184
+    },
+    {
+      "epoch": 0.6708797486432447,
+      "grad_norm": 4.692543522465676,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 28185
+    },
+    {
+      "epoch": 0.6709035513662763,
+      "grad_norm": 2.8194902896860867,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 28186
+    },
+    {
+      "epoch": 0.6709273540893078,
+      "grad_norm": 0.43287708332564373,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 28187
+    },
+    {
+      "epoch": 0.6709511568123393,
+      "grad_norm": 11.054974103118328,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 28188
+    },
+    {
+      "epoch": 0.6709749595353709,
+      "grad_norm": 11.960140387084019,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 28189
+    },
+    {
+      "epoch": 0.6709987622584024,
+      "grad_norm": 4.873404253822399,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 28190
+    },
+    {
+      "epoch": 0.6710225649814339,
+      "grad_norm": 5.1432483574506875,
+      "learning_rate": 2e-05,
+      "loss": 1.1951,
+      "step": 28191
+    },
+    {
+      "epoch": 0.6710463677044654,
+      "grad_norm": 0.5965890613939171,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 28192
+    },
+    {
+      "epoch": 0.671070170427497,
+      "grad_norm": 5.050558169537571,
+      "learning_rate": 2e-05,
+      "loss": 0.6348,
+      "step": 28193
+    },
+    {
+      "epoch": 0.6710939731505284,
+      "grad_norm": 8.108952532232353,
+      "learning_rate": 2e-05,
+      "loss": 1.2634,
+      "step": 28194
+    },
+    {
+      "epoch": 0.6711177758735599,
+      "grad_norm": 7.584328504303277,
+      "learning_rate": 2e-05,
+      "loss": 0.6109,
+      "step": 28195
+    },
+    {
+      "epoch": 0.6711415785965914,
+      "grad_norm": 3.744557963464495,
+      "learning_rate": 2e-05,
+      "loss": 0.8725,
+      "step": 28196
+    },
+    {
+      "epoch": 0.671165381319623,
+      "grad_norm": 5.568928713459572,
+      "learning_rate": 2e-05,
+      "loss": 0.223,
+      "step": 28197
+    },
+    {
+      "epoch": 0.6711891840426545,
+      "grad_norm": 7.932044926416345,
+      "learning_rate": 2e-05,
+      "loss": 0.731,
+      "step": 28198
+    },
+    {
+      "epoch": 0.671212986765686,
+      "grad_norm": 5.180803536115384,
+      "learning_rate": 2e-05,
+      "loss": 0.9378,
+      "step": 28199
+    },
+    {
+      "epoch": 0.6712367894887175,
+      "grad_norm": 7.02391440258443,
+      "learning_rate": 2e-05,
+      "loss": 0.2695,
+      "step": 28200
+    },
+    {
+      "epoch": 0.671260592211749,
+      "grad_norm": 7.1294781252317225,
+      "learning_rate": 2e-05,
+      "loss": 0.1559,
+      "step": 28201
+    },
+    {
+      "epoch": 0.6712843949347805,
+      "grad_norm": 7.7794407387361,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 28202
+    },
+    {
+      "epoch": 0.671308197657812,
+      "grad_norm": 5.0900559478117735,
+      "learning_rate": 2e-05,
+      "loss": 1.1053,
+      "step": 28203
+    },
+    {
+      "epoch": 0.6713320003808436,
+      "grad_norm": 5.2080099840128495,
+      "learning_rate": 2e-05,
+      "loss": 1.4875,
+      "step": 28204
+    },
+    {
+      "epoch": 0.6713558031038751,
+      "grad_norm": 0.5324919494214355,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 28205
+    },
+    {
+      "epoch": 0.6713796058269066,
+      "grad_norm": 7.5932765135124605,
+      "learning_rate": 2e-05,
+      "loss": 0.2107,
+      "step": 28206
+    },
+    {
+      "epoch": 0.6714034085499381,
+      "grad_norm": 1.8746061041273778,
+      "learning_rate": 2e-05,
+      "loss": 0.161,
+      "step": 28207
+    },
+    {
+      "epoch": 0.6714272112729697,
+      "grad_norm": 2.4866218955943813,
+      "learning_rate": 2e-05,
+      "loss": 0.3656,
+      "step": 28208
+    },
+    {
+      "epoch": 0.6714510139960012,
+      "grad_norm": 5.549355298339304,
+      "learning_rate": 2e-05,
+      "loss": 0.7898,
+      "step": 28209
+    },
+    {
+      "epoch": 0.6714748167190326,
+      "grad_norm": 0.06781004404649144,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 28210
+    },
+    {
+      "epoch": 0.6714986194420641,
+      "grad_norm": 1.6965113991345215,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 28211
+    },
+    {
+      "epoch": 0.6715224221650957,
+      "grad_norm": 5.524506753322739,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 28212
+    },
+    {
+      "epoch": 0.6715462248881272,
+      "grad_norm": 4.4704053363116385,
+      "learning_rate": 2e-05,
+      "loss": 0.5529,
+      "step": 28213
+    },
+    {
+      "epoch": 0.6715700276111587,
+      "grad_norm": 3.7632395703906933,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 28214
+    },
+    {
+      "epoch": 0.6715938303341902,
+      "grad_norm": 2.895974288192358,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 28215
+    },
+    {
+      "epoch": 0.6716176330572218,
+      "grad_norm": 6.195048421085983,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 28216
+    },
+    {
+      "epoch": 0.6716414357802533,
+      "grad_norm": 9.101434347215365,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 28217
+    },
+    {
+      "epoch": 0.6716652385032847,
+      "grad_norm": 1.0110815074591273,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 28218
+    },
+    {
+      "epoch": 0.6716890412263163,
+      "grad_norm": 3.88864583633093,
+      "learning_rate": 2e-05,
+      "loss": 0.6135,
+      "step": 28219
+    },
+    {
+      "epoch": 0.6717128439493478,
+      "grad_norm": 8.891904203716773,
+      "learning_rate": 2e-05,
+      "loss": 0.5537,
+      "step": 28220
+    },
+    {
+      "epoch": 0.6717366466723793,
+      "grad_norm": 4.983927796661541,
+      "learning_rate": 2e-05,
+      "loss": 0.2638,
+      "step": 28221
+    },
+    {
+      "epoch": 0.6717604493954108,
+      "grad_norm": 3.2152553156640886,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 28222
+    },
+    {
+      "epoch": 0.6717842521184424,
+      "grad_norm": 3.8096480104857258,
+      "learning_rate": 2e-05,
+      "loss": 0.4095,
+      "step": 28223
+    },
+    {
+      "epoch": 0.6718080548414739,
+      "grad_norm": 1.8919502017236367,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 28224
+    },
+    {
+      "epoch": 0.6718318575645054,
+      "grad_norm": 0.08937193590346151,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28225
+    },
+    {
+      "epoch": 0.6718556602875368,
+      "grad_norm": 1.6984001605808188,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 28226
+    },
+    {
+      "epoch": 0.6718794630105684,
+      "grad_norm": 2.784332852764442,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 28227
+    },
+    {
+      "epoch": 0.6719032657335999,
+      "grad_norm": 0.2523842288614802,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 28228
+    },
+    {
+      "epoch": 0.6719270684566314,
+      "grad_norm": 0.44907666594353596,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 28229
+    },
+    {
+      "epoch": 0.6719508711796629,
+      "grad_norm": 2.9351502710985846,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 28230
+    },
+    {
+      "epoch": 0.6719746739026945,
+      "grad_norm": 0.20413481935869354,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 28231
+    },
+    {
+      "epoch": 0.671998476625726,
+      "grad_norm": 7.070613997175547,
+      "learning_rate": 2e-05,
+      "loss": 1.6306,
+      "step": 28232
+    },
+    {
+      "epoch": 0.6720222793487575,
+      "grad_norm": 0.7178779029397371,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 28233
+    },
+    {
+      "epoch": 0.6720460820717891,
+      "grad_norm": 9.183309838317243,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 28234
+    },
+    {
+      "epoch": 0.6720698847948205,
+      "grad_norm": 4.142926037820829,
+      "learning_rate": 2e-05,
+      "loss": 1.1282,
+      "step": 28235
+    },
+    {
+      "epoch": 0.672093687517852,
+      "grad_norm": 3.4183111152736756,
+      "learning_rate": 2e-05,
+      "loss": 0.2863,
+      "step": 28236
+    },
+    {
+      "epoch": 0.6721174902408835,
+      "grad_norm": 3.687804933818606,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 28237
+    },
+    {
+      "epoch": 0.6721412929639151,
+      "grad_norm": 2.1858405398981766,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 28238
+    },
+    {
+      "epoch": 0.6721650956869466,
+      "grad_norm": 0.6047215634547308,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 28239
+    },
+    {
+      "epoch": 0.6721888984099781,
+      "grad_norm": 1.2368981110245683,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 28240
+    },
+    {
+      "epoch": 0.6722127011330096,
+      "grad_norm": 2.8337971206244363,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 28241
+    },
+    {
+      "epoch": 0.6722365038560412,
+      "grad_norm": 0.8029621685589131,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 28242
+    },
+    {
+      "epoch": 0.6722603065790727,
+      "grad_norm": 1.6296828460895025,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 28243
+    },
+    {
+      "epoch": 0.6722841093021041,
+      "grad_norm": 0.189422483040902,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 28244
+    },
+    {
+      "epoch": 0.6723079120251356,
+      "grad_norm": 1.8469156686302954,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 28245
+    },
+    {
+      "epoch": 0.6723317147481672,
+      "grad_norm": 3.3413128560510557,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 28246
+    },
+    {
+      "epoch": 0.6723555174711987,
+      "grad_norm": 2.1473864440746646,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 28247
+    },
+    {
+      "epoch": 0.6723793201942302,
+      "grad_norm": 8.293415438674597,
+      "learning_rate": 2e-05,
+      "loss": 0.8359,
+      "step": 28248
+    },
+    {
+      "epoch": 0.6724031229172618,
+      "grad_norm": 0.06857631715556203,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 28249
+    },
+    {
+      "epoch": 0.6724269256402933,
+      "grad_norm": 4.774127562023167,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 28250
+    },
+    {
+      "epoch": 0.6724507283633248,
+      "grad_norm": 1.8564182115531365,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 28251
+    },
+    {
+      "epoch": 0.6724745310863562,
+      "grad_norm": 5.727772883368648,
+      "learning_rate": 2e-05,
+      "loss": 0.5109,
+      "step": 28252
+    },
+    {
+      "epoch": 0.6724983338093878,
+      "grad_norm": 7.867985537255765,
+      "learning_rate": 2e-05,
+      "loss": 0.9703,
+      "step": 28253
+    },
+    {
+      "epoch": 0.6725221365324193,
+      "grad_norm": 3.838535753155597,
+      "learning_rate": 2e-05,
+      "loss": 0.4393,
+      "step": 28254
+    },
+    {
+      "epoch": 0.6725459392554508,
+      "grad_norm": 9.158861226249588,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 28255
+    },
+    {
+      "epoch": 0.6725697419784823,
+      "grad_norm": 19.140615623697954,
+      "learning_rate": 2e-05,
+      "loss": 0.3778,
+      "step": 28256
+    },
+    {
+      "epoch": 0.6725935447015139,
+      "grad_norm": 2.3277262849526963,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 28257
+    },
+    {
+      "epoch": 0.6726173474245454,
+      "grad_norm": 3.920220395017467,
+      "learning_rate": 2e-05,
+      "loss": 0.7186,
+      "step": 28258
+    },
+    {
+      "epoch": 0.6726411501475769,
+      "grad_norm": 3.3792388088635255,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 28259
+    },
+    {
+      "epoch": 0.6726649528706083,
+      "grad_norm": 4.556760773677859,
+      "learning_rate": 2e-05,
+      "loss": 0.6212,
+      "step": 28260
+    },
+    {
+      "epoch": 0.6726887555936399,
+      "grad_norm": 3.021662579948256,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 28261
+    },
+    {
+      "epoch": 0.6727125583166714,
+      "grad_norm": 5.008857578273345,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 28262
+    },
+    {
+      "epoch": 0.6727363610397029,
+      "grad_norm": 0.6606971111017323,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 28263
+    },
+    {
+      "epoch": 0.6727601637627345,
+      "grad_norm": 5.2005223970421035,
+      "learning_rate": 2e-05,
+      "loss": 0.529,
+      "step": 28264
+    },
+    {
+      "epoch": 0.672783966485766,
+      "grad_norm": 7.605599959070504,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 28265
+    },
+    {
+      "epoch": 0.6728077692087975,
+      "grad_norm": 3.7212062018051277,
+      "learning_rate": 2e-05,
+      "loss": 0.5702,
+      "step": 28266
+    },
+    {
+      "epoch": 0.672831571931829,
+      "grad_norm": 11.413403504908768,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 28267
+    },
+    {
+      "epoch": 0.6728553746548606,
+      "grad_norm": 30.421621942524876,
+      "learning_rate": 2e-05,
+      "loss": 0.8401,
+      "step": 28268
+    },
+    {
+      "epoch": 0.672879177377892,
+      "grad_norm": 4.357017783263967,
+      "learning_rate": 2e-05,
+      "loss": 0.8911,
+      "step": 28269
+    },
+    {
+      "epoch": 0.6729029801009235,
+      "grad_norm": 2.085741714655458,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 28270
+    },
+    {
+      "epoch": 0.672926782823955,
+      "grad_norm": 2.5366444248066804,
+      "learning_rate": 2e-05,
+      "loss": 0.4023,
+      "step": 28271
+    },
+    {
+      "epoch": 0.6729505855469866,
+      "grad_norm": 0.3295706521158244,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 28272
+    },
+    {
+      "epoch": 0.6729743882700181,
+      "grad_norm": 4.695908280923148,
+      "learning_rate": 2e-05,
+      "loss": 0.1046,
+      "step": 28273
+    },
+    {
+      "epoch": 0.6729981909930496,
+      "grad_norm": 13.697905386627848,
+      "learning_rate": 2e-05,
+      "loss": 0.2752,
+      "step": 28274
+    },
+    {
+      "epoch": 0.6730219937160811,
+      "grad_norm": 0.42409037970710023,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 28275
+    },
+    {
+      "epoch": 0.6730457964391127,
+      "grad_norm": 0.7321007367948592,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 28276
+    },
+    {
+      "epoch": 0.6730695991621441,
+      "grad_norm": 8.402515234238491,
+      "learning_rate": 2e-05,
+      "loss": 0.4297,
+      "step": 28277
+    },
+    {
+      "epoch": 0.6730934018851756,
+      "grad_norm": 6.392969675607877,
+      "learning_rate": 2e-05,
+      "loss": 0.1605,
+      "step": 28278
+    },
+    {
+      "epoch": 0.6731172046082072,
+      "grad_norm": 5.102695816041631,
+      "learning_rate": 2e-05,
+      "loss": 0.4253,
+      "step": 28279
+    },
+    {
+      "epoch": 0.6731410073312387,
+      "grad_norm": 6.5275677191267985,
+      "learning_rate": 2e-05,
+      "loss": 1.432,
+      "step": 28280
+    },
+    {
+      "epoch": 0.6731648100542702,
+      "grad_norm": 19.881768580878717,
+      "learning_rate": 2e-05,
+      "loss": 0.6321,
+      "step": 28281
+    },
+    {
+      "epoch": 0.6731886127773017,
+      "grad_norm": 0.3826446230598738,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 28282
+    },
+    {
+      "epoch": 0.6732124155003333,
+      "grad_norm": 3.4081874486014336,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 28283
+    },
+    {
+      "epoch": 0.6732362182233648,
+      "grad_norm": 3.8327517076114423,
+      "learning_rate": 2e-05,
+      "loss": 0.0722,
+      "step": 28284
+    },
+    {
+      "epoch": 0.6732600209463963,
+      "grad_norm": 4.352639222404787,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 28285
+    },
+    {
+      "epoch": 0.6732838236694277,
+      "grad_norm": 0.9243434831560292,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 28286
+    },
+    {
+      "epoch": 0.6733076263924593,
+      "grad_norm": 1.4428929824504444,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 28287
+    },
+    {
+      "epoch": 0.6733314291154908,
+      "grad_norm": 0.04801840273976002,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 28288
+    },
+    {
+      "epoch": 0.6733552318385223,
+      "grad_norm": 12.340189162922742,
+      "learning_rate": 2e-05,
+      "loss": 0.3667,
+      "step": 28289
+    },
+    {
+      "epoch": 0.6733790345615538,
+      "grad_norm": 12.272343751527155,
+      "learning_rate": 2e-05,
+      "loss": 0.6246,
+      "step": 28290
+    },
+    {
+      "epoch": 0.6734028372845854,
+      "grad_norm": 1.1568415701588899,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 28291
+    },
+    {
+      "epoch": 0.6734266400076169,
+      "grad_norm": 2.001260384376977,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 28292
+    },
+    {
+      "epoch": 0.6734504427306484,
+      "grad_norm": 6.612763245992077,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 28293
+    },
+    {
+      "epoch": 0.67347424545368,
+      "grad_norm": 1.0917313476493464,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 28294
+    },
+    {
+      "epoch": 0.6734980481767114,
+      "grad_norm": 2.8443037270154123,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 28295
+    },
+    {
+      "epoch": 0.6735218508997429,
+      "grad_norm": 4.315345610023733,
+      "learning_rate": 2e-05,
+      "loss": 0.9166,
+      "step": 28296
+    },
+    {
+      "epoch": 0.6735456536227744,
+      "grad_norm": 0.4517200250790166,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 28297
+    },
+    {
+      "epoch": 0.673569456345806,
+      "grad_norm": 4.531868945121623,
+      "learning_rate": 2e-05,
+      "loss": 0.1461,
+      "step": 28298
+    },
+    {
+      "epoch": 0.6735932590688375,
+      "grad_norm": 8.459707586022473,
+      "learning_rate": 2e-05,
+      "loss": 1.0671,
+      "step": 28299
+    },
+    {
+      "epoch": 0.673617061791869,
+      "grad_norm": 1.2799701879213543,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 28300
+    },
+    {
+      "epoch": 0.6736408645149005,
+      "grad_norm": 1.651281122687523,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 28301
+    },
+    {
+      "epoch": 0.6736646672379321,
+      "grad_norm": 5.0617170577804504,
+      "learning_rate": 2e-05,
+      "loss": 0.5231,
+      "step": 28302
+    },
+    {
+      "epoch": 0.6736884699609635,
+      "grad_norm": 4.228498457010466,
+      "learning_rate": 2e-05,
+      "loss": 0.5171,
+      "step": 28303
+    },
+    {
+      "epoch": 0.673712272683995,
+      "grad_norm": 4.453883904772558,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 28304
+    },
+    {
+      "epoch": 0.6737360754070265,
+      "grad_norm": 4.337731413866443,
+      "learning_rate": 2e-05,
+      "loss": 0.6095,
+      "step": 28305
+    },
+    {
+      "epoch": 0.6737598781300581,
+      "grad_norm": 0.5666404919969725,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 28306
+    },
+    {
+      "epoch": 0.6737836808530896,
+      "grad_norm": 6.578128002885398,
+      "learning_rate": 2e-05,
+      "loss": 0.6737,
+      "step": 28307
+    },
+    {
+      "epoch": 0.6738074835761211,
+      "grad_norm": 4.198762628220974,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 28308
+    },
+    {
+      "epoch": 0.6738312862991527,
+      "grad_norm": 4.667701070586914,
+      "learning_rate": 2e-05,
+      "loss": 0.5565,
+      "step": 28309
+    },
+    {
+      "epoch": 0.6738550890221842,
+      "grad_norm": 5.221884934884306,
+      "learning_rate": 2e-05,
+      "loss": 0.5804,
+      "step": 28310
+    },
+    {
+      "epoch": 0.6738788917452156,
+      "grad_norm": 0.060251395303537555,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 28311
+    },
+    {
+      "epoch": 0.6739026944682471,
+      "grad_norm": 1.1008338330351959,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 28312
+    },
+    {
+      "epoch": 0.6739264971912787,
+      "grad_norm": 2.2066927087951065,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 28313
+    },
+    {
+      "epoch": 0.6739502999143102,
+      "grad_norm": 1.7844445260584727,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 28314
+    },
+    {
+      "epoch": 0.6739741026373417,
+      "grad_norm": 2.7649498530603633,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 28315
+    },
+    {
+      "epoch": 0.6739979053603732,
+      "grad_norm": 6.538859229548388,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 28316
+    },
+    {
+      "epoch": 0.6740217080834048,
+      "grad_norm": 4.382335379095999,
+      "learning_rate": 2e-05,
+      "loss": 0.8135,
+      "step": 28317
+    },
+    {
+      "epoch": 0.6740455108064363,
+      "grad_norm": 3.203841467470701,
+      "learning_rate": 2e-05,
+      "loss": 0.5195,
+      "step": 28318
+    },
+    {
+      "epoch": 0.6740693135294678,
+      "grad_norm": 1.020220824880558,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 28319
+    },
+    {
+      "epoch": 0.6740931162524992,
+      "grad_norm": 0.44646981680524306,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 28320
+    },
+    {
+      "epoch": 0.6741169189755308,
+      "grad_norm": 23.355816807785633,
+      "learning_rate": 2e-05,
+      "loss": 0.5026,
+      "step": 28321
+    },
+    {
+      "epoch": 0.6741407216985623,
+      "grad_norm": 1.8201983141918454,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 28322
+    },
+    {
+      "epoch": 0.6741645244215938,
+      "grad_norm": 10.163949612718012,
+      "learning_rate": 2e-05,
+      "loss": 0.2538,
+      "step": 28323
+    },
+    {
+      "epoch": 0.6741883271446254,
+      "grad_norm": 3.7525378272728487,
+      "learning_rate": 2e-05,
+      "loss": 0.4347,
+      "step": 28324
+    },
+    {
+      "epoch": 0.6742121298676569,
+      "grad_norm": 4.209134838617992,
+      "learning_rate": 2e-05,
+      "loss": 0.3837,
+      "step": 28325
+    },
+    {
+      "epoch": 0.6742359325906884,
+      "grad_norm": 3.126441845188481,
+      "learning_rate": 2e-05,
+      "loss": 0.5878,
+      "step": 28326
+    },
+    {
+      "epoch": 0.6742597353137199,
+      "grad_norm": 2.839603884010708,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 28327
+    },
+    {
+      "epoch": 0.6742835380367515,
+      "grad_norm": 3.0487088153391335,
+      "learning_rate": 2e-05,
+      "loss": 0.2553,
+      "step": 28328
+    },
+    {
+      "epoch": 0.6743073407597829,
+      "grad_norm": 3.3579212130655134,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 28329
+    },
+    {
+      "epoch": 0.6743311434828144,
+      "grad_norm": 0.3711204045602311,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 28330
+    },
+    {
+      "epoch": 0.6743549462058459,
+      "grad_norm": 2.1520436026138245,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 28331
+    },
+    {
+      "epoch": 0.6743787489288775,
+      "grad_norm": 2.6382322288356,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 28332
+    },
+    {
+      "epoch": 0.674402551651909,
+      "grad_norm": 3.3767946316855832,
+      "learning_rate": 2e-05,
+      "loss": 0.0856,
+      "step": 28333
+    },
+    {
+      "epoch": 0.6744263543749405,
+      "grad_norm": 0.57555120020273,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 28334
+    },
+    {
+      "epoch": 0.674450157097972,
+      "grad_norm": 7.702162205297356,
+      "learning_rate": 2e-05,
+      "loss": 1.704,
+      "step": 28335
+    },
+    {
+      "epoch": 0.6744739598210036,
+      "grad_norm": 0.1038136635434308,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 28336
+    },
+    {
+      "epoch": 0.674497762544035,
+      "grad_norm": 3.046601247730508,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 28337
+    },
+    {
+      "epoch": 0.6745215652670665,
+      "grad_norm": 1.2804048105894639,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 28338
+    },
+    {
+      "epoch": 0.6745453679900981,
+      "grad_norm": 9.089319393707282,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 28339
+    },
+    {
+      "epoch": 0.6745691707131296,
+      "grad_norm": 0.7815042593327562,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 28340
+    },
+    {
+      "epoch": 0.6745929734361611,
+      "grad_norm": 0.6358568329664206,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 28341
+    },
+    {
+      "epoch": 0.6746167761591926,
+      "grad_norm": 1.5956423654346503,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 28342
+    },
+    {
+      "epoch": 0.6746405788822242,
+      "grad_norm": 6.525399020704509,
+      "learning_rate": 2e-05,
+      "loss": 0.8274,
+      "step": 28343
+    },
+    {
+      "epoch": 0.6746643816052557,
+      "grad_norm": 4.224554852874869,
+      "learning_rate": 2e-05,
+      "loss": 0.4534,
+      "step": 28344
+    },
+    {
+      "epoch": 0.6746881843282871,
+      "grad_norm": 11.774632201852276,
+      "learning_rate": 2e-05,
+      "loss": 1.9184,
+      "step": 28345
+    },
+    {
+      "epoch": 0.6747119870513186,
+      "grad_norm": 0.08108967536771162,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28346
+    },
+    {
+      "epoch": 0.6747357897743502,
+      "grad_norm": 5.187378981669869,
+      "learning_rate": 2e-05,
+      "loss": 1.3253,
+      "step": 28347
+    },
+    {
+      "epoch": 0.6747595924973817,
+      "grad_norm": 9.515152975328352,
+      "learning_rate": 2e-05,
+      "loss": 0.8692,
+      "step": 28348
+    },
+    {
+      "epoch": 0.6747833952204132,
+      "grad_norm": 0.46526625240657976,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 28349
+    },
+    {
+      "epoch": 0.6748071979434447,
+      "grad_norm": 0.27956859396822953,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 28350
+    },
+    {
+      "epoch": 0.6748310006664763,
+      "grad_norm": 5.422341617660883,
+      "learning_rate": 2e-05,
+      "loss": 0.8417,
+      "step": 28351
+    },
+    {
+      "epoch": 0.6748548033895078,
+      "grad_norm": 3.7336175721013785,
+      "learning_rate": 2e-05,
+      "loss": 0.5502,
+      "step": 28352
+    },
+    {
+      "epoch": 0.6748786061125392,
+      "grad_norm": 4.3767777401681265,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 28353
+    },
+    {
+      "epoch": 0.6749024088355708,
+      "grad_norm": 16.06491696647424,
+      "learning_rate": 2e-05,
+      "loss": 1.2759,
+      "step": 28354
+    },
+    {
+      "epoch": 0.6749262115586023,
+      "grad_norm": 9.574187746974275,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 28355
+    },
+    {
+      "epoch": 0.6749500142816338,
+      "grad_norm": 3.5282637972879605,
+      "learning_rate": 2e-05,
+      "loss": 0.4779,
+      "step": 28356
+    },
+    {
+      "epoch": 0.6749738170046653,
+      "grad_norm": 4.394767048267923,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 28357
+    },
+    {
+      "epoch": 0.6749976197276969,
+      "grad_norm": 0.3783960560268433,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 28358
+    },
+    {
+      "epoch": 0.6750214224507284,
+      "grad_norm": 8.250487500337158,
+      "learning_rate": 2e-05,
+      "loss": 0.393,
+      "step": 28359
+    },
+    {
+      "epoch": 0.6750452251737599,
+      "grad_norm": 7.410385235541839,
+      "learning_rate": 2e-05,
+      "loss": 0.6276,
+      "step": 28360
+    },
+    {
+      "epoch": 0.6750690278967914,
+      "grad_norm": 0.24481855787088685,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 28361
+    },
+    {
+      "epoch": 0.675092830619823,
+      "grad_norm": 10.897628581133256,
+      "learning_rate": 2e-05,
+      "loss": 0.7833,
+      "step": 28362
+    },
+    {
+      "epoch": 0.6751166333428544,
+      "grad_norm": 3.6043042425108394,
+      "learning_rate": 2e-05,
+      "loss": 0.5521,
+      "step": 28363
+    },
+    {
+      "epoch": 0.6751404360658859,
+      "grad_norm": 1.6595536096885282,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 28364
+    },
+    {
+      "epoch": 0.6751642387889174,
+      "grad_norm": 1.2906488114502028,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 28365
+    },
+    {
+      "epoch": 0.675188041511949,
+      "grad_norm": 6.496522405850747,
+      "learning_rate": 2e-05,
+      "loss": 0.4711,
+      "step": 28366
+    },
+    {
+      "epoch": 0.6752118442349805,
+      "grad_norm": 9.057845742876461,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 28367
+    },
+    {
+      "epoch": 0.675235646958012,
+      "grad_norm": 0.01504439085910069,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28368
+    },
+    {
+      "epoch": 0.6752594496810436,
+      "grad_norm": 6.518516523827915,
+      "learning_rate": 2e-05,
+      "loss": 0.8537,
+      "step": 28369
+    },
+    {
+      "epoch": 0.675283252404075,
+      "grad_norm": 1.8514311487613015,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 28370
+    },
+    {
+      "epoch": 0.6753070551271065,
+      "grad_norm": 11.75474844754077,
+      "learning_rate": 2e-05,
+      "loss": 0.9387,
+      "step": 28371
+    },
+    {
+      "epoch": 0.675330857850138,
+      "grad_norm": 4.566126368819308,
+      "learning_rate": 2e-05,
+      "loss": 0.7409,
+      "step": 28372
+    },
+    {
+      "epoch": 0.6753546605731696,
+      "grad_norm": 7.247473619039829,
+      "learning_rate": 2e-05,
+      "loss": 1.3004,
+      "step": 28373
+    },
+    {
+      "epoch": 0.6753784632962011,
+      "grad_norm": 0.030513411388128662,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 28374
+    },
+    {
+      "epoch": 0.6754022660192326,
+      "grad_norm": 4.814913516639282,
+      "learning_rate": 2e-05,
+      "loss": 0.4638,
+      "step": 28375
+    },
+    {
+      "epoch": 0.6754260687422641,
+      "grad_norm": 3.5369302082914693,
+      "learning_rate": 2e-05,
+      "loss": 0.505,
+      "step": 28376
+    },
+    {
+      "epoch": 0.6754498714652957,
+      "grad_norm": 4.7714948358212625,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 28377
+    },
+    {
+      "epoch": 0.6754736741883272,
+      "grad_norm": 1.211985443177277,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 28378
+    },
+    {
+      "epoch": 0.6754974769113586,
+      "grad_norm": 9.939780439846514,
+      "learning_rate": 2e-05,
+      "loss": 0.767,
+      "step": 28379
+    },
+    {
+      "epoch": 0.6755212796343901,
+      "grad_norm": 0.5824037930691814,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 28380
+    },
+    {
+      "epoch": 0.6755450823574217,
+      "grad_norm": 1.0883780106582814,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 28381
+    },
+    {
+      "epoch": 0.6755688850804532,
+      "grad_norm": 5.981262459835674,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 28382
+    },
+    {
+      "epoch": 0.6755926878034847,
+      "grad_norm": 2.141409767855585,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 28383
+    },
+    {
+      "epoch": 0.6756164905265162,
+      "grad_norm": 1.889446487864994,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 28384
+    },
+    {
+      "epoch": 0.6756402932495478,
+      "grad_norm": 6.005884734618756,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 28385
+    },
+    {
+      "epoch": 0.6756640959725793,
+      "grad_norm": 11.581150626115821,
+      "learning_rate": 2e-05,
+      "loss": 1.0054,
+      "step": 28386
+    },
+    {
+      "epoch": 0.6756878986956107,
+      "grad_norm": 2.6740918235292743,
+      "learning_rate": 2e-05,
+      "loss": 0.4427,
+      "step": 28387
+    },
+    {
+      "epoch": 0.6757117014186423,
+      "grad_norm": 7.411155958880077,
+      "learning_rate": 2e-05,
+      "loss": 0.8225,
+      "step": 28388
+    },
+    {
+      "epoch": 0.6757355041416738,
+      "grad_norm": 2.7197735754532295,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 28389
+    },
+    {
+      "epoch": 0.6757593068647053,
+      "grad_norm": 0.23586744218797773,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 28390
+    },
+    {
+      "epoch": 0.6757831095877368,
+      "grad_norm": 2.2377979185090027,
+      "learning_rate": 2e-05,
+      "loss": 0.3998,
+      "step": 28391
+    },
+    {
+      "epoch": 0.6758069123107684,
+      "grad_norm": 4.876570166754146,
+      "learning_rate": 2e-05,
+      "loss": 1.1656,
+      "step": 28392
+    },
+    {
+      "epoch": 0.6758307150337999,
+      "grad_norm": 5.276417292296818,
+      "learning_rate": 2e-05,
+      "loss": 0.1914,
+      "step": 28393
+    },
+    {
+      "epoch": 0.6758545177568314,
+      "grad_norm": 10.608990187505587,
+      "learning_rate": 2e-05,
+      "loss": 1.4358,
+      "step": 28394
+    },
+    {
+      "epoch": 0.6758783204798628,
+      "grad_norm": 12.079915479343079,
+      "learning_rate": 2e-05,
+      "loss": 0.6372,
+      "step": 28395
+    },
+    {
+      "epoch": 0.6759021232028944,
+      "grad_norm": 7.2467214021961865,
+      "learning_rate": 2e-05,
+      "loss": 1.1879,
+      "step": 28396
+    },
+    {
+      "epoch": 0.6759259259259259,
+      "grad_norm": 14.348710045337087,
+      "learning_rate": 2e-05,
+      "loss": 1.0058,
+      "step": 28397
+    },
+    {
+      "epoch": 0.6759497286489574,
+      "grad_norm": 6.401564524859686,
+      "learning_rate": 2e-05,
+      "loss": 1.6491,
+      "step": 28398
+    },
+    {
+      "epoch": 0.6759735313719889,
+      "grad_norm": 4.143268987647737,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 28399
+    },
+    {
+      "epoch": 0.6759973340950205,
+      "grad_norm": 0.9170044163978096,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 28400
+    },
+    {
+      "epoch": 0.676021136818052,
+      "grad_norm": 6.767985855628279,
+      "learning_rate": 2e-05,
+      "loss": 0.5267,
+      "step": 28401
+    },
+    {
+      "epoch": 0.6760449395410835,
+      "grad_norm": 2.0409065649722624,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 28402
+    },
+    {
+      "epoch": 0.6760687422641151,
+      "grad_norm": 3.5681887553650253,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 28403
+    },
+    {
+      "epoch": 0.6760925449871465,
+      "grad_norm": 9.81176387625376,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 28404
+    },
+    {
+      "epoch": 0.676116347710178,
+      "grad_norm": 3.3518684015170983,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 28405
+    },
+    {
+      "epoch": 0.6761401504332095,
+      "grad_norm": 5.0950020345738265,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 28406
+    },
+    {
+      "epoch": 0.6761639531562411,
+      "grad_norm": 6.275717596698619,
+      "learning_rate": 2e-05,
+      "loss": 1.466,
+      "step": 28407
+    },
+    {
+      "epoch": 0.6761877558792726,
+      "grad_norm": 4.030866195860169,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 28408
+    },
+    {
+      "epoch": 0.6762115586023041,
+      "grad_norm": 1.3317914078439432,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 28409
+    },
+    {
+      "epoch": 0.6762353613253356,
+      "grad_norm": 1.9342261219684374,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 28410
+    },
+    {
+      "epoch": 0.6762591640483672,
+      "grad_norm": 2.255380127516329,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 28411
+    },
+    {
+      "epoch": 0.6762829667713987,
+      "grad_norm": 5.617504023799874,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 28412
+    },
+    {
+      "epoch": 0.6763067694944301,
+      "grad_norm": 4.744890472193343,
+      "learning_rate": 2e-05,
+      "loss": 1.0937,
+      "step": 28413
+    },
+    {
+      "epoch": 0.6763305722174616,
+      "grad_norm": 0.9207810157849438,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 28414
+    },
+    {
+      "epoch": 0.6763543749404932,
+      "grad_norm": 3.71096497265554,
+      "learning_rate": 2e-05,
+      "loss": 0.8543,
+      "step": 28415
+    },
+    {
+      "epoch": 0.6763781776635247,
+      "grad_norm": 0.4176485037673355,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 28416
+    },
+    {
+      "epoch": 0.6764019803865562,
+      "grad_norm": 3.5199647468713358,
+      "learning_rate": 2e-05,
+      "loss": 0.5585,
+      "step": 28417
+    },
+    {
+      "epoch": 0.6764257831095878,
+      "grad_norm": 9.632440960086166,
+      "learning_rate": 2e-05,
+      "loss": 0.4379,
+      "step": 28418
+    },
+    {
+      "epoch": 0.6764495858326193,
+      "grad_norm": 0.04347082759967356,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 28419
+    },
+    {
+      "epoch": 0.6764733885556508,
+      "grad_norm": 3.8789896485418014,
+      "learning_rate": 2e-05,
+      "loss": 0.7226,
+      "step": 28420
+    },
+    {
+      "epoch": 0.6764971912786822,
+      "grad_norm": 3.320418139730584,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 28421
+    },
+    {
+      "epoch": 0.6765209940017138,
+      "grad_norm": 0.040331957234441045,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 28422
+    },
+    {
+      "epoch": 0.6765447967247453,
+      "grad_norm": 7.573338815278736,
+      "learning_rate": 2e-05,
+      "loss": 0.7084,
+      "step": 28423
+    },
+    {
+      "epoch": 0.6765685994477768,
+      "grad_norm": 2.3097570314687434,
+      "learning_rate": 2e-05,
+      "loss": 0.3271,
+      "step": 28424
+    },
+    {
+      "epoch": 0.6765924021708083,
+      "grad_norm": 3.9379964803668126,
+      "learning_rate": 2e-05,
+      "loss": 0.5422,
+      "step": 28425
+    },
+    {
+      "epoch": 0.6766162048938399,
+      "grad_norm": 3.6204521710624973,
+      "learning_rate": 2e-05,
+      "loss": 0.7742,
+      "step": 28426
+    },
+    {
+      "epoch": 0.6766400076168714,
+      "grad_norm": 5.468419774394212,
+      "learning_rate": 2e-05,
+      "loss": 0.8116,
+      "step": 28427
+    },
+    {
+      "epoch": 0.6766638103399029,
+      "grad_norm": 12.338458089093548,
+      "learning_rate": 2e-05,
+      "loss": 0.4713,
+      "step": 28428
+    },
+    {
+      "epoch": 0.6766876130629343,
+      "grad_norm": 0.4411099942709814,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 28429
+    },
+    {
+      "epoch": 0.6767114157859659,
+      "grad_norm": 6.260475130183233,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 28430
+    },
+    {
+      "epoch": 0.6767352185089974,
+      "grad_norm": 1.2184228998533817,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 28431
+    },
+    {
+      "epoch": 0.6767590212320289,
+      "grad_norm": 4.908383341796548,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 28432
+    },
+    {
+      "epoch": 0.6767828239550605,
+      "grad_norm": 3.7356962096699755,
+      "learning_rate": 2e-05,
+      "loss": 0.7321,
+      "step": 28433
+    },
+    {
+      "epoch": 0.676806626678092,
+      "grad_norm": 3.2244355594189553,
+      "learning_rate": 2e-05,
+      "loss": 0.765,
+      "step": 28434
+    },
+    {
+      "epoch": 0.6768304294011235,
+      "grad_norm": 1.9700990343442015,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 28435
+    },
+    {
+      "epoch": 0.676854232124155,
+      "grad_norm": 2.11181934288244,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 28436
+    },
+    {
+      "epoch": 0.6768780348471866,
+      "grad_norm": 2.8111223092345954,
+      "learning_rate": 2e-05,
+      "loss": 0.2179,
+      "step": 28437
+    },
+    {
+      "epoch": 0.676901837570218,
+      "grad_norm": 2.6152946799264556,
+      "learning_rate": 2e-05,
+      "loss": 0.0808,
+      "step": 28438
+    },
+    {
+      "epoch": 0.6769256402932495,
+      "grad_norm": 0.6078296504622839,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 28439
+    },
+    {
+      "epoch": 0.676949443016281,
+      "grad_norm": 0.34120148095962055,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 28440
+    },
+    {
+      "epoch": 0.6769732457393126,
+      "grad_norm": 0.5829826663095192,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 28441
+    },
+    {
+      "epoch": 0.6769970484623441,
+      "grad_norm": 0.5075435821589147,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 28442
+    },
+    {
+      "epoch": 0.6770208511853756,
+      "grad_norm": 4.920857309388837,
+      "learning_rate": 2e-05,
+      "loss": 0.4813,
+      "step": 28443
+    },
+    {
+      "epoch": 0.6770446539084071,
+      "grad_norm": 3.308321000243692,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 28444
+    },
+    {
+      "epoch": 0.6770684566314387,
+      "grad_norm": 0.3033821125534419,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 28445
+    },
+    {
+      "epoch": 0.6770922593544701,
+      "grad_norm": 5.873828321518053,
+      "learning_rate": 2e-05,
+      "loss": 0.7133,
+      "step": 28446
+    },
+    {
+      "epoch": 0.6771160620775016,
+      "grad_norm": 2.783450022297652,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 28447
+    },
+    {
+      "epoch": 0.6771398648005332,
+      "grad_norm": 1.5207028227085362,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 28448
+    },
+    {
+      "epoch": 0.6771636675235647,
+      "grad_norm": 9.342636105746024,
+      "learning_rate": 2e-05,
+      "loss": 0.3776,
+      "step": 28449
+    },
+    {
+      "epoch": 0.6771874702465962,
+      "grad_norm": 6.99584402307083,
+      "learning_rate": 2e-05,
+      "loss": 1.7847,
+      "step": 28450
+    },
+    {
+      "epoch": 0.6772112729696277,
+      "grad_norm": 8.866387207525245,
+      "learning_rate": 2e-05,
+      "loss": 0.8605,
+      "step": 28451
+    },
+    {
+      "epoch": 0.6772350756926593,
+      "grad_norm": 4.814429720681205,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 28452
+    },
+    {
+      "epoch": 0.6772588784156908,
+      "grad_norm": 4.2611786569245504,
+      "learning_rate": 2e-05,
+      "loss": 0.4714,
+      "step": 28453
+    },
+    {
+      "epoch": 0.6772826811387223,
+      "grad_norm": 6.1965973973562845,
+      "learning_rate": 2e-05,
+      "loss": 1.6038,
+      "step": 28454
+    },
+    {
+      "epoch": 0.6773064838617537,
+      "grad_norm": 0.5637288263207785,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 28455
+    },
+    {
+      "epoch": 0.6773302865847853,
+      "grad_norm": 5.438735594665713,
+      "learning_rate": 2e-05,
+      "loss": 0.5466,
+      "step": 28456
+    },
+    {
+      "epoch": 0.6773540893078168,
+      "grad_norm": 0.7552854747050521,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 28457
+    },
+    {
+      "epoch": 0.6773778920308483,
+      "grad_norm": 4.2360842321966965,
+      "learning_rate": 2e-05,
+      "loss": 0.3278,
+      "step": 28458
+    },
+    {
+      "epoch": 0.6774016947538798,
+      "grad_norm": 14.965557038927306,
+      "learning_rate": 2e-05,
+      "loss": 0.9354,
+      "step": 28459
+    },
+    {
+      "epoch": 0.6774254974769114,
+      "grad_norm": 17.534234814547876,
+      "learning_rate": 2e-05,
+      "loss": 0.446,
+      "step": 28460
+    },
+    {
+      "epoch": 0.6774493001999429,
+      "grad_norm": 4.929141813299501,
+      "learning_rate": 2e-05,
+      "loss": 0.7564,
+      "step": 28461
+    },
+    {
+      "epoch": 0.6774731029229744,
+      "grad_norm": 5.814303861067716,
+      "learning_rate": 2e-05,
+      "loss": 0.9073,
+      "step": 28462
+    },
+    {
+      "epoch": 0.677496905646006,
+      "grad_norm": 2.281506809804567,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 28463
+    },
+    {
+      "epoch": 0.6775207083690374,
+      "grad_norm": 0.33799941767106406,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 28464
+    },
+    {
+      "epoch": 0.6775445110920689,
+      "grad_norm": 2.0451881107969525,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 28465
+    },
+    {
+      "epoch": 0.6775683138151004,
+      "grad_norm": 0.23924918880895008,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 28466
+    },
+    {
+      "epoch": 0.677592116538132,
+      "grad_norm": 38.111133191652634,
+      "learning_rate": 2e-05,
+      "loss": 0.6128,
+      "step": 28467
+    },
+    {
+      "epoch": 0.6776159192611635,
+      "grad_norm": 4.662182218579777,
+      "learning_rate": 2e-05,
+      "loss": 0.8822,
+      "step": 28468
+    },
+    {
+      "epoch": 0.677639721984195,
+      "grad_norm": 0.09397400420099168,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28469
+    },
+    {
+      "epoch": 0.6776635247072265,
+      "grad_norm": 12.041531596958505,
+      "learning_rate": 2e-05,
+      "loss": 1.5578,
+      "step": 28470
+    },
+    {
+      "epoch": 0.6776873274302581,
+      "grad_norm": 0.3336468586447498,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 28471
+    },
+    {
+      "epoch": 0.6777111301532895,
+      "grad_norm": 6.3178660081339535,
+      "learning_rate": 2e-05,
+      "loss": 1.8287,
+      "step": 28472
+    },
+    {
+      "epoch": 0.677734932876321,
+      "grad_norm": 0.28040908050789237,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 28473
+    },
+    {
+      "epoch": 0.6777587355993525,
+      "grad_norm": 5.573886189914077,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 28474
+    },
+    {
+      "epoch": 0.6777825383223841,
+      "grad_norm": 4.5774060962989696,
+      "learning_rate": 2e-05,
+      "loss": 0.117,
+      "step": 28475
+    },
+    {
+      "epoch": 0.6778063410454156,
+      "grad_norm": 0.8456094516174659,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 28476
+    },
+    {
+      "epoch": 0.6778301437684471,
+      "grad_norm": 3.967894558147131,
+      "learning_rate": 2e-05,
+      "loss": 0.9114,
+      "step": 28477
+    },
+    {
+      "epoch": 0.6778539464914787,
+      "grad_norm": 0.0954879455923661,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 28478
+    },
+    {
+      "epoch": 0.6778777492145102,
+      "grad_norm": 2.8224604297720823,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 28479
+    },
+    {
+      "epoch": 0.6779015519375416,
+      "grad_norm": 3.6445867628415525,
+      "learning_rate": 2e-05,
+      "loss": 0.6203,
+      "step": 28480
+    },
+    {
+      "epoch": 0.6779253546605731,
+      "grad_norm": 0.46617771552212095,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 28481
+    },
+    {
+      "epoch": 0.6779491573836047,
+      "grad_norm": 3.182816062482722,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 28482
+    },
+    {
+      "epoch": 0.6779729601066362,
+      "grad_norm": 2.774883214858006,
+      "learning_rate": 2e-05,
+      "loss": 0.4782,
+      "step": 28483
+    },
+    {
+      "epoch": 0.6779967628296677,
+      "grad_norm": 0.08140359326081777,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 28484
+    },
+    {
+      "epoch": 0.6780205655526992,
+      "grad_norm": 3.2323886477076678,
+      "learning_rate": 2e-05,
+      "loss": 0.2354,
+      "step": 28485
+    },
+    {
+      "epoch": 0.6780443682757308,
+      "grad_norm": 3.9128139591045765,
+      "learning_rate": 2e-05,
+      "loss": 0.489,
+      "step": 28486
+    },
+    {
+      "epoch": 0.6780681709987623,
+      "grad_norm": 2.2681568607968785,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 28487
+    },
+    {
+      "epoch": 0.6780919737217938,
+      "grad_norm": 10.049460322515596,
+      "learning_rate": 2e-05,
+      "loss": 0.7696,
+      "step": 28488
+    },
+    {
+      "epoch": 0.6781157764448252,
+      "grad_norm": 1.2799531292736142,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 28489
+    },
+    {
+      "epoch": 0.6781395791678568,
+      "grad_norm": 9.064780685771478,
+      "learning_rate": 2e-05,
+      "loss": 1.1511,
+      "step": 28490
+    },
+    {
+      "epoch": 0.6781633818908883,
+      "grad_norm": 0.13629424142578572,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 28491
+    },
+    {
+      "epoch": 0.6781871846139198,
+      "grad_norm": 8.646247004177864,
+      "learning_rate": 2e-05,
+      "loss": 0.8864,
+      "step": 28492
+    },
+    {
+      "epoch": 0.6782109873369514,
+      "grad_norm": 3.1463145951695304,
+      "learning_rate": 2e-05,
+      "loss": 0.5452,
+      "step": 28493
+    },
+    {
+      "epoch": 0.6782347900599829,
+      "grad_norm": 3.6379052657328383,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 28494
+    },
+    {
+      "epoch": 0.6782585927830144,
+      "grad_norm": 4.680165735950282,
+      "learning_rate": 2e-05,
+      "loss": 0.6149,
+      "step": 28495
+    },
+    {
+      "epoch": 0.6782823955060459,
+      "grad_norm": 27.245286151726976,
+      "learning_rate": 2e-05,
+      "loss": 1.5005,
+      "step": 28496
+    },
+    {
+      "epoch": 0.6783061982290775,
+      "grad_norm": 0.6872677961675128,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 28497
+    },
+    {
+      "epoch": 0.6783300009521089,
+      "grad_norm": 6.4696452588528945,
+      "learning_rate": 2e-05,
+      "loss": 0.7312,
+      "step": 28498
+    },
+    {
+      "epoch": 0.6783538036751404,
+      "grad_norm": 18.21726607110015,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 28499
+    },
+    {
+      "epoch": 0.6783776063981719,
+      "grad_norm": 2.067267288581422,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 28500
+    },
+    {
+      "epoch": 0.6784014091212035,
+      "grad_norm": 0.940806199599403,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 28501
+    },
+    {
+      "epoch": 0.678425211844235,
+      "grad_norm": 12.529035100680568,
+      "learning_rate": 2e-05,
+      "loss": 0.8395,
+      "step": 28502
+    },
+    {
+      "epoch": 0.6784490145672665,
+      "grad_norm": 0.2449909279759286,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 28503
+    },
+    {
+      "epoch": 0.678472817290298,
+      "grad_norm": 19.17319946532002,
+      "learning_rate": 2e-05,
+      "loss": 1.0997,
+      "step": 28504
+    },
+    {
+      "epoch": 0.6784966200133296,
+      "grad_norm": 2.7526533213993356,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 28505
+    },
+    {
+      "epoch": 0.678520422736361,
+      "grad_norm": 1.5290511326941432,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 28506
+    },
+    {
+      "epoch": 0.6785442254593925,
+      "grad_norm": 2.6966625183970154,
+      "learning_rate": 2e-05,
+      "loss": 0.336,
+      "step": 28507
+    },
+    {
+      "epoch": 0.6785680281824241,
+      "grad_norm": 0.6447013803973228,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 28508
+    },
+    {
+      "epoch": 0.6785918309054556,
+      "grad_norm": 33.141825229855094,
+      "learning_rate": 2e-05,
+      "loss": 0.7464,
+      "step": 28509
+    },
+    {
+      "epoch": 0.6786156336284871,
+      "grad_norm": 5.084002744381695,
+      "learning_rate": 2e-05,
+      "loss": 0.9692,
+      "step": 28510
+    },
+    {
+      "epoch": 0.6786394363515186,
+      "grad_norm": 4.401070327676043,
+      "learning_rate": 2e-05,
+      "loss": 0.6725,
+      "step": 28511
+    },
+    {
+      "epoch": 0.6786632390745502,
+      "grad_norm": 6.849223602393061,
+      "learning_rate": 2e-05,
+      "loss": 0.1077,
+      "step": 28512
+    },
+    {
+      "epoch": 0.6786870417975817,
+      "grad_norm": 22.354179807336372,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 28513
+    },
+    {
+      "epoch": 0.6787108445206131,
+      "grad_norm": 0.016118023272397863,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 28514
+    },
+    {
+      "epoch": 0.6787346472436446,
+      "grad_norm": 4.082300528103538,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 28515
+    },
+    {
+      "epoch": 0.6787584499666762,
+      "grad_norm": 3.2125012603035725,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 28516
+    },
+    {
+      "epoch": 0.6787822526897077,
+      "grad_norm": 0.9731834572041071,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 28517
+    },
+    {
+      "epoch": 0.6788060554127392,
+      "grad_norm": 2.4696927775291577,
+      "learning_rate": 2e-05,
+      "loss": 0.347,
+      "step": 28518
+    },
+    {
+      "epoch": 0.6788298581357707,
+      "grad_norm": 31.16375429180921,
+      "learning_rate": 2e-05,
+      "loss": 0.521,
+      "step": 28519
+    },
+    {
+      "epoch": 0.6788536608588023,
+      "grad_norm": 2.2631971208268205,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 28520
+    },
+    {
+      "epoch": 0.6788774635818338,
+      "grad_norm": 2.77189789626682,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 28521
+    },
+    {
+      "epoch": 0.6789012663048652,
+      "grad_norm": 3.6588979664713897,
+      "learning_rate": 2e-05,
+      "loss": 0.6752,
+      "step": 28522
+    },
+    {
+      "epoch": 0.6789250690278968,
+      "grad_norm": 2.680895106470589,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 28523
+    },
+    {
+      "epoch": 0.6789488717509283,
+      "grad_norm": 3.778210629068712,
+      "learning_rate": 2e-05,
+      "loss": 0.5029,
+      "step": 28524
+    },
+    {
+      "epoch": 0.6789726744739598,
+      "grad_norm": 3.45745359287004,
+      "learning_rate": 2e-05,
+      "loss": 0.4866,
+      "step": 28525
+    },
+    {
+      "epoch": 0.6789964771969913,
+      "grad_norm": 1.9229812394571528,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 28526
+    },
+    {
+      "epoch": 0.6790202799200229,
+      "grad_norm": 1.9973874173519865,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 28527
+    },
+    {
+      "epoch": 0.6790440826430544,
+      "grad_norm": 0.2246034220862994,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 28528
+    },
+    {
+      "epoch": 0.6790678853660859,
+      "grad_norm": 0.678364720512134,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 28529
+    },
+    {
+      "epoch": 0.6790916880891174,
+      "grad_norm": 15.208504023498342,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 28530
+    },
+    {
+      "epoch": 0.679115490812149,
+      "grad_norm": 6.278105504966283,
+      "learning_rate": 2e-05,
+      "loss": 0.9136,
+      "step": 28531
+    },
+    {
+      "epoch": 0.6791392935351804,
+      "grad_norm": 1.257804878251049,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 28532
+    },
+    {
+      "epoch": 0.6791630962582119,
+      "grad_norm": 1.5942766712893675,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 28533
+    },
+    {
+      "epoch": 0.6791868989812434,
+      "grad_norm": 3.7601068835392506,
+      "learning_rate": 2e-05,
+      "loss": 0.7721,
+      "step": 28534
+    },
+    {
+      "epoch": 0.679210701704275,
+      "grad_norm": 0.8088433482839176,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 28535
+    },
+    {
+      "epoch": 0.6792345044273065,
+      "grad_norm": 9.232521050426957,
+      "learning_rate": 2e-05,
+      "loss": 0.9841,
+      "step": 28536
+    },
+    {
+      "epoch": 0.679258307150338,
+      "grad_norm": 1.4768834113572344,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 28537
+    },
+    {
+      "epoch": 0.6792821098733696,
+      "grad_norm": 3.262314647416767,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 28538
+    },
+    {
+      "epoch": 0.679305912596401,
+      "grad_norm": 2.9927372260836957,
+      "learning_rate": 2e-05,
+      "loss": 0.5858,
+      "step": 28539
+    },
+    {
+      "epoch": 0.6793297153194325,
+      "grad_norm": 11.953190861976145,
+      "learning_rate": 2e-05,
+      "loss": 0.2682,
+      "step": 28540
+    },
+    {
+      "epoch": 0.679353518042464,
+      "grad_norm": 3.7199586409900105,
+      "learning_rate": 2e-05,
+      "loss": 0.4036,
+      "step": 28541
+    },
+    {
+      "epoch": 0.6793773207654956,
+      "grad_norm": 4.087076738707074,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 28542
+    },
+    {
+      "epoch": 0.6794011234885271,
+      "grad_norm": 3.949044241130004,
+      "learning_rate": 2e-05,
+      "loss": 0.3006,
+      "step": 28543
+    },
+    {
+      "epoch": 0.6794249262115586,
+      "grad_norm": 3.208579259394696,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 28544
+    },
+    {
+      "epoch": 0.6794487289345901,
+      "grad_norm": 3.9827035173084924,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 28545
+    },
+    {
+      "epoch": 0.6794725316576217,
+      "grad_norm": 2.0407355711224526,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 28546
+    },
+    {
+      "epoch": 0.6794963343806532,
+      "grad_norm": 18.114216295306303,
+      "learning_rate": 2e-05,
+      "loss": 0.2787,
+      "step": 28547
+    },
+    {
+      "epoch": 0.6795201371036846,
+      "grad_norm": 2.551282403040491,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 28548
+    },
+    {
+      "epoch": 0.6795439398267161,
+      "grad_norm": 0.16825455014079505,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 28549
+    },
+    {
+      "epoch": 0.6795677425497477,
+      "grad_norm": 0.18745013457619852,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 28550
+    },
+    {
+      "epoch": 0.6795915452727792,
+      "grad_norm": 10.315225818490473,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 28551
+    },
+    {
+      "epoch": 0.6796153479958107,
+      "grad_norm": 0.45475308224782574,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 28552
+    },
+    {
+      "epoch": 0.6796391507188423,
+      "grad_norm": 14.820054663216606,
+      "learning_rate": 2e-05,
+      "loss": 1.4972,
+      "step": 28553
+    },
+    {
+      "epoch": 0.6796629534418738,
+      "grad_norm": 1.1076303907104499,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 28554
+    },
+    {
+      "epoch": 0.6796867561649053,
+      "grad_norm": 5.28485222757357,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 28555
+    },
+    {
+      "epoch": 0.6797105588879367,
+      "grad_norm": 16.664697779467602,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 28556
+    },
+    {
+      "epoch": 0.6797343616109683,
+      "grad_norm": 5.194037984487355,
+      "learning_rate": 2e-05,
+      "loss": 0.5589,
+      "step": 28557
+    },
+    {
+      "epoch": 0.6797581643339998,
+      "grad_norm": 2.6488580174004053,
+      "learning_rate": 2e-05,
+      "loss": 0.3992,
+      "step": 28558
+    },
+    {
+      "epoch": 0.6797819670570313,
+      "grad_norm": 3.297247871328253,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 28559
+    },
+    {
+      "epoch": 0.6798057697800628,
+      "grad_norm": 4.25197529366186,
+      "learning_rate": 2e-05,
+      "loss": 0.5814,
+      "step": 28560
+    },
+    {
+      "epoch": 0.6798295725030944,
+      "grad_norm": 7.1743436516119745,
+      "learning_rate": 2e-05,
+      "loss": 0.1569,
+      "step": 28561
+    },
+    {
+      "epoch": 0.6798533752261259,
+      "grad_norm": 4.757401928863732,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 28562
+    },
+    {
+      "epoch": 0.6798771779491574,
+      "grad_norm": 5.143584912651007,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 28563
+    },
+    {
+      "epoch": 0.6799009806721888,
+      "grad_norm": 1.7496880783794488,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 28564
+    },
+    {
+      "epoch": 0.6799247833952204,
+      "grad_norm": 1.1910194352945807,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 28565
+    },
+    {
+      "epoch": 0.6799485861182519,
+      "grad_norm": 6.803994993041216,
+      "learning_rate": 2e-05,
+      "loss": 1.3853,
+      "step": 28566
+    },
+    {
+      "epoch": 0.6799723888412834,
+      "grad_norm": 4.571991452377337,
+      "learning_rate": 2e-05,
+      "loss": 0.7707,
+      "step": 28567
+    },
+    {
+      "epoch": 0.679996191564315,
+      "grad_norm": 6.41800087759438,
+      "learning_rate": 2e-05,
+      "loss": 1.3809,
+      "step": 28568
+    },
+    {
+      "epoch": 0.6800199942873465,
+      "grad_norm": 7.778170694195723,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 28569
+    },
+    {
+      "epoch": 0.680043797010378,
+      "grad_norm": 65.05527850270853,
+      "learning_rate": 2e-05,
+      "loss": 0.344,
+      "step": 28570
+    },
+    {
+      "epoch": 0.6800675997334095,
+      "grad_norm": 4.360953756503039,
+      "learning_rate": 2e-05,
+      "loss": 0.5902,
+      "step": 28571
+    },
+    {
+      "epoch": 0.6800914024564411,
+      "grad_norm": 0.04891896540995524,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 28572
+    },
+    {
+      "epoch": 0.6801152051794725,
+      "grad_norm": 3.440691090078232,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 28573
+    },
+    {
+      "epoch": 0.680139007902504,
+      "grad_norm": 0.9817895610913991,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 28574
+    },
+    {
+      "epoch": 0.6801628106255355,
+      "grad_norm": 10.584243810154907,
+      "learning_rate": 2e-05,
+      "loss": 0.6343,
+      "step": 28575
+    },
+    {
+      "epoch": 0.6801866133485671,
+      "grad_norm": 10.974050941920572,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 28576
+    },
+    {
+      "epoch": 0.6802104160715986,
+      "grad_norm": 7.690147638811229,
+      "learning_rate": 2e-05,
+      "loss": 0.9183,
+      "step": 28577
+    },
+    {
+      "epoch": 0.6802342187946301,
+      "grad_norm": 4.175713589007624,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 28578
+    },
+    {
+      "epoch": 0.6802580215176616,
+      "grad_norm": 3.05864864405084,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 28579
+    },
+    {
+      "epoch": 0.6802818242406932,
+      "grad_norm": 4.2092715634121705,
+      "learning_rate": 2e-05,
+      "loss": 0.4863,
+      "step": 28580
+    },
+    {
+      "epoch": 0.6803056269637247,
+      "grad_norm": 0.6305472724776776,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 28581
+    },
+    {
+      "epoch": 0.6803294296867561,
+      "grad_norm": 3.834666111163899,
+      "learning_rate": 2e-05,
+      "loss": 0.0881,
+      "step": 28582
+    },
+    {
+      "epoch": 0.6803532324097877,
+      "grad_norm": 2.2091760529703364,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 28583
+    },
+    {
+      "epoch": 0.6803770351328192,
+      "grad_norm": 6.1860885311389255,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 28584
+    },
+    {
+      "epoch": 0.6804008378558507,
+      "grad_norm": 10.912615778279957,
+      "learning_rate": 2e-05,
+      "loss": 0.5698,
+      "step": 28585
+    },
+    {
+      "epoch": 0.6804246405788822,
+      "grad_norm": 3.873919987025824,
+      "learning_rate": 2e-05,
+      "loss": 0.3752,
+      "step": 28586
+    },
+    {
+      "epoch": 0.6804484433019138,
+      "grad_norm": 0.43581904007936556,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 28587
+    },
+    {
+      "epoch": 0.6804722460249453,
+      "grad_norm": 2.481486863539859,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 28588
+    },
+    {
+      "epoch": 0.6804960487479768,
+      "grad_norm": 3.6469463737370806,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 28589
+    },
+    {
+      "epoch": 0.6805198514710082,
+      "grad_norm": 17.401759212532305,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 28590
+    },
+    {
+      "epoch": 0.6805436541940398,
+      "grad_norm": 0.26127052685864205,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 28591
+    },
+    {
+      "epoch": 0.6805674569170713,
+      "grad_norm": 2.576278112438877,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 28592
+    },
+    {
+      "epoch": 0.6805912596401028,
+      "grad_norm": 30.83189972014087,
+      "learning_rate": 2e-05,
+      "loss": 0.3611,
+      "step": 28593
+    },
+    {
+      "epoch": 0.6806150623631343,
+      "grad_norm": 3.633029320935225,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 28594
+    },
+    {
+      "epoch": 0.6806388650861659,
+      "grad_norm": 2.253467862793483,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 28595
+    },
+    {
+      "epoch": 0.6806626678091974,
+      "grad_norm": 0.5508368307382636,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 28596
+    },
+    {
+      "epoch": 0.6806864705322289,
+      "grad_norm": 13.2171530084879,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 28597
+    },
+    {
+      "epoch": 0.6807102732552605,
+      "grad_norm": 12.568909206467973,
+      "learning_rate": 2e-05,
+      "loss": 0.2682,
+      "step": 28598
+    },
+    {
+      "epoch": 0.6807340759782919,
+      "grad_norm": 0.30224288845896297,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 28599
+    },
+    {
+      "epoch": 0.6807578787013234,
+      "grad_norm": 6.676488110070818,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 28600
+    },
+    {
+      "epoch": 0.6807816814243549,
+      "grad_norm": 3.4451042769449423,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 28601
+    },
+    {
+      "epoch": 0.6808054841473865,
+      "grad_norm": 10.435589786250727,
+      "learning_rate": 2e-05,
+      "loss": 0.1878,
+      "step": 28602
+    },
+    {
+      "epoch": 0.680829286870418,
+      "grad_norm": 5.635021778336117,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 28603
+    },
+    {
+      "epoch": 0.6808530895934495,
+      "grad_norm": 12.932885220010254,
+      "learning_rate": 2e-05,
+      "loss": 1.5407,
+      "step": 28604
+    },
+    {
+      "epoch": 0.680876892316481,
+      "grad_norm": 6.009553979325615,
+      "learning_rate": 2e-05,
+      "loss": 0.7044,
+      "step": 28605
+    },
+    {
+      "epoch": 0.6809006950395126,
+      "grad_norm": 1.3777364254062938,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 28606
+    },
+    {
+      "epoch": 0.680924497762544,
+      "grad_norm": 4.788555587172736,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 28607
+    },
+    {
+      "epoch": 0.6809483004855755,
+      "grad_norm": 4.311020272247112,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 28608
+    },
+    {
+      "epoch": 0.680972103208607,
+      "grad_norm": 3.224531078433317,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 28609
+    },
+    {
+      "epoch": 0.6809959059316386,
+      "grad_norm": 41.154062943496825,
+      "learning_rate": 2e-05,
+      "loss": 0.2179,
+      "step": 28610
+    },
+    {
+      "epoch": 0.6810197086546701,
+      "grad_norm": 0.8795802100047291,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 28611
+    },
+    {
+      "epoch": 0.6810435113777016,
+      "grad_norm": 2.954940267311167,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 28612
+    },
+    {
+      "epoch": 0.6810673141007332,
+      "grad_norm": 7.377491714729839,
+      "learning_rate": 2e-05,
+      "loss": 0.1379,
+      "step": 28613
+    },
+    {
+      "epoch": 0.6810911168237647,
+      "grad_norm": 0.15995609848093892,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 28614
+    },
+    {
+      "epoch": 0.6811149195467961,
+      "grad_norm": 2.870327496003856,
+      "learning_rate": 2e-05,
+      "loss": 0.3873,
+      "step": 28615
+    },
+    {
+      "epoch": 0.6811387222698276,
+      "grad_norm": 2.224866204037154,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 28616
+    },
+    {
+      "epoch": 0.6811625249928592,
+      "grad_norm": 0.49544466181928465,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 28617
+    },
+    {
+      "epoch": 0.6811863277158907,
+      "grad_norm": 12.165064769421841,
+      "learning_rate": 2e-05,
+      "loss": 0.2471,
+      "step": 28618
+    },
+    {
+      "epoch": 0.6812101304389222,
+      "grad_norm": 65.00466963531963,
+      "learning_rate": 2e-05,
+      "loss": 1.2952,
+      "step": 28619
+    },
+    {
+      "epoch": 0.6812339331619537,
+      "grad_norm": 43.86453499259807,
+      "learning_rate": 2e-05,
+      "loss": 0.5346,
+      "step": 28620
+    },
+    {
+      "epoch": 0.6812577358849853,
+      "grad_norm": 6.583323560016073,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 28621
+    },
+    {
+      "epoch": 0.6812815386080168,
+      "grad_norm": 1.346349010686053,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 28622
+    },
+    {
+      "epoch": 0.6813053413310483,
+      "grad_norm": 5.770216299579325,
+      "learning_rate": 2e-05,
+      "loss": 0.7362,
+      "step": 28623
+    },
+    {
+      "epoch": 0.6813291440540797,
+      "grad_norm": 5.578513613433518,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 28624
+    },
+    {
+      "epoch": 0.6813529467771113,
+      "grad_norm": 5.083106026163769,
+      "learning_rate": 2e-05,
+      "loss": 0.5764,
+      "step": 28625
+    },
+    {
+      "epoch": 0.6813767495001428,
+      "grad_norm": 4.273775981770388,
+      "learning_rate": 2e-05,
+      "loss": 0.4797,
+      "step": 28626
+    },
+    {
+      "epoch": 0.6814005522231743,
+      "grad_norm": 4.529433375801045,
+      "learning_rate": 2e-05,
+      "loss": 0.5094,
+      "step": 28627
+    },
+    {
+      "epoch": 0.6814243549462059,
+      "grad_norm": 14.840732743954506,
+      "learning_rate": 2e-05,
+      "loss": 0.5368,
+      "step": 28628
+    },
+    {
+      "epoch": 0.6814481576692374,
+      "grad_norm": 0.590519539030504,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 28629
+    },
+    {
+      "epoch": 0.6814719603922689,
+      "grad_norm": 4.768496400011382,
+      "learning_rate": 2e-05,
+      "loss": 0.8255,
+      "step": 28630
+    },
+    {
+      "epoch": 0.6814957631153004,
+      "grad_norm": 1.2569626526987419,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 28631
+    },
+    {
+      "epoch": 0.681519565838332,
+      "grad_norm": 16.501645940296704,
+      "learning_rate": 2e-05,
+      "loss": 0.728,
+      "step": 28632
+    },
+    {
+      "epoch": 0.6815433685613634,
+      "grad_norm": 10.188527015467761,
+      "learning_rate": 2e-05,
+      "loss": 1.0168,
+      "step": 28633
+    },
+    {
+      "epoch": 0.6815671712843949,
+      "grad_norm": 0.0463898424474765,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 28634
+    },
+    {
+      "epoch": 0.6815909740074264,
+      "grad_norm": 8.966116893053872,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 28635
+    },
+    {
+      "epoch": 0.681614776730458,
+      "grad_norm": 4.914135689754819,
+      "learning_rate": 2e-05,
+      "loss": 0.8169,
+      "step": 28636
+    },
+    {
+      "epoch": 0.6816385794534895,
+      "grad_norm": 2.9357515819685993,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 28637
+    },
+    {
+      "epoch": 0.681662382176521,
+      "grad_norm": 3.6019258157301532,
+      "learning_rate": 2e-05,
+      "loss": 0.3941,
+      "step": 28638
+    },
+    {
+      "epoch": 0.6816861848995525,
+      "grad_norm": 13.495336843061931,
+      "learning_rate": 2e-05,
+      "loss": 0.845,
+      "step": 28639
+    },
+    {
+      "epoch": 0.6817099876225841,
+      "grad_norm": 0.8704314325880016,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 28640
+    },
+    {
+      "epoch": 0.6817337903456155,
+      "grad_norm": 1.3605576172598481,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 28641
+    },
+    {
+      "epoch": 0.681757593068647,
+      "grad_norm": 0.4860224972145025,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 28642
+    },
+    {
+      "epoch": 0.6817813957916785,
+      "grad_norm": 0.27729082212785217,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 28643
+    },
+    {
+      "epoch": 0.6818051985147101,
+      "grad_norm": 0.10930509153907234,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 28644
+    },
+    {
+      "epoch": 0.6818290012377416,
+      "grad_norm": 2.4778351150568003,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 28645
+    },
+    {
+      "epoch": 0.6818528039607731,
+      "grad_norm": 3.293534156175036,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 28646
+    },
+    {
+      "epoch": 0.6818766066838047,
+      "grad_norm": 1.86460872347595,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 28647
+    },
+    {
+      "epoch": 0.6819004094068362,
+      "grad_norm": 0.6009711703624652,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 28648
+    },
+    {
+      "epoch": 0.6819242121298676,
+      "grad_norm": 6.7671746788234906,
+      "learning_rate": 2e-05,
+      "loss": 0.7294,
+      "step": 28649
+    },
+    {
+      "epoch": 0.6819480148528991,
+      "grad_norm": 5.388139583978815,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 28650
+    },
+    {
+      "epoch": 0.6819718175759307,
+      "grad_norm": 2.00612329480957,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 28651
+    },
+    {
+      "epoch": 0.6819956202989622,
+      "grad_norm": 7.405218588124362,
+      "learning_rate": 2e-05,
+      "loss": 0.2112,
+      "step": 28652
+    },
+    {
+      "epoch": 0.6820194230219937,
+      "grad_norm": 1.8541914027736406,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 28653
+    },
+    {
+      "epoch": 0.6820432257450252,
+      "grad_norm": 0.11622420597079204,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 28654
+    },
+    {
+      "epoch": 0.6820670284680568,
+      "grad_norm": 3.6503041625522727,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 28655
+    },
+    {
+      "epoch": 0.6820908311910883,
+      "grad_norm": 3.2908619286751564,
+      "learning_rate": 2e-05,
+      "loss": 0.6353,
+      "step": 28656
+    },
+    {
+      "epoch": 0.6821146339141198,
+      "grad_norm": 0.8225257183818606,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 28657
+    },
+    {
+      "epoch": 0.6821384366371512,
+      "grad_norm": 13.71132186404573,
+      "learning_rate": 2e-05,
+      "loss": 0.3489,
+      "step": 28658
+    },
+    {
+      "epoch": 0.6821622393601828,
+      "grad_norm": 1.3431050170863361,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 28659
+    },
+    {
+      "epoch": 0.6821860420832143,
+      "grad_norm": 3.5537991989300073,
+      "learning_rate": 2e-05,
+      "loss": 0.3565,
+      "step": 28660
+    },
+    {
+      "epoch": 0.6822098448062458,
+      "grad_norm": 0.005280485134457233,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 28661
+    },
+    {
+      "epoch": 0.6822336475292774,
+      "grad_norm": 4.270528783259138,
+      "learning_rate": 2e-05,
+      "loss": 0.5397,
+      "step": 28662
+    },
+    {
+      "epoch": 0.6822574502523089,
+      "grad_norm": 10.715158863697788,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 28663
+    },
+    {
+      "epoch": 0.6822812529753404,
+      "grad_norm": 7.955107874050032,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 28664
+    },
+    {
+      "epoch": 0.6823050556983719,
+      "grad_norm": 0.10865086492740166,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 28665
+    },
+    {
+      "epoch": 0.6823288584214035,
+      "grad_norm": 4.941546082990381,
+      "learning_rate": 2e-05,
+      "loss": 0.8427,
+      "step": 28666
+    },
+    {
+      "epoch": 0.6823526611444349,
+      "grad_norm": 1.8613656927016955,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 28667
+    },
+    {
+      "epoch": 0.6823764638674664,
+      "grad_norm": 0.16627834924607568,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 28668
+    },
+    {
+      "epoch": 0.6824002665904979,
+      "grad_norm": 2.5692449385140512,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 28669
+    },
+    {
+      "epoch": 0.6824240693135295,
+      "grad_norm": 14.684834821348115,
+      "learning_rate": 2e-05,
+      "loss": 0.5888,
+      "step": 28670
+    },
+    {
+      "epoch": 0.682447872036561,
+      "grad_norm": 4.176631304167584,
+      "learning_rate": 2e-05,
+      "loss": 0.1798,
+      "step": 28671
+    },
+    {
+      "epoch": 0.6824716747595925,
+      "grad_norm": 3.6191504729293684,
+      "learning_rate": 2e-05,
+      "loss": 0.7909,
+      "step": 28672
+    },
+    {
+      "epoch": 0.682495477482624,
+      "grad_norm": 2.127362434048617,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 28673
+    },
+    {
+      "epoch": 0.6825192802056556,
+      "grad_norm": 2.625515862737137,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 28674
+    },
+    {
+      "epoch": 0.682543082928687,
+      "grad_norm": 3.863991536791408,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 28675
+    },
+    {
+      "epoch": 0.6825668856517185,
+      "grad_norm": 7.147244779137867,
+      "learning_rate": 2e-05,
+      "loss": 0.6106,
+      "step": 28676
+    },
+    {
+      "epoch": 0.6825906883747501,
+      "grad_norm": 19.279903915799377,
+      "learning_rate": 2e-05,
+      "loss": 0.9937,
+      "step": 28677
+    },
+    {
+      "epoch": 0.6826144910977816,
+      "grad_norm": 14.684432807980144,
+      "learning_rate": 2e-05,
+      "loss": 0.8916,
+      "step": 28678
+    },
+    {
+      "epoch": 0.6826382938208131,
+      "grad_norm": 7.9258826497796,
+      "learning_rate": 2e-05,
+      "loss": 1.175,
+      "step": 28679
+    },
+    {
+      "epoch": 0.6826620965438446,
+      "grad_norm": 11.279599085594786,
+      "learning_rate": 2e-05,
+      "loss": 1.1058,
+      "step": 28680
+    },
+    {
+      "epoch": 0.6826858992668762,
+      "grad_norm": 3.837652271603224,
+      "learning_rate": 2e-05,
+      "loss": 0.6308,
+      "step": 28681
+    },
+    {
+      "epoch": 0.6827097019899077,
+      "grad_norm": 3.95900625623962,
+      "learning_rate": 2e-05,
+      "loss": 0.547,
+      "step": 28682
+    },
+    {
+      "epoch": 0.6827335047129391,
+      "grad_norm": 3.312329420777114,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 28683
+    },
+    {
+      "epoch": 0.6827573074359706,
+      "grad_norm": 4.31123537066258,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 28684
+    },
+    {
+      "epoch": 0.6827811101590022,
+      "grad_norm": 11.780979486214848,
+      "learning_rate": 2e-05,
+      "loss": 0.205,
+      "step": 28685
+    },
+    {
+      "epoch": 0.6828049128820337,
+      "grad_norm": 7.875201236223741,
+      "learning_rate": 2e-05,
+      "loss": 0.1941,
+      "step": 28686
+    },
+    {
+      "epoch": 0.6828287156050652,
+      "grad_norm": 6.660310277635937,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 28687
+    },
+    {
+      "epoch": 0.6828525183280967,
+      "grad_norm": 2.5777051453440762,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 28688
+    },
+    {
+      "epoch": 0.6828763210511283,
+      "grad_norm": 5.964404121089591,
+      "learning_rate": 2e-05,
+      "loss": 0.0886,
+      "step": 28689
+    },
+    {
+      "epoch": 0.6829001237741598,
+      "grad_norm": 0.1601419408147955,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 28690
+    },
+    {
+      "epoch": 0.6829239264971912,
+      "grad_norm": 3.9567894201675826,
+      "learning_rate": 2e-05,
+      "loss": 0.5424,
+      "step": 28691
+    },
+    {
+      "epoch": 0.6829477292202228,
+      "grad_norm": 1.8216295556898459,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 28692
+    },
+    {
+      "epoch": 0.6829715319432543,
+      "grad_norm": 1.499896545801667,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 28693
+    },
+    {
+      "epoch": 0.6829953346662858,
+      "grad_norm": 0.27724020571101426,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 28694
+    },
+    {
+      "epoch": 0.6830191373893173,
+      "grad_norm": 47.124451819056915,
+      "learning_rate": 2e-05,
+      "loss": 0.1469,
+      "step": 28695
+    },
+    {
+      "epoch": 0.6830429401123489,
+      "grad_norm": 0.8919372581446682,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 28696
+    },
+    {
+      "epoch": 0.6830667428353804,
+      "grad_norm": 0.319033322026153,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 28697
+    },
+    {
+      "epoch": 0.6830905455584119,
+      "grad_norm": 6.421359623341736,
+      "learning_rate": 2e-05,
+      "loss": 0.6238,
+      "step": 28698
+    },
+    {
+      "epoch": 0.6831143482814434,
+      "grad_norm": 3.9139124135547183,
+      "learning_rate": 2e-05,
+      "loss": 0.7807,
+      "step": 28699
+    },
+    {
+      "epoch": 0.683138151004475,
+      "grad_norm": 4.296913029358109,
+      "learning_rate": 2e-05,
+      "loss": 0.9792,
+      "step": 28700
+    },
+    {
+      "epoch": 0.6831619537275064,
+      "grad_norm": 0.9091688600271588,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 28701
+    },
+    {
+      "epoch": 0.6831857564505379,
+      "grad_norm": 0.8124751714181274,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 28702
+    },
+    {
+      "epoch": 0.6832095591735694,
+      "grad_norm": 0.15851573388047063,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 28703
+    },
+    {
+      "epoch": 0.683233361896601,
+      "grad_norm": 13.093952707782675,
+      "learning_rate": 2e-05,
+      "loss": 1.617,
+      "step": 28704
+    },
+    {
+      "epoch": 0.6832571646196325,
+      "grad_norm": 4.227160561099287,
+      "learning_rate": 2e-05,
+      "loss": 0.3563,
+      "step": 28705
+    },
+    {
+      "epoch": 0.683280967342664,
+      "grad_norm": 3.851360072191623,
+      "learning_rate": 2e-05,
+      "loss": 0.135,
+      "step": 28706
+    },
+    {
+      "epoch": 0.6833047700656956,
+      "grad_norm": 2.1941831823015994,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 28707
+    },
+    {
+      "epoch": 0.683328572788727,
+      "grad_norm": 1.2250203742316776,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 28708
+    },
+    {
+      "epoch": 0.6833523755117585,
+      "grad_norm": 0.3982646546764324,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 28709
+    },
+    {
+      "epoch": 0.68337617823479,
+      "grad_norm": 0.9465746934584829,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 28710
+    },
+    {
+      "epoch": 0.6833999809578216,
+      "grad_norm": 3.0176193979232053,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 28711
+    },
+    {
+      "epoch": 0.6834237836808531,
+      "grad_norm": 7.136924432097084,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 28712
+    },
+    {
+      "epoch": 0.6834475864038846,
+      "grad_norm": 0.8908461477230563,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 28713
+    },
+    {
+      "epoch": 0.6834713891269161,
+      "grad_norm": 2.815694094327389,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 28714
+    },
+    {
+      "epoch": 0.6834951918499477,
+      "grad_norm": 0.10717923969868025,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 28715
+    },
+    {
+      "epoch": 0.6835189945729792,
+      "grad_norm": 2.8606236092069497,
+      "learning_rate": 2e-05,
+      "loss": 0.091,
+      "step": 28716
+    },
+    {
+      "epoch": 0.6835427972960106,
+      "grad_norm": 3.499112533848436,
+      "learning_rate": 2e-05,
+      "loss": 0.1933,
+      "step": 28717
+    },
+    {
+      "epoch": 0.6835666000190421,
+      "grad_norm": 0.10723838900015274,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 28718
+    },
+    {
+      "epoch": 0.6835904027420737,
+      "grad_norm": 0.2601718486679819,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 28719
+    },
+    {
+      "epoch": 0.6836142054651052,
+      "grad_norm": 38.09412539113641,
+      "learning_rate": 2e-05,
+      "loss": 0.341,
+      "step": 28720
+    },
+    {
+      "epoch": 0.6836380081881367,
+      "grad_norm": 3.9183823463419545,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 28721
+    },
+    {
+      "epoch": 0.6836618109111683,
+      "grad_norm": 6.356052881781001,
+      "learning_rate": 2e-05,
+      "loss": 1.1056,
+      "step": 28722
+    },
+    {
+      "epoch": 0.6836856136341998,
+      "grad_norm": 5.631392612041001,
+      "learning_rate": 2e-05,
+      "loss": 0.6027,
+      "step": 28723
+    },
+    {
+      "epoch": 0.6837094163572313,
+      "grad_norm": 2.3200350490802975,
+      "learning_rate": 2e-05,
+      "loss": 0.0742,
+      "step": 28724
+    },
+    {
+      "epoch": 0.6837332190802627,
+      "grad_norm": 6.657589239481836,
+      "learning_rate": 2e-05,
+      "loss": 0.5495,
+      "step": 28725
+    },
+    {
+      "epoch": 0.6837570218032943,
+      "grad_norm": 10.625728361500576,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 28726
+    },
+    {
+      "epoch": 0.6837808245263258,
+      "grad_norm": 2.1311616396482838,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 28727
+    },
+    {
+      "epoch": 0.6838046272493573,
+      "grad_norm": 1.1792427261980265,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 28728
+    },
+    {
+      "epoch": 0.6838284299723888,
+      "grad_norm": 12.167658609226063,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 28729
+    },
+    {
+      "epoch": 0.6838522326954204,
+      "grad_norm": 4.754068602355181,
+      "learning_rate": 2e-05,
+      "loss": 0.8264,
+      "step": 28730
+    },
+    {
+      "epoch": 0.6838760354184519,
+      "grad_norm": 18.911254144345055,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 28731
+    },
+    {
+      "epoch": 0.6838998381414834,
+      "grad_norm": 8.933326334709799,
+      "learning_rate": 2e-05,
+      "loss": 0.9277,
+      "step": 28732
+    },
+    {
+      "epoch": 0.6839236408645148,
+      "grad_norm": 2.6231912754230087,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 28733
+    },
+    {
+      "epoch": 0.6839474435875464,
+      "grad_norm": 2.994259713523311,
+      "learning_rate": 2e-05,
+      "loss": 0.5218,
+      "step": 28734
+    },
+    {
+      "epoch": 0.6839712463105779,
+      "grad_norm": 6.866318931923762,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 28735
+    },
+    {
+      "epoch": 0.6839950490336094,
+      "grad_norm": 1.7162372033866755,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 28736
+    },
+    {
+      "epoch": 0.684018851756641,
+      "grad_norm": 0.5461350269830516,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 28737
+    },
+    {
+      "epoch": 0.6840426544796725,
+      "grad_norm": 5.690168375570783,
+      "learning_rate": 2e-05,
+      "loss": 0.5072,
+      "step": 28738
+    },
+    {
+      "epoch": 0.684066457202704,
+      "grad_norm": 1.6857980739455545,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 28739
+    },
+    {
+      "epoch": 0.6840902599257355,
+      "grad_norm": 0.12258727271251417,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 28740
+    },
+    {
+      "epoch": 0.6841140626487671,
+      "grad_norm": 3.9361734294289494,
+      "learning_rate": 2e-05,
+      "loss": 0.5112,
+      "step": 28741
+    },
+    {
+      "epoch": 0.6841378653717985,
+      "grad_norm": 2.750475420986838,
+      "learning_rate": 2e-05,
+      "loss": 0.3349,
+      "step": 28742
+    },
+    {
+      "epoch": 0.68416166809483,
+      "grad_norm": 4.828840747475715,
+      "learning_rate": 2e-05,
+      "loss": 0.4996,
+      "step": 28743
+    },
+    {
+      "epoch": 0.6841854708178615,
+      "grad_norm": 0.8543095221913509,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 28744
+    },
+    {
+      "epoch": 0.6842092735408931,
+      "grad_norm": 3.5642925829750025,
+      "learning_rate": 2e-05,
+      "loss": 0.3435,
+      "step": 28745
+    },
+    {
+      "epoch": 0.6842330762639246,
+      "grad_norm": 2.546444524528104,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 28746
+    },
+    {
+      "epoch": 0.6842568789869561,
+      "grad_norm": 9.798063170774459,
+      "learning_rate": 2e-05,
+      "loss": 0.2898,
+      "step": 28747
+    },
+    {
+      "epoch": 0.6842806817099876,
+      "grad_norm": 3.4770706900760193,
+      "learning_rate": 2e-05,
+      "loss": 0.4084,
+      "step": 28748
+    },
+    {
+      "epoch": 0.6843044844330192,
+      "grad_norm": 6.216402619733497,
+      "learning_rate": 2e-05,
+      "loss": 0.7431,
+      "step": 28749
+    },
+    {
+      "epoch": 0.6843282871560507,
+      "grad_norm": 0.4689364864010354,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 28750
+    },
+    {
+      "epoch": 0.6843520898790821,
+      "grad_norm": 0.08454476933449553,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 28751
+    },
+    {
+      "epoch": 0.6843758926021137,
+      "grad_norm": 6.744219110254889,
+      "learning_rate": 2e-05,
+      "loss": 0.4228,
+      "step": 28752
+    },
+    {
+      "epoch": 0.6843996953251452,
+      "grad_norm": 3.2750602258152366,
+      "learning_rate": 2e-05,
+      "loss": 0.5472,
+      "step": 28753
+    },
+    {
+      "epoch": 0.6844234980481767,
+      "grad_norm": 0.7390601860628869,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 28754
+    },
+    {
+      "epoch": 0.6844473007712082,
+      "grad_norm": 4.582959203978588,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 28755
+    },
+    {
+      "epoch": 0.6844711034942398,
+      "grad_norm": 13.620006218743326,
+      "learning_rate": 2e-05,
+      "loss": 0.7216,
+      "step": 28756
+    },
+    {
+      "epoch": 0.6844949062172713,
+      "grad_norm": 1.9696076354143952,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 28757
+    },
+    {
+      "epoch": 0.6845187089403028,
+      "grad_norm": 5.542874668374289,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 28758
+    },
+    {
+      "epoch": 0.6845425116633342,
+      "grad_norm": 3.7796704081455132,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 28759
+    },
+    {
+      "epoch": 0.6845663143863658,
+      "grad_norm": 9.425591571945562,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 28760
+    },
+    {
+      "epoch": 0.6845901171093973,
+      "grad_norm": 5.296756977518024,
+      "learning_rate": 2e-05,
+      "loss": 0.3917,
+      "step": 28761
+    },
+    {
+      "epoch": 0.6846139198324288,
+      "grad_norm": 4.332030280967348,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 28762
+    },
+    {
+      "epoch": 0.6846377225554603,
+      "grad_norm": 5.619010255528445,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 28763
+    },
+    {
+      "epoch": 0.6846615252784919,
+      "grad_norm": 3.5620705997066273,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 28764
+    },
+    {
+      "epoch": 0.6846853280015234,
+      "grad_norm": 0.3942025101802431,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 28765
+    },
+    {
+      "epoch": 0.6847091307245549,
+      "grad_norm": 1.0419397361835965,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 28766
+    },
+    {
+      "epoch": 0.6847329334475865,
+      "grad_norm": 0.5228236648395764,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 28767
+    },
+    {
+      "epoch": 0.6847567361706179,
+      "grad_norm": 3.7575670091222837,
+      "learning_rate": 2e-05,
+      "loss": 0.3472,
+      "step": 28768
+    },
+    {
+      "epoch": 0.6847805388936494,
+      "grad_norm": 0.5346930557951716,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 28769
+    },
+    {
+      "epoch": 0.6848043416166809,
+      "grad_norm": 3.710346344920719,
+      "learning_rate": 2e-05,
+      "loss": 0.1916,
+      "step": 28770
+    },
+    {
+      "epoch": 0.6848281443397125,
+      "grad_norm": 0.06228958813566912,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 28771
+    },
+    {
+      "epoch": 0.684851947062744,
+      "grad_norm": 7.539174776894503,
+      "learning_rate": 2e-05,
+      "loss": 1.5364,
+      "step": 28772
+    },
+    {
+      "epoch": 0.6848757497857755,
+      "grad_norm": 2.5276519619751148,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 28773
+    },
+    {
+      "epoch": 0.684899552508807,
+      "grad_norm": 10.446536553029842,
+      "learning_rate": 2e-05,
+      "loss": 0.4103,
+      "step": 28774
+    },
+    {
+      "epoch": 0.6849233552318386,
+      "grad_norm": 4.745874635263456,
+      "learning_rate": 2e-05,
+      "loss": 0.4785,
+      "step": 28775
+    },
+    {
+      "epoch": 0.68494715795487,
+      "grad_norm": 6.52476419616309,
+      "learning_rate": 2e-05,
+      "loss": 0.537,
+      "step": 28776
+    },
+    {
+      "epoch": 0.6849709606779015,
+      "grad_norm": 5.364879004214783,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 28777
+    },
+    {
+      "epoch": 0.684994763400933,
+      "grad_norm": 12.513396473133469,
+      "learning_rate": 2e-05,
+      "loss": 0.3854,
+      "step": 28778
+    },
+    {
+      "epoch": 0.6850185661239646,
+      "grad_norm": 347.42585740110735,
+      "learning_rate": 2e-05,
+      "loss": 0.2899,
+      "step": 28779
+    },
+    {
+      "epoch": 0.6850423688469961,
+      "grad_norm": 4.1569372368494015,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 28780
+    },
+    {
+      "epoch": 0.6850661715700276,
+      "grad_norm": 2.8584724097206498,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 28781
+    },
+    {
+      "epoch": 0.6850899742930592,
+      "grad_norm": 3.399781736117187,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 28782
+    },
+    {
+      "epoch": 0.6851137770160907,
+      "grad_norm": 3.4698801743354357,
+      "learning_rate": 2e-05,
+      "loss": 0.2976,
+      "step": 28783
+    },
+    {
+      "epoch": 0.6851375797391221,
+      "grad_norm": 16.43630401253006,
+      "learning_rate": 2e-05,
+      "loss": 0.4862,
+      "step": 28784
+    },
+    {
+      "epoch": 0.6851613824621536,
+      "grad_norm": 6.191141342447762,
+      "learning_rate": 2e-05,
+      "loss": 0.6202,
+      "step": 28785
+    },
+    {
+      "epoch": 0.6851851851851852,
+      "grad_norm": 13.126554723202178,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 28786
+    },
+    {
+      "epoch": 0.6852089879082167,
+      "grad_norm": 3.6951732677819313,
+      "learning_rate": 2e-05,
+      "loss": 0.5917,
+      "step": 28787
+    },
+    {
+      "epoch": 0.6852327906312482,
+      "grad_norm": 11.156588638374446,
+      "learning_rate": 2e-05,
+      "loss": 0.6581,
+      "step": 28788
+    },
+    {
+      "epoch": 0.6852565933542797,
+      "grad_norm": 15.16276790679468,
+      "learning_rate": 2e-05,
+      "loss": 1.0933,
+      "step": 28789
+    },
+    {
+      "epoch": 0.6852803960773113,
+      "grad_norm": 0.6385191964991248,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 28790
+    },
+    {
+      "epoch": 0.6853041988003428,
+      "grad_norm": 3.2423762166469685,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 28791
+    },
+    {
+      "epoch": 0.6853280015233743,
+      "grad_norm": 4.860139826895289,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 28792
+    },
+    {
+      "epoch": 0.6853518042464057,
+      "grad_norm": 3.503024486758977,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 28793
+    },
+    {
+      "epoch": 0.6853756069694373,
+      "grad_norm": 28.824974734227297,
+      "learning_rate": 2e-05,
+      "loss": 0.9094,
+      "step": 28794
+    },
+    {
+      "epoch": 0.6853994096924688,
+      "grad_norm": 2.9549588472118744,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 28795
+    },
+    {
+      "epoch": 0.6854232124155003,
+      "grad_norm": 0.6424406052555989,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 28796
+    },
+    {
+      "epoch": 0.6854470151385319,
+      "grad_norm": 1.169085967322047,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 28797
+    },
+    {
+      "epoch": 0.6854708178615634,
+      "grad_norm": 3.130769761787938,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 28798
+    },
+    {
+      "epoch": 0.6854946205845949,
+      "grad_norm": 4.398010909801558,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 28799
+    },
+    {
+      "epoch": 0.6855184233076264,
+      "grad_norm": 9.420764681307263,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 28800
+    },
+    {
+      "epoch": 0.685542226030658,
+      "grad_norm": 2.8864587743627554,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 28801
+    },
+    {
+      "epoch": 0.6855660287536894,
+      "grad_norm": 8.388116624765143,
+      "learning_rate": 2e-05,
+      "loss": 0.9403,
+      "step": 28802
+    },
+    {
+      "epoch": 0.6855898314767209,
+      "grad_norm": 6.430962848667349,
+      "learning_rate": 2e-05,
+      "loss": 0.9095,
+      "step": 28803
+    },
+    {
+      "epoch": 0.6856136341997524,
+      "grad_norm": 5.127485250307903,
+      "learning_rate": 2e-05,
+      "loss": 0.8273,
+      "step": 28804
+    },
+    {
+      "epoch": 0.685637436922784,
+      "grad_norm": 2.7015128328295224,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 28805
+    },
+    {
+      "epoch": 0.6856612396458155,
+      "grad_norm": 4.954573898770593,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 28806
+    },
+    {
+      "epoch": 0.685685042368847,
+      "grad_norm": 2.826827080125316,
+      "learning_rate": 2e-05,
+      "loss": 0.3124,
+      "step": 28807
+    },
+    {
+      "epoch": 0.6857088450918785,
+      "grad_norm": 4.263730933831783,
+      "learning_rate": 2e-05,
+      "loss": 0.724,
+      "step": 28808
+    },
+    {
+      "epoch": 0.6857326478149101,
+      "grad_norm": 0.5427886428401526,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 28809
+    },
+    {
+      "epoch": 0.6857564505379415,
+      "grad_norm": 0.1009016167094704,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28810
+    },
+    {
+      "epoch": 0.685780253260973,
+      "grad_norm": 9.467990892270844,
+      "learning_rate": 2e-05,
+      "loss": 0.9465,
+      "step": 28811
+    },
+    {
+      "epoch": 0.6858040559840046,
+      "grad_norm": 10.636547814947003,
+      "learning_rate": 2e-05,
+      "loss": 1.3148,
+      "step": 28812
+    },
+    {
+      "epoch": 0.6858278587070361,
+      "grad_norm": 4.186592337102797,
+      "learning_rate": 2e-05,
+      "loss": 0.8106,
+      "step": 28813
+    },
+    {
+      "epoch": 0.6858516614300676,
+      "grad_norm": 4.629280390405842,
+      "learning_rate": 2e-05,
+      "loss": 1.0021,
+      "step": 28814
+    },
+    {
+      "epoch": 0.6858754641530991,
+      "grad_norm": 0.022320684438501423,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 28815
+    },
+    {
+      "epoch": 0.6858992668761307,
+      "grad_norm": 6.012596147084597,
+      "learning_rate": 2e-05,
+      "loss": 0.7625,
+      "step": 28816
+    },
+    {
+      "epoch": 0.6859230695991622,
+      "grad_norm": 7.785534224511008,
+      "learning_rate": 2e-05,
+      "loss": 0.2763,
+      "step": 28817
+    },
+    {
+      "epoch": 0.6859468723221936,
+      "grad_norm": 1.4975150204302976,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 28818
+    },
+    {
+      "epoch": 0.6859706750452251,
+      "grad_norm": 2.6504289368848593,
+      "learning_rate": 2e-05,
+      "loss": 0.3343,
+      "step": 28819
+    },
+    {
+      "epoch": 0.6859944777682567,
+      "grad_norm": 3.61338454458952,
+      "learning_rate": 2e-05,
+      "loss": 0.5163,
+      "step": 28820
+    },
+    {
+      "epoch": 0.6860182804912882,
+      "grad_norm": 9.531845146322869,
+      "learning_rate": 2e-05,
+      "loss": 1.1183,
+      "step": 28821
+    },
+    {
+      "epoch": 0.6860420832143197,
+      "grad_norm": 10.624842925583984,
+      "learning_rate": 2e-05,
+      "loss": 0.8946,
+      "step": 28822
+    },
+    {
+      "epoch": 0.6860658859373512,
+      "grad_norm": 1.63656745818049,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 28823
+    },
+    {
+      "epoch": 0.6860896886603828,
+      "grad_norm": 17.826804008105697,
+      "learning_rate": 2e-05,
+      "loss": 1.8948,
+      "step": 28824
+    },
+    {
+      "epoch": 0.6861134913834143,
+      "grad_norm": 7.634910220506139,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 28825
+    },
+    {
+      "epoch": 0.6861372941064457,
+      "grad_norm": 7.609951247298917,
+      "learning_rate": 2e-05,
+      "loss": 1.401,
+      "step": 28826
+    },
+    {
+      "epoch": 0.6861610968294773,
+      "grad_norm": 2.871931012546645,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 28827
+    },
+    {
+      "epoch": 0.6861848995525088,
+      "grad_norm": 4.101762012382873,
+      "learning_rate": 2e-05,
+      "loss": 0.7989,
+      "step": 28828
+    },
+    {
+      "epoch": 0.6862087022755403,
+      "grad_norm": 2.3033724335568904,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 28829
+    },
+    {
+      "epoch": 0.6862325049985718,
+      "grad_norm": 5.971977709878996,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 28830
+    },
+    {
+      "epoch": 0.6862563077216034,
+      "grad_norm": 0.6468206153630196,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 28831
+    },
+    {
+      "epoch": 0.6862801104446349,
+      "grad_norm": 12.135253371832206,
+      "learning_rate": 2e-05,
+      "loss": 1.0146,
+      "step": 28832
+    },
+    {
+      "epoch": 0.6863039131676664,
+      "grad_norm": 2.8885054054196506,
+      "learning_rate": 2e-05,
+      "loss": 0.2791,
+      "step": 28833
+    },
+    {
+      "epoch": 0.6863277158906979,
+      "grad_norm": 0.09587897058517286,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28834
+    },
+    {
+      "epoch": 0.6863515186137294,
+      "grad_norm": 138.37490489303792,
+      "learning_rate": 2e-05,
+      "loss": 1.584,
+      "step": 28835
+    },
+    {
+      "epoch": 0.6863753213367609,
+      "grad_norm": 7.788699876936466,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 28836
+    },
+    {
+      "epoch": 0.6863991240597924,
+      "grad_norm": 7.01774262018243,
+      "learning_rate": 2e-05,
+      "loss": 0.9565,
+      "step": 28837
+    },
+    {
+      "epoch": 0.6864229267828239,
+      "grad_norm": 8.50624009415269,
+      "learning_rate": 2e-05,
+      "loss": 0.3533,
+      "step": 28838
+    },
+    {
+      "epoch": 0.6864467295058555,
+      "grad_norm": 7.070829332686292,
+      "learning_rate": 2e-05,
+      "loss": 0.2284,
+      "step": 28839
+    },
+    {
+      "epoch": 0.686470532228887,
+      "grad_norm": 4.631875662055613,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 28840
+    },
+    {
+      "epoch": 0.6864943349519185,
+      "grad_norm": 2.124609806623161,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 28841
+    },
+    {
+      "epoch": 0.6865181376749501,
+      "grad_norm": 1.4927425547881639,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 28842
+    },
+    {
+      "epoch": 0.6865419403979816,
+      "grad_norm": 0.5711444047667328,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 28843
+    },
+    {
+      "epoch": 0.686565743121013,
+      "grad_norm": 0.25044307310218794,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 28844
+    },
+    {
+      "epoch": 0.6865895458440445,
+      "grad_norm": 6.834102518677564,
+      "learning_rate": 2e-05,
+      "loss": 0.0883,
+      "step": 28845
+    },
+    {
+      "epoch": 0.6866133485670761,
+      "grad_norm": 0.7884701521366246,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 28846
+    },
+    {
+      "epoch": 0.6866371512901076,
+      "grad_norm": 4.392603589492638,
+      "learning_rate": 2e-05,
+      "loss": 0.8047,
+      "step": 28847
+    },
+    {
+      "epoch": 0.6866609540131391,
+      "grad_norm": 0.4035845990035934,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 28848
+    },
+    {
+      "epoch": 0.6866847567361706,
+      "grad_norm": 1.511126019003594,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 28849
+    },
+    {
+      "epoch": 0.6867085594592022,
+      "grad_norm": 3.926735220584908,
+      "learning_rate": 2e-05,
+      "loss": 0.2644,
+      "step": 28850
+    },
+    {
+      "epoch": 0.6867323621822337,
+      "grad_norm": 3.4578859961739257,
+      "learning_rate": 2e-05,
+      "loss": 0.3477,
+      "step": 28851
+    },
+    {
+      "epoch": 0.6867561649052651,
+      "grad_norm": 4.579894662332732,
+      "learning_rate": 2e-05,
+      "loss": 1.0016,
+      "step": 28852
+    },
+    {
+      "epoch": 0.6867799676282966,
+      "grad_norm": 1.0553740959494695,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 28853
+    },
+    {
+      "epoch": 0.6868037703513282,
+      "grad_norm": 0.024861663712942423,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28854
+    },
+    {
+      "epoch": 0.6868275730743597,
+      "grad_norm": 2.5476995381690704,
+      "learning_rate": 2e-05,
+      "loss": 0.4409,
+      "step": 28855
+    },
+    {
+      "epoch": 0.6868513757973912,
+      "grad_norm": 1.4133372645367421,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 28856
+    },
+    {
+      "epoch": 0.6868751785204228,
+      "grad_norm": 4.670651441190185,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 28857
+    },
+    {
+      "epoch": 0.6868989812434543,
+      "grad_norm": 6.572690815156285,
+      "learning_rate": 2e-05,
+      "loss": 0.5676,
+      "step": 28858
+    },
+    {
+      "epoch": 0.6869227839664858,
+      "grad_norm": 4.530003651275919,
+      "learning_rate": 2e-05,
+      "loss": 0.8395,
+      "step": 28859
+    },
+    {
+      "epoch": 0.6869465866895172,
+      "grad_norm": 0.2723935182988896,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 28860
+    },
+    {
+      "epoch": 0.6869703894125488,
+      "grad_norm": 0.05423883827996421,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 28861
+    },
+    {
+      "epoch": 0.6869941921355803,
+      "grad_norm": 10.224962296170073,
+      "learning_rate": 2e-05,
+      "loss": 0.5953,
+      "step": 28862
+    },
+    {
+      "epoch": 0.6870179948586118,
+      "grad_norm": 4.366648417449944,
+      "learning_rate": 2e-05,
+      "loss": 0.4402,
+      "step": 28863
+    },
+    {
+      "epoch": 0.6870417975816433,
+      "grad_norm": 3.954581602636494,
+      "learning_rate": 2e-05,
+      "loss": 0.5178,
+      "step": 28864
+    },
+    {
+      "epoch": 0.6870656003046749,
+      "grad_norm": 0.4398821469724394,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 28865
+    },
+    {
+      "epoch": 0.6870894030277064,
+      "grad_norm": 24.130817349624067,
+      "learning_rate": 2e-05,
+      "loss": 0.7089,
+      "step": 28866
+    },
+    {
+      "epoch": 0.6871132057507379,
+      "grad_norm": 3.2994637013134884,
+      "learning_rate": 2e-05,
+      "loss": 0.5608,
+      "step": 28867
+    },
+    {
+      "epoch": 0.6871370084737694,
+      "grad_norm": 7.5121267047093925,
+      "learning_rate": 2e-05,
+      "loss": 0.4431,
+      "step": 28868
+    },
+    {
+      "epoch": 0.687160811196801,
+      "grad_norm": 9.487137098483652,
+      "learning_rate": 2e-05,
+      "loss": 1.1586,
+      "step": 28869
+    },
+    {
+      "epoch": 0.6871846139198324,
+      "grad_norm": 7.704847383506573,
+      "learning_rate": 2e-05,
+      "loss": 0.6844,
+      "step": 28870
+    },
+    {
+      "epoch": 0.6872084166428639,
+      "grad_norm": 4.777329415787229,
+      "learning_rate": 2e-05,
+      "loss": 0.0828,
+      "step": 28871
+    },
+    {
+      "epoch": 0.6872322193658955,
+      "grad_norm": 2.1284116862629907,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 28872
+    },
+    {
+      "epoch": 0.687256022088927,
+      "grad_norm": 0.721954299650894,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 28873
+    },
+    {
+      "epoch": 0.6872798248119585,
+      "grad_norm": 3.591426923323831,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 28874
+    },
+    {
+      "epoch": 0.68730362753499,
+      "grad_norm": 0.06992995196905553,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 28875
+    },
+    {
+      "epoch": 0.6873274302580216,
+      "grad_norm": 6.30554579328198,
+      "learning_rate": 2e-05,
+      "loss": 1.1107,
+      "step": 28876
+    },
+    {
+      "epoch": 0.687351232981053,
+      "grad_norm": 3.5807017968197474,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 28877
+    },
+    {
+      "epoch": 0.6873750357040845,
+      "grad_norm": 2.9762196602547917,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 28878
+    },
+    {
+      "epoch": 0.687398838427116,
+      "grad_norm": 7.013302930574768,
+      "learning_rate": 2e-05,
+      "loss": 1.4683,
+      "step": 28879
+    },
+    {
+      "epoch": 0.6874226411501476,
+      "grad_norm": 2.9361131533438884,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 28880
+    },
+    {
+      "epoch": 0.6874464438731791,
+      "grad_norm": 6.6900003781877535,
+      "learning_rate": 2e-05,
+      "loss": 1.0431,
+      "step": 28881
+    },
+    {
+      "epoch": 0.6874702465962106,
+      "grad_norm": 3.755919609777749,
+      "learning_rate": 2e-05,
+      "loss": 0.7432,
+      "step": 28882
+    },
+    {
+      "epoch": 0.6874940493192421,
+      "grad_norm": 1.1254787084508422,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 28883
+    },
+    {
+      "epoch": 0.6875178520422737,
+      "grad_norm": 5.692635110981289,
+      "learning_rate": 2e-05,
+      "loss": 0.1172,
+      "step": 28884
+    },
+    {
+      "epoch": 0.6875416547653052,
+      "grad_norm": 0.8803508579478773,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 28885
+    },
+    {
+      "epoch": 0.6875654574883366,
+      "grad_norm": 0.07104027216398755,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 28886
+    },
+    {
+      "epoch": 0.6875892602113681,
+      "grad_norm": 3.8977993154691855,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 28887
+    },
+    {
+      "epoch": 0.6876130629343997,
+      "grad_norm": 8.949965405687312,
+      "learning_rate": 2e-05,
+      "loss": 0.9297,
+      "step": 28888
+    },
+    {
+      "epoch": 0.6876368656574312,
+      "grad_norm": 0.09773266521529204,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28889
+    },
+    {
+      "epoch": 0.6876606683804627,
+      "grad_norm": 5.4946285459714055,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 28890
+    },
+    {
+      "epoch": 0.6876844711034943,
+      "grad_norm": 0.274658461530579,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 28891
+    },
+    {
+      "epoch": 0.6877082738265258,
+      "grad_norm": 3.8356872667267847,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 28892
+    },
+    {
+      "epoch": 0.6877320765495573,
+      "grad_norm": 0.8404428213742344,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 28893
+    },
+    {
+      "epoch": 0.6877558792725887,
+      "grad_norm": 1.594680647867401,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 28894
+    },
+    {
+      "epoch": 0.6877796819956203,
+      "grad_norm": 0.16702647955721509,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 28895
+    },
+    {
+      "epoch": 0.6878034847186518,
+      "grad_norm": 4.496897920524037,
+      "learning_rate": 2e-05,
+      "loss": 0.5595,
+      "step": 28896
+    },
+    {
+      "epoch": 0.6878272874416833,
+      "grad_norm": 0.04087629885248203,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28897
+    },
+    {
+      "epoch": 0.6878510901647148,
+      "grad_norm": 16.715375753628877,
+      "learning_rate": 2e-05,
+      "loss": 1.1038,
+      "step": 28898
+    },
+    {
+      "epoch": 0.6878748928877464,
+      "grad_norm": 0.023828357243098747,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 28899
+    },
+    {
+      "epoch": 0.6878986956107779,
+      "grad_norm": 4.224241915693455,
+      "learning_rate": 2e-05,
+      "loss": 0.6389,
+      "step": 28900
+    },
+    {
+      "epoch": 0.6879224983338094,
+      "grad_norm": 2.884461287330833,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 28901
+    },
+    {
+      "epoch": 0.6879463010568408,
+      "grad_norm": 1.6893587222343962,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 28902
+    },
+    {
+      "epoch": 0.6879701037798724,
+      "grad_norm": 0.648163841531085,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 28903
+    },
+    {
+      "epoch": 0.6879939065029039,
+      "grad_norm": 5.229538635037489,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 28904
+    },
+    {
+      "epoch": 0.6880177092259354,
+      "grad_norm": 27.856538736646986,
+      "learning_rate": 2e-05,
+      "loss": 0.6544,
+      "step": 28905
+    },
+    {
+      "epoch": 0.688041511948967,
+      "grad_norm": 2.5626106962240947,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 28906
+    },
+    {
+      "epoch": 0.6880653146719985,
+      "grad_norm": 1.880364909631092,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 28907
+    },
+    {
+      "epoch": 0.68808911739503,
+      "grad_norm": 0.08842178008689121,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 28908
+    },
+    {
+      "epoch": 0.6881129201180615,
+      "grad_norm": 0.2930205889219916,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 28909
+    },
+    {
+      "epoch": 0.6881367228410931,
+      "grad_norm": 4.484334104584648,
+      "learning_rate": 2e-05,
+      "loss": 0.8538,
+      "step": 28910
+    },
+    {
+      "epoch": 0.6881605255641245,
+      "grad_norm": 282.5530616851548,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 28911
+    },
+    {
+      "epoch": 0.688184328287156,
+      "grad_norm": 0.6851415769873562,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 28912
+    },
+    {
+      "epoch": 0.6882081310101875,
+      "grad_norm": 1.6820918359526476,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 28913
+    },
+    {
+      "epoch": 0.6882319337332191,
+      "grad_norm": 0.029721250027325664,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 28914
+    },
+    {
+      "epoch": 0.6882557364562506,
+      "grad_norm": 3.2819633443846765,
+      "learning_rate": 2e-05,
+      "loss": 0.3157,
+      "step": 28915
+    },
+    {
+      "epoch": 0.6882795391792821,
+      "grad_norm": 0.12801244598102238,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 28916
+    },
+    {
+      "epoch": 0.6883033419023136,
+      "grad_norm": 4.868897069698561,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 28917
+    },
+    {
+      "epoch": 0.6883271446253452,
+      "grad_norm": 2.0474990530983117,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 28918
+    },
+    {
+      "epoch": 0.6883509473483767,
+      "grad_norm": 4.976515957950099,
+      "learning_rate": 2e-05,
+      "loss": 0.3921,
+      "step": 28919
+    },
+    {
+      "epoch": 0.6883747500714081,
+      "grad_norm": 3.534152358683409,
+      "learning_rate": 2e-05,
+      "loss": 0.8713,
+      "step": 28920
+    },
+    {
+      "epoch": 0.6883985527944397,
+      "grad_norm": 2.6085812758421274,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 28921
+    },
+    {
+      "epoch": 0.6884223555174712,
+      "grad_norm": 4.751302732592345,
+      "learning_rate": 2e-05,
+      "loss": 0.9382,
+      "step": 28922
+    },
+    {
+      "epoch": 0.6884461582405027,
+      "grad_norm": 3.141648371198827,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 28923
+    },
+    {
+      "epoch": 0.6884699609635342,
+      "grad_norm": 4.050384511904402,
+      "learning_rate": 2e-05,
+      "loss": 0.482,
+      "step": 28924
+    },
+    {
+      "epoch": 0.6884937636865658,
+      "grad_norm": 9.792729112164062,
+      "learning_rate": 2e-05,
+      "loss": 0.3289,
+      "step": 28925
+    },
+    {
+      "epoch": 0.6885175664095973,
+      "grad_norm": 0.15637395040915053,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 28926
+    },
+    {
+      "epoch": 0.6885413691326288,
+      "grad_norm": 0.020281943520083952,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28927
+    },
+    {
+      "epoch": 0.6885651718556602,
+      "grad_norm": 4.4192533848572335,
+      "learning_rate": 2e-05,
+      "loss": 0.4551,
+      "step": 28928
+    },
+    {
+      "epoch": 0.6885889745786918,
+      "grad_norm": 2.8446636768419644,
+      "learning_rate": 2e-05,
+      "loss": 0.5499,
+      "step": 28929
+    },
+    {
+      "epoch": 0.6886127773017233,
+      "grad_norm": 19.59773620591775,
+      "learning_rate": 2e-05,
+      "loss": 0.5419,
+      "step": 28930
+    },
+    {
+      "epoch": 0.6886365800247548,
+      "grad_norm": 9.488991677772065,
+      "learning_rate": 2e-05,
+      "loss": 0.5204,
+      "step": 28931
+    },
+    {
+      "epoch": 0.6886603827477863,
+      "grad_norm": 6.6637408697199705,
+      "learning_rate": 2e-05,
+      "loss": 0.4562,
+      "step": 28932
+    },
+    {
+      "epoch": 0.6886841854708179,
+      "grad_norm": 0.43657420323463175,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 28933
+    },
+    {
+      "epoch": 0.6887079881938494,
+      "grad_norm": 2.8396997986720818,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 28934
+    },
+    {
+      "epoch": 0.6887317909168809,
+      "grad_norm": 4.980383970409527,
+      "learning_rate": 2e-05,
+      "loss": 0.9425,
+      "step": 28935
+    },
+    {
+      "epoch": 0.6887555936399125,
+      "grad_norm": 5.466145768596472,
+      "learning_rate": 2e-05,
+      "loss": 0.5468,
+      "step": 28936
+    },
+    {
+      "epoch": 0.6887793963629439,
+      "grad_norm": 2.4254415074570903,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 28937
+    },
+    {
+      "epoch": 0.6888031990859754,
+      "grad_norm": 5.7861158605598755,
+      "learning_rate": 2e-05,
+      "loss": 0.8816,
+      "step": 28938
+    },
+    {
+      "epoch": 0.6888270018090069,
+      "grad_norm": 3.160011512106332,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 28939
+    },
+    {
+      "epoch": 0.6888508045320385,
+      "grad_norm": 0.10815389233310763,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 28940
+    },
+    {
+      "epoch": 0.68887460725507,
+      "grad_norm": 5.49391219810067,
+      "learning_rate": 2e-05,
+      "loss": 0.7131,
+      "step": 28941
+    },
+    {
+      "epoch": 0.6888984099781015,
+      "grad_norm": 0.1666544453600315,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 28942
+    },
+    {
+      "epoch": 0.688922212701133,
+      "grad_norm": 0.8131332272911447,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 28943
+    },
+    {
+      "epoch": 0.6889460154241646,
+      "grad_norm": 26.64712209887644,
+      "learning_rate": 2e-05,
+      "loss": 0.196,
+      "step": 28944
+    },
+    {
+      "epoch": 0.688969818147196,
+      "grad_norm": 12.124516004822215,
+      "learning_rate": 2e-05,
+      "loss": 1.2282,
+      "step": 28945
+    },
+    {
+      "epoch": 0.6889936208702275,
+      "grad_norm": 0.26337458430342103,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 28946
+    },
+    {
+      "epoch": 0.689017423593259,
+      "grad_norm": 5.311416872407082,
+      "learning_rate": 2e-05,
+      "loss": 0.557,
+      "step": 28947
+    },
+    {
+      "epoch": 0.6890412263162906,
+      "grad_norm": 1.979613950134592,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 28948
+    },
+    {
+      "epoch": 0.6890650290393221,
+      "grad_norm": 2.615097622875723,
+      "learning_rate": 2e-05,
+      "loss": 0.4662,
+      "step": 28949
+    },
+    {
+      "epoch": 0.6890888317623536,
+      "grad_norm": 0.40944261858572567,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 28950
+    },
+    {
+      "epoch": 0.6891126344853852,
+      "grad_norm": 5.670900214348133,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 28951
+    },
+    {
+      "epoch": 0.6891364372084167,
+      "grad_norm": 4.7113229391350595,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 28952
+    },
+    {
+      "epoch": 0.6891602399314481,
+      "grad_norm": 4.8240218975742115,
+      "learning_rate": 2e-05,
+      "loss": 0.6395,
+      "step": 28953
+    },
+    {
+      "epoch": 0.6891840426544796,
+      "grad_norm": 2.616726132668772,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 28954
+    },
+    {
+      "epoch": 0.6892078453775112,
+      "grad_norm": 2.4195998748190473,
+      "learning_rate": 2e-05,
+      "loss": 0.0463,
+      "step": 28955
+    },
+    {
+      "epoch": 0.6892316481005427,
+      "grad_norm": 0.534271493610044,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 28956
+    },
+    {
+      "epoch": 0.6892554508235742,
+      "grad_norm": 0.17235852131812948,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 28957
+    },
+    {
+      "epoch": 0.6892792535466057,
+      "grad_norm": 8.577570279648008,
+      "learning_rate": 2e-05,
+      "loss": 0.9955,
+      "step": 28958
+    },
+    {
+      "epoch": 0.6893030562696373,
+      "grad_norm": 3.571431997020241,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 28959
+    },
+    {
+      "epoch": 0.6893268589926688,
+      "grad_norm": 0.0117050541830611,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 28960
+    },
+    {
+      "epoch": 0.6893506617157003,
+      "grad_norm": 5.487865411537518,
+      "learning_rate": 2e-05,
+      "loss": 1.0252,
+      "step": 28961
+    },
+    {
+      "epoch": 0.6893744644387317,
+      "grad_norm": 2.0439994849503953,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 28962
+    },
+    {
+      "epoch": 0.6893982671617633,
+      "grad_norm": 3.996835993066745,
+      "learning_rate": 2e-05,
+      "loss": 0.7258,
+      "step": 28963
+    },
+    {
+      "epoch": 0.6894220698847948,
+      "grad_norm": 2.9801070561855574,
+      "learning_rate": 2e-05,
+      "loss": 0.3517,
+      "step": 28964
+    },
+    {
+      "epoch": 0.6894458726078263,
+      "grad_norm": 0.1644893165683954,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 28965
+    },
+    {
+      "epoch": 0.6894696753308579,
+      "grad_norm": 3.220029595431465,
+      "learning_rate": 2e-05,
+      "loss": 0.3493,
+      "step": 28966
+    },
+    {
+      "epoch": 0.6894934780538894,
+      "grad_norm": 3.348534960515164,
+      "learning_rate": 2e-05,
+      "loss": 0.3998,
+      "step": 28967
+    },
+    {
+      "epoch": 0.6895172807769209,
+      "grad_norm": 3.3128840908169117,
+      "learning_rate": 2e-05,
+      "loss": 0.535,
+      "step": 28968
+    },
+    {
+      "epoch": 0.6895410834999524,
+      "grad_norm": 0.31991880501767955,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 28969
+    },
+    {
+      "epoch": 0.689564886222984,
+      "grad_norm": 3.737508232704426,
+      "learning_rate": 2e-05,
+      "loss": 0.0738,
+      "step": 28970
+    },
+    {
+      "epoch": 0.6895886889460154,
+      "grad_norm": 9.349352212986204,
+      "learning_rate": 2e-05,
+      "loss": 0.6469,
+      "step": 28971
+    },
+    {
+      "epoch": 0.6896124916690469,
+      "grad_norm": 0.8815304557093091,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 28972
+    },
+    {
+      "epoch": 0.6896362943920784,
+      "grad_norm": 4.326551743588776,
+      "learning_rate": 2e-05,
+      "loss": 0.1293,
+      "step": 28973
+    },
+    {
+      "epoch": 0.68966009711511,
+      "grad_norm": 4.567541527844841,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 28974
+    },
+    {
+      "epoch": 0.6896838998381415,
+      "grad_norm": 0.029845629308420687,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 28975
+    },
+    {
+      "epoch": 0.689707702561173,
+      "grad_norm": 0.2678208674957282,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 28976
+    },
+    {
+      "epoch": 0.6897315052842045,
+      "grad_norm": 10.402345987207267,
+      "learning_rate": 2e-05,
+      "loss": 0.9704,
+      "step": 28977
+    },
+    {
+      "epoch": 0.6897553080072361,
+      "grad_norm": 12.010933325085476,
+      "learning_rate": 2e-05,
+      "loss": 1.6211,
+      "step": 28978
+    },
+    {
+      "epoch": 0.6897791107302675,
+      "grad_norm": 2.5354943367776737,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 28979
+    },
+    {
+      "epoch": 0.689802913453299,
+      "grad_norm": 3.531688659528307,
+      "learning_rate": 2e-05,
+      "loss": 0.4568,
+      "step": 28980
+    },
+    {
+      "epoch": 0.6898267161763306,
+      "grad_norm": 0.642590580102296,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 28981
+    },
+    {
+      "epoch": 0.6898505188993621,
+      "grad_norm": 0.5275489647269543,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 28982
+    },
+    {
+      "epoch": 0.6898743216223936,
+      "grad_norm": 4.713736111488174,
+      "learning_rate": 2e-05,
+      "loss": 1.4177,
+      "step": 28983
+    },
+    {
+      "epoch": 0.6898981243454251,
+      "grad_norm": 6.126268187823148,
+      "learning_rate": 2e-05,
+      "loss": 0.0815,
+      "step": 28984
+    },
+    {
+      "epoch": 0.6899219270684567,
+      "grad_norm": 6.7927304233926735,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 28985
+    },
+    {
+      "epoch": 0.6899457297914882,
+      "grad_norm": 0.16949959694964128,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 28986
+    },
+    {
+      "epoch": 0.6899695325145196,
+      "grad_norm": 4.363514325529489,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 28987
+    },
+    {
+      "epoch": 0.6899933352375511,
+      "grad_norm": 5.344590789994741,
+      "learning_rate": 2e-05,
+      "loss": 0.9882,
+      "step": 28988
+    },
+    {
+      "epoch": 0.6900171379605827,
+      "grad_norm": 7.38704201862459,
+      "learning_rate": 2e-05,
+      "loss": 0.7367,
+      "step": 28989
+    },
+    {
+      "epoch": 0.6900409406836142,
+      "grad_norm": 7.777324681037955,
+      "learning_rate": 2e-05,
+      "loss": 0.8801,
+      "step": 28990
+    },
+    {
+      "epoch": 0.6900647434066457,
+      "grad_norm": 4.829836578868733,
+      "learning_rate": 2e-05,
+      "loss": 0.9217,
+      "step": 28991
+    },
+    {
+      "epoch": 0.6900885461296772,
+      "grad_norm": 6.919613303166923,
+      "learning_rate": 2e-05,
+      "loss": 1.0283,
+      "step": 28992
+    },
+    {
+      "epoch": 0.6901123488527088,
+      "grad_norm": 4.773369055514263,
+      "learning_rate": 2e-05,
+      "loss": 0.7714,
+      "step": 28993
+    },
+    {
+      "epoch": 0.6901361515757403,
+      "grad_norm": 0.016036615166261772,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 28994
+    },
+    {
+      "epoch": 0.6901599542987717,
+      "grad_norm": 0.14162020569485737,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 28995
+    },
+    {
+      "epoch": 0.6901837570218033,
+      "grad_norm": 0.03803491190702125,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 28996
+    },
+    {
+      "epoch": 0.6902075597448348,
+      "grad_norm": 4.598723469300322,
+      "learning_rate": 2e-05,
+      "loss": 1.0828,
+      "step": 28997
+    },
+    {
+      "epoch": 0.6902313624678663,
+      "grad_norm": 4.25963566204606,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 28998
+    },
+    {
+      "epoch": 0.6902551651908978,
+      "grad_norm": 0.3566702393861953,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 28999
+    },
+    {
+      "epoch": 0.6902789679139294,
+      "grad_norm": 4.588467101449099,
+      "learning_rate": 2e-05,
+      "loss": 0.8879,
+      "step": 29000
+    },
+    {
+      "epoch": 0.6903027706369609,
+      "grad_norm": 2.8624277303340495,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 29001
+    },
+    {
+      "epoch": 0.6903265733599924,
+      "grad_norm": 6.167923235747227,
+      "learning_rate": 2e-05,
+      "loss": 0.9129,
+      "step": 29002
+    },
+    {
+      "epoch": 0.6903503760830239,
+      "grad_norm": 12.993849811403576,
+      "learning_rate": 2e-05,
+      "loss": 1.8665,
+      "step": 29003
+    },
+    {
+      "epoch": 0.6903741788060554,
+      "grad_norm": 4.549928083290521,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 29004
+    },
+    {
+      "epoch": 0.6903979815290869,
+      "grad_norm": 0.3402045296694531,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 29005
+    },
+    {
+      "epoch": 0.6904217842521184,
+      "grad_norm": 5.514711372722465,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 29006
+    },
+    {
+      "epoch": 0.6904455869751499,
+      "grad_norm": 14.540492649451151,
+      "learning_rate": 2e-05,
+      "loss": 0.7327,
+      "step": 29007
+    },
+    {
+      "epoch": 0.6904693896981815,
+      "grad_norm": 0.419193628936917,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 29008
+    },
+    {
+      "epoch": 0.690493192421213,
+      "grad_norm": 4.0662545355574755,
+      "learning_rate": 2e-05,
+      "loss": 0.7228,
+      "step": 29009
+    },
+    {
+      "epoch": 0.6905169951442445,
+      "grad_norm": 9.951215195313289,
+      "learning_rate": 2e-05,
+      "loss": 0.3065,
+      "step": 29010
+    },
+    {
+      "epoch": 0.6905407978672761,
+      "grad_norm": 0.14731861940367943,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 29011
+    },
+    {
+      "epoch": 0.6905646005903076,
+      "grad_norm": 1.898265501946642,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 29012
+    },
+    {
+      "epoch": 0.690588403313339,
+      "grad_norm": 3.8525590845654363,
+      "learning_rate": 2e-05,
+      "loss": 0.4098,
+      "step": 29013
+    },
+    {
+      "epoch": 0.6906122060363705,
+      "grad_norm": 5.073475340349202,
+      "learning_rate": 2e-05,
+      "loss": 0.6855,
+      "step": 29014
+    },
+    {
+      "epoch": 0.6906360087594021,
+      "grad_norm": 8.382456579628657,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 29015
+    },
+    {
+      "epoch": 0.6906598114824336,
+      "grad_norm": 6.204078618777322,
+      "learning_rate": 2e-05,
+      "loss": 1.0488,
+      "step": 29016
+    },
+    {
+      "epoch": 0.6906836142054651,
+      "grad_norm": 3.452584042990171,
+      "learning_rate": 2e-05,
+      "loss": 0.3315,
+      "step": 29017
+    },
+    {
+      "epoch": 0.6907074169284966,
+      "grad_norm": 3.4411730699703345,
+      "learning_rate": 2e-05,
+      "loss": 0.4528,
+      "step": 29018
+    },
+    {
+      "epoch": 0.6907312196515282,
+      "grad_norm": 9.362130977671674,
+      "learning_rate": 2e-05,
+      "loss": 0.5914,
+      "step": 29019
+    },
+    {
+      "epoch": 0.6907550223745597,
+      "grad_norm": 17.70647576686076,
+      "learning_rate": 2e-05,
+      "loss": 0.1991,
+      "step": 29020
+    },
+    {
+      "epoch": 0.6907788250975911,
+      "grad_norm": 4.704022000773031,
+      "learning_rate": 2e-05,
+      "loss": 0.5972,
+      "step": 29021
+    },
+    {
+      "epoch": 0.6908026278206226,
+      "grad_norm": 8.924209634305411,
+      "learning_rate": 2e-05,
+      "loss": 0.6095,
+      "step": 29022
+    },
+    {
+      "epoch": 0.6908264305436542,
+      "grad_norm": 4.076387457897926,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 29023
+    },
+    {
+      "epoch": 0.6908502332666857,
+      "grad_norm": 5.48348740857115,
+      "learning_rate": 2e-05,
+      "loss": 0.8782,
+      "step": 29024
+    },
+    {
+      "epoch": 0.6908740359897172,
+      "grad_norm": 3.7150715458358903,
+      "learning_rate": 2e-05,
+      "loss": 0.7568,
+      "step": 29025
+    },
+    {
+      "epoch": 0.6908978387127488,
+      "grad_norm": 0.45569712506072985,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 29026
+    },
+    {
+      "epoch": 0.6909216414357803,
+      "grad_norm": 0.015104912382196906,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 29027
+    },
+    {
+      "epoch": 0.6909454441588118,
+      "grad_norm": 3.60663747751876,
+      "learning_rate": 2e-05,
+      "loss": 0.9275,
+      "step": 29028
+    },
+    {
+      "epoch": 0.6909692468818432,
+      "grad_norm": 35.0244307452904,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 29029
+    },
+    {
+      "epoch": 0.6909930496048748,
+      "grad_norm": 4.061038912792951,
+      "learning_rate": 2e-05,
+      "loss": 0.627,
+      "step": 29030
+    },
+    {
+      "epoch": 0.6910168523279063,
+      "grad_norm": 0.598448131995005,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 29031
+    },
+    {
+      "epoch": 0.6910406550509378,
+      "grad_norm": 7.025247666079902,
+      "learning_rate": 2e-05,
+      "loss": 0.3138,
+      "step": 29032
+    },
+    {
+      "epoch": 0.6910644577739693,
+      "grad_norm": 5.243137228867139,
+      "learning_rate": 2e-05,
+      "loss": 0.8267,
+      "step": 29033
+    },
+    {
+      "epoch": 0.6910882604970009,
+      "grad_norm": 83.05955258229955,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 29034
+    },
+    {
+      "epoch": 0.6911120632200324,
+      "grad_norm": 5.532871495824296,
+      "learning_rate": 2e-05,
+      "loss": 0.9804,
+      "step": 29035
+    },
+    {
+      "epoch": 0.6911358659430639,
+      "grad_norm": 3.8781170360723523,
+      "learning_rate": 2e-05,
+      "loss": 0.7348,
+      "step": 29036
+    },
+    {
+      "epoch": 0.6911596686660954,
+      "grad_norm": 1.9918448883964182,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 29037
+    },
+    {
+      "epoch": 0.691183471389127,
+      "grad_norm": 3.654292103090043,
+      "learning_rate": 2e-05,
+      "loss": 0.6774,
+      "step": 29038
+    },
+    {
+      "epoch": 0.6912072741121584,
+      "grad_norm": 2.9698475771889785,
+      "learning_rate": 2e-05,
+      "loss": 0.4893,
+      "step": 29039
+    },
+    {
+      "epoch": 0.6912310768351899,
+      "grad_norm": 2.824400754926261,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 29040
+    },
+    {
+      "epoch": 0.6912548795582215,
+      "grad_norm": 9.10840046972903,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 29041
+    },
+    {
+      "epoch": 0.691278682281253,
+      "grad_norm": 5.951207227475816,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 29042
+    },
+    {
+      "epoch": 0.6913024850042845,
+      "grad_norm": 5.830811257819962,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 29043
+    },
+    {
+      "epoch": 0.691326287727316,
+      "grad_norm": 0.023579918455565382,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29044
+    },
+    {
+      "epoch": 0.6913500904503476,
+      "grad_norm": 5.327055176756201,
+      "learning_rate": 2e-05,
+      "loss": 0.8657,
+      "step": 29045
+    },
+    {
+      "epoch": 0.691373893173379,
+      "grad_norm": 10.944309010140374,
+      "learning_rate": 2e-05,
+      "loss": 0.293,
+      "step": 29046
+    },
+    {
+      "epoch": 0.6913976958964105,
+      "grad_norm": 1.6841869553303623,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 29047
+    },
+    {
+      "epoch": 0.691421498619442,
+      "grad_norm": 2.857169298801387,
+      "learning_rate": 2e-05,
+      "loss": 0.3749,
+      "step": 29048
+    },
+    {
+      "epoch": 0.6914453013424736,
+      "grad_norm": 1.8153748394918408,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 29049
+    },
+    {
+      "epoch": 0.6914691040655051,
+      "grad_norm": 0.05536274248018322,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 29050
+    },
+    {
+      "epoch": 0.6914929067885366,
+      "grad_norm": 0.027017045557874365,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29051
+    },
+    {
+      "epoch": 0.6915167095115681,
+      "grad_norm": 7.2779637745277235,
+      "learning_rate": 2e-05,
+      "loss": 1.3015,
+      "step": 29052
+    },
+    {
+      "epoch": 0.6915405122345997,
+      "grad_norm": 0.1277522324956269,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29053
+    },
+    {
+      "epoch": 0.6915643149576312,
+      "grad_norm": 3.9102036180761575,
+      "learning_rate": 2e-05,
+      "loss": 0.674,
+      "step": 29054
+    },
+    {
+      "epoch": 0.6915881176806626,
+      "grad_norm": 35.015802375665025,
+      "learning_rate": 2e-05,
+      "loss": 4.6184,
+      "step": 29055
+    },
+    {
+      "epoch": 0.6916119204036942,
+      "grad_norm": 0.22034883253463056,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 29056
+    },
+    {
+      "epoch": 0.6916357231267257,
+      "grad_norm": 4.45721188058721,
+      "learning_rate": 2e-05,
+      "loss": 0.6843,
+      "step": 29057
+    },
+    {
+      "epoch": 0.6916595258497572,
+      "grad_norm": 0.16446426923203844,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 29058
+    },
+    {
+      "epoch": 0.6916833285727887,
+      "grad_norm": 4.482975923762162,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 29059
+    },
+    {
+      "epoch": 0.6917071312958203,
+      "grad_norm": 0.02769214928674618,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29060
+    },
+    {
+      "epoch": 0.6917309340188518,
+      "grad_norm": 3.8966927015723103,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 29061
+    },
+    {
+      "epoch": 0.6917547367418833,
+      "grad_norm": 4.189510060583659,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 29062
+    },
+    {
+      "epoch": 0.6917785394649147,
+      "grad_norm": 4.51491328424393,
+      "learning_rate": 2e-05,
+      "loss": 0.8844,
+      "step": 29063
+    },
+    {
+      "epoch": 0.6918023421879463,
+      "grad_norm": 3.263313419989509,
+      "learning_rate": 2e-05,
+      "loss": 0.6165,
+      "step": 29064
+    },
+    {
+      "epoch": 0.6918261449109778,
+      "grad_norm": 2.3458493858969702,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 29065
+    },
+    {
+      "epoch": 0.6918499476340093,
+      "grad_norm": 4.544777729584745,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 29066
+    },
+    {
+      "epoch": 0.6918737503570408,
+      "grad_norm": 5.308845317583563,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 29067
+    },
+    {
+      "epoch": 0.6918975530800724,
+      "grad_norm": 4.825740178246654,
+      "learning_rate": 2e-05,
+      "loss": 0.6594,
+      "step": 29068
+    },
+    {
+      "epoch": 0.6919213558031039,
+      "grad_norm": 0.04360738341886166,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29069
+    },
+    {
+      "epoch": 0.6919451585261354,
+      "grad_norm": 4.462879583687821,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 29070
+    },
+    {
+      "epoch": 0.691968961249167,
+      "grad_norm": 0.32686912840755067,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 29071
+    },
+    {
+      "epoch": 0.6919927639721984,
+      "grad_norm": 3.4142164981023493,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 29072
+    },
+    {
+      "epoch": 0.6920165666952299,
+      "grad_norm": 11.101440326196396,
+      "learning_rate": 2e-05,
+      "loss": 0.8879,
+      "step": 29073
+    },
+    {
+      "epoch": 0.6920403694182614,
+      "grad_norm": 1.9979553451670924,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 29074
+    },
+    {
+      "epoch": 0.692064172141293,
+      "grad_norm": 2.606976415318034,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 29075
+    },
+    {
+      "epoch": 0.6920879748643245,
+      "grad_norm": 0.16751331912479142,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 29076
+    },
+    {
+      "epoch": 0.692111777587356,
+      "grad_norm": 0.25015713050268723,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 29077
+    },
+    {
+      "epoch": 0.6921355803103875,
+      "grad_norm": 10.552637213645225,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 29078
+    },
+    {
+      "epoch": 0.6921593830334191,
+      "grad_norm": 5.553356099239293,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 29079
+    },
+    {
+      "epoch": 0.6921831857564505,
+      "grad_norm": 3.2459093250434754,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 29080
+    },
+    {
+      "epoch": 0.692206988479482,
+      "grad_norm": 10.245874359744974,
+      "learning_rate": 2e-05,
+      "loss": 0.691,
+      "step": 29081
+    },
+    {
+      "epoch": 0.6922307912025135,
+      "grad_norm": 3.5038844461341068,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 29082
+    },
+    {
+      "epoch": 0.6922545939255451,
+      "grad_norm": 2.276235432642362,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 29083
+    },
+    {
+      "epoch": 0.6922783966485766,
+      "grad_norm": 7.159868411213283,
+      "learning_rate": 2e-05,
+      "loss": 0.464,
+      "step": 29084
+    },
+    {
+      "epoch": 0.6923021993716081,
+      "grad_norm": 2.826152456098662,
+      "learning_rate": 2e-05,
+      "loss": 0.1804,
+      "step": 29085
+    },
+    {
+      "epoch": 0.6923260020946397,
+      "grad_norm": 0.9071207312867913,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 29086
+    },
+    {
+      "epoch": 0.6923498048176712,
+      "grad_norm": 1.8465269156346897,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 29087
+    },
+    {
+      "epoch": 0.6923736075407027,
+      "grad_norm": 3.060855426760457,
+      "learning_rate": 2e-05,
+      "loss": 0.3638,
+      "step": 29088
+    },
+    {
+      "epoch": 0.6923974102637341,
+      "grad_norm": 6.012752991901008,
+      "learning_rate": 2e-05,
+      "loss": 1.1319,
+      "step": 29089
+    },
+    {
+      "epoch": 0.6924212129867657,
+      "grad_norm": 2.168447071216748,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 29090
+    },
+    {
+      "epoch": 0.6924450157097972,
+      "grad_norm": 3.099480691592904,
+      "learning_rate": 2e-05,
+      "loss": 0.4241,
+      "step": 29091
+    },
+    {
+      "epoch": 0.6924688184328287,
+      "grad_norm": 4.404610916523687,
+      "learning_rate": 2e-05,
+      "loss": 0.4721,
+      "step": 29092
+    },
+    {
+      "epoch": 0.6924926211558602,
+      "grad_norm": 7.8208377119799355,
+      "learning_rate": 2e-05,
+      "loss": 0.2308,
+      "step": 29093
+    },
+    {
+      "epoch": 0.6925164238788918,
+      "grad_norm": 5.919670921859466,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 29094
+    },
+    {
+      "epoch": 0.6925402266019233,
+      "grad_norm": 10.622811525973015,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 29095
+    },
+    {
+      "epoch": 0.6925640293249548,
+      "grad_norm": 0.40358713832189064,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 29096
+    },
+    {
+      "epoch": 0.6925878320479862,
+      "grad_norm": 0.3104686191806452,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 29097
+    },
+    {
+      "epoch": 0.6926116347710178,
+      "grad_norm": 4.122279347749108,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 29098
+    },
+    {
+      "epoch": 0.6926354374940493,
+      "grad_norm": 3.2775904533019884,
+      "learning_rate": 2e-05,
+      "loss": 0.433,
+      "step": 29099
+    },
+    {
+      "epoch": 0.6926592402170808,
+      "grad_norm": 2.7553180214474318,
+      "learning_rate": 2e-05,
+      "loss": 0.5356,
+      "step": 29100
+    },
+    {
+      "epoch": 0.6926830429401124,
+      "grad_norm": 0.025385426470459986,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29101
+    },
+    {
+      "epoch": 0.6927068456631439,
+      "grad_norm": 1.6958626410187123,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 29102
+    },
+    {
+      "epoch": 0.6927306483861754,
+      "grad_norm": 0.096937779274906,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 29103
+    },
+    {
+      "epoch": 0.6927544511092069,
+      "grad_norm": 0.9578950131281749,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 29104
+    },
+    {
+      "epoch": 0.6927782538322385,
+      "grad_norm": 0.16226047052779963,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 29105
+    },
+    {
+      "epoch": 0.6928020565552699,
+      "grad_norm": 8.810001324173845,
+      "learning_rate": 2e-05,
+      "loss": 0.9654,
+      "step": 29106
+    },
+    {
+      "epoch": 0.6928258592783014,
+      "grad_norm": 7.437661856821426,
+      "learning_rate": 2e-05,
+      "loss": 0.3363,
+      "step": 29107
+    },
+    {
+      "epoch": 0.6928496620013329,
+      "grad_norm": 10.022968874648221,
+      "learning_rate": 2e-05,
+      "loss": 0.3937,
+      "step": 29108
+    },
+    {
+      "epoch": 0.6928734647243645,
+      "grad_norm": 2.4018899756487855,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 29109
+    },
+    {
+      "epoch": 0.692897267447396,
+      "grad_norm": 0.008779955611779352,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 29110
+    },
+    {
+      "epoch": 0.6929210701704275,
+      "grad_norm": 0.44913702517394644,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 29111
+    },
+    {
+      "epoch": 0.692944872893459,
+      "grad_norm": 2.0804567957545017,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 29112
+    },
+    {
+      "epoch": 0.6929686756164906,
+      "grad_norm": 4.445648881026442,
+      "learning_rate": 2e-05,
+      "loss": 0.2667,
+      "step": 29113
+    },
+    {
+      "epoch": 0.692992478339522,
+      "grad_norm": 5.751012989973223,
+      "learning_rate": 2e-05,
+      "loss": 1.6998,
+      "step": 29114
+    },
+    {
+      "epoch": 0.6930162810625535,
+      "grad_norm": 2.9311860155042915,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 29115
+    },
+    {
+      "epoch": 0.6930400837855851,
+      "grad_norm": 2.09961844551008,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 29116
+    },
+    {
+      "epoch": 0.6930638865086166,
+      "grad_norm": 1.802848873250048,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 29117
+    },
+    {
+      "epoch": 0.6930876892316481,
+      "grad_norm": 7.353803688793522,
+      "learning_rate": 2e-05,
+      "loss": 0.8763,
+      "step": 29118
+    },
+    {
+      "epoch": 0.6931114919546796,
+      "grad_norm": 10.813445563051813,
+      "learning_rate": 2e-05,
+      "loss": 0.996,
+      "step": 29119
+    },
+    {
+      "epoch": 0.6931352946777112,
+      "grad_norm": 0.08138822901557488,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 29120
+    },
+    {
+      "epoch": 0.6931590974007427,
+      "grad_norm": 1.6274748096595062,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 29121
+    },
+    {
+      "epoch": 0.6931829001237741,
+      "grad_norm": 4.197266476141355,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 29122
+    },
+    {
+      "epoch": 0.6932067028468056,
+      "grad_norm": 3.0980680116235253,
+      "learning_rate": 2e-05,
+      "loss": 0.49,
+      "step": 29123
+    },
+    {
+      "epoch": 0.6932305055698372,
+      "grad_norm": 2.168311421506168,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 29124
+    },
+    {
+      "epoch": 0.6932543082928687,
+      "grad_norm": 6.829267884188786,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 29125
+    },
+    {
+      "epoch": 0.6932781110159002,
+      "grad_norm": 4.889880966519807,
+      "learning_rate": 2e-05,
+      "loss": 0.5099,
+      "step": 29126
+    },
+    {
+      "epoch": 0.6933019137389317,
+      "grad_norm": 5.435350749214269,
+      "learning_rate": 2e-05,
+      "loss": 0.223,
+      "step": 29127
+    },
+    {
+      "epoch": 0.6933257164619633,
+      "grad_norm": 6.235962282827062,
+      "learning_rate": 2e-05,
+      "loss": 0.245,
+      "step": 29128
+    },
+    {
+      "epoch": 0.6933495191849948,
+      "grad_norm": 4.444175227620633,
+      "learning_rate": 2e-05,
+      "loss": 0.4166,
+      "step": 29129
+    },
+    {
+      "epoch": 0.6933733219080263,
+      "grad_norm": 5.096295597368838,
+      "learning_rate": 2e-05,
+      "loss": 0.529,
+      "step": 29130
+    },
+    {
+      "epoch": 0.6933971246310578,
+      "grad_norm": 1.302912639375286,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 29131
+    },
+    {
+      "epoch": 0.6934209273540893,
+      "grad_norm": 0.012263417318231646,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 29132
+    },
+    {
+      "epoch": 0.6934447300771208,
+      "grad_norm": 5.134572573845579,
+      "learning_rate": 2e-05,
+      "loss": 0.8201,
+      "step": 29133
+    },
+    {
+      "epoch": 0.6934685328001523,
+      "grad_norm": 4.794263608078876,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 29134
+    },
+    {
+      "epoch": 0.6934923355231839,
+      "grad_norm": 4.279645162747562,
+      "learning_rate": 2e-05,
+      "loss": 0.482,
+      "step": 29135
+    },
+    {
+      "epoch": 0.6935161382462154,
+      "grad_norm": 2.333238381202568,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 29136
+    },
+    {
+      "epoch": 0.6935399409692469,
+      "grad_norm": 3.30798663359407,
+      "learning_rate": 2e-05,
+      "loss": 0.3049,
+      "step": 29137
+    },
+    {
+      "epoch": 0.6935637436922784,
+      "grad_norm": 3.510595672784652,
+      "learning_rate": 2e-05,
+      "loss": 0.6172,
+      "step": 29138
+    },
+    {
+      "epoch": 0.69358754641531,
+      "grad_norm": 0.6311823318885282,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 29139
+    },
+    {
+      "epoch": 0.6936113491383414,
+      "grad_norm": 13.967848140661346,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 29140
+    },
+    {
+      "epoch": 0.6936351518613729,
+      "grad_norm": 7.64300418964506,
+      "learning_rate": 2e-05,
+      "loss": 0.3821,
+      "step": 29141
+    },
+    {
+      "epoch": 0.6936589545844044,
+      "grad_norm": 3.533026750884431,
+      "learning_rate": 2e-05,
+      "loss": 0.3661,
+      "step": 29142
+    },
+    {
+      "epoch": 0.693682757307436,
+      "grad_norm": 8.25639424767276,
+      "learning_rate": 2e-05,
+      "loss": 1.4851,
+      "step": 29143
+    },
+    {
+      "epoch": 0.6937065600304675,
+      "grad_norm": 0.39732778566179905,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 29144
+    },
+    {
+      "epoch": 0.693730362753499,
+      "grad_norm": 5.036392147523831,
+      "learning_rate": 2e-05,
+      "loss": 0.3079,
+      "step": 29145
+    },
+    {
+      "epoch": 0.6937541654765305,
+      "grad_norm": 4.63595063282788,
+      "learning_rate": 2e-05,
+      "loss": 0.4762,
+      "step": 29146
+    },
+    {
+      "epoch": 0.6937779681995621,
+      "grad_norm": 6.225070198816371,
+      "learning_rate": 2e-05,
+      "loss": 1.3593,
+      "step": 29147
+    },
+    {
+      "epoch": 0.6938017709225935,
+      "grad_norm": 4.33876003624885,
+      "learning_rate": 2e-05,
+      "loss": 0.9078,
+      "step": 29148
+    },
+    {
+      "epoch": 0.693825573645625,
+      "grad_norm": 6.0452210449306145,
+      "learning_rate": 2e-05,
+      "loss": 0.6474,
+      "step": 29149
+    },
+    {
+      "epoch": 0.6938493763686566,
+      "grad_norm": 8.347651126251884,
+      "learning_rate": 2e-05,
+      "loss": 0.4377,
+      "step": 29150
+    },
+    {
+      "epoch": 0.6938731790916881,
+      "grad_norm": 6.093931895977069,
+      "learning_rate": 2e-05,
+      "loss": 0.8248,
+      "step": 29151
+    },
+    {
+      "epoch": 0.6938969818147196,
+      "grad_norm": 3.2595479202053474,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 29152
+    },
+    {
+      "epoch": 0.6939207845377511,
+      "grad_norm": 8.007123076169197,
+      "learning_rate": 2e-05,
+      "loss": 1.7588,
+      "step": 29153
+    },
+    {
+      "epoch": 0.6939445872607827,
+      "grad_norm": 1.4261109513667838,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 29154
+    },
+    {
+      "epoch": 0.6939683899838142,
+      "grad_norm": 4.211218273471279,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 29155
+    },
+    {
+      "epoch": 0.6939921927068456,
+      "grad_norm": 3.3638187472426107,
+      "learning_rate": 2e-05,
+      "loss": 0.2688,
+      "step": 29156
+    },
+    {
+      "epoch": 0.6940159954298771,
+      "grad_norm": 0.7940838462772856,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 29157
+    },
+    {
+      "epoch": 0.6940397981529087,
+      "grad_norm": 4.093832538251312,
+      "learning_rate": 2e-05,
+      "loss": 0.8431,
+      "step": 29158
+    },
+    {
+      "epoch": 0.6940636008759402,
+      "grad_norm": 0.9260725333577292,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 29159
+    },
+    {
+      "epoch": 0.6940874035989717,
+      "grad_norm": 3.0829138650231673,
+      "learning_rate": 2e-05,
+      "loss": 0.3174,
+      "step": 29160
+    },
+    {
+      "epoch": 0.6941112063220032,
+      "grad_norm": 0.9907724697760897,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 29161
+    },
+    {
+      "epoch": 0.6941350090450348,
+      "grad_norm": 5.340718614029929,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 29162
+    },
+    {
+      "epoch": 0.6941588117680663,
+      "grad_norm": 7.70543120997498,
+      "learning_rate": 2e-05,
+      "loss": 0.2779,
+      "step": 29163
+    },
+    {
+      "epoch": 0.6941826144910977,
+      "grad_norm": 4.22537054437296,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 29164
+    },
+    {
+      "epoch": 0.6942064172141293,
+      "grad_norm": 2.2156101748420296,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 29165
+    },
+    {
+      "epoch": 0.6942302199371608,
+      "grad_norm": 2.8941077026423265,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 29166
+    },
+    {
+      "epoch": 0.6942540226601923,
+      "grad_norm": 0.8934402017598267,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 29167
+    },
+    {
+      "epoch": 0.6942778253832238,
+      "grad_norm": 1.8145792148403053,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 29168
+    },
+    {
+      "epoch": 0.6943016281062554,
+      "grad_norm": 2.880953216057782,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 29169
+    },
+    {
+      "epoch": 0.6943254308292869,
+      "grad_norm": 0.6865106845173382,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 29170
+    },
+    {
+      "epoch": 0.6943492335523184,
+      "grad_norm": 1.5044143631241338,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 29171
+    },
+    {
+      "epoch": 0.6943730362753499,
+      "grad_norm": 0.07110821087368335,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 29172
+    },
+    {
+      "epoch": 0.6943968389983814,
+      "grad_norm": 0.24783165623589,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 29173
+    },
+    {
+      "epoch": 0.6944206417214129,
+      "grad_norm": 0.5733049232784626,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 29174
+    },
+    {
+      "epoch": 0.6944444444444444,
+      "grad_norm": 5.769387784310668,
+      "learning_rate": 2e-05,
+      "loss": 0.9084,
+      "step": 29175
+    },
+    {
+      "epoch": 0.6944682471674759,
+      "grad_norm": 6.76985672644432,
+      "learning_rate": 2e-05,
+      "loss": 1.5077,
+      "step": 29176
+    },
+    {
+      "epoch": 0.6944920498905075,
+      "grad_norm": 5.613266466129359,
+      "learning_rate": 2e-05,
+      "loss": 0.9699,
+      "step": 29177
+    },
+    {
+      "epoch": 0.694515852613539,
+      "grad_norm": 9.22166485796634,
+      "learning_rate": 2e-05,
+      "loss": 2.0877,
+      "step": 29178
+    },
+    {
+      "epoch": 0.6945396553365705,
+      "grad_norm": 9.330235192941728,
+      "learning_rate": 2e-05,
+      "loss": 0.4655,
+      "step": 29179
+    },
+    {
+      "epoch": 0.6945634580596021,
+      "grad_norm": 0.6563355744881468,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 29180
+    },
+    {
+      "epoch": 0.6945872607826336,
+      "grad_norm": 3.4559556182972004,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 29181
+    },
+    {
+      "epoch": 0.694611063505665,
+      "grad_norm": 0.8359273678701943,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 29182
+    },
+    {
+      "epoch": 0.6946348662286965,
+      "grad_norm": 0.034525922179632594,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29183
+    },
+    {
+      "epoch": 0.6946586689517281,
+      "grad_norm": 1.7587108822047108,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 29184
+    },
+    {
+      "epoch": 0.6946824716747596,
+      "grad_norm": 5.06896233474558,
+      "learning_rate": 2e-05,
+      "loss": 0.442,
+      "step": 29185
+    },
+    {
+      "epoch": 0.6947062743977911,
+      "grad_norm": 11.35651250958836,
+      "learning_rate": 2e-05,
+      "loss": 0.6297,
+      "step": 29186
+    },
+    {
+      "epoch": 0.6947300771208226,
+      "grad_norm": 0.9434543477276995,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 29187
+    },
+    {
+      "epoch": 0.6947538798438542,
+      "grad_norm": 1.8778796023519602,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 29188
+    },
+    {
+      "epoch": 0.6947776825668857,
+      "grad_norm": 2.0706691408189646,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 29189
+    },
+    {
+      "epoch": 0.6948014852899171,
+      "grad_norm": 7.391391730890643,
+      "learning_rate": 2e-05,
+      "loss": 1.1846,
+      "step": 29190
+    },
+    {
+      "epoch": 0.6948252880129486,
+      "grad_norm": 0.12499483450101431,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 29191
+    },
+    {
+      "epoch": 0.6948490907359802,
+      "grad_norm": 7.571693061497146,
+      "learning_rate": 2e-05,
+      "loss": 0.7769,
+      "step": 29192
+    },
+    {
+      "epoch": 0.6948728934590117,
+      "grad_norm": 3.6829955618447796,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 29193
+    },
+    {
+      "epoch": 0.6948966961820432,
+      "grad_norm": 7.35267437522719,
+      "learning_rate": 2e-05,
+      "loss": 1.2399,
+      "step": 29194
+    },
+    {
+      "epoch": 0.6949204989050748,
+      "grad_norm": 6.373092674218273,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 29195
+    },
+    {
+      "epoch": 0.6949443016281063,
+      "grad_norm": 4.8447596564860715,
+      "learning_rate": 2e-05,
+      "loss": 0.8126,
+      "step": 29196
+    },
+    {
+      "epoch": 0.6949681043511378,
+      "grad_norm": 0.46437132087191235,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 29197
+    },
+    {
+      "epoch": 0.6949919070741692,
+      "grad_norm": 0.6717228388705032,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 29198
+    },
+    {
+      "epoch": 0.6950157097972008,
+      "grad_norm": 5.353682235220031,
+      "learning_rate": 2e-05,
+      "loss": 1.1016,
+      "step": 29199
+    },
+    {
+      "epoch": 0.6950395125202323,
+      "grad_norm": 3.8458009809732636,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 29200
+    },
+    {
+      "epoch": 0.6950633152432638,
+      "grad_norm": 9.868965637604274,
+      "learning_rate": 2e-05,
+      "loss": 0.5155,
+      "step": 29201
+    },
+    {
+      "epoch": 0.6950871179662953,
+      "grad_norm": 0.4291930562970443,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 29202
+    },
+    {
+      "epoch": 0.6951109206893269,
+      "grad_norm": 18.592086658382197,
+      "learning_rate": 2e-05,
+      "loss": 0.3126,
+      "step": 29203
+    },
+    {
+      "epoch": 0.6951347234123584,
+      "grad_norm": 5.688988574309753,
+      "learning_rate": 2e-05,
+      "loss": 0.2008,
+      "step": 29204
+    },
+    {
+      "epoch": 0.6951585261353899,
+      "grad_norm": 0.05806187600365201,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 29205
+    },
+    {
+      "epoch": 0.6951823288584214,
+      "grad_norm": 2.8629208440352283,
+      "learning_rate": 2e-05,
+      "loss": 0.4158,
+      "step": 29206
+    },
+    {
+      "epoch": 0.695206131581453,
+      "grad_norm": 4.6355890033776195,
+      "learning_rate": 2e-05,
+      "loss": 0.6038,
+      "step": 29207
+    },
+    {
+      "epoch": 0.6952299343044844,
+      "grad_norm": 1.5211316745042334,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 29208
+    },
+    {
+      "epoch": 0.6952537370275159,
+      "grad_norm": 2.695243093355608,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 29209
+    },
+    {
+      "epoch": 0.6952775397505475,
+      "grad_norm": 2.6345070830716253,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 29210
+    },
+    {
+      "epoch": 0.695301342473579,
+      "grad_norm": 6.884392652016491,
+      "learning_rate": 2e-05,
+      "loss": 1.6278,
+      "step": 29211
+    },
+    {
+      "epoch": 0.6953251451966105,
+      "grad_norm": 2.2569372937696017,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 29212
+    },
+    {
+      "epoch": 0.695348947919642,
+      "grad_norm": 3.6881971430384324,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 29213
+    },
+    {
+      "epoch": 0.6953727506426736,
+      "grad_norm": 0.6046629071675452,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 29214
+    },
+    {
+      "epoch": 0.695396553365705,
+      "grad_norm": 1.9552519927603365,
+      "learning_rate": 2e-05,
+      "loss": 0.136,
+      "step": 29215
+    },
+    {
+      "epoch": 0.6954203560887365,
+      "grad_norm": 14.59166856877769,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 29216
+    },
+    {
+      "epoch": 0.695444158811768,
+      "grad_norm": 8.258864608179364,
+      "learning_rate": 2e-05,
+      "loss": 1.3497,
+      "step": 29217
+    },
+    {
+      "epoch": 0.6954679615347996,
+      "grad_norm": 4.582425339577352,
+      "learning_rate": 2e-05,
+      "loss": 0.8409,
+      "step": 29218
+    },
+    {
+      "epoch": 0.6954917642578311,
+      "grad_norm": 12.518295465599508,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 29219
+    },
+    {
+      "epoch": 0.6955155669808626,
+      "grad_norm": 0.12162265128950513,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 29220
+    },
+    {
+      "epoch": 0.6955393697038941,
+      "grad_norm": 4.228392203386204,
+      "learning_rate": 2e-05,
+      "loss": 0.0963,
+      "step": 29221
+    },
+    {
+      "epoch": 0.6955631724269257,
+      "grad_norm": 5.559171391557557,
+      "learning_rate": 2e-05,
+      "loss": 0.737,
+      "step": 29222
+    },
+    {
+      "epoch": 0.6955869751499572,
+      "grad_norm": 1.0052521736572082,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 29223
+    },
+    {
+      "epoch": 0.6956107778729886,
+      "grad_norm": 4.341389012093409,
+      "learning_rate": 2e-05,
+      "loss": 0.4514,
+      "step": 29224
+    },
+    {
+      "epoch": 0.6956345805960202,
+      "grad_norm": 0.5947024102168279,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 29225
+    },
+    {
+      "epoch": 0.6956583833190517,
+      "grad_norm": 5.176459156974563,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 29226
+    },
+    {
+      "epoch": 0.6956821860420832,
+      "grad_norm": 0.6361976109930808,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 29227
+    },
+    {
+      "epoch": 0.6957059887651147,
+      "grad_norm": 3.313478308417832,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 29228
+    },
+    {
+      "epoch": 0.6957297914881463,
+      "grad_norm": 3.689352307807462,
+      "learning_rate": 2e-05,
+      "loss": 0.6763,
+      "step": 29229
+    },
+    {
+      "epoch": 0.6957535942111778,
+      "grad_norm": 1.4592953862569527,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 29230
+    },
+    {
+      "epoch": 0.6957773969342093,
+      "grad_norm": 2.9735391262765227,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 29231
+    },
+    {
+      "epoch": 0.6958011996572407,
+      "grad_norm": 0.1374109321975586,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 29232
+    },
+    {
+      "epoch": 0.6958250023802723,
+      "grad_norm": 1.9321872249117042,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 29233
+    },
+    {
+      "epoch": 0.6958488051033038,
+      "grad_norm": 4.309303146398139,
+      "learning_rate": 2e-05,
+      "loss": 0.3898,
+      "step": 29234
+    },
+    {
+      "epoch": 0.6958726078263353,
+      "grad_norm": 7.358041410921168,
+      "learning_rate": 2e-05,
+      "loss": 0.6745,
+      "step": 29235
+    },
+    {
+      "epoch": 0.6958964105493668,
+      "grad_norm": 21.53543579957698,
+      "learning_rate": 2e-05,
+      "loss": 0.5596,
+      "step": 29236
+    },
+    {
+      "epoch": 0.6959202132723984,
+      "grad_norm": 2.143126536342295,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 29237
+    },
+    {
+      "epoch": 0.6959440159954299,
+      "grad_norm": 16.56214951129862,
+      "learning_rate": 2e-05,
+      "loss": 0.6364,
+      "step": 29238
+    },
+    {
+      "epoch": 0.6959678187184614,
+      "grad_norm": 4.954640972593096,
+      "learning_rate": 2e-05,
+      "loss": 1.361,
+      "step": 29239
+    },
+    {
+      "epoch": 0.695991621441493,
+      "grad_norm": 1.4663709922838586,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 29240
+    },
+    {
+      "epoch": 0.6960154241645244,
+      "grad_norm": 2.0225270331703413,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 29241
+    },
+    {
+      "epoch": 0.6960392268875559,
+      "grad_norm": 3.2911154268035014,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 29242
+    },
+    {
+      "epoch": 0.6960630296105874,
+      "grad_norm": 3.597904057328731,
+      "learning_rate": 2e-05,
+      "loss": 0.166,
+      "step": 29243
+    },
+    {
+      "epoch": 0.696086832333619,
+      "grad_norm": 3.2332203712851197,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 29244
+    },
+    {
+      "epoch": 0.6961106350566505,
+      "grad_norm": 4.0386887545790096,
+      "learning_rate": 2e-05,
+      "loss": 0.3747,
+      "step": 29245
+    },
+    {
+      "epoch": 0.696134437779682,
+      "grad_norm": 7.298315157579536,
+      "learning_rate": 2e-05,
+      "loss": 0.9478,
+      "step": 29246
+    },
+    {
+      "epoch": 0.6961582405027135,
+      "grad_norm": 2.9852499517203843,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 29247
+    },
+    {
+      "epoch": 0.6961820432257451,
+      "grad_norm": 3.090359716137539,
+      "learning_rate": 2e-05,
+      "loss": 0.3459,
+      "step": 29248
+    },
+    {
+      "epoch": 0.6962058459487765,
+      "grad_norm": 2.7933199100782895,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 29249
+    },
+    {
+      "epoch": 0.696229648671808,
+      "grad_norm": 4.421701908763086,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 29250
+    },
+    {
+      "epoch": 0.6962534513948395,
+      "grad_norm": 1.5196841154064609,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 29251
+    },
+    {
+      "epoch": 0.6962772541178711,
+      "grad_norm": 4.372284681556812,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 29252
+    },
+    {
+      "epoch": 0.6963010568409026,
+      "grad_norm": 2.410426595913227,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 29253
+    },
+    {
+      "epoch": 0.6963248595639341,
+      "grad_norm": 16.154270696210894,
+      "learning_rate": 2e-05,
+      "loss": 0.5654,
+      "step": 29254
+    },
+    {
+      "epoch": 0.6963486622869657,
+      "grad_norm": 2.3100107409658417,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 29255
+    },
+    {
+      "epoch": 0.6963724650099972,
+      "grad_norm": 10.604679916871397,
+      "learning_rate": 2e-05,
+      "loss": 1.3159,
+      "step": 29256
+    },
+    {
+      "epoch": 0.6963962677330287,
+      "grad_norm": 9.18897076734922,
+      "learning_rate": 2e-05,
+      "loss": 0.9194,
+      "step": 29257
+    },
+    {
+      "epoch": 0.6964200704560601,
+      "grad_norm": 3.032517124582002,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 29258
+    },
+    {
+      "epoch": 0.6964438731790917,
+      "grad_norm": 3.821713982852082,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 29259
+    },
+    {
+      "epoch": 0.6964676759021232,
+      "grad_norm": 3.4542398787854594,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 29260
+    },
+    {
+      "epoch": 0.6964914786251547,
+      "grad_norm": 3.528038484515695,
+      "learning_rate": 2e-05,
+      "loss": 0.6438,
+      "step": 29261
+    },
+    {
+      "epoch": 0.6965152813481862,
+      "grad_norm": 4.833368739981902,
+      "learning_rate": 2e-05,
+      "loss": 0.5541,
+      "step": 29262
+    },
+    {
+      "epoch": 0.6965390840712178,
+      "grad_norm": 4.172945694713224,
+      "learning_rate": 2e-05,
+      "loss": 0.6028,
+      "step": 29263
+    },
+    {
+      "epoch": 0.6965628867942493,
+      "grad_norm": 0.17265361028181575,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 29264
+    },
+    {
+      "epoch": 0.6965866895172808,
+      "grad_norm": 3.4527638826735925,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 29265
+    },
+    {
+      "epoch": 0.6966104922403122,
+      "grad_norm": 10.12816505331537,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 29266
+    },
+    {
+      "epoch": 0.6966342949633438,
+      "grad_norm": 1.2793726029990633,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 29267
+    },
+    {
+      "epoch": 0.6966580976863753,
+      "grad_norm": 6.315766293274547,
+      "learning_rate": 2e-05,
+      "loss": 1.1285,
+      "step": 29268
+    },
+    {
+      "epoch": 0.6966819004094068,
+      "grad_norm": 199.23219890356143,
+      "learning_rate": 2e-05,
+      "loss": 1.2137,
+      "step": 29269
+    },
+    {
+      "epoch": 0.6967057031324384,
+      "grad_norm": 0.3955019318781582,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 29270
+    },
+    {
+      "epoch": 0.6967295058554699,
+      "grad_norm": 2.150011644727644,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 29271
+    },
+    {
+      "epoch": 0.6967533085785014,
+      "grad_norm": 0.46799093154250865,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 29272
+    },
+    {
+      "epoch": 0.6967771113015329,
+      "grad_norm": 2.4294347710221422,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 29273
+    },
+    {
+      "epoch": 0.6968009140245645,
+      "grad_norm": 3.7845978367568893,
+      "learning_rate": 2e-05,
+      "loss": 0.1139,
+      "step": 29274
+    },
+    {
+      "epoch": 0.6968247167475959,
+      "grad_norm": 1.8615825481953907,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 29275
+    },
+    {
+      "epoch": 0.6968485194706274,
+      "grad_norm": 5.173990143240546,
+      "learning_rate": 2e-05,
+      "loss": 1.2666,
+      "step": 29276
+    },
+    {
+      "epoch": 0.6968723221936589,
+      "grad_norm": 30.21982554939099,
+      "learning_rate": 2e-05,
+      "loss": 0.205,
+      "step": 29277
+    },
+    {
+      "epoch": 0.6968961249166905,
+      "grad_norm": 0.17473116764495855,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 29278
+    },
+    {
+      "epoch": 0.696919927639722,
+      "grad_norm": 1.5259289517714294,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 29279
+    },
+    {
+      "epoch": 0.6969437303627535,
+      "grad_norm": 5.127182850750702,
+      "learning_rate": 2e-05,
+      "loss": 0.9265,
+      "step": 29280
+    },
+    {
+      "epoch": 0.696967533085785,
+      "grad_norm": 5.9767511031836005,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 29281
+    },
+    {
+      "epoch": 0.6969913358088166,
+      "grad_norm": 11.349011663781585,
+      "learning_rate": 2e-05,
+      "loss": 1.1737,
+      "step": 29282
+    },
+    {
+      "epoch": 0.697015138531848,
+      "grad_norm": 1.9137682815392596,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 29283
+    },
+    {
+      "epoch": 0.6970389412548795,
+      "grad_norm": 4.054450911853537,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 29284
+    },
+    {
+      "epoch": 0.6970627439779111,
+      "grad_norm": 0.143287301419759,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 29285
+    },
+    {
+      "epoch": 0.6970865467009426,
+      "grad_norm": 2.4210286335156415,
+      "learning_rate": 2e-05,
+      "loss": 0.1896,
+      "step": 29286
+    },
+    {
+      "epoch": 0.6971103494239741,
+      "grad_norm": 9.872028354512157,
+      "learning_rate": 2e-05,
+      "loss": 0.7235,
+      "step": 29287
+    },
+    {
+      "epoch": 0.6971341521470056,
+      "grad_norm": 1.9381834901318047,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 29288
+    },
+    {
+      "epoch": 0.6971579548700372,
+      "grad_norm": 0.8558816678658552,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 29289
+    },
+    {
+      "epoch": 0.6971817575930687,
+      "grad_norm": 0.9583472975629329,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 29290
+    },
+    {
+      "epoch": 0.6972055603161001,
+      "grad_norm": 0.20316293956949816,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 29291
+    },
+    {
+      "epoch": 0.6972293630391316,
+      "grad_norm": 1.43349971023068,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 29292
+    },
+    {
+      "epoch": 0.6972531657621632,
+      "grad_norm": 0.09302266895153222,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 29293
+    },
+    {
+      "epoch": 0.6972769684851947,
+      "grad_norm": 2.113291400291717,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 29294
+    },
+    {
+      "epoch": 0.6973007712082262,
+      "grad_norm": 0.26917361005533086,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 29295
+    },
+    {
+      "epoch": 0.6973245739312577,
+      "grad_norm": 0.17886754386037879,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 29296
+    },
+    {
+      "epoch": 0.6973483766542893,
+      "grad_norm": 3.8243530976226485,
+      "learning_rate": 2e-05,
+      "loss": 0.2483,
+      "step": 29297
+    },
+    {
+      "epoch": 0.6973721793773208,
+      "grad_norm": 0.8879361112908305,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 29298
+    },
+    {
+      "epoch": 0.6973959821003523,
+      "grad_norm": 5.384087292525514,
+      "learning_rate": 2e-05,
+      "loss": 0.7438,
+      "step": 29299
+    },
+    {
+      "epoch": 0.6974197848233838,
+      "grad_norm": 6.950561869314332,
+      "learning_rate": 2e-05,
+      "loss": 0.1151,
+      "step": 29300
+    },
+    {
+      "epoch": 0.6974435875464153,
+      "grad_norm": 7.366132586034362,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 29301
+    },
+    {
+      "epoch": 0.6974673902694468,
+      "grad_norm": 0.6364070335604666,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 29302
+    },
+    {
+      "epoch": 0.6974911929924783,
+      "grad_norm": 3.4348874868628196,
+      "learning_rate": 2e-05,
+      "loss": 0.2881,
+      "step": 29303
+    },
+    {
+      "epoch": 0.6975149957155099,
+      "grad_norm": 0.13967783223744853,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 29304
+    },
+    {
+      "epoch": 0.6975387984385414,
+      "grad_norm": 5.154156705429057,
+      "learning_rate": 2e-05,
+      "loss": 0.4412,
+      "step": 29305
+    },
+    {
+      "epoch": 0.6975626011615729,
+      "grad_norm": 7.295486971162945,
+      "learning_rate": 2e-05,
+      "loss": 0.1389,
+      "step": 29306
+    },
+    {
+      "epoch": 0.6975864038846044,
+      "grad_norm": 3.1788496082493194,
+      "learning_rate": 2e-05,
+      "loss": 0.2847,
+      "step": 29307
+    },
+    {
+      "epoch": 0.697610206607636,
+      "grad_norm": 0.3082687375601678,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 29308
+    },
+    {
+      "epoch": 0.6976340093306674,
+      "grad_norm": 5.402993138441536,
+      "learning_rate": 2e-05,
+      "loss": 1.2257,
+      "step": 29309
+    },
+    {
+      "epoch": 0.6976578120536989,
+      "grad_norm": 4.307538676713368,
+      "learning_rate": 2e-05,
+      "loss": 0.4697,
+      "step": 29310
+    },
+    {
+      "epoch": 0.6976816147767304,
+      "grad_norm": 8.58720225297833,
+      "learning_rate": 2e-05,
+      "loss": 0.2179,
+      "step": 29311
+    },
+    {
+      "epoch": 0.697705417499762,
+      "grad_norm": 0.13827106845259488,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29312
+    },
+    {
+      "epoch": 0.6977292202227935,
+      "grad_norm": 9.285413132590527,
+      "learning_rate": 2e-05,
+      "loss": 2.1847,
+      "step": 29313
+    },
+    {
+      "epoch": 0.697753022945825,
+      "grad_norm": 7.952284297829233,
+      "learning_rate": 2e-05,
+      "loss": 0.2847,
+      "step": 29314
+    },
+    {
+      "epoch": 0.6977768256688566,
+      "grad_norm": 0.787152884188389,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 29315
+    },
+    {
+      "epoch": 0.6978006283918881,
+      "grad_norm": 1.9739710300238438,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 29316
+    },
+    {
+      "epoch": 0.6978244311149195,
+      "grad_norm": 2.9708626225483727,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 29317
+    },
+    {
+      "epoch": 0.697848233837951,
+      "grad_norm": 8.412760003106166,
+      "learning_rate": 2e-05,
+      "loss": 0.5865,
+      "step": 29318
+    },
+    {
+      "epoch": 0.6978720365609826,
+      "grad_norm": 7.120183496492815,
+      "learning_rate": 2e-05,
+      "loss": 0.3384,
+      "step": 29319
+    },
+    {
+      "epoch": 0.6978958392840141,
+      "grad_norm": 2.552184625858604,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 29320
+    },
+    {
+      "epoch": 0.6979196420070456,
+      "grad_norm": 1.924610465612682,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 29321
+    },
+    {
+      "epoch": 0.6979434447300771,
+      "grad_norm": 5.580365667518572,
+      "learning_rate": 2e-05,
+      "loss": 1.2579,
+      "step": 29322
+    },
+    {
+      "epoch": 0.6979672474531087,
+      "grad_norm": 13.213842321833113,
+      "learning_rate": 2e-05,
+      "loss": 0.4708,
+      "step": 29323
+    },
+    {
+      "epoch": 0.6979910501761402,
+      "grad_norm": 7.213858149771903,
+      "learning_rate": 2e-05,
+      "loss": 1.9957,
+      "step": 29324
+    },
+    {
+      "epoch": 0.6980148528991716,
+      "grad_norm": 6.599997860055408,
+      "learning_rate": 2e-05,
+      "loss": 0.338,
+      "step": 29325
+    },
+    {
+      "epoch": 0.6980386556222031,
+      "grad_norm": 3.2305832248340307,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 29326
+    },
+    {
+      "epoch": 0.6980624583452347,
+      "grad_norm": 4.20514902742019,
+      "learning_rate": 2e-05,
+      "loss": 0.0885,
+      "step": 29327
+    },
+    {
+      "epoch": 0.6980862610682662,
+      "grad_norm": 9.905004255855497,
+      "learning_rate": 2e-05,
+      "loss": 1.0964,
+      "step": 29328
+    },
+    {
+      "epoch": 0.6981100637912977,
+      "grad_norm": 0.4649483711305876,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 29329
+    },
+    {
+      "epoch": 0.6981338665143293,
+      "grad_norm": 18.32210832248191,
+      "learning_rate": 2e-05,
+      "loss": 0.216,
+      "step": 29330
+    },
+    {
+      "epoch": 0.6981576692373608,
+      "grad_norm": 2.2422726381851725,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 29331
+    },
+    {
+      "epoch": 0.6981814719603923,
+      "grad_norm": 5.357089184166631,
+      "learning_rate": 2e-05,
+      "loss": 0.6425,
+      "step": 29332
+    },
+    {
+      "epoch": 0.6982052746834237,
+      "grad_norm": 0.3036514271924056,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 29333
+    },
+    {
+      "epoch": 0.6982290774064553,
+      "grad_norm": 2.8448567704423695,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 29334
+    },
+    {
+      "epoch": 0.6982528801294868,
+      "grad_norm": 5.1731619500267305,
+      "learning_rate": 2e-05,
+      "loss": 0.7696,
+      "step": 29335
+    },
+    {
+      "epoch": 0.6982766828525183,
+      "grad_norm": 2.2402029806118056,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 29336
+    },
+    {
+      "epoch": 0.6983004855755498,
+      "grad_norm": 0.7273323872032285,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 29337
+    },
+    {
+      "epoch": 0.6983242882985814,
+      "grad_norm": 5.9540769624736,
+      "learning_rate": 2e-05,
+      "loss": 0.114,
+      "step": 29338
+    },
+    {
+      "epoch": 0.6983480910216129,
+      "grad_norm": 5.327029831060804,
+      "learning_rate": 2e-05,
+      "loss": 0.2435,
+      "step": 29339
+    },
+    {
+      "epoch": 0.6983718937446444,
+      "grad_norm": 4.496065856504294,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 29340
+    },
+    {
+      "epoch": 0.6983956964676759,
+      "grad_norm": 3.540795789276213,
+      "learning_rate": 2e-05,
+      "loss": 0.4532,
+      "step": 29341
+    },
+    {
+      "epoch": 0.6984194991907074,
+      "grad_norm": 3.034924096618786,
+      "learning_rate": 2e-05,
+      "loss": 0.539,
+      "step": 29342
+    },
+    {
+      "epoch": 0.6984433019137389,
+      "grad_norm": 26.73850313265425,
+      "learning_rate": 2e-05,
+      "loss": 0.3355,
+      "step": 29343
+    },
+    {
+      "epoch": 0.6984671046367704,
+      "grad_norm": 0.2636866343132172,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 29344
+    },
+    {
+      "epoch": 0.698490907359802,
+      "grad_norm": 0.42407145964130916,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 29345
+    },
+    {
+      "epoch": 0.6985147100828335,
+      "grad_norm": 4.13289017828408,
+      "learning_rate": 2e-05,
+      "loss": 0.5411,
+      "step": 29346
+    },
+    {
+      "epoch": 0.698538512805865,
+      "grad_norm": 1.9358540217680493,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 29347
+    },
+    {
+      "epoch": 0.6985623155288965,
+      "grad_norm": 4.893819352038387,
+      "learning_rate": 2e-05,
+      "loss": 1.0027,
+      "step": 29348
+    },
+    {
+      "epoch": 0.6985861182519281,
+      "grad_norm": 0.332395581926696,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 29349
+    },
+    {
+      "epoch": 0.6986099209749596,
+      "grad_norm": 3.1732938321387,
+      "learning_rate": 2e-05,
+      "loss": 0.5074,
+      "step": 29350
+    },
+    {
+      "epoch": 0.698633723697991,
+      "grad_norm": 0.5244668243451197,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 29351
+    },
+    {
+      "epoch": 0.6986575264210225,
+      "grad_norm": 1.6259662887265502,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 29352
+    },
+    {
+      "epoch": 0.6986813291440541,
+      "grad_norm": 35.98926531756089,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 29353
+    },
+    {
+      "epoch": 0.6987051318670856,
+      "grad_norm": 3.532409558416068,
+      "learning_rate": 2e-05,
+      "loss": 0.5181,
+      "step": 29354
+    },
+    {
+      "epoch": 0.6987289345901171,
+      "grad_norm": 2.053125622356903,
+      "learning_rate": 2e-05,
+      "loss": 0.1987,
+      "step": 29355
+    },
+    {
+      "epoch": 0.6987527373131486,
+      "grad_norm": 11.728421113264426,
+      "learning_rate": 2e-05,
+      "loss": 0.7677,
+      "step": 29356
+    },
+    {
+      "epoch": 0.6987765400361802,
+      "grad_norm": 2.6658359356076704,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 29357
+    },
+    {
+      "epoch": 0.6988003427592117,
+      "grad_norm": 4.871303536744428,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 29358
+    },
+    {
+      "epoch": 0.6988241454822431,
+      "grad_norm": 3.9219660699192977,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 29359
+    },
+    {
+      "epoch": 0.6988479482052747,
+      "grad_norm": 11.063697516623465,
+      "learning_rate": 2e-05,
+      "loss": 1.0331,
+      "step": 29360
+    },
+    {
+      "epoch": 0.6988717509283062,
+      "grad_norm": 1.1990517642929384,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 29361
+    },
+    {
+      "epoch": 0.6988955536513377,
+      "grad_norm": 3.2470749740504994,
+      "learning_rate": 2e-05,
+      "loss": 0.1821,
+      "step": 29362
+    },
+    {
+      "epoch": 0.6989193563743692,
+      "grad_norm": 3.37285417426614,
+      "learning_rate": 2e-05,
+      "loss": 0.1128,
+      "step": 29363
+    },
+    {
+      "epoch": 0.6989431590974008,
+      "grad_norm": 2.8119177791836014,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 29364
+    },
+    {
+      "epoch": 0.6989669618204323,
+      "grad_norm": 1.2068889466163302,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 29365
+    },
+    {
+      "epoch": 0.6989907645434638,
+      "grad_norm": 0.2968965064097844,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 29366
+    },
+    {
+      "epoch": 0.6990145672664952,
+      "grad_norm": 0.3715171160905144,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 29367
+    },
+    {
+      "epoch": 0.6990383699895268,
+      "grad_norm": 2.3470343500487574,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 29368
+    },
+    {
+      "epoch": 0.6990621727125583,
+      "grad_norm": 2.42112913468789,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 29369
+    },
+    {
+      "epoch": 0.6990859754355898,
+      "grad_norm": 0.0702623608908892,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 29370
+    },
+    {
+      "epoch": 0.6991097781586213,
+      "grad_norm": 1.996371629206916,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 29371
+    },
+    {
+      "epoch": 0.6991335808816529,
+      "grad_norm": 11.655164816638598,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 29372
+    },
+    {
+      "epoch": 0.6991573836046844,
+      "grad_norm": 0.13881543987118292,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 29373
+    },
+    {
+      "epoch": 0.6991811863277159,
+      "grad_norm": 0.27389085697355414,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 29374
+    },
+    {
+      "epoch": 0.6992049890507475,
+      "grad_norm": 1.3402841482066956,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 29375
+    },
+    {
+      "epoch": 0.699228791773779,
+      "grad_norm": 3.3168633168921353,
+      "learning_rate": 2e-05,
+      "loss": 0.6721,
+      "step": 29376
+    },
+    {
+      "epoch": 0.6992525944968104,
+      "grad_norm": 6.8310934253088655,
+      "learning_rate": 2e-05,
+      "loss": 0.3798,
+      "step": 29377
+    },
+    {
+      "epoch": 0.6992763972198419,
+      "grad_norm": 10.034238709010607,
+      "learning_rate": 2e-05,
+      "loss": 0.4887,
+      "step": 29378
+    },
+    {
+      "epoch": 0.6993001999428735,
+      "grad_norm": 28.364275607725588,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 29379
+    },
+    {
+      "epoch": 0.699324002665905,
+      "grad_norm": 5.06916636216766,
+      "learning_rate": 2e-05,
+      "loss": 0.7936,
+      "step": 29380
+    },
+    {
+      "epoch": 0.6993478053889365,
+      "grad_norm": 2.6116679501729574,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 29381
+    },
+    {
+      "epoch": 0.699371608111968,
+      "grad_norm": 3.894309411078246,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 29382
+    },
+    {
+      "epoch": 0.6993954108349996,
+      "grad_norm": 4.149182910979996,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 29383
+    },
+    {
+      "epoch": 0.699419213558031,
+      "grad_norm": 3.5245235559264496,
+      "learning_rate": 2e-05,
+      "loss": 0.801,
+      "step": 29384
+    },
+    {
+      "epoch": 0.6994430162810625,
+      "grad_norm": 3.6096297060313347,
+      "learning_rate": 2e-05,
+      "loss": 0.4971,
+      "step": 29385
+    },
+    {
+      "epoch": 0.699466819004094,
+      "grad_norm": 0.5411489817825457,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 29386
+    },
+    {
+      "epoch": 0.6994906217271256,
+      "grad_norm": 12.554378688457314,
+      "learning_rate": 2e-05,
+      "loss": 0.3966,
+      "step": 29387
+    },
+    {
+      "epoch": 0.6995144244501571,
+      "grad_norm": 1.3555919718044298,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 29388
+    },
+    {
+      "epoch": 0.6995382271731886,
+      "grad_norm": 0.08981486643551173,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 29389
+    },
+    {
+      "epoch": 0.6995620298962201,
+      "grad_norm": 0.29483517460669945,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 29390
+    },
+    {
+      "epoch": 0.6995858326192517,
+      "grad_norm": 3.8967464591891017,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 29391
+    },
+    {
+      "epoch": 0.6996096353422832,
+      "grad_norm": 3.102350287515306,
+      "learning_rate": 2e-05,
+      "loss": 0.1943,
+      "step": 29392
+    },
+    {
+      "epoch": 0.6996334380653146,
+      "grad_norm": 22.57901209151979,
+      "learning_rate": 2e-05,
+      "loss": 0.3414,
+      "step": 29393
+    },
+    {
+      "epoch": 0.6996572407883462,
+      "grad_norm": 0.09565068872902906,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 29394
+    },
+    {
+      "epoch": 0.6996810435113777,
+      "grad_norm": 2.6230844216906726,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 29395
+    },
+    {
+      "epoch": 0.6997048462344092,
+      "grad_norm": 0.8160145258474065,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 29396
+    },
+    {
+      "epoch": 0.6997286489574407,
+      "grad_norm": 1.7776950544858467,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 29397
+    },
+    {
+      "epoch": 0.6997524516804723,
+      "grad_norm": 3.997308558360828,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 29398
+    },
+    {
+      "epoch": 0.6997762544035038,
+      "grad_norm": 3.1365330099857043,
+      "learning_rate": 2e-05,
+      "loss": 0.5766,
+      "step": 29399
+    },
+    {
+      "epoch": 0.6998000571265353,
+      "grad_norm": 4.704151800547646,
+      "learning_rate": 2e-05,
+      "loss": 0.942,
+      "step": 29400
+    },
+    {
+      "epoch": 0.6998238598495667,
+      "grad_norm": 3.65945570178085,
+      "learning_rate": 2e-05,
+      "loss": 0.1412,
+      "step": 29401
+    },
+    {
+      "epoch": 0.6998476625725983,
+      "grad_norm": 0.8841626686181753,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 29402
+    },
+    {
+      "epoch": 0.6998714652956298,
+      "grad_norm": 4.993779482631145,
+      "learning_rate": 2e-05,
+      "loss": 0.9368,
+      "step": 29403
+    },
+    {
+      "epoch": 0.6998952680186613,
+      "grad_norm": 3.347456283098267,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 29404
+    },
+    {
+      "epoch": 0.6999190707416928,
+      "grad_norm": 2.49235836967755,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 29405
+    },
+    {
+      "epoch": 0.6999428734647244,
+      "grad_norm": 1.4507294718133346,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 29406
+    },
+    {
+      "epoch": 0.6999666761877559,
+      "grad_norm": 4.5738716749928985,
+      "learning_rate": 2e-05,
+      "loss": 1.3712,
+      "step": 29407
+    },
+    {
+      "epoch": 0.6999904789107874,
+      "grad_norm": 1.5885104420048508,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 29408
+    },
+    {
+      "epoch": 0.700014281633819,
+      "grad_norm": 3.0093649192124463,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 29409
+    },
+    {
+      "epoch": 0.7000380843568504,
+      "grad_norm": 3.768448578502939,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 29410
+    },
+    {
+      "epoch": 0.7000618870798819,
+      "grad_norm": 0.6204499660278034,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 29411
+    },
+    {
+      "epoch": 0.7000856898029134,
+      "grad_norm": 6.827136195555349,
+      "learning_rate": 2e-05,
+      "loss": 0.2989,
+      "step": 29412
+    },
+    {
+      "epoch": 0.700109492525945,
+      "grad_norm": 0.079466499873544,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29413
+    },
+    {
+      "epoch": 0.7001332952489765,
+      "grad_norm": 2.9534054811190575,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 29414
+    },
+    {
+      "epoch": 0.700157097972008,
+      "grad_norm": 12.011510501640384,
+      "learning_rate": 2e-05,
+      "loss": 0.8561,
+      "step": 29415
+    },
+    {
+      "epoch": 0.7001809006950395,
+      "grad_norm": 6.0374825042722255,
+      "learning_rate": 2e-05,
+      "loss": 1.3154,
+      "step": 29416
+    },
+    {
+      "epoch": 0.7002047034180711,
+      "grad_norm": 4.9973890690523195,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 29417
+    },
+    {
+      "epoch": 0.7002285061411025,
+      "grad_norm": 0.9558043459717092,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 29418
+    },
+    {
+      "epoch": 0.700252308864134,
+      "grad_norm": 0.1744642499526753,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 29419
+    },
+    {
+      "epoch": 0.7002761115871655,
+      "grad_norm": 10.711000736488696,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 29420
+    },
+    {
+      "epoch": 0.7002999143101971,
+      "grad_norm": 14.860246995522731,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 29421
+    },
+    {
+      "epoch": 0.7003237170332286,
+      "grad_norm": 1.9278993771001423,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 29422
+    },
+    {
+      "epoch": 0.7003475197562601,
+      "grad_norm": 4.105264295125466,
+      "learning_rate": 2e-05,
+      "loss": 0.6678,
+      "step": 29423
+    },
+    {
+      "epoch": 0.7003713224792917,
+      "grad_norm": 0.6255906593040675,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 29424
+    },
+    {
+      "epoch": 0.7003951252023232,
+      "grad_norm": 3.5532376875870906,
+      "learning_rate": 2e-05,
+      "loss": 0.6101,
+      "step": 29425
+    },
+    {
+      "epoch": 0.7004189279253547,
+      "grad_norm": 3.7092240854003586,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 29426
+    },
+    {
+      "epoch": 0.7004427306483861,
+      "grad_norm": 0.8943310299242101,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 29427
+    },
+    {
+      "epoch": 0.7004665333714177,
+      "grad_norm": 0.8827733475761513,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 29428
+    },
+    {
+      "epoch": 0.7004903360944492,
+      "grad_norm": 0.46310050881998044,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 29429
+    },
+    {
+      "epoch": 0.7005141388174807,
+      "grad_norm": 1.1332169768154843,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 29430
+    },
+    {
+      "epoch": 0.7005379415405122,
+      "grad_norm": 3.7516707075499283,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 29431
+    },
+    {
+      "epoch": 0.7005617442635438,
+      "grad_norm": 3.7534932512339023,
+      "learning_rate": 2e-05,
+      "loss": 0.8828,
+      "step": 29432
+    },
+    {
+      "epoch": 0.7005855469865753,
+      "grad_norm": 5.559272628399871,
+      "learning_rate": 2e-05,
+      "loss": 0.9393,
+      "step": 29433
+    },
+    {
+      "epoch": 0.7006093497096068,
+      "grad_norm": 0.7455090049070255,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 29434
+    },
+    {
+      "epoch": 0.7006331524326382,
+      "grad_norm": 5.003082731474336,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 29435
+    },
+    {
+      "epoch": 0.7006569551556698,
+      "grad_norm": 6.947304237951922,
+      "learning_rate": 2e-05,
+      "loss": 1.8826,
+      "step": 29436
+    },
+    {
+      "epoch": 0.7006807578787013,
+      "grad_norm": 2.6723057734968796,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 29437
+    },
+    {
+      "epoch": 0.7007045606017328,
+      "grad_norm": 4.113825256582888,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 29438
+    },
+    {
+      "epoch": 0.7007283633247644,
+      "grad_norm": 4.892220057901184,
+      "learning_rate": 2e-05,
+      "loss": 0.7026,
+      "step": 29439
+    },
+    {
+      "epoch": 0.7007521660477959,
+      "grad_norm": 2.3208960223316186,
+      "learning_rate": 2e-05,
+      "loss": 0.3136,
+      "step": 29440
+    },
+    {
+      "epoch": 0.7007759687708274,
+      "grad_norm": 4.53595687774337,
+      "learning_rate": 2e-05,
+      "loss": 1.0309,
+      "step": 29441
+    },
+    {
+      "epoch": 0.7007997714938589,
+      "grad_norm": 6.481224592142962,
+      "learning_rate": 2e-05,
+      "loss": 1.0281,
+      "step": 29442
+    },
+    {
+      "epoch": 0.7008235742168905,
+      "grad_norm": 6.690338838025082,
+      "learning_rate": 2e-05,
+      "loss": 1.252,
+      "step": 29443
+    },
+    {
+      "epoch": 0.7008473769399219,
+      "grad_norm": 13.042454563090656,
+      "learning_rate": 2e-05,
+      "loss": 0.4372,
+      "step": 29444
+    },
+    {
+      "epoch": 0.7008711796629534,
+      "grad_norm": 1.7352831681191365,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 29445
+    },
+    {
+      "epoch": 0.7008949823859849,
+      "grad_norm": 3.3939474365795,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 29446
+    },
+    {
+      "epoch": 0.7009187851090165,
+      "grad_norm": 5.570666649234607,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 29447
+    },
+    {
+      "epoch": 0.700942587832048,
+      "grad_norm": 3.159695286135674,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 29448
+    },
+    {
+      "epoch": 0.7009663905550795,
+      "grad_norm": 3.2596258405180545,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 29449
+    },
+    {
+      "epoch": 0.700990193278111,
+      "grad_norm": 4.654673701161277,
+      "learning_rate": 2e-05,
+      "loss": 0.1191,
+      "step": 29450
+    },
+    {
+      "epoch": 0.7010139960011426,
+      "grad_norm": 1.9606057444057818,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 29451
+    },
+    {
+      "epoch": 0.701037798724174,
+      "grad_norm": 2.485204811910648,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 29452
+    },
+    {
+      "epoch": 0.7010616014472055,
+      "grad_norm": 4.0343964685643705,
+      "learning_rate": 2e-05,
+      "loss": 0.5069,
+      "step": 29453
+    },
+    {
+      "epoch": 0.7010854041702371,
+      "grad_norm": 0.43614167000737075,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 29454
+    },
+    {
+      "epoch": 0.7011092068932686,
+      "grad_norm": 0.43303258233742437,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 29455
+    },
+    {
+      "epoch": 0.7011330096163001,
+      "grad_norm": 5.1190871739177135,
+      "learning_rate": 2e-05,
+      "loss": 0.647,
+      "step": 29456
+    },
+    {
+      "epoch": 0.7011568123393316,
+      "grad_norm": 2.036722983402359,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 29457
+    },
+    {
+      "epoch": 0.7011806150623632,
+      "grad_norm": 4.805611946106602,
+      "learning_rate": 2e-05,
+      "loss": 0.9615,
+      "step": 29458
+    },
+    {
+      "epoch": 0.7012044177853947,
+      "grad_norm": 1.1012566540405944,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 29459
+    },
+    {
+      "epoch": 0.7012282205084261,
+      "grad_norm": 3.984720935692778,
+      "learning_rate": 2e-05,
+      "loss": 0.1919,
+      "step": 29460
+    },
+    {
+      "epoch": 0.7012520232314576,
+      "grad_norm": 0.24110244826571,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 29461
+    },
+    {
+      "epoch": 0.7012758259544892,
+      "grad_norm": 11.156365831711428,
+      "learning_rate": 2e-05,
+      "loss": 1.4436,
+      "step": 29462
+    },
+    {
+      "epoch": 0.7012996286775207,
+      "grad_norm": 3.0149120641085467,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 29463
+    },
+    {
+      "epoch": 0.7013234314005522,
+      "grad_norm": 4.591107255817946,
+      "learning_rate": 2e-05,
+      "loss": 0.4831,
+      "step": 29464
+    },
+    {
+      "epoch": 0.7013472341235837,
+      "grad_norm": 0.2101530473347675,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 29465
+    },
+    {
+      "epoch": 0.7013710368466153,
+      "grad_norm": 1.7824548711388324,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 29466
+    },
+    {
+      "epoch": 0.7013948395696468,
+      "grad_norm": 2.1733030934885997,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 29467
+    },
+    {
+      "epoch": 0.7014186422926783,
+      "grad_norm": 1.0161842395714773,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 29468
+    },
+    {
+      "epoch": 0.7014424450157098,
+      "grad_norm": 3.2939076356362413,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 29469
+    },
+    {
+      "epoch": 0.7014662477387413,
+      "grad_norm": 7.199957446715384,
+      "learning_rate": 2e-05,
+      "loss": 0.5081,
+      "step": 29470
+    },
+    {
+      "epoch": 0.7014900504617728,
+      "grad_norm": 4.3301682351311594,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 29471
+    },
+    {
+      "epoch": 0.7015138531848043,
+      "grad_norm": 2.9532866211005406,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 29472
+    },
+    {
+      "epoch": 0.7015376559078359,
+      "grad_norm": 2.513841971176897,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 29473
+    },
+    {
+      "epoch": 0.7015614586308674,
+      "grad_norm": 13.76475021285753,
+      "learning_rate": 2e-05,
+      "loss": 1.2537,
+      "step": 29474
+    },
+    {
+      "epoch": 0.7015852613538989,
+      "grad_norm": 1.7554589998626677,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 29475
+    },
+    {
+      "epoch": 0.7016090640769304,
+      "grad_norm": 6.383226150535895,
+      "learning_rate": 2e-05,
+      "loss": 1.2588,
+      "step": 29476
+    },
+    {
+      "epoch": 0.701632866799962,
+      "grad_norm": 14.04959029307983,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 29477
+    },
+    {
+      "epoch": 0.7016566695229934,
+      "grad_norm": 3.965796794025364,
+      "learning_rate": 2e-05,
+      "loss": 0.7479,
+      "step": 29478
+    },
+    {
+      "epoch": 0.7016804722460249,
+      "grad_norm": 6.069388591905179,
+      "learning_rate": 2e-05,
+      "loss": 1.0044,
+      "step": 29479
+    },
+    {
+      "epoch": 0.7017042749690564,
+      "grad_norm": 7.1976233445015065,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 29480
+    },
+    {
+      "epoch": 0.701728077692088,
+      "grad_norm": 0.022340023831283855,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 29481
+    },
+    {
+      "epoch": 0.7017518804151195,
+      "grad_norm": 0.03973020059050702,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 29482
+    },
+    {
+      "epoch": 0.701775683138151,
+      "grad_norm": 19.91767469567765,
+      "learning_rate": 2e-05,
+      "loss": 0.7655,
+      "step": 29483
+    },
+    {
+      "epoch": 0.7017994858611826,
+      "grad_norm": 0.04523023458212774,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 29484
+    },
+    {
+      "epoch": 0.7018232885842141,
+      "grad_norm": 6.643567074939478,
+      "learning_rate": 2e-05,
+      "loss": 1.1142,
+      "step": 29485
+    },
+    {
+      "epoch": 0.7018470913072455,
+      "grad_norm": 13.490625607266086,
+      "learning_rate": 2e-05,
+      "loss": 0.9362,
+      "step": 29486
+    },
+    {
+      "epoch": 0.701870894030277,
+      "grad_norm": 3.63761149853837,
+      "learning_rate": 2e-05,
+      "loss": 0.466,
+      "step": 29487
+    },
+    {
+      "epoch": 0.7018946967533086,
+      "grad_norm": 5.711603247001089,
+      "learning_rate": 2e-05,
+      "loss": 1.3989,
+      "step": 29488
+    },
+    {
+      "epoch": 0.7019184994763401,
+      "grad_norm": 0.9365520978743725,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 29489
+    },
+    {
+      "epoch": 0.7019423021993716,
+      "grad_norm": 2.4581619522573894,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 29490
+    },
+    {
+      "epoch": 0.7019661049224031,
+      "grad_norm": 9.720459122045817,
+      "learning_rate": 2e-05,
+      "loss": 0.3114,
+      "step": 29491
+    },
+    {
+      "epoch": 0.7019899076454347,
+      "grad_norm": 2.4636787024540574,
+      "learning_rate": 2e-05,
+      "loss": 0.2599,
+      "step": 29492
+    },
+    {
+      "epoch": 0.7020137103684662,
+      "grad_norm": 6.77367332127179,
+      "learning_rate": 2e-05,
+      "loss": 0.8334,
+      "step": 29493
+    },
+    {
+      "epoch": 0.7020375130914976,
+      "grad_norm": 28.43794658740383,
+      "learning_rate": 2e-05,
+      "loss": 0.3941,
+      "step": 29494
+    },
+    {
+      "epoch": 0.7020613158145291,
+      "grad_norm": 5.5167454907451745,
+      "learning_rate": 2e-05,
+      "loss": 0.999,
+      "step": 29495
+    },
+    {
+      "epoch": 0.7020851185375607,
+      "grad_norm": 0.7517002922902698,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 29496
+    },
+    {
+      "epoch": 0.7021089212605922,
+      "grad_norm": 5.687074085033647,
+      "learning_rate": 2e-05,
+      "loss": 0.7939,
+      "step": 29497
+    },
+    {
+      "epoch": 0.7021327239836237,
+      "grad_norm": 4.475254278051778,
+      "learning_rate": 2e-05,
+      "loss": 0.5529,
+      "step": 29498
+    },
+    {
+      "epoch": 0.7021565267066553,
+      "grad_norm": 3.4598004238984403,
+      "learning_rate": 2e-05,
+      "loss": 0.3191,
+      "step": 29499
+    },
+    {
+      "epoch": 0.7021803294296868,
+      "grad_norm": 5.036140672622172,
+      "learning_rate": 2e-05,
+      "loss": 0.5134,
+      "step": 29500
+    },
+    {
+      "epoch": 0.7022041321527183,
+      "grad_norm": 11.807412556255429,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 29501
+    },
+    {
+      "epoch": 0.7022279348757497,
+      "grad_norm": 2.27659673541988,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 29502
+    },
+    {
+      "epoch": 0.7022517375987813,
+      "grad_norm": 0.13968696372063985,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29503
+    },
+    {
+      "epoch": 0.7022755403218128,
+      "grad_norm": 0.043683837421952885,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 29504
+    },
+    {
+      "epoch": 0.7022993430448443,
+      "grad_norm": 9.756519660372733,
+      "learning_rate": 2e-05,
+      "loss": 0.9791,
+      "step": 29505
+    },
+    {
+      "epoch": 0.7023231457678758,
+      "grad_norm": 3.356363745489116,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 29506
+    },
+    {
+      "epoch": 0.7023469484909074,
+      "grad_norm": 4.73620155775487,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 29507
+    },
+    {
+      "epoch": 0.7023707512139389,
+      "grad_norm": 3.0648097908405343,
+      "learning_rate": 2e-05,
+      "loss": 0.6297,
+      "step": 29508
+    },
+    {
+      "epoch": 0.7023945539369704,
+      "grad_norm": 4.124431585315355,
+      "learning_rate": 2e-05,
+      "loss": 0.3965,
+      "step": 29509
+    },
+    {
+      "epoch": 0.7024183566600019,
+      "grad_norm": 22.35750628135873,
+      "learning_rate": 2e-05,
+      "loss": 0.8829,
+      "step": 29510
+    },
+    {
+      "epoch": 0.7024421593830334,
+      "grad_norm": 4.061752073469547,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 29511
+    },
+    {
+      "epoch": 0.7024659621060649,
+      "grad_norm": 1.2398961267655682,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 29512
+    },
+    {
+      "epoch": 0.7024897648290964,
+      "grad_norm": 0.33677886874173896,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 29513
+    },
+    {
+      "epoch": 0.702513567552128,
+      "grad_norm": 4.651497559211934,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 29514
+    },
+    {
+      "epoch": 0.7025373702751595,
+      "grad_norm": 0.8313708746619699,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 29515
+    },
+    {
+      "epoch": 0.702561172998191,
+      "grad_norm": 0.04278465852602413,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29516
+    },
+    {
+      "epoch": 0.7025849757212225,
+      "grad_norm": 2.4313716599377844,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 29517
+    },
+    {
+      "epoch": 0.7026087784442541,
+      "grad_norm": 1.2112487406117056,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 29518
+    },
+    {
+      "epoch": 0.7026325811672856,
+      "grad_norm": 1.9616540142917926,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 29519
+    },
+    {
+      "epoch": 0.702656383890317,
+      "grad_norm": 1.4105439100403503,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 29520
+    },
+    {
+      "epoch": 0.7026801866133485,
+      "grad_norm": 3.6897598771283566,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 29521
+    },
+    {
+      "epoch": 0.7027039893363801,
+      "grad_norm": 0.3525964009286027,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 29522
+    },
+    {
+      "epoch": 0.7027277920594116,
+      "grad_norm": 1.6394143930319276,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 29523
+    },
+    {
+      "epoch": 0.7027515947824431,
+      "grad_norm": 1.8179488476917829,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 29524
+    },
+    {
+      "epoch": 0.7027753975054746,
+      "grad_norm": 5.295433790869406,
+      "learning_rate": 2e-05,
+      "loss": 0.3552,
+      "step": 29525
+    },
+    {
+      "epoch": 0.7027992002285062,
+      "grad_norm": 1.516325020403222,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 29526
+    },
+    {
+      "epoch": 0.7028230029515377,
+      "grad_norm": 3.5781446084419035,
+      "learning_rate": 2e-05,
+      "loss": 0.5686,
+      "step": 29527
+    },
+    {
+      "epoch": 0.7028468056745691,
+      "grad_norm": 4.8224303775560955,
+      "learning_rate": 2e-05,
+      "loss": 0.9423,
+      "step": 29528
+    },
+    {
+      "epoch": 0.7028706083976007,
+      "grad_norm": 8.591798814426403,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 29529
+    },
+    {
+      "epoch": 0.7028944111206322,
+      "grad_norm": 4.820270344070124,
+      "learning_rate": 2e-05,
+      "loss": 0.7369,
+      "step": 29530
+    },
+    {
+      "epoch": 0.7029182138436637,
+      "grad_norm": 0.1056746218453658,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 29531
+    },
+    {
+      "epoch": 0.7029420165666952,
+      "grad_norm": 3.583589469714897,
+      "learning_rate": 2e-05,
+      "loss": 0.5749,
+      "step": 29532
+    },
+    {
+      "epoch": 0.7029658192897268,
+      "grad_norm": 7.511323374134232,
+      "learning_rate": 2e-05,
+      "loss": 0.9147,
+      "step": 29533
+    },
+    {
+      "epoch": 0.7029896220127583,
+      "grad_norm": 0.7817891997758751,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 29534
+    },
+    {
+      "epoch": 0.7030134247357898,
+      "grad_norm": 4.940278454834877,
+      "learning_rate": 2e-05,
+      "loss": 1.1294,
+      "step": 29535
+    },
+    {
+      "epoch": 0.7030372274588212,
+      "grad_norm": 0.38437992443205016,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 29536
+    },
+    {
+      "epoch": 0.7030610301818528,
+      "grad_norm": 2.4327648530648442,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 29537
+    },
+    {
+      "epoch": 0.7030848329048843,
+      "grad_norm": 8.776475672585255,
+      "learning_rate": 2e-05,
+      "loss": 1.3874,
+      "step": 29538
+    },
+    {
+      "epoch": 0.7031086356279158,
+      "grad_norm": 7.29639794473465,
+      "learning_rate": 2e-05,
+      "loss": 0.2414,
+      "step": 29539
+    },
+    {
+      "epoch": 0.7031324383509473,
+      "grad_norm": 1.6994171282465347,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 29540
+    },
+    {
+      "epoch": 0.7031562410739789,
+      "grad_norm": 1.6413887072511835,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 29541
+    },
+    {
+      "epoch": 0.7031800437970104,
+      "grad_norm": 0.7015290726711159,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 29542
+    },
+    {
+      "epoch": 0.7032038465200419,
+      "grad_norm": 6.969611627461973,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 29543
+    },
+    {
+      "epoch": 0.7032276492430735,
+      "grad_norm": 0.5702585980513375,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 29544
+    },
+    {
+      "epoch": 0.703251451966105,
+      "grad_norm": 62.89800892628911,
+      "learning_rate": 2e-05,
+      "loss": 0.2202,
+      "step": 29545
+    },
+    {
+      "epoch": 0.7032752546891364,
+      "grad_norm": 7.536658014308585,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 29546
+    },
+    {
+      "epoch": 0.7032990574121679,
+      "grad_norm": 3.693595384022084,
+      "learning_rate": 2e-05,
+      "loss": 0.4257,
+      "step": 29547
+    },
+    {
+      "epoch": 0.7033228601351995,
+      "grad_norm": 4.991798660760858,
+      "learning_rate": 2e-05,
+      "loss": 0.3254,
+      "step": 29548
+    },
+    {
+      "epoch": 0.703346662858231,
+      "grad_norm": 12.76507190447633,
+      "learning_rate": 2e-05,
+      "loss": 0.8927,
+      "step": 29549
+    },
+    {
+      "epoch": 0.7033704655812625,
+      "grad_norm": 0.7204211259450198,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 29550
+    },
+    {
+      "epoch": 0.703394268304294,
+      "grad_norm": 12.035524309657397,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 29551
+    },
+    {
+      "epoch": 0.7034180710273256,
+      "grad_norm": 13.237142631951667,
+      "learning_rate": 2e-05,
+      "loss": 0.2794,
+      "step": 29552
+    },
+    {
+      "epoch": 0.703441873750357,
+      "grad_norm": 4.377674625776173,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 29553
+    },
+    {
+      "epoch": 0.7034656764733885,
+      "grad_norm": 16.586906236126776,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 29554
+    },
+    {
+      "epoch": 0.70348947919642,
+      "grad_norm": 0.36451420071390517,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 29555
+    },
+    {
+      "epoch": 0.7035132819194516,
+      "grad_norm": 0.6944091295759642,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 29556
+    },
+    {
+      "epoch": 0.7035370846424831,
+      "grad_norm": 8.7361981377083,
+      "learning_rate": 2e-05,
+      "loss": 0.9591,
+      "step": 29557
+    },
+    {
+      "epoch": 0.7035608873655146,
+      "grad_norm": 2.3277859488218895,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 29558
+    },
+    {
+      "epoch": 0.7035846900885462,
+      "grad_norm": 0.11673052886503198,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 29559
+    },
+    {
+      "epoch": 0.7036084928115777,
+      "grad_norm": 6.046877203913037,
+      "learning_rate": 2e-05,
+      "loss": 1.4695,
+      "step": 29560
+    },
+    {
+      "epoch": 0.7036322955346092,
+      "grad_norm": 4.834825579128477,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 29561
+    },
+    {
+      "epoch": 0.7036560982576406,
+      "grad_norm": 0.9522216750897354,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 29562
+    },
+    {
+      "epoch": 0.7036799009806722,
+      "grad_norm": 0.5561963271592198,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 29563
+    },
+    {
+      "epoch": 0.7037037037037037,
+      "grad_norm": 3.7328458251353416,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 29564
+    },
+    {
+      "epoch": 0.7037275064267352,
+      "grad_norm": 3.2361907066694378,
+      "learning_rate": 2e-05,
+      "loss": 0.431,
+      "step": 29565
+    },
+    {
+      "epoch": 0.7037513091497667,
+      "grad_norm": 1.7551172869986222,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 29566
+    },
+    {
+      "epoch": 0.7037751118727983,
+      "grad_norm": 1.6609999173128445,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 29567
+    },
+    {
+      "epoch": 0.7037989145958298,
+      "grad_norm": 8.15119500245866,
+      "learning_rate": 2e-05,
+      "loss": 1.0309,
+      "step": 29568
+    },
+    {
+      "epoch": 0.7038227173188613,
+      "grad_norm": 0.11558842141975674,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 29569
+    },
+    {
+      "epoch": 0.7038465200418927,
+      "grad_norm": 2.836671832486665,
+      "learning_rate": 2e-05,
+      "loss": 0.5535,
+      "step": 29570
+    },
+    {
+      "epoch": 0.7038703227649243,
+      "grad_norm": 0.34588348701729804,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 29571
+    },
+    {
+      "epoch": 0.7038941254879558,
+      "grad_norm": 5.080533056837872,
+      "learning_rate": 2e-05,
+      "loss": 0.8881,
+      "step": 29572
+    },
+    {
+      "epoch": 0.7039179282109873,
+      "grad_norm": 0.8892313174637657,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 29573
+    },
+    {
+      "epoch": 0.7039417309340189,
+      "grad_norm": 7.552720545971651,
+      "learning_rate": 2e-05,
+      "loss": 1.8284,
+      "step": 29574
+    },
+    {
+      "epoch": 0.7039655336570504,
+      "grad_norm": 4.174547686118108,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 29575
+    },
+    {
+      "epoch": 0.7039893363800819,
+      "grad_norm": 2.8432143361935727,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 29576
+    },
+    {
+      "epoch": 0.7040131391031134,
+      "grad_norm": 0.3032679435469187,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 29577
+    },
+    {
+      "epoch": 0.704036941826145,
+      "grad_norm": 3.440941191350133,
+      "learning_rate": 2e-05,
+      "loss": 0.5855,
+      "step": 29578
+    },
+    {
+      "epoch": 0.7040607445491764,
+      "grad_norm": 7.046871724079302,
+      "learning_rate": 2e-05,
+      "loss": 0.5197,
+      "step": 29579
+    },
+    {
+      "epoch": 0.7040845472722079,
+      "grad_norm": 5.461691760006665,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 29580
+    },
+    {
+      "epoch": 0.7041083499952394,
+      "grad_norm": 1.2150772333135604,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 29581
+    },
+    {
+      "epoch": 0.704132152718271,
+      "grad_norm": 9.96206746070675,
+      "learning_rate": 2e-05,
+      "loss": 0.2033,
+      "step": 29582
+    },
+    {
+      "epoch": 0.7041559554413025,
+      "grad_norm": 1.1336534016325264,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 29583
+    },
+    {
+      "epoch": 0.704179758164334,
+      "grad_norm": 10.856525924352523,
+      "learning_rate": 2e-05,
+      "loss": 0.903,
+      "step": 29584
+    },
+    {
+      "epoch": 0.7042035608873655,
+      "grad_norm": 0.49422823659980436,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 29585
+    },
+    {
+      "epoch": 0.7042273636103971,
+      "grad_norm": 5.570065481117922,
+      "learning_rate": 2e-05,
+      "loss": 0.0976,
+      "step": 29586
+    },
+    {
+      "epoch": 0.7042511663334285,
+      "grad_norm": 0.16932609643247795,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 29587
+    },
+    {
+      "epoch": 0.70427496905646,
+      "grad_norm": 2.2730206570436238,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 29588
+    },
+    {
+      "epoch": 0.7042987717794916,
+      "grad_norm": 4.703762097133639,
+      "learning_rate": 2e-05,
+      "loss": 0.7808,
+      "step": 29589
+    },
+    {
+      "epoch": 0.7043225745025231,
+      "grad_norm": 80.62139681747247,
+      "learning_rate": 2e-05,
+      "loss": 0.2699,
+      "step": 29590
+    },
+    {
+      "epoch": 0.7043463772255546,
+      "grad_norm": 5.423367945176758,
+      "learning_rate": 2e-05,
+      "loss": 0.7653,
+      "step": 29591
+    },
+    {
+      "epoch": 0.7043701799485861,
+      "grad_norm": 6.753254963790817,
+      "learning_rate": 2e-05,
+      "loss": 1.0528,
+      "step": 29592
+    },
+    {
+      "epoch": 0.7043939826716177,
+      "grad_norm": 3.318454357671984,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 29593
+    },
+    {
+      "epoch": 0.7044177853946492,
+      "grad_norm": 0.7731665989641552,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 29594
+    },
+    {
+      "epoch": 0.7044415881176807,
+      "grad_norm": 9.899866165517182,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 29595
+    },
+    {
+      "epoch": 0.7044653908407121,
+      "grad_norm": 4.543108907106262,
+      "learning_rate": 2e-05,
+      "loss": 0.7147,
+      "step": 29596
+    },
+    {
+      "epoch": 0.7044891935637437,
+      "grad_norm": 2.685303160054256,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 29597
+    },
+    {
+      "epoch": 0.7045129962867752,
+      "grad_norm": 3.9376074354651203,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 29598
+    },
+    {
+      "epoch": 0.7045367990098067,
+      "grad_norm": 7.984214313583323,
+      "learning_rate": 2e-05,
+      "loss": 0.301,
+      "step": 29599
+    },
+    {
+      "epoch": 0.7045606017328382,
+      "grad_norm": 1.221428672630766,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 29600
+    },
+    {
+      "epoch": 0.7045844044558698,
+      "grad_norm": 6.3212402161233285,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 29601
+    },
+    {
+      "epoch": 0.7046082071789013,
+      "grad_norm": 9.546470369286862,
+      "learning_rate": 2e-05,
+      "loss": 0.2687,
+      "step": 29602
+    },
+    {
+      "epoch": 0.7046320099019328,
+      "grad_norm": 3.6147451929338583,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 29603
+    },
+    {
+      "epoch": 0.7046558126249644,
+      "grad_norm": 6.35222813071648,
+      "learning_rate": 2e-05,
+      "loss": 1.4296,
+      "step": 29604
+    },
+    {
+      "epoch": 0.7046796153479958,
+      "grad_norm": 4.014858218982407,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 29605
+    },
+    {
+      "epoch": 0.7047034180710273,
+      "grad_norm": 0.2302487517542317,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 29606
+    },
+    {
+      "epoch": 0.7047272207940588,
+      "grad_norm": 5.028397302372198,
+      "learning_rate": 2e-05,
+      "loss": 1.022,
+      "step": 29607
+    },
+    {
+      "epoch": 0.7047510235170904,
+      "grad_norm": 4.267118004889514,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 29608
+    },
+    {
+      "epoch": 0.7047748262401219,
+      "grad_norm": 2.4562586691036814,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 29609
+    },
+    {
+      "epoch": 0.7047986289631534,
+      "grad_norm": 0.6430686157237131,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 29610
+    },
+    {
+      "epoch": 0.7048224316861849,
+      "grad_norm": 8.551287878684299,
+      "learning_rate": 2e-05,
+      "loss": 0.1839,
+      "step": 29611
+    },
+    {
+      "epoch": 0.7048462344092165,
+      "grad_norm": 24.545499454365356,
+      "learning_rate": 2e-05,
+      "loss": 0.2707,
+      "step": 29612
+    },
+    {
+      "epoch": 0.7048700371322479,
+      "grad_norm": 4.716121477771128,
+      "learning_rate": 2e-05,
+      "loss": 0.8994,
+      "step": 29613
+    },
+    {
+      "epoch": 0.7048938398552794,
+      "grad_norm": 0.10891871644621735,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29614
+    },
+    {
+      "epoch": 0.7049176425783109,
+      "grad_norm": 3.4670202838549864,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 29615
+    },
+    {
+      "epoch": 0.7049414453013425,
+      "grad_norm": 0.30237782768402943,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 29616
+    },
+    {
+      "epoch": 0.704965248024374,
+      "grad_norm": 3.021534122367408,
+      "learning_rate": 2e-05,
+      "loss": 0.3738,
+      "step": 29617
+    },
+    {
+      "epoch": 0.7049890507474055,
+      "grad_norm": 3.4260635284581484,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 29618
+    },
+    {
+      "epoch": 0.7050128534704371,
+      "grad_norm": 1.9299523084695271,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 29619
+    },
+    {
+      "epoch": 0.7050366561934686,
+      "grad_norm": 10.546007470840724,
+      "learning_rate": 2e-05,
+      "loss": 0.5023,
+      "step": 29620
+    },
+    {
+      "epoch": 0.7050604589165,
+      "grad_norm": 3.2543111018114934,
+      "learning_rate": 2e-05,
+      "loss": 0.4359,
+      "step": 29621
+    },
+    {
+      "epoch": 0.7050842616395315,
+      "grad_norm": 9.193457585564927,
+      "learning_rate": 2e-05,
+      "loss": 0.5511,
+      "step": 29622
+    },
+    {
+      "epoch": 0.7051080643625631,
+      "grad_norm": 4.452084432780847,
+      "learning_rate": 2e-05,
+      "loss": 1.2136,
+      "step": 29623
+    },
+    {
+      "epoch": 0.7051318670855946,
+      "grad_norm": 3.3263403719753555,
+      "learning_rate": 2e-05,
+      "loss": 0.5918,
+      "step": 29624
+    },
+    {
+      "epoch": 0.7051556698086261,
+      "grad_norm": 1.0261614361304674,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 29625
+    },
+    {
+      "epoch": 0.7051794725316576,
+      "grad_norm": 0.029564298376300285,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29626
+    },
+    {
+      "epoch": 0.7052032752546892,
+      "grad_norm": 3.5003998880487845,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 29627
+    },
+    {
+      "epoch": 0.7052270779777207,
+      "grad_norm": 3.8997245549445414,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 29628
+    },
+    {
+      "epoch": 0.7052508807007521,
+      "grad_norm": 1.5223106007696705,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 29629
+    },
+    {
+      "epoch": 0.7052746834237836,
+      "grad_norm": 4.233091577761042,
+      "learning_rate": 2e-05,
+      "loss": 1.0546,
+      "step": 29630
+    },
+    {
+      "epoch": 0.7052984861468152,
+      "grad_norm": 5.271741309657741,
+      "learning_rate": 2e-05,
+      "loss": 1.0663,
+      "step": 29631
+    },
+    {
+      "epoch": 0.7053222888698467,
+      "grad_norm": 4.211738284961605,
+      "learning_rate": 2e-05,
+      "loss": 0.9345,
+      "step": 29632
+    },
+    {
+      "epoch": 0.7053460915928782,
+      "grad_norm": 3.8233009963144418,
+      "learning_rate": 2e-05,
+      "loss": 0.7059,
+      "step": 29633
+    },
+    {
+      "epoch": 0.7053698943159097,
+      "grad_norm": 0.023210868743047192,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 29634
+    },
+    {
+      "epoch": 0.7053936970389413,
+      "grad_norm": 4.955263108089839,
+      "learning_rate": 2e-05,
+      "loss": 1.3353,
+      "step": 29635
+    },
+    {
+      "epoch": 0.7054174997619728,
+      "grad_norm": 0.3626115999466934,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 29636
+    },
+    {
+      "epoch": 0.7054413024850043,
+      "grad_norm": 3.321537134474108,
+      "learning_rate": 2e-05,
+      "loss": 0.2326,
+      "step": 29637
+    },
+    {
+      "epoch": 0.7054651052080358,
+      "grad_norm": 6.114048562977979,
+      "learning_rate": 2e-05,
+      "loss": 0.6592,
+      "step": 29638
+    },
+    {
+      "epoch": 0.7054889079310673,
+      "grad_norm": 3.1543858924831114,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 29639
+    },
+    {
+      "epoch": 0.7055127106540988,
+      "grad_norm": 0.6317036532366447,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 29640
+    },
+    {
+      "epoch": 0.7055365133771303,
+      "grad_norm": 2.4059685274563445,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 29641
+    },
+    {
+      "epoch": 0.7055603161001619,
+      "grad_norm": 3.838536919345042,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 29642
+    },
+    {
+      "epoch": 0.7055841188231934,
+      "grad_norm": 7.78046086688469,
+      "learning_rate": 2e-05,
+      "loss": 1.2233,
+      "step": 29643
+    },
+    {
+      "epoch": 0.7056079215462249,
+      "grad_norm": 1.5591871743238725,
+      "learning_rate": 2e-05,
+      "loss": 0.0894,
+      "step": 29644
+    },
+    {
+      "epoch": 0.7056317242692564,
+      "grad_norm": 1.4416561408329847,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 29645
+    },
+    {
+      "epoch": 0.705655526992288,
+      "grad_norm": 0.026618461978959895,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 29646
+    },
+    {
+      "epoch": 0.7056793297153194,
+      "grad_norm": 2.2615064768009825,
+      "learning_rate": 2e-05,
+      "loss": 0.4167,
+      "step": 29647
+    },
+    {
+      "epoch": 0.7057031324383509,
+      "grad_norm": 0.12610600061442667,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 29648
+    },
+    {
+      "epoch": 0.7057269351613824,
+      "grad_norm": 3.4697911844783977,
+      "learning_rate": 2e-05,
+      "loss": 0.1454,
+      "step": 29649
+    },
+    {
+      "epoch": 0.705750737884414,
+      "grad_norm": 4.76220218584695,
+      "learning_rate": 2e-05,
+      "loss": 0.6405,
+      "step": 29650
+    },
+    {
+      "epoch": 0.7057745406074455,
+      "grad_norm": 2.320430083884633,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 29651
+    },
+    {
+      "epoch": 0.705798343330477,
+      "grad_norm": 1.5325333570699908,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 29652
+    },
+    {
+      "epoch": 0.7058221460535086,
+      "grad_norm": 2.66871957038451,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 29653
+    },
+    {
+      "epoch": 0.70584594877654,
+      "grad_norm": 1.8177043587389194,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 29654
+    },
+    {
+      "epoch": 0.7058697514995715,
+      "grad_norm": 0.2000751999966798,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 29655
+    },
+    {
+      "epoch": 0.705893554222603,
+      "grad_norm": 6.570271523788489,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 29656
+    },
+    {
+      "epoch": 0.7059173569456346,
+      "grad_norm": 3.236419290861022,
+      "learning_rate": 2e-05,
+      "loss": 0.5728,
+      "step": 29657
+    },
+    {
+      "epoch": 0.7059411596686661,
+      "grad_norm": 10.719427807690185,
+      "learning_rate": 2e-05,
+      "loss": 0.8094,
+      "step": 29658
+    },
+    {
+      "epoch": 0.7059649623916976,
+      "grad_norm": 2.9522957653417503,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 29659
+    },
+    {
+      "epoch": 0.7059887651147291,
+      "grad_norm": 3.693736506442558,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 29660
+    },
+    {
+      "epoch": 0.7060125678377607,
+      "grad_norm": 0.10956598804946237,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29661
+    },
+    {
+      "epoch": 0.7060363705607922,
+      "grad_norm": 6.325900753258591,
+      "learning_rate": 2e-05,
+      "loss": 0.5292,
+      "step": 29662
+    },
+    {
+      "epoch": 0.7060601732838236,
+      "grad_norm": 1.0314447723724085,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 29663
+    },
+    {
+      "epoch": 0.7060839760068551,
+      "grad_norm": 4.554752887605375,
+      "learning_rate": 2e-05,
+      "loss": 0.4941,
+      "step": 29664
+    },
+    {
+      "epoch": 0.7061077787298867,
+      "grad_norm": 28.323780600298697,
+      "learning_rate": 2e-05,
+      "loss": 0.4834,
+      "step": 29665
+    },
+    {
+      "epoch": 0.7061315814529182,
+      "grad_norm": 4.701869456930153,
+      "learning_rate": 2e-05,
+      "loss": 0.9001,
+      "step": 29666
+    },
+    {
+      "epoch": 0.7061553841759497,
+      "grad_norm": 0.921470135331984,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 29667
+    },
+    {
+      "epoch": 0.7061791868989813,
+      "grad_norm": 9.320195800707259,
+      "learning_rate": 2e-05,
+      "loss": 0.6387,
+      "step": 29668
+    },
+    {
+      "epoch": 0.7062029896220128,
+      "grad_norm": 0.33196749094304934,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 29669
+    },
+    {
+      "epoch": 0.7062267923450443,
+      "grad_norm": 0.16057061566742858,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 29670
+    },
+    {
+      "epoch": 0.7062505950680757,
+      "grad_norm": 1.9838158379058277,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 29671
+    },
+    {
+      "epoch": 0.7062743977911073,
+      "grad_norm": 15.57344446602117,
+      "learning_rate": 2e-05,
+      "loss": 0.8699,
+      "step": 29672
+    },
+    {
+      "epoch": 0.7062982005141388,
+      "grad_norm": 7.767219638817146,
+      "learning_rate": 2e-05,
+      "loss": 0.601,
+      "step": 29673
+    },
+    {
+      "epoch": 0.7063220032371703,
+      "grad_norm": 0.1437942628497422,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 29674
+    },
+    {
+      "epoch": 0.7063458059602018,
+      "grad_norm": 3.7563945897390707,
+      "learning_rate": 2e-05,
+      "loss": 0.6931,
+      "step": 29675
+    },
+    {
+      "epoch": 0.7063696086832334,
+      "grad_norm": 0.24914309803382986,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 29676
+    },
+    {
+      "epoch": 0.7063934114062649,
+      "grad_norm": 0.041854622927793704,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 29677
+    },
+    {
+      "epoch": 0.7064172141292964,
+      "grad_norm": 1.4592113762385135,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 29678
+    },
+    {
+      "epoch": 0.7064410168523279,
+      "grad_norm": 7.350771792401561,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 29679
+    },
+    {
+      "epoch": 0.7064648195753594,
+      "grad_norm": 1.2444636263154272,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 29680
+    },
+    {
+      "epoch": 0.7064886222983909,
+      "grad_norm": 4.480084523660687,
+      "learning_rate": 2e-05,
+      "loss": 0.5758,
+      "step": 29681
+    },
+    {
+      "epoch": 0.7065124250214224,
+      "grad_norm": 0.9937384955646995,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 29682
+    },
+    {
+      "epoch": 0.706536227744454,
+      "grad_norm": 6.363748068706415,
+      "learning_rate": 2e-05,
+      "loss": 0.6016,
+      "step": 29683
+    },
+    {
+      "epoch": 0.7065600304674855,
+      "grad_norm": 2.410780061579595,
+      "learning_rate": 2e-05,
+      "loss": 0.1877,
+      "step": 29684
+    },
+    {
+      "epoch": 0.706583833190517,
+      "grad_norm": 3.9507119227135865,
+      "learning_rate": 2e-05,
+      "loss": 0.637,
+      "step": 29685
+    },
+    {
+      "epoch": 0.7066076359135485,
+      "grad_norm": 27.054115049655824,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 29686
+    },
+    {
+      "epoch": 0.7066314386365801,
+      "grad_norm": 0.9470129278340454,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 29687
+    },
+    {
+      "epoch": 0.7066552413596116,
+      "grad_norm": 5.882732989267403,
+      "learning_rate": 2e-05,
+      "loss": 0.5341,
+      "step": 29688
+    },
+    {
+      "epoch": 0.706679044082643,
+      "grad_norm": 5.765987112192576,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 29689
+    },
+    {
+      "epoch": 0.7067028468056745,
+      "grad_norm": 0.029835394967884257,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 29690
+    },
+    {
+      "epoch": 0.7067266495287061,
+      "grad_norm": 0.17123291976656524,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 29691
+    },
+    {
+      "epoch": 0.7067504522517376,
+      "grad_norm": 3.409648702510739,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 29692
+    },
+    {
+      "epoch": 0.7067742549747691,
+      "grad_norm": 2.7133900331721432,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 29693
+    },
+    {
+      "epoch": 0.7067980576978006,
+      "grad_norm": 14.597969466600373,
+      "learning_rate": 2e-05,
+      "loss": 0.2956,
+      "step": 29694
+    },
+    {
+      "epoch": 0.7068218604208322,
+      "grad_norm": 3.9018641173330884,
+      "learning_rate": 2e-05,
+      "loss": 0.9323,
+      "step": 29695
+    },
+    {
+      "epoch": 0.7068456631438637,
+      "grad_norm": 3.821084980908415,
+      "learning_rate": 2e-05,
+      "loss": 0.2536,
+      "step": 29696
+    },
+    {
+      "epoch": 0.7068694658668951,
+      "grad_norm": 4.079773920044774,
+      "learning_rate": 2e-05,
+      "loss": 0.6176,
+      "step": 29697
+    },
+    {
+      "epoch": 0.7068932685899267,
+      "grad_norm": 3.8006569123609877,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 29698
+    },
+    {
+      "epoch": 0.7069170713129582,
+      "grad_norm": 3.0763923070680144,
+      "learning_rate": 2e-05,
+      "loss": 0.6134,
+      "step": 29699
+    },
+    {
+      "epoch": 0.7069408740359897,
+      "grad_norm": 16.981242614953153,
+      "learning_rate": 2e-05,
+      "loss": 0.6375,
+      "step": 29700
+    },
+    {
+      "epoch": 0.7069646767590212,
+      "grad_norm": 0.49275784124097544,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 29701
+    },
+    {
+      "epoch": 0.7069884794820528,
+      "grad_norm": 7.941833063576394,
+      "learning_rate": 2e-05,
+      "loss": 0.2992,
+      "step": 29702
+    },
+    {
+      "epoch": 0.7070122822050843,
+      "grad_norm": 8.639527390635328,
+      "learning_rate": 2e-05,
+      "loss": 0.9954,
+      "step": 29703
+    },
+    {
+      "epoch": 0.7070360849281158,
+      "grad_norm": 0.2906418298842969,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 29704
+    },
+    {
+      "epoch": 0.7070598876511472,
+      "grad_norm": 6.520967964982664,
+      "learning_rate": 2e-05,
+      "loss": 0.8683,
+      "step": 29705
+    },
+    {
+      "epoch": 0.7070836903741788,
+      "grad_norm": 1.2575905680840842,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 29706
+    },
+    {
+      "epoch": 0.7071074930972103,
+      "grad_norm": 10.23742499943415,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 29707
+    },
+    {
+      "epoch": 0.7071312958202418,
+      "grad_norm": 3.5440309640714824,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 29708
+    },
+    {
+      "epoch": 0.7071550985432733,
+      "grad_norm": 0.4712301115670615,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 29709
+    },
+    {
+      "epoch": 0.7071789012663049,
+      "grad_norm": 3.6396270285697097,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 29710
+    },
+    {
+      "epoch": 0.7072027039893364,
+      "grad_norm": 0.5657418291926908,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 29711
+    },
+    {
+      "epoch": 0.7072265067123679,
+      "grad_norm": 2.525109063717321,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 29712
+    },
+    {
+      "epoch": 0.7072503094353995,
+      "grad_norm": 3.737652422929157,
+      "learning_rate": 2e-05,
+      "loss": 0.6346,
+      "step": 29713
+    },
+    {
+      "epoch": 0.707274112158431,
+      "grad_norm": 5.979991687377857,
+      "learning_rate": 2e-05,
+      "loss": 1.2538,
+      "step": 29714
+    },
+    {
+      "epoch": 0.7072979148814624,
+      "grad_norm": 7.0909068435265175,
+      "learning_rate": 2e-05,
+      "loss": 0.2518,
+      "step": 29715
+    },
+    {
+      "epoch": 0.7073217176044939,
+      "grad_norm": 14.300253067257962,
+      "learning_rate": 2e-05,
+      "loss": 0.589,
+      "step": 29716
+    },
+    {
+      "epoch": 0.7073455203275255,
+      "grad_norm": 0.2495561617368491,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 29717
+    },
+    {
+      "epoch": 0.707369323050557,
+      "grad_norm": 1.1516110789337028,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 29718
+    },
+    {
+      "epoch": 0.7073931257735885,
+      "grad_norm": 0.47816156640826035,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 29719
+    },
+    {
+      "epoch": 0.70741692849662,
+      "grad_norm": 0.8208980913964108,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 29720
+    },
+    {
+      "epoch": 0.7074407312196516,
+      "grad_norm": 3.8290862750759125,
+      "learning_rate": 2e-05,
+      "loss": 0.7448,
+      "step": 29721
+    },
+    {
+      "epoch": 0.707464533942683,
+      "grad_norm": 0.33052628270554546,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 29722
+    },
+    {
+      "epoch": 0.7074883366657145,
+      "grad_norm": 0.8974463976036737,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 29723
+    },
+    {
+      "epoch": 0.707512139388746,
+      "grad_norm": 0.35144264053834173,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 29724
+    },
+    {
+      "epoch": 0.7075359421117776,
+      "grad_norm": 0.1431599863682004,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 29725
+    },
+    {
+      "epoch": 0.7075597448348091,
+      "grad_norm": 0.9171081358451232,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 29726
+    },
+    {
+      "epoch": 0.7075835475578406,
+      "grad_norm": 5.780926316873403,
+      "learning_rate": 2e-05,
+      "loss": 1.1358,
+      "step": 29727
+    },
+    {
+      "epoch": 0.7076073502808722,
+      "grad_norm": 5.156998471227537,
+      "learning_rate": 2e-05,
+      "loss": 1.0921,
+      "step": 29728
+    },
+    {
+      "epoch": 0.7076311530039037,
+      "grad_norm": 1.7735979620682074,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 29729
+    },
+    {
+      "epoch": 0.7076549557269352,
+      "grad_norm": 9.778261715807904,
+      "learning_rate": 2e-05,
+      "loss": 0.3906,
+      "step": 29730
+    },
+    {
+      "epoch": 0.7076787584499666,
+      "grad_norm": 6.8520842441531515,
+      "learning_rate": 2e-05,
+      "loss": 0.3236,
+      "step": 29731
+    },
+    {
+      "epoch": 0.7077025611729982,
+      "grad_norm": 1.1953858393916228,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 29732
+    },
+    {
+      "epoch": 0.7077263638960297,
+      "grad_norm": 0.7950311461629863,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 29733
+    },
+    {
+      "epoch": 0.7077501666190612,
+      "grad_norm": 2.1233186807423925,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 29734
+    },
+    {
+      "epoch": 0.7077739693420927,
+      "grad_norm": 5.34774499449671,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 29735
+    },
+    {
+      "epoch": 0.7077977720651243,
+      "grad_norm": 8.74929281665465,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 29736
+    },
+    {
+      "epoch": 0.7078215747881558,
+      "grad_norm": 0.7225756093754525,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 29737
+    },
+    {
+      "epoch": 0.7078453775111873,
+      "grad_norm": 5.355282857464201,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 29738
+    },
+    {
+      "epoch": 0.7078691802342187,
+      "grad_norm": 4.263141863999542,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 29739
+    },
+    {
+      "epoch": 0.7078929829572503,
+      "grad_norm": 0.9305653023592951,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 29740
+    },
+    {
+      "epoch": 0.7079167856802818,
+      "grad_norm": 4.81473811313934,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 29741
+    },
+    {
+      "epoch": 0.7079405884033133,
+      "grad_norm": 0.7875731772879977,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 29742
+    },
+    {
+      "epoch": 0.7079643911263449,
+      "grad_norm": 13.501946204730721,
+      "learning_rate": 2e-05,
+      "loss": 1.5265,
+      "step": 29743
+    },
+    {
+      "epoch": 0.7079881938493764,
+      "grad_norm": 2.4062171329935738,
+      "learning_rate": 2e-05,
+      "loss": 0.2439,
+      "step": 29744
+    },
+    {
+      "epoch": 0.7080119965724079,
+      "grad_norm": 2.155054880268653,
+      "learning_rate": 2e-05,
+      "loss": 0.371,
+      "step": 29745
+    },
+    {
+      "epoch": 0.7080357992954394,
+      "grad_norm": 0.008553428774623301,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 29746
+    },
+    {
+      "epoch": 0.708059602018471,
+      "grad_norm": 0.05383003793724329,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29747
+    },
+    {
+      "epoch": 0.7080834047415024,
+      "grad_norm": 0.05187824451598821,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 29748
+    },
+    {
+      "epoch": 0.7081072074645339,
+      "grad_norm": 8.963365697939238,
+      "learning_rate": 2e-05,
+      "loss": 0.9188,
+      "step": 29749
+    },
+    {
+      "epoch": 0.7081310101875654,
+      "grad_norm": 7.49575477643455,
+      "learning_rate": 2e-05,
+      "loss": 0.2757,
+      "step": 29750
+    },
+    {
+      "epoch": 0.708154812910597,
+      "grad_norm": 0.5879040841213905,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 29751
+    },
+    {
+      "epoch": 0.7081786156336285,
+      "grad_norm": 9.292231885738962,
+      "learning_rate": 2e-05,
+      "loss": 1.1951,
+      "step": 29752
+    },
+    {
+      "epoch": 0.70820241835666,
+      "grad_norm": 2.7630941825588136,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 29753
+    },
+    {
+      "epoch": 0.7082262210796915,
+      "grad_norm": 6.200552252584789,
+      "learning_rate": 2e-05,
+      "loss": 0.5423,
+      "step": 29754
+    },
+    {
+      "epoch": 0.7082500238027231,
+      "grad_norm": 8.247253826280694,
+      "learning_rate": 2e-05,
+      "loss": 0.9987,
+      "step": 29755
+    },
+    {
+      "epoch": 0.7082738265257545,
+      "grad_norm": 4.49590778962952,
+      "learning_rate": 2e-05,
+      "loss": 0.8524,
+      "step": 29756
+    },
+    {
+      "epoch": 0.708297629248786,
+      "grad_norm": 5.972342541204273,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 29757
+    },
+    {
+      "epoch": 0.7083214319718176,
+      "grad_norm": 7.001821369201325,
+      "learning_rate": 2e-05,
+      "loss": 0.328,
+      "step": 29758
+    },
+    {
+      "epoch": 0.7083452346948491,
+      "grad_norm": 3.6971127546374905,
+      "learning_rate": 2e-05,
+      "loss": 0.4783,
+      "step": 29759
+    },
+    {
+      "epoch": 0.7083690374178806,
+      "grad_norm": 7.059250235014688,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 29760
+    },
+    {
+      "epoch": 0.7083928401409121,
+      "grad_norm": 9.576368083933039,
+      "learning_rate": 2e-05,
+      "loss": 0.349,
+      "step": 29761
+    },
+    {
+      "epoch": 0.7084166428639437,
+      "grad_norm": 4.569041669527836,
+      "learning_rate": 2e-05,
+      "loss": 0.9212,
+      "step": 29762
+    },
+    {
+      "epoch": 0.7084404455869752,
+      "grad_norm": 2.774178192909619,
+      "learning_rate": 2e-05,
+      "loss": 0.3856,
+      "step": 29763
+    },
+    {
+      "epoch": 0.7084642483100067,
+      "grad_norm": 4.114788157403486,
+      "learning_rate": 2e-05,
+      "loss": 0.4432,
+      "step": 29764
+    },
+    {
+      "epoch": 0.7084880510330381,
+      "grad_norm": 2.963062442623965,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 29765
+    },
+    {
+      "epoch": 0.7085118537560697,
+      "grad_norm": 3.6983098247095882,
+      "learning_rate": 2e-05,
+      "loss": 0.4444,
+      "step": 29766
+    },
+    {
+      "epoch": 0.7085356564791012,
+      "grad_norm": 9.084278518086828,
+      "learning_rate": 2e-05,
+      "loss": 0.8805,
+      "step": 29767
+    },
+    {
+      "epoch": 0.7085594592021327,
+      "grad_norm": 0.2965704274009548,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 29768
+    },
+    {
+      "epoch": 0.7085832619251642,
+      "grad_norm": 4.416069331292728,
+      "learning_rate": 2e-05,
+      "loss": 1.0254,
+      "step": 29769
+    },
+    {
+      "epoch": 0.7086070646481958,
+      "grad_norm": 0.16943847563325898,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 29770
+    },
+    {
+      "epoch": 0.7086308673712273,
+      "grad_norm": 3.6248935198960077,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 29771
+    },
+    {
+      "epoch": 0.7086546700942588,
+      "grad_norm": 7.722943998684629,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 29772
+    },
+    {
+      "epoch": 0.7086784728172904,
+      "grad_norm": 4.3189855918570625,
+      "learning_rate": 2e-05,
+      "loss": 0.1223,
+      "step": 29773
+    },
+    {
+      "epoch": 0.7087022755403218,
+      "grad_norm": 3.151234149914797,
+      "learning_rate": 2e-05,
+      "loss": 0.1141,
+      "step": 29774
+    },
+    {
+      "epoch": 0.7087260782633533,
+      "grad_norm": 1.046808722916822,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 29775
+    },
+    {
+      "epoch": 0.7087498809863848,
+      "grad_norm": 4.660407260869649,
+      "learning_rate": 2e-05,
+      "loss": 0.5706,
+      "step": 29776
+    },
+    {
+      "epoch": 0.7087736837094164,
+      "grad_norm": 0.011906197613737238,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 29777
+    },
+    {
+      "epoch": 0.7087974864324479,
+      "grad_norm": 5.781588667450127,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 29778
+    },
+    {
+      "epoch": 0.7088212891554794,
+      "grad_norm": 5.7597301616600465,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 29779
+    },
+    {
+      "epoch": 0.7088450918785109,
+      "grad_norm": 0.3192201408921311,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 29780
+    },
+    {
+      "epoch": 0.7088688946015425,
+      "grad_norm": 4.478284951592017,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 29781
+    },
+    {
+      "epoch": 0.7088926973245739,
+      "grad_norm": 5.219627135482329,
+      "learning_rate": 2e-05,
+      "loss": 0.5605,
+      "step": 29782
+    },
+    {
+      "epoch": 0.7089165000476054,
+      "grad_norm": 9.552242722046028,
+      "learning_rate": 2e-05,
+      "loss": 0.3443,
+      "step": 29783
+    },
+    {
+      "epoch": 0.7089403027706369,
+      "grad_norm": 7.808526339328059,
+      "learning_rate": 2e-05,
+      "loss": 1.3931,
+      "step": 29784
+    },
+    {
+      "epoch": 0.7089641054936685,
+      "grad_norm": 3.6192640850427247,
+      "learning_rate": 2e-05,
+      "loss": 0.1142,
+      "step": 29785
+    },
+    {
+      "epoch": 0.7089879082167,
+      "grad_norm": 1.9868320203783558,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 29786
+    },
+    {
+      "epoch": 0.7090117109397315,
+      "grad_norm": 2.609775149194156,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 29787
+    },
+    {
+      "epoch": 0.7090355136627631,
+      "grad_norm": 0.08978762345207869,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 29788
+    },
+    {
+      "epoch": 0.7090593163857946,
+      "grad_norm": 3.293146148499698,
+      "learning_rate": 2e-05,
+      "loss": 0.5516,
+      "step": 29789
+    },
+    {
+      "epoch": 0.709083119108826,
+      "grad_norm": 0.9679177479236648,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 29790
+    },
+    {
+      "epoch": 0.7091069218318575,
+      "grad_norm": 6.643744380937737,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 29791
+    },
+    {
+      "epoch": 0.7091307245548891,
+      "grad_norm": 8.295667604451804,
+      "learning_rate": 2e-05,
+      "loss": 1.9822,
+      "step": 29792
+    },
+    {
+      "epoch": 0.7091545272779206,
+      "grad_norm": 9.603939141913742,
+      "learning_rate": 2e-05,
+      "loss": 0.6229,
+      "step": 29793
+    },
+    {
+      "epoch": 0.7091783300009521,
+      "grad_norm": 2.8065418060359786,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 29794
+    },
+    {
+      "epoch": 0.7092021327239836,
+      "grad_norm": 5.237319872545832,
+      "learning_rate": 2e-05,
+      "loss": 0.849,
+      "step": 29795
+    },
+    {
+      "epoch": 0.7092259354470152,
+      "grad_norm": 0.5177379312975768,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 29796
+    },
+    {
+      "epoch": 0.7092497381700467,
+      "grad_norm": 1.2134980799853958,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 29797
+    },
+    {
+      "epoch": 0.7092735408930781,
+      "grad_norm": 5.850352767128667,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 29798
+    },
+    {
+      "epoch": 0.7092973436161096,
+      "grad_norm": 14.480238663769674,
+      "learning_rate": 2e-05,
+      "loss": 0.8651,
+      "step": 29799
+    },
+    {
+      "epoch": 0.7093211463391412,
+      "grad_norm": 5.419505352503624,
+      "learning_rate": 2e-05,
+      "loss": 0.4851,
+      "step": 29800
+    },
+    {
+      "epoch": 0.7093449490621727,
+      "grad_norm": 15.090070777173654,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 29801
+    },
+    {
+      "epoch": 0.7093687517852042,
+      "grad_norm": 0.5560365871546108,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 29802
+    },
+    {
+      "epoch": 0.7093925545082358,
+      "grad_norm": 5.890531216010557,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 29803
+    },
+    {
+      "epoch": 0.7094163572312673,
+      "grad_norm": 2.3728407270062672,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 29804
+    },
+    {
+      "epoch": 0.7094401599542988,
+      "grad_norm": 2.4832089696408683,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 29805
+    },
+    {
+      "epoch": 0.7094639626773303,
+      "grad_norm": 1.8007200465244302,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 29806
+    },
+    {
+      "epoch": 0.7094877654003618,
+      "grad_norm": 6.849128783795548,
+      "learning_rate": 2e-05,
+      "loss": 0.2922,
+      "step": 29807
+    },
+    {
+      "epoch": 0.7095115681233933,
+      "grad_norm": 2.6104244468922073,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 29808
+    },
+    {
+      "epoch": 0.7095353708464248,
+      "grad_norm": 4.244971034410598,
+      "learning_rate": 2e-05,
+      "loss": 0.1924,
+      "step": 29809
+    },
+    {
+      "epoch": 0.7095591735694563,
+      "grad_norm": 2.5116455393917994,
+      "learning_rate": 2e-05,
+      "loss": 0.2987,
+      "step": 29810
+    },
+    {
+      "epoch": 0.7095829762924879,
+      "grad_norm": 4.910063151615992,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 29811
+    },
+    {
+      "epoch": 0.7096067790155194,
+      "grad_norm": 5.253590176485955,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 29812
+    },
+    {
+      "epoch": 0.7096305817385509,
+      "grad_norm": 9.662097390592086,
+      "learning_rate": 2e-05,
+      "loss": 0.4278,
+      "step": 29813
+    },
+    {
+      "epoch": 0.7096543844615824,
+      "grad_norm": 3.584450165331003,
+      "learning_rate": 2e-05,
+      "loss": 0.0381,
+      "step": 29814
+    },
+    {
+      "epoch": 0.709678187184614,
+      "grad_norm": 3.931091443655171,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 29815
+    },
+    {
+      "epoch": 0.7097019899076454,
+      "grad_norm": 4.486461303466702,
+      "learning_rate": 2e-05,
+      "loss": 0.5964,
+      "step": 29816
+    },
+    {
+      "epoch": 0.7097257926306769,
+      "grad_norm": 8.438118713381298,
+      "learning_rate": 2e-05,
+      "loss": 1.9994,
+      "step": 29817
+    },
+    {
+      "epoch": 0.7097495953537085,
+      "grad_norm": 6.890312051571528,
+      "learning_rate": 2e-05,
+      "loss": 0.4463,
+      "step": 29818
+    },
+    {
+      "epoch": 0.70977339807674,
+      "grad_norm": 4.47026800169693,
+      "learning_rate": 2e-05,
+      "loss": 0.9329,
+      "step": 29819
+    },
+    {
+      "epoch": 0.7097972007997715,
+      "grad_norm": 1.3044095283464248,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 29820
+    },
+    {
+      "epoch": 0.709821003522803,
+      "grad_norm": 5.315129522684834,
+      "learning_rate": 2e-05,
+      "loss": 1.2517,
+      "step": 29821
+    },
+    {
+      "epoch": 0.7098448062458346,
+      "grad_norm": 4.564164200837578,
+      "learning_rate": 2e-05,
+      "loss": 0.7081,
+      "step": 29822
+    },
+    {
+      "epoch": 0.709868608968866,
+      "grad_norm": 3.683872535921698,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 29823
+    },
+    {
+      "epoch": 0.7098924116918975,
+      "grad_norm": 0.7083332744178175,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 29824
+    },
+    {
+      "epoch": 0.709916214414929,
+      "grad_norm": 9.563006557158221,
+      "learning_rate": 2e-05,
+      "loss": 0.9932,
+      "step": 29825
+    },
+    {
+      "epoch": 0.7099400171379606,
+      "grad_norm": 0.13842737867390129,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 29826
+    },
+    {
+      "epoch": 0.7099638198609921,
+      "grad_norm": 3.007516627240286,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 29827
+    },
+    {
+      "epoch": 0.7099876225840236,
+      "grad_norm": 5.054069338214586,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 29828
+    },
+    {
+      "epoch": 0.7100114253070551,
+      "grad_norm": 5.632509228752409,
+      "learning_rate": 2e-05,
+      "loss": 0.8984,
+      "step": 29829
+    },
+    {
+      "epoch": 0.7100352280300867,
+      "grad_norm": 7.73380229495442,
+      "learning_rate": 2e-05,
+      "loss": 0.273,
+      "step": 29830
+    },
+    {
+      "epoch": 0.7100590307531182,
+      "grad_norm": 0.9995870170657557,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 29831
+    },
+    {
+      "epoch": 0.7100828334761496,
+      "grad_norm": 0.25675152773151966,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 29832
+    },
+    {
+      "epoch": 0.7101066361991812,
+      "grad_norm": 5.176828831794544,
+      "learning_rate": 2e-05,
+      "loss": 0.6697,
+      "step": 29833
+    },
+    {
+      "epoch": 0.7101304389222127,
+      "grad_norm": 2.28639580850841,
+      "learning_rate": 2e-05,
+      "loss": 0.3276,
+      "step": 29834
+    },
+    {
+      "epoch": 0.7101542416452442,
+      "grad_norm": 4.918049427503968,
+      "learning_rate": 2e-05,
+      "loss": 0.8086,
+      "step": 29835
+    },
+    {
+      "epoch": 0.7101780443682757,
+      "grad_norm": 0.02097124757601355,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 29836
+    },
+    {
+      "epoch": 0.7102018470913073,
+      "grad_norm": 0.034927451212330936,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 29837
+    },
+    {
+      "epoch": 0.7102256498143388,
+      "grad_norm": 3.9743433634667897,
+      "learning_rate": 2e-05,
+      "loss": 0.6737,
+      "step": 29838
+    },
+    {
+      "epoch": 0.7102494525373703,
+      "grad_norm": 0.6408730808675737,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 29839
+    },
+    {
+      "epoch": 0.7102732552604017,
+      "grad_norm": 0.0917283019382931,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 29840
+    },
+    {
+      "epoch": 0.7102970579834333,
+      "grad_norm": 1.3094915304860868,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 29841
+    },
+    {
+      "epoch": 0.7103208607064648,
+      "grad_norm": 4.073173364191441,
+      "learning_rate": 2e-05,
+      "loss": 0.3695,
+      "step": 29842
+    },
+    {
+      "epoch": 0.7103446634294963,
+      "grad_norm": 0.9723313112458476,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 29843
+    },
+    {
+      "epoch": 0.7103684661525278,
+      "grad_norm": 3.103995673093119,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 29844
+    },
+    {
+      "epoch": 0.7103922688755594,
+      "grad_norm": 5.010628187236691,
+      "learning_rate": 2e-05,
+      "loss": 0.5939,
+      "step": 29845
+    },
+    {
+      "epoch": 0.7104160715985909,
+      "grad_norm": 0.020940056786223796,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 29846
+    },
+    {
+      "epoch": 0.7104398743216224,
+      "grad_norm": 2.568335995659527,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 29847
+    },
+    {
+      "epoch": 0.710463677044654,
+      "grad_norm": 2.245836561569969,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 29848
+    },
+    {
+      "epoch": 0.7104874797676854,
+      "grad_norm": 4.448384933215842,
+      "learning_rate": 2e-05,
+      "loss": 0.7718,
+      "step": 29849
+    },
+    {
+      "epoch": 0.7105112824907169,
+      "grad_norm": 0.23453841852755528,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 29850
+    },
+    {
+      "epoch": 0.7105350852137484,
+      "grad_norm": 0.3456872821680252,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 29851
+    },
+    {
+      "epoch": 0.71055888793678,
+      "grad_norm": 0.3607920683376983,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 29852
+    },
+    {
+      "epoch": 0.7105826906598115,
+      "grad_norm": 3.293663253862972,
+      "learning_rate": 2e-05,
+      "loss": 0.4012,
+      "step": 29853
+    },
+    {
+      "epoch": 0.710606493382843,
+      "grad_norm": 4.946942271971994,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 29854
+    },
+    {
+      "epoch": 0.7106302961058745,
+      "grad_norm": 2.3825769237109578,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 29855
+    },
+    {
+      "epoch": 0.7106540988289061,
+      "grad_norm": 0.01681937445451484,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 29856
+    },
+    {
+      "epoch": 0.7106779015519376,
+      "grad_norm": 2.474792753495738,
+      "learning_rate": 2e-05,
+      "loss": 0.29,
+      "step": 29857
+    },
+    {
+      "epoch": 0.710701704274969,
+      "grad_norm": 0.3151008635936842,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 29858
+    },
+    {
+      "epoch": 0.7107255069980005,
+      "grad_norm": 6.3065785086456,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 29859
+    },
+    {
+      "epoch": 0.7107493097210321,
+      "grad_norm": 8.322996118039024,
+      "learning_rate": 2e-05,
+      "loss": 0.5025,
+      "step": 29860
+    },
+    {
+      "epoch": 0.7107731124440636,
+      "grad_norm": 5.468876541601875,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 29861
+    },
+    {
+      "epoch": 0.7107969151670951,
+      "grad_norm": 3.630977591308948,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 29862
+    },
+    {
+      "epoch": 0.7108207178901267,
+      "grad_norm": 8.66517723259199,
+      "learning_rate": 2e-05,
+      "loss": 1.2564,
+      "step": 29863
+    },
+    {
+      "epoch": 0.7108445206131582,
+      "grad_norm": 10.010242641432708,
+      "learning_rate": 2e-05,
+      "loss": 0.4376,
+      "step": 29864
+    },
+    {
+      "epoch": 0.7108683233361897,
+      "grad_norm": 0.057735058314966586,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29865
+    },
+    {
+      "epoch": 0.7108921260592211,
+      "grad_norm": 5.510200608037551,
+      "learning_rate": 2e-05,
+      "loss": 0.7195,
+      "step": 29866
+    },
+    {
+      "epoch": 0.7109159287822527,
+      "grad_norm": 5.215612342624822,
+      "learning_rate": 2e-05,
+      "loss": 0.3963,
+      "step": 29867
+    },
+    {
+      "epoch": 0.7109397315052842,
+      "grad_norm": 14.689577345830035,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 29868
+    },
+    {
+      "epoch": 0.7109635342283157,
+      "grad_norm": 5.9235328915888275,
+      "learning_rate": 2e-05,
+      "loss": 0.2555,
+      "step": 29869
+    },
+    {
+      "epoch": 0.7109873369513472,
+      "grad_norm": 0.11223130758903538,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 29870
+    },
+    {
+      "epoch": 0.7110111396743788,
+      "grad_norm": 9.503485177493832,
+      "learning_rate": 2e-05,
+      "loss": 0.6965,
+      "step": 29871
+    },
+    {
+      "epoch": 0.7110349423974103,
+      "grad_norm": 2.4773061059240185,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 29872
+    },
+    {
+      "epoch": 0.7110587451204418,
+      "grad_norm": 5.133749635341972,
+      "learning_rate": 2e-05,
+      "loss": 0.2687,
+      "step": 29873
+    },
+    {
+      "epoch": 0.7110825478434732,
+      "grad_norm": 4.513341757122501,
+      "learning_rate": 2e-05,
+      "loss": 0.2077,
+      "step": 29874
+    },
+    {
+      "epoch": 0.7111063505665048,
+      "grad_norm": 2.545227784742413,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 29875
+    },
+    {
+      "epoch": 0.7111301532895363,
+      "grad_norm": 0.16716149399584546,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 29876
+    },
+    {
+      "epoch": 0.7111539560125678,
+      "grad_norm": 0.1297871545242033,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 29877
+    },
+    {
+      "epoch": 0.7111777587355994,
+      "grad_norm": 2.577319358507877,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 29878
+    },
+    {
+      "epoch": 0.7112015614586309,
+      "grad_norm": 0.7132623721448093,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 29879
+    },
+    {
+      "epoch": 0.7112253641816624,
+      "grad_norm": 0.13578886828149617,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 29880
+    },
+    {
+      "epoch": 0.7112491669046939,
+      "grad_norm": 6.723259322367592,
+      "learning_rate": 2e-05,
+      "loss": 0.4173,
+      "step": 29881
+    },
+    {
+      "epoch": 0.7112729696277255,
+      "grad_norm": 4.0201442526227,
+      "learning_rate": 2e-05,
+      "loss": 0.4531,
+      "step": 29882
+    },
+    {
+      "epoch": 0.711296772350757,
+      "grad_norm": 3.3177819375909627,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 29883
+    },
+    {
+      "epoch": 0.7113205750737884,
+      "grad_norm": 3.8352518983071664,
+      "learning_rate": 2e-05,
+      "loss": 0.3672,
+      "step": 29884
+    },
+    {
+      "epoch": 0.7113443777968199,
+      "grad_norm": 6.16094307885034,
+      "learning_rate": 2e-05,
+      "loss": 0.1214,
+      "step": 29885
+    },
+    {
+      "epoch": 0.7113681805198515,
+      "grad_norm": 7.055441111854949,
+      "learning_rate": 2e-05,
+      "loss": 0.1692,
+      "step": 29886
+    },
+    {
+      "epoch": 0.711391983242883,
+      "grad_norm": 9.294022126299195,
+      "learning_rate": 2e-05,
+      "loss": 1.212,
+      "step": 29887
+    },
+    {
+      "epoch": 0.7114157859659145,
+      "grad_norm": 0.23926693260994988,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 29888
+    },
+    {
+      "epoch": 0.711439588688946,
+      "grad_norm": 3.0581424024022663,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 29889
+    },
+    {
+      "epoch": 0.7114633914119776,
+      "grad_norm": 2.8660919693170452,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 29890
+    },
+    {
+      "epoch": 0.711487194135009,
+      "grad_norm": 2.3106294106442014,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 29891
+    },
+    {
+      "epoch": 0.7115109968580405,
+      "grad_norm": 11.412908012592004,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 29892
+    },
+    {
+      "epoch": 0.711534799581072,
+      "grad_norm": 3.239131091726582,
+      "learning_rate": 2e-05,
+      "loss": 0.4757,
+      "step": 29893
+    },
+    {
+      "epoch": 0.7115586023041036,
+      "grad_norm": 0.1746957215398346,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 29894
+    },
+    {
+      "epoch": 0.7115824050271351,
+      "grad_norm": 0.019081066534217966,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 29895
+    },
+    {
+      "epoch": 0.7116062077501666,
+      "grad_norm": 0.1952812330389921,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 29896
+    },
+    {
+      "epoch": 0.7116300104731982,
+      "grad_norm": 3.2489370060399443,
+      "learning_rate": 2e-05,
+      "loss": 0.2837,
+      "step": 29897
+    },
+    {
+      "epoch": 0.7116538131962297,
+      "grad_norm": 7.671364476683124,
+      "learning_rate": 2e-05,
+      "loss": 0.9506,
+      "step": 29898
+    },
+    {
+      "epoch": 0.7116776159192612,
+      "grad_norm": 5.4735440512723,
+      "learning_rate": 2e-05,
+      "loss": 0.8894,
+      "step": 29899
+    },
+    {
+      "epoch": 0.7117014186422926,
+      "grad_norm": 2.4088610355063578,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 29900
+    },
+    {
+      "epoch": 0.7117252213653242,
+      "grad_norm": 5.783851089517464,
+      "learning_rate": 2e-05,
+      "loss": 0.817,
+      "step": 29901
+    },
+    {
+      "epoch": 0.7117490240883557,
+      "grad_norm": 0.23003576541019294,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 29902
+    },
+    {
+      "epoch": 0.7117728268113872,
+      "grad_norm": 3.370174304847356,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 29903
+    },
+    {
+      "epoch": 0.7117966295344187,
+      "grad_norm": 0.1137208112754579,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 29904
+    },
+    {
+      "epoch": 0.7118204322574503,
+      "grad_norm": 0.24997089863584732,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 29905
+    },
+    {
+      "epoch": 0.7118442349804818,
+      "grad_norm": 4.157982529274931,
+      "learning_rate": 2e-05,
+      "loss": 0.7383,
+      "step": 29906
+    },
+    {
+      "epoch": 0.7118680377035133,
+      "grad_norm": 0.07968602562042072,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 29907
+    },
+    {
+      "epoch": 0.7118918404265447,
+      "grad_norm": 1.3731391066064806,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 29908
+    },
+    {
+      "epoch": 0.7119156431495763,
+      "grad_norm": 4.588395128819164,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 29909
+    },
+    {
+      "epoch": 0.7119394458726078,
+      "grad_norm": 8.265761902749778,
+      "learning_rate": 2e-05,
+      "loss": 0.8673,
+      "step": 29910
+    },
+    {
+      "epoch": 0.7119632485956393,
+      "grad_norm": 3.576689962589141,
+      "learning_rate": 2e-05,
+      "loss": 0.5443,
+      "step": 29911
+    },
+    {
+      "epoch": 0.7119870513186709,
+      "grad_norm": 11.330076349761816,
+      "learning_rate": 2e-05,
+      "loss": 0.1754,
+      "step": 29912
+    },
+    {
+      "epoch": 0.7120108540417024,
+      "grad_norm": 6.8630617463033845,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 29913
+    },
+    {
+      "epoch": 0.7120346567647339,
+      "grad_norm": 0.44431542497147886,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 29914
+    },
+    {
+      "epoch": 0.7120584594877654,
+      "grad_norm": 0.12811283499821577,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 29915
+    },
+    {
+      "epoch": 0.712082262210797,
+      "grad_norm": 0.9517662253482744,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 29916
+    },
+    {
+      "epoch": 0.7121060649338284,
+      "grad_norm": 8.834565605806764,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 29917
+    },
+    {
+      "epoch": 0.7121298676568599,
+      "grad_norm": 2.567506804230407,
+      "learning_rate": 2e-05,
+      "loss": 0.4255,
+      "step": 29918
+    },
+    {
+      "epoch": 0.7121536703798914,
+      "grad_norm": 1.080138810267409,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 29919
+    },
+    {
+      "epoch": 0.712177473102923,
+      "grad_norm": 3.1341476586462798,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 29920
+    },
+    {
+      "epoch": 0.7122012758259545,
+      "grad_norm": 2.7633252972628517,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 29921
+    },
+    {
+      "epoch": 0.712225078548986,
+      "grad_norm": 2.7021115697595355,
+      "learning_rate": 2e-05,
+      "loss": 0.4462,
+      "step": 29922
+    },
+    {
+      "epoch": 0.7122488812720175,
+      "grad_norm": 2.9284952993826945,
+      "learning_rate": 2e-05,
+      "loss": 0.3577,
+      "step": 29923
+    },
+    {
+      "epoch": 0.7122726839950491,
+      "grad_norm": 6.797467688775596,
+      "learning_rate": 2e-05,
+      "loss": 1.5586,
+      "step": 29924
+    },
+    {
+      "epoch": 0.7122964867180805,
+      "grad_norm": 3.980561100335665,
+      "learning_rate": 2e-05,
+      "loss": 1.0238,
+      "step": 29925
+    },
+    {
+      "epoch": 0.712320289441112,
+      "grad_norm": 3.083735195490098,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 29926
+    },
+    {
+      "epoch": 0.7123440921641436,
+      "grad_norm": 4.08640932969343,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 29927
+    },
+    {
+      "epoch": 0.7123678948871751,
+      "grad_norm": 5.809973421277558,
+      "learning_rate": 2e-05,
+      "loss": 0.8085,
+      "step": 29928
+    },
+    {
+      "epoch": 0.7123916976102066,
+      "grad_norm": 5.641227557956572,
+      "learning_rate": 2e-05,
+      "loss": 1.0249,
+      "step": 29929
+    },
+    {
+      "epoch": 0.7124155003332381,
+      "grad_norm": 4.595285240157165,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 29930
+    },
+    {
+      "epoch": 0.7124393030562697,
+      "grad_norm": 5.198171925818743,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 29931
+    },
+    {
+      "epoch": 0.7124631057793012,
+      "grad_norm": 0.7650011360286666,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 29932
+    },
+    {
+      "epoch": 0.7124869085023326,
+      "grad_norm": 0.5337856684147452,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 29933
+    },
+    {
+      "epoch": 0.7125107112253641,
+      "grad_norm": 2.3025740756430717,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 29934
+    },
+    {
+      "epoch": 0.7125345139483957,
+      "grad_norm": 4.246233614019426,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 29935
+    },
+    {
+      "epoch": 0.7125583166714272,
+      "grad_norm": 5.397778827795114,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 29936
+    },
+    {
+      "epoch": 0.7125821193944587,
+      "grad_norm": 4.574096506793594,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 29937
+    },
+    {
+      "epoch": 0.7126059221174902,
+      "grad_norm": 1.4619997114275487,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 29938
+    },
+    {
+      "epoch": 0.7126297248405218,
+      "grad_norm": 0.09636626474650116,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 29939
+    },
+    {
+      "epoch": 0.7126535275635533,
+      "grad_norm": 5.432441064185106,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 29940
+    },
+    {
+      "epoch": 0.7126773302865848,
+      "grad_norm": 30.754724779275953,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 29941
+    },
+    {
+      "epoch": 0.7127011330096163,
+      "grad_norm": 11.322301083309839,
+      "learning_rate": 2e-05,
+      "loss": 1.6525,
+      "step": 29942
+    },
+    {
+      "epoch": 0.7127249357326478,
+      "grad_norm": 2.3047037069892338,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 29943
+    },
+    {
+      "epoch": 0.7127487384556793,
+      "grad_norm": 14.016957720404434,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 29944
+    },
+    {
+      "epoch": 0.7127725411787108,
+      "grad_norm": 6.409553466219481,
+      "learning_rate": 2e-05,
+      "loss": 0.9911,
+      "step": 29945
+    },
+    {
+      "epoch": 0.7127963439017424,
+      "grad_norm": 0.23683633014724492,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 29946
+    },
+    {
+      "epoch": 0.7128201466247739,
+      "grad_norm": 2.7245350325029025,
+      "learning_rate": 2e-05,
+      "loss": 0.1837,
+      "step": 29947
+    },
+    {
+      "epoch": 0.7128439493478054,
+      "grad_norm": 0.7682654586311138,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 29948
+    },
+    {
+      "epoch": 0.7128677520708369,
+      "grad_norm": 3.1016755192060197,
+      "learning_rate": 2e-05,
+      "loss": 0.3943,
+      "step": 29949
+    },
+    {
+      "epoch": 0.7128915547938685,
+      "grad_norm": 0.3945897301345948,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 29950
+    },
+    {
+      "epoch": 0.7129153575168999,
+      "grad_norm": 4.106771253914104,
+      "learning_rate": 2e-05,
+      "loss": 0.6355,
+      "step": 29951
+    },
+    {
+      "epoch": 0.7129391602399314,
+      "grad_norm": 4.412309733417695,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 29952
+    },
+    {
+      "epoch": 0.7129629629629629,
+      "grad_norm": 1.1358392684113041,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 29953
+    },
+    {
+      "epoch": 0.7129867656859945,
+      "grad_norm": 2.453544170416622,
+      "learning_rate": 2e-05,
+      "loss": 0.2484,
+      "step": 29954
+    },
+    {
+      "epoch": 0.713010568409026,
+      "grad_norm": 1.9269580996679734,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 29955
+    },
+    {
+      "epoch": 0.7130343711320575,
+      "grad_norm": 3.5307547633543326,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 29956
+    },
+    {
+      "epoch": 0.7130581738550891,
+      "grad_norm": 4.4638354782049365,
+      "learning_rate": 2e-05,
+      "loss": 0.1279,
+      "step": 29957
+    },
+    {
+      "epoch": 0.7130819765781206,
+      "grad_norm": 8.090665848037489,
+      "learning_rate": 2e-05,
+      "loss": 0.1701,
+      "step": 29958
+    },
+    {
+      "epoch": 0.713105779301152,
+      "grad_norm": 3.5620756387138544,
+      "learning_rate": 2e-05,
+      "loss": 0.856,
+      "step": 29959
+    },
+    {
+      "epoch": 0.7131295820241835,
+      "grad_norm": 1.3787203120230351,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 29960
+    },
+    {
+      "epoch": 0.7131533847472151,
+      "grad_norm": 10.602039205989287,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 29961
+    },
+    {
+      "epoch": 0.7131771874702466,
+      "grad_norm": 31.823484592253255,
+      "learning_rate": 2e-05,
+      "loss": 0.4194,
+      "step": 29962
+    },
+    {
+      "epoch": 0.7132009901932781,
+      "grad_norm": 1.6092040342175382,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 29963
+    },
+    {
+      "epoch": 0.7132247929163096,
+      "grad_norm": 4.516446064849733,
+      "learning_rate": 2e-05,
+      "loss": 0.7632,
+      "step": 29964
+    },
+    {
+      "epoch": 0.7132485956393412,
+      "grad_norm": 2.4472925325332606,
+      "learning_rate": 2e-05,
+      "loss": 0.3662,
+      "step": 29965
+    },
+    {
+      "epoch": 0.7132723983623727,
+      "grad_norm": 3.77585349645102,
+      "learning_rate": 2e-05,
+      "loss": 0.8227,
+      "step": 29966
+    },
+    {
+      "epoch": 0.7132962010854041,
+      "grad_norm": 0.25725329441536354,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 29967
+    },
+    {
+      "epoch": 0.7133200038084356,
+      "grad_norm": 3.134256353224607,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 29968
+    },
+    {
+      "epoch": 0.7133438065314672,
+      "grad_norm": 5.0947648501051255,
+      "learning_rate": 2e-05,
+      "loss": 0.6996,
+      "step": 29969
+    },
+    {
+      "epoch": 0.7133676092544987,
+      "grad_norm": 6.951388479857678,
+      "learning_rate": 2e-05,
+      "loss": 0.3506,
+      "step": 29970
+    },
+    {
+      "epoch": 0.7133914119775302,
+      "grad_norm": 3.11850860949209,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 29971
+    },
+    {
+      "epoch": 0.7134152147005618,
+      "grad_norm": 1.1886691193532648,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 29972
+    },
+    {
+      "epoch": 0.7134390174235933,
+      "grad_norm": 1.6295068229855565,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 29973
+    },
+    {
+      "epoch": 0.7134628201466248,
+      "grad_norm": 11.681530042093327,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 29974
+    },
+    {
+      "epoch": 0.7134866228696563,
+      "grad_norm": 7.527727150344796,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 29975
+    },
+    {
+      "epoch": 0.7135104255926878,
+      "grad_norm": 0.6054051956745689,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 29976
+    },
+    {
+      "epoch": 0.7135342283157193,
+      "grad_norm": 11.103798039899742,
+      "learning_rate": 2e-05,
+      "loss": 0.617,
+      "step": 29977
+    },
+    {
+      "epoch": 0.7135580310387508,
+      "grad_norm": 0.06596530665665218,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 29978
+    },
+    {
+      "epoch": 0.7135818337617823,
+      "grad_norm": 7.087525163042371,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 29979
+    },
+    {
+      "epoch": 0.7136056364848139,
+      "grad_norm": 3.6382140756203696,
+      "learning_rate": 2e-05,
+      "loss": 0.5702,
+      "step": 29980
+    },
+    {
+      "epoch": 0.7136294392078454,
+      "grad_norm": 0.2055977103210849,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 29981
+    },
+    {
+      "epoch": 0.7136532419308769,
+      "grad_norm": 8.519863164478936,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 29982
+    },
+    {
+      "epoch": 0.7136770446539084,
+      "grad_norm": 5.05149352564965,
+      "learning_rate": 2e-05,
+      "loss": 0.7597,
+      "step": 29983
+    },
+    {
+      "epoch": 0.71370084737694,
+      "grad_norm": 2.204096104587246,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 29984
+    },
+    {
+      "epoch": 0.7137246500999714,
+      "grad_norm": 8.680007313597276,
+      "learning_rate": 2e-05,
+      "loss": 0.755,
+      "step": 29985
+    },
+    {
+      "epoch": 0.7137484528230029,
+      "grad_norm": 1.0803857930776135,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 29986
+    },
+    {
+      "epoch": 0.7137722555460345,
+      "grad_norm": 1.9610953495593026,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 29987
+    },
+    {
+      "epoch": 0.713796058269066,
+      "grad_norm": 5.347718002466287,
+      "learning_rate": 2e-05,
+      "loss": 1.3752,
+      "step": 29988
+    },
+    {
+      "epoch": 0.7138198609920975,
+      "grad_norm": 10.150257205905556,
+      "learning_rate": 2e-05,
+      "loss": 1.0755,
+      "step": 29989
+    },
+    {
+      "epoch": 0.713843663715129,
+      "grad_norm": 4.8302098027666185,
+      "learning_rate": 2e-05,
+      "loss": 0.8468,
+      "step": 29990
+    },
+    {
+      "epoch": 0.7138674664381606,
+      "grad_norm": 2.030820105601148,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 29991
+    },
+    {
+      "epoch": 0.713891269161192,
+      "grad_norm": 11.85617952774399,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 29992
+    },
+    {
+      "epoch": 0.7139150718842235,
+      "grad_norm": 0.16553087855876025,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 29993
+    },
+    {
+      "epoch": 0.713938874607255,
+      "grad_norm": 3.756455369835502,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 29994
+    },
+    {
+      "epoch": 0.7139626773302866,
+      "grad_norm": 4.317260104603092,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 29995
+    },
+    {
+      "epoch": 0.7139864800533181,
+      "grad_norm": 9.590203248511079,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 29996
+    },
+    {
+      "epoch": 0.7140102827763496,
+      "grad_norm": 6.683284512237941,
+      "learning_rate": 2e-05,
+      "loss": 0.1277,
+      "step": 29997
+    },
+    {
+      "epoch": 0.7140340854993811,
+      "grad_norm": 3.529037695961665,
+      "learning_rate": 2e-05,
+      "loss": 0.7269,
+      "step": 29998
+    },
+    {
+      "epoch": 0.7140578882224127,
+      "grad_norm": 1.2930960425459403,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 29999
+    },
+    {
+      "epoch": 0.7140816909454442,
+      "grad_norm": 6.409116912797107,
+      "learning_rate": 2e-05,
+      "loss": 0.882,
+      "step": 30000
+    },
+    {
+      "epoch": 0.7141054936684756,
+      "grad_norm": 3.6841563267453292,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 30001
+    },
+    {
+      "epoch": 0.7141292963915072,
+      "grad_norm": 6.974827959193061,
+      "learning_rate": 2e-05,
+      "loss": 0.1289,
+      "step": 30002
+    },
+    {
+      "epoch": 0.7141530991145387,
+      "grad_norm": 3.295511598446506,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 30003
+    },
+    {
+      "epoch": 0.7141769018375702,
+      "grad_norm": 0.060692978936180364,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 30004
+    },
+    {
+      "epoch": 0.7142007045606017,
+      "grad_norm": 5.428154193953391,
+      "learning_rate": 2e-05,
+      "loss": 0.3919,
+      "step": 30005
+    },
+    {
+      "epoch": 0.7142245072836333,
+      "grad_norm": 5.722798471976925,
+      "learning_rate": 2e-05,
+      "loss": 1.1931,
+      "step": 30006
+    },
+    {
+      "epoch": 0.7142483100066648,
+      "grad_norm": 10.585199285936131,
+      "learning_rate": 2e-05,
+      "loss": 1.3276,
+      "step": 30007
+    },
+    {
+      "epoch": 0.7142721127296963,
+      "grad_norm": 3.282843829860042,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 30008
+    },
+    {
+      "epoch": 0.7142959154527277,
+      "grad_norm": 4.161857569901915,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 30009
+    },
+    {
+      "epoch": 0.7143197181757593,
+      "grad_norm": 2.4805809174961206,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 30010
+    },
+    {
+      "epoch": 0.7143435208987908,
+      "grad_norm": 5.712112076786816,
+      "learning_rate": 2e-05,
+      "loss": 1.7134,
+      "step": 30011
+    },
+    {
+      "epoch": 0.7143673236218223,
+      "grad_norm": 8.275486782567684,
+      "learning_rate": 2e-05,
+      "loss": 0.5562,
+      "step": 30012
+    },
+    {
+      "epoch": 0.7143911263448538,
+      "grad_norm": 1.951897213728986,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 30013
+    },
+    {
+      "epoch": 0.7144149290678854,
+      "grad_norm": 2.1024910664788754,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 30014
+    },
+    {
+      "epoch": 0.7144387317909169,
+      "grad_norm": 4.540908845519757,
+      "learning_rate": 2e-05,
+      "loss": 1.1614,
+      "step": 30015
+    },
+    {
+      "epoch": 0.7144625345139484,
+      "grad_norm": 0.7345948872144038,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 30016
+    },
+    {
+      "epoch": 0.71448633723698,
+      "grad_norm": 3.5418148587106093,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 30017
+    },
+    {
+      "epoch": 0.7145101399600114,
+      "grad_norm": 3.6625535468505896,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 30018
+    },
+    {
+      "epoch": 0.7145339426830429,
+      "grad_norm": 9.794637857276246,
+      "learning_rate": 2e-05,
+      "loss": 1.2105,
+      "step": 30019
+    },
+    {
+      "epoch": 0.7145577454060744,
+      "grad_norm": 0.033574626346149876,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 30020
+    },
+    {
+      "epoch": 0.714581548129106,
+      "grad_norm": 5.307450059138964,
+      "learning_rate": 2e-05,
+      "loss": 0.8829,
+      "step": 30021
+    },
+    {
+      "epoch": 0.7146053508521375,
+      "grad_norm": 2.821658436903219,
+      "learning_rate": 2e-05,
+      "loss": 0.2125,
+      "step": 30022
+    },
+    {
+      "epoch": 0.714629153575169,
+      "grad_norm": 1.120119563534039,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 30023
+    },
+    {
+      "epoch": 0.7146529562982005,
+      "grad_norm": 5.590097274207272,
+      "learning_rate": 2e-05,
+      "loss": 0.7167,
+      "step": 30024
+    },
+    {
+      "epoch": 0.7146767590212321,
+      "grad_norm": 3.8179430303582613,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 30025
+    },
+    {
+      "epoch": 0.7147005617442636,
+      "grad_norm": 3.123216957645012,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 30026
+    },
+    {
+      "epoch": 0.714724364467295,
+      "grad_norm": 0.12234087328879693,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 30027
+    },
+    {
+      "epoch": 0.7147481671903265,
+      "grad_norm": 0.05557966850868057,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 30028
+    },
+    {
+      "epoch": 0.7147719699133581,
+      "grad_norm": 0.942772780017168,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 30029
+    },
+    {
+      "epoch": 0.7147957726363896,
+      "grad_norm": 1.0656730235208527,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 30030
+    },
+    {
+      "epoch": 0.7148195753594211,
+      "grad_norm": 2.562587934100297,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 30031
+    },
+    {
+      "epoch": 0.7148433780824527,
+      "grad_norm": 9.211967147128384,
+      "learning_rate": 2e-05,
+      "loss": 1.0153,
+      "step": 30032
+    },
+    {
+      "epoch": 0.7148671808054842,
+      "grad_norm": 5.10242186415679,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 30033
+    },
+    {
+      "epoch": 0.7148909835285157,
+      "grad_norm": 11.919206570343686,
+      "learning_rate": 2e-05,
+      "loss": 0.9796,
+      "step": 30034
+    },
+    {
+      "epoch": 0.7149147862515471,
+      "grad_norm": 11.687742555963224,
+      "learning_rate": 2e-05,
+      "loss": 0.3298,
+      "step": 30035
+    },
+    {
+      "epoch": 0.7149385889745787,
+      "grad_norm": 6.1883566014101925,
+      "learning_rate": 2e-05,
+      "loss": 0.6228,
+      "step": 30036
+    },
+    {
+      "epoch": 0.7149623916976102,
+      "grad_norm": 3.23189426812146,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 30037
+    },
+    {
+      "epoch": 0.7149861944206417,
+      "grad_norm": 3.9048927917056626,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 30038
+    },
+    {
+      "epoch": 0.7150099971436732,
+      "grad_norm": 0.8661911815791531,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 30039
+    },
+    {
+      "epoch": 0.7150337998667048,
+      "grad_norm": 6.711227498000456,
+      "learning_rate": 2e-05,
+      "loss": 1.371,
+      "step": 30040
+    },
+    {
+      "epoch": 0.7150576025897363,
+      "grad_norm": 11.967516457491303,
+      "learning_rate": 2e-05,
+      "loss": 1.6154,
+      "step": 30041
+    },
+    {
+      "epoch": 0.7150814053127678,
+      "grad_norm": 3.4519207041816538,
+      "learning_rate": 2e-05,
+      "loss": 0.3832,
+      "step": 30042
+    },
+    {
+      "epoch": 0.7151052080357992,
+      "grad_norm": 0.3539012148206513,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 30043
+    },
+    {
+      "epoch": 0.7151290107588308,
+      "grad_norm": 3.6167328361308844,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 30044
+    },
+    {
+      "epoch": 0.7151528134818623,
+      "grad_norm": 1.2912410215395604,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 30045
+    },
+    {
+      "epoch": 0.7151766162048938,
+      "grad_norm": 0.5814356735156516,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 30046
+    },
+    {
+      "epoch": 0.7152004189279254,
+      "grad_norm": 8.229214568733477,
+      "learning_rate": 2e-05,
+      "loss": 0.5233,
+      "step": 30047
+    },
+    {
+      "epoch": 0.7152242216509569,
+      "grad_norm": 4.321100857705044,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 30048
+    },
+    {
+      "epoch": 0.7152480243739884,
+      "grad_norm": 0.06677435809493323,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 30049
+    },
+    {
+      "epoch": 0.7152718270970199,
+      "grad_norm": 4.218996014167644,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 30050
+    },
+    {
+      "epoch": 0.7152956298200515,
+      "grad_norm": 3.9894683725093403,
+      "learning_rate": 2e-05,
+      "loss": 0.8988,
+      "step": 30051
+    },
+    {
+      "epoch": 0.715319432543083,
+      "grad_norm": 6.008546268605964,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 30052
+    },
+    {
+      "epoch": 0.7153432352661144,
+      "grad_norm": 3.229862659552897,
+      "learning_rate": 2e-05,
+      "loss": 0.2036,
+      "step": 30053
+    },
+    {
+      "epoch": 0.7153670379891459,
+      "grad_norm": 10.107381448560352,
+      "learning_rate": 2e-05,
+      "loss": 0.2506,
+      "step": 30054
+    },
+    {
+      "epoch": 0.7153908407121775,
+      "grad_norm": 3.448856575806534,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 30055
+    },
+    {
+      "epoch": 0.715414643435209,
+      "grad_norm": 10.207562605862432,
+      "learning_rate": 2e-05,
+      "loss": 0.2965,
+      "step": 30056
+    },
+    {
+      "epoch": 0.7154384461582405,
+      "grad_norm": 11.534349793725726,
+      "learning_rate": 2e-05,
+      "loss": 0.5762,
+      "step": 30057
+    },
+    {
+      "epoch": 0.715462248881272,
+      "grad_norm": 2.2397425255951364,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 30058
+    },
+    {
+      "epoch": 0.7154860516043036,
+      "grad_norm": 5.038508347951776,
+      "learning_rate": 2e-05,
+      "loss": 0.8711,
+      "step": 30059
+    },
+    {
+      "epoch": 0.715509854327335,
+      "grad_norm": 2.591000210084634,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 30060
+    },
+    {
+      "epoch": 0.7155336570503665,
+      "grad_norm": 9.460738733867334,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 30061
+    },
+    {
+      "epoch": 0.7155574597733981,
+      "grad_norm": 1.8120071043103334,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 30062
+    },
+    {
+      "epoch": 0.7155812624964296,
+      "grad_norm": 1.7050691138906529,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 30063
+    },
+    {
+      "epoch": 0.7156050652194611,
+      "grad_norm": 3.3137683816134538,
+      "learning_rate": 2e-05,
+      "loss": 0.4089,
+      "step": 30064
+    },
+    {
+      "epoch": 0.7156288679424926,
+      "grad_norm": 3.257004503035235,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 30065
+    },
+    {
+      "epoch": 0.7156526706655242,
+      "grad_norm": 9.885789225258517,
+      "learning_rate": 2e-05,
+      "loss": 0.3655,
+      "step": 30066
+    },
+    {
+      "epoch": 0.7156764733885557,
+      "grad_norm": 5.777047556299402,
+      "learning_rate": 2e-05,
+      "loss": 0.6607,
+      "step": 30067
+    },
+    {
+      "epoch": 0.7157002761115872,
+      "grad_norm": 6.490990255330234,
+      "learning_rate": 2e-05,
+      "loss": 1.0365,
+      "step": 30068
+    },
+    {
+      "epoch": 0.7157240788346186,
+      "grad_norm": 4.824902823042773,
+      "learning_rate": 2e-05,
+      "loss": 0.7655,
+      "step": 30069
+    },
+    {
+      "epoch": 0.7157478815576502,
+      "grad_norm": 9.41122647366668,
+      "learning_rate": 2e-05,
+      "loss": 1.2654,
+      "step": 30070
+    },
+    {
+      "epoch": 0.7157716842806817,
+      "grad_norm": 8.630532771571861,
+      "learning_rate": 2e-05,
+      "loss": 1.3474,
+      "step": 30071
+    },
+    {
+      "epoch": 0.7157954870037132,
+      "grad_norm": 6.125868182589039,
+      "learning_rate": 2e-05,
+      "loss": 0.7023,
+      "step": 30072
+    },
+    {
+      "epoch": 0.7158192897267447,
+      "grad_norm": 1.9903004930540524,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 30073
+    },
+    {
+      "epoch": 0.7158430924497763,
+      "grad_norm": 4.99943332325589,
+      "learning_rate": 2e-05,
+      "loss": 1.1456,
+      "step": 30074
+    },
+    {
+      "epoch": 0.7158668951728078,
+      "grad_norm": 0.27779779970588664,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 30075
+    },
+    {
+      "epoch": 0.7158906978958393,
+      "grad_norm": 3.205477874941955,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 30076
+    },
+    {
+      "epoch": 0.7159145006188709,
+      "grad_norm": 0.6532470042646925,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 30077
+    },
+    {
+      "epoch": 0.7159383033419023,
+      "grad_norm": 2.84729673824186,
+      "learning_rate": 2e-05,
+      "loss": 0.4682,
+      "step": 30078
+    },
+    {
+      "epoch": 0.7159621060649338,
+      "grad_norm": 2.688383688284049,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 30079
+    },
+    {
+      "epoch": 0.7159859087879653,
+      "grad_norm": 3.0338125760719366,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 30080
+    },
+    {
+      "epoch": 0.7160097115109969,
+      "grad_norm": 7.919699594122242,
+      "learning_rate": 2e-05,
+      "loss": 0.7023,
+      "step": 30081
+    },
+    {
+      "epoch": 0.7160335142340284,
+      "grad_norm": 3.90796649251995,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 30082
+    },
+    {
+      "epoch": 0.7160573169570599,
+      "grad_norm": 3.0989678179234326,
+      "learning_rate": 2e-05,
+      "loss": 0.6382,
+      "step": 30083
+    },
+    {
+      "epoch": 0.7160811196800914,
+      "grad_norm": 0.5273766353436095,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 30084
+    },
+    {
+      "epoch": 0.716104922403123,
+      "grad_norm": 8.330398571488345,
+      "learning_rate": 2e-05,
+      "loss": 0.3386,
+      "step": 30085
+    },
+    {
+      "epoch": 0.7161287251261544,
+      "grad_norm": 0.2800732692169007,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 30086
+    },
+    {
+      "epoch": 0.7161525278491859,
+      "grad_norm": 6.268390114158684,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 30087
+    },
+    {
+      "epoch": 0.7161763305722174,
+      "grad_norm": 7.892758178213764,
+      "learning_rate": 2e-05,
+      "loss": 1.0344,
+      "step": 30088
+    },
+    {
+      "epoch": 0.716200133295249,
+      "grad_norm": 3.460622528811878,
+      "learning_rate": 2e-05,
+      "loss": 0.542,
+      "step": 30089
+    },
+    {
+      "epoch": 0.7162239360182805,
+      "grad_norm": 1.5927345290937058,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 30090
+    },
+    {
+      "epoch": 0.716247738741312,
+      "grad_norm": 0.14081155711665405,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 30091
+    },
+    {
+      "epoch": 0.7162715414643436,
+      "grad_norm": 3.7558584991338466,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 30092
+    },
+    {
+      "epoch": 0.7162953441873751,
+      "grad_norm": 5.137063189395765,
+      "learning_rate": 2e-05,
+      "loss": 0.5394,
+      "step": 30093
+    },
+    {
+      "epoch": 0.7163191469104065,
+      "grad_norm": 4.948012897186713,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 30094
+    },
+    {
+      "epoch": 0.716342949633438,
+      "grad_norm": 0.36906080414991893,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 30095
+    },
+    {
+      "epoch": 0.7163667523564696,
+      "grad_norm": 3.262765178367194,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 30096
+    },
+    {
+      "epoch": 0.7163905550795011,
+      "grad_norm": 4.487412389913733,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 30097
+    },
+    {
+      "epoch": 0.7164143578025326,
+      "grad_norm": 5.4240762643705205,
+      "learning_rate": 2e-05,
+      "loss": 1.9054,
+      "step": 30098
+    },
+    {
+      "epoch": 0.7164381605255641,
+      "grad_norm": 0.10557277768575707,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 30099
+    },
+    {
+      "epoch": 0.7164619632485957,
+      "grad_norm": 11.59938026744842,
+      "learning_rate": 2e-05,
+      "loss": 0.8628,
+      "step": 30100
+    },
+    {
+      "epoch": 0.7164857659716272,
+      "grad_norm": 0.3990650856983406,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 30101
+    },
+    {
+      "epoch": 0.7165095686946586,
+      "grad_norm": 0.7283693537294672,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 30102
+    },
+    {
+      "epoch": 0.7165333714176901,
+      "grad_norm": 0.3749424377968133,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 30103
+    },
+    {
+      "epoch": 0.7165571741407217,
+      "grad_norm": 4.4032291186446475,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 30104
+    },
+    {
+      "epoch": 0.7165809768637532,
+      "grad_norm": 2.4167550576591426,
+      "learning_rate": 2e-05,
+      "loss": 0.2095,
+      "step": 30105
+    },
+    {
+      "epoch": 0.7166047795867847,
+      "grad_norm": 0.7551833147570403,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 30106
+    },
+    {
+      "epoch": 0.7166285823098163,
+      "grad_norm": 5.326446783575338,
+      "learning_rate": 2e-05,
+      "loss": 0.8929,
+      "step": 30107
+    },
+    {
+      "epoch": 0.7166523850328478,
+      "grad_norm": 2.5955419650307796,
+      "learning_rate": 2e-05,
+      "loss": 0.3634,
+      "step": 30108
+    },
+    {
+      "epoch": 0.7166761877558793,
+      "grad_norm": 3.7105417571444486,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 30109
+    },
+    {
+      "epoch": 0.7166999904789108,
+      "grad_norm": 6.357730272398202,
+      "learning_rate": 2e-05,
+      "loss": 0.2443,
+      "step": 30110
+    },
+    {
+      "epoch": 0.7167237932019423,
+      "grad_norm": 0.14154873137076965,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 30111
+    },
+    {
+      "epoch": 0.7167475959249738,
+      "grad_norm": 0.15034747338733823,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 30112
+    },
+    {
+      "epoch": 0.7167713986480053,
+      "grad_norm": 1.3046380963924518,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 30113
+    },
+    {
+      "epoch": 0.7167952013710368,
+      "grad_norm": 0.19103031211429805,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 30114
+    },
+    {
+      "epoch": 0.7168190040940684,
+      "grad_norm": 2.287101877904586,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 30115
+    },
+    {
+      "epoch": 0.7168428068170999,
+      "grad_norm": 5.161361696639618,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 30116
+    },
+    {
+      "epoch": 0.7168666095401314,
+      "grad_norm": 36.505572690243035,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 30117
+    },
+    {
+      "epoch": 0.7168904122631629,
+      "grad_norm": 3.4379676993983685,
+      "learning_rate": 2e-05,
+      "loss": 0.6088,
+      "step": 30118
+    },
+    {
+      "epoch": 0.7169142149861945,
+      "grad_norm": 5.174063718029049,
+      "learning_rate": 2e-05,
+      "loss": 1.0342,
+      "step": 30119
+    },
+    {
+      "epoch": 0.7169380177092259,
+      "grad_norm": 7.350347135628027,
+      "learning_rate": 2e-05,
+      "loss": 0.845,
+      "step": 30120
+    },
+    {
+      "epoch": 0.7169618204322574,
+      "grad_norm": 7.670991067285769,
+      "learning_rate": 2e-05,
+      "loss": 1.5767,
+      "step": 30121
+    },
+    {
+      "epoch": 0.716985623155289,
+      "grad_norm": 0.6193888082917984,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 30122
+    },
+    {
+      "epoch": 0.7170094258783205,
+      "grad_norm": 0.4858387998113092,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 30123
+    },
+    {
+      "epoch": 0.717033228601352,
+      "grad_norm": 3.71869277356661,
+      "learning_rate": 2e-05,
+      "loss": 0.6771,
+      "step": 30124
+    },
+    {
+      "epoch": 0.7170570313243835,
+      "grad_norm": 3.3602016517597972,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 30125
+    },
+    {
+      "epoch": 0.7170808340474151,
+      "grad_norm": 2.9122854429474665,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 30126
+    },
+    {
+      "epoch": 0.7171046367704466,
+      "grad_norm": 11.359883337650368,
+      "learning_rate": 2e-05,
+      "loss": 0.9876,
+      "step": 30127
+    },
+    {
+      "epoch": 0.717128439493478,
+      "grad_norm": 1.4744502072647456,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 30128
+    },
+    {
+      "epoch": 0.7171522422165095,
+      "grad_norm": 2.094887572169576,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 30129
+    },
+    {
+      "epoch": 0.7171760449395411,
+      "grad_norm": 2.9865478587677856,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 30130
+    },
+    {
+      "epoch": 0.7171998476625726,
+      "grad_norm": 0.41478997619569785,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 30131
+    },
+    {
+      "epoch": 0.7172236503856041,
+      "grad_norm": 0.3563822438872269,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 30132
+    },
+    {
+      "epoch": 0.7172474531086356,
+      "grad_norm": 1.2656874084124303,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 30133
+    },
+    {
+      "epoch": 0.7172712558316672,
+      "grad_norm": 3.9038565162181045,
+      "learning_rate": 2e-05,
+      "loss": 0.5887,
+      "step": 30134
+    },
+    {
+      "epoch": 0.7172950585546987,
+      "grad_norm": 4.855011551968903,
+      "learning_rate": 2e-05,
+      "loss": 0.9903,
+      "step": 30135
+    },
+    {
+      "epoch": 0.7173188612777301,
+      "grad_norm": 1.511822108389325,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 30136
+    },
+    {
+      "epoch": 0.7173426640007616,
+      "grad_norm": 5.125300009683157,
+      "learning_rate": 2e-05,
+      "loss": 0.3446,
+      "step": 30137
+    },
+    {
+      "epoch": 0.7173664667237932,
+      "grad_norm": 4.046200696685891,
+      "learning_rate": 2e-05,
+      "loss": 0.6529,
+      "step": 30138
+    },
+    {
+      "epoch": 0.7173902694468247,
+      "grad_norm": 4.219616174807226,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 30139
+    },
+    {
+      "epoch": 0.7174140721698562,
+      "grad_norm": 3.7144574679636264,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 30140
+    },
+    {
+      "epoch": 0.7174378748928878,
+      "grad_norm": 5.525582683210031,
+      "learning_rate": 2e-05,
+      "loss": 0.6214,
+      "step": 30141
+    },
+    {
+      "epoch": 0.7174616776159193,
+      "grad_norm": 12.911300477051347,
+      "learning_rate": 2e-05,
+      "loss": 1.1344,
+      "step": 30142
+    },
+    {
+      "epoch": 0.7174854803389508,
+      "grad_norm": 2.847947215810033,
+      "learning_rate": 2e-05,
+      "loss": 0.5674,
+      "step": 30143
+    },
+    {
+      "epoch": 0.7175092830619823,
+      "grad_norm": 3.0078835419265326,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 30144
+    },
+    {
+      "epoch": 0.7175330857850138,
+      "grad_norm": 1.7336526524141371,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 30145
+    },
+    {
+      "epoch": 0.7175568885080453,
+      "grad_norm": 3.544751768493086,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 30146
+    },
+    {
+      "epoch": 0.7175806912310768,
+      "grad_norm": 2.901027888734202,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 30147
+    },
+    {
+      "epoch": 0.7176044939541083,
+      "grad_norm": 1.1326346583996023,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 30148
+    },
+    {
+      "epoch": 0.7176282966771399,
+      "grad_norm": 0.5288944027255907,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 30149
+    },
+    {
+      "epoch": 0.7176520994001714,
+      "grad_norm": 0.07161974209255724,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 30150
+    },
+    {
+      "epoch": 0.7176759021232029,
+      "grad_norm": 0.8246528484080361,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 30151
+    },
+    {
+      "epoch": 0.7176997048462344,
+      "grad_norm": 0.03035904316813548,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 30152
+    },
+    {
+      "epoch": 0.717723507569266,
+      "grad_norm": 2.5241711576502013,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 30153
+    },
+    {
+      "epoch": 0.7177473102922974,
+      "grad_norm": 2.2174189246904,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 30154
+    },
+    {
+      "epoch": 0.7177711130153289,
+      "grad_norm": 6.537114562716628,
+      "learning_rate": 2e-05,
+      "loss": 1.0726,
+      "step": 30155
+    },
+    {
+      "epoch": 0.7177949157383605,
+      "grad_norm": 4.014900360524058,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 30156
+    },
+    {
+      "epoch": 0.717818718461392,
+      "grad_norm": 8.345761327664574,
+      "learning_rate": 2e-05,
+      "loss": 0.1473,
+      "step": 30157
+    },
+    {
+      "epoch": 0.7178425211844235,
+      "grad_norm": 0.07186640722699678,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30158
+    },
+    {
+      "epoch": 0.717866323907455,
+      "grad_norm": 3.880264282500556,
+      "learning_rate": 2e-05,
+      "loss": 0.6409,
+      "step": 30159
+    },
+    {
+      "epoch": 0.7178901266304866,
+      "grad_norm": 2.9895155539729865,
+      "learning_rate": 2e-05,
+      "loss": 0.1681,
+      "step": 30160
+    },
+    {
+      "epoch": 0.717913929353518,
+      "grad_norm": 0.6358670921267884,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 30161
+    },
+    {
+      "epoch": 0.7179377320765495,
+      "grad_norm": 2.7377439134864097,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 30162
+    },
+    {
+      "epoch": 0.717961534799581,
+      "grad_norm": 2.784078398152966,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 30163
+    },
+    {
+      "epoch": 0.7179853375226126,
+      "grad_norm": 0.06747123041284232,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30164
+    },
+    {
+      "epoch": 0.7180091402456441,
+      "grad_norm": 5.136102394287044,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 30165
+    },
+    {
+      "epoch": 0.7180329429686756,
+      "grad_norm": 4.535401414975824,
+      "learning_rate": 2e-05,
+      "loss": 0.9919,
+      "step": 30166
+    },
+    {
+      "epoch": 0.7180567456917071,
+      "grad_norm": 3.614513783798037,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 30167
+    },
+    {
+      "epoch": 0.7180805484147387,
+      "grad_norm": 0.7897619158253039,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 30168
+    },
+    {
+      "epoch": 0.7181043511377702,
+      "grad_norm": 0.3707981141829987,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 30169
+    },
+    {
+      "epoch": 0.7181281538608016,
+      "grad_norm": 3.896657225223851,
+      "learning_rate": 2e-05,
+      "loss": 0.8833,
+      "step": 30170
+    },
+    {
+      "epoch": 0.7181519565838332,
+      "grad_norm": 3.4782488268261664,
+      "learning_rate": 2e-05,
+      "loss": 0.2567,
+      "step": 30171
+    },
+    {
+      "epoch": 0.7181757593068647,
+      "grad_norm": 2.1086087184184024,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 30172
+    },
+    {
+      "epoch": 0.7181995620298962,
+      "grad_norm": 2.8830243344377644,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 30173
+    },
+    {
+      "epoch": 0.7182233647529277,
+      "grad_norm": 5.242123755276414,
+      "learning_rate": 2e-05,
+      "loss": 0.9339,
+      "step": 30174
+    },
+    {
+      "epoch": 0.7182471674759593,
+      "grad_norm": 5.265910467000343,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 30175
+    },
+    {
+      "epoch": 0.7182709701989908,
+      "grad_norm": 5.372409110149629,
+      "learning_rate": 2e-05,
+      "loss": 0.7417,
+      "step": 30176
+    },
+    {
+      "epoch": 0.7182947729220223,
+      "grad_norm": 1.0431236433115383,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 30177
+    },
+    {
+      "epoch": 0.7183185756450537,
+      "grad_norm": 0.4100084216795669,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 30178
+    },
+    {
+      "epoch": 0.7183423783680853,
+      "grad_norm": 4.67588210855346,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 30179
+    },
+    {
+      "epoch": 0.7183661810911168,
+      "grad_norm": 29.76888719888347,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 30180
+    },
+    {
+      "epoch": 0.7183899838141483,
+      "grad_norm": 3.872209302853362,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 30181
+    },
+    {
+      "epoch": 0.7184137865371798,
+      "grad_norm": 10.243804225461808,
+      "learning_rate": 2e-05,
+      "loss": 1.0906,
+      "step": 30182
+    },
+    {
+      "epoch": 0.7184375892602114,
+      "grad_norm": 4.631220352657398,
+      "learning_rate": 2e-05,
+      "loss": 0.7404,
+      "step": 30183
+    },
+    {
+      "epoch": 0.7184613919832429,
+      "grad_norm": 2.88109580898493,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 30184
+    },
+    {
+      "epoch": 0.7184851947062744,
+      "grad_norm": 6.0857062445677,
+      "learning_rate": 2e-05,
+      "loss": 0.558,
+      "step": 30185
+    },
+    {
+      "epoch": 0.718508997429306,
+      "grad_norm": 0.3399801928256359,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 30186
+    },
+    {
+      "epoch": 0.7185328001523374,
+      "grad_norm": 0.5890637047565498,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 30187
+    },
+    {
+      "epoch": 0.7185566028753689,
+      "grad_norm": 4.819909756953567,
+      "learning_rate": 2e-05,
+      "loss": 0.6448,
+      "step": 30188
+    },
+    {
+      "epoch": 0.7185804055984004,
+      "grad_norm": 2.3375193210355314,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 30189
+    },
+    {
+      "epoch": 0.718604208321432,
+      "grad_norm": 0.06945620750883887,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30190
+    },
+    {
+      "epoch": 0.7186280110444635,
+      "grad_norm": 17.88545644514158,
+      "learning_rate": 2e-05,
+      "loss": 0.9221,
+      "step": 30191
+    },
+    {
+      "epoch": 0.718651813767495,
+      "grad_norm": 5.98450071139654,
+      "learning_rate": 2e-05,
+      "loss": 0.778,
+      "step": 30192
+    },
+    {
+      "epoch": 0.7186756164905265,
+      "grad_norm": 0.7420430044215194,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 30193
+    },
+    {
+      "epoch": 0.7186994192135581,
+      "grad_norm": 1.5106954191345248,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 30194
+    },
+    {
+      "epoch": 0.7187232219365896,
+      "grad_norm": 0.17218208174645697,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 30195
+    },
+    {
+      "epoch": 0.718747024659621,
+      "grad_norm": 0.20406365024993936,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 30196
+    },
+    {
+      "epoch": 0.7187708273826525,
+      "grad_norm": 5.389955505839036,
+      "learning_rate": 2e-05,
+      "loss": 1.1685,
+      "step": 30197
+    },
+    {
+      "epoch": 0.7187946301056841,
+      "grad_norm": 7.319105872110188,
+      "learning_rate": 2e-05,
+      "loss": 0.3846,
+      "step": 30198
+    },
+    {
+      "epoch": 0.7188184328287156,
+      "grad_norm": 0.6342917271789924,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 30199
+    },
+    {
+      "epoch": 0.7188422355517471,
+      "grad_norm": 4.476491370341058,
+      "learning_rate": 2e-05,
+      "loss": 0.3436,
+      "step": 30200
+    },
+    {
+      "epoch": 0.7188660382747787,
+      "grad_norm": 1.8834269192221873,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 30201
+    },
+    {
+      "epoch": 0.7188898409978102,
+      "grad_norm": 0.4054328041402437,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 30202
+    },
+    {
+      "epoch": 0.7189136437208417,
+      "grad_norm": 0.17367977985962538,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 30203
+    },
+    {
+      "epoch": 0.7189374464438731,
+      "grad_norm": 0.3468183384790743,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 30204
+    },
+    {
+      "epoch": 0.7189612491669047,
+      "grad_norm": 4.558173181298853,
+      "learning_rate": 2e-05,
+      "loss": 0.7586,
+      "step": 30205
+    },
+    {
+      "epoch": 0.7189850518899362,
+      "grad_norm": 3.7249832287716127,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 30206
+    },
+    {
+      "epoch": 0.7190088546129677,
+      "grad_norm": 1.636665519605114,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 30207
+    },
+    {
+      "epoch": 0.7190326573359992,
+      "grad_norm": 2.931164456002537,
+      "learning_rate": 2e-05,
+      "loss": 0.211,
+      "step": 30208
+    },
+    {
+      "epoch": 0.7190564600590308,
+      "grad_norm": 3.136300834630527,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 30209
+    },
+    {
+      "epoch": 0.7190802627820623,
+      "grad_norm": 3.6207878647368013,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 30210
+    },
+    {
+      "epoch": 0.7191040655050938,
+      "grad_norm": 1.0381798265923734,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 30211
+    },
+    {
+      "epoch": 0.7191278682281252,
+      "grad_norm": 0.3900236227320458,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 30212
+    },
+    {
+      "epoch": 0.7191516709511568,
+      "grad_norm": 1.0470193378911574,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 30213
+    },
+    {
+      "epoch": 0.7191754736741883,
+      "grad_norm": 0.08555961060304278,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 30214
+    },
+    {
+      "epoch": 0.7191992763972198,
+      "grad_norm": 3.198842428989272,
+      "learning_rate": 2e-05,
+      "loss": 0.5393,
+      "step": 30215
+    },
+    {
+      "epoch": 0.7192230791202514,
+      "grad_norm": 6.13081271295573,
+      "learning_rate": 2e-05,
+      "loss": 1.0734,
+      "step": 30216
+    },
+    {
+      "epoch": 0.7192468818432829,
+      "grad_norm": 1.740879501907677,
+      "learning_rate": 2e-05,
+      "loss": 0.2103,
+      "step": 30217
+    },
+    {
+      "epoch": 0.7192706845663144,
+      "grad_norm": 0.3750294500438186,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 30218
+    },
+    {
+      "epoch": 0.7192944872893459,
+      "grad_norm": 5.927817804017886,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 30219
+    },
+    {
+      "epoch": 0.7193182900123775,
+      "grad_norm": 3.200685754808418,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 30220
+    },
+    {
+      "epoch": 0.7193420927354089,
+      "grad_norm": 4.287343384952435,
+      "learning_rate": 2e-05,
+      "loss": 0.5211,
+      "step": 30221
+    },
+    {
+      "epoch": 0.7193658954584404,
+      "grad_norm": 8.018850458779772,
+      "learning_rate": 2e-05,
+      "loss": 0.5855,
+      "step": 30222
+    },
+    {
+      "epoch": 0.7193896981814719,
+      "grad_norm": 1.6767829491354924,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 30223
+    },
+    {
+      "epoch": 0.7194135009045035,
+      "grad_norm": 1.413064941395394,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 30224
+    },
+    {
+      "epoch": 0.719437303627535,
+      "grad_norm": 6.125214759946509,
+      "learning_rate": 2e-05,
+      "loss": 0.7799,
+      "step": 30225
+    },
+    {
+      "epoch": 0.7194611063505665,
+      "grad_norm": 2.6155107790971943,
+      "learning_rate": 2e-05,
+      "loss": 0.4434,
+      "step": 30226
+    },
+    {
+      "epoch": 0.719484909073598,
+      "grad_norm": 6.957916035596941,
+      "learning_rate": 2e-05,
+      "loss": 1.5184,
+      "step": 30227
+    },
+    {
+      "epoch": 0.7195087117966296,
+      "grad_norm": 3.2508434716933907,
+      "learning_rate": 2e-05,
+      "loss": 0.7388,
+      "step": 30228
+    },
+    {
+      "epoch": 0.719532514519661,
+      "grad_norm": 0.17122796047317032,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 30229
+    },
+    {
+      "epoch": 0.7195563172426925,
+      "grad_norm": 5.7512075259297655,
+      "learning_rate": 2e-05,
+      "loss": 1.0936,
+      "step": 30230
+    },
+    {
+      "epoch": 0.7195801199657241,
+      "grad_norm": 8.10291121353247,
+      "learning_rate": 2e-05,
+      "loss": 1.0821,
+      "step": 30231
+    },
+    {
+      "epoch": 0.7196039226887556,
+      "grad_norm": 0.7404088432905257,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 30232
+    },
+    {
+      "epoch": 0.7196277254117871,
+      "grad_norm": 3.8391779005541653,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 30233
+    },
+    {
+      "epoch": 0.7196515281348186,
+      "grad_norm": 4.5696179082078965,
+      "learning_rate": 2e-05,
+      "loss": 0.4371,
+      "step": 30234
+    },
+    {
+      "epoch": 0.7196753308578502,
+      "grad_norm": 0.729779141839052,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 30235
+    },
+    {
+      "epoch": 0.7196991335808817,
+      "grad_norm": 2.20072080216844,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 30236
+    },
+    {
+      "epoch": 0.7197229363039132,
+      "grad_norm": 0.4083030357205691,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 30237
+    },
+    {
+      "epoch": 0.7197467390269446,
+      "grad_norm": 4.972390185638398,
+      "learning_rate": 2e-05,
+      "loss": 0.4208,
+      "step": 30238
+    },
+    {
+      "epoch": 0.7197705417499762,
+      "grad_norm": 1.7289964239509652,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 30239
+    },
+    {
+      "epoch": 0.7197943444730077,
+      "grad_norm": 1.2272892811664784,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 30240
+    },
+    {
+      "epoch": 0.7198181471960392,
+      "grad_norm": 5.757220755226581,
+      "learning_rate": 2e-05,
+      "loss": 0.9003,
+      "step": 30241
+    },
+    {
+      "epoch": 0.7198419499190707,
+      "grad_norm": 0.4788423424475197,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 30242
+    },
+    {
+      "epoch": 0.7198657526421023,
+      "grad_norm": 4.196009409961353,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 30243
+    },
+    {
+      "epoch": 0.7198895553651338,
+      "grad_norm": 3.688905731153114,
+      "learning_rate": 2e-05,
+      "loss": 0.4618,
+      "step": 30244
+    },
+    {
+      "epoch": 0.7199133580881653,
+      "grad_norm": 10.838645545134767,
+      "learning_rate": 2e-05,
+      "loss": 0.3936,
+      "step": 30245
+    },
+    {
+      "epoch": 0.7199371608111969,
+      "grad_norm": 0.45973910721992595,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 30246
+    },
+    {
+      "epoch": 0.7199609635342283,
+      "grad_norm": 3.9123391421151523,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 30247
+    },
+    {
+      "epoch": 0.7199847662572598,
+      "grad_norm": 10.016931425442907,
+      "learning_rate": 2e-05,
+      "loss": 0.5201,
+      "step": 30248
+    },
+    {
+      "epoch": 0.7200085689802913,
+      "grad_norm": 3.5672392754319895,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 30249
+    },
+    {
+      "epoch": 0.7200323717033229,
+      "grad_norm": 4.961119192266329,
+      "learning_rate": 2e-05,
+      "loss": 0.4045,
+      "step": 30250
+    },
+    {
+      "epoch": 0.7200561744263544,
+      "grad_norm": 1.3961516310720812,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 30251
+    },
+    {
+      "epoch": 0.7200799771493859,
+      "grad_norm": 9.2410245517626,
+      "learning_rate": 2e-05,
+      "loss": 0.2291,
+      "step": 30252
+    },
+    {
+      "epoch": 0.7201037798724174,
+      "grad_norm": 2.4185548056373656,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 30253
+    },
+    {
+      "epoch": 0.720127582595449,
+      "grad_norm": 6.120912961220581,
+      "learning_rate": 2e-05,
+      "loss": 0.8812,
+      "step": 30254
+    },
+    {
+      "epoch": 0.7201513853184804,
+      "grad_norm": 2.8927766724257986,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 30255
+    },
+    {
+      "epoch": 0.7201751880415119,
+      "grad_norm": 3.961660289251392,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 30256
+    },
+    {
+      "epoch": 0.7201989907645434,
+      "grad_norm": 3.603576806043661,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 30257
+    },
+    {
+      "epoch": 0.720222793487575,
+      "grad_norm": 0.7100685371402875,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 30258
+    },
+    {
+      "epoch": 0.7202465962106065,
+      "grad_norm": 5.003259673691377,
+      "learning_rate": 2e-05,
+      "loss": 0.7422,
+      "step": 30259
+    },
+    {
+      "epoch": 0.720270398933638,
+      "grad_norm": 0.028974369901672528,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 30260
+    },
+    {
+      "epoch": 0.7202942016566696,
+      "grad_norm": 2.3713267518626653,
+      "learning_rate": 2e-05,
+      "loss": 0.2523,
+      "step": 30261
+    },
+    {
+      "epoch": 0.7203180043797011,
+      "grad_norm": 8.72291301995795,
+      "learning_rate": 2e-05,
+      "loss": 0.9818,
+      "step": 30262
+    },
+    {
+      "epoch": 0.7203418071027325,
+      "grad_norm": 0.30711743726892826,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 30263
+    },
+    {
+      "epoch": 0.720365609825764,
+      "grad_norm": 4.744258877465346,
+      "learning_rate": 2e-05,
+      "loss": 0.9497,
+      "step": 30264
+    },
+    {
+      "epoch": 0.7203894125487956,
+      "grad_norm": 4.565344093496375,
+      "learning_rate": 2e-05,
+      "loss": 0.7331,
+      "step": 30265
+    },
+    {
+      "epoch": 0.7204132152718271,
+      "grad_norm": 0.2530549030831006,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 30266
+    },
+    {
+      "epoch": 0.7204370179948586,
+      "grad_norm": 4.568791320607903,
+      "learning_rate": 2e-05,
+      "loss": 0.6236,
+      "step": 30267
+    },
+    {
+      "epoch": 0.7204608207178901,
+      "grad_norm": 3.4544531460055667,
+      "learning_rate": 2e-05,
+      "loss": 0.1929,
+      "step": 30268
+    },
+    {
+      "epoch": 0.7204846234409217,
+      "grad_norm": 2.529170714436491,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 30269
+    },
+    {
+      "epoch": 0.7205084261639532,
+      "grad_norm": 2.2069787546480417,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 30270
+    },
+    {
+      "epoch": 0.7205322288869846,
+      "grad_norm": 4.243108538741526,
+      "learning_rate": 2e-05,
+      "loss": 0.5808,
+      "step": 30271
+    },
+    {
+      "epoch": 0.7205560316100161,
+      "grad_norm": 2.0440823030925244,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 30272
+    },
+    {
+      "epoch": 0.7205798343330477,
+      "grad_norm": 0.062056450850688935,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 30273
+    },
+    {
+      "epoch": 0.7206036370560792,
+      "grad_norm": 2.1812799455711924,
+      "learning_rate": 2e-05,
+      "loss": 0.2126,
+      "step": 30274
+    },
+    {
+      "epoch": 0.7206274397791107,
+      "grad_norm": 3.723609295708034,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 30275
+    },
+    {
+      "epoch": 0.7206512425021423,
+      "grad_norm": 39.83701252880734,
+      "learning_rate": 2e-05,
+      "loss": 0.5119,
+      "step": 30276
+    },
+    {
+      "epoch": 0.7206750452251738,
+      "grad_norm": 0.06699184675997258,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30277
+    },
+    {
+      "epoch": 0.7206988479482053,
+      "grad_norm": 11.914743193115106,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 30278
+    },
+    {
+      "epoch": 0.7207226506712368,
+      "grad_norm": 4.096814440196174,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 30279
+    },
+    {
+      "epoch": 0.7207464533942683,
+      "grad_norm": 0.034595207665056665,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 30280
+    },
+    {
+      "epoch": 0.7207702561172998,
+      "grad_norm": 3.673196000459135,
+      "learning_rate": 2e-05,
+      "loss": 0.3579,
+      "step": 30281
+    },
+    {
+      "epoch": 0.7207940588403313,
+      "grad_norm": 7.974669153249243,
+      "learning_rate": 2e-05,
+      "loss": 0.6104,
+      "step": 30282
+    },
+    {
+      "epoch": 0.7208178615633628,
+      "grad_norm": 1.9022536485858677,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 30283
+    },
+    {
+      "epoch": 0.7208416642863944,
+      "grad_norm": 4.3179706920660905,
+      "learning_rate": 2e-05,
+      "loss": 0.7123,
+      "step": 30284
+    },
+    {
+      "epoch": 0.7208654670094259,
+      "grad_norm": 4.087440621679568,
+      "learning_rate": 2e-05,
+      "loss": 0.5549,
+      "step": 30285
+    },
+    {
+      "epoch": 0.7208892697324574,
+      "grad_norm": 13.050592570784202,
+      "learning_rate": 2e-05,
+      "loss": 0.7034,
+      "step": 30286
+    },
+    {
+      "epoch": 0.7209130724554889,
+      "grad_norm": 2.0865262011737116,
+      "learning_rate": 2e-05,
+      "loss": 0.2003,
+      "step": 30287
+    },
+    {
+      "epoch": 0.7209368751785205,
+      "grad_norm": 4.289173278684465,
+      "learning_rate": 2e-05,
+      "loss": 0.8026,
+      "step": 30288
+    },
+    {
+      "epoch": 0.7209606779015519,
+      "grad_norm": 8.784374625378659,
+      "learning_rate": 2e-05,
+      "loss": 1.5418,
+      "step": 30289
+    },
+    {
+      "epoch": 0.7209844806245834,
+      "grad_norm": 2.242583906756141,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 30290
+    },
+    {
+      "epoch": 0.721008283347615,
+      "grad_norm": 0.21574900536601022,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 30291
+    },
+    {
+      "epoch": 0.7210320860706465,
+      "grad_norm": 4.72575415799351,
+      "learning_rate": 2e-05,
+      "loss": 0.7398,
+      "step": 30292
+    },
+    {
+      "epoch": 0.721055888793678,
+      "grad_norm": 2.1219932946380737,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 30293
+    },
+    {
+      "epoch": 0.7210796915167095,
+      "grad_norm": 0.054478057247335544,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 30294
+    },
+    {
+      "epoch": 0.7211034942397411,
+      "grad_norm": 4.051740327594561,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 30295
+    },
+    {
+      "epoch": 0.7211272969627726,
+      "grad_norm": 1.0202141042664763,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 30296
+    },
+    {
+      "epoch": 0.721151099685804,
+      "grad_norm": 5.447961552291886,
+      "learning_rate": 2e-05,
+      "loss": 0.8203,
+      "step": 30297
+    },
+    {
+      "epoch": 0.7211749024088355,
+      "grad_norm": 7.73596289100102,
+      "learning_rate": 2e-05,
+      "loss": 1.4045,
+      "step": 30298
+    },
+    {
+      "epoch": 0.7211987051318671,
+      "grad_norm": 20.557724017796385,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 30299
+    },
+    {
+      "epoch": 0.7212225078548986,
+      "grad_norm": 29.98747858865827,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 30300
+    },
+    {
+      "epoch": 0.7212463105779301,
+      "grad_norm": 4.343980236679493,
+      "learning_rate": 2e-05,
+      "loss": 0.7947,
+      "step": 30301
+    },
+    {
+      "epoch": 0.7212701133009616,
+      "grad_norm": 1.5435905282837121,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 30302
+    },
+    {
+      "epoch": 0.7212939160239932,
+      "grad_norm": 2.4308945195691187,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 30303
+    },
+    {
+      "epoch": 0.7213177187470247,
+      "grad_norm": 3.240400125494435,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 30304
+    },
+    {
+      "epoch": 0.7213415214700561,
+      "grad_norm": 2.6966272961507536,
+      "learning_rate": 2e-05,
+      "loss": 0.2339,
+      "step": 30305
+    },
+    {
+      "epoch": 0.7213653241930877,
+      "grad_norm": 6.061870549193036,
+      "learning_rate": 2e-05,
+      "loss": 1.2695,
+      "step": 30306
+    },
+    {
+      "epoch": 0.7213891269161192,
+      "grad_norm": 8.801316052270867,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 30307
+    },
+    {
+      "epoch": 0.7214129296391507,
+      "grad_norm": 1.770113566507308,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 30308
+    },
+    {
+      "epoch": 0.7214367323621822,
+      "grad_norm": 2.42363004259365,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 30309
+    },
+    {
+      "epoch": 0.7214605350852138,
+      "grad_norm": 3.9286908793398965,
+      "learning_rate": 2e-05,
+      "loss": 0.4977,
+      "step": 30310
+    },
+    {
+      "epoch": 0.7214843378082453,
+      "grad_norm": 4.378973267201727,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 30311
+    },
+    {
+      "epoch": 0.7215081405312768,
+      "grad_norm": 0.747067430096484,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 30312
+    },
+    {
+      "epoch": 0.7215319432543083,
+      "grad_norm": 0.7290954490644047,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 30313
+    },
+    {
+      "epoch": 0.7215557459773398,
+      "grad_norm": 4.552855663783607,
+      "learning_rate": 2e-05,
+      "loss": 0.5442,
+      "step": 30314
+    },
+    {
+      "epoch": 0.7215795487003713,
+      "grad_norm": 4.450324281196003,
+      "learning_rate": 2e-05,
+      "loss": 0.7584,
+      "step": 30315
+    },
+    {
+      "epoch": 0.7216033514234028,
+      "grad_norm": 2.40118099027384,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 30316
+    },
+    {
+      "epoch": 0.7216271541464343,
+      "grad_norm": 0.8541688999132167,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 30317
+    },
+    {
+      "epoch": 0.7216509568694659,
+      "grad_norm": 5.989152793883749,
+      "learning_rate": 2e-05,
+      "loss": 1.452,
+      "step": 30318
+    },
+    {
+      "epoch": 0.7216747595924974,
+      "grad_norm": 0.9596184023453777,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 30319
+    },
+    {
+      "epoch": 0.7216985623155289,
+      "grad_norm": 3.3907571697737153,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 30320
+    },
+    {
+      "epoch": 0.7217223650385605,
+      "grad_norm": 0.2807220580687433,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 30321
+    },
+    {
+      "epoch": 0.721746167761592,
+      "grad_norm": 0.12131233483577225,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 30322
+    },
+    {
+      "epoch": 0.7217699704846234,
+      "grad_norm": 5.129940513077988,
+      "learning_rate": 2e-05,
+      "loss": 1.0498,
+      "step": 30323
+    },
+    {
+      "epoch": 0.7217937732076549,
+      "grad_norm": 3.821217486801335,
+      "learning_rate": 2e-05,
+      "loss": 0.2651,
+      "step": 30324
+    },
+    {
+      "epoch": 0.7218175759306865,
+      "grad_norm": 0.26598627221189813,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 30325
+    },
+    {
+      "epoch": 0.721841378653718,
+      "grad_norm": 0.5029924616376258,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 30326
+    },
+    {
+      "epoch": 0.7218651813767495,
+      "grad_norm": 5.500218237445341,
+      "learning_rate": 2e-05,
+      "loss": 1.2001,
+      "step": 30327
+    },
+    {
+      "epoch": 0.721888984099781,
+      "grad_norm": 5.969627581634018,
+      "learning_rate": 2e-05,
+      "loss": 1.0639,
+      "step": 30328
+    },
+    {
+      "epoch": 0.7219127868228126,
+      "grad_norm": 5.513897758439197,
+      "learning_rate": 2e-05,
+      "loss": 0.1568,
+      "step": 30329
+    },
+    {
+      "epoch": 0.721936589545844,
+      "grad_norm": 7.0646906713622295,
+      "learning_rate": 2e-05,
+      "loss": 1.4832,
+      "step": 30330
+    },
+    {
+      "epoch": 0.7219603922688755,
+      "grad_norm": 0.0733135424536034,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 30331
+    },
+    {
+      "epoch": 0.721984194991907,
+      "grad_norm": 4.179748159754027,
+      "learning_rate": 2e-05,
+      "loss": 0.4884,
+      "step": 30332
+    },
+    {
+      "epoch": 0.7220079977149386,
+      "grad_norm": 4.188308213438161,
+      "learning_rate": 2e-05,
+      "loss": 0.677,
+      "step": 30333
+    },
+    {
+      "epoch": 0.7220318004379701,
+      "grad_norm": 6.156070471067873,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 30334
+    },
+    {
+      "epoch": 0.7220556031610016,
+      "grad_norm": 0.19645038246171337,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 30335
+    },
+    {
+      "epoch": 0.7220794058840332,
+      "grad_norm": 0.8802350234028893,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 30336
+    },
+    {
+      "epoch": 0.7221032086070647,
+      "grad_norm": 2.949801623847176,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 30337
+    },
+    {
+      "epoch": 0.7221270113300962,
+      "grad_norm": 12.475529998853107,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 30338
+    },
+    {
+      "epoch": 0.7221508140531276,
+      "grad_norm": 0.1356656261090428,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 30339
+    },
+    {
+      "epoch": 0.7221746167761592,
+      "grad_norm": 1.968184178188747,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 30340
+    },
+    {
+      "epoch": 0.7221984194991907,
+      "grad_norm": 5.009483628306587,
+      "learning_rate": 2e-05,
+      "loss": 0.4852,
+      "step": 30341
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 0.42434915808011253,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 30342
+    },
+    {
+      "epoch": 0.7222460249452537,
+      "grad_norm": 2.4548734838729285,
+      "learning_rate": 2e-05,
+      "loss": 0.3023,
+      "step": 30343
+    },
+    {
+      "epoch": 0.7222698276682853,
+      "grad_norm": 8.677541160748643,
+      "learning_rate": 2e-05,
+      "loss": 0.9856,
+      "step": 30344
+    },
+    {
+      "epoch": 0.7222936303913168,
+      "grad_norm": 11.338757173397633,
+      "learning_rate": 2e-05,
+      "loss": 0.2716,
+      "step": 30345
+    },
+    {
+      "epoch": 0.7223174331143483,
+      "grad_norm": 3.9926503893734915,
+      "learning_rate": 2e-05,
+      "loss": 0.3903,
+      "step": 30346
+    },
+    {
+      "epoch": 0.7223412358373797,
+      "grad_norm": 1.1934743417062676,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 30347
+    },
+    {
+      "epoch": 0.7223650385604113,
+      "grad_norm": 1.5168065554305674,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 30348
+    },
+    {
+      "epoch": 0.7223888412834428,
+      "grad_norm": 8.703342796774889,
+      "learning_rate": 2e-05,
+      "loss": 0.4965,
+      "step": 30349
+    },
+    {
+      "epoch": 0.7224126440064743,
+      "grad_norm": 0.013192054610918803,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 30350
+    },
+    {
+      "epoch": 0.7224364467295059,
+      "grad_norm": 2.98175906137643,
+      "learning_rate": 2e-05,
+      "loss": 0.5217,
+      "step": 30351
+    },
+    {
+      "epoch": 0.7224602494525374,
+      "grad_norm": 4.661919363309261,
+      "learning_rate": 2e-05,
+      "loss": 0.8071,
+      "step": 30352
+    },
+    {
+      "epoch": 0.7224840521755689,
+      "grad_norm": 5.916499676776526,
+      "learning_rate": 2e-05,
+      "loss": 0.6348,
+      "step": 30353
+    },
+    {
+      "epoch": 0.7225078548986004,
+      "grad_norm": 3.6809070414681333,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 30354
+    },
+    {
+      "epoch": 0.722531657621632,
+      "grad_norm": 4.411270141950016,
+      "learning_rate": 2e-05,
+      "loss": 0.9979,
+      "step": 30355
+    },
+    {
+      "epoch": 0.7225554603446634,
+      "grad_norm": 1.3025986300736423,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 30356
+    },
+    {
+      "epoch": 0.7225792630676949,
+      "grad_norm": 0.8565979978782712,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 30357
+    },
+    {
+      "epoch": 0.7226030657907264,
+      "grad_norm": 3.4677577883619044,
+      "learning_rate": 2e-05,
+      "loss": 0.4258,
+      "step": 30358
+    },
+    {
+      "epoch": 0.722626868513758,
+      "grad_norm": 10.24190039853294,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 30359
+    },
+    {
+      "epoch": 0.7226506712367895,
+      "grad_norm": 2.2560582110789666,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 30360
+    },
+    {
+      "epoch": 0.722674473959821,
+      "grad_norm": 3.3472571224460537,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 30361
+    },
+    {
+      "epoch": 0.7226982766828525,
+      "grad_norm": 6.200825057336826,
+      "learning_rate": 2e-05,
+      "loss": 0.5276,
+      "step": 30362
+    },
+    {
+      "epoch": 0.7227220794058841,
+      "grad_norm": 0.19134997049676852,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 30363
+    },
+    {
+      "epoch": 0.7227458821289156,
+      "grad_norm": 2.9074188555155205,
+      "learning_rate": 2e-05,
+      "loss": 0.5765,
+      "step": 30364
+    },
+    {
+      "epoch": 0.722769684851947,
+      "grad_norm": 3.1089380181360933,
+      "learning_rate": 2e-05,
+      "loss": 0.3968,
+      "step": 30365
+    },
+    {
+      "epoch": 0.7227934875749786,
+      "grad_norm": 0.8912241827443008,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 30366
+    },
+    {
+      "epoch": 0.7228172902980101,
+      "grad_norm": 14.26964130681639,
+      "learning_rate": 2e-05,
+      "loss": 0.7976,
+      "step": 30367
+    },
+    {
+      "epoch": 0.7228410930210416,
+      "grad_norm": 0.9848134402392935,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 30368
+    },
+    {
+      "epoch": 0.7228648957440731,
+      "grad_norm": 0.09072772764673787,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 30369
+    },
+    {
+      "epoch": 0.7228886984671047,
+      "grad_norm": 1.8700427036553084,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 30370
+    },
+    {
+      "epoch": 0.7229125011901362,
+      "grad_norm": 3.7838485411931493,
+      "learning_rate": 2e-05,
+      "loss": 0.8072,
+      "step": 30371
+    },
+    {
+      "epoch": 0.7229363039131677,
+      "grad_norm": 2.536529094210087,
+      "learning_rate": 2e-05,
+      "loss": 0.1933,
+      "step": 30372
+    },
+    {
+      "epoch": 0.7229601066361991,
+      "grad_norm": 1.711034556532511,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 30373
+    },
+    {
+      "epoch": 0.7229839093592307,
+      "grad_norm": 0.0408258697243847,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 30374
+    },
+    {
+      "epoch": 0.7230077120822622,
+      "grad_norm": 0.08466614509509307,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 30375
+    },
+    {
+      "epoch": 0.7230315148052937,
+      "grad_norm": 0.005232790668139444,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 30376
+    },
+    {
+      "epoch": 0.7230553175283252,
+      "grad_norm": 17.78347077778883,
+      "learning_rate": 2e-05,
+      "loss": 0.3691,
+      "step": 30377
+    },
+    {
+      "epoch": 0.7230791202513568,
+      "grad_norm": 4.886122734186482,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 30378
+    },
+    {
+      "epoch": 0.7231029229743883,
+      "grad_norm": 4.5743601187496346,
+      "learning_rate": 2e-05,
+      "loss": 0.8027,
+      "step": 30379
+    },
+    {
+      "epoch": 0.7231267256974198,
+      "grad_norm": 3.2272948601304683,
+      "learning_rate": 2e-05,
+      "loss": 0.1427,
+      "step": 30380
+    },
+    {
+      "epoch": 0.7231505284204514,
+      "grad_norm": 3.1916037469875738,
+      "learning_rate": 2e-05,
+      "loss": 0.221,
+      "step": 30381
+    },
+    {
+      "epoch": 0.7231743311434828,
+      "grad_norm": 6.596112158404757,
+      "learning_rate": 2e-05,
+      "loss": 0.2882,
+      "step": 30382
+    },
+    {
+      "epoch": 0.7231981338665143,
+      "grad_norm": 2.409274533432111,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 30383
+    },
+    {
+      "epoch": 0.7232219365895458,
+      "grad_norm": 2.373180313544676,
+      "learning_rate": 2e-05,
+      "loss": 0.1193,
+      "step": 30384
+    },
+    {
+      "epoch": 0.7232457393125774,
+      "grad_norm": 0.2562721832808004,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 30385
+    },
+    {
+      "epoch": 0.7232695420356089,
+      "grad_norm": 2.6229142976592525,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 30386
+    },
+    {
+      "epoch": 0.7232933447586404,
+      "grad_norm": 0.14430790978585023,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 30387
+    },
+    {
+      "epoch": 0.7233171474816719,
+      "grad_norm": 0.5518415772261401,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 30388
+    },
+    {
+      "epoch": 0.7233409502047035,
+      "grad_norm": 3.27360180036023,
+      "learning_rate": 2e-05,
+      "loss": 0.3506,
+      "step": 30389
+    },
+    {
+      "epoch": 0.7233647529277349,
+      "grad_norm": 2.9398397269037,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 30390
+    },
+    {
+      "epoch": 0.7233885556507664,
+      "grad_norm": 0.6380636527876483,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 30391
+    },
+    {
+      "epoch": 0.7234123583737979,
+      "grad_norm": 4.127618870054916,
+      "learning_rate": 2e-05,
+      "loss": 1.081,
+      "step": 30392
+    },
+    {
+      "epoch": 0.7234361610968295,
+      "grad_norm": 0.6957321570816533,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 30393
+    },
+    {
+      "epoch": 0.723459963819861,
+      "grad_norm": 0.7258580552170045,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 30394
+    },
+    {
+      "epoch": 0.7234837665428925,
+      "grad_norm": 3.7303117738849574,
+      "learning_rate": 2e-05,
+      "loss": 0.7038,
+      "step": 30395
+    },
+    {
+      "epoch": 0.723507569265924,
+      "grad_norm": 3.514452269146976,
+      "learning_rate": 2e-05,
+      "loss": 0.3459,
+      "step": 30396
+    },
+    {
+      "epoch": 0.7235313719889556,
+      "grad_norm": 1.9539542175082547,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 30397
+    },
+    {
+      "epoch": 0.723555174711987,
+      "grad_norm": 5.920053405372212,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 30398
+    },
+    {
+      "epoch": 0.7235789774350185,
+      "grad_norm": 5.313200399892814,
+      "learning_rate": 2e-05,
+      "loss": 0.8183,
+      "step": 30399
+    },
+    {
+      "epoch": 0.7236027801580501,
+      "grad_norm": 2.622648661076775,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 30400
+    },
+    {
+      "epoch": 0.7236265828810816,
+      "grad_norm": 8.551651271197176,
+      "learning_rate": 2e-05,
+      "loss": 0.3637,
+      "step": 30401
+    },
+    {
+      "epoch": 0.7236503856041131,
+      "grad_norm": 0.20415891366538638,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 30402
+    },
+    {
+      "epoch": 0.7236741883271446,
+      "grad_norm": 7.32513639565335,
+      "learning_rate": 2e-05,
+      "loss": 0.1422,
+      "step": 30403
+    },
+    {
+      "epoch": 0.7236979910501762,
+      "grad_norm": 4.75461351558416,
+      "learning_rate": 2e-05,
+      "loss": 0.5611,
+      "step": 30404
+    },
+    {
+      "epoch": 0.7237217937732077,
+      "grad_norm": 5.352245523173479,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 30405
+    },
+    {
+      "epoch": 0.7237455964962392,
+      "grad_norm": 0.8936644763218292,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 30406
+    },
+    {
+      "epoch": 0.7237693992192706,
+      "grad_norm": 4.0812949978086595,
+      "learning_rate": 2e-05,
+      "loss": 0.3471,
+      "step": 30407
+    },
+    {
+      "epoch": 0.7237932019423022,
+      "grad_norm": 2.277651767461488,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 30408
+    },
+    {
+      "epoch": 0.7238170046653337,
+      "grad_norm": 0.23387400369647565,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 30409
+    },
+    {
+      "epoch": 0.7238408073883652,
+      "grad_norm": 0.4560336695943649,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 30410
+    },
+    {
+      "epoch": 0.7238646101113967,
+      "grad_norm": 0.9942836899284508,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 30411
+    },
+    {
+      "epoch": 0.7238884128344283,
+      "grad_norm": 2.8857341006133654,
+      "learning_rate": 2e-05,
+      "loss": 0.4744,
+      "step": 30412
+    },
+    {
+      "epoch": 0.7239122155574598,
+      "grad_norm": 3.982192676048598,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 30413
+    },
+    {
+      "epoch": 0.7239360182804913,
+      "grad_norm": 10.353947100780083,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 30414
+    },
+    {
+      "epoch": 0.7239598210035229,
+      "grad_norm": 14.473092263007032,
+      "learning_rate": 2e-05,
+      "loss": 0.3051,
+      "step": 30415
+    },
+    {
+      "epoch": 0.7239836237265543,
+      "grad_norm": 0.29974501993024416,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 30416
+    },
+    {
+      "epoch": 0.7240074264495858,
+      "grad_norm": 10.716893351822254,
+      "learning_rate": 2e-05,
+      "loss": 0.661,
+      "step": 30417
+    },
+    {
+      "epoch": 0.7240312291726173,
+      "grad_norm": 2.949642804241611,
+      "learning_rate": 2e-05,
+      "loss": 0.3231,
+      "step": 30418
+    },
+    {
+      "epoch": 0.7240550318956489,
+      "grad_norm": 21.74275680180814,
+      "learning_rate": 2e-05,
+      "loss": 0.2722,
+      "step": 30419
+    },
+    {
+      "epoch": 0.7240788346186804,
+      "grad_norm": 2.757424363504229,
+      "learning_rate": 2e-05,
+      "loss": 0.2552,
+      "step": 30420
+    },
+    {
+      "epoch": 0.7241026373417119,
+      "grad_norm": 0.027067446773587637,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 30421
+    },
+    {
+      "epoch": 0.7241264400647434,
+      "grad_norm": 4.550438742872875,
+      "learning_rate": 2e-05,
+      "loss": 0.4874,
+      "step": 30422
+    },
+    {
+      "epoch": 0.724150242787775,
+      "grad_norm": 5.543926902585854,
+      "learning_rate": 2e-05,
+      "loss": 0.544,
+      "step": 30423
+    },
+    {
+      "epoch": 0.7241740455108064,
+      "grad_norm": 4.702417339299033,
+      "learning_rate": 2e-05,
+      "loss": 1.0448,
+      "step": 30424
+    },
+    {
+      "epoch": 0.7241978482338379,
+      "grad_norm": 6.994658769676035,
+      "learning_rate": 2e-05,
+      "loss": 1.5398,
+      "step": 30425
+    },
+    {
+      "epoch": 0.7242216509568694,
+      "grad_norm": 2.9380672707451168,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 30426
+    },
+    {
+      "epoch": 0.724245453679901,
+      "grad_norm": 4.221852648359472,
+      "learning_rate": 2e-05,
+      "loss": 0.7449,
+      "step": 30427
+    },
+    {
+      "epoch": 0.7242692564029325,
+      "grad_norm": 1.971037389220967,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 30428
+    },
+    {
+      "epoch": 0.724293059125964,
+      "grad_norm": 3.8452501230767178,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 30429
+    },
+    {
+      "epoch": 0.7243168618489956,
+      "grad_norm": 3.193724058454714,
+      "learning_rate": 2e-05,
+      "loss": 0.4709,
+      "step": 30430
+    },
+    {
+      "epoch": 0.7243406645720271,
+      "grad_norm": 4.561480208540127,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 30431
+    },
+    {
+      "epoch": 0.7243644672950585,
+      "grad_norm": 4.9153369095547905,
+      "learning_rate": 2e-05,
+      "loss": 0.5202,
+      "step": 30432
+    },
+    {
+      "epoch": 0.72438827001809,
+      "grad_norm": 3.3906917221735586,
+      "learning_rate": 2e-05,
+      "loss": 0.4542,
+      "step": 30433
+    },
+    {
+      "epoch": 0.7244120727411216,
+      "grad_norm": 4.745666717361973,
+      "learning_rate": 2e-05,
+      "loss": 0.8744,
+      "step": 30434
+    },
+    {
+      "epoch": 0.7244358754641531,
+      "grad_norm": 2.9607385978355727,
+      "learning_rate": 2e-05,
+      "loss": 0.5681,
+      "step": 30435
+    },
+    {
+      "epoch": 0.7244596781871846,
+      "grad_norm": 2.0440043197416142,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 30436
+    },
+    {
+      "epoch": 0.7244834809102161,
+      "grad_norm": 3.4719046010378145,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 30437
+    },
+    {
+      "epoch": 0.7245072836332477,
+      "grad_norm": 1.1044676296485076,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 30438
+    },
+    {
+      "epoch": 0.7245310863562792,
+      "grad_norm": 3.3877250748631242,
+      "learning_rate": 2e-05,
+      "loss": 0.4496,
+      "step": 30439
+    },
+    {
+      "epoch": 0.7245548890793106,
+      "grad_norm": 0.15884293768724586,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 30440
+    },
+    {
+      "epoch": 0.7245786918023421,
+      "grad_norm": 0.08831040782333201,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 30441
+    },
+    {
+      "epoch": 0.7246024945253737,
+      "grad_norm": 0.6655374683287667,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 30442
+    },
+    {
+      "epoch": 0.7246262972484052,
+      "grad_norm": 3.154796331776438,
+      "learning_rate": 2e-05,
+      "loss": 0.31,
+      "step": 30443
+    },
+    {
+      "epoch": 0.7246500999714367,
+      "grad_norm": 0.054221142605194714,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30444
+    },
+    {
+      "epoch": 0.7246739026944683,
+      "grad_norm": 12.144146786401189,
+      "learning_rate": 2e-05,
+      "loss": 1.3542,
+      "step": 30445
+    },
+    {
+      "epoch": 0.7246977054174998,
+      "grad_norm": 5.930636851782153,
+      "learning_rate": 2e-05,
+      "loss": 0.1731,
+      "step": 30446
+    },
+    {
+      "epoch": 0.7247215081405313,
+      "grad_norm": 3.8960430018753414,
+      "learning_rate": 2e-05,
+      "loss": 0.4214,
+      "step": 30447
+    },
+    {
+      "epoch": 0.7247453108635628,
+      "grad_norm": 6.768094756122639,
+      "learning_rate": 2e-05,
+      "loss": 1.0002,
+      "step": 30448
+    },
+    {
+      "epoch": 0.7247691135865943,
+      "grad_norm": 2.7209969869719846,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 30449
+    },
+    {
+      "epoch": 0.7247929163096258,
+      "grad_norm": 1.6789944563136616,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 30450
+    },
+    {
+      "epoch": 0.7248167190326573,
+      "grad_norm": 0.879446305997681,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 30451
+    },
+    {
+      "epoch": 0.7248405217556888,
+      "grad_norm": 4.31725230800224,
+      "learning_rate": 2e-05,
+      "loss": 0.9423,
+      "step": 30452
+    },
+    {
+      "epoch": 0.7248643244787204,
+      "grad_norm": 7.4685889926226325,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 30453
+    },
+    {
+      "epoch": 0.7248881272017519,
+      "grad_norm": 0.21353918420322826,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 30454
+    },
+    {
+      "epoch": 0.7249119299247834,
+      "grad_norm": 3.3336712848569046,
+      "learning_rate": 2e-05,
+      "loss": 0.6216,
+      "step": 30455
+    },
+    {
+      "epoch": 0.7249357326478149,
+      "grad_norm": 3.2296116635905117,
+      "learning_rate": 2e-05,
+      "loss": 0.3452,
+      "step": 30456
+    },
+    {
+      "epoch": 0.7249595353708465,
+      "grad_norm": 2.27288222952452,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 30457
+    },
+    {
+      "epoch": 0.7249833380938779,
+      "grad_norm": 3.357357836924712,
+      "learning_rate": 2e-05,
+      "loss": 0.6734,
+      "step": 30458
+    },
+    {
+      "epoch": 0.7250071408169094,
+      "grad_norm": 0.19379188655921956,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 30459
+    },
+    {
+      "epoch": 0.725030943539941,
+      "grad_norm": 0.5847545969767771,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 30460
+    },
+    {
+      "epoch": 0.7250547462629725,
+      "grad_norm": 9.418672135435086,
+      "learning_rate": 2e-05,
+      "loss": 0.5087,
+      "step": 30461
+    },
+    {
+      "epoch": 0.725078548986004,
+      "grad_norm": 9.290874552443656,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 30462
+    },
+    {
+      "epoch": 0.7251023517090355,
+      "grad_norm": 1.089102959658705,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 30463
+    },
+    {
+      "epoch": 0.7251261544320671,
+      "grad_norm": 30.19157195826813,
+      "learning_rate": 2e-05,
+      "loss": 0.4189,
+      "step": 30464
+    },
+    {
+      "epoch": 0.7251499571550986,
+      "grad_norm": 4.233206430445578,
+      "learning_rate": 2e-05,
+      "loss": 0.871,
+      "step": 30465
+    },
+    {
+      "epoch": 0.72517375987813,
+      "grad_norm": 2.0982063365102896,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 30466
+    },
+    {
+      "epoch": 0.7251975626011615,
+      "grad_norm": 2.9403621703435814,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 30467
+    },
+    {
+      "epoch": 0.7252213653241931,
+      "grad_norm": 0.25071785335204555,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 30468
+    },
+    {
+      "epoch": 0.7252451680472246,
+      "grad_norm": 0.5267782194429397,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 30469
+    },
+    {
+      "epoch": 0.7252689707702561,
+      "grad_norm": 10.109223180623442,
+      "learning_rate": 2e-05,
+      "loss": 1.0775,
+      "step": 30470
+    },
+    {
+      "epoch": 0.7252927734932876,
+      "grad_norm": 0.1456089475121477,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 30471
+    },
+    {
+      "epoch": 0.7253165762163192,
+      "grad_norm": 3.6697261343676097,
+      "learning_rate": 2e-05,
+      "loss": 0.4621,
+      "step": 30472
+    },
+    {
+      "epoch": 0.7253403789393507,
+      "grad_norm": 8.759754834029609,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 30473
+    },
+    {
+      "epoch": 0.7253641816623821,
+      "grad_norm": 2.1523168132537553,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 30474
+    },
+    {
+      "epoch": 0.7253879843854137,
+      "grad_norm": 4.155596317041083,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 30475
+    },
+    {
+      "epoch": 0.7254117871084452,
+      "grad_norm": 12.468053649244665,
+      "learning_rate": 2e-05,
+      "loss": 0.5957,
+      "step": 30476
+    },
+    {
+      "epoch": 0.7254355898314767,
+      "grad_norm": 3.609246507972554,
+      "learning_rate": 2e-05,
+      "loss": 0.1833,
+      "step": 30477
+    },
+    {
+      "epoch": 0.7254593925545082,
+      "grad_norm": 2.8688314323798325,
+      "learning_rate": 2e-05,
+      "loss": 0.5281,
+      "step": 30478
+    },
+    {
+      "epoch": 0.7254831952775398,
+      "grad_norm": 0.42770838420735086,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 30479
+    },
+    {
+      "epoch": 0.7255069980005713,
+      "grad_norm": 15.6407439656842,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 30480
+    },
+    {
+      "epoch": 0.7255308007236028,
+      "grad_norm": 4.3558226032068434,
+      "learning_rate": 2e-05,
+      "loss": 0.8371,
+      "step": 30481
+    },
+    {
+      "epoch": 0.7255546034466342,
+      "grad_norm": 8.356068954169627,
+      "learning_rate": 2e-05,
+      "loss": 0.9374,
+      "step": 30482
+    },
+    {
+      "epoch": 0.7255784061696658,
+      "grad_norm": 5.333914645716016,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 30483
+    },
+    {
+      "epoch": 0.7256022088926973,
+      "grad_norm": 3.4815721258471526,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 30484
+    },
+    {
+      "epoch": 0.7256260116157288,
+      "grad_norm": 4.163500937792924,
+      "learning_rate": 2e-05,
+      "loss": 0.5493,
+      "step": 30485
+    },
+    {
+      "epoch": 0.7256498143387603,
+      "grad_norm": 6.97904803518,
+      "learning_rate": 2e-05,
+      "loss": 1.38,
+      "step": 30486
+    },
+    {
+      "epoch": 0.7256736170617919,
+      "grad_norm": 7.99982148433468,
+      "learning_rate": 2e-05,
+      "loss": 0.4439,
+      "step": 30487
+    },
+    {
+      "epoch": 0.7256974197848234,
+      "grad_norm": 1.6593732945835948,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 30488
+    },
+    {
+      "epoch": 0.7257212225078549,
+      "grad_norm": 1.6874248103668255,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 30489
+    },
+    {
+      "epoch": 0.7257450252308865,
+      "grad_norm": 3.3746242046063983,
+      "learning_rate": 2e-05,
+      "loss": 0.6891,
+      "step": 30490
+    },
+    {
+      "epoch": 0.725768827953918,
+      "grad_norm": 9.990211605316029,
+      "learning_rate": 2e-05,
+      "loss": 2.2036,
+      "step": 30491
+    },
+    {
+      "epoch": 0.7257926306769494,
+      "grad_norm": 11.329298302628976,
+      "learning_rate": 2e-05,
+      "loss": 1.01,
+      "step": 30492
+    },
+    {
+      "epoch": 0.7258164333999809,
+      "grad_norm": 0.0763195201240865,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 30493
+    },
+    {
+      "epoch": 0.7258402361230125,
+      "grad_norm": 0.0463444788370865,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 30494
+    },
+    {
+      "epoch": 0.725864038846044,
+      "grad_norm": 3.383444146039418,
+      "learning_rate": 2e-05,
+      "loss": 0.5173,
+      "step": 30495
+    },
+    {
+      "epoch": 0.7258878415690755,
+      "grad_norm": 0.8981602325695287,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 30496
+    },
+    {
+      "epoch": 0.725911644292107,
+      "grad_norm": 1.4537639670905218,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 30497
+    },
+    {
+      "epoch": 0.7259354470151386,
+      "grad_norm": 2.2798934678824674,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 30498
+    },
+    {
+      "epoch": 0.72595924973817,
+      "grad_norm": 6.84663238518844,
+      "learning_rate": 2e-05,
+      "loss": 0.4556,
+      "step": 30499
+    },
+    {
+      "epoch": 0.7259830524612015,
+      "grad_norm": 0.12597530999189627,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 30500
+    },
+    {
+      "epoch": 0.726006855184233,
+      "grad_norm": 6.092671991911294,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 30501
+    },
+    {
+      "epoch": 0.7260306579072646,
+      "grad_norm": 0.20366618925202706,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 30502
+    },
+    {
+      "epoch": 0.7260544606302961,
+      "grad_norm": 4.458906950118676,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 30503
+    },
+    {
+      "epoch": 0.7260782633533276,
+      "grad_norm": 3.371003155522262,
+      "learning_rate": 2e-05,
+      "loss": 0.3454,
+      "step": 30504
+    },
+    {
+      "epoch": 0.7261020660763592,
+      "grad_norm": 9.948436015465413,
+      "learning_rate": 2e-05,
+      "loss": 0.8026,
+      "step": 30505
+    },
+    {
+      "epoch": 0.7261258687993907,
+      "grad_norm": 11.224020487835423,
+      "learning_rate": 2e-05,
+      "loss": 1.231,
+      "step": 30506
+    },
+    {
+      "epoch": 0.7261496715224222,
+      "grad_norm": 0.7094801566650769,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 30507
+    },
+    {
+      "epoch": 0.7261734742454536,
+      "grad_norm": 3.4854172871306788,
+      "learning_rate": 2e-05,
+      "loss": 0.5171,
+      "step": 30508
+    },
+    {
+      "epoch": 0.7261972769684852,
+      "grad_norm": 4.632451595555017,
+      "learning_rate": 2e-05,
+      "loss": 0.6983,
+      "step": 30509
+    },
+    {
+      "epoch": 0.7262210796915167,
+      "grad_norm": 0.971638255927577,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 30510
+    },
+    {
+      "epoch": 0.7262448824145482,
+      "grad_norm": 2.48773518531859,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 30511
+    },
+    {
+      "epoch": 0.7262686851375797,
+      "grad_norm": 0.039059424667178516,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30512
+    },
+    {
+      "epoch": 0.7262924878606113,
+      "grad_norm": 5.730249050221095,
+      "learning_rate": 2e-05,
+      "loss": 0.7059,
+      "step": 30513
+    },
+    {
+      "epoch": 0.7263162905836428,
+      "grad_norm": 5.362343251393532,
+      "learning_rate": 2e-05,
+      "loss": 0.2329,
+      "step": 30514
+    },
+    {
+      "epoch": 0.7263400933066743,
+      "grad_norm": 8.772598295723395,
+      "learning_rate": 2e-05,
+      "loss": 0.4603,
+      "step": 30515
+    },
+    {
+      "epoch": 0.7263638960297057,
+      "grad_norm": 1.3035117204008109,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 30516
+    },
+    {
+      "epoch": 0.7263876987527373,
+      "grad_norm": 3.8701368884238336,
+      "learning_rate": 2e-05,
+      "loss": 0.648,
+      "step": 30517
+    },
+    {
+      "epoch": 0.7264115014757688,
+      "grad_norm": 4.416196483213915,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 30518
+    },
+    {
+      "epoch": 0.7264353041988003,
+      "grad_norm": 2.308636592446306,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 30519
+    },
+    {
+      "epoch": 0.7264591069218319,
+      "grad_norm": 1.8571053239093658,
+      "learning_rate": 2e-05,
+      "loss": 0.0474,
+      "step": 30520
+    },
+    {
+      "epoch": 0.7264829096448634,
+      "grad_norm": 2.1030250576389116,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 30521
+    },
+    {
+      "epoch": 0.7265067123678949,
+      "grad_norm": 4.941132400903244,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 30522
+    },
+    {
+      "epoch": 0.7265305150909264,
+      "grad_norm": 1.025086200937764,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 30523
+    },
+    {
+      "epoch": 0.726554317813958,
+      "grad_norm": 1.5909617854024332,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 30524
+    },
+    {
+      "epoch": 0.7265781205369894,
+      "grad_norm": 7.1537450491887355,
+      "learning_rate": 2e-05,
+      "loss": 0.9662,
+      "step": 30525
+    },
+    {
+      "epoch": 0.7266019232600209,
+      "grad_norm": 2.664243042868317,
+      "learning_rate": 2e-05,
+      "loss": 0.1905,
+      "step": 30526
+    },
+    {
+      "epoch": 0.7266257259830524,
+      "grad_norm": 0.1950033684412719,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 30527
+    },
+    {
+      "epoch": 0.726649528706084,
+      "grad_norm": 1.9846084204132413,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 30528
+    },
+    {
+      "epoch": 0.7266733314291155,
+      "grad_norm": 13.720992825674939,
+      "learning_rate": 2e-05,
+      "loss": 0.3934,
+      "step": 30529
+    },
+    {
+      "epoch": 0.726697134152147,
+      "grad_norm": 7.740294635779461,
+      "learning_rate": 2e-05,
+      "loss": 0.8257,
+      "step": 30530
+    },
+    {
+      "epoch": 0.7267209368751785,
+      "grad_norm": 0.7696457406327719,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 30531
+    },
+    {
+      "epoch": 0.7267447395982101,
+      "grad_norm": 4.919900819058559,
+      "learning_rate": 2e-05,
+      "loss": 0.6052,
+      "step": 30532
+    },
+    {
+      "epoch": 0.7267685423212416,
+      "grad_norm": 23.600741038760436,
+      "learning_rate": 2e-05,
+      "loss": 0.7852,
+      "step": 30533
+    },
+    {
+      "epoch": 0.726792345044273,
+      "grad_norm": 15.440386193864617,
+      "learning_rate": 2e-05,
+      "loss": 0.9332,
+      "step": 30534
+    },
+    {
+      "epoch": 0.7268161477673046,
+      "grad_norm": 2.7870113528035128,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 30535
+    },
+    {
+      "epoch": 0.7268399504903361,
+      "grad_norm": 4.45183602403411,
+      "learning_rate": 2e-05,
+      "loss": 0.3945,
+      "step": 30536
+    },
+    {
+      "epoch": 0.7268637532133676,
+      "grad_norm": 1.807597833332804,
+      "learning_rate": 2e-05,
+      "loss": 0.047,
+      "step": 30537
+    },
+    {
+      "epoch": 0.7268875559363991,
+      "grad_norm": 3.3624195660838803,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 30538
+    },
+    {
+      "epoch": 0.7269113586594307,
+      "grad_norm": 6.00127901008754,
+      "learning_rate": 2e-05,
+      "loss": 1.3471,
+      "step": 30539
+    },
+    {
+      "epoch": 0.7269351613824622,
+      "grad_norm": 2.009286112651764,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 30540
+    },
+    {
+      "epoch": 0.7269589641054937,
+      "grad_norm": 11.28647596679793,
+      "learning_rate": 2e-05,
+      "loss": 1.506,
+      "step": 30541
+    },
+    {
+      "epoch": 0.7269827668285251,
+      "grad_norm": 0.4605899371509831,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 30542
+    },
+    {
+      "epoch": 0.7270065695515567,
+      "grad_norm": 2.7485964346263994,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 30543
+    },
+    {
+      "epoch": 0.7270303722745882,
+      "grad_norm": 8.437340039100327,
+      "learning_rate": 2e-05,
+      "loss": 0.3393,
+      "step": 30544
+    },
+    {
+      "epoch": 0.7270541749976197,
+      "grad_norm": 3.567901286116259,
+      "learning_rate": 2e-05,
+      "loss": 0.7405,
+      "step": 30545
+    },
+    {
+      "epoch": 0.7270779777206512,
+      "grad_norm": 2.3414618715206688,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 30546
+    },
+    {
+      "epoch": 0.7271017804436828,
+      "grad_norm": 0.6171086621951497,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 30547
+    },
+    {
+      "epoch": 0.7271255831667143,
+      "grad_norm": 1.900406699661599,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 30548
+    },
+    {
+      "epoch": 0.7271493858897458,
+      "grad_norm": 2.9821946250848703,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 30549
+    },
+    {
+      "epoch": 0.7271731886127774,
+      "grad_norm": 6.45330840123329,
+      "learning_rate": 2e-05,
+      "loss": 0.852,
+      "step": 30550
+    },
+    {
+      "epoch": 0.7271969913358088,
+      "grad_norm": 4.955054036644835,
+      "learning_rate": 2e-05,
+      "loss": 0.8305,
+      "step": 30551
+    },
+    {
+      "epoch": 0.7272207940588403,
+      "grad_norm": 10.542472717885165,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 30552
+    },
+    {
+      "epoch": 0.7272445967818718,
+      "grad_norm": 3.4205145269548805,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 30553
+    },
+    {
+      "epoch": 0.7272683995049034,
+      "grad_norm": 3.5374503180007646,
+      "learning_rate": 2e-05,
+      "loss": 0.2265,
+      "step": 30554
+    },
+    {
+      "epoch": 0.7272922022279349,
+      "grad_norm": 3.9973976978205874,
+      "learning_rate": 2e-05,
+      "loss": 0.6312,
+      "step": 30555
+    },
+    {
+      "epoch": 0.7273160049509664,
+      "grad_norm": 6.215585624898624,
+      "learning_rate": 2e-05,
+      "loss": 1.6202,
+      "step": 30556
+    },
+    {
+      "epoch": 0.7273398076739979,
+      "grad_norm": 1.7526069660541328,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 30557
+    },
+    {
+      "epoch": 0.7273636103970295,
+      "grad_norm": 4.2468212807912895,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 30558
+    },
+    {
+      "epoch": 0.7273874131200609,
+      "grad_norm": 3.5196665720223543,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 30559
+    },
+    {
+      "epoch": 0.7274112158430924,
+      "grad_norm": 0.5701218434853788,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 30560
+    },
+    {
+      "epoch": 0.7274350185661239,
+      "grad_norm": 5.833868182982607,
+      "learning_rate": 2e-05,
+      "loss": 0.4365,
+      "step": 30561
+    },
+    {
+      "epoch": 0.7274588212891555,
+      "grad_norm": 4.529978077061427,
+      "learning_rate": 2e-05,
+      "loss": 0.5269,
+      "step": 30562
+    },
+    {
+      "epoch": 0.727482624012187,
+      "grad_norm": 0.10332776228405856,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 30563
+    },
+    {
+      "epoch": 0.7275064267352185,
+      "grad_norm": 14.751312154979807,
+      "learning_rate": 2e-05,
+      "loss": 0.6925,
+      "step": 30564
+    },
+    {
+      "epoch": 0.7275302294582501,
+      "grad_norm": 2.7594490982387625,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 30565
+    },
+    {
+      "epoch": 0.7275540321812816,
+      "grad_norm": 8.758485150104054,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 30566
+    },
+    {
+      "epoch": 0.727577834904313,
+      "grad_norm": 6.537811592272769,
+      "learning_rate": 2e-05,
+      "loss": 0.2409,
+      "step": 30567
+    },
+    {
+      "epoch": 0.7276016376273445,
+      "grad_norm": 13.466415048454676,
+      "learning_rate": 2e-05,
+      "loss": 2.1665,
+      "step": 30568
+    },
+    {
+      "epoch": 0.7276254403503761,
+      "grad_norm": 3.072292251450513,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 30569
+    },
+    {
+      "epoch": 0.7276492430734076,
+      "grad_norm": 3.542070498910922,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 30570
+    },
+    {
+      "epoch": 0.7276730457964391,
+      "grad_norm": 4.8704451463015745,
+      "learning_rate": 2e-05,
+      "loss": 0.7307,
+      "step": 30571
+    },
+    {
+      "epoch": 0.7276968485194706,
+      "grad_norm": 3.021263058826736,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 30572
+    },
+    {
+      "epoch": 0.7277206512425022,
+      "grad_norm": 2.2086150702674208,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 30573
+    },
+    {
+      "epoch": 0.7277444539655337,
+      "grad_norm": 3.959756647026417,
+      "learning_rate": 2e-05,
+      "loss": 0.3773,
+      "step": 30574
+    },
+    {
+      "epoch": 0.7277682566885652,
+      "grad_norm": 0.29351066045590096,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 30575
+    },
+    {
+      "epoch": 0.7277920594115966,
+      "grad_norm": 1.9281005149796866,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 30576
+    },
+    {
+      "epoch": 0.7278158621346282,
+      "grad_norm": 10.937390303306103,
+      "learning_rate": 2e-05,
+      "loss": 1.205,
+      "step": 30577
+    },
+    {
+      "epoch": 0.7278396648576597,
+      "grad_norm": 6.0970934958543,
+      "learning_rate": 2e-05,
+      "loss": 1.7215,
+      "step": 30578
+    },
+    {
+      "epoch": 0.7278634675806912,
+      "grad_norm": 0.30259964992329136,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 30579
+    },
+    {
+      "epoch": 0.7278872703037228,
+      "grad_norm": 0.06541300944728926,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 30580
+    },
+    {
+      "epoch": 0.7279110730267543,
+      "grad_norm": 3.3775438736382632,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 30581
+    },
+    {
+      "epoch": 0.7279348757497858,
+      "grad_norm": 0.4363466056720853,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 30582
+    },
+    {
+      "epoch": 0.7279586784728173,
+      "grad_norm": 3.4226149048900285,
+      "learning_rate": 2e-05,
+      "loss": 0.5711,
+      "step": 30583
+    },
+    {
+      "epoch": 0.7279824811958489,
+      "grad_norm": 0.7387147524197519,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 30584
+    },
+    {
+      "epoch": 0.7280062839188803,
+      "grad_norm": 7.16665001212819,
+      "learning_rate": 2e-05,
+      "loss": 2.1481,
+      "step": 30585
+    },
+    {
+      "epoch": 0.7280300866419118,
+      "grad_norm": 1.8213510057786917,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 30586
+    },
+    {
+      "epoch": 0.7280538893649433,
+      "grad_norm": 7.027314624642513,
+      "learning_rate": 2e-05,
+      "loss": 0.8724,
+      "step": 30587
+    },
+    {
+      "epoch": 0.7280776920879749,
+      "grad_norm": 2.2637012653337276,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 30588
+    },
+    {
+      "epoch": 0.7281014948110064,
+      "grad_norm": 6.013757847576944,
+      "learning_rate": 2e-05,
+      "loss": 0.6856,
+      "step": 30589
+    },
+    {
+      "epoch": 0.7281252975340379,
+      "grad_norm": 3.870199945123824,
+      "learning_rate": 2e-05,
+      "loss": 0.6317,
+      "step": 30590
+    },
+    {
+      "epoch": 0.7281491002570694,
+      "grad_norm": 6.478510190003589,
+      "learning_rate": 2e-05,
+      "loss": 0.1981,
+      "step": 30591
+    },
+    {
+      "epoch": 0.728172902980101,
+      "grad_norm": 3.7828359695779783,
+      "learning_rate": 2e-05,
+      "loss": 0.5498,
+      "step": 30592
+    },
+    {
+      "epoch": 0.7281967057031324,
+      "grad_norm": 4.974575723431421,
+      "learning_rate": 2e-05,
+      "loss": 0.6368,
+      "step": 30593
+    },
+    {
+      "epoch": 0.7282205084261639,
+      "grad_norm": 3.8889964835350117,
+      "learning_rate": 2e-05,
+      "loss": 0.3118,
+      "step": 30594
+    },
+    {
+      "epoch": 0.7282443111491955,
+      "grad_norm": 1.1421365071705716,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 30595
+    },
+    {
+      "epoch": 0.728268113872227,
+      "grad_norm": 1.0781597784105805,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 30596
+    },
+    {
+      "epoch": 0.7282919165952585,
+      "grad_norm": 5.733061062315594,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 30597
+    },
+    {
+      "epoch": 0.72831571931829,
+      "grad_norm": 10.656758018605593,
+      "learning_rate": 2e-05,
+      "loss": 1.7843,
+      "step": 30598
+    },
+    {
+      "epoch": 0.7283395220413216,
+      "grad_norm": 4.075511959648756,
+      "learning_rate": 2e-05,
+      "loss": 0.5808,
+      "step": 30599
+    },
+    {
+      "epoch": 0.7283633247643531,
+      "grad_norm": 0.8330849037296061,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 30600
+    },
+    {
+      "epoch": 0.7283871274873845,
+      "grad_norm": 2.7957037266916904,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 30601
+    },
+    {
+      "epoch": 0.728410930210416,
+      "grad_norm": 0.37027429078334456,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 30602
+    },
+    {
+      "epoch": 0.7284347329334476,
+      "grad_norm": 5.600181279742286,
+      "learning_rate": 2e-05,
+      "loss": 0.9216,
+      "step": 30603
+    },
+    {
+      "epoch": 0.7284585356564791,
+      "grad_norm": 0.4857808872207441,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 30604
+    },
+    {
+      "epoch": 0.7284823383795106,
+      "grad_norm": 0.09843430360201592,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 30605
+    },
+    {
+      "epoch": 0.7285061411025421,
+      "grad_norm": 19.954426348748356,
+      "learning_rate": 2e-05,
+      "loss": 0.4174,
+      "step": 30606
+    },
+    {
+      "epoch": 0.7285299438255737,
+      "grad_norm": 4.947816482104837,
+      "learning_rate": 2e-05,
+      "loss": 1.0546,
+      "step": 30607
+    },
+    {
+      "epoch": 0.7285537465486052,
+      "grad_norm": 3.1328730186779596,
+      "learning_rate": 2e-05,
+      "loss": 0.4231,
+      "step": 30608
+    },
+    {
+      "epoch": 0.7285775492716366,
+      "grad_norm": 5.2245541997210365,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 30609
+    },
+    {
+      "epoch": 0.7286013519946682,
+      "grad_norm": 6.823856307506664,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 30610
+    },
+    {
+      "epoch": 0.7286251547176997,
+      "grad_norm": 1.0469523142657775,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 30611
+    },
+    {
+      "epoch": 0.7286489574407312,
+      "grad_norm": 3.116660581301768,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 30612
+    },
+    {
+      "epoch": 0.7286727601637627,
+      "grad_norm": 0.6716640337980185,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 30613
+    },
+    {
+      "epoch": 0.7286965628867943,
+      "grad_norm": 0.2715148832861581,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 30614
+    },
+    {
+      "epoch": 0.7287203656098258,
+      "grad_norm": 2.3685735312515024,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 30615
+    },
+    {
+      "epoch": 0.7287441683328573,
+      "grad_norm": 2.9503117445610525,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 30616
+    },
+    {
+      "epoch": 0.7287679710558888,
+      "grad_norm": 0.3555274046911975,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 30617
+    },
+    {
+      "epoch": 0.7287917737789203,
+      "grad_norm": 1.8000388099280313,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 30618
+    },
+    {
+      "epoch": 0.7288155765019518,
+      "grad_norm": 2.1372724433292327,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 30619
+    },
+    {
+      "epoch": 0.7288393792249833,
+      "grad_norm": 0.6993468137267732,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 30620
+    },
+    {
+      "epoch": 0.7288631819480148,
+      "grad_norm": 3.5158516334228778,
+      "learning_rate": 2e-05,
+      "loss": 0.4174,
+      "step": 30621
+    },
+    {
+      "epoch": 0.7288869846710464,
+      "grad_norm": 0.7835913910497849,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 30622
+    },
+    {
+      "epoch": 0.7289107873940779,
+      "grad_norm": 4.073555027132853,
+      "learning_rate": 2e-05,
+      "loss": 0.3367,
+      "step": 30623
+    },
+    {
+      "epoch": 0.7289345901171094,
+      "grad_norm": 2.2789209115846707,
+      "learning_rate": 2e-05,
+      "loss": 0.0932,
+      "step": 30624
+    },
+    {
+      "epoch": 0.728958392840141,
+      "grad_norm": 4.214533671668538,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 30625
+    },
+    {
+      "epoch": 0.7289821955631725,
+      "grad_norm": 3.2103784030921267,
+      "learning_rate": 2e-05,
+      "loss": 0.5256,
+      "step": 30626
+    },
+    {
+      "epoch": 0.7290059982862039,
+      "grad_norm": 5.0765880934163485,
+      "learning_rate": 2e-05,
+      "loss": 1.1973,
+      "step": 30627
+    },
+    {
+      "epoch": 0.7290298010092354,
+      "grad_norm": 0.0935155716936666,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30628
+    },
+    {
+      "epoch": 0.729053603732267,
+      "grad_norm": 0.016091240888411774,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 30629
+    },
+    {
+      "epoch": 0.7290774064552985,
+      "grad_norm": 2.446687791321042,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 30630
+    },
+    {
+      "epoch": 0.72910120917833,
+      "grad_norm": 9.30498540337091,
+      "learning_rate": 2e-05,
+      "loss": 0.3535,
+      "step": 30631
+    },
+    {
+      "epoch": 0.7291250119013615,
+      "grad_norm": 3.7019886798774455,
+      "learning_rate": 2e-05,
+      "loss": 0.3876,
+      "step": 30632
+    },
+    {
+      "epoch": 0.7291488146243931,
+      "grad_norm": 0.11327121775037018,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30633
+    },
+    {
+      "epoch": 0.7291726173474246,
+      "grad_norm": 3.628980874961204,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 30634
+    },
+    {
+      "epoch": 0.729196420070456,
+      "grad_norm": 1.1166586177700575,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 30635
+    },
+    {
+      "epoch": 0.7292202227934875,
+      "grad_norm": 4.114569078795159,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 30636
+    },
+    {
+      "epoch": 0.7292440255165191,
+      "grad_norm": 3.609530038597721,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 30637
+    },
+    {
+      "epoch": 0.7292678282395506,
+      "grad_norm": 0.9850063305228983,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 30638
+    },
+    {
+      "epoch": 0.7292916309625821,
+      "grad_norm": 4.915924485691799,
+      "learning_rate": 2e-05,
+      "loss": 0.357,
+      "step": 30639
+    },
+    {
+      "epoch": 0.7293154336856136,
+      "grad_norm": 0.14376934786341045,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 30640
+    },
+    {
+      "epoch": 0.7293392364086452,
+      "grad_norm": 0.37016833602257043,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 30641
+    },
+    {
+      "epoch": 0.7293630391316767,
+      "grad_norm": 2.4619087889590285,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 30642
+    },
+    {
+      "epoch": 0.7293868418547081,
+      "grad_norm": 0.11649828072102897,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 30643
+    },
+    {
+      "epoch": 0.7294106445777397,
+      "grad_norm": 0.06433587141406011,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 30644
+    },
+    {
+      "epoch": 0.7294344473007712,
+      "grad_norm": 1.6012475127491304,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 30645
+    },
+    {
+      "epoch": 0.7294582500238027,
+      "grad_norm": 4.263335318835072,
+      "learning_rate": 2e-05,
+      "loss": 0.7715,
+      "step": 30646
+    },
+    {
+      "epoch": 0.7294820527468342,
+      "grad_norm": 0.09484691269966934,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30647
+    },
+    {
+      "epoch": 0.7295058554698658,
+      "grad_norm": 4.913266706938009,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 30648
+    },
+    {
+      "epoch": 0.7295296581928973,
+      "grad_norm": 1.992332594325936,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 30649
+    },
+    {
+      "epoch": 0.7295534609159288,
+      "grad_norm": 4.017237741747914,
+      "learning_rate": 2e-05,
+      "loss": 0.8588,
+      "step": 30650
+    },
+    {
+      "epoch": 0.7295772636389602,
+      "grad_norm": 1.7067753831622763,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 30651
+    },
+    {
+      "epoch": 0.7296010663619918,
+      "grad_norm": 2.310887980128645,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 30652
+    },
+    {
+      "epoch": 0.7296248690850233,
+      "grad_norm": 0.5905921347095996,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 30653
+    },
+    {
+      "epoch": 0.7296486718080548,
+      "grad_norm": 0.10044981706326163,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 30654
+    },
+    {
+      "epoch": 0.7296724745310863,
+      "grad_norm": 16.362220394123643,
+      "learning_rate": 2e-05,
+      "loss": 0.7871,
+      "step": 30655
+    },
+    {
+      "epoch": 0.7296962772541179,
+      "grad_norm": 8.12489583311148,
+      "learning_rate": 2e-05,
+      "loss": 0.1629,
+      "step": 30656
+    },
+    {
+      "epoch": 0.7297200799771494,
+      "grad_norm": 25.031652643902113,
+      "learning_rate": 2e-05,
+      "loss": 0.6473,
+      "step": 30657
+    },
+    {
+      "epoch": 0.7297438827001809,
+      "grad_norm": 4.790178523294291,
+      "learning_rate": 2e-05,
+      "loss": 0.8946,
+      "step": 30658
+    },
+    {
+      "epoch": 0.7297676854232125,
+      "grad_norm": 4.094925608493028,
+      "learning_rate": 2e-05,
+      "loss": 0.903,
+      "step": 30659
+    },
+    {
+      "epoch": 0.729791488146244,
+      "grad_norm": 2.116231604947785,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 30660
+    },
+    {
+      "epoch": 0.7298152908692754,
+      "grad_norm": 3.690832894202131,
+      "learning_rate": 2e-05,
+      "loss": 0.1242,
+      "step": 30661
+    },
+    {
+      "epoch": 0.7298390935923069,
+      "grad_norm": 3.3018788074021375,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 30662
+    },
+    {
+      "epoch": 0.7298628963153385,
+      "grad_norm": 4.619777630589727,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 30663
+    },
+    {
+      "epoch": 0.72988669903837,
+      "grad_norm": 4.773314609081036,
+      "learning_rate": 2e-05,
+      "loss": 0.6419,
+      "step": 30664
+    },
+    {
+      "epoch": 0.7299105017614015,
+      "grad_norm": 6.604420701782484,
+      "learning_rate": 2e-05,
+      "loss": 0.6318,
+      "step": 30665
+    },
+    {
+      "epoch": 0.729934304484433,
+      "grad_norm": 12.666287639407722,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 30666
+    },
+    {
+      "epoch": 0.7299581072074646,
+      "grad_norm": 2.7675676588592824,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 30667
+    },
+    {
+      "epoch": 0.729981909930496,
+      "grad_norm": 5.8296574714897655,
+      "learning_rate": 2e-05,
+      "loss": 1.4497,
+      "step": 30668
+    },
+    {
+      "epoch": 0.7300057126535275,
+      "grad_norm": 5.302212189760499,
+      "learning_rate": 2e-05,
+      "loss": 0.5622,
+      "step": 30669
+    },
+    {
+      "epoch": 0.730029515376559,
+      "grad_norm": 4.866365194241361,
+      "learning_rate": 2e-05,
+      "loss": 1.1228,
+      "step": 30670
+    },
+    {
+      "epoch": 0.7300533180995906,
+      "grad_norm": 5.767961567273179,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 30671
+    },
+    {
+      "epoch": 0.7300771208226221,
+      "grad_norm": 3.025156094289135,
+      "learning_rate": 2e-05,
+      "loss": 0.2727,
+      "step": 30672
+    },
+    {
+      "epoch": 0.7301009235456536,
+      "grad_norm": 6.365619118475428,
+      "learning_rate": 2e-05,
+      "loss": 1.0713,
+      "step": 30673
+    },
+    {
+      "epoch": 0.7301247262686852,
+      "grad_norm": 2.3237473246971168,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 30674
+    },
+    {
+      "epoch": 0.7301485289917167,
+      "grad_norm": 1.3032073802355344,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 30675
+    },
+    {
+      "epoch": 0.7301723317147482,
+      "grad_norm": 3.539037601198222,
+      "learning_rate": 2e-05,
+      "loss": 0.4681,
+      "step": 30676
+    },
+    {
+      "epoch": 0.7301961344377796,
+      "grad_norm": 0.1442521524572666,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 30677
+    },
+    {
+      "epoch": 0.7302199371608112,
+      "grad_norm": 2.1414095845754946,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 30678
+    },
+    {
+      "epoch": 0.7302437398838427,
+      "grad_norm": 1.588118257029387,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 30679
+    },
+    {
+      "epoch": 0.7302675426068742,
+      "grad_norm": 9.714794247272872,
+      "learning_rate": 2e-05,
+      "loss": 0.8271,
+      "step": 30680
+    },
+    {
+      "epoch": 0.7302913453299057,
+      "grad_norm": 2.1689967585631558,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 30681
+    },
+    {
+      "epoch": 0.7303151480529373,
+      "grad_norm": 4.759773265223254,
+      "learning_rate": 2e-05,
+      "loss": 1.529,
+      "step": 30682
+    },
+    {
+      "epoch": 0.7303389507759688,
+      "grad_norm": 2.7525145340094523,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 30683
+    },
+    {
+      "epoch": 0.7303627534990003,
+      "grad_norm": 2.716290835607361,
+      "learning_rate": 2e-05,
+      "loss": 0.1978,
+      "step": 30684
+    },
+    {
+      "epoch": 0.7303865562220317,
+      "grad_norm": 26.4219870171261,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 30685
+    },
+    {
+      "epoch": 0.7304103589450633,
+      "grad_norm": 1.1264603688667183,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 30686
+    },
+    {
+      "epoch": 0.7304341616680948,
+      "grad_norm": 1.1925381358274336,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 30687
+    },
+    {
+      "epoch": 0.7304579643911263,
+      "grad_norm": 6.480425316415804,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 30688
+    },
+    {
+      "epoch": 0.7304817671141579,
+      "grad_norm": 3.246164718138783,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 30689
+    },
+    {
+      "epoch": 0.7305055698371894,
+      "grad_norm": 4.013038563312576,
+      "learning_rate": 2e-05,
+      "loss": 0.5641,
+      "step": 30690
+    },
+    {
+      "epoch": 0.7305293725602209,
+      "grad_norm": 4.313255371804291,
+      "learning_rate": 2e-05,
+      "loss": 1.2196,
+      "step": 30691
+    },
+    {
+      "epoch": 0.7305531752832524,
+      "grad_norm": 4.651002535198958,
+      "learning_rate": 2e-05,
+      "loss": 0.6273,
+      "step": 30692
+    },
+    {
+      "epoch": 0.730576978006284,
+      "grad_norm": 7.184660486420253,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 30693
+    },
+    {
+      "epoch": 0.7306007807293154,
+      "grad_norm": 10.353260017313817,
+      "learning_rate": 2e-05,
+      "loss": 1.271,
+      "step": 30694
+    },
+    {
+      "epoch": 0.7306245834523469,
+      "grad_norm": 6.227665965327851,
+      "learning_rate": 2e-05,
+      "loss": 0.8917,
+      "step": 30695
+    },
+    {
+      "epoch": 0.7306483861753784,
+      "grad_norm": 4.141770032404356,
+      "learning_rate": 2e-05,
+      "loss": 0.1818,
+      "step": 30696
+    },
+    {
+      "epoch": 0.73067218889841,
+      "grad_norm": 4.145053730409971,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 30697
+    },
+    {
+      "epoch": 0.7306959916214415,
+      "grad_norm": 52.429820471185174,
+      "learning_rate": 2e-05,
+      "loss": 0.1858,
+      "step": 30698
+    },
+    {
+      "epoch": 0.730719794344473,
+      "grad_norm": 1.120552544903034,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 30699
+    },
+    {
+      "epoch": 0.7307435970675045,
+      "grad_norm": 4.605139759463259,
+      "learning_rate": 2e-05,
+      "loss": 0.8505,
+      "step": 30700
+    },
+    {
+      "epoch": 0.7307673997905361,
+      "grad_norm": 13.361482899336538,
+      "learning_rate": 2e-05,
+      "loss": 0.7669,
+      "step": 30701
+    },
+    {
+      "epoch": 0.7307912025135676,
+      "grad_norm": 0.4762694576197957,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 30702
+    },
+    {
+      "epoch": 0.730815005236599,
+      "grad_norm": 1.0029950803451408,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 30703
+    },
+    {
+      "epoch": 0.7308388079596306,
+      "grad_norm": 1.6979320640928148,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 30704
+    },
+    {
+      "epoch": 0.7308626106826621,
+      "grad_norm": 1.2350109060945558,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 30705
+    },
+    {
+      "epoch": 0.7308864134056936,
+      "grad_norm": 1.9732266384811885,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 30706
+    },
+    {
+      "epoch": 0.7309102161287251,
+      "grad_norm": 3.246958580571502,
+      "learning_rate": 2e-05,
+      "loss": 0.3193,
+      "step": 30707
+    },
+    {
+      "epoch": 0.7309340188517567,
+      "grad_norm": 14.309442752469385,
+      "learning_rate": 2e-05,
+      "loss": 0.4327,
+      "step": 30708
+    },
+    {
+      "epoch": 0.7309578215747882,
+      "grad_norm": 1.318262262288033,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 30709
+    },
+    {
+      "epoch": 0.7309816242978197,
+      "grad_norm": 1.015981754167905,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 30710
+    },
+    {
+      "epoch": 0.7310054270208511,
+      "grad_norm": 6.061497121012972,
+      "learning_rate": 2e-05,
+      "loss": 0.4761,
+      "step": 30711
+    },
+    {
+      "epoch": 0.7310292297438827,
+      "grad_norm": 4.893864253325858,
+      "learning_rate": 2e-05,
+      "loss": 0.5134,
+      "step": 30712
+    },
+    {
+      "epoch": 0.7310530324669142,
+      "grad_norm": 0.7786098171688556,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 30713
+    },
+    {
+      "epoch": 0.7310768351899457,
+      "grad_norm": 5.600698947200602,
+      "learning_rate": 2e-05,
+      "loss": 0.6927,
+      "step": 30714
+    },
+    {
+      "epoch": 0.7311006379129772,
+      "grad_norm": 6.72012712393115,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 30715
+    },
+    {
+      "epoch": 0.7311244406360088,
+      "grad_norm": 10.70894850310212,
+      "learning_rate": 2e-05,
+      "loss": 0.5222,
+      "step": 30716
+    },
+    {
+      "epoch": 0.7311482433590403,
+      "grad_norm": 11.882241449920445,
+      "learning_rate": 2e-05,
+      "loss": 0.2519,
+      "step": 30717
+    },
+    {
+      "epoch": 0.7311720460820718,
+      "grad_norm": 0.29740672632797316,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 30718
+    },
+    {
+      "epoch": 0.7311958488051034,
+      "grad_norm": 0.35327744668269234,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 30719
+    },
+    {
+      "epoch": 0.7312196515281348,
+      "grad_norm": 4.845522223152002,
+      "learning_rate": 2e-05,
+      "loss": 0.391,
+      "step": 30720
+    },
+    {
+      "epoch": 0.7312434542511663,
+      "grad_norm": 0.7238606898196593,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 30721
+    },
+    {
+      "epoch": 0.7312672569741978,
+      "grad_norm": 3.6187344879294683,
+      "learning_rate": 2e-05,
+      "loss": 0.2094,
+      "step": 30722
+    },
+    {
+      "epoch": 0.7312910596972294,
+      "grad_norm": 3.6500773269959437,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 30723
+    },
+    {
+      "epoch": 0.7313148624202609,
+      "grad_norm": 6.511426958617574,
+      "learning_rate": 2e-05,
+      "loss": 0.9541,
+      "step": 30724
+    },
+    {
+      "epoch": 0.7313386651432924,
+      "grad_norm": 5.646634950679921,
+      "learning_rate": 2e-05,
+      "loss": 1.3121,
+      "step": 30725
+    },
+    {
+      "epoch": 0.7313624678663239,
+      "grad_norm": 7.79682383585798,
+      "learning_rate": 2e-05,
+      "loss": 0.6801,
+      "step": 30726
+    },
+    {
+      "epoch": 0.7313862705893555,
+      "grad_norm": 0.3231439085470359,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 30727
+    },
+    {
+      "epoch": 0.7314100733123869,
+      "grad_norm": 4.945977246268917,
+      "learning_rate": 2e-05,
+      "loss": 0.7001,
+      "step": 30728
+    },
+    {
+      "epoch": 0.7314338760354184,
+      "grad_norm": 3.1316398637742497,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 30729
+    },
+    {
+      "epoch": 0.7314576787584499,
+      "grad_norm": 3.5281900256375125,
+      "learning_rate": 2e-05,
+      "loss": 0.3262,
+      "step": 30730
+    },
+    {
+      "epoch": 0.7314814814814815,
+      "grad_norm": 0.1841990012585428,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 30731
+    },
+    {
+      "epoch": 0.731505284204513,
+      "grad_norm": 7.486124364284883,
+      "learning_rate": 2e-05,
+      "loss": 0.685,
+      "step": 30732
+    },
+    {
+      "epoch": 0.7315290869275445,
+      "grad_norm": 2.3885136913392064,
+      "learning_rate": 2e-05,
+      "loss": 0.3018,
+      "step": 30733
+    },
+    {
+      "epoch": 0.7315528896505761,
+      "grad_norm": 0.07441540803210889,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30734
+    },
+    {
+      "epoch": 0.7315766923736076,
+      "grad_norm": 5.399661061179358,
+      "learning_rate": 2e-05,
+      "loss": 1.2682,
+      "step": 30735
+    },
+    {
+      "epoch": 0.731600495096639,
+      "grad_norm": 11.102589348891712,
+      "learning_rate": 2e-05,
+      "loss": 0.5534,
+      "step": 30736
+    },
+    {
+      "epoch": 0.7316242978196705,
+      "grad_norm": 0.11058386688840677,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 30737
+    },
+    {
+      "epoch": 0.7316481005427021,
+      "grad_norm": 4.866222736543241,
+      "learning_rate": 2e-05,
+      "loss": 0.6911,
+      "step": 30738
+    },
+    {
+      "epoch": 0.7316719032657336,
+      "grad_norm": 3.56740774078623,
+      "learning_rate": 2e-05,
+      "loss": 0.6552,
+      "step": 30739
+    },
+    {
+      "epoch": 0.7316957059887651,
+      "grad_norm": 0.41006686835639944,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 30740
+    },
+    {
+      "epoch": 0.7317195087117966,
+      "grad_norm": 11.789774929739893,
+      "learning_rate": 2e-05,
+      "loss": 0.423,
+      "step": 30741
+    },
+    {
+      "epoch": 0.7317433114348282,
+      "grad_norm": 5.593982034538269,
+      "learning_rate": 2e-05,
+      "loss": 1.13,
+      "step": 30742
+    },
+    {
+      "epoch": 0.7317671141578597,
+      "grad_norm": 5.620214084136342,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 30743
+    },
+    {
+      "epoch": 0.7317909168808912,
+      "grad_norm": 0.07584367660459719,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 30744
+    },
+    {
+      "epoch": 0.7318147196039226,
+      "grad_norm": 10.309637802229236,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 30745
+    },
+    {
+      "epoch": 0.7318385223269542,
+      "grad_norm": 9.044534015277717,
+      "learning_rate": 2e-05,
+      "loss": 0.4109,
+      "step": 30746
+    },
+    {
+      "epoch": 0.7318623250499857,
+      "grad_norm": 0.3584284733100997,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 30747
+    },
+    {
+      "epoch": 0.7318861277730172,
+      "grad_norm": 9.32130663805188,
+      "learning_rate": 2e-05,
+      "loss": 0.4029,
+      "step": 30748
+    },
+    {
+      "epoch": 0.7319099304960488,
+      "grad_norm": 3.0201745944500344,
+      "learning_rate": 2e-05,
+      "loss": 0.7189,
+      "step": 30749
+    },
+    {
+      "epoch": 0.7319337332190803,
+      "grad_norm": 6.462091833325151,
+      "learning_rate": 2e-05,
+      "loss": 0.7977,
+      "step": 30750
+    },
+    {
+      "epoch": 0.7319575359421118,
+      "grad_norm": 4.22159190647225,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 30751
+    },
+    {
+      "epoch": 0.7319813386651433,
+      "grad_norm": 5.487993120386855,
+      "learning_rate": 2e-05,
+      "loss": 1.061,
+      "step": 30752
+    },
+    {
+      "epoch": 0.7320051413881749,
+      "grad_norm": 2.857384020191168,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 30753
+    },
+    {
+      "epoch": 0.7320289441112063,
+      "grad_norm": 3.232453173718873,
+      "learning_rate": 2e-05,
+      "loss": 0.3921,
+      "step": 30754
+    },
+    {
+      "epoch": 0.7320527468342378,
+      "grad_norm": 2.001647245763846,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 30755
+    },
+    {
+      "epoch": 0.7320765495572693,
+      "grad_norm": 7.175972998820775,
+      "learning_rate": 2e-05,
+      "loss": 0.7954,
+      "step": 30756
+    },
+    {
+      "epoch": 0.7321003522803009,
+      "grad_norm": 3.8731953569088975,
+      "learning_rate": 2e-05,
+      "loss": 0.6027,
+      "step": 30757
+    },
+    {
+      "epoch": 0.7321241550033324,
+      "grad_norm": 8.27432281848606,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 30758
+    },
+    {
+      "epoch": 0.7321479577263639,
+      "grad_norm": 4.1348988029312235,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 30759
+    },
+    {
+      "epoch": 0.7321717604493954,
+      "grad_norm": 3.1034729529615066,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 30760
+    },
+    {
+      "epoch": 0.732195563172427,
+      "grad_norm": 7.098918909921821,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 30761
+    },
+    {
+      "epoch": 0.7322193658954584,
+      "grad_norm": 4.765546857711392,
+      "learning_rate": 2e-05,
+      "loss": 0.6491,
+      "step": 30762
+    },
+    {
+      "epoch": 0.7322431686184899,
+      "grad_norm": 2.8561164136109056,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 30763
+    },
+    {
+      "epoch": 0.7322669713415215,
+      "grad_norm": 1.5267096775931606,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 30764
+    },
+    {
+      "epoch": 0.732290774064553,
+      "grad_norm": 2.412294195132624,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 30765
+    },
+    {
+      "epoch": 0.7323145767875845,
+      "grad_norm": 0.2125535319946734,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 30766
+    },
+    {
+      "epoch": 0.732338379510616,
+      "grad_norm": 0.7750408890723126,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 30767
+    },
+    {
+      "epoch": 0.7323621822336476,
+      "grad_norm": 3.294171486872317,
+      "learning_rate": 2e-05,
+      "loss": 0.588,
+      "step": 30768
+    },
+    {
+      "epoch": 0.7323859849566791,
+      "grad_norm": 0.929777722253386,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 30769
+    },
+    {
+      "epoch": 0.7324097876797105,
+      "grad_norm": 1.3533537821595019,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 30770
+    },
+    {
+      "epoch": 0.732433590402742,
+      "grad_norm": 2.5834292009581072,
+      "learning_rate": 2e-05,
+      "loss": 0.557,
+      "step": 30771
+    },
+    {
+      "epoch": 0.7324573931257736,
+      "grad_norm": 2.8035478345099323,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 30772
+    },
+    {
+      "epoch": 0.7324811958488051,
+      "grad_norm": 4.0592235955146005,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 30773
+    },
+    {
+      "epoch": 0.7325049985718366,
+      "grad_norm": 4.286206360807654,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 30774
+    },
+    {
+      "epoch": 0.7325288012948681,
+      "grad_norm": 4.730550900378767,
+      "learning_rate": 2e-05,
+      "loss": 0.8495,
+      "step": 30775
+    },
+    {
+      "epoch": 0.7325526040178997,
+      "grad_norm": 0.36290155935451723,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 30776
+    },
+    {
+      "epoch": 0.7325764067409312,
+      "grad_norm": 2.5697569406852323,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 30777
+    },
+    {
+      "epoch": 0.7326002094639626,
+      "grad_norm": 0.7261753254346959,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 30778
+    },
+    {
+      "epoch": 0.7326240121869942,
+      "grad_norm": 1.8863969134771204,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 30779
+    },
+    {
+      "epoch": 0.7326478149100257,
+      "grad_norm": 5.263702261187173,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 30780
+    },
+    {
+      "epoch": 0.7326716176330572,
+      "grad_norm": 1.9005316373398953,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 30781
+    },
+    {
+      "epoch": 0.7326954203560887,
+      "grad_norm": 5.403467369249728,
+      "learning_rate": 2e-05,
+      "loss": 1.1781,
+      "step": 30782
+    },
+    {
+      "epoch": 0.7327192230791203,
+      "grad_norm": 0.9898713572076404,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 30783
+    },
+    {
+      "epoch": 0.7327430258021518,
+      "grad_norm": 11.847675362122759,
+      "learning_rate": 2e-05,
+      "loss": 0.3046,
+      "step": 30784
+    },
+    {
+      "epoch": 0.7327668285251833,
+      "grad_norm": 3.1471138027959102,
+      "learning_rate": 2e-05,
+      "loss": 0.7592,
+      "step": 30785
+    },
+    {
+      "epoch": 0.7327906312482148,
+      "grad_norm": 3.7315985567934216,
+      "learning_rate": 2e-05,
+      "loss": 0.5944,
+      "step": 30786
+    },
+    {
+      "epoch": 0.7328144339712463,
+      "grad_norm": 0.4554704936922113,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 30787
+    },
+    {
+      "epoch": 0.7328382366942778,
+      "grad_norm": 7.834213780940928,
+      "learning_rate": 2e-05,
+      "loss": 0.7695,
+      "step": 30788
+    },
+    {
+      "epoch": 0.7328620394173093,
+      "grad_norm": 0.026802996666792107,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 30789
+    },
+    {
+      "epoch": 0.7328858421403408,
+      "grad_norm": 6.5402013910440555,
+      "learning_rate": 2e-05,
+      "loss": 1.6167,
+      "step": 30790
+    },
+    {
+      "epoch": 0.7329096448633724,
+      "grad_norm": 10.138181805473605,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 30791
+    },
+    {
+      "epoch": 0.7329334475864039,
+      "grad_norm": 1.9874660342888835,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 30792
+    },
+    {
+      "epoch": 0.7329572503094354,
+      "grad_norm": 2.688932069536963,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 30793
+    },
+    {
+      "epoch": 0.732981053032467,
+      "grad_norm": 9.936356979271839,
+      "learning_rate": 2e-05,
+      "loss": 1.0289,
+      "step": 30794
+    },
+    {
+      "epoch": 0.7330048557554985,
+      "grad_norm": 14.418866514204295,
+      "learning_rate": 2e-05,
+      "loss": 0.6087,
+      "step": 30795
+    },
+    {
+      "epoch": 0.7330286584785299,
+      "grad_norm": 3.6951549770565437,
+      "learning_rate": 2e-05,
+      "loss": 0.6252,
+      "step": 30796
+    },
+    {
+      "epoch": 0.7330524612015614,
+      "grad_norm": 1.2994746549926726,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 30797
+    },
+    {
+      "epoch": 0.733076263924593,
+      "grad_norm": 8.969743342218228,
+      "learning_rate": 2e-05,
+      "loss": 0.1442,
+      "step": 30798
+    },
+    {
+      "epoch": 0.7331000666476245,
+      "grad_norm": 1.783818000023598,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 30799
+    },
+    {
+      "epoch": 0.733123869370656,
+      "grad_norm": 15.924466571555728,
+      "learning_rate": 2e-05,
+      "loss": 0.5028,
+      "step": 30800
+    },
+    {
+      "epoch": 0.7331476720936875,
+      "grad_norm": 4.055139676187804,
+      "learning_rate": 2e-05,
+      "loss": 0.768,
+      "step": 30801
+    },
+    {
+      "epoch": 0.7331714748167191,
+      "grad_norm": 4.047753835559118,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 30802
+    },
+    {
+      "epoch": 0.7331952775397506,
+      "grad_norm": 1.6769193698461755,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 30803
+    },
+    {
+      "epoch": 0.733219080262782,
+      "grad_norm": 2.589848674655945,
+      "learning_rate": 2e-05,
+      "loss": 0.1046,
+      "step": 30804
+    },
+    {
+      "epoch": 0.7332428829858135,
+      "grad_norm": 3.74968668731529,
+      "learning_rate": 2e-05,
+      "loss": 0.2426,
+      "step": 30805
+    },
+    {
+      "epoch": 0.7332666857088451,
+      "grad_norm": 9.566909303319628,
+      "learning_rate": 2e-05,
+      "loss": 0.6728,
+      "step": 30806
+    },
+    {
+      "epoch": 0.7332904884318766,
+      "grad_norm": 0.19416450122453055,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 30807
+    },
+    {
+      "epoch": 0.7333142911549081,
+      "grad_norm": 3.1375762716899196,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 30808
+    },
+    {
+      "epoch": 0.7333380938779397,
+      "grad_norm": 10.957662846535754,
+      "learning_rate": 2e-05,
+      "loss": 0.9093,
+      "step": 30809
+    },
+    {
+      "epoch": 0.7333618966009712,
+      "grad_norm": 8.66145609648408,
+      "learning_rate": 2e-05,
+      "loss": 1.6791,
+      "step": 30810
+    },
+    {
+      "epoch": 0.7333856993240027,
+      "grad_norm": 2.852888688981283,
+      "learning_rate": 2e-05,
+      "loss": 0.5353,
+      "step": 30811
+    },
+    {
+      "epoch": 0.7334095020470341,
+      "grad_norm": 3.761572770415945,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 30812
+    },
+    {
+      "epoch": 0.7334333047700657,
+      "grad_norm": 6.4112823732750845,
+      "learning_rate": 2e-05,
+      "loss": 0.1453,
+      "step": 30813
+    },
+    {
+      "epoch": 0.7334571074930972,
+      "grad_norm": 0.09182082744188456,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 30814
+    },
+    {
+      "epoch": 0.7334809102161287,
+      "grad_norm": 4.549799607469357,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 30815
+    },
+    {
+      "epoch": 0.7335047129391602,
+      "grad_norm": 4.284327923494987,
+      "learning_rate": 2e-05,
+      "loss": 0.7069,
+      "step": 30816
+    },
+    {
+      "epoch": 0.7335285156621918,
+      "grad_norm": 0.98764706651321,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 30817
+    },
+    {
+      "epoch": 0.7335523183852233,
+      "grad_norm": 15.59841151085071,
+      "learning_rate": 2e-05,
+      "loss": 0.4943,
+      "step": 30818
+    },
+    {
+      "epoch": 0.7335761211082548,
+      "grad_norm": 1.1704107343537506,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 30819
+    },
+    {
+      "epoch": 0.7335999238312862,
+      "grad_norm": 6.523671175499494,
+      "learning_rate": 2e-05,
+      "loss": 0.3765,
+      "step": 30820
+    },
+    {
+      "epoch": 0.7336237265543178,
+      "grad_norm": 5.92889522149889,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 30821
+    },
+    {
+      "epoch": 0.7336475292773493,
+      "grad_norm": 4.296900553377269,
+      "learning_rate": 2e-05,
+      "loss": 0.4645,
+      "step": 30822
+    },
+    {
+      "epoch": 0.7336713320003808,
+      "grad_norm": 4.1615797444556355,
+      "learning_rate": 2e-05,
+      "loss": 0.7166,
+      "step": 30823
+    },
+    {
+      "epoch": 0.7336951347234124,
+      "grad_norm": 4.000469672190089,
+      "learning_rate": 2e-05,
+      "loss": 0.779,
+      "step": 30824
+    },
+    {
+      "epoch": 0.7337189374464439,
+      "grad_norm": 1.0600168417559679,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 30825
+    },
+    {
+      "epoch": 0.7337427401694754,
+      "grad_norm": 0.0826913958411847,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 30826
+    },
+    {
+      "epoch": 0.7337665428925069,
+      "grad_norm": 4.4812852884322085,
+      "learning_rate": 2e-05,
+      "loss": 0.643,
+      "step": 30827
+    },
+    {
+      "epoch": 0.7337903456155385,
+      "grad_norm": 3.7755924480254652,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 30828
+    },
+    {
+      "epoch": 0.73381414833857,
+      "grad_norm": 3.342037481105176,
+      "learning_rate": 2e-05,
+      "loss": 0.6933,
+      "step": 30829
+    },
+    {
+      "epoch": 0.7338379510616014,
+      "grad_norm": 3.660773198865072,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 30830
+    },
+    {
+      "epoch": 0.7338617537846329,
+      "grad_norm": 4.688039933679858,
+      "learning_rate": 2e-05,
+      "loss": 1.1738,
+      "step": 30831
+    },
+    {
+      "epoch": 0.7338855565076645,
+      "grad_norm": 6.209112670148349,
+      "learning_rate": 2e-05,
+      "loss": 0.7967,
+      "step": 30832
+    },
+    {
+      "epoch": 0.733909359230696,
+      "grad_norm": 0.8954104216683157,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 30833
+    },
+    {
+      "epoch": 0.7339331619537275,
+      "grad_norm": 7.532001353864418,
+      "learning_rate": 2e-05,
+      "loss": 1.1142,
+      "step": 30834
+    },
+    {
+      "epoch": 0.733956964676759,
+      "grad_norm": 4.1990688515765555,
+      "learning_rate": 2e-05,
+      "loss": 0.7097,
+      "step": 30835
+    },
+    {
+      "epoch": 0.7339807673997906,
+      "grad_norm": 5.831945479908479,
+      "learning_rate": 2e-05,
+      "loss": 0.8298,
+      "step": 30836
+    },
+    {
+      "epoch": 0.734004570122822,
+      "grad_norm": 0.18688528591167278,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 30837
+    },
+    {
+      "epoch": 0.7340283728458535,
+      "grad_norm": 7.5600388065051165,
+      "learning_rate": 2e-05,
+      "loss": 1.5338,
+      "step": 30838
+    },
+    {
+      "epoch": 0.7340521755688851,
+      "grad_norm": 52.56219213031403,
+      "learning_rate": 2e-05,
+      "loss": 1.9486,
+      "step": 30839
+    },
+    {
+      "epoch": 0.7340759782919166,
+      "grad_norm": 9.682065421199006,
+      "learning_rate": 2e-05,
+      "loss": 0.1686,
+      "step": 30840
+    },
+    {
+      "epoch": 0.7340997810149481,
+      "grad_norm": 2.7004062339114543,
+      "learning_rate": 2e-05,
+      "loss": 0.4744,
+      "step": 30841
+    },
+    {
+      "epoch": 0.7341235837379796,
+      "grad_norm": 0.19001212394841044,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 30842
+    },
+    {
+      "epoch": 0.7341473864610112,
+      "grad_norm": 0.1581561837773875,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 30843
+    },
+    {
+      "epoch": 0.7341711891840427,
+      "grad_norm": 3.178666312088045,
+      "learning_rate": 2e-05,
+      "loss": 0.6643,
+      "step": 30844
+    },
+    {
+      "epoch": 0.7341949919070742,
+      "grad_norm": 8.678675755197863,
+      "learning_rate": 2e-05,
+      "loss": 1.1624,
+      "step": 30845
+    },
+    {
+      "epoch": 0.7342187946301056,
+      "grad_norm": 3.7119715593502587,
+      "learning_rate": 2e-05,
+      "loss": 0.4879,
+      "step": 30846
+    },
+    {
+      "epoch": 0.7342425973531372,
+      "grad_norm": 0.2968138059708106,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 30847
+    },
+    {
+      "epoch": 0.7342664000761687,
+      "grad_norm": 8.148391830950159,
+      "learning_rate": 2e-05,
+      "loss": 0.3358,
+      "step": 30848
+    },
+    {
+      "epoch": 0.7342902027992002,
+      "grad_norm": 19.965408749341783,
+      "learning_rate": 2e-05,
+      "loss": 0.9596,
+      "step": 30849
+    },
+    {
+      "epoch": 0.7343140055222317,
+      "grad_norm": 7.377963939259156,
+      "learning_rate": 2e-05,
+      "loss": 0.1557,
+      "step": 30850
+    },
+    {
+      "epoch": 0.7343378082452633,
+      "grad_norm": 0.4157410663026575,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 30851
+    },
+    {
+      "epoch": 0.7343616109682948,
+      "grad_norm": 0.14267423358977735,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 30852
+    },
+    {
+      "epoch": 0.7343854136913263,
+      "grad_norm": 6.3864803072073055,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 30853
+    },
+    {
+      "epoch": 0.7344092164143579,
+      "grad_norm": 1.66016312815872,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 30854
+    },
+    {
+      "epoch": 0.7344330191373893,
+      "grad_norm": 0.04758476728664176,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 30855
+    },
+    {
+      "epoch": 0.7344568218604208,
+      "grad_norm": 3.7222953296225993,
+      "learning_rate": 2e-05,
+      "loss": 0.4303,
+      "step": 30856
+    },
+    {
+      "epoch": 0.7344806245834523,
+      "grad_norm": 0.5214918686117668,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 30857
+    },
+    {
+      "epoch": 0.7345044273064839,
+      "grad_norm": 3.8789999798082415,
+      "learning_rate": 2e-05,
+      "loss": 0.451,
+      "step": 30858
+    },
+    {
+      "epoch": 0.7345282300295154,
+      "grad_norm": 12.74644690877491,
+      "learning_rate": 2e-05,
+      "loss": 0.7495,
+      "step": 30859
+    },
+    {
+      "epoch": 0.7345520327525469,
+      "grad_norm": 4.708383415561762,
+      "learning_rate": 2e-05,
+      "loss": 0.6903,
+      "step": 30860
+    },
+    {
+      "epoch": 0.7345758354755784,
+      "grad_norm": 3.1057931809380723,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 30861
+    },
+    {
+      "epoch": 0.73459963819861,
+      "grad_norm": 5.029154485562137,
+      "learning_rate": 2e-05,
+      "loss": 0.9475,
+      "step": 30862
+    },
+    {
+      "epoch": 0.7346234409216414,
+      "grad_norm": 0.22866560401878433,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 30863
+    },
+    {
+      "epoch": 0.7346472436446729,
+      "grad_norm": 1.4611714669009532,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 30864
+    },
+    {
+      "epoch": 0.7346710463677044,
+      "grad_norm": 2.5466852557491646,
+      "learning_rate": 2e-05,
+      "loss": 0.3011,
+      "step": 30865
+    },
+    {
+      "epoch": 0.734694849090736,
+      "grad_norm": 0.573136453534983,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 30866
+    },
+    {
+      "epoch": 0.7347186518137675,
+      "grad_norm": 0.13338088337365794,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 30867
+    },
+    {
+      "epoch": 0.734742454536799,
+      "grad_norm": 6.330732037672676,
+      "learning_rate": 2e-05,
+      "loss": 2.1287,
+      "step": 30868
+    },
+    {
+      "epoch": 0.7347662572598306,
+      "grad_norm": 2.8785539589840226,
+      "learning_rate": 2e-05,
+      "loss": 0.2242,
+      "step": 30869
+    },
+    {
+      "epoch": 0.7347900599828621,
+      "grad_norm": 0.17784452907681336,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 30870
+    },
+    {
+      "epoch": 0.7348138627058936,
+      "grad_norm": 0.020811028710042787,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 30871
+    },
+    {
+      "epoch": 0.734837665428925,
+      "grad_norm": 0.39624256982798545,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 30872
+    },
+    {
+      "epoch": 0.7348614681519566,
+      "grad_norm": 2.214534692598194,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 30873
+    },
+    {
+      "epoch": 0.7348852708749881,
+      "grad_norm": 4.07857837885992,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 30874
+    },
+    {
+      "epoch": 0.7349090735980196,
+      "grad_norm": 6.194583265148396,
+      "learning_rate": 2e-05,
+      "loss": 1.1624,
+      "step": 30875
+    },
+    {
+      "epoch": 0.7349328763210511,
+      "grad_norm": 0.8291360871490304,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 30876
+    },
+    {
+      "epoch": 0.7349566790440827,
+      "grad_norm": 4.838403320423874,
+      "learning_rate": 2e-05,
+      "loss": 0.5424,
+      "step": 30877
+    },
+    {
+      "epoch": 0.7349804817671142,
+      "grad_norm": 1.719186397613127,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 30878
+    },
+    {
+      "epoch": 0.7350042844901457,
+      "grad_norm": 3.4935584418176013,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 30879
+    },
+    {
+      "epoch": 0.7350280872131771,
+      "grad_norm": 3.868557858224985,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 30880
+    },
+    {
+      "epoch": 0.7350518899362087,
+      "grad_norm": 6.32969720997347,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 30881
+    },
+    {
+      "epoch": 0.7350756926592402,
+      "grad_norm": 2.9456537544657575,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 30882
+    },
+    {
+      "epoch": 0.7350994953822717,
+      "grad_norm": 0.5202391710263741,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 30883
+    },
+    {
+      "epoch": 0.7351232981053032,
+      "grad_norm": 12.153925947841485,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 30884
+    },
+    {
+      "epoch": 0.7351471008283348,
+      "grad_norm": 0.45457751198189017,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 30885
+    },
+    {
+      "epoch": 0.7351709035513663,
+      "grad_norm": 1.5943639402430352,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 30886
+    },
+    {
+      "epoch": 0.7351947062743978,
+      "grad_norm": 10.054972231683948,
+      "learning_rate": 2e-05,
+      "loss": 0.6682,
+      "step": 30887
+    },
+    {
+      "epoch": 0.7352185089974294,
+      "grad_norm": 1.0555939819956188,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 30888
+    },
+    {
+      "epoch": 0.7352423117204608,
+      "grad_norm": 14.626794331217353,
+      "learning_rate": 2e-05,
+      "loss": 1.0163,
+      "step": 30889
+    },
+    {
+      "epoch": 0.7352661144434923,
+      "grad_norm": 2.581159667075388,
+      "learning_rate": 2e-05,
+      "loss": 0.2957,
+      "step": 30890
+    },
+    {
+      "epoch": 0.7352899171665238,
+      "grad_norm": 5.287676123017034,
+      "learning_rate": 2e-05,
+      "loss": 1.0023,
+      "step": 30891
+    },
+    {
+      "epoch": 0.7353137198895554,
+      "grad_norm": 48.3242289586232,
+      "learning_rate": 2e-05,
+      "loss": 0.2621,
+      "step": 30892
+    },
+    {
+      "epoch": 0.7353375226125869,
+      "grad_norm": 3.790316029250703,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 30893
+    },
+    {
+      "epoch": 0.7353613253356184,
+      "grad_norm": 0.5871088776072435,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 30894
+    },
+    {
+      "epoch": 0.7353851280586499,
+      "grad_norm": 10.138653586187138,
+      "learning_rate": 2e-05,
+      "loss": 0.4758,
+      "step": 30895
+    },
+    {
+      "epoch": 0.7354089307816815,
+      "grad_norm": 5.892717358855742,
+      "learning_rate": 2e-05,
+      "loss": 1.0317,
+      "step": 30896
+    },
+    {
+      "epoch": 0.7354327335047129,
+      "grad_norm": 9.862882566019588,
+      "learning_rate": 2e-05,
+      "loss": 0.5819,
+      "step": 30897
+    },
+    {
+      "epoch": 0.7354565362277444,
+      "grad_norm": 1.0966858998682334,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 30898
+    },
+    {
+      "epoch": 0.7354803389507759,
+      "grad_norm": 3.0520053105061047,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 30899
+    },
+    {
+      "epoch": 0.7355041416738075,
+      "grad_norm": 8.713308952570413,
+      "learning_rate": 2e-05,
+      "loss": 0.7409,
+      "step": 30900
+    },
+    {
+      "epoch": 0.735527944396839,
+      "grad_norm": 5.048315236589704,
+      "learning_rate": 2e-05,
+      "loss": 0.5151,
+      "step": 30901
+    },
+    {
+      "epoch": 0.7355517471198705,
+      "grad_norm": 9.9702863432085,
+      "learning_rate": 2e-05,
+      "loss": 0.3277,
+      "step": 30902
+    },
+    {
+      "epoch": 0.7355755498429021,
+      "grad_norm": 5.782874132916224,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 30903
+    },
+    {
+      "epoch": 0.7355993525659336,
+      "grad_norm": 6.959528209072708,
+      "learning_rate": 2e-05,
+      "loss": 1.3958,
+      "step": 30904
+    },
+    {
+      "epoch": 0.735623155288965,
+      "grad_norm": 3.1915948123212368,
+      "learning_rate": 2e-05,
+      "loss": 0.4727,
+      "step": 30905
+    },
+    {
+      "epoch": 0.7356469580119965,
+      "grad_norm": 3.2595693087249424,
+      "learning_rate": 2e-05,
+      "loss": 0.6329,
+      "step": 30906
+    },
+    {
+      "epoch": 0.7356707607350281,
+      "grad_norm": 0.574167771575107,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 30907
+    },
+    {
+      "epoch": 0.7356945634580596,
+      "grad_norm": 7.19115543028762,
+      "learning_rate": 2e-05,
+      "loss": 0.2427,
+      "step": 30908
+    },
+    {
+      "epoch": 0.7357183661810911,
+      "grad_norm": 0.3807571472794846,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 30909
+    },
+    {
+      "epoch": 0.7357421689041226,
+      "grad_norm": 10.656694582302372,
+      "learning_rate": 2e-05,
+      "loss": 0.7794,
+      "step": 30910
+    },
+    {
+      "epoch": 0.7357659716271542,
+      "grad_norm": 3.009797736916059,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 30911
+    },
+    {
+      "epoch": 0.7357897743501857,
+      "grad_norm": 4.8452291233482665,
+      "learning_rate": 2e-05,
+      "loss": 0.633,
+      "step": 30912
+    },
+    {
+      "epoch": 0.7358135770732172,
+      "grad_norm": 0.291008418909251,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 30913
+    },
+    {
+      "epoch": 0.7358373797962486,
+      "grad_norm": 3.4570149073277077,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 30914
+    },
+    {
+      "epoch": 0.7358611825192802,
+      "grad_norm": 3.851572543780535,
+      "learning_rate": 2e-05,
+      "loss": 0.493,
+      "step": 30915
+    },
+    {
+      "epoch": 0.7358849852423117,
+      "grad_norm": 3.9405264465838874,
+      "learning_rate": 2e-05,
+      "loss": 0.6102,
+      "step": 30916
+    },
+    {
+      "epoch": 0.7359087879653432,
+      "grad_norm": 6.24254039124734,
+      "learning_rate": 2e-05,
+      "loss": 0.143,
+      "step": 30917
+    },
+    {
+      "epoch": 0.7359325906883748,
+      "grad_norm": 4.686455289909953,
+      "learning_rate": 2e-05,
+      "loss": 0.776,
+      "step": 30918
+    },
+    {
+      "epoch": 0.7359563934114063,
+      "grad_norm": 7.435277516129302,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 30919
+    },
+    {
+      "epoch": 0.7359801961344378,
+      "grad_norm": 4.2210204811012995,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 30920
+    },
+    {
+      "epoch": 0.7360039988574693,
+      "grad_norm": 7.605932480861874,
+      "learning_rate": 2e-05,
+      "loss": 0.8629,
+      "step": 30921
+    },
+    {
+      "epoch": 0.7360278015805009,
+      "grad_norm": 6.4314269176987775,
+      "learning_rate": 2e-05,
+      "loss": 0.8072,
+      "step": 30922
+    },
+    {
+      "epoch": 0.7360516043035323,
+      "grad_norm": 5.851254593239748,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 30923
+    },
+    {
+      "epoch": 0.7360754070265638,
+      "grad_norm": 4.462390256957592,
+      "learning_rate": 2e-05,
+      "loss": 0.3747,
+      "step": 30924
+    },
+    {
+      "epoch": 0.7360992097495953,
+      "grad_norm": 3.703848465306363,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 30925
+    },
+    {
+      "epoch": 0.7361230124726269,
+      "grad_norm": 3.6587046528475886,
+      "learning_rate": 2e-05,
+      "loss": 0.6174,
+      "step": 30926
+    },
+    {
+      "epoch": 0.7361468151956584,
+      "grad_norm": 3.4274658631556005,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 30927
+    },
+    {
+      "epoch": 0.7361706179186899,
+      "grad_norm": 5.983231025794482,
+      "learning_rate": 2e-05,
+      "loss": 1.1277,
+      "step": 30928
+    },
+    {
+      "epoch": 0.7361944206417214,
+      "grad_norm": 4.786671061835429,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 30929
+    },
+    {
+      "epoch": 0.736218223364753,
+      "grad_norm": 4.997856769069752,
+      "learning_rate": 2e-05,
+      "loss": 0.646,
+      "step": 30930
+    },
+    {
+      "epoch": 0.7362420260877844,
+      "grad_norm": 7.180295345960745,
+      "learning_rate": 2e-05,
+      "loss": 0.2578,
+      "step": 30931
+    },
+    {
+      "epoch": 0.7362658288108159,
+      "grad_norm": 3.4752338761071773,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 30932
+    },
+    {
+      "epoch": 0.7362896315338475,
+      "grad_norm": 1.5935280855712528,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 30933
+    },
+    {
+      "epoch": 0.736313434256879,
+      "grad_norm": 2.34729759321519,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 30934
+    },
+    {
+      "epoch": 0.7363372369799105,
+      "grad_norm": 7.916994815752713,
+      "learning_rate": 2e-05,
+      "loss": 0.4082,
+      "step": 30935
+    },
+    {
+      "epoch": 0.736361039702942,
+      "grad_norm": 10.21910516761605,
+      "learning_rate": 2e-05,
+      "loss": 0.3263,
+      "step": 30936
+    },
+    {
+      "epoch": 0.7363848424259736,
+      "grad_norm": 0.04349902873697411,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30937
+    },
+    {
+      "epoch": 0.7364086451490051,
+      "grad_norm": 7.25531902420605,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 30938
+    },
+    {
+      "epoch": 0.7364324478720365,
+      "grad_norm": 3.503435905062482,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 30939
+    },
+    {
+      "epoch": 0.736456250595068,
+      "grad_norm": 0.7822203317898113,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 30940
+    },
+    {
+      "epoch": 0.7364800533180996,
+      "grad_norm": 9.696332111072026,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 30941
+    },
+    {
+      "epoch": 0.7365038560411311,
+      "grad_norm": 3.6623777806353535,
+      "learning_rate": 2e-05,
+      "loss": 0.5549,
+      "step": 30942
+    },
+    {
+      "epoch": 0.7365276587641626,
+      "grad_norm": 6.555478501459155,
+      "learning_rate": 2e-05,
+      "loss": 0.9258,
+      "step": 30943
+    },
+    {
+      "epoch": 0.7365514614871941,
+      "grad_norm": 3.626419495426745,
+      "learning_rate": 2e-05,
+      "loss": 0.5081,
+      "step": 30944
+    },
+    {
+      "epoch": 0.7365752642102257,
+      "grad_norm": 0.07739437107181389,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 30945
+    },
+    {
+      "epoch": 0.7365990669332572,
+      "grad_norm": 0.2867849775580382,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 30946
+    },
+    {
+      "epoch": 0.7366228696562886,
+      "grad_norm": 5.647832553066988,
+      "learning_rate": 2e-05,
+      "loss": 0.5183,
+      "step": 30947
+    },
+    {
+      "epoch": 0.7366466723793202,
+      "grad_norm": 4.926575909218413,
+      "learning_rate": 2e-05,
+      "loss": 0.9762,
+      "step": 30948
+    },
+    {
+      "epoch": 0.7366704751023517,
+      "grad_norm": 1.6602605643548025,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 30949
+    },
+    {
+      "epoch": 0.7366942778253832,
+      "grad_norm": 3.7599214108481105,
+      "learning_rate": 2e-05,
+      "loss": 0.3852,
+      "step": 30950
+    },
+    {
+      "epoch": 0.7367180805484147,
+      "grad_norm": 0.008762945670097928,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 30951
+    },
+    {
+      "epoch": 0.7367418832714463,
+      "grad_norm": 2.9852174915532257,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 30952
+    },
+    {
+      "epoch": 0.7367656859944778,
+      "grad_norm": 0.09259860626233118,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 30953
+    },
+    {
+      "epoch": 0.7367894887175093,
+      "grad_norm": 6.471351096937641,
+      "learning_rate": 2e-05,
+      "loss": 0.5791,
+      "step": 30954
+    },
+    {
+      "epoch": 0.7368132914405408,
+      "grad_norm": 5.312120601314495,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 30955
+    },
+    {
+      "epoch": 0.7368370941635723,
+      "grad_norm": 3.945325782799861,
+      "learning_rate": 2e-05,
+      "loss": 0.7189,
+      "step": 30956
+    },
+    {
+      "epoch": 0.7368608968866038,
+      "grad_norm": 3.712025424163706,
+      "learning_rate": 2e-05,
+      "loss": 0.4535,
+      "step": 30957
+    },
+    {
+      "epoch": 0.7368846996096353,
+      "grad_norm": 3.3546058354175967,
+      "learning_rate": 2e-05,
+      "loss": 0.4301,
+      "step": 30958
+    },
+    {
+      "epoch": 0.7369085023326668,
+      "grad_norm": 6.070067576864599,
+      "learning_rate": 2e-05,
+      "loss": 1.1111,
+      "step": 30959
+    },
+    {
+      "epoch": 0.7369323050556984,
+      "grad_norm": 16.76609739463584,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 30960
+    },
+    {
+      "epoch": 0.7369561077787299,
+      "grad_norm": 0.4135275103590931,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 30961
+    },
+    {
+      "epoch": 0.7369799105017614,
+      "grad_norm": 3.885310157614456,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 30962
+    },
+    {
+      "epoch": 0.737003713224793,
+      "grad_norm": 0.035715738244299536,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 30963
+    },
+    {
+      "epoch": 0.7370275159478245,
+      "grad_norm": 11.462724886998016,
+      "learning_rate": 2e-05,
+      "loss": 0.2252,
+      "step": 30964
+    },
+    {
+      "epoch": 0.7370513186708559,
+      "grad_norm": 1.5135601225542787,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 30965
+    },
+    {
+      "epoch": 0.7370751213938874,
+      "grad_norm": 3.024991197450872,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 30966
+    },
+    {
+      "epoch": 0.737098924116919,
+      "grad_norm": 3.5992770429856926,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 30967
+    },
+    {
+      "epoch": 0.7371227268399505,
+      "grad_norm": 0.06822354354044088,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30968
+    },
+    {
+      "epoch": 0.737146529562982,
+      "grad_norm": 4.1400866342168685,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 30969
+    },
+    {
+      "epoch": 0.7371703322860135,
+      "grad_norm": 5.32437219631727,
+      "learning_rate": 2e-05,
+      "loss": 0.2849,
+      "step": 30970
+    },
+    {
+      "epoch": 0.7371941350090451,
+      "grad_norm": 2.7778550611229855,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 30971
+    },
+    {
+      "epoch": 0.7372179377320766,
+      "grad_norm": 1.7291059550901344,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 30972
+    },
+    {
+      "epoch": 0.737241740455108,
+      "grad_norm": 0.9789949471462113,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 30973
+    },
+    {
+      "epoch": 0.7372655431781395,
+      "grad_norm": 12.242191929964132,
+      "learning_rate": 2e-05,
+      "loss": 0.5827,
+      "step": 30974
+    },
+    {
+      "epoch": 0.7372893459011711,
+      "grad_norm": 10.782294238287388,
+      "learning_rate": 2e-05,
+      "loss": 1.592,
+      "step": 30975
+    },
+    {
+      "epoch": 0.7373131486242026,
+      "grad_norm": 0.1918254545393874,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 30976
+    },
+    {
+      "epoch": 0.7373369513472341,
+      "grad_norm": 0.8912702081296009,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 30977
+    },
+    {
+      "epoch": 0.7373607540702657,
+      "grad_norm": 6.899841250092213,
+      "learning_rate": 2e-05,
+      "loss": 0.4813,
+      "step": 30978
+    },
+    {
+      "epoch": 0.7373845567932972,
+      "grad_norm": 4.439409829609312,
+      "learning_rate": 2e-05,
+      "loss": 0.5265,
+      "step": 30979
+    },
+    {
+      "epoch": 0.7374083595163287,
+      "grad_norm": 1.2728540169309095,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 30980
+    },
+    {
+      "epoch": 0.7374321622393601,
+      "grad_norm": 2.1455331449830015,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 30981
+    },
+    {
+      "epoch": 0.7374559649623917,
+      "grad_norm": 4.900772585896086,
+      "learning_rate": 2e-05,
+      "loss": 0.4998,
+      "step": 30982
+    },
+    {
+      "epoch": 0.7374797676854232,
+      "grad_norm": 6.220132471098696,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 30983
+    },
+    {
+      "epoch": 0.7375035704084547,
+      "grad_norm": 6.034856913713187,
+      "learning_rate": 2e-05,
+      "loss": 0.5957,
+      "step": 30984
+    },
+    {
+      "epoch": 0.7375273731314862,
+      "grad_norm": 1.5732875692915287,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 30985
+    },
+    {
+      "epoch": 0.7375511758545178,
+      "grad_norm": 2.6693359387074684,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 30986
+    },
+    {
+      "epoch": 0.7375749785775493,
+      "grad_norm": 5.311825482703262,
+      "learning_rate": 2e-05,
+      "loss": 0.7908,
+      "step": 30987
+    },
+    {
+      "epoch": 0.7375987813005808,
+      "grad_norm": 5.637032756760947,
+      "learning_rate": 2e-05,
+      "loss": 0.6465,
+      "step": 30988
+    },
+    {
+      "epoch": 0.7376225840236122,
+      "grad_norm": 1.4842771190576327,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 30989
+    },
+    {
+      "epoch": 0.7376463867466438,
+      "grad_norm": 6.08330519181279,
+      "learning_rate": 2e-05,
+      "loss": 0.6324,
+      "step": 30990
+    },
+    {
+      "epoch": 0.7376701894696753,
+      "grad_norm": 0.3255331410307892,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 30991
+    },
+    {
+      "epoch": 0.7376939921927068,
+      "grad_norm": 0.05557804419271526,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 30992
+    },
+    {
+      "epoch": 0.7377177949157384,
+      "grad_norm": 4.298773608578315,
+      "learning_rate": 2e-05,
+      "loss": 0.7686,
+      "step": 30993
+    },
+    {
+      "epoch": 0.7377415976387699,
+      "grad_norm": 4.690054277845206,
+      "learning_rate": 2e-05,
+      "loss": 1.0948,
+      "step": 30994
+    },
+    {
+      "epoch": 0.7377654003618014,
+      "grad_norm": 5.163811886430172,
+      "learning_rate": 2e-05,
+      "loss": 1.4362,
+      "step": 30995
+    },
+    {
+      "epoch": 0.7377892030848329,
+      "grad_norm": 16.80477753185597,
+      "learning_rate": 2e-05,
+      "loss": 0.498,
+      "step": 30996
+    },
+    {
+      "epoch": 0.7378130058078645,
+      "grad_norm": 5.665748139767597,
+      "learning_rate": 2e-05,
+      "loss": 1.2115,
+      "step": 30997
+    },
+    {
+      "epoch": 0.737836808530896,
+      "grad_norm": 4.508617736799464,
+      "learning_rate": 2e-05,
+      "loss": 0.1787,
+      "step": 30998
+    },
+    {
+      "epoch": 0.7378606112539274,
+      "grad_norm": 11.671046253272214,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 30999
+    },
+    {
+      "epoch": 0.7378844139769589,
+      "grad_norm": 4.451406516149099,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 31000
+    },
+    {
+      "epoch": 0.7379082166999905,
+      "grad_norm": 0.34633259952240414,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 31001
+    },
+    {
+      "epoch": 0.737932019423022,
+      "grad_norm": 5.32007830883895,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 31002
+    },
+    {
+      "epoch": 0.7379558221460535,
+      "grad_norm": 1.13955567640028,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 31003
+    },
+    {
+      "epoch": 0.737979624869085,
+      "grad_norm": 2.699964048794699,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 31004
+    },
+    {
+      "epoch": 0.7380034275921166,
+      "grad_norm": 1.1788340023049841,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 31005
+    },
+    {
+      "epoch": 0.738027230315148,
+      "grad_norm": 4.472741036152618,
+      "learning_rate": 2e-05,
+      "loss": 0.4823,
+      "step": 31006
+    },
+    {
+      "epoch": 0.7380510330381795,
+      "grad_norm": 1.54003068455734,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 31007
+    },
+    {
+      "epoch": 0.7380748357612111,
+      "grad_norm": 4.957774574749951,
+      "learning_rate": 2e-05,
+      "loss": 0.6337,
+      "step": 31008
+    },
+    {
+      "epoch": 0.7380986384842426,
+      "grad_norm": 2.9772685166124564,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 31009
+    },
+    {
+      "epoch": 0.7381224412072741,
+      "grad_norm": 6.925376917419283,
+      "learning_rate": 2e-05,
+      "loss": 0.1993,
+      "step": 31010
+    },
+    {
+      "epoch": 0.7381462439303056,
+      "grad_norm": 12.962268541257636,
+      "learning_rate": 2e-05,
+      "loss": 0.5414,
+      "step": 31011
+    },
+    {
+      "epoch": 0.7381700466533372,
+      "grad_norm": 2.839399783232563,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 31012
+    },
+    {
+      "epoch": 0.7381938493763687,
+      "grad_norm": 8.128505488402812,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 31013
+    },
+    {
+      "epoch": 0.7382176520994002,
+      "grad_norm": 0.9863231414183637,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 31014
+    },
+    {
+      "epoch": 0.7382414548224316,
+      "grad_norm": 2.2168504181692317,
+      "learning_rate": 2e-05,
+      "loss": 0.0716,
+      "step": 31015
+    },
+    {
+      "epoch": 0.7382652575454632,
+      "grad_norm": 6.922849605002102,
+      "learning_rate": 2e-05,
+      "loss": 0.4482,
+      "step": 31016
+    },
+    {
+      "epoch": 0.7382890602684947,
+      "grad_norm": 0.11919072224978972,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 31017
+    },
+    {
+      "epoch": 0.7383128629915262,
+      "grad_norm": 9.62960815353421,
+      "learning_rate": 2e-05,
+      "loss": 0.7204,
+      "step": 31018
+    },
+    {
+      "epoch": 0.7383366657145577,
+      "grad_norm": 4.970988032640075,
+      "learning_rate": 2e-05,
+      "loss": 0.177,
+      "step": 31019
+    },
+    {
+      "epoch": 0.7383604684375893,
+      "grad_norm": 5.333768000783786,
+      "learning_rate": 2e-05,
+      "loss": 0.9259,
+      "step": 31020
+    },
+    {
+      "epoch": 0.7383842711606208,
+      "grad_norm": 1.6256324084668545,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 31021
+    },
+    {
+      "epoch": 0.7384080738836523,
+      "grad_norm": 3.625101348908719,
+      "learning_rate": 2e-05,
+      "loss": 0.5165,
+      "step": 31022
+    },
+    {
+      "epoch": 0.7384318766066839,
+      "grad_norm": 6.582547773072361,
+      "learning_rate": 2e-05,
+      "loss": 0.4061,
+      "step": 31023
+    },
+    {
+      "epoch": 0.7384556793297153,
+      "grad_norm": 0.14455906213739148,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 31024
+    },
+    {
+      "epoch": 0.7384794820527468,
+      "grad_norm": 4.04656066211313,
+      "learning_rate": 2e-05,
+      "loss": 0.7639,
+      "step": 31025
+    },
+    {
+      "epoch": 0.7385032847757783,
+      "grad_norm": 2.38465679289117,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 31026
+    },
+    {
+      "epoch": 0.7385270874988099,
+      "grad_norm": 6.83564556487959,
+      "learning_rate": 2e-05,
+      "loss": 0.5851,
+      "step": 31027
+    },
+    {
+      "epoch": 0.7385508902218414,
+      "grad_norm": 2.7320816800184367,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 31028
+    },
+    {
+      "epoch": 0.7385746929448729,
+      "grad_norm": 10.988796735159026,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 31029
+    },
+    {
+      "epoch": 0.7385984956679044,
+      "grad_norm": 8.926832209459388,
+      "learning_rate": 2e-05,
+      "loss": 0.3274,
+      "step": 31030
+    },
+    {
+      "epoch": 0.738622298390936,
+      "grad_norm": 2.1427327719517626,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 31031
+    },
+    {
+      "epoch": 0.7386461011139674,
+      "grad_norm": 3.3087825228800556,
+      "learning_rate": 2e-05,
+      "loss": 0.6536,
+      "step": 31032
+    },
+    {
+      "epoch": 0.7386699038369989,
+      "grad_norm": 10.406367836797399,
+      "learning_rate": 2e-05,
+      "loss": 0.851,
+      "step": 31033
+    },
+    {
+      "epoch": 0.7386937065600304,
+      "grad_norm": 13.103484790171896,
+      "learning_rate": 2e-05,
+      "loss": 1.302,
+      "step": 31034
+    },
+    {
+      "epoch": 0.738717509283062,
+      "grad_norm": 1.324751830861578,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 31035
+    },
+    {
+      "epoch": 0.7387413120060935,
+      "grad_norm": 2.1700741673354527,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 31036
+    },
+    {
+      "epoch": 0.738765114729125,
+      "grad_norm": 6.658069834983631,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 31037
+    },
+    {
+      "epoch": 0.7387889174521566,
+      "grad_norm": 10.757090736262823,
+      "learning_rate": 2e-05,
+      "loss": 0.4831,
+      "step": 31038
+    },
+    {
+      "epoch": 0.7388127201751881,
+      "grad_norm": 6.3842087677035275,
+      "learning_rate": 2e-05,
+      "loss": 1.3885,
+      "step": 31039
+    },
+    {
+      "epoch": 0.7388365228982195,
+      "grad_norm": 3.180392727571428,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 31040
+    },
+    {
+      "epoch": 0.738860325621251,
+      "grad_norm": 5.782802072625265,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 31041
+    },
+    {
+      "epoch": 0.7388841283442826,
+      "grad_norm": 0.045757447492970255,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31042
+    },
+    {
+      "epoch": 0.7389079310673141,
+      "grad_norm": 2.8271648155851574,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 31043
+    },
+    {
+      "epoch": 0.7389317337903456,
+      "grad_norm": 1.245141031031444,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 31044
+    },
+    {
+      "epoch": 0.7389555365133771,
+      "grad_norm": 0.8138222387517059,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 31045
+    },
+    {
+      "epoch": 0.7389793392364087,
+      "grad_norm": 3.171695512183444,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 31046
+    },
+    {
+      "epoch": 0.7390031419594402,
+      "grad_norm": 5.816397859683237,
+      "learning_rate": 2e-05,
+      "loss": 1.0196,
+      "step": 31047
+    },
+    {
+      "epoch": 0.7390269446824717,
+      "grad_norm": 2.9644251276114635,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 31048
+    },
+    {
+      "epoch": 0.7390507474055031,
+      "grad_norm": 8.58052771245348,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 31049
+    },
+    {
+      "epoch": 0.7390745501285347,
+      "grad_norm": 0.7636333850314817,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 31050
+    },
+    {
+      "epoch": 0.7390983528515662,
+      "grad_norm": 1.5134789738273011,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 31051
+    },
+    {
+      "epoch": 0.7391221555745977,
+      "grad_norm": 0.35624536672825174,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 31052
+    },
+    {
+      "epoch": 0.7391459582976293,
+      "grad_norm": 12.195105627321428,
+      "learning_rate": 2e-05,
+      "loss": 0.4335,
+      "step": 31053
+    },
+    {
+      "epoch": 0.7391697610206608,
+      "grad_norm": 3.5040307524513934,
+      "learning_rate": 2e-05,
+      "loss": 0.2997,
+      "step": 31054
+    },
+    {
+      "epoch": 0.7391935637436923,
+      "grad_norm": 1.7541135380050208,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 31055
+    },
+    {
+      "epoch": 0.7392173664667238,
+      "grad_norm": 0.26929955577222914,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 31056
+    },
+    {
+      "epoch": 0.7392411691897554,
+      "grad_norm": 29.94273226536172,
+      "learning_rate": 2e-05,
+      "loss": 0.7606,
+      "step": 31057
+    },
+    {
+      "epoch": 0.7392649719127868,
+      "grad_norm": 13.263301200004792,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 31058
+    },
+    {
+      "epoch": 0.7392887746358183,
+      "grad_norm": 4.078016101721073,
+      "learning_rate": 2e-05,
+      "loss": 0.5092,
+      "step": 31059
+    },
+    {
+      "epoch": 0.7393125773588498,
+      "grad_norm": 2.575336310862408,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 31060
+    },
+    {
+      "epoch": 0.7393363800818814,
+      "grad_norm": 3.2557544154830484,
+      "learning_rate": 2e-05,
+      "loss": 0.4907,
+      "step": 31061
+    },
+    {
+      "epoch": 0.7393601828049129,
+      "grad_norm": 2.5510810136300934,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 31062
+    },
+    {
+      "epoch": 0.7393839855279444,
+      "grad_norm": 2.9949035697800688,
+      "learning_rate": 2e-05,
+      "loss": 0.1497,
+      "step": 31063
+    },
+    {
+      "epoch": 0.7394077882509759,
+      "grad_norm": 7.137015153062744,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 31064
+    },
+    {
+      "epoch": 0.7394315909740075,
+      "grad_norm": 0.32930298306648226,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 31065
+    },
+    {
+      "epoch": 0.7394553936970389,
+      "grad_norm": 6.147819642470769,
+      "learning_rate": 2e-05,
+      "loss": 0.4607,
+      "step": 31066
+    },
+    {
+      "epoch": 0.7394791964200704,
+      "grad_norm": 0.9926901566116735,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 31067
+    },
+    {
+      "epoch": 0.739502999143102,
+      "grad_norm": 8.690135112846525,
+      "learning_rate": 2e-05,
+      "loss": 0.1365,
+      "step": 31068
+    },
+    {
+      "epoch": 0.7395268018661335,
+      "grad_norm": 0.7441775658747399,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 31069
+    },
+    {
+      "epoch": 0.739550604589165,
+      "grad_norm": 6.508435720373361,
+      "learning_rate": 2e-05,
+      "loss": 0.2867,
+      "step": 31070
+    },
+    {
+      "epoch": 0.7395744073121965,
+      "grad_norm": 0.4957767018400818,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 31071
+    },
+    {
+      "epoch": 0.7395982100352281,
+      "grad_norm": 0.40553771659946,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 31072
+    },
+    {
+      "epoch": 0.7396220127582596,
+      "grad_norm": 1.0614511341136112,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 31073
+    },
+    {
+      "epoch": 0.739645815481291,
+      "grad_norm": 1.3387555158821238,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 31074
+    },
+    {
+      "epoch": 0.7396696182043225,
+      "grad_norm": 0.008603998729874237,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 31075
+    },
+    {
+      "epoch": 0.7396934209273541,
+      "grad_norm": 12.873580213673952,
+      "learning_rate": 2e-05,
+      "loss": 0.2226,
+      "step": 31076
+    },
+    {
+      "epoch": 0.7397172236503856,
+      "grad_norm": 2.497747868014097,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 31077
+    },
+    {
+      "epoch": 0.7397410263734171,
+      "grad_norm": 6.515763802691074,
+      "learning_rate": 2e-05,
+      "loss": 0.3885,
+      "step": 31078
+    },
+    {
+      "epoch": 0.7397648290964486,
+      "grad_norm": 10.087842441633775,
+      "learning_rate": 2e-05,
+      "loss": 0.1335,
+      "step": 31079
+    },
+    {
+      "epoch": 0.7397886318194802,
+      "grad_norm": 2.309565407108229,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 31080
+    },
+    {
+      "epoch": 0.7398124345425117,
+      "grad_norm": 2.2756476354077586,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 31081
+    },
+    {
+      "epoch": 0.7398362372655432,
+      "grad_norm": 6.477465595421486,
+      "learning_rate": 2e-05,
+      "loss": 1.0681,
+      "step": 31082
+    },
+    {
+      "epoch": 0.7398600399885747,
+      "grad_norm": 2.9209982066780307,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 31083
+    },
+    {
+      "epoch": 0.7398838427116062,
+      "grad_norm": 0.34696937500094105,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 31084
+    },
+    {
+      "epoch": 0.7399076454346377,
+      "grad_norm": 2.2251202427058563,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 31085
+    },
+    {
+      "epoch": 0.7399314481576692,
+      "grad_norm": 7.82258383983545,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 31086
+    },
+    {
+      "epoch": 0.7399552508807008,
+      "grad_norm": 7.26903510950087,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 31087
+    },
+    {
+      "epoch": 0.7399790536037323,
+      "grad_norm": 12.659388262367417,
+      "learning_rate": 2e-05,
+      "loss": 0.5959,
+      "step": 31088
+    },
+    {
+      "epoch": 0.7400028563267638,
+      "grad_norm": 10.200816204106909,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 31089
+    },
+    {
+      "epoch": 0.7400266590497953,
+      "grad_norm": 2.856093206130364,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 31090
+    },
+    {
+      "epoch": 0.7400504617728269,
+      "grad_norm": 9.83595903864105,
+      "learning_rate": 2e-05,
+      "loss": 0.1576,
+      "step": 31091
+    },
+    {
+      "epoch": 0.7400742644958583,
+      "grad_norm": 2.918546841762775,
+      "learning_rate": 2e-05,
+      "loss": 0.3494,
+      "step": 31092
+    },
+    {
+      "epoch": 0.7400980672188898,
+      "grad_norm": 14.093195159244496,
+      "learning_rate": 2e-05,
+      "loss": 0.7336,
+      "step": 31093
+    },
+    {
+      "epoch": 0.7401218699419213,
+      "grad_norm": 5.721489792853404,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 31094
+    },
+    {
+      "epoch": 0.7401456726649529,
+      "grad_norm": 1.7207098191052776,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 31095
+    },
+    {
+      "epoch": 0.7401694753879844,
+      "grad_norm": 2.146373037827694,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 31096
+    },
+    {
+      "epoch": 0.7401932781110159,
+      "grad_norm": 2.871147704312571,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 31097
+    },
+    {
+      "epoch": 0.7402170808340475,
+      "grad_norm": 3.9841827630311113,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 31098
+    },
+    {
+      "epoch": 0.740240883557079,
+      "grad_norm": 1.9168618233759174,
+      "learning_rate": 2e-05,
+      "loss": 0.2015,
+      "step": 31099
+    },
+    {
+      "epoch": 0.7402646862801104,
+      "grad_norm": 2.726801983823675,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 31100
+    },
+    {
+      "epoch": 0.7402884890031419,
+      "grad_norm": 4.198634716354041,
+      "learning_rate": 2e-05,
+      "loss": 0.1041,
+      "step": 31101
+    },
+    {
+      "epoch": 0.7403122917261735,
+      "grad_norm": 4.0846716957082565,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 31102
+    },
+    {
+      "epoch": 0.740336094449205,
+      "grad_norm": 1.3974168650574352,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 31103
+    },
+    {
+      "epoch": 0.7403598971722365,
+      "grad_norm": 5.10153761269876,
+      "learning_rate": 2e-05,
+      "loss": 0.6335,
+      "step": 31104
+    },
+    {
+      "epoch": 0.740383699895268,
+      "grad_norm": 5.409958140226272,
+      "learning_rate": 2e-05,
+      "loss": 0.4945,
+      "step": 31105
+    },
+    {
+      "epoch": 0.7404075026182996,
+      "grad_norm": 1.309161912209593,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 31106
+    },
+    {
+      "epoch": 0.7404313053413311,
+      "grad_norm": 5.267144729726024,
+      "learning_rate": 2e-05,
+      "loss": 0.441,
+      "step": 31107
+    },
+    {
+      "epoch": 0.7404551080643625,
+      "grad_norm": 2.5733783157616394,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 31108
+    },
+    {
+      "epoch": 0.740478910787394,
+      "grad_norm": 3.017834568143089,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 31109
+    },
+    {
+      "epoch": 0.7405027135104256,
+      "grad_norm": 2.9366071132978955,
+      "learning_rate": 2e-05,
+      "loss": 0.2347,
+      "step": 31110
+    },
+    {
+      "epoch": 0.7405265162334571,
+      "grad_norm": 7.921498669313208,
+      "learning_rate": 2e-05,
+      "loss": 0.8322,
+      "step": 31111
+    },
+    {
+      "epoch": 0.7405503189564886,
+      "grad_norm": 0.0277513388503404,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31112
+    },
+    {
+      "epoch": 0.7405741216795202,
+      "grad_norm": 10.765881966463626,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 31113
+    },
+    {
+      "epoch": 0.7405979244025517,
+      "grad_norm": 2.8758437757163486,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 31114
+    },
+    {
+      "epoch": 0.7406217271255832,
+      "grad_norm": 3.4809722595078507,
+      "learning_rate": 2e-05,
+      "loss": 0.5976,
+      "step": 31115
+    },
+    {
+      "epoch": 0.7406455298486146,
+      "grad_norm": 0.3492915490480994,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 31116
+    },
+    {
+      "epoch": 0.7406693325716462,
+      "grad_norm": 4.211812440140541,
+      "learning_rate": 2e-05,
+      "loss": 0.8949,
+      "step": 31117
+    },
+    {
+      "epoch": 0.7406931352946777,
+      "grad_norm": 3.674847128359345,
+      "learning_rate": 2e-05,
+      "loss": 0.2793,
+      "step": 31118
+    },
+    {
+      "epoch": 0.7407169380177092,
+      "grad_norm": 0.1271829299662674,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 31119
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 11.060266262454839,
+      "learning_rate": 2e-05,
+      "loss": 0.3061,
+      "step": 31120
+    },
+    {
+      "epoch": 0.7407645434637723,
+      "grad_norm": 4.782678525237219,
+      "learning_rate": 2e-05,
+      "loss": 0.6884,
+      "step": 31121
+    },
+    {
+      "epoch": 0.7407883461868038,
+      "grad_norm": 0.5721504504093652,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 31122
+    },
+    {
+      "epoch": 0.7408121489098353,
+      "grad_norm": 3.318661721569388,
+      "learning_rate": 2e-05,
+      "loss": 0.5425,
+      "step": 31123
+    },
+    {
+      "epoch": 0.7408359516328668,
+      "grad_norm": 2.735242286832973,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 31124
+    },
+    {
+      "epoch": 0.7408597543558983,
+      "grad_norm": 6.243006417889887,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 31125
+    },
+    {
+      "epoch": 0.7408835570789298,
+      "grad_norm": 6.104098463697764,
+      "learning_rate": 2e-05,
+      "loss": 0.7192,
+      "step": 31126
+    },
+    {
+      "epoch": 0.7409073598019613,
+      "grad_norm": 14.706676283115394,
+      "learning_rate": 2e-05,
+      "loss": 1.4222,
+      "step": 31127
+    },
+    {
+      "epoch": 0.7409311625249929,
+      "grad_norm": 22.512352520331767,
+      "learning_rate": 2e-05,
+      "loss": 0.9672,
+      "step": 31128
+    },
+    {
+      "epoch": 0.7409549652480244,
+      "grad_norm": 3.0837644725755298,
+      "learning_rate": 2e-05,
+      "loss": 0.4914,
+      "step": 31129
+    },
+    {
+      "epoch": 0.7409787679710559,
+      "grad_norm": 2.4429349261294813,
+      "learning_rate": 2e-05,
+      "loss": 0.2709,
+      "step": 31130
+    },
+    {
+      "epoch": 0.7410025706940874,
+      "grad_norm": 6.791586067015362,
+      "learning_rate": 2e-05,
+      "loss": 1.2694,
+      "step": 31131
+    },
+    {
+      "epoch": 0.741026373417119,
+      "grad_norm": 9.67171260425818,
+      "learning_rate": 2e-05,
+      "loss": 0.8238,
+      "step": 31132
+    },
+    {
+      "epoch": 0.7410501761401505,
+      "grad_norm": 6.1158694312246515,
+      "learning_rate": 2e-05,
+      "loss": 0.2541,
+      "step": 31133
+    },
+    {
+      "epoch": 0.7410739788631819,
+      "grad_norm": 2.0135398674005804,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 31134
+    },
+    {
+      "epoch": 0.7410977815862134,
+      "grad_norm": 1.3181145692568075,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 31135
+    },
+    {
+      "epoch": 0.741121584309245,
+      "grad_norm": 0.13513672967304008,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 31136
+    },
+    {
+      "epoch": 0.7411453870322765,
+      "grad_norm": 8.683902163158905,
+      "learning_rate": 2e-05,
+      "loss": 1.001,
+      "step": 31137
+    },
+    {
+      "epoch": 0.741169189755308,
+      "grad_norm": 7.055136480999976,
+      "learning_rate": 2e-05,
+      "loss": 0.1593,
+      "step": 31138
+    },
+    {
+      "epoch": 0.7411929924783395,
+      "grad_norm": 0.18270691653598814,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 31139
+    },
+    {
+      "epoch": 0.7412167952013711,
+      "grad_norm": 12.614839279997813,
+      "learning_rate": 2e-05,
+      "loss": 0.9344,
+      "step": 31140
+    },
+    {
+      "epoch": 0.7412405979244026,
+      "grad_norm": 3.2528321224792363,
+      "learning_rate": 2e-05,
+      "loss": 0.3196,
+      "step": 31141
+    },
+    {
+      "epoch": 0.741264400647434,
+      "grad_norm": 0.4598117644332109,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 31142
+    },
+    {
+      "epoch": 0.7412882033704655,
+      "grad_norm": 0.8170890254686888,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 31143
+    },
+    {
+      "epoch": 0.7413120060934971,
+      "grad_norm": 0.13641628856685892,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 31144
+    },
+    {
+      "epoch": 0.7413358088165286,
+      "grad_norm": 4.542881864764388,
+      "learning_rate": 2e-05,
+      "loss": 0.4053,
+      "step": 31145
+    },
+    {
+      "epoch": 0.7413596115395601,
+      "grad_norm": 4.154887551219233,
+      "learning_rate": 2e-05,
+      "loss": 0.6987,
+      "step": 31146
+    },
+    {
+      "epoch": 0.7413834142625917,
+      "grad_norm": 0.7195478866494786,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 31147
+    },
+    {
+      "epoch": 0.7414072169856232,
+      "grad_norm": 1.162983327676092,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 31148
+    },
+    {
+      "epoch": 0.7414310197086547,
+      "grad_norm": 1.6167778157129096,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 31149
+    },
+    {
+      "epoch": 0.7414548224316861,
+      "grad_norm": 0.027727434300849074,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31150
+    },
+    {
+      "epoch": 0.7414786251547177,
+      "grad_norm": 2.895563410409829,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 31151
+    },
+    {
+      "epoch": 0.7415024278777492,
+      "grad_norm": 12.399598705994961,
+      "learning_rate": 2e-05,
+      "loss": 0.76,
+      "step": 31152
+    },
+    {
+      "epoch": 0.7415262306007807,
+      "grad_norm": 3.4019965230871474,
+      "learning_rate": 2e-05,
+      "loss": 0.6106,
+      "step": 31153
+    },
+    {
+      "epoch": 0.7415500333238122,
+      "grad_norm": 0.060662760655027445,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 31154
+    },
+    {
+      "epoch": 0.7415738360468438,
+      "grad_norm": 12.098393129292228,
+      "learning_rate": 2e-05,
+      "loss": 0.5812,
+      "step": 31155
+    },
+    {
+      "epoch": 0.7415976387698753,
+      "grad_norm": 2.8047751028198595,
+      "learning_rate": 2e-05,
+      "loss": 0.4136,
+      "step": 31156
+    },
+    {
+      "epoch": 0.7416214414929068,
+      "grad_norm": 3.2844022548358978,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 31157
+    },
+    {
+      "epoch": 0.7416452442159382,
+      "grad_norm": 4.136277450467126,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 31158
+    },
+    {
+      "epoch": 0.7416690469389698,
+      "grad_norm": 3.029223966774519,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 31159
+    },
+    {
+      "epoch": 0.7416928496620013,
+      "grad_norm": 12.24973356573642,
+      "learning_rate": 2e-05,
+      "loss": 0.4096,
+      "step": 31160
+    },
+    {
+      "epoch": 0.7417166523850328,
+      "grad_norm": 2.2922337121147454,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 31161
+    },
+    {
+      "epoch": 0.7417404551080644,
+      "grad_norm": 4.823493734751879,
+      "learning_rate": 2e-05,
+      "loss": 0.9886,
+      "step": 31162
+    },
+    {
+      "epoch": 0.7417642578310959,
+      "grad_norm": 4.898917554057977,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 31163
+    },
+    {
+      "epoch": 0.7417880605541274,
+      "grad_norm": 0.053554412825373245,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31164
+    },
+    {
+      "epoch": 0.7418118632771589,
+      "grad_norm": 4.094203673078931,
+      "learning_rate": 2e-05,
+      "loss": 0.7354,
+      "step": 31165
+    },
+    {
+      "epoch": 0.7418356660001905,
+      "grad_norm": 13.276005425462428,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 31166
+    },
+    {
+      "epoch": 0.741859468723222,
+      "grad_norm": 0.03340150371749444,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31167
+    },
+    {
+      "epoch": 0.7418832714462534,
+      "grad_norm": 7.003848364772025,
+      "learning_rate": 2e-05,
+      "loss": 0.7472,
+      "step": 31168
+    },
+    {
+      "epoch": 0.7419070741692849,
+      "grad_norm": 3.2236052423581487,
+      "learning_rate": 2e-05,
+      "loss": 0.7868,
+      "step": 31169
+    },
+    {
+      "epoch": 0.7419308768923165,
+      "grad_norm": 2.7481654209007567,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 31170
+    },
+    {
+      "epoch": 0.741954679615348,
+      "grad_norm": 6.932763706327422,
+      "learning_rate": 2e-05,
+      "loss": 0.3632,
+      "step": 31171
+    },
+    {
+      "epoch": 0.7419784823383795,
+      "grad_norm": 4.964551010829144,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 31172
+    },
+    {
+      "epoch": 0.742002285061411,
+      "grad_norm": 1.071287123796849,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 31173
+    },
+    {
+      "epoch": 0.7420260877844426,
+      "grad_norm": 2.764955442865399,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 31174
+    },
+    {
+      "epoch": 0.742049890507474,
+      "grad_norm": 0.013181490747338676,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 31175
+    },
+    {
+      "epoch": 0.7420736932305055,
+      "grad_norm": 1.376209097013536,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 31176
+    },
+    {
+      "epoch": 0.7420974959535371,
+      "grad_norm": 8.391611118189537,
+      "learning_rate": 2e-05,
+      "loss": 0.0985,
+      "step": 31177
+    },
+    {
+      "epoch": 0.7421212986765686,
+      "grad_norm": 5.324792212062526,
+      "learning_rate": 2e-05,
+      "loss": 0.5866,
+      "step": 31178
+    },
+    {
+      "epoch": 0.7421451013996001,
+      "grad_norm": 1.5393751791274268,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 31179
+    },
+    {
+      "epoch": 0.7421689041226316,
+      "grad_norm": 7.664597781475085,
+      "learning_rate": 2e-05,
+      "loss": 0.7536,
+      "step": 31180
+    },
+    {
+      "epoch": 0.7421927068456632,
+      "grad_norm": 6.3649674436733,
+      "learning_rate": 2e-05,
+      "loss": 1.1093,
+      "step": 31181
+    },
+    {
+      "epoch": 0.7422165095686947,
+      "grad_norm": 4.004685131461711,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 31182
+    },
+    {
+      "epoch": 0.7422403122917262,
+      "grad_norm": 7.851883385442584,
+      "learning_rate": 2e-05,
+      "loss": 0.2675,
+      "step": 31183
+    },
+    {
+      "epoch": 0.7422641150147576,
+      "grad_norm": 5.383552260897166,
+      "learning_rate": 2e-05,
+      "loss": 0.846,
+      "step": 31184
+    },
+    {
+      "epoch": 0.7422879177377892,
+      "grad_norm": 0.9011765266730631,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 31185
+    },
+    {
+      "epoch": 0.7423117204608207,
+      "grad_norm": 1.648453050726575,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 31186
+    },
+    {
+      "epoch": 0.7423355231838522,
+      "grad_norm": 2.4780401215492525,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 31187
+    },
+    {
+      "epoch": 0.7423593259068837,
+      "grad_norm": 7.058442913381857,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 31188
+    },
+    {
+      "epoch": 0.7423831286299153,
+      "grad_norm": 9.106057802376135,
+      "learning_rate": 2e-05,
+      "loss": 0.3668,
+      "step": 31189
+    },
+    {
+      "epoch": 0.7424069313529468,
+      "grad_norm": 4.426785030973949,
+      "learning_rate": 2e-05,
+      "loss": 0.3492,
+      "step": 31190
+    },
+    {
+      "epoch": 0.7424307340759783,
+      "grad_norm": 0.05990008193110114,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31191
+    },
+    {
+      "epoch": 0.7424545367990099,
+      "grad_norm": 0.9177345982308642,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 31192
+    },
+    {
+      "epoch": 0.7424783395220413,
+      "grad_norm": 0.04964130377973324,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 31193
+    },
+    {
+      "epoch": 0.7425021422450728,
+      "grad_norm": 0.5215559372587786,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 31194
+    },
+    {
+      "epoch": 0.7425259449681043,
+      "grad_norm": 0.032320372653698255,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31195
+    },
+    {
+      "epoch": 0.7425497476911359,
+      "grad_norm": 20.687924445144517,
+      "learning_rate": 2e-05,
+      "loss": 1.1927,
+      "step": 31196
+    },
+    {
+      "epoch": 0.7425735504141674,
+      "grad_norm": 7.922717852616971,
+      "learning_rate": 2e-05,
+      "loss": 1.1416,
+      "step": 31197
+    },
+    {
+      "epoch": 0.7425973531371989,
+      "grad_norm": 10.395161859585606,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 31198
+    },
+    {
+      "epoch": 0.7426211558602304,
+      "grad_norm": 4.367353202311885,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 31199
+    },
+    {
+      "epoch": 0.742644958583262,
+      "grad_norm": 3.3410629403454384,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 31200
+    },
+    {
+      "epoch": 0.7426687613062934,
+      "grad_norm": 4.594459646587745,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 31201
+    },
+    {
+      "epoch": 0.7426925640293249,
+      "grad_norm": 5.045292241367963,
+      "learning_rate": 2e-05,
+      "loss": 0.3489,
+      "step": 31202
+    },
+    {
+      "epoch": 0.7427163667523564,
+      "grad_norm": 2.746029625379946,
+      "learning_rate": 2e-05,
+      "loss": 0.1407,
+      "step": 31203
+    },
+    {
+      "epoch": 0.742740169475388,
+      "grad_norm": 6.803192500483788,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 31204
+    },
+    {
+      "epoch": 0.7427639721984195,
+      "grad_norm": 5.7169864080475215,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 31205
+    },
+    {
+      "epoch": 0.742787774921451,
+      "grad_norm": 26.5290687577696,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 31206
+    },
+    {
+      "epoch": 0.7428115776444826,
+      "grad_norm": 1.1164809494644936,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 31207
+    },
+    {
+      "epoch": 0.7428353803675141,
+      "grad_norm": 0.7753934088638912,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 31208
+    },
+    {
+      "epoch": 0.7428591830905455,
+      "grad_norm": 4.876631837848711,
+      "learning_rate": 2e-05,
+      "loss": 0.9453,
+      "step": 31209
+    },
+    {
+      "epoch": 0.742882985813577,
+      "grad_norm": 3.4779045832228155,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 31210
+    },
+    {
+      "epoch": 0.7429067885366086,
+      "grad_norm": 6.809229642727762,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 31211
+    },
+    {
+      "epoch": 0.7429305912596401,
+      "grad_norm": 0.5870307952401731,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 31212
+    },
+    {
+      "epoch": 0.7429543939826716,
+      "grad_norm": 1.8646385279409003,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 31213
+    },
+    {
+      "epoch": 0.7429781967057031,
+      "grad_norm": 1.9848789827494462,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 31214
+    },
+    {
+      "epoch": 0.7430019994287347,
+      "grad_norm": 8.712775630975635,
+      "learning_rate": 2e-05,
+      "loss": 0.3474,
+      "step": 31215
+    },
+    {
+      "epoch": 0.7430258021517662,
+      "grad_norm": 3.800939457295043,
+      "learning_rate": 2e-05,
+      "loss": 0.478,
+      "step": 31216
+    },
+    {
+      "epoch": 0.7430496048747977,
+      "grad_norm": 4.46814064818065,
+      "learning_rate": 2e-05,
+      "loss": 0.7797,
+      "step": 31217
+    },
+    {
+      "epoch": 0.7430734075978291,
+      "grad_norm": 4.17411301414712,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 31218
+    },
+    {
+      "epoch": 0.7430972103208607,
+      "grad_norm": 50.687214440564674,
+      "learning_rate": 2e-05,
+      "loss": 0.402,
+      "step": 31219
+    },
+    {
+      "epoch": 0.7431210130438922,
+      "grad_norm": 5.765536605317896,
+      "learning_rate": 2e-05,
+      "loss": 0.984,
+      "step": 31220
+    },
+    {
+      "epoch": 0.7431448157669237,
+      "grad_norm": 2.799473011198184,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 31221
+    },
+    {
+      "epoch": 0.7431686184899553,
+      "grad_norm": 3.157296584845503,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 31222
+    },
+    {
+      "epoch": 0.7431924212129868,
+      "grad_norm": 1.020706148279419,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 31223
+    },
+    {
+      "epoch": 0.7432162239360183,
+      "grad_norm": 4.793370561998873,
+      "learning_rate": 2e-05,
+      "loss": 1.1654,
+      "step": 31224
+    },
+    {
+      "epoch": 0.7432400266590498,
+      "grad_norm": 13.039746559653997,
+      "learning_rate": 2e-05,
+      "loss": 0.9027,
+      "step": 31225
+    },
+    {
+      "epoch": 0.7432638293820814,
+      "grad_norm": 6.84232898036222,
+      "learning_rate": 2e-05,
+      "loss": 0.7523,
+      "step": 31226
+    },
+    {
+      "epoch": 0.7432876321051128,
+      "grad_norm": 9.243056369501721,
+      "learning_rate": 2e-05,
+      "loss": 0.789,
+      "step": 31227
+    },
+    {
+      "epoch": 0.7433114348281443,
+      "grad_norm": 2.0852825138862627,
+      "learning_rate": 2e-05,
+      "loss": 0.2174,
+      "step": 31228
+    },
+    {
+      "epoch": 0.7433352375511758,
+      "grad_norm": 4.595504867923514,
+      "learning_rate": 2e-05,
+      "loss": 0.8397,
+      "step": 31229
+    },
+    {
+      "epoch": 0.7433590402742074,
+      "grad_norm": 3.4156062267793104,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 31230
+    },
+    {
+      "epoch": 0.7433828429972389,
+      "grad_norm": 0.5701995231230973,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 31231
+    },
+    {
+      "epoch": 0.7434066457202704,
+      "grad_norm": 3.841614624734773,
+      "learning_rate": 2e-05,
+      "loss": 0.5424,
+      "step": 31232
+    },
+    {
+      "epoch": 0.7434304484433019,
+      "grad_norm": 2.1432398809610285,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 31233
+    },
+    {
+      "epoch": 0.7434542511663335,
+      "grad_norm": 6.939905547300137,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 31234
+    },
+    {
+      "epoch": 0.7434780538893649,
+      "grad_norm": 1.9934765419220222,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 31235
+    },
+    {
+      "epoch": 0.7435018566123964,
+      "grad_norm": 0.7194872568046133,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 31236
+    },
+    {
+      "epoch": 0.743525659335428,
+      "grad_norm": 3.641660266974254,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 31237
+    },
+    {
+      "epoch": 0.7435494620584595,
+      "grad_norm": 3.5673954027074513,
+      "learning_rate": 2e-05,
+      "loss": 0.3723,
+      "step": 31238
+    },
+    {
+      "epoch": 0.743573264781491,
+      "grad_norm": 4.629338527337847,
+      "learning_rate": 2e-05,
+      "loss": 0.659,
+      "step": 31239
+    },
+    {
+      "epoch": 0.7435970675045225,
+      "grad_norm": 3.9445913972825055,
+      "learning_rate": 2e-05,
+      "loss": 0.5904,
+      "step": 31240
+    },
+    {
+      "epoch": 0.7436208702275541,
+      "grad_norm": 0.30728502475711533,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 31241
+    },
+    {
+      "epoch": 0.7436446729505856,
+      "grad_norm": 0.951026236185294,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 31242
+    },
+    {
+      "epoch": 0.743668475673617,
+      "grad_norm": 10.265240758482948,
+      "learning_rate": 2e-05,
+      "loss": 0.2808,
+      "step": 31243
+    },
+    {
+      "epoch": 0.7436922783966485,
+      "grad_norm": 9.942784655512973,
+      "learning_rate": 2e-05,
+      "loss": 0.2099,
+      "step": 31244
+    },
+    {
+      "epoch": 0.7437160811196801,
+      "grad_norm": 4.243646119580363,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 31245
+    },
+    {
+      "epoch": 0.7437398838427116,
+      "grad_norm": 6.743382610225325,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 31246
+    },
+    {
+      "epoch": 0.7437636865657431,
+      "grad_norm": 3.6022250195304752,
+      "learning_rate": 2e-05,
+      "loss": 0.6221,
+      "step": 31247
+    },
+    {
+      "epoch": 0.7437874892887746,
+      "grad_norm": 1.6396237480544438,
+      "learning_rate": 2e-05,
+      "loss": 0.1624,
+      "step": 31248
+    },
+    {
+      "epoch": 0.7438112920118062,
+      "grad_norm": 5.95173529099288,
+      "learning_rate": 2e-05,
+      "loss": 1.0952,
+      "step": 31249
+    },
+    {
+      "epoch": 0.7438350947348377,
+      "grad_norm": 16.317030412394455,
+      "learning_rate": 2e-05,
+      "loss": 0.6529,
+      "step": 31250
+    },
+    {
+      "epoch": 0.7438588974578692,
+      "grad_norm": 5.765401065109167,
+      "learning_rate": 2e-05,
+      "loss": 1.3968,
+      "step": 31251
+    },
+    {
+      "epoch": 0.7438827001809007,
+      "grad_norm": 0.11102952910766595,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 31252
+    },
+    {
+      "epoch": 0.7439065029039322,
+      "grad_norm": 0.6396870495478147,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 31253
+    },
+    {
+      "epoch": 0.7439303056269637,
+      "grad_norm": 3.045449135890009,
+      "learning_rate": 2e-05,
+      "loss": 0.5839,
+      "step": 31254
+    },
+    {
+      "epoch": 0.7439541083499952,
+      "grad_norm": 3.9718716369507785,
+      "learning_rate": 2e-05,
+      "loss": 0.8833,
+      "step": 31255
+    },
+    {
+      "epoch": 0.7439779110730268,
+      "grad_norm": 6.02049866494198,
+      "learning_rate": 2e-05,
+      "loss": 0.5105,
+      "step": 31256
+    },
+    {
+      "epoch": 0.7440017137960583,
+      "grad_norm": 0.45496321985308324,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 31257
+    },
+    {
+      "epoch": 0.7440255165190898,
+      "grad_norm": 0.8137657894302862,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 31258
+    },
+    {
+      "epoch": 0.7440493192421213,
+      "grad_norm": 7.986375367685572,
+      "learning_rate": 2e-05,
+      "loss": 0.749,
+      "step": 31259
+    },
+    {
+      "epoch": 0.7440731219651529,
+      "grad_norm": 3.037498080154335,
+      "learning_rate": 2e-05,
+      "loss": 0.357,
+      "step": 31260
+    },
+    {
+      "epoch": 0.7440969246881843,
+      "grad_norm": 1.25143100696779,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 31261
+    },
+    {
+      "epoch": 0.7441207274112158,
+      "grad_norm": 2.867814600512617,
+      "learning_rate": 2e-05,
+      "loss": 0.518,
+      "step": 31262
+    },
+    {
+      "epoch": 0.7441445301342473,
+      "grad_norm": 7.086941305300985,
+      "learning_rate": 2e-05,
+      "loss": 0.2658,
+      "step": 31263
+    },
+    {
+      "epoch": 0.7441683328572789,
+      "grad_norm": 0.09844233117772322,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31264
+    },
+    {
+      "epoch": 0.7441921355803104,
+      "grad_norm": 5.81023000668546,
+      "learning_rate": 2e-05,
+      "loss": 0.6529,
+      "step": 31265
+    },
+    {
+      "epoch": 0.7442159383033419,
+      "grad_norm": 0.21975001243811845,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 31266
+    },
+    {
+      "epoch": 0.7442397410263735,
+      "grad_norm": 6.217392678438081,
+      "learning_rate": 2e-05,
+      "loss": 0.564,
+      "step": 31267
+    },
+    {
+      "epoch": 0.744263543749405,
+      "grad_norm": 1.0480396963591108,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 31268
+    },
+    {
+      "epoch": 0.7442873464724364,
+      "grad_norm": 4.142871549878002,
+      "learning_rate": 2e-05,
+      "loss": 0.6372,
+      "step": 31269
+    },
+    {
+      "epoch": 0.7443111491954679,
+      "grad_norm": 41.81706118714059,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 31270
+    },
+    {
+      "epoch": 0.7443349519184995,
+      "grad_norm": 4.987309708724141,
+      "learning_rate": 2e-05,
+      "loss": 0.6107,
+      "step": 31271
+    },
+    {
+      "epoch": 0.744358754641531,
+      "grad_norm": 0.2700107917156624,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 31272
+    },
+    {
+      "epoch": 0.7443825573645625,
+      "grad_norm": 14.001068353354558,
+      "learning_rate": 2e-05,
+      "loss": 0.5755,
+      "step": 31273
+    },
+    {
+      "epoch": 0.744406360087594,
+      "grad_norm": 7.1429029499508525,
+      "learning_rate": 2e-05,
+      "loss": 0.2126,
+      "step": 31274
+    },
+    {
+      "epoch": 0.7444301628106256,
+      "grad_norm": 3.27908242761502,
+      "learning_rate": 2e-05,
+      "loss": 0.5873,
+      "step": 31275
+    },
+    {
+      "epoch": 0.7444539655336571,
+      "grad_norm": 2.730538262163338,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 31276
+    },
+    {
+      "epoch": 0.7444777682566885,
+      "grad_norm": 6.901031919618345,
+      "learning_rate": 2e-05,
+      "loss": 1.6645,
+      "step": 31277
+    },
+    {
+      "epoch": 0.74450157097972,
+      "grad_norm": 2.428741546040638,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 31278
+    },
+    {
+      "epoch": 0.7445253737027516,
+      "grad_norm": 6.580615983074881,
+      "learning_rate": 2e-05,
+      "loss": 0.1928,
+      "step": 31279
+    },
+    {
+      "epoch": 0.7445491764257831,
+      "grad_norm": 3.0082586088527976,
+      "learning_rate": 2e-05,
+      "loss": 0.2824,
+      "step": 31280
+    },
+    {
+      "epoch": 0.7445729791488146,
+      "grad_norm": 50.266216996228835,
+      "learning_rate": 2e-05,
+      "loss": 0.372,
+      "step": 31281
+    },
+    {
+      "epoch": 0.7445967818718462,
+      "grad_norm": 1.813964729644599,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 31282
+    },
+    {
+      "epoch": 0.7446205845948777,
+      "grad_norm": 3.5106243257798604,
+      "learning_rate": 2e-05,
+      "loss": 0.8194,
+      "step": 31283
+    },
+    {
+      "epoch": 0.7446443873179092,
+      "grad_norm": 2.8562080608915936,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 31284
+    },
+    {
+      "epoch": 0.7446681900409406,
+      "grad_norm": 4.632850650837259,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 31285
+    },
+    {
+      "epoch": 0.7446919927639722,
+      "grad_norm": 7.13536846303466,
+      "learning_rate": 2e-05,
+      "loss": 1.0127,
+      "step": 31286
+    },
+    {
+      "epoch": 0.7447157954870037,
+      "grad_norm": 3.5552966727156177,
+      "learning_rate": 2e-05,
+      "loss": 0.3257,
+      "step": 31287
+    },
+    {
+      "epoch": 0.7447395982100352,
+      "grad_norm": 1.102655560506651,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 31288
+    },
+    {
+      "epoch": 0.7447634009330667,
+      "grad_norm": 2.5998447569500005,
+      "learning_rate": 2e-05,
+      "loss": 0.4944,
+      "step": 31289
+    },
+    {
+      "epoch": 0.7447872036560983,
+      "grad_norm": 8.985131565989072,
+      "learning_rate": 2e-05,
+      "loss": 0.252,
+      "step": 31290
+    },
+    {
+      "epoch": 0.7448110063791298,
+      "grad_norm": 0.44059215014089276,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31291
+    },
+    {
+      "epoch": 0.7448348091021613,
+      "grad_norm": 3.537545394754409,
+      "learning_rate": 2e-05,
+      "loss": 0.1499,
+      "step": 31292
+    },
+    {
+      "epoch": 0.7448586118251928,
+      "grad_norm": 6.685383833743535,
+      "learning_rate": 2e-05,
+      "loss": 1.4547,
+      "step": 31293
+    },
+    {
+      "epoch": 0.7448824145482243,
+      "grad_norm": 0.763293489744409,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 31294
+    },
+    {
+      "epoch": 0.7449062172712558,
+      "grad_norm": 7.648967332664596,
+      "learning_rate": 2e-05,
+      "loss": 0.8506,
+      "step": 31295
+    },
+    {
+      "epoch": 0.7449300199942873,
+      "grad_norm": 0.06220733797701608,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31296
+    },
+    {
+      "epoch": 0.7449538227173189,
+      "grad_norm": 2.203598575568444,
+      "learning_rate": 2e-05,
+      "loss": 0.1791,
+      "step": 31297
+    },
+    {
+      "epoch": 0.7449776254403504,
+      "grad_norm": 5.407722930438402,
+      "learning_rate": 2e-05,
+      "loss": 1.0179,
+      "step": 31298
+    },
+    {
+      "epoch": 0.7450014281633819,
+      "grad_norm": 0.09527083564352302,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 31299
+    },
+    {
+      "epoch": 0.7450252308864134,
+      "grad_norm": 2.2108853230330197,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 31300
+    },
+    {
+      "epoch": 0.745049033609445,
+      "grad_norm": 9.702784101505218,
+      "learning_rate": 2e-05,
+      "loss": 0.66,
+      "step": 31301
+    },
+    {
+      "epoch": 0.7450728363324765,
+      "grad_norm": 4.755528977836281,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 31302
+    },
+    {
+      "epoch": 0.7450966390555079,
+      "grad_norm": 4.995757534825512,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 31303
+    },
+    {
+      "epoch": 0.7451204417785394,
+      "grad_norm": 4.129982400747425,
+      "learning_rate": 2e-05,
+      "loss": 0.6431,
+      "step": 31304
+    },
+    {
+      "epoch": 0.745144244501571,
+      "grad_norm": 2.6935493704121423,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 31305
+    },
+    {
+      "epoch": 0.7451680472246025,
+      "grad_norm": 3.016194592449407,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 31306
+    },
+    {
+      "epoch": 0.745191849947634,
+      "grad_norm": 0.5805114175243686,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 31307
+    },
+    {
+      "epoch": 0.7452156526706655,
+      "grad_norm": 1.7113380930033826,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 31308
+    },
+    {
+      "epoch": 0.7452394553936971,
+      "grad_norm": 23.80159897513842,
+      "learning_rate": 2e-05,
+      "loss": 0.5781,
+      "step": 31309
+    },
+    {
+      "epoch": 0.7452632581167286,
+      "grad_norm": 3.154445189705903,
+      "learning_rate": 2e-05,
+      "loss": 0.7856,
+      "step": 31310
+    },
+    {
+      "epoch": 0.74528706083976,
+      "grad_norm": 3.4590444487819885,
+      "learning_rate": 2e-05,
+      "loss": 0.3464,
+      "step": 31311
+    },
+    {
+      "epoch": 0.7453108635627916,
+      "grad_norm": 4.045475860605456,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 31312
+    },
+    {
+      "epoch": 0.7453346662858231,
+      "grad_norm": 0.3082021812618527,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 31313
+    },
+    {
+      "epoch": 0.7453584690088546,
+      "grad_norm": 15.676911687522697,
+      "learning_rate": 2e-05,
+      "loss": 0.674,
+      "step": 31314
+    },
+    {
+      "epoch": 0.7453822717318861,
+      "grad_norm": 1.9544779550345948,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 31315
+    },
+    {
+      "epoch": 0.7454060744549177,
+      "grad_norm": 4.358884496192148,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 31316
+    },
+    {
+      "epoch": 0.7454298771779492,
+      "grad_norm": 0.6572675094931373,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 31317
+    },
+    {
+      "epoch": 0.7454536799009807,
+      "grad_norm": 1.6959145139385354,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 31318
+    },
+    {
+      "epoch": 0.7454774826240121,
+      "grad_norm": 3.8640916583880247,
+      "learning_rate": 2e-05,
+      "loss": 0.7762,
+      "step": 31319
+    },
+    {
+      "epoch": 0.7455012853470437,
+      "grad_norm": 2.7057590537108758,
+      "learning_rate": 2e-05,
+      "loss": 0.4352,
+      "step": 31320
+    },
+    {
+      "epoch": 0.7455250880700752,
+      "grad_norm": 8.183766898346839,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 31321
+    },
+    {
+      "epoch": 0.7455488907931067,
+      "grad_norm": 7.943115526102756,
+      "learning_rate": 2e-05,
+      "loss": 0.8664,
+      "step": 31322
+    },
+    {
+      "epoch": 0.7455726935161382,
+      "grad_norm": 10.72709721299789,
+      "learning_rate": 2e-05,
+      "loss": 0.2187,
+      "step": 31323
+    },
+    {
+      "epoch": 0.7455964962391698,
+      "grad_norm": 4.792288490658423,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 31324
+    },
+    {
+      "epoch": 0.7456202989622013,
+      "grad_norm": 0.1971194960467997,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 31325
+    },
+    {
+      "epoch": 0.7456441016852328,
+      "grad_norm": 5.2325368828163485,
+      "learning_rate": 2e-05,
+      "loss": 0.5172,
+      "step": 31326
+    },
+    {
+      "epoch": 0.7456679044082644,
+      "grad_norm": 4.067644548301296,
+      "learning_rate": 2e-05,
+      "loss": 0.8927,
+      "step": 31327
+    },
+    {
+      "epoch": 0.7456917071312958,
+      "grad_norm": 7.490007115434764,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 31328
+    },
+    {
+      "epoch": 0.7457155098543273,
+      "grad_norm": 4.654288977395387,
+      "learning_rate": 2e-05,
+      "loss": 0.3916,
+      "step": 31329
+    },
+    {
+      "epoch": 0.7457393125773588,
+      "grad_norm": 0.05500732665132747,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 31330
+    },
+    {
+      "epoch": 0.7457631153003904,
+      "grad_norm": 55.017551151606895,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 31331
+    },
+    {
+      "epoch": 0.7457869180234219,
+      "grad_norm": 16.213125443132913,
+      "learning_rate": 2e-05,
+      "loss": 0.3846,
+      "step": 31332
+    },
+    {
+      "epoch": 0.7458107207464534,
+      "grad_norm": 3.4934132459057503,
+      "learning_rate": 2e-05,
+      "loss": 0.5556,
+      "step": 31333
+    },
+    {
+      "epoch": 0.7458345234694849,
+      "grad_norm": 0.6901828618486632,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 31334
+    },
+    {
+      "epoch": 0.7458583261925165,
+      "grad_norm": 4.584466725378906,
+      "learning_rate": 2e-05,
+      "loss": 0.5923,
+      "step": 31335
+    },
+    {
+      "epoch": 0.745882128915548,
+      "grad_norm": 5.711782862877287,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 31336
+    },
+    {
+      "epoch": 0.7459059316385794,
+      "grad_norm": 2.4621485437878663,
+      "learning_rate": 2e-05,
+      "loss": 0.376,
+      "step": 31337
+    },
+    {
+      "epoch": 0.7459297343616109,
+      "grad_norm": 0.23566454126370362,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 31338
+    },
+    {
+      "epoch": 0.7459535370846425,
+      "grad_norm": 5.122835176817728,
+      "learning_rate": 2e-05,
+      "loss": 0.6153,
+      "step": 31339
+    },
+    {
+      "epoch": 0.745977339807674,
+      "grad_norm": 2.2858266114022023,
+      "learning_rate": 2e-05,
+      "loss": 0.3759,
+      "step": 31340
+    },
+    {
+      "epoch": 0.7460011425307055,
+      "grad_norm": 0.045261493690209396,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31341
+    },
+    {
+      "epoch": 0.7460249452537371,
+      "grad_norm": 7.339925375116187,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 31342
+    },
+    {
+      "epoch": 0.7460487479767686,
+      "grad_norm": 2.0165849279482853,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 31343
+    },
+    {
+      "epoch": 0.7460725506998,
+      "grad_norm": 0.9256998169435007,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 31344
+    },
+    {
+      "epoch": 0.7460963534228315,
+      "grad_norm": 2.088221332837746,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 31345
+    },
+    {
+      "epoch": 0.7461201561458631,
+      "grad_norm": 3.0972246611828496,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 31346
+    },
+    {
+      "epoch": 0.7461439588688946,
+      "grad_norm": 2.4657784423015148,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 31347
+    },
+    {
+      "epoch": 0.7461677615919261,
+      "grad_norm": 0.7244581736689125,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 31348
+    },
+    {
+      "epoch": 0.7461915643149576,
+      "grad_norm": 0.8440981435746154,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 31349
+    },
+    {
+      "epoch": 0.7462153670379892,
+      "grad_norm": 4.790359217451419,
+      "learning_rate": 2e-05,
+      "loss": 0.188,
+      "step": 31350
+    },
+    {
+      "epoch": 0.7462391697610207,
+      "grad_norm": 0.7587431387998044,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 31351
+    },
+    {
+      "epoch": 0.7462629724840522,
+      "grad_norm": 5.2008225218833655,
+      "learning_rate": 2e-05,
+      "loss": 0.675,
+      "step": 31352
+    },
+    {
+      "epoch": 0.7462867752070836,
+      "grad_norm": 0.8693211089863285,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 31353
+    },
+    {
+      "epoch": 0.7463105779301152,
+      "grad_norm": 5.799718606956197,
+      "learning_rate": 2e-05,
+      "loss": 1.0249,
+      "step": 31354
+    },
+    {
+      "epoch": 0.7463343806531467,
+      "grad_norm": 0.27793826953394285,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 31355
+    },
+    {
+      "epoch": 0.7463581833761782,
+      "grad_norm": 3.3156180108177202,
+      "learning_rate": 2e-05,
+      "loss": 0.6132,
+      "step": 31356
+    },
+    {
+      "epoch": 0.7463819860992098,
+      "grad_norm": 2.022081835997423,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 31357
+    },
+    {
+      "epoch": 0.7464057888222413,
+      "grad_norm": 2.035257451617716,
+      "learning_rate": 2e-05,
+      "loss": 0.0826,
+      "step": 31358
+    },
+    {
+      "epoch": 0.7464295915452728,
+      "grad_norm": 3.165122680835799,
+      "learning_rate": 2e-05,
+      "loss": 0.3945,
+      "step": 31359
+    },
+    {
+      "epoch": 0.7464533942683043,
+      "grad_norm": 3.1114769534107247,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 31360
+    },
+    {
+      "epoch": 0.7464771969913359,
+      "grad_norm": 2.3348501338606367,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 31361
+    },
+    {
+      "epoch": 0.7465009997143673,
+      "grad_norm": 3.805327191381134,
+      "learning_rate": 2e-05,
+      "loss": 0.4809,
+      "step": 31362
+    },
+    {
+      "epoch": 0.7465248024373988,
+      "grad_norm": 0.7933066101099974,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 31363
+    },
+    {
+      "epoch": 0.7465486051604303,
+      "grad_norm": 0.44019589460944547,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 31364
+    },
+    {
+      "epoch": 0.7465724078834619,
+      "grad_norm": 1.216097317109457,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 31365
+    },
+    {
+      "epoch": 0.7465962106064934,
+      "grad_norm": 2.970544544299757,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 31366
+    },
+    {
+      "epoch": 0.7466200133295249,
+      "grad_norm": 6.197980114200786,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 31367
+    },
+    {
+      "epoch": 0.7466438160525564,
+      "grad_norm": 0.11938629387375344,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 31368
+    },
+    {
+      "epoch": 0.746667618775588,
+      "grad_norm": 6.147537219689522,
+      "learning_rate": 2e-05,
+      "loss": 0.759,
+      "step": 31369
+    },
+    {
+      "epoch": 0.7466914214986194,
+      "grad_norm": 7.567439906625011,
+      "learning_rate": 2e-05,
+      "loss": 0.6936,
+      "step": 31370
+    },
+    {
+      "epoch": 0.7467152242216509,
+      "grad_norm": 0.7472944602745661,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 31371
+    },
+    {
+      "epoch": 0.7467390269446825,
+      "grad_norm": 0.32189661329820546,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 31372
+    },
+    {
+      "epoch": 0.746762829667714,
+      "grad_norm": 2.993233970095455,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 31373
+    },
+    {
+      "epoch": 0.7467866323907455,
+      "grad_norm": 0.07979117921075238,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31374
+    },
+    {
+      "epoch": 0.746810435113777,
+      "grad_norm": 6.41267566038463,
+      "learning_rate": 2e-05,
+      "loss": 0.788,
+      "step": 31375
+    },
+    {
+      "epoch": 0.7468342378368086,
+      "grad_norm": 3.5869392560170175,
+      "learning_rate": 2e-05,
+      "loss": 0.2294,
+      "step": 31376
+    },
+    {
+      "epoch": 0.7468580405598401,
+      "grad_norm": 0.9720004170148546,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 31377
+    },
+    {
+      "epoch": 0.7468818432828715,
+      "grad_norm": 0.07610114022164693,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 31378
+    },
+    {
+      "epoch": 0.746905646005903,
+      "grad_norm": 0.2997790225941002,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 31379
+    },
+    {
+      "epoch": 0.7469294487289346,
+      "grad_norm": 32.151487476055664,
+      "learning_rate": 2e-05,
+      "loss": 1.4999,
+      "step": 31380
+    },
+    {
+      "epoch": 0.7469532514519661,
+      "grad_norm": 3.271536147793651,
+      "learning_rate": 2e-05,
+      "loss": 0.3971,
+      "step": 31381
+    },
+    {
+      "epoch": 0.7469770541749976,
+      "grad_norm": 4.170461563160506,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 31382
+    },
+    {
+      "epoch": 0.7470008568980291,
+      "grad_norm": 0.011704973751784134,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 31383
+    },
+    {
+      "epoch": 0.7470246596210607,
+      "grad_norm": 9.620104797173877,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 31384
+    },
+    {
+      "epoch": 0.7470484623440922,
+      "grad_norm": 2.975080702684042,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 31385
+    },
+    {
+      "epoch": 0.7470722650671237,
+      "grad_norm": 71.95692789559847,
+      "learning_rate": 2e-05,
+      "loss": 0.3623,
+      "step": 31386
+    },
+    {
+      "epoch": 0.7470960677901551,
+      "grad_norm": 2.6298383511446324,
+      "learning_rate": 2e-05,
+      "loss": 0.4479,
+      "step": 31387
+    },
+    {
+      "epoch": 0.7471198705131867,
+      "grad_norm": 22.026315685240366,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 31388
+    },
+    {
+      "epoch": 0.7471436732362182,
+      "grad_norm": 5.061675567703876,
+      "learning_rate": 2e-05,
+      "loss": 0.5817,
+      "step": 31389
+    },
+    {
+      "epoch": 0.7471674759592497,
+      "grad_norm": 7.749859902853556,
+      "learning_rate": 2e-05,
+      "loss": 0.1968,
+      "step": 31390
+    },
+    {
+      "epoch": 0.7471912786822813,
+      "grad_norm": 1.5696391162289751,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 31391
+    },
+    {
+      "epoch": 0.7472150814053128,
+      "grad_norm": 1.711986930328874,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 31392
+    },
+    {
+      "epoch": 0.7472388841283443,
+      "grad_norm": 4.827567438728339,
+      "learning_rate": 2e-05,
+      "loss": 0.8463,
+      "step": 31393
+    },
+    {
+      "epoch": 0.7472626868513758,
+      "grad_norm": 16.677646912129944,
+      "learning_rate": 2e-05,
+      "loss": 2.161,
+      "step": 31394
+    },
+    {
+      "epoch": 0.7472864895744074,
+      "grad_norm": 17.37302631966038,
+      "learning_rate": 2e-05,
+      "loss": 1.1439,
+      "step": 31395
+    },
+    {
+      "epoch": 0.7473102922974388,
+      "grad_norm": 6.626869965242189,
+      "learning_rate": 2e-05,
+      "loss": 0.9409,
+      "step": 31396
+    },
+    {
+      "epoch": 0.7473340950204703,
+      "grad_norm": 0.39442905333903816,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 31397
+    },
+    {
+      "epoch": 0.7473578977435018,
+      "grad_norm": 3.2435796993861694,
+      "learning_rate": 2e-05,
+      "loss": 0.3177,
+      "step": 31398
+    },
+    {
+      "epoch": 0.7473817004665334,
+      "grad_norm": 0.5777711591815196,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 31399
+    },
+    {
+      "epoch": 0.7474055031895649,
+      "grad_norm": 9.112333433245515,
+      "learning_rate": 2e-05,
+      "loss": 0.8143,
+      "step": 31400
+    },
+    {
+      "epoch": 0.7474293059125964,
+      "grad_norm": 0.9280780607182118,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 31401
+    },
+    {
+      "epoch": 0.7474531086356279,
+      "grad_norm": 0.48267801217775225,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 31402
+    },
+    {
+      "epoch": 0.7474769113586595,
+      "grad_norm": 1.4027826417264464,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 31403
+    },
+    {
+      "epoch": 0.7475007140816909,
+      "grad_norm": 3.8105412742503875,
+      "learning_rate": 2e-05,
+      "loss": 0.6663,
+      "step": 31404
+    },
+    {
+      "epoch": 0.7475245168047224,
+      "grad_norm": 0.027463825996133152,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 31405
+    },
+    {
+      "epoch": 0.747548319527754,
+      "grad_norm": 0.1325431068445251,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 31406
+    },
+    {
+      "epoch": 0.7475721222507855,
+      "grad_norm": 0.8656846637952521,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 31407
+    },
+    {
+      "epoch": 0.747595924973817,
+      "grad_norm": 3.8634935536918,
+      "learning_rate": 2e-05,
+      "loss": 0.4006,
+      "step": 31408
+    },
+    {
+      "epoch": 0.7476197276968485,
+      "grad_norm": 5.820796180245969,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 31409
+    },
+    {
+      "epoch": 0.7476435304198801,
+      "grad_norm": 2.0570993775318476,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 31410
+    },
+    {
+      "epoch": 0.7476673331429116,
+      "grad_norm": 5.067796000010026,
+      "learning_rate": 2e-05,
+      "loss": 0.7661,
+      "step": 31411
+    },
+    {
+      "epoch": 0.747691135865943,
+      "grad_norm": 1.3331647462767078,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 31412
+    },
+    {
+      "epoch": 0.7477149385889745,
+      "grad_norm": 4.638066911368056,
+      "learning_rate": 2e-05,
+      "loss": 0.7365,
+      "step": 31413
+    },
+    {
+      "epoch": 0.7477387413120061,
+      "grad_norm": 10.248827054056123,
+      "learning_rate": 2e-05,
+      "loss": 0.3564,
+      "step": 31414
+    },
+    {
+      "epoch": 0.7477625440350376,
+      "grad_norm": 3.2030831509524837,
+      "learning_rate": 2e-05,
+      "loss": 0.5441,
+      "step": 31415
+    },
+    {
+      "epoch": 0.7477863467580691,
+      "grad_norm": 3.5207823551195676,
+      "learning_rate": 2e-05,
+      "loss": 0.5093,
+      "step": 31416
+    },
+    {
+      "epoch": 0.7478101494811006,
+      "grad_norm": 5.1011962355729485,
+      "learning_rate": 2e-05,
+      "loss": 0.221,
+      "step": 31417
+    },
+    {
+      "epoch": 0.7478339522041322,
+      "grad_norm": 0.4245224293658744,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31418
+    },
+    {
+      "epoch": 0.7478577549271637,
+      "grad_norm": 0.07052130834119394,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 31419
+    },
+    {
+      "epoch": 0.7478815576501952,
+      "grad_norm": 10.728799048965858,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 31420
+    },
+    {
+      "epoch": 0.7479053603732267,
+      "grad_norm": 4.560365269872802,
+      "learning_rate": 2e-05,
+      "loss": 0.7548,
+      "step": 31421
+    },
+    {
+      "epoch": 0.7479291630962582,
+      "grad_norm": 3.0596159590551237,
+      "learning_rate": 2e-05,
+      "loss": 0.2112,
+      "step": 31422
+    },
+    {
+      "epoch": 0.7479529658192897,
+      "grad_norm": 0.07251727807119897,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31423
+    },
+    {
+      "epoch": 0.7479767685423212,
+      "grad_norm": 3.041873984524855,
+      "learning_rate": 2e-05,
+      "loss": 0.2054,
+      "step": 31424
+    },
+    {
+      "epoch": 0.7480005712653528,
+      "grad_norm": 8.748973138975286,
+      "learning_rate": 2e-05,
+      "loss": 0.3551,
+      "step": 31425
+    },
+    {
+      "epoch": 0.7480243739883843,
+      "grad_norm": 0.08042380742175055,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 31426
+    },
+    {
+      "epoch": 0.7480481767114158,
+      "grad_norm": 4.418251267988309,
+      "learning_rate": 2e-05,
+      "loss": 0.3987,
+      "step": 31427
+    },
+    {
+      "epoch": 0.7480719794344473,
+      "grad_norm": 0.15378192327513224,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 31428
+    },
+    {
+      "epoch": 0.7480957821574789,
+      "grad_norm": 10.107861069411564,
+      "learning_rate": 2e-05,
+      "loss": 1.4775,
+      "step": 31429
+    },
+    {
+      "epoch": 0.7481195848805103,
+      "grad_norm": 40.51411110260306,
+      "learning_rate": 2e-05,
+      "loss": 1.2143,
+      "step": 31430
+    },
+    {
+      "epoch": 0.7481433876035418,
+      "grad_norm": 2.762454036533996,
+      "learning_rate": 2e-05,
+      "loss": 0.466,
+      "step": 31431
+    },
+    {
+      "epoch": 0.7481671903265733,
+      "grad_norm": 7.364974445025937,
+      "learning_rate": 2e-05,
+      "loss": 0.3524,
+      "step": 31432
+    },
+    {
+      "epoch": 0.7481909930496049,
+      "grad_norm": 5.376873235127833,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 31433
+    },
+    {
+      "epoch": 0.7482147957726364,
+      "grad_norm": 1.0770743202832023,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 31434
+    },
+    {
+      "epoch": 0.7482385984956679,
+      "grad_norm": 0.06788130538630283,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31435
+    },
+    {
+      "epoch": 0.7482624012186995,
+      "grad_norm": 1.2037254693345596,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 31436
+    },
+    {
+      "epoch": 0.748286203941731,
+      "grad_norm": 16.12532693037605,
+      "learning_rate": 2e-05,
+      "loss": 0.2454,
+      "step": 31437
+    },
+    {
+      "epoch": 0.7483100066647624,
+      "grad_norm": 4.0700119225252465,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 31438
+    },
+    {
+      "epoch": 0.7483338093877939,
+      "grad_norm": 2.5230052105948317,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 31439
+    },
+    {
+      "epoch": 0.7483576121108255,
+      "grad_norm": 0.4772794913574022,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 31440
+    },
+    {
+      "epoch": 0.748381414833857,
+      "grad_norm": 36.30582526346662,
+      "learning_rate": 2e-05,
+      "loss": 0.467,
+      "step": 31441
+    },
+    {
+      "epoch": 0.7484052175568885,
+      "grad_norm": 2.793973774201306,
+      "learning_rate": 2e-05,
+      "loss": 0.2598,
+      "step": 31442
+    },
+    {
+      "epoch": 0.74842902027992,
+      "grad_norm": 2.4202444028301837,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 31443
+    },
+    {
+      "epoch": 0.7484528230029516,
+      "grad_norm": 5.1051899039373,
+      "learning_rate": 2e-05,
+      "loss": 0.6208,
+      "step": 31444
+    },
+    {
+      "epoch": 0.7484766257259831,
+      "grad_norm": 1.083995426317347,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 31445
+    },
+    {
+      "epoch": 0.7485004284490145,
+      "grad_norm": 5.393951295121243,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 31446
+    },
+    {
+      "epoch": 0.748524231172046,
+      "grad_norm": 6.273858231493628,
+      "learning_rate": 2e-05,
+      "loss": 0.2309,
+      "step": 31447
+    },
+    {
+      "epoch": 0.7485480338950776,
+      "grad_norm": 0.18395195546466714,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 31448
+    },
+    {
+      "epoch": 0.7485718366181091,
+      "grad_norm": 0.07995873444097634,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31449
+    },
+    {
+      "epoch": 0.7485956393411406,
+      "grad_norm": 5.976531684653616,
+      "learning_rate": 2e-05,
+      "loss": 0.8146,
+      "step": 31450
+    },
+    {
+      "epoch": 0.7486194420641722,
+      "grad_norm": 3.1518587470841317,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 31451
+    },
+    {
+      "epoch": 0.7486432447872037,
+      "grad_norm": 3.38627061719264,
+      "learning_rate": 2e-05,
+      "loss": 0.6348,
+      "step": 31452
+    },
+    {
+      "epoch": 0.7486670475102352,
+      "grad_norm": 12.746214704714125,
+      "learning_rate": 2e-05,
+      "loss": 1.1148,
+      "step": 31453
+    },
+    {
+      "epoch": 0.7486908502332666,
+      "grad_norm": 20.681825471960906,
+      "learning_rate": 2e-05,
+      "loss": 0.1144,
+      "step": 31454
+    },
+    {
+      "epoch": 0.7487146529562982,
+      "grad_norm": 4.005744873128379,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 31455
+    },
+    {
+      "epoch": 0.7487384556793297,
+      "grad_norm": 0.16914052177698144,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31456
+    },
+    {
+      "epoch": 0.7487622584023612,
+      "grad_norm": 0.14029323175060587,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 31457
+    },
+    {
+      "epoch": 0.7487860611253927,
+      "grad_norm": 0.3270115776801284,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 31458
+    },
+    {
+      "epoch": 0.7488098638484243,
+      "grad_norm": 5.613939078385298,
+      "learning_rate": 2e-05,
+      "loss": 0.4253,
+      "step": 31459
+    },
+    {
+      "epoch": 0.7488336665714558,
+      "grad_norm": 3.0809698143793076,
+      "learning_rate": 2e-05,
+      "loss": 0.6161,
+      "step": 31460
+    },
+    {
+      "epoch": 0.7488574692944873,
+      "grad_norm": 6.253543772547009,
+      "learning_rate": 2e-05,
+      "loss": 0.4954,
+      "step": 31461
+    },
+    {
+      "epoch": 0.7488812720175188,
+      "grad_norm": 2.6459082293024743,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 31462
+    },
+    {
+      "epoch": 0.7489050747405503,
+      "grad_norm": 2.0811675069789666,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 31463
+    },
+    {
+      "epoch": 0.7489288774635818,
+      "grad_norm": 6.56361228830382,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 31464
+    },
+    {
+      "epoch": 0.7489526801866133,
+      "grad_norm": 6.6423617667459975,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 31465
+    },
+    {
+      "epoch": 0.7489764829096449,
+      "grad_norm": 7.60266258115209,
+      "learning_rate": 2e-05,
+      "loss": 0.4182,
+      "step": 31466
+    },
+    {
+      "epoch": 0.7490002856326764,
+      "grad_norm": 3.382082798063295,
+      "learning_rate": 2e-05,
+      "loss": 0.1613,
+      "step": 31467
+    },
+    {
+      "epoch": 0.7490240883557079,
+      "grad_norm": 6.145589186869199,
+      "learning_rate": 2e-05,
+      "loss": 1.1749,
+      "step": 31468
+    },
+    {
+      "epoch": 0.7490478910787394,
+      "grad_norm": 4.115178724606537,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 31469
+    },
+    {
+      "epoch": 0.749071693801771,
+      "grad_norm": 3.1900785264364613,
+      "learning_rate": 2e-05,
+      "loss": 0.3649,
+      "step": 31470
+    },
+    {
+      "epoch": 0.7490954965248025,
+      "grad_norm": 0.24351965024980388,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31471
+    },
+    {
+      "epoch": 0.7491192992478339,
+      "grad_norm": 2.2363613212864473,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 31472
+    },
+    {
+      "epoch": 0.7491431019708654,
+      "grad_norm": 4.449137966080867,
+      "learning_rate": 2e-05,
+      "loss": 0.6858,
+      "step": 31473
+    },
+    {
+      "epoch": 0.749166904693897,
+      "grad_norm": 0.1807721426165287,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 31474
+    },
+    {
+      "epoch": 0.7491907074169285,
+      "grad_norm": 2.198260533864801,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 31475
+    },
+    {
+      "epoch": 0.74921451013996,
+      "grad_norm": 1.8966552118652207,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 31476
+    },
+    {
+      "epoch": 0.7492383128629915,
+      "grad_norm": 5.0883450792487155,
+      "learning_rate": 2e-05,
+      "loss": 1.2042,
+      "step": 31477
+    },
+    {
+      "epoch": 0.7492621155860231,
+      "grad_norm": 4.2344683826320555,
+      "learning_rate": 2e-05,
+      "loss": 0.1607,
+      "step": 31478
+    },
+    {
+      "epoch": 0.7492859183090546,
+      "grad_norm": 3.1074177157098544,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 31479
+    },
+    {
+      "epoch": 0.749309721032086,
+      "grad_norm": 4.788472354852793,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 31480
+    },
+    {
+      "epoch": 0.7493335237551176,
+      "grad_norm": 1.4745326785824353,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 31481
+    },
+    {
+      "epoch": 0.7493573264781491,
+      "grad_norm": 33.578957912275996,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 31482
+    },
+    {
+      "epoch": 0.7493811292011806,
+      "grad_norm": 6.645665048083601,
+      "learning_rate": 2e-05,
+      "loss": 0.4981,
+      "step": 31483
+    },
+    {
+      "epoch": 0.7494049319242121,
+      "grad_norm": 4.103596260756264,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 31484
+    },
+    {
+      "epoch": 0.7494287346472437,
+      "grad_norm": 0.44301136063821206,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 31485
+    },
+    {
+      "epoch": 0.7494525373702752,
+      "grad_norm": 10.990181346558371,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 31486
+    },
+    {
+      "epoch": 0.7494763400933067,
+      "grad_norm": 5.149232990400987,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 31487
+    },
+    {
+      "epoch": 0.7495001428163381,
+      "grad_norm": 0.16252034737337537,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 31488
+    },
+    {
+      "epoch": 0.7495239455393697,
+      "grad_norm": 1.0801937951010359,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 31489
+    },
+    {
+      "epoch": 0.7495477482624012,
+      "grad_norm": 1.4838721943841038,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 31490
+    },
+    {
+      "epoch": 0.7495715509854327,
+      "grad_norm": 0.09140572767244365,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31491
+    },
+    {
+      "epoch": 0.7495953537084642,
+      "grad_norm": 11.657581836412506,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 31492
+    },
+    {
+      "epoch": 0.7496191564314958,
+      "grad_norm": 0.25314123873992994,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 31493
+    },
+    {
+      "epoch": 0.7496429591545273,
+      "grad_norm": 144.96631678841192,
+      "learning_rate": 2e-05,
+      "loss": 1.3011,
+      "step": 31494
+    },
+    {
+      "epoch": 0.7496667618775588,
+      "grad_norm": 6.466989899850041,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 31495
+    },
+    {
+      "epoch": 0.7496905646005904,
+      "grad_norm": 0.1768375433966788,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 31496
+    },
+    {
+      "epoch": 0.7497143673236218,
+      "grad_norm": 0.4890703244454523,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 31497
+    },
+    {
+      "epoch": 0.7497381700466533,
+      "grad_norm": 4.758533019971229,
+      "learning_rate": 2e-05,
+      "loss": 0.3788,
+      "step": 31498
+    },
+    {
+      "epoch": 0.7497619727696848,
+      "grad_norm": 3.8715631323057886,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 31499
+    },
+    {
+      "epoch": 0.7497857754927164,
+      "grad_norm": 5.723544701552999,
+      "learning_rate": 2e-05,
+      "loss": 0.9439,
+      "step": 31500
+    },
+    {
+      "epoch": 0.7498095782157479,
+      "grad_norm": 3.6908430677665627,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 31501
+    },
+    {
+      "epoch": 0.7498333809387794,
+      "grad_norm": 5.9603596329398325,
+      "learning_rate": 2e-05,
+      "loss": 1.1496,
+      "step": 31502
+    },
+    {
+      "epoch": 0.7498571836618109,
+      "grad_norm": 3.412453670563195,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 31503
+    },
+    {
+      "epoch": 0.7498809863848425,
+      "grad_norm": 4.689804303174976,
+      "learning_rate": 2e-05,
+      "loss": 0.9965,
+      "step": 31504
+    },
+    {
+      "epoch": 0.749904789107874,
+      "grad_norm": 0.0928424277797074,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 31505
+    },
+    {
+      "epoch": 0.7499285918309054,
+      "grad_norm": 1.5266619001491093,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 31506
+    },
+    {
+      "epoch": 0.7499523945539369,
+      "grad_norm": 0.603636207646619,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 31507
+    },
+    {
+      "epoch": 0.7499761972769685,
+      "grad_norm": 5.545367178705525,
+      "learning_rate": 2e-05,
+      "loss": 1.1008,
+      "step": 31508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8592693492580127,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 31509
+    },
+    {
+      "epoch": 0.7500238027230315,
+      "grad_norm": 4.506316210420332,
+      "learning_rate": 2e-05,
+      "loss": 0.2833,
+      "step": 31510
+    },
+    {
+      "epoch": 0.7500476054460631,
+      "grad_norm": 9.063771590215243,
+      "learning_rate": 2e-05,
+      "loss": 1.8436,
+      "step": 31511
+    },
+    {
+      "epoch": 0.7500714081690946,
+      "grad_norm": 0.03817595121796805,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 31512
+    },
+    {
+      "epoch": 0.750095210892126,
+      "grad_norm": 52.83554830393421,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 31513
+    },
+    {
+      "epoch": 0.7501190136151575,
+      "grad_norm": 5.47129905405229,
+      "learning_rate": 2e-05,
+      "loss": 1.3631,
+      "step": 31514
+    },
+    {
+      "epoch": 0.7501428163381891,
+      "grad_norm": 46.661662700431435,
+      "learning_rate": 2e-05,
+      "loss": 1.3986,
+      "step": 31515
+    },
+    {
+      "epoch": 0.7501666190612206,
+      "grad_norm": 6.69204201954928,
+      "learning_rate": 2e-05,
+      "loss": 0.8436,
+      "step": 31516
+    },
+    {
+      "epoch": 0.7501904217842521,
+      "grad_norm": 4.6608418895956385,
+      "learning_rate": 2e-05,
+      "loss": 0.2709,
+      "step": 31517
+    },
+    {
+      "epoch": 0.7502142245072836,
+      "grad_norm": 0.7606998347071272,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 31518
+    },
+    {
+      "epoch": 0.7502380272303152,
+      "grad_norm": 0.12214928593976034,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 31519
+    },
+    {
+      "epoch": 0.7502618299533467,
+      "grad_norm": 5.499978910821246,
+      "learning_rate": 2e-05,
+      "loss": 1.1735,
+      "step": 31520
+    },
+    {
+      "epoch": 0.7502856326763782,
+      "grad_norm": 274.82846900267145,
+      "learning_rate": 2e-05,
+      "loss": 0.454,
+      "step": 31521
+    },
+    {
+      "epoch": 0.7503094353994096,
+      "grad_norm": 200.92492364326944,
+      "learning_rate": 2e-05,
+      "loss": 1.5129,
+      "step": 31522
+    },
+    {
+      "epoch": 0.7503332381224412,
+      "grad_norm": 2.7312015170175723,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 31523
+    },
+    {
+      "epoch": 0.7503570408454727,
+      "grad_norm": 7.0651039324165525,
+      "learning_rate": 2e-05,
+      "loss": 0.3377,
+      "step": 31524
+    },
+    {
+      "epoch": 0.7503808435685042,
+      "grad_norm": 2.4575420964234747,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 31525
+    },
+    {
+      "epoch": 0.7504046462915358,
+      "grad_norm": 0.2542886880591309,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 31526
+    },
+    {
+      "epoch": 0.7504284490145673,
+      "grad_norm": 3.897854515857738,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 31527
+    },
+    {
+      "epoch": 0.7504522517375988,
+      "grad_norm": 5.270343549885352,
+      "learning_rate": 2e-05,
+      "loss": 0.883,
+      "step": 31528
+    },
+    {
+      "epoch": 0.7504760544606303,
+      "grad_norm": 0.758026708191712,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 31529
+    },
+    {
+      "epoch": 0.7504998571836619,
+      "grad_norm": 4.159235733366763,
+      "learning_rate": 2e-05,
+      "loss": 0.4385,
+      "step": 31530
+    },
+    {
+      "epoch": 0.7505236599066933,
+      "grad_norm": 2.1567732799828767,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 31531
+    },
+    {
+      "epoch": 0.7505474626297248,
+      "grad_norm": 12.999936668415703,
+      "learning_rate": 2e-05,
+      "loss": 0.6825,
+      "step": 31532
+    },
+    {
+      "epoch": 0.7505712653527563,
+      "grad_norm": 0.8230904520517238,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 31533
+    },
+    {
+      "epoch": 0.7505950680757879,
+      "grad_norm": 6.932541666344651,
+      "learning_rate": 2e-05,
+      "loss": 0.9815,
+      "step": 31534
+    },
+    {
+      "epoch": 0.7506188707988194,
+      "grad_norm": 0.31954266542185406,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 31535
+    },
+    {
+      "epoch": 0.7506426735218509,
+      "grad_norm": 0.9410288808545222,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 31536
+    },
+    {
+      "epoch": 0.7506664762448824,
+      "grad_norm": 2.6627308541735353,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 31537
+    },
+    {
+      "epoch": 0.750690278967914,
+      "grad_norm": 5.429916937922317,
+      "learning_rate": 2e-05,
+      "loss": 1.0458,
+      "step": 31538
+    },
+    {
+      "epoch": 0.7507140816909454,
+      "grad_norm": 3.283114186204969,
+      "learning_rate": 2e-05,
+      "loss": 0.2655,
+      "step": 31539
+    },
+    {
+      "epoch": 0.7507378844139769,
+      "grad_norm": 4.851830055843006,
+      "learning_rate": 2e-05,
+      "loss": 0.869,
+      "step": 31540
+    },
+    {
+      "epoch": 0.7507616871370085,
+      "grad_norm": 3.264831597258166,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 31541
+    },
+    {
+      "epoch": 0.75078548986004,
+      "grad_norm": 1.2630461174765666,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 31542
+    },
+    {
+      "epoch": 0.7508092925830715,
+      "grad_norm": 0.22845941141677073,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 31543
+    },
+    {
+      "epoch": 0.750833095306103,
+      "grad_norm": 1.9180309060201215,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 31544
+    },
+    {
+      "epoch": 0.7508568980291346,
+      "grad_norm": 0.1358660156975456,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 31545
+    },
+    {
+      "epoch": 0.7508807007521661,
+      "grad_norm": 2.1733892340380376,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 31546
+    },
+    {
+      "epoch": 0.7509045034751975,
+      "grad_norm": 1.1414212926037413,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 31547
+    },
+    {
+      "epoch": 0.750928306198229,
+      "grad_norm": 1.0120056277436686,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 31548
+    },
+    {
+      "epoch": 0.7509521089212606,
+      "grad_norm": 0.9476907724658671,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 31549
+    },
+    {
+      "epoch": 0.7509759116442921,
+      "grad_norm": 15.260174499688699,
+      "learning_rate": 2e-05,
+      "loss": 0.5831,
+      "step": 31550
+    },
+    {
+      "epoch": 0.7509997143673236,
+      "grad_norm": 11.997130116438464,
+      "learning_rate": 2e-05,
+      "loss": 0.703,
+      "step": 31551
+    },
+    {
+      "epoch": 0.7510235170903551,
+      "grad_norm": 0.6155042849975355,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 31552
+    },
+    {
+      "epoch": 0.7510473198133867,
+      "grad_norm": 0.20797412689061395,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 31553
+    },
+    {
+      "epoch": 0.7510711225364182,
+      "grad_norm": 5.802740093020592,
+      "learning_rate": 2e-05,
+      "loss": 0.9793,
+      "step": 31554
+    },
+    {
+      "epoch": 0.7510949252594497,
+      "grad_norm": 0.40778219205790706,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 31555
+    },
+    {
+      "epoch": 0.7511187279824812,
+      "grad_norm": 0.4624965538847819,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 31556
+    },
+    {
+      "epoch": 0.7511425307055127,
+      "grad_norm": 6.0931604662195715,
+      "learning_rate": 2e-05,
+      "loss": 0.1977,
+      "step": 31557
+    },
+    {
+      "epoch": 0.7511663334285442,
+      "grad_norm": 1.2505502713301937,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 31558
+    },
+    {
+      "epoch": 0.7511901361515757,
+      "grad_norm": 5.39135423408418,
+      "learning_rate": 2e-05,
+      "loss": 0.9428,
+      "step": 31559
+    },
+    {
+      "epoch": 0.7512139388746073,
+      "grad_norm": 3.315026005659002,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 31560
+    },
+    {
+      "epoch": 0.7512377415976388,
+      "grad_norm": 5.703640059163223,
+      "learning_rate": 2e-05,
+      "loss": 1.5453,
+      "step": 31561
+    },
+    {
+      "epoch": 0.7512615443206703,
+      "grad_norm": 2.2929954704998923,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 31562
+    },
+    {
+      "epoch": 0.7512853470437018,
+      "grad_norm": 4.612235285782111,
+      "learning_rate": 2e-05,
+      "loss": 0.943,
+      "step": 31563
+    },
+    {
+      "epoch": 0.7513091497667334,
+      "grad_norm": 0.3975570060696361,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 31564
+    },
+    {
+      "epoch": 0.7513329524897648,
+      "grad_norm": 13.00421384754027,
+      "learning_rate": 2e-05,
+      "loss": 1.1026,
+      "step": 31565
+    },
+    {
+      "epoch": 0.7513567552127963,
+      "grad_norm": 4.027663616177299,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 31566
+    },
+    {
+      "epoch": 0.7513805579358278,
+      "grad_norm": 0.2739536672846083,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 31567
+    },
+    {
+      "epoch": 0.7514043606588594,
+      "grad_norm": 4.388249820883821,
+      "learning_rate": 2e-05,
+      "loss": 0.9458,
+      "step": 31568
+    },
+    {
+      "epoch": 0.7514281633818909,
+      "grad_norm": 6.573900484155897,
+      "learning_rate": 2e-05,
+      "loss": 0.6163,
+      "step": 31569
+    },
+    {
+      "epoch": 0.7514519661049224,
+      "grad_norm": 6.153952823028163,
+      "learning_rate": 2e-05,
+      "loss": 0.2854,
+      "step": 31570
+    },
+    {
+      "epoch": 0.751475768827954,
+      "grad_norm": 5.523183240652074,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 31571
+    },
+    {
+      "epoch": 0.7514995715509855,
+      "grad_norm": 3.2934074761828245,
+      "learning_rate": 2e-05,
+      "loss": 0.6588,
+      "step": 31572
+    },
+    {
+      "epoch": 0.7515233742740169,
+      "grad_norm": 0.3322918213441424,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 31573
+    },
+    {
+      "epoch": 0.7515471769970484,
+      "grad_norm": 5.131277547139036,
+      "learning_rate": 2e-05,
+      "loss": 0.7583,
+      "step": 31574
+    },
+    {
+      "epoch": 0.75157097972008,
+      "grad_norm": 3.572184488726002,
+      "learning_rate": 2e-05,
+      "loss": 0.4619,
+      "step": 31575
+    },
+    {
+      "epoch": 0.7515947824431115,
+      "grad_norm": 3.2366024842172774,
+      "learning_rate": 2e-05,
+      "loss": 0.2676,
+      "step": 31576
+    },
+    {
+      "epoch": 0.751618585166143,
+      "grad_norm": 0.46920925810213615,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 31577
+    },
+    {
+      "epoch": 0.7516423878891745,
+      "grad_norm": 3.229826349457633,
+      "learning_rate": 2e-05,
+      "loss": 0.7045,
+      "step": 31578
+    },
+    {
+      "epoch": 0.7516661906122061,
+      "grad_norm": 37.0688655932045,
+      "learning_rate": 2e-05,
+      "loss": 0.6104,
+      "step": 31579
+    },
+    {
+      "epoch": 0.7516899933352376,
+      "grad_norm": 4.767783934769401,
+      "learning_rate": 2e-05,
+      "loss": 1.016,
+      "step": 31580
+    },
+    {
+      "epoch": 0.751713796058269,
+      "grad_norm": 1.2678415269683119,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 31581
+    },
+    {
+      "epoch": 0.7517375987813005,
+      "grad_norm": 6.480988724382769,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 31582
+    },
+    {
+      "epoch": 0.7517614015043321,
+      "grad_norm": 11.135966490766664,
+      "learning_rate": 2e-05,
+      "loss": 0.5653,
+      "step": 31583
+    },
+    {
+      "epoch": 0.7517852042273636,
+      "grad_norm": 2.9034597766296715,
+      "learning_rate": 2e-05,
+      "loss": 0.5116,
+      "step": 31584
+    },
+    {
+      "epoch": 0.7518090069503951,
+      "grad_norm": 1.1733384071324129,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 31585
+    },
+    {
+      "epoch": 0.7518328096734267,
+      "grad_norm": 0.04301399651146659,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 31586
+    },
+    {
+      "epoch": 0.7518566123964582,
+      "grad_norm": 0.7057146066287413,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 31587
+    },
+    {
+      "epoch": 0.7518804151194897,
+      "grad_norm": 4.954498873578544,
+      "learning_rate": 2e-05,
+      "loss": 0.7282,
+      "step": 31588
+    },
+    {
+      "epoch": 0.7519042178425211,
+      "grad_norm": 3.6456963235273667,
+      "learning_rate": 2e-05,
+      "loss": 0.7814,
+      "step": 31589
+    },
+    {
+      "epoch": 0.7519280205655527,
+      "grad_norm": 3.3786043434059794,
+      "learning_rate": 2e-05,
+      "loss": 0.4727,
+      "step": 31590
+    },
+    {
+      "epoch": 0.7519518232885842,
+      "grad_norm": 3.2007317127514385,
+      "learning_rate": 2e-05,
+      "loss": 0.4697,
+      "step": 31591
+    },
+    {
+      "epoch": 0.7519756260116157,
+      "grad_norm": 0.04152211134267036,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31592
+    },
+    {
+      "epoch": 0.7519994287346472,
+      "grad_norm": 0.02855875020547455,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31593
+    },
+    {
+      "epoch": 0.7520232314576788,
+      "grad_norm": 5.961272591206373,
+      "learning_rate": 2e-05,
+      "loss": 0.4755,
+      "step": 31594
+    },
+    {
+      "epoch": 0.7520470341807103,
+      "grad_norm": 5.228612572578491,
+      "learning_rate": 2e-05,
+      "loss": 1.3237,
+      "step": 31595
+    },
+    {
+      "epoch": 0.7520708369037418,
+      "grad_norm": 5.145124744040159,
+      "learning_rate": 2e-05,
+      "loss": 0.7519,
+      "step": 31596
+    },
+    {
+      "epoch": 0.7520946396267733,
+      "grad_norm": 4.460736701171326,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 31597
+    },
+    {
+      "epoch": 0.7521184423498048,
+      "grad_norm": 0.7276339619011541,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 31598
+    },
+    {
+      "epoch": 0.7521422450728363,
+      "grad_norm": 4.621606795814689,
+      "learning_rate": 2e-05,
+      "loss": 1.0847,
+      "step": 31599
+    },
+    {
+      "epoch": 0.7521660477958678,
+      "grad_norm": 12.52176094523684,
+      "learning_rate": 2e-05,
+      "loss": 0.1494,
+      "step": 31600
+    },
+    {
+      "epoch": 0.7521898505188994,
+      "grad_norm": 1.3593085097996433,
+      "learning_rate": 2e-05,
+      "loss": 0.0478,
+      "step": 31601
+    },
+    {
+      "epoch": 0.7522136532419309,
+      "grad_norm": 1.5621996881905433,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 31602
+    },
+    {
+      "epoch": 0.7522374559649624,
+      "grad_norm": 24.204597138904997,
+      "learning_rate": 2e-05,
+      "loss": 1.0134,
+      "step": 31603
+    },
+    {
+      "epoch": 0.7522612586879939,
+      "grad_norm": 9.955274793893118,
+      "learning_rate": 2e-05,
+      "loss": 0.8303,
+      "step": 31604
+    },
+    {
+      "epoch": 0.7522850614110255,
+      "grad_norm": 7.2523005034506065,
+      "learning_rate": 2e-05,
+      "loss": 0.7081,
+      "step": 31605
+    },
+    {
+      "epoch": 0.752308864134057,
+      "grad_norm": 4.466522804389508,
+      "learning_rate": 2e-05,
+      "loss": 0.7964,
+      "step": 31606
+    },
+    {
+      "epoch": 0.7523326668570884,
+      "grad_norm": 3.2133491512647288,
+      "learning_rate": 2e-05,
+      "loss": 0.2763,
+      "step": 31607
+    },
+    {
+      "epoch": 0.7523564695801199,
+      "grad_norm": 4.178916283891307,
+      "learning_rate": 2e-05,
+      "loss": 0.598,
+      "step": 31608
+    },
+    {
+      "epoch": 0.7523802723031515,
+      "grad_norm": 8.734201593007672,
+      "learning_rate": 2e-05,
+      "loss": 0.5177,
+      "step": 31609
+    },
+    {
+      "epoch": 0.752404075026183,
+      "grad_norm": 4.7581474567125595,
+      "learning_rate": 2e-05,
+      "loss": 0.5324,
+      "step": 31610
+    },
+    {
+      "epoch": 0.7524278777492145,
+      "grad_norm": 4.4712770339357615,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 31611
+    },
+    {
+      "epoch": 0.752451680472246,
+      "grad_norm": 0.50650783943749,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 31612
+    },
+    {
+      "epoch": 0.7524754831952776,
+      "grad_norm": 3.0477563507957046,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 31613
+    },
+    {
+      "epoch": 0.7524992859183091,
+      "grad_norm": 56.60061608558389,
+      "learning_rate": 2e-05,
+      "loss": 0.2292,
+      "step": 31614
+    },
+    {
+      "epoch": 0.7525230886413405,
+      "grad_norm": 4.731916063914891,
+      "learning_rate": 2e-05,
+      "loss": 0.9713,
+      "step": 31615
+    },
+    {
+      "epoch": 0.7525468913643721,
+      "grad_norm": 0.5734337652698401,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31616
+    },
+    {
+      "epoch": 0.7525706940874036,
+      "grad_norm": 0.3154473082622457,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 31617
+    },
+    {
+      "epoch": 0.7525944968104351,
+      "grad_norm": 6.532434012508605,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 31618
+    },
+    {
+      "epoch": 0.7526182995334666,
+      "grad_norm": 5.239508814642344,
+      "learning_rate": 2e-05,
+      "loss": 0.9975,
+      "step": 31619
+    },
+    {
+      "epoch": 0.7526421022564982,
+      "grad_norm": 1.5983020989506598,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 31620
+    },
+    {
+      "epoch": 0.7526659049795297,
+      "grad_norm": 0.7325955499997947,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 31621
+    },
+    {
+      "epoch": 0.7526897077025612,
+      "grad_norm": 0.06519588440951118,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31622
+    },
+    {
+      "epoch": 0.7527135104255926,
+      "grad_norm": 5.26404884721172,
+      "learning_rate": 2e-05,
+      "loss": 0.5571,
+      "step": 31623
+    },
+    {
+      "epoch": 0.7527373131486242,
+      "grad_norm": 8.882190680398885,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 31624
+    },
+    {
+      "epoch": 0.7527611158716557,
+      "grad_norm": 2.02977701781776,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 31625
+    },
+    {
+      "epoch": 0.7527849185946872,
+      "grad_norm": 4.545714337176319,
+      "learning_rate": 2e-05,
+      "loss": 0.3606,
+      "step": 31626
+    },
+    {
+      "epoch": 0.7528087213177187,
+      "grad_norm": 5.930613682864068,
+      "learning_rate": 2e-05,
+      "loss": 1.3748,
+      "step": 31627
+    },
+    {
+      "epoch": 0.7528325240407503,
+      "grad_norm": 6.625042234474426,
+      "learning_rate": 2e-05,
+      "loss": 0.9937,
+      "step": 31628
+    },
+    {
+      "epoch": 0.7528563267637818,
+      "grad_norm": 10.634757012606023,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 31629
+    },
+    {
+      "epoch": 0.7528801294868133,
+      "grad_norm": 2.660585517253729,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 31630
+    },
+    {
+      "epoch": 0.7529039322098449,
+      "grad_norm": 0.6764993233199685,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 31631
+    },
+    {
+      "epoch": 0.7529277349328763,
+      "grad_norm": 2.8110981893190288,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 31632
+    },
+    {
+      "epoch": 0.7529515376559078,
+      "grad_norm": 3.0965379391616668,
+      "learning_rate": 2e-05,
+      "loss": 0.1184,
+      "step": 31633
+    },
+    {
+      "epoch": 0.7529753403789393,
+      "grad_norm": 0.8706538148314622,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 31634
+    },
+    {
+      "epoch": 0.7529991431019709,
+      "grad_norm": 1.155115461990325,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 31635
+    },
+    {
+      "epoch": 0.7530229458250024,
+      "grad_norm": 5.3489998882640455,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 31636
+    },
+    {
+      "epoch": 0.7530467485480339,
+      "grad_norm": 0.08634500154049908,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 31637
+    },
+    {
+      "epoch": 0.7530705512710654,
+      "grad_norm": 3.526875122507091,
+      "learning_rate": 2e-05,
+      "loss": 0.2827,
+      "step": 31638
+    },
+    {
+      "epoch": 0.753094353994097,
+      "grad_norm": 5.247136965822788,
+      "learning_rate": 2e-05,
+      "loss": 1.0679,
+      "step": 31639
+    },
+    {
+      "epoch": 0.7531181567171285,
+      "grad_norm": 109.3187112135803,
+      "learning_rate": 2e-05,
+      "loss": 0.8151,
+      "step": 31640
+    },
+    {
+      "epoch": 0.7531419594401599,
+      "grad_norm": 4.525154059013395,
+      "learning_rate": 2e-05,
+      "loss": 0.576,
+      "step": 31641
+    },
+    {
+      "epoch": 0.7531657621631914,
+      "grad_norm": 3.4095124986026395,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 31642
+    },
+    {
+      "epoch": 0.753189564886223,
+      "grad_norm": 0.09309188906556214,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31643
+    },
+    {
+      "epoch": 0.7532133676092545,
+      "grad_norm": 6.893445345211265,
+      "learning_rate": 2e-05,
+      "loss": 1.1403,
+      "step": 31644
+    },
+    {
+      "epoch": 0.753237170332286,
+      "grad_norm": 0.14630664601249072,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31645
+    },
+    {
+      "epoch": 0.7532609730553175,
+      "grad_norm": 0.05564598679234274,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31646
+    },
+    {
+      "epoch": 0.7532847757783491,
+      "grad_norm": 0.9496122183661428,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 31647
+    },
+    {
+      "epoch": 0.7533085785013806,
+      "grad_norm": 10.284687465545263,
+      "learning_rate": 2e-05,
+      "loss": 0.8678,
+      "step": 31648
+    },
+    {
+      "epoch": 0.753332381224412,
+      "grad_norm": 3.29755915753663,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 31649
+    },
+    {
+      "epoch": 0.7533561839474436,
+      "grad_norm": 0.06787110381393904,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 31650
+    },
+    {
+      "epoch": 0.7533799866704751,
+      "grad_norm": 2.442467171723052,
+      "learning_rate": 2e-05,
+      "loss": 0.3772,
+      "step": 31651
+    },
+    {
+      "epoch": 0.7534037893935066,
+      "grad_norm": 36.71844026946165,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 31652
+    },
+    {
+      "epoch": 0.7534275921165381,
+      "grad_norm": 2.160519034774867,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 31653
+    },
+    {
+      "epoch": 0.7534513948395697,
+      "grad_norm": 8.225957262652932,
+      "learning_rate": 2e-05,
+      "loss": 0.7955,
+      "step": 31654
+    },
+    {
+      "epoch": 0.7534751975626012,
+      "grad_norm": 4.4241410755514154,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 31655
+    },
+    {
+      "epoch": 0.7534990002856327,
+      "grad_norm": 1.8752403378419211,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 31656
+    },
+    {
+      "epoch": 0.7535228030086641,
+      "grad_norm": 0.2535972087972428,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 31657
+    },
+    {
+      "epoch": 0.7535466057316957,
+      "grad_norm": 2.6016916480476846,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 31658
+    },
+    {
+      "epoch": 0.7535704084547272,
+      "grad_norm": 1.3367659309311029,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 31659
+    },
+    {
+      "epoch": 0.7535942111777587,
+      "grad_norm": 0.4872908980096342,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 31660
+    },
+    {
+      "epoch": 0.7536180139007902,
+      "grad_norm": 0.6762631865851215,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 31661
+    },
+    {
+      "epoch": 0.7536418166238218,
+      "grad_norm": 2.645721685380494,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 31662
+    },
+    {
+      "epoch": 0.7536656193468533,
+      "grad_norm": 0.22663007251611575,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 31663
+    },
+    {
+      "epoch": 0.7536894220698848,
+      "grad_norm": 0.026421954902574804,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31664
+    },
+    {
+      "epoch": 0.7537132247929164,
+      "grad_norm": 10.106886875287596,
+      "learning_rate": 2e-05,
+      "loss": 0.8311,
+      "step": 31665
+    },
+    {
+      "epoch": 0.7537370275159478,
+      "grad_norm": 1.332515454440616,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 31666
+    },
+    {
+      "epoch": 0.7537608302389793,
+      "grad_norm": 8.860786850792461,
+      "learning_rate": 2e-05,
+      "loss": 0.9819,
+      "step": 31667
+    },
+    {
+      "epoch": 0.7537846329620108,
+      "grad_norm": 4.248692217398026,
+      "learning_rate": 2e-05,
+      "loss": 0.5818,
+      "step": 31668
+    },
+    {
+      "epoch": 0.7538084356850424,
+      "grad_norm": 0.5358962929106458,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 31669
+    },
+    {
+      "epoch": 0.7538322384080739,
+      "grad_norm": 4.696647860858465,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 31670
+    },
+    {
+      "epoch": 0.7538560411311054,
+      "grad_norm": 0.04345303481154413,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31671
+    },
+    {
+      "epoch": 0.7538798438541369,
+      "grad_norm": 1.5422200562336807,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 31672
+    },
+    {
+      "epoch": 0.7539036465771685,
+      "grad_norm": 3.5696707230077607,
+      "learning_rate": 2e-05,
+      "loss": 0.3648,
+      "step": 31673
+    },
+    {
+      "epoch": 0.7539274493002,
+      "grad_norm": 0.11348126592590654,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 31674
+    },
+    {
+      "epoch": 0.7539512520232314,
+      "grad_norm": 3.4869378015431525,
+      "learning_rate": 2e-05,
+      "loss": 0.8349,
+      "step": 31675
+    },
+    {
+      "epoch": 0.7539750547462629,
+      "grad_norm": 1.3443857922477598,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 31676
+    },
+    {
+      "epoch": 0.7539988574692945,
+      "grad_norm": 8.712263408688083,
+      "learning_rate": 2e-05,
+      "loss": 1.1583,
+      "step": 31677
+    },
+    {
+      "epoch": 0.754022660192326,
+      "grad_norm": 2.6660725736932496,
+      "learning_rate": 2e-05,
+      "loss": 0.7585,
+      "step": 31678
+    },
+    {
+      "epoch": 0.7540464629153575,
+      "grad_norm": 5.555390397481378,
+      "learning_rate": 2e-05,
+      "loss": 1.358,
+      "step": 31679
+    },
+    {
+      "epoch": 0.7540702656383891,
+      "grad_norm": 3.4897633099845153,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 31680
+    },
+    {
+      "epoch": 0.7540940683614206,
+      "grad_norm": 0.5750111072649586,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31681
+    },
+    {
+      "epoch": 0.754117871084452,
+      "grad_norm": 0.5552383437485146,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 31682
+    },
+    {
+      "epoch": 0.7541416738074835,
+      "grad_norm": 3.647373421507689,
+      "learning_rate": 2e-05,
+      "loss": 0.7056,
+      "step": 31683
+    },
+    {
+      "epoch": 0.7541654765305151,
+      "grad_norm": 2.6197400536719857,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 31684
+    },
+    {
+      "epoch": 0.7541892792535466,
+      "grad_norm": 5.775872018488106,
+      "learning_rate": 2e-05,
+      "loss": 0.7881,
+      "step": 31685
+    },
+    {
+      "epoch": 0.7542130819765781,
+      "grad_norm": 3.1162630981541426,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 31686
+    },
+    {
+      "epoch": 0.7542368846996096,
+      "grad_norm": 7.384829899690661,
+      "learning_rate": 2e-05,
+      "loss": 0.2695,
+      "step": 31687
+    },
+    {
+      "epoch": 0.7542606874226412,
+      "grad_norm": 2.466637450038229,
+      "learning_rate": 2e-05,
+      "loss": 0.257,
+      "step": 31688
+    },
+    {
+      "epoch": 0.7542844901456727,
+      "grad_norm": 2.7260201136238513,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 31689
+    },
+    {
+      "epoch": 0.7543082928687042,
+      "grad_norm": 3.4556517489260066,
+      "learning_rate": 2e-05,
+      "loss": 0.5904,
+      "step": 31690
+    },
+    {
+      "epoch": 0.7543320955917356,
+      "grad_norm": 4.593963458728634,
+      "learning_rate": 2e-05,
+      "loss": 0.7868,
+      "step": 31691
+    },
+    {
+      "epoch": 0.7543558983147672,
+      "grad_norm": 9.124244530875425,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 31692
+    },
+    {
+      "epoch": 0.7543797010377987,
+      "grad_norm": 3.4383040066660264,
+      "learning_rate": 2e-05,
+      "loss": 0.6799,
+      "step": 31693
+    },
+    {
+      "epoch": 0.7544035037608302,
+      "grad_norm": 0.019839135441756674,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 31694
+    },
+    {
+      "epoch": 0.7544273064838618,
+      "grad_norm": 0.2963876169305034,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 31695
+    },
+    {
+      "epoch": 0.7544511092068933,
+      "grad_norm": 4.853557957305459,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 31696
+    },
+    {
+      "epoch": 0.7544749119299248,
+      "grad_norm": 1.9121075241024375,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 31697
+    },
+    {
+      "epoch": 0.7544987146529563,
+      "grad_norm": 0.29231990216701675,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 31698
+    },
+    {
+      "epoch": 0.7545225173759879,
+      "grad_norm": 4.541769495344135,
+      "learning_rate": 2e-05,
+      "loss": 1.0128,
+      "step": 31699
+    },
+    {
+      "epoch": 0.7545463200990193,
+      "grad_norm": 1.1692738041138708,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 31700
+    },
+    {
+      "epoch": 0.7545701228220508,
+      "grad_norm": 12.546797945684034,
+      "learning_rate": 2e-05,
+      "loss": 1.0637,
+      "step": 31701
+    },
+    {
+      "epoch": 0.7545939255450823,
+      "grad_norm": 3.009065626738113,
+      "learning_rate": 2e-05,
+      "loss": 0.1806,
+      "step": 31702
+    },
+    {
+      "epoch": 0.7546177282681139,
+      "grad_norm": 4.910448356022948,
+      "learning_rate": 2e-05,
+      "loss": 0.9439,
+      "step": 31703
+    },
+    {
+      "epoch": 0.7546415309911454,
+      "grad_norm": 3.865964354334686,
+      "learning_rate": 2e-05,
+      "loss": 0.417,
+      "step": 31704
+    },
+    {
+      "epoch": 0.7546653337141769,
+      "grad_norm": 4.754360789036024,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 31705
+    },
+    {
+      "epoch": 0.7546891364372084,
+      "grad_norm": 2.380458353852221,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 31706
+    },
+    {
+      "epoch": 0.75471293916024,
+      "grad_norm": 18.199882596000712,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 31707
+    },
+    {
+      "epoch": 0.7547367418832714,
+      "grad_norm": 3.4750934128791027,
+      "learning_rate": 2e-05,
+      "loss": 0.5647,
+      "step": 31708
+    },
+    {
+      "epoch": 0.7547605446063029,
+      "grad_norm": 5.96687139722163,
+      "learning_rate": 2e-05,
+      "loss": 1.2537,
+      "step": 31709
+    },
+    {
+      "epoch": 0.7547843473293345,
+      "grad_norm": 0.6889748326093765,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 31710
+    },
+    {
+      "epoch": 0.754808150052366,
+      "grad_norm": 4.606732641645523,
+      "learning_rate": 2e-05,
+      "loss": 0.5778,
+      "step": 31711
+    },
+    {
+      "epoch": 0.7548319527753975,
+      "grad_norm": 3.864339265643998,
+      "learning_rate": 2e-05,
+      "loss": 0.445,
+      "step": 31712
+    },
+    {
+      "epoch": 0.754855755498429,
+      "grad_norm": 11.581727719750198,
+      "learning_rate": 2e-05,
+      "loss": 1.1201,
+      "step": 31713
+    },
+    {
+      "epoch": 0.7548795582214606,
+      "grad_norm": 1.477252055750534,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 31714
+    },
+    {
+      "epoch": 0.7549033609444921,
+      "grad_norm": 0.664173343059194,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 31715
+    },
+    {
+      "epoch": 0.7549271636675235,
+      "grad_norm": 16.858878313075728,
+      "learning_rate": 2e-05,
+      "loss": 1.0103,
+      "step": 31716
+    },
+    {
+      "epoch": 0.754950966390555,
+      "grad_norm": 3.532379669937472,
+      "learning_rate": 2e-05,
+      "loss": 0.468,
+      "step": 31717
+    },
+    {
+      "epoch": 0.7549747691135866,
+      "grad_norm": 5.41319285899205,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 31718
+    },
+    {
+      "epoch": 0.7549985718366181,
+      "grad_norm": 4.186396813840218,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 31719
+    },
+    {
+      "epoch": 0.7550223745596496,
+      "grad_norm": 3.0933241847648323,
+      "learning_rate": 2e-05,
+      "loss": 0.5103,
+      "step": 31720
+    },
+    {
+      "epoch": 0.7550461772826811,
+      "grad_norm": 0.7542125665942403,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 31721
+    },
+    {
+      "epoch": 0.7550699800057127,
+      "grad_norm": 2.109005302536323,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 31722
+    },
+    {
+      "epoch": 0.7550937827287442,
+      "grad_norm": 3.6189122918272596,
+      "learning_rate": 2e-05,
+      "loss": 0.194,
+      "step": 31723
+    },
+    {
+      "epoch": 0.7551175854517757,
+      "grad_norm": 4.221268464421715,
+      "learning_rate": 2e-05,
+      "loss": 0.5312,
+      "step": 31724
+    },
+    {
+      "epoch": 0.7551413881748072,
+      "grad_norm": 12.43239173743193,
+      "learning_rate": 2e-05,
+      "loss": 1.3857,
+      "step": 31725
+    },
+    {
+      "epoch": 0.7551651908978387,
+      "grad_norm": 0.051200003139375776,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 31726
+    },
+    {
+      "epoch": 0.7551889936208702,
+      "grad_norm": 3.3247317092449498,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 31727
+    },
+    {
+      "epoch": 0.7552127963439017,
+      "grad_norm": 3.444438500045505,
+      "learning_rate": 2e-05,
+      "loss": 0.5335,
+      "step": 31728
+    },
+    {
+      "epoch": 0.7552365990669333,
+      "grad_norm": 5.984395535861833,
+      "learning_rate": 2e-05,
+      "loss": 0.9973,
+      "step": 31729
+    },
+    {
+      "epoch": 0.7552604017899648,
+      "grad_norm": 4.027517400620579,
+      "learning_rate": 2e-05,
+      "loss": 0.5256,
+      "step": 31730
+    },
+    {
+      "epoch": 0.7552842045129963,
+      "grad_norm": 0.15461684040590176,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 31731
+    },
+    {
+      "epoch": 0.7553080072360278,
+      "grad_norm": 4.044819875552849,
+      "learning_rate": 2e-05,
+      "loss": 0.0682,
+      "step": 31732
+    },
+    {
+      "epoch": 0.7553318099590594,
+      "grad_norm": 1.1421702134994658,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 31733
+    },
+    {
+      "epoch": 0.7553556126820908,
+      "grad_norm": 6.826111234117782,
+      "learning_rate": 2e-05,
+      "loss": 1.3151,
+      "step": 31734
+    },
+    {
+      "epoch": 0.7553794154051223,
+      "grad_norm": 18.00857100170108,
+      "learning_rate": 2e-05,
+      "loss": 0.8265,
+      "step": 31735
+    },
+    {
+      "epoch": 0.7554032181281538,
+      "grad_norm": 11.494497862454002,
+      "learning_rate": 2e-05,
+      "loss": 0.5569,
+      "step": 31736
+    },
+    {
+      "epoch": 0.7554270208511854,
+      "grad_norm": 1.8391527694078127,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 31737
+    },
+    {
+      "epoch": 0.7554508235742169,
+      "grad_norm": 2.1604687166218053,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 31738
+    },
+    {
+      "epoch": 0.7554746262972484,
+      "grad_norm": 12.684435558527804,
+      "learning_rate": 2e-05,
+      "loss": 1.0785,
+      "step": 31739
+    },
+    {
+      "epoch": 0.75549842902028,
+      "grad_norm": 6.122106496566398,
+      "learning_rate": 2e-05,
+      "loss": 0.7734,
+      "step": 31740
+    },
+    {
+      "epoch": 0.7555222317433115,
+      "grad_norm": 3.9017632962005533,
+      "learning_rate": 2e-05,
+      "loss": 0.1364,
+      "step": 31741
+    },
+    {
+      "epoch": 0.7555460344663429,
+      "grad_norm": 0.06036122471599288,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 31742
+    },
+    {
+      "epoch": 0.7555698371893744,
+      "grad_norm": 2.317194045808274,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 31743
+    },
+    {
+      "epoch": 0.755593639912406,
+      "grad_norm": 1.7436731387641358,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 31744
+    },
+    {
+      "epoch": 0.7556174426354375,
+      "grad_norm": 1.4006379877930264,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 31745
+    },
+    {
+      "epoch": 0.755641245358469,
+      "grad_norm": 0.4823585177774886,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 31746
+    },
+    {
+      "epoch": 0.7556650480815005,
+      "grad_norm": 2.8278605127099143,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 31747
+    },
+    {
+      "epoch": 0.7556888508045321,
+      "grad_norm": 3.3224897537668556,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 31748
+    },
+    {
+      "epoch": 0.7557126535275636,
+      "grad_norm": 5.2407623261373635,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 31749
+    },
+    {
+      "epoch": 0.755736456250595,
+      "grad_norm": 0.6448358614981723,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 31750
+    },
+    {
+      "epoch": 0.7557602589736265,
+      "grad_norm": 5.52917790524024,
+      "learning_rate": 2e-05,
+      "loss": 0.6354,
+      "step": 31751
+    },
+    {
+      "epoch": 0.7557840616966581,
+      "grad_norm": 0.23964715890953148,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 31752
+    },
+    {
+      "epoch": 0.7558078644196896,
+      "grad_norm": 6.827063791724641,
+      "learning_rate": 2e-05,
+      "loss": 0.9336,
+      "step": 31753
+    },
+    {
+      "epoch": 0.7558316671427211,
+      "grad_norm": 3.1373564151885125,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 31754
+    },
+    {
+      "epoch": 0.7558554698657527,
+      "grad_norm": 2.481731969152452,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 31755
+    },
+    {
+      "epoch": 0.7558792725887842,
+      "grad_norm": 0.569104627107045,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 31756
+    },
+    {
+      "epoch": 0.7559030753118157,
+      "grad_norm": 4.590393441907662,
+      "learning_rate": 2e-05,
+      "loss": 0.406,
+      "step": 31757
+    },
+    {
+      "epoch": 0.7559268780348471,
+      "grad_norm": 5.793051276532598,
+      "learning_rate": 2e-05,
+      "loss": 1.0404,
+      "step": 31758
+    },
+    {
+      "epoch": 0.7559506807578787,
+      "grad_norm": 7.962564980029104,
+      "learning_rate": 2e-05,
+      "loss": 0.48,
+      "step": 31759
+    },
+    {
+      "epoch": 0.7559744834809102,
+      "grad_norm": 0.6525127527017845,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 31760
+    },
+    {
+      "epoch": 0.7559982862039417,
+      "grad_norm": 10.74594228886316,
+      "learning_rate": 2e-05,
+      "loss": 0.4557,
+      "step": 31761
+    },
+    {
+      "epoch": 0.7560220889269732,
+      "grad_norm": 5.662183032527515,
+      "learning_rate": 2e-05,
+      "loss": 1.0557,
+      "step": 31762
+    },
+    {
+      "epoch": 0.7560458916500048,
+      "grad_norm": 9.05909789424902,
+      "learning_rate": 2e-05,
+      "loss": 0.4339,
+      "step": 31763
+    },
+    {
+      "epoch": 0.7560696943730363,
+      "grad_norm": 4.0972215515284995,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 31764
+    },
+    {
+      "epoch": 0.7560934970960678,
+      "grad_norm": 2.994921322666963,
+      "learning_rate": 2e-05,
+      "loss": 0.0785,
+      "step": 31765
+    },
+    {
+      "epoch": 0.7561172998190993,
+      "grad_norm": 4.864972663745769,
+      "learning_rate": 2e-05,
+      "loss": 1.0627,
+      "step": 31766
+    },
+    {
+      "epoch": 0.7561411025421308,
+      "grad_norm": 4.4506455187844365,
+      "learning_rate": 2e-05,
+      "loss": 0.6867,
+      "step": 31767
+    },
+    {
+      "epoch": 0.7561649052651623,
+      "grad_norm": 0.984579365868127,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 31768
+    },
+    {
+      "epoch": 0.7561887079881938,
+      "grad_norm": 0.3883991808495862,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 31769
+    },
+    {
+      "epoch": 0.7562125107112254,
+      "grad_norm": 1.3955268870309108,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 31770
+    },
+    {
+      "epoch": 0.7562363134342569,
+      "grad_norm": 0.23177526990521688,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 31771
+    },
+    {
+      "epoch": 0.7562601161572884,
+      "grad_norm": 2.092393981435378,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 31772
+    },
+    {
+      "epoch": 0.7562839188803199,
+      "grad_norm": 0.5083479016550859,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 31773
+    },
+    {
+      "epoch": 0.7563077216033515,
+      "grad_norm": 5.583681875879402,
+      "learning_rate": 2e-05,
+      "loss": 0.8742,
+      "step": 31774
+    },
+    {
+      "epoch": 0.756331524326383,
+      "grad_norm": 2.7872611645082292,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 31775
+    },
+    {
+      "epoch": 0.7563553270494144,
+      "grad_norm": 1.6652604444645975,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 31776
+    },
+    {
+      "epoch": 0.7563791297724459,
+      "grad_norm": 3.483620853686835,
+      "learning_rate": 2e-05,
+      "loss": 0.7467,
+      "step": 31777
+    },
+    {
+      "epoch": 0.7564029324954775,
+      "grad_norm": 2.1259437120451725,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 31778
+    },
+    {
+      "epoch": 0.756426735218509,
+      "grad_norm": 3.6435938543701676,
+      "learning_rate": 2e-05,
+      "loss": 0.5626,
+      "step": 31779
+    },
+    {
+      "epoch": 0.7564505379415405,
+      "grad_norm": 8.209812314071371,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 31780
+    },
+    {
+      "epoch": 0.756474340664572,
+      "grad_norm": 2.14689161980772,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 31781
+    },
+    {
+      "epoch": 0.7564981433876036,
+      "grad_norm": 0.7161491838408302,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 31782
+    },
+    {
+      "epoch": 0.7565219461106351,
+      "grad_norm": 3.203505564251218,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 31783
+    },
+    {
+      "epoch": 0.7565457488336665,
+      "grad_norm": 3.3772076299643197,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 31784
+    },
+    {
+      "epoch": 0.7565695515566981,
+      "grad_norm": 6.9185726389681985,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 31785
+    },
+    {
+      "epoch": 0.7565933542797296,
+      "grad_norm": 3.226689641488189,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 31786
+    },
+    {
+      "epoch": 0.7566171570027611,
+      "grad_norm": 4.361534360426363,
+      "learning_rate": 2e-05,
+      "loss": 0.6429,
+      "step": 31787
+    },
+    {
+      "epoch": 0.7566409597257926,
+      "grad_norm": 4.756220100931664,
+      "learning_rate": 2e-05,
+      "loss": 1.2016,
+      "step": 31788
+    },
+    {
+      "epoch": 0.7566647624488242,
+      "grad_norm": 1.494277865184168,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 31789
+    },
+    {
+      "epoch": 0.7566885651718557,
+      "grad_norm": 1.8904055955287784,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 31790
+    },
+    {
+      "epoch": 0.7567123678948872,
+      "grad_norm": 9.296516593123492,
+      "learning_rate": 2e-05,
+      "loss": 0.1293,
+      "step": 31791
+    },
+    {
+      "epoch": 0.7567361706179186,
+      "grad_norm": 5.164344194006146,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 31792
+    },
+    {
+      "epoch": 0.7567599733409502,
+      "grad_norm": 4.385253208878285,
+      "learning_rate": 2e-05,
+      "loss": 1.0522,
+      "step": 31793
+    },
+    {
+      "epoch": 0.7567837760639817,
+      "grad_norm": 4.104491266162938,
+      "learning_rate": 2e-05,
+      "loss": 0.6766,
+      "step": 31794
+    },
+    {
+      "epoch": 0.7568075787870132,
+      "grad_norm": 4.6678268233494515,
+      "learning_rate": 2e-05,
+      "loss": 0.241,
+      "step": 31795
+    },
+    {
+      "epoch": 0.7568313815100447,
+      "grad_norm": 5.352654683520203,
+      "learning_rate": 2e-05,
+      "loss": 0.4118,
+      "step": 31796
+    },
+    {
+      "epoch": 0.7568551842330763,
+      "grad_norm": 2.688014872321738,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 31797
+    },
+    {
+      "epoch": 0.7568789869561078,
+      "grad_norm": 3.317102648656768,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 31798
+    },
+    {
+      "epoch": 0.7569027896791393,
+      "grad_norm": 2.3504012908424143,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 31799
+    },
+    {
+      "epoch": 0.7569265924021709,
+      "grad_norm": 2.437145031009142,
+      "learning_rate": 2e-05,
+      "loss": 0.0804,
+      "step": 31800
+    },
+    {
+      "epoch": 0.7569503951252023,
+      "grad_norm": 5.465156350364977,
+      "learning_rate": 2e-05,
+      "loss": 0.1487,
+      "step": 31801
+    },
+    {
+      "epoch": 0.7569741978482338,
+      "grad_norm": 0.5235852129882788,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 31802
+    },
+    {
+      "epoch": 0.7569980005712653,
+      "grad_norm": 6.346223057116327,
+      "learning_rate": 2e-05,
+      "loss": 0.325,
+      "step": 31803
+    },
+    {
+      "epoch": 0.7570218032942969,
+      "grad_norm": 42.68961107577794,
+      "learning_rate": 2e-05,
+      "loss": 0.8916,
+      "step": 31804
+    },
+    {
+      "epoch": 0.7570456060173284,
+      "grad_norm": 0.6544714451021622,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 31805
+    },
+    {
+      "epoch": 0.7570694087403599,
+      "grad_norm": 11.530522903213727,
+      "learning_rate": 2e-05,
+      "loss": 0.7058,
+      "step": 31806
+    },
+    {
+      "epoch": 0.7570932114633914,
+      "grad_norm": 1.8261539299236342,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 31807
+    },
+    {
+      "epoch": 0.757117014186423,
+      "grad_norm": 4.394422224211212,
+      "learning_rate": 2e-05,
+      "loss": 0.447,
+      "step": 31808
+    },
+    {
+      "epoch": 0.7571408169094545,
+      "grad_norm": 0.15372571080197323,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 31809
+    },
+    {
+      "epoch": 0.7571646196324859,
+      "grad_norm": 3.788380817421682,
+      "learning_rate": 2e-05,
+      "loss": 0.5254,
+      "step": 31810
+    },
+    {
+      "epoch": 0.7571884223555174,
+      "grad_norm": 8.335012779310212,
+      "learning_rate": 2e-05,
+      "loss": 0.5633,
+      "step": 31811
+    },
+    {
+      "epoch": 0.757212225078549,
+      "grad_norm": 0.1942991841368004,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 31812
+    },
+    {
+      "epoch": 0.7572360278015805,
+      "grad_norm": 3.177264900499746,
+      "learning_rate": 2e-05,
+      "loss": 0.188,
+      "step": 31813
+    },
+    {
+      "epoch": 0.757259830524612,
+      "grad_norm": 0.7454009703817381,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 31814
+    },
+    {
+      "epoch": 0.7572836332476436,
+      "grad_norm": 3.743071164180541,
+      "learning_rate": 2e-05,
+      "loss": 0.7557,
+      "step": 31815
+    },
+    {
+      "epoch": 0.7573074359706751,
+      "grad_norm": 4.077013974661714,
+      "learning_rate": 2e-05,
+      "loss": 0.4783,
+      "step": 31816
+    },
+    {
+      "epoch": 0.7573312386937066,
+      "grad_norm": 2.824003658528009,
+      "learning_rate": 2e-05,
+      "loss": 0.3609,
+      "step": 31817
+    },
+    {
+      "epoch": 0.757355041416738,
+      "grad_norm": 4.780452527869212,
+      "learning_rate": 2e-05,
+      "loss": 1.4243,
+      "step": 31818
+    },
+    {
+      "epoch": 0.7573788441397696,
+      "grad_norm": 7.343791297559485,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 31819
+    },
+    {
+      "epoch": 0.7574026468628011,
+      "grad_norm": 4.5809725316287135,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 31820
+    },
+    {
+      "epoch": 0.7574264495858326,
+      "grad_norm": 3.702276545459941,
+      "learning_rate": 2e-05,
+      "loss": 0.4779,
+      "step": 31821
+    },
+    {
+      "epoch": 0.7574502523088641,
+      "grad_norm": 0.31588282959262814,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 31822
+    },
+    {
+      "epoch": 0.7574740550318957,
+      "grad_norm": 1.5605368315043235,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 31823
+    },
+    {
+      "epoch": 0.7574978577549272,
+      "grad_norm": 5.740063368059887,
+      "learning_rate": 2e-05,
+      "loss": 0.3753,
+      "step": 31824
+    },
+    {
+      "epoch": 0.7575216604779587,
+      "grad_norm": 6.484667900982404,
+      "learning_rate": 2e-05,
+      "loss": 1.2997,
+      "step": 31825
+    },
+    {
+      "epoch": 0.7575454632009901,
+      "grad_norm": 3.4338038638770434,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 31826
+    },
+    {
+      "epoch": 0.7575692659240217,
+      "grad_norm": 13.03558940719847,
+      "learning_rate": 2e-05,
+      "loss": 0.2875,
+      "step": 31827
+    },
+    {
+      "epoch": 0.7575930686470532,
+      "grad_norm": 3.384452348065173,
+      "learning_rate": 2e-05,
+      "loss": 0.5629,
+      "step": 31828
+    },
+    {
+      "epoch": 0.7576168713700847,
+      "grad_norm": 8.226459028026246,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 31829
+    },
+    {
+      "epoch": 0.7576406740931163,
+      "grad_norm": 0.20966385616137895,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 31830
+    },
+    {
+      "epoch": 0.7576644768161478,
+      "grad_norm": 10.53923291685381,
+      "learning_rate": 2e-05,
+      "loss": 1.3895,
+      "step": 31831
+    },
+    {
+      "epoch": 0.7576882795391793,
+      "grad_norm": 2.4110959271263517,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 31832
+    },
+    {
+      "epoch": 0.7577120822622108,
+      "grad_norm": 2.479279535639695,
+      "learning_rate": 2e-05,
+      "loss": 0.364,
+      "step": 31833
+    },
+    {
+      "epoch": 0.7577358849852424,
+      "grad_norm": 5.483048388877179,
+      "learning_rate": 2e-05,
+      "loss": 1.2756,
+      "step": 31834
+    },
+    {
+      "epoch": 0.7577596877082738,
+      "grad_norm": 4.2754521277887925,
+      "learning_rate": 2e-05,
+      "loss": 0.9367,
+      "step": 31835
+    },
+    {
+      "epoch": 0.7577834904313053,
+      "grad_norm": 3.3627833666341007,
+      "learning_rate": 2e-05,
+      "loss": 0.4262,
+      "step": 31836
+    },
+    {
+      "epoch": 0.7578072931543368,
+      "grad_norm": 1.1977407998657543,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 31837
+    },
+    {
+      "epoch": 0.7578310958773684,
+      "grad_norm": 0.11181419536988275,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 31838
+    },
+    {
+      "epoch": 0.7578548986003999,
+      "grad_norm": 5.882793885241924,
+      "learning_rate": 2e-05,
+      "loss": 0.9124,
+      "step": 31839
+    },
+    {
+      "epoch": 0.7578787013234314,
+      "grad_norm": 0.4190455048202754,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 31840
+    },
+    {
+      "epoch": 0.7579025040464629,
+      "grad_norm": 3.4041961248612904,
+      "learning_rate": 2e-05,
+      "loss": 0.2909,
+      "step": 31841
+    },
+    {
+      "epoch": 0.7579263067694945,
+      "grad_norm": 3.1073938404637587,
+      "learning_rate": 2e-05,
+      "loss": 0.4135,
+      "step": 31842
+    },
+    {
+      "epoch": 0.757950109492526,
+      "grad_norm": 2.1575570738162573,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 31843
+    },
+    {
+      "epoch": 0.7579739122155574,
+      "grad_norm": 2.267215801517742,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 31844
+    },
+    {
+      "epoch": 0.757997714938589,
+      "grad_norm": 3.625983967463435,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 31845
+    },
+    {
+      "epoch": 0.7580215176616205,
+      "grad_norm": 0.25843389752116847,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 31846
+    },
+    {
+      "epoch": 0.758045320384652,
+      "grad_norm": 2.8923639030055144,
+      "learning_rate": 2e-05,
+      "loss": 0.4142,
+      "step": 31847
+    },
+    {
+      "epoch": 0.7580691231076835,
+      "grad_norm": 3.5873829538665354,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 31848
+    },
+    {
+      "epoch": 0.7580929258307151,
+      "grad_norm": 8.276528171530995,
+      "learning_rate": 2e-05,
+      "loss": 1.5093,
+      "step": 31849
+    },
+    {
+      "epoch": 0.7581167285537466,
+      "grad_norm": 0.07005217011344052,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 31850
+    },
+    {
+      "epoch": 0.758140531276778,
+      "grad_norm": 4.722121423052421,
+      "learning_rate": 2e-05,
+      "loss": 0.7759,
+      "step": 31851
+    },
+    {
+      "epoch": 0.7581643339998095,
+      "grad_norm": 13.34533964089944,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 31852
+    },
+    {
+      "epoch": 0.7581881367228411,
+      "grad_norm": 0.33061627467244264,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 31853
+    },
+    {
+      "epoch": 0.7582119394458726,
+      "grad_norm": 2.794783593982328,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 31854
+    },
+    {
+      "epoch": 0.7582357421689041,
+      "grad_norm": 7.178002499593986,
+      "learning_rate": 2e-05,
+      "loss": 0.7913,
+      "step": 31855
+    },
+    {
+      "epoch": 0.7582595448919356,
+      "grad_norm": 3.52559790028239,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 31856
+    },
+    {
+      "epoch": 0.7582833476149672,
+      "grad_norm": 3.478205781932715,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 31857
+    },
+    {
+      "epoch": 0.7583071503379987,
+      "grad_norm": 2.6698099351069673,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 31858
+    },
+    {
+      "epoch": 0.7583309530610302,
+      "grad_norm": 4.666763092059307,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 31859
+    },
+    {
+      "epoch": 0.7583547557840618,
+      "grad_norm": 4.4301859490070745,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 31860
+    },
+    {
+      "epoch": 0.7583785585070932,
+      "grad_norm": 4.2944993745361595,
+      "learning_rate": 2e-05,
+      "loss": 0.7412,
+      "step": 31861
+    },
+    {
+      "epoch": 0.7584023612301247,
+      "grad_norm": 3.805641822637098,
+      "learning_rate": 2e-05,
+      "loss": 0.7216,
+      "step": 31862
+    },
+    {
+      "epoch": 0.7584261639531562,
+      "grad_norm": 2.3479239138259187,
+      "learning_rate": 2e-05,
+      "loss": 0.1745,
+      "step": 31863
+    },
+    {
+      "epoch": 0.7584499666761878,
+      "grad_norm": 3.2843334897750687,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 31864
+    },
+    {
+      "epoch": 0.7584737693992193,
+      "grad_norm": 0.0640923222085485,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 31865
+    },
+    {
+      "epoch": 0.7584975721222508,
+      "grad_norm": 0.7058826188970022,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 31866
+    },
+    {
+      "epoch": 0.7585213748452823,
+      "grad_norm": 8.165888862025295,
+      "learning_rate": 2e-05,
+      "loss": 1.1326,
+      "step": 31867
+    },
+    {
+      "epoch": 0.7585451775683139,
+      "grad_norm": 0.14098948385536655,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 31868
+    },
+    {
+      "epoch": 0.7585689802913453,
+      "grad_norm": 3.9805151806275156,
+      "learning_rate": 2e-05,
+      "loss": 0.6664,
+      "step": 31869
+    },
+    {
+      "epoch": 0.7585927830143768,
+      "grad_norm": 3.034765908390706,
+      "learning_rate": 2e-05,
+      "loss": 0.526,
+      "step": 31870
+    },
+    {
+      "epoch": 0.7586165857374083,
+      "grad_norm": 3.2909699100543883,
+      "learning_rate": 2e-05,
+      "loss": 0.6363,
+      "step": 31871
+    },
+    {
+      "epoch": 0.7586403884604399,
+      "grad_norm": 3.849517358263246,
+      "learning_rate": 2e-05,
+      "loss": 0.5685,
+      "step": 31872
+    },
+    {
+      "epoch": 0.7586641911834714,
+      "grad_norm": 2.3136335066130416,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 31873
+    },
+    {
+      "epoch": 0.7586879939065029,
+      "grad_norm": 4.949806523197979,
+      "learning_rate": 2e-05,
+      "loss": 0.6938,
+      "step": 31874
+    },
+    {
+      "epoch": 0.7587117966295345,
+      "grad_norm": 1.4738205641512918,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 31875
+    },
+    {
+      "epoch": 0.758735599352566,
+      "grad_norm": 0.5564586777295383,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 31876
+    },
+    {
+      "epoch": 0.7587594020755974,
+      "grad_norm": 3.273270184530386,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 31877
+    },
+    {
+      "epoch": 0.7587832047986289,
+      "grad_norm": 3.174693790057507,
+      "learning_rate": 2e-05,
+      "loss": 0.7391,
+      "step": 31878
+    },
+    {
+      "epoch": 0.7588070075216605,
+      "grad_norm": 0.05728931169967498,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 31879
+    },
+    {
+      "epoch": 0.758830810244692,
+      "grad_norm": 4.709873429260365,
+      "learning_rate": 2e-05,
+      "loss": 1.2559,
+      "step": 31880
+    },
+    {
+      "epoch": 0.7588546129677235,
+      "grad_norm": 5.276756671040855,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 31881
+    },
+    {
+      "epoch": 0.758878415690755,
+      "grad_norm": 0.16296554702808172,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 31882
+    },
+    {
+      "epoch": 0.7589022184137866,
+      "grad_norm": 10.437477754510407,
+      "learning_rate": 2e-05,
+      "loss": 1.0386,
+      "step": 31883
+    },
+    {
+      "epoch": 0.7589260211368181,
+      "grad_norm": 2.2551432824342985,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 31884
+    },
+    {
+      "epoch": 0.7589498238598495,
+      "grad_norm": 5.3545171685934605,
+      "learning_rate": 2e-05,
+      "loss": 0.818,
+      "step": 31885
+    },
+    {
+      "epoch": 0.758973626582881,
+      "grad_norm": 3.878184923552461,
+      "learning_rate": 2e-05,
+      "loss": 0.5229,
+      "step": 31886
+    },
+    {
+      "epoch": 0.7589974293059126,
+      "grad_norm": 0.7464560500694993,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 31887
+    },
+    {
+      "epoch": 0.7590212320289441,
+      "grad_norm": 0.7233521768729421,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 31888
+    },
+    {
+      "epoch": 0.7590450347519756,
+      "grad_norm": 3.2561612530781434,
+      "learning_rate": 2e-05,
+      "loss": 0.2228,
+      "step": 31889
+    },
+    {
+      "epoch": 0.7590688374750071,
+      "grad_norm": 4.252395796788401,
+      "learning_rate": 2e-05,
+      "loss": 0.4563,
+      "step": 31890
+    },
+    {
+      "epoch": 0.7590926401980387,
+      "grad_norm": 7.744945586198412,
+      "learning_rate": 2e-05,
+      "loss": 0.4938,
+      "step": 31891
+    },
+    {
+      "epoch": 0.7591164429210702,
+      "grad_norm": 10.395936826347361,
+      "learning_rate": 2e-05,
+      "loss": 0.3975,
+      "step": 31892
+    },
+    {
+      "epoch": 0.7591402456441017,
+      "grad_norm": 3.298750843851211,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 31893
+    },
+    {
+      "epoch": 0.7591640483671332,
+      "grad_norm": 3.6732857347250696,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 31894
+    },
+    {
+      "epoch": 0.7591878510901647,
+      "grad_norm": 8.48306919265176,
+      "learning_rate": 2e-05,
+      "loss": 0.9534,
+      "step": 31895
+    },
+    {
+      "epoch": 0.7592116538131962,
+      "grad_norm": 6.506303848796495,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 31896
+    },
+    {
+      "epoch": 0.7592354565362277,
+      "grad_norm": 0.20232066580015837,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 31897
+    },
+    {
+      "epoch": 0.7592592592592593,
+      "grad_norm": 2.6789546951798866,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 31898
+    },
+    {
+      "epoch": 0.7592830619822908,
+      "grad_norm": 1.82601481410006,
+      "learning_rate": 2e-05,
+      "loss": 0.1607,
+      "step": 31899
+    },
+    {
+      "epoch": 0.7593068647053223,
+      "grad_norm": 0.7763898918286449,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 31900
+    },
+    {
+      "epoch": 0.7593306674283538,
+      "grad_norm": 0.5156247358833057,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 31901
+    },
+    {
+      "epoch": 0.7593544701513854,
+      "grad_norm": 7.228810584434033,
+      "learning_rate": 2e-05,
+      "loss": 0.1563,
+      "step": 31902
+    },
+    {
+      "epoch": 0.7593782728744168,
+      "grad_norm": 0.023303251229585913,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 31903
+    },
+    {
+      "epoch": 0.7594020755974483,
+      "grad_norm": 1.1899254639766252,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 31904
+    },
+    {
+      "epoch": 0.7594258783204798,
+      "grad_norm": 4.298146051287533,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 31905
+    },
+    {
+      "epoch": 0.7594496810435114,
+      "grad_norm": 0.47435417687333925,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 31906
+    },
+    {
+      "epoch": 0.7594734837665429,
+      "grad_norm": 5.376445701418589,
+      "learning_rate": 2e-05,
+      "loss": 0.1764,
+      "step": 31907
+    },
+    {
+      "epoch": 0.7594972864895744,
+      "grad_norm": 0.7644316555201753,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 31908
+    },
+    {
+      "epoch": 0.759521089212606,
+      "grad_norm": 0.20732739339183406,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 31909
+    },
+    {
+      "epoch": 0.7595448919356375,
+      "grad_norm": 1.986495481425212,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 31910
+    },
+    {
+      "epoch": 0.7595686946586689,
+      "grad_norm": 0.9987384835644983,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 31911
+    },
+    {
+      "epoch": 0.7595924973817004,
+      "grad_norm": 5.609286391567584,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 31912
+    },
+    {
+      "epoch": 0.759616300104732,
+      "grad_norm": 2.880090538629079,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 31913
+    },
+    {
+      "epoch": 0.7596401028277635,
+      "grad_norm": 2.8494381329791207,
+      "learning_rate": 2e-05,
+      "loss": 0.6402,
+      "step": 31914
+    },
+    {
+      "epoch": 0.759663905550795,
+      "grad_norm": 0.9861311474826043,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 31915
+    },
+    {
+      "epoch": 0.7596877082738265,
+      "grad_norm": 14.257451144014173,
+      "learning_rate": 2e-05,
+      "loss": 0.5001,
+      "step": 31916
+    },
+    {
+      "epoch": 0.7597115109968581,
+      "grad_norm": 1.901449471394601,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 31917
+    },
+    {
+      "epoch": 0.7597353137198896,
+      "grad_norm": 0.06132390549623526,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 31918
+    },
+    {
+      "epoch": 0.759759116442921,
+      "grad_norm": 1.7378922014466016,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 31919
+    },
+    {
+      "epoch": 0.7597829191659525,
+      "grad_norm": 0.38466182257291415,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 31920
+    },
+    {
+      "epoch": 0.7598067218889841,
+      "grad_norm": 11.556293664546955,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 31921
+    },
+    {
+      "epoch": 0.7598305246120156,
+      "grad_norm": 3.771460467764327,
+      "learning_rate": 2e-05,
+      "loss": 0.4025,
+      "step": 31922
+    },
+    {
+      "epoch": 0.7598543273350471,
+      "grad_norm": 8.772433809438903,
+      "learning_rate": 2e-05,
+      "loss": 0.545,
+      "step": 31923
+    },
+    {
+      "epoch": 0.7598781300580787,
+      "grad_norm": 2.3504359623377775,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 31924
+    },
+    {
+      "epoch": 0.7599019327811102,
+      "grad_norm": 4.545774769974749,
+      "learning_rate": 2e-05,
+      "loss": 0.6471,
+      "step": 31925
+    },
+    {
+      "epoch": 0.7599257355041417,
+      "grad_norm": 2.233007573490337,
+      "learning_rate": 2e-05,
+      "loss": 0.184,
+      "step": 31926
+    },
+    {
+      "epoch": 0.7599495382271731,
+      "grad_norm": 4.660268214428159,
+      "learning_rate": 2e-05,
+      "loss": 0.587,
+      "step": 31927
+    },
+    {
+      "epoch": 0.7599733409502047,
+      "grad_norm": 4.307614026913646,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 31928
+    },
+    {
+      "epoch": 0.7599971436732362,
+      "grad_norm": 0.36981701703232517,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 31929
+    },
+    {
+      "epoch": 0.7600209463962677,
+      "grad_norm": 4.164088509265139,
+      "learning_rate": 2e-05,
+      "loss": 0.7916,
+      "step": 31930
+    },
+    {
+      "epoch": 0.7600447491192992,
+      "grad_norm": 0.712161203110384,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 31931
+    },
+    {
+      "epoch": 0.7600685518423308,
+      "grad_norm": 6.56573844931378,
+      "learning_rate": 2e-05,
+      "loss": 0.4896,
+      "step": 31932
+    },
+    {
+      "epoch": 0.7600923545653623,
+      "grad_norm": 3.8302550786544156,
+      "learning_rate": 2e-05,
+      "loss": 0.7875,
+      "step": 31933
+    },
+    {
+      "epoch": 0.7601161572883938,
+      "grad_norm": 0.5883140893027841,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 31934
+    },
+    {
+      "epoch": 0.7601399600114253,
+      "grad_norm": 0.2451619223980118,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 31935
+    },
+    {
+      "epoch": 0.7601637627344568,
+      "grad_norm": 0.02446130308369035,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 31936
+    },
+    {
+      "epoch": 0.7601875654574883,
+      "grad_norm": 5.631742874357699,
+      "learning_rate": 2e-05,
+      "loss": 0.3263,
+      "step": 31937
+    },
+    {
+      "epoch": 0.7602113681805198,
+      "grad_norm": 10.452894031822586,
+      "learning_rate": 2e-05,
+      "loss": 1.207,
+      "step": 31938
+    },
+    {
+      "epoch": 0.7602351709035514,
+      "grad_norm": 0.045893005274561235,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31939
+    },
+    {
+      "epoch": 0.7602589736265829,
+      "grad_norm": 0.28148134718841294,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 31940
+    },
+    {
+      "epoch": 0.7602827763496144,
+      "grad_norm": 0.02125221653084236,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 31941
+    },
+    {
+      "epoch": 0.7603065790726459,
+      "grad_norm": 3.5993592018635847,
+      "learning_rate": 2e-05,
+      "loss": 0.864,
+      "step": 31942
+    },
+    {
+      "epoch": 0.7603303817956775,
+      "grad_norm": 4.02476950222978,
+      "learning_rate": 2e-05,
+      "loss": 0.7239,
+      "step": 31943
+    },
+    {
+      "epoch": 0.760354184518709,
+      "grad_norm": 4.978723412669403,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 31944
+    },
+    {
+      "epoch": 0.7603779872417404,
+      "grad_norm": 26.172971082648164,
+      "learning_rate": 2e-05,
+      "loss": 0.1998,
+      "step": 31945
+    },
+    {
+      "epoch": 0.7604017899647719,
+      "grad_norm": 10.140032480112614,
+      "learning_rate": 2e-05,
+      "loss": 0.2522,
+      "step": 31946
+    },
+    {
+      "epoch": 0.7604255926878035,
+      "grad_norm": 2.4365452521631217,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 31947
+    },
+    {
+      "epoch": 0.760449395410835,
+      "grad_norm": 5.903004929248259,
+      "learning_rate": 2e-05,
+      "loss": 1.0688,
+      "step": 31948
+    },
+    {
+      "epoch": 0.7604731981338665,
+      "grad_norm": 3.5406720678880905,
+      "learning_rate": 2e-05,
+      "loss": 0.4771,
+      "step": 31949
+    },
+    {
+      "epoch": 0.760497000856898,
+      "grad_norm": 3.5855904546989397,
+      "learning_rate": 2e-05,
+      "loss": 0.7234,
+      "step": 31950
+    },
+    {
+      "epoch": 0.7605208035799296,
+      "grad_norm": 3.4618781983543387,
+      "learning_rate": 2e-05,
+      "loss": 0.5815,
+      "step": 31951
+    },
+    {
+      "epoch": 0.7605446063029611,
+      "grad_norm": 7.889844250545823,
+      "learning_rate": 2e-05,
+      "loss": 0.1447,
+      "step": 31952
+    },
+    {
+      "epoch": 0.7605684090259925,
+      "grad_norm": 2.1444218116536082,
+      "learning_rate": 2e-05,
+      "loss": 0.1677,
+      "step": 31953
+    },
+    {
+      "epoch": 0.7605922117490241,
+      "grad_norm": 1.290206802711656,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 31954
+    },
+    {
+      "epoch": 0.7606160144720556,
+      "grad_norm": 4.556769651072694,
+      "learning_rate": 2e-05,
+      "loss": 0.401,
+      "step": 31955
+    },
+    {
+      "epoch": 0.7606398171950871,
+      "grad_norm": 7.355551969521583,
+      "learning_rate": 2e-05,
+      "loss": 0.5949,
+      "step": 31956
+    },
+    {
+      "epoch": 0.7606636199181186,
+      "grad_norm": 12.772764203900827,
+      "learning_rate": 2e-05,
+      "loss": 0.298,
+      "step": 31957
+    },
+    {
+      "epoch": 0.7606874226411502,
+      "grad_norm": 0.10623776252814485,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 31958
+    },
+    {
+      "epoch": 0.7607112253641817,
+      "grad_norm": 2.8638183377644966,
+      "learning_rate": 2e-05,
+      "loss": 0.6448,
+      "step": 31959
+    },
+    {
+      "epoch": 0.7607350280872132,
+      "grad_norm": 8.358683978578854,
+      "learning_rate": 2e-05,
+      "loss": 0.347,
+      "step": 31960
+    },
+    {
+      "epoch": 0.7607588308102446,
+      "grad_norm": 2.714247814705683,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 31961
+    },
+    {
+      "epoch": 0.7607826335332762,
+      "grad_norm": 0.13706092513508464,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 31962
+    },
+    {
+      "epoch": 0.7608064362563077,
+      "grad_norm": 0.05485283966827422,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 31963
+    },
+    {
+      "epoch": 0.7608302389793392,
+      "grad_norm": 4.45724215053273,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 31964
+    },
+    {
+      "epoch": 0.7608540417023707,
+      "grad_norm": 0.5834591872847149,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 31965
+    },
+    {
+      "epoch": 0.7608778444254023,
+      "grad_norm": 0.3292443621696835,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 31966
+    },
+    {
+      "epoch": 0.7609016471484338,
+      "grad_norm": 3.6204417693147164,
+      "learning_rate": 2e-05,
+      "loss": 0.3833,
+      "step": 31967
+    },
+    {
+      "epoch": 0.7609254498714653,
+      "grad_norm": 9.72524634616999,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 31968
+    },
+    {
+      "epoch": 0.7609492525944969,
+      "grad_norm": 2.573688447264861,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 31969
+    },
+    {
+      "epoch": 0.7609730553175283,
+      "grad_norm": 4.597898148576239,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 31970
+    },
+    {
+      "epoch": 0.7609968580405598,
+      "grad_norm": 2.940831302537026,
+      "learning_rate": 2e-05,
+      "loss": 0.418,
+      "step": 31971
+    },
+    {
+      "epoch": 0.7610206607635913,
+      "grad_norm": 3.9296487754590923,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 31972
+    },
+    {
+      "epoch": 0.7610444634866229,
+      "grad_norm": 25.741887709971554,
+      "learning_rate": 2e-05,
+      "loss": 0.5968,
+      "step": 31973
+    },
+    {
+      "epoch": 0.7610682662096544,
+      "grad_norm": 29.247246336109946,
+      "learning_rate": 2e-05,
+      "loss": 1.0948,
+      "step": 31974
+    },
+    {
+      "epoch": 0.7610920689326859,
+      "grad_norm": 2.484273237406547,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 31975
+    },
+    {
+      "epoch": 0.7611158716557174,
+      "grad_norm": 2.649912628445632,
+      "learning_rate": 2e-05,
+      "loss": 0.1022,
+      "step": 31976
+    },
+    {
+      "epoch": 0.761139674378749,
+      "grad_norm": 0.22749176742328092,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 31977
+    },
+    {
+      "epoch": 0.7611634771017805,
+      "grad_norm": 0.5055493146155922,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 31978
+    },
+    {
+      "epoch": 0.7611872798248119,
+      "grad_norm": 2.048536371294018,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 31979
+    },
+    {
+      "epoch": 0.7612110825478434,
+      "grad_norm": 3.8275921844590552,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 31980
+    },
+    {
+      "epoch": 0.761234885270875,
+      "grad_norm": 0.8842191393802049,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 31981
+    },
+    {
+      "epoch": 0.7612586879939065,
+      "grad_norm": 5.022036022645318,
+      "learning_rate": 2e-05,
+      "loss": 0.7043,
+      "step": 31982
+    },
+    {
+      "epoch": 0.761282490716938,
+      "grad_norm": 9.258785927168935,
+      "learning_rate": 2e-05,
+      "loss": 0.8299,
+      "step": 31983
+    },
+    {
+      "epoch": 0.7613062934399696,
+      "grad_norm": 2.3300862595548844,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 31984
+    },
+    {
+      "epoch": 0.7613300961630011,
+      "grad_norm": 2.778946917281054,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 31985
+    },
+    {
+      "epoch": 0.7613538988860326,
+      "grad_norm": 3.877931881861652,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 31986
+    },
+    {
+      "epoch": 0.761377701609064,
+      "grad_norm": 3.685120001427455,
+      "learning_rate": 2e-05,
+      "loss": 0.4898,
+      "step": 31987
+    },
+    {
+      "epoch": 0.7614015043320956,
+      "grad_norm": 0.11063921681841182,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 31988
+    },
+    {
+      "epoch": 0.7614253070551271,
+      "grad_norm": 5.863861909029826,
+      "learning_rate": 2e-05,
+      "loss": 0.522,
+      "step": 31989
+    },
+    {
+      "epoch": 0.7614491097781586,
+      "grad_norm": 4.775134872161623,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 31990
+    },
+    {
+      "epoch": 0.7614729125011901,
+      "grad_norm": 14.579619388830857,
+      "learning_rate": 2e-05,
+      "loss": 1.3033,
+      "step": 31991
+    },
+    {
+      "epoch": 0.7614967152242217,
+      "grad_norm": 3.225315494043818,
+      "learning_rate": 2e-05,
+      "loss": 0.472,
+      "step": 31992
+    },
+    {
+      "epoch": 0.7615205179472532,
+      "grad_norm": 4.346297580662069,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 31993
+    },
+    {
+      "epoch": 0.7615443206702847,
+      "grad_norm": 0.05712181907545003,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 31994
+    },
+    {
+      "epoch": 0.7615681233933161,
+      "grad_norm": 35.24955714707577,
+      "learning_rate": 2e-05,
+      "loss": 0.7477,
+      "step": 31995
+    },
+    {
+      "epoch": 0.7615919261163477,
+      "grad_norm": 1.3591227060310942,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 31996
+    },
+    {
+      "epoch": 0.7616157288393792,
+      "grad_norm": 5.170871121825953,
+      "learning_rate": 2e-05,
+      "loss": 1.2779,
+      "step": 31997
+    },
+    {
+      "epoch": 0.7616395315624107,
+      "grad_norm": 2.4539480041149,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 31998
+    },
+    {
+      "epoch": 0.7616633342854423,
+      "grad_norm": 8.227546646154845,
+      "learning_rate": 2e-05,
+      "loss": 0.3933,
+      "step": 31999
+    },
+    {
+      "epoch": 0.7616871370084738,
+      "grad_norm": 2.7390027905442302,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 32000
+    },
+    {
+      "epoch": 0.7617109397315053,
+      "grad_norm": 0.04568306867529743,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 32001
+    },
+    {
+      "epoch": 0.7617347424545368,
+      "grad_norm": 2.2865915748195085,
+      "learning_rate": 2e-05,
+      "loss": 0.2701,
+      "step": 32002
+    },
+    {
+      "epoch": 0.7617585451775684,
+      "grad_norm": 11.186608894982843,
+      "learning_rate": 2e-05,
+      "loss": 0.6737,
+      "step": 32003
+    },
+    {
+      "epoch": 0.7617823479005998,
+      "grad_norm": 6.952506749295442,
+      "learning_rate": 2e-05,
+      "loss": 0.7273,
+      "step": 32004
+    },
+    {
+      "epoch": 0.7618061506236313,
+      "grad_norm": 3.0689706112129973,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 32005
+    },
+    {
+      "epoch": 0.7618299533466628,
+      "grad_norm": 1.4601407118788996,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 32006
+    },
+    {
+      "epoch": 0.7618537560696944,
+      "grad_norm": 3.46048624854207,
+      "learning_rate": 2e-05,
+      "loss": 0.6664,
+      "step": 32007
+    },
+    {
+      "epoch": 0.7618775587927259,
+      "grad_norm": 5.033743104847325,
+      "learning_rate": 2e-05,
+      "loss": 0.8678,
+      "step": 32008
+    },
+    {
+      "epoch": 0.7619013615157574,
+      "grad_norm": 4.1704572936624364,
+      "learning_rate": 2e-05,
+      "loss": 0.7719,
+      "step": 32009
+    },
+    {
+      "epoch": 0.7619251642387889,
+      "grad_norm": 3.8097141334618536,
+      "learning_rate": 2e-05,
+      "loss": 0.4443,
+      "step": 32010
+    },
+    {
+      "epoch": 0.7619489669618205,
+      "grad_norm": 5.750986678098839,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 32011
+    },
+    {
+      "epoch": 0.761972769684852,
+      "grad_norm": 11.157010192108073,
+      "learning_rate": 2e-05,
+      "loss": 1.6277,
+      "step": 32012
+    },
+    {
+      "epoch": 0.7619965724078834,
+      "grad_norm": 10.39428517469874,
+      "learning_rate": 2e-05,
+      "loss": 1.8495,
+      "step": 32013
+    },
+    {
+      "epoch": 0.762020375130915,
+      "grad_norm": 5.150281501477096,
+      "learning_rate": 2e-05,
+      "loss": 0.6175,
+      "step": 32014
+    },
+    {
+      "epoch": 0.7620441778539465,
+      "grad_norm": 6.3181791301504155,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 32015
+    },
+    {
+      "epoch": 0.762067980576978,
+      "grad_norm": 5.7455550311788155,
+      "learning_rate": 2e-05,
+      "loss": 0.2227,
+      "step": 32016
+    },
+    {
+      "epoch": 0.7620917833000095,
+      "grad_norm": 3.007364948125962,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 32017
+    },
+    {
+      "epoch": 0.7621155860230411,
+      "grad_norm": 3.215951298964884,
+      "learning_rate": 2e-05,
+      "loss": 0.1115,
+      "step": 32018
+    },
+    {
+      "epoch": 0.7621393887460726,
+      "grad_norm": 7.915265760786476,
+      "learning_rate": 2e-05,
+      "loss": 0.9699,
+      "step": 32019
+    },
+    {
+      "epoch": 0.762163191469104,
+      "grad_norm": 6.899952205188582,
+      "learning_rate": 2e-05,
+      "loss": 1.6227,
+      "step": 32020
+    },
+    {
+      "epoch": 0.7621869941921355,
+      "grad_norm": 8.327656205056211,
+      "learning_rate": 2e-05,
+      "loss": 1.0473,
+      "step": 32021
+    },
+    {
+      "epoch": 0.7622107969151671,
+      "grad_norm": 3.658226860882206,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 32022
+    },
+    {
+      "epoch": 0.7622345996381986,
+      "grad_norm": 3.1750993574898927,
+      "learning_rate": 2e-05,
+      "loss": 0.6413,
+      "step": 32023
+    },
+    {
+      "epoch": 0.7622584023612301,
+      "grad_norm": 2.5485151002083404,
+      "learning_rate": 2e-05,
+      "loss": 0.2848,
+      "step": 32024
+    },
+    {
+      "epoch": 0.7622822050842616,
+      "grad_norm": 1.6412088490248309,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 32025
+    },
+    {
+      "epoch": 0.7623060078072932,
+      "grad_norm": 5.896241558016578,
+      "learning_rate": 2e-05,
+      "loss": 1.7315,
+      "step": 32026
+    },
+    {
+      "epoch": 0.7623298105303247,
+      "grad_norm": 1.1953974587947946,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 32027
+    },
+    {
+      "epoch": 0.7623536132533562,
+      "grad_norm": 2.2149504119255004,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 32028
+    },
+    {
+      "epoch": 0.7623774159763878,
+      "grad_norm": 6.023450282226081,
+      "learning_rate": 2e-05,
+      "loss": 0.8661,
+      "step": 32029
+    },
+    {
+      "epoch": 0.7624012186994192,
+      "grad_norm": 3.9238941301470587,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 32030
+    },
+    {
+      "epoch": 0.7624250214224507,
+      "grad_norm": 5.474312578047732,
+      "learning_rate": 2e-05,
+      "loss": 1.2524,
+      "step": 32031
+    },
+    {
+      "epoch": 0.7624488241454822,
+      "grad_norm": 1.7309571156262868,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 32032
+    },
+    {
+      "epoch": 0.7624726268685138,
+      "grad_norm": 2.4388480336084104,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 32033
+    },
+    {
+      "epoch": 0.7624964295915453,
+      "grad_norm": 4.310739162315562,
+      "learning_rate": 2e-05,
+      "loss": 0.4171,
+      "step": 32034
+    },
+    {
+      "epoch": 0.7625202323145768,
+      "grad_norm": 4.205978527800985,
+      "learning_rate": 2e-05,
+      "loss": 0.1935,
+      "step": 32035
+    },
+    {
+      "epoch": 0.7625440350376083,
+      "grad_norm": 0.4897009719158115,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 32036
+    },
+    {
+      "epoch": 0.7625678377606399,
+      "grad_norm": 1.1885435649273388,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 32037
+    },
+    {
+      "epoch": 0.7625916404836713,
+      "grad_norm": 1.1923681104059247,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 32038
+    },
+    {
+      "epoch": 0.7626154432067028,
+      "grad_norm": 5.1444468709060045,
+      "learning_rate": 2e-05,
+      "loss": 0.8802,
+      "step": 32039
+    },
+    {
+      "epoch": 0.7626392459297343,
+      "grad_norm": 7.272093804326587,
+      "learning_rate": 2e-05,
+      "loss": 0.3877,
+      "step": 32040
+    },
+    {
+      "epoch": 0.7626630486527659,
+      "grad_norm": 4.889622857725343,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 32041
+    },
+    {
+      "epoch": 0.7626868513757974,
+      "grad_norm": 6.016920404141008,
+      "learning_rate": 2e-05,
+      "loss": 0.6563,
+      "step": 32042
+    },
+    {
+      "epoch": 0.7627106540988289,
+      "grad_norm": 2.7986887303797854,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 32043
+    },
+    {
+      "epoch": 0.7627344568218605,
+      "grad_norm": 5.1915861340292455,
+      "learning_rate": 2e-05,
+      "loss": 0.4672,
+      "step": 32044
+    },
+    {
+      "epoch": 0.762758259544892,
+      "grad_norm": 0.36428032561525514,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 32045
+    },
+    {
+      "epoch": 0.7627820622679234,
+      "grad_norm": 4.3672487185677955,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 32046
+    },
+    {
+      "epoch": 0.7628058649909549,
+      "grad_norm": 8.190589828509507,
+      "learning_rate": 2e-05,
+      "loss": 0.7722,
+      "step": 32047
+    },
+    {
+      "epoch": 0.7628296677139865,
+      "grad_norm": 5.595779640728848,
+      "learning_rate": 2e-05,
+      "loss": 0.1409,
+      "step": 32048
+    },
+    {
+      "epoch": 0.762853470437018,
+      "grad_norm": 19.854539536449618,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 32049
+    },
+    {
+      "epoch": 0.7628772731600495,
+      "grad_norm": 0.10974058791395155,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 32050
+    },
+    {
+      "epoch": 0.762901075883081,
+      "grad_norm": 5.7268530057234335,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 32051
+    },
+    {
+      "epoch": 0.7629248786061126,
+      "grad_norm": 1.9872645349262446,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 32052
+    },
+    {
+      "epoch": 0.7629486813291441,
+      "grad_norm": 3.6319666324785307,
+      "learning_rate": 2e-05,
+      "loss": 0.9303,
+      "step": 32053
+    },
+    {
+      "epoch": 0.7629724840521755,
+      "grad_norm": 4.009070770006765,
+      "learning_rate": 2e-05,
+      "loss": 0.7564,
+      "step": 32054
+    },
+    {
+      "epoch": 0.762996286775207,
+      "grad_norm": 1.7176852994000396,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 32055
+    },
+    {
+      "epoch": 0.7630200894982386,
+      "grad_norm": 2.46999499651337,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 32056
+    },
+    {
+      "epoch": 0.7630438922212701,
+      "grad_norm": 0.18688939687732484,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 32057
+    },
+    {
+      "epoch": 0.7630676949443016,
+      "grad_norm": 3.17885042684978,
+      "learning_rate": 2e-05,
+      "loss": 0.3339,
+      "step": 32058
+    },
+    {
+      "epoch": 0.7630914976673332,
+      "grad_norm": 0.6776288211201079,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 32059
+    },
+    {
+      "epoch": 0.7631153003903647,
+      "grad_norm": 2.0065827440030555,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 32060
+    },
+    {
+      "epoch": 0.7631391031133962,
+      "grad_norm": 2.5878195612492556,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 32061
+    },
+    {
+      "epoch": 0.7631629058364277,
+      "grad_norm": 1.0358156165043204,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 32062
+    },
+    {
+      "epoch": 0.7631867085594592,
+      "grad_norm": 0.5395208558405985,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 32063
+    },
+    {
+      "epoch": 0.7632105112824907,
+      "grad_norm": 2.4399286752092673,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 32064
+    },
+    {
+      "epoch": 0.7632343140055222,
+      "grad_norm": 3.1226094498223107,
+      "learning_rate": 2e-05,
+      "loss": 0.5535,
+      "step": 32065
+    },
+    {
+      "epoch": 0.7632581167285537,
+      "grad_norm": 5.249426819434663,
+      "learning_rate": 2e-05,
+      "loss": 0.9599,
+      "step": 32066
+    },
+    {
+      "epoch": 0.7632819194515853,
+      "grad_norm": 4.815438885419127,
+      "learning_rate": 2e-05,
+      "loss": 0.1947,
+      "step": 32067
+    },
+    {
+      "epoch": 0.7633057221746168,
+      "grad_norm": 0.35598632502595934,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 32068
+    },
+    {
+      "epoch": 0.7633295248976483,
+      "grad_norm": 4.212194994142724,
+      "learning_rate": 2e-05,
+      "loss": 0.463,
+      "step": 32069
+    },
+    {
+      "epoch": 0.7633533276206798,
+      "grad_norm": 4.156678872349461,
+      "learning_rate": 2e-05,
+      "loss": 0.4743,
+      "step": 32070
+    },
+    {
+      "epoch": 0.7633771303437114,
+      "grad_norm": 3.290164925755147,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 32071
+    },
+    {
+      "epoch": 0.7634009330667428,
+      "grad_norm": 4.432746304248049,
+      "learning_rate": 2e-05,
+      "loss": 0.5282,
+      "step": 32072
+    },
+    {
+      "epoch": 0.7634247357897743,
+      "grad_norm": 17.423567365002512,
+      "learning_rate": 2e-05,
+      "loss": 0.3456,
+      "step": 32073
+    },
+    {
+      "epoch": 0.7634485385128059,
+      "grad_norm": 2.068427704797419,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 32074
+    },
+    {
+      "epoch": 0.7634723412358374,
+      "grad_norm": 16.98529579430823,
+      "learning_rate": 2e-05,
+      "loss": 0.8221,
+      "step": 32075
+    },
+    {
+      "epoch": 0.7634961439588689,
+      "grad_norm": 4.567914054073657,
+      "learning_rate": 2e-05,
+      "loss": 0.3006,
+      "step": 32076
+    },
+    {
+      "epoch": 0.7635199466819004,
+      "grad_norm": 0.23751851398511592,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 32077
+    },
+    {
+      "epoch": 0.763543749404932,
+      "grad_norm": 5.768474802755724,
+      "learning_rate": 2e-05,
+      "loss": 1.2673,
+      "step": 32078
+    },
+    {
+      "epoch": 0.7635675521279635,
+      "grad_norm": 3.8730091012199477,
+      "learning_rate": 2e-05,
+      "loss": 1.0027,
+      "step": 32079
+    },
+    {
+      "epoch": 0.7635913548509949,
+      "grad_norm": 5.474113941191126,
+      "learning_rate": 2e-05,
+      "loss": 1.1851,
+      "step": 32080
+    },
+    {
+      "epoch": 0.7636151575740264,
+      "grad_norm": 4.815880578255626,
+      "learning_rate": 2e-05,
+      "loss": 0.642,
+      "step": 32081
+    },
+    {
+      "epoch": 0.763638960297058,
+      "grad_norm": 4.0103537531020255,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 32082
+    },
+    {
+      "epoch": 0.7636627630200895,
+      "grad_norm": 0.17432612931358046,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 32083
+    },
+    {
+      "epoch": 0.763686565743121,
+      "grad_norm": 0.25552308864683476,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 32084
+    },
+    {
+      "epoch": 0.7637103684661525,
+      "grad_norm": 0.10545071186185198,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 32085
+    },
+    {
+      "epoch": 0.7637341711891841,
+      "grad_norm": 0.39354445060991056,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 32086
+    },
+    {
+      "epoch": 0.7637579739122156,
+      "grad_norm": 5.668822401008807,
+      "learning_rate": 2e-05,
+      "loss": 0.7599,
+      "step": 32087
+    },
+    {
+      "epoch": 0.763781776635247,
+      "grad_norm": 2.135331804086557,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 32088
+    },
+    {
+      "epoch": 0.7638055793582786,
+      "grad_norm": 1.1521178483007153,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 32089
+    },
+    {
+      "epoch": 0.7638293820813101,
+      "grad_norm": 1.62377749448652,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 32090
+    },
+    {
+      "epoch": 0.7638531848043416,
+      "grad_norm": 2.137255141739904,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 32091
+    },
+    {
+      "epoch": 0.7638769875273731,
+      "grad_norm": 3.2343830691901823,
+      "learning_rate": 2e-05,
+      "loss": 0.672,
+      "step": 32092
+    },
+    {
+      "epoch": 0.7639007902504047,
+      "grad_norm": 43.82256367842193,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 32093
+    },
+    {
+      "epoch": 0.7639245929734362,
+      "grad_norm": 1.673199464241561,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 32094
+    },
+    {
+      "epoch": 0.7639483956964677,
+      "grad_norm": 3.891144555948534,
+      "learning_rate": 2e-05,
+      "loss": 0.5808,
+      "step": 32095
+    },
+    {
+      "epoch": 0.7639721984194991,
+      "grad_norm": 4.255317242045301,
+      "learning_rate": 2e-05,
+      "loss": 0.665,
+      "step": 32096
+    },
+    {
+      "epoch": 0.7639960011425307,
+      "grad_norm": 17.001281562699948,
+      "learning_rate": 2e-05,
+      "loss": 0.3954,
+      "step": 32097
+    },
+    {
+      "epoch": 0.7640198038655622,
+      "grad_norm": 4.441771508166858,
+      "learning_rate": 2e-05,
+      "loss": 0.7593,
+      "step": 32098
+    },
+    {
+      "epoch": 0.7640436065885937,
+      "grad_norm": 11.214169114363292,
+      "learning_rate": 2e-05,
+      "loss": 1.0314,
+      "step": 32099
+    },
+    {
+      "epoch": 0.7640674093116252,
+      "grad_norm": 0.05034637160844154,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32100
+    },
+    {
+      "epoch": 0.7640912120346568,
+      "grad_norm": 9.510058188145566,
+      "learning_rate": 2e-05,
+      "loss": 0.2948,
+      "step": 32101
+    },
+    {
+      "epoch": 0.7641150147576883,
+      "grad_norm": 5.11593808978674,
+      "learning_rate": 2e-05,
+      "loss": 1.1731,
+      "step": 32102
+    },
+    {
+      "epoch": 0.7641388174807198,
+      "grad_norm": 0.38456825203620126,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32103
+    },
+    {
+      "epoch": 0.7641626202037514,
+      "grad_norm": 0.09803510874624921,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 32104
+    },
+    {
+      "epoch": 0.7641864229267828,
+      "grad_norm": 2.7606942460124917,
+      "learning_rate": 2e-05,
+      "loss": 0.5112,
+      "step": 32105
+    },
+    {
+      "epoch": 0.7642102256498143,
+      "grad_norm": 9.792489133319448,
+      "learning_rate": 2e-05,
+      "loss": 0.8683,
+      "step": 32106
+    },
+    {
+      "epoch": 0.7642340283728458,
+      "grad_norm": 2.838425011316627,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 32107
+    },
+    {
+      "epoch": 0.7642578310958774,
+      "grad_norm": 5.5688913188618905,
+      "learning_rate": 2e-05,
+      "loss": 0.9551,
+      "step": 32108
+    },
+    {
+      "epoch": 0.7642816338189089,
+      "grad_norm": 4.834606019372233,
+      "learning_rate": 2e-05,
+      "loss": 0.6132,
+      "step": 32109
+    },
+    {
+      "epoch": 0.7643054365419404,
+      "grad_norm": 4.572314856258656,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 32110
+    },
+    {
+      "epoch": 0.7643292392649719,
+      "grad_norm": 1.5406259732020742,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 32111
+    },
+    {
+      "epoch": 0.7643530419880035,
+      "grad_norm": 3.1575922250116584,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 32112
+    },
+    {
+      "epoch": 0.764376844711035,
+      "grad_norm": 12.432540727442927,
+      "learning_rate": 2e-05,
+      "loss": 1.4028,
+      "step": 32113
+    },
+    {
+      "epoch": 0.7644006474340664,
+      "grad_norm": 6.5374676592946495,
+      "learning_rate": 2e-05,
+      "loss": 1.2616,
+      "step": 32114
+    },
+    {
+      "epoch": 0.7644244501570979,
+      "grad_norm": 0.2513260286290539,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 32115
+    },
+    {
+      "epoch": 0.7644482528801295,
+      "grad_norm": 0.5966813379969504,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 32116
+    },
+    {
+      "epoch": 0.764472055603161,
+      "grad_norm": 2.231518262336939,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 32117
+    },
+    {
+      "epoch": 0.7644958583261925,
+      "grad_norm": 20.774850751782303,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 32118
+    },
+    {
+      "epoch": 0.7645196610492241,
+      "grad_norm": 3.0991722868746643,
+      "learning_rate": 2e-05,
+      "loss": 0.4523,
+      "step": 32119
+    },
+    {
+      "epoch": 0.7645434637722556,
+      "grad_norm": 10.331508854526689,
+      "learning_rate": 2e-05,
+      "loss": 0.3931,
+      "step": 32120
+    },
+    {
+      "epoch": 0.7645672664952871,
+      "grad_norm": 7.156804704672428,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 32121
+    },
+    {
+      "epoch": 0.7645910692183185,
+      "grad_norm": 15.189460414071982,
+      "learning_rate": 2e-05,
+      "loss": 0.3055,
+      "step": 32122
+    },
+    {
+      "epoch": 0.7646148719413501,
+      "grad_norm": 5.365290657330245,
+      "learning_rate": 2e-05,
+      "loss": 0.5673,
+      "step": 32123
+    },
+    {
+      "epoch": 0.7646386746643816,
+      "grad_norm": 5.47684713728874,
+      "learning_rate": 2e-05,
+      "loss": 0.6407,
+      "step": 32124
+    },
+    {
+      "epoch": 0.7646624773874131,
+      "grad_norm": 2.3521468160139607,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 32125
+    },
+    {
+      "epoch": 0.7646862801104446,
+      "grad_norm": 0.16500272452549944,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 32126
+    },
+    {
+      "epoch": 0.7647100828334762,
+      "grad_norm": 6.169373651625367,
+      "learning_rate": 2e-05,
+      "loss": 0.1025,
+      "step": 32127
+    },
+    {
+      "epoch": 0.7647338855565077,
+      "grad_norm": 3.6057582451617596,
+      "learning_rate": 2e-05,
+      "loss": 0.7376,
+      "step": 32128
+    },
+    {
+      "epoch": 0.7647576882795392,
+      "grad_norm": 2.710136679428716,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 32129
+    },
+    {
+      "epoch": 0.7647814910025706,
+      "grad_norm": 1.4727753528644125,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 32130
+    },
+    {
+      "epoch": 0.7648052937256022,
+      "grad_norm": 2.179256842927403,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 32131
+    },
+    {
+      "epoch": 0.7648290964486337,
+      "grad_norm": 6.329534309988912,
+      "learning_rate": 2e-05,
+      "loss": 1.0367,
+      "step": 32132
+    },
+    {
+      "epoch": 0.7648528991716652,
+      "grad_norm": 8.053562600627384,
+      "learning_rate": 2e-05,
+      "loss": 0.1878,
+      "step": 32133
+    },
+    {
+      "epoch": 0.7648767018946968,
+      "grad_norm": 1.1927795915492605,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 32134
+    },
+    {
+      "epoch": 0.7649005046177283,
+      "grad_norm": 3.1167868445621285,
+      "learning_rate": 2e-05,
+      "loss": 0.5816,
+      "step": 32135
+    },
+    {
+      "epoch": 0.7649243073407598,
+      "grad_norm": 2.177574624269268,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 32136
+    },
+    {
+      "epoch": 0.7649481100637913,
+      "grad_norm": 0.31235468591063775,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32137
+    },
+    {
+      "epoch": 0.7649719127868229,
+      "grad_norm": 6.6188007529036375,
+      "learning_rate": 2e-05,
+      "loss": 0.9448,
+      "step": 32138
+    },
+    {
+      "epoch": 0.7649957155098543,
+      "grad_norm": 11.076893375599301,
+      "learning_rate": 2e-05,
+      "loss": 0.7199,
+      "step": 32139
+    },
+    {
+      "epoch": 0.7650195182328858,
+      "grad_norm": 4.568084543509408,
+      "learning_rate": 2e-05,
+      "loss": 0.5074,
+      "step": 32140
+    },
+    {
+      "epoch": 0.7650433209559173,
+      "grad_norm": 11.423533618649309,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 32141
+    },
+    {
+      "epoch": 0.7650671236789489,
+      "grad_norm": 2.621694068207098,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 32142
+    },
+    {
+      "epoch": 0.7650909264019804,
+      "grad_norm": 1.4765491530019892,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 32143
+    },
+    {
+      "epoch": 0.7651147291250119,
+      "grad_norm": 4.819150248089371,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 32144
+    },
+    {
+      "epoch": 0.7651385318480434,
+      "grad_norm": 1.8007914638540856,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 32145
+    },
+    {
+      "epoch": 0.765162334571075,
+      "grad_norm": 0.04091940557628622,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32146
+    },
+    {
+      "epoch": 0.7651861372941064,
+      "grad_norm": 1.9226071052779827,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 32147
+    },
+    {
+      "epoch": 0.7652099400171379,
+      "grad_norm": 3.2643658112156553,
+      "learning_rate": 2e-05,
+      "loss": 0.4591,
+      "step": 32148
+    },
+    {
+      "epoch": 0.7652337427401694,
+      "grad_norm": 3.2177143064835585,
+      "learning_rate": 2e-05,
+      "loss": 0.1087,
+      "step": 32149
+    },
+    {
+      "epoch": 0.765257545463201,
+      "grad_norm": 0.13980115354966802,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 32150
+    },
+    {
+      "epoch": 0.7652813481862325,
+      "grad_norm": 0.16943960194634988,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 32151
+    },
+    {
+      "epoch": 0.765305150909264,
+      "grad_norm": 6.192946502576193,
+      "learning_rate": 2e-05,
+      "loss": 1.3071,
+      "step": 32152
+    },
+    {
+      "epoch": 0.7653289536322956,
+      "grad_norm": 0.3044239868553738,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 32153
+    },
+    {
+      "epoch": 0.7653527563553271,
+      "grad_norm": 0.2073013483857077,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 32154
+    },
+    {
+      "epoch": 0.7653765590783586,
+      "grad_norm": 4.212093431826789,
+      "learning_rate": 2e-05,
+      "loss": 0.81,
+      "step": 32155
+    },
+    {
+      "epoch": 0.76540036180139,
+      "grad_norm": 2.059072505504404,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 32156
+    },
+    {
+      "epoch": 0.7654241645244216,
+      "grad_norm": 2.7832604123954603,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 32157
+    },
+    {
+      "epoch": 0.7654479672474531,
+      "grad_norm": 30.100877444246937,
+      "learning_rate": 2e-05,
+      "loss": 1.3835,
+      "step": 32158
+    },
+    {
+      "epoch": 0.7654717699704846,
+      "grad_norm": 5.556934746038136,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 32159
+    },
+    {
+      "epoch": 0.7654955726935161,
+      "grad_norm": 0.6289878859610016,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 32160
+    },
+    {
+      "epoch": 0.7655193754165477,
+      "grad_norm": 10.000971162182138,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 32161
+    },
+    {
+      "epoch": 0.7655431781395792,
+      "grad_norm": 10.478678541019697,
+      "learning_rate": 2e-05,
+      "loss": 0.6068,
+      "step": 32162
+    },
+    {
+      "epoch": 0.7655669808626107,
+      "grad_norm": 3.4239007554660597,
+      "learning_rate": 2e-05,
+      "loss": 0.7538,
+      "step": 32163
+    },
+    {
+      "epoch": 0.7655907835856421,
+      "grad_norm": 2.299327628251504,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 32164
+    },
+    {
+      "epoch": 0.7656145863086737,
+      "grad_norm": 4.171269054940101,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 32165
+    },
+    {
+      "epoch": 0.7656383890317052,
+      "grad_norm": 1.1856358363753927,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 32166
+    },
+    {
+      "epoch": 0.7656621917547367,
+      "grad_norm": 8.131344112755283,
+      "learning_rate": 2e-05,
+      "loss": 0.6613,
+      "step": 32167
+    },
+    {
+      "epoch": 0.7656859944777683,
+      "grad_norm": 1.848998886757046,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 32168
+    },
+    {
+      "epoch": 0.7657097972007998,
+      "grad_norm": 4.038280510612146,
+      "learning_rate": 2e-05,
+      "loss": 0.5973,
+      "step": 32169
+    },
+    {
+      "epoch": 0.7657335999238313,
+      "grad_norm": 4.010671294388133,
+      "learning_rate": 2e-05,
+      "loss": 0.6006,
+      "step": 32170
+    },
+    {
+      "epoch": 0.7657574026468628,
+      "grad_norm": 6.426356523172986,
+      "learning_rate": 2e-05,
+      "loss": 1.3574,
+      "step": 32171
+    },
+    {
+      "epoch": 0.7657812053698944,
+      "grad_norm": 3.433968154438046,
+      "learning_rate": 2e-05,
+      "loss": 0.3982,
+      "step": 32172
+    },
+    {
+      "epoch": 0.7658050080929258,
+      "grad_norm": 2.4831208179415203,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 32173
+    },
+    {
+      "epoch": 0.7658288108159573,
+      "grad_norm": 2.439200562448923,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 32174
+    },
+    {
+      "epoch": 0.7658526135389888,
+      "grad_norm": 1.8065366045998796,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 32175
+    },
+    {
+      "epoch": 0.7658764162620204,
+      "grad_norm": 8.213076416958902,
+      "learning_rate": 2e-05,
+      "loss": 0.2714,
+      "step": 32176
+    },
+    {
+      "epoch": 0.7659002189850519,
+      "grad_norm": 7.344839627386878,
+      "learning_rate": 2e-05,
+      "loss": 0.3509,
+      "step": 32177
+    },
+    {
+      "epoch": 0.7659240217080834,
+      "grad_norm": 3.0632272044064277,
+      "learning_rate": 2e-05,
+      "loss": 0.537,
+      "step": 32178
+    },
+    {
+      "epoch": 0.7659478244311149,
+      "grad_norm": 25.464424461387395,
+      "learning_rate": 2e-05,
+      "loss": 0.7456,
+      "step": 32179
+    },
+    {
+      "epoch": 0.7659716271541465,
+      "grad_norm": 0.021511356149428795,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 32180
+    },
+    {
+      "epoch": 0.765995429877178,
+      "grad_norm": 7.170174820303764,
+      "learning_rate": 2e-05,
+      "loss": 0.3025,
+      "step": 32181
+    },
+    {
+      "epoch": 0.7660192326002094,
+      "grad_norm": 3.5639006734133507,
+      "learning_rate": 2e-05,
+      "loss": 0.7444,
+      "step": 32182
+    },
+    {
+      "epoch": 0.766043035323241,
+      "grad_norm": 3.9950026873803277,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 32183
+    },
+    {
+      "epoch": 0.7660668380462725,
+      "grad_norm": 7.7084050700345115,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 32184
+    },
+    {
+      "epoch": 0.766090640769304,
+      "grad_norm": 7.187068129738389,
+      "learning_rate": 2e-05,
+      "loss": 0.353,
+      "step": 32185
+    },
+    {
+      "epoch": 0.7661144434923355,
+      "grad_norm": 4.083314674974587,
+      "learning_rate": 2e-05,
+      "loss": 0.6891,
+      "step": 32186
+    },
+    {
+      "epoch": 0.7661382462153671,
+      "grad_norm": 4.329193944282137,
+      "learning_rate": 2e-05,
+      "loss": 0.1055,
+      "step": 32187
+    },
+    {
+      "epoch": 0.7661620489383986,
+      "grad_norm": 6.997764682927162,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 32188
+    },
+    {
+      "epoch": 0.76618585166143,
+      "grad_norm": 2.963487519231673,
+      "learning_rate": 2e-05,
+      "loss": 0.1528,
+      "step": 32189
+    },
+    {
+      "epoch": 0.7662096543844615,
+      "grad_norm": 4.393091755954049,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 32190
+    },
+    {
+      "epoch": 0.7662334571074931,
+      "grad_norm": 2.757636337594272,
+      "learning_rate": 2e-05,
+      "loss": 0.397,
+      "step": 32191
+    },
+    {
+      "epoch": 0.7662572598305246,
+      "grad_norm": 2.5257839036407144,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 32192
+    },
+    {
+      "epoch": 0.7662810625535561,
+      "grad_norm": 0.6417300830808982,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 32193
+    },
+    {
+      "epoch": 0.7663048652765876,
+      "grad_norm": 4.592550169469902,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 32194
+    },
+    {
+      "epoch": 0.7663286679996192,
+      "grad_norm": 6.619983202848407,
+      "learning_rate": 2e-05,
+      "loss": 1.0607,
+      "step": 32195
+    },
+    {
+      "epoch": 0.7663524707226507,
+      "grad_norm": 2.9230164587939176,
+      "learning_rate": 2e-05,
+      "loss": 0.4764,
+      "step": 32196
+    },
+    {
+      "epoch": 0.7663762734456822,
+      "grad_norm": 0.10441712711501935,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 32197
+    },
+    {
+      "epoch": 0.7664000761687138,
+      "grad_norm": 0.3675259049114894,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 32198
+    },
+    {
+      "epoch": 0.7664238788917452,
+      "grad_norm": 8.767601702569266,
+      "learning_rate": 2e-05,
+      "loss": 0.8733,
+      "step": 32199
+    },
+    {
+      "epoch": 0.7664476816147767,
+      "grad_norm": 3.2045193950901836,
+      "learning_rate": 2e-05,
+      "loss": 0.4664,
+      "step": 32200
+    },
+    {
+      "epoch": 0.7664714843378082,
+      "grad_norm": 3.0889811860248426,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 32201
+    },
+    {
+      "epoch": 0.7664952870608398,
+      "grad_norm": 3.123548405493683,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 32202
+    },
+    {
+      "epoch": 0.7665190897838713,
+      "grad_norm": 5.689204337434844,
+      "learning_rate": 2e-05,
+      "loss": 0.8002,
+      "step": 32203
+    },
+    {
+      "epoch": 0.7665428925069028,
+      "grad_norm": 0.09173239821580759,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 32204
+    },
+    {
+      "epoch": 0.7665666952299343,
+      "grad_norm": 16.5334058726034,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 32205
+    },
+    {
+      "epoch": 0.7665904979529659,
+      "grad_norm": 0.2530230007626812,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 32206
+    },
+    {
+      "epoch": 0.7666143006759973,
+      "grad_norm": 8.068803030174758,
+      "learning_rate": 2e-05,
+      "loss": 0.9506,
+      "step": 32207
+    },
+    {
+      "epoch": 0.7666381033990288,
+      "grad_norm": 2.4782288878832275,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 32208
+    },
+    {
+      "epoch": 0.7666619061220603,
+      "grad_norm": 4.460634475653862,
+      "learning_rate": 2e-05,
+      "loss": 1.0977,
+      "step": 32209
+    },
+    {
+      "epoch": 0.7666857088450919,
+      "grad_norm": 3.2127052444984967,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 32210
+    },
+    {
+      "epoch": 0.7667095115681234,
+      "grad_norm": 2.3064031850323574,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 32211
+    },
+    {
+      "epoch": 0.7667333142911549,
+      "grad_norm": 8.371755216800251,
+      "learning_rate": 2e-05,
+      "loss": 1.1144,
+      "step": 32212
+    },
+    {
+      "epoch": 0.7667571170141865,
+      "grad_norm": 2.476955943025305,
+      "learning_rate": 2e-05,
+      "loss": 0.4313,
+      "step": 32213
+    },
+    {
+      "epoch": 0.766780919737218,
+      "grad_norm": 5.876707197992587,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 32214
+    },
+    {
+      "epoch": 0.7668047224602494,
+      "grad_norm": 3.267707373880471,
+      "learning_rate": 2e-05,
+      "loss": 0.5173,
+      "step": 32215
+    },
+    {
+      "epoch": 0.7668285251832809,
+      "grad_norm": 1.2964749830122568,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 32216
+    },
+    {
+      "epoch": 0.7668523279063125,
+      "grad_norm": 5.10822024521949,
+      "learning_rate": 2e-05,
+      "loss": 0.4291,
+      "step": 32217
+    },
+    {
+      "epoch": 0.766876130629344,
+      "grad_norm": 3.7378335751233753,
+      "learning_rate": 2e-05,
+      "loss": 0.7387,
+      "step": 32218
+    },
+    {
+      "epoch": 0.7668999333523755,
+      "grad_norm": 0.42267928188157,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32219
+    },
+    {
+      "epoch": 0.766923736075407,
+      "grad_norm": 0.23869685989143993,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 32220
+    },
+    {
+      "epoch": 0.7669475387984386,
+      "grad_norm": 4.321432815173572,
+      "learning_rate": 2e-05,
+      "loss": 0.5046,
+      "step": 32221
+    },
+    {
+      "epoch": 0.7669713415214701,
+      "grad_norm": 0.23694204080409795,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 32222
+    },
+    {
+      "epoch": 0.7669951442445015,
+      "grad_norm": 1.4958456899741008,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 32223
+    },
+    {
+      "epoch": 0.767018946967533,
+      "grad_norm": 3.596098525372723,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 32224
+    },
+    {
+      "epoch": 0.7670427496905646,
+      "grad_norm": 3.098214763494987,
+      "learning_rate": 2e-05,
+      "loss": 0.5814,
+      "step": 32225
+    },
+    {
+      "epoch": 0.7670665524135961,
+      "grad_norm": 0.11434657797320638,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32226
+    },
+    {
+      "epoch": 0.7670903551366276,
+      "grad_norm": 15.228694530236531,
+      "learning_rate": 2e-05,
+      "loss": 0.3526,
+      "step": 32227
+    },
+    {
+      "epoch": 0.7671141578596592,
+      "grad_norm": 5.068386358102946,
+      "learning_rate": 2e-05,
+      "loss": 1.5602,
+      "step": 32228
+    },
+    {
+      "epoch": 0.7671379605826907,
+      "grad_norm": 4.088802568889777,
+      "learning_rate": 2e-05,
+      "loss": 0.4514,
+      "step": 32229
+    },
+    {
+      "epoch": 0.7671617633057222,
+      "grad_norm": 4.648362642420936,
+      "learning_rate": 2e-05,
+      "loss": 0.6612,
+      "step": 32230
+    },
+    {
+      "epoch": 0.7671855660287537,
+      "grad_norm": 6.251761956113334,
+      "learning_rate": 2e-05,
+      "loss": 0.3024,
+      "step": 32231
+    },
+    {
+      "epoch": 0.7672093687517852,
+      "grad_norm": 3.1486701323829767,
+      "learning_rate": 2e-05,
+      "loss": 0.6139,
+      "step": 32232
+    },
+    {
+      "epoch": 0.7672331714748167,
+      "grad_norm": 3.717134308061375,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 32233
+    },
+    {
+      "epoch": 0.7672569741978482,
+      "grad_norm": 6.9742138235039555,
+      "learning_rate": 2e-05,
+      "loss": 0.9426,
+      "step": 32234
+    },
+    {
+      "epoch": 0.7672807769208797,
+      "grad_norm": 2.0994000916500246,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 32235
+    },
+    {
+      "epoch": 0.7673045796439113,
+      "grad_norm": 0.5238012374930192,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 32236
+    },
+    {
+      "epoch": 0.7673283823669428,
+      "grad_norm": 4.497966003939575,
+      "learning_rate": 2e-05,
+      "loss": 1.1036,
+      "step": 32237
+    },
+    {
+      "epoch": 0.7673521850899743,
+      "grad_norm": 8.460487396898642,
+      "learning_rate": 2e-05,
+      "loss": 0.8329,
+      "step": 32238
+    },
+    {
+      "epoch": 0.7673759878130058,
+      "grad_norm": 0.10585043079384471,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 32239
+    },
+    {
+      "epoch": 0.7673997905360374,
+      "grad_norm": 2.3612785936763676,
+      "learning_rate": 2e-05,
+      "loss": 0.2865,
+      "step": 32240
+    },
+    {
+      "epoch": 0.7674235932590688,
+      "grad_norm": 4.712324236459102,
+      "learning_rate": 2e-05,
+      "loss": 1.0748,
+      "step": 32241
+    },
+    {
+      "epoch": 0.7674473959821003,
+      "grad_norm": 0.6656344602829467,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 32242
+    },
+    {
+      "epoch": 0.7674711987051319,
+      "grad_norm": 4.666487863968204,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 32243
+    },
+    {
+      "epoch": 0.7674950014281634,
+      "grad_norm": 0.599744508878668,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 32244
+    },
+    {
+      "epoch": 0.7675188041511949,
+      "grad_norm": 21.97015292640956,
+      "learning_rate": 2e-05,
+      "loss": 0.4048,
+      "step": 32245
+    },
+    {
+      "epoch": 0.7675426068742264,
+      "grad_norm": 4.721245686876799,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 32246
+    },
+    {
+      "epoch": 0.767566409597258,
+      "grad_norm": 5.274455130259463,
+      "learning_rate": 2e-05,
+      "loss": 0.4099,
+      "step": 32247
+    },
+    {
+      "epoch": 0.7675902123202895,
+      "grad_norm": 0.305404811312582,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32248
+    },
+    {
+      "epoch": 0.7676140150433209,
+      "grad_norm": 5.799039672333013,
+      "learning_rate": 2e-05,
+      "loss": 0.768,
+      "step": 32249
+    },
+    {
+      "epoch": 0.7676378177663524,
+      "grad_norm": 24.288813579811865,
+      "learning_rate": 2e-05,
+      "loss": 0.1935,
+      "step": 32250
+    },
+    {
+      "epoch": 0.767661620489384,
+      "grad_norm": 5.5137706570444225,
+      "learning_rate": 2e-05,
+      "loss": 0.731,
+      "step": 32251
+    },
+    {
+      "epoch": 0.7676854232124155,
+      "grad_norm": 0.6789560388905523,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 32252
+    },
+    {
+      "epoch": 0.767709225935447,
+      "grad_norm": 2.510796910959371,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 32253
+    },
+    {
+      "epoch": 0.7677330286584785,
+      "grad_norm": 0.7567515685766378,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 32254
+    },
+    {
+      "epoch": 0.7677568313815101,
+      "grad_norm": 6.55041335857154,
+      "learning_rate": 2e-05,
+      "loss": 0.6972,
+      "step": 32255
+    },
+    {
+      "epoch": 0.7677806341045416,
+      "grad_norm": 6.557203006772258,
+      "learning_rate": 2e-05,
+      "loss": 1.1284,
+      "step": 32256
+    },
+    {
+      "epoch": 0.767804436827573,
+      "grad_norm": 0.0374606811988893,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 32257
+    },
+    {
+      "epoch": 0.7678282395506046,
+      "grad_norm": 5.171787768559746,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 32258
+    },
+    {
+      "epoch": 0.7678520422736361,
+      "grad_norm": 9.406771234590254,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 32259
+    },
+    {
+      "epoch": 0.7678758449966676,
+      "grad_norm": 1.1359682426090876,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 32260
+    },
+    {
+      "epoch": 0.7678996477196991,
+      "grad_norm": 11.473016947443101,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 32261
+    },
+    {
+      "epoch": 0.7679234504427307,
+      "grad_norm": 4.480659842098592,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 32262
+    },
+    {
+      "epoch": 0.7679472531657622,
+      "grad_norm": 6.164824415835898,
+      "learning_rate": 2e-05,
+      "loss": 0.1902,
+      "step": 32263
+    },
+    {
+      "epoch": 0.7679710558887937,
+      "grad_norm": 0.8258521181193731,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 32264
+    },
+    {
+      "epoch": 0.7679948586118251,
+      "grad_norm": 6.245361586049782,
+      "learning_rate": 2e-05,
+      "loss": 0.2093,
+      "step": 32265
+    },
+    {
+      "epoch": 0.7680186613348567,
+      "grad_norm": 3.653736800368147,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 32266
+    },
+    {
+      "epoch": 0.7680424640578882,
+      "grad_norm": 3.706254366312642,
+      "learning_rate": 2e-05,
+      "loss": 0.4551,
+      "step": 32267
+    },
+    {
+      "epoch": 0.7680662667809197,
+      "grad_norm": 1.8750135955413083,
+      "learning_rate": 2e-05,
+      "loss": 0.1701,
+      "step": 32268
+    },
+    {
+      "epoch": 0.7680900695039512,
+      "grad_norm": 4.091464525037798,
+      "learning_rate": 2e-05,
+      "loss": 0.7135,
+      "step": 32269
+    },
+    {
+      "epoch": 0.7681138722269828,
+      "grad_norm": 5.568139357474456,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 32270
+    },
+    {
+      "epoch": 0.7681376749500143,
+      "grad_norm": 12.491886739281602,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 32271
+    },
+    {
+      "epoch": 0.7681614776730458,
+      "grad_norm": 6.266439847942224,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 32272
+    },
+    {
+      "epoch": 0.7681852803960774,
+      "grad_norm": 2.125344740402524,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 32273
+    },
+    {
+      "epoch": 0.7682090831191088,
+      "grad_norm": 7.730584307679707,
+      "learning_rate": 2e-05,
+      "loss": 0.785,
+      "step": 32274
+    },
+    {
+      "epoch": 0.7682328858421403,
+      "grad_norm": 5.5883747870637555,
+      "learning_rate": 2e-05,
+      "loss": 1.0286,
+      "step": 32275
+    },
+    {
+      "epoch": 0.7682566885651718,
+      "grad_norm": 3.4499403951686602,
+      "learning_rate": 2e-05,
+      "loss": 0.4189,
+      "step": 32276
+    },
+    {
+      "epoch": 0.7682804912882034,
+      "grad_norm": 0.7763194523715333,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 32277
+    },
+    {
+      "epoch": 0.7683042940112349,
+      "grad_norm": 2.0330944624885854,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 32278
+    },
+    {
+      "epoch": 0.7683280967342664,
+      "grad_norm": 2.40423488783447,
+      "learning_rate": 2e-05,
+      "loss": 0.1694,
+      "step": 32279
+    },
+    {
+      "epoch": 0.7683518994572979,
+      "grad_norm": 7.963177329309694,
+      "learning_rate": 2e-05,
+      "loss": 0.2867,
+      "step": 32280
+    },
+    {
+      "epoch": 0.7683757021803295,
+      "grad_norm": 4.467874504526503,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 32281
+    },
+    {
+      "epoch": 0.768399504903361,
+      "grad_norm": 4.6371240390784125,
+      "learning_rate": 2e-05,
+      "loss": 0.7788,
+      "step": 32282
+    },
+    {
+      "epoch": 0.7684233076263924,
+      "grad_norm": 4.703803209906693,
+      "learning_rate": 2e-05,
+      "loss": 0.6622,
+      "step": 32283
+    },
+    {
+      "epoch": 0.7684471103494239,
+      "grad_norm": 4.344580243126268,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 32284
+    },
+    {
+      "epoch": 0.7684709130724555,
+      "grad_norm": 14.479915482085577,
+      "learning_rate": 2e-05,
+      "loss": 0.8328,
+      "step": 32285
+    },
+    {
+      "epoch": 0.768494715795487,
+      "grad_norm": 2.338326884145443,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 32286
+    },
+    {
+      "epoch": 0.7685185185185185,
+      "grad_norm": 3.300939064183015,
+      "learning_rate": 2e-05,
+      "loss": 0.4962,
+      "step": 32287
+    },
+    {
+      "epoch": 0.7685423212415501,
+      "grad_norm": 0.10792134717321973,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 32288
+    },
+    {
+      "epoch": 0.7685661239645816,
+      "grad_norm": 0.2664851660396618,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 32289
+    },
+    {
+      "epoch": 0.7685899266876131,
+      "grad_norm": 0.03383993542216247,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 32290
+    },
+    {
+      "epoch": 0.7686137294106445,
+      "grad_norm": 8.360986809876755,
+      "learning_rate": 2e-05,
+      "loss": 1.379,
+      "step": 32291
+    },
+    {
+      "epoch": 0.7686375321336761,
+      "grad_norm": 3.5349890384355307,
+      "learning_rate": 2e-05,
+      "loss": 0.9077,
+      "step": 32292
+    },
+    {
+      "epoch": 0.7686613348567076,
+      "grad_norm": 0.44734427837526713,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 32293
+    },
+    {
+      "epoch": 0.7686851375797391,
+      "grad_norm": 1.1147834349622838,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 32294
+    },
+    {
+      "epoch": 0.7687089403027706,
+      "grad_norm": 0.2761429733612209,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32295
+    },
+    {
+      "epoch": 0.7687327430258022,
+      "grad_norm": 4.8513013683111605,
+      "learning_rate": 2e-05,
+      "loss": 0.5134,
+      "step": 32296
+    },
+    {
+      "epoch": 0.7687565457488337,
+      "grad_norm": 4.68120394558561,
+      "learning_rate": 2e-05,
+      "loss": 0.7477,
+      "step": 32297
+    },
+    {
+      "epoch": 0.7687803484718652,
+      "grad_norm": 11.467578473209143,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 32298
+    },
+    {
+      "epoch": 0.7688041511948966,
+      "grad_norm": 4.948307459830128,
+      "learning_rate": 2e-05,
+      "loss": 0.9666,
+      "step": 32299
+    },
+    {
+      "epoch": 0.7688279539179282,
+      "grad_norm": 15.76576990615527,
+      "learning_rate": 2e-05,
+      "loss": 1.0482,
+      "step": 32300
+    },
+    {
+      "epoch": 0.7688517566409597,
+      "grad_norm": 0.9240796292478594,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 32301
+    },
+    {
+      "epoch": 0.7688755593639912,
+      "grad_norm": 8.123568761282536,
+      "learning_rate": 2e-05,
+      "loss": 1.0491,
+      "step": 32302
+    },
+    {
+      "epoch": 0.7688993620870228,
+      "grad_norm": 3.310332451409001,
+      "learning_rate": 2e-05,
+      "loss": 0.4617,
+      "step": 32303
+    },
+    {
+      "epoch": 0.7689231648100543,
+      "grad_norm": 0.7131895431661962,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 32304
+    },
+    {
+      "epoch": 0.7689469675330858,
+      "grad_norm": 5.407762071125316,
+      "learning_rate": 2e-05,
+      "loss": 0.8484,
+      "step": 32305
+    },
+    {
+      "epoch": 0.7689707702561173,
+      "grad_norm": 3.1118150026631195,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 32306
+    },
+    {
+      "epoch": 0.7689945729791489,
+      "grad_norm": 1.610312951940406,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 32307
+    },
+    {
+      "epoch": 0.7690183757021803,
+      "grad_norm": 4.48428209496986,
+      "learning_rate": 2e-05,
+      "loss": 0.3868,
+      "step": 32308
+    },
+    {
+      "epoch": 0.7690421784252118,
+      "grad_norm": 2.900857537480348,
+      "learning_rate": 2e-05,
+      "loss": 0.339,
+      "step": 32309
+    },
+    {
+      "epoch": 0.7690659811482433,
+      "grad_norm": 0.34270297003366307,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 32310
+    },
+    {
+      "epoch": 0.7690897838712749,
+      "grad_norm": 6.649939660275954,
+      "learning_rate": 2e-05,
+      "loss": 0.6201,
+      "step": 32311
+    },
+    {
+      "epoch": 0.7691135865943064,
+      "grad_norm": 4.047440697541914,
+      "learning_rate": 2e-05,
+      "loss": 0.6011,
+      "step": 32312
+    },
+    {
+      "epoch": 0.7691373893173379,
+      "grad_norm": 9.932396319263555,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 32313
+    },
+    {
+      "epoch": 0.7691611920403694,
+      "grad_norm": 0.012985062227204617,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 32314
+    },
+    {
+      "epoch": 0.769184994763401,
+      "grad_norm": 0.6541588010867843,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 32315
+    },
+    {
+      "epoch": 0.7692087974864324,
+      "grad_norm": 5.42401648651098,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 32316
+    },
+    {
+      "epoch": 0.7692326002094639,
+      "grad_norm": 1.9433476898278055,
+      "learning_rate": 2e-05,
+      "loss": 0.1324,
+      "step": 32317
+    },
+    {
+      "epoch": 0.7692564029324955,
+      "grad_norm": 67.751713330851,
+      "learning_rate": 2e-05,
+      "loss": 0.7721,
+      "step": 32318
+    },
+    {
+      "epoch": 0.769280205655527,
+      "grad_norm": 0.7648906444258101,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 32319
+    },
+    {
+      "epoch": 0.7693040083785585,
+      "grad_norm": 4.567951802867424,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 32320
+    },
+    {
+      "epoch": 0.76932781110159,
+      "grad_norm": 3.989953747523494,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 32321
+    },
+    {
+      "epoch": 0.7693516138246216,
+      "grad_norm": 0.07430815147779797,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 32322
+    },
+    {
+      "epoch": 0.7693754165476531,
+      "grad_norm": 4.9582149469511805,
+      "learning_rate": 2e-05,
+      "loss": 1.1866,
+      "step": 32323
+    },
+    {
+      "epoch": 0.7693992192706846,
+      "grad_norm": 5.014275988619022,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 32324
+    },
+    {
+      "epoch": 0.769423021993716,
+      "grad_norm": 0.09415627205203356,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 32325
+    },
+    {
+      "epoch": 0.7694468247167476,
+      "grad_norm": 3.4197419709571895,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 32326
+    },
+    {
+      "epoch": 0.7694706274397791,
+      "grad_norm": 8.663120205823853,
+      "learning_rate": 2e-05,
+      "loss": 1.1482,
+      "step": 32327
+    },
+    {
+      "epoch": 0.7694944301628106,
+      "grad_norm": 3.933203435465093,
+      "learning_rate": 2e-05,
+      "loss": 0.3134,
+      "step": 32328
+    },
+    {
+      "epoch": 0.7695182328858421,
+      "grad_norm": 0.8183077920119695,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 32329
+    },
+    {
+      "epoch": 0.7695420356088737,
+      "grad_norm": 10.275207543324662,
+      "learning_rate": 2e-05,
+      "loss": 0.8482,
+      "step": 32330
+    },
+    {
+      "epoch": 0.7695658383319052,
+      "grad_norm": 3.1510520096616315,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 32331
+    },
+    {
+      "epoch": 0.7695896410549367,
+      "grad_norm": 3.1746530168354883,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 32332
+    },
+    {
+      "epoch": 0.7696134437779683,
+      "grad_norm": 4.768597193629819,
+      "learning_rate": 2e-05,
+      "loss": 0.1584,
+      "step": 32333
+    },
+    {
+      "epoch": 0.7696372465009997,
+      "grad_norm": 3.5231430448457353,
+      "learning_rate": 2e-05,
+      "loss": 0.6334,
+      "step": 32334
+    },
+    {
+      "epoch": 0.7696610492240312,
+      "grad_norm": 15.629845125721051,
+      "learning_rate": 2e-05,
+      "loss": 0.5103,
+      "step": 32335
+    },
+    {
+      "epoch": 0.7696848519470627,
+      "grad_norm": 7.2284547055052055,
+      "learning_rate": 2e-05,
+      "loss": 1.3087,
+      "step": 32336
+    },
+    {
+      "epoch": 0.7697086546700943,
+      "grad_norm": 3.986929231217219,
+      "learning_rate": 2e-05,
+      "loss": 0.6573,
+      "step": 32337
+    },
+    {
+      "epoch": 0.7697324573931258,
+      "grad_norm": 6.297777763805302,
+      "learning_rate": 2e-05,
+      "loss": 0.6907,
+      "step": 32338
+    },
+    {
+      "epoch": 0.7697562601161573,
+      "grad_norm": 7.45280825992345,
+      "learning_rate": 2e-05,
+      "loss": 0.5786,
+      "step": 32339
+    },
+    {
+      "epoch": 0.7697800628391888,
+      "grad_norm": 1.1586914859124362,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 32340
+    },
+    {
+      "epoch": 0.7698038655622204,
+      "grad_norm": 0.730781827867124,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 32341
+    },
+    {
+      "epoch": 0.7698276682852518,
+      "grad_norm": 0.05067281932104173,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 32342
+    },
+    {
+      "epoch": 0.7698514710082833,
+      "grad_norm": 5.442777442797474,
+      "learning_rate": 2e-05,
+      "loss": 0.8719,
+      "step": 32343
+    },
+    {
+      "epoch": 0.7698752737313148,
+      "grad_norm": 4.19242859594964,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 32344
+    },
+    {
+      "epoch": 0.7698990764543464,
+      "grad_norm": 0.027341258000794493,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 32345
+    },
+    {
+      "epoch": 0.7699228791773779,
+      "grad_norm": 3.0372802865381767,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 32346
+    },
+    {
+      "epoch": 0.7699466819004094,
+      "grad_norm": 6.922492027531141,
+      "learning_rate": 2e-05,
+      "loss": 1.0325,
+      "step": 32347
+    },
+    {
+      "epoch": 0.769970484623441,
+      "grad_norm": 3.95798619674159,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 32348
+    },
+    {
+      "epoch": 0.7699942873464725,
+      "grad_norm": 5.187643135233245,
+      "learning_rate": 2e-05,
+      "loss": 0.3126,
+      "step": 32349
+    },
+    {
+      "epoch": 0.770018090069504,
+      "grad_norm": 1.2118335220634486,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 32350
+    },
+    {
+      "epoch": 0.7700418927925354,
+      "grad_norm": 14.566777217846994,
+      "learning_rate": 2e-05,
+      "loss": 2.1443,
+      "step": 32351
+    },
+    {
+      "epoch": 0.770065695515567,
+      "grad_norm": 8.863137384549646,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 32352
+    },
+    {
+      "epoch": 0.7700894982385985,
+      "grad_norm": 2.6856622226359397,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 32353
+    },
+    {
+      "epoch": 0.77011330096163,
+      "grad_norm": 5.887182107804724,
+      "learning_rate": 2e-05,
+      "loss": 0.273,
+      "step": 32354
+    },
+    {
+      "epoch": 0.7701371036846615,
+      "grad_norm": 3.3103748625066056,
+      "learning_rate": 2e-05,
+      "loss": 0.6873,
+      "step": 32355
+    },
+    {
+      "epoch": 0.7701609064076931,
+      "grad_norm": 2.089199636900254,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 32356
+    },
+    {
+      "epoch": 0.7701847091307246,
+      "grad_norm": 3.5092588710380426,
+      "learning_rate": 2e-05,
+      "loss": 0.3594,
+      "step": 32357
+    },
+    {
+      "epoch": 0.770208511853756,
+      "grad_norm": 2.921026298532744,
+      "learning_rate": 2e-05,
+      "loss": 0.3167,
+      "step": 32358
+    },
+    {
+      "epoch": 0.7702323145767875,
+      "grad_norm": 4.760283867315807,
+      "learning_rate": 2e-05,
+      "loss": 0.7804,
+      "step": 32359
+    },
+    {
+      "epoch": 0.7702561172998191,
+      "grad_norm": 0.7446104203019557,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 32360
+    },
+    {
+      "epoch": 0.7702799200228506,
+      "grad_norm": 5.022888574458213,
+      "learning_rate": 2e-05,
+      "loss": 0.9048,
+      "step": 32361
+    },
+    {
+      "epoch": 0.7703037227458821,
+      "grad_norm": 4.592600367462704,
+      "learning_rate": 2e-05,
+      "loss": 0.1415,
+      "step": 32362
+    },
+    {
+      "epoch": 0.7703275254689137,
+      "grad_norm": 4.956680310537663,
+      "learning_rate": 2e-05,
+      "loss": 0.8973,
+      "step": 32363
+    },
+    {
+      "epoch": 0.7703513281919452,
+      "grad_norm": 14.265679217229044,
+      "learning_rate": 2e-05,
+      "loss": 1.3312,
+      "step": 32364
+    },
+    {
+      "epoch": 0.7703751309149767,
+      "grad_norm": 0.732336672883296,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 32365
+    },
+    {
+      "epoch": 0.7703989336380082,
+      "grad_norm": 4.300626711141252,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 32366
+    },
+    {
+      "epoch": 0.7704227363610398,
+      "grad_norm": 0.12979575858540138,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 32367
+    },
+    {
+      "epoch": 0.7704465390840712,
+      "grad_norm": 2.5310279537626146,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 32368
+    },
+    {
+      "epoch": 0.7704703418071027,
+      "grad_norm": 1.6739547375745527,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 32369
+    },
+    {
+      "epoch": 0.7704941445301342,
+      "grad_norm": 10.906290210132331,
+      "learning_rate": 2e-05,
+      "loss": 1.0816,
+      "step": 32370
+    },
+    {
+      "epoch": 0.7705179472531658,
+      "grad_norm": 3.359378884906556,
+      "learning_rate": 2e-05,
+      "loss": 0.1284,
+      "step": 32371
+    },
+    {
+      "epoch": 0.7705417499761973,
+      "grad_norm": 11.881394139390794,
+      "learning_rate": 2e-05,
+      "loss": 1.4402,
+      "step": 32372
+    },
+    {
+      "epoch": 0.7705655526992288,
+      "grad_norm": 8.484751477891365,
+      "learning_rate": 2e-05,
+      "loss": 1.2949,
+      "step": 32373
+    },
+    {
+      "epoch": 0.7705893554222603,
+      "grad_norm": 5.638298594209022,
+      "learning_rate": 2e-05,
+      "loss": 0.1504,
+      "step": 32374
+    },
+    {
+      "epoch": 0.7706131581452919,
+      "grad_norm": 4.418329079164454,
+      "learning_rate": 2e-05,
+      "loss": 0.717,
+      "step": 32375
+    },
+    {
+      "epoch": 0.7706369608683233,
+      "grad_norm": 1.8972919462369193,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 32376
+    },
+    {
+      "epoch": 0.7706607635913548,
+      "grad_norm": 2.1684013931078874,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 32377
+    },
+    {
+      "epoch": 0.7706845663143864,
+      "grad_norm": 2.167811234844607,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 32378
+    },
+    {
+      "epoch": 0.7707083690374179,
+      "grad_norm": 0.424885043239923,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 32379
+    },
+    {
+      "epoch": 0.7707321717604494,
+      "grad_norm": 0.2520294109485656,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 32380
+    },
+    {
+      "epoch": 0.7707559744834809,
+      "grad_norm": 0.03949169946136123,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 32381
+    },
+    {
+      "epoch": 0.7707797772065125,
+      "grad_norm": 10.858024810743853,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 32382
+    },
+    {
+      "epoch": 0.770803579929544,
+      "grad_norm": 6.4859582137676,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 32383
+    },
+    {
+      "epoch": 0.7708273826525754,
+      "grad_norm": 5.4028209554958195,
+      "learning_rate": 2e-05,
+      "loss": 0.5987,
+      "step": 32384
+    },
+    {
+      "epoch": 0.7708511853756069,
+      "grad_norm": 1.2219064585784032,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 32385
+    },
+    {
+      "epoch": 0.7708749880986385,
+      "grad_norm": 3.8452362957631956,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 32386
+    },
+    {
+      "epoch": 0.77089879082167,
+      "grad_norm": 3.1874274031225776,
+      "learning_rate": 2e-05,
+      "loss": 0.4164,
+      "step": 32387
+    },
+    {
+      "epoch": 0.7709225935447015,
+      "grad_norm": 0.18973566947921552,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32388
+    },
+    {
+      "epoch": 0.770946396267733,
+      "grad_norm": 0.7709701236912098,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 32389
+    },
+    {
+      "epoch": 0.7709701989907646,
+      "grad_norm": 14.576765547901898,
+      "learning_rate": 2e-05,
+      "loss": 0.5254,
+      "step": 32390
+    },
+    {
+      "epoch": 0.7709940017137961,
+      "grad_norm": 6.382396055038717,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 32391
+    },
+    {
+      "epoch": 0.7710178044368275,
+      "grad_norm": 0.19109197870098085,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 32392
+    },
+    {
+      "epoch": 0.771041607159859,
+      "grad_norm": 0.12354324754489066,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 32393
+    },
+    {
+      "epoch": 0.7710654098828906,
+      "grad_norm": 0.11677065066386719,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 32394
+    },
+    {
+      "epoch": 0.7710892126059221,
+      "grad_norm": 4.495113782959823,
+      "learning_rate": 2e-05,
+      "loss": 0.7687,
+      "step": 32395
+    },
+    {
+      "epoch": 0.7711130153289536,
+      "grad_norm": 2.8486373190181404,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 32396
+    },
+    {
+      "epoch": 0.7711368180519852,
+      "grad_norm": 3.589516671150886,
+      "learning_rate": 2e-05,
+      "loss": 0.629,
+      "step": 32397
+    },
+    {
+      "epoch": 0.7711606207750167,
+      "grad_norm": 10.359823568103415,
+      "learning_rate": 2e-05,
+      "loss": 0.9538,
+      "step": 32398
+    },
+    {
+      "epoch": 0.7711844234980482,
+      "grad_norm": 2.310057209183912,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 32399
+    },
+    {
+      "epoch": 0.7712082262210797,
+      "grad_norm": 2.0628827260080262,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 32400
+    },
+    {
+      "epoch": 0.7712320289441112,
+      "grad_norm": 0.6258224164984789,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 32401
+    },
+    {
+      "epoch": 0.7712558316671427,
+      "grad_norm": 5.282762770591899,
+      "learning_rate": 2e-05,
+      "loss": 0.7119,
+      "step": 32402
+    },
+    {
+      "epoch": 0.7712796343901742,
+      "grad_norm": 1.9977541737310591,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 32403
+    },
+    {
+      "epoch": 0.7713034371132057,
+      "grad_norm": 1.9627378210735207,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 32404
+    },
+    {
+      "epoch": 0.7713272398362373,
+      "grad_norm": 3.9841163143943508,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 32405
+    },
+    {
+      "epoch": 0.7713510425592688,
+      "grad_norm": 5.382192275948193,
+      "learning_rate": 2e-05,
+      "loss": 0.7633,
+      "step": 32406
+    },
+    {
+      "epoch": 0.7713748452823003,
+      "grad_norm": 0.3641145219912784,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 32407
+    },
+    {
+      "epoch": 0.7713986480053318,
+      "grad_norm": 0.0030946308787906524,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 32408
+    },
+    {
+      "epoch": 0.7714224507283634,
+      "grad_norm": 0.9763590698136763,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 32409
+    },
+    {
+      "epoch": 0.7714462534513948,
+      "grad_norm": 6.268246435181959,
+      "learning_rate": 2e-05,
+      "loss": 0.2525,
+      "step": 32410
+    },
+    {
+      "epoch": 0.7714700561744263,
+      "grad_norm": 6.117506373840835,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 32411
+    },
+    {
+      "epoch": 0.7714938588974579,
+      "grad_norm": 3.505779649394442,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 32412
+    },
+    {
+      "epoch": 0.7715176616204894,
+      "grad_norm": 0.07847007488400189,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 32413
+    },
+    {
+      "epoch": 0.7715414643435209,
+      "grad_norm": 2.8605327326980516,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 32414
+    },
+    {
+      "epoch": 0.7715652670665524,
+      "grad_norm": 0.8717326144372733,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 32415
+    },
+    {
+      "epoch": 0.771589069789584,
+      "grad_norm": 2.9268764594623016,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 32416
+    },
+    {
+      "epoch": 0.7716128725126155,
+      "grad_norm": 3.074052280442677,
+      "learning_rate": 2e-05,
+      "loss": 0.4359,
+      "step": 32417
+    },
+    {
+      "epoch": 0.7716366752356469,
+      "grad_norm": 0.24750861685064904,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 32418
+    },
+    {
+      "epoch": 0.7716604779586784,
+      "grad_norm": 4.643974504329984,
+      "learning_rate": 2e-05,
+      "loss": 0.5121,
+      "step": 32419
+    },
+    {
+      "epoch": 0.77168428068171,
+      "grad_norm": 10.436524657200533,
+      "learning_rate": 2e-05,
+      "loss": 0.5409,
+      "step": 32420
+    },
+    {
+      "epoch": 0.7717080834047415,
+      "grad_norm": 4.309077162861816,
+      "learning_rate": 2e-05,
+      "loss": 0.446,
+      "step": 32421
+    },
+    {
+      "epoch": 0.771731886127773,
+      "grad_norm": 3.425471816945212,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 32422
+    },
+    {
+      "epoch": 0.7717556888508045,
+      "grad_norm": 4.671321310518612,
+      "learning_rate": 2e-05,
+      "loss": 0.8153,
+      "step": 32423
+    },
+    {
+      "epoch": 0.7717794915738361,
+      "grad_norm": 2.0120338951446737,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 32424
+    },
+    {
+      "epoch": 0.7718032942968676,
+      "grad_norm": 6.34084206008279,
+      "learning_rate": 2e-05,
+      "loss": 0.581,
+      "step": 32425
+    },
+    {
+      "epoch": 0.771827097019899,
+      "grad_norm": 7.637348843332478,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 32426
+    },
+    {
+      "epoch": 0.7718508997429306,
+      "grad_norm": 0.9202923891008504,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 32427
+    },
+    {
+      "epoch": 0.7718747024659621,
+      "grad_norm": 2.830514896577687,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 32428
+    },
+    {
+      "epoch": 0.7718985051889936,
+      "grad_norm": 7.7458681120498065,
+      "learning_rate": 2e-05,
+      "loss": 0.4968,
+      "step": 32429
+    },
+    {
+      "epoch": 0.7719223079120251,
+      "grad_norm": 4.537345546389357,
+      "learning_rate": 2e-05,
+      "loss": 0.4709,
+      "step": 32430
+    },
+    {
+      "epoch": 0.7719461106350567,
+      "grad_norm": 2.626384764101649,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 32431
+    },
+    {
+      "epoch": 0.7719699133580882,
+      "grad_norm": 3.979330713862358,
+      "learning_rate": 2e-05,
+      "loss": 0.5625,
+      "step": 32432
+    },
+    {
+      "epoch": 0.7719937160811197,
+      "grad_norm": 0.2024339150565384,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 32433
+    },
+    {
+      "epoch": 0.7720175188041511,
+      "grad_norm": 4.293923861155892,
+      "learning_rate": 2e-05,
+      "loss": 0.8658,
+      "step": 32434
+    },
+    {
+      "epoch": 0.7720413215271827,
+      "grad_norm": 3.031100310870805,
+      "learning_rate": 2e-05,
+      "loss": 0.3295,
+      "step": 32435
+    },
+    {
+      "epoch": 0.7720651242502142,
+      "grad_norm": 2.0005835359530137,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 32436
+    },
+    {
+      "epoch": 0.7720889269732457,
+      "grad_norm": 5.556103247714397,
+      "learning_rate": 2e-05,
+      "loss": 0.1943,
+      "step": 32437
+    },
+    {
+      "epoch": 0.7721127296962772,
+      "grad_norm": 5.630702458332027,
+      "learning_rate": 2e-05,
+      "loss": 0.1105,
+      "step": 32438
+    },
+    {
+      "epoch": 0.7721365324193088,
+      "grad_norm": 2.167085714305583,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 32439
+    },
+    {
+      "epoch": 0.7721603351423403,
+      "grad_norm": 4.066822107439371,
+      "learning_rate": 2e-05,
+      "loss": 0.7799,
+      "step": 32440
+    },
+    {
+      "epoch": 0.7721841378653718,
+      "grad_norm": 3.7770630004441883,
+      "learning_rate": 2e-05,
+      "loss": 0.6679,
+      "step": 32441
+    },
+    {
+      "epoch": 0.7722079405884034,
+      "grad_norm": 0.8374509992056748,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 32442
+    },
+    {
+      "epoch": 0.7722317433114348,
+      "grad_norm": 2.465308473856582,
+      "learning_rate": 2e-05,
+      "loss": 0.5155,
+      "step": 32443
+    },
+    {
+      "epoch": 0.7722555460344663,
+      "grad_norm": 7.986696282182565,
+      "learning_rate": 2e-05,
+      "loss": 0.7797,
+      "step": 32444
+    },
+    {
+      "epoch": 0.7722793487574978,
+      "grad_norm": 6.2444694433275965,
+      "learning_rate": 2e-05,
+      "loss": 0.5782,
+      "step": 32445
+    },
+    {
+      "epoch": 0.7723031514805294,
+      "grad_norm": 2.7946997835600764,
+      "learning_rate": 2e-05,
+      "loss": 0.3562,
+      "step": 32446
+    },
+    {
+      "epoch": 0.7723269542035609,
+      "grad_norm": 3.375454583505648,
+      "learning_rate": 2e-05,
+      "loss": 0.6362,
+      "step": 32447
+    },
+    {
+      "epoch": 0.7723507569265924,
+      "grad_norm": 0.08324778414076518,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 32448
+    },
+    {
+      "epoch": 0.7723745596496239,
+      "grad_norm": 7.913334455079812,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 32449
+    },
+    {
+      "epoch": 0.7723983623726555,
+      "grad_norm": 4.171109198747538,
+      "learning_rate": 2e-05,
+      "loss": 0.6504,
+      "step": 32450
+    },
+    {
+      "epoch": 0.772422165095687,
+      "grad_norm": 3.183491246100817,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 32451
+    },
+    {
+      "epoch": 0.7724459678187184,
+      "grad_norm": 3.657157812708101,
+      "learning_rate": 2e-05,
+      "loss": 0.5367,
+      "step": 32452
+    },
+    {
+      "epoch": 0.7724697705417499,
+      "grad_norm": 0.31443192982430745,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 32453
+    },
+    {
+      "epoch": 0.7724935732647815,
+      "grad_norm": 2.7170062160668143,
+      "learning_rate": 2e-05,
+      "loss": 0.0412,
+      "step": 32454
+    },
+    {
+      "epoch": 0.772517375987813,
+      "grad_norm": 5.818958629466484,
+      "learning_rate": 2e-05,
+      "loss": 1.4288,
+      "step": 32455
+    },
+    {
+      "epoch": 0.7725411787108445,
+      "grad_norm": 10.066547515022013,
+      "learning_rate": 2e-05,
+      "loss": 0.772,
+      "step": 32456
+    },
+    {
+      "epoch": 0.7725649814338761,
+      "grad_norm": 0.10300930634645637,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 32457
+    },
+    {
+      "epoch": 0.7725887841569076,
+      "grad_norm": 0.0489244134335968,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 32458
+    },
+    {
+      "epoch": 0.7726125868799391,
+      "grad_norm": 5.046110063584334,
+      "learning_rate": 2e-05,
+      "loss": 0.1173,
+      "step": 32459
+    },
+    {
+      "epoch": 0.7726363896029705,
+      "grad_norm": 4.102181418879698,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 32460
+    },
+    {
+      "epoch": 0.7726601923260021,
+      "grad_norm": 3.21653929196497,
+      "learning_rate": 2e-05,
+      "loss": 0.1389,
+      "step": 32461
+    },
+    {
+      "epoch": 0.7726839950490336,
+      "grad_norm": 3.3022854780767745,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 32462
+    },
+    {
+      "epoch": 0.7727077977720651,
+      "grad_norm": 4.278900934957072,
+      "learning_rate": 2e-05,
+      "loss": 0.4416,
+      "step": 32463
+    },
+    {
+      "epoch": 0.7727316004950966,
+      "grad_norm": 2.292503177855711,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 32464
+    },
+    {
+      "epoch": 0.7727554032181282,
+      "grad_norm": 0.0878013351166965,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 32465
+    },
+    {
+      "epoch": 0.7727792059411597,
+      "grad_norm": 7.590153623388046,
+      "learning_rate": 2e-05,
+      "loss": 1.0812,
+      "step": 32466
+    },
+    {
+      "epoch": 0.7728030086641912,
+      "grad_norm": 6.357315749927139,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 32467
+    },
+    {
+      "epoch": 0.7728268113872226,
+      "grad_norm": 3.1529079176358152,
+      "learning_rate": 2e-05,
+      "loss": 0.482,
+      "step": 32468
+    },
+    {
+      "epoch": 0.7728506141102542,
+      "grad_norm": 2.5204666852298625,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 32469
+    },
+    {
+      "epoch": 0.7728744168332857,
+      "grad_norm": 1.303789573082772,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 32470
+    },
+    {
+      "epoch": 0.7728982195563172,
+      "grad_norm": 4.029304441784437,
+      "learning_rate": 2e-05,
+      "loss": 0.2954,
+      "step": 32471
+    },
+    {
+      "epoch": 0.7729220222793488,
+      "grad_norm": 1.7141288010713696,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 32472
+    },
+    {
+      "epoch": 0.7729458250023803,
+      "grad_norm": 6.313804889249473,
+      "learning_rate": 2e-05,
+      "loss": 1.2611,
+      "step": 32473
+    },
+    {
+      "epoch": 0.7729696277254118,
+      "grad_norm": 3.3956642175477736,
+      "learning_rate": 2e-05,
+      "loss": 0.2231,
+      "step": 32474
+    },
+    {
+      "epoch": 0.7729934304484433,
+      "grad_norm": 0.3383269979339596,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 32475
+    },
+    {
+      "epoch": 0.7730172331714749,
+      "grad_norm": 2.369206841465773,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 32476
+    },
+    {
+      "epoch": 0.7730410358945063,
+      "grad_norm": 0.17418324709682007,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 32477
+    },
+    {
+      "epoch": 0.7730648386175378,
+      "grad_norm": 10.55492618010271,
+      "learning_rate": 2e-05,
+      "loss": 0.4192,
+      "step": 32478
+    },
+    {
+      "epoch": 0.7730886413405693,
+      "grad_norm": 3.2633513309904734,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 32479
+    },
+    {
+      "epoch": 0.7731124440636009,
+      "grad_norm": 9.472835659783247,
+      "learning_rate": 2e-05,
+      "loss": 0.8026,
+      "step": 32480
+    },
+    {
+      "epoch": 0.7731362467866324,
+      "grad_norm": 2.7302028947671473,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 32481
+    },
+    {
+      "epoch": 0.7731600495096639,
+      "grad_norm": 2.2523738010932313,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 32482
+    },
+    {
+      "epoch": 0.7731838522326954,
+      "grad_norm": 2.0957067100168647,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 32483
+    },
+    {
+      "epoch": 0.773207654955727,
+      "grad_norm": 0.5926624937521864,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 32484
+    },
+    {
+      "epoch": 0.7732314576787584,
+      "grad_norm": 7.315042405321803,
+      "learning_rate": 2e-05,
+      "loss": 0.1015,
+      "step": 32485
+    },
+    {
+      "epoch": 0.7732552604017899,
+      "grad_norm": 2.5281413316196377,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 32486
+    },
+    {
+      "epoch": 0.7732790631248215,
+      "grad_norm": 11.77401526802999,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 32487
+    },
+    {
+      "epoch": 0.773302865847853,
+      "grad_norm": 1.486503110713231,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 32488
+    },
+    {
+      "epoch": 0.7733266685708845,
+      "grad_norm": 2.5092451445424655,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 32489
+    },
+    {
+      "epoch": 0.773350471293916,
+      "grad_norm": 3.2766916141732843,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 32490
+    },
+    {
+      "epoch": 0.7733742740169476,
+      "grad_norm": 6.606420719285647,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 32491
+    },
+    {
+      "epoch": 0.7733980767399791,
+      "grad_norm": 0.12766596559758808,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 32492
+    },
+    {
+      "epoch": 0.7734218794630106,
+      "grad_norm": 7.223206582186887,
+      "learning_rate": 2e-05,
+      "loss": 0.616,
+      "step": 32493
+    },
+    {
+      "epoch": 0.773445682186042,
+      "grad_norm": 2.4762390860205246,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 32494
+    },
+    {
+      "epoch": 0.7734694849090736,
+      "grad_norm": 1.9035555026542537,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 32495
+    },
+    {
+      "epoch": 0.7734932876321051,
+      "grad_norm": 0.02256434021379927,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 32496
+    },
+    {
+      "epoch": 0.7735170903551366,
+      "grad_norm": 2.1869650684822908,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 32497
+    },
+    {
+      "epoch": 0.7735408930781681,
+      "grad_norm": 4.342544171162483,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 32498
+    },
+    {
+      "epoch": 0.7735646958011997,
+      "grad_norm": 3.12344144656585,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 32499
+    },
+    {
+      "epoch": 0.7735884985242312,
+      "grad_norm": 16.977743519144838,
+      "learning_rate": 2e-05,
+      "loss": 0.5353,
+      "step": 32500
+    },
+    {
+      "epoch": 0.7736123012472627,
+      "grad_norm": 5.716454472939212,
+      "learning_rate": 2e-05,
+      "loss": 1.196,
+      "step": 32501
+    },
+    {
+      "epoch": 0.7736361039702943,
+      "grad_norm": 3.658346297143877,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 32502
+    },
+    {
+      "epoch": 0.7736599066933257,
+      "grad_norm": 7.12932290007578,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 32503
+    },
+    {
+      "epoch": 0.7736837094163572,
+      "grad_norm": 1.1612233683112287,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 32504
+    },
+    {
+      "epoch": 0.7737075121393887,
+      "grad_norm": 3.5123997900273394,
+      "learning_rate": 2e-05,
+      "loss": 0.4765,
+      "step": 32505
+    },
+    {
+      "epoch": 0.7737313148624203,
+      "grad_norm": 7.010279127734021,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 32506
+    },
+    {
+      "epoch": 0.7737551175854518,
+      "grad_norm": 5.014833022133252,
+      "learning_rate": 2e-05,
+      "loss": 0.7178,
+      "step": 32507
+    },
+    {
+      "epoch": 0.7737789203084833,
+      "grad_norm": 4.462928730822434,
+      "learning_rate": 2e-05,
+      "loss": 1.0152,
+      "step": 32508
+    },
+    {
+      "epoch": 0.7738027230315148,
+      "grad_norm": 4.0928723770258015,
+      "learning_rate": 2e-05,
+      "loss": 0.4567,
+      "step": 32509
+    },
+    {
+      "epoch": 0.7738265257545464,
+      "grad_norm": 0.06643450384274487,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 32510
+    },
+    {
+      "epoch": 0.7738503284775778,
+      "grad_norm": 2.0587144084706743,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 32511
+    },
+    {
+      "epoch": 0.7738741312006093,
+      "grad_norm": 1.0664750297407706,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 32512
+    },
+    {
+      "epoch": 0.7738979339236408,
+      "grad_norm": 6.23623466118728,
+      "learning_rate": 2e-05,
+      "loss": 0.2201,
+      "step": 32513
+    },
+    {
+      "epoch": 0.7739217366466724,
+      "grad_norm": 0.09947031318961767,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 32514
+    },
+    {
+      "epoch": 0.7739455393697039,
+      "grad_norm": 0.15130832427403074,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 32515
+    },
+    {
+      "epoch": 0.7739693420927354,
+      "grad_norm": 5.455270112700663,
+      "learning_rate": 2e-05,
+      "loss": 1.1508,
+      "step": 32516
+    },
+    {
+      "epoch": 0.773993144815767,
+      "grad_norm": 0.8978955553729471,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 32517
+    },
+    {
+      "epoch": 0.7740169475387985,
+      "grad_norm": 0.6343498217060322,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 32518
+    },
+    {
+      "epoch": 0.77404075026183,
+      "grad_norm": 3.267607857590889,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 32519
+    },
+    {
+      "epoch": 0.7740645529848614,
+      "grad_norm": 4.382039536750063,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 32520
+    },
+    {
+      "epoch": 0.774088355707893,
+      "grad_norm": 6.435383888359409,
+      "learning_rate": 2e-05,
+      "loss": 1.0467,
+      "step": 32521
+    },
+    {
+      "epoch": 0.7741121584309245,
+      "grad_norm": 11.914772517458937,
+      "learning_rate": 2e-05,
+      "loss": 0.6383,
+      "step": 32522
+    },
+    {
+      "epoch": 0.774135961153956,
+      "grad_norm": 2.201272623512866,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 32523
+    },
+    {
+      "epoch": 0.7741597638769875,
+      "grad_norm": 0.32462846385872607,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 32524
+    },
+    {
+      "epoch": 0.7741835666000191,
+      "grad_norm": 4.247627669142443,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 32525
+    },
+    {
+      "epoch": 0.7742073693230506,
+      "grad_norm": 0.12266053290603945,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 32526
+    },
+    {
+      "epoch": 0.774231172046082,
+      "grad_norm": 3.291042357268524,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 32527
+    },
+    {
+      "epoch": 0.7742549747691135,
+      "grad_norm": 4.742713630871142,
+      "learning_rate": 2e-05,
+      "loss": 0.8435,
+      "step": 32528
+    },
+    {
+      "epoch": 0.7742787774921451,
+      "grad_norm": 0.5650325780757751,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 32529
+    },
+    {
+      "epoch": 0.7743025802151766,
+      "grad_norm": 1.178281506537884,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 32530
+    },
+    {
+      "epoch": 0.7743263829382081,
+      "grad_norm": 5.085387755497903,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 32531
+    },
+    {
+      "epoch": 0.7743501856612397,
+      "grad_norm": 4.746888911860154,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 32532
+    },
+    {
+      "epoch": 0.7743739883842712,
+      "grad_norm": 2.2236720155686545,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 32533
+    },
+    {
+      "epoch": 0.7743977911073027,
+      "grad_norm": 4.259087902079076,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 32534
+    },
+    {
+      "epoch": 0.7744215938303342,
+      "grad_norm": 6.831954282482926,
+      "learning_rate": 2e-05,
+      "loss": 1.3244,
+      "step": 32535
+    },
+    {
+      "epoch": 0.7744453965533658,
+      "grad_norm": 10.383812876715078,
+      "learning_rate": 2e-05,
+      "loss": 1.7164,
+      "step": 32536
+    },
+    {
+      "epoch": 0.7744691992763972,
+      "grad_norm": 0.010410993668005079,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 32537
+    },
+    {
+      "epoch": 0.7744930019994287,
+      "grad_norm": 5.612766209549861,
+      "learning_rate": 2e-05,
+      "loss": 0.8181,
+      "step": 32538
+    },
+    {
+      "epoch": 0.7745168047224602,
+      "grad_norm": 9.222722029356408,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 32539
+    },
+    {
+      "epoch": 0.7745406074454918,
+      "grad_norm": 2.4896550158549675,
+      "learning_rate": 2e-05,
+      "loss": 0.4389,
+      "step": 32540
+    },
+    {
+      "epoch": 0.7745644101685233,
+      "grad_norm": 0.005136472474587799,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 32541
+    },
+    {
+      "epoch": 0.7745882128915548,
+      "grad_norm": 6.899282142950421,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 32542
+    },
+    {
+      "epoch": 0.7746120156145863,
+      "grad_norm": 4.113630606113215,
+      "learning_rate": 2e-05,
+      "loss": 0.2661,
+      "step": 32543
+    },
+    {
+      "epoch": 0.7746358183376179,
+      "grad_norm": 6.367081095858624,
+      "learning_rate": 2e-05,
+      "loss": 0.6202,
+      "step": 32544
+    },
+    {
+      "epoch": 0.7746596210606493,
+      "grad_norm": 0.026798025980157177,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 32545
+    },
+    {
+      "epoch": 0.7746834237836808,
+      "grad_norm": 1.5231707828341217,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 32546
+    },
+    {
+      "epoch": 0.7747072265067124,
+      "grad_norm": 0.12108517902268859,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 32547
+    },
+    {
+      "epoch": 0.7747310292297439,
+      "grad_norm": 0.046325414408474784,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32548
+    },
+    {
+      "epoch": 0.7747548319527754,
+      "grad_norm": 2.1752959131377465,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 32549
+    },
+    {
+      "epoch": 0.7747786346758069,
+      "grad_norm": 5.2277620025560925,
+      "learning_rate": 2e-05,
+      "loss": 0.8501,
+      "step": 32550
+    },
+    {
+      "epoch": 0.7748024373988385,
+      "grad_norm": 8.053956924661192,
+      "learning_rate": 2e-05,
+      "loss": 0.711,
+      "step": 32551
+    },
+    {
+      "epoch": 0.77482624012187,
+      "grad_norm": 4.351732986177917,
+      "learning_rate": 2e-05,
+      "loss": 0.4717,
+      "step": 32552
+    },
+    {
+      "epoch": 0.7748500428449014,
+      "grad_norm": 4.778184671386731,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 32553
+    },
+    {
+      "epoch": 0.7748738455679329,
+      "grad_norm": 3.296931013928536,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 32554
+    },
+    {
+      "epoch": 0.7748976482909645,
+      "grad_norm": 0.5562167579348107,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 32555
+    },
+    {
+      "epoch": 0.774921451013996,
+      "grad_norm": 9.323016291969186,
+      "learning_rate": 2e-05,
+      "loss": 2.0192,
+      "step": 32556
+    },
+    {
+      "epoch": 0.7749452537370275,
+      "grad_norm": 4.330661499769809,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 32557
+    },
+    {
+      "epoch": 0.774969056460059,
+      "grad_norm": 7.165494420381518,
+      "learning_rate": 2e-05,
+      "loss": 1.306,
+      "step": 32558
+    },
+    {
+      "epoch": 0.7749928591830906,
+      "grad_norm": 0.006985090918395095,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 32559
+    },
+    {
+      "epoch": 0.7750166619061221,
+      "grad_norm": 3.387826206880003,
+      "learning_rate": 2e-05,
+      "loss": 0.4124,
+      "step": 32560
+    },
+    {
+      "epoch": 0.7750404646291535,
+      "grad_norm": 0.016818323482639834,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 32561
+    },
+    {
+      "epoch": 0.7750642673521851,
+      "grad_norm": 8.107835655859267,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 32562
+    },
+    {
+      "epoch": 0.7750880700752166,
+      "grad_norm": 7.612943164020797,
+      "learning_rate": 2e-05,
+      "loss": 0.1976,
+      "step": 32563
+    },
+    {
+      "epoch": 0.7751118727982481,
+      "grad_norm": 11.62367249181024,
+      "learning_rate": 2e-05,
+      "loss": 0.8458,
+      "step": 32564
+    },
+    {
+      "epoch": 0.7751356755212796,
+      "grad_norm": 4.189554254807164,
+      "learning_rate": 2e-05,
+      "loss": 0.4153,
+      "step": 32565
+    },
+    {
+      "epoch": 0.7751594782443112,
+      "grad_norm": 7.039686436084518,
+      "learning_rate": 2e-05,
+      "loss": 1.657,
+      "step": 32566
+    },
+    {
+      "epoch": 0.7751832809673427,
+      "grad_norm": 2.7255905023116207,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 32567
+    },
+    {
+      "epoch": 0.7752070836903742,
+      "grad_norm": 4.323881568227837,
+      "learning_rate": 2e-05,
+      "loss": 0.0602,
+      "step": 32568
+    },
+    {
+      "epoch": 0.7752308864134057,
+      "grad_norm": 3.9203604326809454,
+      "learning_rate": 2e-05,
+      "loss": 0.5298,
+      "step": 32569
+    },
+    {
+      "epoch": 0.7752546891364372,
+      "grad_norm": 0.12896342427649393,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 32570
+    },
+    {
+      "epoch": 0.7752784918594687,
+      "grad_norm": 14.317607228069939,
+      "learning_rate": 2e-05,
+      "loss": 0.174,
+      "step": 32571
+    },
+    {
+      "epoch": 0.7753022945825002,
+      "grad_norm": 1.3221917910533019,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 32572
+    },
+    {
+      "epoch": 0.7753260973055317,
+      "grad_norm": 0.2329762324387225,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 32573
+    },
+    {
+      "epoch": 0.7753499000285633,
+      "grad_norm": 0.5449959306384798,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32574
+    },
+    {
+      "epoch": 0.7753737027515948,
+      "grad_norm": 3.506155487183756,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 32575
+    },
+    {
+      "epoch": 0.7753975054746263,
+      "grad_norm": 3.648954426696155,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 32576
+    },
+    {
+      "epoch": 0.7754213081976579,
+      "grad_norm": 5.538980712736081,
+      "learning_rate": 2e-05,
+      "loss": 1.3618,
+      "step": 32577
+    },
+    {
+      "epoch": 0.7754451109206894,
+      "grad_norm": 5.098322316205412,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 32578
+    },
+    {
+      "epoch": 0.7754689136437208,
+      "grad_norm": 5.212007048657015,
+      "learning_rate": 2e-05,
+      "loss": 1.1891,
+      "step": 32579
+    },
+    {
+      "epoch": 0.7754927163667523,
+      "grad_norm": 3.318065870027121,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 32580
+    },
+    {
+      "epoch": 0.7755165190897839,
+      "grad_norm": 2.2229471430008663,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 32581
+    },
+    {
+      "epoch": 0.7755403218128154,
+      "grad_norm": 0.8368988120088946,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 32582
+    },
+    {
+      "epoch": 0.7755641245358469,
+      "grad_norm": 5.546483861000185,
+      "learning_rate": 2e-05,
+      "loss": 1.3492,
+      "step": 32583
+    },
+    {
+      "epoch": 0.7755879272588784,
+      "grad_norm": 6.9990478889378895,
+      "learning_rate": 2e-05,
+      "loss": 1.0787,
+      "step": 32584
+    },
+    {
+      "epoch": 0.77561172998191,
+      "grad_norm": 41.62568337580017,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 32585
+    },
+    {
+      "epoch": 0.7756355327049415,
+      "grad_norm": 0.8618117066714189,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 32586
+    },
+    {
+      "epoch": 0.7756593354279729,
+      "grad_norm": 0.6805041982314195,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 32587
+    },
+    {
+      "epoch": 0.7756831381510044,
+      "grad_norm": 3.0761612141378105,
+      "learning_rate": 2e-05,
+      "loss": 0.432,
+      "step": 32588
+    },
+    {
+      "epoch": 0.775706940874036,
+      "grad_norm": 8.803430901320523,
+      "learning_rate": 2e-05,
+      "loss": 0.5342,
+      "step": 32589
+    },
+    {
+      "epoch": 0.7757307435970675,
+      "grad_norm": 5.589800376604402,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 32590
+    },
+    {
+      "epoch": 0.775754546320099,
+      "grad_norm": 13.595032341983545,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 32591
+    },
+    {
+      "epoch": 0.7757783490431306,
+      "grad_norm": 2.8207008137140295,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 32592
+    },
+    {
+      "epoch": 0.7758021517661621,
+      "grad_norm": 1.0947362449989502,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 32593
+    },
+    {
+      "epoch": 0.7758259544891936,
+      "grad_norm": 0.022953177552945007,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 32594
+    },
+    {
+      "epoch": 0.775849757212225,
+      "grad_norm": 2.600911203113616,
+      "learning_rate": 2e-05,
+      "loss": 0.3813,
+      "step": 32595
+    },
+    {
+      "epoch": 0.7758735599352566,
+      "grad_norm": 1.998196266605263,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 32596
+    },
+    {
+      "epoch": 0.7758973626582881,
+      "grad_norm": 0.14068782810980468,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 32597
+    },
+    {
+      "epoch": 0.7759211653813196,
+      "grad_norm": 5.665089689733754,
+      "learning_rate": 2e-05,
+      "loss": 0.5804,
+      "step": 32598
+    },
+    {
+      "epoch": 0.7759449681043511,
+      "grad_norm": 5.83027469556965,
+      "learning_rate": 2e-05,
+      "loss": 0.1683,
+      "step": 32599
+    },
+    {
+      "epoch": 0.7759687708273827,
+      "grad_norm": 1.9500828981659741,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 32600
+    },
+    {
+      "epoch": 0.7759925735504142,
+      "grad_norm": 4.790411841626851,
+      "learning_rate": 2e-05,
+      "loss": 0.9926,
+      "step": 32601
+    },
+    {
+      "epoch": 0.7760163762734457,
+      "grad_norm": 0.1880327591208678,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 32602
+    },
+    {
+      "epoch": 0.7760401789964771,
+      "grad_norm": 2.024201256943681,
+      "learning_rate": 2e-05,
+      "loss": 0.0949,
+      "step": 32603
+    },
+    {
+      "epoch": 0.7760639817195087,
+      "grad_norm": 11.319249844141842,
+      "learning_rate": 2e-05,
+      "loss": 0.5498,
+      "step": 32604
+    },
+    {
+      "epoch": 0.7760877844425402,
+      "grad_norm": 5.102766993551562,
+      "learning_rate": 2e-05,
+      "loss": 0.5789,
+      "step": 32605
+    },
+    {
+      "epoch": 0.7761115871655717,
+      "grad_norm": 3.509272601331493,
+      "learning_rate": 2e-05,
+      "loss": 0.4209,
+      "step": 32606
+    },
+    {
+      "epoch": 0.7761353898886033,
+      "grad_norm": 1.6551094014203593,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 32607
+    },
+    {
+      "epoch": 0.7761591926116348,
+      "grad_norm": 2.9943873519180535,
+      "learning_rate": 2e-05,
+      "loss": 0.434,
+      "step": 32608
+    },
+    {
+      "epoch": 0.7761829953346663,
+      "grad_norm": 2.228373455522996,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 32609
+    },
+    {
+      "epoch": 0.7762067980576978,
+      "grad_norm": 2.474308149413355,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 32610
+    },
+    {
+      "epoch": 0.7762306007807294,
+      "grad_norm": 0.06802024511588853,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 32611
+    },
+    {
+      "epoch": 0.7762544035037608,
+      "grad_norm": 5.8472866921602105,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 32612
+    },
+    {
+      "epoch": 0.7762782062267923,
+      "grad_norm": 26.908382480479823,
+      "learning_rate": 2e-05,
+      "loss": 0.3222,
+      "step": 32613
+    },
+    {
+      "epoch": 0.7763020089498238,
+      "grad_norm": 0.8810328898147027,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 32614
+    },
+    {
+      "epoch": 0.7763258116728554,
+      "grad_norm": 6.531060484853489,
+      "learning_rate": 2e-05,
+      "loss": 0.1224,
+      "step": 32615
+    },
+    {
+      "epoch": 0.7763496143958869,
+      "grad_norm": 8.606475335777967,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 32616
+    },
+    {
+      "epoch": 0.7763734171189184,
+      "grad_norm": 7.191199269498473,
+      "learning_rate": 2e-05,
+      "loss": 0.1613,
+      "step": 32617
+    },
+    {
+      "epoch": 0.7763972198419499,
+      "grad_norm": 7.9207029752018645,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 32618
+    },
+    {
+      "epoch": 0.7764210225649815,
+      "grad_norm": 0.43481554533332667,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 32619
+    },
+    {
+      "epoch": 0.776444825288013,
+      "grad_norm": 3.268262073095025,
+      "learning_rate": 2e-05,
+      "loss": 0.6764,
+      "step": 32620
+    },
+    {
+      "epoch": 0.7764686280110444,
+      "grad_norm": 5.898163145620062,
+      "learning_rate": 2e-05,
+      "loss": 0.9349,
+      "step": 32621
+    },
+    {
+      "epoch": 0.776492430734076,
+      "grad_norm": 3.250504889939043,
+      "learning_rate": 2e-05,
+      "loss": 0.4079,
+      "step": 32622
+    },
+    {
+      "epoch": 0.7765162334571075,
+      "grad_norm": 6.493976371583247,
+      "learning_rate": 2e-05,
+      "loss": 0.4253,
+      "step": 32623
+    },
+    {
+      "epoch": 0.776540036180139,
+      "grad_norm": 3.8403685893812867,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 32624
+    },
+    {
+      "epoch": 0.7765638389031705,
+      "grad_norm": 7.308122028130431,
+      "learning_rate": 2e-05,
+      "loss": 0.4578,
+      "step": 32625
+    },
+    {
+      "epoch": 0.7765876416262021,
+      "grad_norm": 5.020725082696034,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 32626
+    },
+    {
+      "epoch": 0.7766114443492336,
+      "grad_norm": 35.728797195406905,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 32627
+    },
+    {
+      "epoch": 0.7766352470722651,
+      "grad_norm": 5.541840160981008,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 32628
+    },
+    {
+      "epoch": 0.7766590497952965,
+      "grad_norm": 1.918365063246418,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 32629
+    },
+    {
+      "epoch": 0.7766828525183281,
+      "grad_norm": 0.36486283435482636,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 32630
+    },
+    {
+      "epoch": 0.7767066552413596,
+      "grad_norm": 3.261805345508175,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 32631
+    },
+    {
+      "epoch": 0.7767304579643911,
+      "grad_norm": 198.95812750555805,
+      "learning_rate": 2e-05,
+      "loss": 0.5633,
+      "step": 32632
+    },
+    {
+      "epoch": 0.7767542606874226,
+      "grad_norm": 11.834420019901918,
+      "learning_rate": 2e-05,
+      "loss": 0.831,
+      "step": 32633
+    },
+    {
+      "epoch": 0.7767780634104542,
+      "grad_norm": 5.623048831652533,
+      "learning_rate": 2e-05,
+      "loss": 1.164,
+      "step": 32634
+    },
+    {
+      "epoch": 0.7768018661334857,
+      "grad_norm": 0.15118976876316614,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32635
+    },
+    {
+      "epoch": 0.7768256688565172,
+      "grad_norm": 2.743739149681466,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 32636
+    },
+    {
+      "epoch": 0.7768494715795486,
+      "grad_norm": 16.46446042581933,
+      "learning_rate": 2e-05,
+      "loss": 0.4467,
+      "step": 32637
+    },
+    {
+      "epoch": 0.7768732743025802,
+      "grad_norm": 4.982669650703168,
+      "learning_rate": 2e-05,
+      "loss": 0.9867,
+      "step": 32638
+    },
+    {
+      "epoch": 0.7768970770256117,
+      "grad_norm": 0.12460110431334298,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 32639
+    },
+    {
+      "epoch": 0.7769208797486432,
+      "grad_norm": 22.066438256159714,
+      "learning_rate": 2e-05,
+      "loss": 0.3136,
+      "step": 32640
+    },
+    {
+      "epoch": 0.7769446824716748,
+      "grad_norm": 4.706727099649173,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 32641
+    },
+    {
+      "epoch": 0.7769684851947063,
+      "grad_norm": 0.19013042158678892,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 32642
+    },
+    {
+      "epoch": 0.7769922879177378,
+      "grad_norm": 4.134562989665689,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 32643
+    },
+    {
+      "epoch": 0.7770160906407693,
+      "grad_norm": 3.4914365862556824,
+      "learning_rate": 2e-05,
+      "loss": 0.3963,
+      "step": 32644
+    },
+    {
+      "epoch": 0.7770398933638009,
+      "grad_norm": 0.8521837953166295,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 32645
+    },
+    {
+      "epoch": 0.7770636960868323,
+      "grad_norm": 5.212497469530289,
+      "learning_rate": 2e-05,
+      "loss": 0.9157,
+      "step": 32646
+    },
+    {
+      "epoch": 0.7770874988098638,
+      "grad_norm": 3.2238399524348837,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 32647
+    },
+    {
+      "epoch": 0.7771113015328953,
+      "grad_norm": 6.975780371785789,
+      "learning_rate": 2e-05,
+      "loss": 0.6569,
+      "step": 32648
+    },
+    {
+      "epoch": 0.7771351042559269,
+      "grad_norm": 4.333947750698887,
+      "learning_rate": 2e-05,
+      "loss": 0.4906,
+      "step": 32649
+    },
+    {
+      "epoch": 0.7771589069789584,
+      "grad_norm": 3.3057618751694293,
+      "learning_rate": 2e-05,
+      "loss": 0.5304,
+      "step": 32650
+    },
+    {
+      "epoch": 0.7771827097019899,
+      "grad_norm": 6.122849918593002,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 32651
+    },
+    {
+      "epoch": 0.7772065124250214,
+      "grad_norm": 1.7847361979337224,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 32652
+    },
+    {
+      "epoch": 0.777230315148053,
+      "grad_norm": 0.3750561383685642,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 32653
+    },
+    {
+      "epoch": 0.7772541178710844,
+      "grad_norm": 4.108145281706496,
+      "learning_rate": 2e-05,
+      "loss": 0.8353,
+      "step": 32654
+    },
+    {
+      "epoch": 0.7772779205941159,
+      "grad_norm": 3.2849839162094434,
+      "learning_rate": 2e-05,
+      "loss": 0.4512,
+      "step": 32655
+    },
+    {
+      "epoch": 0.7773017233171475,
+      "grad_norm": 6.277760623652587,
+      "learning_rate": 2e-05,
+      "loss": 0.5562,
+      "step": 32656
+    },
+    {
+      "epoch": 0.777325526040179,
+      "grad_norm": 14.616550712965381,
+      "learning_rate": 2e-05,
+      "loss": 0.6749,
+      "step": 32657
+    },
+    {
+      "epoch": 0.7773493287632105,
+      "grad_norm": 11.983884770504911,
+      "learning_rate": 2e-05,
+      "loss": 0.5475,
+      "step": 32658
+    },
+    {
+      "epoch": 0.777373131486242,
+      "grad_norm": 8.684750271018233,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 32659
+    },
+    {
+      "epoch": 0.7773969342092736,
+      "grad_norm": 7.010249599558518,
+      "learning_rate": 2e-05,
+      "loss": 0.1028,
+      "step": 32660
+    },
+    {
+      "epoch": 0.7774207369323051,
+      "grad_norm": 0.6236228150836246,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 32661
+    },
+    {
+      "epoch": 0.7774445396553366,
+      "grad_norm": 2.906730527525864,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 32662
+    },
+    {
+      "epoch": 0.777468342378368,
+      "grad_norm": 0.1721711119528491,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32663
+    },
+    {
+      "epoch": 0.7774921451013996,
+      "grad_norm": 14.281278670640837,
+      "learning_rate": 2e-05,
+      "loss": 0.7449,
+      "step": 32664
+    },
+    {
+      "epoch": 0.7775159478244311,
+      "grad_norm": 1.027058181557434,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 32665
+    },
+    {
+      "epoch": 0.7775397505474626,
+      "grad_norm": 0.4747698299484363,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 32666
+    },
+    {
+      "epoch": 0.7775635532704941,
+      "grad_norm": 23.520473621720296,
+      "learning_rate": 2e-05,
+      "loss": 1.1209,
+      "step": 32667
+    },
+    {
+      "epoch": 0.7775873559935257,
+      "grad_norm": 3.109261911805148,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 32668
+    },
+    {
+      "epoch": 0.7776111587165572,
+      "grad_norm": 0.03132126358451699,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32669
+    },
+    {
+      "epoch": 0.7776349614395887,
+      "grad_norm": 0.9922729783220579,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 32670
+    },
+    {
+      "epoch": 0.7776587641626203,
+      "grad_norm": 1.7515700968329144,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 32671
+    },
+    {
+      "epoch": 0.7776825668856517,
+      "grad_norm": 3.86078592024184,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 32672
+    },
+    {
+      "epoch": 0.7777063696086832,
+      "grad_norm": 7.644596555928478,
+      "learning_rate": 2e-05,
+      "loss": 1.4356,
+      "step": 32673
+    },
+    {
+      "epoch": 0.7777301723317147,
+      "grad_norm": 0.4960676862397163,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 32674
+    },
+    {
+      "epoch": 0.7777539750547463,
+      "grad_norm": 3.1422391833194028,
+      "learning_rate": 2e-05,
+      "loss": 0.4489,
+      "step": 32675
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 4.495387668307381,
+      "learning_rate": 2e-05,
+      "loss": 0.696,
+      "step": 32676
+    },
+    {
+      "epoch": 0.7778015805008093,
+      "grad_norm": 4.167775867936002,
+      "learning_rate": 2e-05,
+      "loss": 1.1204,
+      "step": 32677
+    },
+    {
+      "epoch": 0.7778253832238408,
+      "grad_norm": 4.647660819981936,
+      "learning_rate": 2e-05,
+      "loss": 1.3914,
+      "step": 32678
+    },
+    {
+      "epoch": 0.7778491859468724,
+      "grad_norm": 12.847728910740521,
+      "learning_rate": 2e-05,
+      "loss": 0.2926,
+      "step": 32679
+    },
+    {
+      "epoch": 0.7778729886699038,
+      "grad_norm": 5.545845291893406,
+      "learning_rate": 2e-05,
+      "loss": 1.13,
+      "step": 32680
+    },
+    {
+      "epoch": 0.7778967913929353,
+      "grad_norm": 14.601492846534525,
+      "learning_rate": 2e-05,
+      "loss": 1.4278,
+      "step": 32681
+    },
+    {
+      "epoch": 0.7779205941159668,
+      "grad_norm": 0.1029999283539793,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 32682
+    },
+    {
+      "epoch": 0.7779443968389984,
+      "grad_norm": 0.14033268825995784,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 32683
+    },
+    {
+      "epoch": 0.7779681995620299,
+      "grad_norm": 2.1010760734122957,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 32684
+    },
+    {
+      "epoch": 0.7779920022850614,
+      "grad_norm": 4.893776809654484,
+      "learning_rate": 2e-05,
+      "loss": 0.9487,
+      "step": 32685
+    },
+    {
+      "epoch": 0.778015805008093,
+      "grad_norm": 0.9285670837504991,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 32686
+    },
+    {
+      "epoch": 0.7780396077311245,
+      "grad_norm": 4.472054653497078,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 32687
+    },
+    {
+      "epoch": 0.778063410454156,
+      "grad_norm": 23.761935356980736,
+      "learning_rate": 2e-05,
+      "loss": 1.0872,
+      "step": 32688
+    },
+    {
+      "epoch": 0.7780872131771874,
+      "grad_norm": 1.46612450017492,
+      "learning_rate": 2e-05,
+      "loss": 0.0974,
+      "step": 32689
+    },
+    {
+      "epoch": 0.778111015900219,
+      "grad_norm": 0.3354246180336516,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 32690
+    },
+    {
+      "epoch": 0.7781348186232505,
+      "grad_norm": 6.046179348507732,
+      "learning_rate": 2e-05,
+      "loss": 0.9298,
+      "step": 32691
+    },
+    {
+      "epoch": 0.778158621346282,
+      "grad_norm": 2.4607657000655516,
+      "learning_rate": 2e-05,
+      "loss": 0.3663,
+      "step": 32692
+    },
+    {
+      "epoch": 0.7781824240693135,
+      "grad_norm": 10.879721997697445,
+      "learning_rate": 2e-05,
+      "loss": 1.1908,
+      "step": 32693
+    },
+    {
+      "epoch": 0.7782062267923451,
+      "grad_norm": 4.7347315728598245,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 32694
+    },
+    {
+      "epoch": 0.7782300295153766,
+      "grad_norm": 5.403154991553345,
+      "learning_rate": 2e-05,
+      "loss": 0.8014,
+      "step": 32695
+    },
+    {
+      "epoch": 0.778253832238408,
+      "grad_norm": 0.8011871221869101,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 32696
+    },
+    {
+      "epoch": 0.7782776349614395,
+      "grad_norm": 1.3587156678506402,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 32697
+    },
+    {
+      "epoch": 0.7783014376844711,
+      "grad_norm": 2.7353888703786797,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 32698
+    },
+    {
+      "epoch": 0.7783252404075026,
+      "grad_norm": 1.0321841854433749,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 32699
+    },
+    {
+      "epoch": 0.7783490431305341,
+      "grad_norm": 4.624723097168638,
+      "learning_rate": 2e-05,
+      "loss": 1.0523,
+      "step": 32700
+    },
+    {
+      "epoch": 0.7783728458535657,
+      "grad_norm": 7.523036437551007,
+      "learning_rate": 2e-05,
+      "loss": 0.9134,
+      "step": 32701
+    },
+    {
+      "epoch": 0.7783966485765972,
+      "grad_norm": 1.4076668100438026,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 32702
+    },
+    {
+      "epoch": 0.7784204512996287,
+      "grad_norm": 5.046664159096491,
+      "learning_rate": 2e-05,
+      "loss": 0.8438,
+      "step": 32703
+    },
+    {
+      "epoch": 0.7784442540226602,
+      "grad_norm": 11.538932042287549,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 32704
+    },
+    {
+      "epoch": 0.7784680567456917,
+      "grad_norm": 3.0985675299638666,
+      "learning_rate": 2e-05,
+      "loss": 0.4418,
+      "step": 32705
+    },
+    {
+      "epoch": 0.7784918594687232,
+      "grad_norm": 5.35395551714285,
+      "learning_rate": 2e-05,
+      "loss": 1.0281,
+      "step": 32706
+    },
+    {
+      "epoch": 0.7785156621917547,
+      "grad_norm": 7.440937339347924,
+      "learning_rate": 2e-05,
+      "loss": 0.7819,
+      "step": 32707
+    },
+    {
+      "epoch": 0.7785394649147862,
+      "grad_norm": 4.937841001306537,
+      "learning_rate": 2e-05,
+      "loss": 1.0235,
+      "step": 32708
+    },
+    {
+      "epoch": 0.7785632676378178,
+      "grad_norm": 0.12811821111267513,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 32709
+    },
+    {
+      "epoch": 0.7785870703608493,
+      "grad_norm": 2.318247110614333,
+      "learning_rate": 2e-05,
+      "loss": 0.1863,
+      "step": 32710
+    },
+    {
+      "epoch": 0.7786108730838808,
+      "grad_norm": 2.3405063127909123,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 32711
+    },
+    {
+      "epoch": 0.7786346758069123,
+      "grad_norm": 9.974353920135382,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 32712
+    },
+    {
+      "epoch": 0.7786584785299439,
+      "grad_norm": 3.6753097656537728,
+      "learning_rate": 2e-05,
+      "loss": 0.7985,
+      "step": 32713
+    },
+    {
+      "epoch": 0.7786822812529753,
+      "grad_norm": 6.097173830771579,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 32714
+    },
+    {
+      "epoch": 0.7787060839760068,
+      "grad_norm": 5.483539004036612,
+      "learning_rate": 2e-05,
+      "loss": 0.1187,
+      "step": 32715
+    },
+    {
+      "epoch": 0.7787298866990384,
+      "grad_norm": 16.98409238706536,
+      "learning_rate": 2e-05,
+      "loss": 0.3385,
+      "step": 32716
+    },
+    {
+      "epoch": 0.7787536894220699,
+      "grad_norm": 7.302340331637979,
+      "learning_rate": 2e-05,
+      "loss": 1.2206,
+      "step": 32717
+    },
+    {
+      "epoch": 0.7787774921451014,
+      "grad_norm": 4.746578617589293,
+      "learning_rate": 2e-05,
+      "loss": 1.0051,
+      "step": 32718
+    },
+    {
+      "epoch": 0.7788012948681329,
+      "grad_norm": 3.2535880110735302,
+      "learning_rate": 2e-05,
+      "loss": 0.6125,
+      "step": 32719
+    },
+    {
+      "epoch": 0.7788250975911645,
+      "grad_norm": 6.649692644656783,
+      "learning_rate": 2e-05,
+      "loss": 0.6742,
+      "step": 32720
+    },
+    {
+      "epoch": 0.778848900314196,
+      "grad_norm": 3.2158499185260667,
+      "learning_rate": 2e-05,
+      "loss": 0.4322,
+      "step": 32721
+    },
+    {
+      "epoch": 0.7788727030372274,
+      "grad_norm": 1.5827596429507795,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 32722
+    },
+    {
+      "epoch": 0.7788965057602589,
+      "grad_norm": 3.290244796786095,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 32723
+    },
+    {
+      "epoch": 0.7789203084832905,
+      "grad_norm": 0.3669526145156053,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 32724
+    },
+    {
+      "epoch": 0.778944111206322,
+      "grad_norm": 1.040883693422839,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 32725
+    },
+    {
+      "epoch": 0.7789679139293535,
+      "grad_norm": 1.607322707446452,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 32726
+    },
+    {
+      "epoch": 0.778991716652385,
+      "grad_norm": 1.1571367816617555,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 32727
+    },
+    {
+      "epoch": 0.7790155193754166,
+      "grad_norm": 8.771075366152335,
+      "learning_rate": 2e-05,
+      "loss": 0.4808,
+      "step": 32728
+    },
+    {
+      "epoch": 0.7790393220984481,
+      "grad_norm": 4.025865206915664,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 32729
+    },
+    {
+      "epoch": 0.7790631248214795,
+      "grad_norm": 0.08721696710072757,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 32730
+    },
+    {
+      "epoch": 0.7790869275445111,
+      "grad_norm": 4.599906809015051,
+      "learning_rate": 2e-05,
+      "loss": 0.1916,
+      "step": 32731
+    },
+    {
+      "epoch": 0.7791107302675426,
+      "grad_norm": 0.15173220907480603,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 32732
+    },
+    {
+      "epoch": 0.7791345329905741,
+      "grad_norm": 5.089071571295829,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 32733
+    },
+    {
+      "epoch": 0.7791583357136056,
+      "grad_norm": 1.77517459188777,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 32734
+    },
+    {
+      "epoch": 0.7791821384366372,
+      "grad_norm": 3.317508108608057,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 32735
+    },
+    {
+      "epoch": 0.7792059411596687,
+      "grad_norm": 4.6477233979621,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 32736
+    },
+    {
+      "epoch": 0.7792297438827002,
+      "grad_norm": 4.911245605713145,
+      "learning_rate": 2e-05,
+      "loss": 0.8723,
+      "step": 32737
+    },
+    {
+      "epoch": 0.7792535466057317,
+      "grad_norm": 6.635684927067468,
+      "learning_rate": 2e-05,
+      "loss": 0.9769,
+      "step": 32738
+    },
+    {
+      "epoch": 0.7792773493287632,
+      "grad_norm": 2.2520244494754467,
+      "learning_rate": 2e-05,
+      "loss": 0.046,
+      "step": 32739
+    },
+    {
+      "epoch": 0.7793011520517947,
+      "grad_norm": 2.871999022063798,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 32740
+    },
+    {
+      "epoch": 0.7793249547748262,
+      "grad_norm": 0.7560921159481897,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 32741
+    },
+    {
+      "epoch": 0.7793487574978577,
+      "grad_norm": 3.8887133962729665,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 32742
+    },
+    {
+      "epoch": 0.7793725602208893,
+      "grad_norm": 12.857909211927678,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 32743
+    },
+    {
+      "epoch": 0.7793963629439208,
+      "grad_norm": 1.5863672658363046,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 32744
+    },
+    {
+      "epoch": 0.7794201656669523,
+      "grad_norm": 1.508594282008585,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 32745
+    },
+    {
+      "epoch": 0.7794439683899839,
+      "grad_norm": 0.04166666259255971,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32746
+    },
+    {
+      "epoch": 0.7794677711130154,
+      "grad_norm": 0.9436421053770738,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 32747
+    },
+    {
+      "epoch": 0.7794915738360468,
+      "grad_norm": 2.8219606124497156,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 32748
+    },
+    {
+      "epoch": 0.7795153765590783,
+      "grad_norm": 2.705616892432554,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 32749
+    },
+    {
+      "epoch": 0.7795391792821099,
+      "grad_norm": 10.340512414366268,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 32750
+    },
+    {
+      "epoch": 0.7795629820051414,
+      "grad_norm": 0.6664757847437519,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 32751
+    },
+    {
+      "epoch": 0.7795867847281729,
+      "grad_norm": 0.28349304749513315,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 32752
+    },
+    {
+      "epoch": 0.7796105874512044,
+      "grad_norm": 0.8276655006870842,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 32753
+    },
+    {
+      "epoch": 0.779634390174236,
+      "grad_norm": 0.2247112559763168,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 32754
+    },
+    {
+      "epoch": 0.7796581928972675,
+      "grad_norm": 2.5698175954586966,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 32755
+    },
+    {
+      "epoch": 0.7796819956202989,
+      "grad_norm": 2.6460647891653397,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 32756
+    },
+    {
+      "epoch": 0.7797057983433304,
+      "grad_norm": 0.8657282215600247,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 32757
+    },
+    {
+      "epoch": 0.779729601066362,
+      "grad_norm": 8.18094006443499,
+      "learning_rate": 2e-05,
+      "loss": 1.5096,
+      "step": 32758
+    },
+    {
+      "epoch": 0.7797534037893935,
+      "grad_norm": 5.840815811931992,
+      "learning_rate": 2e-05,
+      "loss": 1.0598,
+      "step": 32759
+    },
+    {
+      "epoch": 0.779777206512425,
+      "grad_norm": 12.047194583194774,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 32760
+    },
+    {
+      "epoch": 0.7798010092354566,
+      "grad_norm": 0.113931438200202,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 32761
+    },
+    {
+      "epoch": 0.7798248119584881,
+      "grad_norm": 9.698080076659044,
+      "learning_rate": 2e-05,
+      "loss": 0.4771,
+      "step": 32762
+    },
+    {
+      "epoch": 0.7798486146815196,
+      "grad_norm": 4.010678433901052,
+      "learning_rate": 2e-05,
+      "loss": 0.6702,
+      "step": 32763
+    },
+    {
+      "epoch": 0.779872417404551,
+      "grad_norm": 0.9937869952475641,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 32764
+    },
+    {
+      "epoch": 0.7798962201275826,
+      "grad_norm": 5.644262053615369,
+      "learning_rate": 2e-05,
+      "loss": 0.5792,
+      "step": 32765
+    },
+    {
+      "epoch": 0.7799200228506141,
+      "grad_norm": 7.259205345956688,
+      "learning_rate": 2e-05,
+      "loss": 0.6674,
+      "step": 32766
+    },
+    {
+      "epoch": 0.7799438255736456,
+      "grad_norm": 0.02635713304201142,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32767
+    },
+    {
+      "epoch": 0.7799676282966771,
+      "grad_norm": 4.211396783319766,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 32768
+    },
+    {
+      "epoch": 0.7799914310197087,
+      "grad_norm": 0.04520752284348628,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 32769
+    },
+    {
+      "epoch": 0.7800152337427402,
+      "grad_norm": 0.05288736371385382,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 32770
+    },
+    {
+      "epoch": 0.7800390364657717,
+      "grad_norm": 4.115918713728596,
+      "learning_rate": 2e-05,
+      "loss": 0.1854,
+      "step": 32771
+    },
+    {
+      "epoch": 0.7800628391888031,
+      "grad_norm": 0.06191876021895952,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 32772
+    },
+    {
+      "epoch": 0.7800866419118347,
+      "grad_norm": 4.0644628441238435,
+      "learning_rate": 2e-05,
+      "loss": 0.6817,
+      "step": 32773
+    },
+    {
+      "epoch": 0.7801104446348662,
+      "grad_norm": 0.5110630242778176,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 32774
+    },
+    {
+      "epoch": 0.7801342473578977,
+      "grad_norm": 4.638583027439547,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 32775
+    },
+    {
+      "epoch": 0.7801580500809293,
+      "grad_norm": 4.282307161872874,
+      "learning_rate": 2e-05,
+      "loss": 0.9345,
+      "step": 32776
+    },
+    {
+      "epoch": 0.7801818528039608,
+      "grad_norm": 3.4326880578179093,
+      "learning_rate": 2e-05,
+      "loss": 0.5893,
+      "step": 32777
+    },
+    {
+      "epoch": 0.7802056555269923,
+      "grad_norm": 3.8131161492497756,
+      "learning_rate": 2e-05,
+      "loss": 0.3936,
+      "step": 32778
+    },
+    {
+      "epoch": 0.7802294582500238,
+      "grad_norm": 4.751428033713429,
+      "learning_rate": 2e-05,
+      "loss": 0.8682,
+      "step": 32779
+    },
+    {
+      "epoch": 0.7802532609730554,
+      "grad_norm": 0.6189818854943212,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 32780
+    },
+    {
+      "epoch": 0.7802770636960868,
+      "grad_norm": 2.437709276443024,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 32781
+    },
+    {
+      "epoch": 0.7803008664191183,
+      "grad_norm": 1.4923606237726739,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 32782
+    },
+    {
+      "epoch": 0.7803246691421498,
+      "grad_norm": 21.815376948756693,
+      "learning_rate": 2e-05,
+      "loss": 0.2188,
+      "step": 32783
+    },
+    {
+      "epoch": 0.7803484718651814,
+      "grad_norm": 4.11755209586775,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 32784
+    },
+    {
+      "epoch": 0.7803722745882129,
+      "grad_norm": 9.145150928753987,
+      "learning_rate": 2e-05,
+      "loss": 1.0254,
+      "step": 32785
+    },
+    {
+      "epoch": 0.7803960773112444,
+      "grad_norm": 3.604220856382969,
+      "learning_rate": 2e-05,
+      "loss": 0.4619,
+      "step": 32786
+    },
+    {
+      "epoch": 0.7804198800342759,
+      "grad_norm": 5.0023162723838785,
+      "learning_rate": 2e-05,
+      "loss": 0.8777,
+      "step": 32787
+    },
+    {
+      "epoch": 0.7804436827573075,
+      "grad_norm": 4.963710973795909,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 32788
+    },
+    {
+      "epoch": 0.780467485480339,
+      "grad_norm": 3.5259331175850455,
+      "learning_rate": 2e-05,
+      "loss": 0.6506,
+      "step": 32789
+    },
+    {
+      "epoch": 0.7804912882033704,
+      "grad_norm": 1.915309744259648,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 32790
+    },
+    {
+      "epoch": 0.780515090926402,
+      "grad_norm": 3.2050714025260394,
+      "learning_rate": 2e-05,
+      "loss": 0.3135,
+      "step": 32791
+    },
+    {
+      "epoch": 0.7805388936494335,
+      "grad_norm": 6.354570938843158,
+      "learning_rate": 2e-05,
+      "loss": 0.7463,
+      "step": 32792
+    },
+    {
+      "epoch": 0.780562696372465,
+      "grad_norm": 0.17075478086630694,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 32793
+    },
+    {
+      "epoch": 0.7805864990954965,
+      "grad_norm": 4.683165287659766,
+      "learning_rate": 2e-05,
+      "loss": 0.6682,
+      "step": 32794
+    },
+    {
+      "epoch": 0.7806103018185281,
+      "grad_norm": 5.21769970100374,
+      "learning_rate": 2e-05,
+      "loss": 1.4818,
+      "step": 32795
+    },
+    {
+      "epoch": 0.7806341045415596,
+      "grad_norm": 2.0052580464417677,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 32796
+    },
+    {
+      "epoch": 0.7806579072645911,
+      "grad_norm": 4.3783664735680725,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 32797
+    },
+    {
+      "epoch": 0.7806817099876225,
+      "grad_norm": 5.424212357262743,
+      "learning_rate": 2e-05,
+      "loss": 1.0881,
+      "step": 32798
+    },
+    {
+      "epoch": 0.7807055127106541,
+      "grad_norm": 0.6724721562306993,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 32799
+    },
+    {
+      "epoch": 0.7807293154336856,
+      "grad_norm": 3.5346118489475016,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 32800
+    },
+    {
+      "epoch": 0.7807531181567171,
+      "grad_norm": 9.190730235249957,
+      "learning_rate": 2e-05,
+      "loss": 0.658,
+      "step": 32801
+    },
+    {
+      "epoch": 0.7807769208797486,
+      "grad_norm": 5.467640016021015,
+      "learning_rate": 2e-05,
+      "loss": 0.2929,
+      "step": 32802
+    },
+    {
+      "epoch": 0.7808007236027802,
+      "grad_norm": 14.910380110048663,
+      "learning_rate": 2e-05,
+      "loss": 1.8203,
+      "step": 32803
+    },
+    {
+      "epoch": 0.7808245263258117,
+      "grad_norm": 6.316428312221066,
+      "learning_rate": 2e-05,
+      "loss": 0.3213,
+      "step": 32804
+    },
+    {
+      "epoch": 0.7808483290488432,
+      "grad_norm": 3.0910534464429253,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 32805
+    },
+    {
+      "epoch": 0.7808721317718748,
+      "grad_norm": 0.07524187837122337,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 32806
+    },
+    {
+      "epoch": 0.7808959344949062,
+      "grad_norm": 3.688024463651525,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 32807
+    },
+    {
+      "epoch": 0.7809197372179377,
+      "grad_norm": 11.397878883369177,
+      "learning_rate": 2e-05,
+      "loss": 1.4951,
+      "step": 32808
+    },
+    {
+      "epoch": 0.7809435399409692,
+      "grad_norm": 6.12867934390048,
+      "learning_rate": 2e-05,
+      "loss": 0.8763,
+      "step": 32809
+    },
+    {
+      "epoch": 0.7809673426640008,
+      "grad_norm": 2.4273800342416334,
+      "learning_rate": 2e-05,
+      "loss": 0.2019,
+      "step": 32810
+    },
+    {
+      "epoch": 0.7809911453870323,
+      "grad_norm": 4.108612030536035,
+      "learning_rate": 2e-05,
+      "loss": 0.6843,
+      "step": 32811
+    },
+    {
+      "epoch": 0.7810149481100638,
+      "grad_norm": 2.5855749926644456,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 32812
+    },
+    {
+      "epoch": 0.7810387508330953,
+      "grad_norm": 4.090406210318663,
+      "learning_rate": 2e-05,
+      "loss": 0.4962,
+      "step": 32813
+    },
+    {
+      "epoch": 0.7810625535561269,
+      "grad_norm": 5.139513479396754,
+      "learning_rate": 2e-05,
+      "loss": 1.2966,
+      "step": 32814
+    },
+    {
+      "epoch": 0.7810863562791583,
+      "grad_norm": 4.05695271612921,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 32815
+    },
+    {
+      "epoch": 0.7811101590021898,
+      "grad_norm": 7.6283012671392,
+      "learning_rate": 2e-05,
+      "loss": 1.4568,
+      "step": 32816
+    },
+    {
+      "epoch": 0.7811339617252213,
+      "grad_norm": 4.84880145987827,
+      "learning_rate": 2e-05,
+      "loss": 0.6352,
+      "step": 32817
+    },
+    {
+      "epoch": 0.7811577644482529,
+      "grad_norm": 4.464414934000629,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 32818
+    },
+    {
+      "epoch": 0.7811815671712844,
+      "grad_norm": 0.2221928419681598,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 32819
+    },
+    {
+      "epoch": 0.7812053698943159,
+      "grad_norm": 4.308770264013967,
+      "learning_rate": 2e-05,
+      "loss": 0.3957,
+      "step": 32820
+    },
+    {
+      "epoch": 0.7812291726173475,
+      "grad_norm": 0.1762682116358501,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 32821
+    },
+    {
+      "epoch": 0.781252975340379,
+      "grad_norm": 6.088295143590219,
+      "learning_rate": 2e-05,
+      "loss": 0.8791,
+      "step": 32822
+    },
+    {
+      "epoch": 0.7812767780634104,
+      "grad_norm": 4.768431062659608,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 32823
+    },
+    {
+      "epoch": 0.7813005807864419,
+      "grad_norm": 4.086542272626576,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 32824
+    },
+    {
+      "epoch": 0.7813243835094735,
+      "grad_norm": 6.1622882659627045,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 32825
+    },
+    {
+      "epoch": 0.781348186232505,
+      "grad_norm": 5.80516784663845,
+      "learning_rate": 2e-05,
+      "loss": 1.134,
+      "step": 32826
+    },
+    {
+      "epoch": 0.7813719889555365,
+      "grad_norm": 0.20578238735015605,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 32827
+    },
+    {
+      "epoch": 0.781395791678568,
+      "grad_norm": 1.6163773788922253,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 32828
+    },
+    {
+      "epoch": 0.7814195944015996,
+      "grad_norm": 1.5950378210131462,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 32829
+    },
+    {
+      "epoch": 0.7814433971246311,
+      "grad_norm": 0.5320331679115463,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 32830
+    },
+    {
+      "epoch": 0.7814671998476626,
+      "grad_norm": 0.028063867689560627,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 32831
+    },
+    {
+      "epoch": 0.781491002570694,
+      "grad_norm": 6.0026453683172845,
+      "learning_rate": 2e-05,
+      "loss": 0.6069,
+      "step": 32832
+    },
+    {
+      "epoch": 0.7815148052937256,
+      "grad_norm": 3.8342843956924453,
+      "learning_rate": 2e-05,
+      "loss": 0.1258,
+      "step": 32833
+    },
+    {
+      "epoch": 0.7815386080167571,
+      "grad_norm": 3.516390402420925,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 32834
+    },
+    {
+      "epoch": 0.7815624107397886,
+      "grad_norm": 13.063710686431358,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 32835
+    },
+    {
+      "epoch": 0.7815862134628202,
+      "grad_norm": 6.70257487203708,
+      "learning_rate": 2e-05,
+      "loss": 1.6271,
+      "step": 32836
+    },
+    {
+      "epoch": 0.7816100161858517,
+      "grad_norm": 3.82872286333448,
+      "learning_rate": 2e-05,
+      "loss": 0.6396,
+      "step": 32837
+    },
+    {
+      "epoch": 0.7816338189088832,
+      "grad_norm": 0.22988876830393248,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 32838
+    },
+    {
+      "epoch": 0.7816576216319147,
+      "grad_norm": 4.46526257485452,
+      "learning_rate": 2e-05,
+      "loss": 0.7197,
+      "step": 32839
+    },
+    {
+      "epoch": 0.7816814243549463,
+      "grad_norm": 5.658595531553834,
+      "learning_rate": 2e-05,
+      "loss": 0.3059,
+      "step": 32840
+    },
+    {
+      "epoch": 0.7817052270779777,
+      "grad_norm": 2.8512266172348566,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 32841
+    },
+    {
+      "epoch": 0.7817290298010092,
+      "grad_norm": 2.9390996715088344,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 32842
+    },
+    {
+      "epoch": 0.7817528325240407,
+      "grad_norm": 0.6327306287428985,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 32843
+    },
+    {
+      "epoch": 0.7817766352470723,
+      "grad_norm": 3.084272231892283,
+      "learning_rate": 2e-05,
+      "loss": 0.0984,
+      "step": 32844
+    },
+    {
+      "epoch": 0.7818004379701038,
+      "grad_norm": 4.656229504848714,
+      "learning_rate": 2e-05,
+      "loss": 0.7758,
+      "step": 32845
+    },
+    {
+      "epoch": 0.7818242406931353,
+      "grad_norm": 6.116774560645489,
+      "learning_rate": 2e-05,
+      "loss": 1.5167,
+      "step": 32846
+    },
+    {
+      "epoch": 0.7818480434161668,
+      "grad_norm": 3.8715342641464074,
+      "learning_rate": 2e-05,
+      "loss": 0.7435,
+      "step": 32847
+    },
+    {
+      "epoch": 0.7818718461391984,
+      "grad_norm": 0.6367053910808277,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 32848
+    },
+    {
+      "epoch": 0.7818956488622298,
+      "grad_norm": 2.776041331534557,
+      "learning_rate": 2e-05,
+      "loss": 0.5108,
+      "step": 32849
+    },
+    {
+      "epoch": 0.7819194515852613,
+      "grad_norm": 2.3868409688836647,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 32850
+    },
+    {
+      "epoch": 0.7819432543082929,
+      "grad_norm": 3.2099078541580535,
+      "learning_rate": 2e-05,
+      "loss": 0.5286,
+      "step": 32851
+    },
+    {
+      "epoch": 0.7819670570313244,
+      "grad_norm": 1.4388629328085274,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 32852
+    },
+    {
+      "epoch": 0.7819908597543559,
+      "grad_norm": 2.1780406116711366,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 32853
+    },
+    {
+      "epoch": 0.7820146624773874,
+      "grad_norm": 0.1262131195904222,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 32854
+    },
+    {
+      "epoch": 0.782038465200419,
+      "grad_norm": 4.578789068872995,
+      "learning_rate": 2e-05,
+      "loss": 0.3771,
+      "step": 32855
+    },
+    {
+      "epoch": 0.7820622679234505,
+      "grad_norm": 5.918367854115682,
+      "learning_rate": 2e-05,
+      "loss": 0.9713,
+      "step": 32856
+    },
+    {
+      "epoch": 0.782086070646482,
+      "grad_norm": 6.409207237440389,
+      "learning_rate": 2e-05,
+      "loss": 1.2462,
+      "step": 32857
+    },
+    {
+      "epoch": 0.7821098733695134,
+      "grad_norm": 23.349069518896712,
+      "learning_rate": 2e-05,
+      "loss": 0.3473,
+      "step": 32858
+    },
+    {
+      "epoch": 0.782133676092545,
+      "grad_norm": 1.6455156589817455,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 32859
+    },
+    {
+      "epoch": 0.7821574788155765,
+      "grad_norm": 0.21359139793941093,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 32860
+    },
+    {
+      "epoch": 0.782181281538608,
+      "grad_norm": 0.15053165289787906,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 32861
+    },
+    {
+      "epoch": 0.7822050842616395,
+      "grad_norm": 7.842700272626062,
+      "learning_rate": 2e-05,
+      "loss": 0.491,
+      "step": 32862
+    },
+    {
+      "epoch": 0.7822288869846711,
+      "grad_norm": 1.9848851896244724,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 32863
+    },
+    {
+      "epoch": 0.7822526897077026,
+      "grad_norm": 1.972742849348438,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 32864
+    },
+    {
+      "epoch": 0.782276492430734,
+      "grad_norm": 0.08154219215664661,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 32865
+    },
+    {
+      "epoch": 0.7823002951537656,
+      "grad_norm": 4.3280137832004755,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 32866
+    },
+    {
+      "epoch": 0.7823240978767971,
+      "grad_norm": 5.603337998222933,
+      "learning_rate": 2e-05,
+      "loss": 0.9172,
+      "step": 32867
+    },
+    {
+      "epoch": 0.7823479005998286,
+      "grad_norm": 0.21228537675344916,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 32868
+    },
+    {
+      "epoch": 0.7823717033228601,
+      "grad_norm": 0.3177182475071255,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 32869
+    },
+    {
+      "epoch": 0.7823955060458917,
+      "grad_norm": 3.6662986195705836,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 32870
+    },
+    {
+      "epoch": 0.7824193087689232,
+      "grad_norm": 7.562947677221506,
+      "learning_rate": 2e-05,
+      "loss": 0.5965,
+      "step": 32871
+    },
+    {
+      "epoch": 0.7824431114919547,
+      "grad_norm": 4.16647966027029,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 32872
+    },
+    {
+      "epoch": 0.7824669142149862,
+      "grad_norm": 2.6177996105548615,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 32873
+    },
+    {
+      "epoch": 0.7824907169380177,
+      "grad_norm": 10.058794276458077,
+      "learning_rate": 2e-05,
+      "loss": 0.7319,
+      "step": 32874
+    },
+    {
+      "epoch": 0.7825145196610492,
+      "grad_norm": 3.207396636598616,
+      "learning_rate": 2e-05,
+      "loss": 0.3193,
+      "step": 32875
+    },
+    {
+      "epoch": 0.7825383223840807,
+      "grad_norm": 10.01368868808815,
+      "learning_rate": 2e-05,
+      "loss": 0.2297,
+      "step": 32876
+    },
+    {
+      "epoch": 0.7825621251071122,
+      "grad_norm": 2.262603956329403,
+      "learning_rate": 2e-05,
+      "loss": 0.1973,
+      "step": 32877
+    },
+    {
+      "epoch": 0.7825859278301438,
+      "grad_norm": 0.026588822058771826,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 32878
+    },
+    {
+      "epoch": 0.7826097305531753,
+      "grad_norm": 1.2671927492439534,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 32879
+    },
+    {
+      "epoch": 0.7826335332762068,
+      "grad_norm": 3.2525445328894897,
+      "learning_rate": 2e-05,
+      "loss": 0.3923,
+      "step": 32880
+    },
+    {
+      "epoch": 0.7826573359992384,
+      "grad_norm": 8.634251220253633,
+      "learning_rate": 2e-05,
+      "loss": 0.667,
+      "step": 32881
+    },
+    {
+      "epoch": 0.7826811387222699,
+      "grad_norm": 2.8065361823457016,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 32882
+    },
+    {
+      "epoch": 0.7827049414453013,
+      "grad_norm": 4.748849370255118,
+      "learning_rate": 2e-05,
+      "loss": 1.5164,
+      "step": 32883
+    },
+    {
+      "epoch": 0.7827287441683328,
+      "grad_norm": 6.1952724827405925,
+      "learning_rate": 2e-05,
+      "loss": 1.2566,
+      "step": 32884
+    },
+    {
+      "epoch": 0.7827525468913644,
+      "grad_norm": 2.052455643784812,
+      "learning_rate": 2e-05,
+      "loss": 0.2821,
+      "step": 32885
+    },
+    {
+      "epoch": 0.7827763496143959,
+      "grad_norm": 0.0447580707784613,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 32886
+    },
+    {
+      "epoch": 0.7828001523374274,
+      "grad_norm": 0.1756876093971729,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 32887
+    },
+    {
+      "epoch": 0.7828239550604589,
+      "grad_norm": 0.2183045642784422,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 32888
+    },
+    {
+      "epoch": 0.7828477577834905,
+      "grad_norm": 9.358756243830243,
+      "learning_rate": 2e-05,
+      "loss": 1.8541,
+      "step": 32889
+    },
+    {
+      "epoch": 0.782871560506522,
+      "grad_norm": 6.265916263861451,
+      "learning_rate": 2e-05,
+      "loss": 0.3763,
+      "step": 32890
+    },
+    {
+      "epoch": 0.7828953632295534,
+      "grad_norm": 11.368201990838243,
+      "learning_rate": 2e-05,
+      "loss": 0.2892,
+      "step": 32891
+    },
+    {
+      "epoch": 0.7829191659525849,
+      "grad_norm": 3.471806490347027,
+      "learning_rate": 2e-05,
+      "loss": 0.565,
+      "step": 32892
+    },
+    {
+      "epoch": 0.7829429686756165,
+      "grad_norm": 3.5060902186124685,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 32893
+    },
+    {
+      "epoch": 0.782966771398648,
+      "grad_norm": 0.304280090605603,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 32894
+    },
+    {
+      "epoch": 0.7829905741216795,
+      "grad_norm": 5.004176680600772,
+      "learning_rate": 2e-05,
+      "loss": 0.815,
+      "step": 32895
+    },
+    {
+      "epoch": 0.783014376844711,
+      "grad_norm": 2.2481712990571148,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 32896
+    },
+    {
+      "epoch": 0.7830381795677426,
+      "grad_norm": 0.07981632573104366,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 32897
+    },
+    {
+      "epoch": 0.7830619822907741,
+      "grad_norm": 2.3180152039245296,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 32898
+    },
+    {
+      "epoch": 0.7830857850138055,
+      "grad_norm": 2.598773531140837,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 32899
+    },
+    {
+      "epoch": 0.7831095877368371,
+      "grad_norm": 1.5310715000788906,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 32900
+    },
+    {
+      "epoch": 0.7831333904598686,
+      "grad_norm": 18.109761362340162,
+      "learning_rate": 2e-05,
+      "loss": 1.4543,
+      "step": 32901
+    },
+    {
+      "epoch": 0.7831571931829001,
+      "grad_norm": 4.105553376527764,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 32902
+    },
+    {
+      "epoch": 0.7831809959059316,
+      "grad_norm": 1.7621141112599368,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 32903
+    },
+    {
+      "epoch": 0.7832047986289632,
+      "grad_norm": 3.225355217008895,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 32904
+    },
+    {
+      "epoch": 0.7832286013519947,
+      "grad_norm": 24.084672634723166,
+      "learning_rate": 2e-05,
+      "loss": 0.368,
+      "step": 32905
+    },
+    {
+      "epoch": 0.7832524040750262,
+      "grad_norm": 5.378607819291883,
+      "learning_rate": 2e-05,
+      "loss": 0.9201,
+      "step": 32906
+    },
+    {
+      "epoch": 0.7832762067980577,
+      "grad_norm": 5.011526989190345,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 32907
+    },
+    {
+      "epoch": 0.7833000095210892,
+      "grad_norm": 2.98323350373836,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 32908
+    },
+    {
+      "epoch": 0.7833238122441207,
+      "grad_norm": 3.678183662899905,
+      "learning_rate": 2e-05,
+      "loss": 0.446,
+      "step": 32909
+    },
+    {
+      "epoch": 0.7833476149671522,
+      "grad_norm": 2.5901538355726985,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 32910
+    },
+    {
+      "epoch": 0.7833714176901837,
+      "grad_norm": 20.699659828924148,
+      "learning_rate": 2e-05,
+      "loss": 0.561,
+      "step": 32911
+    },
+    {
+      "epoch": 0.7833952204132153,
+      "grad_norm": 6.059718458861122,
+      "learning_rate": 2e-05,
+      "loss": 1.3474,
+      "step": 32912
+    },
+    {
+      "epoch": 0.7834190231362468,
+      "grad_norm": 4.694834252030569,
+      "learning_rate": 2e-05,
+      "loss": 0.5995,
+      "step": 32913
+    },
+    {
+      "epoch": 0.7834428258592783,
+      "grad_norm": 6.077716389689319,
+      "learning_rate": 2e-05,
+      "loss": 1.2351,
+      "step": 32914
+    },
+    {
+      "epoch": 0.7834666285823099,
+      "grad_norm": 2.2934365075801897,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 32915
+    },
+    {
+      "epoch": 0.7834904313053414,
+      "grad_norm": 0.22571832923827634,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 32916
+    },
+    {
+      "epoch": 0.7835142340283728,
+      "grad_norm": 1.5613669271476993,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 32917
+    },
+    {
+      "epoch": 0.7835380367514043,
+      "grad_norm": 7.148792117423865,
+      "learning_rate": 2e-05,
+      "loss": 1.1353,
+      "step": 32918
+    },
+    {
+      "epoch": 0.7835618394744359,
+      "grad_norm": 8.387249907765039,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 32919
+    },
+    {
+      "epoch": 0.7835856421974674,
+      "grad_norm": 0.5883311252534056,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 32920
+    },
+    {
+      "epoch": 0.7836094449204989,
+      "grad_norm": 7.108102819219243,
+      "learning_rate": 2e-05,
+      "loss": 0.138,
+      "step": 32921
+    },
+    {
+      "epoch": 0.7836332476435304,
+      "grad_norm": 0.041286064501722994,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 32922
+    },
+    {
+      "epoch": 0.783657050366562,
+      "grad_norm": 1.5293753303289572,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 32923
+    },
+    {
+      "epoch": 0.7836808530895935,
+      "grad_norm": 8.652391455431669,
+      "learning_rate": 2e-05,
+      "loss": 0.6279,
+      "step": 32924
+    },
+    {
+      "epoch": 0.7837046558126249,
+      "grad_norm": 3.5016866442508126,
+      "learning_rate": 2e-05,
+      "loss": 0.4846,
+      "step": 32925
+    },
+    {
+      "epoch": 0.7837284585356564,
+      "grad_norm": 4.567824227939097,
+      "learning_rate": 2e-05,
+      "loss": 0.225,
+      "step": 32926
+    },
+    {
+      "epoch": 0.783752261258688,
+      "grad_norm": 0.06717080621246412,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 32927
+    },
+    {
+      "epoch": 0.7837760639817195,
+      "grad_norm": 4.12735244407029,
+      "learning_rate": 2e-05,
+      "loss": 0.5939,
+      "step": 32928
+    },
+    {
+      "epoch": 0.783799866704751,
+      "grad_norm": 9.288429245789958,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 32929
+    },
+    {
+      "epoch": 0.7838236694277826,
+      "grad_norm": 2.7949558610183933,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 32930
+    },
+    {
+      "epoch": 0.7838474721508141,
+      "grad_norm": 0.03489564196305769,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 32931
+    },
+    {
+      "epoch": 0.7838712748738456,
+      "grad_norm": 1.3576467104527552,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 32932
+    },
+    {
+      "epoch": 0.783895077596877,
+      "grad_norm": 3.891020843435756,
+      "learning_rate": 2e-05,
+      "loss": 0.5927,
+      "step": 32933
+    },
+    {
+      "epoch": 0.7839188803199086,
+      "grad_norm": 8.13462352597533,
+      "learning_rate": 2e-05,
+      "loss": 1.3765,
+      "step": 32934
+    },
+    {
+      "epoch": 0.7839426830429401,
+      "grad_norm": 4.048418579897625,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 32935
+    },
+    {
+      "epoch": 0.7839664857659716,
+      "grad_norm": 2.470015706712294,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 32936
+    },
+    {
+      "epoch": 0.7839902884890031,
+      "grad_norm": 2.301385212762101,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 32937
+    },
+    {
+      "epoch": 0.7840140912120347,
+      "grad_norm": 1.5154283953650576,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 32938
+    },
+    {
+      "epoch": 0.7840378939350662,
+      "grad_norm": 0.7983356514214396,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 32939
+    },
+    {
+      "epoch": 0.7840616966580977,
+      "grad_norm": 5.921978222444276,
+      "learning_rate": 2e-05,
+      "loss": 0.6146,
+      "step": 32940
+    },
+    {
+      "epoch": 0.7840854993811291,
+      "grad_norm": 6.677369257708047,
+      "learning_rate": 2e-05,
+      "loss": 0.8375,
+      "step": 32941
+    },
+    {
+      "epoch": 0.7841093021041607,
+      "grad_norm": 4.941619473291003,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 32942
+    },
+    {
+      "epoch": 0.7841331048271922,
+      "grad_norm": 2.8071631339722933,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 32943
+    },
+    {
+      "epoch": 0.7841569075502237,
+      "grad_norm": 1.4401222921394947,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 32944
+    },
+    {
+      "epoch": 0.7841807102732553,
+      "grad_norm": 0.13277053728727092,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 32945
+    },
+    {
+      "epoch": 0.7842045129962868,
+      "grad_norm": 4.716154464143973,
+      "learning_rate": 2e-05,
+      "loss": 0.8519,
+      "step": 32946
+    },
+    {
+      "epoch": 0.7842283157193183,
+      "grad_norm": 20.0721984809136,
+      "learning_rate": 2e-05,
+      "loss": 1.6273,
+      "step": 32947
+    },
+    {
+      "epoch": 0.7842521184423498,
+      "grad_norm": 2.638719757303521,
+      "learning_rate": 2e-05,
+      "loss": 0.2478,
+      "step": 32948
+    },
+    {
+      "epoch": 0.7842759211653814,
+      "grad_norm": 0.7843116464528337,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 32949
+    },
+    {
+      "epoch": 0.7842997238884128,
+      "grad_norm": 4.809454346449809,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 32950
+    },
+    {
+      "epoch": 0.7843235266114443,
+      "grad_norm": 2.7354733217134357,
+      "learning_rate": 2e-05,
+      "loss": 0.2123,
+      "step": 32951
+    },
+    {
+      "epoch": 0.7843473293344758,
+      "grad_norm": 10.717787390930694,
+      "learning_rate": 2e-05,
+      "loss": 0.8685,
+      "step": 32952
+    },
+    {
+      "epoch": 0.7843711320575074,
+      "grad_norm": 1.1278674509689683,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 32953
+    },
+    {
+      "epoch": 0.7843949347805389,
+      "grad_norm": 6.743705821334919,
+      "learning_rate": 2e-05,
+      "loss": 0.7437,
+      "step": 32954
+    },
+    {
+      "epoch": 0.7844187375035704,
+      "grad_norm": 4.33895343932066,
+      "learning_rate": 2e-05,
+      "loss": 0.4477,
+      "step": 32955
+    },
+    {
+      "epoch": 0.7844425402266019,
+      "grad_norm": 4.231499798359738,
+      "learning_rate": 2e-05,
+      "loss": 0.8301,
+      "step": 32956
+    },
+    {
+      "epoch": 0.7844663429496335,
+      "grad_norm": 2.359670336735519,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 32957
+    },
+    {
+      "epoch": 0.784490145672665,
+      "grad_norm": 1.3866918020126064,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 32958
+    },
+    {
+      "epoch": 0.7845139483956964,
+      "grad_norm": 4.018014250144412,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 32959
+    },
+    {
+      "epoch": 0.784537751118728,
+      "grad_norm": 3.5203867081126012,
+      "learning_rate": 2e-05,
+      "loss": 0.6094,
+      "step": 32960
+    },
+    {
+      "epoch": 0.7845615538417595,
+      "grad_norm": 3.331714648655434,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 32961
+    },
+    {
+      "epoch": 0.784585356564791,
+      "grad_norm": 1.4625785988689535,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 32962
+    },
+    {
+      "epoch": 0.7846091592878225,
+      "grad_norm": 0.12592266748001696,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 32963
+    },
+    {
+      "epoch": 0.7846329620108541,
+      "grad_norm": 4.35324737853754,
+      "learning_rate": 2e-05,
+      "loss": 0.6726,
+      "step": 32964
+    },
+    {
+      "epoch": 0.7846567647338856,
+      "grad_norm": 2.4325059110376195,
+      "learning_rate": 2e-05,
+      "loss": 0.4097,
+      "step": 32965
+    },
+    {
+      "epoch": 0.784680567456917,
+      "grad_norm": 4.300992221041113,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 32966
+    },
+    {
+      "epoch": 0.7847043701799485,
+      "grad_norm": 0.9981077119060123,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 32967
+    },
+    {
+      "epoch": 0.7847281729029801,
+      "grad_norm": 2.9102058235833104,
+      "learning_rate": 2e-05,
+      "loss": 0.4999,
+      "step": 32968
+    },
+    {
+      "epoch": 0.7847519756260116,
+      "grad_norm": 0.9153980780714284,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 32969
+    },
+    {
+      "epoch": 0.7847757783490431,
+      "grad_norm": 1.8681455382724126,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 32970
+    },
+    {
+      "epoch": 0.7847995810720746,
+      "grad_norm": 0.8722471173457076,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 32971
+    },
+    {
+      "epoch": 0.7848233837951062,
+      "grad_norm": 17.184184324050236,
+      "learning_rate": 2e-05,
+      "loss": 0.3509,
+      "step": 32972
+    },
+    {
+      "epoch": 0.7848471865181377,
+      "grad_norm": 6.84004728181305,
+      "learning_rate": 2e-05,
+      "loss": 0.7475,
+      "step": 32973
+    },
+    {
+      "epoch": 0.7848709892411692,
+      "grad_norm": 3.4046573740834747,
+      "learning_rate": 2e-05,
+      "loss": 0.401,
+      "step": 32974
+    },
+    {
+      "epoch": 0.7848947919642008,
+      "grad_norm": 2.78877305009379,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 32975
+    },
+    {
+      "epoch": 0.7849185946872322,
+      "grad_norm": 0.1373163323502111,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 32976
+    },
+    {
+      "epoch": 0.7849423974102637,
+      "grad_norm": 18.386101174299906,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 32977
+    },
+    {
+      "epoch": 0.7849662001332952,
+      "grad_norm": 15.937897605392555,
+      "learning_rate": 2e-05,
+      "loss": 0.7802,
+      "step": 32978
+    },
+    {
+      "epoch": 0.7849900028563268,
+      "grad_norm": 1.5783236701340921,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 32979
+    },
+    {
+      "epoch": 0.7850138055793583,
+      "grad_norm": 4.408695916682616,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 32980
+    },
+    {
+      "epoch": 0.7850376083023898,
+      "grad_norm": 3.161102589211186,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 32981
+    },
+    {
+      "epoch": 0.7850614110254213,
+      "grad_norm": 0.5849745184279139,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 32982
+    },
+    {
+      "epoch": 0.7850852137484529,
+      "grad_norm": 0.9278758871062947,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 32983
+    },
+    {
+      "epoch": 0.7851090164714843,
+      "grad_norm": 20.348226639395975,
+      "learning_rate": 2e-05,
+      "loss": 1.0633,
+      "step": 32984
+    },
+    {
+      "epoch": 0.7851328191945158,
+      "grad_norm": 14.920684690503018,
+      "learning_rate": 2e-05,
+      "loss": 1.3947,
+      "step": 32985
+    },
+    {
+      "epoch": 0.7851566219175473,
+      "grad_norm": 1.6438291867434731,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 32986
+    },
+    {
+      "epoch": 0.7851804246405789,
+      "grad_norm": 7.436132374397093,
+      "learning_rate": 2e-05,
+      "loss": 0.4074,
+      "step": 32987
+    },
+    {
+      "epoch": 0.7852042273636104,
+      "grad_norm": 5.120732500064366,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 32988
+    },
+    {
+      "epoch": 0.7852280300866419,
+      "grad_norm": 28.81965259856366,
+      "learning_rate": 2e-05,
+      "loss": 0.8826,
+      "step": 32989
+    },
+    {
+      "epoch": 0.7852518328096735,
+      "grad_norm": 0.7474207689587961,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 32990
+    },
+    {
+      "epoch": 0.785275635532705,
+      "grad_norm": 13.230274036882134,
+      "learning_rate": 2e-05,
+      "loss": 0.7703,
+      "step": 32991
+    },
+    {
+      "epoch": 0.7852994382557364,
+      "grad_norm": 10.035122202769978,
+      "learning_rate": 2e-05,
+      "loss": 0.7582,
+      "step": 32992
+    },
+    {
+      "epoch": 0.7853232409787679,
+      "grad_norm": 5.224361367781613,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 32993
+    },
+    {
+      "epoch": 0.7853470437017995,
+      "grad_norm": 3.4066248695543333,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 32994
+    },
+    {
+      "epoch": 0.785370846424831,
+      "grad_norm": 1.5367471082384763,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 32995
+    },
+    {
+      "epoch": 0.7853946491478625,
+      "grad_norm": 4.247884704391141,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 32996
+    },
+    {
+      "epoch": 0.785418451870894,
+      "grad_norm": 4.7216596388484975,
+      "learning_rate": 2e-05,
+      "loss": 1.0214,
+      "step": 32997
+    },
+    {
+      "epoch": 0.7854422545939256,
+      "grad_norm": 4.691205060824019,
+      "learning_rate": 2e-05,
+      "loss": 1.2451,
+      "step": 32998
+    },
+    {
+      "epoch": 0.7854660573169571,
+      "grad_norm": 1.1921366630768475,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 32999
+    },
+    {
+      "epoch": 0.7854898600399886,
+      "grad_norm": 7.035634238867774,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 33000
+    },
+    {
+      "epoch": 0.78551366276302,
+      "grad_norm": 9.023815539345588,
+      "learning_rate": 2e-05,
+      "loss": 0.8873,
+      "step": 33001
+    },
+    {
+      "epoch": 0.7855374654860516,
+      "grad_norm": 1.1526946020395734,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 33002
+    },
+    {
+      "epoch": 0.7855612682090831,
+      "grad_norm": 3.3016295073632267,
+      "learning_rate": 2e-05,
+      "loss": 0.3173,
+      "step": 33003
+    },
+    {
+      "epoch": 0.7855850709321146,
+      "grad_norm": 0.054520047425044224,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33004
+    },
+    {
+      "epoch": 0.7856088736551462,
+      "grad_norm": 0.07829496474258474,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 33005
+    },
+    {
+      "epoch": 0.7856326763781777,
+      "grad_norm": 4.494683723794975,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 33006
+    },
+    {
+      "epoch": 0.7856564791012092,
+      "grad_norm": 3.5237645512452587,
+      "learning_rate": 2e-05,
+      "loss": 0.2451,
+      "step": 33007
+    },
+    {
+      "epoch": 0.7856802818242407,
+      "grad_norm": 1.2677786327884977,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 33008
+    },
+    {
+      "epoch": 0.7857040845472723,
+      "grad_norm": 5.035145788494391,
+      "learning_rate": 2e-05,
+      "loss": 0.9735,
+      "step": 33009
+    },
+    {
+      "epoch": 0.7857278872703037,
+      "grad_norm": 1.7871580599483607,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 33010
+    },
+    {
+      "epoch": 0.7857516899933352,
+      "grad_norm": 6.479070157252859,
+      "learning_rate": 2e-05,
+      "loss": 1.0572,
+      "step": 33011
+    },
+    {
+      "epoch": 0.7857754927163667,
+      "grad_norm": 67.45886887099417,
+      "learning_rate": 2e-05,
+      "loss": 1.354,
+      "step": 33012
+    },
+    {
+      "epoch": 0.7857992954393983,
+      "grad_norm": 5.6204905208729725,
+      "learning_rate": 2e-05,
+      "loss": 0.8016,
+      "step": 33013
+    },
+    {
+      "epoch": 0.7858230981624298,
+      "grad_norm": 3.789569453725195,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 33014
+    },
+    {
+      "epoch": 0.7858469008854613,
+      "grad_norm": 0.18252958687907003,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 33015
+    },
+    {
+      "epoch": 0.7858707036084928,
+      "grad_norm": 0.019847351614167914,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 33016
+    },
+    {
+      "epoch": 0.7858945063315244,
+      "grad_norm": 4.740047280762857,
+      "learning_rate": 2e-05,
+      "loss": 0.4495,
+      "step": 33017
+    },
+    {
+      "epoch": 0.7859183090545558,
+      "grad_norm": 2.8901023201456195,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 33018
+    },
+    {
+      "epoch": 0.7859421117775873,
+      "grad_norm": 8.882318532313391,
+      "learning_rate": 2e-05,
+      "loss": 1.0794,
+      "step": 33019
+    },
+    {
+      "epoch": 0.7859659145006189,
+      "grad_norm": 0.2686959297058429,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 33020
+    },
+    {
+      "epoch": 0.7859897172236504,
+      "grad_norm": 7.588651836406866,
+      "learning_rate": 2e-05,
+      "loss": 1.6666,
+      "step": 33021
+    },
+    {
+      "epoch": 0.7860135199466819,
+      "grad_norm": 6.125284707386551,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 33022
+    },
+    {
+      "epoch": 0.7860373226697134,
+      "grad_norm": 0.25185138531982937,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33023
+    },
+    {
+      "epoch": 0.786061125392745,
+      "grad_norm": 1.9356174521741745,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 33024
+    },
+    {
+      "epoch": 0.7860849281157765,
+      "grad_norm": 1.1350557831888572,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 33025
+    },
+    {
+      "epoch": 0.786108730838808,
+      "grad_norm": 0.7863746550448895,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 33026
+    },
+    {
+      "epoch": 0.7861325335618394,
+      "grad_norm": 9.081577683792485,
+      "learning_rate": 2e-05,
+      "loss": 0.3724,
+      "step": 33027
+    },
+    {
+      "epoch": 0.786156336284871,
+      "grad_norm": 17.670788155565234,
+      "learning_rate": 2e-05,
+      "loss": 1.4096,
+      "step": 33028
+    },
+    {
+      "epoch": 0.7861801390079025,
+      "grad_norm": 2.3026530290888916,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 33029
+    },
+    {
+      "epoch": 0.786203941730934,
+      "grad_norm": 35.632921729911985,
+      "learning_rate": 2e-05,
+      "loss": 1.1314,
+      "step": 33030
+    },
+    {
+      "epoch": 0.7862277444539655,
+      "grad_norm": 0.12011048146372115,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 33031
+    },
+    {
+      "epoch": 0.7862515471769971,
+      "grad_norm": 7.12269006633803,
+      "learning_rate": 2e-05,
+      "loss": 0.6996,
+      "step": 33032
+    },
+    {
+      "epoch": 0.7862753499000286,
+      "grad_norm": 4.023624266345304,
+      "learning_rate": 2e-05,
+      "loss": 0.5828,
+      "step": 33033
+    },
+    {
+      "epoch": 0.78629915262306,
+      "grad_norm": 11.482801083602626,
+      "learning_rate": 2e-05,
+      "loss": 0.4182,
+      "step": 33034
+    },
+    {
+      "epoch": 0.7863229553460916,
+      "grad_norm": 4.147477441571501,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 33035
+    },
+    {
+      "epoch": 0.7863467580691231,
+      "grad_norm": 3.322259414094933,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 33036
+    },
+    {
+      "epoch": 0.7863705607921546,
+      "grad_norm": 0.5985550295790976,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 33037
+    },
+    {
+      "epoch": 0.7863943635151861,
+      "grad_norm": 3.599461183525242,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 33038
+    },
+    {
+      "epoch": 0.7864181662382177,
+      "grad_norm": 2.326666671634052,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 33039
+    },
+    {
+      "epoch": 0.7864419689612492,
+      "grad_norm": 1.6997135152806595,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 33040
+    },
+    {
+      "epoch": 0.7864657716842807,
+      "grad_norm": 1.977556913289814,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 33041
+    },
+    {
+      "epoch": 0.7864895744073122,
+      "grad_norm": 4.494523668772376,
+      "learning_rate": 2e-05,
+      "loss": 0.451,
+      "step": 33042
+    },
+    {
+      "epoch": 0.7865133771303437,
+      "grad_norm": 3.792903136072054,
+      "learning_rate": 2e-05,
+      "loss": 0.5299,
+      "step": 33043
+    },
+    {
+      "epoch": 0.7865371798533752,
+      "grad_norm": 4.0432653331631405,
+      "learning_rate": 2e-05,
+      "loss": 0.4941,
+      "step": 33044
+    },
+    {
+      "epoch": 0.7865609825764067,
+      "grad_norm": 1.088605394686157,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 33045
+    },
+    {
+      "epoch": 0.7865847852994382,
+      "grad_norm": 0.5598883217931944,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 33046
+    },
+    {
+      "epoch": 0.7866085880224698,
+      "grad_norm": 1.742791289629462,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 33047
+    },
+    {
+      "epoch": 0.7866323907455013,
+      "grad_norm": 1.6276695517800364,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 33048
+    },
+    {
+      "epoch": 0.7866561934685328,
+      "grad_norm": 3.657255555436978,
+      "learning_rate": 2e-05,
+      "loss": 0.6319,
+      "step": 33049
+    },
+    {
+      "epoch": 0.7866799961915644,
+      "grad_norm": 7.651739076261482,
+      "learning_rate": 2e-05,
+      "loss": 0.4346,
+      "step": 33050
+    },
+    {
+      "epoch": 0.7867037989145959,
+      "grad_norm": 6.047547153761825,
+      "learning_rate": 2e-05,
+      "loss": 0.7081,
+      "step": 33051
+    },
+    {
+      "epoch": 0.7867276016376273,
+      "grad_norm": 4.487431062068576,
+      "learning_rate": 2e-05,
+      "loss": 1.2237,
+      "step": 33052
+    },
+    {
+      "epoch": 0.7867514043606588,
+      "grad_norm": 0.3336889108315315,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 33053
+    },
+    {
+      "epoch": 0.7867752070836904,
+      "grad_norm": 8.715888851456778,
+      "learning_rate": 2e-05,
+      "loss": 0.7481,
+      "step": 33054
+    },
+    {
+      "epoch": 0.7867990098067219,
+      "grad_norm": 9.555821370486136,
+      "learning_rate": 2e-05,
+      "loss": 0.3221,
+      "step": 33055
+    },
+    {
+      "epoch": 0.7868228125297534,
+      "grad_norm": 5.871767185357406,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 33056
+    },
+    {
+      "epoch": 0.7868466152527849,
+      "grad_norm": 4.554483928895815,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 33057
+    },
+    {
+      "epoch": 0.7868704179758165,
+      "grad_norm": 5.145235499089803,
+      "learning_rate": 2e-05,
+      "loss": 1.1519,
+      "step": 33058
+    },
+    {
+      "epoch": 0.786894220698848,
+      "grad_norm": 3.2551692908543606,
+      "learning_rate": 2e-05,
+      "loss": 0.4262,
+      "step": 33059
+    },
+    {
+      "epoch": 0.7869180234218794,
+      "grad_norm": 2.1952916509075515,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 33060
+    },
+    {
+      "epoch": 0.7869418261449109,
+      "grad_norm": 0.12527930616686803,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33061
+    },
+    {
+      "epoch": 0.7869656288679425,
+      "grad_norm": 9.002958175866143,
+      "learning_rate": 2e-05,
+      "loss": 1.0947,
+      "step": 33062
+    },
+    {
+      "epoch": 0.786989431590974,
+      "grad_norm": 1.354444698322153,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 33063
+    },
+    {
+      "epoch": 0.7870132343140055,
+      "grad_norm": 7.74031994311757,
+      "learning_rate": 2e-05,
+      "loss": 1.0539,
+      "step": 33064
+    },
+    {
+      "epoch": 0.7870370370370371,
+      "grad_norm": 2.328338321210727,
+      "learning_rate": 2e-05,
+      "loss": 0.3684,
+      "step": 33065
+    },
+    {
+      "epoch": 0.7870608397600686,
+      "grad_norm": 4.632057534550285,
+      "learning_rate": 2e-05,
+      "loss": 1.0716,
+      "step": 33066
+    },
+    {
+      "epoch": 0.7870846424831001,
+      "grad_norm": 4.683337905325368,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 33067
+    },
+    {
+      "epoch": 0.7871084452061315,
+      "grad_norm": 5.674294262478387,
+      "learning_rate": 2e-05,
+      "loss": 2.0732,
+      "step": 33068
+    },
+    {
+      "epoch": 0.7871322479291631,
+      "grad_norm": 2.819438481010453,
+      "learning_rate": 2e-05,
+      "loss": 0.5519,
+      "step": 33069
+    },
+    {
+      "epoch": 0.7871560506521946,
+      "grad_norm": 2.488115605581004,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 33070
+    },
+    {
+      "epoch": 0.7871798533752261,
+      "grad_norm": 7.946655945918814,
+      "learning_rate": 2e-05,
+      "loss": 0.5222,
+      "step": 33071
+    },
+    {
+      "epoch": 0.7872036560982576,
+      "grad_norm": 3.6359265800198215,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 33072
+    },
+    {
+      "epoch": 0.7872274588212892,
+      "grad_norm": 1.9918756942062565,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 33073
+    },
+    {
+      "epoch": 0.7872512615443207,
+      "grad_norm": 0.12475583582743992,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 33074
+    },
+    {
+      "epoch": 0.7872750642673522,
+      "grad_norm": 1.3026023526986317,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 33075
+    },
+    {
+      "epoch": 0.7872988669903837,
+      "grad_norm": 5.736288747061847,
+      "learning_rate": 2e-05,
+      "loss": 0.5152,
+      "step": 33076
+    },
+    {
+      "epoch": 0.7873226697134152,
+      "grad_norm": 2.8683491142152806,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 33077
+    },
+    {
+      "epoch": 0.7873464724364467,
+      "grad_norm": 3.3094236308083995,
+      "learning_rate": 2e-05,
+      "loss": 0.2396,
+      "step": 33078
+    },
+    {
+      "epoch": 0.7873702751594782,
+      "grad_norm": 0.11410602961502955,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 33079
+    },
+    {
+      "epoch": 0.7873940778825098,
+      "grad_norm": 2.832872620768186,
+      "learning_rate": 2e-05,
+      "loss": 0.3101,
+      "step": 33080
+    },
+    {
+      "epoch": 0.7874178806055413,
+      "grad_norm": 0.8950088072954124,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 33081
+    },
+    {
+      "epoch": 0.7874416833285728,
+      "grad_norm": 10.216798479131581,
+      "learning_rate": 2e-05,
+      "loss": 0.3394,
+      "step": 33082
+    },
+    {
+      "epoch": 0.7874654860516043,
+      "grad_norm": 5.097291830770551,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 33083
+    },
+    {
+      "epoch": 0.7874892887746359,
+      "grad_norm": 1.116407929205563,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 33084
+    },
+    {
+      "epoch": 0.7875130914976674,
+      "grad_norm": 2.3960338899590474,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 33085
+    },
+    {
+      "epoch": 0.7875368942206988,
+      "grad_norm": 6.777229070557038,
+      "learning_rate": 2e-05,
+      "loss": 0.8492,
+      "step": 33086
+    },
+    {
+      "epoch": 0.7875606969437303,
+      "grad_norm": 0.0720758452329567,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 33087
+    },
+    {
+      "epoch": 0.7875844996667619,
+      "grad_norm": 4.278699457746714,
+      "learning_rate": 2e-05,
+      "loss": 0.1103,
+      "step": 33088
+    },
+    {
+      "epoch": 0.7876083023897934,
+      "grad_norm": 0.1826253346476239,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 33089
+    },
+    {
+      "epoch": 0.7876321051128249,
+      "grad_norm": 0.9240548149551535,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 33090
+    },
+    {
+      "epoch": 0.7876559078358564,
+      "grad_norm": 2.7999672571032823,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 33091
+    },
+    {
+      "epoch": 0.787679710558888,
+      "grad_norm": 2.8423520883911313,
+      "learning_rate": 2e-05,
+      "loss": 0.0499,
+      "step": 33092
+    },
+    {
+      "epoch": 0.7877035132819195,
+      "grad_norm": 10.638153820589817,
+      "learning_rate": 2e-05,
+      "loss": 1.1026,
+      "step": 33093
+    },
+    {
+      "epoch": 0.7877273160049509,
+      "grad_norm": 0.8088042961539043,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 33094
+    },
+    {
+      "epoch": 0.7877511187279825,
+      "grad_norm": 4.23336850543424,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 33095
+    },
+    {
+      "epoch": 0.787774921451014,
+      "grad_norm": 15.304702902772126,
+      "learning_rate": 2e-05,
+      "loss": 1.1838,
+      "step": 33096
+    },
+    {
+      "epoch": 0.7877987241740455,
+      "grad_norm": 10.53186558472961,
+      "learning_rate": 2e-05,
+      "loss": 1.3144,
+      "step": 33097
+    },
+    {
+      "epoch": 0.787822526897077,
+      "grad_norm": 7.94570071899999,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 33098
+    },
+    {
+      "epoch": 0.7878463296201086,
+      "grad_norm": 0.6677132386029628,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 33099
+    },
+    {
+      "epoch": 0.7878701323431401,
+      "grad_norm": 5.680032225984237,
+      "learning_rate": 2e-05,
+      "loss": 0.6506,
+      "step": 33100
+    },
+    {
+      "epoch": 0.7878939350661716,
+      "grad_norm": 11.760099457246785,
+      "learning_rate": 2e-05,
+      "loss": 0.8259,
+      "step": 33101
+    },
+    {
+      "epoch": 0.787917737789203,
+      "grad_norm": 2.1310218938527883,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 33102
+    },
+    {
+      "epoch": 0.7879415405122346,
+      "grad_norm": 3.296269239084229,
+      "learning_rate": 2e-05,
+      "loss": 0.7082,
+      "step": 33103
+    },
+    {
+      "epoch": 0.7879653432352661,
+      "grad_norm": 5.019010982149906,
+      "learning_rate": 2e-05,
+      "loss": 0.7761,
+      "step": 33104
+    },
+    {
+      "epoch": 0.7879891459582976,
+      "grad_norm": 0.183563464930137,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 33105
+    },
+    {
+      "epoch": 0.7880129486813291,
+      "grad_norm": 1.9401693967027582,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 33106
+    },
+    {
+      "epoch": 0.7880367514043607,
+      "grad_norm": 0.9933006045572439,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 33107
+    },
+    {
+      "epoch": 0.7880605541273922,
+      "grad_norm": 1.1879541358041106,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 33108
+    },
+    {
+      "epoch": 0.7880843568504237,
+      "grad_norm": 2.0171398582485,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 33109
+    },
+    {
+      "epoch": 0.7881081595734553,
+      "grad_norm": 3.357070400505557,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 33110
+    },
+    {
+      "epoch": 0.7881319622964867,
+      "grad_norm": 1.2878530318849934,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 33111
+    },
+    {
+      "epoch": 0.7881557650195182,
+      "grad_norm": 7.945517743914725,
+      "learning_rate": 2e-05,
+      "loss": 0.2881,
+      "step": 33112
+    },
+    {
+      "epoch": 0.7881795677425497,
+      "grad_norm": 4.487935181174178,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 33113
+    },
+    {
+      "epoch": 0.7882033704655813,
+      "grad_norm": 0.1081377829193013,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 33114
+    },
+    {
+      "epoch": 0.7882271731886128,
+      "grad_norm": 3.3354101730233765,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 33115
+    },
+    {
+      "epoch": 0.7882509759116443,
+      "grad_norm": 2.116329051284164,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 33116
+    },
+    {
+      "epoch": 0.7882747786346758,
+      "grad_norm": 3.748886581390237,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 33117
+    },
+    {
+      "epoch": 0.7882985813577074,
+      "grad_norm": 12.16250579491548,
+      "learning_rate": 2e-05,
+      "loss": 0.3483,
+      "step": 33118
+    },
+    {
+      "epoch": 0.7883223840807388,
+      "grad_norm": 2.463778418436466,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 33119
+    },
+    {
+      "epoch": 0.7883461868037703,
+      "grad_norm": 3.6138360823464084,
+      "learning_rate": 2e-05,
+      "loss": 0.5063,
+      "step": 33120
+    },
+    {
+      "epoch": 0.7883699895268018,
+      "grad_norm": 2.0001985787231447,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 33121
+    },
+    {
+      "epoch": 0.7883937922498334,
+      "grad_norm": 3.3131480950511207,
+      "learning_rate": 2e-05,
+      "loss": 0.7568,
+      "step": 33122
+    },
+    {
+      "epoch": 0.7884175949728649,
+      "grad_norm": 0.3022450870297344,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 33123
+    },
+    {
+      "epoch": 0.7884413976958964,
+      "grad_norm": 0.2617269015696324,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 33124
+    },
+    {
+      "epoch": 0.788465200418928,
+      "grad_norm": 1.0514774889682426,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 33125
+    },
+    {
+      "epoch": 0.7884890031419595,
+      "grad_norm": 4.481835214313369,
+      "learning_rate": 2e-05,
+      "loss": 0.4467,
+      "step": 33126
+    },
+    {
+      "epoch": 0.788512805864991,
+      "grad_norm": 2.6721661788061173,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 33127
+    },
+    {
+      "epoch": 0.7885366085880224,
+      "grad_norm": 0.37753147726140773,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 33128
+    },
+    {
+      "epoch": 0.788560411311054,
+      "grad_norm": 5.65942177773885,
+      "learning_rate": 2e-05,
+      "loss": 0.1883,
+      "step": 33129
+    },
+    {
+      "epoch": 0.7885842140340855,
+      "grad_norm": 4.748847064909019,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 33130
+    },
+    {
+      "epoch": 0.788608016757117,
+      "grad_norm": 3.022545252436005,
+      "learning_rate": 2e-05,
+      "loss": 0.2597,
+      "step": 33131
+    },
+    {
+      "epoch": 0.7886318194801485,
+      "grad_norm": 0.1220166048211619,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 33132
+    },
+    {
+      "epoch": 0.7886556222031801,
+      "grad_norm": 6.119234732239799,
+      "learning_rate": 2e-05,
+      "loss": 0.662,
+      "step": 33133
+    },
+    {
+      "epoch": 0.7886794249262116,
+      "grad_norm": 5.879995471853164,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 33134
+    },
+    {
+      "epoch": 0.788703227649243,
+      "grad_norm": 4.731844517628081,
+      "learning_rate": 2e-05,
+      "loss": 0.7377,
+      "step": 33135
+    },
+    {
+      "epoch": 0.7887270303722745,
+      "grad_norm": 2.9220600297408934,
+      "learning_rate": 2e-05,
+      "loss": 0.0601,
+      "step": 33136
+    },
+    {
+      "epoch": 0.7887508330953061,
+      "grad_norm": 7.282651854979697,
+      "learning_rate": 2e-05,
+      "loss": 0.2326,
+      "step": 33137
+    },
+    {
+      "epoch": 0.7887746358183376,
+      "grad_norm": 7.357899824968779,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 33138
+    },
+    {
+      "epoch": 0.7887984385413691,
+      "grad_norm": 4.4450744082941664,
+      "learning_rate": 2e-05,
+      "loss": 1.0693,
+      "step": 33139
+    },
+    {
+      "epoch": 0.7888222412644006,
+      "grad_norm": 3.080020720478975,
+      "learning_rate": 2e-05,
+      "loss": 0.5785,
+      "step": 33140
+    },
+    {
+      "epoch": 0.7888460439874322,
+      "grad_norm": 0.21813942956017512,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 33141
+    },
+    {
+      "epoch": 0.7888698467104637,
+      "grad_norm": 1.804481972182416,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 33142
+    },
+    {
+      "epoch": 0.7888936494334952,
+      "grad_norm": 4.104089065046675,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 33143
+    },
+    {
+      "epoch": 0.7889174521565268,
+      "grad_norm": 1.3672547772522254,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 33144
+    },
+    {
+      "epoch": 0.7889412548795582,
+      "grad_norm": 10.941675045592888,
+      "learning_rate": 2e-05,
+      "loss": 0.2308,
+      "step": 33145
+    },
+    {
+      "epoch": 0.7889650576025897,
+      "grad_norm": 7.161933559608123,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 33146
+    },
+    {
+      "epoch": 0.7889888603256212,
+      "grad_norm": 5.878564706696289,
+      "learning_rate": 2e-05,
+      "loss": 0.7955,
+      "step": 33147
+    },
+    {
+      "epoch": 0.7890126630486528,
+      "grad_norm": 5.65113833366511,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 33148
+    },
+    {
+      "epoch": 0.7890364657716843,
+      "grad_norm": 0.3509532784393595,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 33149
+    },
+    {
+      "epoch": 0.7890602684947158,
+      "grad_norm": 2.2623218187384593,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 33150
+    },
+    {
+      "epoch": 0.7890840712177473,
+      "grad_norm": 11.39620198513493,
+      "learning_rate": 2e-05,
+      "loss": 0.7961,
+      "step": 33151
+    },
+    {
+      "epoch": 0.7891078739407789,
+      "grad_norm": 3.8720952419555297,
+      "learning_rate": 2e-05,
+      "loss": 0.0762,
+      "step": 33152
+    },
+    {
+      "epoch": 0.7891316766638103,
+      "grad_norm": 2.912066849470592,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 33153
+    },
+    {
+      "epoch": 0.7891554793868418,
+      "grad_norm": 3.0795824658547177,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 33154
+    },
+    {
+      "epoch": 0.7891792821098733,
+      "grad_norm": 9.159023557139426,
+      "learning_rate": 2e-05,
+      "loss": 0.5441,
+      "step": 33155
+    },
+    {
+      "epoch": 0.7892030848329049,
+      "grad_norm": 5.9017466159540435,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 33156
+    },
+    {
+      "epoch": 0.7892268875559364,
+      "grad_norm": 2.558953737915764,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 33157
+    },
+    {
+      "epoch": 0.7892506902789679,
+      "grad_norm": 4.189507395772896,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 33158
+    },
+    {
+      "epoch": 0.7892744930019995,
+      "grad_norm": 5.46946247081912,
+      "learning_rate": 2e-05,
+      "loss": 0.7478,
+      "step": 33159
+    },
+    {
+      "epoch": 0.789298295725031,
+      "grad_norm": 0.12131428058809818,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 33160
+    },
+    {
+      "epoch": 0.7893220984480624,
+      "grad_norm": 7.258051493116355,
+      "learning_rate": 2e-05,
+      "loss": 1.4057,
+      "step": 33161
+    },
+    {
+      "epoch": 0.7893459011710939,
+      "grad_norm": 6.077070470212166,
+      "learning_rate": 2e-05,
+      "loss": 0.4926,
+      "step": 33162
+    },
+    {
+      "epoch": 0.7893697038941255,
+      "grad_norm": 3.367293707505537,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 33163
+    },
+    {
+      "epoch": 0.789393506617157,
+      "grad_norm": 0.7740643793919693,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 33164
+    },
+    {
+      "epoch": 0.7894173093401885,
+      "grad_norm": 9.073841467232304,
+      "learning_rate": 2e-05,
+      "loss": 0.3909,
+      "step": 33165
+    },
+    {
+      "epoch": 0.78944111206322,
+      "grad_norm": 4.599554556473202,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 33166
+    },
+    {
+      "epoch": 0.7894649147862516,
+      "grad_norm": 0.1104019770237863,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 33167
+    },
+    {
+      "epoch": 0.7894887175092831,
+      "grad_norm": 3.4136499644185436,
+      "learning_rate": 2e-05,
+      "loss": 0.4264,
+      "step": 33168
+    },
+    {
+      "epoch": 0.7895125202323146,
+      "grad_norm": 4.528395426036848,
+      "learning_rate": 2e-05,
+      "loss": 0.7378,
+      "step": 33169
+    },
+    {
+      "epoch": 0.789536322955346,
+      "grad_norm": 3.0008409802646634,
+      "learning_rate": 2e-05,
+      "loss": 0.2312,
+      "step": 33170
+    },
+    {
+      "epoch": 0.7895601256783776,
+      "grad_norm": 7.611868640004324,
+      "learning_rate": 2e-05,
+      "loss": 1.3162,
+      "step": 33171
+    },
+    {
+      "epoch": 0.7895839284014091,
+      "grad_norm": 3.1760843402890346,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 33172
+    },
+    {
+      "epoch": 0.7896077311244406,
+      "grad_norm": 1.7180922536821335,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 33173
+    },
+    {
+      "epoch": 0.7896315338474722,
+      "grad_norm": 0.7992111298926436,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 33174
+    },
+    {
+      "epoch": 0.7896553365705037,
+      "grad_norm": 1.0164116035411974,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 33175
+    },
+    {
+      "epoch": 0.7896791392935352,
+      "grad_norm": 3.7533661358528816,
+      "learning_rate": 2e-05,
+      "loss": 0.5636,
+      "step": 33176
+    },
+    {
+      "epoch": 0.7897029420165667,
+      "grad_norm": 6.170918110910234,
+      "learning_rate": 2e-05,
+      "loss": 0.4024,
+      "step": 33177
+    },
+    {
+      "epoch": 0.7897267447395983,
+      "grad_norm": 0.8191136028716481,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 33178
+    },
+    {
+      "epoch": 0.7897505474626297,
+      "grad_norm": 2.9063072716678002,
+      "learning_rate": 2e-05,
+      "loss": 0.3448,
+      "step": 33179
+    },
+    {
+      "epoch": 0.7897743501856612,
+      "grad_norm": 1.1456130485060738,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 33180
+    },
+    {
+      "epoch": 0.7897981529086927,
+      "grad_norm": 1.1135969897050004,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 33181
+    },
+    {
+      "epoch": 0.7898219556317243,
+      "grad_norm": 2.8900295223734775,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 33182
+    },
+    {
+      "epoch": 0.7898457583547558,
+      "grad_norm": 3.5496789907875095,
+      "learning_rate": 2e-05,
+      "loss": 0.6539,
+      "step": 33183
+    },
+    {
+      "epoch": 0.7898695610777873,
+      "grad_norm": 1.205087040499726,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 33184
+    },
+    {
+      "epoch": 0.7898933638008188,
+      "grad_norm": 2.0090595106470857,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 33185
+    },
+    {
+      "epoch": 0.7899171665238504,
+      "grad_norm": 11.117261576589687,
+      "learning_rate": 2e-05,
+      "loss": 0.4857,
+      "step": 33186
+    },
+    {
+      "epoch": 0.7899409692468818,
+      "grad_norm": 3.285836259917957,
+      "learning_rate": 2e-05,
+      "loss": 0.587,
+      "step": 33187
+    },
+    {
+      "epoch": 0.7899647719699133,
+      "grad_norm": 0.4002351150096021,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 33188
+    },
+    {
+      "epoch": 0.7899885746929449,
+      "grad_norm": 6.257908208423731,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 33189
+    },
+    {
+      "epoch": 0.7900123774159764,
+      "grad_norm": 3.7276801942294004,
+      "learning_rate": 2e-05,
+      "loss": 0.7846,
+      "step": 33190
+    },
+    {
+      "epoch": 0.7900361801390079,
+      "grad_norm": 3.457473802570869,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 33191
+    },
+    {
+      "epoch": 0.7900599828620394,
+      "grad_norm": 2.9885577585686156,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 33192
+    },
+    {
+      "epoch": 0.790083785585071,
+      "grad_norm": 7.879258057780447,
+      "learning_rate": 2e-05,
+      "loss": 0.7411,
+      "step": 33193
+    },
+    {
+      "epoch": 0.7901075883081025,
+      "grad_norm": 10.633645861768226,
+      "learning_rate": 2e-05,
+      "loss": 0.925,
+      "step": 33194
+    },
+    {
+      "epoch": 0.790131391031134,
+      "grad_norm": 3.037335786772314,
+      "learning_rate": 2e-05,
+      "loss": 0.4803,
+      "step": 33195
+    },
+    {
+      "epoch": 0.7901551937541654,
+      "grad_norm": 5.437882436272674,
+      "learning_rate": 2e-05,
+      "loss": 0.787,
+      "step": 33196
+    },
+    {
+      "epoch": 0.790178996477197,
+      "grad_norm": 5.741262197203004,
+      "learning_rate": 2e-05,
+      "loss": 0.9826,
+      "step": 33197
+    },
+    {
+      "epoch": 0.7902027992002285,
+      "grad_norm": 0.22019498053399542,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 33198
+    },
+    {
+      "epoch": 0.79022660192326,
+      "grad_norm": 2.3254422850082217,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 33199
+    },
+    {
+      "epoch": 0.7902504046462915,
+      "grad_norm": 4.036746199517349,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 33200
+    },
+    {
+      "epoch": 0.7902742073693231,
+      "grad_norm": 1.5969091358076313,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 33201
+    },
+    {
+      "epoch": 0.7902980100923546,
+      "grad_norm": 0.5204176386721953,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 33202
+    },
+    {
+      "epoch": 0.790321812815386,
+      "grad_norm": 3.0642856101735716,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 33203
+    },
+    {
+      "epoch": 0.7903456155384176,
+      "grad_norm": 2.0377292667226428,
+      "learning_rate": 2e-05,
+      "loss": 0.2121,
+      "step": 33204
+    },
+    {
+      "epoch": 0.7903694182614491,
+      "grad_norm": 0.9456589779299946,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 33205
+    },
+    {
+      "epoch": 0.7903932209844806,
+      "grad_norm": 4.350782511367575,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 33206
+    },
+    {
+      "epoch": 0.7904170237075121,
+      "grad_norm": 4.158203898475317,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 33207
+    },
+    {
+      "epoch": 0.7904408264305437,
+      "grad_norm": 8.750631522329185,
+      "learning_rate": 2e-05,
+      "loss": 1.1175,
+      "step": 33208
+    },
+    {
+      "epoch": 0.7904646291535752,
+      "grad_norm": 3.8468133560072926,
+      "learning_rate": 2e-05,
+      "loss": 0.445,
+      "step": 33209
+    },
+    {
+      "epoch": 0.7904884318766067,
+      "grad_norm": 1.7784410455865245,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 33210
+    },
+    {
+      "epoch": 0.7905122345996382,
+      "grad_norm": 0.27380882780657184,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 33211
+    },
+    {
+      "epoch": 0.7905360373226697,
+      "grad_norm": 2.787645427372441,
+      "learning_rate": 2e-05,
+      "loss": 0.3162,
+      "step": 33212
+    },
+    {
+      "epoch": 0.7905598400457012,
+      "grad_norm": 5.354319097737596,
+      "learning_rate": 2e-05,
+      "loss": 0.4239,
+      "step": 33213
+    },
+    {
+      "epoch": 0.7905836427687327,
+      "grad_norm": 4.19294114080603,
+      "learning_rate": 2e-05,
+      "loss": 1.0235,
+      "step": 33214
+    },
+    {
+      "epoch": 0.7906074454917642,
+      "grad_norm": 1.1815452858591222,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 33215
+    },
+    {
+      "epoch": 0.7906312482147958,
+      "grad_norm": 12.682819557208633,
+      "learning_rate": 2e-05,
+      "loss": 0.8845,
+      "step": 33216
+    },
+    {
+      "epoch": 0.7906550509378273,
+      "grad_norm": 5.479846193241372,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 33217
+    },
+    {
+      "epoch": 0.7906788536608588,
+      "grad_norm": 3.1514034030727367,
+      "learning_rate": 2e-05,
+      "loss": 0.7324,
+      "step": 33218
+    },
+    {
+      "epoch": 0.7907026563838904,
+      "grad_norm": 12.278875281420824,
+      "learning_rate": 2e-05,
+      "loss": 1.0089,
+      "step": 33219
+    },
+    {
+      "epoch": 0.7907264591069219,
+      "grad_norm": 6.503204270828867,
+      "learning_rate": 2e-05,
+      "loss": 0.1367,
+      "step": 33220
+    },
+    {
+      "epoch": 0.7907502618299533,
+      "grad_norm": 2.461079845681799,
+      "learning_rate": 2e-05,
+      "loss": 0.3453,
+      "step": 33221
+    },
+    {
+      "epoch": 0.7907740645529848,
+      "grad_norm": 7.713435169618244,
+      "learning_rate": 2e-05,
+      "loss": 1.1013,
+      "step": 33222
+    },
+    {
+      "epoch": 0.7907978672760164,
+      "grad_norm": 3.3290585603804517,
+      "learning_rate": 2e-05,
+      "loss": 0.3773,
+      "step": 33223
+    },
+    {
+      "epoch": 0.7908216699990479,
+      "grad_norm": 1.3143369321773866,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 33224
+    },
+    {
+      "epoch": 0.7908454727220794,
+      "grad_norm": 2.578806819740912,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 33225
+    },
+    {
+      "epoch": 0.7908692754451109,
+      "grad_norm": 4.060039978139188,
+      "learning_rate": 2e-05,
+      "loss": 0.4999,
+      "step": 33226
+    },
+    {
+      "epoch": 0.7908930781681425,
+      "grad_norm": 1.7103846957987563,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 33227
+    },
+    {
+      "epoch": 0.790916880891174,
+      "grad_norm": 3.4639450206923477,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 33228
+    },
+    {
+      "epoch": 0.7909406836142054,
+      "grad_norm": 9.678591420191154,
+      "learning_rate": 2e-05,
+      "loss": 0.3124,
+      "step": 33229
+    },
+    {
+      "epoch": 0.7909644863372369,
+      "grad_norm": 2.2808819569083667,
+      "learning_rate": 2e-05,
+      "loss": 0.2188,
+      "step": 33230
+    },
+    {
+      "epoch": 0.7909882890602685,
+      "grad_norm": 7.401240616473166,
+      "learning_rate": 2e-05,
+      "loss": 0.8666,
+      "step": 33231
+    },
+    {
+      "epoch": 0.7910120917833,
+      "grad_norm": 4.104898369005815,
+      "learning_rate": 2e-05,
+      "loss": 0.6471,
+      "step": 33232
+    },
+    {
+      "epoch": 0.7910358945063315,
+      "grad_norm": 0.7220971919711353,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 33233
+    },
+    {
+      "epoch": 0.7910596972293631,
+      "grad_norm": 4.374227071721509,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 33234
+    },
+    {
+      "epoch": 0.7910834999523946,
+      "grad_norm": 2.2110191815483966,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 33235
+    },
+    {
+      "epoch": 0.7911073026754261,
+      "grad_norm": 3.267325241667396,
+      "learning_rate": 2e-05,
+      "loss": 0.4478,
+      "step": 33236
+    },
+    {
+      "epoch": 0.7911311053984575,
+      "grad_norm": 6.486764785716501,
+      "learning_rate": 2e-05,
+      "loss": 0.3037,
+      "step": 33237
+    },
+    {
+      "epoch": 0.7911549081214891,
+      "grad_norm": 3.1697025515049924,
+      "learning_rate": 2e-05,
+      "loss": 0.5537,
+      "step": 33238
+    },
+    {
+      "epoch": 0.7911787108445206,
+      "grad_norm": 0.3101654269239799,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 33239
+    },
+    {
+      "epoch": 0.7912025135675521,
+      "grad_norm": 4.849237293590467,
+      "learning_rate": 2e-05,
+      "loss": 0.1393,
+      "step": 33240
+    },
+    {
+      "epoch": 0.7912263162905836,
+      "grad_norm": 0.7360973763954408,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 33241
+    },
+    {
+      "epoch": 0.7912501190136152,
+      "grad_norm": 0.6471961879998399,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 33242
+    },
+    {
+      "epoch": 0.7912739217366467,
+      "grad_norm": 0.17452566964953872,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 33243
+    },
+    {
+      "epoch": 0.7912977244596782,
+      "grad_norm": 0.588372210285849,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 33244
+    },
+    {
+      "epoch": 0.7913215271827096,
+      "grad_norm": 7.006475396603257,
+      "learning_rate": 2e-05,
+      "loss": 0.2392,
+      "step": 33245
+    },
+    {
+      "epoch": 0.7913453299057412,
+      "grad_norm": 4.772886642873153,
+      "learning_rate": 2e-05,
+      "loss": 0.4845,
+      "step": 33246
+    },
+    {
+      "epoch": 0.7913691326287727,
+      "grad_norm": 0.9883906083497209,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 33247
+    },
+    {
+      "epoch": 0.7913929353518042,
+      "grad_norm": 2.6264576767755257,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 33248
+    },
+    {
+      "epoch": 0.7914167380748358,
+      "grad_norm": 0.01726649382087989,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 33249
+    },
+    {
+      "epoch": 0.7914405407978673,
+      "grad_norm": 2.841238781812239,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 33250
+    },
+    {
+      "epoch": 0.7914643435208988,
+      "grad_norm": 7.3694983435245485,
+      "learning_rate": 2e-05,
+      "loss": 0.2094,
+      "step": 33251
+    },
+    {
+      "epoch": 0.7914881462439303,
+      "grad_norm": 4.733681130142287,
+      "learning_rate": 2e-05,
+      "loss": 1.0406,
+      "step": 33252
+    },
+    {
+      "epoch": 0.7915119489669619,
+      "grad_norm": 0.48615523052037235,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 33253
+    },
+    {
+      "epoch": 0.7915357516899933,
+      "grad_norm": 4.277361824465336,
+      "learning_rate": 2e-05,
+      "loss": 0.8384,
+      "step": 33254
+    },
+    {
+      "epoch": 0.7915595544130248,
+      "grad_norm": 140.6273673589305,
+      "learning_rate": 2e-05,
+      "loss": 0.8293,
+      "step": 33255
+    },
+    {
+      "epoch": 0.7915833571360563,
+      "grad_norm": 3.159548683553891,
+      "learning_rate": 2e-05,
+      "loss": 0.805,
+      "step": 33256
+    },
+    {
+      "epoch": 0.7916071598590879,
+      "grad_norm": 0.14242059119877257,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 33257
+    },
+    {
+      "epoch": 0.7916309625821194,
+      "grad_norm": 0.15860340253489374,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33258
+    },
+    {
+      "epoch": 0.7916547653051509,
+      "grad_norm": 3.0504845723062877,
+      "learning_rate": 2e-05,
+      "loss": 0.1116,
+      "step": 33259
+    },
+    {
+      "epoch": 0.7916785680281824,
+      "grad_norm": 5.6806498108956465,
+      "learning_rate": 2e-05,
+      "loss": 0.8368,
+      "step": 33260
+    },
+    {
+      "epoch": 0.791702370751214,
+      "grad_norm": 0.0673457280985315,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33261
+    },
+    {
+      "epoch": 0.7917261734742455,
+      "grad_norm": 3.242142007395702,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 33262
+    },
+    {
+      "epoch": 0.7917499761972769,
+      "grad_norm": 4.82268413276992,
+      "learning_rate": 2e-05,
+      "loss": 0.9351,
+      "step": 33263
+    },
+    {
+      "epoch": 0.7917737789203085,
+      "grad_norm": 0.1895825964421331,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 33264
+    },
+    {
+      "epoch": 0.79179758164334,
+      "grad_norm": 19.978884184797547,
+      "learning_rate": 2e-05,
+      "loss": 0.2361,
+      "step": 33265
+    },
+    {
+      "epoch": 0.7918213843663715,
+      "grad_norm": 5.926978880289406,
+      "learning_rate": 2e-05,
+      "loss": 1.0767,
+      "step": 33266
+    },
+    {
+      "epoch": 0.791845187089403,
+      "grad_norm": 4.562342469870673,
+      "learning_rate": 2e-05,
+      "loss": 0.4364,
+      "step": 33267
+    },
+    {
+      "epoch": 0.7918689898124346,
+      "grad_norm": 1.9900514782619567,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 33268
+    },
+    {
+      "epoch": 0.7918927925354661,
+      "grad_norm": 0.735390955527704,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 33269
+    },
+    {
+      "epoch": 0.7919165952584976,
+      "grad_norm": 8.492226944326543,
+      "learning_rate": 2e-05,
+      "loss": 0.5691,
+      "step": 33270
+    },
+    {
+      "epoch": 0.791940397981529,
+      "grad_norm": 3.9074574187841344,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 33271
+    },
+    {
+      "epoch": 0.7919642007045606,
+      "grad_norm": 6.8032222405141045,
+      "learning_rate": 2e-05,
+      "loss": 0.2207,
+      "step": 33272
+    },
+    {
+      "epoch": 0.7919880034275921,
+      "grad_norm": 11.929925903224511,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 33273
+    },
+    {
+      "epoch": 0.7920118061506236,
+      "grad_norm": 0.6833194286269326,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 33274
+    },
+    {
+      "epoch": 0.7920356088736551,
+      "grad_norm": 0.9507415344808675,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 33275
+    },
+    {
+      "epoch": 0.7920594115966867,
+      "grad_norm": 1.679391254210108,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 33276
+    },
+    {
+      "epoch": 0.7920832143197182,
+      "grad_norm": 4.147330673610802,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 33277
+    },
+    {
+      "epoch": 0.7921070170427497,
+      "grad_norm": 2.7497200388339067,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 33278
+    },
+    {
+      "epoch": 0.7921308197657813,
+      "grad_norm": 3.473540089548675,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 33279
+    },
+    {
+      "epoch": 0.7921546224888127,
+      "grad_norm": 3.4586567828780224,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 33280
+    },
+    {
+      "epoch": 0.7921784252118442,
+      "grad_norm": 5.577376414200137,
+      "learning_rate": 2e-05,
+      "loss": 0.9787,
+      "step": 33281
+    },
+    {
+      "epoch": 0.7922022279348757,
+      "grad_norm": 0.07737980238807378,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33282
+    },
+    {
+      "epoch": 0.7922260306579073,
+      "grad_norm": 0.6207949895182971,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 33283
+    },
+    {
+      "epoch": 0.7922498333809388,
+      "grad_norm": 9.129961510812315,
+      "learning_rate": 2e-05,
+      "loss": 0.4339,
+      "step": 33284
+    },
+    {
+      "epoch": 0.7922736361039703,
+      "grad_norm": 4.633370701299354,
+      "learning_rate": 2e-05,
+      "loss": 0.7482,
+      "step": 33285
+    },
+    {
+      "epoch": 0.7922974388270018,
+      "grad_norm": 3.329073539947262,
+      "learning_rate": 2e-05,
+      "loss": 0.05,
+      "step": 33286
+    },
+    {
+      "epoch": 0.7923212415500334,
+      "grad_norm": 1.3145401141615052,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 33287
+    },
+    {
+      "epoch": 0.7923450442730648,
+      "grad_norm": 0.1915333822082855,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 33288
+    },
+    {
+      "epoch": 0.7923688469960963,
+      "grad_norm": 0.12322111890064036,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 33289
+    },
+    {
+      "epoch": 0.7923926497191278,
+      "grad_norm": 4.124850799365743,
+      "learning_rate": 2e-05,
+      "loss": 0.6398,
+      "step": 33290
+    },
+    {
+      "epoch": 0.7924164524421594,
+      "grad_norm": 2.2805782262212344,
+      "learning_rate": 2e-05,
+      "loss": 0.345,
+      "step": 33291
+    },
+    {
+      "epoch": 0.7924402551651909,
+      "grad_norm": 1.1686116688748454,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 33292
+    },
+    {
+      "epoch": 0.7924640578882224,
+      "grad_norm": 3.3227734288986217,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 33293
+    },
+    {
+      "epoch": 0.792487860611254,
+      "grad_norm": 3.8349240277312764,
+      "learning_rate": 2e-05,
+      "loss": 0.745,
+      "step": 33294
+    },
+    {
+      "epoch": 0.7925116633342855,
+      "grad_norm": 5.412163509596389,
+      "learning_rate": 2e-05,
+      "loss": 0.8482,
+      "step": 33295
+    },
+    {
+      "epoch": 0.792535466057317,
+      "grad_norm": 1.4538401355944899,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 33296
+    },
+    {
+      "epoch": 0.7925592687803484,
+      "grad_norm": 3.2094519544051865,
+      "learning_rate": 2e-05,
+      "loss": 0.4728,
+      "step": 33297
+    },
+    {
+      "epoch": 0.79258307150338,
+      "grad_norm": 5.537275102938715,
+      "learning_rate": 2e-05,
+      "loss": 0.8669,
+      "step": 33298
+    },
+    {
+      "epoch": 0.7926068742264115,
+      "grad_norm": 1.137327504666181,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 33299
+    },
+    {
+      "epoch": 0.792630676949443,
+      "grad_norm": 0.4885295963848933,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 33300
+    },
+    {
+      "epoch": 0.7926544796724745,
+      "grad_norm": 6.776233095876934,
+      "learning_rate": 2e-05,
+      "loss": 0.4241,
+      "step": 33301
+    },
+    {
+      "epoch": 0.7926782823955061,
+      "grad_norm": 3.269831356813479,
+      "learning_rate": 2e-05,
+      "loss": 0.3229,
+      "step": 33302
+    },
+    {
+      "epoch": 0.7927020851185376,
+      "grad_norm": 7.073109475973624,
+      "learning_rate": 2e-05,
+      "loss": 0.7569,
+      "step": 33303
+    },
+    {
+      "epoch": 0.792725887841569,
+      "grad_norm": 0.8593622768676878,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 33304
+    },
+    {
+      "epoch": 0.7927496905646005,
+      "grad_norm": 3.092951246545473,
+      "learning_rate": 2e-05,
+      "loss": 0.7598,
+      "step": 33305
+    },
+    {
+      "epoch": 0.7927734932876321,
+      "grad_norm": 0.07398153823882943,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 33306
+    },
+    {
+      "epoch": 0.7927972960106636,
+      "grad_norm": 0.43479639345110666,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 33307
+    },
+    {
+      "epoch": 0.7928210987336951,
+      "grad_norm": 12.770723022960182,
+      "learning_rate": 2e-05,
+      "loss": 0.7223,
+      "step": 33308
+    },
+    {
+      "epoch": 0.7928449014567267,
+      "grad_norm": 4.521010291027836,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 33309
+    },
+    {
+      "epoch": 0.7928687041797582,
+      "grad_norm": 4.849032245945782,
+      "learning_rate": 2e-05,
+      "loss": 0.6258,
+      "step": 33310
+    },
+    {
+      "epoch": 0.7928925069027897,
+      "grad_norm": 2.0203827144717503,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 33311
+    },
+    {
+      "epoch": 0.7929163096258212,
+      "grad_norm": 2.3052188243245535,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 33312
+    },
+    {
+      "epoch": 0.7929401123488528,
+      "grad_norm": 5.506738204794607,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 33313
+    },
+    {
+      "epoch": 0.7929639150718842,
+      "grad_norm": 6.548867342809718,
+      "learning_rate": 2e-05,
+      "loss": 0.5599,
+      "step": 33314
+    },
+    {
+      "epoch": 0.7929877177949157,
+      "grad_norm": 3.558135432660788,
+      "learning_rate": 2e-05,
+      "loss": 0.5795,
+      "step": 33315
+    },
+    {
+      "epoch": 0.7930115205179472,
+      "grad_norm": 4.905958777571111,
+      "learning_rate": 2e-05,
+      "loss": 0.5921,
+      "step": 33316
+    },
+    {
+      "epoch": 0.7930353232409788,
+      "grad_norm": 3.6371955210286813,
+      "learning_rate": 2e-05,
+      "loss": 0.3277,
+      "step": 33317
+    },
+    {
+      "epoch": 0.7930591259640103,
+      "grad_norm": 0.741628667706765,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 33318
+    },
+    {
+      "epoch": 0.7930829286870418,
+      "grad_norm": 0.2682070079203455,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 33319
+    },
+    {
+      "epoch": 0.7931067314100733,
+      "grad_norm": 1.8565561680111422,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 33320
+    },
+    {
+      "epoch": 0.7931305341331049,
+      "grad_norm": 12.762888091397658,
+      "learning_rate": 2e-05,
+      "loss": 0.9762,
+      "step": 33321
+    },
+    {
+      "epoch": 0.7931543368561363,
+      "grad_norm": 0.49400118953434546,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 33322
+    },
+    {
+      "epoch": 0.7931781395791678,
+      "grad_norm": 4.1417540461429825,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 33323
+    },
+    {
+      "epoch": 0.7932019423021994,
+      "grad_norm": 3.7988036833646968,
+      "learning_rate": 2e-05,
+      "loss": 0.5327,
+      "step": 33324
+    },
+    {
+      "epoch": 0.7932257450252309,
+      "grad_norm": 3.781975005885927,
+      "learning_rate": 2e-05,
+      "loss": 0.6322,
+      "step": 33325
+    },
+    {
+      "epoch": 0.7932495477482624,
+      "grad_norm": 4.963416343492538,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 33326
+    },
+    {
+      "epoch": 0.7932733504712939,
+      "grad_norm": 0.03587160381457741,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33327
+    },
+    {
+      "epoch": 0.7932971531943255,
+      "grad_norm": 2.8122826882806673,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 33328
+    },
+    {
+      "epoch": 0.793320955917357,
+      "grad_norm": 2.23102989207551,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 33329
+    },
+    {
+      "epoch": 0.7933447586403884,
+      "grad_norm": 0.04808580403844998,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33330
+    },
+    {
+      "epoch": 0.7933685613634199,
+      "grad_norm": 20.795350662926477,
+      "learning_rate": 2e-05,
+      "loss": 0.648,
+      "step": 33331
+    },
+    {
+      "epoch": 0.7933923640864515,
+      "grad_norm": 1.387987154941849,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 33332
+    },
+    {
+      "epoch": 0.793416166809483,
+      "grad_norm": 9.085980647317394,
+      "learning_rate": 2e-05,
+      "loss": 1.2617,
+      "step": 33333
+    },
+    {
+      "epoch": 0.7934399695325145,
+      "grad_norm": 13.917626502857924,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 33334
+    },
+    {
+      "epoch": 0.793463772255546,
+      "grad_norm": 0.3414823913806682,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 33335
+    },
+    {
+      "epoch": 0.7934875749785776,
+      "grad_norm": 6.396560855260309,
+      "learning_rate": 2e-05,
+      "loss": 1.1025,
+      "step": 33336
+    },
+    {
+      "epoch": 0.7935113777016091,
+      "grad_norm": 3.858609402889292,
+      "learning_rate": 2e-05,
+      "loss": 0.503,
+      "step": 33337
+    },
+    {
+      "epoch": 0.7935351804246406,
+      "grad_norm": 6.901934122081775,
+      "learning_rate": 2e-05,
+      "loss": 1.5161,
+      "step": 33338
+    },
+    {
+      "epoch": 0.7935589831476721,
+      "grad_norm": 2.3334168628655383,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 33339
+    },
+    {
+      "epoch": 0.7935827858707036,
+      "grad_norm": 3.721620133639724,
+      "learning_rate": 2e-05,
+      "loss": 0.7331,
+      "step": 33340
+    },
+    {
+      "epoch": 0.7936065885937351,
+      "grad_norm": 5.026375153206913,
+      "learning_rate": 2e-05,
+      "loss": 0.6879,
+      "step": 33341
+    },
+    {
+      "epoch": 0.7936303913167666,
+      "grad_norm": 2.4518139040745215,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 33342
+    },
+    {
+      "epoch": 0.7936541940397982,
+      "grad_norm": 5.5537792008766855,
+      "learning_rate": 2e-05,
+      "loss": 0.6638,
+      "step": 33343
+    },
+    {
+      "epoch": 0.7936779967628297,
+      "grad_norm": 3.001073173169564,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 33344
+    },
+    {
+      "epoch": 0.7937017994858612,
+      "grad_norm": 4.4352718340357065,
+      "learning_rate": 2e-05,
+      "loss": 0.7975,
+      "step": 33345
+    },
+    {
+      "epoch": 0.7937256022088927,
+      "grad_norm": 10.941062404485285,
+      "learning_rate": 2e-05,
+      "loss": 1.2512,
+      "step": 33346
+    },
+    {
+      "epoch": 0.7937494049319243,
+      "grad_norm": 4.151994188417034,
+      "learning_rate": 2e-05,
+      "loss": 0.4859,
+      "step": 33347
+    },
+    {
+      "epoch": 0.7937732076549557,
+      "grad_norm": 0.16780017999281222,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 33348
+    },
+    {
+      "epoch": 0.7937970103779872,
+      "grad_norm": 2.7477499248167763,
+      "learning_rate": 2e-05,
+      "loss": 0.4444,
+      "step": 33349
+    },
+    {
+      "epoch": 0.7938208131010187,
+      "grad_norm": 17.984451653077137,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 33350
+    },
+    {
+      "epoch": 0.7938446158240503,
+      "grad_norm": 2.895543000007343,
+      "learning_rate": 2e-05,
+      "loss": 0.1354,
+      "step": 33351
+    },
+    {
+      "epoch": 0.7938684185470818,
+      "grad_norm": 4.688297046977264,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 33352
+    },
+    {
+      "epoch": 0.7938922212701133,
+      "grad_norm": 0.031318159972947346,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 33353
+    },
+    {
+      "epoch": 0.7939160239931449,
+      "grad_norm": 0.17106530624387511,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 33354
+    },
+    {
+      "epoch": 0.7939398267161764,
+      "grad_norm": 0.6161354848176034,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 33355
+    },
+    {
+      "epoch": 0.7939636294392078,
+      "grad_norm": 5.2637091826779985,
+      "learning_rate": 2e-05,
+      "loss": 0.4489,
+      "step": 33356
+    },
+    {
+      "epoch": 0.7939874321622393,
+      "grad_norm": 5.523809106735752,
+      "learning_rate": 2e-05,
+      "loss": 0.5297,
+      "step": 33357
+    },
+    {
+      "epoch": 0.7940112348852709,
+      "grad_norm": 6.1099020465816505,
+      "learning_rate": 2e-05,
+      "loss": 1.5052,
+      "step": 33358
+    },
+    {
+      "epoch": 0.7940350376083024,
+      "grad_norm": 2.392743539766503,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 33359
+    },
+    {
+      "epoch": 0.7940588403313339,
+      "grad_norm": 18.383600862048652,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 33360
+    },
+    {
+      "epoch": 0.7940826430543654,
+      "grad_norm": 10.867403206667984,
+      "learning_rate": 2e-05,
+      "loss": 1.15,
+      "step": 33361
+    },
+    {
+      "epoch": 0.794106445777397,
+      "grad_norm": 0.025176148177109548,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 33362
+    },
+    {
+      "epoch": 0.7941302485004285,
+      "grad_norm": 1.766897679329456,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 33363
+    },
+    {
+      "epoch": 0.79415405122346,
+      "grad_norm": 4.321263760799339,
+      "learning_rate": 2e-05,
+      "loss": 0.6767,
+      "step": 33364
+    },
+    {
+      "epoch": 0.7941778539464914,
+      "grad_norm": 0.2047378847956687,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 33365
+    },
+    {
+      "epoch": 0.794201656669523,
+      "grad_norm": 3.8732820586546652,
+      "learning_rate": 2e-05,
+      "loss": 0.4982,
+      "step": 33366
+    },
+    {
+      "epoch": 0.7942254593925545,
+      "grad_norm": 4.695459482208633,
+      "learning_rate": 2e-05,
+      "loss": 0.4721,
+      "step": 33367
+    },
+    {
+      "epoch": 0.794249262115586,
+      "grad_norm": 3.7089756598592243,
+      "learning_rate": 2e-05,
+      "loss": 0.7524,
+      "step": 33368
+    },
+    {
+      "epoch": 0.7942730648386176,
+      "grad_norm": 2.2985133513594427,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 33369
+    },
+    {
+      "epoch": 0.7942968675616491,
+      "grad_norm": 6.836411607079556,
+      "learning_rate": 2e-05,
+      "loss": 0.4394,
+      "step": 33370
+    },
+    {
+      "epoch": 0.7943206702846806,
+      "grad_norm": 5.446796284275419,
+      "learning_rate": 2e-05,
+      "loss": 0.3337,
+      "step": 33371
+    },
+    {
+      "epoch": 0.794344473007712,
+      "grad_norm": 4.753083428024123,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 33372
+    },
+    {
+      "epoch": 0.7943682757307436,
+      "grad_norm": 1.8896222287335136,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 33373
+    },
+    {
+      "epoch": 0.7943920784537751,
+      "grad_norm": 7.97355032183851,
+      "learning_rate": 2e-05,
+      "loss": 2.3406,
+      "step": 33374
+    },
+    {
+      "epoch": 0.7944158811768066,
+      "grad_norm": 0.1715863460553638,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33375
+    },
+    {
+      "epoch": 0.7944396838998381,
+      "grad_norm": 0.9029417918795257,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 33376
+    },
+    {
+      "epoch": 0.7944634866228697,
+      "grad_norm": 4.798974092826699,
+      "learning_rate": 2e-05,
+      "loss": 0.4555,
+      "step": 33377
+    },
+    {
+      "epoch": 0.7944872893459012,
+      "grad_norm": 8.666778877528342,
+      "learning_rate": 2e-05,
+      "loss": 0.7525,
+      "step": 33378
+    },
+    {
+      "epoch": 0.7945110920689327,
+      "grad_norm": 4.842604507195527,
+      "learning_rate": 2e-05,
+      "loss": 0.419,
+      "step": 33379
+    },
+    {
+      "epoch": 0.7945348947919642,
+      "grad_norm": 5.189772276273437,
+      "learning_rate": 2e-05,
+      "loss": 0.6109,
+      "step": 33380
+    },
+    {
+      "epoch": 0.7945586975149957,
+      "grad_norm": 3.524603582743121,
+      "learning_rate": 2e-05,
+      "loss": 0.5145,
+      "step": 33381
+    },
+    {
+      "epoch": 0.7945825002380272,
+      "grad_norm": 4.711021556707247,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 33382
+    },
+    {
+      "epoch": 0.7946063029610587,
+      "grad_norm": 0.1283079980658026,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33383
+    },
+    {
+      "epoch": 0.7946301056840903,
+      "grad_norm": 4.758868250612337,
+      "learning_rate": 2e-05,
+      "loss": 0.6403,
+      "step": 33384
+    },
+    {
+      "epoch": 0.7946539084071218,
+      "grad_norm": 3.80659374942029,
+      "learning_rate": 2e-05,
+      "loss": 0.2005,
+      "step": 33385
+    },
+    {
+      "epoch": 0.7946777111301533,
+      "grad_norm": 6.525665361674329,
+      "learning_rate": 2e-05,
+      "loss": 1.2026,
+      "step": 33386
+    },
+    {
+      "epoch": 0.7947015138531848,
+      "grad_norm": 8.162928524191953,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 33387
+    },
+    {
+      "epoch": 0.7947253165762164,
+      "grad_norm": 2.486343716708885,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 33388
+    },
+    {
+      "epoch": 0.7947491192992479,
+      "grad_norm": 0.03872218126827567,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 33389
+    },
+    {
+      "epoch": 0.7947729220222793,
+      "grad_norm": 0.587360760028496,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 33390
+    },
+    {
+      "epoch": 0.7947967247453108,
+      "grad_norm": 3.0633252869215224,
+      "learning_rate": 2e-05,
+      "loss": 0.0718,
+      "step": 33391
+    },
+    {
+      "epoch": 0.7948205274683424,
+      "grad_norm": 0.46161247671442135,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 33392
+    },
+    {
+      "epoch": 0.7948443301913739,
+      "grad_norm": 0.2211399659770576,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 33393
+    },
+    {
+      "epoch": 0.7948681329144054,
+      "grad_norm": 2.6541950649915327,
+      "learning_rate": 2e-05,
+      "loss": 0.2752,
+      "step": 33394
+    },
+    {
+      "epoch": 0.7948919356374369,
+      "grad_norm": 0.18122800693957328,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33395
+    },
+    {
+      "epoch": 0.7949157383604685,
+      "grad_norm": 2.980261606988681,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 33396
+    },
+    {
+      "epoch": 0.7949395410835,
+      "grad_norm": 1.5794902223929268,
+      "learning_rate": 2e-05,
+      "loss": 0.0615,
+      "step": 33397
+    },
+    {
+      "epoch": 0.7949633438065314,
+      "grad_norm": 7.178293628890958,
+      "learning_rate": 2e-05,
+      "loss": 0.3083,
+      "step": 33398
+    },
+    {
+      "epoch": 0.7949871465295629,
+      "grad_norm": 17.347440705858553,
+      "learning_rate": 2e-05,
+      "loss": 0.6465,
+      "step": 33399
+    },
+    {
+      "epoch": 0.7950109492525945,
+      "grad_norm": 4.68298432905462,
+      "learning_rate": 2e-05,
+      "loss": 0.6372,
+      "step": 33400
+    },
+    {
+      "epoch": 0.795034751975626,
+      "grad_norm": 0.2570243524460654,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 33401
+    },
+    {
+      "epoch": 0.7950585546986575,
+      "grad_norm": 4.386466153066438,
+      "learning_rate": 2e-05,
+      "loss": 0.4492,
+      "step": 33402
+    },
+    {
+      "epoch": 0.7950823574216891,
+      "grad_norm": 9.86643336324265,
+      "learning_rate": 2e-05,
+      "loss": 1.1844,
+      "step": 33403
+    },
+    {
+      "epoch": 0.7951061601447206,
+      "grad_norm": 5.0751321843544925,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 33404
+    },
+    {
+      "epoch": 0.7951299628677521,
+      "grad_norm": 4.751437050426386,
+      "learning_rate": 2e-05,
+      "loss": 0.6225,
+      "step": 33405
+    },
+    {
+      "epoch": 0.7951537655907835,
+      "grad_norm": 0.8108165816517573,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 33406
+    },
+    {
+      "epoch": 0.7951775683138151,
+      "grad_norm": 1.8487955607071,
+      "learning_rate": 2e-05,
+      "loss": 0.1315,
+      "step": 33407
+    },
+    {
+      "epoch": 0.7952013710368466,
+      "grad_norm": 4.307818269037584,
+      "learning_rate": 2e-05,
+      "loss": 0.3049,
+      "step": 33408
+    },
+    {
+      "epoch": 0.7952251737598781,
+      "grad_norm": 1.2477425335758723,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 33409
+    },
+    {
+      "epoch": 0.7952489764829096,
+      "grad_norm": 7.573343766059104,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 33410
+    },
+    {
+      "epoch": 0.7952727792059412,
+      "grad_norm": 1.135063956778386,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 33411
+    },
+    {
+      "epoch": 0.7952965819289727,
+      "grad_norm": 5.096591644504253,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 33412
+    },
+    {
+      "epoch": 0.7953203846520042,
+      "grad_norm": 3.2178812462859736,
+      "learning_rate": 2e-05,
+      "loss": 0.623,
+      "step": 33413
+    },
+    {
+      "epoch": 0.7953441873750356,
+      "grad_norm": 0.5627765405948464,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 33414
+    },
+    {
+      "epoch": 0.7953679900980672,
+      "grad_norm": 0.20986285720541972,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 33415
+    },
+    {
+      "epoch": 0.7953917928210987,
+      "grad_norm": 7.133412818239025,
+      "learning_rate": 2e-05,
+      "loss": 1.9825,
+      "step": 33416
+    },
+    {
+      "epoch": 0.7954155955441302,
+      "grad_norm": 7.411066590551155,
+      "learning_rate": 2e-05,
+      "loss": 1.2665,
+      "step": 33417
+    },
+    {
+      "epoch": 0.7954393982671618,
+      "grad_norm": 1.2626732347501217,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 33418
+    },
+    {
+      "epoch": 0.7954632009901933,
+      "grad_norm": 2.5307734997502367,
+      "learning_rate": 2e-05,
+      "loss": 0.3366,
+      "step": 33419
+    },
+    {
+      "epoch": 0.7954870037132248,
+      "grad_norm": 3.0062206637388695,
+      "learning_rate": 2e-05,
+      "loss": 0.4231,
+      "step": 33420
+    },
+    {
+      "epoch": 0.7955108064362563,
+      "grad_norm": 6.210368084357041,
+      "learning_rate": 2e-05,
+      "loss": 1.1846,
+      "step": 33421
+    },
+    {
+      "epoch": 0.7955346091592879,
+      "grad_norm": 2.418475056310367,
+      "learning_rate": 2e-05,
+      "loss": 0.3453,
+      "step": 33422
+    },
+    {
+      "epoch": 0.7955584118823193,
+      "grad_norm": 0.960540051571951,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 33423
+    },
+    {
+      "epoch": 0.7955822146053508,
+      "grad_norm": 3.4523003816492657,
+      "learning_rate": 2e-05,
+      "loss": 0.4946,
+      "step": 33424
+    },
+    {
+      "epoch": 0.7956060173283823,
+      "grad_norm": 3.833362543185542,
+      "learning_rate": 2e-05,
+      "loss": 0.5693,
+      "step": 33425
+    },
+    {
+      "epoch": 0.7956298200514139,
+      "grad_norm": 5.34741995066398,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 33426
+    },
+    {
+      "epoch": 0.7956536227744454,
+      "grad_norm": 18.530550976792377,
+      "learning_rate": 2e-05,
+      "loss": 0.1745,
+      "step": 33427
+    },
+    {
+      "epoch": 0.7956774254974769,
+      "grad_norm": 0.11139771505237983,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 33428
+    },
+    {
+      "epoch": 0.7957012282205084,
+      "grad_norm": 5.7305034078917645,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 33429
+    },
+    {
+      "epoch": 0.79572503094354,
+      "grad_norm": 1.230581572413477,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 33430
+    },
+    {
+      "epoch": 0.7957488336665715,
+      "grad_norm": 6.130074554370315,
+      "learning_rate": 2e-05,
+      "loss": 0.7176,
+      "step": 33431
+    },
+    {
+      "epoch": 0.7957726363896029,
+      "grad_norm": 2.8419900179241715,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 33432
+    },
+    {
+      "epoch": 0.7957964391126345,
+      "grad_norm": 4.087973903278471,
+      "learning_rate": 2e-05,
+      "loss": 0.6634,
+      "step": 33433
+    },
+    {
+      "epoch": 0.795820241835666,
+      "grad_norm": 0.6930522477701977,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 33434
+    },
+    {
+      "epoch": 0.7958440445586975,
+      "grad_norm": 2.5080165069378753,
+      "learning_rate": 2e-05,
+      "loss": 0.2795,
+      "step": 33435
+    },
+    {
+      "epoch": 0.795867847281729,
+      "grad_norm": 4.013287882774129,
+      "learning_rate": 2e-05,
+      "loss": 0.8311,
+      "step": 33436
+    },
+    {
+      "epoch": 0.7958916500047606,
+      "grad_norm": 4.005838954754048,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 33437
+    },
+    {
+      "epoch": 0.7959154527277921,
+      "grad_norm": 14.437802159034334,
+      "learning_rate": 2e-05,
+      "loss": 0.7217,
+      "step": 33438
+    },
+    {
+      "epoch": 0.7959392554508236,
+      "grad_norm": 7.708521278661659,
+      "learning_rate": 2e-05,
+      "loss": 1.0172,
+      "step": 33439
+    },
+    {
+      "epoch": 0.795963058173855,
+      "grad_norm": 0.6396015769926582,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 33440
+    },
+    {
+      "epoch": 0.7959868608968866,
+      "grad_norm": 2.5643619687287393,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 33441
+    },
+    {
+      "epoch": 0.7960106636199181,
+      "grad_norm": 1.315351568803902,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 33442
+    },
+    {
+      "epoch": 0.7960344663429496,
+      "grad_norm": 1.6325829067395572,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 33443
+    },
+    {
+      "epoch": 0.7960582690659811,
+      "grad_norm": 2.671656152718868,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 33444
+    },
+    {
+      "epoch": 0.7960820717890127,
+      "grad_norm": 2.500418563215036,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 33445
+    },
+    {
+      "epoch": 0.7961058745120442,
+      "grad_norm": 0.04217986895005255,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33446
+    },
+    {
+      "epoch": 0.7961296772350757,
+      "grad_norm": 9.433692038158572,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 33447
+    },
+    {
+      "epoch": 0.7961534799581073,
+      "grad_norm": 2.5647735286594724,
+      "learning_rate": 2e-05,
+      "loss": 0.3565,
+      "step": 33448
+    },
+    {
+      "epoch": 0.7961772826811387,
+      "grad_norm": 0.27558154600713863,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 33449
+    },
+    {
+      "epoch": 0.7962010854041702,
+      "grad_norm": 13.257318483908024,
+      "learning_rate": 2e-05,
+      "loss": 0.5057,
+      "step": 33450
+    },
+    {
+      "epoch": 0.7962248881272017,
+      "grad_norm": 3.515054918341107,
+      "learning_rate": 2e-05,
+      "loss": 0.7838,
+      "step": 33451
+    },
+    {
+      "epoch": 0.7962486908502333,
+      "grad_norm": 3.0253335605621428,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 33452
+    },
+    {
+      "epoch": 0.7962724935732648,
+      "grad_norm": 6.244012016847561,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 33453
+    },
+    {
+      "epoch": 0.7962962962962963,
+      "grad_norm": 6.465180518313439,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 33454
+    },
+    {
+      "epoch": 0.7963200990193278,
+      "grad_norm": 0.11394898540484447,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 33455
+    },
+    {
+      "epoch": 0.7963439017423594,
+      "grad_norm": 3.7167945823827426,
+      "learning_rate": 2e-05,
+      "loss": 0.9073,
+      "step": 33456
+    },
+    {
+      "epoch": 0.7963677044653908,
+      "grad_norm": 0.6340326960789582,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 33457
+    },
+    {
+      "epoch": 0.7963915071884223,
+      "grad_norm": 5.542251417321648,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 33458
+    },
+    {
+      "epoch": 0.7964153099114538,
+      "grad_norm": 6.0652549379501135,
+      "learning_rate": 2e-05,
+      "loss": 0.5083,
+      "step": 33459
+    },
+    {
+      "epoch": 0.7964391126344854,
+      "grad_norm": 0.2525581112884479,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 33460
+    },
+    {
+      "epoch": 0.7964629153575169,
+      "grad_norm": 10.179588727442397,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 33461
+    },
+    {
+      "epoch": 0.7964867180805484,
+      "grad_norm": 18.238906110912453,
+      "learning_rate": 2e-05,
+      "loss": 0.2676,
+      "step": 33462
+    },
+    {
+      "epoch": 0.79651052080358,
+      "grad_norm": 0.42307542180505797,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33463
+    },
+    {
+      "epoch": 0.7965343235266115,
+      "grad_norm": 0.01972792220141927,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 33464
+    },
+    {
+      "epoch": 0.796558126249643,
+      "grad_norm": 2.765516778550231,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 33465
+    },
+    {
+      "epoch": 0.7965819289726744,
+      "grad_norm": 5.519959076757954,
+      "learning_rate": 2e-05,
+      "loss": 0.4953,
+      "step": 33466
+    },
+    {
+      "epoch": 0.796605731695706,
+      "grad_norm": 5.734391673199084,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 33467
+    },
+    {
+      "epoch": 0.7966295344187375,
+      "grad_norm": 11.067629281565134,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 33468
+    },
+    {
+      "epoch": 0.796653337141769,
+      "grad_norm": 5.031889677605797,
+      "learning_rate": 2e-05,
+      "loss": 0.7964,
+      "step": 33469
+    },
+    {
+      "epoch": 0.7966771398648005,
+      "grad_norm": 4.1781144044182295,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 33470
+    },
+    {
+      "epoch": 0.7967009425878321,
+      "grad_norm": 3.4334339176353064,
+      "learning_rate": 2e-05,
+      "loss": 0.6303,
+      "step": 33471
+    },
+    {
+      "epoch": 0.7967247453108636,
+      "grad_norm": 1.066074530340978,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 33472
+    },
+    {
+      "epoch": 0.796748548033895,
+      "grad_norm": 3.8864672985673203,
+      "learning_rate": 2e-05,
+      "loss": 0.633,
+      "step": 33473
+    },
+    {
+      "epoch": 0.7967723507569265,
+      "grad_norm": 3.232496164773317,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 33474
+    },
+    {
+      "epoch": 0.7967961534799581,
+      "grad_norm": 0.0772321713979179,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33475
+    },
+    {
+      "epoch": 0.7968199562029896,
+      "grad_norm": 0.1784245559713456,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 33476
+    },
+    {
+      "epoch": 0.7968437589260211,
+      "grad_norm": 0.8187916626605619,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 33477
+    },
+    {
+      "epoch": 0.7968675616490527,
+      "grad_norm": 3.631468480342442,
+      "learning_rate": 2e-05,
+      "loss": 0.2535,
+      "step": 33478
+    },
+    {
+      "epoch": 0.7968913643720842,
+      "grad_norm": 0.40931351942444066,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 33479
+    },
+    {
+      "epoch": 0.7969151670951157,
+      "grad_norm": 6.879534657489648,
+      "learning_rate": 2e-05,
+      "loss": 0.9769,
+      "step": 33480
+    },
+    {
+      "epoch": 0.7969389698181472,
+      "grad_norm": 4.9316985927803225,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 33481
+    },
+    {
+      "epoch": 0.7969627725411788,
+      "grad_norm": 3.038052510723953,
+      "learning_rate": 2e-05,
+      "loss": 0.4324,
+      "step": 33482
+    },
+    {
+      "epoch": 0.7969865752642102,
+      "grad_norm": 0.14567365016474365,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 33483
+    },
+    {
+      "epoch": 0.7970103779872417,
+      "grad_norm": 5.072799937034972,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 33484
+    },
+    {
+      "epoch": 0.7970341807102732,
+      "grad_norm": 20.372103722932323,
+      "learning_rate": 2e-05,
+      "loss": 0.6804,
+      "step": 33485
+    },
+    {
+      "epoch": 0.7970579834333048,
+      "grad_norm": 3.6592520938187536,
+      "learning_rate": 2e-05,
+      "loss": 0.577,
+      "step": 33486
+    },
+    {
+      "epoch": 0.7970817861563363,
+      "grad_norm": 6.981119852502103,
+      "learning_rate": 2e-05,
+      "loss": 0.9599,
+      "step": 33487
+    },
+    {
+      "epoch": 0.7971055888793678,
+      "grad_norm": 2.5472228486748385,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 33488
+    },
+    {
+      "epoch": 0.7971293916023993,
+      "grad_norm": 0.1902379622375656,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 33489
+    },
+    {
+      "epoch": 0.7971531943254309,
+      "grad_norm": 4.196899111375585,
+      "learning_rate": 2e-05,
+      "loss": 0.6392,
+      "step": 33490
+    },
+    {
+      "epoch": 0.7971769970484623,
+      "grad_norm": 3.6213628293621087,
+      "learning_rate": 2e-05,
+      "loss": 0.5979,
+      "step": 33491
+    },
+    {
+      "epoch": 0.7972007997714938,
+      "grad_norm": 14.369892657266393,
+      "learning_rate": 2e-05,
+      "loss": 1.9003,
+      "step": 33492
+    },
+    {
+      "epoch": 0.7972246024945254,
+      "grad_norm": 2.6989804477352943,
+      "learning_rate": 2e-05,
+      "loss": 0.2393,
+      "step": 33493
+    },
+    {
+      "epoch": 0.7972484052175569,
+      "grad_norm": 3.434886974113574,
+      "learning_rate": 2e-05,
+      "loss": 0.0612,
+      "step": 33494
+    },
+    {
+      "epoch": 0.7972722079405884,
+      "grad_norm": 3.9419304113735927,
+      "learning_rate": 2e-05,
+      "loss": 0.1165,
+      "step": 33495
+    },
+    {
+      "epoch": 0.7972960106636199,
+      "grad_norm": 7.033411088165973,
+      "learning_rate": 2e-05,
+      "loss": 0.6062,
+      "step": 33496
+    },
+    {
+      "epoch": 0.7973198133866515,
+      "grad_norm": 4.67503652057084,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 33497
+    },
+    {
+      "epoch": 0.797343616109683,
+      "grad_norm": 15.567450983090094,
+      "learning_rate": 2e-05,
+      "loss": 0.3982,
+      "step": 33498
+    },
+    {
+      "epoch": 0.7973674188327144,
+      "grad_norm": 10.853327977728764,
+      "learning_rate": 2e-05,
+      "loss": 1.3749,
+      "step": 33499
+    },
+    {
+      "epoch": 0.7973912215557459,
+      "grad_norm": 2.9833299769952664,
+      "learning_rate": 2e-05,
+      "loss": 0.7525,
+      "step": 33500
+    },
+    {
+      "epoch": 0.7974150242787775,
+      "grad_norm": 18.449183128135978,
+      "learning_rate": 2e-05,
+      "loss": 1.8047,
+      "step": 33501
+    },
+    {
+      "epoch": 0.797438827001809,
+      "grad_norm": 0.3335517215821896,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 33502
+    },
+    {
+      "epoch": 0.7974626297248405,
+      "grad_norm": 3.469880506245131,
+      "learning_rate": 2e-05,
+      "loss": 0.6399,
+      "step": 33503
+    },
+    {
+      "epoch": 0.797486432447872,
+      "grad_norm": 1.3513254147217226,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 33504
+    },
+    {
+      "epoch": 0.7975102351709036,
+      "grad_norm": 3.291443300784892,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 33505
+    },
+    {
+      "epoch": 0.7975340378939351,
+      "grad_norm": 0.36242643885336073,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 33506
+    },
+    {
+      "epoch": 0.7975578406169666,
+      "grad_norm": 14.6483250876904,
+      "learning_rate": 2e-05,
+      "loss": 0.7505,
+      "step": 33507
+    },
+    {
+      "epoch": 0.7975816433399981,
+      "grad_norm": 1.9572221715071523,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 33508
+    },
+    {
+      "epoch": 0.7976054460630296,
+      "grad_norm": 0.3629840563721103,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 33509
+    },
+    {
+      "epoch": 0.7976292487860611,
+      "grad_norm": 5.227739376113001,
+      "learning_rate": 2e-05,
+      "loss": 0.7649,
+      "step": 33510
+    },
+    {
+      "epoch": 0.7976530515090926,
+      "grad_norm": 3.3980568182704047,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 33511
+    },
+    {
+      "epoch": 0.7976768542321242,
+      "grad_norm": 7.836715761666291,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 33512
+    },
+    {
+      "epoch": 0.7977006569551557,
+      "grad_norm": 3.4998894146776864,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 33513
+    },
+    {
+      "epoch": 0.7977244596781872,
+      "grad_norm": 1.1329082528608891,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 33514
+    },
+    {
+      "epoch": 0.7977482624012187,
+      "grad_norm": 0.013238130647601183,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 33515
+    },
+    {
+      "epoch": 0.7977720651242503,
+      "grad_norm": 5.629284536782704,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 33516
+    },
+    {
+      "epoch": 0.7977958678472817,
+      "grad_norm": 4.969516923889462,
+      "learning_rate": 2e-05,
+      "loss": 0.6452,
+      "step": 33517
+    },
+    {
+      "epoch": 0.7978196705703132,
+      "grad_norm": 12.776668922875132,
+      "learning_rate": 2e-05,
+      "loss": 1.0052,
+      "step": 33518
+    },
+    {
+      "epoch": 0.7978434732933447,
+      "grad_norm": 4.516997210363851,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 33519
+    },
+    {
+      "epoch": 0.7978672760163763,
+      "grad_norm": 3.125848578011412,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 33520
+    },
+    {
+      "epoch": 0.7978910787394078,
+      "grad_norm": 0.8175927875300071,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 33521
+    },
+    {
+      "epoch": 0.7979148814624393,
+      "grad_norm": 4.948274724899545,
+      "learning_rate": 2e-05,
+      "loss": 1.5554,
+      "step": 33522
+    },
+    {
+      "epoch": 0.7979386841854709,
+      "grad_norm": 3.9447542949347794,
+      "learning_rate": 2e-05,
+      "loss": 0.5049,
+      "step": 33523
+    },
+    {
+      "epoch": 0.7979624869085024,
+      "grad_norm": 1.0201404910195422,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 33524
+    },
+    {
+      "epoch": 0.7979862896315338,
+      "grad_norm": 0.5723948748572513,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 33525
+    },
+    {
+      "epoch": 0.7980100923545653,
+      "grad_norm": 2.9718454448831126,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 33526
+    },
+    {
+      "epoch": 0.7980338950775969,
+      "grad_norm": 3.824262373467321,
+      "learning_rate": 2e-05,
+      "loss": 0.7016,
+      "step": 33527
+    },
+    {
+      "epoch": 0.7980576978006284,
+      "grad_norm": 5.168092154507829,
+      "learning_rate": 2e-05,
+      "loss": 0.7983,
+      "step": 33528
+    },
+    {
+      "epoch": 0.7980815005236599,
+      "grad_norm": 55.08243011137589,
+      "learning_rate": 2e-05,
+      "loss": 0.7452,
+      "step": 33529
+    },
+    {
+      "epoch": 0.7981053032466914,
+      "grad_norm": 0.16459284300657528,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 33530
+    },
+    {
+      "epoch": 0.798129105969723,
+      "grad_norm": 4.73260079708926,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 33531
+    },
+    {
+      "epoch": 0.7981529086927545,
+      "grad_norm": 0.337349188080743,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 33532
+    },
+    {
+      "epoch": 0.7981767114157859,
+      "grad_norm": 6.539305508427129,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 33533
+    },
+    {
+      "epoch": 0.7982005141388174,
+      "grad_norm": 4.373507838918037,
+      "learning_rate": 2e-05,
+      "loss": 0.2873,
+      "step": 33534
+    },
+    {
+      "epoch": 0.798224316861849,
+      "grad_norm": 0.10543077971100273,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 33535
+    },
+    {
+      "epoch": 0.7982481195848805,
+      "grad_norm": 15.103243722744894,
+      "learning_rate": 2e-05,
+      "loss": 0.4631,
+      "step": 33536
+    },
+    {
+      "epoch": 0.798271922307912,
+      "grad_norm": 4.69332845900189,
+      "learning_rate": 2e-05,
+      "loss": 0.6164,
+      "step": 33537
+    },
+    {
+      "epoch": 0.7982957250309436,
+      "grad_norm": 7.246419326165939,
+      "learning_rate": 2e-05,
+      "loss": 0.8415,
+      "step": 33538
+    },
+    {
+      "epoch": 0.7983195277539751,
+      "grad_norm": 4.588971924627329,
+      "learning_rate": 2e-05,
+      "loss": 0.6485,
+      "step": 33539
+    },
+    {
+      "epoch": 0.7983433304770066,
+      "grad_norm": 0.02898591992765511,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 33540
+    },
+    {
+      "epoch": 0.798367133200038,
+      "grad_norm": 2.6736554441722906,
+      "learning_rate": 2e-05,
+      "loss": 0.3664,
+      "step": 33541
+    },
+    {
+      "epoch": 0.7983909359230696,
+      "grad_norm": 0.05041365510845822,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33542
+    },
+    {
+      "epoch": 0.7984147386461011,
+      "grad_norm": 3.4049347625611017,
+      "learning_rate": 2e-05,
+      "loss": 0.5395,
+      "step": 33543
+    },
+    {
+      "epoch": 0.7984385413691326,
+      "grad_norm": 0.9714665605864864,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 33544
+    },
+    {
+      "epoch": 0.7984623440921641,
+      "grad_norm": 9.945753824596432,
+      "learning_rate": 2e-05,
+      "loss": 1.9563,
+      "step": 33545
+    },
+    {
+      "epoch": 0.7984861468151957,
+      "grad_norm": 3.9933328341504564,
+      "learning_rate": 2e-05,
+      "loss": 0.4385,
+      "step": 33546
+    },
+    {
+      "epoch": 0.7985099495382272,
+      "grad_norm": 2.930664266011394,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 33547
+    },
+    {
+      "epoch": 0.7985337522612587,
+      "grad_norm": 3.8660120240343767,
+      "learning_rate": 2e-05,
+      "loss": 0.5228,
+      "step": 33548
+    },
+    {
+      "epoch": 0.7985575549842902,
+      "grad_norm": 1.1817666058734226,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 33549
+    },
+    {
+      "epoch": 0.7985813577073217,
+      "grad_norm": 6.6235280177678675,
+      "learning_rate": 2e-05,
+      "loss": 1.0507,
+      "step": 33550
+    },
+    {
+      "epoch": 0.7986051604303532,
+      "grad_norm": 2.453360380475486,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 33551
+    },
+    {
+      "epoch": 0.7986289631533847,
+      "grad_norm": 5.062013517031864,
+      "learning_rate": 2e-05,
+      "loss": 0.6449,
+      "step": 33552
+    },
+    {
+      "epoch": 0.7986527658764163,
+      "grad_norm": 4.256672683533561,
+      "learning_rate": 2e-05,
+      "loss": 0.7718,
+      "step": 33553
+    },
+    {
+      "epoch": 0.7986765685994478,
+      "grad_norm": 0.04144394329326698,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33554
+    },
+    {
+      "epoch": 0.7987003713224793,
+      "grad_norm": 4.868031184940995,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 33555
+    },
+    {
+      "epoch": 0.7987241740455108,
+      "grad_norm": 3.322218746842406,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 33556
+    },
+    {
+      "epoch": 0.7987479767685424,
+      "grad_norm": 0.3262032440662551,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33557
+    },
+    {
+      "epoch": 0.7987717794915739,
+      "grad_norm": 2.73706233673813,
+      "learning_rate": 2e-05,
+      "loss": 0.3006,
+      "step": 33558
+    },
+    {
+      "epoch": 0.7987955822146053,
+      "grad_norm": 5.243665346336734,
+      "learning_rate": 2e-05,
+      "loss": 1.1142,
+      "step": 33559
+    },
+    {
+      "epoch": 0.7988193849376368,
+      "grad_norm": 2.7172981323680596,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 33560
+    },
+    {
+      "epoch": 0.7988431876606684,
+      "grad_norm": 6.47482194845762,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 33561
+    },
+    {
+      "epoch": 0.7988669903836999,
+      "grad_norm": 2.986561126812327,
+      "learning_rate": 2e-05,
+      "loss": 0.2121,
+      "step": 33562
+    },
+    {
+      "epoch": 0.7988907931067314,
+      "grad_norm": 0.022313868228410517,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 33563
+    },
+    {
+      "epoch": 0.7989145958297629,
+      "grad_norm": 0.07235672950121151,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33564
+    },
+    {
+      "epoch": 0.7989383985527945,
+      "grad_norm": 0.13860406748544782,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 33565
+    },
+    {
+      "epoch": 0.798962201275826,
+      "grad_norm": 0.04455390624295542,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33566
+    },
+    {
+      "epoch": 0.7989860039988574,
+      "grad_norm": 3.3210176051082185,
+      "learning_rate": 2e-05,
+      "loss": 0.5215,
+      "step": 33567
+    },
+    {
+      "epoch": 0.799009806721889,
+      "grad_norm": 7.253576663629904,
+      "learning_rate": 2e-05,
+      "loss": 0.6308,
+      "step": 33568
+    },
+    {
+      "epoch": 0.7990336094449205,
+      "grad_norm": 3.0395209574961455,
+      "learning_rate": 2e-05,
+      "loss": 0.7104,
+      "step": 33569
+    },
+    {
+      "epoch": 0.799057412167952,
+      "grad_norm": 5.531045954789691,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 33570
+    },
+    {
+      "epoch": 0.7990812148909835,
+      "grad_norm": 7.0321731168458035,
+      "learning_rate": 2e-05,
+      "loss": 1.0629,
+      "step": 33571
+    },
+    {
+      "epoch": 0.7991050176140151,
+      "grad_norm": 4.53354696627747,
+      "learning_rate": 2e-05,
+      "loss": 0.7353,
+      "step": 33572
+    },
+    {
+      "epoch": 0.7991288203370466,
+      "grad_norm": 2.6689867852961724,
+      "learning_rate": 2e-05,
+      "loss": 0.1503,
+      "step": 33573
+    },
+    {
+      "epoch": 0.7991526230600781,
+      "grad_norm": 3.36430530111232,
+      "learning_rate": 2e-05,
+      "loss": 0.6586,
+      "step": 33574
+    },
+    {
+      "epoch": 0.7991764257831095,
+      "grad_norm": 0.5576772576400625,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 33575
+    },
+    {
+      "epoch": 0.7992002285061411,
+      "grad_norm": 3.1949235512099485,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 33576
+    },
+    {
+      "epoch": 0.7992240312291726,
+      "grad_norm": 0.5816931351910852,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33577
+    },
+    {
+      "epoch": 0.7992478339522041,
+      "grad_norm": 6.091679145259865,
+      "learning_rate": 2e-05,
+      "loss": 0.7673,
+      "step": 33578
+    },
+    {
+      "epoch": 0.7992716366752356,
+      "grad_norm": 4.5087035109456295,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 33579
+    },
+    {
+      "epoch": 0.7992954393982672,
+      "grad_norm": 4.422945595406624,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 33580
+    },
+    {
+      "epoch": 0.7993192421212987,
+      "grad_norm": 1.0123957077418586,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 33581
+    },
+    {
+      "epoch": 0.7993430448443302,
+      "grad_norm": 0.061143483621730506,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 33582
+    },
+    {
+      "epoch": 0.7993668475673618,
+      "grad_norm": 0.06210617934193533,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33583
+    },
+    {
+      "epoch": 0.7993906502903932,
+      "grad_norm": 5.930196382445439,
+      "learning_rate": 2e-05,
+      "loss": 1.6271,
+      "step": 33584
+    },
+    {
+      "epoch": 0.7994144530134247,
+      "grad_norm": 6.712367991814214,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 33585
+    },
+    {
+      "epoch": 0.7994382557364562,
+      "grad_norm": 0.9039347782145462,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 33586
+    },
+    {
+      "epoch": 0.7994620584594878,
+      "grad_norm": 3.802797540569907,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 33587
+    },
+    {
+      "epoch": 0.7994858611825193,
+      "grad_norm": 5.402088228132874,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 33588
+    },
+    {
+      "epoch": 0.7995096639055508,
+      "grad_norm": 0.9939461426763868,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 33589
+    },
+    {
+      "epoch": 0.7995334666285823,
+      "grad_norm": 4.8929865241799,
+      "learning_rate": 2e-05,
+      "loss": 0.5918,
+      "step": 33590
+    },
+    {
+      "epoch": 0.7995572693516139,
+      "grad_norm": 5.0592756546797935,
+      "learning_rate": 2e-05,
+      "loss": 0.3933,
+      "step": 33591
+    },
+    {
+      "epoch": 0.7995810720746453,
+      "grad_norm": 0.09673526609896765,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 33592
+    },
+    {
+      "epoch": 0.7996048747976768,
+      "grad_norm": 0.18654825381041973,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 33593
+    },
+    {
+      "epoch": 0.7996286775207083,
+      "grad_norm": 6.7032143643294235,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 33594
+    },
+    {
+      "epoch": 0.7996524802437399,
+      "grad_norm": 0.3664865978236245,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 33595
+    },
+    {
+      "epoch": 0.7996762829667714,
+      "grad_norm": 0.07637360011315501,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33596
+    },
+    {
+      "epoch": 0.7997000856898029,
+      "grad_norm": 1.485318006967147,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 33597
+    },
+    {
+      "epoch": 0.7997238884128345,
+      "grad_norm": 1.9243782259690994,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 33598
+    },
+    {
+      "epoch": 0.799747691135866,
+      "grad_norm": 4.048412715527398,
+      "learning_rate": 2e-05,
+      "loss": 0.6361,
+      "step": 33599
+    },
+    {
+      "epoch": 0.7997714938588975,
+      "grad_norm": 4.509499566679491,
+      "learning_rate": 2e-05,
+      "loss": 0.939,
+      "step": 33600
+    },
+    {
+      "epoch": 0.7997952965819289,
+      "grad_norm": 3.0163781451869327,
+      "learning_rate": 2e-05,
+      "loss": 0.3037,
+      "step": 33601
+    },
+    {
+      "epoch": 0.7998190993049605,
+      "grad_norm": 2.739016268657329,
+      "learning_rate": 2e-05,
+      "loss": 0.5717,
+      "step": 33602
+    },
+    {
+      "epoch": 0.799842902027992,
+      "grad_norm": 3.7505163199655605,
+      "learning_rate": 2e-05,
+      "loss": 0.0562,
+      "step": 33603
+    },
+    {
+      "epoch": 0.7998667047510235,
+      "grad_norm": 0.1520163949460058,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33604
+    },
+    {
+      "epoch": 0.799890507474055,
+      "grad_norm": 18.9141081425576,
+      "learning_rate": 2e-05,
+      "loss": 0.9364,
+      "step": 33605
+    },
+    {
+      "epoch": 0.7999143101970866,
+      "grad_norm": 9.97548001045015,
+      "learning_rate": 2e-05,
+      "loss": 0.5428,
+      "step": 33606
+    },
+    {
+      "epoch": 0.7999381129201181,
+      "grad_norm": 3.159360645482128,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 33607
+    },
+    {
+      "epoch": 0.7999619156431496,
+      "grad_norm": 3.6596597786731624,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 33608
+    },
+    {
+      "epoch": 0.799985718366181,
+      "grad_norm": 2.680237048127707,
+      "learning_rate": 2e-05,
+      "loss": 0.0676,
+      "step": 33609
+    },
+    {
+      "epoch": 0.8000095210892126,
+      "grad_norm": 0.032768017529423335,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33610
+    },
+    {
+      "epoch": 0.8000333238122441,
+      "grad_norm": 0.20042876713930008,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 33611
+    },
+    {
+      "epoch": 0.8000571265352756,
+      "grad_norm": 2.0336649484755505,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 33612
+    },
+    {
+      "epoch": 0.8000809292583072,
+      "grad_norm": 4.761011374016253,
+      "learning_rate": 2e-05,
+      "loss": 0.7377,
+      "step": 33613
+    },
+    {
+      "epoch": 0.8001047319813387,
+      "grad_norm": 3.1531715759062124,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 33614
+    },
+    {
+      "epoch": 0.8001285347043702,
+      "grad_norm": 6.131899533455205,
+      "learning_rate": 2e-05,
+      "loss": 0.3085,
+      "step": 33615
+    },
+    {
+      "epoch": 0.8001523374274017,
+      "grad_norm": 3.1590650685718638,
+      "learning_rate": 2e-05,
+      "loss": 0.081,
+      "step": 33616
+    },
+    {
+      "epoch": 0.8001761401504333,
+      "grad_norm": 2.8296173392933865,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 33617
+    },
+    {
+      "epoch": 0.8001999428734647,
+      "grad_norm": 3.1636153853357456,
+      "learning_rate": 2e-05,
+      "loss": 0.5286,
+      "step": 33618
+    },
+    {
+      "epoch": 0.8002237455964962,
+      "grad_norm": 1.058726385840003,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 33619
+    },
+    {
+      "epoch": 0.8002475483195277,
+      "grad_norm": 0.08192914471908379,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33620
+    },
+    {
+      "epoch": 0.8002713510425593,
+      "grad_norm": 0.15943835449526655,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 33621
+    },
+    {
+      "epoch": 0.8002951537655908,
+      "grad_norm": 0.6021932605399805,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 33622
+    },
+    {
+      "epoch": 0.8003189564886223,
+      "grad_norm": 0.34424537302077657,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 33623
+    },
+    {
+      "epoch": 0.8003427592116538,
+      "grad_norm": 2.1891544185955527,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 33624
+    },
+    {
+      "epoch": 0.8003665619346854,
+      "grad_norm": 0.7923008145158764,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 33625
+    },
+    {
+      "epoch": 0.8003903646577168,
+      "grad_norm": 4.416599361759748,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 33626
+    },
+    {
+      "epoch": 0.8004141673807483,
+      "grad_norm": 0.5624748387323909,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 33627
+    },
+    {
+      "epoch": 0.8004379701037799,
+      "grad_norm": 15.069026941010597,
+      "learning_rate": 2e-05,
+      "loss": 0.2542,
+      "step": 33628
+    },
+    {
+      "epoch": 0.8004617728268114,
+      "grad_norm": 5.561623242586068,
+      "learning_rate": 2e-05,
+      "loss": 0.8493,
+      "step": 33629
+    },
+    {
+      "epoch": 0.8004855755498429,
+      "grad_norm": 1.8847943891499093,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 33630
+    },
+    {
+      "epoch": 0.8005093782728744,
+      "grad_norm": 6.4179066307765495,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 33631
+    },
+    {
+      "epoch": 0.800533180995906,
+      "grad_norm": 0.12573823059698735,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 33632
+    },
+    {
+      "epoch": 0.8005569837189375,
+      "grad_norm": 8.986497284933714,
+      "learning_rate": 2e-05,
+      "loss": 0.5062,
+      "step": 33633
+    },
+    {
+      "epoch": 0.800580786441969,
+      "grad_norm": 2.8864816047513715,
+      "learning_rate": 2e-05,
+      "loss": 0.3252,
+      "step": 33634
+    },
+    {
+      "epoch": 0.8006045891650004,
+      "grad_norm": 3.5503932614618976,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 33635
+    },
+    {
+      "epoch": 0.800628391888032,
+      "grad_norm": 6.475942009234317,
+      "learning_rate": 2e-05,
+      "loss": 1.2066,
+      "step": 33636
+    },
+    {
+      "epoch": 0.8006521946110635,
+      "grad_norm": 2.508976478848207,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 33637
+    },
+    {
+      "epoch": 0.800675997334095,
+      "grad_norm": 0.014828037055284928,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 33638
+    },
+    {
+      "epoch": 0.8006998000571265,
+      "grad_norm": 3.3410260075384177,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 33639
+    },
+    {
+      "epoch": 0.8007236027801581,
+      "grad_norm": 2.53568494236754,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 33640
+    },
+    {
+      "epoch": 0.8007474055031896,
+      "grad_norm": 12.229191667275979,
+      "learning_rate": 2e-05,
+      "loss": 0.6516,
+      "step": 33641
+    },
+    {
+      "epoch": 0.800771208226221,
+      "grad_norm": 0.11179529056382415,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33642
+    },
+    {
+      "epoch": 0.8007950109492525,
+      "grad_norm": 0.06542527844152324,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 33643
+    },
+    {
+      "epoch": 0.8008188136722841,
+      "grad_norm": 0.6170364786880255,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 33644
+    },
+    {
+      "epoch": 0.8008426163953156,
+      "grad_norm": 3.317753094745923,
+      "learning_rate": 2e-05,
+      "loss": 0.2717,
+      "step": 33645
+    },
+    {
+      "epoch": 0.8008664191183471,
+      "grad_norm": 0.1479483404688948,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 33646
+    },
+    {
+      "epoch": 0.8008902218413787,
+      "grad_norm": 6.445777192175346,
+      "learning_rate": 2e-05,
+      "loss": 0.6188,
+      "step": 33647
+    },
+    {
+      "epoch": 0.8009140245644102,
+      "grad_norm": 3.302237170257357,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 33648
+    },
+    {
+      "epoch": 0.8009378272874417,
+      "grad_norm": 0.2670837100635811,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 33649
+    },
+    {
+      "epoch": 0.8009616300104732,
+      "grad_norm": 0.8195936853773533,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 33650
+    },
+    {
+      "epoch": 0.8009854327335048,
+      "grad_norm": 7.79569096657997,
+      "learning_rate": 2e-05,
+      "loss": 0.2111,
+      "step": 33651
+    },
+    {
+      "epoch": 0.8010092354565362,
+      "grad_norm": 3.462485229047789,
+      "learning_rate": 2e-05,
+      "loss": 0.2202,
+      "step": 33652
+    },
+    {
+      "epoch": 0.8010330381795677,
+      "grad_norm": 0.19222741855775688,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 33653
+    },
+    {
+      "epoch": 0.8010568409025992,
+      "grad_norm": 4.013021009863065,
+      "learning_rate": 2e-05,
+      "loss": 0.4181,
+      "step": 33654
+    },
+    {
+      "epoch": 0.8010806436256308,
+      "grad_norm": 4.72801826310068,
+      "learning_rate": 2e-05,
+      "loss": 0.6439,
+      "step": 33655
+    },
+    {
+      "epoch": 0.8011044463486623,
+      "grad_norm": 4.570845361083166,
+      "learning_rate": 2e-05,
+      "loss": 0.54,
+      "step": 33656
+    },
+    {
+      "epoch": 0.8011282490716938,
+      "grad_norm": 4.896913960169473,
+      "learning_rate": 2e-05,
+      "loss": 1.1348,
+      "step": 33657
+    },
+    {
+      "epoch": 0.8011520517947253,
+      "grad_norm": 4.177345849380671,
+      "learning_rate": 2e-05,
+      "loss": 0.8544,
+      "step": 33658
+    },
+    {
+      "epoch": 0.8011758545177569,
+      "grad_norm": 2.6558996961971735,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 33659
+    },
+    {
+      "epoch": 0.8011996572407883,
+      "grad_norm": 4.790728858416505,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 33660
+    },
+    {
+      "epoch": 0.8012234599638198,
+      "grad_norm": 4.864162552647206,
+      "learning_rate": 2e-05,
+      "loss": 0.5871,
+      "step": 33661
+    },
+    {
+      "epoch": 0.8012472626868514,
+      "grad_norm": 7.588978465905032,
+      "learning_rate": 2e-05,
+      "loss": 1.4653,
+      "step": 33662
+    },
+    {
+      "epoch": 0.8012710654098829,
+      "grad_norm": 5.424166469216205,
+      "learning_rate": 2e-05,
+      "loss": 0.2267,
+      "step": 33663
+    },
+    {
+      "epoch": 0.8012948681329144,
+      "grad_norm": 1.005749020762581,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 33664
+    },
+    {
+      "epoch": 0.8013186708559459,
+      "grad_norm": 4.977047825328192,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 33665
+    },
+    {
+      "epoch": 0.8013424735789775,
+      "grad_norm": 2.162402080741796,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 33666
+    },
+    {
+      "epoch": 0.801366276302009,
+      "grad_norm": 0.056765334713926835,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33667
+    },
+    {
+      "epoch": 0.8013900790250404,
+      "grad_norm": 1.5750073806052374,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 33668
+    },
+    {
+      "epoch": 0.8014138817480719,
+      "grad_norm": 1.3633999154280914,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 33669
+    },
+    {
+      "epoch": 0.8014376844711035,
+      "grad_norm": 6.19797478472903,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 33670
+    },
+    {
+      "epoch": 0.801461487194135,
+      "grad_norm": 4.297721066372518,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 33671
+    },
+    {
+      "epoch": 0.8014852899171665,
+      "grad_norm": 5.140742965541677,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 33672
+    },
+    {
+      "epoch": 0.801509092640198,
+      "grad_norm": 0.30701735545076997,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 33673
+    },
+    {
+      "epoch": 0.8015328953632296,
+      "grad_norm": 0.08083552984444488,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33674
+    },
+    {
+      "epoch": 0.8015566980862611,
+      "grad_norm": 2.7919835037434115,
+      "learning_rate": 2e-05,
+      "loss": 0.5719,
+      "step": 33675
+    },
+    {
+      "epoch": 0.8015805008092926,
+      "grad_norm": 4.215007227860287,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 33676
+    },
+    {
+      "epoch": 0.8016043035323241,
+      "grad_norm": 1.192880678233847,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 33677
+    },
+    {
+      "epoch": 0.8016281062553556,
+      "grad_norm": 15.920384649041754,
+      "learning_rate": 2e-05,
+      "loss": 0.8617,
+      "step": 33678
+    },
+    {
+      "epoch": 0.8016519089783871,
+      "grad_norm": 8.70699209943686,
+      "learning_rate": 2e-05,
+      "loss": 0.7882,
+      "step": 33679
+    },
+    {
+      "epoch": 0.8016757117014186,
+      "grad_norm": 6.968126389668611,
+      "learning_rate": 2e-05,
+      "loss": 1.3751,
+      "step": 33680
+    },
+    {
+      "epoch": 0.8016995144244502,
+      "grad_norm": 0.11290173365195291,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 33681
+    },
+    {
+      "epoch": 0.8017233171474817,
+      "grad_norm": 0.420328316752347,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 33682
+    },
+    {
+      "epoch": 0.8017471198705132,
+      "grad_norm": 3.361463324643252,
+      "learning_rate": 2e-05,
+      "loss": 0.2115,
+      "step": 33683
+    },
+    {
+      "epoch": 0.8017709225935447,
+      "grad_norm": 0.45492847052863494,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 33684
+    },
+    {
+      "epoch": 0.8017947253165763,
+      "grad_norm": 7.020502014369184,
+      "learning_rate": 2e-05,
+      "loss": 0.1972,
+      "step": 33685
+    },
+    {
+      "epoch": 0.8018185280396077,
+      "grad_norm": 3.557436185312668,
+      "learning_rate": 2e-05,
+      "loss": 0.6682,
+      "step": 33686
+    },
+    {
+      "epoch": 0.8018423307626392,
+      "grad_norm": 0.1610327296399466,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33687
+    },
+    {
+      "epoch": 0.8018661334856707,
+      "grad_norm": 1.3538739441917524,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 33688
+    },
+    {
+      "epoch": 0.8018899362087023,
+      "grad_norm": 2.9257710491382527,
+      "learning_rate": 2e-05,
+      "loss": 0.3808,
+      "step": 33689
+    },
+    {
+      "epoch": 0.8019137389317338,
+      "grad_norm": 0.984234900246812,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 33690
+    },
+    {
+      "epoch": 0.8019375416547653,
+      "grad_norm": 11.134995687568798,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 33691
+    },
+    {
+      "epoch": 0.8019613443777969,
+      "grad_norm": 3.5443961617868824,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 33692
+    },
+    {
+      "epoch": 0.8019851471008284,
+      "grad_norm": 0.0666002073772066,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 33693
+    },
+    {
+      "epoch": 0.8020089498238598,
+      "grad_norm": 10.301519021786373,
+      "learning_rate": 2e-05,
+      "loss": 1.1247,
+      "step": 33694
+    },
+    {
+      "epoch": 0.8020327525468913,
+      "grad_norm": 0.489092804091285,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 33695
+    },
+    {
+      "epoch": 0.8020565552699229,
+      "grad_norm": 6.663721699980704,
+      "learning_rate": 2e-05,
+      "loss": 0.3262,
+      "step": 33696
+    },
+    {
+      "epoch": 0.8020803579929544,
+      "grad_norm": 10.758239577494264,
+      "learning_rate": 2e-05,
+      "loss": 0.6751,
+      "step": 33697
+    },
+    {
+      "epoch": 0.8021041607159859,
+      "grad_norm": 7.948565155379297,
+      "learning_rate": 2e-05,
+      "loss": 0.6246,
+      "step": 33698
+    },
+    {
+      "epoch": 0.8021279634390174,
+      "grad_norm": 3.125164802794357,
+      "learning_rate": 2e-05,
+      "loss": 0.1411,
+      "step": 33699
+    },
+    {
+      "epoch": 0.802151766162049,
+      "grad_norm": 0.4752425897067598,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 33700
+    },
+    {
+      "epoch": 0.8021755688850805,
+      "grad_norm": 4.5151930354127074,
+      "learning_rate": 2e-05,
+      "loss": 1.2753,
+      "step": 33701
+    },
+    {
+      "epoch": 0.8021993716081119,
+      "grad_norm": 4.391032582489767,
+      "learning_rate": 2e-05,
+      "loss": 0.6779,
+      "step": 33702
+    },
+    {
+      "epoch": 0.8022231743311434,
+      "grad_norm": 0.26430002682158826,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 33703
+    },
+    {
+      "epoch": 0.802246977054175,
+      "grad_norm": 5.436487000268949,
+      "learning_rate": 2e-05,
+      "loss": 0.8283,
+      "step": 33704
+    },
+    {
+      "epoch": 0.8022707797772065,
+      "grad_norm": 0.24183018788124228,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 33705
+    },
+    {
+      "epoch": 0.802294582500238,
+      "grad_norm": 13.454811604097081,
+      "learning_rate": 2e-05,
+      "loss": 0.6885,
+      "step": 33706
+    },
+    {
+      "epoch": 0.8023183852232696,
+      "grad_norm": 5.666904405354348,
+      "learning_rate": 2e-05,
+      "loss": 0.7385,
+      "step": 33707
+    },
+    {
+      "epoch": 0.8023421879463011,
+      "grad_norm": 5.8903544752536146,
+      "learning_rate": 2e-05,
+      "loss": 1.4308,
+      "step": 33708
+    },
+    {
+      "epoch": 0.8023659906693326,
+      "grad_norm": 6.063653193670583,
+      "learning_rate": 2e-05,
+      "loss": 1.2501,
+      "step": 33709
+    },
+    {
+      "epoch": 0.802389793392364,
+      "grad_norm": 3.436994184944313,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 33710
+    },
+    {
+      "epoch": 0.8024135961153956,
+      "grad_norm": 2.0473130101728945,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 33711
+    },
+    {
+      "epoch": 0.8024373988384271,
+      "grad_norm": 2.9475482019781807,
+      "learning_rate": 2e-05,
+      "loss": 0.4784,
+      "step": 33712
+    },
+    {
+      "epoch": 0.8024612015614586,
+      "grad_norm": 0.09487597484574775,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 33713
+    },
+    {
+      "epoch": 0.8024850042844901,
+      "grad_norm": 3.2660818369306965,
+      "learning_rate": 2e-05,
+      "loss": 0.5037,
+      "step": 33714
+    },
+    {
+      "epoch": 0.8025088070075217,
+      "grad_norm": 0.09572399400043048,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33715
+    },
+    {
+      "epoch": 0.8025326097305532,
+      "grad_norm": 5.398746622312322,
+      "learning_rate": 2e-05,
+      "loss": 0.4673,
+      "step": 33716
+    },
+    {
+      "epoch": 0.8025564124535847,
+      "grad_norm": 2.5988316912803073,
+      "learning_rate": 2e-05,
+      "loss": 0.4204,
+      "step": 33717
+    },
+    {
+      "epoch": 0.8025802151766162,
+      "grad_norm": 2.6779387204077008,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 33718
+    },
+    {
+      "epoch": 0.8026040178996477,
+      "grad_norm": 3.99880658776847,
+      "learning_rate": 2e-05,
+      "loss": 0.5709,
+      "step": 33719
+    },
+    {
+      "epoch": 0.8026278206226792,
+      "grad_norm": 0.03839978069103998,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33720
+    },
+    {
+      "epoch": 0.8026516233457107,
+      "grad_norm": 2.0710817789805733,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 33721
+    },
+    {
+      "epoch": 0.8026754260687423,
+      "grad_norm": 1.7650629929537702,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 33722
+    },
+    {
+      "epoch": 0.8026992287917738,
+      "grad_norm": 0.26222067308620245,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 33723
+    },
+    {
+      "epoch": 0.8027230315148053,
+      "grad_norm": 0.06469967753665767,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 33724
+    },
+    {
+      "epoch": 0.8027468342378368,
+      "grad_norm": 0.3107219971604829,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 33725
+    },
+    {
+      "epoch": 0.8027706369608684,
+      "grad_norm": 2.6439694348832274,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 33726
+    },
+    {
+      "epoch": 0.8027944396838999,
+      "grad_norm": 7.155425337196622,
+      "learning_rate": 2e-05,
+      "loss": 0.1172,
+      "step": 33727
+    },
+    {
+      "epoch": 0.8028182424069313,
+      "grad_norm": 7.684505359328159,
+      "learning_rate": 2e-05,
+      "loss": 1.3244,
+      "step": 33728
+    },
+    {
+      "epoch": 0.8028420451299628,
+      "grad_norm": 1.5580938367275525,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 33729
+    },
+    {
+      "epoch": 0.8028658478529944,
+      "grad_norm": 4.293585442690147,
+      "learning_rate": 2e-05,
+      "loss": 0.4559,
+      "step": 33730
+    },
+    {
+      "epoch": 0.8028896505760259,
+      "grad_norm": 14.589286177664626,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 33731
+    },
+    {
+      "epoch": 0.8029134532990574,
+      "grad_norm": 0.3741020365862533,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33732
+    },
+    {
+      "epoch": 0.8029372560220889,
+      "grad_norm": 3.9794733835649434,
+      "learning_rate": 2e-05,
+      "loss": 0.0594,
+      "step": 33733
+    },
+    {
+      "epoch": 0.8029610587451205,
+      "grad_norm": 4.819015547749007,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 33734
+    },
+    {
+      "epoch": 0.802984861468152,
+      "grad_norm": 0.08710289419169873,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 33735
+    },
+    {
+      "epoch": 0.8030086641911834,
+      "grad_norm": 0.06141626207465944,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 33736
+    },
+    {
+      "epoch": 0.803032466914215,
+      "grad_norm": 4.948923244343917,
+      "learning_rate": 2e-05,
+      "loss": 0.5187,
+      "step": 33737
+    },
+    {
+      "epoch": 0.8030562696372465,
+      "grad_norm": 0.37757218643497303,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33738
+    },
+    {
+      "epoch": 0.803080072360278,
+      "grad_norm": 0.1615534062061665,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 33739
+    },
+    {
+      "epoch": 0.8031038750833095,
+      "grad_norm": 2.529502661194828,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 33740
+    },
+    {
+      "epoch": 0.8031276778063411,
+      "grad_norm": 10.776534044635968,
+      "learning_rate": 2e-05,
+      "loss": 0.1021,
+      "step": 33741
+    },
+    {
+      "epoch": 0.8031514805293726,
+      "grad_norm": 21.527842442469474,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 33742
+    },
+    {
+      "epoch": 0.8031752832524041,
+      "grad_norm": 11.951108819012568,
+      "learning_rate": 2e-05,
+      "loss": 0.9072,
+      "step": 33743
+    },
+    {
+      "epoch": 0.8031990859754355,
+      "grad_norm": 2.368664871515179,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 33744
+    },
+    {
+      "epoch": 0.8032228886984671,
+      "grad_norm": 5.109556620695149,
+      "learning_rate": 2e-05,
+      "loss": 0.9055,
+      "step": 33745
+    },
+    {
+      "epoch": 0.8032466914214986,
+      "grad_norm": 0.0956281224910243,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 33746
+    },
+    {
+      "epoch": 0.8032704941445301,
+      "grad_norm": 6.0598427054214925,
+      "learning_rate": 2e-05,
+      "loss": 0.9333,
+      "step": 33747
+    },
+    {
+      "epoch": 0.8032942968675616,
+      "grad_norm": 0.5185623043784031,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 33748
+    },
+    {
+      "epoch": 0.8033180995905932,
+      "grad_norm": 5.324159322814449,
+      "learning_rate": 2e-05,
+      "loss": 0.7852,
+      "step": 33749
+    },
+    {
+      "epoch": 0.8033419023136247,
+      "grad_norm": 8.83369577031726,
+      "learning_rate": 2e-05,
+      "loss": 1.0729,
+      "step": 33750
+    },
+    {
+      "epoch": 0.8033657050366562,
+      "grad_norm": 6.2928502160555695,
+      "learning_rate": 2e-05,
+      "loss": 1.488,
+      "step": 33751
+    },
+    {
+      "epoch": 0.8033895077596878,
+      "grad_norm": 0.04783967050221465,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33752
+    },
+    {
+      "epoch": 0.8034133104827192,
+      "grad_norm": 3.146640427761469,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 33753
+    },
+    {
+      "epoch": 0.8034371132057507,
+      "grad_norm": 8.794284686588533,
+      "learning_rate": 2e-05,
+      "loss": 1.1693,
+      "step": 33754
+    },
+    {
+      "epoch": 0.8034609159287822,
+      "grad_norm": 0.038418253921110895,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33755
+    },
+    {
+      "epoch": 0.8034847186518138,
+      "grad_norm": 2.545937948046792,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 33756
+    },
+    {
+      "epoch": 0.8035085213748453,
+      "grad_norm": 3.8004858982491703,
+      "learning_rate": 2e-05,
+      "loss": 0.2599,
+      "step": 33757
+    },
+    {
+      "epoch": 0.8035323240978768,
+      "grad_norm": 3.6880988412731646,
+      "learning_rate": 2e-05,
+      "loss": 0.606,
+      "step": 33758
+    },
+    {
+      "epoch": 0.8035561268209083,
+      "grad_norm": 11.75421764494458,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 33759
+    },
+    {
+      "epoch": 0.8035799295439399,
+      "grad_norm": 4.660254185093219,
+      "learning_rate": 2e-05,
+      "loss": 0.8237,
+      "step": 33760
+    },
+    {
+      "epoch": 0.8036037322669713,
+      "grad_norm": 0.09071843485580294,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 33761
+    },
+    {
+      "epoch": 0.8036275349900028,
+      "grad_norm": 180.99292967378912,
+      "learning_rate": 2e-05,
+      "loss": 0.9215,
+      "step": 33762
+    },
+    {
+      "epoch": 0.8036513377130343,
+      "grad_norm": 3.2227506018346737,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 33763
+    },
+    {
+      "epoch": 0.8036751404360659,
+      "grad_norm": 20.23745212014087,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 33764
+    },
+    {
+      "epoch": 0.8036989431590974,
+      "grad_norm": 3.365265658834305,
+      "learning_rate": 2e-05,
+      "loss": 0.4476,
+      "step": 33765
+    },
+    {
+      "epoch": 0.8037227458821289,
+      "grad_norm": 4.004763413215044,
+      "learning_rate": 2e-05,
+      "loss": 0.4688,
+      "step": 33766
+    },
+    {
+      "epoch": 0.8037465486051605,
+      "grad_norm": 4.577098009342283,
+      "learning_rate": 2e-05,
+      "loss": 0.1029,
+      "step": 33767
+    },
+    {
+      "epoch": 0.803770351328192,
+      "grad_norm": 2.4849138132215054,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 33768
+    },
+    {
+      "epoch": 0.8037941540512235,
+      "grad_norm": 4.094098087517281,
+      "learning_rate": 2e-05,
+      "loss": 0.4217,
+      "step": 33769
+    },
+    {
+      "epoch": 0.8038179567742549,
+      "grad_norm": 3.1443250082489302,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 33770
+    },
+    {
+      "epoch": 0.8038417594972865,
+      "grad_norm": 6.876372959708657,
+      "learning_rate": 2e-05,
+      "loss": 0.3772,
+      "step": 33771
+    },
+    {
+      "epoch": 0.803865562220318,
+      "grad_norm": 1.5371722859890837,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 33772
+    },
+    {
+      "epoch": 0.8038893649433495,
+      "grad_norm": 7.50096574192317,
+      "learning_rate": 2e-05,
+      "loss": 0.3853,
+      "step": 33773
+    },
+    {
+      "epoch": 0.803913167666381,
+      "grad_norm": 10.552553014997747,
+      "learning_rate": 2e-05,
+      "loss": 2.789,
+      "step": 33774
+    },
+    {
+      "epoch": 0.8039369703894126,
+      "grad_norm": 5.822108165833237,
+      "learning_rate": 2e-05,
+      "loss": 0.8293,
+      "step": 33775
+    },
+    {
+      "epoch": 0.8039607731124441,
+      "grad_norm": 0.2326866448290615,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 33776
+    },
+    {
+      "epoch": 0.8039845758354756,
+      "grad_norm": 3.173265166590501,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 33777
+    },
+    {
+      "epoch": 0.804008378558507,
+      "grad_norm": 0.44399935115644096,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 33778
+    },
+    {
+      "epoch": 0.8040321812815386,
+      "grad_norm": 0.7925031494985693,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 33779
+    },
+    {
+      "epoch": 0.8040559840045701,
+      "grad_norm": 3.5590097242130163,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 33780
+    },
+    {
+      "epoch": 0.8040797867276016,
+      "grad_norm": 9.837413791229498,
+      "learning_rate": 2e-05,
+      "loss": 0.4766,
+      "step": 33781
+    },
+    {
+      "epoch": 0.8041035894506332,
+      "grad_norm": 3.1125040630796805,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 33782
+    },
+    {
+      "epoch": 0.8041273921736647,
+      "grad_norm": 0.6662876658231875,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 33783
+    },
+    {
+      "epoch": 0.8041511948966962,
+      "grad_norm": 9.183675448094291,
+      "learning_rate": 2e-05,
+      "loss": 0.571,
+      "step": 33784
+    },
+    {
+      "epoch": 0.8041749976197277,
+      "grad_norm": 0.04301163423037749,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 33785
+    },
+    {
+      "epoch": 0.8041988003427593,
+      "grad_norm": 9.2157144205498,
+      "learning_rate": 2e-05,
+      "loss": 0.2525,
+      "step": 33786
+    },
+    {
+      "epoch": 0.8042226030657907,
+      "grad_norm": 3.1460221333464777,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 33787
+    },
+    {
+      "epoch": 0.8042464057888222,
+      "grad_norm": 3.3328330298682425,
+      "learning_rate": 2e-05,
+      "loss": 0.6133,
+      "step": 33788
+    },
+    {
+      "epoch": 0.8042702085118537,
+      "grad_norm": 1.3480582304702915,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 33789
+    },
+    {
+      "epoch": 0.8042940112348853,
+      "grad_norm": 5.9556023661458495,
+      "learning_rate": 2e-05,
+      "loss": 0.5561,
+      "step": 33790
+    },
+    {
+      "epoch": 0.8043178139579168,
+      "grad_norm": 2.7790583316609014,
+      "learning_rate": 2e-05,
+      "loss": 0.2405,
+      "step": 33791
+    },
+    {
+      "epoch": 0.8043416166809483,
+      "grad_norm": 0.31314191896047644,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 33792
+    },
+    {
+      "epoch": 0.8043654194039798,
+      "grad_norm": 5.195510497138259,
+      "learning_rate": 2e-05,
+      "loss": 0.6321,
+      "step": 33793
+    },
+    {
+      "epoch": 0.8043892221270114,
+      "grad_norm": 0.07432809801759858,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33794
+    },
+    {
+      "epoch": 0.8044130248500428,
+      "grad_norm": 18.54946766318266,
+      "learning_rate": 2e-05,
+      "loss": 0.5699,
+      "step": 33795
+    },
+    {
+      "epoch": 0.8044368275730743,
+      "grad_norm": 1.118305946647095,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 33796
+    },
+    {
+      "epoch": 0.8044606302961059,
+      "grad_norm": 3.5519176062801048,
+      "learning_rate": 2e-05,
+      "loss": 0.519,
+      "step": 33797
+    },
+    {
+      "epoch": 0.8044844330191374,
+      "grad_norm": 7.629150114367718,
+      "learning_rate": 2e-05,
+      "loss": 0.3339,
+      "step": 33798
+    },
+    {
+      "epoch": 0.8045082357421689,
+      "grad_norm": 0.7013557500844231,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 33799
+    },
+    {
+      "epoch": 0.8045320384652004,
+      "grad_norm": 7.053634256031105,
+      "learning_rate": 2e-05,
+      "loss": 0.9452,
+      "step": 33800
+    },
+    {
+      "epoch": 0.804555841188232,
+      "grad_norm": 3.596240064531142,
+      "learning_rate": 2e-05,
+      "loss": 0.6829,
+      "step": 33801
+    },
+    {
+      "epoch": 0.8045796439112635,
+      "grad_norm": 1.899904327190413,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 33802
+    },
+    {
+      "epoch": 0.804603446634295,
+      "grad_norm": 10.529529451947457,
+      "learning_rate": 2e-05,
+      "loss": 1.6422,
+      "step": 33803
+    },
+    {
+      "epoch": 0.8046272493573264,
+      "grad_norm": 4.340313606138707,
+      "learning_rate": 2e-05,
+      "loss": 0.5696,
+      "step": 33804
+    },
+    {
+      "epoch": 0.804651052080358,
+      "grad_norm": 0.0828465576116325,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 33805
+    },
+    {
+      "epoch": 0.8046748548033895,
+      "grad_norm": 0.03910093340392248,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 33806
+    },
+    {
+      "epoch": 0.804698657526421,
+      "grad_norm": 2.7853181620633687,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 33807
+    },
+    {
+      "epoch": 0.8047224602494525,
+      "grad_norm": 5.00150554715162,
+      "learning_rate": 2e-05,
+      "loss": 1.0264,
+      "step": 33808
+    },
+    {
+      "epoch": 0.8047462629724841,
+      "grad_norm": 3.318757237735967,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 33809
+    },
+    {
+      "epoch": 0.8047700656955156,
+      "grad_norm": 4.331111255664291,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 33810
+    },
+    {
+      "epoch": 0.804793868418547,
+      "grad_norm": 3.8089823145810398,
+      "learning_rate": 2e-05,
+      "loss": 0.7785,
+      "step": 33811
+    },
+    {
+      "epoch": 0.8048176711415786,
+      "grad_norm": 3.3927291022185337,
+      "learning_rate": 2e-05,
+      "loss": 0.4345,
+      "step": 33812
+    },
+    {
+      "epoch": 0.8048414738646101,
+      "grad_norm": 0.14145788262390413,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 33813
+    },
+    {
+      "epoch": 0.8048652765876416,
+      "grad_norm": 0.6090927750217648,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 33814
+    },
+    {
+      "epoch": 0.8048890793106731,
+      "grad_norm": 4.310743454447755,
+      "learning_rate": 2e-05,
+      "loss": 1.3147,
+      "step": 33815
+    },
+    {
+      "epoch": 0.8049128820337047,
+      "grad_norm": 8.490488498447576,
+      "learning_rate": 2e-05,
+      "loss": 1.0738,
+      "step": 33816
+    },
+    {
+      "epoch": 0.8049366847567362,
+      "grad_norm": 1.7291729022099342,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 33817
+    },
+    {
+      "epoch": 0.8049604874797677,
+      "grad_norm": 5.43911914598258,
+      "learning_rate": 2e-05,
+      "loss": 1.6379,
+      "step": 33818
+    },
+    {
+      "epoch": 0.8049842902027992,
+      "grad_norm": 1.714649999777831,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 33819
+    },
+    {
+      "epoch": 0.8050080929258308,
+      "grad_norm": 21.027263205769202,
+      "learning_rate": 2e-05,
+      "loss": 0.9909,
+      "step": 33820
+    },
+    {
+      "epoch": 0.8050318956488622,
+      "grad_norm": 3.0167507378240366,
+      "learning_rate": 2e-05,
+      "loss": 0.275,
+      "step": 33821
+    },
+    {
+      "epoch": 0.8050556983718937,
+      "grad_norm": 8.51172582714588,
+      "learning_rate": 2e-05,
+      "loss": 0.3184,
+      "step": 33822
+    },
+    {
+      "epoch": 0.8050795010949252,
+      "grad_norm": 2.172976139650152,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 33823
+    },
+    {
+      "epoch": 0.8051033038179568,
+      "grad_norm": 0.30374614965588653,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 33824
+    },
+    {
+      "epoch": 0.8051271065409883,
+      "grad_norm": 7.00580348073347,
+      "learning_rate": 2e-05,
+      "loss": 1.0163,
+      "step": 33825
+    },
+    {
+      "epoch": 0.8051509092640198,
+      "grad_norm": 3.352578030132573,
+      "learning_rate": 2e-05,
+      "loss": 0.1455,
+      "step": 33826
+    },
+    {
+      "epoch": 0.8051747119870514,
+      "grad_norm": 2.5091498610026712,
+      "learning_rate": 2e-05,
+      "loss": 0.5265,
+      "step": 33827
+    },
+    {
+      "epoch": 0.8051985147100829,
+      "grad_norm": 3.386512838499168,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 33828
+    },
+    {
+      "epoch": 0.8052223174331143,
+      "grad_norm": 1.0821359362753975,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 33829
+    },
+    {
+      "epoch": 0.8052461201561458,
+      "grad_norm": 0.08574311286386722,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 33830
+    },
+    {
+      "epoch": 0.8052699228791774,
+      "grad_norm": 3.355914188624314,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 33831
+    },
+    {
+      "epoch": 0.8052937256022089,
+      "grad_norm": 2.447190064827669,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 33832
+    },
+    {
+      "epoch": 0.8053175283252404,
+      "grad_norm": 7.711610506391589,
+      "learning_rate": 2e-05,
+      "loss": 0.322,
+      "step": 33833
+    },
+    {
+      "epoch": 0.8053413310482719,
+      "grad_norm": 3.443468672867369,
+      "learning_rate": 2e-05,
+      "loss": 0.561,
+      "step": 33834
+    },
+    {
+      "epoch": 0.8053651337713035,
+      "grad_norm": 0.3011977501222932,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 33835
+    },
+    {
+      "epoch": 0.805388936494335,
+      "grad_norm": 3.30586762398446,
+      "learning_rate": 2e-05,
+      "loss": 0.449,
+      "step": 33836
+    },
+    {
+      "epoch": 0.8054127392173664,
+      "grad_norm": 5.517411224671676,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 33837
+    },
+    {
+      "epoch": 0.8054365419403979,
+      "grad_norm": 1.341409639685263,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 33838
+    },
+    {
+      "epoch": 0.8054603446634295,
+      "grad_norm": 0.9234860723207116,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 33839
+    },
+    {
+      "epoch": 0.805484147386461,
+      "grad_norm": 4.4146150774027415,
+      "learning_rate": 2e-05,
+      "loss": 0.6808,
+      "step": 33840
+    },
+    {
+      "epoch": 0.8055079501094925,
+      "grad_norm": 3.6278715988063053,
+      "learning_rate": 2e-05,
+      "loss": 0.6297,
+      "step": 33841
+    },
+    {
+      "epoch": 0.8055317528325241,
+      "grad_norm": 0.21241736381165238,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 33842
+    },
+    {
+      "epoch": 0.8055555555555556,
+      "grad_norm": 0.03730769150370541,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 33843
+    },
+    {
+      "epoch": 0.8055793582785871,
+      "grad_norm": 9.061596674795442,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 33844
+    },
+    {
+      "epoch": 0.8056031610016186,
+      "grad_norm": 1.071547260767459,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 33845
+    },
+    {
+      "epoch": 0.8056269637246501,
+      "grad_norm": 0.4022146572797071,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 33846
+    },
+    {
+      "epoch": 0.8056507664476816,
+      "grad_norm": 4.461205421833476,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 33847
+    },
+    {
+      "epoch": 0.8056745691707131,
+      "grad_norm": 1.9337577823341674,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 33848
+    },
+    {
+      "epoch": 0.8056983718937446,
+      "grad_norm": 5.342689038695118,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 33849
+    },
+    {
+      "epoch": 0.8057221746167762,
+      "grad_norm": 10.081448492401186,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 33850
+    },
+    {
+      "epoch": 0.8057459773398077,
+      "grad_norm": 8.440030222171423,
+      "learning_rate": 2e-05,
+      "loss": 0.4863,
+      "step": 33851
+    },
+    {
+      "epoch": 0.8057697800628392,
+      "grad_norm": 0.050052783306684896,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 33852
+    },
+    {
+      "epoch": 0.8057935827858707,
+      "grad_norm": 0.12293891944464744,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 33853
+    },
+    {
+      "epoch": 0.8058173855089023,
+      "grad_norm": 2.1119715431378694,
+      "learning_rate": 2e-05,
+      "loss": 0.3199,
+      "step": 33854
+    },
+    {
+      "epoch": 0.8058411882319337,
+      "grad_norm": 3.162003565981948,
+      "learning_rate": 2e-05,
+      "loss": 0.4078,
+      "step": 33855
+    },
+    {
+      "epoch": 0.8058649909549652,
+      "grad_norm": 33.512022645735534,
+      "learning_rate": 2e-05,
+      "loss": 1.0577,
+      "step": 33856
+    },
+    {
+      "epoch": 0.8058887936779968,
+      "grad_norm": 2.5528555011795757,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 33857
+    },
+    {
+      "epoch": 0.8059125964010283,
+      "grad_norm": 9.99421360040987,
+      "learning_rate": 2e-05,
+      "loss": 0.8053,
+      "step": 33858
+    },
+    {
+      "epoch": 0.8059363991240598,
+      "grad_norm": 2.747426503060102,
+      "learning_rate": 2e-05,
+      "loss": 0.1299,
+      "step": 33859
+    },
+    {
+      "epoch": 0.8059602018470913,
+      "grad_norm": 8.513748565787694,
+      "learning_rate": 2e-05,
+      "loss": 0.1665,
+      "step": 33860
+    },
+    {
+      "epoch": 0.8059840045701229,
+      "grad_norm": 3.587917890950441,
+      "learning_rate": 2e-05,
+      "loss": 0.4331,
+      "step": 33861
+    },
+    {
+      "epoch": 0.8060078072931544,
+      "grad_norm": 1.0051891394372527,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 33862
+    },
+    {
+      "epoch": 0.8060316100161858,
+      "grad_norm": 8.405453521456899,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 33863
+    },
+    {
+      "epoch": 0.8060554127392173,
+      "grad_norm": 19.130601488060922,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 33864
+    },
+    {
+      "epoch": 0.8060792154622489,
+      "grad_norm": 5.339645339531851,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 33865
+    },
+    {
+      "epoch": 0.8061030181852804,
+      "grad_norm": 0.07945342911824861,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33866
+    },
+    {
+      "epoch": 0.8061268209083119,
+      "grad_norm": 0.16366864008870813,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 33867
+    },
+    {
+      "epoch": 0.8061506236313434,
+      "grad_norm": 2.9346785291831745,
+      "learning_rate": 2e-05,
+      "loss": 0.4155,
+      "step": 33868
+    },
+    {
+      "epoch": 0.806174426354375,
+      "grad_norm": 5.704254635019863,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 33869
+    },
+    {
+      "epoch": 0.8061982290774065,
+      "grad_norm": 3.856536175783864,
+      "learning_rate": 2e-05,
+      "loss": 0.3821,
+      "step": 33870
+    },
+    {
+      "epoch": 0.8062220318004379,
+      "grad_norm": 3.220177932546921,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 33871
+    },
+    {
+      "epoch": 0.8062458345234695,
+      "grad_norm": 2.2722666289943287,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 33872
+    },
+    {
+      "epoch": 0.806269637246501,
+      "grad_norm": 12.039605129797689,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 33873
+    },
+    {
+      "epoch": 0.8062934399695325,
+      "grad_norm": 5.0954947883913535,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 33874
+    },
+    {
+      "epoch": 0.806317242692564,
+      "grad_norm": 0.15749683212090962,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 33875
+    },
+    {
+      "epoch": 0.8063410454155956,
+      "grad_norm": 5.138505178900778,
+      "learning_rate": 2e-05,
+      "loss": 0.0552,
+      "step": 33876
+    },
+    {
+      "epoch": 0.8063648481386271,
+      "grad_norm": 0.07937046867681136,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 33877
+    },
+    {
+      "epoch": 0.8063886508616586,
+      "grad_norm": 1.330038470869531,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 33878
+    },
+    {
+      "epoch": 0.80641245358469,
+      "grad_norm": 3.084193652595848,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 33879
+    },
+    {
+      "epoch": 0.8064362563077216,
+      "grad_norm": 0.017427944688388645,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 33880
+    },
+    {
+      "epoch": 0.8064600590307531,
+      "grad_norm": 1.9424816763094679,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 33881
+    },
+    {
+      "epoch": 0.8064838617537846,
+      "grad_norm": 0.165906276828639,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33882
+    },
+    {
+      "epoch": 0.8065076644768161,
+      "grad_norm": 0.45479872033704805,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 33883
+    },
+    {
+      "epoch": 0.8065314671998477,
+      "grad_norm": 0.5123569250125912,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 33884
+    },
+    {
+      "epoch": 0.8065552699228792,
+      "grad_norm": 6.936062955977548,
+      "learning_rate": 2e-05,
+      "loss": 0.8548,
+      "step": 33885
+    },
+    {
+      "epoch": 0.8065790726459107,
+      "grad_norm": 2.93095757617533,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 33886
+    },
+    {
+      "epoch": 0.8066028753689422,
+      "grad_norm": 2.6445804222398057,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 33887
+    },
+    {
+      "epoch": 0.8066266780919737,
+      "grad_norm": 1.8460784830677057,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 33888
+    },
+    {
+      "epoch": 0.8066504808150052,
+      "grad_norm": 0.3966880642893735,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 33889
+    },
+    {
+      "epoch": 0.8066742835380367,
+      "grad_norm": 7.1200139084330525,
+      "learning_rate": 2e-05,
+      "loss": 0.3501,
+      "step": 33890
+    },
+    {
+      "epoch": 0.8066980862610683,
+      "grad_norm": 0.2615668643709073,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 33891
+    },
+    {
+      "epoch": 0.8067218889840998,
+      "grad_norm": 0.19465666150664418,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33892
+    },
+    {
+      "epoch": 0.8067456917071313,
+      "grad_norm": 2.790591473006327,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 33893
+    },
+    {
+      "epoch": 0.8067694944301628,
+      "grad_norm": 2.1866330713510016,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 33894
+    },
+    {
+      "epoch": 0.8067932971531944,
+      "grad_norm": 0.04693556489419471,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 33895
+    },
+    {
+      "epoch": 0.8068170998762259,
+      "grad_norm": 12.086980360530744,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 33896
+    },
+    {
+      "epoch": 0.8068409025992573,
+      "grad_norm": 4.49395723714437,
+      "learning_rate": 2e-05,
+      "loss": 0.6055,
+      "step": 33897
+    },
+    {
+      "epoch": 0.8068647053222888,
+      "grad_norm": 1.6938693083526266,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 33898
+    },
+    {
+      "epoch": 0.8068885080453204,
+      "grad_norm": 0.08455434125618784,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 33899
+    },
+    {
+      "epoch": 0.8069123107683519,
+      "grad_norm": 23.027751442591775,
+      "learning_rate": 2e-05,
+      "loss": 0.7411,
+      "step": 33900
+    },
+    {
+      "epoch": 0.8069361134913834,
+      "grad_norm": 2.718420108174245,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 33901
+    },
+    {
+      "epoch": 0.8069599162144149,
+      "grad_norm": 3.5234316561299863,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 33902
+    },
+    {
+      "epoch": 0.8069837189374465,
+      "grad_norm": 4.6373973501954895,
+      "learning_rate": 2e-05,
+      "loss": 0.8327,
+      "step": 33903
+    },
+    {
+      "epoch": 0.807007521660478,
+      "grad_norm": 9.486816395791845,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 33904
+    },
+    {
+      "epoch": 0.8070313243835094,
+      "grad_norm": 2.6495845148553934,
+      "learning_rate": 2e-05,
+      "loss": 0.3395,
+      "step": 33905
+    },
+    {
+      "epoch": 0.807055127106541,
+      "grad_norm": 2.5861049227177304,
+      "learning_rate": 2e-05,
+      "loss": 0.2337,
+      "step": 33906
+    },
+    {
+      "epoch": 0.8070789298295725,
+      "grad_norm": 0.24116474295197934,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33907
+    },
+    {
+      "epoch": 0.807102732552604,
+      "grad_norm": 5.240997343884154,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 33908
+    },
+    {
+      "epoch": 0.8071265352756355,
+      "grad_norm": 6.986787831160797,
+      "learning_rate": 2e-05,
+      "loss": 0.8915,
+      "step": 33909
+    },
+    {
+      "epoch": 0.8071503379986671,
+      "grad_norm": 3.8851054550152138,
+      "learning_rate": 2e-05,
+      "loss": 0.7994,
+      "step": 33910
+    },
+    {
+      "epoch": 0.8071741407216986,
+      "grad_norm": 3.2942334673068343,
+      "learning_rate": 2e-05,
+      "loss": 0.2788,
+      "step": 33911
+    },
+    {
+      "epoch": 0.8071979434447301,
+      "grad_norm": 1.238686913307498,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 33912
+    },
+    {
+      "epoch": 0.8072217461677615,
+      "grad_norm": 8.79309653429347,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 33913
+    },
+    {
+      "epoch": 0.8072455488907931,
+      "grad_norm": 9.11334148591277,
+      "learning_rate": 2e-05,
+      "loss": 1.3693,
+      "step": 33914
+    },
+    {
+      "epoch": 0.8072693516138246,
+      "grad_norm": 2.2360870125402577,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 33915
+    },
+    {
+      "epoch": 0.8072931543368561,
+      "grad_norm": 4.71987557480163,
+      "learning_rate": 2e-05,
+      "loss": 0.8835,
+      "step": 33916
+    },
+    {
+      "epoch": 0.8073169570598876,
+      "grad_norm": 0.19082868196097016,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 33917
+    },
+    {
+      "epoch": 0.8073407597829192,
+      "grad_norm": 4.620909413432674,
+      "learning_rate": 2e-05,
+      "loss": 1.0202,
+      "step": 33918
+    },
+    {
+      "epoch": 0.8073645625059507,
+      "grad_norm": 4.898227592040387,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 33919
+    },
+    {
+      "epoch": 0.8073883652289822,
+      "grad_norm": 0.9810900090047242,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 33920
+    },
+    {
+      "epoch": 0.8074121679520138,
+      "grad_norm": 5.615373183640655,
+      "learning_rate": 2e-05,
+      "loss": 0.9706,
+      "step": 33921
+    },
+    {
+      "epoch": 0.8074359706750452,
+      "grad_norm": 12.4525525771923,
+      "learning_rate": 2e-05,
+      "loss": 0.4192,
+      "step": 33922
+    },
+    {
+      "epoch": 0.8074597733980767,
+      "grad_norm": 5.016065962227514,
+      "learning_rate": 2e-05,
+      "loss": 0.9132,
+      "step": 33923
+    },
+    {
+      "epoch": 0.8074835761211082,
+      "grad_norm": 3.352437903692939,
+      "learning_rate": 2e-05,
+      "loss": 0.8817,
+      "step": 33924
+    },
+    {
+      "epoch": 0.8075073788441398,
+      "grad_norm": 2.200688477200056,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 33925
+    },
+    {
+      "epoch": 0.8075311815671713,
+      "grad_norm": 1.4621238364273585,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 33926
+    },
+    {
+      "epoch": 0.8075549842902028,
+      "grad_norm": 26.265999662739716,
+      "learning_rate": 2e-05,
+      "loss": 0.3893,
+      "step": 33927
+    },
+    {
+      "epoch": 0.8075787870132343,
+      "grad_norm": 0.7677774831952466,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 33928
+    },
+    {
+      "epoch": 0.8076025897362659,
+      "grad_norm": 0.06751644493429103,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 33929
+    },
+    {
+      "epoch": 0.8076263924592973,
+      "grad_norm": 4.4186327631328295,
+      "learning_rate": 2e-05,
+      "loss": 0.1872,
+      "step": 33930
+    },
+    {
+      "epoch": 0.8076501951823288,
+      "grad_norm": 4.725256932289909,
+      "learning_rate": 2e-05,
+      "loss": 0.2808,
+      "step": 33931
+    },
+    {
+      "epoch": 0.8076739979053603,
+      "grad_norm": 0.8772192914412047,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 33932
+    },
+    {
+      "epoch": 0.8076978006283919,
+      "grad_norm": 3.650586259389075,
+      "learning_rate": 2e-05,
+      "loss": 0.6045,
+      "step": 33933
+    },
+    {
+      "epoch": 0.8077216033514234,
+      "grad_norm": 0.16263428629769866,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 33934
+    },
+    {
+      "epoch": 0.8077454060744549,
+      "grad_norm": 0.3214832196196859,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 33935
+    },
+    {
+      "epoch": 0.8077692087974865,
+      "grad_norm": 3.5571564147519372,
+      "learning_rate": 2e-05,
+      "loss": 0.5343,
+      "step": 33936
+    },
+    {
+      "epoch": 0.807793011520518,
+      "grad_norm": 5.167846151878024,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 33937
+    },
+    {
+      "epoch": 0.8078168142435495,
+      "grad_norm": 9.554905439871847,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 33938
+    },
+    {
+      "epoch": 0.8078406169665809,
+      "grad_norm": 0.799944243825404,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 33939
+    },
+    {
+      "epoch": 0.8078644196896125,
+      "grad_norm": 3.977351682381729,
+      "learning_rate": 2e-05,
+      "loss": 0.4826,
+      "step": 33940
+    },
+    {
+      "epoch": 0.807888222412644,
+      "grad_norm": 2.7207243915590085,
+      "learning_rate": 2e-05,
+      "loss": 0.5768,
+      "step": 33941
+    },
+    {
+      "epoch": 0.8079120251356755,
+      "grad_norm": 1.99136090360398,
+      "learning_rate": 2e-05,
+      "loss": 0.1663,
+      "step": 33942
+    },
+    {
+      "epoch": 0.807935827858707,
+      "grad_norm": 7.406747313084935,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 33943
+    },
+    {
+      "epoch": 0.8079596305817386,
+      "grad_norm": 0.1886642842182053,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 33944
+    },
+    {
+      "epoch": 0.8079834333047701,
+      "grad_norm": 0.7649378663733202,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 33945
+    },
+    {
+      "epoch": 0.8080072360278016,
+      "grad_norm": 2.891019723579093,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 33946
+    },
+    {
+      "epoch": 0.808031038750833,
+      "grad_norm": 2.9482836575619604,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 33947
+    },
+    {
+      "epoch": 0.8080548414738646,
+      "grad_norm": 9.524111666084565,
+      "learning_rate": 2e-05,
+      "loss": 1.2392,
+      "step": 33948
+    },
+    {
+      "epoch": 0.8080786441968961,
+      "grad_norm": 4.075891427380811,
+      "learning_rate": 2e-05,
+      "loss": 0.5433,
+      "step": 33949
+    },
+    {
+      "epoch": 0.8081024469199276,
+      "grad_norm": 13.747562713264488,
+      "learning_rate": 2e-05,
+      "loss": 1.474,
+      "step": 33950
+    },
+    {
+      "epoch": 0.8081262496429592,
+      "grad_norm": 0.02074267226114956,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 33951
+    },
+    {
+      "epoch": 0.8081500523659907,
+      "grad_norm": 5.0153101603500865,
+      "learning_rate": 2e-05,
+      "loss": 1.0036,
+      "step": 33952
+    },
+    {
+      "epoch": 0.8081738550890222,
+      "grad_norm": 2.404279850846575,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 33953
+    },
+    {
+      "epoch": 0.8081976578120537,
+      "grad_norm": 7.866342640603503,
+      "learning_rate": 2e-05,
+      "loss": 0.6965,
+      "step": 33954
+    },
+    {
+      "epoch": 0.8082214605350853,
+      "grad_norm": 10.242970017088608,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 33955
+    },
+    {
+      "epoch": 0.8082452632581167,
+      "grad_norm": 0.2384756285299172,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 33956
+    },
+    {
+      "epoch": 0.8082690659811482,
+      "grad_norm": 0.3061416399804477,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 33957
+    },
+    {
+      "epoch": 0.8082928687041797,
+      "grad_norm": 8.905168274443103,
+      "learning_rate": 2e-05,
+      "loss": 0.7579,
+      "step": 33958
+    },
+    {
+      "epoch": 0.8083166714272113,
+      "grad_norm": 0.41475682879508585,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 33959
+    },
+    {
+      "epoch": 0.8083404741502428,
+      "grad_norm": 0.558875341787124,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 33960
+    },
+    {
+      "epoch": 0.8083642768732743,
+      "grad_norm": 5.5459638894241206,
+      "learning_rate": 2e-05,
+      "loss": 0.6935,
+      "step": 33961
+    },
+    {
+      "epoch": 0.8083880795963058,
+      "grad_norm": 0.012452963321750133,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 33962
+    },
+    {
+      "epoch": 0.8084118823193374,
+      "grad_norm": 8.32982618154762,
+      "learning_rate": 2e-05,
+      "loss": 0.3022,
+      "step": 33963
+    },
+    {
+      "epoch": 0.8084356850423688,
+      "grad_norm": 0.37964725380767694,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 33964
+    },
+    {
+      "epoch": 0.8084594877654003,
+      "grad_norm": 26.78942700299698,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 33965
+    },
+    {
+      "epoch": 0.8084832904884319,
+      "grad_norm": 7.56443599072194,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 33966
+    },
+    {
+      "epoch": 0.8085070932114634,
+      "grad_norm": 9.589524752595828,
+      "learning_rate": 2e-05,
+      "loss": 0.4495,
+      "step": 33967
+    },
+    {
+      "epoch": 0.8085308959344949,
+      "grad_norm": 5.1333729083544,
+      "learning_rate": 2e-05,
+      "loss": 0.7927,
+      "step": 33968
+    },
+    {
+      "epoch": 0.8085546986575264,
+      "grad_norm": 5.5513244279077,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 33969
+    },
+    {
+      "epoch": 0.808578501380558,
+      "grad_norm": 3.949006591260855,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 33970
+    },
+    {
+      "epoch": 0.8086023041035895,
+      "grad_norm": 3.0223821893314793,
+      "learning_rate": 2e-05,
+      "loss": 0.7034,
+      "step": 33971
+    },
+    {
+      "epoch": 0.808626106826621,
+      "grad_norm": 2.553050349864669,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 33972
+    },
+    {
+      "epoch": 0.8086499095496524,
+      "grad_norm": 0.3198088699748427,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 33973
+    },
+    {
+      "epoch": 0.808673712272684,
+      "grad_norm": 0.6990719595729318,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 33974
+    },
+    {
+      "epoch": 0.8086975149957155,
+      "grad_norm": 1.514705796915674,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 33975
+    },
+    {
+      "epoch": 0.808721317718747,
+      "grad_norm": 2.3830533287655875,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 33976
+    },
+    {
+      "epoch": 0.8087451204417785,
+      "grad_norm": 5.036783384419453,
+      "learning_rate": 2e-05,
+      "loss": 0.9149,
+      "step": 33977
+    },
+    {
+      "epoch": 0.8087689231648101,
+      "grad_norm": 0.5104770807914336,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 33978
+    },
+    {
+      "epoch": 0.8087927258878416,
+      "grad_norm": 2.6834350682462045,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 33979
+    },
+    {
+      "epoch": 0.808816528610873,
+      "grad_norm": 5.92658856425542,
+      "learning_rate": 2e-05,
+      "loss": 0.2259,
+      "step": 33980
+    },
+    {
+      "epoch": 0.8088403313339046,
+      "grad_norm": 0.5899570868674446,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 33981
+    },
+    {
+      "epoch": 0.8088641340569361,
+      "grad_norm": 4.9979581551134995,
+      "learning_rate": 2e-05,
+      "loss": 0.7324,
+      "step": 33982
+    },
+    {
+      "epoch": 0.8088879367799676,
+      "grad_norm": 6.272782884660351,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 33983
+    },
+    {
+      "epoch": 0.8089117395029991,
+      "grad_norm": 8.272757826462325,
+      "learning_rate": 2e-05,
+      "loss": 0.174,
+      "step": 33984
+    },
+    {
+      "epoch": 0.8089355422260307,
+      "grad_norm": 7.941226035346584,
+      "learning_rate": 2e-05,
+      "loss": 0.4256,
+      "step": 33985
+    },
+    {
+      "epoch": 0.8089593449490622,
+      "grad_norm": 0.29368795143624066,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 33986
+    },
+    {
+      "epoch": 0.8089831476720937,
+      "grad_norm": 5.365886503918524,
+      "learning_rate": 2e-05,
+      "loss": 1.1024,
+      "step": 33987
+    },
+    {
+      "epoch": 0.8090069503951252,
+      "grad_norm": 1.3206932064721493,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 33988
+    },
+    {
+      "epoch": 0.8090307531181568,
+      "grad_norm": 0.014270637269014567,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 33989
+    },
+    {
+      "epoch": 0.8090545558411882,
+      "grad_norm": 3.3699901791324773,
+      "learning_rate": 2e-05,
+      "loss": 0.8283,
+      "step": 33990
+    },
+    {
+      "epoch": 0.8090783585642197,
+      "grad_norm": 4.573958215999616,
+      "learning_rate": 2e-05,
+      "loss": 0.7541,
+      "step": 33991
+    },
+    {
+      "epoch": 0.8091021612872512,
+      "grad_norm": 3.177898594578394,
+      "learning_rate": 2e-05,
+      "loss": 0.5847,
+      "step": 33992
+    },
+    {
+      "epoch": 0.8091259640102828,
+      "grad_norm": 4.977366418370283,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 33993
+    },
+    {
+      "epoch": 0.8091497667333143,
+      "grad_norm": 6.043514698217574,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 33994
+    },
+    {
+      "epoch": 0.8091735694563458,
+      "grad_norm": 2.4974668113438114,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 33995
+    },
+    {
+      "epoch": 0.8091973721793774,
+      "grad_norm": 3.412576753438483,
+      "learning_rate": 2e-05,
+      "loss": 0.3966,
+      "step": 33996
+    },
+    {
+      "epoch": 0.8092211749024089,
+      "grad_norm": 0.10150532634766686,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 33997
+    },
+    {
+      "epoch": 0.8092449776254403,
+      "grad_norm": 2.140848522392339,
+      "learning_rate": 2e-05,
+      "loss": 0.1708,
+      "step": 33998
+    },
+    {
+      "epoch": 0.8092687803484718,
+      "grad_norm": 4.313145069832436,
+      "learning_rate": 2e-05,
+      "loss": 0.6139,
+      "step": 33999
+    },
+    {
+      "epoch": 0.8092925830715034,
+      "grad_norm": 6.247708982486915,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 34000
+    },
+    {
+      "epoch": 0.8093163857945349,
+      "grad_norm": 2.5629293511597533,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 34001
+    },
+    {
+      "epoch": 0.8093401885175664,
+      "grad_norm": 4.974795457077129,
+      "learning_rate": 2e-05,
+      "loss": 1.3724,
+      "step": 34002
+    },
+    {
+      "epoch": 0.8093639912405979,
+      "grad_norm": 4.6991152234948315,
+      "learning_rate": 2e-05,
+      "loss": 1.1363,
+      "step": 34003
+    },
+    {
+      "epoch": 0.8093877939636295,
+      "grad_norm": 5.848133939078279,
+      "learning_rate": 2e-05,
+      "loss": 0.1285,
+      "step": 34004
+    },
+    {
+      "epoch": 0.809411596686661,
+      "grad_norm": 4.316302691685328,
+      "learning_rate": 2e-05,
+      "loss": 0.4954,
+      "step": 34005
+    },
+    {
+      "epoch": 0.8094353994096924,
+      "grad_norm": 1.2555601856510723,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 34006
+    },
+    {
+      "epoch": 0.8094592021327239,
+      "grad_norm": 3.1111903247740447,
+      "learning_rate": 2e-05,
+      "loss": 0.0902,
+      "step": 34007
+    },
+    {
+      "epoch": 0.8094830048557555,
+      "grad_norm": 2.9589964834716005,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 34008
+    },
+    {
+      "epoch": 0.809506807578787,
+      "grad_norm": 0.09639430160475954,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 34009
+    },
+    {
+      "epoch": 0.8095306103018185,
+      "grad_norm": 3.4778854442100173,
+      "learning_rate": 2e-05,
+      "loss": 0.6645,
+      "step": 34010
+    },
+    {
+      "epoch": 0.8095544130248501,
+      "grad_norm": 7.623098081839528,
+      "learning_rate": 2e-05,
+      "loss": 0.9307,
+      "step": 34011
+    },
+    {
+      "epoch": 0.8095782157478816,
+      "grad_norm": 0.7754005813460254,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 34012
+    },
+    {
+      "epoch": 0.8096020184709131,
+      "grad_norm": 5.369322965896262,
+      "learning_rate": 2e-05,
+      "loss": 0.8436,
+      "step": 34013
+    },
+    {
+      "epoch": 0.8096258211939446,
+      "grad_norm": 0.20998333006206474,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 34014
+    },
+    {
+      "epoch": 0.8096496239169761,
+      "grad_norm": 3.518895967755932,
+      "learning_rate": 2e-05,
+      "loss": 0.6235,
+      "step": 34015
+    },
+    {
+      "epoch": 0.8096734266400076,
+      "grad_norm": 2.0040729862631474,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 34016
+    },
+    {
+      "epoch": 0.8096972293630391,
+      "grad_norm": 5.807350775190033,
+      "learning_rate": 2e-05,
+      "loss": 0.5012,
+      "step": 34017
+    },
+    {
+      "epoch": 0.8097210320860706,
+      "grad_norm": 3.118649947116817,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 34018
+    },
+    {
+      "epoch": 0.8097448348091022,
+      "grad_norm": 14.336173513906612,
+      "learning_rate": 2e-05,
+      "loss": 1.1009,
+      "step": 34019
+    },
+    {
+      "epoch": 0.8097686375321337,
+      "grad_norm": 2.815623243914085,
+      "learning_rate": 2e-05,
+      "loss": 0.3924,
+      "step": 34020
+    },
+    {
+      "epoch": 0.8097924402551652,
+      "grad_norm": 7.262577893531332,
+      "learning_rate": 2e-05,
+      "loss": 0.6698,
+      "step": 34021
+    },
+    {
+      "epoch": 0.8098162429781967,
+      "grad_norm": 2.658715890521709,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 34022
+    },
+    {
+      "epoch": 0.8098400457012283,
+      "grad_norm": 2.632229040242584,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 34023
+    },
+    {
+      "epoch": 0.8098638484242597,
+      "grad_norm": 3.1559595390521706,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 34024
+    },
+    {
+      "epoch": 0.8098876511472912,
+      "grad_norm": 5.218864941717926,
+      "learning_rate": 2e-05,
+      "loss": 0.635,
+      "step": 34025
+    },
+    {
+      "epoch": 0.8099114538703228,
+      "grad_norm": 8.39556316454214,
+      "learning_rate": 2e-05,
+      "loss": 0.8277,
+      "step": 34026
+    },
+    {
+      "epoch": 0.8099352565933543,
+      "grad_norm": 8.129909136537389,
+      "learning_rate": 2e-05,
+      "loss": 1.3919,
+      "step": 34027
+    },
+    {
+      "epoch": 0.8099590593163858,
+      "grad_norm": 5.813389998430845,
+      "learning_rate": 2e-05,
+      "loss": 0.9843,
+      "step": 34028
+    },
+    {
+      "epoch": 0.8099828620394173,
+      "grad_norm": 5.241531772602375,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 34029
+    },
+    {
+      "epoch": 0.8100066647624489,
+      "grad_norm": 4.2900800351835375,
+      "learning_rate": 2e-05,
+      "loss": 0.8758,
+      "step": 34030
+    },
+    {
+      "epoch": 0.8100304674854804,
+      "grad_norm": 2.2844024553184785,
+      "learning_rate": 2e-05,
+      "loss": 0.1125,
+      "step": 34031
+    },
+    {
+      "epoch": 0.8100542702085118,
+      "grad_norm": 2.9407432689944204,
+      "learning_rate": 2e-05,
+      "loss": 0.6918,
+      "step": 34032
+    },
+    {
+      "epoch": 0.8100780729315433,
+      "grad_norm": 4.732434931953682,
+      "learning_rate": 2e-05,
+      "loss": 1.3315,
+      "step": 34033
+    },
+    {
+      "epoch": 0.8101018756545749,
+      "grad_norm": 7.520567464410054,
+      "learning_rate": 2e-05,
+      "loss": 0.5259,
+      "step": 34034
+    },
+    {
+      "epoch": 0.8101256783776064,
+      "grad_norm": 0.8126003055203949,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 34035
+    },
+    {
+      "epoch": 0.8101494811006379,
+      "grad_norm": 1.491386273685446,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 34036
+    },
+    {
+      "epoch": 0.8101732838236694,
+      "grad_norm": 3.9734484065276714,
+      "learning_rate": 2e-05,
+      "loss": 0.1192,
+      "step": 34037
+    },
+    {
+      "epoch": 0.810197086546701,
+      "grad_norm": 5.648714367301637,
+      "learning_rate": 2e-05,
+      "loss": 0.8914,
+      "step": 34038
+    },
+    {
+      "epoch": 0.8102208892697325,
+      "grad_norm": 0.22973660957428557,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 34039
+    },
+    {
+      "epoch": 0.8102446919927639,
+      "grad_norm": 11.181080300157147,
+      "learning_rate": 2e-05,
+      "loss": 1.0524,
+      "step": 34040
+    },
+    {
+      "epoch": 0.8102684947157955,
+      "grad_norm": 1.655246001760072,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 34041
+    },
+    {
+      "epoch": 0.810292297438827,
+      "grad_norm": 3.6479706625479027,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 34042
+    },
+    {
+      "epoch": 0.8103161001618585,
+      "grad_norm": 0.7678448713416829,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 34043
+    },
+    {
+      "epoch": 0.81033990288489,
+      "grad_norm": 8.286537719196817,
+      "learning_rate": 2e-05,
+      "loss": 0.8433,
+      "step": 34044
+    },
+    {
+      "epoch": 0.8103637056079216,
+      "grad_norm": 4.046661488186991,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 34045
+    },
+    {
+      "epoch": 0.8103875083309531,
+      "grad_norm": 2.820975849456568,
+      "learning_rate": 2e-05,
+      "loss": 0.4104,
+      "step": 34046
+    },
+    {
+      "epoch": 0.8104113110539846,
+      "grad_norm": 9.089489836468111,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 34047
+    },
+    {
+      "epoch": 0.810435113777016,
+      "grad_norm": 11.202772609957272,
+      "learning_rate": 2e-05,
+      "loss": 0.3049,
+      "step": 34048
+    },
+    {
+      "epoch": 0.8104589165000476,
+      "grad_norm": 6.604452437981909,
+      "learning_rate": 2e-05,
+      "loss": 1.3159,
+      "step": 34049
+    },
+    {
+      "epoch": 0.8104827192230791,
+      "grad_norm": 0.09379561052797727,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 34050
+    },
+    {
+      "epoch": 0.8105065219461106,
+      "grad_norm": 3.21042397273485,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 34051
+    },
+    {
+      "epoch": 0.8105303246691421,
+      "grad_norm": 3.1116300610231473,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 34052
+    },
+    {
+      "epoch": 0.8105541273921737,
+      "grad_norm": 4.269383181460242,
+      "learning_rate": 2e-05,
+      "loss": 0.7543,
+      "step": 34053
+    },
+    {
+      "epoch": 0.8105779301152052,
+      "grad_norm": 1.3330240636029944,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 34054
+    },
+    {
+      "epoch": 0.8106017328382367,
+      "grad_norm": 6.829250346674268,
+      "learning_rate": 2e-05,
+      "loss": 0.2495,
+      "step": 34055
+    },
+    {
+      "epoch": 0.8106255355612683,
+      "grad_norm": 3.271429973074747,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 34056
+    },
+    {
+      "epoch": 0.8106493382842997,
+      "grad_norm": 7.555540248645493,
+      "learning_rate": 2e-05,
+      "loss": 0.7,
+      "step": 34057
+    },
+    {
+      "epoch": 0.8106731410073312,
+      "grad_norm": 2.765331778470568,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 34058
+    },
+    {
+      "epoch": 0.8106969437303627,
+      "grad_norm": 5.425608815117168,
+      "learning_rate": 2e-05,
+      "loss": 1.2603,
+      "step": 34059
+    },
+    {
+      "epoch": 0.8107207464533943,
+      "grad_norm": 11.405130698261175,
+      "learning_rate": 2e-05,
+      "loss": 0.348,
+      "step": 34060
+    },
+    {
+      "epoch": 0.8107445491764258,
+      "grad_norm": 1.8505681142981034,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 34061
+    },
+    {
+      "epoch": 0.8107683518994573,
+      "grad_norm": 0.15093374873525983,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 34062
+    },
+    {
+      "epoch": 0.8107921546224888,
+      "grad_norm": 0.23748524484574052,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 34063
+    },
+    {
+      "epoch": 0.8108159573455204,
+      "grad_norm": 4.86463412067116,
+      "learning_rate": 2e-05,
+      "loss": 0.4471,
+      "step": 34064
+    },
+    {
+      "epoch": 0.8108397600685519,
+      "grad_norm": 6.68316808077398,
+      "learning_rate": 2e-05,
+      "loss": 0.6318,
+      "step": 34065
+    },
+    {
+      "epoch": 0.8108635627915833,
+      "grad_norm": 2.3537778057094414,
+      "learning_rate": 2e-05,
+      "loss": 0.3667,
+      "step": 34066
+    },
+    {
+      "epoch": 0.8108873655146148,
+      "grad_norm": 4.366883658640063,
+      "learning_rate": 2e-05,
+      "loss": 0.6243,
+      "step": 34067
+    },
+    {
+      "epoch": 0.8109111682376464,
+      "grad_norm": 0.1549375452199571,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 34068
+    },
+    {
+      "epoch": 0.8109349709606779,
+      "grad_norm": 5.0067908880176795,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 34069
+    },
+    {
+      "epoch": 0.8109587736837094,
+      "grad_norm": 0.1449728374140068,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 34070
+    },
+    {
+      "epoch": 0.810982576406741,
+      "grad_norm": 2.837309651804149,
+      "learning_rate": 2e-05,
+      "loss": 0.5589,
+      "step": 34071
+    },
+    {
+      "epoch": 0.8110063791297725,
+      "grad_norm": 6.95144633333149,
+      "learning_rate": 2e-05,
+      "loss": 0.2921,
+      "step": 34072
+    },
+    {
+      "epoch": 0.811030181852804,
+      "grad_norm": 1.3977801214232097,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 34073
+    },
+    {
+      "epoch": 0.8110539845758354,
+      "grad_norm": 2.085586971650773,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 34074
+    },
+    {
+      "epoch": 0.811077787298867,
+      "grad_norm": 5.776391224230105,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 34075
+    },
+    {
+      "epoch": 0.8111015900218985,
+      "grad_norm": 1.5797459088383121,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 34076
+    },
+    {
+      "epoch": 0.81112539274493,
+      "grad_norm": 3.701257731262161,
+      "learning_rate": 2e-05,
+      "loss": 0.6542,
+      "step": 34077
+    },
+    {
+      "epoch": 0.8111491954679615,
+      "grad_norm": 6.480669100758946,
+      "learning_rate": 2e-05,
+      "loss": 1.1011,
+      "step": 34078
+    },
+    {
+      "epoch": 0.8111729981909931,
+      "grad_norm": 0.6680257464556629,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 34079
+    },
+    {
+      "epoch": 0.8111968009140246,
+      "grad_norm": 0.20667447142331635,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34080
+    },
+    {
+      "epoch": 0.8112206036370561,
+      "grad_norm": 3.4994285860812018,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 34081
+    },
+    {
+      "epoch": 0.8112444063600875,
+      "grad_norm": 4.6539417282623425,
+      "learning_rate": 2e-05,
+      "loss": 0.9634,
+      "step": 34082
+    },
+    {
+      "epoch": 0.8112682090831191,
+      "grad_norm": 0.36227176145219536,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 34083
+    },
+    {
+      "epoch": 0.8112920118061506,
+      "grad_norm": 0.12321934843356122,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 34084
+    },
+    {
+      "epoch": 0.8113158145291821,
+      "grad_norm": 5.344856408474953,
+      "learning_rate": 2e-05,
+      "loss": 0.8181,
+      "step": 34085
+    },
+    {
+      "epoch": 0.8113396172522137,
+      "grad_norm": 3.1398192278251122,
+      "learning_rate": 2e-05,
+      "loss": 0.1806,
+      "step": 34086
+    },
+    {
+      "epoch": 0.8113634199752452,
+      "grad_norm": 5.225551408083311,
+      "learning_rate": 2e-05,
+      "loss": 0.1629,
+      "step": 34087
+    },
+    {
+      "epoch": 0.8113872226982767,
+      "grad_norm": 7.2992692135247745,
+      "learning_rate": 2e-05,
+      "loss": 0.1735,
+      "step": 34088
+    },
+    {
+      "epoch": 0.8114110254213082,
+      "grad_norm": 0.03403497753035594,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 34089
+    },
+    {
+      "epoch": 0.8114348281443398,
+      "grad_norm": 5.936898077719274,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 34090
+    },
+    {
+      "epoch": 0.8114586308673712,
+      "grad_norm": 4.545540187513092,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 34091
+    },
+    {
+      "epoch": 0.8114824335904027,
+      "grad_norm": 8.651412198477832,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 34092
+    },
+    {
+      "epoch": 0.8115062363134342,
+      "grad_norm": 3.3247706879399397,
+      "learning_rate": 2e-05,
+      "loss": 0.4746,
+      "step": 34093
+    },
+    {
+      "epoch": 0.8115300390364658,
+      "grad_norm": 2.4937852252615396,
+      "learning_rate": 2e-05,
+      "loss": 0.1819,
+      "step": 34094
+    },
+    {
+      "epoch": 0.8115538417594973,
+      "grad_norm": 3.6448252334882008,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 34095
+    },
+    {
+      "epoch": 0.8115776444825288,
+      "grad_norm": 4.858600402437239,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 34096
+    },
+    {
+      "epoch": 0.8116014472055603,
+      "grad_norm": 3.8688688334311245,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 34097
+    },
+    {
+      "epoch": 0.8116252499285919,
+      "grad_norm": 1.6675459684577538,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 34098
+    },
+    {
+      "epoch": 0.8116490526516233,
+      "grad_norm": 3.700723838486645,
+      "learning_rate": 2e-05,
+      "loss": 0.2215,
+      "step": 34099
+    },
+    {
+      "epoch": 0.8116728553746548,
+      "grad_norm": 2.7492026408148615,
+      "learning_rate": 2e-05,
+      "loss": 0.3141,
+      "step": 34100
+    },
+    {
+      "epoch": 0.8116966580976864,
+      "grad_norm": 2.9792833148113895,
+      "learning_rate": 2e-05,
+      "loss": 0.301,
+      "step": 34101
+    },
+    {
+      "epoch": 0.8117204608207179,
+      "grad_norm": 3.0110444373761425,
+      "learning_rate": 2e-05,
+      "loss": 0.5758,
+      "step": 34102
+    },
+    {
+      "epoch": 0.8117442635437494,
+      "grad_norm": 4.478706471715843,
+      "learning_rate": 2e-05,
+      "loss": 0.4675,
+      "step": 34103
+    },
+    {
+      "epoch": 0.8117680662667809,
+      "grad_norm": 6.037086483960755,
+      "learning_rate": 2e-05,
+      "loss": 0.4231,
+      "step": 34104
+    },
+    {
+      "epoch": 0.8117918689898125,
+      "grad_norm": 15.014560401263022,
+      "learning_rate": 2e-05,
+      "loss": 0.521,
+      "step": 34105
+    },
+    {
+      "epoch": 0.811815671712844,
+      "grad_norm": 6.202352481642007,
+      "learning_rate": 2e-05,
+      "loss": 1.1965,
+      "step": 34106
+    },
+    {
+      "epoch": 0.8118394744358755,
+      "grad_norm": 3.9758476937353224,
+      "learning_rate": 2e-05,
+      "loss": 0.8253,
+      "step": 34107
+    },
+    {
+      "epoch": 0.8118632771589069,
+      "grad_norm": 3.9400852278254916,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 34108
+    },
+    {
+      "epoch": 0.8118870798819385,
+      "grad_norm": 0.6898228446065504,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 34109
+    },
+    {
+      "epoch": 0.81191088260497,
+      "grad_norm": 2.7123496710630453,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 34110
+    },
+    {
+      "epoch": 0.8119346853280015,
+      "grad_norm": 1.0242774499024936,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 34111
+    },
+    {
+      "epoch": 0.811958488051033,
+      "grad_norm": 0.10547862276164933,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 34112
+    },
+    {
+      "epoch": 0.8119822907740646,
+      "grad_norm": 2.438789639516391,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 34113
+    },
+    {
+      "epoch": 0.8120060934970961,
+      "grad_norm": 5.534770862034808,
+      "learning_rate": 2e-05,
+      "loss": 0.8121,
+      "step": 34114
+    },
+    {
+      "epoch": 0.8120298962201276,
+      "grad_norm": 3.6554689235781392,
+      "learning_rate": 2e-05,
+      "loss": 0.9466,
+      "step": 34115
+    },
+    {
+      "epoch": 0.8120536989431592,
+      "grad_norm": 6.124018189888053,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 34116
+    },
+    {
+      "epoch": 0.8120775016661906,
+      "grad_norm": 3.343467055021809,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 34117
+    },
+    {
+      "epoch": 0.8121013043892221,
+      "grad_norm": 4.355538192845414,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 34118
+    },
+    {
+      "epoch": 0.8121251071122536,
+      "grad_norm": 4.543941386548544,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 34119
+    },
+    {
+      "epoch": 0.8121489098352852,
+      "grad_norm": 1.037509816088589,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 34120
+    },
+    {
+      "epoch": 0.8121727125583167,
+      "grad_norm": 4.7419731708991835,
+      "learning_rate": 2e-05,
+      "loss": 0.3826,
+      "step": 34121
+    },
+    {
+      "epoch": 0.8121965152813482,
+      "grad_norm": 3.760852066637288,
+      "learning_rate": 2e-05,
+      "loss": 0.4922,
+      "step": 34122
+    },
+    {
+      "epoch": 0.8122203180043797,
+      "grad_norm": 2.934149126272962,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 34123
+    },
+    {
+      "epoch": 0.8122441207274113,
+      "grad_norm": 5.1940383834594925,
+      "learning_rate": 2e-05,
+      "loss": 0.6453,
+      "step": 34124
+    },
+    {
+      "epoch": 0.8122679234504427,
+      "grad_norm": 0.052844635539596806,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34125
+    },
+    {
+      "epoch": 0.8122917261734742,
+      "grad_norm": 4.860993998813654,
+      "learning_rate": 2e-05,
+      "loss": 0.5916,
+      "step": 34126
+    },
+    {
+      "epoch": 0.8123155288965057,
+      "grad_norm": 10.09097800052713,
+      "learning_rate": 2e-05,
+      "loss": 0.5153,
+      "step": 34127
+    },
+    {
+      "epoch": 0.8123393316195373,
+      "grad_norm": 5.425148140401525,
+      "learning_rate": 2e-05,
+      "loss": 1.0247,
+      "step": 34128
+    },
+    {
+      "epoch": 0.8123631343425688,
+      "grad_norm": 3.785365227772587,
+      "learning_rate": 2e-05,
+      "loss": 0.5306,
+      "step": 34129
+    },
+    {
+      "epoch": 0.8123869370656003,
+      "grad_norm": 1.9147228800408487,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 34130
+    },
+    {
+      "epoch": 0.8124107397886319,
+      "grad_norm": 0.3279208371947449,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 34131
+    },
+    {
+      "epoch": 0.8124345425116634,
+      "grad_norm": 5.007727698744682,
+      "learning_rate": 2e-05,
+      "loss": 0.7457,
+      "step": 34132
+    },
+    {
+      "epoch": 0.8124583452346948,
+      "grad_norm": 1.3015480990410004,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 34133
+    },
+    {
+      "epoch": 0.8124821479577263,
+      "grad_norm": 0.7809235728658105,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 34134
+    },
+    {
+      "epoch": 0.8125059506807579,
+      "grad_norm": 4.428599774761008,
+      "learning_rate": 2e-05,
+      "loss": 0.8004,
+      "step": 34135
+    },
+    {
+      "epoch": 0.8125297534037894,
+      "grad_norm": 5.175998439768696,
+      "learning_rate": 2e-05,
+      "loss": 0.9931,
+      "step": 34136
+    },
+    {
+      "epoch": 0.8125535561268209,
+      "grad_norm": 7.859629733621166,
+      "learning_rate": 2e-05,
+      "loss": 0.9633,
+      "step": 34137
+    },
+    {
+      "epoch": 0.8125773588498524,
+      "grad_norm": 1.149280511190532,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 34138
+    },
+    {
+      "epoch": 0.812601161572884,
+      "grad_norm": 2.0343745074881174,
+      "learning_rate": 2e-05,
+      "loss": 0.1399,
+      "step": 34139
+    },
+    {
+      "epoch": 0.8126249642959155,
+      "grad_norm": 0.23675729589061048,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 34140
+    },
+    {
+      "epoch": 0.812648767018947,
+      "grad_norm": 4.038802108386663,
+      "learning_rate": 2e-05,
+      "loss": 0.5911,
+      "step": 34141
+    },
+    {
+      "epoch": 0.8126725697419784,
+      "grad_norm": 2.1318082018266313,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 34142
+    },
+    {
+      "epoch": 0.81269637246501,
+      "grad_norm": 3.757334455553271,
+      "learning_rate": 2e-05,
+      "loss": 0.8326,
+      "step": 34143
+    },
+    {
+      "epoch": 0.8127201751880415,
+      "grad_norm": 0.05683535328731794,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 34144
+    },
+    {
+      "epoch": 0.812743977911073,
+      "grad_norm": 2.509645517185088,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 34145
+    },
+    {
+      "epoch": 0.8127677806341045,
+      "grad_norm": 6.031104531962081,
+      "learning_rate": 2e-05,
+      "loss": 0.1644,
+      "step": 34146
+    },
+    {
+      "epoch": 0.8127915833571361,
+      "grad_norm": 5.59369724679662,
+      "learning_rate": 2e-05,
+      "loss": 0.8515,
+      "step": 34147
+    },
+    {
+      "epoch": 0.8128153860801676,
+      "grad_norm": 3.2280820636361347,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 34148
+    },
+    {
+      "epoch": 0.812839188803199,
+      "grad_norm": 3.5812064493441205,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 34149
+    },
+    {
+      "epoch": 0.8128629915262306,
+      "grad_norm": 4.729332814404845,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 34150
+    },
+    {
+      "epoch": 0.8128867942492621,
+      "grad_norm": 4.448802676465755,
+      "learning_rate": 2e-05,
+      "loss": 0.5885,
+      "step": 34151
+    },
+    {
+      "epoch": 0.8129105969722936,
+      "grad_norm": 2.0225930632942446,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 34152
+    },
+    {
+      "epoch": 0.8129343996953251,
+      "grad_norm": 6.150748288921998,
+      "learning_rate": 2e-05,
+      "loss": 1.5118,
+      "step": 34153
+    },
+    {
+      "epoch": 0.8129582024183567,
+      "grad_norm": 0.09795720546057841,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 34154
+    },
+    {
+      "epoch": 0.8129820051413882,
+      "grad_norm": 3.435457462790231,
+      "learning_rate": 2e-05,
+      "loss": 0.5479,
+      "step": 34155
+    },
+    {
+      "epoch": 0.8130058078644197,
+      "grad_norm": 0.704449322102671,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 34156
+    },
+    {
+      "epoch": 0.8130296105874512,
+      "grad_norm": 4.811696969763429,
+      "learning_rate": 2e-05,
+      "loss": 0.1607,
+      "step": 34157
+    },
+    {
+      "epoch": 0.8130534133104828,
+      "grad_norm": 18.608923666841328,
+      "learning_rate": 2e-05,
+      "loss": 0.6088,
+      "step": 34158
+    },
+    {
+      "epoch": 0.8130772160335142,
+      "grad_norm": 15.74341891827075,
+      "learning_rate": 2e-05,
+      "loss": 0.6513,
+      "step": 34159
+    },
+    {
+      "epoch": 0.8131010187565457,
+      "grad_norm": 0.814057959440477,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 34160
+    },
+    {
+      "epoch": 0.8131248214795772,
+      "grad_norm": 3.15600172904279,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 34161
+    },
+    {
+      "epoch": 0.8131486242026088,
+      "grad_norm": 4.219581746627242,
+      "learning_rate": 2e-05,
+      "loss": 0.9196,
+      "step": 34162
+    },
+    {
+      "epoch": 0.8131724269256403,
+      "grad_norm": 0.2141327607230896,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 34163
+    },
+    {
+      "epoch": 0.8131962296486718,
+      "grad_norm": 1.155730035087391,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 34164
+    },
+    {
+      "epoch": 0.8132200323717034,
+      "grad_norm": 2.229592100358746,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 34165
+    },
+    {
+      "epoch": 0.8132438350947349,
+      "grad_norm": 1.8717540929033614,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 34166
+    },
+    {
+      "epoch": 0.8132676378177663,
+      "grad_norm": 1.177478521439277,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 34167
+    },
+    {
+      "epoch": 0.8132914405407978,
+      "grad_norm": 3.8107722727145497,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 34168
+    },
+    {
+      "epoch": 0.8133152432638294,
+      "grad_norm": 0.5063924574182482,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 34169
+    },
+    {
+      "epoch": 0.8133390459868609,
+      "grad_norm": 0.032955060470979514,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34170
+    },
+    {
+      "epoch": 0.8133628487098924,
+      "grad_norm": 24.122383827025836,
+      "learning_rate": 2e-05,
+      "loss": 0.396,
+      "step": 34171
+    },
+    {
+      "epoch": 0.8133866514329239,
+      "grad_norm": 9.320328440556581,
+      "learning_rate": 2e-05,
+      "loss": 0.2766,
+      "step": 34172
+    },
+    {
+      "epoch": 0.8134104541559555,
+      "grad_norm": 1.648100477633943,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 34173
+    },
+    {
+      "epoch": 0.813434256878987,
+      "grad_norm": 4.45397468361648,
+      "learning_rate": 2e-05,
+      "loss": 0.0706,
+      "step": 34174
+    },
+    {
+      "epoch": 0.8134580596020184,
+      "grad_norm": 10.769467052258804,
+      "learning_rate": 2e-05,
+      "loss": 1.6245,
+      "step": 34175
+    },
+    {
+      "epoch": 0.8134818623250499,
+      "grad_norm": 3.4211926701555186,
+      "learning_rate": 2e-05,
+      "loss": 0.5573,
+      "step": 34176
+    },
+    {
+      "epoch": 0.8135056650480815,
+      "grad_norm": 3.429803625209466,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 34177
+    },
+    {
+      "epoch": 0.813529467771113,
+      "grad_norm": 10.066672748382617,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 34178
+    },
+    {
+      "epoch": 0.8135532704941445,
+      "grad_norm": 0.5757249076479327,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 34179
+    },
+    {
+      "epoch": 0.8135770732171761,
+      "grad_norm": 8.322112346049348,
+      "learning_rate": 2e-05,
+      "loss": 0.8345,
+      "step": 34180
+    },
+    {
+      "epoch": 0.8136008759402076,
+      "grad_norm": 2.511763152221167,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 34181
+    },
+    {
+      "epoch": 0.8136246786632391,
+      "grad_norm": 3.944002022477414,
+      "learning_rate": 2e-05,
+      "loss": 0.3351,
+      "step": 34182
+    },
+    {
+      "epoch": 0.8136484813862706,
+      "grad_norm": 0.13969357623964712,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 34183
+    },
+    {
+      "epoch": 0.8136722841093021,
+      "grad_norm": 4.99732295936424,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 34184
+    },
+    {
+      "epoch": 0.8136960868323336,
+      "grad_norm": 2.526113114860101,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 34185
+    },
+    {
+      "epoch": 0.8137198895553651,
+      "grad_norm": 4.46266534817717,
+      "learning_rate": 2e-05,
+      "loss": 0.6705,
+      "step": 34186
+    },
+    {
+      "epoch": 0.8137436922783966,
+      "grad_norm": 4.12397175625455,
+      "learning_rate": 2e-05,
+      "loss": 0.6126,
+      "step": 34187
+    },
+    {
+      "epoch": 0.8137674950014282,
+      "grad_norm": 6.723315688417311,
+      "learning_rate": 2e-05,
+      "loss": 1.6577,
+      "step": 34188
+    },
+    {
+      "epoch": 0.8137912977244597,
+      "grad_norm": 2.681180690764198,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 34189
+    },
+    {
+      "epoch": 0.8138151004474912,
+      "grad_norm": 0.44294064634687685,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 34190
+    },
+    {
+      "epoch": 0.8138389031705227,
+      "grad_norm": 0.1732192871191436,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 34191
+    },
+    {
+      "epoch": 0.8138627058935543,
+      "grad_norm": 5.241792009607086,
+      "learning_rate": 2e-05,
+      "loss": 0.869,
+      "step": 34192
+    },
+    {
+      "epoch": 0.8138865086165857,
+      "grad_norm": 0.12222631577089872,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 34193
+    },
+    {
+      "epoch": 0.8139103113396172,
+      "grad_norm": 9.652684042466728,
+      "learning_rate": 2e-05,
+      "loss": 0.48,
+      "step": 34194
+    },
+    {
+      "epoch": 0.8139341140626488,
+      "grad_norm": 2.821498886933397,
+      "learning_rate": 2e-05,
+      "loss": 0.5049,
+      "step": 34195
+    },
+    {
+      "epoch": 0.8139579167856803,
+      "grad_norm": 1.085102480829376,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 34196
+    },
+    {
+      "epoch": 0.8139817195087118,
+      "grad_norm": 3.7603117415098244,
+      "learning_rate": 2e-05,
+      "loss": 0.5432,
+      "step": 34197
+    },
+    {
+      "epoch": 0.8140055222317433,
+      "grad_norm": 9.511117234049184,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 34198
+    },
+    {
+      "epoch": 0.8140293249547749,
+      "grad_norm": 3.337249517639195,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 34199
+    },
+    {
+      "epoch": 0.8140531276778064,
+      "grad_norm": 4.936381631803001,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 34200
+    },
+    {
+      "epoch": 0.8140769304008378,
+      "grad_norm": 0.21957983583286725,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 34201
+    },
+    {
+      "epoch": 0.8141007331238693,
+      "grad_norm": 5.363917850661299,
+      "learning_rate": 2e-05,
+      "loss": 0.1132,
+      "step": 34202
+    },
+    {
+      "epoch": 0.8141245358469009,
+      "grad_norm": 5.263857304046047,
+      "learning_rate": 2e-05,
+      "loss": 0.0926,
+      "step": 34203
+    },
+    {
+      "epoch": 0.8141483385699324,
+      "grad_norm": 1.1737892878137866,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 34204
+    },
+    {
+      "epoch": 0.8141721412929639,
+      "grad_norm": 6.075830757576333,
+      "learning_rate": 2e-05,
+      "loss": 0.7359,
+      "step": 34205
+    },
+    {
+      "epoch": 0.8141959440159954,
+      "grad_norm": 0.7972001118269394,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 34206
+    },
+    {
+      "epoch": 0.814219746739027,
+      "grad_norm": 6.011212456594359,
+      "learning_rate": 2e-05,
+      "loss": 0.9872,
+      "step": 34207
+    },
+    {
+      "epoch": 0.8142435494620585,
+      "grad_norm": 3.956967143882668,
+      "learning_rate": 2e-05,
+      "loss": 0.6839,
+      "step": 34208
+    },
+    {
+      "epoch": 0.8142673521850899,
+      "grad_norm": 2.2022993998370044,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 34209
+    },
+    {
+      "epoch": 0.8142911549081215,
+      "grad_norm": 4.638373557147932,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 34210
+    },
+    {
+      "epoch": 0.814314957631153,
+      "grad_norm": 8.863971503264162,
+      "learning_rate": 2e-05,
+      "loss": 0.4431,
+      "step": 34211
+    },
+    {
+      "epoch": 0.8143387603541845,
+      "grad_norm": 5.978085470211233,
+      "learning_rate": 2e-05,
+      "loss": 0.7393,
+      "step": 34212
+    },
+    {
+      "epoch": 0.814362563077216,
+      "grad_norm": 10.637663897187908,
+      "learning_rate": 2e-05,
+      "loss": 1.3035,
+      "step": 34213
+    },
+    {
+      "epoch": 0.8143863658002476,
+      "grad_norm": 1.1396472846732573,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 34214
+    },
+    {
+      "epoch": 0.8144101685232791,
+      "grad_norm": 2.979211929021644,
+      "learning_rate": 2e-05,
+      "loss": 0.2719,
+      "step": 34215
+    },
+    {
+      "epoch": 0.8144339712463106,
+      "grad_norm": 2.4173313862023815,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 34216
+    },
+    {
+      "epoch": 0.814457773969342,
+      "grad_norm": 9.364397559427822,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 34217
+    },
+    {
+      "epoch": 0.8144815766923736,
+      "grad_norm": 0.4872285795752025,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 34218
+    },
+    {
+      "epoch": 0.8145053794154051,
+      "grad_norm": 10.824268195050157,
+      "learning_rate": 2e-05,
+      "loss": 0.3923,
+      "step": 34219
+    },
+    {
+      "epoch": 0.8145291821384366,
+      "grad_norm": 8.472928456158199,
+      "learning_rate": 2e-05,
+      "loss": 0.3502,
+      "step": 34220
+    },
+    {
+      "epoch": 0.8145529848614681,
+      "grad_norm": 5.789342757172055,
+      "learning_rate": 2e-05,
+      "loss": 0.3155,
+      "step": 34221
+    },
+    {
+      "epoch": 0.8145767875844997,
+      "grad_norm": 5.117952127548899,
+      "learning_rate": 2e-05,
+      "loss": 1.1033,
+      "step": 34222
+    },
+    {
+      "epoch": 0.8146005903075312,
+      "grad_norm": 1.1238516144215298,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 34223
+    },
+    {
+      "epoch": 0.8146243930305627,
+      "grad_norm": 1.5432977565796284,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 34224
+    },
+    {
+      "epoch": 0.8146481957535943,
+      "grad_norm": 4.089253583870578,
+      "learning_rate": 2e-05,
+      "loss": 0.4621,
+      "step": 34225
+    },
+    {
+      "epoch": 0.8146719984766257,
+      "grad_norm": 1.9599725157058614,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 34226
+    },
+    {
+      "epoch": 0.8146958011996572,
+      "grad_norm": 4.865146354767191,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 34227
+    },
+    {
+      "epoch": 0.8147196039226887,
+      "grad_norm": 2.3185351552373903,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 34228
+    },
+    {
+      "epoch": 0.8147434066457203,
+      "grad_norm": 0.18373741709327637,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 34229
+    },
+    {
+      "epoch": 0.8147672093687518,
+      "grad_norm": 1.1318331282742062,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 34230
+    },
+    {
+      "epoch": 0.8147910120917833,
+      "grad_norm": 3.7338777054065693,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 34231
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 4.810820628612054,
+      "learning_rate": 2e-05,
+      "loss": 0.4626,
+      "step": 34232
+    },
+    {
+      "epoch": 0.8148386175378464,
+      "grad_norm": 9.343691610519937,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 34233
+    },
+    {
+      "epoch": 0.8148624202608779,
+      "grad_norm": 1.552434310400271,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 34234
+    },
+    {
+      "epoch": 0.8148862229839093,
+      "grad_norm": 6.743258098701897,
+      "learning_rate": 2e-05,
+      "loss": 0.9354,
+      "step": 34235
+    },
+    {
+      "epoch": 0.8149100257069408,
+      "grad_norm": 7.616137283239984,
+      "learning_rate": 2e-05,
+      "loss": 0.2714,
+      "step": 34236
+    },
+    {
+      "epoch": 0.8149338284299724,
+      "grad_norm": 6.115466953267163,
+      "learning_rate": 2e-05,
+      "loss": 0.8727,
+      "step": 34237
+    },
+    {
+      "epoch": 0.8149576311530039,
+      "grad_norm": 3.0046062401395583,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 34238
+    },
+    {
+      "epoch": 0.8149814338760354,
+      "grad_norm": 3.0807532308605126,
+      "learning_rate": 2e-05,
+      "loss": 0.1481,
+      "step": 34239
+    },
+    {
+      "epoch": 0.815005236599067,
+      "grad_norm": 5.544664499235465,
+      "learning_rate": 2e-05,
+      "loss": 0.1782,
+      "step": 34240
+    },
+    {
+      "epoch": 0.8150290393220985,
+      "grad_norm": 10.53103224279401,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 34241
+    },
+    {
+      "epoch": 0.81505284204513,
+      "grad_norm": 5.727037951529235,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 34242
+    },
+    {
+      "epoch": 0.8150766447681614,
+      "grad_norm": 3.163409179218743,
+      "learning_rate": 2e-05,
+      "loss": 0.5446,
+      "step": 34243
+    },
+    {
+      "epoch": 0.815100447491193,
+      "grad_norm": 2.2123724500673743,
+      "learning_rate": 2e-05,
+      "loss": 0.2558,
+      "step": 34244
+    },
+    {
+      "epoch": 0.8151242502142245,
+      "grad_norm": 1.5905832215863176,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 34245
+    },
+    {
+      "epoch": 0.815148052937256,
+      "grad_norm": 0.4630389978273569,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 34246
+    },
+    {
+      "epoch": 0.8151718556602875,
+      "grad_norm": 2.4893436620952683,
+      "learning_rate": 2e-05,
+      "loss": 0.2252,
+      "step": 34247
+    },
+    {
+      "epoch": 0.8151956583833191,
+      "grad_norm": 0.02475718760191618,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34248
+    },
+    {
+      "epoch": 0.8152194611063506,
+      "grad_norm": 2.862883583751576,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 34249
+    },
+    {
+      "epoch": 0.8152432638293821,
+      "grad_norm": 3.8133507012014896,
+      "learning_rate": 2e-05,
+      "loss": 0.0939,
+      "step": 34250
+    },
+    {
+      "epoch": 0.8152670665524135,
+      "grad_norm": 3.1169394318285266,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 34251
+    },
+    {
+      "epoch": 0.8152908692754451,
+      "grad_norm": 1.1671492701706767,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 34252
+    },
+    {
+      "epoch": 0.8153146719984766,
+      "grad_norm": 0.40756632199151305,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 34253
+    },
+    {
+      "epoch": 0.8153384747215081,
+      "grad_norm": 5.460523744505445,
+      "learning_rate": 2e-05,
+      "loss": 0.8868,
+      "step": 34254
+    },
+    {
+      "epoch": 0.8153622774445397,
+      "grad_norm": 0.30599324436650527,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 34255
+    },
+    {
+      "epoch": 0.8153860801675712,
+      "grad_norm": 2.0972845377517633,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 34256
+    },
+    {
+      "epoch": 0.8154098828906027,
+      "grad_norm": 5.175873181055362,
+      "learning_rate": 2e-05,
+      "loss": 0.7363,
+      "step": 34257
+    },
+    {
+      "epoch": 0.8154336856136342,
+      "grad_norm": 8.019423426038205,
+      "learning_rate": 2e-05,
+      "loss": 0.9412,
+      "step": 34258
+    },
+    {
+      "epoch": 0.8154574883366658,
+      "grad_norm": 2.8601553786428378,
+      "learning_rate": 2e-05,
+      "loss": 0.3985,
+      "step": 34259
+    },
+    {
+      "epoch": 0.8154812910596972,
+      "grad_norm": 0.3164640631648575,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 34260
+    },
+    {
+      "epoch": 0.8155050937827287,
+      "grad_norm": 4.175495075863266,
+      "learning_rate": 2e-05,
+      "loss": 0.5458,
+      "step": 34261
+    },
+    {
+      "epoch": 0.8155288965057602,
+      "grad_norm": 2.1136821348303365,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 34262
+    },
+    {
+      "epoch": 0.8155526992287918,
+      "grad_norm": 1.608853116116592,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 34263
+    },
+    {
+      "epoch": 0.8155765019518233,
+      "grad_norm": 2.0102800844485347,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 34264
+    },
+    {
+      "epoch": 0.8156003046748548,
+      "grad_norm": 8.656808521740684,
+      "learning_rate": 2e-05,
+      "loss": 0.9293,
+      "step": 34265
+    },
+    {
+      "epoch": 0.8156241073978863,
+      "grad_norm": 3.129356252257414,
+      "learning_rate": 2e-05,
+      "loss": 0.5582,
+      "step": 34266
+    },
+    {
+      "epoch": 0.8156479101209179,
+      "grad_norm": 2.7758941367555945,
+      "learning_rate": 2e-05,
+      "loss": 0.5876,
+      "step": 34267
+    },
+    {
+      "epoch": 0.8156717128439493,
+      "grad_norm": 4.698509461842105,
+      "learning_rate": 2e-05,
+      "loss": 0.1456,
+      "step": 34268
+    },
+    {
+      "epoch": 0.8156955155669808,
+      "grad_norm": 3.701974982286271,
+      "learning_rate": 2e-05,
+      "loss": 0.5461,
+      "step": 34269
+    },
+    {
+      "epoch": 0.8157193182900124,
+      "grad_norm": 0.2555105245317337,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 34270
+    },
+    {
+      "epoch": 0.8157431210130439,
+      "grad_norm": 4.379936119850266,
+      "learning_rate": 2e-05,
+      "loss": 0.5097,
+      "step": 34271
+    },
+    {
+      "epoch": 0.8157669237360754,
+      "grad_norm": 4.0615738426685954,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 34272
+    },
+    {
+      "epoch": 0.8157907264591069,
+      "grad_norm": 4.942981127131385,
+      "learning_rate": 2e-05,
+      "loss": 0.9561,
+      "step": 34273
+    },
+    {
+      "epoch": 0.8158145291821385,
+      "grad_norm": 2.772242158418496,
+      "learning_rate": 2e-05,
+      "loss": 0.3992,
+      "step": 34274
+    },
+    {
+      "epoch": 0.81583833190517,
+      "grad_norm": 1.5478188637927497,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 34275
+    },
+    {
+      "epoch": 0.8158621346282015,
+      "grad_norm": 7.998771430249629,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 34276
+    },
+    {
+      "epoch": 0.8158859373512329,
+      "grad_norm": 3.5085079634154885,
+      "learning_rate": 2e-05,
+      "loss": 0.0669,
+      "step": 34277
+    },
+    {
+      "epoch": 0.8159097400742645,
+      "grad_norm": 4.566832076367913,
+      "learning_rate": 2e-05,
+      "loss": 0.4364,
+      "step": 34278
+    },
+    {
+      "epoch": 0.815933542797296,
+      "grad_norm": 0.7881896300987795,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 34279
+    },
+    {
+      "epoch": 0.8159573455203275,
+      "grad_norm": 4.83889393967387,
+      "learning_rate": 2e-05,
+      "loss": 1.0639,
+      "step": 34280
+    },
+    {
+      "epoch": 0.815981148243359,
+      "grad_norm": 0.438546926234071,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 34281
+    },
+    {
+      "epoch": 0.8160049509663906,
+      "grad_norm": 0.6280301510907466,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 34282
+    },
+    {
+      "epoch": 0.8160287536894221,
+      "grad_norm": 4.97493676296825,
+      "learning_rate": 2e-05,
+      "loss": 0.0944,
+      "step": 34283
+    },
+    {
+      "epoch": 0.8160525564124536,
+      "grad_norm": 3.0636860189587622,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 34284
+    },
+    {
+      "epoch": 0.8160763591354852,
+      "grad_norm": 2.2294608030391636,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 34285
+    },
+    {
+      "epoch": 0.8161001618585166,
+      "grad_norm": 2.0593853942845826,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 34286
+    },
+    {
+      "epoch": 0.8161239645815481,
+      "grad_norm": 1.5293516234808273,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 34287
+    },
+    {
+      "epoch": 0.8161477673045796,
+      "grad_norm": 8.707725414759206,
+      "learning_rate": 2e-05,
+      "loss": 0.9284,
+      "step": 34288
+    },
+    {
+      "epoch": 0.8161715700276112,
+      "grad_norm": 0.05457290597521039,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34289
+    },
+    {
+      "epoch": 0.8161953727506427,
+      "grad_norm": 2.5170246250411874,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 34290
+    },
+    {
+      "epoch": 0.8162191754736742,
+      "grad_norm": 1.7187806069190097,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 34291
+    },
+    {
+      "epoch": 0.8162429781967057,
+      "grad_norm": 9.982156072388861,
+      "learning_rate": 2e-05,
+      "loss": 0.5778,
+      "step": 34292
+    },
+    {
+      "epoch": 0.8162667809197373,
+      "grad_norm": 1.3605759666246693,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 34293
+    },
+    {
+      "epoch": 0.8162905836427687,
+      "grad_norm": 6.919738053363665,
+      "learning_rate": 2e-05,
+      "loss": 0.4439,
+      "step": 34294
+    },
+    {
+      "epoch": 0.8163143863658002,
+      "grad_norm": 6.688506957640398,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 34295
+    },
+    {
+      "epoch": 0.8163381890888317,
+      "grad_norm": 5.606229229563951,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 34296
+    },
+    {
+      "epoch": 0.8163619918118633,
+      "grad_norm": 1.447678175736155,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 34297
+    },
+    {
+      "epoch": 0.8163857945348948,
+      "grad_norm": 4.79942088597654,
+      "learning_rate": 2e-05,
+      "loss": 0.9037,
+      "step": 34298
+    },
+    {
+      "epoch": 0.8164095972579263,
+      "grad_norm": 0.7090897597362396,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 34299
+    },
+    {
+      "epoch": 0.8164333999809579,
+      "grad_norm": 8.922817521275949,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 34300
+    },
+    {
+      "epoch": 0.8164572027039894,
+      "grad_norm": 3.1151728829230003,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 34301
+    },
+    {
+      "epoch": 0.8164810054270208,
+      "grad_norm": 2.597340080913663,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 34302
+    },
+    {
+      "epoch": 0.8165048081500523,
+      "grad_norm": 0.5017457071919599,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 34303
+    },
+    {
+      "epoch": 0.8165286108730839,
+      "grad_norm": 0.37119322551665934,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 34304
+    },
+    {
+      "epoch": 0.8165524135961154,
+      "grad_norm": 5.680480450924055,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 34305
+    },
+    {
+      "epoch": 0.8165762163191469,
+      "grad_norm": 9.01612608988313,
+      "learning_rate": 2e-05,
+      "loss": 0.1814,
+      "step": 34306
+    },
+    {
+      "epoch": 0.8166000190421784,
+      "grad_norm": 6.747464814640327,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 34307
+    },
+    {
+      "epoch": 0.81662382176521,
+      "grad_norm": 4.399913510502751,
+      "learning_rate": 2e-05,
+      "loss": 0.4761,
+      "step": 34308
+    },
+    {
+      "epoch": 0.8166476244882415,
+      "grad_norm": 10.40420527644825,
+      "learning_rate": 2e-05,
+      "loss": 0.3611,
+      "step": 34309
+    },
+    {
+      "epoch": 0.816671427211273,
+      "grad_norm": 7.585022957770665,
+      "learning_rate": 2e-05,
+      "loss": 0.284,
+      "step": 34310
+    },
+    {
+      "epoch": 0.8166952299343044,
+      "grad_norm": 0.5598619868778878,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 34311
+    },
+    {
+      "epoch": 0.816719032657336,
+      "grad_norm": 3.728523746941686,
+      "learning_rate": 2e-05,
+      "loss": 0.6035,
+      "step": 34312
+    },
+    {
+      "epoch": 0.8167428353803675,
+      "grad_norm": 3.520447049145224,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 34313
+    },
+    {
+      "epoch": 0.816766638103399,
+      "grad_norm": 0.07697918027675238,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 34314
+    },
+    {
+      "epoch": 0.8167904408264306,
+      "grad_norm": 14.190395223761216,
+      "learning_rate": 2e-05,
+      "loss": 1.0298,
+      "step": 34315
+    },
+    {
+      "epoch": 0.8168142435494621,
+      "grad_norm": 0.7893112264669483,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 34316
+    },
+    {
+      "epoch": 0.8168380462724936,
+      "grad_norm": 4.060991693786611,
+      "learning_rate": 2e-05,
+      "loss": 0.7198,
+      "step": 34317
+    },
+    {
+      "epoch": 0.816861848995525,
+      "grad_norm": 5.193007312403601,
+      "learning_rate": 2e-05,
+      "loss": 0.2283,
+      "step": 34318
+    },
+    {
+      "epoch": 0.8168856517185566,
+      "grad_norm": 6.032449945519822,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 34319
+    },
+    {
+      "epoch": 0.8169094544415881,
+      "grad_norm": 2.3921878282730655,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 34320
+    },
+    {
+      "epoch": 0.8169332571646196,
+      "grad_norm": 13.384378180919235,
+      "learning_rate": 2e-05,
+      "loss": 0.5378,
+      "step": 34321
+    },
+    {
+      "epoch": 0.8169570598876511,
+      "grad_norm": 6.244511562980739,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 34322
+    },
+    {
+      "epoch": 0.8169808626106827,
+      "grad_norm": 5.446904651205759,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 34323
+    },
+    {
+      "epoch": 0.8170046653337142,
+      "grad_norm": 2.224172552876222,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 34324
+    },
+    {
+      "epoch": 0.8170284680567457,
+      "grad_norm": 4.299978732120871,
+      "learning_rate": 2e-05,
+      "loss": 0.4126,
+      "step": 34325
+    },
+    {
+      "epoch": 0.8170522707797772,
+      "grad_norm": 0.5767209714588943,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 34326
+    },
+    {
+      "epoch": 0.8170760735028088,
+      "grad_norm": 3.9912826434438444,
+      "learning_rate": 2e-05,
+      "loss": 0.6029,
+      "step": 34327
+    },
+    {
+      "epoch": 0.8170998762258402,
+      "grad_norm": 8.062717392467256,
+      "learning_rate": 2e-05,
+      "loss": 0.8572,
+      "step": 34328
+    },
+    {
+      "epoch": 0.8171236789488717,
+      "grad_norm": 0.07144692074131312,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34329
+    },
+    {
+      "epoch": 0.8171474816719033,
+      "grad_norm": 0.20812589164033196,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 34330
+    },
+    {
+      "epoch": 0.8171712843949348,
+      "grad_norm": 0.025092203009287398,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 34331
+    },
+    {
+      "epoch": 0.8171950871179663,
+      "grad_norm": 3.4457083960201587,
+      "learning_rate": 2e-05,
+      "loss": 0.4882,
+      "step": 34332
+    },
+    {
+      "epoch": 0.8172188898409978,
+      "grad_norm": 3.61405771889035,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 34333
+    },
+    {
+      "epoch": 0.8172426925640294,
+      "grad_norm": 0.7649240044451531,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 34334
+    },
+    {
+      "epoch": 0.8172664952870609,
+      "grad_norm": 5.873863452800408,
+      "learning_rate": 2e-05,
+      "loss": 0.1904,
+      "step": 34335
+    },
+    {
+      "epoch": 0.8172902980100923,
+      "grad_norm": 3.3766658546762147,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 34336
+    },
+    {
+      "epoch": 0.8173141007331238,
+      "grad_norm": 10.889467043194825,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 34337
+    },
+    {
+      "epoch": 0.8173379034561554,
+      "grad_norm": 0.36504142526956324,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 34338
+    },
+    {
+      "epoch": 0.8173617061791869,
+      "grad_norm": 2.096068925563016,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 34339
+    },
+    {
+      "epoch": 0.8173855089022184,
+      "grad_norm": 8.467545011831643,
+      "learning_rate": 2e-05,
+      "loss": 0.3793,
+      "step": 34340
+    },
+    {
+      "epoch": 0.8174093116252499,
+      "grad_norm": 5.239768823503994,
+      "learning_rate": 2e-05,
+      "loss": 0.9515,
+      "step": 34341
+    },
+    {
+      "epoch": 0.8174331143482815,
+      "grad_norm": 6.29308438770457,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 34342
+    },
+    {
+      "epoch": 0.817456917071313,
+      "grad_norm": 10.703100784381387,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 34343
+    },
+    {
+      "epoch": 0.8174807197943444,
+      "grad_norm": 7.487096013183179,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 34344
+    },
+    {
+      "epoch": 0.817504522517376,
+      "grad_norm": 4.354172474534595,
+      "learning_rate": 2e-05,
+      "loss": 0.2318,
+      "step": 34345
+    },
+    {
+      "epoch": 0.8175283252404075,
+      "grad_norm": 3.469225900766446,
+      "learning_rate": 2e-05,
+      "loss": 0.3381,
+      "step": 34346
+    },
+    {
+      "epoch": 0.817552127963439,
+      "grad_norm": 5.214590556590124,
+      "learning_rate": 2e-05,
+      "loss": 0.1159,
+      "step": 34347
+    },
+    {
+      "epoch": 0.8175759306864705,
+      "grad_norm": 10.141913736550883,
+      "learning_rate": 2e-05,
+      "loss": 1.0344,
+      "step": 34348
+    },
+    {
+      "epoch": 0.8175997334095021,
+      "grad_norm": 5.146816328146732,
+      "learning_rate": 2e-05,
+      "loss": 0.6557,
+      "step": 34349
+    },
+    {
+      "epoch": 0.8176235361325336,
+      "grad_norm": 5.699647029660336,
+      "learning_rate": 2e-05,
+      "loss": 0.4493,
+      "step": 34350
+    },
+    {
+      "epoch": 0.8176473388555651,
+      "grad_norm": 7.071445403178235,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 34351
+    },
+    {
+      "epoch": 0.8176711415785965,
+      "grad_norm": 4.535736559637234,
+      "learning_rate": 2e-05,
+      "loss": 0.1803,
+      "step": 34352
+    },
+    {
+      "epoch": 0.8176949443016281,
+      "grad_norm": 1.096041751409973,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 34353
+    },
+    {
+      "epoch": 0.8177187470246596,
+      "grad_norm": 8.40323277301762,
+      "learning_rate": 2e-05,
+      "loss": 0.4832,
+      "step": 34354
+    },
+    {
+      "epoch": 0.8177425497476911,
+      "grad_norm": 3.1502576046770074,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 34355
+    },
+    {
+      "epoch": 0.8177663524707226,
+      "grad_norm": 2.2925602971695738,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 34356
+    },
+    {
+      "epoch": 0.8177901551937542,
+      "grad_norm": 0.2538442565744259,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 34357
+    },
+    {
+      "epoch": 0.8178139579167857,
+      "grad_norm": 2.660004083161662,
+      "learning_rate": 2e-05,
+      "loss": 0.2124,
+      "step": 34358
+    },
+    {
+      "epoch": 0.8178377606398172,
+      "grad_norm": 4.503169701533942,
+      "learning_rate": 2e-05,
+      "loss": 0.882,
+      "step": 34359
+    },
+    {
+      "epoch": 0.8178615633628488,
+      "grad_norm": 0.11660176385576093,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 34360
+    },
+    {
+      "epoch": 0.8178853660858802,
+      "grad_norm": 3.503136555980542,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 34361
+    },
+    {
+      "epoch": 0.8179091688089117,
+      "grad_norm": 3.4791569628303494,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 34362
+    },
+    {
+      "epoch": 0.8179329715319432,
+      "grad_norm": 3.3767263649708315,
+      "learning_rate": 2e-05,
+      "loss": 0.4804,
+      "step": 34363
+    },
+    {
+      "epoch": 0.8179567742549748,
+      "grad_norm": 2.866481184610648,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 34364
+    },
+    {
+      "epoch": 0.8179805769780063,
+      "grad_norm": 5.564247874726756,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 34365
+    },
+    {
+      "epoch": 0.8180043797010378,
+      "grad_norm": 0.12636649827417742,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 34366
+    },
+    {
+      "epoch": 0.8180281824240693,
+      "grad_norm": 5.296381825350148,
+      "learning_rate": 2e-05,
+      "loss": 0.7702,
+      "step": 34367
+    },
+    {
+      "epoch": 0.8180519851471009,
+      "grad_norm": 7.597118545472657,
+      "learning_rate": 2e-05,
+      "loss": 1.7457,
+      "step": 34368
+    },
+    {
+      "epoch": 0.8180757878701324,
+      "grad_norm": 3.5698718877136755,
+      "learning_rate": 2e-05,
+      "loss": 0.571,
+      "step": 34369
+    },
+    {
+      "epoch": 0.8180995905931638,
+      "grad_norm": 7.964391840743104,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 34370
+    },
+    {
+      "epoch": 0.8181233933161953,
+      "grad_norm": 0.819641768928647,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 34371
+    },
+    {
+      "epoch": 0.8181471960392269,
+      "grad_norm": 0.4727443908639654,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 34372
+    },
+    {
+      "epoch": 0.8181709987622584,
+      "grad_norm": 5.221307730061151,
+      "learning_rate": 2e-05,
+      "loss": 0.6692,
+      "step": 34373
+    },
+    {
+      "epoch": 0.8181948014852899,
+      "grad_norm": 9.864436093765333,
+      "learning_rate": 2e-05,
+      "loss": 1.3054,
+      "step": 34374
+    },
+    {
+      "epoch": 0.8182186042083215,
+      "grad_norm": 2.989546014593445,
+      "learning_rate": 2e-05,
+      "loss": 0.1335,
+      "step": 34375
+    },
+    {
+      "epoch": 0.818242406931353,
+      "grad_norm": 0.8315359076705716,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 34376
+    },
+    {
+      "epoch": 0.8182662096543845,
+      "grad_norm": 1.356059850304496,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 34377
+    },
+    {
+      "epoch": 0.8182900123774159,
+      "grad_norm": 3.041577536305949,
+      "learning_rate": 2e-05,
+      "loss": 0.4897,
+      "step": 34378
+    },
+    {
+      "epoch": 0.8183138151004475,
+      "grad_norm": 8.589485759415602,
+      "learning_rate": 2e-05,
+      "loss": 0.1726,
+      "step": 34379
+    },
+    {
+      "epoch": 0.818337617823479,
+      "grad_norm": 0.09952600163096846,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 34380
+    },
+    {
+      "epoch": 0.8183614205465105,
+      "grad_norm": 0.1177910855830914,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 34381
+    },
+    {
+      "epoch": 0.818385223269542,
+      "grad_norm": 0.051920136771633214,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34382
+    },
+    {
+      "epoch": 0.8184090259925736,
+      "grad_norm": 3.8555690236209874,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 34383
+    },
+    {
+      "epoch": 0.8184328287156051,
+      "grad_norm": 1.074929643391574,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 34384
+    },
+    {
+      "epoch": 0.8184566314386366,
+      "grad_norm": 3.0661434816025834,
+      "learning_rate": 2e-05,
+      "loss": 0.3654,
+      "step": 34385
+    },
+    {
+      "epoch": 0.818480434161668,
+      "grad_norm": 5.091659977487271,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 34386
+    },
+    {
+      "epoch": 0.8185042368846996,
+      "grad_norm": 5.92928493569424,
+      "learning_rate": 2e-05,
+      "loss": 1.1473,
+      "step": 34387
+    },
+    {
+      "epoch": 0.8185280396077311,
+      "grad_norm": 4.196631447883932,
+      "learning_rate": 2e-05,
+      "loss": 0.1171,
+      "step": 34388
+    },
+    {
+      "epoch": 0.8185518423307626,
+      "grad_norm": 2.699446553210594,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 34389
+    },
+    {
+      "epoch": 0.8185756450537941,
+      "grad_norm": 4.16269264272792,
+      "learning_rate": 2e-05,
+      "loss": 0.6018,
+      "step": 34390
+    },
+    {
+      "epoch": 0.8185994477768257,
+      "grad_norm": 5.004305206408594,
+      "learning_rate": 2e-05,
+      "loss": 0.9166,
+      "step": 34391
+    },
+    {
+      "epoch": 0.8186232504998572,
+      "grad_norm": 5.0130121351162416,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 34392
+    },
+    {
+      "epoch": 0.8186470532228887,
+      "grad_norm": 7.714734095384478,
+      "learning_rate": 2e-05,
+      "loss": 0.8511,
+      "step": 34393
+    },
+    {
+      "epoch": 0.8186708559459203,
+      "grad_norm": 5.71988019228066,
+      "learning_rate": 2e-05,
+      "loss": 1.2739,
+      "step": 34394
+    },
+    {
+      "epoch": 0.8186946586689517,
+      "grad_norm": 0.06711211774211576,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34395
+    },
+    {
+      "epoch": 0.8187184613919832,
+      "grad_norm": 3.79323492441992,
+      "learning_rate": 2e-05,
+      "loss": 0.8122,
+      "step": 34396
+    },
+    {
+      "epoch": 0.8187422641150147,
+      "grad_norm": 0.18888273468648714,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 34397
+    },
+    {
+      "epoch": 0.8187660668380463,
+      "grad_norm": 2.0721709434010824,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 34398
+    },
+    {
+      "epoch": 0.8187898695610778,
+      "grad_norm": 0.5323162522152249,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 34399
+    },
+    {
+      "epoch": 0.8188136722841093,
+      "grad_norm": 0.034681957579280887,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 34400
+    },
+    {
+      "epoch": 0.8188374750071408,
+      "grad_norm": 9.315491377091602,
+      "learning_rate": 2e-05,
+      "loss": 0.4802,
+      "step": 34401
+    },
+    {
+      "epoch": 0.8188612777301724,
+      "grad_norm": 11.914989057591402,
+      "learning_rate": 2e-05,
+      "loss": 0.5683,
+      "step": 34402
+    },
+    {
+      "epoch": 0.8188850804532039,
+      "grad_norm": 12.602531505395518,
+      "learning_rate": 2e-05,
+      "loss": 0.8919,
+      "step": 34403
+    },
+    {
+      "epoch": 0.8189088831762353,
+      "grad_norm": 2.786216270446108,
+      "learning_rate": 2e-05,
+      "loss": 0.31,
+      "step": 34404
+    },
+    {
+      "epoch": 0.8189326858992668,
+      "grad_norm": 1.919033951042041,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 34405
+    },
+    {
+      "epoch": 0.8189564886222984,
+      "grad_norm": 8.99303393210287,
+      "learning_rate": 2e-05,
+      "loss": 0.4644,
+      "step": 34406
+    },
+    {
+      "epoch": 0.8189802913453299,
+      "grad_norm": 1.4495861502109133,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 34407
+    },
+    {
+      "epoch": 0.8190040940683614,
+      "grad_norm": 0.2945329100106847,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 34408
+    },
+    {
+      "epoch": 0.819027896791393,
+      "grad_norm": 8.262099254260589,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 34409
+    },
+    {
+      "epoch": 0.8190516995144245,
+      "grad_norm": 3.2542443347215735,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 34410
+    },
+    {
+      "epoch": 0.819075502237456,
+      "grad_norm": 7.611043936841155,
+      "learning_rate": 2e-05,
+      "loss": 1.1788,
+      "step": 34411
+    },
+    {
+      "epoch": 0.8190993049604874,
+      "grad_norm": 2.2494634697611287,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 34412
+    },
+    {
+      "epoch": 0.819123107683519,
+      "grad_norm": 6.195479859475667,
+      "learning_rate": 2e-05,
+      "loss": 1.4967,
+      "step": 34413
+    },
+    {
+      "epoch": 0.8191469104065505,
+      "grad_norm": 6.8571381794201685,
+      "learning_rate": 2e-05,
+      "loss": 0.3736,
+      "step": 34414
+    },
+    {
+      "epoch": 0.819170713129582,
+      "grad_norm": 0.05059671683971603,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34415
+    },
+    {
+      "epoch": 0.8191945158526135,
+      "grad_norm": 2.7181806845342043,
+      "learning_rate": 2e-05,
+      "loss": 0.4326,
+      "step": 34416
+    },
+    {
+      "epoch": 0.8192183185756451,
+      "grad_norm": 6.486057254193024,
+      "learning_rate": 2e-05,
+      "loss": 0.5565,
+      "step": 34417
+    },
+    {
+      "epoch": 0.8192421212986766,
+      "grad_norm": 3.7866775437352347,
+      "learning_rate": 2e-05,
+      "loss": 0.76,
+      "step": 34418
+    },
+    {
+      "epoch": 0.8192659240217081,
+      "grad_norm": 3.4812286758080173,
+      "learning_rate": 2e-05,
+      "loss": 0.4803,
+      "step": 34419
+    },
+    {
+      "epoch": 0.8192897267447395,
+      "grad_norm": 3.2015772105985585,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 34420
+    },
+    {
+      "epoch": 0.8193135294677711,
+      "grad_norm": 3.5096311440582464,
+      "learning_rate": 2e-05,
+      "loss": 0.2149,
+      "step": 34421
+    },
+    {
+      "epoch": 0.8193373321908026,
+      "grad_norm": 3.9374457540422667,
+      "learning_rate": 2e-05,
+      "loss": 0.5163,
+      "step": 34422
+    },
+    {
+      "epoch": 0.8193611349138341,
+      "grad_norm": 4.358398895138645,
+      "learning_rate": 2e-05,
+      "loss": 0.4953,
+      "step": 34423
+    },
+    {
+      "epoch": 0.8193849376368657,
+      "grad_norm": 4.3015574070669516,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 34424
+    },
+    {
+      "epoch": 0.8194087403598972,
+      "grad_norm": 8.465230100762968,
+      "learning_rate": 2e-05,
+      "loss": 0.5675,
+      "step": 34425
+    },
+    {
+      "epoch": 0.8194325430829287,
+      "grad_norm": 5.891250572205825,
+      "learning_rate": 2e-05,
+      "loss": 1.0248,
+      "step": 34426
+    },
+    {
+      "epoch": 0.8194563458059602,
+      "grad_norm": 2.1649610542923474,
+      "learning_rate": 2e-05,
+      "loss": 0.3001,
+      "step": 34427
+    },
+    {
+      "epoch": 0.8194801485289918,
+      "grad_norm": 0.8826828421185564,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 34428
+    },
+    {
+      "epoch": 0.8195039512520232,
+      "grad_norm": 6.897191797558977,
+      "learning_rate": 2e-05,
+      "loss": 1.1096,
+      "step": 34429
+    },
+    {
+      "epoch": 0.8195277539750547,
+      "grad_norm": 3.6472150460154844,
+      "learning_rate": 2e-05,
+      "loss": 0.5893,
+      "step": 34430
+    },
+    {
+      "epoch": 0.8195515566980862,
+      "grad_norm": 12.109337851983096,
+      "learning_rate": 2e-05,
+      "loss": 0.6885,
+      "step": 34431
+    },
+    {
+      "epoch": 0.8195753594211178,
+      "grad_norm": 3.5987698678912245,
+      "learning_rate": 2e-05,
+      "loss": 0.3693,
+      "step": 34432
+    },
+    {
+      "epoch": 0.8195991621441493,
+      "grad_norm": 0.8350139307801986,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 34433
+    },
+    {
+      "epoch": 0.8196229648671808,
+      "grad_norm": 1.5662438785554815,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 34434
+    },
+    {
+      "epoch": 0.8196467675902123,
+      "grad_norm": 10.44182587942244,
+      "learning_rate": 2e-05,
+      "loss": 0.5224,
+      "step": 34435
+    },
+    {
+      "epoch": 0.8196705703132439,
+      "grad_norm": 1.2967603662705893,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 34436
+    },
+    {
+      "epoch": 0.8196943730362753,
+      "grad_norm": 2.1483993019180567,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 34437
+    },
+    {
+      "epoch": 0.8197181757593068,
+      "grad_norm": 6.277802479698549,
+      "learning_rate": 2e-05,
+      "loss": 0.1339,
+      "step": 34438
+    },
+    {
+      "epoch": 0.8197419784823384,
+      "grad_norm": 0.1266475656513132,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 34439
+    },
+    {
+      "epoch": 0.8197657812053699,
+      "grad_norm": 1.002884023055435,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 34440
+    },
+    {
+      "epoch": 0.8197895839284014,
+      "grad_norm": 0.11231897289690262,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 34441
+    },
+    {
+      "epoch": 0.8198133866514329,
+      "grad_norm": 2.061394272905398,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 34442
+    },
+    {
+      "epoch": 0.8198371893744645,
+      "grad_norm": 3.061988060081867,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 34443
+    },
+    {
+      "epoch": 0.819860992097496,
+      "grad_norm": 0.06263598878594405,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 34444
+    },
+    {
+      "epoch": 0.8198847948205275,
+      "grad_norm": 4.315103732272204,
+      "learning_rate": 2e-05,
+      "loss": 0.466,
+      "step": 34445
+    },
+    {
+      "epoch": 0.8199085975435589,
+      "grad_norm": 0.03247983140520154,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34446
+    },
+    {
+      "epoch": 0.8199324002665905,
+      "grad_norm": 4.147495050799875,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 34447
+    },
+    {
+      "epoch": 0.819956202989622,
+      "grad_norm": 4.718392833917634,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 34448
+    },
+    {
+      "epoch": 0.8199800057126535,
+      "grad_norm": 12.822758173176611,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 34449
+    },
+    {
+      "epoch": 0.820003808435685,
+      "grad_norm": 6.518023851412971,
+      "learning_rate": 2e-05,
+      "loss": 0.2448,
+      "step": 34450
+    },
+    {
+      "epoch": 0.8200276111587166,
+      "grad_norm": 5.154977222025041,
+      "learning_rate": 2e-05,
+      "loss": 0.1768,
+      "step": 34451
+    },
+    {
+      "epoch": 0.8200514138817481,
+      "grad_norm": 0.739266237005412,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 34452
+    },
+    {
+      "epoch": 0.8200752166047796,
+      "grad_norm": 4.38949863852626,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 34453
+    },
+    {
+      "epoch": 0.8200990193278112,
+      "grad_norm": 3.6430904474484573,
+      "learning_rate": 2e-05,
+      "loss": 0.5177,
+      "step": 34454
+    },
+    {
+      "epoch": 0.8201228220508426,
+      "grad_norm": 0.07271168127608106,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 34455
+    },
+    {
+      "epoch": 0.8201466247738741,
+      "grad_norm": 2.584315797930537,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 34456
+    },
+    {
+      "epoch": 0.8201704274969056,
+      "grad_norm": 5.242111053797682,
+      "learning_rate": 2e-05,
+      "loss": 1.1106,
+      "step": 34457
+    },
+    {
+      "epoch": 0.8201942302199372,
+      "grad_norm": 3.383188471938485,
+      "learning_rate": 2e-05,
+      "loss": 0.4433,
+      "step": 34458
+    },
+    {
+      "epoch": 0.8202180329429687,
+      "grad_norm": 6.800641141555937,
+      "learning_rate": 2e-05,
+      "loss": 0.3382,
+      "step": 34459
+    },
+    {
+      "epoch": 0.8202418356660002,
+      "grad_norm": 2.7836040954430454,
+      "learning_rate": 2e-05,
+      "loss": 0.3224,
+      "step": 34460
+    },
+    {
+      "epoch": 0.8202656383890317,
+      "grad_norm": 5.448787014673727,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 34461
+    },
+    {
+      "epoch": 0.8202894411120633,
+      "grad_norm": 4.07142964730421,
+      "learning_rate": 2e-05,
+      "loss": 0.8296,
+      "step": 34462
+    },
+    {
+      "epoch": 0.8203132438350947,
+      "grad_norm": 3.97293239934443,
+      "learning_rate": 2e-05,
+      "loss": 0.5523,
+      "step": 34463
+    },
+    {
+      "epoch": 0.8203370465581262,
+      "grad_norm": 6.653531531954289,
+      "learning_rate": 2e-05,
+      "loss": 0.6094,
+      "step": 34464
+    },
+    {
+      "epoch": 0.8203608492811577,
+      "grad_norm": 4.506339619973125,
+      "learning_rate": 2e-05,
+      "loss": 0.107,
+      "step": 34465
+    },
+    {
+      "epoch": 0.8203846520041893,
+      "grad_norm": 3.762715883490911,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 34466
+    },
+    {
+      "epoch": 0.8204084547272208,
+      "grad_norm": 4.001083269109439,
+      "learning_rate": 2e-05,
+      "loss": 0.8643,
+      "step": 34467
+    },
+    {
+      "epoch": 0.8204322574502523,
+      "grad_norm": 5.412852818411846,
+      "learning_rate": 2e-05,
+      "loss": 0.6687,
+      "step": 34468
+    },
+    {
+      "epoch": 0.8204560601732839,
+      "grad_norm": 1.4960472969622327,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 34469
+    },
+    {
+      "epoch": 0.8204798628963154,
+      "grad_norm": 17.85068895510357,
+      "learning_rate": 2e-05,
+      "loss": 0.3542,
+      "step": 34470
+    },
+    {
+      "epoch": 0.8205036656193468,
+      "grad_norm": 0.7087549683109544,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 34471
+    },
+    {
+      "epoch": 0.8205274683423783,
+      "grad_norm": 0.006740741920715678,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 34472
+    },
+    {
+      "epoch": 0.8205512710654099,
+      "grad_norm": 6.897319791931333,
+      "learning_rate": 2e-05,
+      "loss": 1.5373,
+      "step": 34473
+    },
+    {
+      "epoch": 0.8205750737884414,
+      "grad_norm": 0.15707595411883732,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 34474
+    },
+    {
+      "epoch": 0.8205988765114729,
+      "grad_norm": 3.5436228379442682,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 34475
+    },
+    {
+      "epoch": 0.8206226792345044,
+      "grad_norm": 6.847242530031926,
+      "learning_rate": 2e-05,
+      "loss": 0.4637,
+      "step": 34476
+    },
+    {
+      "epoch": 0.820646481957536,
+      "grad_norm": 4.506665264216847,
+      "learning_rate": 2e-05,
+      "loss": 0.8935,
+      "step": 34477
+    },
+    {
+      "epoch": 0.8206702846805675,
+      "grad_norm": 2.3132827125836255,
+      "learning_rate": 2e-05,
+      "loss": 0.352,
+      "step": 34478
+    },
+    {
+      "epoch": 0.820694087403599,
+      "grad_norm": 4.391277503688504,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 34479
+    },
+    {
+      "epoch": 0.8207178901266304,
+      "grad_norm": 4.785200716843504,
+      "learning_rate": 2e-05,
+      "loss": 0.5619,
+      "step": 34480
+    },
+    {
+      "epoch": 0.820741692849662,
+      "grad_norm": 18.431207493345397,
+      "learning_rate": 2e-05,
+      "loss": 0.4714,
+      "step": 34481
+    },
+    {
+      "epoch": 0.8207654955726935,
+      "grad_norm": 5.517866774186993,
+      "learning_rate": 2e-05,
+      "loss": 1.6518,
+      "step": 34482
+    },
+    {
+      "epoch": 0.820789298295725,
+      "grad_norm": 5.1729279910641575,
+      "learning_rate": 2e-05,
+      "loss": 0.229,
+      "step": 34483
+    },
+    {
+      "epoch": 0.8208131010187566,
+      "grad_norm": 0.36814174092537255,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 34484
+    },
+    {
+      "epoch": 0.8208369037417881,
+      "grad_norm": 4.294975503349037,
+      "learning_rate": 2e-05,
+      "loss": 0.532,
+      "step": 34485
+    },
+    {
+      "epoch": 0.8208607064648196,
+      "grad_norm": 2.9651005037926215,
+      "learning_rate": 2e-05,
+      "loss": 0.4198,
+      "step": 34486
+    },
+    {
+      "epoch": 0.820884509187851,
+      "grad_norm": 4.879396240379191,
+      "learning_rate": 2e-05,
+      "loss": 0.5757,
+      "step": 34487
+    },
+    {
+      "epoch": 0.8209083119108826,
+      "grad_norm": 7.628424709090002,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 34488
+    },
+    {
+      "epoch": 0.8209321146339141,
+      "grad_norm": 2.9246009961440587,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 34489
+    },
+    {
+      "epoch": 0.8209559173569456,
+      "grad_norm": 0.029665036327374304,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34490
+    },
+    {
+      "epoch": 0.8209797200799771,
+      "grad_norm": 2.375673848470685,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 34491
+    },
+    {
+      "epoch": 0.8210035228030087,
+      "grad_norm": 4.6770505640489075,
+      "learning_rate": 2e-05,
+      "loss": 0.6668,
+      "step": 34492
+    },
+    {
+      "epoch": 0.8210273255260402,
+      "grad_norm": 0.5521078748186298,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 34493
+    },
+    {
+      "epoch": 0.8210511282490717,
+      "grad_norm": 0.11349639191804485,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 34494
+    },
+    {
+      "epoch": 0.8210749309721032,
+      "grad_norm": 2.6836453337800408,
+      "learning_rate": 2e-05,
+      "loss": 0.2969,
+      "step": 34495
+    },
+    {
+      "epoch": 0.8210987336951348,
+      "grad_norm": 2.443035979194932,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 34496
+    },
+    {
+      "epoch": 0.8211225364181662,
+      "grad_norm": 11.886543042479927,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 34497
+    },
+    {
+      "epoch": 0.8211463391411977,
+      "grad_norm": 7.302508905859415,
+      "learning_rate": 2e-05,
+      "loss": 1.6012,
+      "step": 34498
+    },
+    {
+      "epoch": 0.8211701418642293,
+      "grad_norm": 0.9564407053048866,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 34499
+    },
+    {
+      "epoch": 0.8211939445872608,
+      "grad_norm": 2.096544297666591,
+      "learning_rate": 2e-05,
+      "loss": 0.3435,
+      "step": 34500
+    },
+    {
+      "epoch": 0.8212177473102923,
+      "grad_norm": 4.068444260804708,
+      "learning_rate": 2e-05,
+      "loss": 0.4946,
+      "step": 34501
+    },
+    {
+      "epoch": 0.8212415500333238,
+      "grad_norm": 3.262730876234357,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 34502
+    },
+    {
+      "epoch": 0.8212653527563554,
+      "grad_norm": 5.870839650339211,
+      "learning_rate": 2e-05,
+      "loss": 0.9193,
+      "step": 34503
+    },
+    {
+      "epoch": 0.8212891554793869,
+      "grad_norm": 6.151121259579856,
+      "learning_rate": 2e-05,
+      "loss": 1.2315,
+      "step": 34504
+    },
+    {
+      "epoch": 0.8213129582024183,
+      "grad_norm": 5.8273084465318155,
+      "learning_rate": 2e-05,
+      "loss": 0.7199,
+      "step": 34505
+    },
+    {
+      "epoch": 0.8213367609254498,
+      "grad_norm": 4.588419735954105,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 34506
+    },
+    {
+      "epoch": 0.8213605636484814,
+      "grad_norm": 6.38493653411564,
+      "learning_rate": 2e-05,
+      "loss": 0.1804,
+      "step": 34507
+    },
+    {
+      "epoch": 0.8213843663715129,
+      "grad_norm": 1.2260365150441603,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 34508
+    },
+    {
+      "epoch": 0.8214081690945444,
+      "grad_norm": 6.426652783941923,
+      "learning_rate": 2e-05,
+      "loss": 0.1421,
+      "step": 34509
+    },
+    {
+      "epoch": 0.8214319718175759,
+      "grad_norm": 4.946125390116158,
+      "learning_rate": 2e-05,
+      "loss": 1.2815,
+      "step": 34510
+    },
+    {
+      "epoch": 0.8214557745406075,
+      "grad_norm": 5.707054871203918,
+      "learning_rate": 2e-05,
+      "loss": 1.3089,
+      "step": 34511
+    },
+    {
+      "epoch": 0.821479577263639,
+      "grad_norm": 5.117868028993157,
+      "learning_rate": 2e-05,
+      "loss": 0.7336,
+      "step": 34512
+    },
+    {
+      "epoch": 0.8215033799866704,
+      "grad_norm": 0.10687999457813738,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 34513
+    },
+    {
+      "epoch": 0.821527182709702,
+      "grad_norm": 6.771664078587319,
+      "learning_rate": 2e-05,
+      "loss": 0.2803,
+      "step": 34514
+    },
+    {
+      "epoch": 0.8215509854327335,
+      "grad_norm": 3.8514969525753444,
+      "learning_rate": 2e-05,
+      "loss": 0.4613,
+      "step": 34515
+    },
+    {
+      "epoch": 0.821574788155765,
+      "grad_norm": 2.578101581509356,
+      "learning_rate": 2e-05,
+      "loss": 0.368,
+      "step": 34516
+    },
+    {
+      "epoch": 0.8215985908787965,
+      "grad_norm": 4.025331468912593,
+      "learning_rate": 2e-05,
+      "loss": 0.6221,
+      "step": 34517
+    },
+    {
+      "epoch": 0.8216223936018281,
+      "grad_norm": 0.7423420472189928,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 34518
+    },
+    {
+      "epoch": 0.8216461963248596,
+      "grad_norm": 2.831038614384232,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 34519
+    },
+    {
+      "epoch": 0.8216699990478911,
+      "grad_norm": 6.442680781827629,
+      "learning_rate": 2e-05,
+      "loss": 0.2172,
+      "step": 34520
+    },
+    {
+      "epoch": 0.8216938017709225,
+      "grad_norm": 7.093154196574958,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 34521
+    },
+    {
+      "epoch": 0.8217176044939541,
+      "grad_norm": 7.7929712462991105,
+      "learning_rate": 2e-05,
+      "loss": 1.8496,
+      "step": 34522
+    },
+    {
+      "epoch": 0.8217414072169856,
+      "grad_norm": 4.31077458493544,
+      "learning_rate": 2e-05,
+      "loss": 0.6111,
+      "step": 34523
+    },
+    {
+      "epoch": 0.8217652099400171,
+      "grad_norm": 3.0886706402228667,
+      "learning_rate": 2e-05,
+      "loss": 0.1717,
+      "step": 34524
+    },
+    {
+      "epoch": 0.8217890126630486,
+      "grad_norm": 25.89203666374524,
+      "learning_rate": 2e-05,
+      "loss": 0.614,
+      "step": 34525
+    },
+    {
+      "epoch": 0.8218128153860802,
+      "grad_norm": 1.9149454467879188,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 34526
+    },
+    {
+      "epoch": 0.8218366181091117,
+      "grad_norm": 3.9669612054416152,
+      "learning_rate": 2e-05,
+      "loss": 0.1099,
+      "step": 34527
+    },
+    {
+      "epoch": 0.8218604208321432,
+      "grad_norm": 1.002250209585,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 34528
+    },
+    {
+      "epoch": 0.8218842235551748,
+      "grad_norm": 3.607116940288522,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 34529
+    },
+    {
+      "epoch": 0.8219080262782062,
+      "grad_norm": 1.1005007138862823,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 34530
+    },
+    {
+      "epoch": 0.8219318290012377,
+      "grad_norm": 5.212710612864032,
+      "learning_rate": 2e-05,
+      "loss": 0.7969,
+      "step": 34531
+    },
+    {
+      "epoch": 0.8219556317242692,
+      "grad_norm": 7.06967203104561,
+      "learning_rate": 2e-05,
+      "loss": 0.362,
+      "step": 34532
+    },
+    {
+      "epoch": 0.8219794344473008,
+      "grad_norm": 1.0230839068762643,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 34533
+    },
+    {
+      "epoch": 0.8220032371703323,
+      "grad_norm": 0.24658934177893566,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 34534
+    },
+    {
+      "epoch": 0.8220270398933638,
+      "grad_norm": 3.2221173152120817,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 34535
+    },
+    {
+      "epoch": 0.8220508426163953,
+      "grad_norm": 4.323135868963113,
+      "learning_rate": 2e-05,
+      "loss": 0.3281,
+      "step": 34536
+    },
+    {
+      "epoch": 0.8220746453394269,
+      "grad_norm": 6.45059180620629,
+      "learning_rate": 2e-05,
+      "loss": 0.6978,
+      "step": 34537
+    },
+    {
+      "epoch": 0.8220984480624584,
+      "grad_norm": 0.10801270817972125,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 34538
+    },
+    {
+      "epoch": 0.8221222507854898,
+      "grad_norm": 5.935268924814399,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 34539
+    },
+    {
+      "epoch": 0.8221460535085213,
+      "grad_norm": 1.8812987589605366,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 34540
+    },
+    {
+      "epoch": 0.8221698562315529,
+      "grad_norm": 0.6199883639210405,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 34541
+    },
+    {
+      "epoch": 0.8221936589545844,
+      "grad_norm": 2.2346117557486482,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 34542
+    },
+    {
+      "epoch": 0.8222174616776159,
+      "grad_norm": 0.06813668494816646,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 34543
+    },
+    {
+      "epoch": 0.8222412644006475,
+      "grad_norm": 8.198319307173191,
+      "learning_rate": 2e-05,
+      "loss": 0.8821,
+      "step": 34544
+    },
+    {
+      "epoch": 0.822265067123679,
+      "grad_norm": 14.303640149859667,
+      "learning_rate": 2e-05,
+      "loss": 0.2696,
+      "step": 34545
+    },
+    {
+      "epoch": 0.8222888698467105,
+      "grad_norm": 3.192599692940782,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 34546
+    },
+    {
+      "epoch": 0.8223126725697419,
+      "grad_norm": 2.1175365372697765,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 34547
+    },
+    {
+      "epoch": 0.8223364752927735,
+      "grad_norm": 0.12662963374134234,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 34548
+    },
+    {
+      "epoch": 0.822360278015805,
+      "grad_norm": 4.256577340938135,
+      "learning_rate": 2e-05,
+      "loss": 0.6636,
+      "step": 34549
+    },
+    {
+      "epoch": 0.8223840807388365,
+      "grad_norm": 0.13886079157448863,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 34550
+    },
+    {
+      "epoch": 0.822407883461868,
+      "grad_norm": 5.531812690436746,
+      "learning_rate": 2e-05,
+      "loss": 0.8699,
+      "step": 34551
+    },
+    {
+      "epoch": 0.8224316861848996,
+      "grad_norm": 0.9348788242113246,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 34552
+    },
+    {
+      "epoch": 0.8224554889079311,
+      "grad_norm": 7.559561325466683,
+      "learning_rate": 2e-05,
+      "loss": 0.6471,
+      "step": 34553
+    },
+    {
+      "epoch": 0.8224792916309626,
+      "grad_norm": 3.086432706415598,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 34554
+    },
+    {
+      "epoch": 0.822503094353994,
+      "grad_norm": 6.056696444826327,
+      "learning_rate": 2e-05,
+      "loss": 0.4404,
+      "step": 34555
+    },
+    {
+      "epoch": 0.8225268970770256,
+      "grad_norm": 0.1838682073198536,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 34556
+    },
+    {
+      "epoch": 0.8225506998000571,
+      "grad_norm": 0.17710323783943743,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 34557
+    },
+    {
+      "epoch": 0.8225745025230886,
+      "grad_norm": 2.1367583850128504,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 34558
+    },
+    {
+      "epoch": 0.8225983052461202,
+      "grad_norm": 3.5729159285585235,
+      "learning_rate": 2e-05,
+      "loss": 0.3904,
+      "step": 34559
+    },
+    {
+      "epoch": 0.8226221079691517,
+      "grad_norm": 6.13956399154382,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 34560
+    },
+    {
+      "epoch": 0.8226459106921832,
+      "grad_norm": 2.357123145199756,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 34561
+    },
+    {
+      "epoch": 0.8226697134152147,
+      "grad_norm": 4.677453634163923,
+      "learning_rate": 2e-05,
+      "loss": 0.5588,
+      "step": 34562
+    },
+    {
+      "epoch": 0.8226935161382463,
+      "grad_norm": 0.15065604857651754,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 34563
+    },
+    {
+      "epoch": 0.8227173188612777,
+      "grad_norm": 2.5755822601942833,
+      "learning_rate": 2e-05,
+      "loss": 0.2742,
+      "step": 34564
+    },
+    {
+      "epoch": 0.8227411215843092,
+      "grad_norm": 4.678904592193374,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 34565
+    },
+    {
+      "epoch": 0.8227649243073407,
+      "grad_norm": 3.9542820936161216,
+      "learning_rate": 2e-05,
+      "loss": 0.369,
+      "step": 34566
+    },
+    {
+      "epoch": 0.8227887270303723,
+      "grad_norm": 0.5555574456934231,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 34567
+    },
+    {
+      "epoch": 0.8228125297534038,
+      "grad_norm": 0.15327513558060069,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34568
+    },
+    {
+      "epoch": 0.8228363324764353,
+      "grad_norm": 3.7390711110957247,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 34569
+    },
+    {
+      "epoch": 0.8228601351994668,
+      "grad_norm": 8.034436649271465,
+      "learning_rate": 2e-05,
+      "loss": 0.3542,
+      "step": 34570
+    },
+    {
+      "epoch": 0.8228839379224984,
+      "grad_norm": 0.2503744078573854,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 34571
+    },
+    {
+      "epoch": 0.8229077406455299,
+      "grad_norm": 0.04539596747541778,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 34572
+    },
+    {
+      "epoch": 0.8229315433685613,
+      "grad_norm": 4.905288771795802,
+      "learning_rate": 2e-05,
+      "loss": 0.9613,
+      "step": 34573
+    },
+    {
+      "epoch": 0.8229553460915929,
+      "grad_norm": 11.928174490708207,
+      "learning_rate": 2e-05,
+      "loss": 0.53,
+      "step": 34574
+    },
+    {
+      "epoch": 0.8229791488146244,
+      "grad_norm": 9.809994272128735,
+      "learning_rate": 2e-05,
+      "loss": 0.2382,
+      "step": 34575
+    },
+    {
+      "epoch": 0.8230029515376559,
+      "grad_norm": 1.3967981251707653,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 34576
+    },
+    {
+      "epoch": 0.8230267542606874,
+      "grad_norm": 4.041042766369093,
+      "learning_rate": 2e-05,
+      "loss": 0.022,
+      "step": 34577
+    },
+    {
+      "epoch": 0.823050556983719,
+      "grad_norm": 3.000552132455327,
+      "learning_rate": 2e-05,
+      "loss": 0.5464,
+      "step": 34578
+    },
+    {
+      "epoch": 0.8230743597067505,
+      "grad_norm": 2.7558101414001737,
+      "learning_rate": 2e-05,
+      "loss": 0.615,
+      "step": 34579
+    },
+    {
+      "epoch": 0.823098162429782,
+      "grad_norm": 7.609908108183525,
+      "learning_rate": 2e-05,
+      "loss": 0.6944,
+      "step": 34580
+    },
+    {
+      "epoch": 0.8231219651528134,
+      "grad_norm": 3.0853425076845022,
+      "learning_rate": 2e-05,
+      "loss": 0.6171,
+      "step": 34581
+    },
+    {
+      "epoch": 0.823145767875845,
+      "grad_norm": 0.16770199476156264,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 34582
+    },
+    {
+      "epoch": 0.8231695705988765,
+      "grad_norm": 8.773835923233687,
+      "learning_rate": 2e-05,
+      "loss": 0.9542,
+      "step": 34583
+    },
+    {
+      "epoch": 0.823193373321908,
+      "grad_norm": 18.728969045591707,
+      "learning_rate": 2e-05,
+      "loss": 0.5726,
+      "step": 34584
+    },
+    {
+      "epoch": 0.8232171760449395,
+      "grad_norm": 2.0827780007661714,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 34585
+    },
+    {
+      "epoch": 0.8232409787679711,
+      "grad_norm": 0.1563297386976635,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 34586
+    },
+    {
+      "epoch": 0.8232647814910026,
+      "grad_norm": 10.827866766946288,
+      "learning_rate": 2e-05,
+      "loss": 0.6904,
+      "step": 34587
+    },
+    {
+      "epoch": 0.8232885842140341,
+      "grad_norm": 5.202485820613104,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 34588
+    },
+    {
+      "epoch": 0.8233123869370657,
+      "grad_norm": 4.830839053285826,
+      "learning_rate": 2e-05,
+      "loss": 0.8319,
+      "step": 34589
+    },
+    {
+      "epoch": 0.8233361896600971,
+      "grad_norm": 6.6636123995778025,
+      "learning_rate": 2e-05,
+      "loss": 0.1932,
+      "step": 34590
+    },
+    {
+      "epoch": 0.8233599923831286,
+      "grad_norm": 7.838062003903264,
+      "learning_rate": 2e-05,
+      "loss": 0.5845,
+      "step": 34591
+    },
+    {
+      "epoch": 0.8233837951061601,
+      "grad_norm": 6.398275042462192,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 34592
+    },
+    {
+      "epoch": 0.8234075978291917,
+      "grad_norm": 8.494741805951483,
+      "learning_rate": 2e-05,
+      "loss": 0.9696,
+      "step": 34593
+    },
+    {
+      "epoch": 0.8234314005522232,
+      "grad_norm": 0.38216801708615816,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 34594
+    },
+    {
+      "epoch": 0.8234552032752547,
+      "grad_norm": 3.5294736366867276,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 34595
+    },
+    {
+      "epoch": 0.8234790059982862,
+      "grad_norm": 0.3826349591320134,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 34596
+    },
+    {
+      "epoch": 0.8235028087213178,
+      "grad_norm": 24.597370146402724,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 34597
+    },
+    {
+      "epoch": 0.8235266114443492,
+      "grad_norm": 2.238936140481443,
+      "learning_rate": 2e-05,
+      "loss": 0.3525,
+      "step": 34598
+    },
+    {
+      "epoch": 0.8235504141673807,
+      "grad_norm": 10.329415541578749,
+      "learning_rate": 2e-05,
+      "loss": 0.5564,
+      "step": 34599
+    },
+    {
+      "epoch": 0.8235742168904122,
+      "grad_norm": 25.53183556955316,
+      "learning_rate": 2e-05,
+      "loss": 0.5978,
+      "step": 34600
+    },
+    {
+      "epoch": 0.8235980196134438,
+      "grad_norm": 0.05531649993018039,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 34601
+    },
+    {
+      "epoch": 0.8236218223364753,
+      "grad_norm": 13.044199938727665,
+      "learning_rate": 2e-05,
+      "loss": 0.6363,
+      "step": 34602
+    },
+    {
+      "epoch": 0.8236456250595068,
+      "grad_norm": 1.6869740198456402,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 34603
+    },
+    {
+      "epoch": 0.8236694277825384,
+      "grad_norm": 2.625729349219003,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 34604
+    },
+    {
+      "epoch": 0.8236932305055699,
+      "grad_norm": 0.7762617668456879,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 34605
+    },
+    {
+      "epoch": 0.8237170332286013,
+      "grad_norm": 0.3267336893347426,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 34606
+    },
+    {
+      "epoch": 0.8237408359516328,
+      "grad_norm": 14.98840658069119,
+      "learning_rate": 2e-05,
+      "loss": 0.2726,
+      "step": 34607
+    },
+    {
+      "epoch": 0.8237646386746644,
+      "grad_norm": 6.201400082394927,
+      "learning_rate": 2e-05,
+      "loss": 0.4645,
+      "step": 34608
+    },
+    {
+      "epoch": 0.8237884413976959,
+      "grad_norm": 3.630689368308282,
+      "learning_rate": 2e-05,
+      "loss": 0.4818,
+      "step": 34609
+    },
+    {
+      "epoch": 0.8238122441207274,
+      "grad_norm": 6.113916404144677,
+      "learning_rate": 2e-05,
+      "loss": 1.3856,
+      "step": 34610
+    },
+    {
+      "epoch": 0.8238360468437589,
+      "grad_norm": 0.018286613074283302,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 34611
+    },
+    {
+      "epoch": 0.8238598495667905,
+      "grad_norm": 0.5107016220500498,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 34612
+    },
+    {
+      "epoch": 0.823883652289822,
+      "grad_norm": 3.7394761018742417,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 34613
+    },
+    {
+      "epoch": 0.8239074550128535,
+      "grad_norm": 9.263527930815222,
+      "learning_rate": 2e-05,
+      "loss": 0.4778,
+      "step": 34614
+    },
+    {
+      "epoch": 0.8239312577358849,
+      "grad_norm": 2.779000593012812,
+      "learning_rate": 2e-05,
+      "loss": 0.112,
+      "step": 34615
+    },
+    {
+      "epoch": 0.8239550604589165,
+      "grad_norm": 0.2928051152554578,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 34616
+    },
+    {
+      "epoch": 0.823978863181948,
+      "grad_norm": 0.041087726870022226,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34617
+    },
+    {
+      "epoch": 0.8240026659049795,
+      "grad_norm": 3.454068720986767,
+      "learning_rate": 2e-05,
+      "loss": 0.3932,
+      "step": 34618
+    },
+    {
+      "epoch": 0.8240264686280111,
+      "grad_norm": 2.936506992553722,
+      "learning_rate": 2e-05,
+      "loss": 0.2584,
+      "step": 34619
+    },
+    {
+      "epoch": 0.8240502713510426,
+      "grad_norm": 1.1294851363193745,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 34620
+    },
+    {
+      "epoch": 0.8240740740740741,
+      "grad_norm": 7.117807414105984,
+      "learning_rate": 2e-05,
+      "loss": 0.6899,
+      "step": 34621
+    },
+    {
+      "epoch": 0.8240978767971056,
+      "grad_norm": 7.179893749127658,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 34622
+    },
+    {
+      "epoch": 0.8241216795201372,
+      "grad_norm": 9.99412456253419,
+      "learning_rate": 2e-05,
+      "loss": 0.2188,
+      "step": 34623
+    },
+    {
+      "epoch": 0.8241454822431686,
+      "grad_norm": 3.867454912796862,
+      "learning_rate": 2e-05,
+      "loss": 0.3782,
+      "step": 34624
+    },
+    {
+      "epoch": 0.8241692849662001,
+      "grad_norm": 0.3008844554527534,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 34625
+    },
+    {
+      "epoch": 0.8241930876892316,
+      "grad_norm": 7.9341623952938125,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 34626
+    },
+    {
+      "epoch": 0.8242168904122632,
+      "grad_norm": 0.1290473326618681,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34627
+    },
+    {
+      "epoch": 0.8242406931352947,
+      "grad_norm": 5.057915304084796,
+      "learning_rate": 2e-05,
+      "loss": 0.7876,
+      "step": 34628
+    },
+    {
+      "epoch": 0.8242644958583262,
+      "grad_norm": 6.434930767954409,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 34629
+    },
+    {
+      "epoch": 0.8242882985813577,
+      "grad_norm": 0.04587851923309602,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 34630
+    },
+    {
+      "epoch": 0.8243121013043893,
+      "grad_norm": 4.48219269392296,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 34631
+    },
+    {
+      "epoch": 0.8243359040274207,
+      "grad_norm": 0.9938751901498963,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 34632
+    },
+    {
+      "epoch": 0.8243597067504522,
+      "grad_norm": 2.214097632298962,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 34633
+    },
+    {
+      "epoch": 0.8243835094734838,
+      "grad_norm": 8.037296173666439,
+      "learning_rate": 2e-05,
+      "loss": 0.9778,
+      "step": 34634
+    },
+    {
+      "epoch": 0.8244073121965153,
+      "grad_norm": 0.7404086933891355,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 34635
+    },
+    {
+      "epoch": 0.8244311149195468,
+      "grad_norm": 0.06426999909995214,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34636
+    },
+    {
+      "epoch": 0.8244549176425783,
+      "grad_norm": 0.1477642373321606,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34637
+    },
+    {
+      "epoch": 0.8244787203656099,
+      "grad_norm": 0.2027840091152493,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34638
+    },
+    {
+      "epoch": 0.8245025230886414,
+      "grad_norm": 2.320648943416545,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 34639
+    },
+    {
+      "epoch": 0.8245263258116728,
+      "grad_norm": 8.143643266016555,
+      "learning_rate": 2e-05,
+      "loss": 1.1468,
+      "step": 34640
+    },
+    {
+      "epoch": 0.8245501285347043,
+      "grad_norm": 8.534545773856108,
+      "learning_rate": 2e-05,
+      "loss": 0.1942,
+      "step": 34641
+    },
+    {
+      "epoch": 0.8245739312577359,
+      "grad_norm": 1.1907055159150466,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 34642
+    },
+    {
+      "epoch": 0.8245977339807674,
+      "grad_norm": 0.07864844505001434,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 34643
+    },
+    {
+      "epoch": 0.8246215367037989,
+      "grad_norm": 0.07338003123916165,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 34644
+    },
+    {
+      "epoch": 0.8246453394268304,
+      "grad_norm": 8.466310620000936,
+      "learning_rate": 2e-05,
+      "loss": 0.3883,
+      "step": 34645
+    },
+    {
+      "epoch": 0.824669142149862,
+      "grad_norm": 0.5219773510704585,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 34646
+    },
+    {
+      "epoch": 0.8246929448728935,
+      "grad_norm": 4.025033626493869,
+      "learning_rate": 2e-05,
+      "loss": 1.0372,
+      "step": 34647
+    },
+    {
+      "epoch": 0.824716747595925,
+      "grad_norm": 2.05230937326438,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 34648
+    },
+    {
+      "epoch": 0.8247405503189564,
+      "grad_norm": 2.7683924184422066,
+      "learning_rate": 2e-05,
+      "loss": 0.6534,
+      "step": 34649
+    },
+    {
+      "epoch": 0.824764353041988,
+      "grad_norm": 3.6778366583090274,
+      "learning_rate": 2e-05,
+      "loss": 0.5947,
+      "step": 34650
+    },
+    {
+      "epoch": 0.8247881557650195,
+      "grad_norm": 2.4864628259773425,
+      "learning_rate": 2e-05,
+      "loss": 0.1347,
+      "step": 34651
+    },
+    {
+      "epoch": 0.824811958488051,
+      "grad_norm": 3.8169079386336353,
+      "learning_rate": 2e-05,
+      "loss": 0.1305,
+      "step": 34652
+    },
+    {
+      "epoch": 0.8248357612110826,
+      "grad_norm": 5.7488095315888845,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 34653
+    },
+    {
+      "epoch": 0.8248595639341141,
+      "grad_norm": 1.187432363006717,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 34654
+    },
+    {
+      "epoch": 0.8248833666571456,
+      "grad_norm": 3.6468796654981315,
+      "learning_rate": 2e-05,
+      "loss": 0.5053,
+      "step": 34655
+    },
+    {
+      "epoch": 0.824907169380177,
+      "grad_norm": 5.414394262786908,
+      "learning_rate": 2e-05,
+      "loss": 0.758,
+      "step": 34656
+    },
+    {
+      "epoch": 0.8249309721032086,
+      "grad_norm": 3.415059190736964,
+      "learning_rate": 2e-05,
+      "loss": 0.5156,
+      "step": 34657
+    },
+    {
+      "epoch": 0.8249547748262401,
+      "grad_norm": 9.759376516558076,
+      "learning_rate": 2e-05,
+      "loss": 0.3741,
+      "step": 34658
+    },
+    {
+      "epoch": 0.8249785775492716,
+      "grad_norm": 15.689500956607088,
+      "learning_rate": 2e-05,
+      "loss": 1.5782,
+      "step": 34659
+    },
+    {
+      "epoch": 0.8250023802723031,
+      "grad_norm": 3.4525443670702636,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 34660
+    },
+    {
+      "epoch": 0.8250261829953347,
+      "grad_norm": 3.370532492303396,
+      "learning_rate": 2e-05,
+      "loss": 0.436,
+      "step": 34661
+    },
+    {
+      "epoch": 0.8250499857183662,
+      "grad_norm": 1.5929296607723769,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 34662
+    },
+    {
+      "epoch": 0.8250737884413977,
+      "grad_norm": 2.765456822704506,
+      "learning_rate": 2e-05,
+      "loss": 0.1609,
+      "step": 34663
+    },
+    {
+      "epoch": 0.8250975911644292,
+      "grad_norm": 4.137634379926473,
+      "learning_rate": 2e-05,
+      "loss": 0.0621,
+      "step": 34664
+    },
+    {
+      "epoch": 0.8251213938874608,
+      "grad_norm": 0.8621489999562779,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 34665
+    },
+    {
+      "epoch": 0.8251451966104922,
+      "grad_norm": 5.299139626303952,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 34666
+    },
+    {
+      "epoch": 0.8251689993335237,
+      "grad_norm": 4.0603698348396895,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 34667
+    },
+    {
+      "epoch": 0.8251928020565553,
+      "grad_norm": 1.6304756214355702,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 34668
+    },
+    {
+      "epoch": 0.8252166047795868,
+      "grad_norm": 2.0263130117800916,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 34669
+    },
+    {
+      "epoch": 0.8252404075026183,
+      "grad_norm": 2.5869717825080256,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 34670
+    },
+    {
+      "epoch": 0.8252642102256498,
+      "grad_norm": 2.8071592948012083,
+      "learning_rate": 2e-05,
+      "loss": 0.0504,
+      "step": 34671
+    },
+    {
+      "epoch": 0.8252880129486814,
+      "grad_norm": 2.135656079709035,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 34672
+    },
+    {
+      "epoch": 0.8253118156717129,
+      "grad_norm": 0.5990315646301879,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 34673
+    },
+    {
+      "epoch": 0.8253356183947443,
+      "grad_norm": 2.2284301074266657,
+      "learning_rate": 2e-05,
+      "loss": 0.2465,
+      "step": 34674
+    },
+    {
+      "epoch": 0.8253594211177758,
+      "grad_norm": 4.548787766686362,
+      "learning_rate": 2e-05,
+      "loss": 0.0859,
+      "step": 34675
+    },
+    {
+      "epoch": 0.8253832238408074,
+      "grad_norm": 3.8473949230141318,
+      "learning_rate": 2e-05,
+      "loss": 0.4074,
+      "step": 34676
+    },
+    {
+      "epoch": 0.8254070265638389,
+      "grad_norm": 1.2590055373097075,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 34677
+    },
+    {
+      "epoch": 0.8254308292868704,
+      "grad_norm": 4.946653330822379,
+      "learning_rate": 2e-05,
+      "loss": 0.3843,
+      "step": 34678
+    },
+    {
+      "epoch": 0.8254546320099019,
+      "grad_norm": 4.917290164723843,
+      "learning_rate": 2e-05,
+      "loss": 0.7956,
+      "step": 34679
+    },
+    {
+      "epoch": 0.8254784347329335,
+      "grad_norm": 1.3742766779575515,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 34680
+    },
+    {
+      "epoch": 0.825502237455965,
+      "grad_norm": 1.379079783066413,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 34681
+    },
+    {
+      "epoch": 0.8255260401789964,
+      "grad_norm": 2.9529435991750215,
+      "learning_rate": 2e-05,
+      "loss": 0.4239,
+      "step": 34682
+    },
+    {
+      "epoch": 0.825549842902028,
+      "grad_norm": 0.026425620648143567,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34683
+    },
+    {
+      "epoch": 0.8255736456250595,
+      "grad_norm": 2.6682799242197905,
+      "learning_rate": 2e-05,
+      "loss": 0.4826,
+      "step": 34684
+    },
+    {
+      "epoch": 0.825597448348091,
+      "grad_norm": 4.252120566665559,
+      "learning_rate": 2e-05,
+      "loss": 0.9424,
+      "step": 34685
+    },
+    {
+      "epoch": 0.8256212510711225,
+      "grad_norm": 7.291194113013943,
+      "learning_rate": 2e-05,
+      "loss": 0.6895,
+      "step": 34686
+    },
+    {
+      "epoch": 0.8256450537941541,
+      "grad_norm": 5.997104841889178,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 34687
+    },
+    {
+      "epoch": 0.8256688565171856,
+      "grad_norm": 4.224614874070502,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 34688
+    },
+    {
+      "epoch": 0.8256926592402171,
+      "grad_norm": 0.1337713380819293,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 34689
+    },
+    {
+      "epoch": 0.8257164619632485,
+      "grad_norm": 0.8897647957900993,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 34690
+    },
+    {
+      "epoch": 0.8257402646862801,
+      "grad_norm": 0.7132279471444878,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 34691
+    },
+    {
+      "epoch": 0.8257640674093116,
+      "grad_norm": 1.0216045148178863,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 34692
+    },
+    {
+      "epoch": 0.8257878701323431,
+      "grad_norm": 4.243950959768679,
+      "learning_rate": 2e-05,
+      "loss": 0.0836,
+      "step": 34693
+    },
+    {
+      "epoch": 0.8258116728553746,
+      "grad_norm": 0.07507174843805305,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 34694
+    },
+    {
+      "epoch": 0.8258354755784062,
+      "grad_norm": 0.18106145369027674,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34695
+    },
+    {
+      "epoch": 0.8258592783014377,
+      "grad_norm": 0.6382228605234734,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 34696
+    },
+    {
+      "epoch": 0.8258830810244692,
+      "grad_norm": 0.04621663142234592,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34697
+    },
+    {
+      "epoch": 0.8259068837475008,
+      "grad_norm": 6.306052148239148,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 34698
+    },
+    {
+      "epoch": 0.8259306864705322,
+      "grad_norm": 2.6062384396170755,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 34699
+    },
+    {
+      "epoch": 0.8259544891935637,
+      "grad_norm": 6.5723769227375515,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 34700
+    },
+    {
+      "epoch": 0.8259782919165952,
+      "grad_norm": 0.4028069078957542,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 34701
+    },
+    {
+      "epoch": 0.8260020946396268,
+      "grad_norm": 3.597776754834365,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 34702
+    },
+    {
+      "epoch": 0.8260258973626583,
+      "grad_norm": 0.01775657292813396,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 34703
+    },
+    {
+      "epoch": 0.8260497000856898,
+      "grad_norm": 0.1861838985088461,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 34704
+    },
+    {
+      "epoch": 0.8260735028087213,
+      "grad_norm": 3.3445127995236295,
+      "learning_rate": 2e-05,
+      "loss": 0.8003,
+      "step": 34705
+    },
+    {
+      "epoch": 0.8260973055317529,
+      "grad_norm": 0.2747042649408352,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 34706
+    },
+    {
+      "epoch": 0.8261211082547844,
+      "grad_norm": 1.4966896515900445,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 34707
+    },
+    {
+      "epoch": 0.8261449109778158,
+      "grad_norm": 3.111593570431418,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 34708
+    },
+    {
+      "epoch": 0.8261687137008473,
+      "grad_norm": 1.30618509386808,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 34709
+    },
+    {
+      "epoch": 0.8261925164238789,
+      "grad_norm": 3.382682759914546,
+      "learning_rate": 2e-05,
+      "loss": 0.6807,
+      "step": 34710
+    },
+    {
+      "epoch": 0.8262163191469104,
+      "grad_norm": 3.3771853630949122,
+      "learning_rate": 2e-05,
+      "loss": 0.6233,
+      "step": 34711
+    },
+    {
+      "epoch": 0.8262401218699419,
+      "grad_norm": 0.017148120232420863,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 34712
+    },
+    {
+      "epoch": 0.8262639245929735,
+      "grad_norm": 4.3761927702806975,
+      "learning_rate": 2e-05,
+      "loss": 1.0798,
+      "step": 34713
+    },
+    {
+      "epoch": 0.826287727316005,
+      "grad_norm": 5.35151772159483,
+      "learning_rate": 2e-05,
+      "loss": 1.6107,
+      "step": 34714
+    },
+    {
+      "epoch": 0.8263115300390365,
+      "grad_norm": 6.486603979990264,
+      "learning_rate": 2e-05,
+      "loss": 0.2388,
+      "step": 34715
+    },
+    {
+      "epoch": 0.8263353327620679,
+      "grad_norm": 3.756588278615203,
+      "learning_rate": 2e-05,
+      "loss": 0.6408,
+      "step": 34716
+    },
+    {
+      "epoch": 0.8263591354850995,
+      "grad_norm": 6.3921506454631905,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 34717
+    },
+    {
+      "epoch": 0.826382938208131,
+      "grad_norm": 0.007535781627633524,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 34718
+    },
+    {
+      "epoch": 0.8264067409311625,
+      "grad_norm": 57.66639937961836,
+      "learning_rate": 2e-05,
+      "loss": 0.2268,
+      "step": 34719
+    },
+    {
+      "epoch": 0.826430543654194,
+      "grad_norm": 8.62701043873791,
+      "learning_rate": 2e-05,
+      "loss": 0.2615,
+      "step": 34720
+    },
+    {
+      "epoch": 0.8264543463772256,
+      "grad_norm": 1.00264705085586,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 34721
+    },
+    {
+      "epoch": 0.8264781491002571,
+      "grad_norm": 3.6171036371500707,
+      "learning_rate": 2e-05,
+      "loss": 0.6359,
+      "step": 34722
+    },
+    {
+      "epoch": 0.8265019518232886,
+      "grad_norm": 1.739813400009677,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 34723
+    },
+    {
+      "epoch": 0.82652575454632,
+      "grad_norm": 0.04150189528307255,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34724
+    },
+    {
+      "epoch": 0.8265495572693516,
+      "grad_norm": 5.578040658171721,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 34725
+    },
+    {
+      "epoch": 0.8265733599923831,
+      "grad_norm": 8.358981273148038,
+      "learning_rate": 2e-05,
+      "loss": 2.3096,
+      "step": 34726
+    },
+    {
+      "epoch": 0.8265971627154146,
+      "grad_norm": 4.86733278388099,
+      "learning_rate": 2e-05,
+      "loss": 0.6754,
+      "step": 34727
+    },
+    {
+      "epoch": 0.8266209654384462,
+      "grad_norm": 3.123640651323706,
+      "learning_rate": 2e-05,
+      "loss": 0.7683,
+      "step": 34728
+    },
+    {
+      "epoch": 0.8266447681614777,
+      "grad_norm": 3.7269088032071283,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 34729
+    },
+    {
+      "epoch": 0.8266685708845092,
+      "grad_norm": 5.32041676159123,
+      "learning_rate": 2e-05,
+      "loss": 1.0528,
+      "step": 34730
+    },
+    {
+      "epoch": 0.8266923736075407,
+      "grad_norm": 4.808318349601084,
+      "learning_rate": 2e-05,
+      "loss": 0.8919,
+      "step": 34731
+    },
+    {
+      "epoch": 0.8267161763305723,
+      "grad_norm": 0.015193200207031162,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 34732
+    },
+    {
+      "epoch": 0.8267399790536037,
+      "grad_norm": 6.497678864521574,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 34733
+    },
+    {
+      "epoch": 0.8267637817766352,
+      "grad_norm": 4.277540325865572,
+      "learning_rate": 2e-05,
+      "loss": 0.1599,
+      "step": 34734
+    },
+    {
+      "epoch": 0.8267875844996667,
+      "grad_norm": 4.460629206806414,
+      "learning_rate": 2e-05,
+      "loss": 1.0946,
+      "step": 34735
+    },
+    {
+      "epoch": 0.8268113872226983,
+      "grad_norm": 3.5128665538081987,
+      "learning_rate": 2e-05,
+      "loss": 0.6015,
+      "step": 34736
+    },
+    {
+      "epoch": 0.8268351899457298,
+      "grad_norm": 3.737558355702299,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 34737
+    },
+    {
+      "epoch": 0.8268589926687613,
+      "grad_norm": 7.201292243364863,
+      "learning_rate": 2e-05,
+      "loss": 0.2601,
+      "step": 34738
+    },
+    {
+      "epoch": 0.8268827953917928,
+      "grad_norm": 0.5959058074395995,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 34739
+    },
+    {
+      "epoch": 0.8269065981148244,
+      "grad_norm": 4.769489035873176,
+      "learning_rate": 2e-05,
+      "loss": 1.2912,
+      "step": 34740
+    },
+    {
+      "epoch": 0.8269304008378559,
+      "grad_norm": 2.949073706553259,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 34741
+    },
+    {
+      "epoch": 0.8269542035608873,
+      "grad_norm": 15.239017377560659,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 34742
+    },
+    {
+      "epoch": 0.8269780062839189,
+      "grad_norm": 1.832843412290018,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 34743
+    },
+    {
+      "epoch": 0.8270018090069504,
+      "grad_norm": 3.5779770836500444,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 34744
+    },
+    {
+      "epoch": 0.8270256117299819,
+      "grad_norm": 3.361618839623628,
+      "learning_rate": 2e-05,
+      "loss": 0.3859,
+      "step": 34745
+    },
+    {
+      "epoch": 0.8270494144530134,
+      "grad_norm": 0.14365456497949652,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 34746
+    },
+    {
+      "epoch": 0.827073217176045,
+      "grad_norm": 0.6846338567214177,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 34747
+    },
+    {
+      "epoch": 0.8270970198990765,
+      "grad_norm": 0.8812992014698616,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 34748
+    },
+    {
+      "epoch": 0.827120822622108,
+      "grad_norm": 5.32151577877999,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 34749
+    },
+    {
+      "epoch": 0.8271446253451394,
+      "grad_norm": 4.300922129129732,
+      "learning_rate": 2e-05,
+      "loss": 0.9403,
+      "step": 34750
+    },
+    {
+      "epoch": 0.827168428068171,
+      "grad_norm": 0.5834179163428455,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 34751
+    },
+    {
+      "epoch": 0.8271922307912025,
+      "grad_norm": 0.14548844541459335,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 34752
+    },
+    {
+      "epoch": 0.827216033514234,
+      "grad_norm": 5.391399751005081,
+      "learning_rate": 2e-05,
+      "loss": 0.9384,
+      "step": 34753
+    },
+    {
+      "epoch": 0.8272398362372655,
+      "grad_norm": 0.17983197844946525,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 34754
+    },
+    {
+      "epoch": 0.8272636389602971,
+      "grad_norm": 0.23095338657923528,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 34755
+    },
+    {
+      "epoch": 0.8272874416833286,
+      "grad_norm": 8.415299924041658,
+      "learning_rate": 2e-05,
+      "loss": 0.9244,
+      "step": 34756
+    },
+    {
+      "epoch": 0.8273112444063601,
+      "grad_norm": 0.5591937985279616,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 34757
+    },
+    {
+      "epoch": 0.8273350471293917,
+      "grad_norm": 5.962650380017528,
+      "learning_rate": 2e-05,
+      "loss": 1.5619,
+      "step": 34758
+    },
+    {
+      "epoch": 0.8273588498524231,
+      "grad_norm": 4.149344668581759,
+      "learning_rate": 2e-05,
+      "loss": 0.6674,
+      "step": 34759
+    },
+    {
+      "epoch": 0.8273826525754546,
+      "grad_norm": 2.6541476953253706,
+      "learning_rate": 2e-05,
+      "loss": 0.33,
+      "step": 34760
+    },
+    {
+      "epoch": 0.8274064552984861,
+      "grad_norm": 4.412261959028678,
+      "learning_rate": 2e-05,
+      "loss": 0.6778,
+      "step": 34761
+    },
+    {
+      "epoch": 0.8274302580215177,
+      "grad_norm": 3.2913071548198363,
+      "learning_rate": 2e-05,
+      "loss": 0.6078,
+      "step": 34762
+    },
+    {
+      "epoch": 0.8274540607445492,
+      "grad_norm": 8.095003708754323,
+      "learning_rate": 2e-05,
+      "loss": 0.1203,
+      "step": 34763
+    },
+    {
+      "epoch": 0.8274778634675807,
+      "grad_norm": 4.949887836914703,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 34764
+    },
+    {
+      "epoch": 0.8275016661906122,
+      "grad_norm": 0.02984348902725708,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34765
+    },
+    {
+      "epoch": 0.8275254689136438,
+      "grad_norm": 4.035468244285635,
+      "learning_rate": 2e-05,
+      "loss": 0.1824,
+      "step": 34766
+    },
+    {
+      "epoch": 0.8275492716366752,
+      "grad_norm": 6.371492043139653,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 34767
+    },
+    {
+      "epoch": 0.8275730743597067,
+      "grad_norm": 6.691130763210181,
+      "learning_rate": 2e-05,
+      "loss": 1.3832,
+      "step": 34768
+    },
+    {
+      "epoch": 0.8275968770827382,
+      "grad_norm": 5.459666333053407,
+      "learning_rate": 2e-05,
+      "loss": 0.122,
+      "step": 34769
+    },
+    {
+      "epoch": 0.8276206798057698,
+      "grad_norm": 1.15921857081496,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 34770
+    },
+    {
+      "epoch": 0.8276444825288013,
+      "grad_norm": 0.04993217267808152,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34771
+    },
+    {
+      "epoch": 0.8276682852518328,
+      "grad_norm": 25.78352345397572,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 34772
+    },
+    {
+      "epoch": 0.8276920879748644,
+      "grad_norm": 4.603891102126508,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 34773
+    },
+    {
+      "epoch": 0.8277158906978959,
+      "grad_norm": 3.4258832129193832,
+      "learning_rate": 2e-05,
+      "loss": 0.5629,
+      "step": 34774
+    },
+    {
+      "epoch": 0.8277396934209273,
+      "grad_norm": 0.1247955303979868,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 34775
+    },
+    {
+      "epoch": 0.8277634961439588,
+      "grad_norm": 0.6682184235947863,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 34776
+    },
+    {
+      "epoch": 0.8277872988669904,
+      "grad_norm": 0.2370960967435845,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 34777
+    },
+    {
+      "epoch": 0.8278111015900219,
+      "grad_norm": 0.0394578599267431,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34778
+    },
+    {
+      "epoch": 0.8278349043130534,
+      "grad_norm": 5.472580308046808,
+      "learning_rate": 2e-05,
+      "loss": 1.1964,
+      "step": 34779
+    },
+    {
+      "epoch": 0.8278587070360849,
+      "grad_norm": 0.07519378904687096,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34780
+    },
+    {
+      "epoch": 0.8278825097591165,
+      "grad_norm": 0.6227970020547178,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 34781
+    },
+    {
+      "epoch": 0.827906312482148,
+      "grad_norm": 4.602905827852107,
+      "learning_rate": 2e-05,
+      "loss": 0.1527,
+      "step": 34782
+    },
+    {
+      "epoch": 0.8279301152051795,
+      "grad_norm": 4.5377996125026,
+      "learning_rate": 2e-05,
+      "loss": 0.6543,
+      "step": 34783
+    },
+    {
+      "epoch": 0.8279539179282109,
+      "grad_norm": 4.6313120638166545,
+      "learning_rate": 2e-05,
+      "loss": 0.6348,
+      "step": 34784
+    },
+    {
+      "epoch": 0.8279777206512425,
+      "grad_norm": 5.181317958760624,
+      "learning_rate": 2e-05,
+      "loss": 0.8184,
+      "step": 34785
+    },
+    {
+      "epoch": 0.828001523374274,
+      "grad_norm": 0.5304866538119013,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 34786
+    },
+    {
+      "epoch": 0.8280253260973055,
+      "grad_norm": 4.157791697227482,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 34787
+    },
+    {
+      "epoch": 0.8280491288203371,
+      "grad_norm": 3.586971403420952,
+      "learning_rate": 2e-05,
+      "loss": 0.72,
+      "step": 34788
+    },
+    {
+      "epoch": 0.8280729315433686,
+      "grad_norm": 16.013112775387533,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 34789
+    },
+    {
+      "epoch": 0.8280967342664001,
+      "grad_norm": 4.204765549053069,
+      "learning_rate": 2e-05,
+      "loss": 0.9873,
+      "step": 34790
+    },
+    {
+      "epoch": 0.8281205369894316,
+      "grad_norm": 0.5058992409490154,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 34791
+    },
+    {
+      "epoch": 0.8281443397124632,
+      "grad_norm": 0.03896267565949326,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34792
+    },
+    {
+      "epoch": 0.8281681424354946,
+      "grad_norm": 6.940302923530041,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 34793
+    },
+    {
+      "epoch": 0.8281919451585261,
+      "grad_norm": 4.8868682669959425,
+      "learning_rate": 2e-05,
+      "loss": 0.0821,
+      "step": 34794
+    },
+    {
+      "epoch": 0.8282157478815576,
+      "grad_norm": 2.4697568904102454,
+      "learning_rate": 2e-05,
+      "loss": 0.3881,
+      "step": 34795
+    },
+    {
+      "epoch": 0.8282395506045892,
+      "grad_norm": 4.233785711768015,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 34796
+    },
+    {
+      "epoch": 0.8282633533276207,
+      "grad_norm": 1.897274877812488,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 34797
+    },
+    {
+      "epoch": 0.8282871560506522,
+      "grad_norm": 5.577150254605855,
+      "learning_rate": 2e-05,
+      "loss": 0.8783,
+      "step": 34798
+    },
+    {
+      "epoch": 0.8283109587736837,
+      "grad_norm": 2.8936612577262273,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 34799
+    },
+    {
+      "epoch": 0.8283347614967153,
+      "grad_norm": 2.8191984506636967,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 34800
+    },
+    {
+      "epoch": 0.8283585642197467,
+      "grad_norm": 4.59605266366133,
+      "learning_rate": 2e-05,
+      "loss": 0.1085,
+      "step": 34801
+    },
+    {
+      "epoch": 0.8283823669427782,
+      "grad_norm": 3.0007972958768923,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 34802
+    },
+    {
+      "epoch": 0.8284061696658098,
+      "grad_norm": 10.3780539587089,
+      "learning_rate": 2e-05,
+      "loss": 1.3466,
+      "step": 34803
+    },
+    {
+      "epoch": 0.8284299723888413,
+      "grad_norm": 3.0312559989940255,
+      "learning_rate": 2e-05,
+      "loss": 0.5861,
+      "step": 34804
+    },
+    {
+      "epoch": 0.8284537751118728,
+      "grad_norm": 0.15544915849276672,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 34805
+    },
+    {
+      "epoch": 0.8284775778349043,
+      "grad_norm": 11.660181933997691,
+      "learning_rate": 2e-05,
+      "loss": 0.3129,
+      "step": 34806
+    },
+    {
+      "epoch": 0.8285013805579359,
+      "grad_norm": 5.441844125176175,
+      "learning_rate": 2e-05,
+      "loss": 1.0501,
+      "step": 34807
+    },
+    {
+      "epoch": 0.8285251832809674,
+      "grad_norm": 3.539984601110388,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 34808
+    },
+    {
+      "epoch": 0.8285489860039988,
+      "grad_norm": 1.4565320230741363,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 34809
+    },
+    {
+      "epoch": 0.8285727887270303,
+      "grad_norm": 1.1775343208516242,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 34810
+    },
+    {
+      "epoch": 0.8285965914500619,
+      "grad_norm": 8.456338018548863,
+      "learning_rate": 2e-05,
+      "loss": 0.8379,
+      "step": 34811
+    },
+    {
+      "epoch": 0.8286203941730934,
+      "grad_norm": 11.749834285581649,
+      "learning_rate": 2e-05,
+      "loss": 0.1451,
+      "step": 34812
+    },
+    {
+      "epoch": 0.8286441968961249,
+      "grad_norm": 4.539427902824267,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 34813
+    },
+    {
+      "epoch": 0.8286679996191564,
+      "grad_norm": 0.748551378108219,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 34814
+    },
+    {
+      "epoch": 0.828691802342188,
+      "grad_norm": 3.7342649069321343,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 34815
+    },
+    {
+      "epoch": 0.8287156050652195,
+      "grad_norm": 1.3780268281466983,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 34816
+    },
+    {
+      "epoch": 0.828739407788251,
+      "grad_norm": 0.05402830723214483,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34817
+    },
+    {
+      "epoch": 0.8287632105112825,
+      "grad_norm": 12.83261001273157,
+      "learning_rate": 2e-05,
+      "loss": 1.7562,
+      "step": 34818
+    },
+    {
+      "epoch": 0.828787013234314,
+      "grad_norm": 5.137147909012041,
+      "learning_rate": 2e-05,
+      "loss": 0.6214,
+      "step": 34819
+    },
+    {
+      "epoch": 0.8288108159573455,
+      "grad_norm": 0.9469081496441822,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 34820
+    },
+    {
+      "epoch": 0.828834618680377,
+      "grad_norm": 3.383010336218804,
+      "learning_rate": 2e-05,
+      "loss": 0.1564,
+      "step": 34821
+    },
+    {
+      "epoch": 0.8288584214034086,
+      "grad_norm": 3.105011351548716,
+      "learning_rate": 2e-05,
+      "loss": 0.771,
+      "step": 34822
+    },
+    {
+      "epoch": 0.8288822241264401,
+      "grad_norm": 5.091899025581681,
+      "learning_rate": 2e-05,
+      "loss": 1.6495,
+      "step": 34823
+    },
+    {
+      "epoch": 0.8289060268494716,
+      "grad_norm": 24.106052702438326,
+      "learning_rate": 2e-05,
+      "loss": 0.7043,
+      "step": 34824
+    },
+    {
+      "epoch": 0.828929829572503,
+      "grad_norm": 4.544473431372365,
+      "learning_rate": 2e-05,
+      "loss": 0.9332,
+      "step": 34825
+    },
+    {
+      "epoch": 0.8289536322955346,
+      "grad_norm": 5.721021656933929,
+      "learning_rate": 2e-05,
+      "loss": 0.4188,
+      "step": 34826
+    },
+    {
+      "epoch": 0.8289774350185661,
+      "grad_norm": 5.164488940393252,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 34827
+    },
+    {
+      "epoch": 0.8290012377415976,
+      "grad_norm": 0.27996956734246864,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 34828
+    },
+    {
+      "epoch": 0.8290250404646291,
+      "grad_norm": 5.413893181162016,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 34829
+    },
+    {
+      "epoch": 0.8290488431876607,
+      "grad_norm": 9.007230311396363,
+      "learning_rate": 2e-05,
+      "loss": 0.313,
+      "step": 34830
+    },
+    {
+      "epoch": 0.8290726459106922,
+      "grad_norm": 0.44544789380229244,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 34831
+    },
+    {
+      "epoch": 0.8290964486337237,
+      "grad_norm": 1.4531278175239324,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 34832
+    },
+    {
+      "epoch": 0.8291202513567553,
+      "grad_norm": 2.9606787131854277,
+      "learning_rate": 2e-05,
+      "loss": 0.2178,
+      "step": 34833
+    },
+    {
+      "epoch": 0.8291440540797868,
+      "grad_norm": 0.7247187303546315,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 34834
+    },
+    {
+      "epoch": 0.8291678568028182,
+      "grad_norm": 0.13003356232793184,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 34835
+    },
+    {
+      "epoch": 0.8291916595258497,
+      "grad_norm": 2.8447643506733105,
+      "learning_rate": 2e-05,
+      "loss": 0.1498,
+      "step": 34836
+    },
+    {
+      "epoch": 0.8292154622488813,
+      "grad_norm": 0.0802855398454301,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 34837
+    },
+    {
+      "epoch": 0.8292392649719128,
+      "grad_norm": 0.10763065080663527,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 34838
+    },
+    {
+      "epoch": 0.8292630676949443,
+      "grad_norm": 5.250591131710049,
+      "learning_rate": 2e-05,
+      "loss": 0.8163,
+      "step": 34839
+    },
+    {
+      "epoch": 0.8292868704179758,
+      "grad_norm": 5.218841169614541,
+      "learning_rate": 2e-05,
+      "loss": 0.0526,
+      "step": 34840
+    },
+    {
+      "epoch": 0.8293106731410074,
+      "grad_norm": 2.5666885125407637,
+      "learning_rate": 2e-05,
+      "loss": 0.1361,
+      "step": 34841
+    },
+    {
+      "epoch": 0.8293344758640389,
+      "grad_norm": 5.025285909555364,
+      "learning_rate": 2e-05,
+      "loss": 1.0491,
+      "step": 34842
+    },
+    {
+      "epoch": 0.8293582785870703,
+      "grad_norm": 2.470745189530009,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 34843
+    },
+    {
+      "epoch": 0.8293820813101018,
+      "grad_norm": 2.8282803577464426,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 34844
+    },
+    {
+      "epoch": 0.8294058840331334,
+      "grad_norm": 4.4449170695320115,
+      "learning_rate": 2e-05,
+      "loss": 1.3762,
+      "step": 34845
+    },
+    {
+      "epoch": 0.8294296867561649,
+      "grad_norm": 0.15850877620206313,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 34846
+    },
+    {
+      "epoch": 0.8294534894791964,
+      "grad_norm": 8.998353317231143,
+      "learning_rate": 2e-05,
+      "loss": 0.7066,
+      "step": 34847
+    },
+    {
+      "epoch": 0.829477292202228,
+      "grad_norm": 13.034829285903845,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 34848
+    },
+    {
+      "epoch": 0.8295010949252595,
+      "grad_norm": 3.917691122515098,
+      "learning_rate": 2e-05,
+      "loss": 0.0436,
+      "step": 34849
+    },
+    {
+      "epoch": 0.829524897648291,
+      "grad_norm": 0.42552615059756177,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 34850
+    },
+    {
+      "epoch": 0.8295487003713224,
+      "grad_norm": 4.828362166653521,
+      "learning_rate": 2e-05,
+      "loss": 1.374,
+      "step": 34851
+    },
+    {
+      "epoch": 0.829572503094354,
+      "grad_norm": 0.4496029873879273,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 34852
+    },
+    {
+      "epoch": 0.8295963058173855,
+      "grad_norm": 1.370841289702024,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 34853
+    },
+    {
+      "epoch": 0.829620108540417,
+      "grad_norm": 8.155618846562536,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 34854
+    },
+    {
+      "epoch": 0.8296439112634485,
+      "grad_norm": 0.7437612981169169,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 34855
+    },
+    {
+      "epoch": 0.8296677139864801,
+      "grad_norm": 4.361237478737487,
+      "learning_rate": 2e-05,
+      "loss": 0.9642,
+      "step": 34856
+    },
+    {
+      "epoch": 0.8296915167095116,
+      "grad_norm": 1.8775291451729847,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 34857
+    },
+    {
+      "epoch": 0.8297153194325431,
+      "grad_norm": 66.10831235162186,
+      "learning_rate": 2e-05,
+      "loss": 0.6621,
+      "step": 34858
+    },
+    {
+      "epoch": 0.8297391221555745,
+      "grad_norm": 6.095684875688015,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 34859
+    },
+    {
+      "epoch": 0.8297629248786061,
+      "grad_norm": 1.298524435394228,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 34860
+    },
+    {
+      "epoch": 0.8297867276016376,
+      "grad_norm": 3.399077640664857,
+      "learning_rate": 2e-05,
+      "loss": 0.5958,
+      "step": 34861
+    },
+    {
+      "epoch": 0.8298105303246691,
+      "grad_norm": 6.490017012439375,
+      "learning_rate": 2e-05,
+      "loss": 1.1593,
+      "step": 34862
+    },
+    {
+      "epoch": 0.8298343330477007,
+      "grad_norm": 2.7773866965447778,
+      "learning_rate": 2e-05,
+      "loss": 0.6592,
+      "step": 34863
+    },
+    {
+      "epoch": 0.8298581357707322,
+      "grad_norm": 3.6299376280175024,
+      "learning_rate": 2e-05,
+      "loss": 0.4697,
+      "step": 34864
+    },
+    {
+      "epoch": 0.8298819384937637,
+      "grad_norm": 0.01272239039582859,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 34865
+    },
+    {
+      "epoch": 0.8299057412167952,
+      "grad_norm": 0.059893466141183696,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 34866
+    },
+    {
+      "epoch": 0.8299295439398268,
+      "grad_norm": 13.735384205249195,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 34867
+    },
+    {
+      "epoch": 0.8299533466628582,
+      "grad_norm": 0.10068463460502973,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 34868
+    },
+    {
+      "epoch": 0.8299771493858897,
+      "grad_norm": 3.0842848942611125,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 34869
+    },
+    {
+      "epoch": 0.8300009521089212,
+      "grad_norm": 27.852837395361327,
+      "learning_rate": 2e-05,
+      "loss": 0.7728,
+      "step": 34870
+    },
+    {
+      "epoch": 0.8300247548319528,
+      "grad_norm": 6.157742102197007,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 34871
+    },
+    {
+      "epoch": 0.8300485575549843,
+      "grad_norm": 0.04608465678683469,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 34872
+    },
+    {
+      "epoch": 0.8300723602780158,
+      "grad_norm": 5.513325892854091,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 34873
+    },
+    {
+      "epoch": 0.8300961630010473,
+      "grad_norm": 5.4874015519434085,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 34874
+    },
+    {
+      "epoch": 0.8301199657240789,
+      "grad_norm": 1.093875244197372,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 34875
+    },
+    {
+      "epoch": 0.8301437684471104,
+      "grad_norm": 4.979673133872245,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 34876
+    },
+    {
+      "epoch": 0.8301675711701418,
+      "grad_norm": 9.775664267447514,
+      "learning_rate": 2e-05,
+      "loss": 0.9327,
+      "step": 34877
+    },
+    {
+      "epoch": 0.8301913738931734,
+      "grad_norm": 6.394198140353866,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 34878
+    },
+    {
+      "epoch": 0.8302151766162049,
+      "grad_norm": 3.5577265011383994,
+      "learning_rate": 2e-05,
+      "loss": 0.8535,
+      "step": 34879
+    },
+    {
+      "epoch": 0.8302389793392364,
+      "grad_norm": 14.791356838041532,
+      "learning_rate": 2e-05,
+      "loss": 0.3816,
+      "step": 34880
+    },
+    {
+      "epoch": 0.8302627820622679,
+      "grad_norm": 4.050636279015453,
+      "learning_rate": 2e-05,
+      "loss": 0.6011,
+      "step": 34881
+    },
+    {
+      "epoch": 0.8302865847852995,
+      "grad_norm": 3.998124494334268,
+      "learning_rate": 2e-05,
+      "loss": 1.0419,
+      "step": 34882
+    },
+    {
+      "epoch": 0.830310387508331,
+      "grad_norm": 0.6914237246674233,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 34883
+    },
+    {
+      "epoch": 0.8303341902313625,
+      "grad_norm": 0.7820741867042205,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 34884
+    },
+    {
+      "epoch": 0.8303579929543939,
+      "grad_norm": 4.14201982414297,
+      "learning_rate": 2e-05,
+      "loss": 0.0701,
+      "step": 34885
+    },
+    {
+      "epoch": 0.8303817956774255,
+      "grad_norm": 0.24073125324348355,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 34886
+    },
+    {
+      "epoch": 0.830405598400457,
+      "grad_norm": 4.437956593122114,
+      "learning_rate": 2e-05,
+      "loss": 0.9002,
+      "step": 34887
+    },
+    {
+      "epoch": 0.8304294011234885,
+      "grad_norm": 0.9119134351923341,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 34888
+    },
+    {
+      "epoch": 0.83045320384652,
+      "grad_norm": 3.8838987165532872,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 34889
+    },
+    {
+      "epoch": 0.8304770065695516,
+      "grad_norm": 3.428786163452345,
+      "learning_rate": 2e-05,
+      "loss": 0.5714,
+      "step": 34890
+    },
+    {
+      "epoch": 0.8305008092925831,
+      "grad_norm": 15.451809101211616,
+      "learning_rate": 2e-05,
+      "loss": 0.6274,
+      "step": 34891
+    },
+    {
+      "epoch": 0.8305246120156146,
+      "grad_norm": 4.234980148212594,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 34892
+    },
+    {
+      "epoch": 0.830548414738646,
+      "grad_norm": 0.7900689850406333,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 34893
+    },
+    {
+      "epoch": 0.8305722174616776,
+      "grad_norm": 2.464630111434751,
+      "learning_rate": 2e-05,
+      "loss": 0.1646,
+      "step": 34894
+    },
+    {
+      "epoch": 0.8305960201847091,
+      "grad_norm": 5.301509477579287,
+      "learning_rate": 2e-05,
+      "loss": 1.3588,
+      "step": 34895
+    },
+    {
+      "epoch": 0.8306198229077406,
+      "grad_norm": 0.2637366138004257,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 34896
+    },
+    {
+      "epoch": 0.8306436256307722,
+      "grad_norm": 3.6715762257338898,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 34897
+    },
+    {
+      "epoch": 0.8306674283538037,
+      "grad_norm": 3.0593169653184544,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 34898
+    },
+    {
+      "epoch": 0.8306912310768352,
+      "grad_norm": 14.171275525908785,
+      "learning_rate": 2e-05,
+      "loss": 0.3893,
+      "step": 34899
+    },
+    {
+      "epoch": 0.8307150337998667,
+      "grad_norm": 2.7312229171769005,
+      "learning_rate": 2e-05,
+      "loss": 0.1971,
+      "step": 34900
+    },
+    {
+      "epoch": 0.8307388365228983,
+      "grad_norm": 1.2302199174009576,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 34901
+    },
+    {
+      "epoch": 0.8307626392459297,
+      "grad_norm": 0.9135688842249989,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 34902
+    },
+    {
+      "epoch": 0.8307864419689612,
+      "grad_norm": 5.886666605673047,
+      "learning_rate": 2e-05,
+      "loss": 0.5562,
+      "step": 34903
+    },
+    {
+      "epoch": 0.8308102446919927,
+      "grad_norm": 2.8891808810599198,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 34904
+    },
+    {
+      "epoch": 0.8308340474150243,
+      "grad_norm": 0.5842984214457805,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 34905
+    },
+    {
+      "epoch": 0.8308578501380558,
+      "grad_norm": 0.1814608304669239,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 34906
+    },
+    {
+      "epoch": 0.8308816528610873,
+      "grad_norm": 0.3498825130562966,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 34907
+    },
+    {
+      "epoch": 0.8309054555841188,
+      "grad_norm": 4.918198966780578,
+      "learning_rate": 2e-05,
+      "loss": 0.7577,
+      "step": 34908
+    },
+    {
+      "epoch": 0.8309292583071504,
+      "grad_norm": 5.454995594541922,
+      "learning_rate": 2e-05,
+      "loss": 0.2024,
+      "step": 34909
+    },
+    {
+      "epoch": 0.8309530610301818,
+      "grad_norm": 5.899077216622211,
+      "learning_rate": 2e-05,
+      "loss": 1.1611,
+      "step": 34910
+    },
+    {
+      "epoch": 0.8309768637532133,
+      "grad_norm": 1.7731033239072274,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 34911
+    },
+    {
+      "epoch": 0.8310006664762449,
+      "grad_norm": 6.370696633956906,
+      "learning_rate": 2e-05,
+      "loss": 0.6995,
+      "step": 34912
+    },
+    {
+      "epoch": 0.8310244691992764,
+      "grad_norm": 0.3044982227625958,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 34913
+    },
+    {
+      "epoch": 0.8310482719223079,
+      "grad_norm": 0.5927220015421345,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 34914
+    },
+    {
+      "epoch": 0.8310720746453394,
+      "grad_norm": 8.46642871983722,
+      "learning_rate": 2e-05,
+      "loss": 0.1328,
+      "step": 34915
+    },
+    {
+      "epoch": 0.831095877368371,
+      "grad_norm": 1.6045894566132075,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 34916
+    },
+    {
+      "epoch": 0.8311196800914025,
+      "grad_norm": 1.2593692045918592,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 34917
+    },
+    {
+      "epoch": 0.831143482814434,
+      "grad_norm": 5.024447485924617,
+      "learning_rate": 2e-05,
+      "loss": 0.5453,
+      "step": 34918
+    },
+    {
+      "epoch": 0.8311672855374654,
+      "grad_norm": 1.5366443937157823,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 34919
+    },
+    {
+      "epoch": 0.831191088260497,
+      "grad_norm": 1.3535291091400912,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 34920
+    },
+    {
+      "epoch": 0.8312148909835285,
+      "grad_norm": 3.904370733880657,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 34921
+    },
+    {
+      "epoch": 0.83123869370656,
+      "grad_norm": 19.366902861642238,
+      "learning_rate": 2e-05,
+      "loss": 0.4236,
+      "step": 34922
+    },
+    {
+      "epoch": 0.8312624964295915,
+      "grad_norm": 6.386095308677367,
+      "learning_rate": 2e-05,
+      "loss": 1.361,
+      "step": 34923
+    },
+    {
+      "epoch": 0.8312862991526231,
+      "grad_norm": 1.2414325973920874,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 34924
+    },
+    {
+      "epoch": 0.8313101018756546,
+      "grad_norm": 0.02974783959955368,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 34925
+    },
+    {
+      "epoch": 0.8313339045986861,
+      "grad_norm": 3.9511209709572097,
+      "learning_rate": 2e-05,
+      "loss": 0.6902,
+      "step": 34926
+    },
+    {
+      "epoch": 0.8313577073217177,
+      "grad_norm": 3.5088253120302197,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 34927
+    },
+    {
+      "epoch": 0.8313815100447491,
+      "grad_norm": 2.9792082788232213,
+      "learning_rate": 2e-05,
+      "loss": 0.1296,
+      "step": 34928
+    },
+    {
+      "epoch": 0.8314053127677806,
+      "grad_norm": 5.307767249795857,
+      "learning_rate": 2e-05,
+      "loss": 0.9471,
+      "step": 34929
+    },
+    {
+      "epoch": 0.8314291154908121,
+      "grad_norm": 2.302419912203291,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 34930
+    },
+    {
+      "epoch": 0.8314529182138437,
+      "grad_norm": 4.8511124262864005,
+      "learning_rate": 2e-05,
+      "loss": 0.7551,
+      "step": 34931
+    },
+    {
+      "epoch": 0.8314767209368752,
+      "grad_norm": 0.011470852885727299,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 34932
+    },
+    {
+      "epoch": 0.8315005236599067,
+      "grad_norm": 4.592440185938865,
+      "learning_rate": 2e-05,
+      "loss": 0.8001,
+      "step": 34933
+    },
+    {
+      "epoch": 0.8315243263829382,
+      "grad_norm": 3.230083000286738,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 34934
+    },
+    {
+      "epoch": 0.8315481291059698,
+      "grad_norm": 2.6287517596794148,
+      "learning_rate": 2e-05,
+      "loss": 0.5512,
+      "step": 34935
+    },
+    {
+      "epoch": 0.8315719318290012,
+      "grad_norm": 9.135687535353213,
+      "learning_rate": 2e-05,
+      "loss": 0.2694,
+      "step": 34936
+    },
+    {
+      "epoch": 0.8315957345520327,
+      "grad_norm": 0.027456219869332357,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 34937
+    },
+    {
+      "epoch": 0.8316195372750642,
+      "grad_norm": 2.2082225134443645,
+      "learning_rate": 2e-05,
+      "loss": 0.1634,
+      "step": 34938
+    },
+    {
+      "epoch": 0.8316433399980958,
+      "grad_norm": 5.897814303515573,
+      "learning_rate": 2e-05,
+      "loss": 1.0743,
+      "step": 34939
+    },
+    {
+      "epoch": 0.8316671427211273,
+      "grad_norm": 5.957169288069174,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 34940
+    },
+    {
+      "epoch": 0.8316909454441588,
+      "grad_norm": 4.2266173520051495,
+      "learning_rate": 2e-05,
+      "loss": 0.4159,
+      "step": 34941
+    },
+    {
+      "epoch": 0.8317147481671904,
+      "grad_norm": 2.999180525691215,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 34942
+    },
+    {
+      "epoch": 0.8317385508902219,
+      "grad_norm": 0.30758980981461564,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 34943
+    },
+    {
+      "epoch": 0.8317623536132533,
+      "grad_norm": 0.5865573871060379,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 34944
+    },
+    {
+      "epoch": 0.8317861563362848,
+      "grad_norm": 4.312809972924596,
+      "learning_rate": 2e-05,
+      "loss": 0.0683,
+      "step": 34945
+    },
+    {
+      "epoch": 0.8318099590593164,
+      "grad_norm": 8.392750850831039,
+      "learning_rate": 2e-05,
+      "loss": 1.1714,
+      "step": 34946
+    },
+    {
+      "epoch": 0.8318337617823479,
+      "grad_norm": 9.041589782915139,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 34947
+    },
+    {
+      "epoch": 0.8318575645053794,
+      "grad_norm": 2.42362398692289,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 34948
+    },
+    {
+      "epoch": 0.8318813672284109,
+      "grad_norm": 5.253773509324127,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 34949
+    },
+    {
+      "epoch": 0.8319051699514425,
+      "grad_norm": 10.883370751482548,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 34950
+    },
+    {
+      "epoch": 0.831928972674474,
+      "grad_norm": 2.485256682210411,
+      "learning_rate": 2e-05,
+      "loss": 0.2118,
+      "step": 34951
+    },
+    {
+      "epoch": 0.8319527753975055,
+      "grad_norm": 0.6720934710935421,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 34952
+    },
+    {
+      "epoch": 0.8319765781205369,
+      "grad_norm": 6.384983910460364,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 34953
+    },
+    {
+      "epoch": 0.8320003808435685,
+      "grad_norm": 3.8016013321948674,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 34954
+    },
+    {
+      "epoch": 0.8320241835666,
+      "grad_norm": 8.43699105217878,
+      "learning_rate": 2e-05,
+      "loss": 0.6551,
+      "step": 34955
+    },
+    {
+      "epoch": 0.8320479862896315,
+      "grad_norm": 1.6546390907016937,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 34956
+    },
+    {
+      "epoch": 0.8320717890126631,
+      "grad_norm": 0.021589480590237253,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 34957
+    },
+    {
+      "epoch": 0.8320955917356946,
+      "grad_norm": 3.4843506602185457,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 34958
+    },
+    {
+      "epoch": 0.8321193944587261,
+      "grad_norm": 5.907494406870397,
+      "learning_rate": 2e-05,
+      "loss": 0.5704,
+      "step": 34959
+    },
+    {
+      "epoch": 0.8321431971817576,
+      "grad_norm": 10.296169320479764,
+      "learning_rate": 2e-05,
+      "loss": 0.5421,
+      "step": 34960
+    },
+    {
+      "epoch": 0.8321669999047892,
+      "grad_norm": 2.836110567946078,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 34961
+    },
+    {
+      "epoch": 0.8321908026278206,
+      "grad_norm": 3.0855574745448475,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 34962
+    },
+    {
+      "epoch": 0.8322146053508521,
+      "grad_norm": 2.701538176867573,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 34963
+    },
+    {
+      "epoch": 0.8322384080738836,
+      "grad_norm": 10.576097965033586,
+      "learning_rate": 2e-05,
+      "loss": 0.5881,
+      "step": 34964
+    },
+    {
+      "epoch": 0.8322622107969152,
+      "grad_norm": 1.0453473892586922,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 34965
+    },
+    {
+      "epoch": 0.8322860135199467,
+      "grad_norm": 4.276519679760419,
+      "learning_rate": 2e-05,
+      "loss": 0.7554,
+      "step": 34966
+    },
+    {
+      "epoch": 0.8323098162429782,
+      "grad_norm": 0.005538506030188054,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 34967
+    },
+    {
+      "epoch": 0.8323336189660097,
+      "grad_norm": 7.047262774092327,
+      "learning_rate": 2e-05,
+      "loss": 1.1008,
+      "step": 34968
+    },
+    {
+      "epoch": 0.8323574216890413,
+      "grad_norm": 8.084603484295034,
+      "learning_rate": 2e-05,
+      "loss": 1.645,
+      "step": 34969
+    },
+    {
+      "epoch": 0.8323812244120727,
+      "grad_norm": 7.233951548094422,
+      "learning_rate": 2e-05,
+      "loss": 1.9325,
+      "step": 34970
+    },
+    {
+      "epoch": 0.8324050271351042,
+      "grad_norm": 6.408244077352254,
+      "learning_rate": 2e-05,
+      "loss": 0.8017,
+      "step": 34971
+    },
+    {
+      "epoch": 0.8324288298581358,
+      "grad_norm": 4.160198904207658,
+      "learning_rate": 2e-05,
+      "loss": 1.0482,
+      "step": 34972
+    },
+    {
+      "epoch": 0.8324526325811673,
+      "grad_norm": 3.8600401614350175,
+      "learning_rate": 2e-05,
+      "loss": 0.8171,
+      "step": 34973
+    },
+    {
+      "epoch": 0.8324764353041988,
+      "grad_norm": 7.2153358214733,
+      "learning_rate": 2e-05,
+      "loss": 0.4444,
+      "step": 34974
+    },
+    {
+      "epoch": 0.8325002380272303,
+      "grad_norm": 3.8521852721060545,
+      "learning_rate": 2e-05,
+      "loss": 0.8628,
+      "step": 34975
+    },
+    {
+      "epoch": 0.8325240407502619,
+      "grad_norm": 3.888448139354399,
+      "learning_rate": 2e-05,
+      "loss": 0.5958,
+      "step": 34976
+    },
+    {
+      "epoch": 0.8325478434732934,
+      "grad_norm": 2.2548320512045543,
+      "learning_rate": 2e-05,
+      "loss": 0.0861,
+      "step": 34977
+    },
+    {
+      "epoch": 0.8325716461963248,
+      "grad_norm": 6.756741223544515,
+      "learning_rate": 2e-05,
+      "loss": 1.295,
+      "step": 34978
+    },
+    {
+      "epoch": 0.8325954489193563,
+      "grad_norm": 1.6757532268686368,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 34979
+    },
+    {
+      "epoch": 0.8326192516423879,
+      "grad_norm": 3.8377049214685868,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 34980
+    },
+    {
+      "epoch": 0.8326430543654194,
+      "grad_norm": 0.22131345942092534,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 34981
+    },
+    {
+      "epoch": 0.8326668570884509,
+      "grad_norm": 0.04832968471666483,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 34982
+    },
+    {
+      "epoch": 0.8326906598114824,
+      "grad_norm": 11.653753458306587,
+      "learning_rate": 2e-05,
+      "loss": 1.3953,
+      "step": 34983
+    },
+    {
+      "epoch": 0.832714462534514,
+      "grad_norm": 5.456779468115641,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 34984
+    },
+    {
+      "epoch": 0.8327382652575455,
+      "grad_norm": 0.13312310845788222,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 34985
+    },
+    {
+      "epoch": 0.832762067980577,
+      "grad_norm": 16.994486590755212,
+      "learning_rate": 2e-05,
+      "loss": 0.5844,
+      "step": 34986
+    },
+    {
+      "epoch": 0.8327858707036085,
+      "grad_norm": 9.501810918315932,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 34987
+    },
+    {
+      "epoch": 0.83280967342664,
+      "grad_norm": 2.551180458048312,
+      "learning_rate": 2e-05,
+      "loss": 0.3404,
+      "step": 34988
+    },
+    {
+      "epoch": 0.8328334761496715,
+      "grad_norm": 3.586209113047905,
+      "learning_rate": 2e-05,
+      "loss": 0.3415,
+      "step": 34989
+    },
+    {
+      "epoch": 0.832857278872703,
+      "grad_norm": 5.0810386859908565,
+      "learning_rate": 2e-05,
+      "loss": 0.592,
+      "step": 34990
+    },
+    {
+      "epoch": 0.8328810815957346,
+      "grad_norm": 2.74353562312985,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 34991
+    },
+    {
+      "epoch": 0.8329048843187661,
+      "grad_norm": 0.047318756695463246,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 34992
+    },
+    {
+      "epoch": 0.8329286870417976,
+      "grad_norm": 6.266571336596156,
+      "learning_rate": 2e-05,
+      "loss": 1.6876,
+      "step": 34993
+    },
+    {
+      "epoch": 0.832952489764829,
+      "grad_norm": 1.9860818266611044,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 34994
+    },
+    {
+      "epoch": 0.8329762924878606,
+      "grad_norm": 4.207500162197661,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 34995
+    },
+    {
+      "epoch": 0.8330000952108921,
+      "grad_norm": 4.018646433985485,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 34996
+    },
+    {
+      "epoch": 0.8330238979339236,
+      "grad_norm": 0.16563938598324676,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 34997
+    },
+    {
+      "epoch": 0.8330477006569551,
+      "grad_norm": 2.4090827813993236,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 34998
+    },
+    {
+      "epoch": 0.8330715033799867,
+      "grad_norm": 4.267956441527289,
+      "learning_rate": 2e-05,
+      "loss": 0.9545,
+      "step": 34999
+    },
+    {
+      "epoch": 0.8330953061030182,
+      "grad_norm": 4.199788215349882,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 35000
+    },
+    {
+      "epoch": 0.8331191088260497,
+      "grad_norm": 2.0780040126477517,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 35001
+    },
+    {
+      "epoch": 0.8331429115490813,
+      "grad_norm": 4.697101956074193,
+      "learning_rate": 2e-05,
+      "loss": 0.9042,
+      "step": 35002
+    },
+    {
+      "epoch": 0.8331667142721128,
+      "grad_norm": 3.8141193654112877,
+      "learning_rate": 2e-05,
+      "loss": 0.7619,
+      "step": 35003
+    },
+    {
+      "epoch": 0.8331905169951442,
+      "grad_norm": 0.08981886865932259,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 35004
+    },
+    {
+      "epoch": 0.8332143197181757,
+      "grad_norm": 3.4417980008052225,
+      "learning_rate": 2e-05,
+      "loss": 0.5159,
+      "step": 35005
+    },
+    {
+      "epoch": 0.8332381224412073,
+      "grad_norm": 2.2306443701318655,
+      "learning_rate": 2e-05,
+      "loss": 0.3427,
+      "step": 35006
+    },
+    {
+      "epoch": 0.8332619251642388,
+      "grad_norm": 5.01136074845053,
+      "learning_rate": 2e-05,
+      "loss": 0.891,
+      "step": 35007
+    },
+    {
+      "epoch": 0.8332857278872703,
+      "grad_norm": 1.3273465596307086,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 35008
+    },
+    {
+      "epoch": 0.8333095306103018,
+      "grad_norm": 0.0677261489158296,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 35009
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.7729986491431219,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 35010
+    },
+    {
+      "epoch": 0.8333571360563649,
+      "grad_norm": 5.848677540995941,
+      "learning_rate": 2e-05,
+      "loss": 1.0018,
+      "step": 35011
+    },
+    {
+      "epoch": 0.8333809387793963,
+      "grad_norm": 7.2550821096091775,
+      "learning_rate": 2e-05,
+      "loss": 1.716,
+      "step": 35012
+    },
+    {
+      "epoch": 0.8334047415024278,
+      "grad_norm": 6.168634499252147,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 35013
+    },
+    {
+      "epoch": 0.8334285442254594,
+      "grad_norm": 2.283612654471176,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 35014
+    },
+    {
+      "epoch": 0.8334523469484909,
+      "grad_norm": 1.882005957803684,
+      "learning_rate": 2e-05,
+      "loss": 0.2739,
+      "step": 35015
+    },
+    {
+      "epoch": 0.8334761496715224,
+      "grad_norm": 4.5436729292653695,
+      "learning_rate": 2e-05,
+      "loss": 0.468,
+      "step": 35016
+    },
+    {
+      "epoch": 0.833499952394554,
+      "grad_norm": 1.598187366955006,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 35017
+    },
+    {
+      "epoch": 0.8335237551175855,
+      "grad_norm": 3.442355563871035,
+      "learning_rate": 2e-05,
+      "loss": 0.5229,
+      "step": 35018
+    },
+    {
+      "epoch": 0.833547557840617,
+      "grad_norm": 3.962096203002435,
+      "learning_rate": 2e-05,
+      "loss": 0.6285,
+      "step": 35019
+    },
+    {
+      "epoch": 0.8335713605636484,
+      "grad_norm": 3.1637293612060926,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 35020
+    },
+    {
+      "epoch": 0.83359516328668,
+      "grad_norm": 0.1639614491431043,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 35021
+    },
+    {
+      "epoch": 0.8336189660097115,
+      "grad_norm": 5.293557213791957,
+      "learning_rate": 2e-05,
+      "loss": 0.1538,
+      "step": 35022
+    },
+    {
+      "epoch": 0.833642768732743,
+      "grad_norm": 4.1286700313932165,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 35023
+    },
+    {
+      "epoch": 0.8336665714557745,
+      "grad_norm": 5.855635872581709,
+      "learning_rate": 2e-05,
+      "loss": 1.6896,
+      "step": 35024
+    },
+    {
+      "epoch": 0.8336903741788061,
+      "grad_norm": 12.724329591753772,
+      "learning_rate": 2e-05,
+      "loss": 1.6956,
+      "step": 35025
+    },
+    {
+      "epoch": 0.8337141769018376,
+      "grad_norm": 2.9417692225512035,
+      "learning_rate": 2e-05,
+      "loss": 0.0475,
+      "step": 35026
+    },
+    {
+      "epoch": 0.8337379796248691,
+      "grad_norm": 0.061651870704611166,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35027
+    },
+    {
+      "epoch": 0.8337617823479005,
+      "grad_norm": 0.039592452972587885,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35028
+    },
+    {
+      "epoch": 0.8337855850709321,
+      "grad_norm": 12.790130898024344,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 35029
+    },
+    {
+      "epoch": 0.8338093877939636,
+      "grad_norm": 2.282028357065372,
+      "learning_rate": 2e-05,
+      "loss": 0.055,
+      "step": 35030
+    },
+    {
+      "epoch": 0.8338331905169951,
+      "grad_norm": 8.752221449855178,
+      "learning_rate": 2e-05,
+      "loss": 1.4489,
+      "step": 35031
+    },
+    {
+      "epoch": 0.8338569932400267,
+      "grad_norm": 0.5240033544074996,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 35032
+    },
+    {
+      "epoch": 0.8338807959630582,
+      "grad_norm": 6.085932833470613,
+      "learning_rate": 2e-05,
+      "loss": 1.607,
+      "step": 35033
+    },
+    {
+      "epoch": 0.8339045986860897,
+      "grad_norm": 0.21454487674371986,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 35034
+    },
+    {
+      "epoch": 0.8339284014091212,
+      "grad_norm": 6.192487335148347,
+      "learning_rate": 2e-05,
+      "loss": 0.7801,
+      "step": 35035
+    },
+    {
+      "epoch": 0.8339522041321528,
+      "grad_norm": 4.693059011306082,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 35036
+    },
+    {
+      "epoch": 0.8339760068551842,
+      "grad_norm": 0.27831253408727225,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 35037
+    },
+    {
+      "epoch": 0.8339998095782157,
+      "grad_norm": 12.061168963642578,
+      "learning_rate": 2e-05,
+      "loss": 1.4044,
+      "step": 35038
+    },
+    {
+      "epoch": 0.8340236123012472,
+      "grad_norm": 16.259818796141435,
+      "learning_rate": 2e-05,
+      "loss": 0.2837,
+      "step": 35039
+    },
+    {
+      "epoch": 0.8340474150242788,
+      "grad_norm": 7.13344534324909,
+      "learning_rate": 2e-05,
+      "loss": 0.7088,
+      "step": 35040
+    },
+    {
+      "epoch": 0.8340712177473103,
+      "grad_norm": 6.602129703339678,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 35041
+    },
+    {
+      "epoch": 0.8340950204703418,
+      "grad_norm": 0.24972045585395536,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 35042
+    },
+    {
+      "epoch": 0.8341188231933733,
+      "grad_norm": 0.26432288357594064,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 35043
+    },
+    {
+      "epoch": 0.8341426259164049,
+      "grad_norm": 6.791734145168925,
+      "learning_rate": 2e-05,
+      "loss": 0.2598,
+      "step": 35044
+    },
+    {
+      "epoch": 0.8341664286394364,
+      "grad_norm": 0.06521425837247037,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 35045
+    },
+    {
+      "epoch": 0.8341902313624678,
+      "grad_norm": 0.055982524833339896,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 35046
+    },
+    {
+      "epoch": 0.8342140340854994,
+      "grad_norm": 1.9318229930138817,
+      "learning_rate": 2e-05,
+      "loss": 0.2718,
+      "step": 35047
+    },
+    {
+      "epoch": 0.8342378368085309,
+      "grad_norm": 1.056176091547161,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 35048
+    },
+    {
+      "epoch": 0.8342616395315624,
+      "grad_norm": 8.09963975817961,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 35049
+    },
+    {
+      "epoch": 0.8342854422545939,
+      "grad_norm": 5.838182921452733,
+      "learning_rate": 2e-05,
+      "loss": 0.4515,
+      "step": 35050
+    },
+    {
+      "epoch": 0.8343092449776255,
+      "grad_norm": 8.979287964865291,
+      "learning_rate": 2e-05,
+      "loss": 0.5296,
+      "step": 35051
+    },
+    {
+      "epoch": 0.834333047700657,
+      "grad_norm": 3.639946981579868,
+      "learning_rate": 2e-05,
+      "loss": 0.0577,
+      "step": 35052
+    },
+    {
+      "epoch": 0.8343568504236885,
+      "grad_norm": 8.363319411141996,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 35053
+    },
+    {
+      "epoch": 0.8343806531467199,
+      "grad_norm": 4.060667803157819,
+      "learning_rate": 2e-05,
+      "loss": 0.4488,
+      "step": 35054
+    },
+    {
+      "epoch": 0.8344044558697515,
+      "grad_norm": 3.3549108803869045,
+      "learning_rate": 2e-05,
+      "loss": 0.625,
+      "step": 35055
+    },
+    {
+      "epoch": 0.834428258592783,
+      "grad_norm": 0.2326714624171134,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 35056
+    },
+    {
+      "epoch": 0.8344520613158145,
+      "grad_norm": 7.60124529902602,
+      "learning_rate": 2e-05,
+      "loss": 0.3454,
+      "step": 35057
+    },
+    {
+      "epoch": 0.834475864038846,
+      "grad_norm": 5.303513247342298,
+      "learning_rate": 2e-05,
+      "loss": 0.9967,
+      "step": 35058
+    },
+    {
+      "epoch": 0.8344996667618776,
+      "grad_norm": 9.13383856664693,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 35059
+    },
+    {
+      "epoch": 0.8345234694849091,
+      "grad_norm": 5.219577997944323,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 35060
+    },
+    {
+      "epoch": 0.8345472722079406,
+      "grad_norm": 6.89786580911735,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 35061
+    },
+    {
+      "epoch": 0.8345710749309722,
+      "grad_norm": 2.938200169308249,
+      "learning_rate": 2e-05,
+      "loss": 0.0719,
+      "step": 35062
+    },
+    {
+      "epoch": 0.8345948776540036,
+      "grad_norm": 0.7372385112976836,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 35063
+    },
+    {
+      "epoch": 0.8346186803770351,
+      "grad_norm": 5.366846392940591,
+      "learning_rate": 2e-05,
+      "loss": 0.4839,
+      "step": 35064
+    },
+    {
+      "epoch": 0.8346424831000666,
+      "grad_norm": 4.651289929727918,
+      "learning_rate": 2e-05,
+      "loss": 0.6883,
+      "step": 35065
+    },
+    {
+      "epoch": 0.8346662858230982,
+      "grad_norm": 3.8456749619893698,
+      "learning_rate": 2e-05,
+      "loss": 0.5631,
+      "step": 35066
+    },
+    {
+      "epoch": 0.8346900885461297,
+      "grad_norm": 7.0954477659631925,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 35067
+    },
+    {
+      "epoch": 0.8347138912691612,
+      "grad_norm": 0.005638058695886588,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 35068
+    },
+    {
+      "epoch": 0.8347376939921927,
+      "grad_norm": 6.133080638794991,
+      "learning_rate": 2e-05,
+      "loss": 0.9011,
+      "step": 35069
+    },
+    {
+      "epoch": 0.8347614967152243,
+      "grad_norm": 4.293692612883156,
+      "learning_rate": 2e-05,
+      "loss": 0.1064,
+      "step": 35070
+    },
+    {
+      "epoch": 0.8347852994382557,
+      "grad_norm": 0.4963989618767265,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 35071
+    },
+    {
+      "epoch": 0.8348091021612872,
+      "grad_norm": 11.727669051712885,
+      "learning_rate": 2e-05,
+      "loss": 0.4094,
+      "step": 35072
+    },
+    {
+      "epoch": 0.8348329048843187,
+      "grad_norm": 6.378134170004997,
+      "learning_rate": 2e-05,
+      "loss": 1.1937,
+      "step": 35073
+    },
+    {
+      "epoch": 0.8348567076073503,
+      "grad_norm": 2.6381993628788405,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 35074
+    },
+    {
+      "epoch": 0.8348805103303818,
+      "grad_norm": 2.3780002217827705,
+      "learning_rate": 2e-05,
+      "loss": 0.0639,
+      "step": 35075
+    },
+    {
+      "epoch": 0.8349043130534133,
+      "grad_norm": 3.908956501851681,
+      "learning_rate": 2e-05,
+      "loss": 0.5996,
+      "step": 35076
+    },
+    {
+      "epoch": 0.8349281157764449,
+      "grad_norm": 3.0379094911331292,
+      "learning_rate": 2e-05,
+      "loss": 0.4509,
+      "step": 35077
+    },
+    {
+      "epoch": 0.8349519184994764,
+      "grad_norm": 3.7155188223363647,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 35078
+    },
+    {
+      "epoch": 0.8349757212225078,
+      "grad_norm": 10.977109958766013,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 35079
+    },
+    {
+      "epoch": 0.8349995239455393,
+      "grad_norm": 4.9176562752220585,
+      "learning_rate": 2e-05,
+      "loss": 1.0418,
+      "step": 35080
+    },
+    {
+      "epoch": 0.8350233266685709,
+      "grad_norm": 0.36881896797254105,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 35081
+    },
+    {
+      "epoch": 0.8350471293916024,
+      "grad_norm": 3.4612744322070417,
+      "learning_rate": 2e-05,
+      "loss": 0.5541,
+      "step": 35082
+    },
+    {
+      "epoch": 0.8350709321146339,
+      "grad_norm": 3.4411605356689225,
+      "learning_rate": 2e-05,
+      "loss": 0.3816,
+      "step": 35083
+    },
+    {
+      "epoch": 0.8350947348376654,
+      "grad_norm": 4.726222969631016,
+      "learning_rate": 2e-05,
+      "loss": 0.6537,
+      "step": 35084
+    },
+    {
+      "epoch": 0.835118537560697,
+      "grad_norm": 0.10599549086867824,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35085
+    },
+    {
+      "epoch": 0.8351423402837285,
+      "grad_norm": 0.2629497825392819,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 35086
+    },
+    {
+      "epoch": 0.83516614300676,
+      "grad_norm": 3.167308127052327,
+      "learning_rate": 2e-05,
+      "loss": 0.3988,
+      "step": 35087
+    },
+    {
+      "epoch": 0.8351899457297914,
+      "grad_norm": 5.945607948633856,
+      "learning_rate": 2e-05,
+      "loss": 0.7789,
+      "step": 35088
+    },
+    {
+      "epoch": 0.835213748452823,
+      "grad_norm": 3.5835174722374195,
+      "learning_rate": 2e-05,
+      "loss": 0.5381,
+      "step": 35089
+    },
+    {
+      "epoch": 0.8352375511758545,
+      "grad_norm": 0.6409499514862085,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 35090
+    },
+    {
+      "epoch": 0.835261353898886,
+      "grad_norm": 0.14275605007150272,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35091
+    },
+    {
+      "epoch": 0.8352851566219176,
+      "grad_norm": 7.9601316684915835,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 35092
+    },
+    {
+      "epoch": 0.8353089593449491,
+      "grad_norm": 3.4290539639707207,
+      "learning_rate": 2e-05,
+      "loss": 0.1383,
+      "step": 35093
+    },
+    {
+      "epoch": 0.8353327620679806,
+      "grad_norm": 2.8815469298110026,
+      "learning_rate": 2e-05,
+      "loss": 0.3597,
+      "step": 35094
+    },
+    {
+      "epoch": 0.8353565647910121,
+      "grad_norm": 2.4443060088379154,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 35095
+    },
+    {
+      "epoch": 0.8353803675140437,
+      "grad_norm": 10.08699214061854,
+      "learning_rate": 2e-05,
+      "loss": 0.2148,
+      "step": 35096
+    },
+    {
+      "epoch": 0.8354041702370751,
+      "grad_norm": 3.7661451677138804,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 35097
+    },
+    {
+      "epoch": 0.8354279729601066,
+      "grad_norm": 0.2830925186537231,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 35098
+    },
+    {
+      "epoch": 0.8354517756831381,
+      "grad_norm": 1.440071363398912,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 35099
+    },
+    {
+      "epoch": 0.8354755784061697,
+      "grad_norm": 6.7630220425934775,
+      "learning_rate": 2e-05,
+      "loss": 0.6791,
+      "step": 35100
+    },
+    {
+      "epoch": 0.8354993811292012,
+      "grad_norm": 0.5915442778803998,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 35101
+    },
+    {
+      "epoch": 0.8355231838522327,
+      "grad_norm": 6.369120471633287,
+      "learning_rate": 2e-05,
+      "loss": 0.217,
+      "step": 35102
+    },
+    {
+      "epoch": 0.8355469865752642,
+      "grad_norm": 0.03619829158354915,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 35103
+    },
+    {
+      "epoch": 0.8355707892982958,
+      "grad_norm": 2.6618579334683474,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 35104
+    },
+    {
+      "epoch": 0.8355945920213272,
+      "grad_norm": 0.2718538598862656,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 35105
+    },
+    {
+      "epoch": 0.8356183947443587,
+      "grad_norm": 5.5328433507086086,
+      "learning_rate": 2e-05,
+      "loss": 0.6808,
+      "step": 35106
+    },
+    {
+      "epoch": 0.8356421974673903,
+      "grad_norm": 0.4932390915325083,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 35107
+    },
+    {
+      "epoch": 0.8356660001904218,
+      "grad_norm": 4.6455730591298545,
+      "learning_rate": 2e-05,
+      "loss": 0.7919,
+      "step": 35108
+    },
+    {
+      "epoch": 0.8356898029134533,
+      "grad_norm": 2.330091535506465,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 35109
+    },
+    {
+      "epoch": 0.8357136056364848,
+      "grad_norm": 0.04340416101525488,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 35110
+    },
+    {
+      "epoch": 0.8357374083595164,
+      "grad_norm": 5.91137817784313,
+      "learning_rate": 2e-05,
+      "loss": 0.0668,
+      "step": 35111
+    },
+    {
+      "epoch": 0.8357612110825479,
+      "grad_norm": 1.403810041734869,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 35112
+    },
+    {
+      "epoch": 0.8357850138055793,
+      "grad_norm": 5.549462321098797,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 35113
+    },
+    {
+      "epoch": 0.8358088165286108,
+      "grad_norm": 3.6610044178232597,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 35114
+    },
+    {
+      "epoch": 0.8358326192516424,
+      "grad_norm": 10.535682337841044,
+      "learning_rate": 2e-05,
+      "loss": 0.4173,
+      "step": 35115
+    },
+    {
+      "epoch": 0.8358564219746739,
+      "grad_norm": 6.294884355585942,
+      "learning_rate": 2e-05,
+      "loss": 0.3366,
+      "step": 35116
+    },
+    {
+      "epoch": 0.8358802246977054,
+      "grad_norm": 11.256965736494893,
+      "learning_rate": 2e-05,
+      "loss": 2.0165,
+      "step": 35117
+    },
+    {
+      "epoch": 0.8359040274207369,
+      "grad_norm": 6.315732535830571,
+      "learning_rate": 2e-05,
+      "loss": 0.1822,
+      "step": 35118
+    },
+    {
+      "epoch": 0.8359278301437685,
+      "grad_norm": 2.35055225452598,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 35119
+    },
+    {
+      "epoch": 0.8359516328668,
+      "grad_norm": 2.8403818522249664,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 35120
+    },
+    {
+      "epoch": 0.8359754355898315,
+      "grad_norm": 2.2965304551573973,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 35121
+    },
+    {
+      "epoch": 0.835999238312863,
+      "grad_norm": 8.393902752056084,
+      "learning_rate": 2e-05,
+      "loss": 2.0618,
+      "step": 35122
+    },
+    {
+      "epoch": 0.8360230410358945,
+      "grad_norm": 9.452057874466865,
+      "learning_rate": 2e-05,
+      "loss": 1.1003,
+      "step": 35123
+    },
+    {
+      "epoch": 0.836046843758926,
+      "grad_norm": 4.386717648307445,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 35124
+    },
+    {
+      "epoch": 0.8360706464819575,
+      "grad_norm": 0.6691794672337861,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 35125
+    },
+    {
+      "epoch": 0.8360944492049891,
+      "grad_norm": 0.07083109093487959,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35126
+    },
+    {
+      "epoch": 0.8361182519280206,
+      "grad_norm": 4.102778956658577,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 35127
+    },
+    {
+      "epoch": 0.8361420546510521,
+      "grad_norm": 0.04690137415314142,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35128
+    },
+    {
+      "epoch": 0.8361658573740836,
+      "grad_norm": 5.063611624102541,
+      "learning_rate": 2e-05,
+      "loss": 0.0775,
+      "step": 35129
+    },
+    {
+      "epoch": 0.8361896600971152,
+      "grad_norm": 2.7480493017235608,
+      "learning_rate": 2e-05,
+      "loss": 0.0619,
+      "step": 35130
+    },
+    {
+      "epoch": 0.8362134628201466,
+      "grad_norm": 6.9007041444436386,
+      "learning_rate": 2e-05,
+      "loss": 0.1186,
+      "step": 35131
+    },
+    {
+      "epoch": 0.8362372655431781,
+      "grad_norm": 0.6955897645539736,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 35132
+    },
+    {
+      "epoch": 0.8362610682662096,
+      "grad_norm": 2.674859284259627,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 35133
+    },
+    {
+      "epoch": 0.8362848709892412,
+      "grad_norm": 3.5490440765466014,
+      "learning_rate": 2e-05,
+      "loss": 0.6881,
+      "step": 35134
+    },
+    {
+      "epoch": 0.8363086737122727,
+      "grad_norm": 0.14179065357719353,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 35135
+    },
+    {
+      "epoch": 0.8363324764353042,
+      "grad_norm": 4.26627702146822,
+      "learning_rate": 2e-05,
+      "loss": 0.6152,
+      "step": 35136
+    },
+    {
+      "epoch": 0.8363562791583357,
+      "grad_norm": 3.8621820081537908,
+      "learning_rate": 2e-05,
+      "loss": 0.6551,
+      "step": 35137
+    },
+    {
+      "epoch": 0.8363800818813673,
+      "grad_norm": 9.366430858891317,
+      "learning_rate": 2e-05,
+      "loss": 0.9593,
+      "step": 35138
+    },
+    {
+      "epoch": 0.8364038846043987,
+      "grad_norm": 0.022147744108463176,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 35139
+    },
+    {
+      "epoch": 0.8364276873274302,
+      "grad_norm": 4.749030759739586,
+      "learning_rate": 2e-05,
+      "loss": 0.464,
+      "step": 35140
+    },
+    {
+      "epoch": 0.8364514900504618,
+      "grad_norm": 4.4859009111717905,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 35141
+    },
+    {
+      "epoch": 0.8364752927734933,
+      "grad_norm": 0.34144450411995647,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 35142
+    },
+    {
+      "epoch": 0.8364990954965248,
+      "grad_norm": 6.934224546899111,
+      "learning_rate": 2e-05,
+      "loss": 0.2344,
+      "step": 35143
+    },
+    {
+      "epoch": 0.8365228982195563,
+      "grad_norm": 2.139159027763182,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 35144
+    },
+    {
+      "epoch": 0.8365467009425879,
+      "grad_norm": 2.0876487328239257,
+      "learning_rate": 2e-05,
+      "loss": 0.351,
+      "step": 35145
+    },
+    {
+      "epoch": 0.8365705036656194,
+      "grad_norm": 0.6396322569262969,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 35146
+    },
+    {
+      "epoch": 0.8365943063886508,
+      "grad_norm": 3.3072398157689866,
+      "learning_rate": 2e-05,
+      "loss": 0.4211,
+      "step": 35147
+    },
+    {
+      "epoch": 0.8366181091116823,
+      "grad_norm": 0.24996206134975898,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 35148
+    },
+    {
+      "epoch": 0.8366419118347139,
+      "grad_norm": 7.077295529353012,
+      "learning_rate": 2e-05,
+      "loss": 0.6372,
+      "step": 35149
+    },
+    {
+      "epoch": 0.8366657145577454,
+      "grad_norm": 4.259534848859857,
+      "learning_rate": 2e-05,
+      "loss": 0.0873,
+      "step": 35150
+    },
+    {
+      "epoch": 0.8366895172807769,
+      "grad_norm": 0.3174394979438809,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 35151
+    },
+    {
+      "epoch": 0.8367133200038084,
+      "grad_norm": 1.2248898326697182,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 35152
+    },
+    {
+      "epoch": 0.83673712272684,
+      "grad_norm": 1.802002973371245,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 35153
+    },
+    {
+      "epoch": 0.8367609254498715,
+      "grad_norm": 2.0117129556198172,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 35154
+    },
+    {
+      "epoch": 0.836784728172903,
+      "grad_norm": 4.623302172617354,
+      "learning_rate": 2e-05,
+      "loss": 0.8709,
+      "step": 35155
+    },
+    {
+      "epoch": 0.8368085308959345,
+      "grad_norm": 4.199039376160961,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 35156
+    },
+    {
+      "epoch": 0.836832333618966,
+      "grad_norm": 3.5925024780095995,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 35157
+    },
+    {
+      "epoch": 0.8368561363419975,
+      "grad_norm": 4.298964572615879,
+      "learning_rate": 2e-05,
+      "loss": 0.4486,
+      "step": 35158
+    },
+    {
+      "epoch": 0.836879939065029,
+      "grad_norm": 0.4555353413261352,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 35159
+    },
+    {
+      "epoch": 0.8369037417880606,
+      "grad_norm": 3.7759177820106635,
+      "learning_rate": 2e-05,
+      "loss": 0.6292,
+      "step": 35160
+    },
+    {
+      "epoch": 0.8369275445110921,
+      "grad_norm": 3.383619368639247,
+      "learning_rate": 2e-05,
+      "loss": 0.1866,
+      "step": 35161
+    },
+    {
+      "epoch": 0.8369513472341236,
+      "grad_norm": 4.307465430863214,
+      "learning_rate": 2e-05,
+      "loss": 0.7406,
+      "step": 35162
+    },
+    {
+      "epoch": 0.836975149957155,
+      "grad_norm": 3.147106305893566,
+      "learning_rate": 2e-05,
+      "loss": 0.415,
+      "step": 35163
+    },
+    {
+      "epoch": 0.8369989526801866,
+      "grad_norm": 5.325178911177077,
+      "learning_rate": 2e-05,
+      "loss": 1.0702,
+      "step": 35164
+    },
+    {
+      "epoch": 0.8370227554032181,
+      "grad_norm": 3.5571775634824747,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 35165
+    },
+    {
+      "epoch": 0.8370465581262496,
+      "grad_norm": 3.045834205732649,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 35166
+    },
+    {
+      "epoch": 0.8370703608492811,
+      "grad_norm": 0.21695294522748865,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 35167
+    },
+    {
+      "epoch": 0.8370941635723127,
+      "grad_norm": 3.60577517490583,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 35168
+    },
+    {
+      "epoch": 0.8371179662953442,
+      "grad_norm": 4.601409893974782,
+      "learning_rate": 2e-05,
+      "loss": 1.085,
+      "step": 35169
+    },
+    {
+      "epoch": 0.8371417690183757,
+      "grad_norm": 2.4573140610557793,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 35170
+    },
+    {
+      "epoch": 0.8371655717414073,
+      "grad_norm": 3.8616246872186983,
+      "learning_rate": 2e-05,
+      "loss": 0.5633,
+      "step": 35171
+    },
+    {
+      "epoch": 0.8371893744644388,
+      "grad_norm": 0.042174873556401725,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35172
+    },
+    {
+      "epoch": 0.8372131771874702,
+      "grad_norm": 26.3024772504026,
+      "learning_rate": 2e-05,
+      "loss": 1.1123,
+      "step": 35173
+    },
+    {
+      "epoch": 0.8372369799105017,
+      "grad_norm": 0.6904876904123869,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 35174
+    },
+    {
+      "epoch": 0.8372607826335333,
+      "grad_norm": 10.563491581909892,
+      "learning_rate": 2e-05,
+      "loss": 0.5229,
+      "step": 35175
+    },
+    {
+      "epoch": 0.8372845853565648,
+      "grad_norm": 1.5721023912197123,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 35176
+    },
+    {
+      "epoch": 0.8373083880795963,
+      "grad_norm": 1.3581092097258505,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 35177
+    },
+    {
+      "epoch": 0.8373321908026278,
+      "grad_norm": 0.7095099964556362,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 35178
+    },
+    {
+      "epoch": 0.8373559935256594,
+      "grad_norm": 0.0556928008943643,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35179
+    },
+    {
+      "epoch": 0.8373797962486909,
+      "grad_norm": 6.433862515292583,
+      "learning_rate": 2e-05,
+      "loss": 1.8415,
+      "step": 35180
+    },
+    {
+      "epoch": 0.8374035989717223,
+      "grad_norm": 3.1408398230011088,
+      "learning_rate": 2e-05,
+      "loss": 0.373,
+      "step": 35181
+    },
+    {
+      "epoch": 0.8374274016947538,
+      "grad_norm": 16.375570742474032,
+      "learning_rate": 2e-05,
+      "loss": 1.0598,
+      "step": 35182
+    },
+    {
+      "epoch": 0.8374512044177854,
+      "grad_norm": 5.834335282966012,
+      "learning_rate": 2e-05,
+      "loss": 1.5943,
+      "step": 35183
+    },
+    {
+      "epoch": 0.8374750071408169,
+      "grad_norm": 1.316218302824011,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 35184
+    },
+    {
+      "epoch": 0.8374988098638484,
+      "grad_norm": 0.15804765571708285,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35185
+    },
+    {
+      "epoch": 0.83752261258688,
+      "grad_norm": 9.516854710950884,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 35186
+    },
+    {
+      "epoch": 0.8375464153099115,
+      "grad_norm": 2.2206136158144885,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 35187
+    },
+    {
+      "epoch": 0.837570218032943,
+      "grad_norm": 2.341407505286855,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 35188
+    },
+    {
+      "epoch": 0.8375940207559744,
+      "grad_norm": 6.727111284809729,
+      "learning_rate": 2e-05,
+      "loss": 0.2663,
+      "step": 35189
+    },
+    {
+      "epoch": 0.837617823479006,
+      "grad_norm": 3.552682598927861,
+      "learning_rate": 2e-05,
+      "loss": 0.3088,
+      "step": 35190
+    },
+    {
+      "epoch": 0.8376416262020375,
+      "grad_norm": 0.3882421492548331,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 35191
+    },
+    {
+      "epoch": 0.837665428925069,
+      "grad_norm": 10.279045520717062,
+      "learning_rate": 2e-05,
+      "loss": 0.7174,
+      "step": 35192
+    },
+    {
+      "epoch": 0.8376892316481005,
+      "grad_norm": 3.960475979481236,
+      "learning_rate": 2e-05,
+      "loss": 0.599,
+      "step": 35193
+    },
+    {
+      "epoch": 0.8377130343711321,
+      "grad_norm": 0.08611340491185623,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35194
+    },
+    {
+      "epoch": 0.8377368370941636,
+      "grad_norm": 4.134662510433441,
+      "learning_rate": 2e-05,
+      "loss": 1.1039,
+      "step": 35195
+    },
+    {
+      "epoch": 0.8377606398171951,
+      "grad_norm": 0.49996858977989267,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 35196
+    },
+    {
+      "epoch": 0.8377844425402265,
+      "grad_norm": 7.919331048787747,
+      "learning_rate": 2e-05,
+      "loss": 0.4877,
+      "step": 35197
+    },
+    {
+      "epoch": 0.8378082452632581,
+      "grad_norm": 10.230240636455434,
+      "learning_rate": 2e-05,
+      "loss": 0.2171,
+      "step": 35198
+    },
+    {
+      "epoch": 0.8378320479862896,
+      "grad_norm": 6.701041398506733,
+      "learning_rate": 2e-05,
+      "loss": 0.7357,
+      "step": 35199
+    },
+    {
+      "epoch": 0.8378558507093211,
+      "grad_norm": 3.9130382754150506,
+      "learning_rate": 2e-05,
+      "loss": 0.2977,
+      "step": 35200
+    },
+    {
+      "epoch": 0.8378796534323527,
+      "grad_norm": 0.01698678192495418,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 35201
+    },
+    {
+      "epoch": 0.8379034561553842,
+      "grad_norm": 3.4055979851805955,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 35202
+    },
+    {
+      "epoch": 0.8379272588784157,
+      "grad_norm": 6.693461197376322,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 35203
+    },
+    {
+      "epoch": 0.8379510616014472,
+      "grad_norm": 0.15261780378656517,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 35204
+    },
+    {
+      "epoch": 0.8379748643244788,
+      "grad_norm": 3.6063222481658777,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 35205
+    },
+    {
+      "epoch": 0.8379986670475102,
+      "grad_norm": 2.043524214571854,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 35206
+    },
+    {
+      "epoch": 0.8380224697705417,
+      "grad_norm": 2.140423213733256,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 35207
+    },
+    {
+      "epoch": 0.8380462724935732,
+      "grad_norm": 4.503301566925431,
+      "learning_rate": 2e-05,
+      "loss": 0.1366,
+      "step": 35208
+    },
+    {
+      "epoch": 0.8380700752166048,
+      "grad_norm": 3.9945211649819257,
+      "learning_rate": 2e-05,
+      "loss": 0.7458,
+      "step": 35209
+    },
+    {
+      "epoch": 0.8380938779396363,
+      "grad_norm": 7.630623893861726,
+      "learning_rate": 2e-05,
+      "loss": 0.3487,
+      "step": 35210
+    },
+    {
+      "epoch": 0.8381176806626678,
+      "grad_norm": 9.500343369197251,
+      "learning_rate": 2e-05,
+      "loss": 0.731,
+      "step": 35211
+    },
+    {
+      "epoch": 0.8381414833856993,
+      "grad_norm": 11.08217545129299,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 35212
+    },
+    {
+      "epoch": 0.8381652861087309,
+      "grad_norm": 4.252064035980076,
+      "learning_rate": 2e-05,
+      "loss": 0.5094,
+      "step": 35213
+    },
+    {
+      "epoch": 0.8381890888317624,
+      "grad_norm": 14.89557867651014,
+      "learning_rate": 2e-05,
+      "loss": 0.1229,
+      "step": 35214
+    },
+    {
+      "epoch": 0.8382128915547938,
+      "grad_norm": 0.5482223263705287,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 35215
+    },
+    {
+      "epoch": 0.8382366942778254,
+      "grad_norm": 5.193472431212395,
+      "learning_rate": 2e-05,
+      "loss": 0.5237,
+      "step": 35216
+    },
+    {
+      "epoch": 0.8382604970008569,
+      "grad_norm": 6.502713417796903,
+      "learning_rate": 2e-05,
+      "loss": 0.2727,
+      "step": 35217
+    },
+    {
+      "epoch": 0.8382842997238884,
+      "grad_norm": 46.85730652832142,
+      "learning_rate": 2e-05,
+      "loss": 1.5225,
+      "step": 35218
+    },
+    {
+      "epoch": 0.8383081024469199,
+      "grad_norm": 36.89425521649978,
+      "learning_rate": 2e-05,
+      "loss": 1.5528,
+      "step": 35219
+    },
+    {
+      "epoch": 0.8383319051699515,
+      "grad_norm": 6.081269601657347,
+      "learning_rate": 2e-05,
+      "loss": 0.9628,
+      "step": 35220
+    },
+    {
+      "epoch": 0.838355707892983,
+      "grad_norm": 0.753245429300435,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35221
+    },
+    {
+      "epoch": 0.8383795106160145,
+      "grad_norm": 2.1615028073782185,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 35222
+    },
+    {
+      "epoch": 0.8384033133390459,
+      "grad_norm": 3.649615224267104,
+      "learning_rate": 2e-05,
+      "loss": 0.3104,
+      "step": 35223
+    },
+    {
+      "epoch": 0.8384271160620775,
+      "grad_norm": 0.05435584729458659,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 35224
+    },
+    {
+      "epoch": 0.838450918785109,
+      "grad_norm": 2.5263159297491353,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 35225
+    },
+    {
+      "epoch": 0.8384747215081405,
+      "grad_norm": 0.42787088591463507,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 35226
+    },
+    {
+      "epoch": 0.838498524231172,
+      "grad_norm": 3.893744960681871,
+      "learning_rate": 2e-05,
+      "loss": 0.6378,
+      "step": 35227
+    },
+    {
+      "epoch": 0.8385223269542036,
+      "grad_norm": 2.521878602189771,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 35228
+    },
+    {
+      "epoch": 0.8385461296772351,
+      "grad_norm": 0.09244357276225248,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 35229
+    },
+    {
+      "epoch": 0.8385699324002666,
+      "grad_norm": 8.056443137503583,
+      "learning_rate": 2e-05,
+      "loss": 0.7028,
+      "step": 35230
+    },
+    {
+      "epoch": 0.8385937351232982,
+      "grad_norm": 6.542673083327613,
+      "learning_rate": 2e-05,
+      "loss": 1.0868,
+      "step": 35231
+    },
+    {
+      "epoch": 0.8386175378463296,
+      "grad_norm": 3.168481317350915,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 35232
+    },
+    {
+      "epoch": 0.8386413405693611,
+      "grad_norm": 3.9607102515362422,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 35233
+    },
+    {
+      "epoch": 0.8386651432923926,
+      "grad_norm": 0.0642288990143975,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35234
+    },
+    {
+      "epoch": 0.8386889460154242,
+      "grad_norm": 5.241280525077762,
+      "learning_rate": 2e-05,
+      "loss": 0.5727,
+      "step": 35235
+    },
+    {
+      "epoch": 0.8387127487384557,
+      "grad_norm": 3.1514182834219295,
+      "learning_rate": 2e-05,
+      "loss": 0.8413,
+      "step": 35236
+    },
+    {
+      "epoch": 0.8387365514614872,
+      "grad_norm": 4.995586092283102,
+      "learning_rate": 2e-05,
+      "loss": 0.8361,
+      "step": 35237
+    },
+    {
+      "epoch": 0.8387603541845187,
+      "grad_norm": 1.4346750030145572,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 35238
+    },
+    {
+      "epoch": 0.8387841569075503,
+      "grad_norm": 5.463914640797852,
+      "learning_rate": 2e-05,
+      "loss": 0.3607,
+      "step": 35239
+    },
+    {
+      "epoch": 0.8388079596305817,
+      "grad_norm": 0.31113400887417836,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 35240
+    },
+    {
+      "epoch": 0.8388317623536132,
+      "grad_norm": 0.3061039216615284,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 35241
+    },
+    {
+      "epoch": 0.8388555650766447,
+      "grad_norm": 0.7075659729210559,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 35242
+    },
+    {
+      "epoch": 0.8388793677996763,
+      "grad_norm": 0.5096806966590406,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 35243
+    },
+    {
+      "epoch": 0.8389031705227078,
+      "grad_norm": 9.692116657368627,
+      "learning_rate": 2e-05,
+      "loss": 1.9982,
+      "step": 35244
+    },
+    {
+      "epoch": 0.8389269732457393,
+      "grad_norm": 3.5734535043498563,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 35245
+    },
+    {
+      "epoch": 0.8389507759687709,
+      "grad_norm": 4.894419046845663,
+      "learning_rate": 2e-05,
+      "loss": 0.6936,
+      "step": 35246
+    },
+    {
+      "epoch": 0.8389745786918024,
+      "grad_norm": 0.1417406357484351,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35247
+    },
+    {
+      "epoch": 0.8389983814148338,
+      "grad_norm": 3.141754237102535,
+      "learning_rate": 2e-05,
+      "loss": 0.295,
+      "step": 35248
+    },
+    {
+      "epoch": 0.8390221841378653,
+      "grad_norm": 0.5719269515772004,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 35249
+    },
+    {
+      "epoch": 0.8390459868608969,
+      "grad_norm": 12.82328440154306,
+      "learning_rate": 2e-05,
+      "loss": 0.9291,
+      "step": 35250
+    },
+    {
+      "epoch": 0.8390697895839284,
+      "grad_norm": 0.828769840544615,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 35251
+    },
+    {
+      "epoch": 0.8390935923069599,
+      "grad_norm": 3.066101784045677,
+      "learning_rate": 2e-05,
+      "loss": 0.0373,
+      "step": 35252
+    },
+    {
+      "epoch": 0.8391173950299914,
+      "grad_norm": 0.9313034326952174,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 35253
+    },
+    {
+      "epoch": 0.839141197753023,
+      "grad_norm": 5.967671840605108,
+      "learning_rate": 2e-05,
+      "loss": 0.9493,
+      "step": 35254
+    },
+    {
+      "epoch": 0.8391650004760545,
+      "grad_norm": 0.10051961481224402,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 35255
+    },
+    {
+      "epoch": 0.839188803199086,
+      "grad_norm": 3.7591649825621056,
+      "learning_rate": 2e-05,
+      "loss": 0.8321,
+      "step": 35256
+    },
+    {
+      "epoch": 0.8392126059221174,
+      "grad_norm": 2.6764386284970487,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 35257
+    },
+    {
+      "epoch": 0.839236408645149,
+      "grad_norm": 3.387916307406401,
+      "learning_rate": 2e-05,
+      "loss": 0.4333,
+      "step": 35258
+    },
+    {
+      "epoch": 0.8392602113681805,
+      "grad_norm": 3.3397734104596184,
+      "learning_rate": 2e-05,
+      "loss": 0.1875,
+      "step": 35259
+    },
+    {
+      "epoch": 0.839284014091212,
+      "grad_norm": 3.3152742408833284,
+      "learning_rate": 2e-05,
+      "loss": 0.4205,
+      "step": 35260
+    },
+    {
+      "epoch": 0.8393078168142436,
+      "grad_norm": 3.2538829475012063,
+      "learning_rate": 2e-05,
+      "loss": 0.5827,
+      "step": 35261
+    },
+    {
+      "epoch": 0.8393316195372751,
+      "grad_norm": 11.2700274322717,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 35262
+    },
+    {
+      "epoch": 0.8393554222603066,
+      "grad_norm": 11.503302867229861,
+      "learning_rate": 2e-05,
+      "loss": 0.8342,
+      "step": 35263
+    },
+    {
+      "epoch": 0.8393792249833381,
+      "grad_norm": 3.092371941902061,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 35264
+    },
+    {
+      "epoch": 0.8394030277063697,
+      "grad_norm": 4.9476907926835265,
+      "learning_rate": 2e-05,
+      "loss": 0.9393,
+      "step": 35265
+    },
+    {
+      "epoch": 0.8394268304294011,
+      "grad_norm": 4.664520373073307,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 35266
+    },
+    {
+      "epoch": 0.8394506331524326,
+      "grad_norm": 2.7811665900369644,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 35267
+    },
+    {
+      "epoch": 0.8394744358754641,
+      "grad_norm": 9.604190123049403,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 35268
+    },
+    {
+      "epoch": 0.8394982385984957,
+      "grad_norm": 16.541510656897724,
+      "learning_rate": 2e-05,
+      "loss": 0.8724,
+      "step": 35269
+    },
+    {
+      "epoch": 0.8395220413215272,
+      "grad_norm": 10.700757541724872,
+      "learning_rate": 2e-05,
+      "loss": 0.284,
+      "step": 35270
+    },
+    {
+      "epoch": 0.8395458440445587,
+      "grad_norm": 9.86895672333303,
+      "learning_rate": 2e-05,
+      "loss": 0.5406,
+      "step": 35271
+    },
+    {
+      "epoch": 0.8395696467675902,
+      "grad_norm": 0.08994106530261925,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 35272
+    },
+    {
+      "epoch": 0.8395934494906218,
+      "grad_norm": 4.311642770217177,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 35273
+    },
+    {
+      "epoch": 0.8396172522136532,
+      "grad_norm": 9.54597909202718,
+      "learning_rate": 2e-05,
+      "loss": 0.1464,
+      "step": 35274
+    },
+    {
+      "epoch": 0.8396410549366847,
+      "grad_norm": 0.5282830671124659,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 35275
+    },
+    {
+      "epoch": 0.8396648576597163,
+      "grad_norm": 0.2037382065617179,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 35276
+    },
+    {
+      "epoch": 0.8396886603827478,
+      "grad_norm": 3.4176608630422582,
+      "learning_rate": 2e-05,
+      "loss": 0.6631,
+      "step": 35277
+    },
+    {
+      "epoch": 0.8397124631057793,
+      "grad_norm": 5.043634629256938,
+      "learning_rate": 2e-05,
+      "loss": 0.6734,
+      "step": 35278
+    },
+    {
+      "epoch": 0.8397362658288108,
+      "grad_norm": 3.9167885493649846,
+      "learning_rate": 2e-05,
+      "loss": 0.5685,
+      "step": 35279
+    },
+    {
+      "epoch": 0.8397600685518424,
+      "grad_norm": 8.030247967463326,
+      "learning_rate": 2e-05,
+      "loss": 2.3846,
+      "step": 35280
+    },
+    {
+      "epoch": 0.8397838712748739,
+      "grad_norm": 2.939947013097482,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 35281
+    },
+    {
+      "epoch": 0.8398076739979053,
+      "grad_norm": 6.579095116044693,
+      "learning_rate": 2e-05,
+      "loss": 0.8424,
+      "step": 35282
+    },
+    {
+      "epoch": 0.8398314767209368,
+      "grad_norm": 0.9323610966044642,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 35283
+    },
+    {
+      "epoch": 0.8398552794439684,
+      "grad_norm": 5.234745831035505,
+      "learning_rate": 2e-05,
+      "loss": 0.8072,
+      "step": 35284
+    },
+    {
+      "epoch": 0.8398790821669999,
+      "grad_norm": 0.10948889006418736,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35285
+    },
+    {
+      "epoch": 0.8399028848900314,
+      "grad_norm": 0.05132835185168436,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35286
+    },
+    {
+      "epoch": 0.8399266876130629,
+      "grad_norm": 2.085751231361111,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 35287
+    },
+    {
+      "epoch": 0.8399504903360945,
+      "grad_norm": 3.9586730787835256,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 35288
+    },
+    {
+      "epoch": 0.839974293059126,
+      "grad_norm": 4.030765795861279,
+      "learning_rate": 2e-05,
+      "loss": 0.5899,
+      "step": 35289
+    },
+    {
+      "epoch": 0.8399980957821575,
+      "grad_norm": 0.6514004499599237,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 35290
+    },
+    {
+      "epoch": 0.840021898505189,
+      "grad_norm": 0.7217503961039357,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 35291
+    },
+    {
+      "epoch": 0.8400457012282205,
+      "grad_norm": 6.848575322042909,
+      "learning_rate": 2e-05,
+      "loss": 0.0788,
+      "step": 35292
+    },
+    {
+      "epoch": 0.840069503951252,
+      "grad_norm": 8.314108697226466,
+      "learning_rate": 2e-05,
+      "loss": 0.3903,
+      "step": 35293
+    },
+    {
+      "epoch": 0.8400933066742835,
+      "grad_norm": 0.18156487475282432,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 35294
+    },
+    {
+      "epoch": 0.8401171093973151,
+      "grad_norm": 1.7932560971341145,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 35295
+    },
+    {
+      "epoch": 0.8401409121203466,
+      "grad_norm": 23.173880727043446,
+      "learning_rate": 2e-05,
+      "loss": 0.708,
+      "step": 35296
+    },
+    {
+      "epoch": 0.8401647148433781,
+      "grad_norm": 4.111368801777755,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 35297
+    },
+    {
+      "epoch": 0.8401885175664096,
+      "grad_norm": 1.057482436267191,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 35298
+    },
+    {
+      "epoch": 0.8402123202894412,
+      "grad_norm": 0.20765367714645222,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 35299
+    },
+    {
+      "epoch": 0.8402361230124726,
+      "grad_norm": 0.6673070780354804,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 35300
+    },
+    {
+      "epoch": 0.8402599257355041,
+      "grad_norm": 4.972082555776114,
+      "learning_rate": 2e-05,
+      "loss": 0.8453,
+      "step": 35301
+    },
+    {
+      "epoch": 0.8402837284585356,
+      "grad_norm": 1.7192369036293111,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 35302
+    },
+    {
+      "epoch": 0.8403075311815672,
+      "grad_norm": 0.4790264934140664,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 35303
+    },
+    {
+      "epoch": 0.8403313339045987,
+      "grad_norm": 17.03128991963419,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 35304
+    },
+    {
+      "epoch": 0.8403551366276302,
+      "grad_norm": 3.7199159523590213,
+      "learning_rate": 2e-05,
+      "loss": 0.7685,
+      "step": 35305
+    },
+    {
+      "epoch": 0.8403789393506618,
+      "grad_norm": 3.2603610617085703,
+      "learning_rate": 2e-05,
+      "loss": 0.4302,
+      "step": 35306
+    },
+    {
+      "epoch": 0.8404027420736933,
+      "grad_norm": 6.41576237082299,
+      "learning_rate": 2e-05,
+      "loss": 1.1731,
+      "step": 35307
+    },
+    {
+      "epoch": 0.8404265447967247,
+      "grad_norm": 0.27650633029601523,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 35308
+    },
+    {
+      "epoch": 0.8404503475197562,
+      "grad_norm": 1.2390357611639866,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 35309
+    },
+    {
+      "epoch": 0.8404741502427878,
+      "grad_norm": 4.510658044299411,
+      "learning_rate": 2e-05,
+      "loss": 0.1721,
+      "step": 35310
+    },
+    {
+      "epoch": 0.8404979529658193,
+      "grad_norm": 4.643044993202854,
+      "learning_rate": 2e-05,
+      "loss": 1.1542,
+      "step": 35311
+    },
+    {
+      "epoch": 0.8405217556888508,
+      "grad_norm": 1.5966813080372755,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 35312
+    },
+    {
+      "epoch": 0.8405455584118823,
+      "grad_norm": 0.1293579937633114,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 35313
+    },
+    {
+      "epoch": 0.8405693611349139,
+      "grad_norm": 5.629970627254644,
+      "learning_rate": 2e-05,
+      "loss": 0.1341,
+      "step": 35314
+    },
+    {
+      "epoch": 0.8405931638579454,
+      "grad_norm": 0.18415040704236052,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 35315
+    },
+    {
+      "epoch": 0.8406169665809768,
+      "grad_norm": 2.3346340999058834,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 35316
+    },
+    {
+      "epoch": 0.8406407693040083,
+      "grad_norm": 5.259643206765027,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 35317
+    },
+    {
+      "epoch": 0.8406645720270399,
+      "grad_norm": 0.004301587333153896,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 35318
+    },
+    {
+      "epoch": 0.8406883747500714,
+      "grad_norm": 8.899809102551663,
+      "learning_rate": 2e-05,
+      "loss": 0.5426,
+      "step": 35319
+    },
+    {
+      "epoch": 0.8407121774731029,
+      "grad_norm": 1.2128738809697561,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 35320
+    },
+    {
+      "epoch": 0.8407359801961345,
+      "grad_norm": 2.6705969690096905,
+      "learning_rate": 2e-05,
+      "loss": 0.5721,
+      "step": 35321
+    },
+    {
+      "epoch": 0.840759782919166,
+      "grad_norm": 8.52039863062911,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 35322
+    },
+    {
+      "epoch": 0.8407835856421975,
+      "grad_norm": 13.305793225025063,
+      "learning_rate": 2e-05,
+      "loss": 0.8423,
+      "step": 35323
+    },
+    {
+      "epoch": 0.840807388365229,
+      "grad_norm": 0.1598669976984874,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35324
+    },
+    {
+      "epoch": 0.8408311910882605,
+      "grad_norm": 4.441477794583399,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 35325
+    },
+    {
+      "epoch": 0.840854993811292,
+      "grad_norm": 2.6438560759287215,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 35326
+    },
+    {
+      "epoch": 0.8408787965343235,
+      "grad_norm": 0.30510657739342895,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35327
+    },
+    {
+      "epoch": 0.840902599257355,
+      "grad_norm": 2.071153156797251,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 35328
+    },
+    {
+      "epoch": 0.8409264019803866,
+      "grad_norm": 0.24547603003050333,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 35329
+    },
+    {
+      "epoch": 0.8409502047034181,
+      "grad_norm": 3.7980626692926074,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 35330
+    },
+    {
+      "epoch": 0.8409740074264496,
+      "grad_norm": 8.738800803925105,
+      "learning_rate": 2e-05,
+      "loss": 0.6732,
+      "step": 35331
+    },
+    {
+      "epoch": 0.840997810149481,
+      "grad_norm": 2.388934834208994,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 35332
+    },
+    {
+      "epoch": 0.8410216128725126,
+      "grad_norm": 2.3460838027698636,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 35333
+    },
+    {
+      "epoch": 0.8410454155955441,
+      "grad_norm": 0.3902338029854889,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 35334
+    },
+    {
+      "epoch": 0.8410692183185756,
+      "grad_norm": 3.714704016408179,
+      "learning_rate": 2e-05,
+      "loss": 0.3079,
+      "step": 35335
+    },
+    {
+      "epoch": 0.8410930210416072,
+      "grad_norm": 0.11465025154988254,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 35336
+    },
+    {
+      "epoch": 0.8411168237646387,
+      "grad_norm": 12.891050931337768,
+      "learning_rate": 2e-05,
+      "loss": 0.6141,
+      "step": 35337
+    },
+    {
+      "epoch": 0.8411406264876702,
+      "grad_norm": 3.863405195929984,
+      "learning_rate": 2e-05,
+      "loss": 0.1307,
+      "step": 35338
+    },
+    {
+      "epoch": 0.8411644292107017,
+      "grad_norm": 7.713235925538853,
+      "learning_rate": 2e-05,
+      "loss": 0.7897,
+      "step": 35339
+    },
+    {
+      "epoch": 0.8411882319337333,
+      "grad_norm": 4.375372759337933,
+      "learning_rate": 2e-05,
+      "loss": 0.6477,
+      "step": 35340
+    },
+    {
+      "epoch": 0.8412120346567648,
+      "grad_norm": 3.1464766401280495,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 35341
+    },
+    {
+      "epoch": 0.8412358373797962,
+      "grad_norm": 0.09139405600986392,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35342
+    },
+    {
+      "epoch": 0.8412596401028277,
+      "grad_norm": 1.2306941767120783,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 35343
+    },
+    {
+      "epoch": 0.8412834428258593,
+      "grad_norm": 0.6926059112993379,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 35344
+    },
+    {
+      "epoch": 0.8413072455488908,
+      "grad_norm": 7.017402144795785,
+      "learning_rate": 2e-05,
+      "loss": 0.4329,
+      "step": 35345
+    },
+    {
+      "epoch": 0.8413310482719223,
+      "grad_norm": 0.4872266330062203,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 35346
+    },
+    {
+      "epoch": 0.8413548509949538,
+      "grad_norm": 7.875432220487568,
+      "learning_rate": 2e-05,
+      "loss": 1.1202,
+      "step": 35347
+    },
+    {
+      "epoch": 0.8413786537179854,
+      "grad_norm": 0.18128520984146482,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 35348
+    },
+    {
+      "epoch": 0.8414024564410169,
+      "grad_norm": 20.55783631408362,
+      "learning_rate": 2e-05,
+      "loss": 0.5116,
+      "step": 35349
+    },
+    {
+      "epoch": 0.8414262591640483,
+      "grad_norm": 3.417167422500094,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 35350
+    },
+    {
+      "epoch": 0.8414500618870799,
+      "grad_norm": 0.0450145021011022,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 35351
+    },
+    {
+      "epoch": 0.8414738646101114,
+      "grad_norm": 4.361196653022841,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 35352
+    },
+    {
+      "epoch": 0.8414976673331429,
+      "grad_norm": 1.980345707227013,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 35353
+    },
+    {
+      "epoch": 0.8415214700561744,
+      "grad_norm": 0.3277454617161575,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 35354
+    },
+    {
+      "epoch": 0.841545272779206,
+      "grad_norm": 16.579781920466026,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 35355
+    },
+    {
+      "epoch": 0.8415690755022375,
+      "grad_norm": 9.140705086004955,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 35356
+    },
+    {
+      "epoch": 0.841592878225269,
+      "grad_norm": 0.4810343618369237,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 35357
+    },
+    {
+      "epoch": 0.8416166809483004,
+      "grad_norm": 7.132734058667287,
+      "learning_rate": 2e-05,
+      "loss": 0.8786,
+      "step": 35358
+    },
+    {
+      "epoch": 0.841640483671332,
+      "grad_norm": 0.03778114580545457,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35359
+    },
+    {
+      "epoch": 0.8416642863943635,
+      "grad_norm": 15.450695728945114,
+      "learning_rate": 2e-05,
+      "loss": 0.7618,
+      "step": 35360
+    },
+    {
+      "epoch": 0.841688089117395,
+      "grad_norm": 1.5373935412314197,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 35361
+    },
+    {
+      "epoch": 0.8417118918404265,
+      "grad_norm": 29.787321117522605,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 35362
+    },
+    {
+      "epoch": 0.8417356945634581,
+      "grad_norm": 0.8010700972783432,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 35363
+    },
+    {
+      "epoch": 0.8417594972864896,
+      "grad_norm": 27.51977716724553,
+      "learning_rate": 2e-05,
+      "loss": 0.6127,
+      "step": 35364
+    },
+    {
+      "epoch": 0.8417833000095211,
+      "grad_norm": 4.275127858020413,
+      "learning_rate": 2e-05,
+      "loss": 0.6335,
+      "step": 35365
+    },
+    {
+      "epoch": 0.8418071027325527,
+      "grad_norm": 3.0146158653818005,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 35366
+    },
+    {
+      "epoch": 0.8418309054555841,
+      "grad_norm": 0.15379919698955652,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 35367
+    },
+    {
+      "epoch": 0.8418547081786156,
+      "grad_norm": 11.509780750092139,
+      "learning_rate": 2e-05,
+      "loss": 0.4649,
+      "step": 35368
+    },
+    {
+      "epoch": 0.8418785109016471,
+      "grad_norm": 0.1448455885821658,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35369
+    },
+    {
+      "epoch": 0.8419023136246787,
+      "grad_norm": 0.06733543293482461,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35370
+    },
+    {
+      "epoch": 0.8419261163477102,
+      "grad_norm": 0.044945668814246734,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35371
+    },
+    {
+      "epoch": 0.8419499190707417,
+      "grad_norm": 1.728366562900575,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 35372
+    },
+    {
+      "epoch": 0.8419737217937732,
+      "grad_norm": 0.27570363563091654,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 35373
+    },
+    {
+      "epoch": 0.8419975245168048,
+      "grad_norm": 4.879926638223905,
+      "learning_rate": 2e-05,
+      "loss": 0.2919,
+      "step": 35374
+    },
+    {
+      "epoch": 0.8420213272398362,
+      "grad_norm": 0.46144973591611727,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 35375
+    },
+    {
+      "epoch": 0.8420451299628677,
+      "grad_norm": 4.788693547224831,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 35376
+    },
+    {
+      "epoch": 0.8420689326858992,
+      "grad_norm": 3.3781029260814988,
+      "learning_rate": 2e-05,
+      "loss": 0.2217,
+      "step": 35377
+    },
+    {
+      "epoch": 0.8420927354089308,
+      "grad_norm": 1.1350384984504103,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 35378
+    },
+    {
+      "epoch": 0.8421165381319623,
+      "grad_norm": 3.026640770895234,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 35379
+    },
+    {
+      "epoch": 0.8421403408549938,
+      "grad_norm": 2.7046006134817553,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 35380
+    },
+    {
+      "epoch": 0.8421641435780254,
+      "grad_norm": 3.242295184649764,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 35381
+    },
+    {
+      "epoch": 0.8421879463010569,
+      "grad_norm": 1.0630228533197967,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 35382
+    },
+    {
+      "epoch": 0.8422117490240884,
+      "grad_norm": 5.049511954987685,
+      "learning_rate": 2e-05,
+      "loss": 0.6976,
+      "step": 35383
+    },
+    {
+      "epoch": 0.8422355517471198,
+      "grad_norm": 4.417321567707487,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 35384
+    },
+    {
+      "epoch": 0.8422593544701514,
+      "grad_norm": 22.115731896367997,
+      "learning_rate": 2e-05,
+      "loss": 0.4282,
+      "step": 35385
+    },
+    {
+      "epoch": 0.8422831571931829,
+      "grad_norm": 84.52516669313437,
+      "learning_rate": 2e-05,
+      "loss": 1.0729,
+      "step": 35386
+    },
+    {
+      "epoch": 0.8423069599162144,
+      "grad_norm": 8.224505626206328,
+      "learning_rate": 2e-05,
+      "loss": 0.1957,
+      "step": 35387
+    },
+    {
+      "epoch": 0.8423307626392459,
+      "grad_norm": 5.020325165922754,
+      "learning_rate": 2e-05,
+      "loss": 0.6764,
+      "step": 35388
+    },
+    {
+      "epoch": 0.8423545653622775,
+      "grad_norm": 0.18103241674027215,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 35389
+    },
+    {
+      "epoch": 0.842378368085309,
+      "grad_norm": 5.2147429216668515,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 35390
+    },
+    {
+      "epoch": 0.8424021708083405,
+      "grad_norm": 0.1789398963140252,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 35391
+    },
+    {
+      "epoch": 0.8424259735313719,
+      "grad_norm": 0.7779993853501157,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 35392
+    },
+    {
+      "epoch": 0.8424497762544035,
+      "grad_norm": 7.899524453531659,
+      "learning_rate": 2e-05,
+      "loss": 0.242,
+      "step": 35393
+    },
+    {
+      "epoch": 0.842473578977435,
+      "grad_norm": 3.1106585948522376,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 35394
+    },
+    {
+      "epoch": 0.8424973817004665,
+      "grad_norm": 0.20821793947858905,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35395
+    },
+    {
+      "epoch": 0.842521184423498,
+      "grad_norm": 6.830975902912387,
+      "learning_rate": 2e-05,
+      "loss": 0.1846,
+      "step": 35396
+    },
+    {
+      "epoch": 0.8425449871465296,
+      "grad_norm": 3.4812056370930504,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 35397
+    },
+    {
+      "epoch": 0.8425687898695611,
+      "grad_norm": 0.523561405988535,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 35398
+    },
+    {
+      "epoch": 0.8425925925925926,
+      "grad_norm": 4.373570413578524,
+      "learning_rate": 2e-05,
+      "loss": 0.2909,
+      "step": 35399
+    },
+    {
+      "epoch": 0.8426163953156242,
+      "grad_norm": 9.626262713259075,
+      "learning_rate": 2e-05,
+      "loss": 1.2193,
+      "step": 35400
+    },
+    {
+      "epoch": 0.8426401980386556,
+      "grad_norm": 6.423073175689761,
+      "learning_rate": 2e-05,
+      "loss": 1.5993,
+      "step": 35401
+    },
+    {
+      "epoch": 0.8426640007616871,
+      "grad_norm": 4.171187406270332,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 35402
+    },
+    {
+      "epoch": 0.8426878034847186,
+      "grad_norm": 5.0444183849763595,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 35403
+    },
+    {
+      "epoch": 0.8427116062077502,
+      "grad_norm": 4.216679766075694,
+      "learning_rate": 2e-05,
+      "loss": 0.7052,
+      "step": 35404
+    },
+    {
+      "epoch": 0.8427354089307817,
+      "grad_norm": 0.0739034365500918,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35405
+    },
+    {
+      "epoch": 0.8427592116538132,
+      "grad_norm": 0.019252355735939076,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 35406
+    },
+    {
+      "epoch": 0.8427830143768447,
+      "grad_norm": 2.620155220967372,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 35407
+    },
+    {
+      "epoch": 0.8428068170998763,
+      "grad_norm": 0.3999127371832391,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 35408
+    },
+    {
+      "epoch": 0.8428306198229077,
+      "grad_norm": 1.3742213970556516,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 35409
+    },
+    {
+      "epoch": 0.8428544225459392,
+      "grad_norm": 3.69921124384324,
+      "learning_rate": 2e-05,
+      "loss": 0.3838,
+      "step": 35410
+    },
+    {
+      "epoch": 0.8428782252689707,
+      "grad_norm": 2.5551652236328724,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 35411
+    },
+    {
+      "epoch": 0.8429020279920023,
+      "grad_norm": 3.4760671064370237,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 35412
+    },
+    {
+      "epoch": 0.8429258307150338,
+      "grad_norm": 3.804446399662031,
+      "learning_rate": 2e-05,
+      "loss": 1.0153,
+      "step": 35413
+    },
+    {
+      "epoch": 0.8429496334380653,
+      "grad_norm": 2.985384676809881,
+      "learning_rate": 2e-05,
+      "loss": 0.8077,
+      "step": 35414
+    },
+    {
+      "epoch": 0.8429734361610969,
+      "grad_norm": 0.06735874980344463,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 35415
+    },
+    {
+      "epoch": 0.8429972388841284,
+      "grad_norm": 5.172478105656099,
+      "learning_rate": 2e-05,
+      "loss": 0.1374,
+      "step": 35416
+    },
+    {
+      "epoch": 0.8430210416071598,
+      "grad_norm": 2.540384088498134,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 35417
+    },
+    {
+      "epoch": 0.8430448443301913,
+      "grad_norm": 1.4842122747545141,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 35418
+    },
+    {
+      "epoch": 0.8430686470532229,
+      "grad_norm": 14.424432126806071,
+      "learning_rate": 2e-05,
+      "loss": 0.7241,
+      "step": 35419
+    },
+    {
+      "epoch": 0.8430924497762544,
+      "grad_norm": 1.4804565878256797,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 35420
+    },
+    {
+      "epoch": 0.8431162524992859,
+      "grad_norm": 1.9999736852922645,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 35421
+    },
+    {
+      "epoch": 0.8431400552223174,
+      "grad_norm": 0.2924077421837491,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 35422
+    },
+    {
+      "epoch": 0.843163857945349,
+      "grad_norm": 0.99670276738252,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 35423
+    },
+    {
+      "epoch": 0.8431876606683805,
+      "grad_norm": 4.895662466070142,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 35424
+    },
+    {
+      "epoch": 0.843211463391412,
+      "grad_norm": 5.633965825118639,
+      "learning_rate": 2e-05,
+      "loss": 1.3023,
+      "step": 35425
+    },
+    {
+      "epoch": 0.8432352661144434,
+      "grad_norm": 2.8730103393671804,
+      "learning_rate": 2e-05,
+      "loss": 0.4076,
+      "step": 35426
+    },
+    {
+      "epoch": 0.843259068837475,
+      "grad_norm": 0.46840143683871416,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 35427
+    },
+    {
+      "epoch": 0.8432828715605065,
+      "grad_norm": 3.9873514004818023,
+      "learning_rate": 2e-05,
+      "loss": 0.7801,
+      "step": 35428
+    },
+    {
+      "epoch": 0.843306674283538,
+      "grad_norm": 14.768749359080745,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 35429
+    },
+    {
+      "epoch": 0.8433304770065696,
+      "grad_norm": 2.3695991693357032,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 35430
+    },
+    {
+      "epoch": 0.8433542797296011,
+      "grad_norm": 4.530242842354708,
+      "learning_rate": 2e-05,
+      "loss": 0.764,
+      "step": 35431
+    },
+    {
+      "epoch": 0.8433780824526326,
+      "grad_norm": 0.7611498012628449,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 35432
+    },
+    {
+      "epoch": 0.8434018851756641,
+      "grad_norm": 0.03040527287573101,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 35433
+    },
+    {
+      "epoch": 0.8434256878986957,
+      "grad_norm": 1.2559399301884815,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 35434
+    },
+    {
+      "epoch": 0.8434494906217271,
+      "grad_norm": 0.1101794801279591,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 35435
+    },
+    {
+      "epoch": 0.8434732933447586,
+      "grad_norm": 3.7090130835120525,
+      "learning_rate": 2e-05,
+      "loss": 0.3804,
+      "step": 35436
+    },
+    {
+      "epoch": 0.8434970960677901,
+      "grad_norm": 2.711712026009955,
+      "learning_rate": 2e-05,
+      "loss": 0.3927,
+      "step": 35437
+    },
+    {
+      "epoch": 0.8435208987908217,
+      "grad_norm": 3.161155976209215,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 35438
+    },
+    {
+      "epoch": 0.8435447015138532,
+      "grad_norm": 9.02372237078442,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 35439
+    },
+    {
+      "epoch": 0.8435685042368847,
+      "grad_norm": 0.030184945881318943,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 35440
+    },
+    {
+      "epoch": 0.8435923069599162,
+      "grad_norm": 0.15756139529022345,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 35441
+    },
+    {
+      "epoch": 0.8436161096829478,
+      "grad_norm": 4.258975103100216,
+      "learning_rate": 2e-05,
+      "loss": 0.9372,
+      "step": 35442
+    },
+    {
+      "epoch": 0.8436399124059792,
+      "grad_norm": 2.833350402907891,
+      "learning_rate": 2e-05,
+      "loss": 0.5202,
+      "step": 35443
+    },
+    {
+      "epoch": 0.8436637151290107,
+      "grad_norm": 7.254213265798549,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 35444
+    },
+    {
+      "epoch": 0.8436875178520423,
+      "grad_norm": 0.21567369634476932,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 35445
+    },
+    {
+      "epoch": 0.8437113205750738,
+      "grad_norm": 5.195840402156241,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 35446
+    },
+    {
+      "epoch": 0.8437351232981053,
+      "grad_norm": 3.1755966035653302,
+      "learning_rate": 2e-05,
+      "loss": 0.1314,
+      "step": 35447
+    },
+    {
+      "epoch": 0.8437589260211368,
+      "grad_norm": 7.241872857374433,
+      "learning_rate": 2e-05,
+      "loss": 0.9514,
+      "step": 35448
+    },
+    {
+      "epoch": 0.8437827287441684,
+      "grad_norm": 0.09583096975042596,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 35449
+    },
+    {
+      "epoch": 0.8438065314671999,
+      "grad_norm": 5.256954611217794,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 35450
+    },
+    {
+      "epoch": 0.8438303341902313,
+      "grad_norm": 0.0623399935033361,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35451
+    },
+    {
+      "epoch": 0.8438541369132628,
+      "grad_norm": 10.065062781200945,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 35452
+    },
+    {
+      "epoch": 0.8438779396362944,
+      "grad_norm": 4.962291994097503,
+      "learning_rate": 2e-05,
+      "loss": 0.092,
+      "step": 35453
+    },
+    {
+      "epoch": 0.8439017423593259,
+      "grad_norm": 5.364681601741936,
+      "learning_rate": 2e-05,
+      "loss": 0.7082,
+      "step": 35454
+    },
+    {
+      "epoch": 0.8439255450823574,
+      "grad_norm": 3.6713354193377072,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 35455
+    },
+    {
+      "epoch": 0.8439493478053889,
+      "grad_norm": 8.550191106150418,
+      "learning_rate": 2e-05,
+      "loss": 0.8551,
+      "step": 35456
+    },
+    {
+      "epoch": 0.8439731505284205,
+      "grad_norm": 723.0321345045178,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 35457
+    },
+    {
+      "epoch": 0.843996953251452,
+      "grad_norm": 2.756541960034534,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 35458
+    },
+    {
+      "epoch": 0.8440207559744834,
+      "grad_norm": 5.190946263719342,
+      "learning_rate": 2e-05,
+      "loss": 0.1117,
+      "step": 35459
+    },
+    {
+      "epoch": 0.844044558697515,
+      "grad_norm": 43.54499011043265,
+      "learning_rate": 2e-05,
+      "loss": 1.4906,
+      "step": 35460
+    },
+    {
+      "epoch": 0.8440683614205465,
+      "grad_norm": 9.593047406182508,
+      "learning_rate": 2e-05,
+      "loss": 1.0764,
+      "step": 35461
+    },
+    {
+      "epoch": 0.844092164143578,
+      "grad_norm": 2622.096319117462,
+      "learning_rate": 2e-05,
+      "loss": 1.2966,
+      "step": 35462
+    },
+    {
+      "epoch": 0.8441159668666095,
+      "grad_norm": 5.570547887352988,
+      "learning_rate": 2e-05,
+      "loss": 0.7677,
+      "step": 35463
+    },
+    {
+      "epoch": 0.8441397695896411,
+      "grad_norm": 25.783501453539866,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 35464
+    },
+    {
+      "epoch": 0.8441635723126726,
+      "grad_norm": 4.9710649353890854,
+      "learning_rate": 2e-05,
+      "loss": 0.218,
+      "step": 35465
+    },
+    {
+      "epoch": 0.8441873750357041,
+      "grad_norm": 7.718256514014906,
+      "learning_rate": 2e-05,
+      "loss": 1.6027,
+      "step": 35466
+    },
+    {
+      "epoch": 0.8442111777587356,
+      "grad_norm": 6.524939876477969,
+      "learning_rate": 2e-05,
+      "loss": 0.2027,
+      "step": 35467
+    },
+    {
+      "epoch": 0.8442349804817671,
+      "grad_norm": 101.94817899291789,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 35468
+    },
+    {
+      "epoch": 0.8442587832047986,
+      "grad_norm": 4.723631047362309,
+      "learning_rate": 2e-05,
+      "loss": 0.5404,
+      "step": 35469
+    },
+    {
+      "epoch": 0.8442825859278301,
+      "grad_norm": 8.573991692448436,
+      "learning_rate": 2e-05,
+      "loss": 0.7623,
+      "step": 35470
+    },
+    {
+      "epoch": 0.8443063886508616,
+      "grad_norm": 1.9429561367586816,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 35471
+    },
+    {
+      "epoch": 0.8443301913738932,
+      "grad_norm": 2.948382123683992,
+      "learning_rate": 2e-05,
+      "loss": 0.2764,
+      "step": 35472
+    },
+    {
+      "epoch": 0.8443539940969247,
+      "grad_norm": 2.263619436890732,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 35473
+    },
+    {
+      "epoch": 0.8443777968199562,
+      "grad_norm": 0.46388427684095274,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 35474
+    },
+    {
+      "epoch": 0.8444015995429878,
+      "grad_norm": 3.0688745056952897,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 35475
+    },
+    {
+      "epoch": 0.8444254022660193,
+      "grad_norm": 0.09554212474394576,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35476
+    },
+    {
+      "epoch": 0.8444492049890507,
+      "grad_norm": 360.7415769247765,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 35477
+    },
+    {
+      "epoch": 0.8444730077120822,
+      "grad_norm": 0.4086661198376228,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 35478
+    },
+    {
+      "epoch": 0.8444968104351138,
+      "grad_norm": 2.7958423589272754,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 35479
+    },
+    {
+      "epoch": 0.8445206131581453,
+      "grad_norm": 113.32966831438824,
+      "learning_rate": 2e-05,
+      "loss": 0.7199,
+      "step": 35480
+    },
+    {
+      "epoch": 0.8445444158811768,
+      "grad_norm": 13.37387313101295,
+      "learning_rate": 2e-05,
+      "loss": 0.4387,
+      "step": 35481
+    },
+    {
+      "epoch": 0.8445682186042083,
+      "grad_norm": 0.15564632556812039,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 35482
+    },
+    {
+      "epoch": 0.8445920213272399,
+      "grad_norm": 16.401892410548438,
+      "learning_rate": 2e-05,
+      "loss": 1.0259,
+      "step": 35483
+    },
+    {
+      "epoch": 0.8446158240502714,
+      "grad_norm": 3.290366236279881,
+      "learning_rate": 2e-05,
+      "loss": 0.5052,
+      "step": 35484
+    },
+    {
+      "epoch": 0.8446396267733028,
+      "grad_norm": 0.5269026630934691,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 35485
+    },
+    {
+      "epoch": 0.8446634294963343,
+      "grad_norm": 7.748301911462888,
+      "learning_rate": 2e-05,
+      "loss": 0.5179,
+      "step": 35486
+    },
+    {
+      "epoch": 0.8446872322193659,
+      "grad_norm": 102.0695082882324,
+      "learning_rate": 2e-05,
+      "loss": 1.224,
+      "step": 35487
+    },
+    {
+      "epoch": 0.8447110349423974,
+      "grad_norm": 325.9092325082873,
+      "learning_rate": 2e-05,
+      "loss": 3.2017,
+      "step": 35488
+    },
+    {
+      "epoch": 0.8447348376654289,
+      "grad_norm": 0.9461730808266382,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 35489
+    },
+    {
+      "epoch": 0.8447586403884605,
+      "grad_norm": 2.170067478686278,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 35490
+    },
+    {
+      "epoch": 0.844782443111492,
+      "grad_norm": 0.015044434922119728,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 35491
+    },
+    {
+      "epoch": 0.8448062458345235,
+      "grad_norm": 2.716274162929299,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 35492
+    },
+    {
+      "epoch": 0.844830048557555,
+      "grad_norm": 95.95236065090663,
+      "learning_rate": 2e-05,
+      "loss": 2.2209,
+      "step": 35493
+    },
+    {
+      "epoch": 0.8448538512805865,
+      "grad_norm": 2.4835343406875094,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 35494
+    },
+    {
+      "epoch": 0.844877654003618,
+      "grad_norm": 0.8174073572985084,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 35495
+    },
+    {
+      "epoch": 0.8449014567266495,
+      "grad_norm": 4.671777085380757,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 35496
+    },
+    {
+      "epoch": 0.844925259449681,
+      "grad_norm": 261.10687855267076,
+      "learning_rate": 2e-05,
+      "loss": 0.742,
+      "step": 35497
+    },
+    {
+      "epoch": 0.8449490621727126,
+      "grad_norm": 0.82128454389073,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 35498
+    },
+    {
+      "epoch": 0.8449728648957441,
+      "grad_norm": 41.78708059680778,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 35499
+    },
+    {
+      "epoch": 0.8449966676187756,
+      "grad_norm": 28.059492820264857,
+      "learning_rate": 2e-05,
+      "loss": 1.1466,
+      "step": 35500
+    },
+    {
+      "epoch": 0.845020470341807,
+      "grad_norm": 190.33168831894582,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 35501
+    },
+    {
+      "epoch": 0.8450442730648386,
+      "grad_norm": 362.3443599070432,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 35502
+    },
+    {
+      "epoch": 0.8450680757878701,
+      "grad_norm": 2391.9012061828475,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 35503
+    },
+    {
+      "epoch": 0.8450918785109016,
+      "grad_norm": 647.5419202410436,
+      "learning_rate": 2e-05,
+      "loss": 0.5498,
+      "step": 35504
+    },
+    {
+      "epoch": 0.8451156812339332,
+      "grad_norm": 3369.680490097395,
+      "learning_rate": 2e-05,
+      "loss": 2.3678,
+      "step": 35505
+    },
+    {
+      "epoch": 0.8451394839569647,
+      "grad_norm": 165.86384440652358,
+      "learning_rate": 2e-05,
+      "loss": 1.7987,
+      "step": 35506
+    },
+    {
+      "epoch": 0.8451632866799962,
+      "grad_norm": 12.933104440907268,
+      "learning_rate": 2e-05,
+      "loss": 0.1781,
+      "step": 35507
+    },
+    {
+      "epoch": 0.8451870894030277,
+      "grad_norm": 4219.098140415834,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 35508
+    },
+    {
+      "epoch": 0.8452108921260593,
+      "grad_norm": 112.41502018191545,
+      "learning_rate": 2e-05,
+      "loss": 0.2344,
+      "step": 35509
+    },
+    {
+      "epoch": 0.8452346948490908,
+      "grad_norm": 96.71887428121637,
+      "learning_rate": 2e-05,
+      "loss": 2.3,
+      "step": 35510
+    },
+    {
+      "epoch": 0.8452584975721222,
+      "grad_norm": 29.8163541497544,
+      "learning_rate": 2e-05,
+      "loss": 0.2858,
+      "step": 35511
+    },
+    {
+      "epoch": 0.8452823002951537,
+      "grad_norm": 7.782740076995675,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 35512
+    },
+    {
+      "epoch": 0.8453061030181853,
+      "grad_norm": 1059.5074402202692,
+      "learning_rate": 2e-05,
+      "loss": 1.4536,
+      "step": 35513
+    },
+    {
+      "epoch": 0.8453299057412168,
+      "grad_norm": 461.94375892189424,
+      "learning_rate": 2e-05,
+      "loss": 0.1245,
+      "step": 35514
+    },
+    {
+      "epoch": 0.8453537084642483,
+      "grad_norm": 21.314894962259057,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 35515
+    },
+    {
+      "epoch": 0.8453775111872798,
+      "grad_norm": 69.13218134370014,
+      "learning_rate": 2e-05,
+      "loss": 0.1181,
+      "step": 35516
+    },
+    {
+      "epoch": 0.8454013139103114,
+      "grad_norm": 357.6689887928119,
+      "learning_rate": 2e-05,
+      "loss": 2.2226,
+      "step": 35517
+    },
+    {
+      "epoch": 0.8454251166333429,
+      "grad_norm": 113.54061180728678,
+      "learning_rate": 2e-05,
+      "loss": 0.5888,
+      "step": 35518
+    },
+    {
+      "epoch": 0.8454489193563743,
+      "grad_norm": 148.3640090807332,
+      "learning_rate": 2e-05,
+      "loss": 0.8506,
+      "step": 35519
+    },
+    {
+      "epoch": 0.8454727220794059,
+      "grad_norm": 306.9727186633615,
+      "learning_rate": 2e-05,
+      "loss": 0.513,
+      "step": 35520
+    },
+    {
+      "epoch": 0.8454965248024374,
+      "grad_norm": 13.63976898831346,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 35521
+    },
+    {
+      "epoch": 0.8455203275254689,
+      "grad_norm": 149.97636632159737,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 35522
+    },
+    {
+      "epoch": 0.8455441302485004,
+      "grad_norm": 68.10254582385168,
+      "learning_rate": 2e-05,
+      "loss": 1.2917,
+      "step": 35523
+    },
+    {
+      "epoch": 0.845567932971532,
+      "grad_norm": 6.872374508460735,
+      "learning_rate": 2e-05,
+      "loss": 0.723,
+      "step": 35524
+    },
+    {
+      "epoch": 0.8455917356945635,
+      "grad_norm": 4474.791587731165,
+      "learning_rate": 2e-05,
+      "loss": 0.6452,
+      "step": 35525
+    },
+    {
+      "epoch": 0.845615538417595,
+      "grad_norm": 1850.7569440665357,
+      "learning_rate": 2e-05,
+      "loss": 0.5464,
+      "step": 35526
+    },
+    {
+      "epoch": 0.8456393411406264,
+      "grad_norm": 69.99512139127154,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 35527
+    },
+    {
+      "epoch": 0.845663143863658,
+      "grad_norm": 2064.7495760584607,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 35528
+    },
+    {
+      "epoch": 0.8456869465866895,
+      "grad_norm": 50.92024268850606,
+      "learning_rate": 2e-05,
+      "loss": 1.0526,
+      "step": 35529
+    },
+    {
+      "epoch": 0.845710749309721,
+      "grad_norm": 781.2075754610387,
+      "learning_rate": 2e-05,
+      "loss": 0.5722,
+      "step": 35530
+    },
+    {
+      "epoch": 0.8457345520327525,
+      "grad_norm": 207.89270524703662,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 35531
+    },
+    {
+      "epoch": 0.8457583547557841,
+      "grad_norm": 2443.868733725957,
+      "learning_rate": 2e-05,
+      "loss": 0.8615,
+      "step": 35532
+    },
+    {
+      "epoch": 0.8457821574788156,
+      "grad_norm": 299.2601772452288,
+      "learning_rate": 2e-05,
+      "loss": 0.2075,
+      "step": 35533
+    },
+    {
+      "epoch": 0.8458059602018471,
+      "grad_norm": 585.221747722083,
+      "learning_rate": 2e-05,
+      "loss": 0.2207,
+      "step": 35534
+    },
+    {
+      "epoch": 0.8458297629248787,
+      "grad_norm": 827.579476076395,
+      "learning_rate": 2e-05,
+      "loss": 0.1726,
+      "step": 35535
+    },
+    {
+      "epoch": 0.8458535656479101,
+      "grad_norm": 34.507705879898246,
+      "learning_rate": 2e-05,
+      "loss": 1.0392,
+      "step": 35536
+    },
+    {
+      "epoch": 0.8458773683709416,
+      "grad_norm": 22.897548672922188,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 35537
+    },
+    {
+      "epoch": 0.8459011710939731,
+      "grad_norm": 4.776187971371472,
+      "learning_rate": 2e-05,
+      "loss": 0.6896,
+      "step": 35538
+    },
+    {
+      "epoch": 0.8459249738170047,
+      "grad_norm": 15.648459345369591,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 35539
+    },
+    {
+      "epoch": 0.8459487765400362,
+      "grad_norm": 110.27316064405136,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 35540
+    },
+    {
+      "epoch": 0.8459725792630677,
+      "grad_norm": 1727.4634903646736,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 35541
+    },
+    {
+      "epoch": 0.8459963819860992,
+      "grad_norm": 18.371899939385756,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 35542
+    },
+    {
+      "epoch": 0.8460201847091308,
+      "grad_norm": 6.803863040038823,
+      "learning_rate": 2e-05,
+      "loss": 0.9709,
+      "step": 35543
+    },
+    {
+      "epoch": 0.8460439874321622,
+      "grad_norm": 138.7197207047804,
+      "learning_rate": 2e-05,
+      "loss": 0.8629,
+      "step": 35544
+    },
+    {
+      "epoch": 0.8460677901551937,
+      "grad_norm": 22.014742055883094,
+      "learning_rate": 2e-05,
+      "loss": 0.6606,
+      "step": 35545
+    },
+    {
+      "epoch": 0.8460915928782252,
+      "grad_norm": 966.5369808996552,
+      "learning_rate": 2e-05,
+      "loss": 0.929,
+      "step": 35546
+    },
+    {
+      "epoch": 0.8461153956012568,
+      "grad_norm": 1492.5119266550603,
+      "learning_rate": 2e-05,
+      "loss": 0.5716,
+      "step": 35547
+    },
+    {
+      "epoch": 0.8461391983242883,
+      "grad_norm": 1688.0974740129636,
+      "learning_rate": 2e-05,
+      "loss": 1.4544,
+      "step": 35548
+    },
+    {
+      "epoch": 0.8461630010473198,
+      "grad_norm": 12.578389399801525,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 35549
+    },
+    {
+      "epoch": 0.8461868037703514,
+      "grad_norm": 1130.8503512134964,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 35550
+    },
+    {
+      "epoch": 0.8462106064933829,
+      "grad_norm": 92.39988318280334,
+      "learning_rate": 2e-05,
+      "loss": 0.1462,
+      "step": 35551
+    },
+    {
+      "epoch": 0.8462344092164144,
+      "grad_norm": 7.004703793885525,
+      "learning_rate": 2e-05,
+      "loss": 0.3508,
+      "step": 35552
+    },
+    {
+      "epoch": 0.8462582119394458,
+      "grad_norm": 1052.1433374860792,
+      "learning_rate": 2e-05,
+      "loss": 0.8645,
+      "step": 35553
+    },
+    {
+      "epoch": 0.8462820146624774,
+      "grad_norm": 347.66055111843286,
+      "learning_rate": 2e-05,
+      "loss": 0.1897,
+      "step": 35554
+    },
+    {
+      "epoch": 0.8463058173855089,
+      "grad_norm": 27.292306413068278,
+      "learning_rate": 2e-05,
+      "loss": 0.4644,
+      "step": 35555
+    },
+    {
+      "epoch": 0.8463296201085404,
+      "grad_norm": 5.091695909062158,
+      "learning_rate": 2e-05,
+      "loss": 0.1397,
+      "step": 35556
+    },
+    {
+      "epoch": 0.8463534228315719,
+      "grad_norm": 32.265812926542075,
+      "learning_rate": 2e-05,
+      "loss": 0.2707,
+      "step": 35557
+    },
+    {
+      "epoch": 0.8463772255546035,
+      "grad_norm": 1238.9943960229461,
+      "learning_rate": 2e-05,
+      "loss": 1.69,
+      "step": 35558
+    },
+    {
+      "epoch": 0.846401028277635,
+      "grad_norm": 332.3322886128656,
+      "learning_rate": 2e-05,
+      "loss": 1.8608,
+      "step": 35559
+    },
+    {
+      "epoch": 0.8464248310006665,
+      "grad_norm": 142.08426493939817,
+      "learning_rate": 2e-05,
+      "loss": 2.1014,
+      "step": 35560
+    },
+    {
+      "epoch": 0.8464486337236979,
+      "grad_norm": 7.652735147951967,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 35561
+    },
+    {
+      "epoch": 0.8464724364467295,
+      "grad_norm": 18.931230360514856,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 35562
+    },
+    {
+      "epoch": 0.846496239169761,
+      "grad_norm": 82.91931961987152,
+      "learning_rate": 2e-05,
+      "loss": 0.3323,
+      "step": 35563
+    },
+    {
+      "epoch": 0.8465200418927925,
+      "grad_norm": 158.27264772398868,
+      "learning_rate": 2e-05,
+      "loss": 0.5435,
+      "step": 35564
+    },
+    {
+      "epoch": 0.8465438446158241,
+      "grad_norm": 191.43855810977888,
+      "learning_rate": 2e-05,
+      "loss": 1.5655,
+      "step": 35565
+    },
+    {
+      "epoch": 0.8465676473388556,
+      "grad_norm": 68.02956332200787,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 35566
+    },
+    {
+      "epoch": 0.8465914500618871,
+      "grad_norm": 65.73144687854457,
+      "learning_rate": 2e-05,
+      "loss": 0.312,
+      "step": 35567
+    },
+    {
+      "epoch": 0.8466152527849186,
+      "grad_norm": 498.8010645396273,
+      "learning_rate": 2e-05,
+      "loss": 0.2395,
+      "step": 35568
+    },
+    {
+      "epoch": 0.8466390555079502,
+      "grad_norm": 353.99321591503036,
+      "learning_rate": 2e-05,
+      "loss": 1.1728,
+      "step": 35569
+    },
+    {
+      "epoch": 0.8466628582309816,
+      "grad_norm": 188.63993375451813,
+      "learning_rate": 2e-05,
+      "loss": 0.4425,
+      "step": 35570
+    },
+    {
+      "epoch": 0.8466866609540131,
+      "grad_norm": 559.4047069934935,
+      "learning_rate": 2e-05,
+      "loss": 1.6482,
+      "step": 35571
+    },
+    {
+      "epoch": 0.8467104636770446,
+      "grad_norm": 4395.935886854347,
+      "learning_rate": 2e-05,
+      "loss": 2.5609,
+      "step": 35572
+    },
+    {
+      "epoch": 0.8467342664000762,
+      "grad_norm": 341.15339925364566,
+      "learning_rate": 2e-05,
+      "loss": 2.1862,
+      "step": 35573
+    },
+    {
+      "epoch": 0.8467580691231077,
+      "grad_norm": 2691.420526070473,
+      "learning_rate": 2e-05,
+      "loss": 0.9471,
+      "step": 35574
+    },
+    {
+      "epoch": 0.8467818718461392,
+      "grad_norm": 39.752207507101566,
+      "learning_rate": 2e-05,
+      "loss": 2.0596,
+      "step": 35575
+    },
+    {
+      "epoch": 0.8468056745691707,
+      "grad_norm": 622.1252337085037,
+      "learning_rate": 2e-05,
+      "loss": 3.6907,
+      "step": 35576
+    },
+    {
+      "epoch": 0.8468294772922023,
+      "grad_norm": 720.2683022182995,
+      "learning_rate": 2e-05,
+      "loss": 3.3203,
+      "step": 35577
+    },
+    {
+      "epoch": 0.8468532800152337,
+      "grad_norm": 42.88095741450501,
+      "learning_rate": 2e-05,
+      "loss": 3.6495,
+      "step": 35578
+    },
+    {
+      "epoch": 0.8468770827382652,
+      "grad_norm": 246.92139424034085,
+      "learning_rate": 2e-05,
+      "loss": 2.1794,
+      "step": 35579
+    },
+    {
+      "epoch": 0.8469008854612968,
+      "grad_norm": 853.6156768733031,
+      "learning_rate": 2e-05,
+      "loss": 6.7888,
+      "step": 35580
+    },
+    {
+      "epoch": 0.8469246881843283,
+      "grad_norm": 745.7985809059169,
+      "learning_rate": 2e-05,
+      "loss": 3.4946,
+      "step": 35581
+    },
+    {
+      "epoch": 0.8469484909073598,
+      "grad_norm": 753.8932458127136,
+      "learning_rate": 2e-05,
+      "loss": 4.4561,
+      "step": 35582
+    },
+    {
+      "epoch": 0.8469722936303913,
+      "grad_norm": 1459.548175858215,
+      "learning_rate": 2e-05,
+      "loss": 4.4622,
+      "step": 35583
+    },
+    {
+      "epoch": 0.8469960963534229,
+      "grad_norm": 77.82080222066287,
+      "learning_rate": 2e-05,
+      "loss": 2.7396,
+      "step": 35584
+    },
+    {
+      "epoch": 0.8470198990764544,
+      "grad_norm": 105.46149408519368,
+      "learning_rate": 2e-05,
+      "loss": 1.0628,
+      "step": 35585
+    },
+    {
+      "epoch": 0.8470437017994858,
+      "grad_norm": 92.5499862149275,
+      "learning_rate": 2e-05,
+      "loss": 2.5081,
+      "step": 35586
+    },
+    {
+      "epoch": 0.8470675045225173,
+      "grad_norm": 46.35415556430697,
+      "learning_rate": 2e-05,
+      "loss": 1.1746,
+      "step": 35587
+    },
+    {
+      "epoch": 0.8470913072455489,
+      "grad_norm": 545.0257223375736,
+      "learning_rate": 2e-05,
+      "loss": 3.4203,
+      "step": 35588
+    },
+    {
+      "epoch": 0.8471151099685804,
+      "grad_norm": 446.69018891396337,
+      "learning_rate": 2e-05,
+      "loss": 1.6128,
+      "step": 35589
+    },
+    {
+      "epoch": 0.8471389126916119,
+      "grad_norm": 146.37426876396236,
+      "learning_rate": 2e-05,
+      "loss": 1.9094,
+      "step": 35590
+    },
+    {
+      "epoch": 0.8471627154146434,
+      "grad_norm": 420.15680276162476,
+      "learning_rate": 2e-05,
+      "loss": 0.6941,
+      "step": 35591
+    },
+    {
+      "epoch": 0.847186518137675,
+      "grad_norm": 349.2096315295971,
+      "learning_rate": 2e-05,
+      "loss": 0.9251,
+      "step": 35592
+    },
+    {
+      "epoch": 0.8472103208607065,
+      "grad_norm": 340.8164545616075,
+      "learning_rate": 2e-05,
+      "loss": 0.6034,
+      "step": 35593
+    },
+    {
+      "epoch": 0.847234123583738,
+      "grad_norm": 26.35693941662577,
+      "learning_rate": 2e-05,
+      "loss": 0.4539,
+      "step": 35594
+    },
+    {
+      "epoch": 0.8472579263067695,
+      "grad_norm": 193.89305184553382,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 35595
+    },
+    {
+      "epoch": 0.847281729029801,
+      "grad_norm": 125.66877383693388,
+      "learning_rate": 2e-05,
+      "loss": 0.3576,
+      "step": 35596
+    },
+    {
+      "epoch": 0.8473055317528325,
+      "grad_norm": 87.91807811114577,
+      "learning_rate": 2e-05,
+      "loss": 1.4352,
+      "step": 35597
+    },
+    {
+      "epoch": 0.847329334475864,
+      "grad_norm": 59.9499353627725,
+      "learning_rate": 2e-05,
+      "loss": 1.4803,
+      "step": 35598
+    },
+    {
+      "epoch": 0.8473531371988956,
+      "grad_norm": 58.683150946514445,
+      "learning_rate": 2e-05,
+      "loss": 0.5897,
+      "step": 35599
+    },
+    {
+      "epoch": 0.8473769399219271,
+      "grad_norm": 52.069025744557436,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 35600
+    },
+    {
+      "epoch": 0.8474007426449586,
+      "grad_norm": 80.27942940373156,
+      "learning_rate": 2e-05,
+      "loss": 0.5759,
+      "step": 35601
+    },
+    {
+      "epoch": 0.8474245453679901,
+      "grad_norm": 635.6496023427948,
+      "learning_rate": 2e-05,
+      "loss": 0.2767,
+      "step": 35602
+    },
+    {
+      "epoch": 0.8474483480910217,
+      "grad_norm": 257.6829802481284,
+      "learning_rate": 2e-05,
+      "loss": 1.6792,
+      "step": 35603
+    },
+    {
+      "epoch": 0.8474721508140531,
+      "grad_norm": 656.7848457070662,
+      "learning_rate": 2e-05,
+      "loss": 1.0775,
+      "step": 35604
+    },
+    {
+      "epoch": 0.8474959535370846,
+      "grad_norm": 25.525885283233055,
+      "learning_rate": 2e-05,
+      "loss": 1.7623,
+      "step": 35605
+    },
+    {
+      "epoch": 0.8475197562601161,
+      "grad_norm": 61.10422887207875,
+      "learning_rate": 2e-05,
+      "loss": 1.4187,
+      "step": 35606
+    },
+    {
+      "epoch": 0.8475435589831477,
+      "grad_norm": 38.51745708233571,
+      "learning_rate": 2e-05,
+      "loss": 1.2906,
+      "step": 35607
+    },
+    {
+      "epoch": 0.8475673617061792,
+      "grad_norm": 30.4317186982014,
+      "learning_rate": 2e-05,
+      "loss": 1.2434,
+      "step": 35608
+    },
+    {
+      "epoch": 0.8475911644292107,
+      "grad_norm": 106.3723098219779,
+      "learning_rate": 2e-05,
+      "loss": 1.8513,
+      "step": 35609
+    },
+    {
+      "epoch": 0.8476149671522423,
+      "grad_norm": 479.1656198498968,
+      "learning_rate": 2e-05,
+      "loss": 1.0663,
+      "step": 35610
+    },
+    {
+      "epoch": 0.8476387698752738,
+      "grad_norm": 1087.8061982266865,
+      "learning_rate": 2e-05,
+      "loss": 2.4353,
+      "step": 35611
+    },
+    {
+      "epoch": 0.8476625725983052,
+      "grad_norm": 62.797300528457946,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 35612
+    },
+    {
+      "epoch": 0.8476863753213367,
+      "grad_norm": 77.95983572068836,
+      "learning_rate": 2e-05,
+      "loss": 2.1208,
+      "step": 35613
+    },
+    {
+      "epoch": 0.8477101780443683,
+      "grad_norm": 24.799330685582976,
+      "learning_rate": 2e-05,
+      "loss": 1.013,
+      "step": 35614
+    },
+    {
+      "epoch": 0.8477339807673998,
+      "grad_norm": 1362.1726013369919,
+      "learning_rate": 2e-05,
+      "loss": 1.7051,
+      "step": 35615
+    },
+    {
+      "epoch": 0.8477577834904313,
+      "grad_norm": 13.676424426685061,
+      "learning_rate": 2e-05,
+      "loss": 0.8751,
+      "step": 35616
+    },
+    {
+      "epoch": 0.8477815862134628,
+      "grad_norm": 89.01385316762666,
+      "learning_rate": 2e-05,
+      "loss": 1.2522,
+      "step": 35617
+    },
+    {
+      "epoch": 0.8478053889364944,
+      "grad_norm": 62.41452853339475,
+      "learning_rate": 2e-05,
+      "loss": 1.7345,
+      "step": 35618
+    },
+    {
+      "epoch": 0.8478291916595259,
+      "grad_norm": 132.5928102607787,
+      "learning_rate": 2e-05,
+      "loss": 1.995,
+      "step": 35619
+    },
+    {
+      "epoch": 0.8478529943825573,
+      "grad_norm": 95.34185707677791,
+      "learning_rate": 2e-05,
+      "loss": 1.9119,
+      "step": 35620
+    },
+    {
+      "epoch": 0.8478767971055888,
+      "grad_norm": 464.7074419454092,
+      "learning_rate": 2e-05,
+      "loss": 1.3256,
+      "step": 35621
+    },
+    {
+      "epoch": 0.8479005998286204,
+      "grad_norm": 45.84371736158324,
+      "learning_rate": 2e-05,
+      "loss": 0.7971,
+      "step": 35622
+    },
+    {
+      "epoch": 0.8479244025516519,
+      "grad_norm": 10.990374485512572,
+      "learning_rate": 2e-05,
+      "loss": 0.5304,
+      "step": 35623
+    },
+    {
+      "epoch": 0.8479482052746834,
+      "grad_norm": 135.1931627361323,
+      "learning_rate": 2e-05,
+      "loss": 0.3502,
+      "step": 35624
+    },
+    {
+      "epoch": 0.847972007997715,
+      "grad_norm": 23.999387933795685,
+      "learning_rate": 2e-05,
+      "loss": 0.8786,
+      "step": 35625
+    },
+    {
+      "epoch": 0.8479958107207465,
+      "grad_norm": 28.47804260278079,
+      "learning_rate": 2e-05,
+      "loss": 0.5016,
+      "step": 35626
+    },
+    {
+      "epoch": 0.848019613443778,
+      "grad_norm": 8.580012812468881,
+      "learning_rate": 2e-05,
+      "loss": 0.2606,
+      "step": 35627
+    },
+    {
+      "epoch": 0.8480434161668094,
+      "grad_norm": 14.76346710869836,
+      "learning_rate": 2e-05,
+      "loss": 0.5317,
+      "step": 35628
+    },
+    {
+      "epoch": 0.848067218889841,
+      "grad_norm": 11.652933183268182,
+      "learning_rate": 2e-05,
+      "loss": 0.5784,
+      "step": 35629
+    },
+    {
+      "epoch": 0.8480910216128725,
+      "grad_norm": 12.873162116218525,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 35630
+    },
+    {
+      "epoch": 0.848114824335904,
+      "grad_norm": 3.7379065081119243,
+      "learning_rate": 2e-05,
+      "loss": 0.3841,
+      "step": 35631
+    },
+    {
+      "epoch": 0.8481386270589355,
+      "grad_norm": 10.251434555896472,
+      "learning_rate": 2e-05,
+      "loss": 0.5835,
+      "step": 35632
+    },
+    {
+      "epoch": 0.8481624297819671,
+      "grad_norm": 12.141986737866517,
+      "learning_rate": 2e-05,
+      "loss": 0.599,
+      "step": 35633
+    },
+    {
+      "epoch": 0.8481862325049986,
+      "grad_norm": 20.539564843112608,
+      "learning_rate": 2e-05,
+      "loss": 0.398,
+      "step": 35634
+    },
+    {
+      "epoch": 0.8482100352280301,
+      "grad_norm": 26.075851683113523,
+      "learning_rate": 2e-05,
+      "loss": 0.6826,
+      "step": 35635
+    },
+    {
+      "epoch": 0.8482338379510616,
+      "grad_norm": 11.092049343744417,
+      "learning_rate": 2e-05,
+      "loss": 1.058,
+      "step": 35636
+    },
+    {
+      "epoch": 0.8482576406740931,
+      "grad_norm": 39.961745189480894,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 35637
+    },
+    {
+      "epoch": 0.8482814433971246,
+      "grad_norm": 7.013492386538715,
+      "learning_rate": 2e-05,
+      "loss": 0.4147,
+      "step": 35638
+    },
+    {
+      "epoch": 0.8483052461201561,
+      "grad_norm": 6.39436836387652,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 35639
+    },
+    {
+      "epoch": 0.8483290488431876,
+      "grad_norm": 9.95627394738687,
+      "learning_rate": 2e-05,
+      "loss": 0.3022,
+      "step": 35640
+    },
+    {
+      "epoch": 0.8483528515662192,
+      "grad_norm": 20.682825777872058,
+      "learning_rate": 2e-05,
+      "loss": 0.1935,
+      "step": 35641
+    },
+    {
+      "epoch": 0.8483766542892507,
+      "grad_norm": 10.931466267836765,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 35642
+    },
+    {
+      "epoch": 0.8484004570122822,
+      "grad_norm": 17.472024645273997,
+      "learning_rate": 2e-05,
+      "loss": 0.5421,
+      "step": 35643
+    },
+    {
+      "epoch": 0.8484242597353138,
+      "grad_norm": 21.30442221351765,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 35644
+    },
+    {
+      "epoch": 0.8484480624583453,
+      "grad_norm": 11.452763230545449,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 35645
+    },
+    {
+      "epoch": 0.8484718651813767,
+      "grad_norm": 10.177154284683288,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 35646
+    },
+    {
+      "epoch": 0.8484956679044082,
+      "grad_norm": 3.4683879196545786,
+      "learning_rate": 2e-05,
+      "loss": 0.297,
+      "step": 35647
+    },
+    {
+      "epoch": 0.8485194706274398,
+      "grad_norm": 10.496795625442761,
+      "learning_rate": 2e-05,
+      "loss": 0.0532,
+      "step": 35648
+    },
+    {
+      "epoch": 0.8485432733504713,
+      "grad_norm": 12.454484956046787,
+      "learning_rate": 2e-05,
+      "loss": 1.3003,
+      "step": 35649
+    },
+    {
+      "epoch": 0.8485670760735028,
+      "grad_norm": 2.1984129546644593,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 35650
+    },
+    {
+      "epoch": 0.8485908787965343,
+      "grad_norm": 3.941378063231446,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 35651
+    },
+    {
+      "epoch": 0.8486146815195659,
+      "grad_norm": 7.763445338412707,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 35652
+    },
+    {
+      "epoch": 0.8486384842425974,
+      "grad_norm": 5.245896858885037,
+      "learning_rate": 2e-05,
+      "loss": 0.8744,
+      "step": 35653
+    },
+    {
+      "epoch": 0.8486622869656288,
+      "grad_norm": 7.244339289522278,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 35654
+    },
+    {
+      "epoch": 0.8486860896886603,
+      "grad_norm": 5.771655632993185,
+      "learning_rate": 2e-05,
+      "loss": 0.4295,
+      "step": 35655
+    },
+    {
+      "epoch": 0.8487098924116919,
+      "grad_norm": 16.568385454654205,
+      "learning_rate": 2e-05,
+      "loss": 0.2315,
+      "step": 35656
+    },
+    {
+      "epoch": 0.8487336951347234,
+      "grad_norm": 27.806164757427208,
+      "learning_rate": 2e-05,
+      "loss": 0.8754,
+      "step": 35657
+    },
+    {
+      "epoch": 0.8487574978577549,
+      "grad_norm": 4.610869028356837,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 35658
+    },
+    {
+      "epoch": 0.8487813005807865,
+      "grad_norm": 3.5479302789496656,
+      "learning_rate": 2e-05,
+      "loss": 0.546,
+      "step": 35659
+    },
+    {
+      "epoch": 0.848805103303818,
+      "grad_norm": 8.680373241459447,
+      "learning_rate": 2e-05,
+      "loss": 0.2689,
+      "step": 35660
+    },
+    {
+      "epoch": 0.8488289060268495,
+      "grad_norm": 15.316767612982533,
+      "learning_rate": 2e-05,
+      "loss": 0.7682,
+      "step": 35661
+    },
+    {
+      "epoch": 0.848852708749881,
+      "grad_norm": 8.73341954805439,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 35662
+    },
+    {
+      "epoch": 0.8488765114729125,
+      "grad_norm": 3.015597152582503,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 35663
+    },
+    {
+      "epoch": 0.848900314195944,
+      "grad_norm": 7.3938846138801,
+      "learning_rate": 2e-05,
+      "loss": 0.9094,
+      "step": 35664
+    },
+    {
+      "epoch": 0.8489241169189755,
+      "grad_norm": 1.262209274896943,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 35665
+    },
+    {
+      "epoch": 0.848947919642007,
+      "grad_norm": 6.01096847435334,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 35666
+    },
+    {
+      "epoch": 0.8489717223650386,
+      "grad_norm": 10.027025059599644,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 35667
+    },
+    {
+      "epoch": 0.8489955250880701,
+      "grad_norm": 77.17367837793236,
+      "learning_rate": 2e-05,
+      "loss": 1.055,
+      "step": 35668
+    },
+    {
+      "epoch": 0.8490193278111016,
+      "grad_norm": 4.693478581332965,
+      "learning_rate": 2e-05,
+      "loss": 0.2984,
+      "step": 35669
+    },
+    {
+      "epoch": 0.849043130534133,
+      "grad_norm": 4.905593287015947,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 35670
+    },
+    {
+      "epoch": 0.8490669332571646,
+      "grad_norm": 10.803599657573587,
+      "learning_rate": 2e-05,
+      "loss": 0.538,
+      "step": 35671
+    },
+    {
+      "epoch": 0.8490907359801961,
+      "grad_norm": 3.3260537175274414,
+      "learning_rate": 2e-05,
+      "loss": 0.8084,
+      "step": 35672
+    },
+    {
+      "epoch": 0.8491145387032276,
+      "grad_norm": 4.634897088774519,
+      "learning_rate": 2e-05,
+      "loss": 0.6892,
+      "step": 35673
+    },
+    {
+      "epoch": 0.8491383414262592,
+      "grad_norm": 1.6760328763242796,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 35674
+    },
+    {
+      "epoch": 0.8491621441492907,
+      "grad_norm": 0.5856047234086308,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 35675
+    },
+    {
+      "epoch": 0.8491859468723222,
+      "grad_norm": 10.549471767893442,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 35676
+    },
+    {
+      "epoch": 0.8492097495953537,
+      "grad_norm": 3.704829294219346,
+      "learning_rate": 2e-05,
+      "loss": 0.6085,
+      "step": 35677
+    },
+    {
+      "epoch": 0.8492335523183853,
+      "grad_norm": 2.078536060057521,
+      "learning_rate": 2e-05,
+      "loss": 0.156,
+      "step": 35678
+    },
+    {
+      "epoch": 0.8492573550414168,
+      "grad_norm": 187.03880901563465,
+      "learning_rate": 2e-05,
+      "loss": 0.2515,
+      "step": 35679
+    },
+    {
+      "epoch": 0.8492811577644482,
+      "grad_norm": 1.3286473829652825,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 35680
+    },
+    {
+      "epoch": 0.8493049604874797,
+      "grad_norm": 6.993105368836919,
+      "learning_rate": 2e-05,
+      "loss": 1.1359,
+      "step": 35681
+    },
+    {
+      "epoch": 0.8493287632105113,
+      "grad_norm": 6.084545879700891,
+      "learning_rate": 2e-05,
+      "loss": 1.7559,
+      "step": 35682
+    },
+    {
+      "epoch": 0.8493525659335428,
+      "grad_norm": 3.9237743746748586,
+      "learning_rate": 2e-05,
+      "loss": 0.2348,
+      "step": 35683
+    },
+    {
+      "epoch": 0.8493763686565743,
+      "grad_norm": 2.7154793557792662,
+      "learning_rate": 2e-05,
+      "loss": 0.3514,
+      "step": 35684
+    },
+    {
+      "epoch": 0.8494001713796058,
+      "grad_norm": 6.534064610695676,
+      "learning_rate": 2e-05,
+      "loss": 0.2479,
+      "step": 35685
+    },
+    {
+      "epoch": 0.8494239741026374,
+      "grad_norm": 0.608797233082344,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 35686
+    },
+    {
+      "epoch": 0.8494477768256689,
+      "grad_norm": 0.1395431263432334,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 35687
+    },
+    {
+      "epoch": 0.8494715795487003,
+      "grad_norm": 9.761971079430463,
+      "learning_rate": 2e-05,
+      "loss": 0.6149,
+      "step": 35688
+    },
+    {
+      "epoch": 0.8494953822717319,
+      "grad_norm": 6.5984844235528035,
+      "learning_rate": 2e-05,
+      "loss": 0.7255,
+      "step": 35689
+    },
+    {
+      "epoch": 0.8495191849947634,
+      "grad_norm": 12.050899887766297,
+      "learning_rate": 2e-05,
+      "loss": 0.6921,
+      "step": 35690
+    },
+    {
+      "epoch": 0.8495429877177949,
+      "grad_norm": 7.945746528058154,
+      "learning_rate": 2e-05,
+      "loss": 0.6077,
+      "step": 35691
+    },
+    {
+      "epoch": 0.8495667904408264,
+      "grad_norm": 7.970787930011744,
+      "learning_rate": 2e-05,
+      "loss": 0.622,
+      "step": 35692
+    },
+    {
+      "epoch": 0.849590593163858,
+      "grad_norm": 4.13166533667524,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 35693
+    },
+    {
+      "epoch": 0.8496143958868895,
+      "grad_norm": 20.108605068949803,
+      "learning_rate": 2e-05,
+      "loss": 0.8944,
+      "step": 35694
+    },
+    {
+      "epoch": 0.849638198609921,
+      "grad_norm": 10.488644894053328,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 35695
+    },
+    {
+      "epoch": 0.8496620013329524,
+      "grad_norm": 0.433737634089241,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 35696
+    },
+    {
+      "epoch": 0.849685804055984,
+      "grad_norm": 7.746415837919186,
+      "learning_rate": 2e-05,
+      "loss": 0.7924,
+      "step": 35697
+    },
+    {
+      "epoch": 0.8497096067790155,
+      "grad_norm": 22.385583249179486,
+      "learning_rate": 2e-05,
+      "loss": 0.3721,
+      "step": 35698
+    },
+    {
+      "epoch": 0.849733409502047,
+      "grad_norm": 2.7131155626554166,
+      "learning_rate": 2e-05,
+      "loss": 0.5109,
+      "step": 35699
+    },
+    {
+      "epoch": 0.8497572122250785,
+      "grad_norm": 5.667585647464316,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 35700
+    },
+    {
+      "epoch": 0.8497810149481101,
+      "grad_norm": 4.550667535431153,
+      "learning_rate": 2e-05,
+      "loss": 0.3095,
+      "step": 35701
+    },
+    {
+      "epoch": 0.8498048176711416,
+      "grad_norm": 4.887305762089129,
+      "learning_rate": 2e-05,
+      "loss": 0.7981,
+      "step": 35702
+    },
+    {
+      "epoch": 0.8498286203941731,
+      "grad_norm": 4.864848973914428,
+      "learning_rate": 2e-05,
+      "loss": 0.7378,
+      "step": 35703
+    },
+    {
+      "epoch": 0.8498524231172047,
+      "grad_norm": 41.72351281892166,
+      "learning_rate": 2e-05,
+      "loss": 0.4759,
+      "step": 35704
+    },
+    {
+      "epoch": 0.8498762258402361,
+      "grad_norm": 0.021885096180692568,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 35705
+    },
+    {
+      "epoch": 0.8499000285632676,
+      "grad_norm": 6.265408795833489,
+      "learning_rate": 2e-05,
+      "loss": 0.9725,
+      "step": 35706
+    },
+    {
+      "epoch": 0.8499238312862991,
+      "grad_norm": 6.732690313762484,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 35707
+    },
+    {
+      "epoch": 0.8499476340093307,
+      "grad_norm": 9.301164887100622,
+      "learning_rate": 2e-05,
+      "loss": 0.6889,
+      "step": 35708
+    },
+    {
+      "epoch": 0.8499714367323622,
+      "grad_norm": 3.907902403080544,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 35709
+    },
+    {
+      "epoch": 0.8499952394553937,
+      "grad_norm": 4.722862430732655,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 35710
+    },
+    {
+      "epoch": 0.8500190421784252,
+      "grad_norm": 5.182974596285988,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 35711
+    },
+    {
+      "epoch": 0.8500428449014568,
+      "grad_norm": 4.222900347493974,
+      "learning_rate": 2e-05,
+      "loss": 0.3371,
+      "step": 35712
+    },
+    {
+      "epoch": 0.8500666476244882,
+      "grad_norm": 8.12823205232988,
+      "learning_rate": 2e-05,
+      "loss": 0.2437,
+      "step": 35713
+    },
+    {
+      "epoch": 0.8500904503475197,
+      "grad_norm": 17.332802421249845,
+      "learning_rate": 2e-05,
+      "loss": 0.1795,
+      "step": 35714
+    },
+    {
+      "epoch": 0.8501142530705512,
+      "grad_norm": 5.287774336186828,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 35715
+    },
+    {
+      "epoch": 0.8501380557935828,
+      "grad_norm": 4.151622851276079,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 35716
+    },
+    {
+      "epoch": 0.8501618585166143,
+      "grad_norm": 9.263943830119196,
+      "learning_rate": 2e-05,
+      "loss": 0.8964,
+      "step": 35717
+    },
+    {
+      "epoch": 0.8501856612396458,
+      "grad_norm": 7.986232247446805,
+      "learning_rate": 2e-05,
+      "loss": 0.9669,
+      "step": 35718
+    },
+    {
+      "epoch": 0.8502094639626774,
+      "grad_norm": 2.9587112454907003,
+      "learning_rate": 2e-05,
+      "loss": 0.259,
+      "step": 35719
+    },
+    {
+      "epoch": 0.8502332666857089,
+      "grad_norm": 0.09299728897166633,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 35720
+    },
+    {
+      "epoch": 0.8502570694087404,
+      "grad_norm": 5.513642115143511,
+      "learning_rate": 2e-05,
+      "loss": 0.207,
+      "step": 35721
+    },
+    {
+      "epoch": 0.8502808721317718,
+      "grad_norm": 0.24034419940875332,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 35722
+    },
+    {
+      "epoch": 0.8503046748548034,
+      "grad_norm": 0.0695224450197683,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 35723
+    },
+    {
+      "epoch": 0.8503284775778349,
+      "grad_norm": 24.46639068063607,
+      "learning_rate": 2e-05,
+      "loss": 0.3692,
+      "step": 35724
+    },
+    {
+      "epoch": 0.8503522803008664,
+      "grad_norm": 2.795043786246313,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 35725
+    },
+    {
+      "epoch": 0.8503760830238979,
+      "grad_norm": 4.822496948540524,
+      "learning_rate": 2e-05,
+      "loss": 1.0883,
+      "step": 35726
+    },
+    {
+      "epoch": 0.8503998857469295,
+      "grad_norm": 5.898871846488123,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 35727
+    },
+    {
+      "epoch": 0.850423688469961,
+      "grad_norm": 3.577391900288347,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 35728
+    },
+    {
+      "epoch": 0.8504474911929925,
+      "grad_norm": 5.406931717211712,
+      "learning_rate": 2e-05,
+      "loss": 0.7426,
+      "step": 35729
+    },
+    {
+      "epoch": 0.8504712939160239,
+      "grad_norm": 7.341459855967654,
+      "learning_rate": 2e-05,
+      "loss": 0.899,
+      "step": 35730
+    },
+    {
+      "epoch": 0.8504950966390555,
+      "grad_norm": 7.0697404417113505,
+      "learning_rate": 2e-05,
+      "loss": 0.2563,
+      "step": 35731
+    },
+    {
+      "epoch": 0.850518899362087,
+      "grad_norm": 1.321392896447928,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 35732
+    },
+    {
+      "epoch": 0.8505427020851185,
+      "grad_norm": 3.0646878273726315,
+      "learning_rate": 2e-05,
+      "loss": 0.1156,
+      "step": 35733
+    },
+    {
+      "epoch": 0.8505665048081501,
+      "grad_norm": 5.568914252147857,
+      "learning_rate": 2e-05,
+      "loss": 0.9321,
+      "step": 35734
+    },
+    {
+      "epoch": 0.8505903075311816,
+      "grad_norm": 3.361335310357277,
+      "learning_rate": 2e-05,
+      "loss": 0.0452,
+      "step": 35735
+    },
+    {
+      "epoch": 0.8506141102542131,
+      "grad_norm": 5.4024869161786855,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 35736
+    },
+    {
+      "epoch": 0.8506379129772446,
+      "grad_norm": 8.50054798894736,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 35737
+    },
+    {
+      "epoch": 0.8506617157002762,
+      "grad_norm": 5.092700485530595,
+      "learning_rate": 2e-05,
+      "loss": 0.2554,
+      "step": 35738
+    },
+    {
+      "epoch": 0.8506855184233076,
+      "grad_norm": 4.291051967574991,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 35739
+    },
+    {
+      "epoch": 0.8507093211463391,
+      "grad_norm": 2.5431005501607227,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 35740
+    },
+    {
+      "epoch": 0.8507331238693706,
+      "grad_norm": 5.856524353212216,
+      "learning_rate": 2e-05,
+      "loss": 0.8155,
+      "step": 35741
+    },
+    {
+      "epoch": 0.8507569265924022,
+      "grad_norm": 4.868258061653026,
+      "learning_rate": 2e-05,
+      "loss": 0.9787,
+      "step": 35742
+    },
+    {
+      "epoch": 0.8507807293154337,
+      "grad_norm": 0.15758552185089053,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 35743
+    },
+    {
+      "epoch": 0.8508045320384652,
+      "grad_norm": 6.024597315405381,
+      "learning_rate": 2e-05,
+      "loss": 0.2299,
+      "step": 35744
+    },
+    {
+      "epoch": 0.8508283347614967,
+      "grad_norm": 10.433556290644045,
+      "learning_rate": 2e-05,
+      "loss": 0.4541,
+      "step": 35745
+    },
+    {
+      "epoch": 0.8508521374845283,
+      "grad_norm": 7.311616634264794,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 35746
+    },
+    {
+      "epoch": 0.8508759402075597,
+      "grad_norm": 5.652501546709284,
+      "learning_rate": 2e-05,
+      "loss": 0.2185,
+      "step": 35747
+    },
+    {
+      "epoch": 0.8508997429305912,
+      "grad_norm": 2.951932837452349,
+      "learning_rate": 2e-05,
+      "loss": 0.3145,
+      "step": 35748
+    },
+    {
+      "epoch": 0.8509235456536228,
+      "grad_norm": 3.2172823932287797,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 35749
+    },
+    {
+      "epoch": 0.8509473483766543,
+      "grad_norm": 4.562084132073066,
+      "learning_rate": 2e-05,
+      "loss": 0.1742,
+      "step": 35750
+    },
+    {
+      "epoch": 0.8509711510996858,
+      "grad_norm": 2.7609318682152604,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 35751
+    },
+    {
+      "epoch": 0.8509949538227173,
+      "grad_norm": 5.481586110522438,
+      "learning_rate": 2e-05,
+      "loss": 1.0375,
+      "step": 35752
+    },
+    {
+      "epoch": 0.8510187565457489,
+      "grad_norm": 1.9929287139875471,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 35753
+    },
+    {
+      "epoch": 0.8510425592687804,
+      "grad_norm": 0.9928327219584901,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 35754
+    },
+    {
+      "epoch": 0.8510663619918118,
+      "grad_norm": 7.160167778731434,
+      "learning_rate": 2e-05,
+      "loss": 1.042,
+      "step": 35755
+    },
+    {
+      "epoch": 0.8510901647148433,
+      "grad_norm": 51.475396401163,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 35756
+    },
+    {
+      "epoch": 0.8511139674378749,
+      "grad_norm": 6.611251795186783,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 35757
+    },
+    {
+      "epoch": 0.8511377701609064,
+      "grad_norm": 23.07065821485301,
+      "learning_rate": 2e-05,
+      "loss": 0.5818,
+      "step": 35758
+    },
+    {
+      "epoch": 0.8511615728839379,
+      "grad_norm": 5.517375369842186,
+      "learning_rate": 2e-05,
+      "loss": 1.2721,
+      "step": 35759
+    },
+    {
+      "epoch": 0.8511853756069694,
+      "grad_norm": 5.156755112968738,
+      "learning_rate": 2e-05,
+      "loss": 0.2059,
+      "step": 35760
+    },
+    {
+      "epoch": 0.851209178330001,
+      "grad_norm": 7.545674248708933,
+      "learning_rate": 2e-05,
+      "loss": 0.6619,
+      "step": 35761
+    },
+    {
+      "epoch": 0.8512329810530325,
+      "grad_norm": 8.508569416189514,
+      "learning_rate": 2e-05,
+      "loss": 1.1797,
+      "step": 35762
+    },
+    {
+      "epoch": 0.851256783776064,
+      "grad_norm": 6.473181280854027,
+      "learning_rate": 2e-05,
+      "loss": 0.4673,
+      "step": 35763
+    },
+    {
+      "epoch": 0.8512805864990955,
+      "grad_norm": 22.577431445072342,
+      "learning_rate": 2e-05,
+      "loss": 0.5234,
+      "step": 35764
+    },
+    {
+      "epoch": 0.851304389222127,
+      "grad_norm": 5.714029040755994,
+      "learning_rate": 2e-05,
+      "loss": 0.2838,
+      "step": 35765
+    },
+    {
+      "epoch": 0.8513281919451585,
+      "grad_norm": 0.19569851317036197,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 35766
+    },
+    {
+      "epoch": 0.85135199466819,
+      "grad_norm": 9.359507869946698,
+      "learning_rate": 2e-05,
+      "loss": 0.1407,
+      "step": 35767
+    },
+    {
+      "epoch": 0.8513757973912216,
+      "grad_norm": 2.2840805877693526,
+      "learning_rate": 2e-05,
+      "loss": 0.4087,
+      "step": 35768
+    },
+    {
+      "epoch": 0.8513996001142531,
+      "grad_norm": 0.1301418526257088,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 35769
+    },
+    {
+      "epoch": 0.8514234028372846,
+      "grad_norm": 4.075943983204211,
+      "learning_rate": 2e-05,
+      "loss": 0.4796,
+      "step": 35770
+    },
+    {
+      "epoch": 0.8514472055603161,
+      "grad_norm": 4.76852219340549,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 35771
+    },
+    {
+      "epoch": 0.8514710082833477,
+      "grad_norm": 1.8730301920539276,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 35772
+    },
+    {
+      "epoch": 0.8514948110063791,
+      "grad_norm": 2.190796999594958,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 35773
+    },
+    {
+      "epoch": 0.8515186137294106,
+      "grad_norm": 5.282847605150324,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 35774
+    },
+    {
+      "epoch": 0.8515424164524421,
+      "grad_norm": 4.790925066785325,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 35775
+    },
+    {
+      "epoch": 0.8515662191754737,
+      "grad_norm": 5.361194401744035,
+      "learning_rate": 2e-05,
+      "loss": 0.1433,
+      "step": 35776
+    },
+    {
+      "epoch": 0.8515900218985052,
+      "grad_norm": 3.6202006959575113,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 35777
+    },
+    {
+      "epoch": 0.8516138246215367,
+      "grad_norm": 3.8853335873940527,
+      "learning_rate": 2e-05,
+      "loss": 0.5277,
+      "step": 35778
+    },
+    {
+      "epoch": 0.8516376273445683,
+      "grad_norm": 12.791430048211788,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 35779
+    },
+    {
+      "epoch": 0.8516614300675998,
+      "grad_norm": 4.488615633479516,
+      "learning_rate": 2e-05,
+      "loss": 1.136,
+      "step": 35780
+    },
+    {
+      "epoch": 0.8516852327906312,
+      "grad_norm": 1.8317188303303313,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 35781
+    },
+    {
+      "epoch": 0.8517090355136627,
+      "grad_norm": 8.447061713972793,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 35782
+    },
+    {
+      "epoch": 0.8517328382366943,
+      "grad_norm": 2.1093800625233983,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 35783
+    },
+    {
+      "epoch": 0.8517566409597258,
+      "grad_norm": 5.361092889888861,
+      "learning_rate": 2e-05,
+      "loss": 0.7032,
+      "step": 35784
+    },
+    {
+      "epoch": 0.8517804436827573,
+      "grad_norm": 5.54577893420669,
+      "learning_rate": 2e-05,
+      "loss": 0.7209,
+      "step": 35785
+    },
+    {
+      "epoch": 0.8518042464057888,
+      "grad_norm": 3.716851372838343,
+      "learning_rate": 2e-05,
+      "loss": 0.0995,
+      "step": 35786
+    },
+    {
+      "epoch": 0.8518280491288204,
+      "grad_norm": 0.03654706328151077,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35787
+    },
+    {
+      "epoch": 0.8518518518518519,
+      "grad_norm": 3.698793079595934,
+      "learning_rate": 2e-05,
+      "loss": 0.4417,
+      "step": 35788
+    },
+    {
+      "epoch": 0.8518756545748833,
+      "grad_norm": 4.090754282467082,
+      "learning_rate": 2e-05,
+      "loss": 0.5886,
+      "step": 35789
+    },
+    {
+      "epoch": 0.8518994572979148,
+      "grad_norm": 3.8918190505584005,
+      "learning_rate": 2e-05,
+      "loss": 0.3869,
+      "step": 35790
+    },
+    {
+      "epoch": 0.8519232600209464,
+      "grad_norm": 1.5086246893595199,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 35791
+    },
+    {
+      "epoch": 0.8519470627439779,
+      "grad_norm": 3.177137768987081,
+      "learning_rate": 2e-05,
+      "loss": 0.2401,
+      "step": 35792
+    },
+    {
+      "epoch": 0.8519708654670094,
+      "grad_norm": 9.928601657586118,
+      "learning_rate": 2e-05,
+      "loss": 0.5908,
+      "step": 35793
+    },
+    {
+      "epoch": 0.851994668190041,
+      "grad_norm": 15.440424331261369,
+      "learning_rate": 2e-05,
+      "loss": 0.0712,
+      "step": 35794
+    },
+    {
+      "epoch": 0.8520184709130725,
+      "grad_norm": 0.2574984348977341,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 35795
+    },
+    {
+      "epoch": 0.852042273636104,
+      "grad_norm": 2.3736798475527983,
+      "learning_rate": 2e-05,
+      "loss": 0.032,
+      "step": 35796
+    },
+    {
+      "epoch": 0.8520660763591354,
+      "grad_norm": 3.561180036642988,
+      "learning_rate": 2e-05,
+      "loss": 0.6142,
+      "step": 35797
+    },
+    {
+      "epoch": 0.852089879082167,
+      "grad_norm": 4.3858349555172165,
+      "learning_rate": 2e-05,
+      "loss": 0.8723,
+      "step": 35798
+    },
+    {
+      "epoch": 0.8521136818051985,
+      "grad_norm": 0.028283727119683095,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 35799
+    },
+    {
+      "epoch": 0.85213748452823,
+      "grad_norm": 5.521523970245344,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 35800
+    },
+    {
+      "epoch": 0.8521612872512615,
+      "grad_norm": 3.21402493833662,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 35801
+    },
+    {
+      "epoch": 0.8521850899742931,
+      "grad_norm": 1.2473550989084021,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 35802
+    },
+    {
+      "epoch": 0.8522088926973246,
+      "grad_norm": 0.30198770160176597,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 35803
+    },
+    {
+      "epoch": 0.8522326954203561,
+      "grad_norm": 78.51510306081498,
+      "learning_rate": 2e-05,
+      "loss": 1.6556,
+      "step": 35804
+    },
+    {
+      "epoch": 0.8522564981433876,
+      "grad_norm": 2.358929230022464,
+      "learning_rate": 2e-05,
+      "loss": 0.2684,
+      "step": 35805
+    },
+    {
+      "epoch": 0.8522803008664191,
+      "grad_norm": 2.1440260214735365,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 35806
+    },
+    {
+      "epoch": 0.8523041035894506,
+      "grad_norm": 8.466752161151334,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 35807
+    },
+    {
+      "epoch": 0.8523279063124821,
+      "grad_norm": 1.0083246278535054,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 35808
+    },
+    {
+      "epoch": 0.8523517090355137,
+      "grad_norm": 3.3477537124732035,
+      "learning_rate": 2e-05,
+      "loss": 0.2734,
+      "step": 35809
+    },
+    {
+      "epoch": 0.8523755117585452,
+      "grad_norm": 3.6033419570785914,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 35810
+    },
+    {
+      "epoch": 0.8523993144815767,
+      "grad_norm": 6.504087924932483,
+      "learning_rate": 2e-05,
+      "loss": 0.1707,
+      "step": 35811
+    },
+    {
+      "epoch": 0.8524231172046082,
+      "grad_norm": 0.6565375961160431,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 35812
+    },
+    {
+      "epoch": 0.8524469199276398,
+      "grad_norm": 4.735403882750182,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 35813
+    },
+    {
+      "epoch": 0.8524707226506713,
+      "grad_norm": 4.578441583020845,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 35814
+    },
+    {
+      "epoch": 0.8524945253737027,
+      "grad_norm": 7.782654269199091,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 35815
+    },
+    {
+      "epoch": 0.8525183280967342,
+      "grad_norm": 3.443023920833685,
+      "learning_rate": 2e-05,
+      "loss": 0.4683,
+      "step": 35816
+    },
+    {
+      "epoch": 0.8525421308197658,
+      "grad_norm": 59.707251377064935,
+      "learning_rate": 2e-05,
+      "loss": 0.8832,
+      "step": 35817
+    },
+    {
+      "epoch": 0.8525659335427973,
+      "grad_norm": 3.485862950631696,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 35818
+    },
+    {
+      "epoch": 0.8525897362658288,
+      "grad_norm": 1.184556205807178,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 35819
+    },
+    {
+      "epoch": 0.8526135389888603,
+      "grad_norm": 3.0182836202722774,
+      "learning_rate": 2e-05,
+      "loss": 0.344,
+      "step": 35820
+    },
+    {
+      "epoch": 0.8526373417118919,
+      "grad_norm": 6.343120573977408,
+      "learning_rate": 2e-05,
+      "loss": 0.8706,
+      "step": 35821
+    },
+    {
+      "epoch": 0.8526611444349234,
+      "grad_norm": 0.2882598729909376,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 35822
+    },
+    {
+      "epoch": 0.8526849471579548,
+      "grad_norm": 0.767132555815057,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 35823
+    },
+    {
+      "epoch": 0.8527087498809864,
+      "grad_norm": 4.6473278278985255,
+      "learning_rate": 2e-05,
+      "loss": 0.9629,
+      "step": 35824
+    },
+    {
+      "epoch": 0.8527325526040179,
+      "grad_norm": 0.0583193305509967,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 35825
+    },
+    {
+      "epoch": 0.8527563553270494,
+      "grad_norm": 3.3861745580290754,
+      "learning_rate": 2e-05,
+      "loss": 0.7801,
+      "step": 35826
+    },
+    {
+      "epoch": 0.8527801580500809,
+      "grad_norm": 2.2342012135108757,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 35827
+    },
+    {
+      "epoch": 0.8528039607731125,
+      "grad_norm": 4.0805341166073426,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 35828
+    },
+    {
+      "epoch": 0.852827763496144,
+      "grad_norm": 0.052867555183649946,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35829
+    },
+    {
+      "epoch": 0.8528515662191755,
+      "grad_norm": 3.803648923200782,
+      "learning_rate": 2e-05,
+      "loss": 0.642,
+      "step": 35830
+    },
+    {
+      "epoch": 0.852875368942207,
+      "grad_norm": 0.16042458302070983,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35831
+    },
+    {
+      "epoch": 0.8528991716652385,
+      "grad_norm": 0.9922018130302468,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 35832
+    },
+    {
+      "epoch": 0.85292297438827,
+      "grad_norm": 3.0276586816773534,
+      "learning_rate": 2e-05,
+      "loss": 0.4465,
+      "step": 35833
+    },
+    {
+      "epoch": 0.8529467771113015,
+      "grad_norm": 3.169015044060284,
+      "learning_rate": 2e-05,
+      "loss": 0.1513,
+      "step": 35834
+    },
+    {
+      "epoch": 0.852970579834333,
+      "grad_norm": 1.08263184357256,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 35835
+    },
+    {
+      "epoch": 0.8529943825573646,
+      "grad_norm": 6.873769348697375,
+      "learning_rate": 2e-05,
+      "loss": 1.3635,
+      "step": 35836
+    },
+    {
+      "epoch": 0.8530181852803961,
+      "grad_norm": 3.1024554586098,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 35837
+    },
+    {
+      "epoch": 0.8530419880034276,
+      "grad_norm": 6.242582579800711,
+      "learning_rate": 2e-05,
+      "loss": 0.7484,
+      "step": 35838
+    },
+    {
+      "epoch": 0.8530657907264592,
+      "grad_norm": 8.287576337913773,
+      "learning_rate": 2e-05,
+      "loss": 1.8023,
+      "step": 35839
+    },
+    {
+      "epoch": 0.8530895934494906,
+      "grad_norm": 3.3682975195678213,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 35840
+    },
+    {
+      "epoch": 0.8531133961725221,
+      "grad_norm": 0.47245009111396546,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 35841
+    },
+    {
+      "epoch": 0.8531371988955536,
+      "grad_norm": 0.0650558929887159,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35842
+    },
+    {
+      "epoch": 0.8531610016185852,
+      "grad_norm": 4.794234510740682,
+      "learning_rate": 2e-05,
+      "loss": 0.7275,
+      "step": 35843
+    },
+    {
+      "epoch": 0.8531848043416167,
+      "grad_norm": 4.397446200833623,
+      "learning_rate": 2e-05,
+      "loss": 0.4095,
+      "step": 35844
+    },
+    {
+      "epoch": 0.8532086070646482,
+      "grad_norm": 0.09898289366552501,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35845
+    },
+    {
+      "epoch": 0.8532324097876797,
+      "grad_norm": 0.15230598481968693,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 35846
+    },
+    {
+      "epoch": 0.8532562125107113,
+      "grad_norm": 3.17646180348861,
+      "learning_rate": 2e-05,
+      "loss": 0.162,
+      "step": 35847
+    },
+    {
+      "epoch": 0.8532800152337428,
+      "grad_norm": 2.44008425685053,
+      "learning_rate": 2e-05,
+      "loss": 0.1467,
+      "step": 35848
+    },
+    {
+      "epoch": 0.8533038179567742,
+      "grad_norm": 0.43655484426211605,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 35849
+    },
+    {
+      "epoch": 0.8533276206798057,
+      "grad_norm": 3.429001039916662,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 35850
+    },
+    {
+      "epoch": 0.8533514234028373,
+      "grad_norm": 0.28901633918065156,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 35851
+    },
+    {
+      "epoch": 0.8533752261258688,
+      "grad_norm": 3.0530861492927794,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 35852
+    },
+    {
+      "epoch": 0.8533990288489003,
+      "grad_norm": 0.26491014114369615,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 35853
+    },
+    {
+      "epoch": 0.8534228315719319,
+      "grad_norm": 2.947160340752982,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 35854
+    },
+    {
+      "epoch": 0.8534466342949634,
+      "grad_norm": 0.3569172488629581,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 35855
+    },
+    {
+      "epoch": 0.8534704370179949,
+      "grad_norm": 3.930510673358674,
+      "learning_rate": 2e-05,
+      "loss": 0.0765,
+      "step": 35856
+    },
+    {
+      "epoch": 0.8534942397410263,
+      "grad_norm": 1.1794801608583632,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 35857
+    },
+    {
+      "epoch": 0.8535180424640579,
+      "grad_norm": 0.8144539889983922,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 35858
+    },
+    {
+      "epoch": 0.8535418451870894,
+      "grad_norm": 3.26526633040729,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 35859
+    },
+    {
+      "epoch": 0.8535656479101209,
+      "grad_norm": 2.6789980897011203,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 35860
+    },
+    {
+      "epoch": 0.8535894506331524,
+      "grad_norm": 1.4229200799497463,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 35861
+    },
+    {
+      "epoch": 0.853613253356184,
+      "grad_norm": 146.01704202024865,
+      "learning_rate": 2e-05,
+      "loss": 0.5168,
+      "step": 35862
+    },
+    {
+      "epoch": 0.8536370560792155,
+      "grad_norm": 4.020853515794057,
+      "learning_rate": 2e-05,
+      "loss": 0.5831,
+      "step": 35863
+    },
+    {
+      "epoch": 0.853660858802247,
+      "grad_norm": 0.1710917916489378,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 35864
+    },
+    {
+      "epoch": 0.8536846615252784,
+      "grad_norm": 0.6375572668719401,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 35865
+    },
+    {
+      "epoch": 0.85370846424831,
+      "grad_norm": 5.669248689587499,
+      "learning_rate": 2e-05,
+      "loss": 1.1955,
+      "step": 35866
+    },
+    {
+      "epoch": 0.8537322669713415,
+      "grad_norm": 2.6225980746343107,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 35867
+    },
+    {
+      "epoch": 0.853756069694373,
+      "grad_norm": 8.218049164300925,
+      "learning_rate": 2e-05,
+      "loss": 0.8228,
+      "step": 35868
+    },
+    {
+      "epoch": 0.8537798724174046,
+      "grad_norm": 3.379396159934147,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 35869
+    },
+    {
+      "epoch": 0.8538036751404361,
+      "grad_norm": 0.3913101384312916,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 35870
+    },
+    {
+      "epoch": 0.8538274778634676,
+      "grad_norm": 0.18584081368802244,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 35871
+    },
+    {
+      "epoch": 0.8538512805864991,
+      "grad_norm": 0.07943007785879812,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 35872
+    },
+    {
+      "epoch": 0.8538750833095307,
+      "grad_norm": 50.04299228152577,
+      "learning_rate": 2e-05,
+      "loss": 1.6805,
+      "step": 35873
+    },
+    {
+      "epoch": 0.8538988860325621,
+      "grad_norm": 1.8018732182141834,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 35874
+    },
+    {
+      "epoch": 0.8539226887555936,
+      "grad_norm": 2.495362953570815,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 35875
+    },
+    {
+      "epoch": 0.8539464914786251,
+      "grad_norm": 5.46692028466544,
+      "learning_rate": 2e-05,
+      "loss": 0.9989,
+      "step": 35876
+    },
+    {
+      "epoch": 0.8539702942016567,
+      "grad_norm": 0.5551794181057215,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 35877
+    },
+    {
+      "epoch": 0.8539940969246882,
+      "grad_norm": 1.96354951972026,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 35878
+    },
+    {
+      "epoch": 0.8540178996477197,
+      "grad_norm": 7.980657874229646,
+      "learning_rate": 2e-05,
+      "loss": 0.3428,
+      "step": 35879
+    },
+    {
+      "epoch": 0.8540417023707512,
+      "grad_norm": 6.049479485570169,
+      "learning_rate": 2e-05,
+      "loss": 1.1214,
+      "step": 35880
+    },
+    {
+      "epoch": 0.8540655050937828,
+      "grad_norm": 20.201265888596218,
+      "learning_rate": 2e-05,
+      "loss": 0.9455,
+      "step": 35881
+    },
+    {
+      "epoch": 0.8540893078168142,
+      "grad_norm": 30.083372882286103,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 35882
+    },
+    {
+      "epoch": 0.8541131105398457,
+      "grad_norm": 0.9627537576445262,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 35883
+    },
+    {
+      "epoch": 0.8541369132628773,
+      "grad_norm": 7.834571915105341,
+      "learning_rate": 2e-05,
+      "loss": 1.1958,
+      "step": 35884
+    },
+    {
+      "epoch": 0.8541607159859088,
+      "grad_norm": 26.648894491542976,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 35885
+    },
+    {
+      "epoch": 0.8541845187089403,
+      "grad_norm": 3.1948688809051218,
+      "learning_rate": 2e-05,
+      "loss": 0.4493,
+      "step": 35886
+    },
+    {
+      "epoch": 0.8542083214319718,
+      "grad_norm": 3.9888933510228344,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 35887
+    },
+    {
+      "epoch": 0.8542321241550034,
+      "grad_norm": 8.045092622334364,
+      "learning_rate": 2e-05,
+      "loss": 0.1307,
+      "step": 35888
+    },
+    {
+      "epoch": 0.8542559268780349,
+      "grad_norm": 5.869737101261544,
+      "learning_rate": 2e-05,
+      "loss": 0.5771,
+      "step": 35889
+    },
+    {
+      "epoch": 0.8542797296010664,
+      "grad_norm": 5.345913856606384,
+      "learning_rate": 2e-05,
+      "loss": 1.5019,
+      "step": 35890
+    },
+    {
+      "epoch": 0.8543035323240978,
+      "grad_norm": 5.009881692527562,
+      "learning_rate": 2e-05,
+      "loss": 0.7431,
+      "step": 35891
+    },
+    {
+      "epoch": 0.8543273350471294,
+      "grad_norm": 1.7268232999548832,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 35892
+    },
+    {
+      "epoch": 0.8543511377701609,
+      "grad_norm": 3.270352487158499,
+      "learning_rate": 2e-05,
+      "loss": 0.2854,
+      "step": 35893
+    },
+    {
+      "epoch": 0.8543749404931924,
+      "grad_norm": 18.127802954979078,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 35894
+    },
+    {
+      "epoch": 0.8543987432162239,
+      "grad_norm": 0.4836280602981043,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 35895
+    },
+    {
+      "epoch": 0.8544225459392555,
+      "grad_norm": 2.9157428454093908,
+      "learning_rate": 2e-05,
+      "loss": 0.1749,
+      "step": 35896
+    },
+    {
+      "epoch": 0.854446348662287,
+      "grad_norm": 25.7267050157432,
+      "learning_rate": 2e-05,
+      "loss": 0.8572,
+      "step": 35897
+    },
+    {
+      "epoch": 0.8544701513853185,
+      "grad_norm": 4.355744197423241,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 35898
+    },
+    {
+      "epoch": 0.8544939541083499,
+      "grad_norm": 3.7717283506137744,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 35899
+    },
+    {
+      "epoch": 0.8545177568313815,
+      "grad_norm": 31.211376510295395,
+      "learning_rate": 2e-05,
+      "loss": 0.6327,
+      "step": 35900
+    },
+    {
+      "epoch": 0.854541559554413,
+      "grad_norm": 0.4750241549183072,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 35901
+    },
+    {
+      "epoch": 0.8545653622774445,
+      "grad_norm": 4.289187835174228,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 35902
+    },
+    {
+      "epoch": 0.8545891650004761,
+      "grad_norm": 6.436738852186791,
+      "learning_rate": 2e-05,
+      "loss": 1.2266,
+      "step": 35903
+    },
+    {
+      "epoch": 0.8546129677235076,
+      "grad_norm": 8.466773510079177,
+      "learning_rate": 2e-05,
+      "loss": 1.61,
+      "step": 35904
+    },
+    {
+      "epoch": 0.8546367704465391,
+      "grad_norm": 67.127583749749,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 35905
+    },
+    {
+      "epoch": 0.8546605731695706,
+      "grad_norm": 4.853331807827661,
+      "learning_rate": 2e-05,
+      "loss": 0.8371,
+      "step": 35906
+    },
+    {
+      "epoch": 0.8546843758926022,
+      "grad_norm": 0.36161607747095903,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 35907
+    },
+    {
+      "epoch": 0.8547081786156336,
+      "grad_norm": 2.852142470948946,
+      "learning_rate": 2e-05,
+      "loss": 0.3963,
+      "step": 35908
+    },
+    {
+      "epoch": 0.8547319813386651,
+      "grad_norm": 4.229181437884729,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 35909
+    },
+    {
+      "epoch": 0.8547557840616966,
+      "grad_norm": 0.7514874336665056,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 35910
+    },
+    {
+      "epoch": 0.8547795867847282,
+      "grad_norm": 11.627620157199939,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 35911
+    },
+    {
+      "epoch": 0.8548033895077597,
+      "grad_norm": 2.653735449383182,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 35912
+    },
+    {
+      "epoch": 0.8548271922307912,
+      "grad_norm": 0.2617021859527528,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 35913
+    },
+    {
+      "epoch": 0.8548509949538227,
+      "grad_norm": 0.03622116624181481,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 35914
+    },
+    {
+      "epoch": 0.8548747976768543,
+      "grad_norm": 5.192152430782925,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 35915
+    },
+    {
+      "epoch": 0.8548986003998857,
+      "grad_norm": 2.6493238596491575,
+      "learning_rate": 2e-05,
+      "loss": 0.4432,
+      "step": 35916
+    },
+    {
+      "epoch": 0.8549224031229172,
+      "grad_norm": 11.720128111775855,
+      "learning_rate": 2e-05,
+      "loss": 0.8325,
+      "step": 35917
+    },
+    {
+      "epoch": 0.8549462058459488,
+      "grad_norm": 0.03292474926254971,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 35918
+    },
+    {
+      "epoch": 0.8549700085689803,
+      "grad_norm": 2.2547726214360795,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 35919
+    },
+    {
+      "epoch": 0.8549938112920118,
+      "grad_norm": 11.075680963771827,
+      "learning_rate": 2e-05,
+      "loss": 0.4395,
+      "step": 35920
+    },
+    {
+      "epoch": 0.8550176140150433,
+      "grad_norm": 10.766010625087482,
+      "learning_rate": 2e-05,
+      "loss": 1.649,
+      "step": 35921
+    },
+    {
+      "epoch": 0.8550414167380749,
+      "grad_norm": 2.0683903257181595,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 35922
+    },
+    {
+      "epoch": 0.8550652194611064,
+      "grad_norm": 3.410282797944316,
+      "learning_rate": 2e-05,
+      "loss": 0.7288,
+      "step": 35923
+    },
+    {
+      "epoch": 0.8550890221841378,
+      "grad_norm": 0.7447293706287679,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 35924
+    },
+    {
+      "epoch": 0.8551128249071693,
+      "grad_norm": 4.46576205076601,
+      "learning_rate": 2e-05,
+      "loss": 0.5007,
+      "step": 35925
+    },
+    {
+      "epoch": 0.8551366276302009,
+      "grad_norm": 3.9957112646498683,
+      "learning_rate": 2e-05,
+      "loss": 0.5706,
+      "step": 35926
+    },
+    {
+      "epoch": 0.8551604303532324,
+      "grad_norm": 2.8167167348002975,
+      "learning_rate": 2e-05,
+      "loss": 0.0681,
+      "step": 35927
+    },
+    {
+      "epoch": 0.8551842330762639,
+      "grad_norm": 6.791760732690987,
+      "learning_rate": 2e-05,
+      "loss": 1.0248,
+      "step": 35928
+    },
+    {
+      "epoch": 0.8552080357992954,
+      "grad_norm": 10.69056697733569,
+      "learning_rate": 2e-05,
+      "loss": 0.2207,
+      "step": 35929
+    },
+    {
+      "epoch": 0.855231838522327,
+      "grad_norm": 6.444362016844604,
+      "learning_rate": 2e-05,
+      "loss": 1.0445,
+      "step": 35930
+    },
+    {
+      "epoch": 0.8552556412453585,
+      "grad_norm": 4.537541967941502,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 35931
+    },
+    {
+      "epoch": 0.85527944396839,
+      "grad_norm": 0.20675105352687603,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 35932
+    },
+    {
+      "epoch": 0.8553032466914215,
+      "grad_norm": 5.2096419429773695,
+      "learning_rate": 2e-05,
+      "loss": 0.7306,
+      "step": 35933
+    },
+    {
+      "epoch": 0.855327049414453,
+      "grad_norm": 4.246331935161523,
+      "learning_rate": 2e-05,
+      "loss": 0.5547,
+      "step": 35934
+    },
+    {
+      "epoch": 0.8553508521374845,
+      "grad_norm": 5.0001366785472054,
+      "learning_rate": 2e-05,
+      "loss": 0.672,
+      "step": 35935
+    },
+    {
+      "epoch": 0.855374654860516,
+      "grad_norm": 0.10136076673858499,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 35936
+    },
+    {
+      "epoch": 0.8553984575835476,
+      "grad_norm": 0.4309940373673097,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 35937
+    },
+    {
+      "epoch": 0.8554222603065791,
+      "grad_norm": 5.512183637925696,
+      "learning_rate": 2e-05,
+      "loss": 0.6015,
+      "step": 35938
+    },
+    {
+      "epoch": 0.8554460630296106,
+      "grad_norm": 7.200174600760565,
+      "learning_rate": 2e-05,
+      "loss": 0.4614,
+      "step": 35939
+    },
+    {
+      "epoch": 0.8554698657526421,
+      "grad_norm": 4.786092425220238,
+      "learning_rate": 2e-05,
+      "loss": 0.4362,
+      "step": 35940
+    },
+    {
+      "epoch": 0.8554936684756737,
+      "grad_norm": 5.429405549275892,
+      "learning_rate": 2e-05,
+      "loss": 0.1252,
+      "step": 35941
+    },
+    {
+      "epoch": 0.8555174711987051,
+      "grad_norm": 4.449511039326634,
+      "learning_rate": 2e-05,
+      "loss": 0.3234,
+      "step": 35942
+    },
+    {
+      "epoch": 0.8555412739217366,
+      "grad_norm": 2.6428504155001438,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 35943
+    },
+    {
+      "epoch": 0.8555650766447681,
+      "grad_norm": 0.0258866737336715,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 35944
+    },
+    {
+      "epoch": 0.8555888793677997,
+      "grad_norm": 10.273565346246256,
+      "learning_rate": 2e-05,
+      "loss": 0.2397,
+      "step": 35945
+    },
+    {
+      "epoch": 0.8556126820908312,
+      "grad_norm": 6.791073602951203,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 35946
+    },
+    {
+      "epoch": 0.8556364848138627,
+      "grad_norm": 2.0808741485467555,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 35947
+    },
+    {
+      "epoch": 0.8556602875368943,
+      "grad_norm": 0.5908673879749341,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 35948
+    },
+    {
+      "epoch": 0.8556840902599258,
+      "grad_norm": 2.448896313005554,
+      "learning_rate": 2e-05,
+      "loss": 0.2693,
+      "step": 35949
+    },
+    {
+      "epoch": 0.8557078929829572,
+      "grad_norm": 3.1723474719488984,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 35950
+    },
+    {
+      "epoch": 0.8557316957059887,
+      "grad_norm": 0.9845359058775437,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 35951
+    },
+    {
+      "epoch": 0.8557554984290203,
+      "grad_norm": 12.72959083179797,
+      "learning_rate": 2e-05,
+      "loss": 0.5863,
+      "step": 35952
+    },
+    {
+      "epoch": 0.8557793011520518,
+      "grad_norm": 13.208909875673827,
+      "learning_rate": 2e-05,
+      "loss": 0.2573,
+      "step": 35953
+    },
+    {
+      "epoch": 0.8558031038750833,
+      "grad_norm": 0.13547632528667897,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 35954
+    },
+    {
+      "epoch": 0.8558269065981148,
+      "grad_norm": 6.7896242710346915,
+      "learning_rate": 2e-05,
+      "loss": 0.4966,
+      "step": 35955
+    },
+    {
+      "epoch": 0.8558507093211464,
+      "grad_norm": 32.94363103310251,
+      "learning_rate": 2e-05,
+      "loss": 1.4153,
+      "step": 35956
+    },
+    {
+      "epoch": 0.8558745120441779,
+      "grad_norm": 0.40987257755942513,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 35957
+    },
+    {
+      "epoch": 0.8558983147672093,
+      "grad_norm": 20.226465927395264,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 35958
+    },
+    {
+      "epoch": 0.8559221174902408,
+      "grad_norm": 3.8647820895743275,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 35959
+    },
+    {
+      "epoch": 0.8559459202132724,
+      "grad_norm": 5.852025889221333,
+      "learning_rate": 2e-05,
+      "loss": 0.6835,
+      "step": 35960
+    },
+    {
+      "epoch": 0.8559697229363039,
+      "grad_norm": 6.685352498425318,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 35961
+    },
+    {
+      "epoch": 0.8559935256593354,
+      "grad_norm": 8.578774841996328,
+      "learning_rate": 2e-05,
+      "loss": 1.216,
+      "step": 35962
+    },
+    {
+      "epoch": 0.856017328382367,
+      "grad_norm": 0.391723710481133,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 35963
+    },
+    {
+      "epoch": 0.8560411311053985,
+      "grad_norm": 4.701049207551873,
+      "learning_rate": 2e-05,
+      "loss": 1.308,
+      "step": 35964
+    },
+    {
+      "epoch": 0.85606493382843,
+      "grad_norm": 5.85188176989616,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 35965
+    },
+    {
+      "epoch": 0.8560887365514614,
+      "grad_norm": 2.945104693829694,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 35966
+    },
+    {
+      "epoch": 0.856112539274493,
+      "grad_norm": 2.9813920178728366,
+      "learning_rate": 2e-05,
+      "loss": 0.3291,
+      "step": 35967
+    },
+    {
+      "epoch": 0.8561363419975245,
+      "grad_norm": 0.32856351825718283,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 35968
+    },
+    {
+      "epoch": 0.856160144720556,
+      "grad_norm": 3.353611369395085,
+      "learning_rate": 2e-05,
+      "loss": 0.1126,
+      "step": 35969
+    },
+    {
+      "epoch": 0.8561839474435875,
+      "grad_norm": 3.7519692758794023,
+      "learning_rate": 2e-05,
+      "loss": 0.7706,
+      "step": 35970
+    },
+    {
+      "epoch": 0.8562077501666191,
+      "grad_norm": 4.244473154184945,
+      "learning_rate": 2e-05,
+      "loss": 0.2556,
+      "step": 35971
+    },
+    {
+      "epoch": 0.8562315528896506,
+      "grad_norm": 7.8053493470341175,
+      "learning_rate": 2e-05,
+      "loss": 1.0181,
+      "step": 35972
+    },
+    {
+      "epoch": 0.8562553556126821,
+      "grad_norm": 3.583668955150406,
+      "learning_rate": 2e-05,
+      "loss": 0.6225,
+      "step": 35973
+    },
+    {
+      "epoch": 0.8562791583357136,
+      "grad_norm": 2.6921396498343952,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 35974
+    },
+    {
+      "epoch": 0.8563029610587451,
+      "grad_norm": 5.214773499213732,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 35975
+    },
+    {
+      "epoch": 0.8563267637817766,
+      "grad_norm": 6.433628463102852,
+      "learning_rate": 2e-05,
+      "loss": 0.7263,
+      "step": 35976
+    },
+    {
+      "epoch": 0.8563505665048081,
+      "grad_norm": 5.205108348768422,
+      "learning_rate": 2e-05,
+      "loss": 0.4458,
+      "step": 35977
+    },
+    {
+      "epoch": 0.8563743692278397,
+      "grad_norm": 2.3195262371471945,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 35978
+    },
+    {
+      "epoch": 0.8563981719508712,
+      "grad_norm": 4.900333201962547,
+      "learning_rate": 2e-05,
+      "loss": 0.6781,
+      "step": 35979
+    },
+    {
+      "epoch": 0.8564219746739027,
+      "grad_norm": 1.942079341955677,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 35980
+    },
+    {
+      "epoch": 0.8564457773969342,
+      "grad_norm": 12.27353231138112,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 35981
+    },
+    {
+      "epoch": 0.8564695801199658,
+      "grad_norm": 24.50480104086391,
+      "learning_rate": 2e-05,
+      "loss": 0.8285,
+      "step": 35982
+    },
+    {
+      "epoch": 0.8564933828429973,
+      "grad_norm": 4.691886034005262,
+      "learning_rate": 2e-05,
+      "loss": 0.0593,
+      "step": 35983
+    },
+    {
+      "epoch": 0.8565171855660287,
+      "grad_norm": 3.3026542514857393,
+      "learning_rate": 2e-05,
+      "loss": 0.4981,
+      "step": 35984
+    },
+    {
+      "epoch": 0.8565409882890602,
+      "grad_norm": 10.947451583403723,
+      "learning_rate": 2e-05,
+      "loss": 0.1984,
+      "step": 35985
+    },
+    {
+      "epoch": 0.8565647910120918,
+      "grad_norm": 3.9354693070720157,
+      "learning_rate": 2e-05,
+      "loss": 0.1189,
+      "step": 35986
+    },
+    {
+      "epoch": 0.8565885937351233,
+      "grad_norm": 9.836759330268992,
+      "learning_rate": 2e-05,
+      "loss": 0.9114,
+      "step": 35987
+    },
+    {
+      "epoch": 0.8566123964581548,
+      "grad_norm": 5.7751359270838565,
+      "learning_rate": 2e-05,
+      "loss": 1.2949,
+      "step": 35988
+    },
+    {
+      "epoch": 0.8566361991811863,
+      "grad_norm": 3.979856558549668,
+      "learning_rate": 2e-05,
+      "loss": 0.8607,
+      "step": 35989
+    },
+    {
+      "epoch": 0.8566600019042179,
+      "grad_norm": 0.30367429879124974,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 35990
+    },
+    {
+      "epoch": 0.8566838046272494,
+      "grad_norm": 2.4235110681933656,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 35991
+    },
+    {
+      "epoch": 0.8567076073502808,
+      "grad_norm": 4.180260737011417,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 35992
+    },
+    {
+      "epoch": 0.8567314100733124,
+      "grad_norm": 9.696656668139093,
+      "learning_rate": 2e-05,
+      "loss": 0.1188,
+      "step": 35993
+    },
+    {
+      "epoch": 0.8567552127963439,
+      "grad_norm": 10.65958620150895,
+      "learning_rate": 2e-05,
+      "loss": 0.276,
+      "step": 35994
+    },
+    {
+      "epoch": 0.8567790155193754,
+      "grad_norm": 3.186099031996047,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 35995
+    },
+    {
+      "epoch": 0.8568028182424069,
+      "grad_norm": 6.7538971214046795,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 35996
+    },
+    {
+      "epoch": 0.8568266209654385,
+      "grad_norm": 5.4958021486779955,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 35997
+    },
+    {
+      "epoch": 0.85685042368847,
+      "grad_norm": 1.1624274654226636,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 35998
+    },
+    {
+      "epoch": 0.8568742264115015,
+      "grad_norm": 0.15366308605474568,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 35999
+    },
+    {
+      "epoch": 0.856898029134533,
+      "grad_norm": 11.678221617611694,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 36000
+    },
+    {
+      "epoch": 0.8569218318575645,
+      "grad_norm": 0.41188230296000256,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 36001
+    },
+    {
+      "epoch": 0.856945634580596,
+      "grad_norm": 4.811789717388302,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 36002
+    },
+    {
+      "epoch": 0.8569694373036275,
+      "grad_norm": 3.6717233627140184,
+      "learning_rate": 2e-05,
+      "loss": 0.4022,
+      "step": 36003
+    },
+    {
+      "epoch": 0.856993240026659,
+      "grad_norm": 3.2870126560256,
+      "learning_rate": 2e-05,
+      "loss": 0.2707,
+      "step": 36004
+    },
+    {
+      "epoch": 0.8570170427496906,
+      "grad_norm": 0.5070677894424047,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 36005
+    },
+    {
+      "epoch": 0.8570408454727221,
+      "grad_norm": 4.294049547595779,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 36006
+    },
+    {
+      "epoch": 0.8570646481957536,
+      "grad_norm": 10.105675239286127,
+      "learning_rate": 2e-05,
+      "loss": 0.7927,
+      "step": 36007
+    },
+    {
+      "epoch": 0.8570884509187852,
+      "grad_norm": 0.5064991462778399,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 36008
+    },
+    {
+      "epoch": 0.8571122536418166,
+      "grad_norm": 6.1629422029418075,
+      "learning_rate": 2e-05,
+      "loss": 0.0998,
+      "step": 36009
+    },
+    {
+      "epoch": 0.8571360563648481,
+      "grad_norm": 3.0665467267137863,
+      "learning_rate": 2e-05,
+      "loss": 0.4169,
+      "step": 36010
+    },
+    {
+      "epoch": 0.8571598590878796,
+      "grad_norm": 8.017173510096034,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 36011
+    },
+    {
+      "epoch": 0.8571836618109112,
+      "grad_norm": 3.5878372217704064,
+      "learning_rate": 2e-05,
+      "loss": 0.5095,
+      "step": 36012
+    },
+    {
+      "epoch": 0.8572074645339427,
+      "grad_norm": 0.40150742836484127,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 36013
+    },
+    {
+      "epoch": 0.8572312672569742,
+      "grad_norm": 0.885195480526753,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 36014
+    },
+    {
+      "epoch": 0.8572550699800057,
+      "grad_norm": 0.07080747358422941,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36015
+    },
+    {
+      "epoch": 0.8572788727030373,
+      "grad_norm": 3.555225357500371,
+      "learning_rate": 2e-05,
+      "loss": 0.3786,
+      "step": 36016
+    },
+    {
+      "epoch": 0.8573026754260687,
+      "grad_norm": 2.7093986011072317,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 36017
+    },
+    {
+      "epoch": 0.8573264781491002,
+      "grad_norm": 0.44221836761325145,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 36018
+    },
+    {
+      "epoch": 0.8573502808721317,
+      "grad_norm": 0.34535836050609403,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 36019
+    },
+    {
+      "epoch": 0.8573740835951633,
+      "grad_norm": 7.18913424138101,
+      "learning_rate": 2e-05,
+      "loss": 0.2116,
+      "step": 36020
+    },
+    {
+      "epoch": 0.8573978863181948,
+      "grad_norm": 4.380972384107119,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 36021
+    },
+    {
+      "epoch": 0.8574216890412263,
+      "grad_norm": 1.635696149857184,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 36022
+    },
+    {
+      "epoch": 0.8574454917642579,
+      "grad_norm": 5.56192439535736,
+      "learning_rate": 2e-05,
+      "loss": 0.9463,
+      "step": 36023
+    },
+    {
+      "epoch": 0.8574692944872894,
+      "grad_norm": 5.4505751854851345,
+      "learning_rate": 2e-05,
+      "loss": 0.1345,
+      "step": 36024
+    },
+    {
+      "epoch": 0.8574930972103209,
+      "grad_norm": 2.8071959669864466,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 36025
+    },
+    {
+      "epoch": 0.8575168999333523,
+      "grad_norm": 4.440353568447781,
+      "learning_rate": 2e-05,
+      "loss": 0.6702,
+      "step": 36026
+    },
+    {
+      "epoch": 0.8575407026563839,
+      "grad_norm": 1.6127846884054007,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 36027
+    },
+    {
+      "epoch": 0.8575645053794154,
+      "grad_norm": 7.833675297209439,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 36028
+    },
+    {
+      "epoch": 0.8575883081024469,
+      "grad_norm": 3.3695034646822797,
+      "learning_rate": 2e-05,
+      "loss": 0.3878,
+      "step": 36029
+    },
+    {
+      "epoch": 0.8576121108254784,
+      "grad_norm": 11.382810577237164,
+      "learning_rate": 2e-05,
+      "loss": 0.2079,
+      "step": 36030
+    },
+    {
+      "epoch": 0.85763591354851,
+      "grad_norm": 0.0260768987433642,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36031
+    },
+    {
+      "epoch": 0.8576597162715415,
+      "grad_norm": 8.666002775307913,
+      "learning_rate": 2e-05,
+      "loss": 1.1317,
+      "step": 36032
+    },
+    {
+      "epoch": 0.857683518994573,
+      "grad_norm": 0.17490604123883577,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36033
+    },
+    {
+      "epoch": 0.8577073217176044,
+      "grad_norm": 2.283581465559888,
+      "learning_rate": 2e-05,
+      "loss": 0.0835,
+      "step": 36034
+    },
+    {
+      "epoch": 0.857731124440636,
+      "grad_norm": 0.5836636639692225,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 36035
+    },
+    {
+      "epoch": 0.8577549271636675,
+      "grad_norm": 9.063616715247855,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 36036
+    },
+    {
+      "epoch": 0.857778729886699,
+      "grad_norm": 0.08124622785443673,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 36037
+    },
+    {
+      "epoch": 0.8578025326097306,
+      "grad_norm": 0.5346218531741576,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 36038
+    },
+    {
+      "epoch": 0.8578263353327621,
+      "grad_norm": 2.971580473036763,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 36039
+    },
+    {
+      "epoch": 0.8578501380557936,
+      "grad_norm": 13.759345464209744,
+      "learning_rate": 2e-05,
+      "loss": 0.5638,
+      "step": 36040
+    },
+    {
+      "epoch": 0.8578739407788251,
+      "grad_norm": 3.1234567553636894,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 36041
+    },
+    {
+      "epoch": 0.8578977435018567,
+      "grad_norm": 1.023993211817236,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 36042
+    },
+    {
+      "epoch": 0.8579215462248881,
+      "grad_norm": 0.021560217704202456,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36043
+    },
+    {
+      "epoch": 0.8579453489479196,
+      "grad_norm": 0.4790028236192468,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 36044
+    },
+    {
+      "epoch": 0.8579691516709511,
+      "grad_norm": 0.1911823374028348,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 36045
+    },
+    {
+      "epoch": 0.8579929543939827,
+      "grad_norm": 1.5482498288905582,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 36046
+    },
+    {
+      "epoch": 0.8580167571170142,
+      "grad_norm": 0.06663855215851391,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 36047
+    },
+    {
+      "epoch": 0.8580405598400457,
+      "grad_norm": 4.398351516332909,
+      "learning_rate": 2e-05,
+      "loss": 0.0714,
+      "step": 36048
+    },
+    {
+      "epoch": 0.8580643625630772,
+      "grad_norm": 7.507561348161152,
+      "learning_rate": 2e-05,
+      "loss": 1.7544,
+      "step": 36049
+    },
+    {
+      "epoch": 0.8580881652861088,
+      "grad_norm": 5.390405655611806,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 36050
+    },
+    {
+      "epoch": 0.8581119680091402,
+      "grad_norm": 2.5458502260862055,
+      "learning_rate": 2e-05,
+      "loss": 0.1298,
+      "step": 36051
+    },
+    {
+      "epoch": 0.8581357707321717,
+      "grad_norm": 4.556700195458029,
+      "learning_rate": 2e-05,
+      "loss": 0.6954,
+      "step": 36052
+    },
+    {
+      "epoch": 0.8581595734552033,
+      "grad_norm": 3.419618833978675,
+      "learning_rate": 2e-05,
+      "loss": 0.4838,
+      "step": 36053
+    },
+    {
+      "epoch": 0.8581833761782348,
+      "grad_norm": 6.2054289369322575,
+      "learning_rate": 2e-05,
+      "loss": 0.7725,
+      "step": 36054
+    },
+    {
+      "epoch": 0.8582071789012663,
+      "grad_norm": 1.0961714993995462,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 36055
+    },
+    {
+      "epoch": 0.8582309816242978,
+      "grad_norm": 29.76152923937987,
+      "learning_rate": 2e-05,
+      "loss": 0.2531,
+      "step": 36056
+    },
+    {
+      "epoch": 0.8582547843473294,
+      "grad_norm": 0.09222600582470866,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 36057
+    },
+    {
+      "epoch": 0.8582785870703609,
+      "grad_norm": 1.3848240852929332,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 36058
+    },
+    {
+      "epoch": 0.8583023897933924,
+      "grad_norm": 0.20465970260614477,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 36059
+    },
+    {
+      "epoch": 0.8583261925164238,
+      "grad_norm": 4.2323001208176585,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 36060
+    },
+    {
+      "epoch": 0.8583499952394554,
+      "grad_norm": 4.610251253085797,
+      "learning_rate": 2e-05,
+      "loss": 0.438,
+      "step": 36061
+    },
+    {
+      "epoch": 0.8583737979624869,
+      "grad_norm": 0.41337141289513224,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 36062
+    },
+    {
+      "epoch": 0.8583976006855184,
+      "grad_norm": 3.9403021195650663,
+      "learning_rate": 2e-05,
+      "loss": 0.7074,
+      "step": 36063
+    },
+    {
+      "epoch": 0.8584214034085499,
+      "grad_norm": 9.655886125037359,
+      "learning_rate": 2e-05,
+      "loss": 1.235,
+      "step": 36064
+    },
+    {
+      "epoch": 0.8584452061315815,
+      "grad_norm": 6.947882776265889,
+      "learning_rate": 2e-05,
+      "loss": 0.8824,
+      "step": 36065
+    },
+    {
+      "epoch": 0.858469008854613,
+      "grad_norm": 3.608117497272895,
+      "learning_rate": 2e-05,
+      "loss": 0.5522,
+      "step": 36066
+    },
+    {
+      "epoch": 0.8584928115776445,
+      "grad_norm": 4.354038028701154,
+      "learning_rate": 2e-05,
+      "loss": 0.4511,
+      "step": 36067
+    },
+    {
+      "epoch": 0.858516614300676,
+      "grad_norm": 2.807781435011461,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 36068
+    },
+    {
+      "epoch": 0.8585404170237075,
+      "grad_norm": 1.289417600878753,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 36069
+    },
+    {
+      "epoch": 0.858564219746739,
+      "grad_norm": 2.422913004518313,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 36070
+    },
+    {
+      "epoch": 0.8585880224697705,
+      "grad_norm": 10.545945351137808,
+      "learning_rate": 2e-05,
+      "loss": 0.9579,
+      "step": 36071
+    },
+    {
+      "epoch": 0.8586118251928021,
+      "grad_norm": 19.37525117643513,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 36072
+    },
+    {
+      "epoch": 0.8586356279158336,
+      "grad_norm": 2.829809915344043,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 36073
+    },
+    {
+      "epoch": 0.8586594306388651,
+      "grad_norm": 6.656841002711409,
+      "learning_rate": 2e-05,
+      "loss": 1.0178,
+      "step": 36074
+    },
+    {
+      "epoch": 0.8586832333618966,
+      "grad_norm": 2.9180945392392985,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 36075
+    },
+    {
+      "epoch": 0.8587070360849282,
+      "grad_norm": 0.06043325498732343,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36076
+    },
+    {
+      "epoch": 0.8587308388079596,
+      "grad_norm": 4.309182233490547,
+      "learning_rate": 2e-05,
+      "loss": 0.5242,
+      "step": 36077
+    },
+    {
+      "epoch": 0.8587546415309911,
+      "grad_norm": 0.23999773412428707,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 36078
+    },
+    {
+      "epoch": 0.8587784442540226,
+      "grad_norm": 3.819660771658572,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 36079
+    },
+    {
+      "epoch": 0.8588022469770542,
+      "grad_norm": 9.934413299021244,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 36080
+    },
+    {
+      "epoch": 0.8588260497000857,
+      "grad_norm": 10.220162380609999,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 36081
+    },
+    {
+      "epoch": 0.8588498524231172,
+      "grad_norm": 5.161970918537873,
+      "learning_rate": 2e-05,
+      "loss": 0.7288,
+      "step": 36082
+    },
+    {
+      "epoch": 0.8588736551461488,
+      "grad_norm": 2.981014875325638,
+      "learning_rate": 2e-05,
+      "loss": 0.4526,
+      "step": 36083
+    },
+    {
+      "epoch": 0.8588974578691803,
+      "grad_norm": 5.823503671169003,
+      "learning_rate": 2e-05,
+      "loss": 0.1695,
+      "step": 36084
+    },
+    {
+      "epoch": 0.8589212605922117,
+      "grad_norm": 6.1271710758189055,
+      "learning_rate": 2e-05,
+      "loss": 0.169,
+      "step": 36085
+    },
+    {
+      "epoch": 0.8589450633152432,
+      "grad_norm": 20.931385189981622,
+      "learning_rate": 2e-05,
+      "loss": 1.3454,
+      "step": 36086
+    },
+    {
+      "epoch": 0.8589688660382748,
+      "grad_norm": 4.000645334206715,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 36087
+    },
+    {
+      "epoch": 0.8589926687613063,
+      "grad_norm": 6.029333890581257,
+      "learning_rate": 2e-05,
+      "loss": 0.5203,
+      "step": 36088
+    },
+    {
+      "epoch": 0.8590164714843378,
+      "grad_norm": 0.22881799050206628,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 36089
+    },
+    {
+      "epoch": 0.8590402742073693,
+      "grad_norm": 5.001130705882926,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 36090
+    },
+    {
+      "epoch": 0.8590640769304009,
+      "grad_norm": 0.3879498947105884,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 36091
+    },
+    {
+      "epoch": 0.8590878796534324,
+      "grad_norm": 3.0767560240967584,
+      "learning_rate": 2e-05,
+      "loss": 0.4504,
+      "step": 36092
+    },
+    {
+      "epoch": 0.8591116823764638,
+      "grad_norm": 18.04138154436041,
+      "learning_rate": 2e-05,
+      "loss": 0.5318,
+      "step": 36093
+    },
+    {
+      "epoch": 0.8591354850994953,
+      "grad_norm": 7.255726276595558,
+      "learning_rate": 2e-05,
+      "loss": 0.101,
+      "step": 36094
+    },
+    {
+      "epoch": 0.8591592878225269,
+      "grad_norm": 9.937511765291395,
+      "learning_rate": 2e-05,
+      "loss": 0.0757,
+      "step": 36095
+    },
+    {
+      "epoch": 0.8591830905455584,
+      "grad_norm": 0.5248859744675796,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 36096
+    },
+    {
+      "epoch": 0.8592068932685899,
+      "grad_norm": 0.6324994734245949,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 36097
+    },
+    {
+      "epoch": 0.8592306959916215,
+      "grad_norm": 4.652358651115957,
+      "learning_rate": 2e-05,
+      "loss": 0.9991,
+      "step": 36098
+    },
+    {
+      "epoch": 0.859254498714653,
+      "grad_norm": 19.241687660408797,
+      "learning_rate": 2e-05,
+      "loss": 0.6984,
+      "step": 36099
+    },
+    {
+      "epoch": 0.8592783014376845,
+      "grad_norm": 4.685553651435736,
+      "learning_rate": 2e-05,
+      "loss": 0.7917,
+      "step": 36100
+    },
+    {
+      "epoch": 0.859302104160716,
+      "grad_norm": 4.888642783653765,
+      "learning_rate": 2e-05,
+      "loss": 0.9961,
+      "step": 36101
+    },
+    {
+      "epoch": 0.8593259068837475,
+      "grad_norm": 9.09447072652213,
+      "learning_rate": 2e-05,
+      "loss": 2.7757,
+      "step": 36102
+    },
+    {
+      "epoch": 0.859349709606779,
+      "grad_norm": 1.5561149751407966,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 36103
+    },
+    {
+      "epoch": 0.8593735123298105,
+      "grad_norm": 0.749194655122041,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 36104
+    },
+    {
+      "epoch": 0.859397315052842,
+      "grad_norm": 1.3999528164753798,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 36105
+    },
+    {
+      "epoch": 0.8594211177758736,
+      "grad_norm": 0.025175214147509044,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36106
+    },
+    {
+      "epoch": 0.8594449204989051,
+      "grad_norm": 11.002901887564787,
+      "learning_rate": 2e-05,
+      "loss": 0.1344,
+      "step": 36107
+    },
+    {
+      "epoch": 0.8594687232219366,
+      "grad_norm": 0.0951592875069036,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 36108
+    },
+    {
+      "epoch": 0.859492525944968,
+      "grad_norm": 0.6775586905644155,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 36109
+    },
+    {
+      "epoch": 0.8595163286679997,
+      "grad_norm": 9.055881628773859,
+      "learning_rate": 2e-05,
+      "loss": 0.506,
+      "step": 36110
+    },
+    {
+      "epoch": 0.8595401313910311,
+      "grad_norm": 2.2744479768381254,
+      "learning_rate": 2e-05,
+      "loss": 0.0503,
+      "step": 36111
+    },
+    {
+      "epoch": 0.8595639341140626,
+      "grad_norm": 7.420263798750519,
+      "learning_rate": 2e-05,
+      "loss": 1.4975,
+      "step": 36112
+    },
+    {
+      "epoch": 0.8595877368370942,
+      "grad_norm": 4.8441517869042,
+      "learning_rate": 2e-05,
+      "loss": 0.0959,
+      "step": 36113
+    },
+    {
+      "epoch": 0.8596115395601257,
+      "grad_norm": 3.1442776081753983,
+      "learning_rate": 2e-05,
+      "loss": 0.2858,
+      "step": 36114
+    },
+    {
+      "epoch": 0.8596353422831572,
+      "grad_norm": 4.12996366083489,
+      "learning_rate": 2e-05,
+      "loss": 0.6852,
+      "step": 36115
+    },
+    {
+      "epoch": 0.8596591450061887,
+      "grad_norm": 5.56842418413664,
+      "learning_rate": 2e-05,
+      "loss": 0.8204,
+      "step": 36116
+    },
+    {
+      "epoch": 0.8596829477292203,
+      "grad_norm": 16.19461171727809,
+      "learning_rate": 2e-05,
+      "loss": 0.6846,
+      "step": 36117
+    },
+    {
+      "epoch": 0.8597067504522518,
+      "grad_norm": 4.0939819863746285,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 36118
+    },
+    {
+      "epoch": 0.8597305531752832,
+      "grad_norm": 3.7397234211837143,
+      "learning_rate": 2e-05,
+      "loss": 0.0967,
+      "step": 36119
+    },
+    {
+      "epoch": 0.8597543558983147,
+      "grad_norm": 7.938059303129566,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 36120
+    },
+    {
+      "epoch": 0.8597781586213463,
+      "grad_norm": 17.553150654953875,
+      "learning_rate": 2e-05,
+      "loss": 0.9966,
+      "step": 36121
+    },
+    {
+      "epoch": 0.8598019613443778,
+      "grad_norm": 2.8576923407953934,
+      "learning_rate": 2e-05,
+      "loss": 0.5841,
+      "step": 36122
+    },
+    {
+      "epoch": 0.8598257640674093,
+      "grad_norm": 13.19272311429545,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 36123
+    },
+    {
+      "epoch": 0.8598495667904408,
+      "grad_norm": 3.8264945719538357,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 36124
+    },
+    {
+      "epoch": 0.8598733695134724,
+      "grad_norm": 6.226294319705705,
+      "learning_rate": 2e-05,
+      "loss": 0.7349,
+      "step": 36125
+    },
+    {
+      "epoch": 0.8598971722365039,
+      "grad_norm": 2.089558090565617,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 36126
+    },
+    {
+      "epoch": 0.8599209749595353,
+      "grad_norm": 15.28747632682624,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 36127
+    },
+    {
+      "epoch": 0.8599447776825669,
+      "grad_norm": 0.13660121512355727,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36128
+    },
+    {
+      "epoch": 0.8599685804055984,
+      "grad_norm": 35.40284966625911,
+      "learning_rate": 2e-05,
+      "loss": 1.3271,
+      "step": 36129
+    },
+    {
+      "epoch": 0.8599923831286299,
+      "grad_norm": 0.2688754325615941,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 36130
+    },
+    {
+      "epoch": 0.8600161858516614,
+      "grad_norm": 3.550249460412248,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 36131
+    },
+    {
+      "epoch": 0.860039988574693,
+      "grad_norm": 2.3360306889279547,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 36132
+    },
+    {
+      "epoch": 0.8600637912977245,
+      "grad_norm": 2.335819019046017,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 36133
+    },
+    {
+      "epoch": 0.860087594020756,
+      "grad_norm": 1.5310580873104347,
+      "learning_rate": 2e-05,
+      "loss": 0.0386,
+      "step": 36134
+    },
+    {
+      "epoch": 0.8601113967437874,
+      "grad_norm": 3.764232407165155,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 36135
+    },
+    {
+      "epoch": 0.860135199466819,
+      "grad_norm": 3.842999997693189,
+      "learning_rate": 2e-05,
+      "loss": 0.6832,
+      "step": 36136
+    },
+    {
+      "epoch": 0.8601590021898505,
+      "grad_norm": 20.953241338686766,
+      "learning_rate": 2e-05,
+      "loss": 1.3343,
+      "step": 36137
+    },
+    {
+      "epoch": 0.860182804912882,
+      "grad_norm": 0.1362468757852964,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 36138
+    },
+    {
+      "epoch": 0.8602066076359135,
+      "grad_norm": 0.9732639321097977,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 36139
+    },
+    {
+      "epoch": 0.8602304103589451,
+      "grad_norm": 1.9462846697295748,
+      "learning_rate": 2e-05,
+      "loss": 0.0492,
+      "step": 36140
+    },
+    {
+      "epoch": 0.8602542130819766,
+      "grad_norm": 2.2638779296963314,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 36141
+    },
+    {
+      "epoch": 0.8602780158050081,
+      "grad_norm": 2.7165982153604453,
+      "learning_rate": 2e-05,
+      "loss": 0.6687,
+      "step": 36142
+    },
+    {
+      "epoch": 0.8603018185280396,
+      "grad_norm": 3.0857552601245377,
+      "learning_rate": 2e-05,
+      "loss": 0.3448,
+      "step": 36143
+    },
+    {
+      "epoch": 0.8603256212510711,
+      "grad_norm": 0.15779508956548605,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 36144
+    },
+    {
+      "epoch": 0.8603494239741026,
+      "grad_norm": 4.7049832903383875,
+      "learning_rate": 2e-05,
+      "loss": 0.6625,
+      "step": 36145
+    },
+    {
+      "epoch": 0.8603732266971341,
+      "grad_norm": 3.687192511900848,
+      "learning_rate": 2e-05,
+      "loss": 0.3005,
+      "step": 36146
+    },
+    {
+      "epoch": 0.8603970294201657,
+      "grad_norm": 3.110709988210748,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 36147
+    },
+    {
+      "epoch": 0.8604208321431972,
+      "grad_norm": 2.565929110692102,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 36148
+    },
+    {
+      "epoch": 0.8604446348662287,
+      "grad_norm": 4.074607180878294,
+      "learning_rate": 2e-05,
+      "loss": 0.1779,
+      "step": 36149
+    },
+    {
+      "epoch": 0.8604684375892602,
+      "grad_norm": 9.497203011563927,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 36150
+    },
+    {
+      "epoch": 0.8604922403122918,
+      "grad_norm": 2.4018076299636486,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 36151
+    },
+    {
+      "epoch": 0.8605160430353233,
+      "grad_norm": 2.926464009154083,
+      "learning_rate": 2e-05,
+      "loss": 0.0539,
+      "step": 36152
+    },
+    {
+      "epoch": 0.8605398457583547,
+      "grad_norm": 0.046421158732559545,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 36153
+    },
+    {
+      "epoch": 0.8605636484813862,
+      "grad_norm": 0.47640419248223975,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 36154
+    },
+    {
+      "epoch": 0.8605874512044178,
+      "grad_norm": 6.352874373998828,
+      "learning_rate": 2e-05,
+      "loss": 0.4951,
+      "step": 36155
+    },
+    {
+      "epoch": 0.8606112539274493,
+      "grad_norm": 4.019338168491239,
+      "learning_rate": 2e-05,
+      "loss": 0.5948,
+      "step": 36156
+    },
+    {
+      "epoch": 0.8606350566504808,
+      "grad_norm": 4.983535978449819,
+      "learning_rate": 2e-05,
+      "loss": 1.2132,
+      "step": 36157
+    },
+    {
+      "epoch": 0.8606588593735123,
+      "grad_norm": 2.637340952018299,
+      "learning_rate": 2e-05,
+      "loss": 0.4387,
+      "step": 36158
+    },
+    {
+      "epoch": 0.8606826620965439,
+      "grad_norm": 3.146318751752406,
+      "learning_rate": 2e-05,
+      "loss": 0.0636,
+      "step": 36159
+    },
+    {
+      "epoch": 0.8607064648195754,
+      "grad_norm": 2.962319971254827,
+      "learning_rate": 2e-05,
+      "loss": 0.4572,
+      "step": 36160
+    },
+    {
+      "epoch": 0.8607302675426068,
+      "grad_norm": 3.8867951058037478,
+      "learning_rate": 2e-05,
+      "loss": 0.7085,
+      "step": 36161
+    },
+    {
+      "epoch": 0.8607540702656384,
+      "grad_norm": 0.36257586626813054,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 36162
+    },
+    {
+      "epoch": 0.8607778729886699,
+      "grad_norm": 13.846730030100169,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 36163
+    },
+    {
+      "epoch": 0.8608016757117014,
+      "grad_norm": 0.0773779419408584,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 36164
+    },
+    {
+      "epoch": 0.8608254784347329,
+      "grad_norm": 5.799438679010969,
+      "learning_rate": 2e-05,
+      "loss": 0.2283,
+      "step": 36165
+    },
+    {
+      "epoch": 0.8608492811577645,
+      "grad_norm": 1.1578269069155769,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 36166
+    },
+    {
+      "epoch": 0.860873083880796,
+      "grad_norm": 1.020274468507894,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 36167
+    },
+    {
+      "epoch": 0.8608968866038275,
+      "grad_norm": 20.29631800842084,
+      "learning_rate": 2e-05,
+      "loss": 0.8129,
+      "step": 36168
+    },
+    {
+      "epoch": 0.860920689326859,
+      "grad_norm": 9.189095310048705,
+      "learning_rate": 2e-05,
+      "loss": 0.3116,
+      "step": 36169
+    },
+    {
+      "epoch": 0.8609444920498905,
+      "grad_norm": 4.436335106857873,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 36170
+    },
+    {
+      "epoch": 0.860968294772922,
+      "grad_norm": 2.321464326015162,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 36171
+    },
+    {
+      "epoch": 0.8609920974959535,
+      "grad_norm": 4.37591450757244,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 36172
+    },
+    {
+      "epoch": 0.861015900218985,
+      "grad_norm": 4.937422342077838,
+      "learning_rate": 2e-05,
+      "loss": 0.6677,
+      "step": 36173
+    },
+    {
+      "epoch": 0.8610397029420166,
+      "grad_norm": 5.459219639609067,
+      "learning_rate": 2e-05,
+      "loss": 0.6563,
+      "step": 36174
+    },
+    {
+      "epoch": 0.8610635056650481,
+      "grad_norm": 0.2411615989757579,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 36175
+    },
+    {
+      "epoch": 0.8610873083880796,
+      "grad_norm": 1.5495947205735094,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 36176
+    },
+    {
+      "epoch": 0.8611111111111112,
+      "grad_norm": 3.864877161428676,
+      "learning_rate": 2e-05,
+      "loss": 0.3048,
+      "step": 36177
+    },
+    {
+      "epoch": 0.8611349138341426,
+      "grad_norm": 2.592880914031397,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 36178
+    },
+    {
+      "epoch": 0.8611587165571741,
+      "grad_norm": 4.017183644780284,
+      "learning_rate": 2e-05,
+      "loss": 0.8113,
+      "step": 36179
+    },
+    {
+      "epoch": 0.8611825192802056,
+      "grad_norm": 0.04732511704331158,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 36180
+    },
+    {
+      "epoch": 0.8612063220032372,
+      "grad_norm": 2.4884142989708624,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 36181
+    },
+    {
+      "epoch": 0.8612301247262687,
+      "grad_norm": 0.8979256384305188,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 36182
+    },
+    {
+      "epoch": 0.8612539274493002,
+      "grad_norm": 2.543353620555958,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 36183
+    },
+    {
+      "epoch": 0.8612777301723317,
+      "grad_norm": 17.73191360292682,
+      "learning_rate": 2e-05,
+      "loss": 0.4569,
+      "step": 36184
+    },
+    {
+      "epoch": 0.8613015328953633,
+      "grad_norm": 1.0343377231788238,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 36185
+    },
+    {
+      "epoch": 0.8613253356183947,
+      "grad_norm": 2.5171270040170053,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 36186
+    },
+    {
+      "epoch": 0.8613491383414262,
+      "grad_norm": 1.9601402478289824,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 36187
+    },
+    {
+      "epoch": 0.8613729410644577,
+      "grad_norm": 0.3293187084694548,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 36188
+    },
+    {
+      "epoch": 0.8613967437874893,
+      "grad_norm": 27.787849488434574,
+      "learning_rate": 2e-05,
+      "loss": 1.2338,
+      "step": 36189
+    },
+    {
+      "epoch": 0.8614205465105208,
+      "grad_norm": 1.1048380312724992,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 36190
+    },
+    {
+      "epoch": 0.8614443492335523,
+      "grad_norm": 16.84292347675915,
+      "learning_rate": 2e-05,
+      "loss": 0.4019,
+      "step": 36191
+    },
+    {
+      "epoch": 0.8614681519565839,
+      "grad_norm": 4.495327720024364,
+      "learning_rate": 2e-05,
+      "loss": 0.6407,
+      "step": 36192
+    },
+    {
+      "epoch": 0.8614919546796154,
+      "grad_norm": 24.24241392026506,
+      "learning_rate": 2e-05,
+      "loss": 1.16,
+      "step": 36193
+    },
+    {
+      "epoch": 0.8615157574026469,
+      "grad_norm": 4.330629773501857,
+      "learning_rate": 2e-05,
+      "loss": 0.7992,
+      "step": 36194
+    },
+    {
+      "epoch": 0.8615395601256783,
+      "grad_norm": 7.343518197590532,
+      "learning_rate": 2e-05,
+      "loss": 0.9789,
+      "step": 36195
+    },
+    {
+      "epoch": 0.8615633628487099,
+      "grad_norm": 5.868345522867723,
+      "learning_rate": 2e-05,
+      "loss": 0.7652,
+      "step": 36196
+    },
+    {
+      "epoch": 0.8615871655717414,
+      "grad_norm": 9.06078668222071,
+      "learning_rate": 2e-05,
+      "loss": 0.6166,
+      "step": 36197
+    },
+    {
+      "epoch": 0.8616109682947729,
+      "grad_norm": 6.201707783513593,
+      "learning_rate": 2e-05,
+      "loss": 1.5086,
+      "step": 36198
+    },
+    {
+      "epoch": 0.8616347710178044,
+      "grad_norm": 0.12172475770749763,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 36199
+    },
+    {
+      "epoch": 0.861658573740836,
+      "grad_norm": 3.983848979790293,
+      "learning_rate": 2e-05,
+      "loss": 0.4878,
+      "step": 36200
+    },
+    {
+      "epoch": 0.8616823764638675,
+      "grad_norm": 0.4459713613694037,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 36201
+    },
+    {
+      "epoch": 0.861706179186899,
+      "grad_norm": 1.142877358048581,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 36202
+    },
+    {
+      "epoch": 0.8617299819099304,
+      "grad_norm": 14.272911843803671,
+      "learning_rate": 2e-05,
+      "loss": 1.5575,
+      "step": 36203
+    },
+    {
+      "epoch": 0.861753784632962,
+      "grad_norm": 5.367959655326051,
+      "learning_rate": 2e-05,
+      "loss": 0.1175,
+      "step": 36204
+    },
+    {
+      "epoch": 0.8617775873559935,
+      "grad_norm": 0.30278297215748273,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 36205
+    },
+    {
+      "epoch": 0.861801390079025,
+      "grad_norm": 2.329168575123335,
+      "learning_rate": 2e-05,
+      "loss": 0.3908,
+      "step": 36206
+    },
+    {
+      "epoch": 0.8618251928020566,
+      "grad_norm": 2.5684104444024056,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 36207
+    },
+    {
+      "epoch": 0.8618489955250881,
+      "grad_norm": 2.731419749623375,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 36208
+    },
+    {
+      "epoch": 0.8618727982481196,
+      "grad_norm": 3.1959931651115077,
+      "learning_rate": 2e-05,
+      "loss": 0.4703,
+      "step": 36209
+    },
+    {
+      "epoch": 0.8618966009711511,
+      "grad_norm": 8.83303259969136,
+      "learning_rate": 2e-05,
+      "loss": 0.598,
+      "step": 36210
+    },
+    {
+      "epoch": 0.8619204036941827,
+      "grad_norm": 3.1255471096131706,
+      "learning_rate": 2e-05,
+      "loss": 0.5669,
+      "step": 36211
+    },
+    {
+      "epoch": 0.8619442064172141,
+      "grad_norm": 0.4976294044403711,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 36212
+    },
+    {
+      "epoch": 0.8619680091402456,
+      "grad_norm": 1.5398855469525985,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 36213
+    },
+    {
+      "epoch": 0.8619918118632771,
+      "grad_norm": 5.528651246883543,
+      "learning_rate": 2e-05,
+      "loss": 0.4035,
+      "step": 36214
+    },
+    {
+      "epoch": 0.8620156145863087,
+      "grad_norm": 6.7759495018741465,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 36215
+    },
+    {
+      "epoch": 0.8620394173093402,
+      "grad_norm": 6.975997730950824,
+      "learning_rate": 2e-05,
+      "loss": 1.7885,
+      "step": 36216
+    },
+    {
+      "epoch": 0.8620632200323717,
+      "grad_norm": 16.600381004370202,
+      "learning_rate": 2e-05,
+      "loss": 0.3386,
+      "step": 36217
+    },
+    {
+      "epoch": 0.8620870227554032,
+      "grad_norm": 0.9429844557647048,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 36218
+    },
+    {
+      "epoch": 0.8621108254784348,
+      "grad_norm": 0.7296005740443916,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 36219
+    },
+    {
+      "epoch": 0.8621346282014662,
+      "grad_norm": 4.557103297359375,
+      "learning_rate": 2e-05,
+      "loss": 1.0463,
+      "step": 36220
+    },
+    {
+      "epoch": 0.8621584309244977,
+      "grad_norm": 5.072589113570847,
+      "learning_rate": 2e-05,
+      "loss": 0.1584,
+      "step": 36221
+    },
+    {
+      "epoch": 0.8621822336475293,
+      "grad_norm": 0.20974130565575697,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 36222
+    },
+    {
+      "epoch": 0.8622060363705608,
+      "grad_norm": 10.731055049701288,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 36223
+    },
+    {
+      "epoch": 0.8622298390935923,
+      "grad_norm": 0.011771561777735218,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 36224
+    },
+    {
+      "epoch": 0.8622536418166238,
+      "grad_norm": 2.5244754713921393,
+      "learning_rate": 2e-05,
+      "loss": 0.2307,
+      "step": 36225
+    },
+    {
+      "epoch": 0.8622774445396554,
+      "grad_norm": 6.124862634419528,
+      "learning_rate": 2e-05,
+      "loss": 0.6704,
+      "step": 36226
+    },
+    {
+      "epoch": 0.8623012472626869,
+      "grad_norm": 0.07201382324867457,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 36227
+    },
+    {
+      "epoch": 0.8623250499857184,
+      "grad_norm": 5.146381907695536,
+      "learning_rate": 2e-05,
+      "loss": 0.7642,
+      "step": 36228
+    },
+    {
+      "epoch": 0.8623488527087498,
+      "grad_norm": 1.2297788477816503,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 36229
+    },
+    {
+      "epoch": 0.8623726554317814,
+      "grad_norm": 20.59682467776704,
+      "learning_rate": 2e-05,
+      "loss": 0.1372,
+      "step": 36230
+    },
+    {
+      "epoch": 0.8623964581548129,
+      "grad_norm": 6.463751269996751,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 36231
+    },
+    {
+      "epoch": 0.8624202608778444,
+      "grad_norm": 11.129328664595464,
+      "learning_rate": 2e-05,
+      "loss": 1.3666,
+      "step": 36232
+    },
+    {
+      "epoch": 0.8624440636008759,
+      "grad_norm": 1.0638496310352863,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 36233
+    },
+    {
+      "epoch": 0.8624678663239075,
+      "grad_norm": 2.3144524565371074,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 36234
+    },
+    {
+      "epoch": 0.862491669046939,
+      "grad_norm": 7.204010002372502,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 36235
+    },
+    {
+      "epoch": 0.8625154717699705,
+      "grad_norm": 2.804129681325245,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 36236
+    },
+    {
+      "epoch": 0.862539274493002,
+      "grad_norm": 3.825178836829496,
+      "learning_rate": 2e-05,
+      "loss": 0.8696,
+      "step": 36237
+    },
+    {
+      "epoch": 0.8625630772160335,
+      "grad_norm": 0.03770828334351731,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 36238
+    },
+    {
+      "epoch": 0.862586879939065,
+      "grad_norm": 2.968111105670589,
+      "learning_rate": 2e-05,
+      "loss": 0.5704,
+      "step": 36239
+    },
+    {
+      "epoch": 0.8626106826620965,
+      "grad_norm": 4.591492025984788,
+      "learning_rate": 2e-05,
+      "loss": 0.6876,
+      "step": 36240
+    },
+    {
+      "epoch": 0.8626344853851281,
+      "grad_norm": 0.06675816779432554,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 36241
+    },
+    {
+      "epoch": 0.8626582881081596,
+      "grad_norm": 3.309631305922026,
+      "learning_rate": 2e-05,
+      "loss": 0.4336,
+      "step": 36242
+    },
+    {
+      "epoch": 0.8626820908311911,
+      "grad_norm": 8.584944631003356,
+      "learning_rate": 2e-05,
+      "loss": 1.2995,
+      "step": 36243
+    },
+    {
+      "epoch": 0.8627058935542226,
+      "grad_norm": 3.1742957567146024,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 36244
+    },
+    {
+      "epoch": 0.8627296962772542,
+      "grad_norm": 4.861694189025564,
+      "learning_rate": 2e-05,
+      "loss": 0.6755,
+      "step": 36245
+    },
+    {
+      "epoch": 0.8627534990002856,
+      "grad_norm": 0.003349971111763994,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 36246
+    },
+    {
+      "epoch": 0.8627773017233171,
+      "grad_norm": 3.371391174536634,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 36247
+    },
+    {
+      "epoch": 0.8628011044463486,
+      "grad_norm": 2.607958714419176,
+      "learning_rate": 2e-05,
+      "loss": 0.4052,
+      "step": 36248
+    },
+    {
+      "epoch": 0.8628249071693802,
+      "grad_norm": 0.31390877347374363,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36249
+    },
+    {
+      "epoch": 0.8628487098924117,
+      "grad_norm": 5.762181878995875,
+      "learning_rate": 2e-05,
+      "loss": 0.714,
+      "step": 36250
+    },
+    {
+      "epoch": 0.8628725126154432,
+      "grad_norm": 5.426402550134944,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 36251
+    },
+    {
+      "epoch": 0.8628963153384748,
+      "grad_norm": 3.43695820960567,
+      "learning_rate": 2e-05,
+      "loss": 0.3139,
+      "step": 36252
+    },
+    {
+      "epoch": 0.8629201180615063,
+      "grad_norm": 0.6608283181811153,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 36253
+    },
+    {
+      "epoch": 0.8629439207845377,
+      "grad_norm": 0.4246543360253178,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 36254
+    },
+    {
+      "epoch": 0.8629677235075692,
+      "grad_norm": 5.290177028567443,
+      "learning_rate": 2e-05,
+      "loss": 1.2151,
+      "step": 36255
+    },
+    {
+      "epoch": 0.8629915262306008,
+      "grad_norm": 16.66467452680626,
+      "learning_rate": 2e-05,
+      "loss": 0.3781,
+      "step": 36256
+    },
+    {
+      "epoch": 0.8630153289536323,
+      "grad_norm": 5.908302516680597,
+      "learning_rate": 2e-05,
+      "loss": 1.2882,
+      "step": 36257
+    },
+    {
+      "epoch": 0.8630391316766638,
+      "grad_norm": 0.20079706967540165,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 36258
+    },
+    {
+      "epoch": 0.8630629343996953,
+      "grad_norm": 5.359502690442665,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 36259
+    },
+    {
+      "epoch": 0.8630867371227269,
+      "grad_norm": 4.1138554303201555,
+      "learning_rate": 2e-05,
+      "loss": 0.4569,
+      "step": 36260
+    },
+    {
+      "epoch": 0.8631105398457584,
+      "grad_norm": 59.17873782217985,
+      "learning_rate": 2e-05,
+      "loss": 0.3583,
+      "step": 36261
+    },
+    {
+      "epoch": 0.8631343425687898,
+      "grad_norm": 7.289861952097331,
+      "learning_rate": 2e-05,
+      "loss": 0.6461,
+      "step": 36262
+    },
+    {
+      "epoch": 0.8631581452918213,
+      "grad_norm": 6.323186042947515,
+      "learning_rate": 2e-05,
+      "loss": 0.9433,
+      "step": 36263
+    },
+    {
+      "epoch": 0.8631819480148529,
+      "grad_norm": 1.7168849408230917,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 36264
+    },
+    {
+      "epoch": 0.8632057507378844,
+      "grad_norm": 5.873090863486159,
+      "learning_rate": 2e-05,
+      "loss": 0.6629,
+      "step": 36265
+    },
+    {
+      "epoch": 0.8632295534609159,
+      "grad_norm": 1.0947886755437424,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 36266
+    },
+    {
+      "epoch": 0.8632533561839475,
+      "grad_norm": 0.7400904366376498,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 36267
+    },
+    {
+      "epoch": 0.863277158906979,
+      "grad_norm": 1.9177410737499876,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 36268
+    },
+    {
+      "epoch": 0.8633009616300105,
+      "grad_norm": 4.502484060605704,
+      "learning_rate": 2e-05,
+      "loss": 0.6336,
+      "step": 36269
+    },
+    {
+      "epoch": 0.863324764353042,
+      "grad_norm": 11.77090778054456,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 36270
+    },
+    {
+      "epoch": 0.8633485670760735,
+      "grad_norm": 3.494006901367899,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 36271
+    },
+    {
+      "epoch": 0.863372369799105,
+      "grad_norm": 9.377109402082576,
+      "learning_rate": 2e-05,
+      "loss": 0.1655,
+      "step": 36272
+    },
+    {
+      "epoch": 0.8633961725221365,
+      "grad_norm": 1.309939288500583,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 36273
+    },
+    {
+      "epoch": 0.863419975245168,
+      "grad_norm": 5.6722092652376395,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 36274
+    },
+    {
+      "epoch": 0.8634437779681996,
+      "grad_norm": 4.0487347262579245,
+      "learning_rate": 2e-05,
+      "loss": 0.5194,
+      "step": 36275
+    },
+    {
+      "epoch": 0.8634675806912311,
+      "grad_norm": 3.5142580078547865,
+      "learning_rate": 2e-05,
+      "loss": 0.6956,
+      "step": 36276
+    },
+    {
+      "epoch": 0.8634913834142626,
+      "grad_norm": 0.08021611770638423,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 36277
+    },
+    {
+      "epoch": 0.863515186137294,
+      "grad_norm": 4.612590334539514,
+      "learning_rate": 2e-05,
+      "loss": 0.5706,
+      "step": 36278
+    },
+    {
+      "epoch": 0.8635389888603257,
+      "grad_norm": 3.1903758972075202,
+      "learning_rate": 2e-05,
+      "loss": 0.262,
+      "step": 36279
+    },
+    {
+      "epoch": 0.8635627915833571,
+      "grad_norm": 9.198754525551342,
+      "learning_rate": 2e-05,
+      "loss": 2.0905,
+      "step": 36280
+    },
+    {
+      "epoch": 0.8635865943063886,
+      "grad_norm": 3.518891387973406,
+      "learning_rate": 2e-05,
+      "loss": 0.2333,
+      "step": 36281
+    },
+    {
+      "epoch": 0.8636103970294202,
+      "grad_norm": 3.518582784456748,
+      "learning_rate": 2e-05,
+      "loss": 0.5529,
+      "step": 36282
+    },
+    {
+      "epoch": 0.8636341997524517,
+      "grad_norm": 0.2668829579104195,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 36283
+    },
+    {
+      "epoch": 0.8636580024754832,
+      "grad_norm": 2.695081408217004,
+      "learning_rate": 2e-05,
+      "loss": 0.3054,
+      "step": 36284
+    },
+    {
+      "epoch": 0.8636818051985147,
+      "grad_norm": 6.439540609675753,
+      "learning_rate": 2e-05,
+      "loss": 0.5532,
+      "step": 36285
+    },
+    {
+      "epoch": 0.8637056079215463,
+      "grad_norm": 3.0037500627820273,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 36286
+    },
+    {
+      "epoch": 0.8637294106445778,
+      "grad_norm": 4.7279320844004635,
+      "learning_rate": 2e-05,
+      "loss": 0.2759,
+      "step": 36287
+    },
+    {
+      "epoch": 0.8637532133676092,
+      "grad_norm": 4.790963635668994,
+      "learning_rate": 2e-05,
+      "loss": 0.7193,
+      "step": 36288
+    },
+    {
+      "epoch": 0.8637770160906407,
+      "grad_norm": 0.41465476341189045,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 36289
+    },
+    {
+      "epoch": 0.8638008188136723,
+      "grad_norm": 10.779594258453916,
+      "learning_rate": 2e-05,
+      "loss": 0.7355,
+      "step": 36290
+    },
+    {
+      "epoch": 0.8638246215367038,
+      "grad_norm": 4.630566150577197,
+      "learning_rate": 2e-05,
+      "loss": 0.2582,
+      "step": 36291
+    },
+    {
+      "epoch": 0.8638484242597353,
+      "grad_norm": 29.50758699863491,
+      "learning_rate": 2e-05,
+      "loss": 1.5294,
+      "step": 36292
+    },
+    {
+      "epoch": 0.8638722269827668,
+      "grad_norm": 0.7752322646386426,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 36293
+    },
+    {
+      "epoch": 0.8638960297057984,
+      "grad_norm": 0.1656785975106883,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 36294
+    },
+    {
+      "epoch": 0.8639198324288299,
+      "grad_norm": 0.14929389693986375,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36295
+    },
+    {
+      "epoch": 0.8639436351518613,
+      "grad_norm": 28.424403378206215,
+      "learning_rate": 2e-05,
+      "loss": 0.6297,
+      "step": 36296
+    },
+    {
+      "epoch": 0.8639674378748929,
+      "grad_norm": 4.3901358334245035,
+      "learning_rate": 2e-05,
+      "loss": 0.8994,
+      "step": 36297
+    },
+    {
+      "epoch": 0.8639912405979244,
+      "grad_norm": 4.425704288681115,
+      "learning_rate": 2e-05,
+      "loss": 1.3108,
+      "step": 36298
+    },
+    {
+      "epoch": 0.8640150433209559,
+      "grad_norm": 3.487253091745929,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 36299
+    },
+    {
+      "epoch": 0.8640388460439874,
+      "grad_norm": 4.862653880999262,
+      "learning_rate": 2e-05,
+      "loss": 0.4619,
+      "step": 36300
+    },
+    {
+      "epoch": 0.864062648767019,
+      "grad_norm": 7.0973328822972865,
+      "learning_rate": 2e-05,
+      "loss": 0.2689,
+      "step": 36301
+    },
+    {
+      "epoch": 0.8640864514900505,
+      "grad_norm": 1.2366410064391118,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 36302
+    },
+    {
+      "epoch": 0.864110254213082,
+      "grad_norm": 2.293616735398725,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 36303
+    },
+    {
+      "epoch": 0.8641340569361134,
+      "grad_norm": 1.846417049104402,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 36304
+    },
+    {
+      "epoch": 0.864157859659145,
+      "grad_norm": 12.601108008820354,
+      "learning_rate": 2e-05,
+      "loss": 0.9968,
+      "step": 36305
+    },
+    {
+      "epoch": 0.8641816623821765,
+      "grad_norm": 4.959983248531126,
+      "learning_rate": 2e-05,
+      "loss": 0.7663,
+      "step": 36306
+    },
+    {
+      "epoch": 0.864205465105208,
+      "grad_norm": 4.571268144224658,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 36307
+    },
+    {
+      "epoch": 0.8642292678282395,
+      "grad_norm": 5.726244523019586,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 36308
+    },
+    {
+      "epoch": 0.8642530705512711,
+      "grad_norm": 3.143262688457565,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 36309
+    },
+    {
+      "epoch": 0.8642768732743026,
+      "grad_norm": 0.13239061427411808,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 36310
+    },
+    {
+      "epoch": 0.8643006759973341,
+      "grad_norm": 0.4460264830026638,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 36311
+    },
+    {
+      "epoch": 0.8643244787203657,
+      "grad_norm": 3.0114040033454272,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 36312
+    },
+    {
+      "epoch": 0.8643482814433971,
+      "grad_norm": 6.164286642045479,
+      "learning_rate": 2e-05,
+      "loss": 1.6394,
+      "step": 36313
+    },
+    {
+      "epoch": 0.8643720841664286,
+      "grad_norm": 7.5927140790055185,
+      "learning_rate": 2e-05,
+      "loss": 0.4727,
+      "step": 36314
+    },
+    {
+      "epoch": 0.8643958868894601,
+      "grad_norm": 5.194398988203715,
+      "learning_rate": 2e-05,
+      "loss": 0.7815,
+      "step": 36315
+    },
+    {
+      "epoch": 0.8644196896124917,
+      "grad_norm": 0.14701964661985154,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 36316
+    },
+    {
+      "epoch": 0.8644434923355232,
+      "grad_norm": 3.179484884119077,
+      "learning_rate": 2e-05,
+      "loss": 0.8367,
+      "step": 36317
+    },
+    {
+      "epoch": 0.8644672950585547,
+      "grad_norm": 3.0573443315473035,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 36318
+    },
+    {
+      "epoch": 0.8644910977815862,
+      "grad_norm": 16.893266037655522,
+      "learning_rate": 2e-05,
+      "loss": 0.9477,
+      "step": 36319
+    },
+    {
+      "epoch": 0.8645149005046178,
+      "grad_norm": 3.5185785909585787,
+      "learning_rate": 2e-05,
+      "loss": 0.5096,
+      "step": 36320
+    },
+    {
+      "epoch": 0.8645387032276493,
+      "grad_norm": 2.1215738725077946,
+      "learning_rate": 2e-05,
+      "loss": 0.0399,
+      "step": 36321
+    },
+    {
+      "epoch": 0.8645625059506807,
+      "grad_norm": 7.150140018582568,
+      "learning_rate": 2e-05,
+      "loss": 0.3702,
+      "step": 36322
+    },
+    {
+      "epoch": 0.8645863086737122,
+      "grad_norm": 4.834794665296837,
+      "learning_rate": 2e-05,
+      "loss": 0.8962,
+      "step": 36323
+    },
+    {
+      "epoch": 0.8646101113967438,
+      "grad_norm": 3.322864884540463,
+      "learning_rate": 2e-05,
+      "loss": 0.5211,
+      "step": 36324
+    },
+    {
+      "epoch": 0.8646339141197753,
+      "grad_norm": 5.899698032948008,
+      "learning_rate": 2e-05,
+      "loss": 0.6722,
+      "step": 36325
+    },
+    {
+      "epoch": 0.8646577168428068,
+      "grad_norm": 5.031034000599188,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 36326
+    },
+    {
+      "epoch": 0.8646815195658384,
+      "grad_norm": 2.35986024967966,
+      "learning_rate": 2e-05,
+      "loss": 0.3534,
+      "step": 36327
+    },
+    {
+      "epoch": 0.8647053222888699,
+      "grad_norm": 13.839320182559973,
+      "learning_rate": 2e-05,
+      "loss": 0.6064,
+      "step": 36328
+    },
+    {
+      "epoch": 0.8647291250119014,
+      "grad_norm": 2.2427448743713674,
+      "learning_rate": 2e-05,
+      "loss": 0.1248,
+      "step": 36329
+    },
+    {
+      "epoch": 0.8647529277349328,
+      "grad_norm": 0.8679447290204623,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 36330
+    },
+    {
+      "epoch": 0.8647767304579644,
+      "grad_norm": 4.337935003384305,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 36331
+    },
+    {
+      "epoch": 0.8648005331809959,
+      "grad_norm": 3.7332309674557034,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 36332
+    },
+    {
+      "epoch": 0.8648243359040274,
+      "grad_norm": 2.5444556412889363,
+      "learning_rate": 2e-05,
+      "loss": 0.06,
+      "step": 36333
+    },
+    {
+      "epoch": 0.8648481386270589,
+      "grad_norm": 0.6919354919525569,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 36334
+    },
+    {
+      "epoch": 0.8648719413500905,
+      "grad_norm": 2.81602794561522,
+      "learning_rate": 2e-05,
+      "loss": 0.2322,
+      "step": 36335
+    },
+    {
+      "epoch": 0.864895744073122,
+      "grad_norm": 6.266440381021138,
+      "learning_rate": 2e-05,
+      "loss": 0.2307,
+      "step": 36336
+    },
+    {
+      "epoch": 0.8649195467961535,
+      "grad_norm": 6.152639336971847,
+      "learning_rate": 2e-05,
+      "loss": 1.0535,
+      "step": 36337
+    },
+    {
+      "epoch": 0.864943349519185,
+      "grad_norm": 4.488883546622141,
+      "learning_rate": 2e-05,
+      "loss": 0.2534,
+      "step": 36338
+    },
+    {
+      "epoch": 0.8649671522422165,
+      "grad_norm": 0.753057774616247,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 36339
+    },
+    {
+      "epoch": 0.864990954965248,
+      "grad_norm": 11.814231774639241,
+      "learning_rate": 2e-05,
+      "loss": 0.9238,
+      "step": 36340
+    },
+    {
+      "epoch": 0.8650147576882795,
+      "grad_norm": 3.418069808364775,
+      "learning_rate": 2e-05,
+      "loss": 0.6138,
+      "step": 36341
+    },
+    {
+      "epoch": 0.8650385604113111,
+      "grad_norm": 0.13120018269109665,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 36342
+    },
+    {
+      "epoch": 0.8650623631343426,
+      "grad_norm": 6.315419016468323,
+      "learning_rate": 2e-05,
+      "loss": 1.171,
+      "step": 36343
+    },
+    {
+      "epoch": 0.8650861658573741,
+      "grad_norm": 4.06857605241794,
+      "learning_rate": 2e-05,
+      "loss": 0.5919,
+      "step": 36344
+    },
+    {
+      "epoch": 0.8651099685804056,
+      "grad_norm": 1.0658005932076013,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 36345
+    },
+    {
+      "epoch": 0.8651337713034372,
+      "grad_norm": 0.22596333716899267,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 36346
+    },
+    {
+      "epoch": 0.8651575740264686,
+      "grad_norm": 16.895969497164945,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 36347
+    },
+    {
+      "epoch": 0.8651813767495001,
+      "grad_norm": 0.24016562509685693,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 36348
+    },
+    {
+      "epoch": 0.8652051794725316,
+      "grad_norm": 4.439849894112373,
+      "learning_rate": 2e-05,
+      "loss": 0.2736,
+      "step": 36349
+    },
+    {
+      "epoch": 0.8652289821955632,
+      "grad_norm": 4.08329096968654,
+      "learning_rate": 2e-05,
+      "loss": 0.5288,
+      "step": 36350
+    },
+    {
+      "epoch": 0.8652527849185947,
+      "grad_norm": 9.58746992544892,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 36351
+    },
+    {
+      "epoch": 0.8652765876416262,
+      "grad_norm": 0.329016131365902,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 36352
+    },
+    {
+      "epoch": 0.8653003903646577,
+      "grad_norm": 1.1545598324361237,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 36353
+    },
+    {
+      "epoch": 0.8653241930876893,
+      "grad_norm": 3.0551852497921286,
+      "learning_rate": 2e-05,
+      "loss": 0.4342,
+      "step": 36354
+    },
+    {
+      "epoch": 0.8653479958107207,
+      "grad_norm": 0.3878823563323843,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 36355
+    },
+    {
+      "epoch": 0.8653717985337522,
+      "grad_norm": 9.341363188035581,
+      "learning_rate": 2e-05,
+      "loss": 0.4178,
+      "step": 36356
+    },
+    {
+      "epoch": 0.8653956012567838,
+      "grad_norm": 3.8177406605513657,
+      "learning_rate": 2e-05,
+      "loss": 0.5276,
+      "step": 36357
+    },
+    {
+      "epoch": 0.8654194039798153,
+      "grad_norm": 2.488743731433257,
+      "learning_rate": 2e-05,
+      "loss": 0.2671,
+      "step": 36358
+    },
+    {
+      "epoch": 0.8654432067028468,
+      "grad_norm": 1.3841675631614927,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 36359
+    },
+    {
+      "epoch": 0.8654670094258783,
+      "grad_norm": 4.076470720777101,
+      "learning_rate": 2e-05,
+      "loss": 0.6947,
+      "step": 36360
+    },
+    {
+      "epoch": 0.8654908121489099,
+      "grad_norm": 2.6766644455049526,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 36361
+    },
+    {
+      "epoch": 0.8655146148719414,
+      "grad_norm": 1.3915711725819209,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 36362
+    },
+    {
+      "epoch": 0.8655384175949729,
+      "grad_norm": 2.827302563105401,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 36363
+    },
+    {
+      "epoch": 0.8655622203180043,
+      "grad_norm": 0.37770510674287605,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 36364
+    },
+    {
+      "epoch": 0.8655860230410359,
+      "grad_norm": 5.517160449359562,
+      "learning_rate": 2e-05,
+      "loss": 1.053,
+      "step": 36365
+    },
+    {
+      "epoch": 0.8656098257640674,
+      "grad_norm": 2.6744245756535405,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 36366
+    },
+    {
+      "epoch": 0.8656336284870989,
+      "grad_norm": 13.163405515819845,
+      "learning_rate": 2e-05,
+      "loss": 0.1917,
+      "step": 36367
+    },
+    {
+      "epoch": 0.8656574312101304,
+      "grad_norm": 0.1160809834617041,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36368
+    },
+    {
+      "epoch": 0.865681233933162,
+      "grad_norm": 14.480161872607193,
+      "learning_rate": 2e-05,
+      "loss": 0.6929,
+      "step": 36369
+    },
+    {
+      "epoch": 0.8657050366561935,
+      "grad_norm": 5.794287629339558,
+      "learning_rate": 2e-05,
+      "loss": 1.1257,
+      "step": 36370
+    },
+    {
+      "epoch": 0.865728839379225,
+      "grad_norm": 3.6363407158189043,
+      "learning_rate": 2e-05,
+      "loss": 0.3027,
+      "step": 36371
+    },
+    {
+      "epoch": 0.8657526421022566,
+      "grad_norm": 2.8951978234137976,
+      "learning_rate": 2e-05,
+      "loss": 0.1232,
+      "step": 36372
+    },
+    {
+      "epoch": 0.865776444825288,
+      "grad_norm": 5.17095526915249,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 36373
+    },
+    {
+      "epoch": 0.8658002475483195,
+      "grad_norm": 1.7233231242415914,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 36374
+    },
+    {
+      "epoch": 0.865824050271351,
+      "grad_norm": 0.07640427026543423,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36375
+    },
+    {
+      "epoch": 0.8658478529943826,
+      "grad_norm": 0.0915721599929691,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 36376
+    },
+    {
+      "epoch": 0.8658716557174141,
+      "grad_norm": 3.8018412153752927,
+      "learning_rate": 2e-05,
+      "loss": 0.729,
+      "step": 36377
+    },
+    {
+      "epoch": 0.8658954584404456,
+      "grad_norm": 22.389322225525923,
+      "learning_rate": 2e-05,
+      "loss": 0.1898,
+      "step": 36378
+    },
+    {
+      "epoch": 0.8659192611634771,
+      "grad_norm": 5.107852619615018,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 36379
+    },
+    {
+      "epoch": 0.8659430638865087,
+      "grad_norm": 3.597753023171409,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 36380
+    },
+    {
+      "epoch": 0.8659668666095401,
+      "grad_norm": 3.489870727868312,
+      "learning_rate": 2e-05,
+      "loss": 0.5211,
+      "step": 36381
+    },
+    {
+      "epoch": 0.8659906693325716,
+      "grad_norm": 19.362949789131992,
+      "learning_rate": 2e-05,
+      "loss": 0.8059,
+      "step": 36382
+    },
+    {
+      "epoch": 0.8660144720556031,
+      "grad_norm": 0.08224778425100401,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 36383
+    },
+    {
+      "epoch": 0.8660382747786347,
+      "grad_norm": 6.480203000585137,
+      "learning_rate": 2e-05,
+      "loss": 0.6028,
+      "step": 36384
+    },
+    {
+      "epoch": 0.8660620775016662,
+      "grad_norm": 0.017195014198985285,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36385
+    },
+    {
+      "epoch": 0.8660858802246977,
+      "grad_norm": 23.0543428771101,
+      "learning_rate": 2e-05,
+      "loss": 0.7671,
+      "step": 36386
+    },
+    {
+      "epoch": 0.8661096829477292,
+      "grad_norm": 6.978724337774518,
+      "learning_rate": 2e-05,
+      "loss": 0.0955,
+      "step": 36387
+    },
+    {
+      "epoch": 0.8661334856707608,
+      "grad_norm": 2.059681995595473,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 36388
+    },
+    {
+      "epoch": 0.8661572883937922,
+      "grad_norm": 1.334044109989297,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 36389
+    },
+    {
+      "epoch": 0.8661810911168237,
+      "grad_norm": 1.7083673035054425,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 36390
+    },
+    {
+      "epoch": 0.8662048938398553,
+      "grad_norm": 8.500342481735847,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 36391
+    },
+    {
+      "epoch": 0.8662286965628868,
+      "grad_norm": 2.7658820765859313,
+      "learning_rate": 2e-05,
+      "loss": 0.3352,
+      "step": 36392
+    },
+    {
+      "epoch": 0.8662524992859183,
+      "grad_norm": 3.7623849273524357,
+      "learning_rate": 2e-05,
+      "loss": 0.6617,
+      "step": 36393
+    },
+    {
+      "epoch": 0.8662763020089498,
+      "grad_norm": 7.506302883029014,
+      "learning_rate": 2e-05,
+      "loss": 0.6151,
+      "step": 36394
+    },
+    {
+      "epoch": 0.8663001047319814,
+      "grad_norm": 0.02760028975368459,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 36395
+    },
+    {
+      "epoch": 0.8663239074550129,
+      "grad_norm": 3.8905343093998437,
+      "learning_rate": 2e-05,
+      "loss": 0.0853,
+      "step": 36396
+    },
+    {
+      "epoch": 0.8663477101780444,
+      "grad_norm": 1.0402650582498822,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 36397
+    },
+    {
+      "epoch": 0.8663715129010758,
+      "grad_norm": 4.060471619621193,
+      "learning_rate": 2e-05,
+      "loss": 0.91,
+      "step": 36398
+    },
+    {
+      "epoch": 0.8663953156241074,
+      "grad_norm": 3.386034625867967,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 36399
+    },
+    {
+      "epoch": 0.8664191183471389,
+      "grad_norm": 1.289325732274304,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 36400
+    },
+    {
+      "epoch": 0.8664429210701704,
+      "grad_norm": 2.496561708024851,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 36401
+    },
+    {
+      "epoch": 0.8664667237932019,
+      "grad_norm": 4.5639361353735195,
+      "learning_rate": 2e-05,
+      "loss": 0.5379,
+      "step": 36402
+    },
+    {
+      "epoch": 0.8664905265162335,
+      "grad_norm": 0.5615214841679749,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 36403
+    },
+    {
+      "epoch": 0.866514329239265,
+      "grad_norm": 4.939794660290135,
+      "learning_rate": 2e-05,
+      "loss": 1.0867,
+      "step": 36404
+    },
+    {
+      "epoch": 0.8665381319622965,
+      "grad_norm": 10.082843207654246,
+      "learning_rate": 2e-05,
+      "loss": 0.9892,
+      "step": 36405
+    },
+    {
+      "epoch": 0.866561934685328,
+      "grad_norm": 8.222526578480386,
+      "learning_rate": 2e-05,
+      "loss": 0.6078,
+      "step": 36406
+    },
+    {
+      "epoch": 0.8665857374083595,
+      "grad_norm": 3.83121386691066,
+      "learning_rate": 2e-05,
+      "loss": 0.6871,
+      "step": 36407
+    },
+    {
+      "epoch": 0.866609540131391,
+      "grad_norm": 1.0356349319824276,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 36408
+    },
+    {
+      "epoch": 0.8666333428544225,
+      "grad_norm": 4.220160544975298,
+      "learning_rate": 2e-05,
+      "loss": 0.8732,
+      "step": 36409
+    },
+    {
+      "epoch": 0.8666571455774541,
+      "grad_norm": 0.1438886812695031,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 36410
+    },
+    {
+      "epoch": 0.8666809483004856,
+      "grad_norm": 4.950216232064276,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 36411
+    },
+    {
+      "epoch": 0.8667047510235171,
+      "grad_norm": 4.963015455023114,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 36412
+    },
+    {
+      "epoch": 0.8667285537465486,
+      "grad_norm": 10.752160645369976,
+      "learning_rate": 2e-05,
+      "loss": 1.0111,
+      "step": 36413
+    },
+    {
+      "epoch": 0.8667523564695802,
+      "grad_norm": 1.789735851046215,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 36414
+    },
+    {
+      "epoch": 0.8667761591926116,
+      "grad_norm": 3.7891417077318357,
+      "learning_rate": 2e-05,
+      "loss": 0.8116,
+      "step": 36415
+    },
+    {
+      "epoch": 0.8667999619156431,
+      "grad_norm": 3.664573703123361,
+      "learning_rate": 2e-05,
+      "loss": 0.6201,
+      "step": 36416
+    },
+    {
+      "epoch": 0.8668237646386746,
+      "grad_norm": 0.6433763766844546,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 36417
+    },
+    {
+      "epoch": 0.8668475673617062,
+      "grad_norm": 2.2652561208171518,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 36418
+    },
+    {
+      "epoch": 0.8668713700847377,
+      "grad_norm": 0.09731466656072042,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36419
+    },
+    {
+      "epoch": 0.8668951728077692,
+      "grad_norm": 2.6467558217427416,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 36420
+    },
+    {
+      "epoch": 0.8669189755308008,
+      "grad_norm": 3.8897521421230703,
+      "learning_rate": 2e-05,
+      "loss": 0.9128,
+      "step": 36421
+    },
+    {
+      "epoch": 0.8669427782538323,
+      "grad_norm": 7.133344758149175,
+      "learning_rate": 2e-05,
+      "loss": 1.3471,
+      "step": 36422
+    },
+    {
+      "epoch": 0.8669665809768637,
+      "grad_norm": 4.593140122665637,
+      "learning_rate": 2e-05,
+      "loss": 0.5379,
+      "step": 36423
+    },
+    {
+      "epoch": 0.8669903836998952,
+      "grad_norm": 3.3588523662125174,
+      "learning_rate": 2e-05,
+      "loss": 0.169,
+      "step": 36424
+    },
+    {
+      "epoch": 0.8670141864229268,
+      "grad_norm": 3.551992014685399,
+      "learning_rate": 2e-05,
+      "loss": 0.095,
+      "step": 36425
+    },
+    {
+      "epoch": 0.8670379891459583,
+      "grad_norm": 3.7821916142331116,
+      "learning_rate": 2e-05,
+      "loss": 0.5231,
+      "step": 36426
+    },
+    {
+      "epoch": 0.8670617918689898,
+      "grad_norm": 3.188776393718036,
+      "learning_rate": 2e-05,
+      "loss": 0.6765,
+      "step": 36427
+    },
+    {
+      "epoch": 0.8670855945920213,
+      "grad_norm": 4.20434094386821,
+      "learning_rate": 2e-05,
+      "loss": 0.6612,
+      "step": 36428
+    },
+    {
+      "epoch": 0.8671093973150529,
+      "grad_norm": 3.1188638557830815,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 36429
+    },
+    {
+      "epoch": 0.8671332000380844,
+      "grad_norm": 0.1715669845771293,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 36430
+    },
+    {
+      "epoch": 0.8671570027611158,
+      "grad_norm": 3.5933926403809893,
+      "learning_rate": 2e-05,
+      "loss": 0.786,
+      "step": 36431
+    },
+    {
+      "epoch": 0.8671808054841473,
+      "grad_norm": 3.6251086712682974,
+      "learning_rate": 2e-05,
+      "loss": 0.7821,
+      "step": 36432
+    },
+    {
+      "epoch": 0.8672046082071789,
+      "grad_norm": 5.20745313825175,
+      "learning_rate": 2e-05,
+      "loss": 0.8211,
+      "step": 36433
+    },
+    {
+      "epoch": 0.8672284109302104,
+      "grad_norm": 27.218985074334423,
+      "learning_rate": 2e-05,
+      "loss": 0.232,
+      "step": 36434
+    },
+    {
+      "epoch": 0.8672522136532419,
+      "grad_norm": 1.7594308239151724,
+      "learning_rate": 2e-05,
+      "loss": 0.1434,
+      "step": 36435
+    },
+    {
+      "epoch": 0.8672760163762735,
+      "grad_norm": 5.829059039025137,
+      "learning_rate": 2e-05,
+      "loss": 0.5749,
+      "step": 36436
+    },
+    {
+      "epoch": 0.867299819099305,
+      "grad_norm": 0.019203073852444363,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36437
+    },
+    {
+      "epoch": 0.8673236218223365,
+      "grad_norm": 0.21513424385716473,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 36438
+    },
+    {
+      "epoch": 0.867347424545368,
+      "grad_norm": 9.804600003828067,
+      "learning_rate": 2e-05,
+      "loss": 1.1263,
+      "step": 36439
+    },
+    {
+      "epoch": 0.8673712272683995,
+      "grad_norm": 0.03406321601651682,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36440
+    },
+    {
+      "epoch": 0.867395029991431,
+      "grad_norm": 2.5684122000061023,
+      "learning_rate": 2e-05,
+      "loss": 0.369,
+      "step": 36441
+    },
+    {
+      "epoch": 0.8674188327144625,
+      "grad_norm": 2.6068425144547556,
+      "learning_rate": 2e-05,
+      "loss": 0.3473,
+      "step": 36442
+    },
+    {
+      "epoch": 0.867442635437494,
+      "grad_norm": 0.15753131236712495,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 36443
+    },
+    {
+      "epoch": 0.8674664381605256,
+      "grad_norm": 0.30330804274217016,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 36444
+    },
+    {
+      "epoch": 0.8674902408835571,
+      "grad_norm": 9.694420470675574,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 36445
+    },
+    {
+      "epoch": 0.8675140436065886,
+      "grad_norm": 0.04538575367729404,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36446
+    },
+    {
+      "epoch": 0.86753784632962,
+      "grad_norm": 0.18093976156391628,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 36447
+    },
+    {
+      "epoch": 0.8675616490526517,
+      "grad_norm": 1.4720245098456215,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 36448
+    },
+    {
+      "epoch": 0.8675854517756831,
+      "grad_norm": 10.60753248203875,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 36449
+    },
+    {
+      "epoch": 0.8676092544987146,
+      "grad_norm": 4.5641295605758465,
+      "learning_rate": 2e-05,
+      "loss": 0.5131,
+      "step": 36450
+    },
+    {
+      "epoch": 0.8676330572217462,
+      "grad_norm": 0.9191488981558626,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 36451
+    },
+    {
+      "epoch": 0.8676568599447777,
+      "grad_norm": 3.499214632923529,
+      "learning_rate": 2e-05,
+      "loss": 0.4193,
+      "step": 36452
+    },
+    {
+      "epoch": 0.8676806626678092,
+      "grad_norm": 1.1363619107594078,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 36453
+    },
+    {
+      "epoch": 0.8677044653908407,
+      "grad_norm": 6.726179103949737,
+      "learning_rate": 2e-05,
+      "loss": 1.1094,
+      "step": 36454
+    },
+    {
+      "epoch": 0.8677282681138723,
+      "grad_norm": 3.8401746175678846,
+      "learning_rate": 2e-05,
+      "loss": 0.8228,
+      "step": 36455
+    },
+    {
+      "epoch": 0.8677520708369038,
+      "grad_norm": 0.19422372387287307,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 36456
+    },
+    {
+      "epoch": 0.8677758735599352,
+      "grad_norm": 0.08941636781476492,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36457
+    },
+    {
+      "epoch": 0.8677996762829667,
+      "grad_norm": 0.32072694542989644,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 36458
+    },
+    {
+      "epoch": 0.8678234790059983,
+      "grad_norm": 5.482680838824431,
+      "learning_rate": 2e-05,
+      "loss": 1.3927,
+      "step": 36459
+    },
+    {
+      "epoch": 0.8678472817290298,
+      "grad_norm": 16.85269544019289,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 36460
+    },
+    {
+      "epoch": 0.8678710844520613,
+      "grad_norm": 0.028214023747238987,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 36461
+    },
+    {
+      "epoch": 0.8678948871750928,
+      "grad_norm": 4.859208536210833,
+      "learning_rate": 2e-05,
+      "loss": 0.8968,
+      "step": 36462
+    },
+    {
+      "epoch": 0.8679186898981244,
+      "grad_norm": 0.6045145048198602,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 36463
+    },
+    {
+      "epoch": 0.8679424926211559,
+      "grad_norm": 2.915895155547365,
+      "learning_rate": 2e-05,
+      "loss": 0.411,
+      "step": 36464
+    },
+    {
+      "epoch": 0.8679662953441873,
+      "grad_norm": 5.010469353817329,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 36465
+    },
+    {
+      "epoch": 0.8679900980672189,
+      "grad_norm": 3.5507877290860557,
+      "learning_rate": 2e-05,
+      "loss": 0.3629,
+      "step": 36466
+    },
+    {
+      "epoch": 0.8680139007902504,
+      "grad_norm": 4.968401347765186,
+      "learning_rate": 2e-05,
+      "loss": 0.3329,
+      "step": 36467
+    },
+    {
+      "epoch": 0.8680377035132819,
+      "grad_norm": 1.3446988896276917,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 36468
+    },
+    {
+      "epoch": 0.8680615062363134,
+      "grad_norm": 2.8209102929370804,
+      "learning_rate": 2e-05,
+      "loss": 0.4998,
+      "step": 36469
+    },
+    {
+      "epoch": 0.868085308959345,
+      "grad_norm": 0.26614119843518697,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 36470
+    },
+    {
+      "epoch": 0.8681091116823765,
+      "grad_norm": 3.0824703214818188,
+      "learning_rate": 2e-05,
+      "loss": 0.6404,
+      "step": 36471
+    },
+    {
+      "epoch": 0.868132914405408,
+      "grad_norm": 8.54519317683096,
+      "learning_rate": 2e-05,
+      "loss": 0.2952,
+      "step": 36472
+    },
+    {
+      "epoch": 0.8681567171284394,
+      "grad_norm": 0.06820816373780819,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36473
+    },
+    {
+      "epoch": 0.868180519851471,
+      "grad_norm": 8.547852800156164,
+      "learning_rate": 2e-05,
+      "loss": 0.3871,
+      "step": 36474
+    },
+    {
+      "epoch": 0.8682043225745025,
+      "grad_norm": 3.7835882865456267,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 36475
+    },
+    {
+      "epoch": 0.868228125297534,
+      "grad_norm": 0.13186441890232928,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36476
+    },
+    {
+      "epoch": 0.8682519280205655,
+      "grad_norm": 0.7342397090760144,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 36477
+    },
+    {
+      "epoch": 0.8682757307435971,
+      "grad_norm": 6.99762595106617,
+      "learning_rate": 2e-05,
+      "loss": 0.6081,
+      "step": 36478
+    },
+    {
+      "epoch": 0.8682995334666286,
+      "grad_norm": 1.5791666330500873,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 36479
+    },
+    {
+      "epoch": 0.8683233361896601,
+      "grad_norm": 3.743902577195125,
+      "learning_rate": 2e-05,
+      "loss": 0.7183,
+      "step": 36480
+    },
+    {
+      "epoch": 0.8683471389126917,
+      "grad_norm": 3.7084809879810146,
+      "learning_rate": 2e-05,
+      "loss": 0.5363,
+      "step": 36481
+    },
+    {
+      "epoch": 0.8683709416357231,
+      "grad_norm": 13.591398175886773,
+      "learning_rate": 2e-05,
+      "loss": 0.4941,
+      "step": 36482
+    },
+    {
+      "epoch": 0.8683947443587546,
+      "grad_norm": 3.142450087565549,
+      "learning_rate": 2e-05,
+      "loss": 0.3281,
+      "step": 36483
+    },
+    {
+      "epoch": 0.8684185470817861,
+      "grad_norm": 0.2262437190560984,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 36484
+    },
+    {
+      "epoch": 0.8684423498048177,
+      "grad_norm": 1.3576658196295726,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 36485
+    },
+    {
+      "epoch": 0.8684661525278492,
+      "grad_norm": 3.663705517688931,
+      "learning_rate": 2e-05,
+      "loss": 0.5695,
+      "step": 36486
+    },
+    {
+      "epoch": 0.8684899552508807,
+      "grad_norm": 3.1330268383564857,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 36487
+    },
+    {
+      "epoch": 0.8685137579739122,
+      "grad_norm": 4.154317476142662,
+      "learning_rate": 2e-05,
+      "loss": 0.6606,
+      "step": 36488
+    },
+    {
+      "epoch": 0.8685375606969438,
+      "grad_norm": 3.0960178801390716,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 36489
+    },
+    {
+      "epoch": 0.8685613634199753,
+      "grad_norm": 2.9694406878963173,
+      "learning_rate": 2e-05,
+      "loss": 0.4563,
+      "step": 36490
+    },
+    {
+      "epoch": 0.8685851661430067,
+      "grad_norm": 3.6515186049673356,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 36491
+    },
+    {
+      "epoch": 0.8686089688660382,
+      "grad_norm": 3.175874384344581,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 36492
+    },
+    {
+      "epoch": 0.8686327715890698,
+      "grad_norm": 0.08806540766820965,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 36493
+    },
+    {
+      "epoch": 0.8686565743121013,
+      "grad_norm": 0.5363757570767508,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 36494
+    },
+    {
+      "epoch": 0.8686803770351328,
+      "grad_norm": 12.099620524149305,
+      "learning_rate": 2e-05,
+      "loss": 0.5651,
+      "step": 36495
+    },
+    {
+      "epoch": 0.8687041797581644,
+      "grad_norm": 2.834815408517706,
+      "learning_rate": 2e-05,
+      "loss": 0.0876,
+      "step": 36496
+    },
+    {
+      "epoch": 0.8687279824811959,
+      "grad_norm": 2.5962172553260867,
+      "learning_rate": 2e-05,
+      "loss": 0.2104,
+      "step": 36497
+    },
+    {
+      "epoch": 0.8687517852042274,
+      "grad_norm": 1.6452414195309208,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 36498
+    },
+    {
+      "epoch": 0.8687755879272588,
+      "grad_norm": 6.417810721350697,
+      "learning_rate": 2e-05,
+      "loss": 0.5874,
+      "step": 36499
+    },
+    {
+      "epoch": 0.8687993906502904,
+      "grad_norm": 9.114341883559495,
+      "learning_rate": 2e-05,
+      "loss": 0.399,
+      "step": 36500
+    },
+    {
+      "epoch": 0.8688231933733219,
+      "grad_norm": 0.7163447373787514,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 36501
+    },
+    {
+      "epoch": 0.8688469960963534,
+      "grad_norm": 0.8013263090712552,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 36502
+    },
+    {
+      "epoch": 0.8688707988193849,
+      "grad_norm": 2.855529920074768,
+      "learning_rate": 2e-05,
+      "loss": 0.3673,
+      "step": 36503
+    },
+    {
+      "epoch": 0.8688946015424165,
+      "grad_norm": 5.671894545146232,
+      "learning_rate": 2e-05,
+      "loss": 1.0214,
+      "step": 36504
+    },
+    {
+      "epoch": 0.868918404265448,
+      "grad_norm": 4.009735654588863,
+      "learning_rate": 2e-05,
+      "loss": 0.6202,
+      "step": 36505
+    },
+    {
+      "epoch": 0.8689422069884795,
+      "grad_norm": 2.6043935499662485,
+      "learning_rate": 2e-05,
+      "loss": 0.0348,
+      "step": 36506
+    },
+    {
+      "epoch": 0.868966009711511,
+      "grad_norm": 7.17748888029559,
+      "learning_rate": 2e-05,
+      "loss": 0.2539,
+      "step": 36507
+    },
+    {
+      "epoch": 0.8689898124345425,
+      "grad_norm": 4.349945894207034,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 36508
+    },
+    {
+      "epoch": 0.869013615157574,
+      "grad_norm": 7.019330732558374,
+      "learning_rate": 2e-05,
+      "loss": 0.1458,
+      "step": 36509
+    },
+    {
+      "epoch": 0.8690374178806055,
+      "grad_norm": 15.89068425481476,
+      "learning_rate": 2e-05,
+      "loss": 1.4559,
+      "step": 36510
+    },
+    {
+      "epoch": 0.8690612206036371,
+      "grad_norm": 4.460537971805605,
+      "learning_rate": 2e-05,
+      "loss": 0.8971,
+      "step": 36511
+    },
+    {
+      "epoch": 0.8690850233266686,
+      "grad_norm": 1.0973236212248534,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 36512
+    },
+    {
+      "epoch": 0.8691088260497001,
+      "grad_norm": 0.06291615666653737,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 36513
+    },
+    {
+      "epoch": 0.8691326287727316,
+      "grad_norm": 2.501005186515366,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 36514
+    },
+    {
+      "epoch": 0.8691564314957632,
+      "grad_norm": 8.402573037265821,
+      "learning_rate": 2e-05,
+      "loss": 0.395,
+      "step": 36515
+    },
+    {
+      "epoch": 0.8691802342187946,
+      "grad_norm": 0.18239375426113172,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 36516
+    },
+    {
+      "epoch": 0.8692040369418261,
+      "grad_norm": 1.287593547854178,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 36517
+    },
+    {
+      "epoch": 0.8692278396648576,
+      "grad_norm": 0.013146040102104016,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 36518
+    },
+    {
+      "epoch": 0.8692516423878892,
+      "grad_norm": 4.307323184239471,
+      "learning_rate": 2e-05,
+      "loss": 0.725,
+      "step": 36519
+    },
+    {
+      "epoch": 0.8692754451109207,
+      "grad_norm": 3.5750163930062326,
+      "learning_rate": 2e-05,
+      "loss": 0.324,
+      "step": 36520
+    },
+    {
+      "epoch": 0.8692992478339522,
+      "grad_norm": 3.541796227299319,
+      "learning_rate": 2e-05,
+      "loss": 0.4669,
+      "step": 36521
+    },
+    {
+      "epoch": 0.8693230505569837,
+      "grad_norm": 4.216408140242372,
+      "learning_rate": 2e-05,
+      "loss": 0.5281,
+      "step": 36522
+    },
+    {
+      "epoch": 0.8693468532800153,
+      "grad_norm": 2.261242477244789,
+      "learning_rate": 2e-05,
+      "loss": 0.2681,
+      "step": 36523
+    },
+    {
+      "epoch": 0.8693706560030467,
+      "grad_norm": 13.89223319369799,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 36524
+    },
+    {
+      "epoch": 0.8693944587260782,
+      "grad_norm": 2.5144931546607103,
+      "learning_rate": 2e-05,
+      "loss": 0.1489,
+      "step": 36525
+    },
+    {
+      "epoch": 0.8694182614491098,
+      "grad_norm": 0.4002138858220513,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 36526
+    },
+    {
+      "epoch": 0.8694420641721413,
+      "grad_norm": 3.443073751577206,
+      "learning_rate": 2e-05,
+      "loss": 0.5245,
+      "step": 36527
+    },
+    {
+      "epoch": 0.8694658668951728,
+      "grad_norm": 0.036876424225070065,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 36528
+    },
+    {
+      "epoch": 0.8694896696182043,
+      "grad_norm": 0.28351150480425164,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36529
+    },
+    {
+      "epoch": 0.8695134723412359,
+      "grad_norm": 3.5434424691867017,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 36530
+    },
+    {
+      "epoch": 0.8695372750642674,
+      "grad_norm": 0.19137960596322218,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 36531
+    },
+    {
+      "epoch": 0.8695610777872989,
+      "grad_norm": 0.6462608000706889,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 36532
+    },
+    {
+      "epoch": 0.8695848805103303,
+      "grad_norm": 0.22187967237977196,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 36533
+    },
+    {
+      "epoch": 0.8696086832333619,
+      "grad_norm": 2.8740871331358204,
+      "learning_rate": 2e-05,
+      "loss": 0.1912,
+      "step": 36534
+    },
+    {
+      "epoch": 0.8696324859563934,
+      "grad_norm": 2.89560891709632,
+      "learning_rate": 2e-05,
+      "loss": 0.0724,
+      "step": 36535
+    },
+    {
+      "epoch": 0.8696562886794249,
+      "grad_norm": 3.3754688710516185,
+      "learning_rate": 2e-05,
+      "loss": 0.6062,
+      "step": 36536
+    },
+    {
+      "epoch": 0.8696800914024564,
+      "grad_norm": 4.250407057492638,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 36537
+    },
+    {
+      "epoch": 0.869703894125488,
+      "grad_norm": 4.720002492597119,
+      "learning_rate": 2e-05,
+      "loss": 0.6547,
+      "step": 36538
+    },
+    {
+      "epoch": 0.8697276968485195,
+      "grad_norm": 1.4334145601560642,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 36539
+    },
+    {
+      "epoch": 0.869751499571551,
+      "grad_norm": 6.421409445505583,
+      "learning_rate": 2e-05,
+      "loss": 0.3751,
+      "step": 36540
+    },
+    {
+      "epoch": 0.8697753022945826,
+      "grad_norm": 4.7448789160571625,
+      "learning_rate": 2e-05,
+      "loss": 0.3903,
+      "step": 36541
+    },
+    {
+      "epoch": 0.869799105017614,
+      "grad_norm": 2.646078977365753,
+      "learning_rate": 2e-05,
+      "loss": 0.0754,
+      "step": 36542
+    },
+    {
+      "epoch": 0.8698229077406455,
+      "grad_norm": 0.11433837017089006,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36543
+    },
+    {
+      "epoch": 0.869846710463677,
+      "grad_norm": 6.190254064570633,
+      "learning_rate": 2e-05,
+      "loss": 0.6347,
+      "step": 36544
+    },
+    {
+      "epoch": 0.8698705131867086,
+      "grad_norm": 8.162415341463516,
+      "learning_rate": 2e-05,
+      "loss": 0.3076,
+      "step": 36545
+    },
+    {
+      "epoch": 0.8698943159097401,
+      "grad_norm": 0.46242619375931127,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36546
+    },
+    {
+      "epoch": 0.8699181186327716,
+      "grad_norm": 2.8213620897198055,
+      "learning_rate": 2e-05,
+      "loss": 0.4358,
+      "step": 36547
+    },
+    {
+      "epoch": 0.8699419213558031,
+      "grad_norm": 5.522618856521311,
+      "learning_rate": 2e-05,
+      "loss": 0.8616,
+      "step": 36548
+    },
+    {
+      "epoch": 0.8699657240788347,
+      "grad_norm": 3.9256014807507063,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 36549
+    },
+    {
+      "epoch": 0.8699895268018661,
+      "grad_norm": 7.3393707845034655,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 36550
+    },
+    {
+      "epoch": 0.8700133295248976,
+      "grad_norm": 1.577586267855324,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 36551
+    },
+    {
+      "epoch": 0.8700371322479291,
+      "grad_norm": 4.023223937135621,
+      "learning_rate": 2e-05,
+      "loss": 0.3964,
+      "step": 36552
+    },
+    {
+      "epoch": 0.8700609349709607,
+      "grad_norm": 10.736320524368915,
+      "learning_rate": 2e-05,
+      "loss": 1.556,
+      "step": 36553
+    },
+    {
+      "epoch": 0.8700847376939922,
+      "grad_norm": 4.507891384654534,
+      "learning_rate": 2e-05,
+      "loss": 0.2599,
+      "step": 36554
+    },
+    {
+      "epoch": 0.8701085404170237,
+      "grad_norm": 5.002997871124016,
+      "learning_rate": 2e-05,
+      "loss": 0.1058,
+      "step": 36555
+    },
+    {
+      "epoch": 0.8701323431400553,
+      "grad_norm": 0.030191675594164535,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 36556
+    },
+    {
+      "epoch": 0.8701561458630868,
+      "grad_norm": 0.07441314892073606,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 36557
+    },
+    {
+      "epoch": 0.8701799485861182,
+      "grad_norm": 0.49880931853135946,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 36558
+    },
+    {
+      "epoch": 0.8702037513091497,
+      "grad_norm": 0.19844975550778113,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 36559
+    },
+    {
+      "epoch": 0.8702275540321813,
+      "grad_norm": 0.0024857920414941637,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 36560
+    },
+    {
+      "epoch": 0.8702513567552128,
+      "grad_norm": 0.9704928783546849,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 36561
+    },
+    {
+      "epoch": 0.8702751594782443,
+      "grad_norm": 0.8066961618422268,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 36562
+    },
+    {
+      "epoch": 0.8702989622012758,
+      "grad_norm": 4.74221665582741,
+      "learning_rate": 2e-05,
+      "loss": 1.0338,
+      "step": 36563
+    },
+    {
+      "epoch": 0.8703227649243074,
+      "grad_norm": 4.6821462187657845,
+      "learning_rate": 2e-05,
+      "loss": 0.6543,
+      "step": 36564
+    },
+    {
+      "epoch": 0.8703465676473389,
+      "grad_norm": 3.026590858792061,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 36565
+    },
+    {
+      "epoch": 0.8703703703703703,
+      "grad_norm": 11.705073284935713,
+      "learning_rate": 2e-05,
+      "loss": 0.5057,
+      "step": 36566
+    },
+    {
+      "epoch": 0.8703941730934018,
+      "grad_norm": 3.3449501914506645,
+      "learning_rate": 2e-05,
+      "loss": 0.3832,
+      "step": 36567
+    },
+    {
+      "epoch": 0.8704179758164334,
+      "grad_norm": 2.5663035290996947,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 36568
+    },
+    {
+      "epoch": 0.8704417785394649,
+      "grad_norm": 3.1916417344044654,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 36569
+    },
+    {
+      "epoch": 0.8704655812624964,
+      "grad_norm": 0.18468842188770018,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 36570
+    },
+    {
+      "epoch": 0.870489383985528,
+      "grad_norm": 8.313413479301042,
+      "learning_rate": 2e-05,
+      "loss": 1.3944,
+      "step": 36571
+    },
+    {
+      "epoch": 0.8705131867085595,
+      "grad_norm": 4.381522581244384,
+      "learning_rate": 2e-05,
+      "loss": 0.7892,
+      "step": 36572
+    },
+    {
+      "epoch": 0.870536989431591,
+      "grad_norm": 0.18756475427706382,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36573
+    },
+    {
+      "epoch": 0.8705607921546225,
+      "grad_norm": 4.478331190438616,
+      "learning_rate": 2e-05,
+      "loss": 0.1113,
+      "step": 36574
+    },
+    {
+      "epoch": 0.870584594877654,
+      "grad_norm": 4.158338277248468,
+      "learning_rate": 2e-05,
+      "loss": 0.4925,
+      "step": 36575
+    },
+    {
+      "epoch": 0.8706083976006855,
+      "grad_norm": 7.234322006391134,
+      "learning_rate": 2e-05,
+      "loss": 0.3037,
+      "step": 36576
+    },
+    {
+      "epoch": 0.870632200323717,
+      "grad_norm": 2.5802935790383823,
+      "learning_rate": 2e-05,
+      "loss": 0.2822,
+      "step": 36577
+    },
+    {
+      "epoch": 0.8706560030467485,
+      "grad_norm": 3.2268630317359555,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 36578
+    },
+    {
+      "epoch": 0.8706798057697801,
+      "grad_norm": 5.071518388044123,
+      "learning_rate": 2e-05,
+      "loss": 1.2262,
+      "step": 36579
+    },
+    {
+      "epoch": 0.8707036084928116,
+      "grad_norm": 1.4347293147281863,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 36580
+    },
+    {
+      "epoch": 0.8707274112158431,
+      "grad_norm": 2.3167340490639234,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 36581
+    },
+    {
+      "epoch": 0.8707512139388746,
+      "grad_norm": 3.644633197591363,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 36582
+    },
+    {
+      "epoch": 0.8707750166619062,
+      "grad_norm": 0.27603786350044446,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 36583
+    },
+    {
+      "epoch": 0.8707988193849376,
+      "grad_norm": 0.049638389970128406,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 36584
+    },
+    {
+      "epoch": 0.8708226221079691,
+      "grad_norm": 3.257686357808283,
+      "learning_rate": 2e-05,
+      "loss": 0.4655,
+      "step": 36585
+    },
+    {
+      "epoch": 0.8708464248310007,
+      "grad_norm": 1.3525677379036074,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 36586
+    },
+    {
+      "epoch": 0.8708702275540322,
+      "grad_norm": 7.959873272749453,
+      "learning_rate": 2e-05,
+      "loss": 0.2956,
+      "step": 36587
+    },
+    {
+      "epoch": 0.8708940302770637,
+      "grad_norm": 3.548037659121775,
+      "learning_rate": 2e-05,
+      "loss": 0.4218,
+      "step": 36588
+    },
+    {
+      "epoch": 0.8709178330000952,
+      "grad_norm": 4.892452376748356,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 36589
+    },
+    {
+      "epoch": 0.8709416357231268,
+      "grad_norm": 5.194791103823215,
+      "learning_rate": 2e-05,
+      "loss": 0.9878,
+      "step": 36590
+    },
+    {
+      "epoch": 0.8709654384461583,
+      "grad_norm": 0.908172294155997,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 36591
+    },
+    {
+      "epoch": 0.8709892411691897,
+      "grad_norm": 2.3826862354973093,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 36592
+    },
+    {
+      "epoch": 0.8710130438922212,
+      "grad_norm": 0.025123741792660734,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36593
+    },
+    {
+      "epoch": 0.8710368466152528,
+      "grad_norm": 2.266896385034878,
+      "learning_rate": 2e-05,
+      "loss": 0.2472,
+      "step": 36594
+    },
+    {
+      "epoch": 0.8710606493382843,
+      "grad_norm": 3.823661726473304,
+      "learning_rate": 2e-05,
+      "loss": 0.2572,
+      "step": 36595
+    },
+    {
+      "epoch": 0.8710844520613158,
+      "grad_norm": 3.9772357874294655,
+      "learning_rate": 2e-05,
+      "loss": 0.9516,
+      "step": 36596
+    },
+    {
+      "epoch": 0.8711082547843473,
+      "grad_norm": 7.770281981030185,
+      "learning_rate": 2e-05,
+      "loss": 1.0603,
+      "step": 36597
+    },
+    {
+      "epoch": 0.8711320575073789,
+      "grad_norm": 4.332361331616047,
+      "learning_rate": 2e-05,
+      "loss": 0.8263,
+      "step": 36598
+    },
+    {
+      "epoch": 0.8711558602304104,
+      "grad_norm": 1.8546665952935932,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 36599
+    },
+    {
+      "epoch": 0.8711796629534418,
+      "grad_norm": 5.171386410335707,
+      "learning_rate": 2e-05,
+      "loss": 0.9776,
+      "step": 36600
+    },
+    {
+      "epoch": 0.8712034656764734,
+      "grad_norm": 2.5057835627091456,
+      "learning_rate": 2e-05,
+      "loss": 0.2581,
+      "step": 36601
+    },
+    {
+      "epoch": 0.8712272683995049,
+      "grad_norm": 4.884401375683296,
+      "learning_rate": 2e-05,
+      "loss": 0.6738,
+      "step": 36602
+    },
+    {
+      "epoch": 0.8712510711225364,
+      "grad_norm": 2.83722998287029,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 36603
+    },
+    {
+      "epoch": 0.8712748738455679,
+      "grad_norm": 0.12331614870520251,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 36604
+    },
+    {
+      "epoch": 0.8712986765685995,
+      "grad_norm": 5.888913445704014,
+      "learning_rate": 2e-05,
+      "loss": 0.2476,
+      "step": 36605
+    },
+    {
+      "epoch": 0.871322479291631,
+      "grad_norm": 4.649632427135083,
+      "learning_rate": 2e-05,
+      "loss": 0.0458,
+      "step": 36606
+    },
+    {
+      "epoch": 0.8713462820146625,
+      "grad_norm": 4.421334095599124,
+      "learning_rate": 2e-05,
+      "loss": 0.7067,
+      "step": 36607
+    },
+    {
+      "epoch": 0.871370084737694,
+      "grad_norm": 4.483963196364134,
+      "learning_rate": 2e-05,
+      "loss": 1.3983,
+      "step": 36608
+    },
+    {
+      "epoch": 0.8713938874607255,
+      "grad_norm": 5.040629735310116,
+      "learning_rate": 2e-05,
+      "loss": 0.9775,
+      "step": 36609
+    },
+    {
+      "epoch": 0.871417690183757,
+      "grad_norm": 0.010454300881496273,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 36610
+    },
+    {
+      "epoch": 0.8714414929067885,
+      "grad_norm": 0.35693540811788604,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 36611
+    },
+    {
+      "epoch": 0.87146529562982,
+      "grad_norm": 6.739165531496049,
+      "learning_rate": 2e-05,
+      "loss": 0.1477,
+      "step": 36612
+    },
+    {
+      "epoch": 0.8714890983528516,
+      "grad_norm": 0.3348048793050478,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 36613
+    },
+    {
+      "epoch": 0.8715129010758831,
+      "grad_norm": 0.5240976098935007,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 36614
+    },
+    {
+      "epoch": 0.8715367037989146,
+      "grad_norm": 0.852188278272655,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 36615
+    },
+    {
+      "epoch": 0.8715605065219462,
+      "grad_norm": 0.32284748009782815,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 36616
+    },
+    {
+      "epoch": 0.8715843092449777,
+      "grad_norm": 3.7776500595148157,
+      "learning_rate": 2e-05,
+      "loss": 0.18,
+      "step": 36617
+    },
+    {
+      "epoch": 0.8716081119680091,
+      "grad_norm": 3.7792895109124425,
+      "learning_rate": 2e-05,
+      "loss": 0.3758,
+      "step": 36618
+    },
+    {
+      "epoch": 0.8716319146910406,
+      "grad_norm": 3.545303825231892,
+      "learning_rate": 2e-05,
+      "loss": 0.2939,
+      "step": 36619
+    },
+    {
+      "epoch": 0.8716557174140722,
+      "grad_norm": 2.7643704312475847,
+      "learning_rate": 2e-05,
+      "loss": 0.102,
+      "step": 36620
+    },
+    {
+      "epoch": 0.8716795201371037,
+      "grad_norm": 5.47101764195152,
+      "learning_rate": 2e-05,
+      "loss": 1.3724,
+      "step": 36621
+    },
+    {
+      "epoch": 0.8717033228601352,
+      "grad_norm": 6.4418414640273705,
+      "learning_rate": 2e-05,
+      "loss": 0.0704,
+      "step": 36622
+    },
+    {
+      "epoch": 0.8717271255831667,
+      "grad_norm": 5.029515833539297,
+      "learning_rate": 2e-05,
+      "loss": 0.9806,
+      "step": 36623
+    },
+    {
+      "epoch": 0.8717509283061983,
+      "grad_norm": 5.175167656925728,
+      "learning_rate": 2e-05,
+      "loss": 0.781,
+      "step": 36624
+    },
+    {
+      "epoch": 0.8717747310292298,
+      "grad_norm": 6.588916502686478,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 36625
+    },
+    {
+      "epoch": 0.8717985337522612,
+      "grad_norm": 3.0329125040793636,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 36626
+    },
+    {
+      "epoch": 0.8718223364752927,
+      "grad_norm": 3.042306174186334,
+      "learning_rate": 2e-05,
+      "loss": 0.3825,
+      "step": 36627
+    },
+    {
+      "epoch": 0.8718461391983243,
+      "grad_norm": 4.414163605871788,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 36628
+    },
+    {
+      "epoch": 0.8718699419213558,
+      "grad_norm": 0.3272103895748323,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 36629
+    },
+    {
+      "epoch": 0.8718937446443873,
+      "grad_norm": 6.151339977537682,
+      "learning_rate": 2e-05,
+      "loss": 0.9235,
+      "step": 36630
+    },
+    {
+      "epoch": 0.8719175473674189,
+      "grad_norm": 4.2997188136229765,
+      "learning_rate": 2e-05,
+      "loss": 0.6229,
+      "step": 36631
+    },
+    {
+      "epoch": 0.8719413500904504,
+      "grad_norm": 0.5163582786553961,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 36632
+    },
+    {
+      "epoch": 0.8719651528134819,
+      "grad_norm": 3.244446536839616,
+      "learning_rate": 2e-05,
+      "loss": 0.4611,
+      "step": 36633
+    },
+    {
+      "epoch": 0.8719889555365133,
+      "grad_norm": 3.02016115646493,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 36634
+    },
+    {
+      "epoch": 0.8720127582595449,
+      "grad_norm": 1.033318143515265,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 36635
+    },
+    {
+      "epoch": 0.8720365609825764,
+      "grad_norm": 6.956510705208647,
+      "learning_rate": 2e-05,
+      "loss": 1.2733,
+      "step": 36636
+    },
+    {
+      "epoch": 0.8720603637056079,
+      "grad_norm": 5.067156065670732,
+      "learning_rate": 2e-05,
+      "loss": 0.8774,
+      "step": 36637
+    },
+    {
+      "epoch": 0.8720841664286394,
+      "grad_norm": 4.695924756659963,
+      "learning_rate": 2e-05,
+      "loss": 1.1831,
+      "step": 36638
+    },
+    {
+      "epoch": 0.872107969151671,
+      "grad_norm": 3.3469347066021866,
+      "learning_rate": 2e-05,
+      "loss": 0.5989,
+      "step": 36639
+    },
+    {
+      "epoch": 0.8721317718747025,
+      "grad_norm": 0.019437518322033023,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 36640
+    },
+    {
+      "epoch": 0.872155574597734,
+      "grad_norm": 11.776343681500885,
+      "learning_rate": 2e-05,
+      "loss": 0.4354,
+      "step": 36641
+    },
+    {
+      "epoch": 0.8721793773207654,
+      "grad_norm": 3.310101265553643,
+      "learning_rate": 2e-05,
+      "loss": 0.3462,
+      "step": 36642
+    },
+    {
+      "epoch": 0.872203180043797,
+      "grad_norm": 0.07784579723061173,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36643
+    },
+    {
+      "epoch": 0.8722269827668285,
+      "grad_norm": 0.03936695428520174,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36644
+    },
+    {
+      "epoch": 0.87225078548986,
+      "grad_norm": 4.029996871437194,
+      "learning_rate": 2e-05,
+      "loss": 0.6783,
+      "step": 36645
+    },
+    {
+      "epoch": 0.8722745882128915,
+      "grad_norm": 0.015369395128980931,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36646
+    },
+    {
+      "epoch": 0.8722983909359231,
+      "grad_norm": 0.7147713992233533,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 36647
+    },
+    {
+      "epoch": 0.8723221936589546,
+      "grad_norm": 0.056641414477696175,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 36648
+    },
+    {
+      "epoch": 0.8723459963819861,
+      "grad_norm": 3.41492881008842,
+      "learning_rate": 2e-05,
+      "loss": 0.316,
+      "step": 36649
+    },
+    {
+      "epoch": 0.8723697991050177,
+      "grad_norm": 3.206376014249379,
+      "learning_rate": 2e-05,
+      "loss": 0.42,
+      "step": 36650
+    },
+    {
+      "epoch": 0.8723936018280491,
+      "grad_norm": 20.02881707883686,
+      "learning_rate": 2e-05,
+      "loss": 0.2525,
+      "step": 36651
+    },
+    {
+      "epoch": 0.8724174045510806,
+      "grad_norm": 0.1665929961679822,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 36652
+    },
+    {
+      "epoch": 0.8724412072741121,
+      "grad_norm": 3.941228262056002,
+      "learning_rate": 2e-05,
+      "loss": 0.4843,
+      "step": 36653
+    },
+    {
+      "epoch": 0.8724650099971437,
+      "grad_norm": 0.2688562088360509,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36654
+    },
+    {
+      "epoch": 0.8724888127201752,
+      "grad_norm": 1.0914870536256434,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 36655
+    },
+    {
+      "epoch": 0.8725126154432067,
+      "grad_norm": 7.683017936374539,
+      "learning_rate": 2e-05,
+      "loss": 1.1936,
+      "step": 36656
+    },
+    {
+      "epoch": 0.8725364181662382,
+      "grad_norm": 2.910984534096839,
+      "learning_rate": 2e-05,
+      "loss": 0.3257,
+      "step": 36657
+    },
+    {
+      "epoch": 0.8725602208892698,
+      "grad_norm": 0.026239561491401053,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36658
+    },
+    {
+      "epoch": 0.8725840236123013,
+      "grad_norm": 0.026862030551461136,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36659
+    },
+    {
+      "epoch": 0.8726078263353327,
+      "grad_norm": 3.1957725627315257,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 36660
+    },
+    {
+      "epoch": 0.8726316290583642,
+      "grad_norm": 0.15390682326329785,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 36661
+    },
+    {
+      "epoch": 0.8726554317813958,
+      "grad_norm": 7.339478443366335,
+      "learning_rate": 2e-05,
+      "loss": 0.3295,
+      "step": 36662
+    },
+    {
+      "epoch": 0.8726792345044273,
+      "grad_norm": 5.303757140648669,
+      "learning_rate": 2e-05,
+      "loss": 0.924,
+      "step": 36663
+    },
+    {
+      "epoch": 0.8727030372274588,
+      "grad_norm": 2.5048915083343815,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 36664
+    },
+    {
+      "epoch": 0.8727268399504904,
+      "grad_norm": 1.100170929502993,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 36665
+    },
+    {
+      "epoch": 0.8727506426735219,
+      "grad_norm": 0.7107405293911864,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 36666
+    },
+    {
+      "epoch": 0.8727744453965534,
+      "grad_norm": 3.321625469897396,
+      "learning_rate": 2e-05,
+      "loss": 0.0643,
+      "step": 36667
+    },
+    {
+      "epoch": 0.8727982481195848,
+      "grad_norm": 3.4442370862495286,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 36668
+    },
+    {
+      "epoch": 0.8728220508426164,
+      "grad_norm": 1.9941848828155981,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 36669
+    },
+    {
+      "epoch": 0.8728458535656479,
+      "grad_norm": 2.7898271809601565,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 36670
+    },
+    {
+      "epoch": 0.8728696562886794,
+      "grad_norm": 4.722057663224356,
+      "learning_rate": 2e-05,
+      "loss": 0.7099,
+      "step": 36671
+    },
+    {
+      "epoch": 0.8728934590117109,
+      "grad_norm": 0.7595210533315666,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 36672
+    },
+    {
+      "epoch": 0.8729172617347425,
+      "grad_norm": 0.15939607813720116,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36673
+    },
+    {
+      "epoch": 0.872941064457774,
+      "grad_norm": 0.1314729835190359,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 36674
+    },
+    {
+      "epoch": 0.8729648671808055,
+      "grad_norm": 0.1943177248092934,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 36675
+    },
+    {
+      "epoch": 0.872988669903837,
+      "grad_norm": 5.739755495183311,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 36676
+    },
+    {
+      "epoch": 0.8730124726268685,
+      "grad_norm": 4.94758396708697,
+      "learning_rate": 2e-05,
+      "loss": 0.9505,
+      "step": 36677
+    },
+    {
+      "epoch": 0.8730362753499,
+      "grad_norm": 0.7324126068873371,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 36678
+    },
+    {
+      "epoch": 0.8730600780729315,
+      "grad_norm": 5.366287961905949,
+      "learning_rate": 2e-05,
+      "loss": 0.6675,
+      "step": 36679
+    },
+    {
+      "epoch": 0.8730838807959631,
+      "grad_norm": 3.317484979026185,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 36680
+    },
+    {
+      "epoch": 0.8731076835189946,
+      "grad_norm": 4.459040484404352,
+      "learning_rate": 2e-05,
+      "loss": 0.5373,
+      "step": 36681
+    },
+    {
+      "epoch": 0.8731314862420261,
+      "grad_norm": 10.33840682833871,
+      "learning_rate": 2e-05,
+      "loss": 0.6513,
+      "step": 36682
+    },
+    {
+      "epoch": 0.8731552889650576,
+      "grad_norm": 3.1150849164807712,
+      "learning_rate": 2e-05,
+      "loss": 0.7239,
+      "step": 36683
+    },
+    {
+      "epoch": 0.8731790916880892,
+      "grad_norm": 15.381703284094968,
+      "learning_rate": 2e-05,
+      "loss": 0.8103,
+      "step": 36684
+    },
+    {
+      "epoch": 0.8732028944111206,
+      "grad_norm": 14.840407779146963,
+      "learning_rate": 2e-05,
+      "loss": 0.8169,
+      "step": 36685
+    },
+    {
+      "epoch": 0.8732266971341521,
+      "grad_norm": 0.08975604026636524,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36686
+    },
+    {
+      "epoch": 0.8732504998571836,
+      "grad_norm": 0.045030210067624865,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 36687
+    },
+    {
+      "epoch": 0.8732743025802152,
+      "grad_norm": 1.1521352445072328,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 36688
+    },
+    {
+      "epoch": 0.8732981053032467,
+      "grad_norm": 10.411104533450901,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 36689
+    },
+    {
+      "epoch": 0.8733219080262782,
+      "grad_norm": 7.965561563440725,
+      "learning_rate": 2e-05,
+      "loss": 0.2721,
+      "step": 36690
+    },
+    {
+      "epoch": 0.8733457107493097,
+      "grad_norm": 6.82032987063771,
+      "learning_rate": 2e-05,
+      "loss": 0.4063,
+      "step": 36691
+    },
+    {
+      "epoch": 0.8733695134723413,
+      "grad_norm": 2.7177315678404566,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 36692
+    },
+    {
+      "epoch": 0.8733933161953727,
+      "grad_norm": 7.466874418142576,
+      "learning_rate": 2e-05,
+      "loss": 1.3182,
+      "step": 36693
+    },
+    {
+      "epoch": 0.8734171189184042,
+      "grad_norm": 1.8352412582666737,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 36694
+    },
+    {
+      "epoch": 0.8734409216414358,
+      "grad_norm": 0.025844335652671583,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 36695
+    },
+    {
+      "epoch": 0.8734647243644673,
+      "grad_norm": 3.032832781766703,
+      "learning_rate": 2e-05,
+      "loss": 0.3466,
+      "step": 36696
+    },
+    {
+      "epoch": 0.8734885270874988,
+      "grad_norm": 3.528306636183356,
+      "learning_rate": 2e-05,
+      "loss": 0.4245,
+      "step": 36697
+    },
+    {
+      "epoch": 0.8735123298105303,
+      "grad_norm": 5.5556179911974475,
+      "learning_rate": 2e-05,
+      "loss": 0.6677,
+      "step": 36698
+    },
+    {
+      "epoch": 0.8735361325335619,
+      "grad_norm": 3.069403536327108,
+      "learning_rate": 2e-05,
+      "loss": 0.3801,
+      "step": 36699
+    },
+    {
+      "epoch": 0.8735599352565934,
+      "grad_norm": 1.873549132811284,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 36700
+    },
+    {
+      "epoch": 0.8735837379796249,
+      "grad_norm": 2.5950721105791796,
+      "learning_rate": 2e-05,
+      "loss": 0.1204,
+      "step": 36701
+    },
+    {
+      "epoch": 0.8736075407026563,
+      "grad_norm": 0.33746305982539304,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36702
+    },
+    {
+      "epoch": 0.8736313434256879,
+      "grad_norm": 5.94508881655793,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 36703
+    },
+    {
+      "epoch": 0.8736551461487194,
+      "grad_norm": 0.3990143483571595,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 36704
+    },
+    {
+      "epoch": 0.8736789488717509,
+      "grad_norm": 0.5549496748527056,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36705
+    },
+    {
+      "epoch": 0.8737027515947824,
+      "grad_norm": 10.944314620551257,
+      "learning_rate": 2e-05,
+      "loss": 1.215,
+      "step": 36706
+    },
+    {
+      "epoch": 0.873726554317814,
+      "grad_norm": 5.103168684775721,
+      "learning_rate": 2e-05,
+      "loss": 0.7474,
+      "step": 36707
+    },
+    {
+      "epoch": 0.8737503570408455,
+      "grad_norm": 0.013515246530078325,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36708
+    },
+    {
+      "epoch": 0.873774159763877,
+      "grad_norm": 0.332947784329566,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 36709
+    },
+    {
+      "epoch": 0.8737979624869086,
+      "grad_norm": 0.8114322148972662,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 36710
+    },
+    {
+      "epoch": 0.87382176520994,
+      "grad_norm": 7.864285593021907,
+      "learning_rate": 2e-05,
+      "loss": 0.3057,
+      "step": 36711
+    },
+    {
+      "epoch": 0.8738455679329715,
+      "grad_norm": 0.6528742660850477,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 36712
+    },
+    {
+      "epoch": 0.873869370656003,
+      "grad_norm": 0.017413255462112816,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36713
+    },
+    {
+      "epoch": 0.8738931733790346,
+      "grad_norm": 4.368368480200669,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 36714
+    },
+    {
+      "epoch": 0.8739169761020661,
+      "grad_norm": 2.7556766759225018,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 36715
+    },
+    {
+      "epoch": 0.8739407788250976,
+      "grad_norm": 5.983316663711567,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 36716
+    },
+    {
+      "epoch": 0.8739645815481291,
+      "grad_norm": 4.723964401388916,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 36717
+    },
+    {
+      "epoch": 0.8739883842711607,
+      "grad_norm": 4.163297746040248,
+      "learning_rate": 2e-05,
+      "loss": 0.6879,
+      "step": 36718
+    },
+    {
+      "epoch": 0.8740121869941921,
+      "grad_norm": 4.115676032300307,
+      "learning_rate": 2e-05,
+      "loss": 0.8634,
+      "step": 36719
+    },
+    {
+      "epoch": 0.8740359897172236,
+      "grad_norm": 5.5135337902846855,
+      "learning_rate": 2e-05,
+      "loss": 0.8599,
+      "step": 36720
+    },
+    {
+      "epoch": 0.8740597924402551,
+      "grad_norm": 3.577140503634842,
+      "learning_rate": 2e-05,
+      "loss": 0.5736,
+      "step": 36721
+    },
+    {
+      "epoch": 0.8740835951632867,
+      "grad_norm": 6.073993998148141,
+      "learning_rate": 2e-05,
+      "loss": 0.9929,
+      "step": 36722
+    },
+    {
+      "epoch": 0.8741073978863182,
+      "grad_norm": 6.404822898830265,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 36723
+    },
+    {
+      "epoch": 0.8741312006093497,
+      "grad_norm": 0.03165289917607258,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 36724
+    },
+    {
+      "epoch": 0.8741550033323813,
+      "grad_norm": 3.710444180239146,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 36725
+    },
+    {
+      "epoch": 0.8741788060554128,
+      "grad_norm": 6.141945098744578,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 36726
+    },
+    {
+      "epoch": 0.8742026087784442,
+      "grad_norm": 3.252029015109447,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 36727
+    },
+    {
+      "epoch": 0.8742264115014757,
+      "grad_norm": 2.7167951366620744,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 36728
+    },
+    {
+      "epoch": 0.8742502142245073,
+      "grad_norm": 5.849268772432536,
+      "learning_rate": 2e-05,
+      "loss": 1.2218,
+      "step": 36729
+    },
+    {
+      "epoch": 0.8742740169475388,
+      "grad_norm": 0.5822225884930079,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 36730
+    },
+    {
+      "epoch": 0.8742978196705703,
+      "grad_norm": 2.8383933017240492,
+      "learning_rate": 2e-05,
+      "loss": 0.3087,
+      "step": 36731
+    },
+    {
+      "epoch": 0.8743216223936018,
+      "grad_norm": 1.1668577198215377,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 36732
+    },
+    {
+      "epoch": 0.8743454251166334,
+      "grad_norm": 2.790590877376707,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 36733
+    },
+    {
+      "epoch": 0.8743692278396649,
+      "grad_norm": 0.15813913815701433,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36734
+    },
+    {
+      "epoch": 0.8743930305626963,
+      "grad_norm": 4.328940708740073,
+      "learning_rate": 2e-05,
+      "loss": 0.0763,
+      "step": 36735
+    },
+    {
+      "epoch": 0.8744168332857278,
+      "grad_norm": 10.654322266549414,
+      "learning_rate": 2e-05,
+      "loss": 0.1249,
+      "step": 36736
+    },
+    {
+      "epoch": 0.8744406360087594,
+      "grad_norm": 2.2803009213569174,
+      "learning_rate": 2e-05,
+      "loss": 0.4328,
+      "step": 36737
+    },
+    {
+      "epoch": 0.8744644387317909,
+      "grad_norm": 3.267466330625991,
+      "learning_rate": 2e-05,
+      "loss": 0.4194,
+      "step": 36738
+    },
+    {
+      "epoch": 0.8744882414548224,
+      "grad_norm": 0.06204007726650914,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 36739
+    },
+    {
+      "epoch": 0.874512044177854,
+      "grad_norm": 14.188607340713503,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 36740
+    },
+    {
+      "epoch": 0.8745358469008855,
+      "grad_norm": 11.213673828195134,
+      "learning_rate": 2e-05,
+      "loss": 0.9412,
+      "step": 36741
+    },
+    {
+      "epoch": 0.874559649623917,
+      "grad_norm": 4.407594121562272,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 36742
+    },
+    {
+      "epoch": 0.8745834523469485,
+      "grad_norm": 4.983713459139604,
+      "learning_rate": 2e-05,
+      "loss": 0.1377,
+      "step": 36743
+    },
+    {
+      "epoch": 0.87460725506998,
+      "grad_norm": 1.5075227116426038,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 36744
+    },
+    {
+      "epoch": 0.8746310577930115,
+      "grad_norm": 7.4511920200484125,
+      "learning_rate": 2e-05,
+      "loss": 0.7286,
+      "step": 36745
+    },
+    {
+      "epoch": 0.874654860516043,
+      "grad_norm": 6.5503114788396175,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 36746
+    },
+    {
+      "epoch": 0.8746786632390745,
+      "grad_norm": 5.285019272422661,
+      "learning_rate": 2e-05,
+      "loss": 1.5406,
+      "step": 36747
+    },
+    {
+      "epoch": 0.8747024659621061,
+      "grad_norm": 0.33827598401590825,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 36748
+    },
+    {
+      "epoch": 0.8747262686851376,
+      "grad_norm": 0.1599704599750702,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36749
+    },
+    {
+      "epoch": 0.8747500714081691,
+      "grad_norm": 0.05008137952079285,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 36750
+    },
+    {
+      "epoch": 0.8747738741312006,
+      "grad_norm": 0.4566934373253386,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 36751
+    },
+    {
+      "epoch": 0.8747976768542322,
+      "grad_norm": 0.7694058221904148,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 36752
+    },
+    {
+      "epoch": 0.8748214795772636,
+      "grad_norm": 8.298461710423553,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 36753
+    },
+    {
+      "epoch": 0.8748452823002951,
+      "grad_norm": 0.6226136172647465,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 36754
+    },
+    {
+      "epoch": 0.8748690850233267,
+      "grad_norm": 4.783651469835753,
+      "learning_rate": 2e-05,
+      "loss": 0.8843,
+      "step": 36755
+    },
+    {
+      "epoch": 0.8748928877463582,
+      "grad_norm": 12.291014226614859,
+      "learning_rate": 2e-05,
+      "loss": 0.5943,
+      "step": 36756
+    },
+    {
+      "epoch": 0.8749166904693897,
+      "grad_norm": 4.566113732500526,
+      "learning_rate": 2e-05,
+      "loss": 0.55,
+      "step": 36757
+    },
+    {
+      "epoch": 0.8749404931924212,
+      "grad_norm": 0.12442381308260352,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 36758
+    },
+    {
+      "epoch": 0.8749642959154528,
+      "grad_norm": 4.45278642523029,
+      "learning_rate": 2e-05,
+      "loss": 1.2755,
+      "step": 36759
+    },
+    {
+      "epoch": 0.8749880986384843,
+      "grad_norm": 5.510050100541299,
+      "learning_rate": 2e-05,
+      "loss": 0.8928,
+      "step": 36760
+    },
+    {
+      "epoch": 0.8750119013615157,
+      "grad_norm": 5.88105484416442,
+      "learning_rate": 2e-05,
+      "loss": 0.1011,
+      "step": 36761
+    },
+    {
+      "epoch": 0.8750357040845472,
+      "grad_norm": 5.1098337172624015,
+      "learning_rate": 2e-05,
+      "loss": 1.1978,
+      "step": 36762
+    },
+    {
+      "epoch": 0.8750595068075788,
+      "grad_norm": 0.09266355054007953,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36763
+    },
+    {
+      "epoch": 0.8750833095306103,
+      "grad_norm": 0.048274898161918586,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 36764
+    },
+    {
+      "epoch": 0.8751071122536418,
+      "grad_norm": 0.11105704678516407,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36765
+    },
+    {
+      "epoch": 0.8751309149766733,
+      "grad_norm": 3.6826397501572177,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 36766
+    },
+    {
+      "epoch": 0.8751547176997049,
+      "grad_norm": 0.29655493228238883,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 36767
+    },
+    {
+      "epoch": 0.8751785204227364,
+      "grad_norm": 5.441092533072239,
+      "learning_rate": 2e-05,
+      "loss": 1.0617,
+      "step": 36768
+    },
+    {
+      "epoch": 0.8752023231457678,
+      "grad_norm": 2.5008332417195414,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 36769
+    },
+    {
+      "epoch": 0.8752261258687994,
+      "grad_norm": 3.9137227587245937,
+      "learning_rate": 2e-05,
+      "loss": 0.7686,
+      "step": 36770
+    },
+    {
+      "epoch": 0.8752499285918309,
+      "grad_norm": 3.977548439582822,
+      "learning_rate": 2e-05,
+      "loss": 0.8491,
+      "step": 36771
+    },
+    {
+      "epoch": 0.8752737313148624,
+      "grad_norm": 11.016431921793886,
+      "learning_rate": 2e-05,
+      "loss": 0.6354,
+      "step": 36772
+    },
+    {
+      "epoch": 0.8752975340378939,
+      "grad_norm": 3.7878608437381023,
+      "learning_rate": 2e-05,
+      "loss": 0.8129,
+      "step": 36773
+    },
+    {
+      "epoch": 0.8753213367609255,
+      "grad_norm": 3.2655271036160474,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 36774
+    },
+    {
+      "epoch": 0.875345139483957,
+      "grad_norm": 4.176928263861687,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 36775
+    },
+    {
+      "epoch": 0.8753689422069885,
+      "grad_norm": 5.307428431825972,
+      "learning_rate": 2e-05,
+      "loss": 1.2316,
+      "step": 36776
+    },
+    {
+      "epoch": 0.87539274493002,
+      "grad_norm": 3.2989546804261667,
+      "learning_rate": 2e-05,
+      "loss": 0.4767,
+      "step": 36777
+    },
+    {
+      "epoch": 0.8754165476530515,
+      "grad_norm": 0.216388189485113,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 36778
+    },
+    {
+      "epoch": 0.875440350376083,
+      "grad_norm": 3.301777263332285,
+      "learning_rate": 2e-05,
+      "loss": 0.4025,
+      "step": 36779
+    },
+    {
+      "epoch": 0.8754641530991145,
+      "grad_norm": 0.6437362115290275,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 36780
+    },
+    {
+      "epoch": 0.875487955822146,
+      "grad_norm": 5.34141275444135,
+      "learning_rate": 2e-05,
+      "loss": 0.6149,
+      "step": 36781
+    },
+    {
+      "epoch": 0.8755117585451776,
+      "grad_norm": 4.089386950726453,
+      "learning_rate": 2e-05,
+      "loss": 0.2161,
+      "step": 36782
+    },
+    {
+      "epoch": 0.8755355612682091,
+      "grad_norm": 1.8992459131518662,
+      "learning_rate": 2e-05,
+      "loss": 0.0216,
+      "step": 36783
+    },
+    {
+      "epoch": 0.8755593639912406,
+      "grad_norm": 14.668275126969107,
+      "learning_rate": 2e-05,
+      "loss": 1.222,
+      "step": 36784
+    },
+    {
+      "epoch": 0.8755831667142722,
+      "grad_norm": 4.735704125579973,
+      "learning_rate": 2e-05,
+      "loss": 0.7388,
+      "step": 36785
+    },
+    {
+      "epoch": 0.8756069694373037,
+      "grad_norm": 0.6173245841433205,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 36786
+    },
+    {
+      "epoch": 0.8756307721603351,
+      "grad_norm": 10.976677120739293,
+      "learning_rate": 2e-05,
+      "loss": 1.0118,
+      "step": 36787
+    },
+    {
+      "epoch": 0.8756545748833666,
+      "grad_norm": 4.689479660516267,
+      "learning_rate": 2e-05,
+      "loss": 0.5343,
+      "step": 36788
+    },
+    {
+      "epoch": 0.8756783776063982,
+      "grad_norm": 13.143188574003311,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 36789
+    },
+    {
+      "epoch": 0.8757021803294297,
+      "grad_norm": 2.979288325748791,
+      "learning_rate": 2e-05,
+      "loss": 0.3152,
+      "step": 36790
+    },
+    {
+      "epoch": 0.8757259830524612,
+      "grad_norm": 3.84496011704384,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 36791
+    },
+    {
+      "epoch": 0.8757497857754927,
+      "grad_norm": 1.557259839715303,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 36792
+    },
+    {
+      "epoch": 0.8757735884985243,
+      "grad_norm": 16.14027125514236,
+      "learning_rate": 2e-05,
+      "loss": 0.7375,
+      "step": 36793
+    },
+    {
+      "epoch": 0.8757973912215558,
+      "grad_norm": 2.7838095284572293,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 36794
+    },
+    {
+      "epoch": 0.8758211939445872,
+      "grad_norm": 10.760328756391813,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 36795
+    },
+    {
+      "epoch": 0.8758449966676187,
+      "grad_norm": 2.0595758947106693,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 36796
+    },
+    {
+      "epoch": 0.8758687993906503,
+      "grad_norm": 0.06777754401668044,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 36797
+    },
+    {
+      "epoch": 0.8758926021136818,
+      "grad_norm": 2.5819429540522423,
+      "learning_rate": 2e-05,
+      "loss": 0.3621,
+      "step": 36798
+    },
+    {
+      "epoch": 0.8759164048367133,
+      "grad_norm": 0.025393093472921933,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 36799
+    },
+    {
+      "epoch": 0.8759402075597449,
+      "grad_norm": 1.6721582016044356,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 36800
+    },
+    {
+      "epoch": 0.8759640102827764,
+      "grad_norm": 3.477375239306548,
+      "learning_rate": 2e-05,
+      "loss": 0.1823,
+      "step": 36801
+    },
+    {
+      "epoch": 0.8759878130058079,
+      "grad_norm": 3.273388367652183,
+      "learning_rate": 2e-05,
+      "loss": 0.6619,
+      "step": 36802
+    },
+    {
+      "epoch": 0.8760116157288393,
+      "grad_norm": 11.280301738233362,
+      "learning_rate": 2e-05,
+      "loss": 0.3477,
+      "step": 36803
+    },
+    {
+      "epoch": 0.8760354184518709,
+      "grad_norm": 0.5707336345322717,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 36804
+    },
+    {
+      "epoch": 0.8760592211749024,
+      "grad_norm": 0.2863198048217272,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 36805
+    },
+    {
+      "epoch": 0.8760830238979339,
+      "grad_norm": 0.17833515701112063,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 36806
+    },
+    {
+      "epoch": 0.8761068266209654,
+      "grad_norm": 8.27505589182059,
+      "learning_rate": 2e-05,
+      "loss": 0.8577,
+      "step": 36807
+    },
+    {
+      "epoch": 0.876130629343997,
+      "grad_norm": 0.19781241522198045,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 36808
+    },
+    {
+      "epoch": 0.8761544320670285,
+      "grad_norm": 0.042134158738118925,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 36809
+    },
+    {
+      "epoch": 0.87617823479006,
+      "grad_norm": 6.314440624895498,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 36810
+    },
+    {
+      "epoch": 0.8762020375130914,
+      "grad_norm": 1.5156422987642502,
+      "learning_rate": 2e-05,
+      "loss": 0.1162,
+      "step": 36811
+    },
+    {
+      "epoch": 0.876225840236123,
+      "grad_norm": 3.7849090368726963,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 36812
+    },
+    {
+      "epoch": 0.8762496429591545,
+      "grad_norm": 11.761889619588098,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 36813
+    },
+    {
+      "epoch": 0.876273445682186,
+      "grad_norm": 11.212075895042672,
+      "learning_rate": 2e-05,
+      "loss": 1.0083,
+      "step": 36814
+    },
+    {
+      "epoch": 0.8762972484052176,
+      "grad_norm": 7.698917357371685,
+      "learning_rate": 2e-05,
+      "loss": 0.0922,
+      "step": 36815
+    },
+    {
+      "epoch": 0.8763210511282491,
+      "grad_norm": 2.899204408541871,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 36816
+    },
+    {
+      "epoch": 0.8763448538512806,
+      "grad_norm": 2.6040519516779472,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 36817
+    },
+    {
+      "epoch": 0.8763686565743121,
+      "grad_norm": 5.617326259676728,
+      "learning_rate": 2e-05,
+      "loss": 0.1278,
+      "step": 36818
+    },
+    {
+      "epoch": 0.8763924592973437,
+      "grad_norm": 6.472852858132638,
+      "learning_rate": 2e-05,
+      "loss": 0.3616,
+      "step": 36819
+    },
+    {
+      "epoch": 0.8764162620203751,
+      "grad_norm": 3.3395999084390593,
+      "learning_rate": 2e-05,
+      "loss": 0.5521,
+      "step": 36820
+    },
+    {
+      "epoch": 0.8764400647434066,
+      "grad_norm": 5.206061262070265,
+      "learning_rate": 2e-05,
+      "loss": 0.5336,
+      "step": 36821
+    },
+    {
+      "epoch": 0.8764638674664381,
+      "grad_norm": 0.033493614041727675,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 36822
+    },
+    {
+      "epoch": 0.8764876701894697,
+      "grad_norm": 6.272456670878909,
+      "learning_rate": 2e-05,
+      "loss": 0.0938,
+      "step": 36823
+    },
+    {
+      "epoch": 0.8765114729125012,
+      "grad_norm": 3.396974127714361,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 36824
+    },
+    {
+      "epoch": 0.8765352756355327,
+      "grad_norm": 9.090979757735687,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 36825
+    },
+    {
+      "epoch": 0.8765590783585642,
+      "grad_norm": 4.8861946283174325,
+      "learning_rate": 2e-05,
+      "loss": 0.7803,
+      "step": 36826
+    },
+    {
+      "epoch": 0.8765828810815958,
+      "grad_norm": 1.4696991412216223,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 36827
+    },
+    {
+      "epoch": 0.8766066838046273,
+      "grad_norm": 4.1339401414734835,
+      "learning_rate": 2e-05,
+      "loss": 0.8636,
+      "step": 36828
+    },
+    {
+      "epoch": 0.8766304865276587,
+      "grad_norm": 0.090556393376954,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 36829
+    },
+    {
+      "epoch": 0.8766542892506903,
+      "grad_norm": 7.387275929016173,
+      "learning_rate": 2e-05,
+      "loss": 0.622,
+      "step": 36830
+    },
+    {
+      "epoch": 0.8766780919737218,
+      "grad_norm": 5.576629069737893,
+      "learning_rate": 2e-05,
+      "loss": 1.3926,
+      "step": 36831
+    },
+    {
+      "epoch": 0.8767018946967533,
+      "grad_norm": 7.138557675374927,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 36832
+    },
+    {
+      "epoch": 0.8767256974197848,
+      "grad_norm": 2.6942784090537377,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 36833
+    },
+    {
+      "epoch": 0.8767495001428164,
+      "grad_norm": 4.699740306832436,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 36834
+    },
+    {
+      "epoch": 0.8767733028658479,
+      "grad_norm": 5.310642367909546,
+      "learning_rate": 2e-05,
+      "loss": 0.7391,
+      "step": 36835
+    },
+    {
+      "epoch": 0.8767971055888794,
+      "grad_norm": 0.5444068402336164,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 36836
+    },
+    {
+      "epoch": 0.8768209083119108,
+      "grad_norm": 4.6046229230217515,
+      "learning_rate": 2e-05,
+      "loss": 0.7156,
+      "step": 36837
+    },
+    {
+      "epoch": 0.8768447110349424,
+      "grad_norm": 8.78726719613331,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 36838
+    },
+    {
+      "epoch": 0.8768685137579739,
+      "grad_norm": 7.228313724171865,
+      "learning_rate": 2e-05,
+      "loss": 0.9448,
+      "step": 36839
+    },
+    {
+      "epoch": 0.8768923164810054,
+      "grad_norm": 5.293439171073273,
+      "learning_rate": 2e-05,
+      "loss": 1.2233,
+      "step": 36840
+    },
+    {
+      "epoch": 0.8769161192040369,
+      "grad_norm": 10.445634184120612,
+      "learning_rate": 2e-05,
+      "loss": 0.349,
+      "step": 36841
+    },
+    {
+      "epoch": 0.8769399219270685,
+      "grad_norm": 4.982166448107588,
+      "learning_rate": 2e-05,
+      "loss": 0.6769,
+      "step": 36842
+    },
+    {
+      "epoch": 0.8769637246501,
+      "grad_norm": 6.457671371241047,
+      "learning_rate": 2e-05,
+      "loss": 0.2011,
+      "step": 36843
+    },
+    {
+      "epoch": 0.8769875273731315,
+      "grad_norm": 6.357548715172366,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 36844
+    },
+    {
+      "epoch": 0.877011330096163,
+      "grad_norm": 1.8405263246501788,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 36845
+    },
+    {
+      "epoch": 0.8770351328191945,
+      "grad_norm": 2.5613186225316396,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 36846
+    },
+    {
+      "epoch": 0.877058935542226,
+      "grad_norm": 3.004996945168401,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 36847
+    },
+    {
+      "epoch": 0.8770827382652575,
+      "grad_norm": 6.258700307978134,
+      "learning_rate": 2e-05,
+      "loss": 0.7049,
+      "step": 36848
+    },
+    {
+      "epoch": 0.8771065409882891,
+      "grad_norm": 0.2861982640607908,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 36849
+    },
+    {
+      "epoch": 0.8771303437113206,
+      "grad_norm": 5.517554078931571,
+      "learning_rate": 2e-05,
+      "loss": 1.3132,
+      "step": 36850
+    },
+    {
+      "epoch": 0.8771541464343521,
+      "grad_norm": 0.2743448309759208,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 36851
+    },
+    {
+      "epoch": 0.8771779491573836,
+      "grad_norm": 6.486067611036331,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 36852
+    },
+    {
+      "epoch": 0.8772017518804152,
+      "grad_norm": 3.6686389591988684,
+      "learning_rate": 2e-05,
+      "loss": 0.7892,
+      "step": 36853
+    },
+    {
+      "epoch": 0.8772255546034466,
+      "grad_norm": 2.069996859858869,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 36854
+    },
+    {
+      "epoch": 0.8772493573264781,
+      "grad_norm": 4.418216547366475,
+      "learning_rate": 2e-05,
+      "loss": 0.3855,
+      "step": 36855
+    },
+    {
+      "epoch": 0.8772731600495096,
+      "grad_norm": 3.483089452009392,
+      "learning_rate": 2e-05,
+      "loss": 0.7491,
+      "step": 36856
+    },
+    {
+      "epoch": 0.8772969627725412,
+      "grad_norm": 3.1175386283384343,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 36857
+    },
+    {
+      "epoch": 0.8773207654955727,
+      "grad_norm": 5.389536548493456,
+      "learning_rate": 2e-05,
+      "loss": 1.2317,
+      "step": 36858
+    },
+    {
+      "epoch": 0.8773445682186042,
+      "grad_norm": 1.4659821612117245,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 36859
+    },
+    {
+      "epoch": 0.8773683709416358,
+      "grad_norm": 2.797301606023109,
+      "learning_rate": 2e-05,
+      "loss": 0.4947,
+      "step": 36860
+    },
+    {
+      "epoch": 0.8773921736646673,
+      "grad_norm": 5.050338398712322,
+      "learning_rate": 2e-05,
+      "loss": 0.6089,
+      "step": 36861
+    },
+    {
+      "epoch": 0.8774159763876987,
+      "grad_norm": 5.556449745207864,
+      "learning_rate": 2e-05,
+      "loss": 0.1511,
+      "step": 36862
+    },
+    {
+      "epoch": 0.8774397791107302,
+      "grad_norm": 1.085336483684764,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 36863
+    },
+    {
+      "epoch": 0.8774635818337618,
+      "grad_norm": 7.416453503748104,
+      "learning_rate": 2e-05,
+      "loss": 0.5328,
+      "step": 36864
+    },
+    {
+      "epoch": 0.8774873845567933,
+      "grad_norm": 2.920794625220426,
+      "learning_rate": 2e-05,
+      "loss": 0.5704,
+      "step": 36865
+    },
+    {
+      "epoch": 0.8775111872798248,
+      "grad_norm": 2.1007997271933894,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 36866
+    },
+    {
+      "epoch": 0.8775349900028563,
+      "grad_norm": 1.8185843542275701,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 36867
+    },
+    {
+      "epoch": 0.8775587927258879,
+      "grad_norm": 3.7889173605143878,
+      "learning_rate": 2e-05,
+      "loss": 0.8102,
+      "step": 36868
+    },
+    {
+      "epoch": 0.8775825954489194,
+      "grad_norm": 0.7318710703189155,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 36869
+    },
+    {
+      "epoch": 0.8776063981719509,
+      "grad_norm": 5.504926020221404,
+      "learning_rate": 2e-05,
+      "loss": 0.2549,
+      "step": 36870
+    },
+    {
+      "epoch": 0.8776302008949823,
+      "grad_norm": 0.25265028935425454,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 36871
+    },
+    {
+      "epoch": 0.8776540036180139,
+      "grad_norm": 9.788692233176398,
+      "learning_rate": 2e-05,
+      "loss": 0.2373,
+      "step": 36872
+    },
+    {
+      "epoch": 0.8776778063410454,
+      "grad_norm": 30.342772352489447,
+      "learning_rate": 2e-05,
+      "loss": 0.8425,
+      "step": 36873
+    },
+    {
+      "epoch": 0.8777016090640769,
+      "grad_norm": 2.246232728553719,
+      "learning_rate": 2e-05,
+      "loss": 0.1239,
+      "step": 36874
+    },
+    {
+      "epoch": 0.8777254117871085,
+      "grad_norm": 3.618803029041722,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 36875
+    },
+    {
+      "epoch": 0.87774921451014,
+      "grad_norm": 2.9558054364161745,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 36876
+    },
+    {
+      "epoch": 0.8777730172331715,
+      "grad_norm": 2.242220511034648,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 36877
+    },
+    {
+      "epoch": 0.877796819956203,
+      "grad_norm": 1.961534382489636,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 36878
+    },
+    {
+      "epoch": 0.8778206226792346,
+      "grad_norm": 16.268322864258046,
+      "learning_rate": 2e-05,
+      "loss": 1.8288,
+      "step": 36879
+    },
+    {
+      "epoch": 0.877844425402266,
+      "grad_norm": 3.4149033157678392,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 36880
+    },
+    {
+      "epoch": 0.8778682281252975,
+      "grad_norm": 0.6937277054497826,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 36881
+    },
+    {
+      "epoch": 0.877892030848329,
+      "grad_norm": 5.3001368696620785,
+      "learning_rate": 2e-05,
+      "loss": 0.2589,
+      "step": 36882
+    },
+    {
+      "epoch": 0.8779158335713606,
+      "grad_norm": 4.410512540017873,
+      "learning_rate": 2e-05,
+      "loss": 0.3287,
+      "step": 36883
+    },
+    {
+      "epoch": 0.8779396362943921,
+      "grad_norm": 0.6424703482713535,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 36884
+    },
+    {
+      "epoch": 0.8779634390174236,
+      "grad_norm": 6.815542737961392,
+      "learning_rate": 2e-05,
+      "loss": 0.2856,
+      "step": 36885
+    },
+    {
+      "epoch": 0.8779872417404551,
+      "grad_norm": 11.29580554886867,
+      "learning_rate": 2e-05,
+      "loss": 0.8476,
+      "step": 36886
+    },
+    {
+      "epoch": 0.8780110444634867,
+      "grad_norm": 19.62800283041115,
+      "learning_rate": 2e-05,
+      "loss": 0.5049,
+      "step": 36887
+    },
+    {
+      "epoch": 0.8780348471865181,
+      "grad_norm": 7.811211391759708,
+      "learning_rate": 2e-05,
+      "loss": 1.489,
+      "step": 36888
+    },
+    {
+      "epoch": 0.8780586499095496,
+      "grad_norm": 1.8744926921685476,
+      "learning_rate": 2e-05,
+      "loss": 0.0443,
+      "step": 36889
+    },
+    {
+      "epoch": 0.8780824526325811,
+      "grad_norm": 5.203563862076432,
+      "learning_rate": 2e-05,
+      "loss": 0.9318,
+      "step": 36890
+    },
+    {
+      "epoch": 0.8781062553556127,
+      "grad_norm": 3.5088269153861398,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 36891
+    },
+    {
+      "epoch": 0.8781300580786442,
+      "grad_norm": 5.789335938910607,
+      "learning_rate": 2e-05,
+      "loss": 0.8995,
+      "step": 36892
+    },
+    {
+      "epoch": 0.8781538608016757,
+      "grad_norm": 0.9214228825334126,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 36893
+    },
+    {
+      "epoch": 0.8781776635247073,
+      "grad_norm": 3.7583041433539885,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 36894
+    },
+    {
+      "epoch": 0.8782014662477388,
+      "grad_norm": 0.5894222173159327,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 36895
+    },
+    {
+      "epoch": 0.8782252689707702,
+      "grad_norm": 0.0434676554975568,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 36896
+    },
+    {
+      "epoch": 0.8782490716938017,
+      "grad_norm": 0.22528015122534908,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 36897
+    },
+    {
+      "epoch": 0.8782728744168333,
+      "grad_norm": 1.6702484724217328,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 36898
+    },
+    {
+      "epoch": 0.8782966771398648,
+      "grad_norm": 6.487744866341394,
+      "learning_rate": 2e-05,
+      "loss": 0.7478,
+      "step": 36899
+    },
+    {
+      "epoch": 0.8783204798628963,
+      "grad_norm": 0.1482902062775206,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 36900
+    },
+    {
+      "epoch": 0.8783442825859278,
+      "grad_norm": 6.3265823989397765,
+      "learning_rate": 2e-05,
+      "loss": 1.0219,
+      "step": 36901
+    },
+    {
+      "epoch": 0.8783680853089594,
+      "grad_norm": 3.2347685733100016,
+      "learning_rate": 2e-05,
+      "loss": 0.8512,
+      "step": 36902
+    },
+    {
+      "epoch": 0.8783918880319909,
+      "grad_norm": 0.12826727721815942,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 36903
+    },
+    {
+      "epoch": 0.8784156907550223,
+      "grad_norm": 7.358570443261873,
+      "learning_rate": 2e-05,
+      "loss": 0.4878,
+      "step": 36904
+    },
+    {
+      "epoch": 0.8784394934780538,
+      "grad_norm": 11.034920528075736,
+      "learning_rate": 2e-05,
+      "loss": 1.4334,
+      "step": 36905
+    },
+    {
+      "epoch": 0.8784632962010854,
+      "grad_norm": 22.656675645134477,
+      "learning_rate": 2e-05,
+      "loss": 0.8287,
+      "step": 36906
+    },
+    {
+      "epoch": 0.8784870989241169,
+      "grad_norm": 1.0047193369587355,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 36907
+    },
+    {
+      "epoch": 0.8785109016471484,
+      "grad_norm": 0.06990310844462214,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 36908
+    },
+    {
+      "epoch": 0.87853470437018,
+      "grad_norm": 6.564869246350722,
+      "learning_rate": 2e-05,
+      "loss": 0.1202,
+      "step": 36909
+    },
+    {
+      "epoch": 0.8785585070932115,
+      "grad_norm": 1.4506927667220166,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 36910
+    },
+    {
+      "epoch": 0.878582309816243,
+      "grad_norm": 10.17531880251226,
+      "learning_rate": 2e-05,
+      "loss": 0.9692,
+      "step": 36911
+    },
+    {
+      "epoch": 0.8786061125392745,
+      "grad_norm": 1.2337256764538322,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 36912
+    },
+    {
+      "epoch": 0.878629915262306,
+      "grad_norm": 2.4671367821176307,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 36913
+    },
+    {
+      "epoch": 0.8786537179853375,
+      "grad_norm": 2.359450582709513,
+      "learning_rate": 2e-05,
+      "loss": 0.0506,
+      "step": 36914
+    },
+    {
+      "epoch": 0.878677520708369,
+      "grad_norm": 0.2563374698505843,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 36915
+    },
+    {
+      "epoch": 0.8787013234314005,
+      "grad_norm": 3.0597656300393083,
+      "learning_rate": 2e-05,
+      "loss": 0.5587,
+      "step": 36916
+    },
+    {
+      "epoch": 0.8787251261544321,
+      "grad_norm": 2.949129483938671,
+      "learning_rate": 2e-05,
+      "loss": 0.4823,
+      "step": 36917
+    },
+    {
+      "epoch": 0.8787489288774636,
+      "grad_norm": 5.260227947322991,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 36918
+    },
+    {
+      "epoch": 0.8787727316004951,
+      "grad_norm": 4.424394292831709,
+      "learning_rate": 2e-05,
+      "loss": 0.8175,
+      "step": 36919
+    },
+    {
+      "epoch": 0.8787965343235266,
+      "grad_norm": 0.4624484473016281,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 36920
+    },
+    {
+      "epoch": 0.8788203370465582,
+      "grad_norm": 2.83508232498983,
+      "learning_rate": 2e-05,
+      "loss": 0.3532,
+      "step": 36921
+    },
+    {
+      "epoch": 0.8788441397695896,
+      "grad_norm": 1.755202922082068,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 36922
+    },
+    {
+      "epoch": 0.8788679424926211,
+      "grad_norm": 0.10296664700847438,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 36923
+    },
+    {
+      "epoch": 0.8788917452156527,
+      "grad_norm": 1.130354830324667,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 36924
+    },
+    {
+      "epoch": 0.8789155479386842,
+      "grad_norm": 5.419834321068761,
+      "learning_rate": 2e-05,
+      "loss": 0.9609,
+      "step": 36925
+    },
+    {
+      "epoch": 0.8789393506617157,
+      "grad_norm": 3.856008542244154,
+      "learning_rate": 2e-05,
+      "loss": 0.5014,
+      "step": 36926
+    },
+    {
+      "epoch": 0.8789631533847472,
+      "grad_norm": 2.785936797893677,
+      "learning_rate": 2e-05,
+      "loss": 0.0595,
+      "step": 36927
+    },
+    {
+      "epoch": 0.8789869561077788,
+      "grad_norm": 3.2352874897657626,
+      "learning_rate": 2e-05,
+      "loss": 0.4925,
+      "step": 36928
+    },
+    {
+      "epoch": 0.8790107588308103,
+      "grad_norm": 0.10060530821972063,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 36929
+    },
+    {
+      "epoch": 0.8790345615538417,
+      "grad_norm": 4.450837908473101,
+      "learning_rate": 2e-05,
+      "loss": 0.5026,
+      "step": 36930
+    },
+    {
+      "epoch": 0.8790583642768732,
+      "grad_norm": 10.80756425261052,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 36931
+    },
+    {
+      "epoch": 0.8790821669999048,
+      "grad_norm": 0.37863442973739914,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 36932
+    },
+    {
+      "epoch": 0.8791059697229363,
+      "grad_norm": 0.3489668039923401,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 36933
+    },
+    {
+      "epoch": 0.8791297724459678,
+      "grad_norm": 10.633990480955102,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 36934
+    },
+    {
+      "epoch": 0.8791535751689993,
+      "grad_norm": 3.485667041351472,
+      "learning_rate": 2e-05,
+      "loss": 0.7541,
+      "step": 36935
+    },
+    {
+      "epoch": 0.8791773778920309,
+      "grad_norm": 3.8331354959562995,
+      "learning_rate": 2e-05,
+      "loss": 0.5212,
+      "step": 36936
+    },
+    {
+      "epoch": 0.8792011806150624,
+      "grad_norm": 1.1046200357602804,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 36937
+    },
+    {
+      "epoch": 0.8792249833380938,
+      "grad_norm": 0.0856524474489517,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 36938
+    },
+    {
+      "epoch": 0.8792487860611254,
+      "grad_norm": 0.9538140650176531,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 36939
+    },
+    {
+      "epoch": 0.8792725887841569,
+      "grad_norm": 1.520694434937211,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 36940
+    },
+    {
+      "epoch": 0.8792963915071884,
+      "grad_norm": 3.0021202790743686,
+      "learning_rate": 2e-05,
+      "loss": 0.4016,
+      "step": 36941
+    },
+    {
+      "epoch": 0.8793201942302199,
+      "grad_norm": 2.8451496739108726,
+      "learning_rate": 2e-05,
+      "loss": 0.2768,
+      "step": 36942
+    },
+    {
+      "epoch": 0.8793439969532515,
+      "grad_norm": 0.013745195742141794,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36943
+    },
+    {
+      "epoch": 0.879367799676283,
+      "grad_norm": 0.09016729029102963,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 36944
+    },
+    {
+      "epoch": 0.8793916023993145,
+      "grad_norm": 3.9222095399171106,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 36945
+    },
+    {
+      "epoch": 0.879415405122346,
+      "grad_norm": 19.733071681900455,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 36946
+    },
+    {
+      "epoch": 0.8794392078453775,
+      "grad_norm": 3.5426770609187392,
+      "learning_rate": 2e-05,
+      "loss": 0.4642,
+      "step": 36947
+    },
+    {
+      "epoch": 0.879463010568409,
+      "grad_norm": 0.7910164508156409,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 36948
+    },
+    {
+      "epoch": 0.8794868132914405,
+      "grad_norm": 4.844699204646334,
+      "learning_rate": 2e-05,
+      "loss": 0.5144,
+      "step": 36949
+    },
+    {
+      "epoch": 0.879510616014472,
+      "grad_norm": 5.90833072783586,
+      "learning_rate": 2e-05,
+      "loss": 0.7251,
+      "step": 36950
+    },
+    {
+      "epoch": 0.8795344187375036,
+      "grad_norm": 10.38774847479212,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 36951
+    },
+    {
+      "epoch": 0.8795582214605351,
+      "grad_norm": 2.2822096073215343,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 36952
+    },
+    {
+      "epoch": 0.8795820241835666,
+      "grad_norm": 4.694630459708338,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 36953
+    },
+    {
+      "epoch": 0.8796058269065982,
+      "grad_norm": 4.043679058439298,
+      "learning_rate": 2e-05,
+      "loss": 0.4956,
+      "step": 36954
+    },
+    {
+      "epoch": 0.8796296296296297,
+      "grad_norm": 7.140726562789088,
+      "learning_rate": 2e-05,
+      "loss": 0.9169,
+      "step": 36955
+    },
+    {
+      "epoch": 0.8796534323526611,
+      "grad_norm": 2.217040547980014,
+      "learning_rate": 2e-05,
+      "loss": 0.1999,
+      "step": 36956
+    },
+    {
+      "epoch": 0.8796772350756926,
+      "grad_norm": 5.01092609990313,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 36957
+    },
+    {
+      "epoch": 0.8797010377987242,
+      "grad_norm": 3.966119587290832,
+      "learning_rate": 2e-05,
+      "loss": 0.9469,
+      "step": 36958
+    },
+    {
+      "epoch": 0.8797248405217557,
+      "grad_norm": 3.970123288853627,
+      "learning_rate": 2e-05,
+      "loss": 0.5613,
+      "step": 36959
+    },
+    {
+      "epoch": 0.8797486432447872,
+      "grad_norm": 7.481976478463895,
+      "learning_rate": 2e-05,
+      "loss": 1.9433,
+      "step": 36960
+    },
+    {
+      "epoch": 0.8797724459678187,
+      "grad_norm": 0.1292571843213501,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 36961
+    },
+    {
+      "epoch": 0.8797962486908503,
+      "grad_norm": 1.9693903243146385,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 36962
+    },
+    {
+      "epoch": 0.8798200514138818,
+      "grad_norm": 2.5536694069837425,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 36963
+    },
+    {
+      "epoch": 0.8798438541369132,
+      "grad_norm": 25.85110504862496,
+      "learning_rate": 2e-05,
+      "loss": 1.4502,
+      "step": 36964
+    },
+    {
+      "epoch": 0.8798676568599447,
+      "grad_norm": 4.982504292325681,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 36965
+    },
+    {
+      "epoch": 0.8798914595829763,
+      "grad_norm": 0.1741188544873642,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 36966
+    },
+    {
+      "epoch": 0.8799152623060078,
+      "grad_norm": 11.780925860263794,
+      "learning_rate": 2e-05,
+      "loss": 0.7004,
+      "step": 36967
+    },
+    {
+      "epoch": 0.8799390650290393,
+      "grad_norm": 0.010839607040014838,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 36968
+    },
+    {
+      "epoch": 0.8799628677520709,
+      "grad_norm": 0.11913573465475946,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 36969
+    },
+    {
+      "epoch": 0.8799866704751024,
+      "grad_norm": 5.244817272525145,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 36970
+    },
+    {
+      "epoch": 0.8800104731981339,
+      "grad_norm": 5.269148516699665,
+      "learning_rate": 2e-05,
+      "loss": 0.4869,
+      "step": 36971
+    },
+    {
+      "epoch": 0.8800342759211653,
+      "grad_norm": 6.797550789024731,
+      "learning_rate": 2e-05,
+      "loss": 2.3048,
+      "step": 36972
+    },
+    {
+      "epoch": 0.8800580786441969,
+      "grad_norm": 0.12093480237269254,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 36973
+    },
+    {
+      "epoch": 0.8800818813672284,
+      "grad_norm": 2.5522843260756294,
+      "learning_rate": 2e-05,
+      "loss": 0.379,
+      "step": 36974
+    },
+    {
+      "epoch": 0.8801056840902599,
+      "grad_norm": 0.19690096167538967,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 36975
+    },
+    {
+      "epoch": 0.8801294868132914,
+      "grad_norm": 0.12475197270198267,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 36976
+    },
+    {
+      "epoch": 0.880153289536323,
+      "grad_norm": 4.1605026934270875,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 36977
+    },
+    {
+      "epoch": 0.8801770922593545,
+      "grad_norm": 5.238596003554426,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 36978
+    },
+    {
+      "epoch": 0.880200894982386,
+      "grad_norm": 2.494542286298569,
+      "learning_rate": 2e-05,
+      "loss": 0.0528,
+      "step": 36979
+    },
+    {
+      "epoch": 0.8802246977054174,
+      "grad_norm": 2.196033861285962,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 36980
+    },
+    {
+      "epoch": 0.880248500428449,
+      "grad_norm": 0.48274949665804134,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 36981
+    },
+    {
+      "epoch": 0.8802723031514805,
+      "grad_norm": 4.4525507696727775,
+      "learning_rate": 2e-05,
+      "loss": 0.5274,
+      "step": 36982
+    },
+    {
+      "epoch": 0.880296105874512,
+      "grad_norm": 0.280139212195295,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 36983
+    },
+    {
+      "epoch": 0.8803199085975436,
+      "grad_norm": 0.18736288615507962,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 36984
+    },
+    {
+      "epoch": 0.8803437113205751,
+      "grad_norm": 3.1505482255911734,
+      "learning_rate": 2e-05,
+      "loss": 0.369,
+      "step": 36985
+    },
+    {
+      "epoch": 0.8803675140436066,
+      "grad_norm": 0.1334870361597834,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 36986
+    },
+    {
+      "epoch": 0.8803913167666381,
+      "grad_norm": 2.355181390435297,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 36987
+    },
+    {
+      "epoch": 0.8804151194896697,
+      "grad_norm": 6.052637254970047,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 36988
+    },
+    {
+      "epoch": 0.8804389222127011,
+      "grad_norm": 9.989443720145344,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 36989
+    },
+    {
+      "epoch": 0.8804627249357326,
+      "grad_norm": 5.351235759486704,
+      "learning_rate": 2e-05,
+      "loss": 0.7188,
+      "step": 36990
+    },
+    {
+      "epoch": 0.8804865276587641,
+      "grad_norm": 0.27112235634625315,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 36991
+    },
+    {
+      "epoch": 0.8805103303817957,
+      "grad_norm": 3.4642816746542344,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 36992
+    },
+    {
+      "epoch": 0.8805341331048272,
+      "grad_norm": 4.171151064909699,
+      "learning_rate": 2e-05,
+      "loss": 0.441,
+      "step": 36993
+    },
+    {
+      "epoch": 0.8805579358278587,
+      "grad_norm": 8.365683597679249,
+      "learning_rate": 2e-05,
+      "loss": 0.2448,
+      "step": 36994
+    },
+    {
+      "epoch": 0.8805817385508902,
+      "grad_norm": 4.724327072869331,
+      "learning_rate": 2e-05,
+      "loss": 1.1445,
+      "step": 36995
+    },
+    {
+      "epoch": 0.8806055412739218,
+      "grad_norm": 7.742450754762049,
+      "learning_rate": 2e-05,
+      "loss": 0.7281,
+      "step": 36996
+    },
+    {
+      "epoch": 0.8806293439969533,
+      "grad_norm": 0.28250618308101266,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 36997
+    },
+    {
+      "epoch": 0.8806531467199847,
+      "grad_norm": 12.757601407871821,
+      "learning_rate": 2e-05,
+      "loss": 0.8496,
+      "step": 36998
+    },
+    {
+      "epoch": 0.8806769494430163,
+      "grad_norm": 8.610679842045458,
+      "learning_rate": 2e-05,
+      "loss": 0.1829,
+      "step": 36999
+    },
+    {
+      "epoch": 0.8807007521660478,
+      "grad_norm": 0.49652836632480496,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 37000
+    },
+    {
+      "epoch": 0.8807245548890793,
+      "grad_norm": 3.7602330003098587,
+      "learning_rate": 2e-05,
+      "loss": 0.8628,
+      "step": 37001
+    },
+    {
+      "epoch": 0.8807483576121108,
+      "grad_norm": 4.08769337363828,
+      "learning_rate": 2e-05,
+      "loss": 0.7328,
+      "step": 37002
+    },
+    {
+      "epoch": 0.8807721603351424,
+      "grad_norm": 5.930877419677425,
+      "learning_rate": 2e-05,
+      "loss": 0.2228,
+      "step": 37003
+    },
+    {
+      "epoch": 0.8807959630581739,
+      "grad_norm": 5.245842224924583,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 37004
+    },
+    {
+      "epoch": 0.8808197657812054,
+      "grad_norm": 0.06251339801473203,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37005
+    },
+    {
+      "epoch": 0.8808435685042368,
+      "grad_norm": 1.8713343915733507,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 37006
+    },
+    {
+      "epoch": 0.8808673712272684,
+      "grad_norm": 7.013275696443347,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 37007
+    },
+    {
+      "epoch": 0.8808911739502999,
+      "grad_norm": 1.0052512441925936,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 37008
+    },
+    {
+      "epoch": 0.8809149766733314,
+      "grad_norm": 2.254957353325219,
+      "learning_rate": 2e-05,
+      "loss": 0.2784,
+      "step": 37009
+    },
+    {
+      "epoch": 0.8809387793963629,
+      "grad_norm": 9.534922442551691,
+      "learning_rate": 2e-05,
+      "loss": 1.0121,
+      "step": 37010
+    },
+    {
+      "epoch": 0.8809625821193945,
+      "grad_norm": 10.633039514117664,
+      "learning_rate": 2e-05,
+      "loss": 0.5551,
+      "step": 37011
+    },
+    {
+      "epoch": 0.880986384842426,
+      "grad_norm": 9.791534842037919,
+      "learning_rate": 2e-05,
+      "loss": 0.1651,
+      "step": 37012
+    },
+    {
+      "epoch": 0.8810101875654575,
+      "grad_norm": 3.3823182308923596,
+      "learning_rate": 2e-05,
+      "loss": 0.8169,
+      "step": 37013
+    },
+    {
+      "epoch": 0.881033990288489,
+      "grad_norm": 0.059110867611186224,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37014
+    },
+    {
+      "epoch": 0.8810577930115205,
+      "grad_norm": 3.767368905451569,
+      "learning_rate": 2e-05,
+      "loss": 0.7551,
+      "step": 37015
+    },
+    {
+      "epoch": 0.881081595734552,
+      "grad_norm": 2.5727610722965664,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 37016
+    },
+    {
+      "epoch": 0.8811053984575835,
+      "grad_norm": 3.0359090477556845,
+      "learning_rate": 2e-05,
+      "loss": 0.2526,
+      "step": 37017
+    },
+    {
+      "epoch": 0.8811292011806151,
+      "grad_norm": 7.160736929816768,
+      "learning_rate": 2e-05,
+      "loss": 0.7403,
+      "step": 37018
+    },
+    {
+      "epoch": 0.8811530039036466,
+      "grad_norm": 0.5864112178250597,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37019
+    },
+    {
+      "epoch": 0.8811768066266781,
+      "grad_norm": 0.15462075726106242,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 37020
+    },
+    {
+      "epoch": 0.8812006093497096,
+      "grad_norm": 5.070142386173689,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 37021
+    },
+    {
+      "epoch": 0.8812244120727412,
+      "grad_norm": 4.3064905936300395,
+      "learning_rate": 2e-05,
+      "loss": 0.6331,
+      "step": 37022
+    },
+    {
+      "epoch": 0.8812482147957726,
+      "grad_norm": 0.051120236863170125,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37023
+    },
+    {
+      "epoch": 0.8812720175188041,
+      "grad_norm": 4.2017016166160355,
+      "learning_rate": 2e-05,
+      "loss": 0.818,
+      "step": 37024
+    },
+    {
+      "epoch": 0.8812958202418356,
+      "grad_norm": 5.879740917449853,
+      "learning_rate": 2e-05,
+      "loss": 0.3091,
+      "step": 37025
+    },
+    {
+      "epoch": 0.8813196229648672,
+      "grad_norm": 0.29083032130861,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 37026
+    },
+    {
+      "epoch": 0.8813434256878987,
+      "grad_norm": 0.3394870575227939,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 37027
+    },
+    {
+      "epoch": 0.8813672284109302,
+      "grad_norm": 7.530879818809476,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 37028
+    },
+    {
+      "epoch": 0.8813910311339618,
+      "grad_norm": 3.2071312592836576,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 37029
+    },
+    {
+      "epoch": 0.8814148338569933,
+      "grad_norm": 0.13865178902243808,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 37030
+    },
+    {
+      "epoch": 0.8814386365800247,
+      "grad_norm": 2.796353348934589,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 37031
+    },
+    {
+      "epoch": 0.8814624393030562,
+      "grad_norm": 0.03291786732975058,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37032
+    },
+    {
+      "epoch": 0.8814862420260878,
+      "grad_norm": 1.4912898364817864,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 37033
+    },
+    {
+      "epoch": 0.8815100447491193,
+      "grad_norm": 2.486969121650779,
+      "learning_rate": 2e-05,
+      "loss": 0.0556,
+      "step": 37034
+    },
+    {
+      "epoch": 0.8815338474721508,
+      "grad_norm": 3.934764266289514,
+      "learning_rate": 2e-05,
+      "loss": 0.6309,
+      "step": 37035
+    },
+    {
+      "epoch": 0.8815576501951823,
+      "grad_norm": 4.050314602442034,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 37036
+    },
+    {
+      "epoch": 0.8815814529182139,
+      "grad_norm": 4.06044985303781,
+      "learning_rate": 2e-05,
+      "loss": 0.6623,
+      "step": 37037
+    },
+    {
+      "epoch": 0.8816052556412454,
+      "grad_norm": 5.22177647052493,
+      "learning_rate": 2e-05,
+      "loss": 0.8807,
+      "step": 37038
+    },
+    {
+      "epoch": 0.8816290583642769,
+      "grad_norm": 8.404469398260314,
+      "learning_rate": 2e-05,
+      "loss": 0.3053,
+      "step": 37039
+    },
+    {
+      "epoch": 0.8816528610873083,
+      "grad_norm": 0.2603842119733633,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 37040
+    },
+    {
+      "epoch": 0.8816766638103399,
+      "grad_norm": 3.4392722041913233,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 37041
+    },
+    {
+      "epoch": 0.8817004665333714,
+      "grad_norm": 0.0907550908587215,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 37042
+    },
+    {
+      "epoch": 0.8817242692564029,
+      "grad_norm": 7.917550172033194,
+      "learning_rate": 2e-05,
+      "loss": 1.3732,
+      "step": 37043
+    },
+    {
+      "epoch": 0.8817480719794345,
+      "grad_norm": 5.705584789521307,
+      "learning_rate": 2e-05,
+      "loss": 0.4112,
+      "step": 37044
+    },
+    {
+      "epoch": 0.881771874702466,
+      "grad_norm": 3.21351120996688,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 37045
+    },
+    {
+      "epoch": 0.8817956774254975,
+      "grad_norm": 2.916800843861913,
+      "learning_rate": 2e-05,
+      "loss": 0.0688,
+      "step": 37046
+    },
+    {
+      "epoch": 0.881819480148529,
+      "grad_norm": 2.2095836537529343,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 37047
+    },
+    {
+      "epoch": 0.8818432828715606,
+      "grad_norm": 9.169406932865675,
+      "learning_rate": 2e-05,
+      "loss": 0.474,
+      "step": 37048
+    },
+    {
+      "epoch": 0.881867085594592,
+      "grad_norm": 0.33104816379072505,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 37049
+    },
+    {
+      "epoch": 0.8818908883176235,
+      "grad_norm": 4.849723300578299,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 37050
+    },
+    {
+      "epoch": 0.881914691040655,
+      "grad_norm": 4.683214811399736,
+      "learning_rate": 2e-05,
+      "loss": 0.9553,
+      "step": 37051
+    },
+    {
+      "epoch": 0.8819384937636866,
+      "grad_norm": 8.83508579563882,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 37052
+    },
+    {
+      "epoch": 0.8819622964867181,
+      "grad_norm": 0.08809794439578161,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37053
+    },
+    {
+      "epoch": 0.8819860992097496,
+      "grad_norm": 0.5535520138281166,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 37054
+    },
+    {
+      "epoch": 0.8820099019327811,
+      "grad_norm": 4.980721610986981,
+      "learning_rate": 2e-05,
+      "loss": 0.9668,
+      "step": 37055
+    },
+    {
+      "epoch": 0.8820337046558127,
+      "grad_norm": 4.492747622396682,
+      "learning_rate": 2e-05,
+      "loss": 0.6775,
+      "step": 37056
+    },
+    {
+      "epoch": 0.8820575073788441,
+      "grad_norm": 0.09400652208474218,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 37057
+    },
+    {
+      "epoch": 0.8820813101018756,
+      "grad_norm": 6.672710048103746,
+      "learning_rate": 2e-05,
+      "loss": 0.9468,
+      "step": 37058
+    },
+    {
+      "epoch": 0.8821051128249072,
+      "grad_norm": 0.07109074980044824,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 37059
+    },
+    {
+      "epoch": 0.8821289155479387,
+      "grad_norm": 2.710516686845368,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 37060
+    },
+    {
+      "epoch": 0.8821527182709702,
+      "grad_norm": 5.2308874573963795,
+      "learning_rate": 2e-05,
+      "loss": 0.6874,
+      "step": 37061
+    },
+    {
+      "epoch": 0.8821765209940017,
+      "grad_norm": 3.7675171478536704,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 37062
+    },
+    {
+      "epoch": 0.8822003237170333,
+      "grad_norm": 1.9765798430235544,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 37063
+    },
+    {
+      "epoch": 0.8822241264400648,
+      "grad_norm": 5.252216280391004,
+      "learning_rate": 2e-05,
+      "loss": 0.1614,
+      "step": 37064
+    },
+    {
+      "epoch": 0.8822479291630962,
+      "grad_norm": 4.2306005151124975,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 37065
+    },
+    {
+      "epoch": 0.8822717318861277,
+      "grad_norm": 0.1958868755802459,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 37066
+    },
+    {
+      "epoch": 0.8822955346091593,
+      "grad_norm": 4.054488546952375,
+      "learning_rate": 2e-05,
+      "loss": 0.6109,
+      "step": 37067
+    },
+    {
+      "epoch": 0.8823193373321908,
+      "grad_norm": 5.528789434690169,
+      "learning_rate": 2e-05,
+      "loss": 1.3407,
+      "step": 37068
+    },
+    {
+      "epoch": 0.8823431400552223,
+      "grad_norm": 0.17067267456242544,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37069
+    },
+    {
+      "epoch": 0.8823669427782538,
+      "grad_norm": 2.6529278526401074,
+      "learning_rate": 2e-05,
+      "loss": 0.3149,
+      "step": 37070
+    },
+    {
+      "epoch": 0.8823907455012854,
+      "grad_norm": 0.9705572945170476,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 37071
+    },
+    {
+      "epoch": 0.8824145482243169,
+      "grad_norm": 3.7085427636096715,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 37072
+    },
+    {
+      "epoch": 0.8824383509473483,
+      "grad_norm": 0.6015932247537713,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 37073
+    },
+    {
+      "epoch": 0.88246215367038,
+      "grad_norm": 7.170167730368912,
+      "learning_rate": 2e-05,
+      "loss": 0.4006,
+      "step": 37074
+    },
+    {
+      "epoch": 0.8824859563934114,
+      "grad_norm": 0.05923666063684396,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37075
+    },
+    {
+      "epoch": 0.8825097591164429,
+      "grad_norm": 1.26998633755016,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 37076
+    },
+    {
+      "epoch": 0.8825335618394744,
+      "grad_norm": 0.8784165929383267,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 37077
+    },
+    {
+      "epoch": 0.882557364562506,
+      "grad_norm": 2.811375261269046,
+      "learning_rate": 2e-05,
+      "loss": 0.3222,
+      "step": 37078
+    },
+    {
+      "epoch": 0.8825811672855375,
+      "grad_norm": 5.001736803970406,
+      "learning_rate": 2e-05,
+      "loss": 1.0336,
+      "step": 37079
+    },
+    {
+      "epoch": 0.882604970008569,
+      "grad_norm": 3.6428318674838316,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 37080
+    },
+    {
+      "epoch": 0.8826287727316005,
+      "grad_norm": 0.15331761715904482,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 37081
+    },
+    {
+      "epoch": 0.882652575454632,
+      "grad_norm": 6.2556813341819515,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 37082
+    },
+    {
+      "epoch": 0.8826763781776635,
+      "grad_norm": 3.6208149116657116,
+      "learning_rate": 2e-05,
+      "loss": 0.6403,
+      "step": 37083
+    },
+    {
+      "epoch": 0.882700180900695,
+      "grad_norm": 8.669940294522602,
+      "learning_rate": 2e-05,
+      "loss": 0.7281,
+      "step": 37084
+    },
+    {
+      "epoch": 0.8827239836237265,
+      "grad_norm": 0.13644089229801734,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 37085
+    },
+    {
+      "epoch": 0.8827477863467581,
+      "grad_norm": 0.13992405220489063,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 37086
+    },
+    {
+      "epoch": 0.8827715890697896,
+      "grad_norm": 0.8389599550437242,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 37087
+    },
+    {
+      "epoch": 0.8827953917928211,
+      "grad_norm": 0.18384843467418877,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 37088
+    },
+    {
+      "epoch": 0.8828191945158527,
+      "grad_norm": 0.5195881341849286,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37089
+    },
+    {
+      "epoch": 0.8828429972388842,
+      "grad_norm": 0.002308845721475603,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 37090
+    },
+    {
+      "epoch": 0.8828667999619156,
+      "grad_norm": 5.034334247745317,
+      "learning_rate": 2e-05,
+      "loss": 1.0892,
+      "step": 37091
+    },
+    {
+      "epoch": 0.8828906026849471,
+      "grad_norm": 4.925842668585053,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 37092
+    },
+    {
+      "epoch": 0.8829144054079787,
+      "grad_norm": 25.74449494813252,
+      "learning_rate": 2e-05,
+      "loss": 0.0777,
+      "step": 37093
+    },
+    {
+      "epoch": 0.8829382081310102,
+      "grad_norm": 6.560991541741786,
+      "learning_rate": 2e-05,
+      "loss": 0.821,
+      "step": 37094
+    },
+    {
+      "epoch": 0.8829620108540417,
+      "grad_norm": 2.7890793547694983,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 37095
+    },
+    {
+      "epoch": 0.8829858135770732,
+      "grad_norm": 6.000917743790107,
+      "learning_rate": 2e-05,
+      "loss": 1.4512,
+      "step": 37096
+    },
+    {
+      "epoch": 0.8830096163001048,
+      "grad_norm": 4.058179279254222,
+      "learning_rate": 2e-05,
+      "loss": 0.076,
+      "step": 37097
+    },
+    {
+      "epoch": 0.8830334190231363,
+      "grad_norm": 14.26249348410282,
+      "learning_rate": 2e-05,
+      "loss": 0.8378,
+      "step": 37098
+    },
+    {
+      "epoch": 0.8830572217461677,
+      "grad_norm": 2.1270728630179008,
+      "learning_rate": 2e-05,
+      "loss": 0.1915,
+      "step": 37099
+    },
+    {
+      "epoch": 0.8830810244691992,
+      "grad_norm": 5.144888805365407,
+      "learning_rate": 2e-05,
+      "loss": 0.5892,
+      "step": 37100
+    },
+    {
+      "epoch": 0.8831048271922308,
+      "grad_norm": 3.371412414009411,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 37101
+    },
+    {
+      "epoch": 0.8831286299152623,
+      "grad_norm": 12.653516221636755,
+      "learning_rate": 2e-05,
+      "loss": 1.805,
+      "step": 37102
+    },
+    {
+      "epoch": 0.8831524326382938,
+      "grad_norm": 5.613512096967161,
+      "learning_rate": 2e-05,
+      "loss": 0.9852,
+      "step": 37103
+    },
+    {
+      "epoch": 0.8831762353613254,
+      "grad_norm": 1.8676354358828249,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 37104
+    },
+    {
+      "epoch": 0.8832000380843569,
+      "grad_norm": 3.1766957267339757,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 37105
+    },
+    {
+      "epoch": 0.8832238408073884,
+      "grad_norm": 2.6731785639287855,
+      "learning_rate": 2e-05,
+      "loss": 0.3749,
+      "step": 37106
+    },
+    {
+      "epoch": 0.8832476435304198,
+      "grad_norm": 0.24592723089880017,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 37107
+    },
+    {
+      "epoch": 0.8832714462534514,
+      "grad_norm": 3.735523625336929,
+      "learning_rate": 2e-05,
+      "loss": 0.5303,
+      "step": 37108
+    },
+    {
+      "epoch": 0.8832952489764829,
+      "grad_norm": 0.08723724606944616,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 37109
+    },
+    {
+      "epoch": 0.8833190516995144,
+      "grad_norm": 7.126960678329319,
+      "learning_rate": 2e-05,
+      "loss": 0.9392,
+      "step": 37110
+    },
+    {
+      "epoch": 0.8833428544225459,
+      "grad_norm": 0.0622681039337472,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 37111
+    },
+    {
+      "epoch": 0.8833666571455775,
+      "grad_norm": 0.20800675884420153,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 37112
+    },
+    {
+      "epoch": 0.883390459868609,
+      "grad_norm": 3.109399839116923,
+      "learning_rate": 2e-05,
+      "loss": 0.5165,
+      "step": 37113
+    },
+    {
+      "epoch": 0.8834142625916405,
+      "grad_norm": 5.100016289642461,
+      "learning_rate": 2e-05,
+      "loss": 1.1094,
+      "step": 37114
+    },
+    {
+      "epoch": 0.883438065314672,
+      "grad_norm": 5.675980288716497,
+      "learning_rate": 2e-05,
+      "loss": 0.795,
+      "step": 37115
+    },
+    {
+      "epoch": 0.8834618680377035,
+      "grad_norm": 0.16428139417874957,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37116
+    },
+    {
+      "epoch": 0.883485670760735,
+      "grad_norm": 0.010128109271725475,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37117
+    },
+    {
+      "epoch": 0.8835094734837665,
+      "grad_norm": 0.10805164836383019,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 37118
+    },
+    {
+      "epoch": 0.8835332762067981,
+      "grad_norm": 3.795793909465379,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 37119
+    },
+    {
+      "epoch": 0.8835570789298296,
+      "grad_norm": 4.606960308676363,
+      "learning_rate": 2e-05,
+      "loss": 0.8485,
+      "step": 37120
+    },
+    {
+      "epoch": 0.8835808816528611,
+      "grad_norm": 2.636433098133571,
+      "learning_rate": 2e-05,
+      "loss": 0.082,
+      "step": 37121
+    },
+    {
+      "epoch": 0.8836046843758926,
+      "grad_norm": 10.556651101183736,
+      "learning_rate": 2e-05,
+      "loss": 0.9253,
+      "step": 37122
+    },
+    {
+      "epoch": 0.8836284870989242,
+      "grad_norm": 2.001136823352144,
+      "learning_rate": 2e-05,
+      "loss": 0.1295,
+      "step": 37123
+    },
+    {
+      "epoch": 0.8836522898219556,
+      "grad_norm": 6.929135225775988,
+      "learning_rate": 2e-05,
+      "loss": 0.3868,
+      "step": 37124
+    },
+    {
+      "epoch": 0.8836760925449871,
+      "grad_norm": 4.098100554073966,
+      "learning_rate": 2e-05,
+      "loss": 0.6014,
+      "step": 37125
+    },
+    {
+      "epoch": 0.8836998952680186,
+      "grad_norm": 3.362209737802091,
+      "learning_rate": 2e-05,
+      "loss": 0.5453,
+      "step": 37126
+    },
+    {
+      "epoch": 0.8837236979910502,
+      "grad_norm": 3.70269227521428,
+      "learning_rate": 2e-05,
+      "loss": 0.4358,
+      "step": 37127
+    },
+    {
+      "epoch": 0.8837475007140817,
+      "grad_norm": 3.015410187176325,
+      "learning_rate": 2e-05,
+      "loss": 0.467,
+      "step": 37128
+    },
+    {
+      "epoch": 0.8837713034371132,
+      "grad_norm": 3.5974800710372814,
+      "learning_rate": 2e-05,
+      "loss": 0.6371,
+      "step": 37129
+    },
+    {
+      "epoch": 0.8837951061601447,
+      "grad_norm": 2.0703873020230703,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 37130
+    },
+    {
+      "epoch": 0.8838189088831763,
+      "grad_norm": 2.344817512086128,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 37131
+    },
+    {
+      "epoch": 0.8838427116062078,
+      "grad_norm": 4.7860595935816175,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 37132
+    },
+    {
+      "epoch": 0.8838665143292392,
+      "grad_norm": 7.055118765610328,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 37133
+    },
+    {
+      "epoch": 0.8838903170522708,
+      "grad_norm": 4.634022924521822,
+      "learning_rate": 2e-05,
+      "loss": 0.5427,
+      "step": 37134
+    },
+    {
+      "epoch": 0.8839141197753023,
+      "grad_norm": 7.538854834565799,
+      "learning_rate": 2e-05,
+      "loss": 0.8107,
+      "step": 37135
+    },
+    {
+      "epoch": 0.8839379224983338,
+      "grad_norm": 5.228000950374765,
+      "learning_rate": 2e-05,
+      "loss": 1.6431,
+      "step": 37136
+    },
+    {
+      "epoch": 0.8839617252213653,
+      "grad_norm": 16.498042766371668,
+      "learning_rate": 2e-05,
+      "loss": 1.1441,
+      "step": 37137
+    },
+    {
+      "epoch": 0.8839855279443969,
+      "grad_norm": 1.8616704774934558,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 37138
+    },
+    {
+      "epoch": 0.8840093306674284,
+      "grad_norm": 5.645193127916103,
+      "learning_rate": 2e-05,
+      "loss": 1.0252,
+      "step": 37139
+    },
+    {
+      "epoch": 0.8840331333904599,
+      "grad_norm": 4.713944974508951,
+      "learning_rate": 2e-05,
+      "loss": 0.0996,
+      "step": 37140
+    },
+    {
+      "epoch": 0.8840569361134913,
+      "grad_norm": 0.5289456954452656,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37141
+    },
+    {
+      "epoch": 0.8840807388365229,
+      "grad_norm": 1.3653703122244643,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 37142
+    },
+    {
+      "epoch": 0.8841045415595544,
+      "grad_norm": 5.3531248991817435,
+      "learning_rate": 2e-05,
+      "loss": 0.8611,
+      "step": 37143
+    },
+    {
+      "epoch": 0.8841283442825859,
+      "grad_norm": 1.708048817925774,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 37144
+    },
+    {
+      "epoch": 0.8841521470056174,
+      "grad_norm": 6.460562927388587,
+      "learning_rate": 2e-05,
+      "loss": 1.0672,
+      "step": 37145
+    },
+    {
+      "epoch": 0.884175949728649,
+      "grad_norm": 1.4720745142685006,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 37146
+    },
+    {
+      "epoch": 0.8841997524516805,
+      "grad_norm": 0.17802616502044544,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 37147
+    },
+    {
+      "epoch": 0.884223555174712,
+      "grad_norm": 16.782522092790757,
+      "learning_rate": 2e-05,
+      "loss": 0.6601,
+      "step": 37148
+    },
+    {
+      "epoch": 0.8842473578977434,
+      "grad_norm": 13.877765400903426,
+      "learning_rate": 2e-05,
+      "loss": 0.7755,
+      "step": 37149
+    },
+    {
+      "epoch": 0.884271160620775,
+      "grad_norm": 2.5976410125389444,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 37150
+    },
+    {
+      "epoch": 0.8842949633438065,
+      "grad_norm": 3.2380275143915487,
+      "learning_rate": 2e-05,
+      "loss": 0.5481,
+      "step": 37151
+    },
+    {
+      "epoch": 0.884318766066838,
+      "grad_norm": 4.5704999079244795,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 37152
+    },
+    {
+      "epoch": 0.8843425687898696,
+      "grad_norm": 11.972786335064534,
+      "learning_rate": 2e-05,
+      "loss": 1.8897,
+      "step": 37153
+    },
+    {
+      "epoch": 0.8843663715129011,
+      "grad_norm": 113.32566642804204,
+      "learning_rate": 2e-05,
+      "loss": 0.1625,
+      "step": 37154
+    },
+    {
+      "epoch": 0.8843901742359326,
+      "grad_norm": 0.48802281185470814,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 37155
+    },
+    {
+      "epoch": 0.8844139769589641,
+      "grad_norm": 0.24336090825526643,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 37156
+    },
+    {
+      "epoch": 0.8844377796819957,
+      "grad_norm": 5.848361382414104,
+      "learning_rate": 2e-05,
+      "loss": 0.9661,
+      "step": 37157
+    },
+    {
+      "epoch": 0.8844615824050271,
+      "grad_norm": 3.5357607114965455,
+      "learning_rate": 2e-05,
+      "loss": 0.5438,
+      "step": 37158
+    },
+    {
+      "epoch": 0.8844853851280586,
+      "grad_norm": 0.3145002132200688,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 37159
+    },
+    {
+      "epoch": 0.8845091878510901,
+      "grad_norm": 4.148611748971404,
+      "learning_rate": 2e-05,
+      "loss": 0.9506,
+      "step": 37160
+    },
+    {
+      "epoch": 0.8845329905741217,
+      "grad_norm": 14.379907260329421,
+      "learning_rate": 2e-05,
+      "loss": 0.8788,
+      "step": 37161
+    },
+    {
+      "epoch": 0.8845567932971532,
+      "grad_norm": 1.547477164745573,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 37162
+    },
+    {
+      "epoch": 0.8845805960201847,
+      "grad_norm": 12.588439343631652,
+      "learning_rate": 2e-05,
+      "loss": 0.8777,
+      "step": 37163
+    },
+    {
+      "epoch": 0.8846043987432162,
+      "grad_norm": 1.9927860641186335,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 37164
+    },
+    {
+      "epoch": 0.8846282014662478,
+      "grad_norm": 7.417459434625378,
+      "learning_rate": 2e-05,
+      "loss": 0.2006,
+      "step": 37165
+    },
+    {
+      "epoch": 0.8846520041892793,
+      "grad_norm": 0.24261330277781748,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 37166
+    },
+    {
+      "epoch": 0.8846758069123107,
+      "grad_norm": 4.603936819494111,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 37167
+    },
+    {
+      "epoch": 0.8846996096353423,
+      "grad_norm": 35.66337739509434,
+      "learning_rate": 2e-05,
+      "loss": 0.4049,
+      "step": 37168
+    },
+    {
+      "epoch": 0.8847234123583738,
+      "grad_norm": 4.335598342521899,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 37169
+    },
+    {
+      "epoch": 0.8847472150814053,
+      "grad_norm": 5.102382510327236,
+      "learning_rate": 2e-05,
+      "loss": 0.5838,
+      "step": 37170
+    },
+    {
+      "epoch": 0.8847710178044368,
+      "grad_norm": 3.660300766461658,
+      "learning_rate": 2e-05,
+      "loss": 0.3762,
+      "step": 37171
+    },
+    {
+      "epoch": 0.8847948205274684,
+      "grad_norm": 6.7930333369715505,
+      "learning_rate": 2e-05,
+      "loss": 1.0942,
+      "step": 37172
+    },
+    {
+      "epoch": 0.8848186232504999,
+      "grad_norm": 0.0321552846166479,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37173
+    },
+    {
+      "epoch": 0.8848424259735314,
+      "grad_norm": 3.9196842928109423,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 37174
+    },
+    {
+      "epoch": 0.8848662286965628,
+      "grad_norm": 4.651883240746257,
+      "learning_rate": 2e-05,
+      "loss": 0.5932,
+      "step": 37175
+    },
+    {
+      "epoch": 0.8848900314195944,
+      "grad_norm": 0.21197838862221455,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 37176
+    },
+    {
+      "epoch": 0.8849138341426259,
+      "grad_norm": 0.06946118164453026,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37177
+    },
+    {
+      "epoch": 0.8849376368656574,
+      "grad_norm": 4.144727673747375,
+      "learning_rate": 2e-05,
+      "loss": 0.8568,
+      "step": 37178
+    },
+    {
+      "epoch": 0.8849614395886889,
+      "grad_norm": 7.315400893552517,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 37179
+    },
+    {
+      "epoch": 0.8849852423117205,
+      "grad_norm": 0.9933578945824797,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 37180
+    },
+    {
+      "epoch": 0.885009045034752,
+      "grad_norm": 4.602053820590527,
+      "learning_rate": 2e-05,
+      "loss": 0.5477,
+      "step": 37181
+    },
+    {
+      "epoch": 0.8850328477577835,
+      "grad_norm": 4.001086412827426,
+      "learning_rate": 2e-05,
+      "loss": 0.6275,
+      "step": 37182
+    },
+    {
+      "epoch": 0.885056650480815,
+      "grad_norm": 8.713265116225449,
+      "learning_rate": 2e-05,
+      "loss": 0.8114,
+      "step": 37183
+    },
+    {
+      "epoch": 0.8850804532038465,
+      "grad_norm": 0.9836166707296179,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 37184
+    },
+    {
+      "epoch": 0.885104255926878,
+      "grad_norm": 6.758926452638374,
+      "learning_rate": 2e-05,
+      "loss": 0.2656,
+      "step": 37185
+    },
+    {
+      "epoch": 0.8851280586499095,
+      "grad_norm": 6.233677730992905,
+      "learning_rate": 2e-05,
+      "loss": 1.4693,
+      "step": 37186
+    },
+    {
+      "epoch": 0.8851518613729411,
+      "grad_norm": 3.4569994027974205,
+      "learning_rate": 2e-05,
+      "loss": 0.5559,
+      "step": 37187
+    },
+    {
+      "epoch": 0.8851756640959726,
+      "grad_norm": 2.9441403417391867,
+      "learning_rate": 2e-05,
+      "loss": 0.641,
+      "step": 37188
+    },
+    {
+      "epoch": 0.8851994668190041,
+      "grad_norm": 0.17470933145758163,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 37189
+    },
+    {
+      "epoch": 0.8852232695420356,
+      "grad_norm": 4.083814068292584,
+      "learning_rate": 2e-05,
+      "loss": 0.9199,
+      "step": 37190
+    },
+    {
+      "epoch": 0.8852470722650672,
+      "grad_norm": 3.098248443307445,
+      "learning_rate": 2e-05,
+      "loss": 0.6696,
+      "step": 37191
+    },
+    {
+      "epoch": 0.8852708749880986,
+      "grad_norm": 0.11357147130996686,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 37192
+    },
+    {
+      "epoch": 0.8852946777111301,
+      "grad_norm": 0.6058160234019215,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 37193
+    },
+    {
+      "epoch": 0.8853184804341616,
+      "grad_norm": 0.6372144348798546,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 37194
+    },
+    {
+      "epoch": 0.8853422831571932,
+      "grad_norm": 15.448956961724768,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 37195
+    },
+    {
+      "epoch": 0.8853660858802247,
+      "grad_norm": 1.4928013027751619,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 37196
+    },
+    {
+      "epoch": 0.8853898886032562,
+      "grad_norm": 1.4681690428313126,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 37197
+    },
+    {
+      "epoch": 0.8854136913262878,
+      "grad_norm": 0.383675245026956,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37198
+    },
+    {
+      "epoch": 0.8854374940493193,
+      "grad_norm": 0.04100929290332333,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37199
+    },
+    {
+      "epoch": 0.8854612967723507,
+      "grad_norm": 9.107676190079383,
+      "learning_rate": 2e-05,
+      "loss": 0.2864,
+      "step": 37200
+    },
+    {
+      "epoch": 0.8854850994953822,
+      "grad_norm": 0.05875014569457652,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37201
+    },
+    {
+      "epoch": 0.8855089022184138,
+      "grad_norm": 1.0931218309196689,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 37202
+    },
+    {
+      "epoch": 0.8855327049414453,
+      "grad_norm": 6.111777551437954,
+      "learning_rate": 2e-05,
+      "loss": 1.7168,
+      "step": 37203
+    },
+    {
+      "epoch": 0.8855565076644768,
+      "grad_norm": 7.217185095305649,
+      "learning_rate": 2e-05,
+      "loss": 1.5326,
+      "step": 37204
+    },
+    {
+      "epoch": 0.8855803103875083,
+      "grad_norm": 0.2602648046314516,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37205
+    },
+    {
+      "epoch": 0.8856041131105399,
+      "grad_norm": 8.756383181237737,
+      "learning_rate": 2e-05,
+      "loss": 0.3544,
+      "step": 37206
+    },
+    {
+      "epoch": 0.8856279158335714,
+      "grad_norm": 0.38814249495051933,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 37207
+    },
+    {
+      "epoch": 0.8856517185566029,
+      "grad_norm": 3.5935568431683964,
+      "learning_rate": 2e-05,
+      "loss": 0.6334,
+      "step": 37208
+    },
+    {
+      "epoch": 0.8856755212796343,
+      "grad_norm": 2.3267283793820686,
+      "learning_rate": 2e-05,
+      "loss": 0.1491,
+      "step": 37209
+    },
+    {
+      "epoch": 0.8856993240026659,
+      "grad_norm": 4.380774944012703,
+      "learning_rate": 2e-05,
+      "loss": 0.6255,
+      "step": 37210
+    },
+    {
+      "epoch": 0.8857231267256974,
+      "grad_norm": 2.2523300269637785,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 37211
+    },
+    {
+      "epoch": 0.8857469294487289,
+      "grad_norm": 3.054451341594099,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 37212
+    },
+    {
+      "epoch": 0.8857707321717605,
+      "grad_norm": 5.877356014727922,
+      "learning_rate": 2e-05,
+      "loss": 1.469,
+      "step": 37213
+    },
+    {
+      "epoch": 0.885794534894792,
+      "grad_norm": 2.8043187669386205,
+      "learning_rate": 2e-05,
+      "loss": 0.8041,
+      "step": 37214
+    },
+    {
+      "epoch": 0.8858183376178235,
+      "grad_norm": 0.03633424925606276,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37215
+    },
+    {
+      "epoch": 0.885842140340855,
+      "grad_norm": 0.06828022868908055,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 37216
+    },
+    {
+      "epoch": 0.8858659430638866,
+      "grad_norm": 1.7832247465978555,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 37217
+    },
+    {
+      "epoch": 0.885889745786918,
+      "grad_norm": 18.758906892235,
+      "learning_rate": 2e-05,
+      "loss": 1.0201,
+      "step": 37218
+    },
+    {
+      "epoch": 0.8859135485099495,
+      "grad_norm": 2.2776232441137783,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 37219
+    },
+    {
+      "epoch": 0.885937351232981,
+      "grad_norm": 0.07153181456527398,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 37220
+    },
+    {
+      "epoch": 0.8859611539560126,
+      "grad_norm": 5.021043295975411,
+      "learning_rate": 2e-05,
+      "loss": 0.5065,
+      "step": 37221
+    },
+    {
+      "epoch": 0.8859849566790441,
+      "grad_norm": 12.6295981947933,
+      "learning_rate": 2e-05,
+      "loss": 0.2629,
+      "step": 37222
+    },
+    {
+      "epoch": 0.8860087594020756,
+      "grad_norm": 5.211446526105718,
+      "learning_rate": 2e-05,
+      "loss": 0.8201,
+      "step": 37223
+    },
+    {
+      "epoch": 0.8860325621251071,
+      "grad_norm": 0.023593213802870695,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37224
+    },
+    {
+      "epoch": 0.8860563648481387,
+      "grad_norm": 1.6075857192013245,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 37225
+    },
+    {
+      "epoch": 0.8860801675711701,
+      "grad_norm": 4.307290097221306,
+      "learning_rate": 2e-05,
+      "loss": 0.869,
+      "step": 37226
+    },
+    {
+      "epoch": 0.8861039702942016,
+      "grad_norm": 1.286283134195473,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 37227
+    },
+    {
+      "epoch": 0.8861277730172332,
+      "grad_norm": 0.7174850788548991,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 37228
+    },
+    {
+      "epoch": 0.8861515757402647,
+      "grad_norm": 3.9452803712178137,
+      "learning_rate": 2e-05,
+      "loss": 0.7074,
+      "step": 37229
+    },
+    {
+      "epoch": 0.8861753784632962,
+      "grad_norm": 11.437961514707252,
+      "learning_rate": 2e-05,
+      "loss": 0.5894,
+      "step": 37230
+    },
+    {
+      "epoch": 0.8861991811863277,
+      "grad_norm": 3.44747301286591,
+      "learning_rate": 2e-05,
+      "loss": 0.8633,
+      "step": 37231
+    },
+    {
+      "epoch": 0.8862229839093593,
+      "grad_norm": 6.041518585710629,
+      "learning_rate": 2e-05,
+      "loss": 0.9432,
+      "step": 37232
+    },
+    {
+      "epoch": 0.8862467866323908,
+      "grad_norm": 1.245594383951171,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 37233
+    },
+    {
+      "epoch": 0.8862705893554222,
+      "grad_norm": 1.5594654129122139,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 37234
+    },
+    {
+      "epoch": 0.8862943920784537,
+      "grad_norm": 7.614105819126007,
+      "learning_rate": 2e-05,
+      "loss": 0.2197,
+      "step": 37235
+    },
+    {
+      "epoch": 0.8863181948014853,
+      "grad_norm": 8.723378532000886,
+      "learning_rate": 2e-05,
+      "loss": 0.7286,
+      "step": 37236
+    },
+    {
+      "epoch": 0.8863419975245168,
+      "grad_norm": 3.1420737217169754,
+      "learning_rate": 2e-05,
+      "loss": 0.5829,
+      "step": 37237
+    },
+    {
+      "epoch": 0.8863658002475483,
+      "grad_norm": 8.118416236289331,
+      "learning_rate": 2e-05,
+      "loss": 0.4925,
+      "step": 37238
+    },
+    {
+      "epoch": 0.8863896029705798,
+      "grad_norm": 0.9107789316942331,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 37239
+    },
+    {
+      "epoch": 0.8864134056936114,
+      "grad_norm": 5.863034045205138,
+      "learning_rate": 2e-05,
+      "loss": 0.6904,
+      "step": 37240
+    },
+    {
+      "epoch": 0.8864372084166429,
+      "grad_norm": 2.3740747745668007,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 37241
+    },
+    {
+      "epoch": 0.8864610111396743,
+      "grad_norm": 5.754371449048505,
+      "learning_rate": 2e-05,
+      "loss": 0.5897,
+      "step": 37242
+    },
+    {
+      "epoch": 0.886484813862706,
+      "grad_norm": 25.545394162383374,
+      "learning_rate": 2e-05,
+      "loss": 0.2644,
+      "step": 37243
+    },
+    {
+      "epoch": 0.8865086165857374,
+      "grad_norm": 3.6705422327443067,
+      "learning_rate": 2e-05,
+      "loss": 0.879,
+      "step": 37244
+    },
+    {
+      "epoch": 0.8865324193087689,
+      "grad_norm": 6.797732744593797,
+      "learning_rate": 2e-05,
+      "loss": 1.5358,
+      "step": 37245
+    },
+    {
+      "epoch": 0.8865562220318004,
+      "grad_norm": 14.28073709293105,
+      "learning_rate": 2e-05,
+      "loss": 0.4894,
+      "step": 37246
+    },
+    {
+      "epoch": 0.886580024754832,
+      "grad_norm": 1.735276369202212,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 37247
+    },
+    {
+      "epoch": 0.8866038274778635,
+      "grad_norm": 7.86572260469714,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 37248
+    },
+    {
+      "epoch": 0.886627630200895,
+      "grad_norm": 3.473106145409982,
+      "learning_rate": 2e-05,
+      "loss": 0.2614,
+      "step": 37249
+    },
+    {
+      "epoch": 0.8866514329239265,
+      "grad_norm": 4.600035835429728,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 37250
+    },
+    {
+      "epoch": 0.886675235646958,
+      "grad_norm": 0.024197576373729195,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37251
+    },
+    {
+      "epoch": 0.8866990383699895,
+      "grad_norm": 1.4415927086824964,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 37252
+    },
+    {
+      "epoch": 0.886722841093021,
+      "grad_norm": 0.9989498520849492,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 37253
+    },
+    {
+      "epoch": 0.8867466438160525,
+      "grad_norm": 3.3711030311412737,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 37254
+    },
+    {
+      "epoch": 0.8867704465390841,
+      "grad_norm": 0.15929732469989377,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 37255
+    },
+    {
+      "epoch": 0.8867942492621156,
+      "grad_norm": 2.980816714261964,
+      "learning_rate": 2e-05,
+      "loss": 0.4125,
+      "step": 37256
+    },
+    {
+      "epoch": 0.8868180519851471,
+      "grad_norm": 2.225340646895965,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 37257
+    },
+    {
+      "epoch": 0.8868418547081787,
+      "grad_norm": 2.4788884884200693,
+      "learning_rate": 2e-05,
+      "loss": 0.5149,
+      "step": 37258
+    },
+    {
+      "epoch": 0.8868656574312102,
+      "grad_norm": 0.3893023812355601,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 37259
+    },
+    {
+      "epoch": 0.8868894601542416,
+      "grad_norm": 5.538550624760774,
+      "learning_rate": 2e-05,
+      "loss": 1.3054,
+      "step": 37260
+    },
+    {
+      "epoch": 0.8869132628772731,
+      "grad_norm": 0.6383172789967128,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 37261
+    },
+    {
+      "epoch": 0.8869370656003047,
+      "grad_norm": 0.049418230244273824,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37262
+    },
+    {
+      "epoch": 0.8869608683233362,
+      "grad_norm": 2.8118873773461637,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 37263
+    },
+    {
+      "epoch": 0.8869846710463677,
+      "grad_norm": 2.901328826451787,
+      "learning_rate": 2e-05,
+      "loss": 0.5034,
+      "step": 37264
+    },
+    {
+      "epoch": 0.8870084737693992,
+      "grad_norm": 10.79763147574456,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 37265
+    },
+    {
+      "epoch": 0.8870322764924308,
+      "grad_norm": 4.095046209215288,
+      "learning_rate": 2e-05,
+      "loss": 0.8045,
+      "step": 37266
+    },
+    {
+      "epoch": 0.8870560792154623,
+      "grad_norm": 6.291600359497142,
+      "learning_rate": 2e-05,
+      "loss": 0.7042,
+      "step": 37267
+    },
+    {
+      "epoch": 0.8870798819384937,
+      "grad_norm": 4.64678946686255,
+      "learning_rate": 2e-05,
+      "loss": 0.9327,
+      "step": 37268
+    },
+    {
+      "epoch": 0.8871036846615252,
+      "grad_norm": 1.823257517885386,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 37269
+    },
+    {
+      "epoch": 0.8871274873845568,
+      "grad_norm": 3.369431560710916,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 37270
+    },
+    {
+      "epoch": 0.8871512901075883,
+      "grad_norm": 5.091951812635078,
+      "learning_rate": 2e-05,
+      "loss": 1.3922,
+      "step": 37271
+    },
+    {
+      "epoch": 0.8871750928306198,
+      "grad_norm": 9.543282208266161,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 37272
+    },
+    {
+      "epoch": 0.8871988955536514,
+      "grad_norm": 1.4808383440168515,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 37273
+    },
+    {
+      "epoch": 0.8872226982766829,
+      "grad_norm": 2.970037030659197,
+      "learning_rate": 2e-05,
+      "loss": 0.6568,
+      "step": 37274
+    },
+    {
+      "epoch": 0.8872465009997144,
+      "grad_norm": 10.81000331336184,
+      "learning_rate": 2e-05,
+      "loss": 0.3612,
+      "step": 37275
+    },
+    {
+      "epoch": 0.8872703037227458,
+      "grad_norm": 0.09281127878613173,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 37276
+    },
+    {
+      "epoch": 0.8872941064457774,
+      "grad_norm": 5.758315552616927,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 37277
+    },
+    {
+      "epoch": 0.8873179091688089,
+      "grad_norm": 10.787284339761571,
+      "learning_rate": 2e-05,
+      "loss": 0.5116,
+      "step": 37278
+    },
+    {
+      "epoch": 0.8873417118918404,
+      "grad_norm": 5.6684250705129795,
+      "learning_rate": 2e-05,
+      "loss": 1.0131,
+      "step": 37279
+    },
+    {
+      "epoch": 0.8873655146148719,
+      "grad_norm": 4.540326319400911,
+      "learning_rate": 2e-05,
+      "loss": 0.7659,
+      "step": 37280
+    },
+    {
+      "epoch": 0.8873893173379035,
+      "grad_norm": 0.7768158908042935,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 37281
+    },
+    {
+      "epoch": 0.887413120060935,
+      "grad_norm": 3.9932057305020865,
+      "learning_rate": 2e-05,
+      "loss": 0.5908,
+      "step": 37282
+    },
+    {
+      "epoch": 0.8874369227839665,
+      "grad_norm": 0.1701082654798097,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 37283
+    },
+    {
+      "epoch": 0.887460725506998,
+      "grad_norm": 3.076473663336662,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 37284
+    },
+    {
+      "epoch": 0.8874845282300295,
+      "grad_norm": 3.5093596901231905,
+      "learning_rate": 2e-05,
+      "loss": 0.3525,
+      "step": 37285
+    },
+    {
+      "epoch": 0.887508330953061,
+      "grad_norm": 0.05989708019923111,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 37286
+    },
+    {
+      "epoch": 0.8875321336760925,
+      "grad_norm": 5.105258701173464,
+      "learning_rate": 2e-05,
+      "loss": 0.9052,
+      "step": 37287
+    },
+    {
+      "epoch": 0.8875559363991241,
+      "grad_norm": 7.366865266131574,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 37288
+    },
+    {
+      "epoch": 0.8875797391221556,
+      "grad_norm": 0.8165212435767225,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 37289
+    },
+    {
+      "epoch": 0.8876035418451871,
+      "grad_norm": 7.733558547928967,
+      "learning_rate": 2e-05,
+      "loss": 0.8053,
+      "step": 37290
+    },
+    {
+      "epoch": 0.8876273445682186,
+      "grad_norm": 2.1301943567506827,
+      "learning_rate": 2e-05,
+      "loss": 0.2503,
+      "step": 37291
+    },
+    {
+      "epoch": 0.8876511472912502,
+      "grad_norm": 4.697569896227744,
+      "learning_rate": 2e-05,
+      "loss": 0.6407,
+      "step": 37292
+    },
+    {
+      "epoch": 0.8876749500142816,
+      "grad_norm": 0.10179580882681676,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 37293
+    },
+    {
+      "epoch": 0.8876987527373131,
+      "grad_norm": 4.572495781456489,
+      "learning_rate": 2e-05,
+      "loss": 0.5332,
+      "step": 37294
+    },
+    {
+      "epoch": 0.8877225554603446,
+      "grad_norm": 9.825446535943234,
+      "learning_rate": 2e-05,
+      "loss": 0.6722,
+      "step": 37295
+    },
+    {
+      "epoch": 0.8877463581833762,
+      "grad_norm": 1.2131476362143674,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 37296
+    },
+    {
+      "epoch": 0.8877701609064077,
+      "grad_norm": 6.151781748749836,
+      "learning_rate": 2e-05,
+      "loss": 0.4034,
+      "step": 37297
+    },
+    {
+      "epoch": 0.8877939636294392,
+      "grad_norm": 3.20924697888052,
+      "learning_rate": 2e-05,
+      "loss": 0.611,
+      "step": 37298
+    },
+    {
+      "epoch": 0.8878177663524707,
+      "grad_norm": 3.4999209621456315,
+      "learning_rate": 2e-05,
+      "loss": 0.2129,
+      "step": 37299
+    },
+    {
+      "epoch": 0.8878415690755023,
+      "grad_norm": 0.13730588030966712,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 37300
+    },
+    {
+      "epoch": 0.8878653717985338,
+      "grad_norm": 2.0915117775269882,
+      "learning_rate": 2e-05,
+      "loss": 0.0419,
+      "step": 37301
+    },
+    {
+      "epoch": 0.8878891745215652,
+      "grad_norm": 0.39598685793695315,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 37302
+    },
+    {
+      "epoch": 0.8879129772445968,
+      "grad_norm": 0.6124031847870617,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 37303
+    },
+    {
+      "epoch": 0.8879367799676283,
+      "grad_norm": 1.4176787937351734,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 37304
+    },
+    {
+      "epoch": 0.8879605826906598,
+      "grad_norm": 0.20143584331306832,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 37305
+    },
+    {
+      "epoch": 0.8879843854136913,
+      "grad_norm": 0.21255653317283732,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 37306
+    },
+    {
+      "epoch": 0.8880081881367229,
+      "grad_norm": 4.7732870138893,
+      "learning_rate": 2e-05,
+      "loss": 1.0389,
+      "step": 37307
+    },
+    {
+      "epoch": 0.8880319908597544,
+      "grad_norm": 0.9283265545914636,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 37308
+    },
+    {
+      "epoch": 0.8880557935827859,
+      "grad_norm": 6.921039508939612,
+      "learning_rate": 2e-05,
+      "loss": 0.141,
+      "step": 37309
+    },
+    {
+      "epoch": 0.8880795963058173,
+      "grad_norm": 2.136365225469003,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 37310
+    },
+    {
+      "epoch": 0.8881033990288489,
+      "grad_norm": 2.745107204576427,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 37311
+    },
+    {
+      "epoch": 0.8881272017518804,
+      "grad_norm": 0.24033287296749078,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 37312
+    },
+    {
+      "epoch": 0.8881510044749119,
+      "grad_norm": 3.5188218499159345,
+      "learning_rate": 2e-05,
+      "loss": 0.5058,
+      "step": 37313
+    },
+    {
+      "epoch": 0.8881748071979434,
+      "grad_norm": 10.131160678398027,
+      "learning_rate": 2e-05,
+      "loss": 0.7309,
+      "step": 37314
+    },
+    {
+      "epoch": 0.888198609920975,
+      "grad_norm": 2.3018876680385847,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 37315
+    },
+    {
+      "epoch": 0.8882224126440065,
+      "grad_norm": 14.33430070832638,
+      "learning_rate": 2e-05,
+      "loss": 0.6394,
+      "step": 37316
+    },
+    {
+      "epoch": 0.888246215367038,
+      "grad_norm": 0.049168733095688054,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37317
+    },
+    {
+      "epoch": 0.8882700180900696,
+      "grad_norm": 4.907874528978389,
+      "learning_rate": 2e-05,
+      "loss": 0.4876,
+      "step": 37318
+    },
+    {
+      "epoch": 0.888293820813101,
+      "grad_norm": 0.02248003287779398,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37319
+    },
+    {
+      "epoch": 0.8883176235361325,
+      "grad_norm": 0.7204506503825153,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37320
+    },
+    {
+      "epoch": 0.888341426259164,
+      "grad_norm": 4.968966571749875,
+      "learning_rate": 2e-05,
+      "loss": 0.1266,
+      "step": 37321
+    },
+    {
+      "epoch": 0.8883652289821956,
+      "grad_norm": 16.158646143385745,
+      "learning_rate": 2e-05,
+      "loss": 0.2002,
+      "step": 37322
+    },
+    {
+      "epoch": 0.8883890317052271,
+      "grad_norm": 0.5045761659175428,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37323
+    },
+    {
+      "epoch": 0.8884128344282586,
+      "grad_norm": 4.7006811675696225,
+      "learning_rate": 2e-05,
+      "loss": 0.8301,
+      "step": 37324
+    },
+    {
+      "epoch": 0.8884366371512901,
+      "grad_norm": 0.19479673592796062,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 37325
+    },
+    {
+      "epoch": 0.8884604398743217,
+      "grad_norm": 3.9517728596852373,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 37326
+    },
+    {
+      "epoch": 0.8884842425973531,
+      "grad_norm": 0.35096323779676764,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 37327
+    },
+    {
+      "epoch": 0.8885080453203846,
+      "grad_norm": 0.7155482436188529,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 37328
+    },
+    {
+      "epoch": 0.8885318480434161,
+      "grad_norm": 3.586929630797204,
+      "learning_rate": 2e-05,
+      "loss": 0.1248,
+      "step": 37329
+    },
+    {
+      "epoch": 0.8885556507664477,
+      "grad_norm": 16.773409436373555,
+      "learning_rate": 2e-05,
+      "loss": 0.3832,
+      "step": 37330
+    },
+    {
+      "epoch": 0.8885794534894792,
+      "grad_norm": 5.374706395994853,
+      "learning_rate": 2e-05,
+      "loss": 1.1909,
+      "step": 37331
+    },
+    {
+      "epoch": 0.8886032562125107,
+      "grad_norm": 3.928367547936497,
+      "learning_rate": 2e-05,
+      "loss": 0.063,
+      "step": 37332
+    },
+    {
+      "epoch": 0.8886270589355423,
+      "grad_norm": 0.6462920667145903,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 37333
+    },
+    {
+      "epoch": 0.8886508616585738,
+      "grad_norm": 3.1420467927436526,
+      "learning_rate": 2e-05,
+      "loss": 0.693,
+      "step": 37334
+    },
+    {
+      "epoch": 0.8886746643816053,
+      "grad_norm": 2.7060325060712347,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 37335
+    },
+    {
+      "epoch": 0.8886984671046367,
+      "grad_norm": 3.1754942747422503,
+      "learning_rate": 2e-05,
+      "loss": 0.6414,
+      "step": 37336
+    },
+    {
+      "epoch": 0.8887222698276683,
+      "grad_norm": 4.843113489436414,
+      "learning_rate": 2e-05,
+      "loss": 1.1519,
+      "step": 37337
+    },
+    {
+      "epoch": 0.8887460725506998,
+      "grad_norm": 1.4108250452312674,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 37338
+    },
+    {
+      "epoch": 0.8887698752737313,
+      "grad_norm": 4.8376523147852595,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 37339
+    },
+    {
+      "epoch": 0.8887936779967628,
+      "grad_norm": 2.049238103260386,
+      "learning_rate": 2e-05,
+      "loss": 0.246,
+      "step": 37340
+    },
+    {
+      "epoch": 0.8888174807197944,
+      "grad_norm": 4.277958678499015,
+      "learning_rate": 2e-05,
+      "loss": 0.744,
+      "step": 37341
+    },
+    {
+      "epoch": 0.8888412834428259,
+      "grad_norm": 2.6981341877079803,
+      "learning_rate": 2e-05,
+      "loss": 0.2364,
+      "step": 37342
+    },
+    {
+      "epoch": 0.8888650861658574,
+      "grad_norm": 3.3110538831699476,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 37343
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 4.347488553200752,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 37344
+    },
+    {
+      "epoch": 0.8889126916119204,
+      "grad_norm": 0.014348333354613527,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 37345
+    },
+    {
+      "epoch": 0.8889364943349519,
+      "grad_norm": 3.63196733795341,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 37346
+    },
+    {
+      "epoch": 0.8889602970579834,
+      "grad_norm": 0.4911391179657475,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 37347
+    },
+    {
+      "epoch": 0.888984099781015,
+      "grad_norm": 14.331209203796021,
+      "learning_rate": 2e-05,
+      "loss": 1.1679,
+      "step": 37348
+    },
+    {
+      "epoch": 0.8890079025040465,
+      "grad_norm": 1.0246381692718272,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 37349
+    },
+    {
+      "epoch": 0.889031705227078,
+      "grad_norm": 3.1322746573258433,
+      "learning_rate": 2e-05,
+      "loss": 0.5667,
+      "step": 37350
+    },
+    {
+      "epoch": 0.8890555079501095,
+      "grad_norm": 2.563924241529141,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 37351
+    },
+    {
+      "epoch": 0.889079310673141,
+      "grad_norm": 0.5522969916679754,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 37352
+    },
+    {
+      "epoch": 0.8891031133961725,
+      "grad_norm": 5.2281727444340795,
+      "learning_rate": 2e-05,
+      "loss": 0.6455,
+      "step": 37353
+    },
+    {
+      "epoch": 0.889126916119204,
+      "grad_norm": 2.1839080941425304,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 37354
+    },
+    {
+      "epoch": 0.8891507188422355,
+      "grad_norm": 6.807605807230886,
+      "learning_rate": 2e-05,
+      "loss": 0.8995,
+      "step": 37355
+    },
+    {
+      "epoch": 0.8891745215652671,
+      "grad_norm": 13.854688822675353,
+      "learning_rate": 2e-05,
+      "loss": 0.2125,
+      "step": 37356
+    },
+    {
+      "epoch": 0.8891983242882986,
+      "grad_norm": 10.127413092565885,
+      "learning_rate": 2e-05,
+      "loss": 0.2247,
+      "step": 37357
+    },
+    {
+      "epoch": 0.8892221270113301,
+      "grad_norm": 3.2053878817586954,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 37358
+    },
+    {
+      "epoch": 0.8892459297343616,
+      "grad_norm": 2.800937869976763,
+      "learning_rate": 2e-05,
+      "loss": 0.4964,
+      "step": 37359
+    },
+    {
+      "epoch": 0.8892697324573932,
+      "grad_norm": 34.04183227915159,
+      "learning_rate": 2e-05,
+      "loss": 0.3894,
+      "step": 37360
+    },
+    {
+      "epoch": 0.8892935351804246,
+      "grad_norm": 11.212978676187413,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 37361
+    },
+    {
+      "epoch": 0.8893173379034561,
+      "grad_norm": 0.5375026828100978,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37362
+    },
+    {
+      "epoch": 0.8893411406264877,
+      "grad_norm": 2.3799331103602865,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 37363
+    },
+    {
+      "epoch": 0.8893649433495192,
+      "grad_norm": 0.012341658294877132,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 37364
+    },
+    {
+      "epoch": 0.8893887460725507,
+      "grad_norm": 5.818265817109696,
+      "learning_rate": 2e-05,
+      "loss": 0.1426,
+      "step": 37365
+    },
+    {
+      "epoch": 0.8894125487955822,
+      "grad_norm": 4.435785450290043,
+      "learning_rate": 2e-05,
+      "loss": 0.793,
+      "step": 37366
+    },
+    {
+      "epoch": 0.8894363515186138,
+      "grad_norm": 6.718178466314192,
+      "learning_rate": 2e-05,
+      "loss": 1.2552,
+      "step": 37367
+    },
+    {
+      "epoch": 0.8894601542416453,
+      "grad_norm": 5.829445453324391,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 37368
+    },
+    {
+      "epoch": 0.8894839569646767,
+      "grad_norm": 0.6756620662932387,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 37369
+    },
+    {
+      "epoch": 0.8895077596877082,
+      "grad_norm": 0.03590514980490167,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37370
+    },
+    {
+      "epoch": 0.8895315624107398,
+      "grad_norm": 6.373825012366704,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 37371
+    },
+    {
+      "epoch": 0.8895553651337713,
+      "grad_norm": 0.4556906176825587,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 37372
+    },
+    {
+      "epoch": 0.8895791678568028,
+      "grad_norm": 16.11333815513454,
+      "learning_rate": 2e-05,
+      "loss": 0.7102,
+      "step": 37373
+    },
+    {
+      "epoch": 0.8896029705798343,
+      "grad_norm": 3.4241119599002987,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 37374
+    },
+    {
+      "epoch": 0.8896267733028659,
+      "grad_norm": 0.06794234628115545,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 37375
+    },
+    {
+      "epoch": 0.8896505760258974,
+      "grad_norm": 25.910852509674527,
+      "learning_rate": 2e-05,
+      "loss": 1.4755,
+      "step": 37376
+    },
+    {
+      "epoch": 0.8896743787489289,
+      "grad_norm": 16.243312720329325,
+      "learning_rate": 2e-05,
+      "loss": 0.1921,
+      "step": 37377
+    },
+    {
+      "epoch": 0.8896981814719604,
+      "grad_norm": 2.88964728651013,
+      "learning_rate": 2e-05,
+      "loss": 0.4305,
+      "step": 37378
+    },
+    {
+      "epoch": 0.8897219841949919,
+      "grad_norm": 2.0335770603866705,
+      "learning_rate": 2e-05,
+      "loss": 0.181,
+      "step": 37379
+    },
+    {
+      "epoch": 0.8897457869180234,
+      "grad_norm": 2.2553563311847262,
+      "learning_rate": 2e-05,
+      "loss": 0.2049,
+      "step": 37380
+    },
+    {
+      "epoch": 0.8897695896410549,
+      "grad_norm": 4.093647299602257,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 37381
+    },
+    {
+      "epoch": 0.8897933923640865,
+      "grad_norm": 4.622383380664061,
+      "learning_rate": 2e-05,
+      "loss": 0.2932,
+      "step": 37382
+    },
+    {
+      "epoch": 0.889817195087118,
+      "grad_norm": 10.098014402049754,
+      "learning_rate": 2e-05,
+      "loss": 0.5894,
+      "step": 37383
+    },
+    {
+      "epoch": 0.8898409978101495,
+      "grad_norm": 0.14083207061628222,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 37384
+    },
+    {
+      "epoch": 0.889864800533181,
+      "grad_norm": 5.591950962195791,
+      "learning_rate": 2e-05,
+      "loss": 1.661,
+      "step": 37385
+    },
+    {
+      "epoch": 0.8898886032562126,
+      "grad_norm": 8.434772440704593,
+      "learning_rate": 2e-05,
+      "loss": 0.7108,
+      "step": 37386
+    },
+    {
+      "epoch": 0.889912405979244,
+      "grad_norm": 6.383889149118629,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 37387
+    },
+    {
+      "epoch": 0.8899362087022755,
+      "grad_norm": 2.6598885373061516,
+      "learning_rate": 2e-05,
+      "loss": 0.3337,
+      "step": 37388
+    },
+    {
+      "epoch": 0.889960011425307,
+      "grad_norm": 0.049594841749313535,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37389
+    },
+    {
+      "epoch": 0.8899838141483386,
+      "grad_norm": 3.2695150810984948,
+      "learning_rate": 2e-05,
+      "loss": 0.5539,
+      "step": 37390
+    },
+    {
+      "epoch": 0.8900076168713701,
+      "grad_norm": 1.4188806713877615,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 37391
+    },
+    {
+      "epoch": 0.8900314195944016,
+      "grad_norm": 4.019878613537881,
+      "learning_rate": 2e-05,
+      "loss": 0.4316,
+      "step": 37392
+    },
+    {
+      "epoch": 0.8900552223174331,
+      "grad_norm": 1.1406192530505193,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 37393
+    },
+    {
+      "epoch": 0.8900790250404647,
+      "grad_norm": 2.999351351864973,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 37394
+    },
+    {
+      "epoch": 0.8901028277634961,
+      "grad_norm": 0.015513944332960402,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37395
+    },
+    {
+      "epoch": 0.8901266304865276,
+      "grad_norm": 3.0729432829178256,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 37396
+    },
+    {
+      "epoch": 0.8901504332095592,
+      "grad_norm": 8.933150519189628,
+      "learning_rate": 2e-05,
+      "loss": 0.1654,
+      "step": 37397
+    },
+    {
+      "epoch": 0.8901742359325907,
+      "grad_norm": 3.5495849426358896,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 37398
+    },
+    {
+      "epoch": 0.8901980386556222,
+      "grad_norm": 0.3306592760513737,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 37399
+    },
+    {
+      "epoch": 0.8902218413786537,
+      "grad_norm": 1.8658133733539228,
+      "learning_rate": 2e-05,
+      "loss": 0.1683,
+      "step": 37400
+    },
+    {
+      "epoch": 0.8902456441016853,
+      "grad_norm": 0.030498090464390257,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37401
+    },
+    {
+      "epoch": 0.8902694468247168,
+      "grad_norm": 0.05239939597893758,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37402
+    },
+    {
+      "epoch": 0.8902932495477482,
+      "grad_norm": 4.459476134087775,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 37403
+    },
+    {
+      "epoch": 0.8903170522707797,
+      "grad_norm": 3.0687216618508066,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 37404
+    },
+    {
+      "epoch": 0.8903408549938113,
+      "grad_norm": 1.6175183644179896,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 37405
+    },
+    {
+      "epoch": 0.8903646577168428,
+      "grad_norm": 9.975075811401394,
+      "learning_rate": 2e-05,
+      "loss": 0.2259,
+      "step": 37406
+    },
+    {
+      "epoch": 0.8903884604398743,
+      "grad_norm": 2.916277502550099,
+      "learning_rate": 2e-05,
+      "loss": 0.1535,
+      "step": 37407
+    },
+    {
+      "epoch": 0.8904122631629058,
+      "grad_norm": 0.04682824004343991,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37408
+    },
+    {
+      "epoch": 0.8904360658859374,
+      "grad_norm": 10.150963669867917,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 37409
+    },
+    {
+      "epoch": 0.8904598686089689,
+      "grad_norm": 6.807567046672148,
+      "learning_rate": 2e-05,
+      "loss": 1.0651,
+      "step": 37410
+    },
+    {
+      "epoch": 0.8904836713320003,
+      "grad_norm": 2.262491854817243,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 37411
+    },
+    {
+      "epoch": 0.890507474055032,
+      "grad_norm": 3.090435478431742,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 37412
+    },
+    {
+      "epoch": 0.8905312767780634,
+      "grad_norm": 4.47494877751957,
+      "learning_rate": 2e-05,
+      "loss": 0.7053,
+      "step": 37413
+    },
+    {
+      "epoch": 0.8905550795010949,
+      "grad_norm": 42.76644229005864,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 37414
+    },
+    {
+      "epoch": 0.8905788822241264,
+      "grad_norm": 1.2407207818754038,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 37415
+    },
+    {
+      "epoch": 0.890602684947158,
+      "grad_norm": 12.514372892018578,
+      "learning_rate": 2e-05,
+      "loss": 0.9164,
+      "step": 37416
+    },
+    {
+      "epoch": 0.8906264876701895,
+      "grad_norm": 0.2242111440743852,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37417
+    },
+    {
+      "epoch": 0.890650290393221,
+      "grad_norm": 3.509076607327087,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 37418
+    },
+    {
+      "epoch": 0.8906740931162525,
+      "grad_norm": 0.5357911855873096,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 37419
+    },
+    {
+      "epoch": 0.890697895839284,
+      "grad_norm": 0.06662030893493309,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37420
+    },
+    {
+      "epoch": 0.8907216985623155,
+      "grad_norm": 4.643286959326888,
+      "learning_rate": 2e-05,
+      "loss": 0.6531,
+      "step": 37421
+    },
+    {
+      "epoch": 0.890745501285347,
+      "grad_norm": 3.7678889039581924,
+      "learning_rate": 2e-05,
+      "loss": 0.3746,
+      "step": 37422
+    },
+    {
+      "epoch": 0.8907693040083785,
+      "grad_norm": 2.659323257974982,
+      "learning_rate": 2e-05,
+      "loss": 0.4085,
+      "step": 37423
+    },
+    {
+      "epoch": 0.8907931067314101,
+      "grad_norm": 6.940777989577391,
+      "learning_rate": 2e-05,
+      "loss": 0.6759,
+      "step": 37424
+    },
+    {
+      "epoch": 0.8908169094544416,
+      "grad_norm": 12.440377381199392,
+      "learning_rate": 2e-05,
+      "loss": 2.407,
+      "step": 37425
+    },
+    {
+      "epoch": 0.8908407121774731,
+      "grad_norm": 5.6846630358438315,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 37426
+    },
+    {
+      "epoch": 0.8908645149005047,
+      "grad_norm": 0.18753739851700704,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 37427
+    },
+    {
+      "epoch": 0.8908883176235362,
+      "grad_norm": 10.338128576943152,
+      "learning_rate": 2e-05,
+      "loss": 1.4726,
+      "step": 37428
+    },
+    {
+      "epoch": 0.8909121203465676,
+      "grad_norm": 5.143282345981769,
+      "learning_rate": 2e-05,
+      "loss": 0.5095,
+      "step": 37429
+    },
+    {
+      "epoch": 0.8909359230695991,
+      "grad_norm": 0.2784368148935471,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 37430
+    },
+    {
+      "epoch": 0.8909597257926307,
+      "grad_norm": 0.02044717408671846,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 37431
+    },
+    {
+      "epoch": 0.8909835285156622,
+      "grad_norm": 10.877276751430855,
+      "learning_rate": 2e-05,
+      "loss": 0.231,
+      "step": 37432
+    },
+    {
+      "epoch": 0.8910073312386937,
+      "grad_norm": 5.0619300101218,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 37433
+    },
+    {
+      "epoch": 0.8910311339617252,
+      "grad_norm": 0.5191407503879368,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 37434
+    },
+    {
+      "epoch": 0.8910549366847568,
+      "grad_norm": 2.7105227723476037,
+      "learning_rate": 2e-05,
+      "loss": 0.5767,
+      "step": 37435
+    },
+    {
+      "epoch": 0.8910787394077883,
+      "grad_norm": 4.353377226278104,
+      "learning_rate": 2e-05,
+      "loss": 0.6186,
+      "step": 37436
+    },
+    {
+      "epoch": 0.8911025421308197,
+      "grad_norm": 8.498033821440723,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 37437
+    },
+    {
+      "epoch": 0.8911263448538512,
+      "grad_norm": 3.18445160328273,
+      "learning_rate": 2e-05,
+      "loss": 0.5893,
+      "step": 37438
+    },
+    {
+      "epoch": 0.8911501475768828,
+      "grad_norm": 11.581889398543476,
+      "learning_rate": 2e-05,
+      "loss": 1.2795,
+      "step": 37439
+    },
+    {
+      "epoch": 0.8911739502999143,
+      "grad_norm": 5.279865164592027,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 37440
+    },
+    {
+      "epoch": 0.8911977530229458,
+      "grad_norm": 0.13580333737490255,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 37441
+    },
+    {
+      "epoch": 0.8912215557459774,
+      "grad_norm": 0.4965571799681581,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 37442
+    },
+    {
+      "epoch": 0.8912453584690089,
+      "grad_norm": 6.543578535382716,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 37443
+    },
+    {
+      "epoch": 0.8912691611920404,
+      "grad_norm": 2.07969593658862,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 37444
+    },
+    {
+      "epoch": 0.8912929639150718,
+      "grad_norm": 6.491347234078053,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 37445
+    },
+    {
+      "epoch": 0.8913167666381034,
+      "grad_norm": 0.006100713515651612,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 37446
+    },
+    {
+      "epoch": 0.8913405693611349,
+      "grad_norm": 2.8124060563437636,
+      "learning_rate": 2e-05,
+      "loss": 0.1043,
+      "step": 37447
+    },
+    {
+      "epoch": 0.8913643720841664,
+      "grad_norm": 3.2462500394082308,
+      "learning_rate": 2e-05,
+      "loss": 0.5341,
+      "step": 37448
+    },
+    {
+      "epoch": 0.8913881748071979,
+      "grad_norm": 0.7050292628493177,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37449
+    },
+    {
+      "epoch": 0.8914119775302295,
+      "grad_norm": 1.6414035797794218,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 37450
+    },
+    {
+      "epoch": 0.891435780253261,
+      "grad_norm": 2.5339491931610048,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 37451
+    },
+    {
+      "epoch": 0.8914595829762925,
+      "grad_norm": 5.228223138038997,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 37452
+    },
+    {
+      "epoch": 0.891483385699324,
+      "grad_norm": 2.9399822272844935,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 37453
+    },
+    {
+      "epoch": 0.8915071884223555,
+      "grad_norm": 7.547362553768585,
+      "learning_rate": 2e-05,
+      "loss": 0.7783,
+      "step": 37454
+    },
+    {
+      "epoch": 0.891530991145387,
+      "grad_norm": 5.1493382843732425,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 37455
+    },
+    {
+      "epoch": 0.8915547938684185,
+      "grad_norm": 5.468412166621149,
+      "learning_rate": 2e-05,
+      "loss": 0.7907,
+      "step": 37456
+    },
+    {
+      "epoch": 0.8915785965914501,
+      "grad_norm": 0.008223221301069571,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37457
+    },
+    {
+      "epoch": 0.8916023993144816,
+      "grad_norm": 7.818951224457662,
+      "learning_rate": 2e-05,
+      "loss": 0.8819,
+      "step": 37458
+    },
+    {
+      "epoch": 0.8916262020375131,
+      "grad_norm": 0.05200578351229685,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37459
+    },
+    {
+      "epoch": 0.8916500047605446,
+      "grad_norm": 5.117691785173205,
+      "learning_rate": 2e-05,
+      "loss": 0.9134,
+      "step": 37460
+    },
+    {
+      "epoch": 0.8916738074835762,
+      "grad_norm": 0.5611413440632154,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 37461
+    },
+    {
+      "epoch": 0.8916976102066076,
+      "grad_norm": 1.9642816984178322,
+      "learning_rate": 2e-05,
+      "loss": 0.2952,
+      "step": 37462
+    },
+    {
+      "epoch": 0.8917214129296391,
+      "grad_norm": 0.20511812206977723,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 37463
+    },
+    {
+      "epoch": 0.8917452156526706,
+      "grad_norm": 2.9371502002041714,
+      "learning_rate": 2e-05,
+      "loss": 0.1413,
+      "step": 37464
+    },
+    {
+      "epoch": 0.8917690183757022,
+      "grad_norm": 3.856637072003606,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 37465
+    },
+    {
+      "epoch": 0.8917928210987337,
+      "grad_norm": 5.5844816504146015,
+      "learning_rate": 2e-05,
+      "loss": 1.0612,
+      "step": 37466
+    },
+    {
+      "epoch": 0.8918166238217652,
+      "grad_norm": 0.03586028814371342,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37467
+    },
+    {
+      "epoch": 0.8918404265447967,
+      "grad_norm": 2.590145377124242,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 37468
+    },
+    {
+      "epoch": 0.8918642292678283,
+      "grad_norm": 4.510975187816035,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 37469
+    },
+    {
+      "epoch": 0.8918880319908598,
+      "grad_norm": 0.07234184268907805,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 37470
+    },
+    {
+      "epoch": 0.8919118347138912,
+      "grad_norm": 1.5858027887215336,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 37471
+    },
+    {
+      "epoch": 0.8919356374369228,
+      "grad_norm": 3.4837001707017565,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 37472
+    },
+    {
+      "epoch": 0.8919594401599543,
+      "grad_norm": 4.792013078084624,
+      "learning_rate": 2e-05,
+      "loss": 0.7351,
+      "step": 37473
+    },
+    {
+      "epoch": 0.8919832428829858,
+      "grad_norm": 0.02908678003086284,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37474
+    },
+    {
+      "epoch": 0.8920070456060173,
+      "grad_norm": 2.9722769500730664,
+      "learning_rate": 2e-05,
+      "loss": 0.2178,
+      "step": 37475
+    },
+    {
+      "epoch": 0.8920308483290489,
+      "grad_norm": 5.713200703635209,
+      "learning_rate": 2e-05,
+      "loss": 0.8608,
+      "step": 37476
+    },
+    {
+      "epoch": 0.8920546510520804,
+      "grad_norm": 0.5100693764190287,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 37477
+    },
+    {
+      "epoch": 0.8920784537751119,
+      "grad_norm": 4.390866162451745,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 37478
+    },
+    {
+      "epoch": 0.8921022564981433,
+      "grad_norm": 0.3179127766879447,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 37479
+    },
+    {
+      "epoch": 0.8921260592211749,
+      "grad_norm": 2.2839731259542804,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 37480
+    },
+    {
+      "epoch": 0.8921498619442064,
+      "grad_norm": 0.06181503198831144,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37481
+    },
+    {
+      "epoch": 0.8921736646672379,
+      "grad_norm": 7.801174394155785,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 37482
+    },
+    {
+      "epoch": 0.8921974673902694,
+      "grad_norm": 0.5062083805992643,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 37483
+    },
+    {
+      "epoch": 0.892221270113301,
+      "grad_norm": 4.47612000014934,
+      "learning_rate": 2e-05,
+      "loss": 0.7577,
+      "step": 37484
+    },
+    {
+      "epoch": 0.8922450728363325,
+      "grad_norm": 3.125145299345425,
+      "learning_rate": 2e-05,
+      "loss": 0.4138,
+      "step": 37485
+    },
+    {
+      "epoch": 0.892268875559364,
+      "grad_norm": 3.2643948674080363,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 37486
+    },
+    {
+      "epoch": 0.8922926782823956,
+      "grad_norm": 11.548152445094606,
+      "learning_rate": 2e-05,
+      "loss": 0.2481,
+      "step": 37487
+    },
+    {
+      "epoch": 0.892316481005427,
+      "grad_norm": 3.679915248290285,
+      "learning_rate": 2e-05,
+      "loss": 0.2237,
+      "step": 37488
+    },
+    {
+      "epoch": 0.8923402837284585,
+      "grad_norm": 0.16399327746262185,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37489
+    },
+    {
+      "epoch": 0.89236408645149,
+      "grad_norm": 7.026834875403453,
+      "learning_rate": 2e-05,
+      "loss": 0.2906,
+      "step": 37490
+    },
+    {
+      "epoch": 0.8923878891745216,
+      "grad_norm": 2.621335681586933,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 37491
+    },
+    {
+      "epoch": 0.8924116918975531,
+      "grad_norm": 0.005023780148495057,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37492
+    },
+    {
+      "epoch": 0.8924354946205846,
+      "grad_norm": 0.23258315784955433,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37493
+    },
+    {
+      "epoch": 0.8924592973436161,
+      "grad_norm": 3.803428381496175,
+      "learning_rate": 2e-05,
+      "loss": 0.5787,
+      "step": 37494
+    },
+    {
+      "epoch": 0.8924831000666477,
+      "grad_norm": 4.830286591471812,
+      "learning_rate": 2e-05,
+      "loss": 0.8639,
+      "step": 37495
+    },
+    {
+      "epoch": 0.8925069027896791,
+      "grad_norm": 2.0397131937334416,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 37496
+    },
+    {
+      "epoch": 0.8925307055127106,
+      "grad_norm": 0.030970952294666586,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37497
+    },
+    {
+      "epoch": 0.8925545082357421,
+      "grad_norm": 5.666776128908643,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 37498
+    },
+    {
+      "epoch": 0.8925783109587737,
+      "grad_norm": 9.16777648480242,
+      "learning_rate": 2e-05,
+      "loss": 0.7887,
+      "step": 37499
+    },
+    {
+      "epoch": 0.8926021136818052,
+      "grad_norm": 5.063448372575232,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 37500
+    },
+    {
+      "epoch": 0.8926259164048367,
+      "grad_norm": 1.8715737325082875,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 37501
+    },
+    {
+      "epoch": 0.8926497191278683,
+      "grad_norm": 3.7451152195307422,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 37502
+    },
+    {
+      "epoch": 0.8926735218508998,
+      "grad_norm": 1.113855695872724,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 37503
+    },
+    {
+      "epoch": 0.8926973245739313,
+      "grad_norm": 0.4277175348618624,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 37504
+    },
+    {
+      "epoch": 0.8927211272969627,
+      "grad_norm": 3.589946915175956,
+      "learning_rate": 2e-05,
+      "loss": 0.4319,
+      "step": 37505
+    },
+    {
+      "epoch": 0.8927449300199943,
+      "grad_norm": 5.012801906507809,
+      "learning_rate": 2e-05,
+      "loss": 0.752,
+      "step": 37506
+    },
+    {
+      "epoch": 0.8927687327430258,
+      "grad_norm": 0.9285857295401165,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 37507
+    },
+    {
+      "epoch": 0.8927925354660573,
+      "grad_norm": 2.65942587137836,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 37508
+    },
+    {
+      "epoch": 0.8928163381890888,
+      "grad_norm": 4.22243867898025,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 37509
+    },
+    {
+      "epoch": 0.8928401409121204,
+      "grad_norm": 1.893982145862436,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 37510
+    },
+    {
+      "epoch": 0.8928639436351519,
+      "grad_norm": 5.297598692953885,
+      "learning_rate": 2e-05,
+      "loss": 0.3752,
+      "step": 37511
+    },
+    {
+      "epoch": 0.8928877463581834,
+      "grad_norm": 4.5108226166640515,
+      "learning_rate": 2e-05,
+      "loss": 0.6271,
+      "step": 37512
+    },
+    {
+      "epoch": 0.8929115490812148,
+      "grad_norm": 10.53152892402895,
+      "learning_rate": 2e-05,
+      "loss": 1.0395,
+      "step": 37513
+    },
+    {
+      "epoch": 0.8929353518042464,
+      "grad_norm": 2.770849974587233,
+      "learning_rate": 2e-05,
+      "loss": 0.0685,
+      "step": 37514
+    },
+    {
+      "epoch": 0.8929591545272779,
+      "grad_norm": 17.33807564909407,
+      "learning_rate": 2e-05,
+      "loss": 0.9003,
+      "step": 37515
+    },
+    {
+      "epoch": 0.8929829572503094,
+      "grad_norm": 3.6207429238646425,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 37516
+    },
+    {
+      "epoch": 0.893006759973341,
+      "grad_norm": 5.507957083246212,
+      "learning_rate": 2e-05,
+      "loss": 0.664,
+      "step": 37517
+    },
+    {
+      "epoch": 0.8930305626963725,
+      "grad_norm": 13.615843970661283,
+      "learning_rate": 2e-05,
+      "loss": 1.2053,
+      "step": 37518
+    },
+    {
+      "epoch": 0.893054365419404,
+      "grad_norm": 2.6704748846386046,
+      "learning_rate": 2e-05,
+      "loss": 0.3828,
+      "step": 37519
+    },
+    {
+      "epoch": 0.8930781681424355,
+      "grad_norm": 4.996423717998926,
+      "learning_rate": 2e-05,
+      "loss": 0.8882,
+      "step": 37520
+    },
+    {
+      "epoch": 0.893101970865467,
+      "grad_norm": 0.25948011199896753,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 37521
+    },
+    {
+      "epoch": 0.8931257735884985,
+      "grad_norm": 1.9232957385045588,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 37522
+    },
+    {
+      "epoch": 0.89314957631153,
+      "grad_norm": 5.2537859750927165,
+      "learning_rate": 2e-05,
+      "loss": 0.6032,
+      "step": 37523
+    },
+    {
+      "epoch": 0.8931733790345615,
+      "grad_norm": 0.08646789689504714,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37524
+    },
+    {
+      "epoch": 0.8931971817575931,
+      "grad_norm": 5.7730488054709035,
+      "learning_rate": 2e-05,
+      "loss": 1.061,
+      "step": 37525
+    },
+    {
+      "epoch": 0.8932209844806246,
+      "grad_norm": 4.372871928311049,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 37526
+    },
+    {
+      "epoch": 0.8932447872036561,
+      "grad_norm": 0.5632805142530978,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 37527
+    },
+    {
+      "epoch": 0.8932685899266876,
+      "grad_norm": 4.250308960375066,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 37528
+    },
+    {
+      "epoch": 0.8932923926497192,
+      "grad_norm": 0.011331181843761383,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37529
+    },
+    {
+      "epoch": 0.8933161953727506,
+      "grad_norm": 0.6487277410170209,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 37530
+    },
+    {
+      "epoch": 0.8933399980957821,
+      "grad_norm": 0.02429863513381071,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37531
+    },
+    {
+      "epoch": 0.8933638008188137,
+      "grad_norm": 0.42551715449695016,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 37532
+    },
+    {
+      "epoch": 0.8933876035418452,
+      "grad_norm": 1.761050598744178,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 37533
+    },
+    {
+      "epoch": 0.8934114062648767,
+      "grad_norm": 4.367370718312852,
+      "learning_rate": 2e-05,
+      "loss": 0.4137,
+      "step": 37534
+    },
+    {
+      "epoch": 0.8934352089879082,
+      "grad_norm": 4.3279663066439715,
+      "learning_rate": 2e-05,
+      "loss": 1.3319,
+      "step": 37535
+    },
+    {
+      "epoch": 0.8934590117109398,
+      "grad_norm": 56.936371712000394,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 37536
+    },
+    {
+      "epoch": 0.8934828144339713,
+      "grad_norm": 6.131450119350955,
+      "learning_rate": 2e-05,
+      "loss": 0.885,
+      "step": 37537
+    },
+    {
+      "epoch": 0.8935066171570027,
+      "grad_norm": 2.5705912827430186,
+      "learning_rate": 2e-05,
+      "loss": 0.1812,
+      "step": 37538
+    },
+    {
+      "epoch": 0.8935304198800342,
+      "grad_norm": 0.5884944511425635,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 37539
+    },
+    {
+      "epoch": 0.8935542226030658,
+      "grad_norm": 0.19678627736448578,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 37540
+    },
+    {
+      "epoch": 0.8935780253260973,
+      "grad_norm": 0.6912941629591705,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37541
+    },
+    {
+      "epoch": 0.8936018280491288,
+      "grad_norm": 2.821605988869825,
+      "learning_rate": 2e-05,
+      "loss": 0.4098,
+      "step": 37542
+    },
+    {
+      "epoch": 0.8936256307721603,
+      "grad_norm": 6.874046659927965,
+      "learning_rate": 2e-05,
+      "loss": 1.0941,
+      "step": 37543
+    },
+    {
+      "epoch": 0.8936494334951919,
+      "grad_norm": 2.8982535385754034,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 37544
+    },
+    {
+      "epoch": 0.8936732362182234,
+      "grad_norm": 0.033376532998734575,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37545
+    },
+    {
+      "epoch": 0.8936970389412549,
+      "grad_norm": 0.7519241845594374,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 37546
+    },
+    {
+      "epoch": 0.8937208416642864,
+      "grad_norm": 2.154370541396983,
+      "learning_rate": 2e-05,
+      "loss": 0.0814,
+      "step": 37547
+    },
+    {
+      "epoch": 0.8937446443873179,
+      "grad_norm": 4.147461346265436,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 37548
+    },
+    {
+      "epoch": 0.8937684471103494,
+      "grad_norm": 18.242252224328194,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 37549
+    },
+    {
+      "epoch": 0.8937922498333809,
+      "grad_norm": 4.269920215653466,
+      "learning_rate": 2e-05,
+      "loss": 0.6556,
+      "step": 37550
+    },
+    {
+      "epoch": 0.8938160525564125,
+      "grad_norm": 5.910552484225775,
+      "learning_rate": 2e-05,
+      "loss": 0.775,
+      "step": 37551
+    },
+    {
+      "epoch": 0.893839855279444,
+      "grad_norm": 5.539858171082999,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 37552
+    },
+    {
+      "epoch": 0.8938636580024755,
+      "grad_norm": 1.0687759623466844,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 37553
+    },
+    {
+      "epoch": 0.893887460725507,
+      "grad_norm": 1.2131476447136442,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 37554
+    },
+    {
+      "epoch": 0.8939112634485386,
+      "grad_norm": 4.843930580444369,
+      "learning_rate": 2e-05,
+      "loss": 0.6255,
+      "step": 37555
+    },
+    {
+      "epoch": 0.89393506617157,
+      "grad_norm": 1.2672890185918047,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 37556
+    },
+    {
+      "epoch": 0.8939588688946015,
+      "grad_norm": 0.5283664268731335,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 37557
+    },
+    {
+      "epoch": 0.893982671617633,
+      "grad_norm": 10.707581557755836,
+      "learning_rate": 2e-05,
+      "loss": 0.0845,
+      "step": 37558
+    },
+    {
+      "epoch": 0.8940064743406646,
+      "grad_norm": 4.380367393946242,
+      "learning_rate": 2e-05,
+      "loss": 0.5439,
+      "step": 37559
+    },
+    {
+      "epoch": 0.8940302770636961,
+      "grad_norm": 0.29173384383913153,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 37560
+    },
+    {
+      "epoch": 0.8940540797867276,
+      "grad_norm": 6.694594508702508,
+      "learning_rate": 2e-05,
+      "loss": 0.8095,
+      "step": 37561
+    },
+    {
+      "epoch": 0.8940778825097592,
+      "grad_norm": 4.555774419903093,
+      "learning_rate": 2e-05,
+      "loss": 0.7122,
+      "step": 37562
+    },
+    {
+      "epoch": 0.8941016852327907,
+      "grad_norm": 0.35661793280878507,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 37563
+    },
+    {
+      "epoch": 0.8941254879558221,
+      "grad_norm": 3.8776696379465276,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 37564
+    },
+    {
+      "epoch": 0.8941492906788536,
+      "grad_norm": 2.641076773512441,
+      "learning_rate": 2e-05,
+      "loss": 0.2848,
+      "step": 37565
+    },
+    {
+      "epoch": 0.8941730934018852,
+      "grad_norm": 11.586816318920615,
+      "learning_rate": 2e-05,
+      "loss": 2.5366,
+      "step": 37566
+    },
+    {
+      "epoch": 0.8941968961249167,
+      "grad_norm": 3.2584150222689456,
+      "learning_rate": 2e-05,
+      "loss": 0.3632,
+      "step": 37567
+    },
+    {
+      "epoch": 0.8942206988479482,
+      "grad_norm": 3.2810123190465315,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 37568
+    },
+    {
+      "epoch": 0.8942445015709797,
+      "grad_norm": 3.2205180313281025,
+      "learning_rate": 2e-05,
+      "loss": 0.6583,
+      "step": 37569
+    },
+    {
+      "epoch": 0.8942683042940113,
+      "grad_norm": 0.06937946662612383,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37570
+    },
+    {
+      "epoch": 0.8942921070170428,
+      "grad_norm": 5.9616426370106055,
+      "learning_rate": 2e-05,
+      "loss": 1.2859,
+      "step": 37571
+    },
+    {
+      "epoch": 0.8943159097400742,
+      "grad_norm": 4.915110063109055,
+      "learning_rate": 2e-05,
+      "loss": 1.1236,
+      "step": 37572
+    },
+    {
+      "epoch": 0.8943397124631057,
+      "grad_norm": 3.5497430400810503,
+      "learning_rate": 2e-05,
+      "loss": 0.3911,
+      "step": 37573
+    },
+    {
+      "epoch": 0.8943635151861373,
+      "grad_norm": 3.396008220773408,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 37574
+    },
+    {
+      "epoch": 0.8943873179091688,
+      "grad_norm": 3.282240366381652,
+      "learning_rate": 2e-05,
+      "loss": 0.5454,
+      "step": 37575
+    },
+    {
+      "epoch": 0.8944111206322003,
+      "grad_norm": 6.839049653415704,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 37576
+    },
+    {
+      "epoch": 0.8944349233552319,
+      "grad_norm": 7.282107025806991,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 37577
+    },
+    {
+      "epoch": 0.8944587260782634,
+      "grad_norm": 2.7345932810441216,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 37578
+    },
+    {
+      "epoch": 0.8944825288012949,
+      "grad_norm": 16.948088096305746,
+      "learning_rate": 2e-05,
+      "loss": 0.2384,
+      "step": 37579
+    },
+    {
+      "epoch": 0.8945063315243263,
+      "grad_norm": 4.25822481515019,
+      "learning_rate": 2e-05,
+      "loss": 0.4762,
+      "step": 37580
+    },
+    {
+      "epoch": 0.8945301342473579,
+      "grad_norm": 5.10193820913004,
+      "learning_rate": 2e-05,
+      "loss": 0.8114,
+      "step": 37581
+    },
+    {
+      "epoch": 0.8945539369703894,
+      "grad_norm": 4.238119937047163,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 37582
+    },
+    {
+      "epoch": 0.8945777396934209,
+      "grad_norm": 6.066833635049739,
+      "learning_rate": 2e-05,
+      "loss": 0.5755,
+      "step": 37583
+    },
+    {
+      "epoch": 0.8946015424164524,
+      "grad_norm": 3.8624964071952417,
+      "learning_rate": 2e-05,
+      "loss": 0.5973,
+      "step": 37584
+    },
+    {
+      "epoch": 0.894625345139484,
+      "grad_norm": 5.0134530973576945,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 37585
+    },
+    {
+      "epoch": 0.8946491478625155,
+      "grad_norm": 1.8294422505259804,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 37586
+    },
+    {
+      "epoch": 0.894672950585547,
+      "grad_norm": 0.043323118264472535,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37587
+    },
+    {
+      "epoch": 0.8946967533085785,
+      "grad_norm": 10.688140051818932,
+      "learning_rate": 2e-05,
+      "loss": 0.5011,
+      "step": 37588
+    },
+    {
+      "epoch": 0.89472055603161,
+      "grad_norm": 15.070873491167326,
+      "learning_rate": 2e-05,
+      "loss": 1.1098,
+      "step": 37589
+    },
+    {
+      "epoch": 0.8947443587546415,
+      "grad_norm": 16.092313984116025,
+      "learning_rate": 2e-05,
+      "loss": 1.0522,
+      "step": 37590
+    },
+    {
+      "epoch": 0.894768161477673,
+      "grad_norm": 0.34304215156172757,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 37591
+    },
+    {
+      "epoch": 0.8947919642007046,
+      "grad_norm": 4.94200771633913,
+      "learning_rate": 2e-05,
+      "loss": 0.9622,
+      "step": 37592
+    },
+    {
+      "epoch": 0.8948157669237361,
+      "grad_norm": 1.2171186032994685,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 37593
+    },
+    {
+      "epoch": 0.8948395696467676,
+      "grad_norm": 0.09014733387053774,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37594
+    },
+    {
+      "epoch": 0.8948633723697991,
+      "grad_norm": 5.593864223875627,
+      "learning_rate": 2e-05,
+      "loss": 0.8304,
+      "step": 37595
+    },
+    {
+      "epoch": 0.8948871750928307,
+      "grad_norm": 6.664264197405372,
+      "learning_rate": 2e-05,
+      "loss": 0.5336,
+      "step": 37596
+    },
+    {
+      "epoch": 0.8949109778158622,
+      "grad_norm": 0.17278024733119374,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37597
+    },
+    {
+      "epoch": 0.8949347805388936,
+      "grad_norm": 3.323780760016217,
+      "learning_rate": 2e-05,
+      "loss": 0.6109,
+      "step": 37598
+    },
+    {
+      "epoch": 0.8949585832619251,
+      "grad_norm": 5.032559937530913,
+      "learning_rate": 2e-05,
+      "loss": 0.1272,
+      "step": 37599
+    },
+    {
+      "epoch": 0.8949823859849567,
+      "grad_norm": 0.1791282634477308,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37600
+    },
+    {
+      "epoch": 0.8950061887079882,
+      "grad_norm": 0.769556735373522,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 37601
+    },
+    {
+      "epoch": 0.8950299914310197,
+      "grad_norm": 2.335051573578881,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 37602
+    },
+    {
+      "epoch": 0.8950537941540512,
+      "grad_norm": 0.14737146372755544,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 37603
+    },
+    {
+      "epoch": 0.8950775968770828,
+      "grad_norm": 8.916299856341137,
+      "learning_rate": 2e-05,
+      "loss": 0.8157,
+      "step": 37604
+    },
+    {
+      "epoch": 0.8951013996001143,
+      "grad_norm": 10.006278724543725,
+      "learning_rate": 2e-05,
+      "loss": 0.2455,
+      "step": 37605
+    },
+    {
+      "epoch": 0.8951252023231457,
+      "grad_norm": 0.01076838828743192,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37606
+    },
+    {
+      "epoch": 0.8951490050461773,
+      "grad_norm": 4.867853600721258,
+      "learning_rate": 2e-05,
+      "loss": 0.1768,
+      "step": 37607
+    },
+    {
+      "epoch": 0.8951728077692088,
+      "grad_norm": 3.691579993331643,
+      "learning_rate": 2e-05,
+      "loss": 0.5111,
+      "step": 37608
+    },
+    {
+      "epoch": 0.8951966104922403,
+      "grad_norm": 1.4104429838379884,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 37609
+    },
+    {
+      "epoch": 0.8952204132152718,
+      "grad_norm": 4.614343316175005,
+      "learning_rate": 2e-05,
+      "loss": 0.439,
+      "step": 37610
+    },
+    {
+      "epoch": 0.8952442159383034,
+      "grad_norm": 0.6209531219519322,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 37611
+    },
+    {
+      "epoch": 0.8952680186613349,
+      "grad_norm": 6.0827930848132725,
+      "learning_rate": 2e-05,
+      "loss": 1.4671,
+      "step": 37612
+    },
+    {
+      "epoch": 0.8952918213843664,
+      "grad_norm": 0.009288889486120343,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37613
+    },
+    {
+      "epoch": 0.8953156241073978,
+      "grad_norm": 3.398086553289297,
+      "learning_rate": 2e-05,
+      "loss": 0.5507,
+      "step": 37614
+    },
+    {
+      "epoch": 0.8953394268304294,
+      "grad_norm": 4.995900231514657,
+      "learning_rate": 2e-05,
+      "loss": 0.4267,
+      "step": 37615
+    },
+    {
+      "epoch": 0.8953632295534609,
+      "grad_norm": 18.326617867879488,
+      "learning_rate": 2e-05,
+      "loss": 0.7758,
+      "step": 37616
+    },
+    {
+      "epoch": 0.8953870322764924,
+      "grad_norm": 14.097516729615581,
+      "learning_rate": 2e-05,
+      "loss": 0.4723,
+      "step": 37617
+    },
+    {
+      "epoch": 0.8954108349995239,
+      "grad_norm": 3.359935118279008,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 37618
+    },
+    {
+      "epoch": 0.8954346377225555,
+      "grad_norm": 0.0637511285630751,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 37619
+    },
+    {
+      "epoch": 0.895458440445587,
+      "grad_norm": 0.14310690091624131,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 37620
+    },
+    {
+      "epoch": 0.8954822431686185,
+      "grad_norm": 7.406662358903585,
+      "learning_rate": 2e-05,
+      "loss": 0.5547,
+      "step": 37621
+    },
+    {
+      "epoch": 0.8955060458916501,
+      "grad_norm": 4.58924923756846,
+      "learning_rate": 2e-05,
+      "loss": 0.6739,
+      "step": 37622
+    },
+    {
+      "epoch": 0.8955298486146815,
+      "grad_norm": 3.3420008256158003,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 37623
+    },
+    {
+      "epoch": 0.895553651337713,
+      "grad_norm": 5.197244468928725,
+      "learning_rate": 2e-05,
+      "loss": 1.1906,
+      "step": 37624
+    },
+    {
+      "epoch": 0.8955774540607445,
+      "grad_norm": 23.884556358699708,
+      "learning_rate": 2e-05,
+      "loss": 1.6618,
+      "step": 37625
+    },
+    {
+      "epoch": 0.8956012567837761,
+      "grad_norm": 0.40506835463708524,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 37626
+    },
+    {
+      "epoch": 0.8956250595068076,
+      "grad_norm": 5.379027476888883,
+      "learning_rate": 2e-05,
+      "loss": 0.1641,
+      "step": 37627
+    },
+    {
+      "epoch": 0.8956488622298391,
+      "grad_norm": 28.722883519760664,
+      "learning_rate": 2e-05,
+      "loss": 1.5641,
+      "step": 37628
+    },
+    {
+      "epoch": 0.8956726649528706,
+      "grad_norm": 0.05035251921003797,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37629
+    },
+    {
+      "epoch": 0.8956964676759022,
+      "grad_norm": 0.0055592075832561925,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 37630
+    },
+    {
+      "epoch": 0.8957202703989336,
+      "grad_norm": 4.715504852150637,
+      "learning_rate": 2e-05,
+      "loss": 0.8398,
+      "step": 37631
+    },
+    {
+      "epoch": 0.8957440731219651,
+      "grad_norm": 1.8379424582834198,
+      "learning_rate": 2e-05,
+      "loss": 0.1656,
+      "step": 37632
+    },
+    {
+      "epoch": 0.8957678758449966,
+      "grad_norm": 2.151667626339696,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 37633
+    },
+    {
+      "epoch": 0.8957916785680282,
+      "grad_norm": 1.75773775442395,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 37634
+    },
+    {
+      "epoch": 0.8958154812910597,
+      "grad_norm": 5.099786598319569,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 37635
+    },
+    {
+      "epoch": 0.8958392840140912,
+      "grad_norm": 10.577312144001139,
+      "learning_rate": 2e-05,
+      "loss": 0.7427,
+      "step": 37636
+    },
+    {
+      "epoch": 0.8958630867371227,
+      "grad_norm": 48.63162049172804,
+      "learning_rate": 2e-05,
+      "loss": 0.4868,
+      "step": 37637
+    },
+    {
+      "epoch": 0.8958868894601543,
+      "grad_norm": 11.91618280794068,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 37638
+    },
+    {
+      "epoch": 0.8959106921831858,
+      "grad_norm": 3.148566549979311,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 37639
+    },
+    {
+      "epoch": 0.8959344949062172,
+      "grad_norm": 1.4925310726864094,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 37640
+    },
+    {
+      "epoch": 0.8959582976292488,
+      "grad_norm": 0.06914040374394098,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37641
+    },
+    {
+      "epoch": 0.8959821003522803,
+      "grad_norm": 0.06337132466957712,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 37642
+    },
+    {
+      "epoch": 0.8960059030753118,
+      "grad_norm": 5.11497758592653,
+      "learning_rate": 2e-05,
+      "loss": 0.1743,
+      "step": 37643
+    },
+    {
+      "epoch": 0.8960297057983433,
+      "grad_norm": 3.7021459624587485,
+      "learning_rate": 2e-05,
+      "loss": 0.8553,
+      "step": 37644
+    },
+    {
+      "epoch": 0.8960535085213749,
+      "grad_norm": 2.957283529165347,
+      "learning_rate": 2e-05,
+      "loss": 0.3908,
+      "step": 37645
+    },
+    {
+      "epoch": 0.8960773112444064,
+      "grad_norm": 6.175340402773613,
+      "learning_rate": 2e-05,
+      "loss": 0.7805,
+      "step": 37646
+    },
+    {
+      "epoch": 0.8961011139674379,
+      "grad_norm": 0.021869998627413553,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37647
+    },
+    {
+      "epoch": 0.8961249166904693,
+      "grad_norm": 0.02482632718454173,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 37648
+    },
+    {
+      "epoch": 0.8961487194135009,
+      "grad_norm": 2.8835494979975023,
+      "learning_rate": 2e-05,
+      "loss": 0.2827,
+      "step": 37649
+    },
+    {
+      "epoch": 0.8961725221365324,
+      "grad_norm": 8.951239332727,
+      "learning_rate": 2e-05,
+      "loss": 1.689,
+      "step": 37650
+    },
+    {
+      "epoch": 0.8961963248595639,
+      "grad_norm": 4.724291154124647,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 37651
+    },
+    {
+      "epoch": 0.8962201275825954,
+      "grad_norm": 1.4995787172438177,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 37652
+    },
+    {
+      "epoch": 0.896243930305627,
+      "grad_norm": 8.411435881314047,
+      "learning_rate": 2e-05,
+      "loss": 0.6653,
+      "step": 37653
+    },
+    {
+      "epoch": 0.8962677330286585,
+      "grad_norm": 1.8990142950915507,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 37654
+    },
+    {
+      "epoch": 0.89629153575169,
+      "grad_norm": 7.174495746175758,
+      "learning_rate": 2e-05,
+      "loss": 0.4418,
+      "step": 37655
+    },
+    {
+      "epoch": 0.8963153384747216,
+      "grad_norm": 4.107514531536098,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 37656
+    },
+    {
+      "epoch": 0.896339141197753,
+      "grad_norm": 5.0661670000203225,
+      "learning_rate": 2e-05,
+      "loss": 1.2449,
+      "step": 37657
+    },
+    {
+      "epoch": 0.8963629439207845,
+      "grad_norm": 4.899096970862906,
+      "learning_rate": 2e-05,
+      "loss": 0.9776,
+      "step": 37658
+    },
+    {
+      "epoch": 0.896386746643816,
+      "grad_norm": 12.008608717526181,
+      "learning_rate": 2e-05,
+      "loss": 0.3755,
+      "step": 37659
+    },
+    {
+      "epoch": 0.8964105493668476,
+      "grad_norm": 6.113430610550194,
+      "learning_rate": 2e-05,
+      "loss": 0.18,
+      "step": 37660
+    },
+    {
+      "epoch": 0.8964343520898791,
+      "grad_norm": 14.895672002050516,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 37661
+    },
+    {
+      "epoch": 0.8964581548129106,
+      "grad_norm": 1.1773499599296628,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 37662
+    },
+    {
+      "epoch": 0.8964819575359421,
+      "grad_norm": 2.016640208183541,
+      "learning_rate": 2e-05,
+      "loss": 0.1887,
+      "step": 37663
+    },
+    {
+      "epoch": 0.8965057602589737,
+      "grad_norm": 0.45352053575238266,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 37664
+    },
+    {
+      "epoch": 0.8965295629820051,
+      "grad_norm": 2.8804253595427203,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 37665
+    },
+    {
+      "epoch": 0.8965533657050366,
+      "grad_norm": 11.963199635014453,
+      "learning_rate": 2e-05,
+      "loss": 0.2195,
+      "step": 37666
+    },
+    {
+      "epoch": 0.8965771684280681,
+      "grad_norm": 4.965630585769953,
+      "learning_rate": 2e-05,
+      "loss": 0.1161,
+      "step": 37667
+    },
+    {
+      "epoch": 0.8966009711510997,
+      "grad_norm": 2.343961261186443,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 37668
+    },
+    {
+      "epoch": 0.8966247738741312,
+      "grad_norm": 0.13929360205853378,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 37669
+    },
+    {
+      "epoch": 0.8966485765971627,
+      "grad_norm": 0.037057654116995895,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37670
+    },
+    {
+      "epoch": 0.8966723793201943,
+      "grad_norm": 0.24263391947833865,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 37671
+    },
+    {
+      "epoch": 0.8966961820432258,
+      "grad_norm": 3.0007981519773144,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 37672
+    },
+    {
+      "epoch": 0.8967199847662572,
+      "grad_norm": 0.02301268620537118,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 37673
+    },
+    {
+      "epoch": 0.8967437874892887,
+      "grad_norm": 0.1142570871447064,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 37674
+    },
+    {
+      "epoch": 0.8967675902123203,
+      "grad_norm": 3.124658435991269,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 37675
+    },
+    {
+      "epoch": 0.8967913929353518,
+      "grad_norm": 6.573454809660727,
+      "learning_rate": 2e-05,
+      "loss": 1.8544,
+      "step": 37676
+    },
+    {
+      "epoch": 0.8968151956583833,
+      "grad_norm": 0.04445329312841427,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 37677
+    },
+    {
+      "epoch": 0.8968389983814148,
+      "grad_norm": 3.935923701163246,
+      "learning_rate": 2e-05,
+      "loss": 0.6992,
+      "step": 37678
+    },
+    {
+      "epoch": 0.8968628011044464,
+      "grad_norm": 3.950686218115241,
+      "learning_rate": 2e-05,
+      "loss": 0.6447,
+      "step": 37679
+    },
+    {
+      "epoch": 0.8968866038274779,
+      "grad_norm": 7.962461105310484,
+      "learning_rate": 2e-05,
+      "loss": 0.1521,
+      "step": 37680
+    },
+    {
+      "epoch": 0.8969104065505094,
+      "grad_norm": 1.2714149940907222,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 37681
+    },
+    {
+      "epoch": 0.8969342092735408,
+      "grad_norm": 2.888823669604499,
+      "learning_rate": 2e-05,
+      "loss": 0.6656,
+      "step": 37682
+    },
+    {
+      "epoch": 0.8969580119965724,
+      "grad_norm": 7.816942577925926,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 37683
+    },
+    {
+      "epoch": 0.8969818147196039,
+      "grad_norm": 0.051468557963633324,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37684
+    },
+    {
+      "epoch": 0.8970056174426354,
+      "grad_norm": 3.633566357855458,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 37685
+    },
+    {
+      "epoch": 0.897029420165667,
+      "grad_norm": 0.37331618853491744,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 37686
+    },
+    {
+      "epoch": 0.8970532228886985,
+      "grad_norm": 0.6561781614419925,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37687
+    },
+    {
+      "epoch": 0.89707702561173,
+      "grad_norm": 3.527751064630529,
+      "learning_rate": 2e-05,
+      "loss": 0.3319,
+      "step": 37688
+    },
+    {
+      "epoch": 0.8971008283347615,
+      "grad_norm": 5.149620001000997,
+      "learning_rate": 2e-05,
+      "loss": 1.1267,
+      "step": 37689
+    },
+    {
+      "epoch": 0.897124631057793,
+      "grad_norm": 0.2881998497891175,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 37690
+    },
+    {
+      "epoch": 0.8971484337808245,
+      "grad_norm": 7.852717359636523,
+      "learning_rate": 2e-05,
+      "loss": 0.7109,
+      "step": 37691
+    },
+    {
+      "epoch": 0.897172236503856,
+      "grad_norm": 5.02670354153069,
+      "learning_rate": 2e-05,
+      "loss": 0.17,
+      "step": 37692
+    },
+    {
+      "epoch": 0.8971960392268875,
+      "grad_norm": 3.1813716320922163,
+      "learning_rate": 2e-05,
+      "loss": 0.4236,
+      "step": 37693
+    },
+    {
+      "epoch": 0.8972198419499191,
+      "grad_norm": 0.04655919918619708,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 37694
+    },
+    {
+      "epoch": 0.8972436446729506,
+      "grad_norm": 14.619323652859393,
+      "learning_rate": 2e-05,
+      "loss": 1.4972,
+      "step": 37695
+    },
+    {
+      "epoch": 0.8972674473959821,
+      "grad_norm": 10.542432838368274,
+      "learning_rate": 2e-05,
+      "loss": 1.4638,
+      "step": 37696
+    },
+    {
+      "epoch": 0.8972912501190136,
+      "grad_norm": 1.3569368947092098,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 37697
+    },
+    {
+      "epoch": 0.8973150528420452,
+      "grad_norm": 0.31657396351888034,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 37698
+    },
+    {
+      "epoch": 0.8973388555650766,
+      "grad_norm": 0.9162653892743167,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 37699
+    },
+    {
+      "epoch": 0.8973626582881081,
+      "grad_norm": 3.612204613580503,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 37700
+    },
+    {
+      "epoch": 0.8973864610111397,
+      "grad_norm": 0.6985848937738062,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 37701
+    },
+    {
+      "epoch": 0.8974102637341712,
+      "grad_norm": 23.154317827467604,
+      "learning_rate": 2e-05,
+      "loss": 0.1808,
+      "step": 37702
+    },
+    {
+      "epoch": 0.8974340664572027,
+      "grad_norm": 3.0315296196401595,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 37703
+    },
+    {
+      "epoch": 0.8974578691802342,
+      "grad_norm": 2.0721952219466973,
+      "learning_rate": 2e-05,
+      "loss": 0.11,
+      "step": 37704
+    },
+    {
+      "epoch": 0.8974816719032658,
+      "grad_norm": 0.023793606880599793,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 37705
+    },
+    {
+      "epoch": 0.8975054746262973,
+      "grad_norm": 2.6987088881971735,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 37706
+    },
+    {
+      "epoch": 0.8975292773493287,
+      "grad_norm": 3.2079041718979804,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 37707
+    },
+    {
+      "epoch": 0.8975530800723602,
+      "grad_norm": 5.955637098171811,
+      "learning_rate": 2e-05,
+      "loss": 0.2431,
+      "step": 37708
+    },
+    {
+      "epoch": 0.8975768827953918,
+      "grad_norm": 3.196353211513089,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 37709
+    },
+    {
+      "epoch": 0.8976006855184233,
+      "grad_norm": 8.412167284949572,
+      "learning_rate": 2e-05,
+      "loss": 1.1497,
+      "step": 37710
+    },
+    {
+      "epoch": 0.8976244882414548,
+      "grad_norm": 1.4407753919619009,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 37711
+    },
+    {
+      "epoch": 0.8976482909644863,
+      "grad_norm": 0.30412844965458,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37712
+    },
+    {
+      "epoch": 0.8976720936875179,
+      "grad_norm": 0.07344983174352365,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 37713
+    },
+    {
+      "epoch": 0.8976958964105494,
+      "grad_norm": 9.430912449433537,
+      "learning_rate": 2e-05,
+      "loss": 0.8449,
+      "step": 37714
+    },
+    {
+      "epoch": 0.8977196991335809,
+      "grad_norm": 5.885917475446355,
+      "learning_rate": 2e-05,
+      "loss": 0.8591,
+      "step": 37715
+    },
+    {
+      "epoch": 0.8977435018566124,
+      "grad_norm": 1.6372132469316842,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 37716
+    },
+    {
+      "epoch": 0.8977673045796439,
+      "grad_norm": 9.038775500473529,
+      "learning_rate": 2e-05,
+      "loss": 1.7869,
+      "step": 37717
+    },
+    {
+      "epoch": 0.8977911073026754,
+      "grad_norm": 4.97988992498081,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 37718
+    },
+    {
+      "epoch": 0.8978149100257069,
+      "grad_norm": 3.6307874310042925,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 37719
+    },
+    {
+      "epoch": 0.8978387127487385,
+      "grad_norm": 0.2900952830055561,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 37720
+    },
+    {
+      "epoch": 0.89786251547177,
+      "grad_norm": 17.54474113141466,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 37721
+    },
+    {
+      "epoch": 0.8978863181948015,
+      "grad_norm": 0.9185844521997089,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 37722
+    },
+    {
+      "epoch": 0.897910120917833,
+      "grad_norm": 5.177147202448712,
+      "learning_rate": 2e-05,
+      "loss": 1.506,
+      "step": 37723
+    },
+    {
+      "epoch": 0.8979339236408646,
+      "grad_norm": 8.078578050494952,
+      "learning_rate": 2e-05,
+      "loss": 0.8508,
+      "step": 37724
+    },
+    {
+      "epoch": 0.897957726363896,
+      "grad_norm": 2.320372083433521,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 37725
+    },
+    {
+      "epoch": 0.8979815290869275,
+      "grad_norm": 0.16686748379487862,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37726
+    },
+    {
+      "epoch": 0.898005331809959,
+      "grad_norm": 0.4880133594430952,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 37727
+    },
+    {
+      "epoch": 0.8980291345329906,
+      "grad_norm": 4.599715319753231,
+      "learning_rate": 2e-05,
+      "loss": 0.4891,
+      "step": 37728
+    },
+    {
+      "epoch": 0.8980529372560221,
+      "grad_norm": 2.3833575246031993,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 37729
+    },
+    {
+      "epoch": 0.8980767399790536,
+      "grad_norm": 6.866156070087999,
+      "learning_rate": 2e-05,
+      "loss": 0.2776,
+      "step": 37730
+    },
+    {
+      "epoch": 0.8981005427020852,
+      "grad_norm": 3.870258049539195,
+      "learning_rate": 2e-05,
+      "loss": 0.6454,
+      "step": 37731
+    },
+    {
+      "epoch": 0.8981243454251167,
+      "grad_norm": 2.6998823352263384,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 37732
+    },
+    {
+      "epoch": 0.8981481481481481,
+      "grad_norm": 10.412367370141766,
+      "learning_rate": 2e-05,
+      "loss": 1.4053,
+      "step": 37733
+    },
+    {
+      "epoch": 0.8981719508711796,
+      "grad_norm": 7.210195447098537,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 37734
+    },
+    {
+      "epoch": 0.8981957535942112,
+      "grad_norm": 0.02121766558211904,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 37735
+    },
+    {
+      "epoch": 0.8982195563172427,
+      "grad_norm": 4.672108580370066,
+      "learning_rate": 2e-05,
+      "loss": 0.061,
+      "step": 37736
+    },
+    {
+      "epoch": 0.8982433590402742,
+      "grad_norm": 1.1389869535733435,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 37737
+    },
+    {
+      "epoch": 0.8982671617633057,
+      "grad_norm": 1.1172616373158144,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 37738
+    },
+    {
+      "epoch": 0.8982909644863373,
+      "grad_norm": 1.259380114243613,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 37739
+    },
+    {
+      "epoch": 0.8983147672093688,
+      "grad_norm": 5.952313874963119,
+      "learning_rate": 2e-05,
+      "loss": 1.1013,
+      "step": 37740
+    },
+    {
+      "epoch": 0.8983385699324002,
+      "grad_norm": 4.148319114798166,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 37741
+    },
+    {
+      "epoch": 0.8983623726554317,
+      "grad_norm": 2.0964453228214444,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 37742
+    },
+    {
+      "epoch": 0.8983861753784633,
+      "grad_norm": 3.137998042415748,
+      "learning_rate": 2e-05,
+      "loss": 0.4474,
+      "step": 37743
+    },
+    {
+      "epoch": 0.8984099781014948,
+      "grad_norm": 1.9354332007103565,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 37744
+    },
+    {
+      "epoch": 0.8984337808245263,
+      "grad_norm": 3.435279063697223,
+      "learning_rate": 2e-05,
+      "loss": 0.49,
+      "step": 37745
+    },
+    {
+      "epoch": 0.8984575835475579,
+      "grad_norm": 43.27050118179932,
+      "learning_rate": 2e-05,
+      "loss": 0.2887,
+      "step": 37746
+    },
+    {
+      "epoch": 0.8984813862705894,
+      "grad_norm": 4.590008091962133,
+      "learning_rate": 2e-05,
+      "loss": 0.6707,
+      "step": 37747
+    },
+    {
+      "epoch": 0.8985051889936209,
+      "grad_norm": 0.10456255611134582,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 37748
+    },
+    {
+      "epoch": 0.8985289917166523,
+      "grad_norm": 2.427268268581365,
+      "learning_rate": 2e-05,
+      "loss": 0.2779,
+      "step": 37749
+    },
+    {
+      "epoch": 0.8985527944396839,
+      "grad_norm": 0.05595365486517813,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 37750
+    },
+    {
+      "epoch": 0.8985765971627154,
+      "grad_norm": 4.141617740507698,
+      "learning_rate": 2e-05,
+      "loss": 0.6884,
+      "step": 37751
+    },
+    {
+      "epoch": 0.8986003998857469,
+      "grad_norm": 0.012766524972377437,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37752
+    },
+    {
+      "epoch": 0.8986242026087784,
+      "grad_norm": 1.6193763214705252,
+      "learning_rate": 2e-05,
+      "loss": 0.0214,
+      "step": 37753
+    },
+    {
+      "epoch": 0.89864800533181,
+      "grad_norm": 0.2880119785841563,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 37754
+    },
+    {
+      "epoch": 0.8986718080548415,
+      "grad_norm": 6.085395859448203,
+      "learning_rate": 2e-05,
+      "loss": 0.2466,
+      "step": 37755
+    },
+    {
+      "epoch": 0.898695610777873,
+      "grad_norm": 17.95895273011027,
+      "learning_rate": 2e-05,
+      "loss": 0.6444,
+      "step": 37756
+    },
+    {
+      "epoch": 0.8987194135009045,
+      "grad_norm": 1.7820645578246537,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 37757
+    },
+    {
+      "epoch": 0.898743216223936,
+      "grad_norm": 2.227851199252908,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 37758
+    },
+    {
+      "epoch": 0.8987670189469675,
+      "grad_norm": 4.472182773757094,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 37759
+    },
+    {
+      "epoch": 0.898790821669999,
+      "grad_norm": 6.075509967686776,
+      "learning_rate": 2e-05,
+      "loss": 0.676,
+      "step": 37760
+    },
+    {
+      "epoch": 0.8988146243930306,
+      "grad_norm": 0.2536392274710697,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 37761
+    },
+    {
+      "epoch": 0.8988384271160621,
+      "grad_norm": 3.2211319370263274,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 37762
+    },
+    {
+      "epoch": 0.8988622298390936,
+      "grad_norm": 4.742771989883807,
+      "learning_rate": 2e-05,
+      "loss": 0.754,
+      "step": 37763
+    },
+    {
+      "epoch": 0.8988860325621251,
+      "grad_norm": 0.7328282047720777,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 37764
+    },
+    {
+      "epoch": 0.8989098352851567,
+      "grad_norm": 6.266568663432139,
+      "learning_rate": 2e-05,
+      "loss": 1.2511,
+      "step": 37765
+    },
+    {
+      "epoch": 0.8989336380081882,
+      "grad_norm": 4.050975713382134,
+      "learning_rate": 2e-05,
+      "loss": 0.6687,
+      "step": 37766
+    },
+    {
+      "epoch": 0.8989574407312196,
+      "grad_norm": 10.630022449036437,
+      "learning_rate": 2e-05,
+      "loss": 0.8555,
+      "step": 37767
+    },
+    {
+      "epoch": 0.8989812434542511,
+      "grad_norm": 9.070378868111115,
+      "learning_rate": 2e-05,
+      "loss": 0.7526,
+      "step": 37768
+    },
+    {
+      "epoch": 0.8990050461772827,
+      "grad_norm": 12.86232088849006,
+      "learning_rate": 2e-05,
+      "loss": 0.8048,
+      "step": 37769
+    },
+    {
+      "epoch": 0.8990288489003142,
+      "grad_norm": 4.422011544500652,
+      "learning_rate": 2e-05,
+      "loss": 1.0092,
+      "step": 37770
+    },
+    {
+      "epoch": 0.8990526516233457,
+      "grad_norm": 10.241838046750132,
+      "learning_rate": 2e-05,
+      "loss": 1.6014,
+      "step": 37771
+    },
+    {
+      "epoch": 0.8990764543463772,
+      "grad_norm": 6.334029026142752,
+      "learning_rate": 2e-05,
+      "loss": 0.3585,
+      "step": 37772
+    },
+    {
+      "epoch": 0.8991002570694088,
+      "grad_norm": 6.275274019846185,
+      "learning_rate": 2e-05,
+      "loss": 1.2525,
+      "step": 37773
+    },
+    {
+      "epoch": 0.8991240597924403,
+      "grad_norm": 7.308178116634004,
+      "learning_rate": 2e-05,
+      "loss": 0.6692,
+      "step": 37774
+    },
+    {
+      "epoch": 0.8991478625154717,
+      "grad_norm": 17.701242314642066,
+      "learning_rate": 2e-05,
+      "loss": 0.1958,
+      "step": 37775
+    },
+    {
+      "epoch": 0.8991716652385033,
+      "grad_norm": 2.989685675939505,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 37776
+    },
+    {
+      "epoch": 0.8991954679615348,
+      "grad_norm": 2.0286068002093978,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 37777
+    },
+    {
+      "epoch": 0.8992192706845663,
+      "grad_norm": 6.534668703345313,
+      "learning_rate": 2e-05,
+      "loss": 1.2959,
+      "step": 37778
+    },
+    {
+      "epoch": 0.8992430734075978,
+      "grad_norm": 3.211014307249651,
+      "learning_rate": 2e-05,
+      "loss": 0.3143,
+      "step": 37779
+    },
+    {
+      "epoch": 0.8992668761306294,
+      "grad_norm": 4.918525170083801,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 37780
+    },
+    {
+      "epoch": 0.8992906788536609,
+      "grad_norm": 1.7857639607429565,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 37781
+    },
+    {
+      "epoch": 0.8993144815766924,
+      "grad_norm": 1.9506742660924983,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 37782
+    },
+    {
+      "epoch": 0.8993382842997238,
+      "grad_norm": 2.729654257929698,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 37783
+    },
+    {
+      "epoch": 0.8993620870227554,
+      "grad_norm": 3.0882939611823494,
+      "learning_rate": 2e-05,
+      "loss": 0.4032,
+      "step": 37784
+    },
+    {
+      "epoch": 0.8993858897457869,
+      "grad_norm": 2.794386832020748,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 37785
+    },
+    {
+      "epoch": 0.8994096924688184,
+      "grad_norm": 3.679253595635368,
+      "learning_rate": 2e-05,
+      "loss": 0.4828,
+      "step": 37786
+    },
+    {
+      "epoch": 0.8994334951918499,
+      "grad_norm": 6.090131333820619,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 37787
+    },
+    {
+      "epoch": 0.8994572979148815,
+      "grad_norm": 2.426951378617457,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 37788
+    },
+    {
+      "epoch": 0.899481100637913,
+      "grad_norm": 2.4629998874036816,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 37789
+    },
+    {
+      "epoch": 0.8995049033609445,
+      "grad_norm": 0.5805948559273755,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 37790
+    },
+    {
+      "epoch": 0.8995287060839761,
+      "grad_norm": 0.32086018456965526,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 37791
+    },
+    {
+      "epoch": 0.8995525088070075,
+      "grad_norm": 24.69463293336274,
+      "learning_rate": 2e-05,
+      "loss": 0.8404,
+      "step": 37792
+    },
+    {
+      "epoch": 0.899576311530039,
+      "grad_norm": 0.8648609747863304,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 37793
+    },
+    {
+      "epoch": 0.8996001142530705,
+      "grad_norm": 3.756602844457796,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 37794
+    },
+    {
+      "epoch": 0.8996239169761021,
+      "grad_norm": 12.737646346908846,
+      "learning_rate": 2e-05,
+      "loss": 0.4572,
+      "step": 37795
+    },
+    {
+      "epoch": 0.8996477196991336,
+      "grad_norm": 7.4929745662010205,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 37796
+    },
+    {
+      "epoch": 0.8996715224221651,
+      "grad_norm": 2.373715870040146,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 37797
+    },
+    {
+      "epoch": 0.8996953251451966,
+      "grad_norm": 4.487651332117351,
+      "learning_rate": 2e-05,
+      "loss": 0.5565,
+      "step": 37798
+    },
+    {
+      "epoch": 0.8997191278682282,
+      "grad_norm": 3.1292183654871217,
+      "learning_rate": 2e-05,
+      "loss": 0.3517,
+      "step": 37799
+    },
+    {
+      "epoch": 0.8997429305912596,
+      "grad_norm": 4.864801880320454,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 37800
+    },
+    {
+      "epoch": 0.8997667333142911,
+      "grad_norm": 0.8797146473305891,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 37801
+    },
+    {
+      "epoch": 0.8997905360373226,
+      "grad_norm": 12.508847381812476,
+      "learning_rate": 2e-05,
+      "loss": 0.4203,
+      "step": 37802
+    },
+    {
+      "epoch": 0.8998143387603542,
+      "grad_norm": 4.1459491713745695,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 37803
+    },
+    {
+      "epoch": 0.8998381414833857,
+      "grad_norm": 7.793141703170801,
+      "learning_rate": 2e-05,
+      "loss": 0.3489,
+      "step": 37804
+    },
+    {
+      "epoch": 0.8998619442064172,
+      "grad_norm": 0.22145411664344902,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 37805
+    },
+    {
+      "epoch": 0.8998857469294488,
+      "grad_norm": 0.9272356341527161,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 37806
+    },
+    {
+      "epoch": 0.8999095496524803,
+      "grad_norm": 3.5870017074694283,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 37807
+    },
+    {
+      "epoch": 0.8999333523755118,
+      "grad_norm": 0.6744585077680884,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 37808
+    },
+    {
+      "epoch": 0.8999571550985432,
+      "grad_norm": 3.9603972273972823,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 37809
+    },
+    {
+      "epoch": 0.8999809578215748,
+      "grad_norm": 7.633357979702536,
+      "learning_rate": 2e-05,
+      "loss": 1.2791,
+      "step": 37810
+    },
+    {
+      "epoch": 0.9000047605446063,
+      "grad_norm": 6.452941467796022,
+      "learning_rate": 2e-05,
+      "loss": 1.5666,
+      "step": 37811
+    },
+    {
+      "epoch": 0.9000285632676378,
+      "grad_norm": 2.60616214698068,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 37812
+    },
+    {
+      "epoch": 0.9000523659906693,
+      "grad_norm": 0.8812963589695904,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 37813
+    },
+    {
+      "epoch": 0.9000761687137009,
+      "grad_norm": 3.3590997937392224,
+      "learning_rate": 2e-05,
+      "loss": 0.5407,
+      "step": 37814
+    },
+    {
+      "epoch": 0.9000999714367324,
+      "grad_norm": 3.2031543337844286,
+      "learning_rate": 2e-05,
+      "loss": 0.1463,
+      "step": 37815
+    },
+    {
+      "epoch": 0.9001237741597639,
+      "grad_norm": 0.2701929908803506,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 37816
+    },
+    {
+      "epoch": 0.9001475768827953,
+      "grad_norm": 4.4992587645074975,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 37817
+    },
+    {
+      "epoch": 0.9001713796058269,
+      "grad_norm": 5.113470539989718,
+      "learning_rate": 2e-05,
+      "loss": 0.4257,
+      "step": 37818
+    },
+    {
+      "epoch": 0.9001951823288584,
+      "grad_norm": 10.484808665952936,
+      "learning_rate": 2e-05,
+      "loss": 0.3907,
+      "step": 37819
+    },
+    {
+      "epoch": 0.9002189850518899,
+      "grad_norm": 0.4712396915793002,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 37820
+    },
+    {
+      "epoch": 0.9002427877749215,
+      "grad_norm": 0.051528116216172905,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 37821
+    },
+    {
+      "epoch": 0.900266590497953,
+      "grad_norm": 10.509486020660903,
+      "learning_rate": 2e-05,
+      "loss": 0.1127,
+      "step": 37822
+    },
+    {
+      "epoch": 0.9002903932209845,
+      "grad_norm": 0.16227113647898556,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 37823
+    },
+    {
+      "epoch": 0.900314195944016,
+      "grad_norm": 0.32990033294058335,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 37824
+    },
+    {
+      "epoch": 0.9003379986670476,
+      "grad_norm": 3.468591939931311,
+      "learning_rate": 2e-05,
+      "loss": 0.6693,
+      "step": 37825
+    },
+    {
+      "epoch": 0.900361801390079,
+      "grad_norm": 0.38497324686979223,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 37826
+    },
+    {
+      "epoch": 0.9003856041131105,
+      "grad_norm": 3.945817701061705,
+      "learning_rate": 2e-05,
+      "loss": 0.6475,
+      "step": 37827
+    },
+    {
+      "epoch": 0.900409406836142,
+      "grad_norm": 8.695840886011245,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 37828
+    },
+    {
+      "epoch": 0.9004332095591736,
+      "grad_norm": 3.0141094923851126,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 37829
+    },
+    {
+      "epoch": 0.9004570122822051,
+      "grad_norm": 0.639737553572209,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 37830
+    },
+    {
+      "epoch": 0.9004808150052366,
+      "grad_norm": 4.8543295132134645,
+      "learning_rate": 2e-05,
+      "loss": 0.7181,
+      "step": 37831
+    },
+    {
+      "epoch": 0.9005046177282681,
+      "grad_norm": 4.430616278884742,
+      "learning_rate": 2e-05,
+      "loss": 1.1865,
+      "step": 37832
+    },
+    {
+      "epoch": 0.9005284204512997,
+      "grad_norm": 4.541245894934103,
+      "learning_rate": 2e-05,
+      "loss": 0.1643,
+      "step": 37833
+    },
+    {
+      "epoch": 0.9005522231743311,
+      "grad_norm": 5.41145817160176,
+      "learning_rate": 2e-05,
+      "loss": 1.195,
+      "step": 37834
+    },
+    {
+      "epoch": 0.9005760258973626,
+      "grad_norm": 7.692912346841025,
+      "learning_rate": 2e-05,
+      "loss": 0.24,
+      "step": 37835
+    },
+    {
+      "epoch": 0.9005998286203942,
+      "grad_norm": 3.5797238384868044,
+      "learning_rate": 2e-05,
+      "loss": 0.3431,
+      "step": 37836
+    },
+    {
+      "epoch": 0.9006236313434257,
+      "grad_norm": 2.5017411150768973,
+      "learning_rate": 2e-05,
+      "loss": 0.3663,
+      "step": 37837
+    },
+    {
+      "epoch": 0.9006474340664572,
+      "grad_norm": 8.41459327160969,
+      "learning_rate": 2e-05,
+      "loss": 0.7079,
+      "step": 37838
+    },
+    {
+      "epoch": 0.9006712367894887,
+      "grad_norm": 1.4954355249856892,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 37839
+    },
+    {
+      "epoch": 0.9006950395125203,
+      "grad_norm": 2.4218467455097477,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 37840
+    },
+    {
+      "epoch": 0.9007188422355518,
+      "grad_norm": 3.123222857557113,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 37841
+    },
+    {
+      "epoch": 0.9007426449585832,
+      "grad_norm": 1.781754615882068,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 37842
+    },
+    {
+      "epoch": 0.9007664476816147,
+      "grad_norm": 4.135887813932417,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 37843
+    },
+    {
+      "epoch": 0.9007902504046463,
+      "grad_norm": 0.13895929024928177,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 37844
+    },
+    {
+      "epoch": 0.9008140531276778,
+      "grad_norm": 4.097659044976629,
+      "learning_rate": 2e-05,
+      "loss": 0.4623,
+      "step": 37845
+    },
+    {
+      "epoch": 0.9008378558507093,
+      "grad_norm": 13.433892185417513,
+      "learning_rate": 2e-05,
+      "loss": 0.9867,
+      "step": 37846
+    },
+    {
+      "epoch": 0.9008616585737408,
+      "grad_norm": 0.9163863554253547,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 37847
+    },
+    {
+      "epoch": 0.9008854612967724,
+      "grad_norm": 0.04270538921907167,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 37848
+    },
+    {
+      "epoch": 0.9009092640198039,
+      "grad_norm": 0.1770605789251588,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 37849
+    },
+    {
+      "epoch": 0.9009330667428354,
+      "grad_norm": 5.8845514146399545,
+      "learning_rate": 2e-05,
+      "loss": 1.3116,
+      "step": 37850
+    },
+    {
+      "epoch": 0.900956869465867,
+      "grad_norm": 8.435343646981538,
+      "learning_rate": 2e-05,
+      "loss": 1.0099,
+      "step": 37851
+    },
+    {
+      "epoch": 0.9009806721888984,
+      "grad_norm": 2.9061682254333223,
+      "learning_rate": 2e-05,
+      "loss": 0.5278,
+      "step": 37852
+    },
+    {
+      "epoch": 0.9010044749119299,
+      "grad_norm": 4.777069298527444,
+      "learning_rate": 2e-05,
+      "loss": 0.652,
+      "step": 37853
+    },
+    {
+      "epoch": 0.9010282776349614,
+      "grad_norm": 33.69486326852075,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 37854
+    },
+    {
+      "epoch": 0.901052080357993,
+      "grad_norm": 0.07543910192573189,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 37855
+    },
+    {
+      "epoch": 0.9010758830810245,
+      "grad_norm": 7.855269104475531,
+      "learning_rate": 2e-05,
+      "loss": 0.3934,
+      "step": 37856
+    },
+    {
+      "epoch": 0.901099685804056,
+      "grad_norm": 3.796171712457554,
+      "learning_rate": 2e-05,
+      "loss": 0.4136,
+      "step": 37857
+    },
+    {
+      "epoch": 0.9011234885270875,
+      "grad_norm": 4.239801333300408,
+      "learning_rate": 2e-05,
+      "loss": 0.1506,
+      "step": 37858
+    },
+    {
+      "epoch": 0.901147291250119,
+      "grad_norm": 2.699011859199682,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 37859
+    },
+    {
+      "epoch": 0.9011710939731505,
+      "grad_norm": 0.2567875188363088,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 37860
+    },
+    {
+      "epoch": 0.901194896696182,
+      "grad_norm": 0.19091882802127544,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 37861
+    },
+    {
+      "epoch": 0.9012186994192135,
+      "grad_norm": 9.856742888803172,
+      "learning_rate": 2e-05,
+      "loss": 0.3533,
+      "step": 37862
+    },
+    {
+      "epoch": 0.9012425021422451,
+      "grad_norm": 14.734004859857201,
+      "learning_rate": 2e-05,
+      "loss": 1.2132,
+      "step": 37863
+    },
+    {
+      "epoch": 0.9012663048652766,
+      "grad_norm": 4.038795256875213,
+      "learning_rate": 2e-05,
+      "loss": 0.4146,
+      "step": 37864
+    },
+    {
+      "epoch": 0.9012901075883081,
+      "grad_norm": 5.7285960056698935,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 37865
+    },
+    {
+      "epoch": 0.9013139103113397,
+      "grad_norm": 8.985716261877235,
+      "learning_rate": 2e-05,
+      "loss": 1.4401,
+      "step": 37866
+    },
+    {
+      "epoch": 0.9013377130343712,
+      "grad_norm": 0.017970708520475862,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 37867
+    },
+    {
+      "epoch": 0.9013615157574026,
+      "grad_norm": 5.155596264512733,
+      "learning_rate": 2e-05,
+      "loss": 0.2079,
+      "step": 37868
+    },
+    {
+      "epoch": 0.9013853184804341,
+      "grad_norm": 0.05345251808930594,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 37869
+    },
+    {
+      "epoch": 0.9014091212034657,
+      "grad_norm": 0.23377361985659156,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 37870
+    },
+    {
+      "epoch": 0.9014329239264972,
+      "grad_norm": 3.722700299277904,
+      "learning_rate": 2e-05,
+      "loss": 0.5976,
+      "step": 37871
+    },
+    {
+      "epoch": 0.9014567266495287,
+      "grad_norm": 4.293114583251024,
+      "learning_rate": 2e-05,
+      "loss": 0.7049,
+      "step": 37872
+    },
+    {
+      "epoch": 0.9014805293725602,
+      "grad_norm": 0.5116798234511618,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 37873
+    },
+    {
+      "epoch": 0.9015043320955918,
+      "grad_norm": 1.2039429733363953,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 37874
+    },
+    {
+      "epoch": 0.9015281348186233,
+      "grad_norm": 8.8451649761625,
+      "learning_rate": 2e-05,
+      "loss": 0.8029,
+      "step": 37875
+    },
+    {
+      "epoch": 0.9015519375416547,
+      "grad_norm": 0.5500756768453864,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 37876
+    },
+    {
+      "epoch": 0.9015757402646862,
+      "grad_norm": 2.729474679251901,
+      "learning_rate": 2e-05,
+      "loss": 0.0895,
+      "step": 37877
+    },
+    {
+      "epoch": 0.9015995429877178,
+      "grad_norm": 5.538701348479474,
+      "learning_rate": 2e-05,
+      "loss": 0.8609,
+      "step": 37878
+    },
+    {
+      "epoch": 0.9016233457107493,
+      "grad_norm": 6.710400766222352,
+      "learning_rate": 2e-05,
+      "loss": 0.2709,
+      "step": 37879
+    },
+    {
+      "epoch": 0.9016471484337808,
+      "grad_norm": 4.962470136099526,
+      "learning_rate": 2e-05,
+      "loss": 1.2731,
+      "step": 37880
+    },
+    {
+      "epoch": 0.9016709511568124,
+      "grad_norm": 4.614188952627054,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 37881
+    },
+    {
+      "epoch": 0.9016947538798439,
+      "grad_norm": 4.427150946153108,
+      "learning_rate": 2e-05,
+      "loss": 0.1502,
+      "step": 37882
+    },
+    {
+      "epoch": 0.9017185566028754,
+      "grad_norm": 6.997115694742614,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 37883
+    },
+    {
+      "epoch": 0.9017423593259069,
+      "grad_norm": 14.288942384213117,
+      "learning_rate": 2e-05,
+      "loss": 0.5393,
+      "step": 37884
+    },
+    {
+      "epoch": 0.9017661620489384,
+      "grad_norm": 9.12835341446536,
+      "learning_rate": 2e-05,
+      "loss": 0.283,
+      "step": 37885
+    },
+    {
+      "epoch": 0.9017899647719699,
+      "grad_norm": 2.5462858323817517,
+      "learning_rate": 2e-05,
+      "loss": 0.4923,
+      "step": 37886
+    },
+    {
+      "epoch": 0.9018137674950014,
+      "grad_norm": 6.791870366939372,
+      "learning_rate": 2e-05,
+      "loss": 1.8041,
+      "step": 37887
+    },
+    {
+      "epoch": 0.9018375702180329,
+      "grad_norm": 3.1671752778685303,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 37888
+    },
+    {
+      "epoch": 0.9018613729410645,
+      "grad_norm": 1.2000387629050493,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 37889
+    },
+    {
+      "epoch": 0.901885175664096,
+      "grad_norm": 2.5191788736057763,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 37890
+    },
+    {
+      "epoch": 0.9019089783871275,
+      "grad_norm": 2.4326641113132377,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 37891
+    },
+    {
+      "epoch": 0.901932781110159,
+      "grad_norm": 4.3259598550033855,
+      "learning_rate": 2e-05,
+      "loss": 0.1708,
+      "step": 37892
+    },
+    {
+      "epoch": 0.9019565838331906,
+      "grad_norm": 8.886416257147083,
+      "learning_rate": 2e-05,
+      "loss": 0.4996,
+      "step": 37893
+    },
+    {
+      "epoch": 0.901980386556222,
+      "grad_norm": 0.3534881750083058,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 37894
+    },
+    {
+      "epoch": 0.9020041892792535,
+      "grad_norm": 1.9852487756822392,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 37895
+    },
+    {
+      "epoch": 0.902027992002285,
+      "grad_norm": 0.10796226859232781,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37896
+    },
+    {
+      "epoch": 0.9020517947253166,
+      "grad_norm": 0.11799914959066554,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 37897
+    },
+    {
+      "epoch": 0.9020755974483481,
+      "grad_norm": 2.3480880210703616,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 37898
+    },
+    {
+      "epoch": 0.9020994001713796,
+      "grad_norm": 3.9028585826215227,
+      "learning_rate": 2e-05,
+      "loss": 0.4587,
+      "step": 37899
+    },
+    {
+      "epoch": 0.9021232028944112,
+      "grad_norm": 1.5739622669147757,
+      "learning_rate": 2e-05,
+      "loss": 0.0327,
+      "step": 37900
+    },
+    {
+      "epoch": 0.9021470056174427,
+      "grad_norm": 0.08186887807092814,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 37901
+    },
+    {
+      "epoch": 0.9021708083404741,
+      "grad_norm": 6.027642920470178,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 37902
+    },
+    {
+      "epoch": 0.9021946110635056,
+      "grad_norm": 0.9528855075587255,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 37903
+    },
+    {
+      "epoch": 0.9022184137865372,
+      "grad_norm": 0.3444565485673718,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 37904
+    },
+    {
+      "epoch": 0.9022422165095687,
+      "grad_norm": 1.989014027420446,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 37905
+    },
+    {
+      "epoch": 0.9022660192326002,
+      "grad_norm": 3.4229512481461555,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 37906
+    },
+    {
+      "epoch": 0.9022898219556317,
+      "grad_norm": 3.797703604598774,
+      "learning_rate": 2e-05,
+      "loss": 0.1109,
+      "step": 37907
+    },
+    {
+      "epoch": 0.9023136246786633,
+      "grad_norm": 3.1236463993782215,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 37908
+    },
+    {
+      "epoch": 0.9023374274016948,
+      "grad_norm": 2.764221035119799,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 37909
+    },
+    {
+      "epoch": 0.9023612301247262,
+      "grad_norm": 13.502276282909007,
+      "learning_rate": 2e-05,
+      "loss": 1.0755,
+      "step": 37910
+    },
+    {
+      "epoch": 0.9023850328477577,
+      "grad_norm": 2.0677688378089223,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 37911
+    },
+    {
+      "epoch": 0.9024088355707893,
+      "grad_norm": 4.299240347216963,
+      "learning_rate": 2e-05,
+      "loss": 0.084,
+      "step": 37912
+    },
+    {
+      "epoch": 0.9024326382938208,
+      "grad_norm": 3.6620873559423726,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 37913
+    },
+    {
+      "epoch": 0.9024564410168523,
+      "grad_norm": 4.864032619059551,
+      "learning_rate": 2e-05,
+      "loss": 0.4762,
+      "step": 37914
+    },
+    {
+      "epoch": 0.9024802437398839,
+      "grad_norm": 0.4488995340757437,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 37915
+    },
+    {
+      "epoch": 0.9025040464629154,
+      "grad_norm": 3.6837706829276056,
+      "learning_rate": 2e-05,
+      "loss": 0.6605,
+      "step": 37916
+    },
+    {
+      "epoch": 0.9025278491859469,
+      "grad_norm": 4.748824875301211,
+      "learning_rate": 2e-05,
+      "loss": 1.148,
+      "step": 37917
+    },
+    {
+      "epoch": 0.9025516519089783,
+      "grad_norm": 4.226185001074465,
+      "learning_rate": 2e-05,
+      "loss": 1.025,
+      "step": 37918
+    },
+    {
+      "epoch": 0.9025754546320099,
+      "grad_norm": 0.22585001384413841,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 37919
+    },
+    {
+      "epoch": 0.9025992573550414,
+      "grad_norm": 5.922230750512294,
+      "learning_rate": 2e-05,
+      "loss": 0.6197,
+      "step": 37920
+    },
+    {
+      "epoch": 0.9026230600780729,
+      "grad_norm": 1.593189752570819,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 37921
+    },
+    {
+      "epoch": 0.9026468628011044,
+      "grad_norm": 0.23409736360753616,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 37922
+    },
+    {
+      "epoch": 0.902670665524136,
+      "grad_norm": 1.3899382159156302,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 37923
+    },
+    {
+      "epoch": 0.9026944682471675,
+      "grad_norm": 3.727520160882272,
+      "learning_rate": 2e-05,
+      "loss": 0.5388,
+      "step": 37924
+    },
+    {
+      "epoch": 0.902718270970199,
+      "grad_norm": 0.14068777855571915,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 37925
+    },
+    {
+      "epoch": 0.9027420736932305,
+      "grad_norm": 0.37404977464328687,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 37926
+    },
+    {
+      "epoch": 0.902765876416262,
+      "grad_norm": 3.568261914913301,
+      "learning_rate": 2e-05,
+      "loss": 0.9534,
+      "step": 37927
+    },
+    {
+      "epoch": 0.9027896791392935,
+      "grad_norm": 6.012907547483125,
+      "learning_rate": 2e-05,
+      "loss": 0.7354,
+      "step": 37928
+    },
+    {
+      "epoch": 0.902813481862325,
+      "grad_norm": 1.6120299090056913,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 37929
+    },
+    {
+      "epoch": 0.9028372845853566,
+      "grad_norm": 4.198195710897765,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 37930
+    },
+    {
+      "epoch": 0.9028610873083881,
+      "grad_norm": 5.630359580760932,
+      "learning_rate": 2e-05,
+      "loss": 0.4513,
+      "step": 37931
+    },
+    {
+      "epoch": 0.9028848900314196,
+      "grad_norm": 0.5753362368150591,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 37932
+    },
+    {
+      "epoch": 0.9029086927544511,
+      "grad_norm": 0.07928380813502853,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 37933
+    },
+    {
+      "epoch": 0.9029324954774827,
+      "grad_norm": 5.95401271981178,
+      "learning_rate": 2e-05,
+      "loss": 0.8226,
+      "step": 37934
+    },
+    {
+      "epoch": 0.9029562982005142,
+      "grad_norm": 4.102440895113832,
+      "learning_rate": 2e-05,
+      "loss": 0.5809,
+      "step": 37935
+    },
+    {
+      "epoch": 0.9029801009235456,
+      "grad_norm": 2.156454076668672,
+      "learning_rate": 2e-05,
+      "loss": 0.3154,
+      "step": 37936
+    },
+    {
+      "epoch": 0.9030039036465771,
+      "grad_norm": 10.400478275189311,
+      "learning_rate": 2e-05,
+      "loss": 0.818,
+      "step": 37937
+    },
+    {
+      "epoch": 0.9030277063696087,
+      "grad_norm": 1.4269796576066693,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 37938
+    },
+    {
+      "epoch": 0.9030515090926402,
+      "grad_norm": 0.059522589170510495,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 37939
+    },
+    {
+      "epoch": 0.9030753118156717,
+      "grad_norm": 0.40925046041150115,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 37940
+    },
+    {
+      "epoch": 0.9030991145387032,
+      "grad_norm": 1.902599619169768,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 37941
+    },
+    {
+      "epoch": 0.9031229172617348,
+      "grad_norm": 9.196029449111325,
+      "learning_rate": 2e-05,
+      "loss": 0.8853,
+      "step": 37942
+    },
+    {
+      "epoch": 0.9031467199847663,
+      "grad_norm": 4.675303260602535,
+      "learning_rate": 2e-05,
+      "loss": 0.5137,
+      "step": 37943
+    },
+    {
+      "epoch": 0.9031705227077977,
+      "grad_norm": 2.8322169889121116,
+      "learning_rate": 2e-05,
+      "loss": 0.6447,
+      "step": 37944
+    },
+    {
+      "epoch": 0.9031943254308293,
+      "grad_norm": 4.590221727057565,
+      "learning_rate": 2e-05,
+      "loss": 0.2429,
+      "step": 37945
+    },
+    {
+      "epoch": 0.9032181281538608,
+      "grad_norm": 4.238026075591932,
+      "learning_rate": 2e-05,
+      "loss": 0.7117,
+      "step": 37946
+    },
+    {
+      "epoch": 0.9032419308768923,
+      "grad_norm": 0.4634760198007301,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 37947
+    },
+    {
+      "epoch": 0.9032657335999238,
+      "grad_norm": 0.14590965479315923,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 37948
+    },
+    {
+      "epoch": 0.9032895363229554,
+      "grad_norm": 5.329842080697104,
+      "learning_rate": 2e-05,
+      "loss": 0.4485,
+      "step": 37949
+    },
+    {
+      "epoch": 0.9033133390459869,
+      "grad_norm": 0.39040648042118775,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 37950
+    },
+    {
+      "epoch": 0.9033371417690184,
+      "grad_norm": 5.539999260899076,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 37951
+    },
+    {
+      "epoch": 0.9033609444920498,
+      "grad_norm": 3.0141924185866924,
+      "learning_rate": 2e-05,
+      "loss": 0.4964,
+      "step": 37952
+    },
+    {
+      "epoch": 0.9033847472150814,
+      "grad_norm": 1.8725921509862697,
+      "learning_rate": 2e-05,
+      "loss": 0.0331,
+      "step": 37953
+    },
+    {
+      "epoch": 0.9034085499381129,
+      "grad_norm": 9.178973879976134,
+      "learning_rate": 2e-05,
+      "loss": 0.7137,
+      "step": 37954
+    },
+    {
+      "epoch": 0.9034323526611444,
+      "grad_norm": 3.057643505121602,
+      "learning_rate": 2e-05,
+      "loss": 0.5009,
+      "step": 37955
+    },
+    {
+      "epoch": 0.9034561553841759,
+      "grad_norm": 4.436681827124827,
+      "learning_rate": 2e-05,
+      "loss": 0.4035,
+      "step": 37956
+    },
+    {
+      "epoch": 0.9034799581072075,
+      "grad_norm": 0.11723704604859661,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 37957
+    },
+    {
+      "epoch": 0.903503760830239,
+      "grad_norm": 4.4586238164383225,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 37958
+    },
+    {
+      "epoch": 0.9035275635532705,
+      "grad_norm": 3.140146201335026,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 37959
+    },
+    {
+      "epoch": 0.9035513662763021,
+      "grad_norm": 4.675875627699895,
+      "learning_rate": 2e-05,
+      "loss": 1.2175,
+      "step": 37960
+    },
+    {
+      "epoch": 0.9035751689993335,
+      "grad_norm": 2.578356594325981,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 37961
+    },
+    {
+      "epoch": 0.903598971722365,
+      "grad_norm": 0.10370319211030832,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 37962
+    },
+    {
+      "epoch": 0.9036227744453965,
+      "grad_norm": 0.1422647844323455,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 37963
+    },
+    {
+      "epoch": 0.9036465771684281,
+      "grad_norm": 0.009588879650247913,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 37964
+    },
+    {
+      "epoch": 0.9036703798914596,
+      "grad_norm": 2.0810156155285906,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 37965
+    },
+    {
+      "epoch": 0.9036941826144911,
+      "grad_norm": 2.49000643524093,
+      "learning_rate": 2e-05,
+      "loss": 0.5485,
+      "step": 37966
+    },
+    {
+      "epoch": 0.9037179853375226,
+      "grad_norm": 10.952355982745466,
+      "learning_rate": 2e-05,
+      "loss": 0.4988,
+      "step": 37967
+    },
+    {
+      "epoch": 0.9037417880605542,
+      "grad_norm": 0.8347573368238497,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 37968
+    },
+    {
+      "epoch": 0.9037655907835856,
+      "grad_norm": 0.298300437062216,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 37969
+    },
+    {
+      "epoch": 0.9037893935066171,
+      "grad_norm": 0.03587354104032487,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 37970
+    },
+    {
+      "epoch": 0.9038131962296486,
+      "grad_norm": 17.686851491924486,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 37971
+    },
+    {
+      "epoch": 0.9038369989526802,
+      "grad_norm": 0.136194993978469,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 37972
+    },
+    {
+      "epoch": 0.9038608016757117,
+      "grad_norm": 4.846029594662246,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 37973
+    },
+    {
+      "epoch": 0.9038846043987432,
+      "grad_norm": 5.5572317292427735,
+      "learning_rate": 2e-05,
+      "loss": 1.377,
+      "step": 37974
+    },
+    {
+      "epoch": 0.9039084071217748,
+      "grad_norm": 23.10677068366011,
+      "learning_rate": 2e-05,
+      "loss": 1.8735,
+      "step": 37975
+    },
+    {
+      "epoch": 0.9039322098448063,
+      "grad_norm": 3.8129674483695135,
+      "learning_rate": 2e-05,
+      "loss": 0.8151,
+      "step": 37976
+    },
+    {
+      "epoch": 0.9039560125678378,
+      "grad_norm": 3.4672891082812534,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 37977
+    },
+    {
+      "epoch": 0.9039798152908692,
+      "grad_norm": 1.504431315384824,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 37978
+    },
+    {
+      "epoch": 0.9040036180139008,
+      "grad_norm": 0.4765040988978228,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 37979
+    },
+    {
+      "epoch": 0.9040274207369323,
+      "grad_norm": 1.9719230660308908,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 37980
+    },
+    {
+      "epoch": 0.9040512234599638,
+      "grad_norm": 0.7236441277981983,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 37981
+    },
+    {
+      "epoch": 0.9040750261829953,
+      "grad_norm": 6.419336110235665,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 37982
+    },
+    {
+      "epoch": 0.9040988289060269,
+      "grad_norm": 6.215842194720268,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 37983
+    },
+    {
+      "epoch": 0.9041226316290584,
+      "grad_norm": 1.8360955739733449,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 37984
+    },
+    {
+      "epoch": 0.9041464343520899,
+      "grad_norm": 3.6456566163715918,
+      "learning_rate": 2e-05,
+      "loss": 0.45,
+      "step": 37985
+    },
+    {
+      "epoch": 0.9041702370751213,
+      "grad_norm": 4.979107286671004,
+      "learning_rate": 2e-05,
+      "loss": 0.0677,
+      "step": 37986
+    },
+    {
+      "epoch": 0.9041940397981529,
+      "grad_norm": 7.094499730566177,
+      "learning_rate": 2e-05,
+      "loss": 1.1469,
+      "step": 37987
+    },
+    {
+      "epoch": 0.9042178425211844,
+      "grad_norm": 0.6049790867220758,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37988
+    },
+    {
+      "epoch": 0.9042416452442159,
+      "grad_norm": 0.09963459782091576,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 37989
+    },
+    {
+      "epoch": 0.9042654479672475,
+      "grad_norm": 5.615146873435039,
+      "learning_rate": 2e-05,
+      "loss": 1.2525,
+      "step": 37990
+    },
+    {
+      "epoch": 0.904289250690279,
+      "grad_norm": 0.23527818667058142,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 37991
+    },
+    {
+      "epoch": 0.9043130534133105,
+      "grad_norm": 0.957266026386806,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 37992
+    },
+    {
+      "epoch": 0.904336856136342,
+      "grad_norm": 0.1262304902022657,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 37993
+    },
+    {
+      "epoch": 0.9043606588593736,
+      "grad_norm": 4.344517862181419,
+      "learning_rate": 2e-05,
+      "loss": 0.4467,
+      "step": 37994
+    },
+    {
+      "epoch": 0.904384461582405,
+      "grad_norm": 9.245906096926086,
+      "learning_rate": 2e-05,
+      "loss": 0.8802,
+      "step": 37995
+    },
+    {
+      "epoch": 0.9044082643054365,
+      "grad_norm": 3.4880403792314363,
+      "learning_rate": 2e-05,
+      "loss": 0.5986,
+      "step": 37996
+    },
+    {
+      "epoch": 0.904432067028468,
+      "grad_norm": 3.077606320836611,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 37997
+    },
+    {
+      "epoch": 0.9044558697514996,
+      "grad_norm": 3.883575563858699,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 37998
+    },
+    {
+      "epoch": 0.9044796724745311,
+      "grad_norm": 1.9873520908340812,
+      "learning_rate": 2e-05,
+      "loss": 0.0525,
+      "step": 37999
+    },
+    {
+      "epoch": 0.9045034751975626,
+      "grad_norm": 9.208603780667122,
+      "learning_rate": 2e-05,
+      "loss": 0.212,
+      "step": 38000
+    },
+    {
+      "epoch": 0.9045272779205941,
+      "grad_norm": 4.198171305551776,
+      "learning_rate": 2e-05,
+      "loss": 0.3798,
+      "step": 38001
+    },
+    {
+      "epoch": 0.9045510806436257,
+      "grad_norm": 1.9037787191203945,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 38002
+    },
+    {
+      "epoch": 0.9045748833666571,
+      "grad_norm": 5.323327260652918,
+      "learning_rate": 2e-05,
+      "loss": 0.7656,
+      "step": 38003
+    },
+    {
+      "epoch": 0.9045986860896886,
+      "grad_norm": 8.690616527268759,
+      "learning_rate": 2e-05,
+      "loss": 0.1761,
+      "step": 38004
+    },
+    {
+      "epoch": 0.9046224888127202,
+      "grad_norm": 3.8990325017208205,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 38005
+    },
+    {
+      "epoch": 0.9046462915357517,
+      "grad_norm": 2.531417628789335,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 38006
+    },
+    {
+      "epoch": 0.9046700942587832,
+      "grad_norm": 6.154462434941215,
+      "learning_rate": 2e-05,
+      "loss": 0.9958,
+      "step": 38007
+    },
+    {
+      "epoch": 0.9046938969818147,
+      "grad_norm": 4.1615213587645385,
+      "learning_rate": 2e-05,
+      "loss": 0.8803,
+      "step": 38008
+    },
+    {
+      "epoch": 0.9047176997048463,
+      "grad_norm": 2.8284762993186483,
+      "learning_rate": 2e-05,
+      "loss": 0.4408,
+      "step": 38009
+    },
+    {
+      "epoch": 0.9047415024278778,
+      "grad_norm": 4.57229348209127,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 38010
+    },
+    {
+      "epoch": 0.9047653051509092,
+      "grad_norm": 8.947570886111501,
+      "learning_rate": 2e-05,
+      "loss": 1.1493,
+      "step": 38011
+    },
+    {
+      "epoch": 0.9047891078739407,
+      "grad_norm": 0.17531440346487367,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 38012
+    },
+    {
+      "epoch": 0.9048129105969723,
+      "grad_norm": 0.032580163592743794,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 38013
+    },
+    {
+      "epoch": 0.9048367133200038,
+      "grad_norm": 3.052442055772779,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 38014
+    },
+    {
+      "epoch": 0.9048605160430353,
+      "grad_norm": 6.693252563059113,
+      "learning_rate": 2e-05,
+      "loss": 1.4781,
+      "step": 38015
+    },
+    {
+      "epoch": 0.9048843187660668,
+      "grad_norm": 7.304409276696658,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 38016
+    },
+    {
+      "epoch": 0.9049081214890984,
+      "grad_norm": 2.2264597003941997,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 38017
+    },
+    {
+      "epoch": 0.9049319242121299,
+      "grad_norm": 6.76482620791067,
+      "learning_rate": 2e-05,
+      "loss": 0.7327,
+      "step": 38018
+    },
+    {
+      "epoch": 0.9049557269351614,
+      "grad_norm": 2.997419074149043,
+      "learning_rate": 2e-05,
+      "loss": 0.5263,
+      "step": 38019
+    },
+    {
+      "epoch": 0.904979529658193,
+      "grad_norm": 0.14931989387667124,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 38020
+    },
+    {
+      "epoch": 0.9050033323812244,
+      "grad_norm": 3.815387192499021,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 38021
+    },
+    {
+      "epoch": 0.9050271351042559,
+      "grad_norm": 3.355369766861274,
+      "learning_rate": 2e-05,
+      "loss": 0.5791,
+      "step": 38022
+    },
+    {
+      "epoch": 0.9050509378272874,
+      "grad_norm": 5.398318392453177,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 38023
+    },
+    {
+      "epoch": 0.905074740550319,
+      "grad_norm": 2.134400334632587,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 38024
+    },
+    {
+      "epoch": 0.9050985432733505,
+      "grad_norm": 12.464541306468057,
+      "learning_rate": 2e-05,
+      "loss": 0.2532,
+      "step": 38025
+    },
+    {
+      "epoch": 0.905122345996382,
+      "grad_norm": 4.564681481306898,
+      "learning_rate": 2e-05,
+      "loss": 1.5103,
+      "step": 38026
+    },
+    {
+      "epoch": 0.9051461487194135,
+      "grad_norm": 5.421266248701099,
+      "learning_rate": 2e-05,
+      "loss": 0.9622,
+      "step": 38027
+    },
+    {
+      "epoch": 0.905169951442445,
+      "grad_norm": 6.076767156441265,
+      "learning_rate": 2e-05,
+      "loss": 0.7441,
+      "step": 38028
+    },
+    {
+      "epoch": 0.9051937541654765,
+      "grad_norm": 20.511560469382196,
+      "learning_rate": 2e-05,
+      "loss": 0.4507,
+      "step": 38029
+    },
+    {
+      "epoch": 0.905217556888508,
+      "grad_norm": 1.6750876517519946,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 38030
+    },
+    {
+      "epoch": 0.9052413596115395,
+      "grad_norm": 4.522573778997284,
+      "learning_rate": 2e-05,
+      "loss": 0.8876,
+      "step": 38031
+    },
+    {
+      "epoch": 0.9052651623345711,
+      "grad_norm": 0.07046317812752219,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38032
+    },
+    {
+      "epoch": 0.9052889650576026,
+      "grad_norm": 3.1314416038910737,
+      "learning_rate": 2e-05,
+      "loss": 0.4104,
+      "step": 38033
+    },
+    {
+      "epoch": 0.9053127677806341,
+      "grad_norm": 2.022107437898214,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 38034
+    },
+    {
+      "epoch": 0.9053365705036657,
+      "grad_norm": 4.449914089662824,
+      "learning_rate": 2e-05,
+      "loss": 0.7211,
+      "step": 38035
+    },
+    {
+      "epoch": 0.9053603732266972,
+      "grad_norm": 0.04144181014357803,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 38036
+    },
+    {
+      "epoch": 0.9053841759497286,
+      "grad_norm": 3.6252010310121445,
+      "learning_rate": 2e-05,
+      "loss": 0.739,
+      "step": 38037
+    },
+    {
+      "epoch": 0.9054079786727601,
+      "grad_norm": 5.102854840637295,
+      "learning_rate": 2e-05,
+      "loss": 0.8899,
+      "step": 38038
+    },
+    {
+      "epoch": 0.9054317813957917,
+      "grad_norm": 0.3642951443098703,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 38039
+    },
+    {
+      "epoch": 0.9054555841188232,
+      "grad_norm": 5.135808578652914,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 38040
+    },
+    {
+      "epoch": 0.9054793868418547,
+      "grad_norm": 0.2390789677352696,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 38041
+    },
+    {
+      "epoch": 0.9055031895648862,
+      "grad_norm": 9.149996591885227,
+      "learning_rate": 2e-05,
+      "loss": 0.44,
+      "step": 38042
+    },
+    {
+      "epoch": 0.9055269922879178,
+      "grad_norm": 12.874965075129962,
+      "learning_rate": 2e-05,
+      "loss": 0.5293,
+      "step": 38043
+    },
+    {
+      "epoch": 0.9055507950109493,
+      "grad_norm": 13.524716568547314,
+      "learning_rate": 2e-05,
+      "loss": 0.9844,
+      "step": 38044
+    },
+    {
+      "epoch": 0.9055745977339807,
+      "grad_norm": 0.13268385168373423,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 38045
+    },
+    {
+      "epoch": 0.9055984004570122,
+      "grad_norm": 4.8712477868466495,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 38046
+    },
+    {
+      "epoch": 0.9056222031800438,
+      "grad_norm": 4.216267433623901,
+      "learning_rate": 2e-05,
+      "loss": 0.4133,
+      "step": 38047
+    },
+    {
+      "epoch": 0.9056460059030753,
+      "grad_norm": 0.9039917377920587,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 38048
+    },
+    {
+      "epoch": 0.9056698086261068,
+      "grad_norm": 2.6123609666745105,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 38049
+    },
+    {
+      "epoch": 0.9056936113491384,
+      "grad_norm": 9.940159618194809,
+      "learning_rate": 2e-05,
+      "loss": 0.6615,
+      "step": 38050
+    },
+    {
+      "epoch": 0.9057174140721699,
+      "grad_norm": 8.266904770114206,
+      "learning_rate": 2e-05,
+      "loss": 0.3693,
+      "step": 38051
+    },
+    {
+      "epoch": 0.9057412167952014,
+      "grad_norm": 1.352145896974257,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 38052
+    },
+    {
+      "epoch": 0.9057650195182329,
+      "grad_norm": 3.84524957397497,
+      "learning_rate": 2e-05,
+      "loss": 0.1031,
+      "step": 38053
+    },
+    {
+      "epoch": 0.9057888222412644,
+      "grad_norm": 3.477879573333653,
+      "learning_rate": 2e-05,
+      "loss": 0.652,
+      "step": 38054
+    },
+    {
+      "epoch": 0.9058126249642959,
+      "grad_norm": 0.06344417710237522,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38055
+    },
+    {
+      "epoch": 0.9058364276873274,
+      "grad_norm": 5.971922167451533,
+      "learning_rate": 2e-05,
+      "loss": 0.9656,
+      "step": 38056
+    },
+    {
+      "epoch": 0.9058602304103589,
+      "grad_norm": 5.008411442140898,
+      "learning_rate": 2e-05,
+      "loss": 0.3516,
+      "step": 38057
+    },
+    {
+      "epoch": 0.9058840331333905,
+      "grad_norm": 9.596649400766328,
+      "learning_rate": 2e-05,
+      "loss": 0.1709,
+      "step": 38058
+    },
+    {
+      "epoch": 0.905907835856422,
+      "grad_norm": 10.844504346735333,
+      "learning_rate": 2e-05,
+      "loss": 0.961,
+      "step": 38059
+    },
+    {
+      "epoch": 0.9059316385794535,
+      "grad_norm": 3.713982344919742,
+      "learning_rate": 2e-05,
+      "loss": 0.4343,
+      "step": 38060
+    },
+    {
+      "epoch": 0.905955441302485,
+      "grad_norm": 2.607357055702117,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 38061
+    },
+    {
+      "epoch": 0.9059792440255166,
+      "grad_norm": 2.149459313658788,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 38062
+    },
+    {
+      "epoch": 0.906003046748548,
+      "grad_norm": 5.369583745663762,
+      "learning_rate": 2e-05,
+      "loss": 0.7274,
+      "step": 38063
+    },
+    {
+      "epoch": 0.9060268494715795,
+      "grad_norm": 6.789184033109503,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 38064
+    },
+    {
+      "epoch": 0.9060506521946111,
+      "grad_norm": 9.329094467858994,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 38065
+    },
+    {
+      "epoch": 0.9060744549176426,
+      "grad_norm": 0.020233552414571757,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 38066
+    },
+    {
+      "epoch": 0.9060982576406741,
+      "grad_norm": 2.5340129003641305,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 38067
+    },
+    {
+      "epoch": 0.9061220603637056,
+      "grad_norm": 1.9636170312979904,
+      "learning_rate": 2e-05,
+      "loss": 0.0632,
+      "step": 38068
+    },
+    {
+      "epoch": 0.9061458630867372,
+      "grad_norm": 1.2761962282770702,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 38069
+    },
+    {
+      "epoch": 0.9061696658097687,
+      "grad_norm": 0.07849482995214839,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 38070
+    },
+    {
+      "epoch": 0.9061934685328001,
+      "grad_norm": 0.3302967531529929,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 38071
+    },
+    {
+      "epoch": 0.9062172712558316,
+      "grad_norm": 4.17502383152305,
+      "learning_rate": 2e-05,
+      "loss": 0.9299,
+      "step": 38072
+    },
+    {
+      "epoch": 0.9062410739788632,
+      "grad_norm": 3.7478856577575463,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 38073
+    },
+    {
+      "epoch": 0.9062648767018947,
+      "grad_norm": 3.3558259698843314,
+      "learning_rate": 2e-05,
+      "loss": 0.5252,
+      "step": 38074
+    },
+    {
+      "epoch": 0.9062886794249262,
+      "grad_norm": 1.77868850606156,
+      "learning_rate": 2e-05,
+      "loss": 0.1446,
+      "step": 38075
+    },
+    {
+      "epoch": 0.9063124821479577,
+      "grad_norm": 4.719008527488954,
+      "learning_rate": 2e-05,
+      "loss": 0.7415,
+      "step": 38076
+    },
+    {
+      "epoch": 0.9063362848709893,
+      "grad_norm": 3.337438313857689,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 38077
+    },
+    {
+      "epoch": 0.9063600875940208,
+      "grad_norm": 2.100599813794568,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 38078
+    },
+    {
+      "epoch": 0.9063838903170522,
+      "grad_norm": 2.199596813991975,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 38079
+    },
+    {
+      "epoch": 0.9064076930400838,
+      "grad_norm": 6.270600979924932,
+      "learning_rate": 2e-05,
+      "loss": 1.0267,
+      "step": 38080
+    },
+    {
+      "epoch": 0.9064314957631153,
+      "grad_norm": 9.819932394094852,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 38081
+    },
+    {
+      "epoch": 0.9064552984861468,
+      "grad_norm": 2.196817444150774,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 38082
+    },
+    {
+      "epoch": 0.9064791012091783,
+      "grad_norm": 0.5503796049718322,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 38083
+    },
+    {
+      "epoch": 0.9065029039322099,
+      "grad_norm": 5.410288965093083,
+      "learning_rate": 2e-05,
+      "loss": 1.1689,
+      "step": 38084
+    },
+    {
+      "epoch": 0.9065267066552414,
+      "grad_norm": 9.491700009524296,
+      "learning_rate": 2e-05,
+      "loss": 0.4659,
+      "step": 38085
+    },
+    {
+      "epoch": 0.9065505093782729,
+      "grad_norm": 8.315310483284723,
+      "learning_rate": 2e-05,
+      "loss": 0.8253,
+      "step": 38086
+    },
+    {
+      "epoch": 0.9065743121013043,
+      "grad_norm": 0.09451786921078853,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38087
+    },
+    {
+      "epoch": 0.9065981148243359,
+      "grad_norm": 2.2562532731200244,
+      "learning_rate": 2e-05,
+      "loss": 0.2251,
+      "step": 38088
+    },
+    {
+      "epoch": 0.9066219175473674,
+      "grad_norm": 2.2057274075423994,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 38089
+    },
+    {
+      "epoch": 0.9066457202703989,
+      "grad_norm": 2.3705085599392004,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 38090
+    },
+    {
+      "epoch": 0.9066695229934304,
+      "grad_norm": 5.427170037694908,
+      "learning_rate": 2e-05,
+      "loss": 0.7484,
+      "step": 38091
+    },
+    {
+      "epoch": 0.906693325716462,
+      "grad_norm": 3.983735101547132,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 38092
+    },
+    {
+      "epoch": 0.9067171284394935,
+      "grad_norm": 5.95009127348779,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 38093
+    },
+    {
+      "epoch": 0.906740931162525,
+      "grad_norm": 4.016687200286175,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 38094
+    },
+    {
+      "epoch": 0.9067647338855566,
+      "grad_norm": 13.904989220466788,
+      "learning_rate": 2e-05,
+      "loss": 0.4331,
+      "step": 38095
+    },
+    {
+      "epoch": 0.906788536608588,
+      "grad_norm": 3.558415703158032,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 38096
+    },
+    {
+      "epoch": 0.9068123393316195,
+      "grad_norm": 2.0408466366857687,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 38097
+    },
+    {
+      "epoch": 0.906836142054651,
+      "grad_norm": 3.257906650286898,
+      "learning_rate": 2e-05,
+      "loss": 0.5627,
+      "step": 38098
+    },
+    {
+      "epoch": 0.9068599447776826,
+      "grad_norm": 2.573389253733202,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 38099
+    },
+    {
+      "epoch": 0.9068837475007141,
+      "grad_norm": 4.8979625048712006,
+      "learning_rate": 2e-05,
+      "loss": 0.9054,
+      "step": 38100
+    },
+    {
+      "epoch": 0.9069075502237456,
+      "grad_norm": 0.7686102163785901,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 38101
+    },
+    {
+      "epoch": 0.9069313529467771,
+      "grad_norm": 2.0116275393316805,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 38102
+    },
+    {
+      "epoch": 0.9069551556698087,
+      "grad_norm": 1.5347585173358862,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 38103
+    },
+    {
+      "epoch": 0.9069789583928402,
+      "grad_norm": 0.8189670014321601,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 38104
+    },
+    {
+      "epoch": 0.9070027611158716,
+      "grad_norm": 3.388744403595248,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 38105
+    },
+    {
+      "epoch": 0.9070265638389031,
+      "grad_norm": 4.69757159622613,
+      "learning_rate": 2e-05,
+      "loss": 0.5985,
+      "step": 38106
+    },
+    {
+      "epoch": 0.9070503665619347,
+      "grad_norm": 0.23176575408275174,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 38107
+    },
+    {
+      "epoch": 0.9070741692849662,
+      "grad_norm": 0.34837842216260373,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 38108
+    },
+    {
+      "epoch": 0.9070979720079977,
+      "grad_norm": 3.6988820174619716,
+      "learning_rate": 2e-05,
+      "loss": 0.1183,
+      "step": 38109
+    },
+    {
+      "epoch": 0.9071217747310293,
+      "grad_norm": 6.163776630640143,
+      "learning_rate": 2e-05,
+      "loss": 0.776,
+      "step": 38110
+    },
+    {
+      "epoch": 0.9071455774540608,
+      "grad_norm": 0.5753316108072053,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 38111
+    },
+    {
+      "epoch": 0.9071693801770923,
+      "grad_norm": 2.6064456730411143,
+      "learning_rate": 2e-05,
+      "loss": 0.414,
+      "step": 38112
+    },
+    {
+      "epoch": 0.9071931829001237,
+      "grad_norm": 3.922524045739443,
+      "learning_rate": 2e-05,
+      "loss": 0.5522,
+      "step": 38113
+    },
+    {
+      "epoch": 0.9072169856231553,
+      "grad_norm": 1.0623280010468192,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 38114
+    },
+    {
+      "epoch": 0.9072407883461868,
+      "grad_norm": 3.3005760979147074,
+      "learning_rate": 2e-05,
+      "loss": 0.6046,
+      "step": 38115
+    },
+    {
+      "epoch": 0.9072645910692183,
+      "grad_norm": 12.183657452412891,
+      "learning_rate": 2e-05,
+      "loss": 0.5899,
+      "step": 38116
+    },
+    {
+      "epoch": 0.9072883937922498,
+      "grad_norm": 0.16528575763906436,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 38117
+    },
+    {
+      "epoch": 0.9073121965152814,
+      "grad_norm": 4.681822979427348,
+      "learning_rate": 2e-05,
+      "loss": 0.8172,
+      "step": 38118
+    },
+    {
+      "epoch": 0.9073359992383129,
+      "grad_norm": 0.1209226162641101,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 38119
+    },
+    {
+      "epoch": 0.9073598019613444,
+      "grad_norm": 5.317837012117837,
+      "learning_rate": 2e-05,
+      "loss": 0.0557,
+      "step": 38120
+    },
+    {
+      "epoch": 0.9073836046843758,
+      "grad_norm": 0.11774717485372334,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 38121
+    },
+    {
+      "epoch": 0.9074074074074074,
+      "grad_norm": 0.834745209629297,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 38122
+    },
+    {
+      "epoch": 0.9074312101304389,
+      "grad_norm": 5.179152278148791,
+      "learning_rate": 2e-05,
+      "loss": 0.536,
+      "step": 38123
+    },
+    {
+      "epoch": 0.9074550128534704,
+      "grad_norm": 0.07706383534632229,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38124
+    },
+    {
+      "epoch": 0.907478815576502,
+      "grad_norm": 0.08105173526327038,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 38125
+    },
+    {
+      "epoch": 0.9075026182995335,
+      "grad_norm": 5.380210420230425,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 38126
+    },
+    {
+      "epoch": 0.907526421022565,
+      "grad_norm": 2.769880092293798,
+      "learning_rate": 2e-05,
+      "loss": 0.0291,
+      "step": 38127
+    },
+    {
+      "epoch": 0.9075502237455965,
+      "grad_norm": 2.4800745720965907,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 38128
+    },
+    {
+      "epoch": 0.9075740264686281,
+      "grad_norm": 10.54162163168928,
+      "learning_rate": 2e-05,
+      "loss": 0.6122,
+      "step": 38129
+    },
+    {
+      "epoch": 0.9075978291916595,
+      "grad_norm": 7.059995733911547,
+      "learning_rate": 2e-05,
+      "loss": 0.5143,
+      "step": 38130
+    },
+    {
+      "epoch": 0.907621631914691,
+      "grad_norm": 2.742207676395694,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 38131
+    },
+    {
+      "epoch": 0.9076454346377225,
+      "grad_norm": 6.931326636643578,
+      "learning_rate": 2e-05,
+      "loss": 0.121,
+      "step": 38132
+    },
+    {
+      "epoch": 0.9076692373607541,
+      "grad_norm": 4.532402152172652,
+      "learning_rate": 2e-05,
+      "loss": 0.8191,
+      "step": 38133
+    },
+    {
+      "epoch": 0.9076930400837856,
+      "grad_norm": 0.2571201739985128,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 38134
+    },
+    {
+      "epoch": 0.9077168428068171,
+      "grad_norm": 6.987919878057003,
+      "learning_rate": 2e-05,
+      "loss": 1.4026,
+      "step": 38135
+    },
+    {
+      "epoch": 0.9077406455298486,
+      "grad_norm": 0.3461091543529827,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 38136
+    },
+    {
+      "epoch": 0.9077644482528802,
+      "grad_norm": 3.036216943953014,
+      "learning_rate": 2e-05,
+      "loss": 0.6052,
+      "step": 38137
+    },
+    {
+      "epoch": 0.9077882509759116,
+      "grad_norm": 7.917055742334378,
+      "learning_rate": 2e-05,
+      "loss": 0.2561,
+      "step": 38138
+    },
+    {
+      "epoch": 0.9078120536989431,
+      "grad_norm": 3.4112694792261835,
+      "learning_rate": 2e-05,
+      "loss": 0.0447,
+      "step": 38139
+    },
+    {
+      "epoch": 0.9078358564219746,
+      "grad_norm": 3.1118989618393207,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 38140
+    },
+    {
+      "epoch": 0.9078596591450062,
+      "grad_norm": 3.8224770636236647,
+      "learning_rate": 2e-05,
+      "loss": 0.8157,
+      "step": 38141
+    },
+    {
+      "epoch": 0.9078834618680377,
+      "grad_norm": 2.120286639479474,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 38142
+    },
+    {
+      "epoch": 0.9079072645910692,
+      "grad_norm": 18.427964683627124,
+      "learning_rate": 2e-05,
+      "loss": 0.8914,
+      "step": 38143
+    },
+    {
+      "epoch": 0.9079310673141008,
+      "grad_norm": 4.8938411655888645,
+      "learning_rate": 2e-05,
+      "loss": 0.5247,
+      "step": 38144
+    },
+    {
+      "epoch": 0.9079548700371323,
+      "grad_norm": 1.0378871417991733,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 38145
+    },
+    {
+      "epoch": 0.9079786727601638,
+      "grad_norm": 3.2931189683874273,
+      "learning_rate": 2e-05,
+      "loss": 0.3936,
+      "step": 38146
+    },
+    {
+      "epoch": 0.9080024754831952,
+      "grad_norm": 19.647770109498737,
+      "learning_rate": 2e-05,
+      "loss": 0.9458,
+      "step": 38147
+    },
+    {
+      "epoch": 0.9080262782062268,
+      "grad_norm": 0.15293314433869448,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 38148
+    },
+    {
+      "epoch": 0.9080500809292583,
+      "grad_norm": 3.066956755022017,
+      "learning_rate": 2e-05,
+      "loss": 0.4946,
+      "step": 38149
+    },
+    {
+      "epoch": 0.9080738836522898,
+      "grad_norm": 1.7638170002053835,
+      "learning_rate": 2e-05,
+      "loss": 0.0929,
+      "step": 38150
+    },
+    {
+      "epoch": 0.9080976863753213,
+      "grad_norm": 6.088913277465626,
+      "learning_rate": 2e-05,
+      "loss": 0.8867,
+      "step": 38151
+    },
+    {
+      "epoch": 0.9081214890983529,
+      "grad_norm": 3.7229369343290086,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 38152
+    },
+    {
+      "epoch": 0.9081452918213844,
+      "grad_norm": 0.03966782996344366,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38153
+    },
+    {
+      "epoch": 0.9081690945444159,
+      "grad_norm": 4.183261178195173,
+      "learning_rate": 2e-05,
+      "loss": 0.1114,
+      "step": 38154
+    },
+    {
+      "epoch": 0.9081928972674473,
+      "grad_norm": 7.451366146064657,
+      "learning_rate": 2e-05,
+      "loss": 0.4242,
+      "step": 38155
+    },
+    {
+      "epoch": 0.9082166999904789,
+      "grad_norm": 2.956790759490175,
+      "learning_rate": 2e-05,
+      "loss": 0.0879,
+      "step": 38156
+    },
+    {
+      "epoch": 0.9082405027135104,
+      "grad_norm": 0.6376402451020793,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 38157
+    },
+    {
+      "epoch": 0.9082643054365419,
+      "grad_norm": 0.21703481841222733,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 38158
+    },
+    {
+      "epoch": 0.9082881081595735,
+      "grad_norm": 0.12155116514169018,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 38159
+    },
+    {
+      "epoch": 0.908311910882605,
+      "grad_norm": 1.0173575067356386,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 38160
+    },
+    {
+      "epoch": 0.9083357136056365,
+      "grad_norm": 1.5103206978756458,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 38161
+    },
+    {
+      "epoch": 0.908359516328668,
+      "grad_norm": 8.193705151505391,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 38162
+    },
+    {
+      "epoch": 0.9083833190516996,
+      "grad_norm": 0.8179654476799056,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 38163
+    },
+    {
+      "epoch": 0.908407121774731,
+      "grad_norm": 6.484675970747004,
+      "learning_rate": 2e-05,
+      "loss": 1.1336,
+      "step": 38164
+    },
+    {
+      "epoch": 0.9084309244977625,
+      "grad_norm": 3.294216631960938,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 38165
+    },
+    {
+      "epoch": 0.908454727220794,
+      "grad_norm": 0.3081239064893027,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 38166
+    },
+    {
+      "epoch": 0.9084785299438256,
+      "grad_norm": 1.8725778243714177,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 38167
+    },
+    {
+      "epoch": 0.9085023326668571,
+      "grad_norm": 4.64512687964152,
+      "learning_rate": 2e-05,
+      "loss": 0.7949,
+      "step": 38168
+    },
+    {
+      "epoch": 0.9085261353898886,
+      "grad_norm": 3.6399145473924586,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 38169
+    },
+    {
+      "epoch": 0.9085499381129201,
+      "grad_norm": 0.02924898479437035,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 38170
+    },
+    {
+      "epoch": 0.9085737408359517,
+      "grad_norm": 2.6216085881464677,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 38171
+    },
+    {
+      "epoch": 0.9085975435589831,
+      "grad_norm": 2.5778237751822033,
+      "learning_rate": 2e-05,
+      "loss": 0.0646,
+      "step": 38172
+    },
+    {
+      "epoch": 0.9086213462820146,
+      "grad_norm": 4.5014370430232695,
+      "learning_rate": 2e-05,
+      "loss": 0.8125,
+      "step": 38173
+    },
+    {
+      "epoch": 0.9086451490050462,
+      "grad_norm": 8.148413313194725,
+      "learning_rate": 2e-05,
+      "loss": 0.8439,
+      "step": 38174
+    },
+    {
+      "epoch": 0.9086689517280777,
+      "grad_norm": 4.518903824114568,
+      "learning_rate": 2e-05,
+      "loss": 1.1434,
+      "step": 38175
+    },
+    {
+      "epoch": 0.9086927544511092,
+      "grad_norm": 0.1330188535328824,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 38176
+    },
+    {
+      "epoch": 0.9087165571741407,
+      "grad_norm": 5.245644639986188,
+      "learning_rate": 2e-05,
+      "loss": 0.6274,
+      "step": 38177
+    },
+    {
+      "epoch": 0.9087403598971723,
+      "grad_norm": 3.7055295040887377,
+      "learning_rate": 2e-05,
+      "loss": 0.1056,
+      "step": 38178
+    },
+    {
+      "epoch": 0.9087641626202038,
+      "grad_norm": 5.990386776813256,
+      "learning_rate": 2e-05,
+      "loss": 0.1406,
+      "step": 38179
+    },
+    {
+      "epoch": 0.9087879653432352,
+      "grad_norm": 2.365333266465231,
+      "learning_rate": 2e-05,
+      "loss": 0.0495,
+      "step": 38180
+    },
+    {
+      "epoch": 0.9088117680662667,
+      "grad_norm": 0.16138904185348826,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38181
+    },
+    {
+      "epoch": 0.9088355707892983,
+      "grad_norm": 5.23543129662502,
+      "learning_rate": 2e-05,
+      "loss": 1.1003,
+      "step": 38182
+    },
+    {
+      "epoch": 0.9088593735123298,
+      "grad_norm": 0.9360406948365163,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 38183
+    },
+    {
+      "epoch": 0.9088831762353613,
+      "grad_norm": 0.4084050727237497,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 38184
+    },
+    {
+      "epoch": 0.9089069789583928,
+      "grad_norm": 14.005228537147817,
+      "learning_rate": 2e-05,
+      "loss": 1.2339,
+      "step": 38185
+    },
+    {
+      "epoch": 0.9089307816814244,
+      "grad_norm": 8.725243841729252,
+      "learning_rate": 2e-05,
+      "loss": 0.7555,
+      "step": 38186
+    },
+    {
+      "epoch": 0.9089545844044559,
+      "grad_norm": 0.046027819477218526,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 38187
+    },
+    {
+      "epoch": 0.9089783871274874,
+      "grad_norm": 1.41537229911737,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 38188
+    },
+    {
+      "epoch": 0.909002189850519,
+      "grad_norm": 0.08584264822740359,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38189
+    },
+    {
+      "epoch": 0.9090259925735504,
+      "grad_norm": 4.925863225297225,
+      "learning_rate": 2e-05,
+      "loss": 0.3537,
+      "step": 38190
+    },
+    {
+      "epoch": 0.9090497952965819,
+      "grad_norm": 0.34992423233271014,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 38191
+    },
+    {
+      "epoch": 0.9090735980196134,
+      "grad_norm": 8.30058497302502,
+      "learning_rate": 2e-05,
+      "loss": 0.5474,
+      "step": 38192
+    },
+    {
+      "epoch": 0.909097400742645,
+      "grad_norm": 0.07899202002031767,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 38193
+    },
+    {
+      "epoch": 0.9091212034656765,
+      "grad_norm": 6.191572527758487,
+      "learning_rate": 2e-05,
+      "loss": 0.766,
+      "step": 38194
+    },
+    {
+      "epoch": 0.909145006188708,
+      "grad_norm": 9.007271840772644,
+      "learning_rate": 2e-05,
+      "loss": 0.9308,
+      "step": 38195
+    },
+    {
+      "epoch": 0.9091688089117395,
+      "grad_norm": 0.664216427966069,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 38196
+    },
+    {
+      "epoch": 0.909192611634771,
+      "grad_norm": 2.0651577044267144,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 38197
+    },
+    {
+      "epoch": 0.9092164143578025,
+      "grad_norm": 3.721685021382257,
+      "learning_rate": 2e-05,
+      "loss": 0.5572,
+      "step": 38198
+    },
+    {
+      "epoch": 0.909240217080834,
+      "grad_norm": 6.019658009857722,
+      "learning_rate": 2e-05,
+      "loss": 0.7179,
+      "step": 38199
+    },
+    {
+      "epoch": 0.9092640198038655,
+      "grad_norm": 24.53519904227397,
+      "learning_rate": 2e-05,
+      "loss": 0.782,
+      "step": 38200
+    },
+    {
+      "epoch": 0.9092878225268971,
+      "grad_norm": 5.590499413073674,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 38201
+    },
+    {
+      "epoch": 0.9093116252499286,
+      "grad_norm": 0.060769965433911874,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 38202
+    },
+    {
+      "epoch": 0.9093354279729601,
+      "grad_norm": 6.217660822098375,
+      "learning_rate": 2e-05,
+      "loss": 1.5669,
+      "step": 38203
+    },
+    {
+      "epoch": 0.9093592306959917,
+      "grad_norm": 6.8763221394028,
+      "learning_rate": 2e-05,
+      "loss": 0.5351,
+      "step": 38204
+    },
+    {
+      "epoch": 0.9093830334190232,
+      "grad_norm": 0.7723511569021193,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 38205
+    },
+    {
+      "epoch": 0.9094068361420546,
+      "grad_norm": 5.758715451073192,
+      "learning_rate": 2e-05,
+      "loss": 1.0404,
+      "step": 38206
+    },
+    {
+      "epoch": 0.9094306388650861,
+      "grad_norm": 2.682095635594155,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 38207
+    },
+    {
+      "epoch": 0.9094544415881177,
+      "grad_norm": 3.475457084661601,
+      "learning_rate": 2e-05,
+      "loss": 0.0767,
+      "step": 38208
+    },
+    {
+      "epoch": 0.9094782443111492,
+      "grad_norm": 8.287852224003704,
+      "learning_rate": 2e-05,
+      "loss": 0.5668,
+      "step": 38209
+    },
+    {
+      "epoch": 0.9095020470341807,
+      "grad_norm": 6.832749039366021,
+      "learning_rate": 2e-05,
+      "loss": 0.2968,
+      "step": 38210
+    },
+    {
+      "epoch": 0.9095258497572122,
+      "grad_norm": 6.115710727167249,
+      "learning_rate": 2e-05,
+      "loss": 1.0486,
+      "step": 38211
+    },
+    {
+      "epoch": 0.9095496524802438,
+      "grad_norm": 3.7602969436237057,
+      "learning_rate": 2e-05,
+      "loss": 0.8252,
+      "step": 38212
+    },
+    {
+      "epoch": 0.9095734552032753,
+      "grad_norm": 5.152133981470203,
+      "learning_rate": 2e-05,
+      "loss": 0.5769,
+      "step": 38213
+    },
+    {
+      "epoch": 0.9095972579263067,
+      "grad_norm": 7.22084233380647,
+      "learning_rate": 2e-05,
+      "loss": 0.6816,
+      "step": 38214
+    },
+    {
+      "epoch": 0.9096210606493382,
+      "grad_norm": 3.8390263914740066,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 38215
+    },
+    {
+      "epoch": 0.9096448633723698,
+      "grad_norm": 0.2040761436307632,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 38216
+    },
+    {
+      "epoch": 0.9096686660954013,
+      "grad_norm": 7.852726003152247,
+      "learning_rate": 2e-05,
+      "loss": 0.2331,
+      "step": 38217
+    },
+    {
+      "epoch": 0.9096924688184328,
+      "grad_norm": 0.6390970130008884,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 38218
+    },
+    {
+      "epoch": 0.9097162715414644,
+      "grad_norm": 2.7173459289072124,
+      "learning_rate": 2e-05,
+      "loss": 0.0393,
+      "step": 38219
+    },
+    {
+      "epoch": 0.9097400742644959,
+      "grad_norm": 3.9798096559152683,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 38220
+    },
+    {
+      "epoch": 0.9097638769875274,
+      "grad_norm": 2.6801220985899827,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 38221
+    },
+    {
+      "epoch": 0.9097876797105588,
+      "grad_norm": 2.2842670773244587,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 38222
+    },
+    {
+      "epoch": 0.9098114824335904,
+      "grad_norm": 6.162664504981144,
+      "learning_rate": 2e-05,
+      "loss": 1.5145,
+      "step": 38223
+    },
+    {
+      "epoch": 0.9098352851566219,
+      "grad_norm": 7.279091856362469,
+      "learning_rate": 2e-05,
+      "loss": 0.4495,
+      "step": 38224
+    },
+    {
+      "epoch": 0.9098590878796534,
+      "grad_norm": 2.6106523382327005,
+      "learning_rate": 2e-05,
+      "loss": 0.29,
+      "step": 38225
+    },
+    {
+      "epoch": 0.9098828906026849,
+      "grad_norm": 0.18373016558494965,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 38226
+    },
+    {
+      "epoch": 0.9099066933257165,
+      "grad_norm": 1.6330986170313346,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 38227
+    },
+    {
+      "epoch": 0.909930496048748,
+      "grad_norm": 1.4135631600748069,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 38228
+    },
+    {
+      "epoch": 0.9099542987717795,
+      "grad_norm": 6.132012384409411,
+      "learning_rate": 2e-05,
+      "loss": 1.6682,
+      "step": 38229
+    },
+    {
+      "epoch": 0.909978101494811,
+      "grad_norm": 5.524896082842142,
+      "learning_rate": 2e-05,
+      "loss": 0.2434,
+      "step": 38230
+    },
+    {
+      "epoch": 0.9100019042178425,
+      "grad_norm": 5.401166338096674,
+      "learning_rate": 2e-05,
+      "loss": 0.7196,
+      "step": 38231
+    },
+    {
+      "epoch": 0.910025706940874,
+      "grad_norm": 1.9346088183333745,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 38232
+    },
+    {
+      "epoch": 0.9100495096639055,
+      "grad_norm": 3.8350890426263096,
+      "learning_rate": 2e-05,
+      "loss": 0.2936,
+      "step": 38233
+    },
+    {
+      "epoch": 0.9100733123869371,
+      "grad_norm": 11.759987664773119,
+      "learning_rate": 2e-05,
+      "loss": 0.7397,
+      "step": 38234
+    },
+    {
+      "epoch": 0.9100971151099686,
+      "grad_norm": 3.874335189824789,
+      "learning_rate": 2e-05,
+      "loss": 0.4274,
+      "step": 38235
+    },
+    {
+      "epoch": 0.9101209178330001,
+      "grad_norm": 1.1791940114204502,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 38236
+    },
+    {
+      "epoch": 0.9101447205560316,
+      "grad_norm": 3.7951585695189274,
+      "learning_rate": 2e-05,
+      "loss": 0.7255,
+      "step": 38237
+    },
+    {
+      "epoch": 0.9101685232790632,
+      "grad_norm": 5.4624630658884685,
+      "learning_rate": 2e-05,
+      "loss": 1.0003,
+      "step": 38238
+    },
+    {
+      "epoch": 0.9101923260020947,
+      "grad_norm": 4.710041715642162,
+      "learning_rate": 2e-05,
+      "loss": 0.5335,
+      "step": 38239
+    },
+    {
+      "epoch": 0.9102161287251261,
+      "grad_norm": 2.898229981700883,
+      "learning_rate": 2e-05,
+      "loss": 0.3619,
+      "step": 38240
+    },
+    {
+      "epoch": 0.9102399314481576,
+      "grad_norm": 4.901086729458532,
+      "learning_rate": 2e-05,
+      "loss": 1.1646,
+      "step": 38241
+    },
+    {
+      "epoch": 0.9102637341711892,
+      "grad_norm": 0.12735846669543432,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 38242
+    },
+    {
+      "epoch": 0.9102875368942207,
+      "grad_norm": 0.09494638200360873,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 38243
+    },
+    {
+      "epoch": 0.9103113396172522,
+      "grad_norm": 4.4122634281355895,
+      "learning_rate": 2e-05,
+      "loss": 0.9461,
+      "step": 38244
+    },
+    {
+      "epoch": 0.9103351423402837,
+      "grad_norm": 0.5636780627339462,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 38245
+    },
+    {
+      "epoch": 0.9103589450633153,
+      "grad_norm": 5.368514493498851,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 38246
+    },
+    {
+      "epoch": 0.9103827477863468,
+      "grad_norm": 1.9658660625496718,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 38247
+    },
+    {
+      "epoch": 0.9104065505093782,
+      "grad_norm": 8.468395922404008,
+      "learning_rate": 2e-05,
+      "loss": 0.9866,
+      "step": 38248
+    },
+    {
+      "epoch": 0.9104303532324098,
+      "grad_norm": 3.2169336424010186,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 38249
+    },
+    {
+      "epoch": 0.9104541559554413,
+      "grad_norm": 4.920576416463669,
+      "learning_rate": 2e-05,
+      "loss": 1.3499,
+      "step": 38250
+    },
+    {
+      "epoch": 0.9104779586784728,
+      "grad_norm": 6.555202194302872,
+      "learning_rate": 2e-05,
+      "loss": 0.3311,
+      "step": 38251
+    },
+    {
+      "epoch": 0.9105017614015043,
+      "grad_norm": 6.9348603521022705,
+      "learning_rate": 2e-05,
+      "loss": 0.9968,
+      "step": 38252
+    },
+    {
+      "epoch": 0.9105255641245359,
+      "grad_norm": 9.153193581548104,
+      "learning_rate": 2e-05,
+      "loss": 0.2315,
+      "step": 38253
+    },
+    {
+      "epoch": 0.9105493668475674,
+      "grad_norm": 0.21468461741817296,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 38254
+    },
+    {
+      "epoch": 0.9105731695705989,
+      "grad_norm": 0.29777513168413605,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 38255
+    },
+    {
+      "epoch": 0.9105969722936303,
+      "grad_norm": 2.4459893303376368,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 38256
+    },
+    {
+      "epoch": 0.9106207750166619,
+      "grad_norm": 4.140171897122071,
+      "learning_rate": 2e-05,
+      "loss": 0.6793,
+      "step": 38257
+    },
+    {
+      "epoch": 0.9106445777396934,
+      "grad_norm": 4.394160443177966,
+      "learning_rate": 2e-05,
+      "loss": 0.6709,
+      "step": 38258
+    },
+    {
+      "epoch": 0.9106683804627249,
+      "grad_norm": 0.6976156067845714,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 38259
+    },
+    {
+      "epoch": 0.9106921831857564,
+      "grad_norm": 4.547171734823422,
+      "learning_rate": 2e-05,
+      "loss": 0.9215,
+      "step": 38260
+    },
+    {
+      "epoch": 0.910715985908788,
+      "grad_norm": 6.519708309695663,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 38261
+    },
+    {
+      "epoch": 0.9107397886318195,
+      "grad_norm": 4.272996529662591,
+      "learning_rate": 2e-05,
+      "loss": 0.9266,
+      "step": 38262
+    },
+    {
+      "epoch": 0.910763591354851,
+      "grad_norm": 5.365094045678181,
+      "learning_rate": 2e-05,
+      "loss": 0.3073,
+      "step": 38263
+    },
+    {
+      "epoch": 0.9107873940778826,
+      "grad_norm": 0.5729994963380906,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 38264
+    },
+    {
+      "epoch": 0.910811196800914,
+      "grad_norm": 7.293125874805525,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 38265
+    },
+    {
+      "epoch": 0.9108349995239455,
+      "grad_norm": 4.204612322952607,
+      "learning_rate": 2e-05,
+      "loss": 0.6685,
+      "step": 38266
+    },
+    {
+      "epoch": 0.910858802246977,
+      "grad_norm": 2.766533486419343,
+      "learning_rate": 2e-05,
+      "loss": 0.2281,
+      "step": 38267
+    },
+    {
+      "epoch": 0.9108826049700086,
+      "grad_norm": 2.7227891109046674,
+      "learning_rate": 2e-05,
+      "loss": 0.3465,
+      "step": 38268
+    },
+    {
+      "epoch": 0.9109064076930401,
+      "grad_norm": 0.14223143079112402,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 38269
+    },
+    {
+      "epoch": 0.9109302104160716,
+      "grad_norm": 5.223237887741678,
+      "learning_rate": 2e-05,
+      "loss": 0.1858,
+      "step": 38270
+    },
+    {
+      "epoch": 0.9109540131391031,
+      "grad_norm": 7.6216031668800746,
+      "learning_rate": 2e-05,
+      "loss": 1.246,
+      "step": 38271
+    },
+    {
+      "epoch": 0.9109778158621347,
+      "grad_norm": 0.28578704936121213,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 38272
+    },
+    {
+      "epoch": 0.9110016185851662,
+      "grad_norm": 4.3526036977089095,
+      "learning_rate": 2e-05,
+      "loss": 0.522,
+      "step": 38273
+    },
+    {
+      "epoch": 0.9110254213081976,
+      "grad_norm": 0.8545559641011962,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 38274
+    },
+    {
+      "epoch": 0.9110492240312291,
+      "grad_norm": 2.7509186643692836,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 38275
+    },
+    {
+      "epoch": 0.9110730267542607,
+      "grad_norm": 4.978206495462622,
+      "learning_rate": 2e-05,
+      "loss": 0.5505,
+      "step": 38276
+    },
+    {
+      "epoch": 0.9110968294772922,
+      "grad_norm": 4.881745641193669,
+      "learning_rate": 2e-05,
+      "loss": 0.7703,
+      "step": 38277
+    },
+    {
+      "epoch": 0.9111206322003237,
+      "grad_norm": 2.7328872331632206,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 38278
+    },
+    {
+      "epoch": 0.9111444349233553,
+      "grad_norm": 4.451354248111437,
+      "learning_rate": 2e-05,
+      "loss": 0.4409,
+      "step": 38279
+    },
+    {
+      "epoch": 0.9111682376463868,
+      "grad_norm": 2.9796437733136814,
+      "learning_rate": 2e-05,
+      "loss": 0.3165,
+      "step": 38280
+    },
+    {
+      "epoch": 0.9111920403694183,
+      "grad_norm": 8.635003408841616,
+      "learning_rate": 2e-05,
+      "loss": 0.4468,
+      "step": 38281
+    },
+    {
+      "epoch": 0.9112158430924497,
+      "grad_norm": 2.504454160042329,
+      "learning_rate": 2e-05,
+      "loss": 0.2635,
+      "step": 38282
+    },
+    {
+      "epoch": 0.9112396458154813,
+      "grad_norm": 1.735219460010725,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 38283
+    },
+    {
+      "epoch": 0.9112634485385128,
+      "grad_norm": 1.4481507060405894,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 38284
+    },
+    {
+      "epoch": 0.9112872512615443,
+      "grad_norm": 4.293397822290934,
+      "learning_rate": 2e-05,
+      "loss": 0.0698,
+      "step": 38285
+    },
+    {
+      "epoch": 0.9113110539845758,
+      "grad_norm": 0.2123840524047776,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 38286
+    },
+    {
+      "epoch": 0.9113348567076074,
+      "grad_norm": 4.741617135875854,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 38287
+    },
+    {
+      "epoch": 0.9113586594306389,
+      "grad_norm": 0.09942893933903117,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38288
+    },
+    {
+      "epoch": 0.9113824621536704,
+      "grad_norm": 5.377602175193968,
+      "learning_rate": 2e-05,
+      "loss": 0.1038,
+      "step": 38289
+    },
+    {
+      "epoch": 0.9114062648767018,
+      "grad_norm": 0.6125147877185091,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 38290
+    },
+    {
+      "epoch": 0.9114300675997334,
+      "grad_norm": 4.363483665360822,
+      "learning_rate": 2e-05,
+      "loss": 0.4398,
+      "step": 38291
+    },
+    {
+      "epoch": 0.9114538703227649,
+      "grad_norm": 3.2881249092257487,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 38292
+    },
+    {
+      "epoch": 0.9114776730457964,
+      "grad_norm": 0.19711953224642137,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 38293
+    },
+    {
+      "epoch": 0.911501475768828,
+      "grad_norm": 2.06798275045984,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 38294
+    },
+    {
+      "epoch": 0.9115252784918595,
+      "grad_norm": 1.517519836679476,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 38295
+    },
+    {
+      "epoch": 0.911549081214891,
+      "grad_norm": 0.057139292810634845,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 38296
+    },
+    {
+      "epoch": 0.9115728839379225,
+      "grad_norm": 4.956924302453794,
+      "learning_rate": 2e-05,
+      "loss": 0.8036,
+      "step": 38297
+    },
+    {
+      "epoch": 0.9115966866609541,
+      "grad_norm": 4.871681244032512,
+      "learning_rate": 2e-05,
+      "loss": 0.372,
+      "step": 38298
+    },
+    {
+      "epoch": 0.9116204893839855,
+      "grad_norm": 0.4225016112897293,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 38299
+    },
+    {
+      "epoch": 0.911644292107017,
+      "grad_norm": 3.6111408865114116,
+      "learning_rate": 2e-05,
+      "loss": 0.6883,
+      "step": 38300
+    },
+    {
+      "epoch": 0.9116680948300485,
+      "grad_norm": 7.854087109795036,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 38301
+    },
+    {
+      "epoch": 0.9116918975530801,
+      "grad_norm": 12.55088321262526,
+      "learning_rate": 2e-05,
+      "loss": 0.5304,
+      "step": 38302
+    },
+    {
+      "epoch": 0.9117157002761116,
+      "grad_norm": 7.651315306608165,
+      "learning_rate": 2e-05,
+      "loss": 0.8249,
+      "step": 38303
+    },
+    {
+      "epoch": 0.9117395029991431,
+      "grad_norm": 3.1300047026180517,
+      "learning_rate": 2e-05,
+      "loss": 0.6254,
+      "step": 38304
+    },
+    {
+      "epoch": 0.9117633057221746,
+      "grad_norm": 5.062340380063564,
+      "learning_rate": 2e-05,
+      "loss": 0.7147,
+      "step": 38305
+    },
+    {
+      "epoch": 0.9117871084452062,
+      "grad_norm": 1.1224578940267917,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 38306
+    },
+    {
+      "epoch": 0.9118109111682376,
+      "grad_norm": 7.266336042268523,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 38307
+    },
+    {
+      "epoch": 0.9118347138912691,
+      "grad_norm": 4.918210382969826,
+      "learning_rate": 2e-05,
+      "loss": 0.3075,
+      "step": 38308
+    },
+    {
+      "epoch": 0.9118585166143007,
+      "grad_norm": 5.8516752675387815,
+      "learning_rate": 2e-05,
+      "loss": 0.9809,
+      "step": 38309
+    },
+    {
+      "epoch": 0.9118823193373322,
+      "grad_norm": 1.9936422330929893,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 38310
+    },
+    {
+      "epoch": 0.9119061220603637,
+      "grad_norm": 4.156901508654289,
+      "learning_rate": 2e-05,
+      "loss": 0.4199,
+      "step": 38311
+    },
+    {
+      "epoch": 0.9119299247833952,
+      "grad_norm": 82.70398946970221,
+      "learning_rate": 2e-05,
+      "loss": 0.9568,
+      "step": 38312
+    },
+    {
+      "epoch": 0.9119537275064268,
+      "grad_norm": 4.356271368383315,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 38313
+    },
+    {
+      "epoch": 0.9119775302294583,
+      "grad_norm": 4.653001883022443,
+      "learning_rate": 2e-05,
+      "loss": 0.5422,
+      "step": 38314
+    },
+    {
+      "epoch": 0.9120013329524898,
+      "grad_norm": 2.7862158977483347,
+      "learning_rate": 2e-05,
+      "loss": 0.5048,
+      "step": 38315
+    },
+    {
+      "epoch": 0.9120251356755212,
+      "grad_norm": 8.812908523337404,
+      "learning_rate": 2e-05,
+      "loss": 0.1057,
+      "step": 38316
+    },
+    {
+      "epoch": 0.9120489383985528,
+      "grad_norm": 2.475046478146604,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 38317
+    },
+    {
+      "epoch": 0.9120727411215843,
+      "grad_norm": 0.15840050084133084,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 38318
+    },
+    {
+      "epoch": 0.9120965438446158,
+      "grad_norm": 2.528092641832547,
+      "learning_rate": 2e-05,
+      "loss": 0.0527,
+      "step": 38319
+    },
+    {
+      "epoch": 0.9121203465676473,
+      "grad_norm": 0.26323894431440115,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 38320
+    },
+    {
+      "epoch": 0.9121441492906789,
+      "grad_norm": 9.625277040248983,
+      "learning_rate": 2e-05,
+      "loss": 0.8504,
+      "step": 38321
+    },
+    {
+      "epoch": 0.9121679520137104,
+      "grad_norm": 0.07934819593411076,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38322
+    },
+    {
+      "epoch": 0.9121917547367419,
+      "grad_norm": 7.230892064439528,
+      "learning_rate": 2e-05,
+      "loss": 0.338,
+      "step": 38323
+    },
+    {
+      "epoch": 0.9122155574597735,
+      "grad_norm": 6.733495172158999,
+      "learning_rate": 2e-05,
+      "loss": 0.8122,
+      "step": 38324
+    },
+    {
+      "epoch": 0.9122393601828049,
+      "grad_norm": 3.2945906124986015,
+      "learning_rate": 2e-05,
+      "loss": 0.5818,
+      "step": 38325
+    },
+    {
+      "epoch": 0.9122631629058364,
+      "grad_norm": 0.21832439347845298,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38326
+    },
+    {
+      "epoch": 0.9122869656288679,
+      "grad_norm": 2.4440780219697142,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 38327
+    },
+    {
+      "epoch": 0.9123107683518995,
+      "grad_norm": 0.27202198823180135,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 38328
+    },
+    {
+      "epoch": 0.912334571074931,
+      "grad_norm": 1.1853563189837328,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 38329
+    },
+    {
+      "epoch": 0.9123583737979625,
+      "grad_norm": 3.3068315863759175,
+      "learning_rate": 2e-05,
+      "loss": 0.7047,
+      "step": 38330
+    },
+    {
+      "epoch": 0.912382176520994,
+      "grad_norm": 0.5840421765118282,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 38331
+    },
+    {
+      "epoch": 0.9124059792440256,
+      "grad_norm": 5.486754052056268,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 38332
+    },
+    {
+      "epoch": 0.912429781967057,
+      "grad_norm": 5.411844003753838,
+      "learning_rate": 2e-05,
+      "loss": 0.6846,
+      "step": 38333
+    },
+    {
+      "epoch": 0.9124535846900885,
+      "grad_norm": 0.10848457046703211,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38334
+    },
+    {
+      "epoch": 0.91247738741312,
+      "grad_norm": 13.731526551004423,
+      "learning_rate": 2e-05,
+      "loss": 0.2363,
+      "step": 38335
+    },
+    {
+      "epoch": 0.9125011901361516,
+      "grad_norm": 30.555531216221652,
+      "learning_rate": 2e-05,
+      "loss": 0.7347,
+      "step": 38336
+    },
+    {
+      "epoch": 0.9125249928591831,
+      "grad_norm": 3.449294772246859,
+      "learning_rate": 2e-05,
+      "loss": 0.951,
+      "step": 38337
+    },
+    {
+      "epoch": 0.9125487955822146,
+      "grad_norm": 4.6740937769557265,
+      "learning_rate": 2e-05,
+      "loss": 0.9157,
+      "step": 38338
+    },
+    {
+      "epoch": 0.9125725983052462,
+      "grad_norm": 7.65994668987998,
+      "learning_rate": 2e-05,
+      "loss": 0.0604,
+      "step": 38339
+    },
+    {
+      "epoch": 0.9125964010282777,
+      "grad_norm": 3.6917253071830287,
+      "learning_rate": 2e-05,
+      "loss": 0.8384,
+      "step": 38340
+    },
+    {
+      "epoch": 0.9126202037513091,
+      "grad_norm": 0.8389575658991977,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 38341
+    },
+    {
+      "epoch": 0.9126440064743406,
+      "grad_norm": 4.174392003564653,
+      "learning_rate": 2e-05,
+      "loss": 0.1411,
+      "step": 38342
+    },
+    {
+      "epoch": 0.9126678091973722,
+      "grad_norm": 4.195710118775337,
+      "learning_rate": 2e-05,
+      "loss": 0.5702,
+      "step": 38343
+    },
+    {
+      "epoch": 0.9126916119204037,
+      "grad_norm": 7.202613580488152,
+      "learning_rate": 2e-05,
+      "loss": 0.4675,
+      "step": 38344
+    },
+    {
+      "epoch": 0.9127154146434352,
+      "grad_norm": 5.847928877669668,
+      "learning_rate": 2e-05,
+      "loss": 1.0398,
+      "step": 38345
+    },
+    {
+      "epoch": 0.9127392173664667,
+      "grad_norm": 0.3532749867427542,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 38346
+    },
+    {
+      "epoch": 0.9127630200894983,
+      "grad_norm": 6.348643288734759,
+      "learning_rate": 2e-05,
+      "loss": 0.1181,
+      "step": 38347
+    },
+    {
+      "epoch": 0.9127868228125298,
+      "grad_norm": 5.862210486003823,
+      "learning_rate": 2e-05,
+      "loss": 1.5781,
+      "step": 38348
+    },
+    {
+      "epoch": 0.9128106255355612,
+      "grad_norm": 4.746110020892259,
+      "learning_rate": 2e-05,
+      "loss": 0.677,
+      "step": 38349
+    },
+    {
+      "epoch": 0.9128344282585927,
+      "grad_norm": 8.411066789045387,
+      "learning_rate": 2e-05,
+      "loss": 0.5482,
+      "step": 38350
+    },
+    {
+      "epoch": 0.9128582309816243,
+      "grad_norm": 16.450752058376256,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 38351
+    },
+    {
+      "epoch": 0.9128820337046558,
+      "grad_norm": 2.2067041728257557,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 38352
+    },
+    {
+      "epoch": 0.9129058364276873,
+      "grad_norm": 6.2458604267219755,
+      "learning_rate": 2e-05,
+      "loss": 1.3254,
+      "step": 38353
+    },
+    {
+      "epoch": 0.9129296391507189,
+      "grad_norm": 5.924396633491743,
+      "learning_rate": 2e-05,
+      "loss": 1.1519,
+      "step": 38354
+    },
+    {
+      "epoch": 0.9129534418737504,
+      "grad_norm": 1.1030254843766545,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 38355
+    },
+    {
+      "epoch": 0.9129772445967819,
+      "grad_norm": 3.5559840460338723,
+      "learning_rate": 2e-05,
+      "loss": 0.439,
+      "step": 38356
+    },
+    {
+      "epoch": 0.9130010473198134,
+      "grad_norm": 4.729843443604808,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 38357
+    },
+    {
+      "epoch": 0.913024850042845,
+      "grad_norm": 2.3894755576750843,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 38358
+    },
+    {
+      "epoch": 0.9130486527658764,
+      "grad_norm": 2.228263379597193,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 38359
+    },
+    {
+      "epoch": 0.9130724554889079,
+      "grad_norm": 1.3299705955634875,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 38360
+    },
+    {
+      "epoch": 0.9130962582119394,
+      "grad_norm": 2.7604406911566137,
+      "learning_rate": 2e-05,
+      "loss": 0.3035,
+      "step": 38361
+    },
+    {
+      "epoch": 0.913120060934971,
+      "grad_norm": 4.031454691759107,
+      "learning_rate": 2e-05,
+      "loss": 0.449,
+      "step": 38362
+    },
+    {
+      "epoch": 0.9131438636580025,
+      "grad_norm": 3.10876477119625,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 38363
+    },
+    {
+      "epoch": 0.913167666381034,
+      "grad_norm": 3.2655153526595755,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 38364
+    },
+    {
+      "epoch": 0.9131914691040655,
+      "grad_norm": 8.873416513330667,
+      "learning_rate": 2e-05,
+      "loss": 0.8427,
+      "step": 38365
+    },
+    {
+      "epoch": 0.913215271827097,
+      "grad_norm": 0.33487456846976693,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 38366
+    },
+    {
+      "epoch": 0.9132390745501285,
+      "grad_norm": 0.9135623364112518,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 38367
+    },
+    {
+      "epoch": 0.91326287727316,
+      "grad_norm": 2.1139309811290667,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 38368
+    },
+    {
+      "epoch": 0.9132866799961916,
+      "grad_norm": 8.63771933449074,
+      "learning_rate": 2e-05,
+      "loss": 1.0861,
+      "step": 38369
+    },
+    {
+      "epoch": 0.9133104827192231,
+      "grad_norm": 6.002629109935521,
+      "learning_rate": 2e-05,
+      "loss": 0.9238,
+      "step": 38370
+    },
+    {
+      "epoch": 0.9133342854422546,
+      "grad_norm": 5.2753691601931285,
+      "learning_rate": 2e-05,
+      "loss": 0.135,
+      "step": 38371
+    },
+    {
+      "epoch": 0.9133580881652861,
+      "grad_norm": 0.4551156265674191,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 38372
+    },
+    {
+      "epoch": 0.9133818908883177,
+      "grad_norm": 1.5728503301763297,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 38373
+    },
+    {
+      "epoch": 0.9134056936113492,
+      "grad_norm": 3.918908803447503,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 38374
+    },
+    {
+      "epoch": 0.9134294963343806,
+      "grad_norm": 0.16191135734757528,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 38375
+    },
+    {
+      "epoch": 0.9134532990574121,
+      "grad_norm": 0.6464563256733081,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 38376
+    },
+    {
+      "epoch": 0.9134771017804437,
+      "grad_norm": 5.760130950535819,
+      "learning_rate": 2e-05,
+      "loss": 0.2926,
+      "step": 38377
+    },
+    {
+      "epoch": 0.9135009045034752,
+      "grad_norm": 2.3418993175803844,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 38378
+    },
+    {
+      "epoch": 0.9135247072265067,
+      "grad_norm": 2.855516528516749,
+      "learning_rate": 2e-05,
+      "loss": 0.4234,
+      "step": 38379
+    },
+    {
+      "epoch": 0.9135485099495382,
+      "grad_norm": 5.208910076143153,
+      "learning_rate": 2e-05,
+      "loss": 0.5705,
+      "step": 38380
+    },
+    {
+      "epoch": 0.9135723126725698,
+      "grad_norm": 10.659665077996406,
+      "learning_rate": 2e-05,
+      "loss": 0.637,
+      "step": 38381
+    },
+    {
+      "epoch": 0.9135961153956013,
+      "grad_norm": 25.05236153242109,
+      "learning_rate": 2e-05,
+      "loss": 0.4134,
+      "step": 38382
+    },
+    {
+      "epoch": 0.9136199181186327,
+      "grad_norm": 0.026985216656489456,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 38383
+    },
+    {
+      "epoch": 0.9136437208416643,
+      "grad_norm": 0.37991407409541095,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38384
+    },
+    {
+      "epoch": 0.9136675235646958,
+      "grad_norm": 3.8700170809067793,
+      "learning_rate": 2e-05,
+      "loss": 0.1052,
+      "step": 38385
+    },
+    {
+      "epoch": 0.9136913262877273,
+      "grad_norm": 6.405545999010011,
+      "learning_rate": 2e-05,
+      "loss": 0.7107,
+      "step": 38386
+    },
+    {
+      "epoch": 0.9137151290107588,
+      "grad_norm": 0.09282963051929963,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 38387
+    },
+    {
+      "epoch": 0.9137389317337904,
+      "grad_norm": 2.6392053386463963,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 38388
+    },
+    {
+      "epoch": 0.9137627344568219,
+      "grad_norm": 8.301012795340243,
+      "learning_rate": 2e-05,
+      "loss": 2.0991,
+      "step": 38389
+    },
+    {
+      "epoch": 0.9137865371798534,
+      "grad_norm": 7.244725751034893,
+      "learning_rate": 2e-05,
+      "loss": 0.0789,
+      "step": 38390
+    },
+    {
+      "epoch": 0.9138103399028848,
+      "grad_norm": 4.0734029779701935,
+      "learning_rate": 2e-05,
+      "loss": 0.548,
+      "step": 38391
+    },
+    {
+      "epoch": 0.9138341426259164,
+      "grad_norm": 11.552958281105642,
+      "learning_rate": 2e-05,
+      "loss": 1.6445,
+      "step": 38392
+    },
+    {
+      "epoch": 0.9138579453489479,
+      "grad_norm": 0.6164161151712727,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 38393
+    },
+    {
+      "epoch": 0.9138817480719794,
+      "grad_norm": 13.061402240163298,
+      "learning_rate": 2e-05,
+      "loss": 0.4536,
+      "step": 38394
+    },
+    {
+      "epoch": 0.9139055507950109,
+      "grad_norm": 8.02627916726305,
+      "learning_rate": 2e-05,
+      "loss": 1.7701,
+      "step": 38395
+    },
+    {
+      "epoch": 0.9139293535180425,
+      "grad_norm": 2.952781878676338,
+      "learning_rate": 2e-05,
+      "loss": 0.0581,
+      "step": 38396
+    },
+    {
+      "epoch": 0.913953156241074,
+      "grad_norm": 0.6512267983307836,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 38397
+    },
+    {
+      "epoch": 0.9139769589641055,
+      "grad_norm": 5.662109008460937,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 38398
+    },
+    {
+      "epoch": 0.914000761687137,
+      "grad_norm": 0.053208792236228195,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38399
+    },
+    {
+      "epoch": 0.9140245644101685,
+      "grad_norm": 4.831190671447378,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 38400
+    },
+    {
+      "epoch": 0.9140483671332,
+      "grad_norm": 1.7297084163214984,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 38401
+    },
+    {
+      "epoch": 0.9140721698562315,
+      "grad_norm": 3.1709296026524982,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 38402
+    },
+    {
+      "epoch": 0.9140959725792631,
+      "grad_norm": 3.6239624461200117,
+      "learning_rate": 2e-05,
+      "loss": 0.5251,
+      "step": 38403
+    },
+    {
+      "epoch": 0.9141197753022946,
+      "grad_norm": 4.381126091557544,
+      "learning_rate": 2e-05,
+      "loss": 0.8066,
+      "step": 38404
+    },
+    {
+      "epoch": 0.9141435780253261,
+      "grad_norm": 12.464735753236095,
+      "learning_rate": 2e-05,
+      "loss": 1.2253,
+      "step": 38405
+    },
+    {
+      "epoch": 0.9141673807483576,
+      "grad_norm": 6.315540277598241,
+      "learning_rate": 2e-05,
+      "loss": 0.8958,
+      "step": 38406
+    },
+    {
+      "epoch": 0.9141911834713892,
+      "grad_norm": 3.7145970148966985,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 38407
+    },
+    {
+      "epoch": 0.9142149861944207,
+      "grad_norm": 22.687426422895914,
+      "learning_rate": 2e-05,
+      "loss": 0.4271,
+      "step": 38408
+    },
+    {
+      "epoch": 0.9142387889174521,
+      "grad_norm": 1.854235822661644,
+      "learning_rate": 2e-05,
+      "loss": 0.0827,
+      "step": 38409
+    },
+    {
+      "epoch": 0.9142625916404836,
+      "grad_norm": 1.4493399114711047,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 38410
+    },
+    {
+      "epoch": 0.9142863943635152,
+      "grad_norm": 0.2734492727985376,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 38411
+    },
+    {
+      "epoch": 0.9143101970865467,
+      "grad_norm": 2.0752834985445467,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 38412
+    },
+    {
+      "epoch": 0.9143339998095782,
+      "grad_norm": 13.327805612769442,
+      "learning_rate": 2e-05,
+      "loss": 0.1575,
+      "step": 38413
+    },
+    {
+      "epoch": 0.9143578025326097,
+      "grad_norm": 3.6356450199448083,
+      "learning_rate": 2e-05,
+      "loss": 0.9155,
+      "step": 38414
+    },
+    {
+      "epoch": 0.9143816052556413,
+      "grad_norm": 4.897162058154419,
+      "learning_rate": 2e-05,
+      "loss": 1.3578,
+      "step": 38415
+    },
+    {
+      "epoch": 0.9144054079786728,
+      "grad_norm": 4.5267907518178925,
+      "learning_rate": 2e-05,
+      "loss": 0.6339,
+      "step": 38416
+    },
+    {
+      "epoch": 0.9144292107017042,
+      "grad_norm": 1.2035997509607228,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 38417
+    },
+    {
+      "epoch": 0.9144530134247358,
+      "grad_norm": 6.000699735184181,
+      "learning_rate": 2e-05,
+      "loss": 0.8497,
+      "step": 38418
+    },
+    {
+      "epoch": 0.9144768161477673,
+      "grad_norm": 8.124904596111994,
+      "learning_rate": 2e-05,
+      "loss": 0.3599,
+      "step": 38419
+    },
+    {
+      "epoch": 0.9145006188707988,
+      "grad_norm": 0.44522520556886463,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 38420
+    },
+    {
+      "epoch": 0.9145244215938303,
+      "grad_norm": 3.185578412068352,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 38421
+    },
+    {
+      "epoch": 0.9145482243168619,
+      "grad_norm": 9.499035276966795,
+      "learning_rate": 2e-05,
+      "loss": 0.4841,
+      "step": 38422
+    },
+    {
+      "epoch": 0.9145720270398934,
+      "grad_norm": 0.16969990500293067,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 38423
+    },
+    {
+      "epoch": 0.9145958297629249,
+      "grad_norm": 4.6512973340839805,
+      "learning_rate": 2e-05,
+      "loss": 0.096,
+      "step": 38424
+    },
+    {
+      "epoch": 0.9146196324859563,
+      "grad_norm": 15.656786813046336,
+      "learning_rate": 2e-05,
+      "loss": 0.7492,
+      "step": 38425
+    },
+    {
+      "epoch": 0.9146434352089879,
+      "grad_norm": 0.25935095344147985,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 38426
+    },
+    {
+      "epoch": 0.9146672379320194,
+      "grad_norm": 5.086080487746947,
+      "learning_rate": 2e-05,
+      "loss": 0.8091,
+      "step": 38427
+    },
+    {
+      "epoch": 0.9146910406550509,
+      "grad_norm": 4.6979524115103795,
+      "learning_rate": 2e-05,
+      "loss": 0.7153,
+      "step": 38428
+    },
+    {
+      "epoch": 0.9147148433780824,
+      "grad_norm": 0.004138300711907682,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 38429
+    },
+    {
+      "epoch": 0.914738646101114,
+      "grad_norm": 4.329398764157836,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 38430
+    },
+    {
+      "epoch": 0.9147624488241455,
+      "grad_norm": 5.160409042673265,
+      "learning_rate": 2e-05,
+      "loss": 1.2335,
+      "step": 38431
+    },
+    {
+      "epoch": 0.914786251547177,
+      "grad_norm": 2.538542136705792,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 38432
+    },
+    {
+      "epoch": 0.9148100542702086,
+      "grad_norm": 3.675601013608986,
+      "learning_rate": 2e-05,
+      "loss": 1.069,
+      "step": 38433
+    },
+    {
+      "epoch": 0.91483385699324,
+      "grad_norm": 4.3016070659767704,
+      "learning_rate": 2e-05,
+      "loss": 0.6288,
+      "step": 38434
+    },
+    {
+      "epoch": 0.9148576597162715,
+      "grad_norm": 0.3854702169488699,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 38435
+    },
+    {
+      "epoch": 0.914881462439303,
+      "grad_norm": 2.190235591384474,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 38436
+    },
+    {
+      "epoch": 0.9149052651623346,
+      "grad_norm": 2.3160082790272236,
+      "learning_rate": 2e-05,
+      "loss": 0.0493,
+      "step": 38437
+    },
+    {
+      "epoch": 0.9149290678853661,
+      "grad_norm": 3.1620813587253362,
+      "learning_rate": 2e-05,
+      "loss": 0.5292,
+      "step": 38438
+    },
+    {
+      "epoch": 0.9149528706083976,
+      "grad_norm": 0.054365619103120495,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 38439
+    },
+    {
+      "epoch": 0.9149766733314291,
+      "grad_norm": 4.964624478562738,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 38440
+    },
+    {
+      "epoch": 0.9150004760544607,
+      "grad_norm": 1.5845499696386223,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 38441
+    },
+    {
+      "epoch": 0.9150242787774922,
+      "grad_norm": 2.7603614640032648,
+      "learning_rate": 2e-05,
+      "loss": 0.3831,
+      "step": 38442
+    },
+    {
+      "epoch": 0.9150480815005236,
+      "grad_norm": 0.38929952883223956,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 38443
+    },
+    {
+      "epoch": 0.9150718842235551,
+      "grad_norm": 0.23037450633368775,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 38444
+    },
+    {
+      "epoch": 0.9150956869465867,
+      "grad_norm": 0.12655710606816734,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 38445
+    },
+    {
+      "epoch": 0.9151194896696182,
+      "grad_norm": 28.130697457521496,
+      "learning_rate": 2e-05,
+      "loss": 0.6746,
+      "step": 38446
+    },
+    {
+      "epoch": 0.9151432923926497,
+      "grad_norm": 2.6948861851271944,
+      "learning_rate": 2e-05,
+      "loss": 0.0549,
+      "step": 38447
+    },
+    {
+      "epoch": 0.9151670951156813,
+      "grad_norm": 8.268589569870233,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 38448
+    },
+    {
+      "epoch": 0.9151908978387128,
+      "grad_norm": 4.397938833947391,
+      "learning_rate": 2e-05,
+      "loss": 0.696,
+      "step": 38449
+    },
+    {
+      "epoch": 0.9152147005617443,
+      "grad_norm": 4.585047456738576,
+      "learning_rate": 2e-05,
+      "loss": 0.2272,
+      "step": 38450
+    },
+    {
+      "epoch": 0.9152385032847757,
+      "grad_norm": 4.753468408370111,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 38451
+    },
+    {
+      "epoch": 0.9152623060078073,
+      "grad_norm": 0.5024326725693694,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 38452
+    },
+    {
+      "epoch": 0.9152861087308388,
+      "grad_norm": 4.514685441391185,
+      "learning_rate": 2e-05,
+      "loss": 0.6773,
+      "step": 38453
+    },
+    {
+      "epoch": 0.9153099114538703,
+      "grad_norm": 4.021189239363753,
+      "learning_rate": 2e-05,
+      "loss": 0.4801,
+      "step": 38454
+    },
+    {
+      "epoch": 0.9153337141769018,
+      "grad_norm": 3.7433388203590825,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 38455
+    },
+    {
+      "epoch": 0.9153575168999334,
+      "grad_norm": 19.156831203472077,
+      "learning_rate": 2e-05,
+      "loss": 0.5602,
+      "step": 38456
+    },
+    {
+      "epoch": 0.9153813196229649,
+      "grad_norm": 1.3307314507474053,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 38457
+    },
+    {
+      "epoch": 0.9154051223459964,
+      "grad_norm": 2.7029269877335316,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 38458
+    },
+    {
+      "epoch": 0.9154289250690278,
+      "grad_norm": 1.3959281454283918,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 38459
+    },
+    {
+      "epoch": 0.9154527277920594,
+      "grad_norm": 3.946591148830651,
+      "learning_rate": 2e-05,
+      "loss": 0.7799,
+      "step": 38460
+    },
+    {
+      "epoch": 0.9154765305150909,
+      "grad_norm": 4.176917117893064,
+      "learning_rate": 2e-05,
+      "loss": 0.626,
+      "step": 38461
+    },
+    {
+      "epoch": 0.9155003332381224,
+      "grad_norm": 2.8155215592869327,
+      "learning_rate": 2e-05,
+      "loss": 0.3727,
+      "step": 38462
+    },
+    {
+      "epoch": 0.915524135961154,
+      "grad_norm": 7.121489112211958,
+      "learning_rate": 2e-05,
+      "loss": 0.2806,
+      "step": 38463
+    },
+    {
+      "epoch": 0.9155479386841855,
+      "grad_norm": 0.10875027221479694,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 38464
+    },
+    {
+      "epoch": 0.915571741407217,
+      "grad_norm": 3.0313628321606054,
+      "learning_rate": 2e-05,
+      "loss": 0.3075,
+      "step": 38465
+    },
+    {
+      "epoch": 0.9155955441302485,
+      "grad_norm": 0.8968042114860273,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 38466
+    },
+    {
+      "epoch": 0.9156193468532801,
+      "grad_norm": 3.826340340983602,
+      "learning_rate": 2e-05,
+      "loss": 0.5184,
+      "step": 38467
+    },
+    {
+      "epoch": 0.9156431495763115,
+      "grad_norm": 11.62836211162146,
+      "learning_rate": 2e-05,
+      "loss": 0.6121,
+      "step": 38468
+    },
+    {
+      "epoch": 0.915666952299343,
+      "grad_norm": 1.6266556180613918,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 38469
+    },
+    {
+      "epoch": 0.9156907550223745,
+      "grad_norm": 7.264810388294504,
+      "learning_rate": 2e-05,
+      "loss": 0.2712,
+      "step": 38470
+    },
+    {
+      "epoch": 0.9157145577454061,
+      "grad_norm": 10.821274628769018,
+      "learning_rate": 2e-05,
+      "loss": 1.4185,
+      "step": 38471
+    },
+    {
+      "epoch": 0.9157383604684376,
+      "grad_norm": 0.2418658094126479,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 38472
+    },
+    {
+      "epoch": 0.9157621631914691,
+      "grad_norm": 3.601069751793295,
+      "learning_rate": 2e-05,
+      "loss": 0.0801,
+      "step": 38473
+    },
+    {
+      "epoch": 0.9157859659145006,
+      "grad_norm": 4.328643285260661,
+      "learning_rate": 2e-05,
+      "loss": 0.8223,
+      "step": 38474
+    },
+    {
+      "epoch": 0.9158097686375322,
+      "grad_norm": 16.427714537169056,
+      "learning_rate": 2e-05,
+      "loss": 0.5248,
+      "step": 38475
+    },
+    {
+      "epoch": 0.9158335713605636,
+      "grad_norm": 0.07653864751847593,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38476
+    },
+    {
+      "epoch": 0.9158573740835951,
+      "grad_norm": 9.97219424671055,
+      "learning_rate": 2e-05,
+      "loss": 1.1761,
+      "step": 38477
+    },
+    {
+      "epoch": 0.9158811768066267,
+      "grad_norm": 7.885290567137755,
+      "learning_rate": 2e-05,
+      "loss": 2.1244,
+      "step": 38478
+    },
+    {
+      "epoch": 0.9159049795296582,
+      "grad_norm": 4.995188432789644,
+      "learning_rate": 2e-05,
+      "loss": 0.5118,
+      "step": 38479
+    },
+    {
+      "epoch": 0.9159287822526897,
+      "grad_norm": 1.0143040459715353,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 38480
+    },
+    {
+      "epoch": 0.9159525849757212,
+      "grad_norm": 2.118597958510224,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 38481
+    },
+    {
+      "epoch": 0.9159763876987528,
+      "grad_norm": 3.213844584868008,
+      "learning_rate": 2e-05,
+      "loss": 0.4275,
+      "step": 38482
+    },
+    {
+      "epoch": 0.9160001904217843,
+      "grad_norm": 1.028106124391077,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 38483
+    },
+    {
+      "epoch": 0.9160239931448158,
+      "grad_norm": 0.13605846981182468,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 38484
+    },
+    {
+      "epoch": 0.9160477958678472,
+      "grad_norm": 2.5917984034113184,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 38485
+    },
+    {
+      "epoch": 0.9160715985908788,
+      "grad_norm": 5.1838446220635666,
+      "learning_rate": 2e-05,
+      "loss": 1.2643,
+      "step": 38486
+    },
+    {
+      "epoch": 0.9160954013139103,
+      "grad_norm": 3.923554983902355,
+      "learning_rate": 2e-05,
+      "loss": 0.6367,
+      "step": 38487
+    },
+    {
+      "epoch": 0.9161192040369418,
+      "grad_norm": 3.5077325985999668,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 38488
+    },
+    {
+      "epoch": 0.9161430067599733,
+      "grad_norm": 2.6998715559567437,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 38489
+    },
+    {
+      "epoch": 0.9161668094830049,
+      "grad_norm": 5.187210164396615,
+      "learning_rate": 2e-05,
+      "loss": 0.4558,
+      "step": 38490
+    },
+    {
+      "epoch": 0.9161906122060364,
+      "grad_norm": 3.8401169701643663,
+      "learning_rate": 2e-05,
+      "loss": 0.5838,
+      "step": 38491
+    },
+    {
+      "epoch": 0.9162144149290679,
+      "grad_norm": 1.4672091261866298,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 38492
+    },
+    {
+      "epoch": 0.9162382176520995,
+      "grad_norm": 4.534561693339088,
+      "learning_rate": 2e-05,
+      "loss": 0.5373,
+      "step": 38493
+    },
+    {
+      "epoch": 0.9162620203751309,
+      "grad_norm": 1.017253888519923,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 38494
+    },
+    {
+      "epoch": 0.9162858230981624,
+      "grad_norm": 12.571221768188895,
+      "learning_rate": 2e-05,
+      "loss": 1.0878,
+      "step": 38495
+    },
+    {
+      "epoch": 0.9163096258211939,
+      "grad_norm": 5.86166017239295,
+      "learning_rate": 2e-05,
+      "loss": 0.746,
+      "step": 38496
+    },
+    {
+      "epoch": 0.9163334285442255,
+      "grad_norm": 2.836939290312561,
+      "learning_rate": 2e-05,
+      "loss": 0.2205,
+      "step": 38497
+    },
+    {
+      "epoch": 0.916357231267257,
+      "grad_norm": 3.1474931566331685,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 38498
+    },
+    {
+      "epoch": 0.9163810339902885,
+      "grad_norm": 0.7531192766470082,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 38499
+    },
+    {
+      "epoch": 0.91640483671332,
+      "grad_norm": 0.12453830648013747,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 38500
+    },
+    {
+      "epoch": 0.9164286394363516,
+      "grad_norm": 1.5279416952924318,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 38501
+    },
+    {
+      "epoch": 0.916452442159383,
+      "grad_norm": 5.504048072543643,
+      "learning_rate": 2e-05,
+      "loss": 1.2581,
+      "step": 38502
+    },
+    {
+      "epoch": 0.9164762448824145,
+      "grad_norm": 2.8792641196232998,
+      "learning_rate": 2e-05,
+      "loss": 0.4876,
+      "step": 38503
+    },
+    {
+      "epoch": 0.916500047605446,
+      "grad_norm": 3.9748504286038417,
+      "learning_rate": 2e-05,
+      "loss": 0.4647,
+      "step": 38504
+    },
+    {
+      "epoch": 0.9165238503284776,
+      "grad_norm": 4.23701277839917,
+      "learning_rate": 2e-05,
+      "loss": 0.7656,
+      "step": 38505
+    },
+    {
+      "epoch": 0.9165476530515091,
+      "grad_norm": 6.09593600486975,
+      "learning_rate": 2e-05,
+      "loss": 0.6635,
+      "step": 38506
+    },
+    {
+      "epoch": 0.9165714557745406,
+      "grad_norm": 36.07808363866478,
+      "learning_rate": 2e-05,
+      "loss": 0.1312,
+      "step": 38507
+    },
+    {
+      "epoch": 0.9165952584975722,
+      "grad_norm": 3.41987563067977,
+      "learning_rate": 2e-05,
+      "loss": 0.2517,
+      "step": 38508
+    },
+    {
+      "epoch": 0.9166190612206037,
+      "grad_norm": 0.2803934468800587,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 38509
+    },
+    {
+      "epoch": 0.9166428639436351,
+      "grad_norm": 4.272463302494627,
+      "learning_rate": 2e-05,
+      "loss": 0.6852,
+      "step": 38510
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 3.1404891158059813,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 38511
+    },
+    {
+      "epoch": 0.9166904693896982,
+      "grad_norm": 2.165356936585235,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 38512
+    },
+    {
+      "epoch": 0.9167142721127297,
+      "grad_norm": 8.58823825973066,
+      "learning_rate": 2e-05,
+      "loss": 0.6926,
+      "step": 38513
+    },
+    {
+      "epoch": 0.9167380748357612,
+      "grad_norm": 5.518670541345351,
+      "learning_rate": 2e-05,
+      "loss": 1.1738,
+      "step": 38514
+    },
+    {
+      "epoch": 0.9167618775587927,
+      "grad_norm": 3.5494173755666116,
+      "learning_rate": 2e-05,
+      "loss": 0.3116,
+      "step": 38515
+    },
+    {
+      "epoch": 0.9167856802818243,
+      "grad_norm": 5.7824555475109936,
+      "learning_rate": 2e-05,
+      "loss": 0.9897,
+      "step": 38516
+    },
+    {
+      "epoch": 0.9168094830048558,
+      "grad_norm": 3.5098527295760946,
+      "learning_rate": 2e-05,
+      "loss": 0.4417,
+      "step": 38517
+    },
+    {
+      "epoch": 0.9168332857278872,
+      "grad_norm": 0.5209237359533768,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 38518
+    },
+    {
+      "epoch": 0.9168570884509187,
+      "grad_norm": 4.750501744636362,
+      "learning_rate": 2e-05,
+      "loss": 0.7865,
+      "step": 38519
+    },
+    {
+      "epoch": 0.9168808911739503,
+      "grad_norm": 0.2894013915074184,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 38520
+    },
+    {
+      "epoch": 0.9169046938969818,
+      "grad_norm": 3.7317144789863868,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 38521
+    },
+    {
+      "epoch": 0.9169284966200133,
+      "grad_norm": 8.793681001373148,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 38522
+    },
+    {
+      "epoch": 0.9169522993430449,
+      "grad_norm": 4.208917582824764,
+      "learning_rate": 2e-05,
+      "loss": 0.6924,
+      "step": 38523
+    },
+    {
+      "epoch": 0.9169761020660764,
+      "grad_norm": 4.092585550909729,
+      "learning_rate": 2e-05,
+      "loss": 0.4459,
+      "step": 38524
+    },
+    {
+      "epoch": 0.9169999047891079,
+      "grad_norm": 7.0083417292351315,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 38525
+    },
+    {
+      "epoch": 0.9170237075121394,
+      "grad_norm": 2.6485994812187044,
+      "learning_rate": 2e-05,
+      "loss": 0.2545,
+      "step": 38526
+    },
+    {
+      "epoch": 0.917047510235171,
+      "grad_norm": 3.60004455664099,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 38527
+    },
+    {
+      "epoch": 0.9170713129582024,
+      "grad_norm": 4.433946404373703,
+      "learning_rate": 2e-05,
+      "loss": 0.1344,
+      "step": 38528
+    },
+    {
+      "epoch": 0.9170951156812339,
+      "grad_norm": 0.5455466520591233,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 38529
+    },
+    {
+      "epoch": 0.9171189184042654,
+      "grad_norm": 3.7339118048459707,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 38530
+    },
+    {
+      "epoch": 0.917142721127297,
+      "grad_norm": 0.5592157891105437,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 38531
+    },
+    {
+      "epoch": 0.9171665238503285,
+      "grad_norm": 3.1013175110715796,
+      "learning_rate": 2e-05,
+      "loss": 0.4649,
+      "step": 38532
+    },
+    {
+      "epoch": 0.91719032657336,
+      "grad_norm": 0.2042349954020573,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 38533
+    },
+    {
+      "epoch": 0.9172141292963915,
+      "grad_norm": 2.828753433922226,
+      "learning_rate": 2e-05,
+      "loss": 0.3699,
+      "step": 38534
+    },
+    {
+      "epoch": 0.917237932019423,
+      "grad_norm": 4.736137836950142,
+      "learning_rate": 2e-05,
+      "loss": 0.6961,
+      "step": 38535
+    },
+    {
+      "epoch": 0.9172617347424545,
+      "grad_norm": 10.707412796736897,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 38536
+    },
+    {
+      "epoch": 0.917285537465486,
+      "grad_norm": 0.228063251122681,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 38537
+    },
+    {
+      "epoch": 0.9173093401885176,
+      "grad_norm": 2.8719470408535543,
+      "learning_rate": 2e-05,
+      "loss": 0.5276,
+      "step": 38538
+    },
+    {
+      "epoch": 0.9173331429115491,
+      "grad_norm": 0.9671049268999969,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 38539
+    },
+    {
+      "epoch": 0.9173569456345806,
+      "grad_norm": 0.04588586087834566,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38540
+    },
+    {
+      "epoch": 0.9173807483576121,
+      "grad_norm": 1.3776345478721503,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 38541
+    },
+    {
+      "epoch": 0.9174045510806437,
+      "grad_norm": 0.2067908082371984,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 38542
+    },
+    {
+      "epoch": 0.9174283538036752,
+      "grad_norm": 8.855540676411085,
+      "learning_rate": 2e-05,
+      "loss": 0.1222,
+      "step": 38543
+    },
+    {
+      "epoch": 0.9174521565267066,
+      "grad_norm": 3.9341227672310346,
+      "learning_rate": 2e-05,
+      "loss": 0.6007,
+      "step": 38544
+    },
+    {
+      "epoch": 0.9174759592497381,
+      "grad_norm": 7.319913774813897,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 38545
+    },
+    {
+      "epoch": 0.9174997619727697,
+      "grad_norm": 3.0993472361403382,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 38546
+    },
+    {
+      "epoch": 0.9175235646958012,
+      "grad_norm": 5.644708365655038,
+      "learning_rate": 2e-05,
+      "loss": 0.6778,
+      "step": 38547
+    },
+    {
+      "epoch": 0.9175473674188327,
+      "grad_norm": 1.779307544972619,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 38548
+    },
+    {
+      "epoch": 0.9175711701418642,
+      "grad_norm": 3.3871601738784904,
+      "learning_rate": 2e-05,
+      "loss": 0.0878,
+      "step": 38549
+    },
+    {
+      "epoch": 0.9175949728648958,
+      "grad_norm": 6.313996985325637,
+      "learning_rate": 2e-05,
+      "loss": 0.5487,
+      "step": 38550
+    },
+    {
+      "epoch": 0.9176187755879273,
+      "grad_norm": 0.078276544024183,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 38551
+    },
+    {
+      "epoch": 0.9176425783109587,
+      "grad_norm": 4.798944173291937,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 38552
+    },
+    {
+      "epoch": 0.9176663810339903,
+      "grad_norm": 0.22483725803412413,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 38553
+    },
+    {
+      "epoch": 0.9176901837570218,
+      "grad_norm": 0.039268025616397806,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 38554
+    },
+    {
+      "epoch": 0.9177139864800533,
+      "grad_norm": 5.630555178551614,
+      "learning_rate": 2e-05,
+      "loss": 0.3865,
+      "step": 38555
+    },
+    {
+      "epoch": 0.9177377892030848,
+      "grad_norm": 3.08968268028414,
+      "learning_rate": 2e-05,
+      "loss": 0.2345,
+      "step": 38556
+    },
+    {
+      "epoch": 0.9177615919261164,
+      "grad_norm": 0.780041787116896,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 38557
+    },
+    {
+      "epoch": 0.9177853946491479,
+      "grad_norm": 0.26229497543449787,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 38558
+    },
+    {
+      "epoch": 0.9178091973721794,
+      "grad_norm": 2.186289645327642,
+      "learning_rate": 2e-05,
+      "loss": 0.313,
+      "step": 38559
+    },
+    {
+      "epoch": 0.9178330000952108,
+      "grad_norm": 3.8804745683726325,
+      "learning_rate": 2e-05,
+      "loss": 0.6172,
+      "step": 38560
+    },
+    {
+      "epoch": 0.9178568028182424,
+      "grad_norm": 0.29503555496694017,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38561
+    },
+    {
+      "epoch": 0.9178806055412739,
+      "grad_norm": 3.3369037594609297,
+      "learning_rate": 2e-05,
+      "loss": 0.3711,
+      "step": 38562
+    },
+    {
+      "epoch": 0.9179044082643054,
+      "grad_norm": 2.167584148228943,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 38563
+    },
+    {
+      "epoch": 0.9179282109873369,
+      "grad_norm": 3.952118458467338,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 38564
+    },
+    {
+      "epoch": 0.9179520137103685,
+      "grad_norm": 11.39922695161813,
+      "learning_rate": 2e-05,
+      "loss": 1.3202,
+      "step": 38565
+    },
+    {
+      "epoch": 0.9179758164334,
+      "grad_norm": 2.3478963485183595,
+      "learning_rate": 2e-05,
+      "loss": 0.2068,
+      "step": 38566
+    },
+    {
+      "epoch": 0.9179996191564315,
+      "grad_norm": 0.05672849719318355,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 38567
+    },
+    {
+      "epoch": 0.9180234218794631,
+      "grad_norm": 1.9966381715757424,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 38568
+    },
+    {
+      "epoch": 0.9180472246024945,
+      "grad_norm": 3.0657941476621002,
+      "learning_rate": 2e-05,
+      "loss": 0.3221,
+      "step": 38569
+    },
+    {
+      "epoch": 0.918071027325526,
+      "grad_norm": 1.7999284136893485,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 38570
+    },
+    {
+      "epoch": 0.9180948300485575,
+      "grad_norm": 3.7288289341451506,
+      "learning_rate": 2e-05,
+      "loss": 0.1199,
+      "step": 38571
+    },
+    {
+      "epoch": 0.9181186327715891,
+      "grad_norm": 15.551140557095458,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 38572
+    },
+    {
+      "epoch": 0.9181424354946206,
+      "grad_norm": 2.3121360882220996,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 38573
+    },
+    {
+      "epoch": 0.9181662382176521,
+      "grad_norm": 0.011321375635875687,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 38574
+    },
+    {
+      "epoch": 0.9181900409406836,
+      "grad_norm": 3.2143027187275246,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 38575
+    },
+    {
+      "epoch": 0.9182138436637152,
+      "grad_norm": 0.9039212608734544,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 38576
+    },
+    {
+      "epoch": 0.9182376463867467,
+      "grad_norm": 6.560635083658891,
+      "learning_rate": 2e-05,
+      "loss": 0.6277,
+      "step": 38577
+    },
+    {
+      "epoch": 0.9182614491097781,
+      "grad_norm": 0.20393933282649385,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 38578
+    },
+    {
+      "epoch": 0.9182852518328096,
+      "grad_norm": 6.032219594032401,
+      "learning_rate": 2e-05,
+      "loss": 0.8642,
+      "step": 38579
+    },
+    {
+      "epoch": 0.9183090545558412,
+      "grad_norm": 0.012537162805743914,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 38580
+    },
+    {
+      "epoch": 0.9183328572788727,
+      "grad_norm": 13.23986292418746,
+      "learning_rate": 2e-05,
+      "loss": 1.5234,
+      "step": 38581
+    },
+    {
+      "epoch": 0.9183566600019042,
+      "grad_norm": 0.014695845461925946,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 38582
+    },
+    {
+      "epoch": 0.9183804627249358,
+      "grad_norm": 0.05782760573816562,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 38583
+    },
+    {
+      "epoch": 0.9184042654479673,
+      "grad_norm": 3.2039853680691532,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 38584
+    },
+    {
+      "epoch": 0.9184280681709988,
+      "grad_norm": 1.005289738658147,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 38585
+    },
+    {
+      "epoch": 0.9184518708940302,
+      "grad_norm": 0.15192797637872713,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 38586
+    },
+    {
+      "epoch": 0.9184756736170618,
+      "grad_norm": 1.9970218217743543,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 38587
+    },
+    {
+      "epoch": 0.9184994763400933,
+      "grad_norm": 2.358252836896262,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 38588
+    },
+    {
+      "epoch": 0.9185232790631248,
+      "grad_norm": 0.11137877395574951,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38589
+    },
+    {
+      "epoch": 0.9185470817861563,
+      "grad_norm": 24.730213195264604,
+      "learning_rate": 2e-05,
+      "loss": 1.5542,
+      "step": 38590
+    },
+    {
+      "epoch": 0.9185708845091879,
+      "grad_norm": 3.0914368063173026,
+      "learning_rate": 2e-05,
+      "loss": 0.6993,
+      "step": 38591
+    },
+    {
+      "epoch": 0.9185946872322194,
+      "grad_norm": 7.664019235722611,
+      "learning_rate": 2e-05,
+      "loss": 1.0517,
+      "step": 38592
+    },
+    {
+      "epoch": 0.9186184899552509,
+      "grad_norm": 6.726519057011916,
+      "learning_rate": 2e-05,
+      "loss": 0.7479,
+      "step": 38593
+    },
+    {
+      "epoch": 0.9186422926782823,
+      "grad_norm": 3.268705717022353,
+      "learning_rate": 2e-05,
+      "loss": 0.1949,
+      "step": 38594
+    },
+    {
+      "epoch": 0.9186660954013139,
+      "grad_norm": 3.6376223311881413,
+      "learning_rate": 2e-05,
+      "loss": 0.6215,
+      "step": 38595
+    },
+    {
+      "epoch": 0.9186898981243454,
+      "grad_norm": 9.847567730857133,
+      "learning_rate": 2e-05,
+      "loss": 1.0953,
+      "step": 38596
+    },
+    {
+      "epoch": 0.9187137008473769,
+      "grad_norm": 6.964741234529438,
+      "learning_rate": 2e-05,
+      "loss": 1.3593,
+      "step": 38597
+    },
+    {
+      "epoch": 0.9187375035704085,
+      "grad_norm": 12.567736956785867,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 38598
+    },
+    {
+      "epoch": 0.91876130629344,
+      "grad_norm": 0.06405669150281534,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38599
+    },
+    {
+      "epoch": 0.9187851090164715,
+      "grad_norm": 3.271434665161585,
+      "learning_rate": 2e-05,
+      "loss": 0.2691,
+      "step": 38600
+    },
+    {
+      "epoch": 0.918808911739503,
+      "grad_norm": 1.5804515647320678,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 38601
+    },
+    {
+      "epoch": 0.9188327144625346,
+      "grad_norm": 3.982439710328878,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 38602
+    },
+    {
+      "epoch": 0.918856517185566,
+      "grad_norm": 3.8606159320282005,
+      "learning_rate": 2e-05,
+      "loss": 0.578,
+      "step": 38603
+    },
+    {
+      "epoch": 0.9188803199085975,
+      "grad_norm": 3.428053180819696,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 38604
+    },
+    {
+      "epoch": 0.918904122631629,
+      "grad_norm": 4.201475920473723,
+      "learning_rate": 2e-05,
+      "loss": 0.6827,
+      "step": 38605
+    },
+    {
+      "epoch": 0.9189279253546606,
+      "grad_norm": 0.2896836295454939,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38606
+    },
+    {
+      "epoch": 0.9189517280776921,
+      "grad_norm": 0.13026349937337903,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 38607
+    },
+    {
+      "epoch": 0.9189755308007236,
+      "grad_norm": 4.210104758972794,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 38608
+    },
+    {
+      "epoch": 0.9189993335237551,
+      "grad_norm": 3.927722113791388,
+      "learning_rate": 2e-05,
+      "loss": 0.6975,
+      "step": 38609
+    },
+    {
+      "epoch": 0.9190231362467867,
+      "grad_norm": 5.140871800862335,
+      "learning_rate": 2e-05,
+      "loss": 0.1984,
+      "step": 38610
+    },
+    {
+      "epoch": 0.9190469389698182,
+      "grad_norm": 5.968920751343959,
+      "learning_rate": 2e-05,
+      "loss": 0.2658,
+      "step": 38611
+    },
+    {
+      "epoch": 0.9190707416928496,
+      "grad_norm": 0.5158058703722463,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 38612
+    },
+    {
+      "epoch": 0.9190945444158812,
+      "grad_norm": 5.0378627063615085,
+      "learning_rate": 2e-05,
+      "loss": 0.1305,
+      "step": 38613
+    },
+    {
+      "epoch": 0.9191183471389127,
+      "grad_norm": 7.849556515163043,
+      "learning_rate": 2e-05,
+      "loss": 0.2396,
+      "step": 38614
+    },
+    {
+      "epoch": 0.9191421498619442,
+      "grad_norm": 0.04908444575554171,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 38615
+    },
+    {
+      "epoch": 0.9191659525849757,
+      "grad_norm": 6.428589481156849,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 38616
+    },
+    {
+      "epoch": 0.9191897553080073,
+      "grad_norm": 3.0095871869324617,
+      "learning_rate": 2e-05,
+      "loss": 0.0354,
+      "step": 38617
+    },
+    {
+      "epoch": 0.9192135580310388,
+      "grad_norm": 2.4327969940436134,
+      "learning_rate": 2e-05,
+      "loss": 0.3391,
+      "step": 38618
+    },
+    {
+      "epoch": 0.9192373607540703,
+      "grad_norm": 16.962792461078827,
+      "learning_rate": 2e-05,
+      "loss": 1.7423,
+      "step": 38619
+    },
+    {
+      "epoch": 0.9192611634771017,
+      "grad_norm": 3.1688199931939773,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 38620
+    },
+    {
+      "epoch": 0.9192849662001333,
+      "grad_norm": 4.15427472774291,
+      "learning_rate": 2e-05,
+      "loss": 0.5661,
+      "step": 38621
+    },
+    {
+      "epoch": 0.9193087689231648,
+      "grad_norm": 4.0663012857286,
+      "learning_rate": 2e-05,
+      "loss": 0.9219,
+      "step": 38622
+    },
+    {
+      "epoch": 0.9193325716461963,
+      "grad_norm": 3.0532375109341436,
+      "learning_rate": 2e-05,
+      "loss": 0.6347,
+      "step": 38623
+    },
+    {
+      "epoch": 0.9193563743692278,
+      "grad_norm": 5.529711959521245,
+      "learning_rate": 2e-05,
+      "loss": 0.318,
+      "step": 38624
+    },
+    {
+      "epoch": 0.9193801770922594,
+      "grad_norm": 3.883962466733609,
+      "learning_rate": 2e-05,
+      "loss": 0.5102,
+      "step": 38625
+    },
+    {
+      "epoch": 0.9194039798152909,
+      "grad_norm": 1.0870361335543512,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 38626
+    },
+    {
+      "epoch": 0.9194277825383224,
+      "grad_norm": 14.621584953866455,
+      "learning_rate": 2e-05,
+      "loss": 0.9597,
+      "step": 38627
+    },
+    {
+      "epoch": 0.919451585261354,
+      "grad_norm": 2.3137079460750627,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 38628
+    },
+    {
+      "epoch": 0.9194753879843854,
+      "grad_norm": 1.3826051651409226,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 38629
+    },
+    {
+      "epoch": 0.9194991907074169,
+      "grad_norm": 3.6494543743965933,
+      "learning_rate": 2e-05,
+      "loss": 0.4601,
+      "step": 38630
+    },
+    {
+      "epoch": 0.9195229934304484,
+      "grad_norm": 5.682441071889753,
+      "learning_rate": 2e-05,
+      "loss": 0.8781,
+      "step": 38631
+    },
+    {
+      "epoch": 0.91954679615348,
+      "grad_norm": 0.07598566962361473,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38632
+    },
+    {
+      "epoch": 0.9195705988765115,
+      "grad_norm": 5.365713318732938,
+      "learning_rate": 2e-05,
+      "loss": 0.5667,
+      "step": 38633
+    },
+    {
+      "epoch": 0.919594401599543,
+      "grad_norm": 0.1251604403900471,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 38634
+    },
+    {
+      "epoch": 0.9196182043225745,
+      "grad_norm": 9.901212149704811,
+      "learning_rate": 2e-05,
+      "loss": 0.541,
+      "step": 38635
+    },
+    {
+      "epoch": 0.9196420070456061,
+      "grad_norm": 2.2140344519866098,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 38636
+    },
+    {
+      "epoch": 0.9196658097686375,
+      "grad_norm": 1.6819007097640193,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 38637
+    },
+    {
+      "epoch": 0.919689612491669,
+      "grad_norm": 2.6789783667642864,
+      "learning_rate": 2e-05,
+      "loss": 0.3213,
+      "step": 38638
+    },
+    {
+      "epoch": 0.9197134152147005,
+      "grad_norm": 0.07389804942378396,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38639
+    },
+    {
+      "epoch": 0.9197372179377321,
+      "grad_norm": 3.497463585282278,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 38640
+    },
+    {
+      "epoch": 0.9197610206607636,
+      "grad_norm": 4.148254487137623,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 38641
+    },
+    {
+      "epoch": 0.9197848233837951,
+      "grad_norm": 0.2806269426505603,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 38642
+    },
+    {
+      "epoch": 0.9198086261068266,
+      "grad_norm": 7.217238485457039,
+      "learning_rate": 2e-05,
+      "loss": 0.8222,
+      "step": 38643
+    },
+    {
+      "epoch": 0.9198324288298582,
+      "grad_norm": 1.5405170656562497,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 38644
+    },
+    {
+      "epoch": 0.9198562315528896,
+      "grad_norm": 5.755573244519754,
+      "learning_rate": 2e-05,
+      "loss": 0.3917,
+      "step": 38645
+    },
+    {
+      "epoch": 0.9198800342759211,
+      "grad_norm": 2.7452136625400487,
+      "learning_rate": 2e-05,
+      "loss": 0.5835,
+      "step": 38646
+    },
+    {
+      "epoch": 0.9199038369989527,
+      "grad_norm": 1.8546289046615831,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 38647
+    },
+    {
+      "epoch": 0.9199276397219842,
+      "grad_norm": 3.5342358021905484,
+      "learning_rate": 2e-05,
+      "loss": 0.0778,
+      "step": 38648
+    },
+    {
+      "epoch": 0.9199514424450157,
+      "grad_norm": 7.186755925900432,
+      "learning_rate": 2e-05,
+      "loss": 0.971,
+      "step": 38649
+    },
+    {
+      "epoch": 0.9199752451680472,
+      "grad_norm": 4.95549059502597,
+      "learning_rate": 2e-05,
+      "loss": 0.1855,
+      "step": 38650
+    },
+    {
+      "epoch": 0.9199990478910788,
+      "grad_norm": 3.159177508868725,
+      "learning_rate": 2e-05,
+      "loss": 0.5997,
+      "step": 38651
+    },
+    {
+      "epoch": 0.9200228506141103,
+      "grad_norm": 3.8408295485023665,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 38652
+    },
+    {
+      "epoch": 0.9200466533371418,
+      "grad_norm": 0.022476443721189183,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 38653
+    },
+    {
+      "epoch": 0.9200704560601732,
+      "grad_norm": 9.298082953291232,
+      "learning_rate": 2e-05,
+      "loss": 0.2023,
+      "step": 38654
+    },
+    {
+      "epoch": 0.9200942587832048,
+      "grad_norm": 0.7784802302708145,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 38655
+    },
+    {
+      "epoch": 0.9201180615062363,
+      "grad_norm": 0.03214316079070413,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 38656
+    },
+    {
+      "epoch": 0.9201418642292678,
+      "grad_norm": 4.286132196850231,
+      "learning_rate": 2e-05,
+      "loss": 0.8633,
+      "step": 38657
+    },
+    {
+      "epoch": 0.9201656669522993,
+      "grad_norm": 2.61341268108658,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 38658
+    },
+    {
+      "epoch": 0.9201894696753309,
+      "grad_norm": 0.1004978053730138,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 38659
+    },
+    {
+      "epoch": 0.9202132723983624,
+      "grad_norm": 0.2778456763903238,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 38660
+    },
+    {
+      "epoch": 0.9202370751213939,
+      "grad_norm": 6.6486204168179714,
+      "learning_rate": 2e-05,
+      "loss": 0.5392,
+      "step": 38661
+    },
+    {
+      "epoch": 0.9202608778444255,
+      "grad_norm": 4.164219357822124,
+      "learning_rate": 2e-05,
+      "loss": 0.2848,
+      "step": 38662
+    },
+    {
+      "epoch": 0.9202846805674569,
+      "grad_norm": 0.03158797989770465,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 38663
+    },
+    {
+      "epoch": 0.9203084832904884,
+      "grad_norm": 4.657544717171115,
+      "learning_rate": 2e-05,
+      "loss": 0.4863,
+      "step": 38664
+    },
+    {
+      "epoch": 0.9203322860135199,
+      "grad_norm": 5.457311653703904,
+      "learning_rate": 2e-05,
+      "loss": 0.5875,
+      "step": 38665
+    },
+    {
+      "epoch": 0.9203560887365515,
+      "grad_norm": 4.124309758103546,
+      "learning_rate": 2e-05,
+      "loss": 0.5794,
+      "step": 38666
+    },
+    {
+      "epoch": 0.920379891459583,
+      "grad_norm": 2.554905822758776,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 38667
+    },
+    {
+      "epoch": 0.9204036941826145,
+      "grad_norm": 3.4956362774398912,
+      "learning_rate": 2e-05,
+      "loss": 0.5914,
+      "step": 38668
+    },
+    {
+      "epoch": 0.920427496905646,
+      "grad_norm": 3.6005913681487978,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 38669
+    },
+    {
+      "epoch": 0.9204512996286776,
+      "grad_norm": 8.81137835177189,
+      "learning_rate": 2e-05,
+      "loss": 0.4622,
+      "step": 38670
+    },
+    {
+      "epoch": 0.920475102351709,
+      "grad_norm": 0.3755024468914118,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 38671
+    },
+    {
+      "epoch": 0.9204989050747405,
+      "grad_norm": 4.551170653898644,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 38672
+    },
+    {
+      "epoch": 0.920522707797772,
+      "grad_norm": 3.10487133166284,
+      "learning_rate": 2e-05,
+      "loss": 0.389,
+      "step": 38673
+    },
+    {
+      "epoch": 0.9205465105208036,
+      "grad_norm": 2.7396763640009247,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 38674
+    },
+    {
+      "epoch": 0.9205703132438351,
+      "grad_norm": 4.229038220805743,
+      "learning_rate": 2e-05,
+      "loss": 0.552,
+      "step": 38675
+    },
+    {
+      "epoch": 0.9205941159668666,
+      "grad_norm": 4.186881861311692,
+      "learning_rate": 2e-05,
+      "loss": 0.7861,
+      "step": 38676
+    },
+    {
+      "epoch": 0.9206179186898982,
+      "grad_norm": 0.033533388519613835,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 38677
+    },
+    {
+      "epoch": 0.9206417214129297,
+      "grad_norm": 2.0849944533114213,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 38678
+    },
+    {
+      "epoch": 0.9206655241359611,
+      "grad_norm": 1.1047972577131346,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 38679
+    },
+    {
+      "epoch": 0.9206893268589926,
+      "grad_norm": 3.1117816355327568,
+      "learning_rate": 2e-05,
+      "loss": 0.2147,
+      "step": 38680
+    },
+    {
+      "epoch": 0.9207131295820242,
+      "grad_norm": 4.681545093807169,
+      "learning_rate": 2e-05,
+      "loss": 0.0829,
+      "step": 38681
+    },
+    {
+      "epoch": 0.9207369323050557,
+      "grad_norm": 4.283824683932672,
+      "learning_rate": 2e-05,
+      "loss": 0.7874,
+      "step": 38682
+    },
+    {
+      "epoch": 0.9207607350280872,
+      "grad_norm": 5.044932581714413,
+      "learning_rate": 2e-05,
+      "loss": 1.0199,
+      "step": 38683
+    },
+    {
+      "epoch": 0.9207845377511187,
+      "grad_norm": 5.387164966638888,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 38684
+    },
+    {
+      "epoch": 0.9208083404741503,
+      "grad_norm": 0.5060209363873271,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 38685
+    },
+    {
+      "epoch": 0.9208321431971818,
+      "grad_norm": 4.0837284938103595,
+      "learning_rate": 2e-05,
+      "loss": 0.7415,
+      "step": 38686
+    },
+    {
+      "epoch": 0.9208559459202132,
+      "grad_norm": 1.8430627914326723,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 38687
+    },
+    {
+      "epoch": 0.9208797486432447,
+      "grad_norm": 10.766813374646219,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 38688
+    },
+    {
+      "epoch": 0.9209035513662763,
+      "grad_norm": 2.731880691860105,
+      "learning_rate": 2e-05,
+      "loss": 0.4163,
+      "step": 38689
+    },
+    {
+      "epoch": 0.9209273540893078,
+      "grad_norm": 3.1914733183742583,
+      "learning_rate": 2e-05,
+      "loss": 0.1696,
+      "step": 38690
+    },
+    {
+      "epoch": 0.9209511568123393,
+      "grad_norm": 0.2882767995078442,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 38691
+    },
+    {
+      "epoch": 0.9209749595353709,
+      "grad_norm": 2.984430070899918,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 38692
+    },
+    {
+      "epoch": 0.9209987622584024,
+      "grad_norm": 0.02482622610626425,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 38693
+    },
+    {
+      "epoch": 0.9210225649814339,
+      "grad_norm": 0.34810229505110646,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 38694
+    },
+    {
+      "epoch": 0.9210463677044654,
+      "grad_norm": 0.1733048651663909,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 38695
+    },
+    {
+      "epoch": 0.921070170427497,
+      "grad_norm": 0.13292646566586924,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 38696
+    },
+    {
+      "epoch": 0.9210939731505284,
+      "grad_norm": 12.329431173575076,
+      "learning_rate": 2e-05,
+      "loss": 0.9232,
+      "step": 38697
+    },
+    {
+      "epoch": 0.9211177758735599,
+      "grad_norm": 5.378179282730501,
+      "learning_rate": 2e-05,
+      "loss": 0.1784,
+      "step": 38698
+    },
+    {
+      "epoch": 0.9211415785965914,
+      "grad_norm": 1.7564253632635167,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 38699
+    },
+    {
+      "epoch": 0.921165381319623,
+      "grad_norm": 5.94011930518345,
+      "learning_rate": 2e-05,
+      "loss": 0.3589,
+      "step": 38700
+    },
+    {
+      "epoch": 0.9211891840426545,
+      "grad_norm": 1.0928586832512621,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 38701
+    },
+    {
+      "epoch": 0.921212986765686,
+      "grad_norm": 3.0497095722316376,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 38702
+    },
+    {
+      "epoch": 0.9212367894887175,
+      "grad_norm": 1.2072144700250378,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 38703
+    },
+    {
+      "epoch": 0.921260592211749,
+      "grad_norm": 1.9092534042576361,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 38704
+    },
+    {
+      "epoch": 0.9212843949347805,
+      "grad_norm": 4.181416432457089,
+      "learning_rate": 2e-05,
+      "loss": 0.5876,
+      "step": 38705
+    },
+    {
+      "epoch": 0.921308197657812,
+      "grad_norm": 2.9977864002923953,
+      "learning_rate": 2e-05,
+      "loss": 0.3219,
+      "step": 38706
+    },
+    {
+      "epoch": 0.9213320003808436,
+      "grad_norm": 8.431746981034442,
+      "learning_rate": 2e-05,
+      "loss": 0.2386,
+      "step": 38707
+    },
+    {
+      "epoch": 0.9213558031038751,
+      "grad_norm": 2.69305538895921,
+      "learning_rate": 2e-05,
+      "loss": 0.3898,
+      "step": 38708
+    },
+    {
+      "epoch": 0.9213796058269066,
+      "grad_norm": 0.762346704226238,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 38709
+    },
+    {
+      "epoch": 0.9214034085499381,
+      "grad_norm": 1.493722159485168,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 38710
+    },
+    {
+      "epoch": 0.9214272112729697,
+      "grad_norm": 0.7195907696586584,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 38711
+    },
+    {
+      "epoch": 0.9214510139960012,
+      "grad_norm": 0.02320463014856544,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 38712
+    },
+    {
+      "epoch": 0.9214748167190326,
+      "grad_norm": 3.8947850874411944,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 38713
+    },
+    {
+      "epoch": 0.9214986194420641,
+      "grad_norm": 6.597425299895962,
+      "learning_rate": 2e-05,
+      "loss": 0.2362,
+      "step": 38714
+    },
+    {
+      "epoch": 0.9215224221650957,
+      "grad_norm": 3.279491440662956,
+      "learning_rate": 2e-05,
+      "loss": 0.401,
+      "step": 38715
+    },
+    {
+      "epoch": 0.9215462248881272,
+      "grad_norm": 2.5562747690915018,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 38716
+    },
+    {
+      "epoch": 0.9215700276111587,
+      "grad_norm": 16.198339756661642,
+      "learning_rate": 2e-05,
+      "loss": 0.5783,
+      "step": 38717
+    },
+    {
+      "epoch": 0.9215938303341902,
+      "grad_norm": 4.38831330612419,
+      "learning_rate": 2e-05,
+      "loss": 0.5475,
+      "step": 38718
+    },
+    {
+      "epoch": 0.9216176330572218,
+      "grad_norm": 31.344037208402717,
+      "learning_rate": 2e-05,
+      "loss": 0.5409,
+      "step": 38719
+    },
+    {
+      "epoch": 0.9216414357802533,
+      "grad_norm": 2.063065491816449,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 38720
+    },
+    {
+      "epoch": 0.9216652385032847,
+      "grad_norm": 2.9491270123763247,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 38721
+    },
+    {
+      "epoch": 0.9216890412263163,
+      "grad_norm": 11.570104101212548,
+      "learning_rate": 2e-05,
+      "loss": 1.3642,
+      "step": 38722
+    },
+    {
+      "epoch": 0.9217128439493478,
+      "grad_norm": 0.03711239633214326,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 38723
+    },
+    {
+      "epoch": 0.9217366466723793,
+      "grad_norm": 6.780841241255357,
+      "learning_rate": 2e-05,
+      "loss": 0.1122,
+      "step": 38724
+    },
+    {
+      "epoch": 0.9217604493954108,
+      "grad_norm": 0.8069552506107788,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 38725
+    },
+    {
+      "epoch": 0.9217842521184424,
+      "grad_norm": 25.92464977801424,
+      "learning_rate": 2e-05,
+      "loss": 0.7458,
+      "step": 38726
+    },
+    {
+      "epoch": 0.9218080548414739,
+      "grad_norm": 9.855537986340279,
+      "learning_rate": 2e-05,
+      "loss": 0.1136,
+      "step": 38727
+    },
+    {
+      "epoch": 0.9218318575645054,
+      "grad_norm": 4.109546436071387,
+      "learning_rate": 2e-05,
+      "loss": 0.1161,
+      "step": 38728
+    },
+    {
+      "epoch": 0.9218556602875368,
+      "grad_norm": 2.1051897607932766,
+      "learning_rate": 2e-05,
+      "loss": 0.2513,
+      "step": 38729
+    },
+    {
+      "epoch": 0.9218794630105684,
+      "grad_norm": 12.365871755171218,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 38730
+    },
+    {
+      "epoch": 0.9219032657335999,
+      "grad_norm": 2.309302080821907,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 38731
+    },
+    {
+      "epoch": 0.9219270684566314,
+      "grad_norm": 14.509346520937973,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 38732
+    },
+    {
+      "epoch": 0.9219508711796629,
+      "grad_norm": 0.5605054552692947,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 38733
+    },
+    {
+      "epoch": 0.9219746739026945,
+      "grad_norm": 2.357471143063421,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 38734
+    },
+    {
+      "epoch": 0.921998476625726,
+      "grad_norm": 8.047563812022227,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 38735
+    },
+    {
+      "epoch": 0.9220222793487575,
+      "grad_norm": 2.795760539263999,
+      "learning_rate": 2e-05,
+      "loss": 0.2213,
+      "step": 38736
+    },
+    {
+      "epoch": 0.9220460820717891,
+      "grad_norm": 3.378902150793634,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 38737
+    },
+    {
+      "epoch": 0.9220698847948205,
+      "grad_norm": 2.3206761925344472,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 38738
+    },
+    {
+      "epoch": 0.922093687517852,
+      "grad_norm": 3.0807422510108147,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 38739
+    },
+    {
+      "epoch": 0.9221174902408835,
+      "grad_norm": 8.322224165040499,
+      "learning_rate": 2e-05,
+      "loss": 0.4456,
+      "step": 38740
+    },
+    {
+      "epoch": 0.9221412929639151,
+      "grad_norm": 6.764286413034268,
+      "learning_rate": 2e-05,
+      "loss": 0.3748,
+      "step": 38741
+    },
+    {
+      "epoch": 0.9221650956869466,
+      "grad_norm": 4.89753404499245,
+      "learning_rate": 2e-05,
+      "loss": 1.0567,
+      "step": 38742
+    },
+    {
+      "epoch": 0.9221888984099781,
+      "grad_norm": 5.833277824669455,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 38743
+    },
+    {
+      "epoch": 0.9222127011330096,
+      "grad_norm": 0.08313758976460714,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38744
+    },
+    {
+      "epoch": 0.9222365038560412,
+      "grad_norm": 0.1345154987328537,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 38745
+    },
+    {
+      "epoch": 0.9222603065790727,
+      "grad_norm": 1.9410372968970167,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 38746
+    },
+    {
+      "epoch": 0.9222841093021041,
+      "grad_norm": 0.06065945162838028,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38747
+    },
+    {
+      "epoch": 0.9223079120251356,
+      "grad_norm": 3.4544847158577965,
+      "learning_rate": 2e-05,
+      "loss": 0.0867,
+      "step": 38748
+    },
+    {
+      "epoch": 0.9223317147481672,
+      "grad_norm": 0.2801509901538717,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 38749
+    },
+    {
+      "epoch": 0.9223555174711987,
+      "grad_norm": 3.085474085458406,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 38750
+    },
+    {
+      "epoch": 0.9223793201942302,
+      "grad_norm": 0.43456739907019615,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 38751
+    },
+    {
+      "epoch": 0.9224031229172618,
+      "grad_norm": 2.507935026796001,
+      "learning_rate": 2e-05,
+      "loss": 0.0454,
+      "step": 38752
+    },
+    {
+      "epoch": 0.9224269256402933,
+      "grad_norm": 3.1307205431743492,
+      "learning_rate": 2e-05,
+      "loss": 0.5685,
+      "step": 38753
+    },
+    {
+      "epoch": 0.9224507283633248,
+      "grad_norm": 2.2391312754697172,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 38754
+    },
+    {
+      "epoch": 0.9224745310863562,
+      "grad_norm": 0.465203441693679,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 38755
+    },
+    {
+      "epoch": 0.9224983338093878,
+      "grad_norm": 6.728358358034254,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 38756
+    },
+    {
+      "epoch": 0.9225221365324193,
+      "grad_norm": 4.04829193743554,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 38757
+    },
+    {
+      "epoch": 0.9225459392554508,
+      "grad_norm": 1.9098733827215522,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 38758
+    },
+    {
+      "epoch": 0.9225697419784823,
+      "grad_norm": 5.82664338728756,
+      "learning_rate": 2e-05,
+      "loss": 0.3821,
+      "step": 38759
+    },
+    {
+      "epoch": 0.9225935447015139,
+      "grad_norm": 0.17361458052462114,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 38760
+    },
+    {
+      "epoch": 0.9226173474245454,
+      "grad_norm": 14.222548484600974,
+      "learning_rate": 2e-05,
+      "loss": 1.1022,
+      "step": 38761
+    },
+    {
+      "epoch": 0.9226411501475769,
+      "grad_norm": 0.30402312218463945,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 38762
+    },
+    {
+      "epoch": 0.9226649528706083,
+      "grad_norm": 0.3366879564151997,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 38763
+    },
+    {
+      "epoch": 0.9226887555936399,
+      "grad_norm": 3.98314275255827,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 38764
+    },
+    {
+      "epoch": 0.9227125583166714,
+      "grad_norm": 3.290255723929214,
+      "learning_rate": 2e-05,
+      "loss": 0.2125,
+      "step": 38765
+    },
+    {
+      "epoch": 0.9227363610397029,
+      "grad_norm": 2.323585434872294,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 38766
+    },
+    {
+      "epoch": 0.9227601637627345,
+      "grad_norm": 3.1419183804966426,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 38767
+    },
+    {
+      "epoch": 0.922783966485766,
+      "grad_norm": 7.520368518233931,
+      "learning_rate": 2e-05,
+      "loss": 0.2238,
+      "step": 38768
+    },
+    {
+      "epoch": 0.9228077692087975,
+      "grad_norm": 1.013636792819434,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 38769
+    },
+    {
+      "epoch": 0.922831571931829,
+      "grad_norm": 3.6396359765607396,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 38770
+    },
+    {
+      "epoch": 0.9228553746548606,
+      "grad_norm": 1.3373515052244243,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 38771
+    },
+    {
+      "epoch": 0.922879177377892,
+      "grad_norm": 1.9467536045446647,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 38772
+    },
+    {
+      "epoch": 0.9229029801009235,
+      "grad_norm": 4.044076425842809,
+      "learning_rate": 2e-05,
+      "loss": 0.5503,
+      "step": 38773
+    },
+    {
+      "epoch": 0.922926782823955,
+      "grad_norm": 0.044493207969653926,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 38774
+    },
+    {
+      "epoch": 0.9229505855469866,
+      "grad_norm": 3.793078292128359,
+      "learning_rate": 2e-05,
+      "loss": 0.1002,
+      "step": 38775
+    },
+    {
+      "epoch": 0.9229743882700181,
+      "grad_norm": 2.5010709071966666,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 38776
+    },
+    {
+      "epoch": 0.9229981909930496,
+      "grad_norm": 0.242512800760805,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 38777
+    },
+    {
+      "epoch": 0.9230219937160811,
+      "grad_norm": 0.4662046985940218,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 38778
+    },
+    {
+      "epoch": 0.9230457964391127,
+      "grad_norm": 0.05812910229017529,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 38779
+    },
+    {
+      "epoch": 0.9230695991621441,
+      "grad_norm": 8.518681655926166,
+      "learning_rate": 2e-05,
+      "loss": 0.292,
+      "step": 38780
+    },
+    {
+      "epoch": 0.9230934018851756,
+      "grad_norm": 0.44908596084921293,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 38781
+    },
+    {
+      "epoch": 0.9231172046082072,
+      "grad_norm": 10.504561254895718,
+      "learning_rate": 2e-05,
+      "loss": 0.6768,
+      "step": 38782
+    },
+    {
+      "epoch": 0.9231410073312387,
+      "grad_norm": 3.8245241196811146,
+      "learning_rate": 2e-05,
+      "loss": 0.2096,
+      "step": 38783
+    },
+    {
+      "epoch": 0.9231648100542702,
+      "grad_norm": 3.3405068868311623,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 38784
+    },
+    {
+      "epoch": 0.9231886127773017,
+      "grad_norm": 4.855392765170348,
+      "learning_rate": 2e-05,
+      "loss": 0.3725,
+      "step": 38785
+    },
+    {
+      "epoch": 0.9232124155003333,
+      "grad_norm": 2.703766876057953,
+      "learning_rate": 2e-05,
+      "loss": 0.5538,
+      "step": 38786
+    },
+    {
+      "epoch": 0.9232362182233648,
+      "grad_norm": 0.0442355842227466,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 38787
+    },
+    {
+      "epoch": 0.9232600209463963,
+      "grad_norm": 2.170009319180804,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 38788
+    },
+    {
+      "epoch": 0.9232838236694277,
+      "grad_norm": 3.5296228931863767,
+      "learning_rate": 2e-05,
+      "loss": 0.496,
+      "step": 38789
+    },
+    {
+      "epoch": 0.9233076263924593,
+      "grad_norm": 5.7822082155825525,
+      "learning_rate": 2e-05,
+      "loss": 1.528,
+      "step": 38790
+    },
+    {
+      "epoch": 0.9233314291154908,
+      "grad_norm": 0.1975799013395957,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 38791
+    },
+    {
+      "epoch": 0.9233552318385223,
+      "grad_norm": 2.4368706846039325,
+      "learning_rate": 2e-05,
+      "loss": 0.4275,
+      "step": 38792
+    },
+    {
+      "epoch": 0.9233790345615538,
+      "grad_norm": 4.904490445200776,
+      "learning_rate": 2e-05,
+      "loss": 1.055,
+      "step": 38793
+    },
+    {
+      "epoch": 0.9234028372845854,
+      "grad_norm": 2.638042822603261,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 38794
+    },
+    {
+      "epoch": 0.9234266400076169,
+      "grad_norm": 3.0760784397066034,
+      "learning_rate": 2e-05,
+      "loss": 0.7169,
+      "step": 38795
+    },
+    {
+      "epoch": 0.9234504427306484,
+      "grad_norm": 11.037675226078088,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 38796
+    },
+    {
+      "epoch": 0.92347424545368,
+      "grad_norm": 12.838700081458821,
+      "learning_rate": 2e-05,
+      "loss": 0.408,
+      "step": 38797
+    },
+    {
+      "epoch": 0.9234980481767114,
+      "grad_norm": 1.7846808893975243,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 38798
+    },
+    {
+      "epoch": 0.9235218508997429,
+      "grad_norm": 24.880808289342554,
+      "learning_rate": 2e-05,
+      "loss": 0.5612,
+      "step": 38799
+    },
+    {
+      "epoch": 0.9235456536227744,
+      "grad_norm": 101.49992552458501,
+      "learning_rate": 2e-05,
+      "loss": 0.1178,
+      "step": 38800
+    },
+    {
+      "epoch": 0.923569456345806,
+      "grad_norm": 4.645986560882511,
+      "learning_rate": 2e-05,
+      "loss": 1.0797,
+      "step": 38801
+    },
+    {
+      "epoch": 0.9235932590688375,
+      "grad_norm": 5.602505838506073,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 38802
+    },
+    {
+      "epoch": 0.923617061791869,
+      "grad_norm": 0.18565285013287747,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 38803
+    },
+    {
+      "epoch": 0.9236408645149005,
+      "grad_norm": 5.511305507535134,
+      "learning_rate": 2e-05,
+      "loss": 0.9936,
+      "step": 38804
+    },
+    {
+      "epoch": 0.9236646672379321,
+      "grad_norm": 6.862621375835772,
+      "learning_rate": 2e-05,
+      "loss": 1.1397,
+      "step": 38805
+    },
+    {
+      "epoch": 0.9236884699609635,
+      "grad_norm": 4.85711374225622,
+      "learning_rate": 2e-05,
+      "loss": 1.0164,
+      "step": 38806
+    },
+    {
+      "epoch": 0.923712272683995,
+      "grad_norm": 1.6309140067591776,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 38807
+    },
+    {
+      "epoch": 0.9237360754070265,
+      "grad_norm": 3.9688450840183283,
+      "learning_rate": 2e-05,
+      "loss": 0.6062,
+      "step": 38808
+    },
+    {
+      "epoch": 0.9237598781300581,
+      "grad_norm": 0.1970870991682691,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 38809
+    },
+    {
+      "epoch": 0.9237836808530896,
+      "grad_norm": 1.6016156566701254,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 38810
+    },
+    {
+      "epoch": 0.9238074835761211,
+      "grad_norm": 1.0148791417773153,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 38811
+    },
+    {
+      "epoch": 0.9238312862991527,
+      "grad_norm": 1.3520462199278152,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 38812
+    },
+    {
+      "epoch": 0.9238550890221842,
+      "grad_norm": 0.37779431582929524,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 38813
+    },
+    {
+      "epoch": 0.9238788917452156,
+      "grad_norm": 5.507993354886761,
+      "learning_rate": 2e-05,
+      "loss": 0.718,
+      "step": 38814
+    },
+    {
+      "epoch": 0.9239026944682471,
+      "grad_norm": 1.5352361563279955,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 38815
+    },
+    {
+      "epoch": 0.9239264971912787,
+      "grad_norm": 0.5117856380743274,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 38816
+    },
+    {
+      "epoch": 0.9239502999143102,
+      "grad_norm": 6.6891070821133045,
+      "learning_rate": 2e-05,
+      "loss": 0.237,
+      "step": 38817
+    },
+    {
+      "epoch": 0.9239741026373417,
+      "grad_norm": 5.5201582482491505,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 38818
+    },
+    {
+      "epoch": 0.9239979053603732,
+      "grad_norm": 7.036586110507729,
+      "learning_rate": 2e-05,
+      "loss": 0.5762,
+      "step": 38819
+    },
+    {
+      "epoch": 0.9240217080834048,
+      "grad_norm": 3.2739372228136925,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 38820
+    },
+    {
+      "epoch": 0.9240455108064363,
+      "grad_norm": 2.583570217921387,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 38821
+    },
+    {
+      "epoch": 0.9240693135294678,
+      "grad_norm": 0.16487753361196916,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 38822
+    },
+    {
+      "epoch": 0.9240931162524992,
+      "grad_norm": 4.111986333523313,
+      "learning_rate": 2e-05,
+      "loss": 0.4592,
+      "step": 38823
+    },
+    {
+      "epoch": 0.9241169189755308,
+      "grad_norm": 3.8735842308176514,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 38824
+    },
+    {
+      "epoch": 0.9241407216985623,
+      "grad_norm": 0.8481262045806139,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 38825
+    },
+    {
+      "epoch": 0.9241645244215938,
+      "grad_norm": 3.2378978821599778,
+      "learning_rate": 2e-05,
+      "loss": 0.1209,
+      "step": 38826
+    },
+    {
+      "epoch": 0.9241883271446254,
+      "grad_norm": 2.7935122163039154,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 38827
+    },
+    {
+      "epoch": 0.9242121298676569,
+      "grad_norm": 5.327186990563625,
+      "learning_rate": 2e-05,
+      "loss": 0.8832,
+      "step": 38828
+    },
+    {
+      "epoch": 0.9242359325906884,
+      "grad_norm": 6.572302800542367,
+      "learning_rate": 2e-05,
+      "loss": 1.0574,
+      "step": 38829
+    },
+    {
+      "epoch": 0.9242597353137199,
+      "grad_norm": 3.835996455621727,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 38830
+    },
+    {
+      "epoch": 0.9242835380367515,
+      "grad_norm": 7.461139593248891,
+      "learning_rate": 2e-05,
+      "loss": 0.3484,
+      "step": 38831
+    },
+    {
+      "epoch": 0.9243073407597829,
+      "grad_norm": 2.611080009839147,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 38832
+    },
+    {
+      "epoch": 0.9243311434828144,
+      "grad_norm": 6.0122210443479975,
+      "learning_rate": 2e-05,
+      "loss": 0.7616,
+      "step": 38833
+    },
+    {
+      "epoch": 0.9243549462058459,
+      "grad_norm": 18.190066777224498,
+      "learning_rate": 2e-05,
+      "loss": 0.827,
+      "step": 38834
+    },
+    {
+      "epoch": 0.9243787489288775,
+      "grad_norm": 3.9687873679762826,
+      "learning_rate": 2e-05,
+      "loss": 0.8098,
+      "step": 38835
+    },
+    {
+      "epoch": 0.924402551651909,
+      "grad_norm": 3.880377029334822,
+      "learning_rate": 2e-05,
+      "loss": 0.7445,
+      "step": 38836
+    },
+    {
+      "epoch": 0.9244263543749405,
+      "grad_norm": 3.604415744619156,
+      "learning_rate": 2e-05,
+      "loss": 0.3338,
+      "step": 38837
+    },
+    {
+      "epoch": 0.924450157097972,
+      "grad_norm": 0.8552641640199887,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 38838
+    },
+    {
+      "epoch": 0.9244739598210036,
+      "grad_norm": 0.8063145200942453,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 38839
+    },
+    {
+      "epoch": 0.924497762544035,
+      "grad_norm": 0.29276635690631947,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38840
+    },
+    {
+      "epoch": 0.9245215652670665,
+      "grad_norm": 4.547632423502918,
+      "learning_rate": 2e-05,
+      "loss": 1.1908,
+      "step": 38841
+    },
+    {
+      "epoch": 0.9245453679900981,
+      "grad_norm": 0.05636587909204969,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 38842
+    },
+    {
+      "epoch": 0.9245691707131296,
+      "grad_norm": 5.659464495750127,
+      "learning_rate": 2e-05,
+      "loss": 0.4804,
+      "step": 38843
+    },
+    {
+      "epoch": 0.9245929734361611,
+      "grad_norm": 0.09784557125819487,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38844
+    },
+    {
+      "epoch": 0.9246167761591926,
+      "grad_norm": 1.06569535424922,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 38845
+    },
+    {
+      "epoch": 0.9246405788822242,
+      "grad_norm": 5.475604451970736,
+      "learning_rate": 2e-05,
+      "loss": 0.1237,
+      "step": 38846
+    },
+    {
+      "epoch": 0.9246643816052557,
+      "grad_norm": 4.015828442474302,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 38847
+    },
+    {
+      "epoch": 0.9246881843282871,
+      "grad_norm": 30.5562026389592,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 38848
+    },
+    {
+      "epoch": 0.9247119870513186,
+      "grad_norm": 5.060451302740433,
+      "learning_rate": 2e-05,
+      "loss": 0.3528,
+      "step": 38849
+    },
+    {
+      "epoch": 0.9247357897743502,
+      "grad_norm": 0.06204727329721153,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38850
+    },
+    {
+      "epoch": 0.9247595924973817,
+      "grad_norm": 0.03507821175151481,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38851
+    },
+    {
+      "epoch": 0.9247833952204132,
+      "grad_norm": 5.310206174560947,
+      "learning_rate": 2e-05,
+      "loss": 1.0255,
+      "step": 38852
+    },
+    {
+      "epoch": 0.9248071979434447,
+      "grad_norm": 2.1815474890059634,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 38853
+    },
+    {
+      "epoch": 0.9248310006664763,
+      "grad_norm": 1.4540306636509033,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 38854
+    },
+    {
+      "epoch": 0.9248548033895078,
+      "grad_norm": 6.508963060407412,
+      "learning_rate": 2e-05,
+      "loss": 0.621,
+      "step": 38855
+    },
+    {
+      "epoch": 0.9248786061125392,
+      "grad_norm": 12.523470487683426,
+      "learning_rate": 2e-05,
+      "loss": 0.8812,
+      "step": 38856
+    },
+    {
+      "epoch": 0.9249024088355708,
+      "grad_norm": 41.05186579488153,
+      "learning_rate": 2e-05,
+      "loss": 1.0566,
+      "step": 38857
+    },
+    {
+      "epoch": 0.9249262115586023,
+      "grad_norm": 1.6841296003087693,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 38858
+    },
+    {
+      "epoch": 0.9249500142816338,
+      "grad_norm": 3.5038230210476047,
+      "learning_rate": 2e-05,
+      "loss": 0.2924,
+      "step": 38859
+    },
+    {
+      "epoch": 0.9249738170046653,
+      "grad_norm": 19.07105695449224,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 38860
+    },
+    {
+      "epoch": 0.9249976197276969,
+      "grad_norm": 0.27085999654759446,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 38861
+    },
+    {
+      "epoch": 0.9250214224507284,
+      "grad_norm": 9.254137688348777,
+      "learning_rate": 2e-05,
+      "loss": 0.7172,
+      "step": 38862
+    },
+    {
+      "epoch": 0.9250452251737599,
+      "grad_norm": 9.914963710628355,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 38863
+    },
+    {
+      "epoch": 0.9250690278967914,
+      "grad_norm": 3.839036468286201,
+      "learning_rate": 2e-05,
+      "loss": 0.6473,
+      "step": 38864
+    },
+    {
+      "epoch": 0.925092830619823,
+      "grad_norm": 1.556482862093967,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 38865
+    },
+    {
+      "epoch": 0.9251166333428544,
+      "grad_norm": 1.947846064899909,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 38866
+    },
+    {
+      "epoch": 0.9251404360658859,
+      "grad_norm": 4.1610139673823605,
+      "learning_rate": 2e-05,
+      "loss": 0.0912,
+      "step": 38867
+    },
+    {
+      "epoch": 0.9251642387889174,
+      "grad_norm": 3.476550838992059,
+      "learning_rate": 2e-05,
+      "loss": 0.0735,
+      "step": 38868
+    },
+    {
+      "epoch": 0.925188041511949,
+      "grad_norm": 1.423742768876644,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 38869
+    },
+    {
+      "epoch": 0.9252118442349805,
+      "grad_norm": 12.321456313898244,
+      "learning_rate": 2e-05,
+      "loss": 0.6549,
+      "step": 38870
+    },
+    {
+      "epoch": 0.925235646958012,
+      "grad_norm": 3.1856442204639057,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 38871
+    },
+    {
+      "epoch": 0.9252594496810436,
+      "grad_norm": 0.21109364624863727,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 38872
+    },
+    {
+      "epoch": 0.925283252404075,
+      "grad_norm": 16.873591528701997,
+      "learning_rate": 2e-05,
+      "loss": 0.8058,
+      "step": 38873
+    },
+    {
+      "epoch": 0.9253070551271065,
+      "grad_norm": 0.1857160871327881,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 38874
+    },
+    {
+      "epoch": 0.925330857850138,
+      "grad_norm": 0.8331904225520051,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 38875
+    },
+    {
+      "epoch": 0.9253546605731696,
+      "grad_norm": 2.29899891587873,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 38876
+    },
+    {
+      "epoch": 0.9253784632962011,
+      "grad_norm": 3.5132533325394215,
+      "learning_rate": 2e-05,
+      "loss": 0.4138,
+      "step": 38877
+    },
+    {
+      "epoch": 0.9254022660192326,
+      "grad_norm": 5.105661048771214,
+      "learning_rate": 2e-05,
+      "loss": 1.1608,
+      "step": 38878
+    },
+    {
+      "epoch": 0.9254260687422641,
+      "grad_norm": 2.2435786225941396,
+      "learning_rate": 2e-05,
+      "loss": 0.2455,
+      "step": 38879
+    },
+    {
+      "epoch": 0.9254498714652957,
+      "grad_norm": 1.8182068314896336,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 38880
+    },
+    {
+      "epoch": 0.9254736741883272,
+      "grad_norm": 2.7698067600351712,
+      "learning_rate": 2e-05,
+      "loss": 0.0237,
+      "step": 38881
+    },
+    {
+      "epoch": 0.9254974769113586,
+      "grad_norm": 1.1437191419061565,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 38882
+    },
+    {
+      "epoch": 0.9255212796343901,
+      "grad_norm": 4.548535951387426,
+      "learning_rate": 2e-05,
+      "loss": 0.8084,
+      "step": 38883
+    },
+    {
+      "epoch": 0.9255450823574217,
+      "grad_norm": 2.8211126948397243,
+      "learning_rate": 2e-05,
+      "loss": 0.3102,
+      "step": 38884
+    },
+    {
+      "epoch": 0.9255688850804532,
+      "grad_norm": 10.574091964331874,
+      "learning_rate": 2e-05,
+      "loss": 1.0049,
+      "step": 38885
+    },
+    {
+      "epoch": 0.9255926878034847,
+      "grad_norm": 6.589642994331408,
+      "learning_rate": 2e-05,
+      "loss": 0.1982,
+      "step": 38886
+    },
+    {
+      "epoch": 0.9256164905265162,
+      "grad_norm": 1.3042152026192562,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 38887
+    },
+    {
+      "epoch": 0.9256402932495478,
+      "grad_norm": 6.071338778774606,
+      "learning_rate": 2e-05,
+      "loss": 1.0137,
+      "step": 38888
+    },
+    {
+      "epoch": 0.9256640959725793,
+      "grad_norm": 2.7557997049388874,
+      "learning_rate": 2e-05,
+      "loss": 0.3322,
+      "step": 38889
+    },
+    {
+      "epoch": 0.9256878986956107,
+      "grad_norm": 2.5940728527881567,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 38890
+    },
+    {
+      "epoch": 0.9257117014186423,
+      "grad_norm": 4.244075852351258,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 38891
+    },
+    {
+      "epoch": 0.9257355041416738,
+      "grad_norm": 7.828432327728589,
+      "learning_rate": 2e-05,
+      "loss": 0.9033,
+      "step": 38892
+    },
+    {
+      "epoch": 0.9257593068647053,
+      "grad_norm": 3.517203740962532,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 38893
+    },
+    {
+      "epoch": 0.9257831095877368,
+      "grad_norm": 0.030148771813566316,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 38894
+    },
+    {
+      "epoch": 0.9258069123107684,
+      "grad_norm": 1.860092970210053,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 38895
+    },
+    {
+      "epoch": 0.9258307150337999,
+      "grad_norm": 3.3797962503860277,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 38896
+    },
+    {
+      "epoch": 0.9258545177568314,
+      "grad_norm": 0.24200005020905618,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 38897
+    },
+    {
+      "epoch": 0.9258783204798628,
+      "grad_norm": 4.734511529849912,
+      "learning_rate": 2e-05,
+      "loss": 0.5919,
+      "step": 38898
+    },
+    {
+      "epoch": 0.9259021232028944,
+      "grad_norm": 12.294855192124649,
+      "learning_rate": 2e-05,
+      "loss": 1.3811,
+      "step": 38899
+    },
+    {
+      "epoch": 0.9259259259259259,
+      "grad_norm": 7.917575202771382,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 38900
+    },
+    {
+      "epoch": 0.9259497286489574,
+      "grad_norm": 5.735594620764539,
+      "learning_rate": 2e-05,
+      "loss": 1.0012,
+      "step": 38901
+    },
+    {
+      "epoch": 0.9259735313719889,
+      "grad_norm": 0.48013125223886133,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 38902
+    },
+    {
+      "epoch": 0.9259973340950205,
+      "grad_norm": 3.539996824353546,
+      "learning_rate": 2e-05,
+      "loss": 0.289,
+      "step": 38903
+    },
+    {
+      "epoch": 0.926021136818052,
+      "grad_norm": 2.256150798878931,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 38904
+    },
+    {
+      "epoch": 0.9260449395410835,
+      "grad_norm": 0.07167004778396147,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38905
+    },
+    {
+      "epoch": 0.9260687422641151,
+      "grad_norm": 3.2108313616978266,
+      "learning_rate": 2e-05,
+      "loss": 0.154,
+      "step": 38906
+    },
+    {
+      "epoch": 0.9260925449871465,
+      "grad_norm": 1.0516096539210518,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 38907
+    },
+    {
+      "epoch": 0.926116347710178,
+      "grad_norm": 0.11705549514461885,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38908
+    },
+    {
+      "epoch": 0.9261401504332095,
+      "grad_norm": 4.153167715579917,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 38909
+    },
+    {
+      "epoch": 0.9261639531562411,
+      "grad_norm": 3.089886860615295,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 38910
+    },
+    {
+      "epoch": 0.9261877558792726,
+      "grad_norm": 10.25704216741709,
+      "learning_rate": 2e-05,
+      "loss": 1.2564,
+      "step": 38911
+    },
+    {
+      "epoch": 0.9262115586023041,
+      "grad_norm": 2.072423140239851,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 38912
+    },
+    {
+      "epoch": 0.9262353613253356,
+      "grad_norm": 4.056394754370022,
+      "learning_rate": 2e-05,
+      "loss": 0.5458,
+      "step": 38913
+    },
+    {
+      "epoch": 0.9262591640483672,
+      "grad_norm": 0.8646612070685472,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 38914
+    },
+    {
+      "epoch": 0.9262829667713987,
+      "grad_norm": 3.2389887713724552,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 38915
+    },
+    {
+      "epoch": 0.9263067694944301,
+      "grad_norm": 5.914202568173365,
+      "learning_rate": 2e-05,
+      "loss": 0.5487,
+      "step": 38916
+    },
+    {
+      "epoch": 0.9263305722174616,
+      "grad_norm": 1.3927475290037021,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 38917
+    },
+    {
+      "epoch": 0.9263543749404932,
+      "grad_norm": 0.22730729990990337,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 38918
+    },
+    {
+      "epoch": 0.9263781776635247,
+      "grad_norm": 1.7012791759932713,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 38919
+    },
+    {
+      "epoch": 0.9264019803865562,
+      "grad_norm": 6.469185900293087,
+      "learning_rate": 2e-05,
+      "loss": 0.5084,
+      "step": 38920
+    },
+    {
+      "epoch": 0.9264257831095878,
+      "grad_norm": 4.83078593386975,
+      "learning_rate": 2e-05,
+      "loss": 1.2591,
+      "step": 38921
+    },
+    {
+      "epoch": 0.9264495858326193,
+      "grad_norm": 4.975975284710763,
+      "learning_rate": 2e-05,
+      "loss": 0.4911,
+      "step": 38922
+    },
+    {
+      "epoch": 0.9264733885556508,
+      "grad_norm": 18.967328788980968,
+      "learning_rate": 2e-05,
+      "loss": 0.7716,
+      "step": 38923
+    },
+    {
+      "epoch": 0.9264971912786822,
+      "grad_norm": 0.64554280634537,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 38924
+    },
+    {
+      "epoch": 0.9265209940017138,
+      "grad_norm": 3.37953983861864,
+      "learning_rate": 2e-05,
+      "loss": 0.5389,
+      "step": 38925
+    },
+    {
+      "epoch": 0.9265447967247453,
+      "grad_norm": 3.8276604347326915,
+      "learning_rate": 2e-05,
+      "loss": 0.4066,
+      "step": 38926
+    },
+    {
+      "epoch": 0.9265685994477768,
+      "grad_norm": 0.11955181902934699,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38927
+    },
+    {
+      "epoch": 0.9265924021708083,
+      "grad_norm": 0.07347953942511896,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38928
+    },
+    {
+      "epoch": 0.9266162048938399,
+      "grad_norm": 2.40415308209821,
+      "learning_rate": 2e-05,
+      "loss": 0.0338,
+      "step": 38929
+    },
+    {
+      "epoch": 0.9266400076168714,
+      "grad_norm": 3.196313888826077,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 38930
+    },
+    {
+      "epoch": 0.9266638103399029,
+      "grad_norm": 0.9016608406062839,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 38931
+    },
+    {
+      "epoch": 0.9266876130629343,
+      "grad_norm": 0.044710442214600105,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 38932
+    },
+    {
+      "epoch": 0.9267114157859659,
+      "grad_norm": 6.271209040472793,
+      "learning_rate": 2e-05,
+      "loss": 0.9434,
+      "step": 38933
+    },
+    {
+      "epoch": 0.9267352185089974,
+      "grad_norm": 5.190074964727768,
+      "learning_rate": 2e-05,
+      "loss": 0.1623,
+      "step": 38934
+    },
+    {
+      "epoch": 0.9267590212320289,
+      "grad_norm": 7.492432965630401,
+      "learning_rate": 2e-05,
+      "loss": 0.8118,
+      "step": 38935
+    },
+    {
+      "epoch": 0.9267828239550605,
+      "grad_norm": 5.483042726210185,
+      "learning_rate": 2e-05,
+      "loss": 0.3988,
+      "step": 38936
+    },
+    {
+      "epoch": 0.926806626678092,
+      "grad_norm": 3.2610796721435955,
+      "learning_rate": 2e-05,
+      "loss": 0.4471,
+      "step": 38937
+    },
+    {
+      "epoch": 0.9268304294011235,
+      "grad_norm": 3.6270937208430314,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 38938
+    },
+    {
+      "epoch": 0.926854232124155,
+      "grad_norm": 7.184628298867964,
+      "learning_rate": 2e-05,
+      "loss": 0.4807,
+      "step": 38939
+    },
+    {
+      "epoch": 0.9268780348471866,
+      "grad_norm": 0.08444677066567895,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 38940
+    },
+    {
+      "epoch": 0.926901837570218,
+      "grad_norm": 0.20970456218145955,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 38941
+    },
+    {
+      "epoch": 0.9269256402932495,
+      "grad_norm": 6.126800084213956,
+      "learning_rate": 2e-05,
+      "loss": 1.5344,
+      "step": 38942
+    },
+    {
+      "epoch": 0.926949443016281,
+      "grad_norm": 11.537297707173556,
+      "learning_rate": 2e-05,
+      "loss": 0.5923,
+      "step": 38943
+    },
+    {
+      "epoch": 0.9269732457393126,
+      "grad_norm": 0.059039247645061806,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 38944
+    },
+    {
+      "epoch": 0.9269970484623441,
+      "grad_norm": 2.6126293616140135,
+      "learning_rate": 2e-05,
+      "loss": 0.6145,
+      "step": 38945
+    },
+    {
+      "epoch": 0.9270208511853756,
+      "grad_norm": 4.089434097891276,
+      "learning_rate": 2e-05,
+      "loss": 0.0746,
+      "step": 38946
+    },
+    {
+      "epoch": 0.9270446539084071,
+      "grad_norm": 0.5951544089703057,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 38947
+    },
+    {
+      "epoch": 0.9270684566314387,
+      "grad_norm": 7.9477887654503405,
+      "learning_rate": 2e-05,
+      "loss": 0.5771,
+      "step": 38948
+    },
+    {
+      "epoch": 0.9270922593544701,
+      "grad_norm": 12.76920339112719,
+      "learning_rate": 2e-05,
+      "loss": 1.1437,
+      "step": 38949
+    },
+    {
+      "epoch": 0.9271160620775016,
+      "grad_norm": 0.38600209695378984,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 38950
+    },
+    {
+      "epoch": 0.9271398648005332,
+      "grad_norm": 1.8104843566172524,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 38951
+    },
+    {
+      "epoch": 0.9271636675235647,
+      "grad_norm": 14.412622327478603,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 38952
+    },
+    {
+      "epoch": 0.9271874702465962,
+      "grad_norm": 1.9160037290419585,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 38953
+    },
+    {
+      "epoch": 0.9272112729696277,
+      "grad_norm": 1.4017157709741357,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 38954
+    },
+    {
+      "epoch": 0.9272350756926593,
+      "grad_norm": 4.3655957704401755,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 38955
+    },
+    {
+      "epoch": 0.9272588784156908,
+      "grad_norm": 0.31647891575669834,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 38956
+    },
+    {
+      "epoch": 0.9272826811387223,
+      "grad_norm": 3.463271437960978,
+      "learning_rate": 2e-05,
+      "loss": 0.8004,
+      "step": 38957
+    },
+    {
+      "epoch": 0.9273064838617537,
+      "grad_norm": 4.615524440877517,
+      "learning_rate": 2e-05,
+      "loss": 0.7555,
+      "step": 38958
+    },
+    {
+      "epoch": 0.9273302865847853,
+      "grad_norm": 3.2632495141293316,
+      "learning_rate": 2e-05,
+      "loss": 0.5626,
+      "step": 38959
+    },
+    {
+      "epoch": 0.9273540893078168,
+      "grad_norm": 6.08296384092066,
+      "learning_rate": 2e-05,
+      "loss": 1.067,
+      "step": 38960
+    },
+    {
+      "epoch": 0.9273778920308483,
+      "grad_norm": 3.4841188323230057,
+      "learning_rate": 2e-05,
+      "loss": 0.3398,
+      "step": 38961
+    },
+    {
+      "epoch": 0.9274016947538798,
+      "grad_norm": 0.5879683684608742,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 38962
+    },
+    {
+      "epoch": 0.9274254974769114,
+      "grad_norm": 2.8283599616276085,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 38963
+    },
+    {
+      "epoch": 0.9274493001999429,
+      "grad_norm": 5.554270571234378,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 38964
+    },
+    {
+      "epoch": 0.9274731029229744,
+      "grad_norm": 3.5753239805395385,
+      "learning_rate": 2e-05,
+      "loss": 0.7662,
+      "step": 38965
+    },
+    {
+      "epoch": 0.927496905646006,
+      "grad_norm": 0.6305594031481498,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 38966
+    },
+    {
+      "epoch": 0.9275207083690374,
+      "grad_norm": 0.7324458997929955,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 38967
+    },
+    {
+      "epoch": 0.9275445110920689,
+      "grad_norm": 0.40635863796399796,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 38968
+    },
+    {
+      "epoch": 0.9275683138151004,
+      "grad_norm": 4.44497104807511,
+      "learning_rate": 2e-05,
+      "loss": 0.6076,
+      "step": 38969
+    },
+    {
+      "epoch": 0.927592116538132,
+      "grad_norm": 0.9535705458755601,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 38970
+    },
+    {
+      "epoch": 0.9276159192611635,
+      "grad_norm": 7.684517718542328,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 38971
+    },
+    {
+      "epoch": 0.927639721984195,
+      "grad_norm": 0.6131739869682893,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 38972
+    },
+    {
+      "epoch": 0.9276635247072265,
+      "grad_norm": 4.030433827339828,
+      "learning_rate": 2e-05,
+      "loss": 0.583,
+      "step": 38973
+    },
+    {
+      "epoch": 0.9276873274302581,
+      "grad_norm": 0.16419728921349108,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 38974
+    },
+    {
+      "epoch": 0.9277111301532895,
+      "grad_norm": 0.7195703659458399,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 38975
+    },
+    {
+      "epoch": 0.927734932876321,
+      "grad_norm": 6.490943878132398,
+      "learning_rate": 2e-05,
+      "loss": 1.1289,
+      "step": 38976
+    },
+    {
+      "epoch": 0.9277587355993525,
+      "grad_norm": 1.1776335936234787,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 38977
+    },
+    {
+      "epoch": 0.9277825383223841,
+      "grad_norm": 0.04259855346725675,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 38978
+    },
+    {
+      "epoch": 0.9278063410454156,
+      "grad_norm": 3.8987716671396995,
+      "learning_rate": 2e-05,
+      "loss": 0.7339,
+      "step": 38979
+    },
+    {
+      "epoch": 0.9278301437684471,
+      "grad_norm": 0.028796779750340923,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 38980
+    },
+    {
+      "epoch": 0.9278539464914787,
+      "grad_norm": 5.3190728872360955,
+      "learning_rate": 2e-05,
+      "loss": 0.9866,
+      "step": 38981
+    },
+    {
+      "epoch": 0.9278777492145102,
+      "grad_norm": 0.23823485018313137,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 38982
+    },
+    {
+      "epoch": 0.9279015519375416,
+      "grad_norm": 2.576218234183597,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 38983
+    },
+    {
+      "epoch": 0.9279253546605731,
+      "grad_norm": 0.09076568558614029,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 38984
+    },
+    {
+      "epoch": 0.9279491573836047,
+      "grad_norm": 0.026100090595778763,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 38985
+    },
+    {
+      "epoch": 0.9279729601066362,
+      "grad_norm": 4.054567583425669,
+      "learning_rate": 2e-05,
+      "loss": 0.5304,
+      "step": 38986
+    },
+    {
+      "epoch": 0.9279967628296677,
+      "grad_norm": 11.877150347444944,
+      "learning_rate": 2e-05,
+      "loss": 1.2875,
+      "step": 38987
+    },
+    {
+      "epoch": 0.9280205655526992,
+      "grad_norm": 6.424099038170401,
+      "learning_rate": 2e-05,
+      "loss": 0.3254,
+      "step": 38988
+    },
+    {
+      "epoch": 0.9280443682757308,
+      "grad_norm": 1.3219837310271896,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 38989
+    },
+    {
+      "epoch": 0.9280681709987623,
+      "grad_norm": 0.5034770487393615,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 38990
+    },
+    {
+      "epoch": 0.9280919737217938,
+      "grad_norm": 2.121689802060782,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 38991
+    },
+    {
+      "epoch": 0.9281157764448252,
+      "grad_norm": 3.2416936240681054,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 38992
+    },
+    {
+      "epoch": 0.9281395791678568,
+      "grad_norm": 3.376909131187448,
+      "learning_rate": 2e-05,
+      "loss": 0.6239,
+      "step": 38993
+    },
+    {
+      "epoch": 0.9281633818908883,
+      "grad_norm": 6.403801553255645,
+      "learning_rate": 2e-05,
+      "loss": 0.9217,
+      "step": 38994
+    },
+    {
+      "epoch": 0.9281871846139198,
+      "grad_norm": 0.04811577051145656,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 38995
+    },
+    {
+      "epoch": 0.9282109873369514,
+      "grad_norm": 7.621472518164628,
+      "learning_rate": 2e-05,
+      "loss": 0.1084,
+      "step": 38996
+    },
+    {
+      "epoch": 0.9282347900599829,
+      "grad_norm": 0.19797240368899005,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 38997
+    },
+    {
+      "epoch": 0.9282585927830144,
+      "grad_norm": 0.34848628623163935,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 38998
+    },
+    {
+      "epoch": 0.9282823955060459,
+      "grad_norm": 1.301856915557417,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 38999
+    },
+    {
+      "epoch": 0.9283061982290775,
+      "grad_norm": 4.784567158010269,
+      "learning_rate": 2e-05,
+      "loss": 0.8875,
+      "step": 39000
+    },
+    {
+      "epoch": 0.9283300009521089,
+      "grad_norm": 3.252002148905316,
+      "learning_rate": 2e-05,
+      "loss": 0.4863,
+      "step": 39001
+    },
+    {
+      "epoch": 0.9283538036751404,
+      "grad_norm": 0.19820202527452724,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 39002
+    },
+    {
+      "epoch": 0.9283776063981719,
+      "grad_norm": 6.9531660039244825,
+      "learning_rate": 2e-05,
+      "loss": 1.672,
+      "step": 39003
+    },
+    {
+      "epoch": 0.9284014091212035,
+      "grad_norm": 9.498303153467875,
+      "learning_rate": 2e-05,
+      "loss": 1.3746,
+      "step": 39004
+    },
+    {
+      "epoch": 0.928425211844235,
+      "grad_norm": 3.4786089088181633,
+      "learning_rate": 2e-05,
+      "loss": 0.7121,
+      "step": 39005
+    },
+    {
+      "epoch": 0.9284490145672665,
+      "grad_norm": 3.846118005488079,
+      "learning_rate": 2e-05,
+      "loss": 0.4482,
+      "step": 39006
+    },
+    {
+      "epoch": 0.928472817290298,
+      "grad_norm": 3.639599950894461,
+      "learning_rate": 2e-05,
+      "loss": 0.7927,
+      "step": 39007
+    },
+    {
+      "epoch": 0.9284966200133296,
+      "grad_norm": 8.382938053467356,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 39008
+    },
+    {
+      "epoch": 0.928520422736361,
+      "grad_norm": 5.1720396672423785,
+      "learning_rate": 2e-05,
+      "loss": 0.16,
+      "step": 39009
+    },
+    {
+      "epoch": 0.9285442254593925,
+      "grad_norm": 0.03047029046314402,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 39010
+    },
+    {
+      "epoch": 0.9285680281824241,
+      "grad_norm": 2.35381572061768,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 39011
+    },
+    {
+      "epoch": 0.9285918309054556,
+      "grad_norm": 1.8449499072578661,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 39012
+    },
+    {
+      "epoch": 0.9286156336284871,
+      "grad_norm": 4.772057178313351,
+      "learning_rate": 2e-05,
+      "loss": 1.1132,
+      "step": 39013
+    },
+    {
+      "epoch": 0.9286394363515186,
+      "grad_norm": 0.09835027702930013,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 39014
+    },
+    {
+      "epoch": 0.9286632390745502,
+      "grad_norm": 3.0825353986942434,
+      "learning_rate": 2e-05,
+      "loss": 0.1533,
+      "step": 39015
+    },
+    {
+      "epoch": 0.9286870417975817,
+      "grad_norm": 8.016241481745856,
+      "learning_rate": 2e-05,
+      "loss": 0.6397,
+      "step": 39016
+    },
+    {
+      "epoch": 0.9287108445206131,
+      "grad_norm": 10.11611401302428,
+      "learning_rate": 2e-05,
+      "loss": 1.8383,
+      "step": 39017
+    },
+    {
+      "epoch": 0.9287346472436446,
+      "grad_norm": 4.619248699412956,
+      "learning_rate": 2e-05,
+      "loss": 0.6373,
+      "step": 39018
+    },
+    {
+      "epoch": 0.9287584499666762,
+      "grad_norm": 0.6727976322849998,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 39019
+    },
+    {
+      "epoch": 0.9287822526897077,
+      "grad_norm": 6.233313305642093,
+      "learning_rate": 2e-05,
+      "loss": 0.5608,
+      "step": 39020
+    },
+    {
+      "epoch": 0.9288060554127392,
+      "grad_norm": 3.4605294866573226,
+      "learning_rate": 2e-05,
+      "loss": 0.5526,
+      "step": 39021
+    },
+    {
+      "epoch": 0.9288298581357707,
+      "grad_norm": 1.9279045776116523,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 39022
+    },
+    {
+      "epoch": 0.9288536608588023,
+      "grad_norm": 8.902414749428736,
+      "learning_rate": 2e-05,
+      "loss": 1.1098,
+      "step": 39023
+    },
+    {
+      "epoch": 0.9288774635818338,
+      "grad_norm": 1.9315036143499813,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 39024
+    },
+    {
+      "epoch": 0.9289012663048652,
+      "grad_norm": 6.658524759310014,
+      "learning_rate": 2e-05,
+      "loss": 0.808,
+      "step": 39025
+    },
+    {
+      "epoch": 0.9289250690278968,
+      "grad_norm": 3.7930346719162875,
+      "learning_rate": 2e-05,
+      "loss": 0.7149,
+      "step": 39026
+    },
+    {
+      "epoch": 0.9289488717509283,
+      "grad_norm": 4.553611124843013,
+      "learning_rate": 2e-05,
+      "loss": 0.6162,
+      "step": 39027
+    },
+    {
+      "epoch": 0.9289726744739598,
+      "grad_norm": 5.984848289705183,
+      "learning_rate": 2e-05,
+      "loss": 0.7965,
+      "step": 39028
+    },
+    {
+      "epoch": 0.9289964771969913,
+      "grad_norm": 0.03562419413506864,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39029
+    },
+    {
+      "epoch": 0.9290202799200229,
+      "grad_norm": 3.213525492887881,
+      "learning_rate": 2e-05,
+      "loss": 0.321,
+      "step": 39030
+    },
+    {
+      "epoch": 0.9290440826430544,
+      "grad_norm": 9.85657249286236,
+      "learning_rate": 2e-05,
+      "loss": 0.8155,
+      "step": 39031
+    },
+    {
+      "epoch": 0.9290678853660859,
+      "grad_norm": 0.029915725004502246,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39032
+    },
+    {
+      "epoch": 0.9290916880891174,
+      "grad_norm": 6.224766883769403,
+      "learning_rate": 2e-05,
+      "loss": 0.1479,
+      "step": 39033
+    },
+    {
+      "epoch": 0.929115490812149,
+      "grad_norm": 3.8626801378045954,
+      "learning_rate": 2e-05,
+      "loss": 0.5713,
+      "step": 39034
+    },
+    {
+      "epoch": 0.9291392935351804,
+      "grad_norm": 6.192155812593335,
+      "learning_rate": 2e-05,
+      "loss": 0.7113,
+      "step": 39035
+    },
+    {
+      "epoch": 0.9291630962582119,
+      "grad_norm": 7.140748340273997,
+      "learning_rate": 2e-05,
+      "loss": 1.5017,
+      "step": 39036
+    },
+    {
+      "epoch": 0.9291868989812434,
+      "grad_norm": 1.923923691772932,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 39037
+    },
+    {
+      "epoch": 0.929210701704275,
+      "grad_norm": 2.9178720116159895,
+      "learning_rate": 2e-05,
+      "loss": 0.5731,
+      "step": 39038
+    },
+    {
+      "epoch": 0.9292345044273065,
+      "grad_norm": 0.029773044865617843,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39039
+    },
+    {
+      "epoch": 0.929258307150338,
+      "grad_norm": 1.0641550891863891,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 39040
+    },
+    {
+      "epoch": 0.9292821098733696,
+      "grad_norm": 0.7924398620759712,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 39041
+    },
+    {
+      "epoch": 0.929305912596401,
+      "grad_norm": 3.134087064048027,
+      "learning_rate": 2e-05,
+      "loss": 0.0497,
+      "step": 39042
+    },
+    {
+      "epoch": 0.9293297153194325,
+      "grad_norm": 2.724634668460151,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 39043
+    },
+    {
+      "epoch": 0.929353518042464,
+      "grad_norm": 4.7539512564140285,
+      "learning_rate": 2e-05,
+      "loss": 0.9108,
+      "step": 39044
+    },
+    {
+      "epoch": 0.9293773207654956,
+      "grad_norm": 3.011303595901135,
+      "learning_rate": 2e-05,
+      "loss": 0.3732,
+      "step": 39045
+    },
+    {
+      "epoch": 0.9294011234885271,
+      "grad_norm": 0.4979702559804675,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 39046
+    },
+    {
+      "epoch": 0.9294249262115586,
+      "grad_norm": 4.7955124882019815,
+      "learning_rate": 2e-05,
+      "loss": 0.0744,
+      "step": 39047
+    },
+    {
+      "epoch": 0.9294487289345901,
+      "grad_norm": 0.41361788049620907,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 39048
+    },
+    {
+      "epoch": 0.9294725316576217,
+      "grad_norm": 0.682216399146061,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 39049
+    },
+    {
+      "epoch": 0.9294963343806532,
+      "grad_norm": 2.5172842793395525,
+      "learning_rate": 2e-05,
+      "loss": 0.4832,
+      "step": 39050
+    },
+    {
+      "epoch": 0.9295201371036846,
+      "grad_norm": 0.32974873256195747,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 39051
+    },
+    {
+      "epoch": 0.9295439398267161,
+      "grad_norm": 1.0475601808590693,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 39052
+    },
+    {
+      "epoch": 0.9295677425497477,
+      "grad_norm": 1.5946107650036836,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 39053
+    },
+    {
+      "epoch": 0.9295915452727792,
+      "grad_norm": 1.4867452064118158,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 39054
+    },
+    {
+      "epoch": 0.9296153479958107,
+      "grad_norm": 0.047937008703459384,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 39055
+    },
+    {
+      "epoch": 0.9296391507188423,
+      "grad_norm": 8.977276590172023,
+      "learning_rate": 2e-05,
+      "loss": 0.1757,
+      "step": 39056
+    },
+    {
+      "epoch": 0.9296629534418738,
+      "grad_norm": 3.8862196872294392,
+      "learning_rate": 2e-05,
+      "loss": 0.947,
+      "step": 39057
+    },
+    {
+      "epoch": 0.9296867561649053,
+      "grad_norm": 7.457883952767069,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 39058
+    },
+    {
+      "epoch": 0.9297105588879367,
+      "grad_norm": 0.028114755062730288,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39059
+    },
+    {
+      "epoch": 0.9297343616109683,
+      "grad_norm": 3.2456967953442106,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 39060
+    },
+    {
+      "epoch": 0.9297581643339998,
+      "grad_norm": 4.055830539129065,
+      "learning_rate": 2e-05,
+      "loss": 0.6623,
+      "step": 39061
+    },
+    {
+      "epoch": 0.9297819670570313,
+      "grad_norm": 1.7126260864018965,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 39062
+    },
+    {
+      "epoch": 0.9298057697800628,
+      "grad_norm": 1.6485509752354206,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 39063
+    },
+    {
+      "epoch": 0.9298295725030944,
+      "grad_norm": 3.900082953556974,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 39064
+    },
+    {
+      "epoch": 0.9298533752261259,
+      "grad_norm": 5.575898205654714,
+      "learning_rate": 2e-05,
+      "loss": 0.6584,
+      "step": 39065
+    },
+    {
+      "epoch": 0.9298771779491574,
+      "grad_norm": 0.23442911201732744,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 39066
+    },
+    {
+      "epoch": 0.9299009806721888,
+      "grad_norm": 1.8193177206930575,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 39067
+    },
+    {
+      "epoch": 0.9299247833952204,
+      "grad_norm": 4.411374999454535,
+      "learning_rate": 2e-05,
+      "loss": 0.4029,
+      "step": 39068
+    },
+    {
+      "epoch": 0.9299485861182519,
+      "grad_norm": 3.8405926507488335,
+      "learning_rate": 2e-05,
+      "loss": 0.68,
+      "step": 39069
+    },
+    {
+      "epoch": 0.9299723888412834,
+      "grad_norm": 0.4018559785703116,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 39070
+    },
+    {
+      "epoch": 0.929996191564315,
+      "grad_norm": 0.511738388675034,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 39071
+    },
+    {
+      "epoch": 0.9300199942873465,
+      "grad_norm": 0.6393502194466404,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 39072
+    },
+    {
+      "epoch": 0.930043797010378,
+      "grad_norm": 3.67213859661903,
+      "learning_rate": 2e-05,
+      "loss": 0.2492,
+      "step": 39073
+    },
+    {
+      "epoch": 0.9300675997334095,
+      "grad_norm": 20.92164655281675,
+      "learning_rate": 2e-05,
+      "loss": 0.9522,
+      "step": 39074
+    },
+    {
+      "epoch": 0.9300914024564411,
+      "grad_norm": 6.371098512641763,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 39075
+    },
+    {
+      "epoch": 0.9301152051794725,
+      "grad_norm": 8.220588110237328,
+      "learning_rate": 2e-05,
+      "loss": 0.4867,
+      "step": 39076
+    },
+    {
+      "epoch": 0.930139007902504,
+      "grad_norm": 3.4609022370981894,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 39077
+    },
+    {
+      "epoch": 0.9301628106255355,
+      "grad_norm": 1.892136677461419,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 39078
+    },
+    {
+      "epoch": 0.9301866133485671,
+      "grad_norm": 5.407631363682782,
+      "learning_rate": 2e-05,
+      "loss": 0.716,
+      "step": 39079
+    },
+    {
+      "epoch": 0.9302104160715986,
+      "grad_norm": 5.237413030467239,
+      "learning_rate": 2e-05,
+      "loss": 0.7391,
+      "step": 39080
+    },
+    {
+      "epoch": 0.9302342187946301,
+      "grad_norm": 3.0450546337038977,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 39081
+    },
+    {
+      "epoch": 0.9302580215176616,
+      "grad_norm": 3.9229221572875175,
+      "learning_rate": 2e-05,
+      "loss": 0.4475,
+      "step": 39082
+    },
+    {
+      "epoch": 0.9302818242406932,
+      "grad_norm": 3.8661822046395065,
+      "learning_rate": 2e-05,
+      "loss": 0.8627,
+      "step": 39083
+    },
+    {
+      "epoch": 0.9303056269637247,
+      "grad_norm": 0.3035978155055729,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 39084
+    },
+    {
+      "epoch": 0.9303294296867561,
+      "grad_norm": 16.6772749237589,
+      "learning_rate": 2e-05,
+      "loss": 0.5016,
+      "step": 39085
+    },
+    {
+      "epoch": 0.9303532324097877,
+      "grad_norm": 0.0374236824036046,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39086
+    },
+    {
+      "epoch": 0.9303770351328192,
+      "grad_norm": 3.2983845937304324,
+      "learning_rate": 2e-05,
+      "loss": 0.4809,
+      "step": 39087
+    },
+    {
+      "epoch": 0.9304008378558507,
+      "grad_norm": 0.5776965515104462,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 39088
+    },
+    {
+      "epoch": 0.9304246405788822,
+      "grad_norm": 17.54342564019468,
+      "learning_rate": 2e-05,
+      "loss": 0.381,
+      "step": 39089
+    },
+    {
+      "epoch": 0.9304484433019138,
+      "grad_norm": 2.8985116742456314,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 39090
+    },
+    {
+      "epoch": 0.9304722460249453,
+      "grad_norm": 4.085390520634832,
+      "learning_rate": 2e-05,
+      "loss": 0.6769,
+      "step": 39091
+    },
+    {
+      "epoch": 0.9304960487479768,
+      "grad_norm": 7.102431606778525,
+      "learning_rate": 2e-05,
+      "loss": 0.3284,
+      "step": 39092
+    },
+    {
+      "epoch": 0.9305198514710082,
+      "grad_norm": 1.5961753771790024,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 39093
+    },
+    {
+      "epoch": 0.9305436541940398,
+      "grad_norm": 0.03446089633639923,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39094
+    },
+    {
+      "epoch": 0.9305674569170713,
+      "grad_norm": 0.020849573946681542,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39095
+    },
+    {
+      "epoch": 0.9305912596401028,
+      "grad_norm": 6.490293449756395,
+      "learning_rate": 2e-05,
+      "loss": 0.6791,
+      "step": 39096
+    },
+    {
+      "epoch": 0.9306150623631343,
+      "grad_norm": 7.577838576960817,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 39097
+    },
+    {
+      "epoch": 0.9306388650861659,
+      "grad_norm": 3.7352793301063243,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 39098
+    },
+    {
+      "epoch": 0.9306626678091974,
+      "grad_norm": 3.5612287286951827,
+      "learning_rate": 2e-05,
+      "loss": 0.8269,
+      "step": 39099
+    },
+    {
+      "epoch": 0.9306864705322289,
+      "grad_norm": 2.380121338279752,
+      "learning_rate": 2e-05,
+      "loss": 0.0333,
+      "step": 39100
+    },
+    {
+      "epoch": 0.9307102732552605,
+      "grad_norm": 2.929122846845209,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 39101
+    },
+    {
+      "epoch": 0.9307340759782919,
+      "grad_norm": 2.9578724895183033,
+      "learning_rate": 2e-05,
+      "loss": 0.4965,
+      "step": 39102
+    },
+    {
+      "epoch": 0.9307578787013234,
+      "grad_norm": 0.2919757983127444,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 39103
+    },
+    {
+      "epoch": 0.9307816814243549,
+      "grad_norm": 1.5609301647624971,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 39104
+    },
+    {
+      "epoch": 0.9308054841473865,
+      "grad_norm": 1.1518602646727805,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 39105
+    },
+    {
+      "epoch": 0.930829286870418,
+      "grad_norm": 5.703438218111593,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 39106
+    },
+    {
+      "epoch": 0.9308530895934495,
+      "grad_norm": 0.3165596330929196,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 39107
+    },
+    {
+      "epoch": 0.930876892316481,
+      "grad_norm": 0.06664781464748205,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39108
+    },
+    {
+      "epoch": 0.9309006950395126,
+      "grad_norm": 10.188112611043447,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 39109
+    },
+    {
+      "epoch": 0.930924497762544,
+      "grad_norm": 5.692463149122114,
+      "learning_rate": 2e-05,
+      "loss": 0.6783,
+      "step": 39110
+    },
+    {
+      "epoch": 0.9309483004855755,
+      "grad_norm": 0.41288696614789405,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 39111
+    },
+    {
+      "epoch": 0.930972103208607,
+      "grad_norm": 2.0662244968354258,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 39112
+    },
+    {
+      "epoch": 0.9309959059316386,
+      "grad_norm": 4.16592270145388,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 39113
+    },
+    {
+      "epoch": 0.9310197086546701,
+      "grad_norm": 5.681236700081603,
+      "learning_rate": 2e-05,
+      "loss": 0.762,
+      "step": 39114
+    },
+    {
+      "epoch": 0.9310435113777016,
+      "grad_norm": 2.997680528134682,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 39115
+    },
+    {
+      "epoch": 0.9310673141007332,
+      "grad_norm": 2.305156485363066,
+      "learning_rate": 2e-05,
+      "loss": 0.394,
+      "step": 39116
+    },
+    {
+      "epoch": 0.9310911168237647,
+      "grad_norm": 8.229989754792891,
+      "learning_rate": 2e-05,
+      "loss": 1.1273,
+      "step": 39117
+    },
+    {
+      "epoch": 0.9311149195467961,
+      "grad_norm": 2.3294267316208264,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 39118
+    },
+    {
+      "epoch": 0.9311387222698276,
+      "grad_norm": 6.541359741276948,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 39119
+    },
+    {
+      "epoch": 0.9311625249928592,
+      "grad_norm": 0.4107346436807199,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 39120
+    },
+    {
+      "epoch": 0.9311863277158907,
+      "grad_norm": 3.4649701932384747,
+      "learning_rate": 2e-05,
+      "loss": 0.4919,
+      "step": 39121
+    },
+    {
+      "epoch": 0.9312101304389222,
+      "grad_norm": 28.372342727187032,
+      "learning_rate": 2e-05,
+      "loss": 0.8121,
+      "step": 39122
+    },
+    {
+      "epoch": 0.9312339331619537,
+      "grad_norm": 3.185747107565477,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 39123
+    },
+    {
+      "epoch": 0.9312577358849853,
+      "grad_norm": 1.4910032430046258,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 39124
+    },
+    {
+      "epoch": 0.9312815386080168,
+      "grad_norm": 0.05544112356832145,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39125
+    },
+    {
+      "epoch": 0.9313053413310483,
+      "grad_norm": 3.1666999386885784,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 39126
+    },
+    {
+      "epoch": 0.9313291440540797,
+      "grad_norm": 7.634854592935084,
+      "learning_rate": 2e-05,
+      "loss": 1.0532,
+      "step": 39127
+    },
+    {
+      "epoch": 0.9313529467771113,
+      "grad_norm": 5.3067744044863785,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 39128
+    },
+    {
+      "epoch": 0.9313767495001428,
+      "grad_norm": 4.584186490941309,
+      "learning_rate": 2e-05,
+      "loss": 0.0809,
+      "step": 39129
+    },
+    {
+      "epoch": 0.9314005522231743,
+      "grad_norm": 0.09143748508974055,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39130
+    },
+    {
+      "epoch": 0.9314243549462059,
+      "grad_norm": 2.9819231270809556,
+      "learning_rate": 2e-05,
+      "loss": 0.2517,
+      "step": 39131
+    },
+    {
+      "epoch": 0.9314481576692374,
+      "grad_norm": 4.971546944286351,
+      "learning_rate": 2e-05,
+      "loss": 0.4788,
+      "step": 39132
+    },
+    {
+      "epoch": 0.9314719603922689,
+      "grad_norm": 6.551861151002089,
+      "learning_rate": 2e-05,
+      "loss": 1.2618,
+      "step": 39133
+    },
+    {
+      "epoch": 0.9314957631153004,
+      "grad_norm": 3.3267849675249224,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 39134
+    },
+    {
+      "epoch": 0.931519565838332,
+      "grad_norm": 3.698998848776033,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 39135
+    },
+    {
+      "epoch": 0.9315433685613634,
+      "grad_norm": 12.523807609339414,
+      "learning_rate": 2e-05,
+      "loss": 0.4989,
+      "step": 39136
+    },
+    {
+      "epoch": 0.9315671712843949,
+      "grad_norm": 2.257908137866933,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 39137
+    },
+    {
+      "epoch": 0.9315909740074264,
+      "grad_norm": 0.4420515771924325,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 39138
+    },
+    {
+      "epoch": 0.931614776730458,
+      "grad_norm": 4.2350478666274505,
+      "learning_rate": 2e-05,
+      "loss": 1.0598,
+      "step": 39139
+    },
+    {
+      "epoch": 0.9316385794534895,
+      "grad_norm": 3.399588801019135,
+      "learning_rate": 2e-05,
+      "loss": 0.3714,
+      "step": 39140
+    },
+    {
+      "epoch": 0.931662382176521,
+      "grad_norm": 5.869109135808603,
+      "learning_rate": 2e-05,
+      "loss": 0.2507,
+      "step": 39141
+    },
+    {
+      "epoch": 0.9316861848995525,
+      "grad_norm": 4.686234773453889,
+      "learning_rate": 2e-05,
+      "loss": 0.3208,
+      "step": 39142
+    },
+    {
+      "epoch": 0.9317099876225841,
+      "grad_norm": 0.11591193121513661,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 39143
+    },
+    {
+      "epoch": 0.9317337903456155,
+      "grad_norm": 2.668661166770047,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 39144
+    },
+    {
+      "epoch": 0.931757593068647,
+      "grad_norm": 0.06645931877143493,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39145
+    },
+    {
+      "epoch": 0.9317813957916785,
+      "grad_norm": 5.554425251348316,
+      "learning_rate": 2e-05,
+      "loss": 0.7183,
+      "step": 39146
+    },
+    {
+      "epoch": 0.9318051985147101,
+      "grad_norm": 2.179153622435618,
+      "learning_rate": 2e-05,
+      "loss": 0.0545,
+      "step": 39147
+    },
+    {
+      "epoch": 0.9318290012377416,
+      "grad_norm": 3.7483149628500017,
+      "learning_rate": 2e-05,
+      "loss": 0.3897,
+      "step": 39148
+    },
+    {
+      "epoch": 0.9318528039607731,
+      "grad_norm": 2.227060456063295,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 39149
+    },
+    {
+      "epoch": 0.9318766066838047,
+      "grad_norm": 1.2030980031369023,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 39150
+    },
+    {
+      "epoch": 0.9319004094068362,
+      "grad_norm": 2.4824984231180003,
+      "learning_rate": 2e-05,
+      "loss": 0.4628,
+      "step": 39151
+    },
+    {
+      "epoch": 0.9319242121298676,
+      "grad_norm": 4.858121253780652,
+      "learning_rate": 2e-05,
+      "loss": 0.6767,
+      "step": 39152
+    },
+    {
+      "epoch": 0.9319480148528991,
+      "grad_norm": 4.7643205665602935,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 39153
+    },
+    {
+      "epoch": 0.9319718175759307,
+      "grad_norm": 4.671897345254389,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 39154
+    },
+    {
+      "epoch": 0.9319956202989622,
+      "grad_norm": 9.92822512554423,
+      "learning_rate": 2e-05,
+      "loss": 0.6574,
+      "step": 39155
+    },
+    {
+      "epoch": 0.9320194230219937,
+      "grad_norm": 3.152092732622288,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 39156
+    },
+    {
+      "epoch": 0.9320432257450252,
+      "grad_norm": 1.268491192499407,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 39157
+    },
+    {
+      "epoch": 0.9320670284680568,
+      "grad_norm": 3.4255028623249424,
+      "learning_rate": 2e-05,
+      "loss": 0.9311,
+      "step": 39158
+    },
+    {
+      "epoch": 0.9320908311910883,
+      "grad_norm": 4.759392420841264,
+      "learning_rate": 2e-05,
+      "loss": 0.7469,
+      "step": 39159
+    },
+    {
+      "epoch": 0.9321146339141198,
+      "grad_norm": 2.0318954158474702,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 39160
+    },
+    {
+      "epoch": 0.9321384366371512,
+      "grad_norm": 3.392552711058173,
+      "learning_rate": 2e-05,
+      "loss": 0.473,
+      "step": 39161
+    },
+    {
+      "epoch": 0.9321622393601828,
+      "grad_norm": 6.693154441510576,
+      "learning_rate": 2e-05,
+      "loss": 1.283,
+      "step": 39162
+    },
+    {
+      "epoch": 0.9321860420832143,
+      "grad_norm": 0.2024536047223643,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 39163
+    },
+    {
+      "epoch": 0.9322098448062458,
+      "grad_norm": 0.1733267706923728,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 39164
+    },
+    {
+      "epoch": 0.9322336475292774,
+      "grad_norm": 11.208507263199774,
+      "learning_rate": 2e-05,
+      "loss": 0.1563,
+      "step": 39165
+    },
+    {
+      "epoch": 0.9322574502523089,
+      "grad_norm": 0.8496921220292242,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 39166
+    },
+    {
+      "epoch": 0.9322812529753404,
+      "grad_norm": 3.6225222116943208,
+      "learning_rate": 2e-05,
+      "loss": 0.8026,
+      "step": 39167
+    },
+    {
+      "epoch": 0.9323050556983719,
+      "grad_norm": 3.5645259133588563,
+      "learning_rate": 2e-05,
+      "loss": 0.5092,
+      "step": 39168
+    },
+    {
+      "epoch": 0.9323288584214035,
+      "grad_norm": 0.2223559286014371,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39169
+    },
+    {
+      "epoch": 0.9323526611444349,
+      "grad_norm": 6.620014336953427,
+      "learning_rate": 2e-05,
+      "loss": 0.2628,
+      "step": 39170
+    },
+    {
+      "epoch": 0.9323764638674664,
+      "grad_norm": 9.726921422649621,
+      "learning_rate": 2e-05,
+      "loss": 0.9205,
+      "step": 39171
+    },
+    {
+      "epoch": 0.9324002665904979,
+      "grad_norm": 3.589204845111209,
+      "learning_rate": 2e-05,
+      "loss": 0.333,
+      "step": 39172
+    },
+    {
+      "epoch": 0.9324240693135295,
+      "grad_norm": 4.156066022994001,
+      "learning_rate": 2e-05,
+      "loss": 0.7008,
+      "step": 39173
+    },
+    {
+      "epoch": 0.932447872036561,
+      "grad_norm": 3.4222278573584255,
+      "learning_rate": 2e-05,
+      "loss": 0.0511,
+      "step": 39174
+    },
+    {
+      "epoch": 0.9324716747595925,
+      "grad_norm": 0.27745897011275356,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 39175
+    },
+    {
+      "epoch": 0.932495477482624,
+      "grad_norm": 0.8517727938950136,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 39176
+    },
+    {
+      "epoch": 0.9325192802056556,
+      "grad_norm": 2.3153506551495515,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 39177
+    },
+    {
+      "epoch": 0.932543082928687,
+      "grad_norm": 2.138689535212276,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 39178
+    },
+    {
+      "epoch": 0.9325668856517185,
+      "grad_norm": 0.801499242770403,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 39179
+    },
+    {
+      "epoch": 0.9325906883747501,
+      "grad_norm": 0.6862118150255704,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 39180
+    },
+    {
+      "epoch": 0.9326144910977816,
+      "grad_norm": 9.046940016492695,
+      "learning_rate": 2e-05,
+      "loss": 1.2374,
+      "step": 39181
+    },
+    {
+      "epoch": 0.9326382938208131,
+      "grad_norm": 4.843594421210597,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 39182
+    },
+    {
+      "epoch": 0.9326620965438446,
+      "grad_norm": 2.4675598649270323,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 39183
+    },
+    {
+      "epoch": 0.9326858992668762,
+      "grad_norm": 0.3596032070530481,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 39184
+    },
+    {
+      "epoch": 0.9327097019899077,
+      "grad_norm": 4.7906488350942125,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 39185
+    },
+    {
+      "epoch": 0.9327335047129391,
+      "grad_norm": 0.33045442690131377,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 39186
+    },
+    {
+      "epoch": 0.9327573074359706,
+      "grad_norm": 13.519322354994276,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 39187
+    },
+    {
+      "epoch": 0.9327811101590022,
+      "grad_norm": 11.37286736046362,
+      "learning_rate": 2e-05,
+      "loss": 0.2303,
+      "step": 39188
+    },
+    {
+      "epoch": 0.9328049128820337,
+      "grad_norm": 2.768723692740827,
+      "learning_rate": 2e-05,
+      "loss": 0.4769,
+      "step": 39189
+    },
+    {
+      "epoch": 0.9328287156050652,
+      "grad_norm": 0.8434446083154651,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 39190
+    },
+    {
+      "epoch": 0.9328525183280967,
+      "grad_norm": 8.387568303687159,
+      "learning_rate": 2e-05,
+      "loss": 0.5541,
+      "step": 39191
+    },
+    {
+      "epoch": 0.9328763210511283,
+      "grad_norm": 0.06133069878466471,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39192
+    },
+    {
+      "epoch": 0.9329001237741598,
+      "grad_norm": 4.832945974711742,
+      "learning_rate": 2e-05,
+      "loss": 0.5886,
+      "step": 39193
+    },
+    {
+      "epoch": 0.9329239264971912,
+      "grad_norm": 0.12501158867765777,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 39194
+    },
+    {
+      "epoch": 0.9329477292202228,
+      "grad_norm": 4.129484087146261,
+      "learning_rate": 2e-05,
+      "loss": 0.9018,
+      "step": 39195
+    },
+    {
+      "epoch": 0.9329715319432543,
+      "grad_norm": 6.68806461626929,
+      "learning_rate": 2e-05,
+      "loss": 1.2583,
+      "step": 39196
+    },
+    {
+      "epoch": 0.9329953346662858,
+      "grad_norm": 0.10778376108707186,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 39197
+    },
+    {
+      "epoch": 0.9330191373893173,
+      "grad_norm": 0.00589431567676481,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 39198
+    },
+    {
+      "epoch": 0.9330429401123489,
+      "grad_norm": 6.004554675195282,
+      "learning_rate": 2e-05,
+      "loss": 0.7417,
+      "step": 39199
+    },
+    {
+      "epoch": 0.9330667428353804,
+      "grad_norm": 3.7203485953030593,
+      "learning_rate": 2e-05,
+      "loss": 0.3162,
+      "step": 39200
+    },
+    {
+      "epoch": 0.9330905455584119,
+      "grad_norm": 0.12212200226159368,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39201
+    },
+    {
+      "epoch": 0.9331143482814434,
+      "grad_norm": 3.6799583187511464,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 39202
+    },
+    {
+      "epoch": 0.933138151004475,
+      "grad_norm": 1.1680710636792913,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 39203
+    },
+    {
+      "epoch": 0.9331619537275064,
+      "grad_norm": 3.526482552816014,
+      "learning_rate": 2e-05,
+      "loss": 0.4827,
+      "step": 39204
+    },
+    {
+      "epoch": 0.9331857564505379,
+      "grad_norm": 4.24170466656456,
+      "learning_rate": 2e-05,
+      "loss": 0.5941,
+      "step": 39205
+    },
+    {
+      "epoch": 0.9332095591735694,
+      "grad_norm": 0.5111851693691178,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 39206
+    },
+    {
+      "epoch": 0.933233361896601,
+      "grad_norm": 2.7879814073938616,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 39207
+    },
+    {
+      "epoch": 0.9332571646196325,
+      "grad_norm": 17.61666249205153,
+      "learning_rate": 2e-05,
+      "loss": 0.9546,
+      "step": 39208
+    },
+    {
+      "epoch": 0.933280967342664,
+      "grad_norm": 0.13511965190734093,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 39209
+    },
+    {
+      "epoch": 0.9333047700656956,
+      "grad_norm": 0.09182738975118679,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39210
+    },
+    {
+      "epoch": 0.933328572788727,
+      "grad_norm": 3.07326651997257,
+      "learning_rate": 2e-05,
+      "loss": 0.3391,
+      "step": 39211
+    },
+    {
+      "epoch": 0.9333523755117585,
+      "grad_norm": 8.415667604037274,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 39212
+    },
+    {
+      "epoch": 0.93337617823479,
+      "grad_norm": 0.44494153514130363,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 39213
+    },
+    {
+      "epoch": 0.9333999809578216,
+      "grad_norm": 3.9136038562188853,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 39214
+    },
+    {
+      "epoch": 0.9334237836808531,
+      "grad_norm": 10.285865288077746,
+      "learning_rate": 2e-05,
+      "loss": 0.3515,
+      "step": 39215
+    },
+    {
+      "epoch": 0.9334475864038846,
+      "grad_norm": 0.01576875876900739,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 39216
+    },
+    {
+      "epoch": 0.9334713891269161,
+      "grad_norm": 5.11363337543483,
+      "learning_rate": 2e-05,
+      "loss": 0.5532,
+      "step": 39217
+    },
+    {
+      "epoch": 0.9334951918499477,
+      "grad_norm": 44.648539357090485,
+      "learning_rate": 2e-05,
+      "loss": 0.7807,
+      "step": 39218
+    },
+    {
+      "epoch": 0.9335189945729792,
+      "grad_norm": 4.984730500895901,
+      "learning_rate": 2e-05,
+      "loss": 0.776,
+      "step": 39219
+    },
+    {
+      "epoch": 0.9335427972960106,
+      "grad_norm": 4.175270577596727,
+      "learning_rate": 2e-05,
+      "loss": 0.4067,
+      "step": 39220
+    },
+    {
+      "epoch": 0.9335666000190421,
+      "grad_norm": 0.05957410233410952,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 39221
+    },
+    {
+      "epoch": 0.9335904027420737,
+      "grad_norm": 2.2782521219691247,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 39222
+    },
+    {
+      "epoch": 0.9336142054651052,
+      "grad_norm": 8.646943685216714,
+      "learning_rate": 2e-05,
+      "loss": 0.7977,
+      "step": 39223
+    },
+    {
+      "epoch": 0.9336380081881367,
+      "grad_norm": 0.14578902157771104,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 39224
+    },
+    {
+      "epoch": 0.9336618109111683,
+      "grad_norm": 3.5862363785087004,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 39225
+    },
+    {
+      "epoch": 0.9336856136341998,
+      "grad_norm": 7.049048355529697,
+      "learning_rate": 2e-05,
+      "loss": 1.1738,
+      "step": 39226
+    },
+    {
+      "epoch": 0.9337094163572313,
+      "grad_norm": 3.374526988145762,
+      "learning_rate": 2e-05,
+      "loss": 0.0865,
+      "step": 39227
+    },
+    {
+      "epoch": 0.9337332190802627,
+      "grad_norm": 2.309208328020879,
+      "learning_rate": 2e-05,
+      "loss": 0.2889,
+      "step": 39228
+    },
+    {
+      "epoch": 0.9337570218032943,
+      "grad_norm": 3.379823883715403,
+      "learning_rate": 2e-05,
+      "loss": 0.7082,
+      "step": 39229
+    },
+    {
+      "epoch": 0.9337808245263258,
+      "grad_norm": 7.26771238373268,
+      "learning_rate": 2e-05,
+      "loss": 1.4562,
+      "step": 39230
+    },
+    {
+      "epoch": 0.9338046272493573,
+      "grad_norm": 0.05528166627737961,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 39231
+    },
+    {
+      "epoch": 0.9338284299723888,
+      "grad_norm": 0.11343744575903808,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39232
+    },
+    {
+      "epoch": 0.9338522326954204,
+      "grad_norm": 5.157169319308133,
+      "learning_rate": 2e-05,
+      "loss": 0.113,
+      "step": 39233
+    },
+    {
+      "epoch": 0.9338760354184519,
+      "grad_norm": 5.148783287446727,
+      "learning_rate": 2e-05,
+      "loss": 1.069,
+      "step": 39234
+    },
+    {
+      "epoch": 0.9338998381414834,
+      "grad_norm": 10.68622258307872,
+      "learning_rate": 2e-05,
+      "loss": 0.47,
+      "step": 39235
+    },
+    {
+      "epoch": 0.9339236408645148,
+      "grad_norm": 4.533480316098036,
+      "learning_rate": 2e-05,
+      "loss": 0.9352,
+      "step": 39236
+    },
+    {
+      "epoch": 0.9339474435875464,
+      "grad_norm": 4.4860297622109755,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 39237
+    },
+    {
+      "epoch": 0.9339712463105779,
+      "grad_norm": 11.007341351669055,
+      "learning_rate": 2e-05,
+      "loss": 1.2505,
+      "step": 39238
+    },
+    {
+      "epoch": 0.9339950490336094,
+      "grad_norm": 2.512205493846447,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 39239
+    },
+    {
+      "epoch": 0.934018851756641,
+      "grad_norm": 1.8769011808508302,
+      "learning_rate": 2e-05,
+      "loss": 0.131,
+      "step": 39240
+    },
+    {
+      "epoch": 0.9340426544796725,
+      "grad_norm": 8.416853465303305,
+      "learning_rate": 2e-05,
+      "loss": 0.5044,
+      "step": 39241
+    },
+    {
+      "epoch": 0.934066457202704,
+      "grad_norm": 5.078005279263555,
+      "learning_rate": 2e-05,
+      "loss": 1.4704,
+      "step": 39242
+    },
+    {
+      "epoch": 0.9340902599257355,
+      "grad_norm": 5.105946103103986,
+      "learning_rate": 2e-05,
+      "loss": 1.1168,
+      "step": 39243
+    },
+    {
+      "epoch": 0.9341140626487671,
+      "grad_norm": 0.07994488661992759,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39244
+    },
+    {
+      "epoch": 0.9341378653717985,
+      "grad_norm": 6.143163158934203,
+      "learning_rate": 2e-05,
+      "loss": 0.4419,
+      "step": 39245
+    },
+    {
+      "epoch": 0.93416166809483,
+      "grad_norm": 0.12048427648871879,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 39246
+    },
+    {
+      "epoch": 0.9341854708178615,
+      "grad_norm": 2.5314322406626673,
+      "learning_rate": 2e-05,
+      "loss": 0.2495,
+      "step": 39247
+    },
+    {
+      "epoch": 0.9342092735408931,
+      "grad_norm": 0.7609451755469623,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 39248
+    },
+    {
+      "epoch": 0.9342330762639246,
+      "grad_norm": 2.8039074488529914,
+      "learning_rate": 2e-05,
+      "loss": 0.5775,
+      "step": 39249
+    },
+    {
+      "epoch": 0.9342568789869561,
+      "grad_norm": 3.756766061973903,
+      "learning_rate": 2e-05,
+      "loss": 0.7287,
+      "step": 39250
+    },
+    {
+      "epoch": 0.9342806817099876,
+      "grad_norm": 3.2233741055282215,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 39251
+    },
+    {
+      "epoch": 0.9343044844330192,
+      "grad_norm": 3.210083996969938,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 39252
+    },
+    {
+      "epoch": 0.9343282871560507,
+      "grad_norm": 1.3524053719588705,
+      "learning_rate": 2e-05,
+      "loss": 0.0205,
+      "step": 39253
+    },
+    {
+      "epoch": 0.9343520898790821,
+      "grad_norm": 0.3638156079011293,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39254
+    },
+    {
+      "epoch": 0.9343758926021137,
+      "grad_norm": 2.2497618582777363,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 39255
+    },
+    {
+      "epoch": 0.9343996953251452,
+      "grad_norm": 0.0428695909133063,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39256
+    },
+    {
+      "epoch": 0.9344234980481767,
+      "grad_norm": 0.18551169615644444,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 39257
+    },
+    {
+      "epoch": 0.9344473007712082,
+      "grad_norm": 0.17478929608451546,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 39258
+    },
+    {
+      "epoch": 0.9344711034942398,
+      "grad_norm": 8.177142020327127,
+      "learning_rate": 2e-05,
+      "loss": 1.3549,
+      "step": 39259
+    },
+    {
+      "epoch": 0.9344949062172713,
+      "grad_norm": 2.1611765399464486,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 39260
+    },
+    {
+      "epoch": 0.9345187089403028,
+      "grad_norm": 2.1406173470898393,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 39261
+    },
+    {
+      "epoch": 0.9345425116633342,
+      "grad_norm": 5.116389575335508,
+      "learning_rate": 2e-05,
+      "loss": 0.1627,
+      "step": 39262
+    },
+    {
+      "epoch": 0.9345663143863658,
+      "grad_norm": 5.403424261325384,
+      "learning_rate": 2e-05,
+      "loss": 0.679,
+      "step": 39263
+    },
+    {
+      "epoch": 0.9345901171093973,
+      "grad_norm": 5.5455768184888905,
+      "learning_rate": 2e-05,
+      "loss": 0.7929,
+      "step": 39264
+    },
+    {
+      "epoch": 0.9346139198324288,
+      "grad_norm": 15.038640783382615,
+      "learning_rate": 2e-05,
+      "loss": 1.0411,
+      "step": 39265
+    },
+    {
+      "epoch": 0.9346377225554603,
+      "grad_norm": 3.566021700640646,
+      "learning_rate": 2e-05,
+      "loss": 0.4478,
+      "step": 39266
+    },
+    {
+      "epoch": 0.9346615252784919,
+      "grad_norm": 0.389032659684542,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 39267
+    },
+    {
+      "epoch": 0.9346853280015234,
+      "grad_norm": 4.167102722060396,
+      "learning_rate": 2e-05,
+      "loss": 0.6818,
+      "step": 39268
+    },
+    {
+      "epoch": 0.9347091307245549,
+      "grad_norm": 7.590357299388814,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 39269
+    },
+    {
+      "epoch": 0.9347329334475865,
+      "grad_norm": 4.166900263169537,
+      "learning_rate": 2e-05,
+      "loss": 0.5626,
+      "step": 39270
+    },
+    {
+      "epoch": 0.9347567361706179,
+      "grad_norm": 3.39453963409294,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 39271
+    },
+    {
+      "epoch": 0.9347805388936494,
+      "grad_norm": 0.7366925999163859,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 39272
+    },
+    {
+      "epoch": 0.9348043416166809,
+      "grad_norm": 0.5881872193897429,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 39273
+    },
+    {
+      "epoch": 0.9348281443397125,
+      "grad_norm": 2.7942405411405757,
+      "learning_rate": 2e-05,
+      "loss": 0.5386,
+      "step": 39274
+    },
+    {
+      "epoch": 0.934851947062744,
+      "grad_norm": 3.436752414009138,
+      "learning_rate": 2e-05,
+      "loss": 0.5992,
+      "step": 39275
+    },
+    {
+      "epoch": 0.9348757497857755,
+      "grad_norm": 6.565558367082753,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 39276
+    },
+    {
+      "epoch": 0.934899552508807,
+      "grad_norm": 2.710423726702192,
+      "learning_rate": 2e-05,
+      "loss": 0.5942,
+      "step": 39277
+    },
+    {
+      "epoch": 0.9349233552318386,
+      "grad_norm": 5.263895628580839,
+      "learning_rate": 2e-05,
+      "loss": 0.8767,
+      "step": 39278
+    },
+    {
+      "epoch": 0.93494715795487,
+      "grad_norm": 0.27077711866967735,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 39279
+    },
+    {
+      "epoch": 0.9349709606779015,
+      "grad_norm": 0.08392906824735151,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39280
+    },
+    {
+      "epoch": 0.934994763400933,
+      "grad_norm": 5.161541133623826,
+      "learning_rate": 2e-05,
+      "loss": 0.2206,
+      "step": 39281
+    },
+    {
+      "epoch": 0.9350185661239646,
+      "grad_norm": 3.9252714015195833,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 39282
+    },
+    {
+      "epoch": 0.9350423688469961,
+      "grad_norm": 0.6681818012141137,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 39283
+    },
+    {
+      "epoch": 0.9350661715700276,
+      "grad_norm": 13.020236211708202,
+      "learning_rate": 2e-05,
+      "loss": 0.7259,
+      "step": 39284
+    },
+    {
+      "epoch": 0.9350899742930592,
+      "grad_norm": 10.062855082626884,
+      "learning_rate": 2e-05,
+      "loss": 0.5358,
+      "step": 39285
+    },
+    {
+      "epoch": 0.9351137770160907,
+      "grad_norm": 1.067867864626863,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 39286
+    },
+    {
+      "epoch": 0.9351375797391221,
+      "grad_norm": 0.9011499223310702,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 39287
+    },
+    {
+      "epoch": 0.9351613824621536,
+      "grad_norm": 0.07896467703322443,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 39288
+    },
+    {
+      "epoch": 0.9351851851851852,
+      "grad_norm": 1.337669028429145,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 39289
+    },
+    {
+      "epoch": 0.9352089879082167,
+      "grad_norm": 7.717359948668948,
+      "learning_rate": 2e-05,
+      "loss": 0.6849,
+      "step": 39290
+    },
+    {
+      "epoch": 0.9352327906312482,
+      "grad_norm": 3.734885198473883,
+      "learning_rate": 2e-05,
+      "loss": 0.6806,
+      "step": 39291
+    },
+    {
+      "epoch": 0.9352565933542797,
+      "grad_norm": 6.833807914789676,
+      "learning_rate": 2e-05,
+      "loss": 0.535,
+      "step": 39292
+    },
+    {
+      "epoch": 0.9352803960773113,
+      "grad_norm": 0.8220037376376753,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 39293
+    },
+    {
+      "epoch": 0.9353041988003428,
+      "grad_norm": 6.63190163031393,
+      "learning_rate": 2e-05,
+      "loss": 0.6117,
+      "step": 39294
+    },
+    {
+      "epoch": 0.9353280015233743,
+      "grad_norm": 3.3487049228949837,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 39295
+    },
+    {
+      "epoch": 0.9353518042464057,
+      "grad_norm": 6.08263586330996,
+      "learning_rate": 2e-05,
+      "loss": 0.4422,
+      "step": 39296
+    },
+    {
+      "epoch": 0.9353756069694373,
+      "grad_norm": 17.02774309854441,
+      "learning_rate": 2e-05,
+      "loss": 0.504,
+      "step": 39297
+    },
+    {
+      "epoch": 0.9353994096924688,
+      "grad_norm": 3.593325317150065,
+      "learning_rate": 2e-05,
+      "loss": 0.6837,
+      "step": 39298
+    },
+    {
+      "epoch": 0.9354232124155003,
+      "grad_norm": 0.13289374127129558,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 39299
+    },
+    {
+      "epoch": 0.9354470151385319,
+      "grad_norm": 4.113592572880918,
+      "learning_rate": 2e-05,
+      "loss": 0.6059,
+      "step": 39300
+    },
+    {
+      "epoch": 0.9354708178615634,
+      "grad_norm": 0.7322286239400638,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 39301
+    },
+    {
+      "epoch": 0.9354946205845949,
+      "grad_norm": 5.984030530436894,
+      "learning_rate": 2e-05,
+      "loss": 0.8954,
+      "step": 39302
+    },
+    {
+      "epoch": 0.9355184233076264,
+      "grad_norm": 8.25696985270972,
+      "learning_rate": 2e-05,
+      "loss": 1.6029,
+      "step": 39303
+    },
+    {
+      "epoch": 0.935542226030658,
+      "grad_norm": 3.556833935277497,
+      "learning_rate": 2e-05,
+      "loss": 0.04,
+      "step": 39304
+    },
+    {
+      "epoch": 0.9355660287536894,
+      "grad_norm": 0.015528118142267545,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 39305
+    },
+    {
+      "epoch": 0.9355898314767209,
+      "grad_norm": 2.163246328257503,
+      "learning_rate": 2e-05,
+      "loss": 0.2031,
+      "step": 39306
+    },
+    {
+      "epoch": 0.9356136341997524,
+      "grad_norm": 16.54494709957264,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 39307
+    },
+    {
+      "epoch": 0.935637436922784,
+      "grad_norm": 2.336538605034694,
+      "learning_rate": 2e-05,
+      "loss": 0.0943,
+      "step": 39308
+    },
+    {
+      "epoch": 0.9356612396458155,
+      "grad_norm": 3.1216173554929694,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 39309
+    },
+    {
+      "epoch": 0.935685042368847,
+      "grad_norm": 5.816767685235307,
+      "learning_rate": 2e-05,
+      "loss": 0.3451,
+      "step": 39310
+    },
+    {
+      "epoch": 0.9357088450918785,
+      "grad_norm": 0.14789060308501406,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 39311
+    },
+    {
+      "epoch": 0.9357326478149101,
+      "grad_norm": 1.3587827562996384,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 39312
+    },
+    {
+      "epoch": 0.9357564505379415,
+      "grad_norm": 0.763643625106304,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 39313
+    },
+    {
+      "epoch": 0.935780253260973,
+      "grad_norm": 1.4165453521824416,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 39314
+    },
+    {
+      "epoch": 0.9358040559840046,
+      "grad_norm": 2.3297474990656797,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 39315
+    },
+    {
+      "epoch": 0.9358278587070361,
+      "grad_norm": 5.13022228426782,
+      "learning_rate": 2e-05,
+      "loss": 0.1203,
+      "step": 39316
+    },
+    {
+      "epoch": 0.9358516614300676,
+      "grad_norm": 6.964810026200941,
+      "learning_rate": 2e-05,
+      "loss": 1.034,
+      "step": 39317
+    },
+    {
+      "epoch": 0.9358754641530991,
+      "grad_norm": 3.588683133564865,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 39318
+    },
+    {
+      "epoch": 0.9358992668761307,
+      "grad_norm": 6.338707572362798,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 39319
+    },
+    {
+      "epoch": 0.9359230695991622,
+      "grad_norm": 2.160676407554405,
+      "learning_rate": 2e-05,
+      "loss": 0.078,
+      "step": 39320
+    },
+    {
+      "epoch": 0.9359468723221936,
+      "grad_norm": 5.721051475263399,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 39321
+    },
+    {
+      "epoch": 0.9359706750452251,
+      "grad_norm": 0.425253059742116,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 39322
+    },
+    {
+      "epoch": 0.9359944777682567,
+      "grad_norm": 0.07550926716921867,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39323
+    },
+    {
+      "epoch": 0.9360182804912882,
+      "grad_norm": 2.5471604365835594,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 39324
+    },
+    {
+      "epoch": 0.9360420832143197,
+      "grad_norm": 3.480999577585237,
+      "learning_rate": 2e-05,
+      "loss": 0.5922,
+      "step": 39325
+    },
+    {
+      "epoch": 0.9360658859373512,
+      "grad_norm": 3.7528973791543803,
+      "learning_rate": 2e-05,
+      "loss": 0.3437,
+      "step": 39326
+    },
+    {
+      "epoch": 0.9360896886603828,
+      "grad_norm": 3.9691830521589795,
+      "learning_rate": 2e-05,
+      "loss": 0.5076,
+      "step": 39327
+    },
+    {
+      "epoch": 0.9361134913834143,
+      "grad_norm": 3.049326227259303,
+      "learning_rate": 2e-05,
+      "loss": 0.6958,
+      "step": 39328
+    },
+    {
+      "epoch": 0.9361372941064457,
+      "grad_norm": 4.45596797425047,
+      "learning_rate": 2e-05,
+      "loss": 0.4588,
+      "step": 39329
+    },
+    {
+      "epoch": 0.9361610968294773,
+      "grad_norm": 10.886430610832225,
+      "learning_rate": 2e-05,
+      "loss": 0.3004,
+      "step": 39330
+    },
+    {
+      "epoch": 0.9361848995525088,
+      "grad_norm": 3.9539292333835707,
+      "learning_rate": 2e-05,
+      "loss": 0.5804,
+      "step": 39331
+    },
+    {
+      "epoch": 0.9362087022755403,
+      "grad_norm": 20.947905590514853,
+      "learning_rate": 2e-05,
+      "loss": 0.5686,
+      "step": 39332
+    },
+    {
+      "epoch": 0.9362325049985718,
+      "grad_norm": 2.9123773429141986,
+      "learning_rate": 2e-05,
+      "loss": 0.2096,
+      "step": 39333
+    },
+    {
+      "epoch": 0.9362563077216034,
+      "grad_norm": 4.704220774414718,
+      "learning_rate": 2e-05,
+      "loss": 0.7517,
+      "step": 39334
+    },
+    {
+      "epoch": 0.9362801104446349,
+      "grad_norm": 8.746423973162951,
+      "learning_rate": 2e-05,
+      "loss": 0.6623,
+      "step": 39335
+    },
+    {
+      "epoch": 0.9363039131676664,
+      "grad_norm": 0.13354764339244174,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 39336
+    },
+    {
+      "epoch": 0.9363277158906979,
+      "grad_norm": 5.799312115018149,
+      "learning_rate": 2e-05,
+      "loss": 1.0858,
+      "step": 39337
+    },
+    {
+      "epoch": 0.9363515186137294,
+      "grad_norm": 7.512799572248595,
+      "learning_rate": 2e-05,
+      "loss": 0.1938,
+      "step": 39338
+    },
+    {
+      "epoch": 0.9363753213367609,
+      "grad_norm": 9.49464225929863,
+      "learning_rate": 2e-05,
+      "loss": 0.3167,
+      "step": 39339
+    },
+    {
+      "epoch": 0.9363991240597924,
+      "grad_norm": 1.2811857334093757,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 39340
+    },
+    {
+      "epoch": 0.9364229267828239,
+      "grad_norm": 2.249438671233556,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 39341
+    },
+    {
+      "epoch": 0.9364467295058555,
+      "grad_norm": 3.494922029331008,
+      "learning_rate": 2e-05,
+      "loss": 0.1658,
+      "step": 39342
+    },
+    {
+      "epoch": 0.936470532228887,
+      "grad_norm": 3.5322300254058216,
+      "learning_rate": 2e-05,
+      "loss": 0.8104,
+      "step": 39343
+    },
+    {
+      "epoch": 0.9364943349519185,
+      "grad_norm": 0.06641403428580339,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39344
+    },
+    {
+      "epoch": 0.9365181376749501,
+      "grad_norm": 4.075638508555318,
+      "learning_rate": 2e-05,
+      "loss": 0.7777,
+      "step": 39345
+    },
+    {
+      "epoch": 0.9365419403979816,
+      "grad_norm": 4.945910489337289,
+      "learning_rate": 2e-05,
+      "loss": 0.3211,
+      "step": 39346
+    },
+    {
+      "epoch": 0.936565743121013,
+      "grad_norm": 2.177999380851464,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 39347
+    },
+    {
+      "epoch": 0.9365895458440445,
+      "grad_norm": 0.0766293807966278,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 39348
+    },
+    {
+      "epoch": 0.9366133485670761,
+      "grad_norm": 4.073320067696051,
+      "learning_rate": 2e-05,
+      "loss": 0.7715,
+      "step": 39349
+    },
+    {
+      "epoch": 0.9366371512901076,
+      "grad_norm": 4.419666672947327,
+      "learning_rate": 2e-05,
+      "loss": 0.7056,
+      "step": 39350
+    },
+    {
+      "epoch": 0.9366609540131391,
+      "grad_norm": 8.040739321953845,
+      "learning_rate": 2e-05,
+      "loss": 1.0471,
+      "step": 39351
+    },
+    {
+      "epoch": 0.9366847567361706,
+      "grad_norm": 2.0332336675981457,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 39352
+    },
+    {
+      "epoch": 0.9367085594592022,
+      "grad_norm": 3.3514498087892597,
+      "learning_rate": 2e-05,
+      "loss": 0.68,
+      "step": 39353
+    },
+    {
+      "epoch": 0.9367323621822337,
+      "grad_norm": 1.3716758902451012,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 39354
+    },
+    {
+      "epoch": 0.9367561649052651,
+      "grad_norm": 3.6923860266060102,
+      "learning_rate": 2e-05,
+      "loss": 0.5983,
+      "step": 39355
+    },
+    {
+      "epoch": 0.9367799676282966,
+      "grad_norm": 1.6869110020967404,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 39356
+    },
+    {
+      "epoch": 0.9368037703513282,
+      "grad_norm": 2.575290536478073,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 39357
+    },
+    {
+      "epoch": 0.9368275730743597,
+      "grad_norm": 2.150568928673077,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 39358
+    },
+    {
+      "epoch": 0.9368513757973912,
+      "grad_norm": 6.068904205909309,
+      "learning_rate": 2e-05,
+      "loss": 1.1223,
+      "step": 39359
+    },
+    {
+      "epoch": 0.9368751785204228,
+      "grad_norm": 0.2769907448099125,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 39360
+    },
+    {
+      "epoch": 0.9368989812434543,
+      "grad_norm": 29.396112355592063,
+      "learning_rate": 2e-05,
+      "loss": 0.8251,
+      "step": 39361
+    },
+    {
+      "epoch": 0.9369227839664858,
+      "grad_norm": 0.24288609408033665,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 39362
+    },
+    {
+      "epoch": 0.9369465866895172,
+      "grad_norm": 9.547890735958077,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 39363
+    },
+    {
+      "epoch": 0.9369703894125488,
+      "grad_norm": 2.999813185586659,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 39364
+    },
+    {
+      "epoch": 0.9369941921355803,
+      "grad_norm": 0.0323113246303433,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39365
+    },
+    {
+      "epoch": 0.9370179948586118,
+      "grad_norm": 2.5643127333997144,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 39366
+    },
+    {
+      "epoch": 0.9370417975816433,
+      "grad_norm": 7.210679836347853,
+      "learning_rate": 2e-05,
+      "loss": 0.5712,
+      "step": 39367
+    },
+    {
+      "epoch": 0.9370656003046749,
+      "grad_norm": 0.10455711454648528,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39368
+    },
+    {
+      "epoch": 0.9370894030277064,
+      "grad_norm": 3.831102944512222,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 39369
+    },
+    {
+      "epoch": 0.9371132057507379,
+      "grad_norm": 0.07303886649536905,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39370
+    },
+    {
+      "epoch": 0.9371370084737694,
+      "grad_norm": 1.463619192702415,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 39371
+    },
+    {
+      "epoch": 0.937160811196801,
+      "grad_norm": 2.4153355929458646,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 39372
+    },
+    {
+      "epoch": 0.9371846139198324,
+      "grad_norm": 0.43091335956470056,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 39373
+    },
+    {
+      "epoch": 0.9372084166428639,
+      "grad_norm": 4.853643658321729,
+      "learning_rate": 2e-05,
+      "loss": 0.8006,
+      "step": 39374
+    },
+    {
+      "epoch": 0.9372322193658955,
+      "grad_norm": 2.914272529580369,
+      "learning_rate": 2e-05,
+      "loss": 0.3383,
+      "step": 39375
+    },
+    {
+      "epoch": 0.937256022088927,
+      "grad_norm": 0.4479101274025606,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 39376
+    },
+    {
+      "epoch": 0.9372798248119585,
+      "grad_norm": 5.470090290428674,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 39377
+    },
+    {
+      "epoch": 0.93730362753499,
+      "grad_norm": 0.059495955779341685,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39378
+    },
+    {
+      "epoch": 0.9373274302580216,
+      "grad_norm": 4.253906560520227,
+      "learning_rate": 2e-05,
+      "loss": 0.6292,
+      "step": 39379
+    },
+    {
+      "epoch": 0.937351232981053,
+      "grad_norm": 8.600442872897009,
+      "learning_rate": 2e-05,
+      "loss": 1.274,
+      "step": 39380
+    },
+    {
+      "epoch": 0.9373750357040845,
+      "grad_norm": 11.710245390531105,
+      "learning_rate": 2e-05,
+      "loss": 0.2646,
+      "step": 39381
+    },
+    {
+      "epoch": 0.937398838427116,
+      "grad_norm": 6.297184453078198,
+      "learning_rate": 2e-05,
+      "loss": 0.631,
+      "step": 39382
+    },
+    {
+      "epoch": 0.9374226411501476,
+      "grad_norm": 3.439506374718588,
+      "learning_rate": 2e-05,
+      "loss": 0.561,
+      "step": 39383
+    },
+    {
+      "epoch": 0.9374464438731791,
+      "grad_norm": 4.806919487863739,
+      "learning_rate": 2e-05,
+      "loss": 0.363,
+      "step": 39384
+    },
+    {
+      "epoch": 0.9374702465962106,
+      "grad_norm": 0.39278038311326974,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39385
+    },
+    {
+      "epoch": 0.9374940493192421,
+      "grad_norm": 4.154837823814847,
+      "learning_rate": 2e-05,
+      "loss": 0.5285,
+      "step": 39386
+    },
+    {
+      "epoch": 0.9375178520422737,
+      "grad_norm": 0.0446374570188334,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 39387
+    },
+    {
+      "epoch": 0.9375416547653052,
+      "grad_norm": 6.250969807341739,
+      "learning_rate": 2e-05,
+      "loss": 0.9503,
+      "step": 39388
+    },
+    {
+      "epoch": 0.9375654574883366,
+      "grad_norm": 12.117138353972747,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 39389
+    },
+    {
+      "epoch": 0.9375892602113681,
+      "grad_norm": 4.720568641670269,
+      "learning_rate": 2e-05,
+      "loss": 0.8551,
+      "step": 39390
+    },
+    {
+      "epoch": 0.9376130629343997,
+      "grad_norm": 0.24746720946350448,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 39391
+    },
+    {
+      "epoch": 0.9376368656574312,
+      "grad_norm": 0.0411047137919362,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39392
+    },
+    {
+      "epoch": 0.9376606683804627,
+      "grad_norm": 3.952490039158817,
+      "learning_rate": 2e-05,
+      "loss": 0.4327,
+      "step": 39393
+    },
+    {
+      "epoch": 0.9376844711034943,
+      "grad_norm": 3.9262943900271408,
+      "learning_rate": 2e-05,
+      "loss": 0.6974,
+      "step": 39394
+    },
+    {
+      "epoch": 0.9377082738265258,
+      "grad_norm": 3.342344558424553,
+      "learning_rate": 2e-05,
+      "loss": 0.5123,
+      "step": 39395
+    },
+    {
+      "epoch": 0.9377320765495573,
+      "grad_norm": 0.7018248176126896,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 39396
+    },
+    {
+      "epoch": 0.9377558792725887,
+      "grad_norm": 4.27883241313649,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 39397
+    },
+    {
+      "epoch": 0.9377796819956203,
+      "grad_norm": 0.2113801397650845,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 39398
+    },
+    {
+      "epoch": 0.9378034847186518,
+      "grad_norm": 2.8544295734857825,
+      "learning_rate": 2e-05,
+      "loss": 0.4462,
+      "step": 39399
+    },
+    {
+      "epoch": 0.9378272874416833,
+      "grad_norm": 3.3287898700357474,
+      "learning_rate": 2e-05,
+      "loss": 0.4487,
+      "step": 39400
+    },
+    {
+      "epoch": 0.9378510901647148,
+      "grad_norm": 5.623693374693356,
+      "learning_rate": 2e-05,
+      "loss": 0.234,
+      "step": 39401
+    },
+    {
+      "epoch": 0.9378748928877464,
+      "grad_norm": 6.595936070476638,
+      "learning_rate": 2e-05,
+      "loss": 0.7709,
+      "step": 39402
+    },
+    {
+      "epoch": 0.9378986956107779,
+      "grad_norm": 3.259149086798903,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 39403
+    },
+    {
+      "epoch": 0.9379224983338094,
+      "grad_norm": 4.596514000962044,
+      "learning_rate": 2e-05,
+      "loss": 1.5057,
+      "step": 39404
+    },
+    {
+      "epoch": 0.9379463010568408,
+      "grad_norm": 0.20175038473456586,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 39405
+    },
+    {
+      "epoch": 0.9379701037798724,
+      "grad_norm": 4.2834044984882444,
+      "learning_rate": 2e-05,
+      "loss": 0.7715,
+      "step": 39406
+    },
+    {
+      "epoch": 0.9379939065029039,
+      "grad_norm": 3.301145858583789,
+      "learning_rate": 2e-05,
+      "loss": 0.0247,
+      "step": 39407
+    },
+    {
+      "epoch": 0.9380177092259354,
+      "grad_norm": 0.20733227558617573,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39408
+    },
+    {
+      "epoch": 0.938041511948967,
+      "grad_norm": 6.766977467332312,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 39409
+    },
+    {
+      "epoch": 0.9380653146719985,
+      "grad_norm": 3.7832099337792213,
+      "learning_rate": 2e-05,
+      "loss": 0.8004,
+      "step": 39410
+    },
+    {
+      "epoch": 0.93808911739503,
+      "grad_norm": 5.248842377574822,
+      "learning_rate": 2e-05,
+      "loss": 0.5005,
+      "step": 39411
+    },
+    {
+      "epoch": 0.9381129201180615,
+      "grad_norm": 6.617101914278148,
+      "learning_rate": 2e-05,
+      "loss": 0.2175,
+      "step": 39412
+    },
+    {
+      "epoch": 0.9381367228410931,
+      "grad_norm": 3.472227762553736,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 39413
+    },
+    {
+      "epoch": 0.9381605255641245,
+      "grad_norm": 0.45595440199925935,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 39414
+    },
+    {
+      "epoch": 0.938184328287156,
+      "grad_norm": 11.93935041868332,
+      "learning_rate": 2e-05,
+      "loss": 0.7932,
+      "step": 39415
+    },
+    {
+      "epoch": 0.9382081310101875,
+      "grad_norm": 23.1319791992383,
+      "learning_rate": 2e-05,
+      "loss": 0.2705,
+      "step": 39416
+    },
+    {
+      "epoch": 0.9382319337332191,
+      "grad_norm": 2.429744559345691,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 39417
+    },
+    {
+      "epoch": 0.9382557364562506,
+      "grad_norm": 9.531822326088964,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 39418
+    },
+    {
+      "epoch": 0.9382795391792821,
+      "grad_norm": 0.1029902167590507,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39419
+    },
+    {
+      "epoch": 0.9383033419023136,
+      "grad_norm": 0.36405741181602647,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 39420
+    },
+    {
+      "epoch": 0.9383271446253452,
+      "grad_norm": 0.06287535994766019,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39421
+    },
+    {
+      "epoch": 0.9383509473483767,
+      "grad_norm": 0.31467803911808556,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 39422
+    },
+    {
+      "epoch": 0.9383747500714081,
+      "grad_norm": 2.90442351583512,
+      "learning_rate": 2e-05,
+      "loss": 0.1362,
+      "step": 39423
+    },
+    {
+      "epoch": 0.9383985527944397,
+      "grad_norm": 19.498873781464525,
+      "learning_rate": 2e-05,
+      "loss": 1.0208,
+      "step": 39424
+    },
+    {
+      "epoch": 0.9384223555174712,
+      "grad_norm": 0.25181072061225906,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39425
+    },
+    {
+      "epoch": 0.9384461582405027,
+      "grad_norm": 76.06601286188595,
+      "learning_rate": 2e-05,
+      "loss": 0.3031,
+      "step": 39426
+    },
+    {
+      "epoch": 0.9384699609635342,
+      "grad_norm": 8.95541794997993,
+      "learning_rate": 2e-05,
+      "loss": 0.5143,
+      "step": 39427
+    },
+    {
+      "epoch": 0.9384937636865658,
+      "grad_norm": 11.319673259770912,
+      "learning_rate": 2e-05,
+      "loss": 0.8207,
+      "step": 39428
+    },
+    {
+      "epoch": 0.9385175664095973,
+      "grad_norm": 0.013288901524871844,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 39429
+    },
+    {
+      "epoch": 0.9385413691326288,
+      "grad_norm": 2.543114154343906,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 39430
+    },
+    {
+      "epoch": 0.9385651718556602,
+      "grad_norm": 0.04387138090491493,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39431
+    },
+    {
+      "epoch": 0.9385889745786918,
+      "grad_norm": 11.765539588972482,
+      "learning_rate": 2e-05,
+      "loss": 0.766,
+      "step": 39432
+    },
+    {
+      "epoch": 0.9386127773017233,
+      "grad_norm": 2.4710488522911858,
+      "learning_rate": 2e-05,
+      "loss": 0.5732,
+      "step": 39433
+    },
+    {
+      "epoch": 0.9386365800247548,
+      "grad_norm": 5.468904437311217,
+      "learning_rate": 2e-05,
+      "loss": 0.1907,
+      "step": 39434
+    },
+    {
+      "epoch": 0.9386603827477863,
+      "grad_norm": 2.6737696261686548,
+      "learning_rate": 2e-05,
+      "loss": 0.4386,
+      "step": 39435
+    },
+    {
+      "epoch": 0.9386841854708179,
+      "grad_norm": 5.636588210143263,
+      "learning_rate": 2e-05,
+      "loss": 1.3186,
+      "step": 39436
+    },
+    {
+      "epoch": 0.9387079881938494,
+      "grad_norm": 3.1937703931858077,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 39437
+    },
+    {
+      "epoch": 0.9387317909168809,
+      "grad_norm": 0.5350019257851106,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 39438
+    },
+    {
+      "epoch": 0.9387555936399125,
+      "grad_norm": 23.75535084348002,
+      "learning_rate": 2e-05,
+      "loss": 1.1085,
+      "step": 39439
+    },
+    {
+      "epoch": 0.9387793963629439,
+      "grad_norm": 0.3807520524161558,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 39440
+    },
+    {
+      "epoch": 0.9388031990859754,
+      "grad_norm": 4.2540873820874925,
+      "learning_rate": 2e-05,
+      "loss": 0.5096,
+      "step": 39441
+    },
+    {
+      "epoch": 0.9388270018090069,
+      "grad_norm": 0.984259478733288,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 39442
+    },
+    {
+      "epoch": 0.9388508045320385,
+      "grad_norm": 5.232347009884196,
+      "learning_rate": 2e-05,
+      "loss": 0.9569,
+      "step": 39443
+    },
+    {
+      "epoch": 0.93887460725507,
+      "grad_norm": 2.8203379241253215,
+      "learning_rate": 2e-05,
+      "loss": 0.2706,
+      "step": 39444
+    },
+    {
+      "epoch": 0.9388984099781015,
+      "grad_norm": 4.551648405602671,
+      "learning_rate": 2e-05,
+      "loss": 1.1373,
+      "step": 39445
+    },
+    {
+      "epoch": 0.938922212701133,
+      "grad_norm": 6.929391427124746,
+      "learning_rate": 2e-05,
+      "loss": 0.2191,
+      "step": 39446
+    },
+    {
+      "epoch": 0.9389460154241646,
+      "grad_norm": 13.03992037822241,
+      "learning_rate": 2e-05,
+      "loss": 0.4752,
+      "step": 39447
+    },
+    {
+      "epoch": 0.938969818147196,
+      "grad_norm": 15.694223302643035,
+      "learning_rate": 2e-05,
+      "loss": 0.3245,
+      "step": 39448
+    },
+    {
+      "epoch": 0.9389936208702275,
+      "grad_norm": 4.6192379446271925,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 39449
+    },
+    {
+      "epoch": 0.939017423593259,
+      "grad_norm": 3.030492425168841,
+      "learning_rate": 2e-05,
+      "loss": 0.6526,
+      "step": 39450
+    },
+    {
+      "epoch": 0.9390412263162906,
+      "grad_norm": 0.8896180470754742,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 39451
+    },
+    {
+      "epoch": 0.9390650290393221,
+      "grad_norm": 0.5368941961429765,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 39452
+    },
+    {
+      "epoch": 0.9390888317623536,
+      "grad_norm": 0.697852380253184,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 39453
+    },
+    {
+      "epoch": 0.9391126344853852,
+      "grad_norm": 2.598527628471661,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 39454
+    },
+    {
+      "epoch": 0.9391364372084167,
+      "grad_norm": 9.493730664474333,
+      "learning_rate": 2e-05,
+      "loss": 0.7948,
+      "step": 39455
+    },
+    {
+      "epoch": 0.9391602399314481,
+      "grad_norm": 18.824524555114813,
+      "learning_rate": 2e-05,
+      "loss": 0.4009,
+      "step": 39456
+    },
+    {
+      "epoch": 0.9391840426544796,
+      "grad_norm": 3.3722095158254044,
+      "learning_rate": 2e-05,
+      "loss": 0.0977,
+      "step": 39457
+    },
+    {
+      "epoch": 0.9392078453775112,
+      "grad_norm": 3.155965831524056,
+      "learning_rate": 2e-05,
+      "loss": 0.6139,
+      "step": 39458
+    },
+    {
+      "epoch": 0.9392316481005427,
+      "grad_norm": 3.081892436230323,
+      "learning_rate": 2e-05,
+      "loss": 0.1034,
+      "step": 39459
+    },
+    {
+      "epoch": 0.9392554508235742,
+      "grad_norm": 4.152217758599186,
+      "learning_rate": 2e-05,
+      "loss": 0.5077,
+      "step": 39460
+    },
+    {
+      "epoch": 0.9392792535466057,
+      "grad_norm": 4.322385044821832,
+      "learning_rate": 2e-05,
+      "loss": 0.7966,
+      "step": 39461
+    },
+    {
+      "epoch": 0.9393030562696373,
+      "grad_norm": 2.7728930969831955,
+      "learning_rate": 2e-05,
+      "loss": 0.3155,
+      "step": 39462
+    },
+    {
+      "epoch": 0.9393268589926688,
+      "grad_norm": 3.2655737315370836,
+      "learning_rate": 2e-05,
+      "loss": 0.5126,
+      "step": 39463
+    },
+    {
+      "epoch": 0.9393506617157003,
+      "grad_norm": 0.37648188628475954,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 39464
+    },
+    {
+      "epoch": 0.9393744644387317,
+      "grad_norm": 5.256961279309983,
+      "learning_rate": 2e-05,
+      "loss": 0.7593,
+      "step": 39465
+    },
+    {
+      "epoch": 0.9393982671617633,
+      "grad_norm": 0.12551523948673612,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 39466
+    },
+    {
+      "epoch": 0.9394220698847948,
+      "grad_norm": 2.884903396529143,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 39467
+    },
+    {
+      "epoch": 0.9394458726078263,
+      "grad_norm": 9.410577916411171,
+      "learning_rate": 2e-05,
+      "loss": 1.4632,
+      "step": 39468
+    },
+    {
+      "epoch": 0.9394696753308579,
+      "grad_norm": 3.9266734396567915,
+      "learning_rate": 2e-05,
+      "loss": 0.6414,
+      "step": 39469
+    },
+    {
+      "epoch": 0.9394934780538894,
+      "grad_norm": 0.03664027299108394,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39470
+    },
+    {
+      "epoch": 0.9395172807769209,
+      "grad_norm": 5.798324643253519,
+      "learning_rate": 2e-05,
+      "loss": 1.0727,
+      "step": 39471
+    },
+    {
+      "epoch": 0.9395410834999524,
+      "grad_norm": 2.754017135786344,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 39472
+    },
+    {
+      "epoch": 0.939564886222984,
+      "grad_norm": 11.539099375078896,
+      "learning_rate": 2e-05,
+      "loss": 1.0525,
+      "step": 39473
+    },
+    {
+      "epoch": 0.9395886889460154,
+      "grad_norm": 0.8205928473977468,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 39474
+    },
+    {
+      "epoch": 0.9396124916690469,
+      "grad_norm": 0.12481868406375933,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39475
+    },
+    {
+      "epoch": 0.9396362943920784,
+      "grad_norm": 4.3817922589502665,
+      "learning_rate": 2e-05,
+      "loss": 0.7538,
+      "step": 39476
+    },
+    {
+      "epoch": 0.93966009711511,
+      "grad_norm": 0.08508914892437748,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39477
+    },
+    {
+      "epoch": 0.9396838998381415,
+      "grad_norm": 0.06782894452710778,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39478
+    },
+    {
+      "epoch": 0.939707702561173,
+      "grad_norm": 5.958486802880567,
+      "learning_rate": 2e-05,
+      "loss": 0.6583,
+      "step": 39479
+    },
+    {
+      "epoch": 0.9397315052842045,
+      "grad_norm": 1.9510527305336904,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 39480
+    },
+    {
+      "epoch": 0.9397553080072361,
+      "grad_norm": 1.9529785625425322,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 39481
+    },
+    {
+      "epoch": 0.9397791107302675,
+      "grad_norm": 0.03172679516623565,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 39482
+    },
+    {
+      "epoch": 0.939802913453299,
+      "grad_norm": 3.8345571567563184,
+      "learning_rate": 2e-05,
+      "loss": 0.6686,
+      "step": 39483
+    },
+    {
+      "epoch": 0.9398267161763306,
+      "grad_norm": 3.414811607058311,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 39484
+    },
+    {
+      "epoch": 0.9398505188993621,
+      "grad_norm": 4.188997188903048,
+      "learning_rate": 2e-05,
+      "loss": 0.0672,
+      "step": 39485
+    },
+    {
+      "epoch": 0.9398743216223936,
+      "grad_norm": 0.147992897167393,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 39486
+    },
+    {
+      "epoch": 0.9398981243454251,
+      "grad_norm": 2.6157720549321426,
+      "learning_rate": 2e-05,
+      "loss": 0.0728,
+      "step": 39487
+    },
+    {
+      "epoch": 0.9399219270684567,
+      "grad_norm": 2.2122594981370165,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 39488
+    },
+    {
+      "epoch": 0.9399457297914882,
+      "grad_norm": 3.6531224181396826,
+      "learning_rate": 2e-05,
+      "loss": 0.4583,
+      "step": 39489
+    },
+    {
+      "epoch": 0.9399695325145196,
+      "grad_norm": 4.9910167338820175,
+      "learning_rate": 2e-05,
+      "loss": 0.9008,
+      "step": 39490
+    },
+    {
+      "epoch": 0.9399933352375511,
+      "grad_norm": 2.6428644831159738,
+      "learning_rate": 2e-05,
+      "loss": 0.247,
+      "step": 39491
+    },
+    {
+      "epoch": 0.9400171379605827,
+      "grad_norm": 4.3844314850881405,
+      "learning_rate": 2e-05,
+      "loss": 0.6169,
+      "step": 39492
+    },
+    {
+      "epoch": 0.9400409406836142,
+      "grad_norm": 4.1029096789048936,
+      "learning_rate": 2e-05,
+      "loss": 0.5769,
+      "step": 39493
+    },
+    {
+      "epoch": 0.9400647434066457,
+      "grad_norm": 2.4744713968935397,
+      "learning_rate": 2e-05,
+      "loss": 0.2735,
+      "step": 39494
+    },
+    {
+      "epoch": 0.9400885461296772,
+      "grad_norm": 2.477118427603146,
+      "learning_rate": 2e-05,
+      "loss": 0.3414,
+      "step": 39495
+    },
+    {
+      "epoch": 0.9401123488527088,
+      "grad_norm": 3.0500791410062176,
+      "learning_rate": 2e-05,
+      "loss": 0.1126,
+      "step": 39496
+    },
+    {
+      "epoch": 0.9401361515757403,
+      "grad_norm": 11.478930377448231,
+      "learning_rate": 2e-05,
+      "loss": 0.7042,
+      "step": 39497
+    },
+    {
+      "epoch": 0.9401599542987717,
+      "grad_norm": 0.009463739699119095,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 39498
+    },
+    {
+      "epoch": 0.9401837570218033,
+      "grad_norm": 2.8564974605694404,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 39499
+    },
+    {
+      "epoch": 0.9402075597448348,
+      "grad_norm": 3.1111121191657003,
+      "learning_rate": 2e-05,
+      "loss": 0.4197,
+      "step": 39500
+    },
+    {
+      "epoch": 0.9402313624678663,
+      "grad_norm": 0.5824883654666623,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 39501
+    },
+    {
+      "epoch": 0.9402551651908978,
+      "grad_norm": 0.08058808213997046,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 39502
+    },
+    {
+      "epoch": 0.9402789679139294,
+      "grad_norm": 9.66072773176989,
+      "learning_rate": 2e-05,
+      "loss": 0.4466,
+      "step": 39503
+    },
+    {
+      "epoch": 0.9403027706369609,
+      "grad_norm": 1.4379953825529488,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 39504
+    },
+    {
+      "epoch": 0.9403265733599924,
+      "grad_norm": 1.6150790246643896,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 39505
+    },
+    {
+      "epoch": 0.9403503760830239,
+      "grad_norm": 19.430487431975862,
+      "learning_rate": 2e-05,
+      "loss": 1.3137,
+      "step": 39506
+    },
+    {
+      "epoch": 0.9403741788060554,
+      "grad_norm": 0.010486309148367696,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 39507
+    },
+    {
+      "epoch": 0.9403979815290869,
+      "grad_norm": 1.3675834790699402,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 39508
+    },
+    {
+      "epoch": 0.9404217842521184,
+      "grad_norm": 1.834032073372552,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 39509
+    },
+    {
+      "epoch": 0.9404455869751499,
+      "grad_norm": 3.03043450161556,
+      "learning_rate": 2e-05,
+      "loss": 0.3384,
+      "step": 39510
+    },
+    {
+      "epoch": 0.9404693896981815,
+      "grad_norm": 4.613475190639297,
+      "learning_rate": 2e-05,
+      "loss": 0.1006,
+      "step": 39511
+    },
+    {
+      "epoch": 0.940493192421213,
+      "grad_norm": 2.8440541001973956,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 39512
+    },
+    {
+      "epoch": 0.9405169951442445,
+      "grad_norm": 0.054261347039084376,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39513
+    },
+    {
+      "epoch": 0.9405407978672761,
+      "grad_norm": 10.095684675995294,
+      "learning_rate": 2e-05,
+      "loss": 1.1007,
+      "step": 39514
+    },
+    {
+      "epoch": 0.9405646005903076,
+      "grad_norm": 2.65524288453656,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 39515
+    },
+    {
+      "epoch": 0.940588403313339,
+      "grad_norm": 15.305017736225235,
+      "learning_rate": 2e-05,
+      "loss": 0.4167,
+      "step": 39516
+    },
+    {
+      "epoch": 0.9406122060363705,
+      "grad_norm": 1.0930186235869306,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 39517
+    },
+    {
+      "epoch": 0.9406360087594021,
+      "grad_norm": 2.135152295091131,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 39518
+    },
+    {
+      "epoch": 0.9406598114824336,
+      "grad_norm": 5.767803814058525,
+      "learning_rate": 2e-05,
+      "loss": 0.0982,
+      "step": 39519
+    },
+    {
+      "epoch": 0.9406836142054651,
+      "grad_norm": 6.622815620896263,
+      "learning_rate": 2e-05,
+      "loss": 0.7968,
+      "step": 39520
+    },
+    {
+      "epoch": 0.9407074169284966,
+      "grad_norm": 6.301186394070231,
+      "learning_rate": 2e-05,
+      "loss": 0.394,
+      "step": 39521
+    },
+    {
+      "epoch": 0.9407312196515282,
+      "grad_norm": 9.01626980103188,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 39522
+    },
+    {
+      "epoch": 0.9407550223745597,
+      "grad_norm": 5.2959325515619895,
+      "learning_rate": 2e-05,
+      "loss": 0.0982,
+      "step": 39523
+    },
+    {
+      "epoch": 0.9407788250975911,
+      "grad_norm": 3.90002225486909,
+      "learning_rate": 2e-05,
+      "loss": 0.1276,
+      "step": 39524
+    },
+    {
+      "epoch": 0.9408026278206226,
+      "grad_norm": 2.731994195667454,
+      "learning_rate": 2e-05,
+      "loss": 0.2179,
+      "step": 39525
+    },
+    {
+      "epoch": 0.9408264305436542,
+      "grad_norm": 2.8265575778443504,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 39526
+    },
+    {
+      "epoch": 0.9408502332666857,
+      "grad_norm": 4.767547436181711,
+      "learning_rate": 2e-05,
+      "loss": 1.0764,
+      "step": 39527
+    },
+    {
+      "epoch": 0.9408740359897172,
+      "grad_norm": 10.099938533765364,
+      "learning_rate": 2e-05,
+      "loss": 0.2632,
+      "step": 39528
+    },
+    {
+      "epoch": 0.9408978387127488,
+      "grad_norm": 1.9214730010485466,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 39529
+    },
+    {
+      "epoch": 0.9409216414357803,
+      "grad_norm": 16.101378556536773,
+      "learning_rate": 2e-05,
+      "loss": 0.6203,
+      "step": 39530
+    },
+    {
+      "epoch": 0.9409454441588118,
+      "grad_norm": 4.066939914918847,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 39531
+    },
+    {
+      "epoch": 0.9409692468818432,
+      "grad_norm": 37.33478216979286,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 39532
+    },
+    {
+      "epoch": 0.9409930496048748,
+      "grad_norm": 4.301931135972773,
+      "learning_rate": 2e-05,
+      "loss": 0.5206,
+      "step": 39533
+    },
+    {
+      "epoch": 0.9410168523279063,
+      "grad_norm": 0.00782703184355487,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 39534
+    },
+    {
+      "epoch": 0.9410406550509378,
+      "grad_norm": 5.138600979418508,
+      "learning_rate": 2e-05,
+      "loss": 0.2352,
+      "step": 39535
+    },
+    {
+      "epoch": 0.9410644577739693,
+      "grad_norm": 3.2886697560531353,
+      "learning_rate": 2e-05,
+      "loss": 0.3301,
+      "step": 39536
+    },
+    {
+      "epoch": 0.9410882604970009,
+      "grad_norm": 0.01667367676299706,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 39537
+    },
+    {
+      "epoch": 0.9411120632200324,
+      "grad_norm": 0.5800144038396325,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 39538
+    },
+    {
+      "epoch": 0.9411358659430639,
+      "grad_norm": 3.073316604336401,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 39539
+    },
+    {
+      "epoch": 0.9411596686660954,
+      "grad_norm": 0.2826068409965282,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 39540
+    },
+    {
+      "epoch": 0.941183471389127,
+      "grad_norm": 2.871270304266316,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 39541
+    },
+    {
+      "epoch": 0.9412072741121584,
+      "grad_norm": 2.263869432459603,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 39542
+    },
+    {
+      "epoch": 0.9412310768351899,
+      "grad_norm": 0.2629092938896052,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 39543
+    },
+    {
+      "epoch": 0.9412548795582215,
+      "grad_norm": 6.622273032723381,
+      "learning_rate": 2e-05,
+      "loss": 0.1405,
+      "step": 39544
+    },
+    {
+      "epoch": 0.941278682281253,
+      "grad_norm": 7.210984533153938,
+      "learning_rate": 2e-05,
+      "loss": 0.2228,
+      "step": 39545
+    },
+    {
+      "epoch": 0.9413024850042845,
+      "grad_norm": 13.579755956611391,
+      "learning_rate": 2e-05,
+      "loss": 0.659,
+      "step": 39546
+    },
+    {
+      "epoch": 0.941326287727316,
+      "grad_norm": 3.7342164286789075,
+      "learning_rate": 2e-05,
+      "loss": 0.7347,
+      "step": 39547
+    },
+    {
+      "epoch": 0.9413500904503476,
+      "grad_norm": 1.267263176442831,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 39548
+    },
+    {
+      "epoch": 0.941373893173379,
+      "grad_norm": 3.1167102702419553,
+      "learning_rate": 2e-05,
+      "loss": 0.305,
+      "step": 39549
+    },
+    {
+      "epoch": 0.9413976958964105,
+      "grad_norm": 0.03054579371070324,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39550
+    },
+    {
+      "epoch": 0.941421498619442,
+      "grad_norm": 8.31878587012249,
+      "learning_rate": 2e-05,
+      "loss": 1.276,
+      "step": 39551
+    },
+    {
+      "epoch": 0.9414453013424736,
+      "grad_norm": 5.787853786639522,
+      "learning_rate": 2e-05,
+      "loss": 0.2589,
+      "step": 39552
+    },
+    {
+      "epoch": 0.9414691040655051,
+      "grad_norm": 0.4593641549921255,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 39553
+    },
+    {
+      "epoch": 0.9414929067885366,
+      "grad_norm": 3.821372698176485,
+      "learning_rate": 2e-05,
+      "loss": 0.5803,
+      "step": 39554
+    },
+    {
+      "epoch": 0.9415167095115681,
+      "grad_norm": 6.206193880784766,
+      "learning_rate": 2e-05,
+      "loss": 0.1338,
+      "step": 39555
+    },
+    {
+      "epoch": 0.9415405122345997,
+      "grad_norm": 2.1504237647533504,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 39556
+    },
+    {
+      "epoch": 0.9415643149576312,
+      "grad_norm": 0.26514518917903035,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 39557
+    },
+    {
+      "epoch": 0.9415881176806626,
+      "grad_norm": 3.958122189259253,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 39558
+    },
+    {
+      "epoch": 0.9416119204036942,
+      "grad_norm": 1.706605813305811,
+      "learning_rate": 2e-05,
+      "loss": 0.0934,
+      "step": 39559
+    },
+    {
+      "epoch": 0.9416357231267257,
+      "grad_norm": 6.308186472843848,
+      "learning_rate": 2e-05,
+      "loss": 0.7016,
+      "step": 39560
+    },
+    {
+      "epoch": 0.9416595258497572,
+      "grad_norm": 3.1484613979647453,
+      "learning_rate": 2e-05,
+      "loss": 0.0752,
+      "step": 39561
+    },
+    {
+      "epoch": 0.9416833285727887,
+      "grad_norm": 2.8192411961744694,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 39562
+    },
+    {
+      "epoch": 0.9417071312958203,
+      "grad_norm": 1.8039189254525556,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 39563
+    },
+    {
+      "epoch": 0.9417309340188518,
+      "grad_norm": 4.068939840499772,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 39564
+    },
+    {
+      "epoch": 0.9417547367418833,
+      "grad_norm": 2.733769398675847,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 39565
+    },
+    {
+      "epoch": 0.9417785394649147,
+      "grad_norm": 2.4324722764119326,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 39566
+    },
+    {
+      "epoch": 0.9418023421879463,
+      "grad_norm": 0.06993131097101643,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 39567
+    },
+    {
+      "epoch": 0.9418261449109778,
+      "grad_norm": 0.13114392664511304,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 39568
+    },
+    {
+      "epoch": 0.9418499476340093,
+      "grad_norm": 7.106604655806961,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 39569
+    },
+    {
+      "epoch": 0.9418737503570408,
+      "grad_norm": 1.8312783247502138,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 39570
+    },
+    {
+      "epoch": 0.9418975530800724,
+      "grad_norm": 4.4816093867497075,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 39571
+    },
+    {
+      "epoch": 0.9419213558031039,
+      "grad_norm": 4.138569722847951,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 39572
+    },
+    {
+      "epoch": 0.9419451585261354,
+      "grad_norm": 3.696732881696724,
+      "learning_rate": 2e-05,
+      "loss": 0.1586,
+      "step": 39573
+    },
+    {
+      "epoch": 0.941968961249167,
+      "grad_norm": 6.4955363902489935,
+      "learning_rate": 2e-05,
+      "loss": 1.5837,
+      "step": 39574
+    },
+    {
+      "epoch": 0.9419927639721984,
+      "grad_norm": 17.44618683018893,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 39575
+    },
+    {
+      "epoch": 0.9420165666952299,
+      "grad_norm": 9.429336474601767,
+      "learning_rate": 2e-05,
+      "loss": 0.248,
+      "step": 39576
+    },
+    {
+      "epoch": 0.9420403694182614,
+      "grad_norm": 0.15908094603996978,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39577
+    },
+    {
+      "epoch": 0.942064172141293,
+      "grad_norm": 2.7191965708134767,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 39578
+    },
+    {
+      "epoch": 0.9420879748643245,
+      "grad_norm": 2.626907745854732,
+      "learning_rate": 2e-05,
+      "loss": 0.2381,
+      "step": 39579
+    },
+    {
+      "epoch": 0.942111777587356,
+      "grad_norm": 2.6277894775892667,
+      "learning_rate": 2e-05,
+      "loss": 0.0547,
+      "step": 39580
+    },
+    {
+      "epoch": 0.9421355803103875,
+      "grad_norm": 3.2577313091476268,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 39581
+    },
+    {
+      "epoch": 0.9421593830334191,
+      "grad_norm": 0.22754759026268995,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 39582
+    },
+    {
+      "epoch": 0.9421831857564505,
+      "grad_norm": 9.159994934919471,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 39583
+    },
+    {
+      "epoch": 0.942206988479482,
+      "grad_norm": 0.05117666211407226,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 39584
+    },
+    {
+      "epoch": 0.9422307912025135,
+      "grad_norm": 5.264366576952554,
+      "learning_rate": 2e-05,
+      "loss": 0.5635,
+      "step": 39585
+    },
+    {
+      "epoch": 0.9422545939255451,
+      "grad_norm": 0.8600242742331905,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 39586
+    },
+    {
+      "epoch": 0.9422783966485766,
+      "grad_norm": 5.369595462125695,
+      "learning_rate": 2e-05,
+      "loss": 0.3042,
+      "step": 39587
+    },
+    {
+      "epoch": 0.9423021993716081,
+      "grad_norm": 0.339833632096658,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 39588
+    },
+    {
+      "epoch": 0.9423260020946397,
+      "grad_norm": 10.455230593461149,
+      "learning_rate": 2e-05,
+      "loss": 1.3988,
+      "step": 39589
+    },
+    {
+      "epoch": 0.9423498048176712,
+      "grad_norm": 0.44750252820497066,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 39590
+    },
+    {
+      "epoch": 0.9423736075407027,
+      "grad_norm": 0.040174309016952084,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39591
+    },
+    {
+      "epoch": 0.9423974102637341,
+      "grad_norm": 3.6218443952081407,
+      "learning_rate": 2e-05,
+      "loss": 0.6425,
+      "step": 39592
+    },
+    {
+      "epoch": 0.9424212129867657,
+      "grad_norm": 3.9603236582188486,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 39593
+    },
+    {
+      "epoch": 0.9424450157097972,
+      "grad_norm": 2.7522441231022023,
+      "learning_rate": 2e-05,
+      "loss": 0.6216,
+      "step": 39594
+    },
+    {
+      "epoch": 0.9424688184328287,
+      "grad_norm": 3.9385701359322662,
+      "learning_rate": 2e-05,
+      "loss": 0.6151,
+      "step": 39595
+    },
+    {
+      "epoch": 0.9424926211558602,
+      "grad_norm": 2.8759438642631228,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 39596
+    },
+    {
+      "epoch": 0.9425164238788918,
+      "grad_norm": 5.308954214350011,
+      "learning_rate": 2e-05,
+      "loss": 0.7098,
+      "step": 39597
+    },
+    {
+      "epoch": 0.9425402266019233,
+      "grad_norm": 3.107383054845609,
+      "learning_rate": 2e-05,
+      "loss": 0.4346,
+      "step": 39598
+    },
+    {
+      "epoch": 0.9425640293249548,
+      "grad_norm": 4.775985588857655,
+      "learning_rate": 2e-05,
+      "loss": 0.2076,
+      "step": 39599
+    },
+    {
+      "epoch": 0.9425878320479862,
+      "grad_norm": 9.81390359045706,
+      "learning_rate": 2e-05,
+      "loss": 0.4543,
+      "step": 39600
+    },
+    {
+      "epoch": 0.9426116347710178,
+      "grad_norm": 7.9362313213696645,
+      "learning_rate": 2e-05,
+      "loss": 0.3853,
+      "step": 39601
+    },
+    {
+      "epoch": 0.9426354374940493,
+      "grad_norm": 7.028045056838314,
+      "learning_rate": 2e-05,
+      "loss": 0.2324,
+      "step": 39602
+    },
+    {
+      "epoch": 0.9426592402170808,
+      "grad_norm": 2.928267663103489,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 39603
+    },
+    {
+      "epoch": 0.9426830429401124,
+      "grad_norm": 4.701179799241858,
+      "learning_rate": 2e-05,
+      "loss": 0.2199,
+      "step": 39604
+    },
+    {
+      "epoch": 0.9427068456631439,
+      "grad_norm": 4.001129096346734,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 39605
+    },
+    {
+      "epoch": 0.9427306483861754,
+      "grad_norm": 7.211046935585836,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 39606
+    },
+    {
+      "epoch": 0.9427544511092069,
+      "grad_norm": 11.6424491149052,
+      "learning_rate": 2e-05,
+      "loss": 0.1495,
+      "step": 39607
+    },
+    {
+      "epoch": 0.9427782538322385,
+      "grad_norm": 4.957154588763003,
+      "learning_rate": 2e-05,
+      "loss": 0.6014,
+      "step": 39608
+    },
+    {
+      "epoch": 0.9428020565552699,
+      "grad_norm": 3.995489291373026,
+      "learning_rate": 2e-05,
+      "loss": 0.8134,
+      "step": 39609
+    },
+    {
+      "epoch": 0.9428258592783014,
+      "grad_norm": 0.1271088797890296,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 39610
+    },
+    {
+      "epoch": 0.9428496620013329,
+      "grad_norm": 4.785082189787335,
+      "learning_rate": 2e-05,
+      "loss": 0.9279,
+      "step": 39611
+    },
+    {
+      "epoch": 0.9428734647243645,
+      "grad_norm": 6.265787230760628,
+      "learning_rate": 2e-05,
+      "loss": 0.4127,
+      "step": 39612
+    },
+    {
+      "epoch": 0.942897267447396,
+      "grad_norm": 6.502509072578445,
+      "learning_rate": 2e-05,
+      "loss": 1.5524,
+      "step": 39613
+    },
+    {
+      "epoch": 0.9429210701704275,
+      "grad_norm": 0.7283623687893331,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 39614
+    },
+    {
+      "epoch": 0.942944872893459,
+      "grad_norm": 7.089423728481674,
+      "learning_rate": 2e-05,
+      "loss": 0.9866,
+      "step": 39615
+    },
+    {
+      "epoch": 0.9429686756164906,
+      "grad_norm": 2.564379050167567,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 39616
+    },
+    {
+      "epoch": 0.942992478339522,
+      "grad_norm": 9.576518147164249,
+      "learning_rate": 2e-05,
+      "loss": 0.9318,
+      "step": 39617
+    },
+    {
+      "epoch": 0.9430162810625535,
+      "grad_norm": 3.5785747943558666,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 39618
+    },
+    {
+      "epoch": 0.9430400837855851,
+      "grad_norm": 3.830540414701102,
+      "learning_rate": 2e-05,
+      "loss": 0.7033,
+      "step": 39619
+    },
+    {
+      "epoch": 0.9430638865086166,
+      "grad_norm": 0.48023083384889087,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 39620
+    },
+    {
+      "epoch": 0.9430876892316481,
+      "grad_norm": 4.193808648136894,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 39621
+    },
+    {
+      "epoch": 0.9431114919546796,
+      "grad_norm": 6.0758856960159155,
+      "learning_rate": 2e-05,
+      "loss": 1.1724,
+      "step": 39622
+    },
+    {
+      "epoch": 0.9431352946777112,
+      "grad_norm": 0.7861456905784818,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 39623
+    },
+    {
+      "epoch": 0.9431590974007427,
+      "grad_norm": 0.14147155634091158,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 39624
+    },
+    {
+      "epoch": 0.9431829001237741,
+      "grad_norm": 3.7201025187434413,
+      "learning_rate": 2e-05,
+      "loss": 0.2561,
+      "step": 39625
+    },
+    {
+      "epoch": 0.9432067028468056,
+      "grad_norm": 0.24187953681359356,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 39626
+    },
+    {
+      "epoch": 0.9432305055698372,
+      "grad_norm": 3.845625826131274,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 39627
+    },
+    {
+      "epoch": 0.9432543082928687,
+      "grad_norm": 11.17450331026427,
+      "learning_rate": 2e-05,
+      "loss": 1.0312,
+      "step": 39628
+    },
+    {
+      "epoch": 0.9432781110159002,
+      "grad_norm": 0.3545666382627597,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 39629
+    },
+    {
+      "epoch": 0.9433019137389317,
+      "grad_norm": 0.6620153905086527,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 39630
+    },
+    {
+      "epoch": 0.9433257164619633,
+      "grad_norm": 4.022888992662634,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 39631
+    },
+    {
+      "epoch": 0.9433495191849948,
+      "grad_norm": 6.2830023571393605,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 39632
+    },
+    {
+      "epoch": 0.9433733219080263,
+      "grad_norm": 4.197980870534735,
+      "learning_rate": 2e-05,
+      "loss": 0.7694,
+      "step": 39633
+    },
+    {
+      "epoch": 0.9433971246310578,
+      "grad_norm": 4.26022185018367,
+      "learning_rate": 2e-05,
+      "loss": 0.7114,
+      "step": 39634
+    },
+    {
+      "epoch": 0.9434209273540893,
+      "grad_norm": 0.41363333779296485,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 39635
+    },
+    {
+      "epoch": 0.9434447300771208,
+      "grad_norm": 4.329892300968025,
+      "learning_rate": 2e-05,
+      "loss": 0.597,
+      "step": 39636
+    },
+    {
+      "epoch": 0.9434685328001523,
+      "grad_norm": 2.975700696299261,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 39637
+    },
+    {
+      "epoch": 0.9434923355231839,
+      "grad_norm": 5.868175317677766,
+      "learning_rate": 2e-05,
+      "loss": 1.6306,
+      "step": 39638
+    },
+    {
+      "epoch": 0.9435161382462154,
+      "grad_norm": 0.554271761459785,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 39639
+    },
+    {
+      "epoch": 0.9435399409692469,
+      "grad_norm": 3.900103663292704,
+      "learning_rate": 2e-05,
+      "loss": 0.4987,
+      "step": 39640
+    },
+    {
+      "epoch": 0.9435637436922784,
+      "grad_norm": 0.7444148486350747,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 39641
+    },
+    {
+      "epoch": 0.94358754641531,
+      "grad_norm": 0.2339818554373981,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 39642
+    },
+    {
+      "epoch": 0.9436113491383414,
+      "grad_norm": 4.174632629587523,
+      "learning_rate": 2e-05,
+      "loss": 0.5707,
+      "step": 39643
+    },
+    {
+      "epoch": 0.9436351518613729,
+      "grad_norm": 1.1333423135657417,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 39644
+    },
+    {
+      "epoch": 0.9436589545844044,
+      "grad_norm": 0.1355985599735802,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 39645
+    },
+    {
+      "epoch": 0.943682757307436,
+      "grad_norm": 4.5346603437796,
+      "learning_rate": 2e-05,
+      "loss": 1.155,
+      "step": 39646
+    },
+    {
+      "epoch": 0.9437065600304675,
+      "grad_norm": 0.047861048924002794,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 39647
+    },
+    {
+      "epoch": 0.943730362753499,
+      "grad_norm": 12.336337433651742,
+      "learning_rate": 2e-05,
+      "loss": 0.8317,
+      "step": 39648
+    },
+    {
+      "epoch": 0.9437541654765305,
+      "grad_norm": 4.491469108999492,
+      "learning_rate": 2e-05,
+      "loss": 0.0982,
+      "step": 39649
+    },
+    {
+      "epoch": 0.9437779681995621,
+      "grad_norm": 0.011599636011282742,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 39650
+    },
+    {
+      "epoch": 0.9438017709225935,
+      "grad_norm": 0.06351354330964237,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 39651
+    },
+    {
+      "epoch": 0.943825573645625,
+      "grad_norm": 4.235870445024058,
+      "learning_rate": 2e-05,
+      "loss": 0.7065,
+      "step": 39652
+    },
+    {
+      "epoch": 0.9438493763686566,
+      "grad_norm": 2.9394878501622315,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 39653
+    },
+    {
+      "epoch": 0.9438731790916881,
+      "grad_norm": 9.376969181424077,
+      "learning_rate": 2e-05,
+      "loss": 0.3576,
+      "step": 39654
+    },
+    {
+      "epoch": 0.9438969818147196,
+      "grad_norm": 4.969779019647457,
+      "learning_rate": 2e-05,
+      "loss": 0.7342,
+      "step": 39655
+    },
+    {
+      "epoch": 0.9439207845377511,
+      "grad_norm": 1.9191172608601592,
+      "learning_rate": 2e-05,
+      "loss": 0.0964,
+      "step": 39656
+    },
+    {
+      "epoch": 0.9439445872607827,
+      "grad_norm": 0.23479909115157052,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39657
+    },
+    {
+      "epoch": 0.9439683899838142,
+      "grad_norm": 3.63412006312261,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 39658
+    },
+    {
+      "epoch": 0.9439921927068456,
+      "grad_norm": 0.11626711889873541,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 39659
+    },
+    {
+      "epoch": 0.9440159954298771,
+      "grad_norm": 0.4485918391249169,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 39660
+    },
+    {
+      "epoch": 0.9440397981529087,
+      "grad_norm": 6.114244328265265,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 39661
+    },
+    {
+      "epoch": 0.9440636008759402,
+      "grad_norm": 0.26677176279371007,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 39662
+    },
+    {
+      "epoch": 0.9440874035989717,
+      "grad_norm": 29.297085346832812,
+      "learning_rate": 2e-05,
+      "loss": 0.1839,
+      "step": 39663
+    },
+    {
+      "epoch": 0.9441112063220032,
+      "grad_norm": 3.3736281190982718,
+      "learning_rate": 2e-05,
+      "loss": 0.0844,
+      "step": 39664
+    },
+    {
+      "epoch": 0.9441350090450348,
+      "grad_norm": 3.821342833968347,
+      "learning_rate": 2e-05,
+      "loss": 0.5964,
+      "step": 39665
+    },
+    {
+      "epoch": 0.9441588117680663,
+      "grad_norm": 3.784305813064905,
+      "learning_rate": 2e-05,
+      "loss": 0.7197,
+      "step": 39666
+    },
+    {
+      "epoch": 0.9441826144910977,
+      "grad_norm": 0.9770887349695653,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 39667
+    },
+    {
+      "epoch": 0.9442064172141293,
+      "grad_norm": 0.02005556150116041,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39668
+    },
+    {
+      "epoch": 0.9442302199371608,
+      "grad_norm": 9.428475030871141,
+      "learning_rate": 2e-05,
+      "loss": 0.815,
+      "step": 39669
+    },
+    {
+      "epoch": 0.9442540226601923,
+      "grad_norm": 20.823930996501588,
+      "learning_rate": 2e-05,
+      "loss": 1.3189,
+      "step": 39670
+    },
+    {
+      "epoch": 0.9442778253832238,
+      "grad_norm": 0.06297789316248846,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39671
+    },
+    {
+      "epoch": 0.9443016281062554,
+      "grad_norm": 0.17672720449927196,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39672
+    },
+    {
+      "epoch": 0.9443254308292869,
+      "grad_norm": 4.5630602619248455,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 39673
+    },
+    {
+      "epoch": 0.9443492335523184,
+      "grad_norm": 3.5927911727188526,
+      "learning_rate": 2e-05,
+      "loss": 0.4017,
+      "step": 39674
+    },
+    {
+      "epoch": 0.9443730362753499,
+      "grad_norm": 2.692642686280746,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 39675
+    },
+    {
+      "epoch": 0.9443968389983814,
+      "grad_norm": 2.694707236887506,
+      "learning_rate": 2e-05,
+      "loss": 0.4295,
+      "step": 39676
+    },
+    {
+      "epoch": 0.9444206417214129,
+      "grad_norm": 4.893588629606966,
+      "learning_rate": 2e-05,
+      "loss": 1.4303,
+      "step": 39677
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 0.06244390705322335,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 39678
+    },
+    {
+      "epoch": 0.9444682471674759,
+      "grad_norm": 5.708692682677663,
+      "learning_rate": 2e-05,
+      "loss": 1.2586,
+      "step": 39679
+    },
+    {
+      "epoch": 0.9444920498905075,
+      "grad_norm": 0.9920666569161266,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 39680
+    },
+    {
+      "epoch": 0.944515852613539,
+      "grad_norm": 10.584832771733861,
+      "learning_rate": 2e-05,
+      "loss": 1.4167,
+      "step": 39681
+    },
+    {
+      "epoch": 0.9445396553365705,
+      "grad_norm": 3.843560233691552,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 39682
+    },
+    {
+      "epoch": 0.9445634580596021,
+      "grad_norm": 15.257718637374644,
+      "learning_rate": 2e-05,
+      "loss": 0.7464,
+      "step": 39683
+    },
+    {
+      "epoch": 0.9445872607826336,
+      "grad_norm": 0.8201752156848663,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 39684
+    },
+    {
+      "epoch": 0.944611063505665,
+      "grad_norm": 0.20212858121527258,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 39685
+    },
+    {
+      "epoch": 0.9446348662286965,
+      "grad_norm": 5.0684759067063485,
+      "learning_rate": 2e-05,
+      "loss": 0.1718,
+      "step": 39686
+    },
+    {
+      "epoch": 0.9446586689517281,
+      "grad_norm": 3.361579585154736,
+      "learning_rate": 2e-05,
+      "loss": 0.7657,
+      "step": 39687
+    },
+    {
+      "epoch": 0.9446824716747596,
+      "grad_norm": 2.4166925771252066,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 39688
+    },
+    {
+      "epoch": 0.9447062743977911,
+      "grad_norm": 4.468949643149729,
+      "learning_rate": 2e-05,
+      "loss": 0.1612,
+      "step": 39689
+    },
+    {
+      "epoch": 0.9447300771208226,
+      "grad_norm": 18.713904825832184,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 39690
+    },
+    {
+      "epoch": 0.9447538798438542,
+      "grad_norm": 4.453307629269857,
+      "learning_rate": 2e-05,
+      "loss": 0.15,
+      "step": 39691
+    },
+    {
+      "epoch": 0.9447776825668857,
+      "grad_norm": 1.0396249889287201,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 39692
+    },
+    {
+      "epoch": 0.9448014852899171,
+      "grad_norm": 2.1438278691179407,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 39693
+    },
+    {
+      "epoch": 0.9448252880129486,
+      "grad_norm": 4.6641030547484545,
+      "learning_rate": 2e-05,
+      "loss": 0.1648,
+      "step": 39694
+    },
+    {
+      "epoch": 0.9448490907359802,
+      "grad_norm": 5.135823666870807,
+      "learning_rate": 2e-05,
+      "loss": 0.918,
+      "step": 39695
+    },
+    {
+      "epoch": 0.9448728934590117,
+      "grad_norm": 4.4809244922459905,
+      "learning_rate": 2e-05,
+      "loss": 0.1691,
+      "step": 39696
+    },
+    {
+      "epoch": 0.9448966961820432,
+      "grad_norm": 48.61005611088141,
+      "learning_rate": 2e-05,
+      "loss": 2.1615,
+      "step": 39697
+    },
+    {
+      "epoch": 0.9449204989050748,
+      "grad_norm": 0.0780564394704555,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39698
+    },
+    {
+      "epoch": 0.9449443016281063,
+      "grad_norm": 5.773571893244676,
+      "learning_rate": 2e-05,
+      "loss": 0.9473,
+      "step": 39699
+    },
+    {
+      "epoch": 0.9449681043511378,
+      "grad_norm": 201.90487062462336,
+      "learning_rate": 2e-05,
+      "loss": 0.344,
+      "step": 39700
+    },
+    {
+      "epoch": 0.9449919070741692,
+      "grad_norm": 1.0219887918023536,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 39701
+    },
+    {
+      "epoch": 0.9450157097972008,
+      "grad_norm": 0.19332266158165923,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 39702
+    },
+    {
+      "epoch": 0.9450395125202323,
+      "grad_norm": 0.4980896147049657,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 39703
+    },
+    {
+      "epoch": 0.9450633152432638,
+      "grad_norm": 9.927948196193656,
+      "learning_rate": 2e-05,
+      "loss": 0.3767,
+      "step": 39704
+    },
+    {
+      "epoch": 0.9450871179662953,
+      "grad_norm": 7.999017698919671,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 39705
+    },
+    {
+      "epoch": 0.9451109206893269,
+      "grad_norm": 3.28381391841755,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 39706
+    },
+    {
+      "epoch": 0.9451347234123584,
+      "grad_norm": 1.5957001144831255,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 39707
+    },
+    {
+      "epoch": 0.9451585261353899,
+      "grad_norm": 4.021324385026598,
+      "learning_rate": 2e-05,
+      "loss": 0.5055,
+      "step": 39708
+    },
+    {
+      "epoch": 0.9451823288584214,
+      "grad_norm": 4.340314122973042,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 39709
+    },
+    {
+      "epoch": 0.945206131581453,
+      "grad_norm": 11.253802980299485,
+      "learning_rate": 2e-05,
+      "loss": 0.7667,
+      "step": 39710
+    },
+    {
+      "epoch": 0.9452299343044844,
+      "grad_norm": 1.036995557040589,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 39711
+    },
+    {
+      "epoch": 0.9452537370275159,
+      "grad_norm": 1.7411285584726244,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 39712
+    },
+    {
+      "epoch": 0.9452775397505475,
+      "grad_norm": 3.349734982717546,
+      "learning_rate": 2e-05,
+      "loss": 0.7193,
+      "step": 39713
+    },
+    {
+      "epoch": 0.945301342473579,
+      "grad_norm": 0.3200071940151932,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 39714
+    },
+    {
+      "epoch": 0.9453251451966105,
+      "grad_norm": 10.416619939107516,
+      "learning_rate": 2e-05,
+      "loss": 0.5509,
+      "step": 39715
+    },
+    {
+      "epoch": 0.945348947919642,
+      "grad_norm": 4.772386133885726,
+      "learning_rate": 2e-05,
+      "loss": 0.5904,
+      "step": 39716
+    },
+    {
+      "epoch": 0.9453727506426736,
+      "grad_norm": 0.9870313408128,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 39717
+    },
+    {
+      "epoch": 0.945396553365705,
+      "grad_norm": 3.2779304727385177,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 39718
+    },
+    {
+      "epoch": 0.9454203560887365,
+      "grad_norm": 4.204228352038746,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 39719
+    },
+    {
+      "epoch": 0.945444158811768,
+      "grad_norm": 4.726826075292028,
+      "learning_rate": 2e-05,
+      "loss": 0.6115,
+      "step": 39720
+    },
+    {
+      "epoch": 0.9454679615347996,
+      "grad_norm": 2.236367856711863,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 39721
+    },
+    {
+      "epoch": 0.9454917642578311,
+      "grad_norm": 7.3164219200726395,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 39722
+    },
+    {
+      "epoch": 0.9455155669808626,
+      "grad_norm": 2.680392250778066,
+      "learning_rate": 2e-05,
+      "loss": 0.356,
+      "step": 39723
+    },
+    {
+      "epoch": 0.9455393697038941,
+      "grad_norm": 15.437959984685765,
+      "learning_rate": 2e-05,
+      "loss": 0.9609,
+      "step": 39724
+    },
+    {
+      "epoch": 0.9455631724269257,
+      "grad_norm": 1.2996435483490412,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 39725
+    },
+    {
+      "epoch": 0.9455869751499572,
+      "grad_norm": 5.466717469154345,
+      "learning_rate": 2e-05,
+      "loss": 0.8332,
+      "step": 39726
+    },
+    {
+      "epoch": 0.9456107778729886,
+      "grad_norm": 7.395433718237537,
+      "learning_rate": 2e-05,
+      "loss": 1.307,
+      "step": 39727
+    },
+    {
+      "epoch": 0.9456345805960202,
+      "grad_norm": 3.5417568631133567,
+      "learning_rate": 2e-05,
+      "loss": 0.3723,
+      "step": 39728
+    },
+    {
+      "epoch": 0.9456583833190517,
+      "grad_norm": 5.167032481991539,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 39729
+    },
+    {
+      "epoch": 0.9456821860420832,
+      "grad_norm": 4.404092926992682,
+      "learning_rate": 2e-05,
+      "loss": 0.7276,
+      "step": 39730
+    },
+    {
+      "epoch": 0.9457059887651147,
+      "grad_norm": 5.433990305204126,
+      "learning_rate": 2e-05,
+      "loss": 0.4291,
+      "step": 39731
+    },
+    {
+      "epoch": 0.9457297914881463,
+      "grad_norm": 6.063767905363082,
+      "learning_rate": 2e-05,
+      "loss": 0.4229,
+      "step": 39732
+    },
+    {
+      "epoch": 0.9457535942111778,
+      "grad_norm": 0.022659103189440834,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39733
+    },
+    {
+      "epoch": 0.9457773969342093,
+      "grad_norm": 5.878862120642528,
+      "learning_rate": 2e-05,
+      "loss": 0.1287,
+      "step": 39734
+    },
+    {
+      "epoch": 0.9458011996572407,
+      "grad_norm": 3.018891793322782,
+      "learning_rate": 2e-05,
+      "loss": 0.3211,
+      "step": 39735
+    },
+    {
+      "epoch": 0.9458250023802723,
+      "grad_norm": 0.9677351464169619,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 39736
+    },
+    {
+      "epoch": 0.9458488051033038,
+      "grad_norm": 2.701509872512159,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 39737
+    },
+    {
+      "epoch": 0.9458726078263353,
+      "grad_norm": 2.7673697803989095,
+      "learning_rate": 2e-05,
+      "loss": 0.3008,
+      "step": 39738
+    },
+    {
+      "epoch": 0.9458964105493668,
+      "grad_norm": 1.7740909913758298,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 39739
+    },
+    {
+      "epoch": 0.9459202132723984,
+      "grad_norm": 2.0030764900377935,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 39740
+    },
+    {
+      "epoch": 0.9459440159954299,
+      "grad_norm": 2.7875593365556055,
+      "learning_rate": 2e-05,
+      "loss": 0.6184,
+      "step": 39741
+    },
+    {
+      "epoch": 0.9459678187184614,
+      "grad_norm": 2.6310122100707263,
+      "learning_rate": 2e-05,
+      "loss": 0.5489,
+      "step": 39742
+    },
+    {
+      "epoch": 0.945991621441493,
+      "grad_norm": 4.302548281827922,
+      "learning_rate": 2e-05,
+      "loss": 0.1718,
+      "step": 39743
+    },
+    {
+      "epoch": 0.9460154241645244,
+      "grad_norm": 1.2728398466902977,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 39744
+    },
+    {
+      "epoch": 0.9460392268875559,
+      "grad_norm": 0.006019159743586933,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 39745
+    },
+    {
+      "epoch": 0.9460630296105874,
+      "grad_norm": 2.3217303737547765,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 39746
+    },
+    {
+      "epoch": 0.946086832333619,
+      "grad_norm": 1.6052472042275239,
+      "learning_rate": 2e-05,
+      "loss": 0.0869,
+      "step": 39747
+    },
+    {
+      "epoch": 0.9461106350566505,
+      "grad_norm": 7.9071339296903735,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 39748
+    },
+    {
+      "epoch": 0.946134437779682,
+      "grad_norm": 8.033062469391497,
+      "learning_rate": 2e-05,
+      "loss": 0.5196,
+      "step": 39749
+    },
+    {
+      "epoch": 0.9461582405027135,
+      "grad_norm": 6.162583045767644,
+      "learning_rate": 2e-05,
+      "loss": 1.3241,
+      "step": 39750
+    },
+    {
+      "epoch": 0.9461820432257451,
+      "grad_norm": 5.975550070143133,
+      "learning_rate": 2e-05,
+      "loss": 0.1273,
+      "step": 39751
+    },
+    {
+      "epoch": 0.9462058459487765,
+      "grad_norm": 9.788036466639165,
+      "learning_rate": 2e-05,
+      "loss": 1.1543,
+      "step": 39752
+    },
+    {
+      "epoch": 0.946229648671808,
+      "grad_norm": 0.32862553810372713,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 39753
+    },
+    {
+      "epoch": 0.9462534513948395,
+      "grad_norm": 3.338893724824461,
+      "learning_rate": 2e-05,
+      "loss": 0.6862,
+      "step": 39754
+    },
+    {
+      "epoch": 0.9462772541178711,
+      "grad_norm": 3.6936445877843167,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 39755
+    },
+    {
+      "epoch": 0.9463010568409026,
+      "grad_norm": 10.815324365938178,
+      "learning_rate": 2e-05,
+      "loss": 1.0737,
+      "step": 39756
+    },
+    {
+      "epoch": 0.9463248595639341,
+      "grad_norm": 0.13157234130966455,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 39757
+    },
+    {
+      "epoch": 0.9463486622869657,
+      "grad_norm": 4.795716539675519,
+      "learning_rate": 2e-05,
+      "loss": 0.4653,
+      "step": 39758
+    },
+    {
+      "epoch": 0.9463724650099972,
+      "grad_norm": 0.20338437477525356,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 39759
+    },
+    {
+      "epoch": 0.9463962677330287,
+      "grad_norm": 12.668068789389894,
+      "learning_rate": 2e-05,
+      "loss": 0.4863,
+      "step": 39760
+    },
+    {
+      "epoch": 0.9464200704560601,
+      "grad_norm": 0.3626368710136869,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 39761
+    },
+    {
+      "epoch": 0.9464438731790917,
+      "grad_norm": 7.087847422303895,
+      "learning_rate": 2e-05,
+      "loss": 0.2109,
+      "step": 39762
+    },
+    {
+      "epoch": 0.9464676759021232,
+      "grad_norm": 7.063279256137452,
+      "learning_rate": 2e-05,
+      "loss": 1.6356,
+      "step": 39763
+    },
+    {
+      "epoch": 0.9464914786251547,
+      "grad_norm": 4.077256768957138,
+      "learning_rate": 2e-05,
+      "loss": 0.4318,
+      "step": 39764
+    },
+    {
+      "epoch": 0.9465152813481862,
+      "grad_norm": 0.1766557804939645,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 39765
+    },
+    {
+      "epoch": 0.9465390840712178,
+      "grad_norm": 8.557127351678854,
+      "learning_rate": 2e-05,
+      "loss": 0.9862,
+      "step": 39766
+    },
+    {
+      "epoch": 0.9465628867942493,
+      "grad_norm": 4.827137724499692,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 39767
+    },
+    {
+      "epoch": 0.9465866895172808,
+      "grad_norm": 9.958744303608407,
+      "learning_rate": 2e-05,
+      "loss": 0.8173,
+      "step": 39768
+    },
+    {
+      "epoch": 0.9466104922403122,
+      "grad_norm": 9.061584604635286,
+      "learning_rate": 2e-05,
+      "loss": 0.3807,
+      "step": 39769
+    },
+    {
+      "epoch": 0.9466342949633438,
+      "grad_norm": 6.70731489361888,
+      "learning_rate": 2e-05,
+      "loss": 1.4547,
+      "step": 39770
+    },
+    {
+      "epoch": 0.9466580976863753,
+      "grad_norm": 0.8175990896560956,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 39771
+    },
+    {
+      "epoch": 0.9466819004094068,
+      "grad_norm": 3.1091561499336886,
+      "learning_rate": 2e-05,
+      "loss": 0.4072,
+      "step": 39772
+    },
+    {
+      "epoch": 0.9467057031324384,
+      "grad_norm": 0.12114359215646574,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 39773
+    },
+    {
+      "epoch": 0.9467295058554699,
+      "grad_norm": 3.9897966679063357,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 39774
+    },
+    {
+      "epoch": 0.9467533085785014,
+      "grad_norm": 5.2679999845946135,
+      "learning_rate": 2e-05,
+      "loss": 1.1431,
+      "step": 39775
+    },
+    {
+      "epoch": 0.9467771113015329,
+      "grad_norm": 5.341851581701245,
+      "learning_rate": 2e-05,
+      "loss": 0.9842,
+      "step": 39776
+    },
+    {
+      "epoch": 0.9468009140245645,
+      "grad_norm": 5.3633391788191895,
+      "learning_rate": 2e-05,
+      "loss": 0.6894,
+      "step": 39777
+    },
+    {
+      "epoch": 0.9468247167475959,
+      "grad_norm": 3.0541741689331756,
+      "learning_rate": 2e-05,
+      "loss": 0.4253,
+      "step": 39778
+    },
+    {
+      "epoch": 0.9468485194706274,
+      "grad_norm": 0.864040250685293,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 39779
+    },
+    {
+      "epoch": 0.9468723221936589,
+      "grad_norm": 1.8765221527003368,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 39780
+    },
+    {
+      "epoch": 0.9468961249166905,
+      "grad_norm": 5.052743251914782,
+      "learning_rate": 2e-05,
+      "loss": 0.1044,
+      "step": 39781
+    },
+    {
+      "epoch": 0.946919927639722,
+      "grad_norm": 2.6517065856352353,
+      "learning_rate": 2e-05,
+      "loss": 0.2772,
+      "step": 39782
+    },
+    {
+      "epoch": 0.9469437303627535,
+      "grad_norm": 0.06955237315869786,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39783
+    },
+    {
+      "epoch": 0.946967533085785,
+      "grad_norm": 0.431614720759351,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 39784
+    },
+    {
+      "epoch": 0.9469913358088166,
+      "grad_norm": 1.7529380435441904,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 39785
+    },
+    {
+      "epoch": 0.947015138531848,
+      "grad_norm": 0.02075848962454974,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 39786
+    },
+    {
+      "epoch": 0.9470389412548795,
+      "grad_norm": 0.3415180385852597,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 39787
+    },
+    {
+      "epoch": 0.9470627439779111,
+      "grad_norm": 3.997731419243149,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 39788
+    },
+    {
+      "epoch": 0.9470865467009426,
+      "grad_norm": 12.774477794548723,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 39789
+    },
+    {
+      "epoch": 0.9471103494239741,
+      "grad_norm": 5.038164262092501,
+      "learning_rate": 2e-05,
+      "loss": 0.6649,
+      "step": 39790
+    },
+    {
+      "epoch": 0.9471341521470056,
+      "grad_norm": 0.8810313275407107,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 39791
+    },
+    {
+      "epoch": 0.9471579548700372,
+      "grad_norm": 0.5255625877314594,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 39792
+    },
+    {
+      "epoch": 0.9471817575930687,
+      "grad_norm": 7.745688671538633,
+      "learning_rate": 2e-05,
+      "loss": 1.3304,
+      "step": 39793
+    },
+    {
+      "epoch": 0.9472055603161001,
+      "grad_norm": 0.0949204800544284,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 39794
+    },
+    {
+      "epoch": 0.9472293630391316,
+      "grad_norm": 0.9880160030243894,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 39795
+    },
+    {
+      "epoch": 0.9472531657621632,
+      "grad_norm": 0.3201198795256423,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 39796
+    },
+    {
+      "epoch": 0.9472769684851947,
+      "grad_norm": 3.9728844789592817,
+      "learning_rate": 2e-05,
+      "loss": 0.0871,
+      "step": 39797
+    },
+    {
+      "epoch": 0.9473007712082262,
+      "grad_norm": 2.9796021574476836,
+      "learning_rate": 2e-05,
+      "loss": 0.0703,
+      "step": 39798
+    },
+    {
+      "epoch": 0.9473245739312577,
+      "grad_norm": 0.092256560389527,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 39799
+    },
+    {
+      "epoch": 0.9473483766542893,
+      "grad_norm": 0.12885149986718644,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 39800
+    },
+    {
+      "epoch": 0.9473721793773208,
+      "grad_norm": 1.4643736547043897,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 39801
+    },
+    {
+      "epoch": 0.9473959821003523,
+      "grad_norm": 5.335747539498704,
+      "learning_rate": 2e-05,
+      "loss": 0.6848,
+      "step": 39802
+    },
+    {
+      "epoch": 0.9474197848233838,
+      "grad_norm": 4.11114918462079,
+      "learning_rate": 2e-05,
+      "loss": 0.5656,
+      "step": 39803
+    },
+    {
+      "epoch": 0.9474435875464153,
+      "grad_norm": 13.161133572635471,
+      "learning_rate": 2e-05,
+      "loss": 0.2503,
+      "step": 39804
+    },
+    {
+      "epoch": 0.9474673902694468,
+      "grad_norm": 0.41452188769869014,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 39805
+    },
+    {
+      "epoch": 0.9474911929924783,
+      "grad_norm": 2.3240902994330526,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 39806
+    },
+    {
+      "epoch": 0.9475149957155099,
+      "grad_norm": 4.951836859355253,
+      "learning_rate": 2e-05,
+      "loss": 1.0655,
+      "step": 39807
+    },
+    {
+      "epoch": 0.9475387984385414,
+      "grad_norm": 6.095919972153408,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 39808
+    },
+    {
+      "epoch": 0.9475626011615729,
+      "grad_norm": 4.839488475577738,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 39809
+    },
+    {
+      "epoch": 0.9475864038846044,
+      "grad_norm": 4.594588214486882,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 39810
+    },
+    {
+      "epoch": 0.947610206607636,
+      "grad_norm": 0.9588661485286024,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 39811
+    },
+    {
+      "epoch": 0.9476340093306674,
+      "grad_norm": 1.026132459817983,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 39812
+    },
+    {
+      "epoch": 0.9476578120536989,
+      "grad_norm": 5.416981833497189,
+      "learning_rate": 2e-05,
+      "loss": 0.7986,
+      "step": 39813
+    },
+    {
+      "epoch": 0.9476816147767304,
+      "grad_norm": 0.21246857101606345,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 39814
+    },
+    {
+      "epoch": 0.947705417499762,
+      "grad_norm": 4.92496854111644,
+      "learning_rate": 2e-05,
+      "loss": 1.1208,
+      "step": 39815
+    },
+    {
+      "epoch": 0.9477292202227935,
+      "grad_norm": 3.743289133082773,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 39816
+    },
+    {
+      "epoch": 0.947753022945825,
+      "grad_norm": 7.15150298636149,
+      "learning_rate": 2e-05,
+      "loss": 0.1498,
+      "step": 39817
+    },
+    {
+      "epoch": 0.9477768256688566,
+      "grad_norm": 6.994692419970719,
+      "learning_rate": 2e-05,
+      "loss": 0.0927,
+      "step": 39818
+    },
+    {
+      "epoch": 0.9478006283918881,
+      "grad_norm": 0.20324449356101593,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 39819
+    },
+    {
+      "epoch": 0.9478244311149195,
+      "grad_norm": 1.1885679988603985,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 39820
+    },
+    {
+      "epoch": 0.947848233837951,
+      "grad_norm": 20.899140265271818,
+      "learning_rate": 2e-05,
+      "loss": 0.6462,
+      "step": 39821
+    },
+    {
+      "epoch": 0.9478720365609826,
+      "grad_norm": 3.0907260398104626,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 39822
+    },
+    {
+      "epoch": 0.9478958392840141,
+      "grad_norm": 3.30258362669364,
+      "learning_rate": 2e-05,
+      "loss": 0.3587,
+      "step": 39823
+    },
+    {
+      "epoch": 0.9479196420070456,
+      "grad_norm": 3.4820599818581774,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 39824
+    },
+    {
+      "epoch": 0.9479434447300771,
+      "grad_norm": 0.2250821991541007,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 39825
+    },
+    {
+      "epoch": 0.9479672474531087,
+      "grad_norm": 1.3874268774227136,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 39826
+    },
+    {
+      "epoch": 0.9479910501761402,
+      "grad_norm": 4.690479104772691,
+      "learning_rate": 2e-05,
+      "loss": 0.9546,
+      "step": 39827
+    },
+    {
+      "epoch": 0.9480148528991716,
+      "grad_norm": 7.104054277348764,
+      "learning_rate": 2e-05,
+      "loss": 1.3513,
+      "step": 39828
+    },
+    {
+      "epoch": 0.9480386556222031,
+      "grad_norm": 2.2500462957709106,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 39829
+    },
+    {
+      "epoch": 0.9480624583452347,
+      "grad_norm": 5.4669929005577895,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 39830
+    },
+    {
+      "epoch": 0.9480862610682662,
+      "grad_norm": 0.9893140986949024,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 39831
+    },
+    {
+      "epoch": 0.9481100637912977,
+      "grad_norm": 7.694939081462895,
+      "learning_rate": 2e-05,
+      "loss": 0.2238,
+      "step": 39832
+    },
+    {
+      "epoch": 0.9481338665143293,
+      "grad_norm": 4.856662933304713,
+      "learning_rate": 2e-05,
+      "loss": 1.2871,
+      "step": 39833
+    },
+    {
+      "epoch": 0.9481576692373608,
+      "grad_norm": 7.301761639513441,
+      "learning_rate": 2e-05,
+      "loss": 0.1318,
+      "step": 39834
+    },
+    {
+      "epoch": 0.9481814719603923,
+      "grad_norm": 0.12342897132005137,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 39835
+    },
+    {
+      "epoch": 0.9482052746834237,
+      "grad_norm": 3.9195407369001436,
+      "learning_rate": 2e-05,
+      "loss": 0.1111,
+      "step": 39836
+    },
+    {
+      "epoch": 0.9482290774064553,
+      "grad_norm": 4.189455066958452,
+      "learning_rate": 2e-05,
+      "loss": 0.9578,
+      "step": 39837
+    },
+    {
+      "epoch": 0.9482528801294868,
+      "grad_norm": 9.271460417925649,
+      "learning_rate": 2e-05,
+      "loss": 0.5337,
+      "step": 39838
+    },
+    {
+      "epoch": 0.9482766828525183,
+      "grad_norm": 3.4882812374520795,
+      "learning_rate": 2e-05,
+      "loss": 0.3493,
+      "step": 39839
+    },
+    {
+      "epoch": 0.9483004855755498,
+      "grad_norm": 5.210531132762225,
+      "learning_rate": 2e-05,
+      "loss": 0.6024,
+      "step": 39840
+    },
+    {
+      "epoch": 0.9483242882985814,
+      "grad_norm": 3.2613193147762405,
+      "learning_rate": 2e-05,
+      "loss": 0.5113,
+      "step": 39841
+    },
+    {
+      "epoch": 0.9483480910216129,
+      "grad_norm": 20.822756011408355,
+      "learning_rate": 2e-05,
+      "loss": 0.2841,
+      "step": 39842
+    },
+    {
+      "epoch": 0.9483718937446444,
+      "grad_norm": 2.4010464808687777,
+      "learning_rate": 2e-05,
+      "loss": 0.5052,
+      "step": 39843
+    },
+    {
+      "epoch": 0.9483956964676759,
+      "grad_norm": 8.787507035448984,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 39844
+    },
+    {
+      "epoch": 0.9484194991907074,
+      "grad_norm": 0.02581692104294665,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39845
+    },
+    {
+      "epoch": 0.9484433019137389,
+      "grad_norm": 4.823321412003185,
+      "learning_rate": 2e-05,
+      "loss": 0.7259,
+      "step": 39846
+    },
+    {
+      "epoch": 0.9484671046367704,
+      "grad_norm": 5.028703343546609,
+      "learning_rate": 2e-05,
+      "loss": 1.0134,
+      "step": 39847
+    },
+    {
+      "epoch": 0.948490907359802,
+      "grad_norm": 2.7709754841926206,
+      "learning_rate": 2e-05,
+      "loss": 0.2356,
+      "step": 39848
+    },
+    {
+      "epoch": 0.9485147100828335,
+      "grad_norm": 0.9732621140754234,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 39849
+    },
+    {
+      "epoch": 0.948538512805865,
+      "grad_norm": 0.7782732831590209,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 39850
+    },
+    {
+      "epoch": 0.9485623155288965,
+      "grad_norm": 2.916188543883671,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 39851
+    },
+    {
+      "epoch": 0.9485861182519281,
+      "grad_norm": 2.5009908049334,
+      "learning_rate": 2e-05,
+      "loss": 0.0362,
+      "step": 39852
+    },
+    {
+      "epoch": 0.9486099209749596,
+      "grad_norm": 0.10834691531739522,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39853
+    },
+    {
+      "epoch": 0.948633723697991,
+      "grad_norm": 0.04341928645429312,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 39854
+    },
+    {
+      "epoch": 0.9486575264210225,
+      "grad_norm": 0.5440822916820248,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 39855
+    },
+    {
+      "epoch": 0.9486813291440541,
+      "grad_norm": 4.202162310476375,
+      "learning_rate": 2e-05,
+      "loss": 0.6803,
+      "step": 39856
+    },
+    {
+      "epoch": 0.9487051318670856,
+      "grad_norm": 7.37586502932808,
+      "learning_rate": 2e-05,
+      "loss": 0.2742,
+      "step": 39857
+    },
+    {
+      "epoch": 0.9487289345901171,
+      "grad_norm": 4.7880092999368635,
+      "learning_rate": 2e-05,
+      "loss": 0.6403,
+      "step": 39858
+    },
+    {
+      "epoch": 0.9487527373131486,
+      "grad_norm": 0.3383316203825893,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 39859
+    },
+    {
+      "epoch": 0.9487765400361802,
+      "grad_norm": 0.4230573483395477,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 39860
+    },
+    {
+      "epoch": 0.9488003427592117,
+      "grad_norm": 5.032482382996927,
+      "learning_rate": 2e-05,
+      "loss": 0.6689,
+      "step": 39861
+    },
+    {
+      "epoch": 0.9488241454822431,
+      "grad_norm": 3.2622470900090454,
+      "learning_rate": 2e-05,
+      "loss": 0.0776,
+      "step": 39862
+    },
+    {
+      "epoch": 0.9488479482052747,
+      "grad_norm": 5.175568856767048,
+      "learning_rate": 2e-05,
+      "loss": 0.3501,
+      "step": 39863
+    },
+    {
+      "epoch": 0.9488717509283062,
+      "grad_norm": 2.621554973133852,
+      "learning_rate": 2e-05,
+      "loss": 0.5236,
+      "step": 39864
+    },
+    {
+      "epoch": 0.9488955536513377,
+      "grad_norm": 8.620440616083414,
+      "learning_rate": 2e-05,
+      "loss": 0.3975,
+      "step": 39865
+    },
+    {
+      "epoch": 0.9489193563743692,
+      "grad_norm": 8.374078608873273,
+      "learning_rate": 2e-05,
+      "loss": 0.1246,
+      "step": 39866
+    },
+    {
+      "epoch": 0.9489431590974008,
+      "grad_norm": 1.843055036589575,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 39867
+    },
+    {
+      "epoch": 0.9489669618204323,
+      "grad_norm": 1.9227993395003775,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 39868
+    },
+    {
+      "epoch": 0.9489907645434638,
+      "grad_norm": 2.331940267658217,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 39869
+    },
+    {
+      "epoch": 0.9490145672664952,
+      "grad_norm": 3.0862890682244397,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 39870
+    },
+    {
+      "epoch": 0.9490383699895268,
+      "grad_norm": 5.75182666448482,
+      "learning_rate": 2e-05,
+      "loss": 0.161,
+      "step": 39871
+    },
+    {
+      "epoch": 0.9490621727125583,
+      "grad_norm": 0.05181121748827503,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 39872
+    },
+    {
+      "epoch": 0.9490859754355898,
+      "grad_norm": 4.722984339684453,
+      "learning_rate": 2e-05,
+      "loss": 0.1439,
+      "step": 39873
+    },
+    {
+      "epoch": 0.9491097781586213,
+      "grad_norm": 1.9114174872232592,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 39874
+    },
+    {
+      "epoch": 0.9491335808816529,
+      "grad_norm": 0.03456733272853734,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39875
+    },
+    {
+      "epoch": 0.9491573836046844,
+      "grad_norm": 0.0023779508390048784,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 39876
+    },
+    {
+      "epoch": 0.9491811863277159,
+      "grad_norm": 0.8026734733724837,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 39877
+    },
+    {
+      "epoch": 0.9492049890507475,
+      "grad_norm": 0.3178494435132138,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 39878
+    },
+    {
+      "epoch": 0.949228791773779,
+      "grad_norm": 0.2756474785768114,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 39879
+    },
+    {
+      "epoch": 0.9492525944968104,
+      "grad_norm": 0.3782143049387503,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 39880
+    },
+    {
+      "epoch": 0.9492763972198419,
+      "grad_norm": 26.78880372561721,
+      "learning_rate": 2e-05,
+      "loss": 0.5129,
+      "step": 39881
+    },
+    {
+      "epoch": 0.9493001999428735,
+      "grad_norm": 24.506002513934856,
+      "learning_rate": 2e-05,
+      "loss": 0.524,
+      "step": 39882
+    },
+    {
+      "epoch": 0.949324002665905,
+      "grad_norm": 3.0532334739503195,
+      "learning_rate": 2e-05,
+      "loss": 0.0887,
+      "step": 39883
+    },
+    {
+      "epoch": 0.9493478053889365,
+      "grad_norm": 4.861306482125202,
+      "learning_rate": 2e-05,
+      "loss": 1.123,
+      "step": 39884
+    },
+    {
+      "epoch": 0.949371608111968,
+      "grad_norm": 0.23623150905996326,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 39885
+    },
+    {
+      "epoch": 0.9493954108349996,
+      "grad_norm": 0.01196044665077505,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 39886
+    },
+    {
+      "epoch": 0.949419213558031,
+      "grad_norm": 4.718900828281805,
+      "learning_rate": 2e-05,
+      "loss": 0.5573,
+      "step": 39887
+    },
+    {
+      "epoch": 0.9494430162810625,
+      "grad_norm": 0.9413614587707807,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 39888
+    },
+    {
+      "epoch": 0.949466819004094,
+      "grad_norm": 3.4084763109468272,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 39889
+    },
+    {
+      "epoch": 0.9494906217271256,
+      "grad_norm": 3.6137772724703887,
+      "learning_rate": 2e-05,
+      "loss": 0.4753,
+      "step": 39890
+    },
+    {
+      "epoch": 0.9495144244501571,
+      "grad_norm": 4.478098997828083,
+      "learning_rate": 2e-05,
+      "loss": 0.983,
+      "step": 39891
+    },
+    {
+      "epoch": 0.9495382271731886,
+      "grad_norm": 9.446517934970261,
+      "learning_rate": 2e-05,
+      "loss": 0.0806,
+      "step": 39892
+    },
+    {
+      "epoch": 0.9495620298962201,
+      "grad_norm": 1.9355822482938303,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 39893
+    },
+    {
+      "epoch": 0.9495858326192517,
+      "grad_norm": 11.528512355043413,
+      "learning_rate": 2e-05,
+      "loss": 0.5668,
+      "step": 39894
+    },
+    {
+      "epoch": 0.9496096353422832,
+      "grad_norm": 2.00026475924849,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 39895
+    },
+    {
+      "epoch": 0.9496334380653146,
+      "grad_norm": 2.933320920661324,
+      "learning_rate": 2e-05,
+      "loss": 0.6752,
+      "step": 39896
+    },
+    {
+      "epoch": 0.9496572407883462,
+      "grad_norm": 3.5788249521690685,
+      "learning_rate": 2e-05,
+      "loss": 0.1397,
+      "step": 39897
+    },
+    {
+      "epoch": 0.9496810435113777,
+      "grad_norm": 2.9058248851141757,
+      "learning_rate": 2e-05,
+      "loss": 0.0797,
+      "step": 39898
+    },
+    {
+      "epoch": 0.9497048462344092,
+      "grad_norm": 1.8068931984541503,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 39899
+    },
+    {
+      "epoch": 0.9497286489574407,
+      "grad_norm": 0.27329511422581265,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 39900
+    },
+    {
+      "epoch": 0.9497524516804723,
+      "grad_norm": 5.4816053067698896,
+      "learning_rate": 2e-05,
+      "loss": 0.7852,
+      "step": 39901
+    },
+    {
+      "epoch": 0.9497762544035038,
+      "grad_norm": 0.10963128649371638,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 39902
+    },
+    {
+      "epoch": 0.9498000571265353,
+      "grad_norm": 0.06547454588877676,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39903
+    },
+    {
+      "epoch": 0.9498238598495667,
+      "grad_norm": 2.6841645340962397,
+      "learning_rate": 2e-05,
+      "loss": 0.3926,
+      "step": 39904
+    },
+    {
+      "epoch": 0.9498476625725983,
+      "grad_norm": 1.2681279417613407,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 39905
+    },
+    {
+      "epoch": 0.9498714652956298,
+      "grad_norm": 6.329858510534198,
+      "learning_rate": 2e-05,
+      "loss": 1.0905,
+      "step": 39906
+    },
+    {
+      "epoch": 0.9498952680186613,
+      "grad_norm": 4.199480733851265,
+      "learning_rate": 2e-05,
+      "loss": 0.7275,
+      "step": 39907
+    },
+    {
+      "epoch": 0.9499190707416928,
+      "grad_norm": 20.05881503142466,
+      "learning_rate": 2e-05,
+      "loss": 0.3401,
+      "step": 39908
+    },
+    {
+      "epoch": 0.9499428734647244,
+      "grad_norm": 9.84256107453651,
+      "learning_rate": 2e-05,
+      "loss": 0.1304,
+      "step": 39909
+    },
+    {
+      "epoch": 0.9499666761877559,
+      "grad_norm": 7.356687307447152,
+      "learning_rate": 2e-05,
+      "loss": 1.0315,
+      "step": 39910
+    },
+    {
+      "epoch": 0.9499904789107874,
+      "grad_norm": 7.577203238599404,
+      "learning_rate": 2e-05,
+      "loss": 1.6175,
+      "step": 39911
+    },
+    {
+      "epoch": 0.950014281633819,
+      "grad_norm": 8.035720716059592,
+      "learning_rate": 2e-05,
+      "loss": 0.408,
+      "step": 39912
+    },
+    {
+      "epoch": 0.9500380843568504,
+      "grad_norm": 2.2956237793629857,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 39913
+    },
+    {
+      "epoch": 0.9500618870798819,
+      "grad_norm": 2.532986079794253,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 39914
+    },
+    {
+      "epoch": 0.9500856898029134,
+      "grad_norm": 2.529995122738967,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 39915
+    },
+    {
+      "epoch": 0.950109492525945,
+      "grad_norm": 5.6615644194226595,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 39916
+    },
+    {
+      "epoch": 0.9501332952489765,
+      "grad_norm": 1.5666696942017042,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 39917
+    },
+    {
+      "epoch": 0.950157097972008,
+      "grad_norm": 11.58294711797589,
+      "learning_rate": 2e-05,
+      "loss": 0.6748,
+      "step": 39918
+    },
+    {
+      "epoch": 0.9501809006950395,
+      "grad_norm": 0.028695081305002507,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39919
+    },
+    {
+      "epoch": 0.9502047034180711,
+      "grad_norm": 0.023681760265829573,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 39920
+    },
+    {
+      "epoch": 0.9502285061411025,
+      "grad_norm": 11.869557824167988,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 39921
+    },
+    {
+      "epoch": 0.950252308864134,
+      "grad_norm": 0.5623693486434125,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 39922
+    },
+    {
+      "epoch": 0.9502761115871655,
+      "grad_norm": 5.587506437338111,
+      "learning_rate": 2e-05,
+      "loss": 0.3781,
+      "step": 39923
+    },
+    {
+      "epoch": 0.9502999143101971,
+      "grad_norm": 4.018363669404817,
+      "learning_rate": 2e-05,
+      "loss": 0.7215,
+      "step": 39924
+    },
+    {
+      "epoch": 0.9503237170332286,
+      "grad_norm": 1.3527742417321444,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 39925
+    },
+    {
+      "epoch": 0.9503475197562601,
+      "grad_norm": 2.617989111078547,
+      "learning_rate": 2e-05,
+      "loss": 0.4462,
+      "step": 39926
+    },
+    {
+      "epoch": 0.9503713224792917,
+      "grad_norm": 0.37460385914485494,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 39927
+    },
+    {
+      "epoch": 0.9503951252023232,
+      "grad_norm": 4.6446272200526755,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 39928
+    },
+    {
+      "epoch": 0.9504189279253547,
+      "grad_norm": 26.82612317241772,
+      "learning_rate": 2e-05,
+      "loss": 1.9193,
+      "step": 39929
+    },
+    {
+      "epoch": 0.9504427306483861,
+      "grad_norm": 6.229175628789089,
+      "learning_rate": 2e-05,
+      "loss": 0.164,
+      "step": 39930
+    },
+    {
+      "epoch": 0.9504665333714177,
+      "grad_norm": 0.06183186680298095,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39931
+    },
+    {
+      "epoch": 0.9504903360944492,
+      "grad_norm": 0.20729883548747988,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 39932
+    },
+    {
+      "epoch": 0.9505141388174807,
+      "grad_norm": 1.515958497657607,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 39933
+    },
+    {
+      "epoch": 0.9505379415405122,
+      "grad_norm": 1.5027254545900457,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 39934
+    },
+    {
+      "epoch": 0.9505617442635438,
+      "grad_norm": 3.2259722718835397,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 39935
+    },
+    {
+      "epoch": 0.9505855469865753,
+      "grad_norm": 0.00834251948326354,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 39936
+    },
+    {
+      "epoch": 0.9506093497096068,
+      "grad_norm": 2.193950807815705,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 39937
+    },
+    {
+      "epoch": 0.9506331524326382,
+      "grad_norm": 4.971327870325575,
+      "learning_rate": 2e-05,
+      "loss": 0.2358,
+      "step": 39938
+    },
+    {
+      "epoch": 0.9506569551556698,
+      "grad_norm": 3.8523711126184663,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 39939
+    },
+    {
+      "epoch": 0.9506807578787013,
+      "grad_norm": 3.020297181754265,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 39940
+    },
+    {
+      "epoch": 0.9507045606017328,
+      "grad_norm": 3.7378957818518246,
+      "learning_rate": 2e-05,
+      "loss": 0.7262,
+      "step": 39941
+    },
+    {
+      "epoch": 0.9507283633247644,
+      "grad_norm": 3.2429072221074495,
+      "learning_rate": 2e-05,
+      "loss": 0.3557,
+      "step": 39942
+    },
+    {
+      "epoch": 0.9507521660477959,
+      "grad_norm": 0.06195877631071922,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 39943
+    },
+    {
+      "epoch": 0.9507759687708274,
+      "grad_norm": 0.3750198226915471,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 39944
+    },
+    {
+      "epoch": 0.9507997714938589,
+      "grad_norm": 9.037306311681828,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 39945
+    },
+    {
+      "epoch": 0.9508235742168905,
+      "grad_norm": 3.64263188358861,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 39946
+    },
+    {
+      "epoch": 0.9508473769399219,
+      "grad_norm": 7.330119777533223,
+      "learning_rate": 2e-05,
+      "loss": 0.7817,
+      "step": 39947
+    },
+    {
+      "epoch": 0.9508711796629534,
+      "grad_norm": 3.866925582981644,
+      "learning_rate": 2e-05,
+      "loss": 0.421,
+      "step": 39948
+    },
+    {
+      "epoch": 0.9508949823859849,
+      "grad_norm": 2.5450041172870135,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 39949
+    },
+    {
+      "epoch": 0.9509187851090165,
+      "grad_norm": 1.045983716038146,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 39950
+    },
+    {
+      "epoch": 0.950942587832048,
+      "grad_norm": 4.238387356479284,
+      "learning_rate": 2e-05,
+      "loss": 0.0453,
+      "step": 39951
+    },
+    {
+      "epoch": 0.9509663905550795,
+      "grad_norm": 7.169836778136607,
+      "learning_rate": 2e-05,
+      "loss": 1.1125,
+      "step": 39952
+    },
+    {
+      "epoch": 0.950990193278111,
+      "grad_norm": 6.012725226692191,
+      "learning_rate": 2e-05,
+      "loss": 0.4185,
+      "step": 39953
+    },
+    {
+      "epoch": 0.9510139960011426,
+      "grad_norm": 7.294297337908873,
+      "learning_rate": 2e-05,
+      "loss": 0.2924,
+      "step": 39954
+    },
+    {
+      "epoch": 0.951037798724174,
+      "grad_norm": 0.7360732969803782,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 39955
+    },
+    {
+      "epoch": 0.9510616014472055,
+      "grad_norm": 9.659368351592006,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 39956
+    },
+    {
+      "epoch": 0.9510854041702371,
+      "grad_norm": 2.5017798863760237,
+      "learning_rate": 2e-05,
+      "loss": 0.3404,
+      "step": 39957
+    },
+    {
+      "epoch": 0.9511092068932686,
+      "grad_norm": 4.087661809247035,
+      "learning_rate": 2e-05,
+      "loss": 0.5749,
+      "step": 39958
+    },
+    {
+      "epoch": 0.9511330096163001,
+      "grad_norm": 5.631813769792523,
+      "learning_rate": 2e-05,
+      "loss": 1.005,
+      "step": 39959
+    },
+    {
+      "epoch": 0.9511568123393316,
+      "grad_norm": 0.45305674270970103,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 39960
+    },
+    {
+      "epoch": 0.9511806150623632,
+      "grad_norm": 3.2717164319702268,
+      "learning_rate": 2e-05,
+      "loss": 0.1759,
+      "step": 39961
+    },
+    {
+      "epoch": 0.9512044177853947,
+      "grad_norm": 4.855233154282109,
+      "learning_rate": 2e-05,
+      "loss": 0.4012,
+      "step": 39962
+    },
+    {
+      "epoch": 0.9512282205084261,
+      "grad_norm": 13.307969048646314,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 39963
+    },
+    {
+      "epoch": 0.9512520232314576,
+      "grad_norm": 0.0793113939556136,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 39964
+    },
+    {
+      "epoch": 0.9512758259544892,
+      "grad_norm": 5.310451751600212,
+      "learning_rate": 2e-05,
+      "loss": 0.9221,
+      "step": 39965
+    },
+    {
+      "epoch": 0.9512996286775207,
+      "grad_norm": 0.3810766566328815,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 39966
+    },
+    {
+      "epoch": 0.9513234314005522,
+      "grad_norm": 10.088422418037903,
+      "learning_rate": 2e-05,
+      "loss": 2.0688,
+      "step": 39967
+    },
+    {
+      "epoch": 0.9513472341235837,
+      "grad_norm": 11.45228946761188,
+      "learning_rate": 2e-05,
+      "loss": 0.3851,
+      "step": 39968
+    },
+    {
+      "epoch": 0.9513710368466153,
+      "grad_norm": 0.5815526661468232,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 39969
+    },
+    {
+      "epoch": 0.9513948395696468,
+      "grad_norm": 0.4242539799385141,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 39970
+    },
+    {
+      "epoch": 0.9514186422926783,
+      "grad_norm": 0.6581005280505032,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 39971
+    },
+    {
+      "epoch": 0.9514424450157098,
+      "grad_norm": 4.075204864630292,
+      "learning_rate": 2e-05,
+      "loss": 0.7083,
+      "step": 39972
+    },
+    {
+      "epoch": 0.9514662477387413,
+      "grad_norm": 8.280644463587675,
+      "learning_rate": 2e-05,
+      "loss": 0.9917,
+      "step": 39973
+    },
+    {
+      "epoch": 0.9514900504617728,
+      "grad_norm": 3.5508338603623124,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 39974
+    },
+    {
+      "epoch": 0.9515138531848043,
+      "grad_norm": 5.647893486621476,
+      "learning_rate": 2e-05,
+      "loss": 1.2221,
+      "step": 39975
+    },
+    {
+      "epoch": 0.9515376559078359,
+      "grad_norm": 3.625455718119658,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 39976
+    },
+    {
+      "epoch": 0.9515614586308674,
+      "grad_norm": 0.029953704944423103,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39977
+    },
+    {
+      "epoch": 0.9515852613538989,
+      "grad_norm": 0.16176519880719914,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 39978
+    },
+    {
+      "epoch": 0.9516090640769304,
+      "grad_norm": 2.159870382818275,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 39979
+    },
+    {
+      "epoch": 0.951632866799962,
+      "grad_norm": 5.685904796266295,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 39980
+    },
+    {
+      "epoch": 0.9516566695229934,
+      "grad_norm": 0.03679678753303575,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 39981
+    },
+    {
+      "epoch": 0.9516804722460249,
+      "grad_norm": 3.2264864620707256,
+      "learning_rate": 2e-05,
+      "loss": 0.374,
+      "step": 39982
+    },
+    {
+      "epoch": 0.9517042749690564,
+      "grad_norm": 0.20022171585907195,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 39983
+    },
+    {
+      "epoch": 0.951728077692088,
+      "grad_norm": 2.700302265061597,
+      "learning_rate": 2e-05,
+      "loss": 0.1835,
+      "step": 39984
+    },
+    {
+      "epoch": 0.9517518804151195,
+      "grad_norm": 3.997222440936065,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 39985
+    },
+    {
+      "epoch": 0.951775683138151,
+      "grad_norm": 0.005377652463274714,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 39986
+    },
+    {
+      "epoch": 0.9517994858611826,
+      "grad_norm": 5.63795901512498,
+      "learning_rate": 2e-05,
+      "loss": 0.5029,
+      "step": 39987
+    },
+    {
+      "epoch": 0.9518232885842141,
+      "grad_norm": 0.22403164699119948,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 39988
+    },
+    {
+      "epoch": 0.9518470913072455,
+      "grad_norm": 2.1334973858395228,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 39989
+    },
+    {
+      "epoch": 0.951870894030277,
+      "grad_norm": 4.040876807543493,
+      "learning_rate": 2e-05,
+      "loss": 0.7116,
+      "step": 39990
+    },
+    {
+      "epoch": 0.9518946967533086,
+      "grad_norm": 0.032198211066578795,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 39991
+    },
+    {
+      "epoch": 0.9519184994763401,
+      "grad_norm": 4.991534792216363,
+      "learning_rate": 2e-05,
+      "loss": 0.995,
+      "step": 39992
+    },
+    {
+      "epoch": 0.9519423021993716,
+      "grad_norm": 1.1492696206399964,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 39993
+    },
+    {
+      "epoch": 0.9519661049224031,
+      "grad_norm": 6.205272353788158,
+      "learning_rate": 2e-05,
+      "loss": 0.2131,
+      "step": 39994
+    },
+    {
+      "epoch": 0.9519899076454347,
+      "grad_norm": 1.7492002116811085,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 39995
+    },
+    {
+      "epoch": 0.9520137103684662,
+      "grad_norm": 2.471763748657995,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 39996
+    },
+    {
+      "epoch": 0.9520375130914976,
+      "grad_norm": 5.979993451114626,
+      "learning_rate": 2e-05,
+      "loss": 0.9344,
+      "step": 39997
+    },
+    {
+      "epoch": 0.9520613158145291,
+      "grad_norm": 4.885273020353307,
+      "learning_rate": 2e-05,
+      "loss": 0.8536,
+      "step": 39998
+    },
+    {
+      "epoch": 0.9520851185375607,
+      "grad_norm": 1.8753263059042722,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 39999
+    },
+    {
+      "epoch": 0.9521089212605922,
+      "grad_norm": 1.8683919263712874,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 40000
+    },
+    {
+      "epoch": 0.9521327239836237,
+      "grad_norm": 3.8895277930440044,
+      "learning_rate": 2e-05,
+      "loss": 0.2307,
+      "step": 40001
+    },
+    {
+      "epoch": 0.9521565267066553,
+      "grad_norm": 0.412059048405203,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 40002
+    },
+    {
+      "epoch": 0.9521803294296868,
+      "grad_norm": 11.779990830552846,
+      "learning_rate": 2e-05,
+      "loss": 1.1894,
+      "step": 40003
+    },
+    {
+      "epoch": 0.9522041321527183,
+      "grad_norm": 8.668901887619466,
+      "learning_rate": 2e-05,
+      "loss": 0.4334,
+      "step": 40004
+    },
+    {
+      "epoch": 0.9522279348757497,
+      "grad_norm": 1.7276562307778585,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 40005
+    },
+    {
+      "epoch": 0.9522517375987813,
+      "grad_norm": 2.9921937066817796,
+      "learning_rate": 2e-05,
+      "loss": 0.523,
+      "step": 40006
+    },
+    {
+      "epoch": 0.9522755403218128,
+      "grad_norm": 0.19064282061322949,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 40007
+    },
+    {
+      "epoch": 0.9522993430448443,
+      "grad_norm": 0.035341362552065075,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40008
+    },
+    {
+      "epoch": 0.9523231457678758,
+      "grad_norm": 14.101539989319267,
+      "learning_rate": 2e-05,
+      "loss": 0.4376,
+      "step": 40009
+    },
+    {
+      "epoch": 0.9523469484909074,
+      "grad_norm": 4.375656009096326,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 40010
+    },
+    {
+      "epoch": 0.9523707512139389,
+      "grad_norm": 0.20411681760970019,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 40011
+    },
+    {
+      "epoch": 0.9523945539369704,
+      "grad_norm": 8.730644555304359,
+      "learning_rate": 2e-05,
+      "loss": 0.495,
+      "step": 40012
+    },
+    {
+      "epoch": 0.9524183566600019,
+      "grad_norm": 4.0765649967273045,
+      "learning_rate": 2e-05,
+      "loss": 0.4348,
+      "step": 40013
+    },
+    {
+      "epoch": 0.9524421593830334,
+      "grad_norm": 0.027574200024311036,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40014
+    },
+    {
+      "epoch": 0.9524659621060649,
+      "grad_norm": 11.048078724293303,
+      "learning_rate": 2e-05,
+      "loss": 0.6382,
+      "step": 40015
+    },
+    {
+      "epoch": 0.9524897648290964,
+      "grad_norm": 9.930345020404127,
+      "learning_rate": 2e-05,
+      "loss": 0.127,
+      "step": 40016
+    },
+    {
+      "epoch": 0.952513567552128,
+      "grad_norm": 0.8092094402629902,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 40017
+    },
+    {
+      "epoch": 0.9525373702751595,
+      "grad_norm": 4.960843690694684,
+      "learning_rate": 2e-05,
+      "loss": 0.1976,
+      "step": 40018
+    },
+    {
+      "epoch": 0.952561172998191,
+      "grad_norm": 5.397997982760314,
+      "learning_rate": 2e-05,
+      "loss": 0.796,
+      "step": 40019
+    },
+    {
+      "epoch": 0.9525849757212225,
+      "grad_norm": 3.9652248639429373,
+      "learning_rate": 2e-05,
+      "loss": 0.2571,
+      "step": 40020
+    },
+    {
+      "epoch": 0.9526087784442541,
+      "grad_norm": 5.288903832983528,
+      "learning_rate": 2e-05,
+      "loss": 0.5415,
+      "step": 40021
+    },
+    {
+      "epoch": 0.9526325811672856,
+      "grad_norm": 4.397097395801043,
+      "learning_rate": 2e-05,
+      "loss": 1.0479,
+      "step": 40022
+    },
+    {
+      "epoch": 0.952656383890317,
+      "grad_norm": 0.05483795725353286,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40023
+    },
+    {
+      "epoch": 0.9526801866133485,
+      "grad_norm": 9.887756614241978,
+      "learning_rate": 2e-05,
+      "loss": 0.24,
+      "step": 40024
+    },
+    {
+      "epoch": 0.9527039893363801,
+      "grad_norm": 5.126708231883211,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 40025
+    },
+    {
+      "epoch": 0.9527277920594116,
+      "grad_norm": 2.9706362094342973,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 40026
+    },
+    {
+      "epoch": 0.9527515947824431,
+      "grad_norm": 4.045678810652318,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 40027
+    },
+    {
+      "epoch": 0.9527753975054746,
+      "grad_norm": 0.9339156919931866,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 40028
+    },
+    {
+      "epoch": 0.9527992002285062,
+      "grad_norm": 5.704933928270383,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 40029
+    },
+    {
+      "epoch": 0.9528230029515377,
+      "grad_norm": 4.572234535941931,
+      "learning_rate": 2e-05,
+      "loss": 0.9851,
+      "step": 40030
+    },
+    {
+      "epoch": 0.9528468056745691,
+      "grad_norm": 6.869295243110394,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 40031
+    },
+    {
+      "epoch": 0.9528706083976007,
+      "grad_norm": 4.188833134828039,
+      "learning_rate": 2e-05,
+      "loss": 0.045,
+      "step": 40032
+    },
+    {
+      "epoch": 0.9528944111206322,
+      "grad_norm": 20.341941877307086,
+      "learning_rate": 2e-05,
+      "loss": 0.9312,
+      "step": 40033
+    },
+    {
+      "epoch": 0.9529182138436637,
+      "grad_norm": 3.3263979964151837,
+      "learning_rate": 2e-05,
+      "loss": 0.4552,
+      "step": 40034
+    },
+    {
+      "epoch": 0.9529420165666952,
+      "grad_norm": 4.991132272062749,
+      "learning_rate": 2e-05,
+      "loss": 0.9753,
+      "step": 40035
+    },
+    {
+      "epoch": 0.9529658192897268,
+      "grad_norm": 8.058315968768538,
+      "learning_rate": 2e-05,
+      "loss": 0.4552,
+      "step": 40036
+    },
+    {
+      "epoch": 0.9529896220127583,
+      "grad_norm": 2.971223806507348,
+      "learning_rate": 2e-05,
+      "loss": 0.3448,
+      "step": 40037
+    },
+    {
+      "epoch": 0.9530134247357898,
+      "grad_norm": 5.22917064200349,
+      "learning_rate": 2e-05,
+      "loss": 0.9512,
+      "step": 40038
+    },
+    {
+      "epoch": 0.9530372274588212,
+      "grad_norm": 2.4604392691767254,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 40039
+    },
+    {
+      "epoch": 0.9530610301818528,
+      "grad_norm": 3.7666269172358513,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 40040
+    },
+    {
+      "epoch": 0.9530848329048843,
+      "grad_norm": 15.148059564582075,
+      "learning_rate": 2e-05,
+      "loss": 1.01,
+      "step": 40041
+    },
+    {
+      "epoch": 0.9531086356279158,
+      "grad_norm": 4.57748757448982,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 40042
+    },
+    {
+      "epoch": 0.9531324383509473,
+      "grad_norm": 14.043364786453898,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 40043
+    },
+    {
+      "epoch": 0.9531562410739789,
+      "grad_norm": 1.5136599099936112,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 40044
+    },
+    {
+      "epoch": 0.9531800437970104,
+      "grad_norm": 0.9556595914510834,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 40045
+    },
+    {
+      "epoch": 0.9532038465200419,
+      "grad_norm": 1.6237640289751034,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 40046
+    },
+    {
+      "epoch": 0.9532276492430735,
+      "grad_norm": 0.10285792157840348,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 40047
+    },
+    {
+      "epoch": 0.953251451966105,
+      "grad_norm": 2.321603001371283,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 40048
+    },
+    {
+      "epoch": 0.9532752546891364,
+      "grad_norm": 0.03639623684035599,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40049
+    },
+    {
+      "epoch": 0.9532990574121679,
+      "grad_norm": 2.2834664968048903,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 40050
+    },
+    {
+      "epoch": 0.9533228601351995,
+      "grad_norm": 0.08083063200752774,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 40051
+    },
+    {
+      "epoch": 0.953346662858231,
+      "grad_norm": 2.9225221467524554,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 40052
+    },
+    {
+      "epoch": 0.9533704655812625,
+      "grad_norm": 5.6291863904118635,
+      "learning_rate": 2e-05,
+      "loss": 0.6644,
+      "step": 40053
+    },
+    {
+      "epoch": 0.953394268304294,
+      "grad_norm": 2.16677726193327,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 40054
+    },
+    {
+      "epoch": 0.9534180710273256,
+      "grad_norm": 1.797673510589994,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 40055
+    },
+    {
+      "epoch": 0.953441873750357,
+      "grad_norm": 1.1683452081711543,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 40056
+    },
+    {
+      "epoch": 0.9534656764733885,
+      "grad_norm": 4.088963416297762,
+      "learning_rate": 2e-05,
+      "loss": 0.693,
+      "step": 40057
+    },
+    {
+      "epoch": 0.95348947919642,
+      "grad_norm": 1.486095767777324,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 40058
+    },
+    {
+      "epoch": 0.9535132819194516,
+      "grad_norm": 0.25291522604764366,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 40059
+    },
+    {
+      "epoch": 0.9535370846424831,
+      "grad_norm": 9.079136275005697,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 40060
+    },
+    {
+      "epoch": 0.9535608873655146,
+      "grad_norm": 3.3547283820528513,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 40061
+    },
+    {
+      "epoch": 0.9535846900885462,
+      "grad_norm": 0.024790047979224305,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40062
+    },
+    {
+      "epoch": 0.9536084928115777,
+      "grad_norm": 5.055641074481456,
+      "learning_rate": 2e-05,
+      "loss": 0.5772,
+      "step": 40063
+    },
+    {
+      "epoch": 0.9536322955346092,
+      "grad_norm": 5.798362312664779,
+      "learning_rate": 2e-05,
+      "loss": 0.8036,
+      "step": 40064
+    },
+    {
+      "epoch": 0.9536560982576406,
+      "grad_norm": 27.711322437429104,
+      "learning_rate": 2e-05,
+      "loss": 0.4275,
+      "step": 40065
+    },
+    {
+      "epoch": 0.9536799009806722,
+      "grad_norm": 2.676688448131387,
+      "learning_rate": 2e-05,
+      "loss": 0.4741,
+      "step": 40066
+    },
+    {
+      "epoch": 0.9537037037037037,
+      "grad_norm": 4.462390334662508,
+      "learning_rate": 2e-05,
+      "loss": 0.8833,
+      "step": 40067
+    },
+    {
+      "epoch": 0.9537275064267352,
+      "grad_norm": 0.03415394437763075,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40068
+    },
+    {
+      "epoch": 0.9537513091497667,
+      "grad_norm": 32.70517863594894,
+      "learning_rate": 2e-05,
+      "loss": 1.8868,
+      "step": 40069
+    },
+    {
+      "epoch": 0.9537751118727983,
+      "grad_norm": 0.2913100363374073,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 40070
+    },
+    {
+      "epoch": 0.9537989145958298,
+      "grad_norm": 5.8788977202051536,
+      "learning_rate": 2e-05,
+      "loss": 0.1262,
+      "step": 40071
+    },
+    {
+      "epoch": 0.9538227173188613,
+      "grad_norm": 11.843781537398417,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 40072
+    },
+    {
+      "epoch": 0.9538465200418927,
+      "grad_norm": 0.39617019664673986,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 40073
+    },
+    {
+      "epoch": 0.9538703227649243,
+      "grad_norm": 35.14488680881118,
+      "learning_rate": 2e-05,
+      "loss": 1.1704,
+      "step": 40074
+    },
+    {
+      "epoch": 0.9538941254879558,
+      "grad_norm": 0.8081678970937693,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 40075
+    },
+    {
+      "epoch": 0.9539179282109873,
+      "grad_norm": 5.079250279844167,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 40076
+    },
+    {
+      "epoch": 0.9539417309340189,
+      "grad_norm": 3.42199005532831,
+      "learning_rate": 2e-05,
+      "loss": 0.3966,
+      "step": 40077
+    },
+    {
+      "epoch": 0.9539655336570504,
+      "grad_norm": 17.05077252793584,
+      "learning_rate": 2e-05,
+      "loss": 0.4251,
+      "step": 40078
+    },
+    {
+      "epoch": 0.9539893363800819,
+      "grad_norm": 0.041057434496547744,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40079
+    },
+    {
+      "epoch": 0.9540131391031134,
+      "grad_norm": 4.075852672018497,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 40080
+    },
+    {
+      "epoch": 0.954036941826145,
+      "grad_norm": 5.800188373660791,
+      "learning_rate": 2e-05,
+      "loss": 0.7589,
+      "step": 40081
+    },
+    {
+      "epoch": 0.9540607445491764,
+      "grad_norm": 14.48403005795757,
+      "learning_rate": 2e-05,
+      "loss": 0.1443,
+      "step": 40082
+    },
+    {
+      "epoch": 0.9540845472722079,
+      "grad_norm": 3.8795217883883613,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 40083
+    },
+    {
+      "epoch": 0.9541083499952394,
+      "grad_norm": 16.20498126910121,
+      "learning_rate": 2e-05,
+      "loss": 1.359,
+      "step": 40084
+    },
+    {
+      "epoch": 0.954132152718271,
+      "grad_norm": 0.03450682624468081,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40085
+    },
+    {
+      "epoch": 0.9541559554413025,
+      "grad_norm": 5.955437663011451,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 40086
+    },
+    {
+      "epoch": 0.954179758164334,
+      "grad_norm": 0.3385498773663746,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 40087
+    },
+    {
+      "epoch": 0.9542035608873655,
+      "grad_norm": 3.307589080264454,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 40088
+    },
+    {
+      "epoch": 0.9542273636103971,
+      "grad_norm": 0.09983610664982082,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 40089
+    },
+    {
+      "epoch": 0.9542511663334285,
+      "grad_norm": 4.400729625322612,
+      "learning_rate": 2e-05,
+      "loss": 0.6948,
+      "step": 40090
+    },
+    {
+      "epoch": 0.95427496905646,
+      "grad_norm": 6.241076791720763,
+      "learning_rate": 2e-05,
+      "loss": 0.2461,
+      "step": 40091
+    },
+    {
+      "epoch": 0.9542987717794916,
+      "grad_norm": 3.345015898136533,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 40092
+    },
+    {
+      "epoch": 0.9543225745025231,
+      "grad_norm": 0.8351788154997914,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 40093
+    },
+    {
+      "epoch": 0.9543463772255546,
+      "grad_norm": 0.13633368156803513,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 40094
+    },
+    {
+      "epoch": 0.9543701799485861,
+      "grad_norm": 1.3295541043559356,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 40095
+    },
+    {
+      "epoch": 0.9543939826716177,
+      "grad_norm": 2.7362724175802646,
+      "learning_rate": 2e-05,
+      "loss": 0.5261,
+      "step": 40096
+    },
+    {
+      "epoch": 0.9544177853946492,
+      "grad_norm": 3.072289572412963,
+      "learning_rate": 2e-05,
+      "loss": 0.5937,
+      "step": 40097
+    },
+    {
+      "epoch": 0.9544415881176807,
+      "grad_norm": 0.14872129698839312,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 40098
+    },
+    {
+      "epoch": 0.9544653908407121,
+      "grad_norm": 2.4466069275096163,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 40099
+    },
+    {
+      "epoch": 0.9544891935637437,
+      "grad_norm": 7.389003039503729,
+      "learning_rate": 2e-05,
+      "loss": 0.9298,
+      "step": 40100
+    },
+    {
+      "epoch": 0.9545129962867752,
+      "grad_norm": 7.822836297862359,
+      "learning_rate": 2e-05,
+      "loss": 0.3061,
+      "step": 40101
+    },
+    {
+      "epoch": 0.9545367990098067,
+      "grad_norm": 1.4806857181052475,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 40102
+    },
+    {
+      "epoch": 0.9545606017328382,
+      "grad_norm": 2.9132030794618484,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 40103
+    },
+    {
+      "epoch": 0.9545844044558698,
+      "grad_norm": 1.4002269581516886,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 40104
+    },
+    {
+      "epoch": 0.9546082071789013,
+      "grad_norm": 1.189432282515808,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 40105
+    },
+    {
+      "epoch": 0.9546320099019328,
+      "grad_norm": 4.972775647588384,
+      "learning_rate": 2e-05,
+      "loss": 0.1037,
+      "step": 40106
+    },
+    {
+      "epoch": 0.9546558126249644,
+      "grad_norm": 3.3490281775594513,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 40107
+    },
+    {
+      "epoch": 0.9546796153479958,
+      "grad_norm": 3.230467189445909,
+      "learning_rate": 2e-05,
+      "loss": 0.4558,
+      "step": 40108
+    },
+    {
+      "epoch": 0.9547034180710273,
+      "grad_norm": 1.9319550222000454,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 40109
+    },
+    {
+      "epoch": 0.9547272207940588,
+      "grad_norm": 0.32826513988949096,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 40110
+    },
+    {
+      "epoch": 0.9547510235170904,
+      "grad_norm": 16.13055048096264,
+      "learning_rate": 2e-05,
+      "loss": 0.4706,
+      "step": 40111
+    },
+    {
+      "epoch": 0.9547748262401219,
+      "grad_norm": 2.466989765429501,
+      "learning_rate": 2e-05,
+      "loss": 0.5187,
+      "step": 40112
+    },
+    {
+      "epoch": 0.9547986289631534,
+      "grad_norm": 0.03271337664157592,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40113
+    },
+    {
+      "epoch": 0.9548224316861849,
+      "grad_norm": 3.7648928111830275,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 40114
+    },
+    {
+      "epoch": 0.9548462344092165,
+      "grad_norm": 2.8090961988860967,
+      "learning_rate": 2e-05,
+      "loss": 0.4104,
+      "step": 40115
+    },
+    {
+      "epoch": 0.9548700371322479,
+      "grad_norm": 4.359677664765205,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 40116
+    },
+    {
+      "epoch": 0.9548938398552794,
+      "grad_norm": 8.268854526912502,
+      "learning_rate": 2e-05,
+      "loss": 1.1637,
+      "step": 40117
+    },
+    {
+      "epoch": 0.9549176425783109,
+      "grad_norm": 6.899828324363522,
+      "learning_rate": 2e-05,
+      "loss": 0.7847,
+      "step": 40118
+    },
+    {
+      "epoch": 0.9549414453013425,
+      "grad_norm": 4.011884902169065,
+      "learning_rate": 2e-05,
+      "loss": 0.0313,
+      "step": 40119
+    },
+    {
+      "epoch": 0.954965248024374,
+      "grad_norm": 3.8654385452436215,
+      "learning_rate": 2e-05,
+      "loss": 0.4145,
+      "step": 40120
+    },
+    {
+      "epoch": 0.9549890507474055,
+      "grad_norm": 3.772998715608576,
+      "learning_rate": 2e-05,
+      "loss": 0.6041,
+      "step": 40121
+    },
+    {
+      "epoch": 0.9550128534704371,
+      "grad_norm": 2.244567116361235,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 40122
+    },
+    {
+      "epoch": 0.9550366561934686,
+      "grad_norm": 0.04132563140503689,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40123
+    },
+    {
+      "epoch": 0.9550604589165,
+      "grad_norm": 1.2744989827242126,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 40124
+    },
+    {
+      "epoch": 0.9550842616395315,
+      "grad_norm": 1.5273528363746607,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 40125
+    },
+    {
+      "epoch": 0.9551080643625631,
+      "grad_norm": 0.9156728786207916,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 40126
+    },
+    {
+      "epoch": 0.9551318670855946,
+      "grad_norm": 8.83880926597255,
+      "learning_rate": 2e-05,
+      "loss": 1.0127,
+      "step": 40127
+    },
+    {
+      "epoch": 0.9551556698086261,
+      "grad_norm": 1.4140098351895805,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 40128
+    },
+    {
+      "epoch": 0.9551794725316576,
+      "grad_norm": 1.0632615966187033,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 40129
+    },
+    {
+      "epoch": 0.9552032752546892,
+      "grad_norm": 0.14881541897121994,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 40130
+    },
+    {
+      "epoch": 0.9552270779777207,
+      "grad_norm": 8.597750222833351,
+      "learning_rate": 2e-05,
+      "loss": 1.0233,
+      "step": 40131
+    },
+    {
+      "epoch": 0.9552508807007521,
+      "grad_norm": 1.0529330552194276,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 40132
+    },
+    {
+      "epoch": 0.9552746834237836,
+      "grad_norm": 1.4176172122944493,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 40133
+    },
+    {
+      "epoch": 0.9552984861468152,
+      "grad_norm": 0.016479758973830225,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40134
+    },
+    {
+      "epoch": 0.9553222888698467,
+      "grad_norm": 0.02311937195302502,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40135
+    },
+    {
+      "epoch": 0.9553460915928782,
+      "grad_norm": 10.533745805887229,
+      "learning_rate": 2e-05,
+      "loss": 0.4284,
+      "step": 40136
+    },
+    {
+      "epoch": 0.9553698943159097,
+      "grad_norm": 11.062247466056439,
+      "learning_rate": 2e-05,
+      "loss": 0.2859,
+      "step": 40137
+    },
+    {
+      "epoch": 0.9553936970389413,
+      "grad_norm": 14.661681738892195,
+      "learning_rate": 2e-05,
+      "loss": 0.0933,
+      "step": 40138
+    },
+    {
+      "epoch": 0.9554174997619728,
+      "grad_norm": 0.12656014591057066,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 40139
+    },
+    {
+      "epoch": 0.9554413024850043,
+      "grad_norm": 2.9298805866356523,
+      "learning_rate": 2e-05,
+      "loss": 0.1083,
+      "step": 40140
+    },
+    {
+      "epoch": 0.9554651052080358,
+      "grad_norm": 0.3229160831977943,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 40141
+    },
+    {
+      "epoch": 0.9554889079310673,
+      "grad_norm": 3.976052705944483,
+      "learning_rate": 2e-05,
+      "loss": 1.1779,
+      "step": 40142
+    },
+    {
+      "epoch": 0.9555127106540988,
+      "grad_norm": 7.399518976792421,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 40143
+    },
+    {
+      "epoch": 0.9555365133771303,
+      "grad_norm": 29.85973829217428,
+      "learning_rate": 2e-05,
+      "loss": 0.6954,
+      "step": 40144
+    },
+    {
+      "epoch": 0.9555603161001619,
+      "grad_norm": 13.079779244246048,
+      "learning_rate": 2e-05,
+      "loss": 1.2728,
+      "step": 40145
+    },
+    {
+      "epoch": 0.9555841188231934,
+      "grad_norm": 3.2643910389696242,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 40146
+    },
+    {
+      "epoch": 0.9556079215462249,
+      "grad_norm": 0.02130558095096029,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40147
+    },
+    {
+      "epoch": 0.9556317242692564,
+      "grad_norm": 0.3361657931180336,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 40148
+    },
+    {
+      "epoch": 0.955655526992288,
+      "grad_norm": 0.2672975051466634,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 40149
+    },
+    {
+      "epoch": 0.9556793297153194,
+      "grad_norm": 0.07866523683907879,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 40150
+    },
+    {
+      "epoch": 0.9557031324383509,
+      "grad_norm": 6.121555253930255,
+      "learning_rate": 2e-05,
+      "loss": 0.2601,
+      "step": 40151
+    },
+    {
+      "epoch": 0.9557269351613824,
+      "grad_norm": 0.006638750201573594,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 40152
+    },
+    {
+      "epoch": 0.955750737884414,
+      "grad_norm": 0.05726201145451143,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40153
+    },
+    {
+      "epoch": 0.9557745406074455,
+      "grad_norm": 0.047918869608714464,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40154
+    },
+    {
+      "epoch": 0.955798343330477,
+      "grad_norm": 3.200203981940515,
+      "learning_rate": 2e-05,
+      "loss": 0.3459,
+      "step": 40155
+    },
+    {
+      "epoch": 0.9558221460535086,
+      "grad_norm": 9.57224480027901,
+      "learning_rate": 2e-05,
+      "loss": 0.589,
+      "step": 40156
+    },
+    {
+      "epoch": 0.95584594877654,
+      "grad_norm": 24.46702574904273,
+      "learning_rate": 2e-05,
+      "loss": 1.5516,
+      "step": 40157
+    },
+    {
+      "epoch": 0.9558697514995715,
+      "grad_norm": 3.0519239497520774,
+      "learning_rate": 2e-05,
+      "loss": 0.1003,
+      "step": 40158
+    },
+    {
+      "epoch": 0.955893554222603,
+      "grad_norm": 3.3313031809046487,
+      "learning_rate": 2e-05,
+      "loss": 0.5569,
+      "step": 40159
+    },
+    {
+      "epoch": 0.9559173569456346,
+      "grad_norm": 24.73698909139981,
+      "learning_rate": 2e-05,
+      "loss": 0.5243,
+      "step": 40160
+    },
+    {
+      "epoch": 0.9559411596686661,
+      "grad_norm": 0.1372858619940044,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 40161
+    },
+    {
+      "epoch": 0.9559649623916976,
+      "grad_norm": 4.943168462072311,
+      "learning_rate": 2e-05,
+      "loss": 1.0754,
+      "step": 40162
+    },
+    {
+      "epoch": 0.9559887651147291,
+      "grad_norm": 1.486126137375937,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 40163
+    },
+    {
+      "epoch": 0.9560125678377607,
+      "grad_norm": 2.284911265751517,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 40164
+    },
+    {
+      "epoch": 0.9560363705607922,
+      "grad_norm": 14.679827540683874,
+      "learning_rate": 2e-05,
+      "loss": 0.4836,
+      "step": 40165
+    },
+    {
+      "epoch": 0.9560601732838236,
+      "grad_norm": 0.4100498672513061,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 40166
+    },
+    {
+      "epoch": 0.9560839760068551,
+      "grad_norm": 4.7586622651818375,
+      "learning_rate": 2e-05,
+      "loss": 0.1288,
+      "step": 40167
+    },
+    {
+      "epoch": 0.9561077787298867,
+      "grad_norm": 0.04759262600872138,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40168
+    },
+    {
+      "epoch": 0.9561315814529182,
+      "grad_norm": 7.768239714022877,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 40169
+    },
+    {
+      "epoch": 0.9561553841759497,
+      "grad_norm": 5.315628214035083,
+      "learning_rate": 2e-05,
+      "loss": 0.6951,
+      "step": 40170
+    },
+    {
+      "epoch": 0.9561791868989813,
+      "grad_norm": 9.129972199682983,
+      "learning_rate": 2e-05,
+      "loss": 0.47,
+      "step": 40171
+    },
+    {
+      "epoch": 0.9562029896220128,
+      "grad_norm": 5.011987803173975,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 40172
+    },
+    {
+      "epoch": 0.9562267923450443,
+      "grad_norm": 3.6614987119567513,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 40173
+    },
+    {
+      "epoch": 0.9562505950680757,
+      "grad_norm": 3.790727577285722,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 40174
+    },
+    {
+      "epoch": 0.9562743977911073,
+      "grad_norm": 3.1123739073913677,
+      "learning_rate": 2e-05,
+      "loss": 0.4332,
+      "step": 40175
+    },
+    {
+      "epoch": 0.9562982005141388,
+      "grad_norm": 0.8568586364328857,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 40176
+    },
+    {
+      "epoch": 0.9563220032371703,
+      "grad_norm": 7.10630260312944,
+      "learning_rate": 2e-05,
+      "loss": 1.1569,
+      "step": 40177
+    },
+    {
+      "epoch": 0.9563458059602018,
+      "grad_norm": 0.04824497093656371,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40178
+    },
+    {
+      "epoch": 0.9563696086832334,
+      "grad_norm": 3.217967030223514,
+      "learning_rate": 2e-05,
+      "loss": 0.5915,
+      "step": 40179
+    },
+    {
+      "epoch": 0.9563934114062649,
+      "grad_norm": 6.162406307471172,
+      "learning_rate": 2e-05,
+      "loss": 0.6915,
+      "step": 40180
+    },
+    {
+      "epoch": 0.9564172141292964,
+      "grad_norm": 0.13583542275145633,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 40181
+    },
+    {
+      "epoch": 0.9564410168523279,
+      "grad_norm": 2.4454819778274923,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 40182
+    },
+    {
+      "epoch": 0.9564648195753594,
+      "grad_norm": 5.252904244620527,
+      "learning_rate": 2e-05,
+      "loss": 0.9333,
+      "step": 40183
+    },
+    {
+      "epoch": 0.9564886222983909,
+      "grad_norm": 8.004087841007603,
+      "learning_rate": 2e-05,
+      "loss": 0.6365,
+      "step": 40184
+    },
+    {
+      "epoch": 0.9565124250214224,
+      "grad_norm": 2.9192831597236535,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 40185
+    },
+    {
+      "epoch": 0.956536227744454,
+      "grad_norm": 1.6740397005050462,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 40186
+    },
+    {
+      "epoch": 0.9565600304674855,
+      "grad_norm": 3.0024789733607693,
+      "learning_rate": 2e-05,
+      "loss": 0.37,
+      "step": 40187
+    },
+    {
+      "epoch": 0.956583833190517,
+      "grad_norm": 2.0161425638933004,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 40188
+    },
+    {
+      "epoch": 0.9566076359135485,
+      "grad_norm": 7.192113384727459,
+      "learning_rate": 2e-05,
+      "loss": 1.7563,
+      "step": 40189
+    },
+    {
+      "epoch": 0.9566314386365801,
+      "grad_norm": 9.095424152963801,
+      "learning_rate": 2e-05,
+      "loss": 0.554,
+      "step": 40190
+    },
+    {
+      "epoch": 0.9566552413596116,
+      "grad_norm": 6.613680177993006,
+      "learning_rate": 2e-05,
+      "loss": 1.6384,
+      "step": 40191
+    },
+    {
+      "epoch": 0.956679044082643,
+      "grad_norm": 3.0859618811723966,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 40192
+    },
+    {
+      "epoch": 0.9567028468056745,
+      "grad_norm": 2.9927281940327255,
+      "learning_rate": 2e-05,
+      "loss": 0.5309,
+      "step": 40193
+    },
+    {
+      "epoch": 0.9567266495287061,
+      "grad_norm": 5.1115293132040405,
+      "learning_rate": 2e-05,
+      "loss": 0.2355,
+      "step": 40194
+    },
+    {
+      "epoch": 0.9567504522517376,
+      "grad_norm": 0.2802869385307838,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 40195
+    },
+    {
+      "epoch": 0.9567742549747691,
+      "grad_norm": 4.1878285501029975,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 40196
+    },
+    {
+      "epoch": 0.9567980576978006,
+      "grad_norm": 78.68379717670321,
+      "learning_rate": 2e-05,
+      "loss": 1.2436,
+      "step": 40197
+    },
+    {
+      "epoch": 0.9568218604208322,
+      "grad_norm": 3.8743564821212324,
+      "learning_rate": 2e-05,
+      "loss": 0.5715,
+      "step": 40198
+    },
+    {
+      "epoch": 0.9568456631438637,
+      "grad_norm": 6.780009415220026,
+      "learning_rate": 2e-05,
+      "loss": 1.056,
+      "step": 40199
+    },
+    {
+      "epoch": 0.9568694658668951,
+      "grad_norm": 0.13562470037583266,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 40200
+    },
+    {
+      "epoch": 0.9568932685899267,
+      "grad_norm": 3.3959801359731765,
+      "learning_rate": 2e-05,
+      "loss": 0.0857,
+      "step": 40201
+    },
+    {
+      "epoch": 0.9569170713129582,
+      "grad_norm": 3.6595147084994544,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 40202
+    },
+    {
+      "epoch": 0.9569408740359897,
+      "grad_norm": 1.6575674265378875,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 40203
+    },
+    {
+      "epoch": 0.9569646767590212,
+      "grad_norm": 8.035221749871978,
+      "learning_rate": 2e-05,
+      "loss": 0.5949,
+      "step": 40204
+    },
+    {
+      "epoch": 0.9569884794820528,
+      "grad_norm": 6.975083923171384,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 40205
+    },
+    {
+      "epoch": 0.9570122822050843,
+      "grad_norm": 0.12642434181968387,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 40206
+    },
+    {
+      "epoch": 0.9570360849281158,
+      "grad_norm": 4.33270297882095,
+      "learning_rate": 2e-05,
+      "loss": 1.0674,
+      "step": 40207
+    },
+    {
+      "epoch": 0.9570598876511472,
+      "grad_norm": 3.465803804569857,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 40208
+    },
+    {
+      "epoch": 0.9570836903741788,
+      "grad_norm": 2.904406105066063,
+      "learning_rate": 2e-05,
+      "loss": 0.426,
+      "step": 40209
+    },
+    {
+      "epoch": 0.9571074930972103,
+      "grad_norm": 2.3513014946275557,
+      "learning_rate": 2e-05,
+      "loss": 0.2656,
+      "step": 40210
+    },
+    {
+      "epoch": 0.9571312958202418,
+      "grad_norm": 3.143854416647035,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 40211
+    },
+    {
+      "epoch": 0.9571550985432733,
+      "grad_norm": 0.9833103873482338,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 40212
+    },
+    {
+      "epoch": 0.9571789012663049,
+      "grad_norm": 4.676850558583796,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 40213
+    },
+    {
+      "epoch": 0.9572027039893364,
+      "grad_norm": 3.6613448766273518,
+      "learning_rate": 2e-05,
+      "loss": 0.6548,
+      "step": 40214
+    },
+    {
+      "epoch": 0.9572265067123679,
+      "grad_norm": 1.0549809172120577,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 40215
+    },
+    {
+      "epoch": 0.9572503094353995,
+      "grad_norm": 2.644280837736686,
+      "learning_rate": 2e-05,
+      "loss": 0.4348,
+      "step": 40216
+    },
+    {
+      "epoch": 0.957274112158431,
+      "grad_norm": 3.261953744635958,
+      "learning_rate": 2e-05,
+      "loss": 0.0709,
+      "step": 40217
+    },
+    {
+      "epoch": 0.9572979148814624,
+      "grad_norm": 8.638498550226132,
+      "learning_rate": 2e-05,
+      "loss": 1.0498,
+      "step": 40218
+    },
+    {
+      "epoch": 0.9573217176044939,
+      "grad_norm": 3.1239071903840365,
+      "learning_rate": 2e-05,
+      "loss": 0.6565,
+      "step": 40219
+    },
+    {
+      "epoch": 0.9573455203275255,
+      "grad_norm": 0.31664165303235925,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40220
+    },
+    {
+      "epoch": 0.957369323050557,
+      "grad_norm": 9.759979862520611,
+      "learning_rate": 2e-05,
+      "loss": 0.2749,
+      "step": 40221
+    },
+    {
+      "epoch": 0.9573931257735885,
+      "grad_norm": 0.07332500539229729,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40222
+    },
+    {
+      "epoch": 0.95741692849662,
+      "grad_norm": 0.20906201000344476,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 40223
+    },
+    {
+      "epoch": 0.9574407312196516,
+      "grad_norm": 0.10100339413455718,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40224
+    },
+    {
+      "epoch": 0.957464533942683,
+      "grad_norm": 6.764876525990579,
+      "learning_rate": 2e-05,
+      "loss": 0.0947,
+      "step": 40225
+    },
+    {
+      "epoch": 0.9574883366657145,
+      "grad_norm": 7.302322961906086,
+      "learning_rate": 2e-05,
+      "loss": 0.2747,
+      "step": 40226
+    },
+    {
+      "epoch": 0.957512139388746,
+      "grad_norm": 2.2089849610941323,
+      "learning_rate": 2e-05,
+      "loss": 0.2048,
+      "step": 40227
+    },
+    {
+      "epoch": 0.9575359421117776,
+      "grad_norm": 3.8721200084962786,
+      "learning_rate": 2e-05,
+      "loss": 0.769,
+      "step": 40228
+    },
+    {
+      "epoch": 0.9575597448348091,
+      "grad_norm": 6.095407367324788,
+      "learning_rate": 2e-05,
+      "loss": 0.4208,
+      "step": 40229
+    },
+    {
+      "epoch": 0.9575835475578406,
+      "grad_norm": 5.363328925366745,
+      "learning_rate": 2e-05,
+      "loss": 0.1869,
+      "step": 40230
+    },
+    {
+      "epoch": 0.9576073502808722,
+      "grad_norm": 0.15478406646586326,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 40231
+    },
+    {
+      "epoch": 0.9576311530039037,
+      "grad_norm": 4.617172767990175,
+      "learning_rate": 2e-05,
+      "loss": 0.359,
+      "step": 40232
+    },
+    {
+      "epoch": 0.9576549557269352,
+      "grad_norm": 6.884244672656654,
+      "learning_rate": 2e-05,
+      "loss": 0.6906,
+      "step": 40233
+    },
+    {
+      "epoch": 0.9576787584499666,
+      "grad_norm": 5.3593496613620015,
+      "learning_rate": 2e-05,
+      "loss": 0.8706,
+      "step": 40234
+    },
+    {
+      "epoch": 0.9577025611729982,
+      "grad_norm": 4.233182527774918,
+      "learning_rate": 2e-05,
+      "loss": 0.5396,
+      "step": 40235
+    },
+    {
+      "epoch": 0.9577263638960297,
+      "grad_norm": 3.3970609619098875,
+      "learning_rate": 2e-05,
+      "loss": 0.0518,
+      "step": 40236
+    },
+    {
+      "epoch": 0.9577501666190612,
+      "grad_norm": 5.189389300023483,
+      "learning_rate": 2e-05,
+      "loss": 0.6069,
+      "step": 40237
+    },
+    {
+      "epoch": 0.9577739693420927,
+      "grad_norm": 2.6399756705339654,
+      "learning_rate": 2e-05,
+      "loss": 0.218,
+      "step": 40238
+    },
+    {
+      "epoch": 0.9577977720651243,
+      "grad_norm": 1.0535268153622095,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 40239
+    },
+    {
+      "epoch": 0.9578215747881558,
+      "grad_norm": 5.055685170680962,
+      "learning_rate": 2e-05,
+      "loss": 1.0869,
+      "step": 40240
+    },
+    {
+      "epoch": 0.9578453775111873,
+      "grad_norm": 0.12567043701995748,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40241
+    },
+    {
+      "epoch": 0.9578691802342187,
+      "grad_norm": 4.284802389164559,
+      "learning_rate": 2e-05,
+      "loss": 0.0551,
+      "step": 40242
+    },
+    {
+      "epoch": 0.9578929829572503,
+      "grad_norm": 6.641901248047675,
+      "learning_rate": 2e-05,
+      "loss": 0.6183,
+      "step": 40243
+    },
+    {
+      "epoch": 0.9579167856802818,
+      "grad_norm": 3.0114475362959903,
+      "learning_rate": 2e-05,
+      "loss": 0.5508,
+      "step": 40244
+    },
+    {
+      "epoch": 0.9579405884033133,
+      "grad_norm": 4.831507024584868,
+      "learning_rate": 2e-05,
+      "loss": 0.6647,
+      "step": 40245
+    },
+    {
+      "epoch": 0.9579643911263449,
+      "grad_norm": 9.142588252158463,
+      "learning_rate": 2e-05,
+      "loss": 0.435,
+      "step": 40246
+    },
+    {
+      "epoch": 0.9579881938493764,
+      "grad_norm": 0.006902106156328896,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 40247
+    },
+    {
+      "epoch": 0.9580119965724079,
+      "grad_norm": 0.6171863953611259,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 40248
+    },
+    {
+      "epoch": 0.9580357992954394,
+      "grad_norm": 3.81108078891025,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 40249
+    },
+    {
+      "epoch": 0.958059602018471,
+      "grad_norm": 0.06274773983727147,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40250
+    },
+    {
+      "epoch": 0.9580834047415024,
+      "grad_norm": 3.490725622452724,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 40251
+    },
+    {
+      "epoch": 0.9581072074645339,
+      "grad_norm": 0.011793919591100608,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40252
+    },
+    {
+      "epoch": 0.9581310101875654,
+      "grad_norm": 4.772173203547606,
+      "learning_rate": 2e-05,
+      "loss": 0.4587,
+      "step": 40253
+    },
+    {
+      "epoch": 0.958154812910597,
+      "grad_norm": 0.8279787055166004,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 40254
+    },
+    {
+      "epoch": 0.9581786156336285,
+      "grad_norm": 4.869641676747478,
+      "learning_rate": 2e-05,
+      "loss": 0.7987,
+      "step": 40255
+    },
+    {
+      "epoch": 0.95820241835666,
+      "grad_norm": 0.7874405137610612,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 40256
+    },
+    {
+      "epoch": 0.9582262210796915,
+      "grad_norm": 4.293015720324795,
+      "learning_rate": 2e-05,
+      "loss": 0.3713,
+      "step": 40257
+    },
+    {
+      "epoch": 0.9582500238027231,
+      "grad_norm": 7.353664202337291,
+      "learning_rate": 2e-05,
+      "loss": 0.8354,
+      "step": 40258
+    },
+    {
+      "epoch": 0.9582738265257545,
+      "grad_norm": 15.021203274159838,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 40259
+    },
+    {
+      "epoch": 0.958297629248786,
+      "grad_norm": 0.2857743910955529,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 40260
+    },
+    {
+      "epoch": 0.9583214319718176,
+      "grad_norm": 0.48779252571009907,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 40261
+    },
+    {
+      "epoch": 0.9583452346948491,
+      "grad_norm": 1.441132963490848,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 40262
+    },
+    {
+      "epoch": 0.9583690374178806,
+      "grad_norm": 11.235825959314305,
+      "learning_rate": 2e-05,
+      "loss": 1.2244,
+      "step": 40263
+    },
+    {
+      "epoch": 0.9583928401409121,
+      "grad_norm": 1.7212144747960556,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 40264
+    },
+    {
+      "epoch": 0.9584166428639437,
+      "grad_norm": 3.140930276584545,
+      "learning_rate": 2e-05,
+      "loss": 0.5957,
+      "step": 40265
+    },
+    {
+      "epoch": 0.9584404455869752,
+      "grad_norm": 6.446685653397416,
+      "learning_rate": 2e-05,
+      "loss": 0.1403,
+      "step": 40266
+    },
+    {
+      "epoch": 0.9584642483100067,
+      "grad_norm": 4.0119366544261625,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 40267
+    },
+    {
+      "epoch": 0.9584880510330381,
+      "grad_norm": 4.846750461572115,
+      "learning_rate": 2e-05,
+      "loss": 0.9753,
+      "step": 40268
+    },
+    {
+      "epoch": 0.9585118537560697,
+      "grad_norm": 2.482986579546529,
+      "learning_rate": 2e-05,
+      "loss": 0.294,
+      "step": 40269
+    },
+    {
+      "epoch": 0.9585356564791012,
+      "grad_norm": 3.806424849219619,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 40270
+    },
+    {
+      "epoch": 0.9585594592021327,
+      "grad_norm": 4.9838739394262355,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 40271
+    },
+    {
+      "epoch": 0.9585832619251642,
+      "grad_norm": 10.270238839877186,
+      "learning_rate": 2e-05,
+      "loss": 1.5263,
+      "step": 40272
+    },
+    {
+      "epoch": 0.9586070646481958,
+      "grad_norm": 0.14770329028531193,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 40273
+    },
+    {
+      "epoch": 0.9586308673712273,
+      "grad_norm": 0.5245412535884412,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 40274
+    },
+    {
+      "epoch": 0.9586546700942588,
+      "grad_norm": 2.2160850056406427,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 40275
+    },
+    {
+      "epoch": 0.9586784728172904,
+      "grad_norm": 5.2017920180170885,
+      "learning_rate": 2e-05,
+      "loss": 0.7935,
+      "step": 40276
+    },
+    {
+      "epoch": 0.9587022755403218,
+      "grad_norm": 0.8558913654972017,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 40277
+    },
+    {
+      "epoch": 0.9587260782633533,
+      "grad_norm": 3.62959853019774,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 40278
+    },
+    {
+      "epoch": 0.9587498809863848,
+      "grad_norm": 0.009351960657178087,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 40279
+    },
+    {
+      "epoch": 0.9587736837094164,
+      "grad_norm": 3.292738131759747,
+      "learning_rate": 2e-05,
+      "loss": 0.2925,
+      "step": 40280
+    },
+    {
+      "epoch": 0.9587974864324479,
+      "grad_norm": 0.4056833912334779,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 40281
+    },
+    {
+      "epoch": 0.9588212891554794,
+      "grad_norm": 4.294669538394556,
+      "learning_rate": 2e-05,
+      "loss": 1.018,
+      "step": 40282
+    },
+    {
+      "epoch": 0.9588450918785109,
+      "grad_norm": 0.10818142762442513,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 40283
+    },
+    {
+      "epoch": 0.9588688946015425,
+      "grad_norm": 0.09437605894101332,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 40284
+    },
+    {
+      "epoch": 0.9588926973245739,
+      "grad_norm": 0.48634271626398434,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 40285
+    },
+    {
+      "epoch": 0.9589165000476054,
+      "grad_norm": 0.13518076362366777,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40286
+    },
+    {
+      "epoch": 0.9589403027706369,
+      "grad_norm": 2.998988318149395,
+      "learning_rate": 2e-05,
+      "loss": 0.5916,
+      "step": 40287
+    },
+    {
+      "epoch": 0.9589641054936685,
+      "grad_norm": 10.764668954918944,
+      "learning_rate": 2e-05,
+      "loss": 0.6197,
+      "step": 40288
+    },
+    {
+      "epoch": 0.9589879082167,
+      "grad_norm": 1.0629275583298878,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 40289
+    },
+    {
+      "epoch": 0.9590117109397315,
+      "grad_norm": 0.30111637382328477,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 40290
+    },
+    {
+      "epoch": 0.9590355136627631,
+      "grad_norm": 0.95850408788764,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 40291
+    },
+    {
+      "epoch": 0.9590593163857946,
+      "grad_norm": 2.9713792039736777,
+      "learning_rate": 2e-05,
+      "loss": 0.5058,
+      "step": 40292
+    },
+    {
+      "epoch": 0.959083119108826,
+      "grad_norm": 3.3684614191080446,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 40293
+    },
+    {
+      "epoch": 0.9591069218318575,
+      "grad_norm": 0.03374029212129451,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40294
+    },
+    {
+      "epoch": 0.9591307245548891,
+      "grad_norm": 5.581167630767536,
+      "learning_rate": 2e-05,
+      "loss": 0.8492,
+      "step": 40295
+    },
+    {
+      "epoch": 0.9591545272779206,
+      "grad_norm": 12.21289420484328,
+      "learning_rate": 2e-05,
+      "loss": 1.6132,
+      "step": 40296
+    },
+    {
+      "epoch": 0.9591783300009521,
+      "grad_norm": 6.9938051279572635,
+      "learning_rate": 2e-05,
+      "loss": 0.7039,
+      "step": 40297
+    },
+    {
+      "epoch": 0.9592021327239836,
+      "grad_norm": 3.5368201587956425,
+      "learning_rate": 2e-05,
+      "loss": 0.4568,
+      "step": 40298
+    },
+    {
+      "epoch": 0.9592259354470152,
+      "grad_norm": 4.416782946086689,
+      "learning_rate": 2e-05,
+      "loss": 0.9144,
+      "step": 40299
+    },
+    {
+      "epoch": 0.9592497381700467,
+      "grad_norm": 3.859617204955407,
+      "learning_rate": 2e-05,
+      "loss": 0.747,
+      "step": 40300
+    },
+    {
+      "epoch": 0.9592735408930781,
+      "grad_norm": 0.41699903394796145,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40301
+    },
+    {
+      "epoch": 0.9592973436161096,
+      "grad_norm": 11.858726711556585,
+      "learning_rate": 2e-05,
+      "loss": 1.5378,
+      "step": 40302
+    },
+    {
+      "epoch": 0.9593211463391412,
+      "grad_norm": 10.326856739838227,
+      "learning_rate": 2e-05,
+      "loss": 1.1893,
+      "step": 40303
+    },
+    {
+      "epoch": 0.9593449490621727,
+      "grad_norm": 10.712685757324293,
+      "learning_rate": 2e-05,
+      "loss": 0.5531,
+      "step": 40304
+    },
+    {
+      "epoch": 0.9593687517852042,
+      "grad_norm": 0.5737478389082237,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 40305
+    },
+    {
+      "epoch": 0.9593925545082358,
+      "grad_norm": 3.8858321021783153,
+      "learning_rate": 2e-05,
+      "loss": 0.6286,
+      "step": 40306
+    },
+    {
+      "epoch": 0.9594163572312673,
+      "grad_norm": 22.474123238875343,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 40307
+    },
+    {
+      "epoch": 0.9594401599542988,
+      "grad_norm": 0.43379035010035133,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 40308
+    },
+    {
+      "epoch": 0.9594639626773303,
+      "grad_norm": 3.7576574508423444,
+      "learning_rate": 2e-05,
+      "loss": 0.0786,
+      "step": 40309
+    },
+    {
+      "epoch": 0.9594877654003618,
+      "grad_norm": 0.37130376220325545,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 40310
+    },
+    {
+      "epoch": 0.9595115681233933,
+      "grad_norm": 9.566292243050093,
+      "learning_rate": 2e-05,
+      "loss": 1.3111,
+      "step": 40311
+    },
+    {
+      "epoch": 0.9595353708464248,
+      "grad_norm": 0.04402826988158942,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40312
+    },
+    {
+      "epoch": 0.9595591735694563,
+      "grad_norm": 3.2655511963107813,
+      "learning_rate": 2e-05,
+      "loss": 0.6637,
+      "step": 40313
+    },
+    {
+      "epoch": 0.9595829762924879,
+      "grad_norm": 8.04216254765509,
+      "learning_rate": 2e-05,
+      "loss": 0.1,
+      "step": 40314
+    },
+    {
+      "epoch": 0.9596067790155194,
+      "grad_norm": 3.335779023390419,
+      "learning_rate": 2e-05,
+      "loss": 0.4918,
+      "step": 40315
+    },
+    {
+      "epoch": 0.9596305817385509,
+      "grad_norm": 4.367681609461286,
+      "learning_rate": 2e-05,
+      "loss": 0.6963,
+      "step": 40316
+    },
+    {
+      "epoch": 0.9596543844615824,
+      "grad_norm": 0.10722592923944682,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40317
+    },
+    {
+      "epoch": 0.959678187184614,
+      "grad_norm": 13.300258225736375,
+      "learning_rate": 2e-05,
+      "loss": 0.1266,
+      "step": 40318
+    },
+    {
+      "epoch": 0.9597019899076454,
+      "grad_norm": 4.784101907251637,
+      "learning_rate": 2e-05,
+      "loss": 1.4842,
+      "step": 40319
+    },
+    {
+      "epoch": 0.9597257926306769,
+      "grad_norm": 0.10164403112993708,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 40320
+    },
+    {
+      "epoch": 0.9597495953537085,
+      "grad_norm": 6.132851243484964,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 40321
+    },
+    {
+      "epoch": 0.95977339807674,
+      "grad_norm": 5.2913039953700896,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 40322
+    },
+    {
+      "epoch": 0.9597972007997715,
+      "grad_norm": 0.0071861493647995215,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 40323
+    },
+    {
+      "epoch": 0.959821003522803,
+      "grad_norm": 4.669474019627238,
+      "learning_rate": 2e-05,
+      "loss": 0.8163,
+      "step": 40324
+    },
+    {
+      "epoch": 0.9598448062458346,
+      "grad_norm": 4.902443572841984,
+      "learning_rate": 2e-05,
+      "loss": 0.0498,
+      "step": 40325
+    },
+    {
+      "epoch": 0.959868608968866,
+      "grad_norm": 12.829223063052984,
+      "learning_rate": 2e-05,
+      "loss": 0.9452,
+      "step": 40326
+    },
+    {
+      "epoch": 0.9598924116918975,
+      "grad_norm": 2.4861281900469234,
+      "learning_rate": 2e-05,
+      "loss": 0.1255,
+      "step": 40327
+    },
+    {
+      "epoch": 0.959916214414929,
+      "grad_norm": 5.047804154637769,
+      "learning_rate": 2e-05,
+      "loss": 0.855,
+      "step": 40328
+    },
+    {
+      "epoch": 0.9599400171379606,
+      "grad_norm": 9.563236213107531,
+      "learning_rate": 2e-05,
+      "loss": 0.8491,
+      "step": 40329
+    },
+    {
+      "epoch": 0.9599638198609921,
+      "grad_norm": 4.012832333627328,
+      "learning_rate": 2e-05,
+      "loss": 0.7595,
+      "step": 40330
+    },
+    {
+      "epoch": 0.9599876225840236,
+      "grad_norm": 6.650684508075699,
+      "learning_rate": 2e-05,
+      "loss": 0.7509,
+      "step": 40331
+    },
+    {
+      "epoch": 0.9600114253070551,
+      "grad_norm": 0.016314862411159453,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40332
+    },
+    {
+      "epoch": 0.9600352280300867,
+      "grad_norm": 3.3845375440428658,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 40333
+    },
+    {
+      "epoch": 0.9600590307531182,
+      "grad_norm": 2.4087305648605115,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 40334
+    },
+    {
+      "epoch": 0.9600828334761496,
+      "grad_norm": 0.6479771041202445,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 40335
+    },
+    {
+      "epoch": 0.9601066361991812,
+      "grad_norm": 3.420978795200669,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 40336
+    },
+    {
+      "epoch": 0.9601304389222127,
+      "grad_norm": 0.010051025470072585,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 40337
+    },
+    {
+      "epoch": 0.9601542416452442,
+      "grad_norm": 3.0357247494794235,
+      "learning_rate": 2e-05,
+      "loss": 0.1974,
+      "step": 40338
+    },
+    {
+      "epoch": 0.9601780443682757,
+      "grad_norm": 8.463544162392234,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 40339
+    },
+    {
+      "epoch": 0.9602018470913073,
+      "grad_norm": 4.864739006160767,
+      "learning_rate": 2e-05,
+      "loss": 1.1642,
+      "step": 40340
+    },
+    {
+      "epoch": 0.9602256498143388,
+      "grad_norm": 5.799488823027494,
+      "learning_rate": 2e-05,
+      "loss": 0.2526,
+      "step": 40341
+    },
+    {
+      "epoch": 0.9602494525373703,
+      "grad_norm": 0.2425133713417813,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 40342
+    },
+    {
+      "epoch": 0.9602732552604017,
+      "grad_norm": 2.283492027612464,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 40343
+    },
+    {
+      "epoch": 0.9602970579834333,
+      "grad_norm": 5.902633383669402,
+      "learning_rate": 2e-05,
+      "loss": 1.4862,
+      "step": 40344
+    },
+    {
+      "epoch": 0.9603208607064648,
+      "grad_norm": 0.17458450562318142,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 40345
+    },
+    {
+      "epoch": 0.9603446634294963,
+      "grad_norm": 4.992232822510093,
+      "learning_rate": 2e-05,
+      "loss": 0.2023,
+      "step": 40346
+    },
+    {
+      "epoch": 0.9603684661525278,
+      "grad_norm": 4.634276966466889,
+      "learning_rate": 2e-05,
+      "loss": 1.1554,
+      "step": 40347
+    },
+    {
+      "epoch": 0.9603922688755594,
+      "grad_norm": 12.727296213440603,
+      "learning_rate": 2e-05,
+      "loss": 0.3669,
+      "step": 40348
+    },
+    {
+      "epoch": 0.9604160715985909,
+      "grad_norm": 0.1241046536365381,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 40349
+    },
+    {
+      "epoch": 0.9604398743216224,
+      "grad_norm": 2.4796789327581426,
+      "learning_rate": 2e-05,
+      "loss": 0.041,
+      "step": 40350
+    },
+    {
+      "epoch": 0.960463677044654,
+      "grad_norm": 0.15796986249555212,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 40351
+    },
+    {
+      "epoch": 0.9604874797676854,
+      "grad_norm": 3.1671460156515954,
+      "learning_rate": 2e-05,
+      "loss": 0.5857,
+      "step": 40352
+    },
+    {
+      "epoch": 0.9605112824907169,
+      "grad_norm": 0.8156348697143548,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 40353
+    },
+    {
+      "epoch": 0.9605350852137484,
+      "grad_norm": 1.9005133188516823,
+      "learning_rate": 2e-05,
+      "loss": 0.0261,
+      "step": 40354
+    },
+    {
+      "epoch": 0.96055888793678,
+      "grad_norm": 0.7468730846590281,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 40355
+    },
+    {
+      "epoch": 0.9605826906598115,
+      "grad_norm": 6.703730566394162,
+      "learning_rate": 2e-05,
+      "loss": 0.4058,
+      "step": 40356
+    },
+    {
+      "epoch": 0.960606493382843,
+      "grad_norm": 5.407142389098329,
+      "learning_rate": 2e-05,
+      "loss": 0.9283,
+      "step": 40357
+    },
+    {
+      "epoch": 0.9606302961058745,
+      "grad_norm": 4.314490065276481,
+      "learning_rate": 2e-05,
+      "loss": 0.8556,
+      "step": 40358
+    },
+    {
+      "epoch": 0.9606540988289061,
+      "grad_norm": 2.3182394335831598,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 40359
+    },
+    {
+      "epoch": 0.9606779015519376,
+      "grad_norm": 6.8163483039120045,
+      "learning_rate": 2e-05,
+      "loss": 1.4635,
+      "step": 40360
+    },
+    {
+      "epoch": 0.960701704274969,
+      "grad_norm": 6.9831790729401035,
+      "learning_rate": 2e-05,
+      "loss": 0.2652,
+      "step": 40361
+    },
+    {
+      "epoch": 0.9607255069980005,
+      "grad_norm": 0.04114877102201348,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40362
+    },
+    {
+      "epoch": 0.9607493097210321,
+      "grad_norm": 0.6087601834003147,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 40363
+    },
+    {
+      "epoch": 0.9607731124440636,
+      "grad_norm": 5.793437509149154,
+      "learning_rate": 2e-05,
+      "loss": 0.7073,
+      "step": 40364
+    },
+    {
+      "epoch": 0.9607969151670951,
+      "grad_norm": 0.7807493286105112,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 40365
+    },
+    {
+      "epoch": 0.9608207178901267,
+      "grad_norm": 0.011013372380695136,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 40366
+    },
+    {
+      "epoch": 0.9608445206131582,
+      "grad_norm": 5.315418183137252,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 40367
+    },
+    {
+      "epoch": 0.9608683233361897,
+      "grad_norm": 7.927722315508009,
+      "learning_rate": 2e-05,
+      "loss": 0.2401,
+      "step": 40368
+    },
+    {
+      "epoch": 0.9608921260592211,
+      "grad_norm": 0.15953403346988398,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 40369
+    },
+    {
+      "epoch": 0.9609159287822527,
+      "grad_norm": 9.04911501115931,
+      "learning_rate": 2e-05,
+      "loss": 0.7507,
+      "step": 40370
+    },
+    {
+      "epoch": 0.9609397315052842,
+      "grad_norm": 2.983921574948408,
+      "learning_rate": 2e-05,
+      "loss": 0.0915,
+      "step": 40371
+    },
+    {
+      "epoch": 0.9609635342283157,
+      "grad_norm": 3.720334858484325,
+      "learning_rate": 2e-05,
+      "loss": 0.4025,
+      "step": 40372
+    },
+    {
+      "epoch": 0.9609873369513472,
+      "grad_norm": 0.7896345923714843,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 40373
+    },
+    {
+      "epoch": 0.9610111396743788,
+      "grad_norm": 0.02220697581037343,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40374
+    },
+    {
+      "epoch": 0.9610349423974103,
+      "grad_norm": 4.9364167375617605,
+      "learning_rate": 2e-05,
+      "loss": 0.1866,
+      "step": 40375
+    },
+    {
+      "epoch": 0.9610587451204418,
+      "grad_norm": 2.9756887039423883,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 40376
+    },
+    {
+      "epoch": 0.9610825478434732,
+      "grad_norm": 6.369470172170138,
+      "learning_rate": 2e-05,
+      "loss": 0.301,
+      "step": 40377
+    },
+    {
+      "epoch": 0.9611063505665048,
+      "grad_norm": 2.8635566375684354,
+      "learning_rate": 2e-05,
+      "loss": 0.4041,
+      "step": 40378
+    },
+    {
+      "epoch": 0.9611301532895363,
+      "grad_norm": 0.32093825797290987,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 40379
+    },
+    {
+      "epoch": 0.9611539560125678,
+      "grad_norm": 0.5255117884039819,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 40380
+    },
+    {
+      "epoch": 0.9611777587355994,
+      "grad_norm": 3.205006074024903,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 40381
+    },
+    {
+      "epoch": 0.9612015614586309,
+      "grad_norm": 3.7359405499827174,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 40382
+    },
+    {
+      "epoch": 0.9612253641816624,
+      "grad_norm": 3.822128701534243,
+      "learning_rate": 2e-05,
+      "loss": 0.1256,
+      "step": 40383
+    },
+    {
+      "epoch": 0.9612491669046939,
+      "grad_norm": 1.2275245278273277,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 40384
+    },
+    {
+      "epoch": 0.9612729696277255,
+      "grad_norm": 4.800174080649713,
+      "learning_rate": 2e-05,
+      "loss": 0.6904,
+      "step": 40385
+    },
+    {
+      "epoch": 0.961296772350757,
+      "grad_norm": 2.730413367457504,
+      "learning_rate": 2e-05,
+      "loss": 0.1247,
+      "step": 40386
+    },
+    {
+      "epoch": 0.9613205750737884,
+      "grad_norm": 3.017826086361207,
+      "learning_rate": 2e-05,
+      "loss": 0.2302,
+      "step": 40387
+    },
+    {
+      "epoch": 0.9613443777968199,
+      "grad_norm": 2.3592565255478113,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 40388
+    },
+    {
+      "epoch": 0.9613681805198515,
+      "grad_norm": 5.4004164311898135,
+      "learning_rate": 2e-05,
+      "loss": 0.2034,
+      "step": 40389
+    },
+    {
+      "epoch": 0.961391983242883,
+      "grad_norm": 5.239001128916497,
+      "learning_rate": 2e-05,
+      "loss": 0.1967,
+      "step": 40390
+    },
+    {
+      "epoch": 0.9614157859659145,
+      "grad_norm": 7.871375159904081,
+      "learning_rate": 2e-05,
+      "loss": 0.3752,
+      "step": 40391
+    },
+    {
+      "epoch": 0.961439588688946,
+      "grad_norm": 5.79389301825716,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 40392
+    },
+    {
+      "epoch": 0.9614633914119776,
+      "grad_norm": 2.6726795818971594,
+      "learning_rate": 2e-05,
+      "loss": 0.0705,
+      "step": 40393
+    },
+    {
+      "epoch": 0.961487194135009,
+      "grad_norm": 3.672263366916846,
+      "learning_rate": 2e-05,
+      "loss": 0.0988,
+      "step": 40394
+    },
+    {
+      "epoch": 0.9615109968580405,
+      "grad_norm": 0.5713852170431711,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 40395
+    },
+    {
+      "epoch": 0.961534799581072,
+      "grad_norm": 1.746205096270346,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 40396
+    },
+    {
+      "epoch": 0.9615586023041036,
+      "grad_norm": 2.860693035059001,
+      "learning_rate": 2e-05,
+      "loss": 0.3497,
+      "step": 40397
+    },
+    {
+      "epoch": 0.9615824050271351,
+      "grad_norm": 0.1746620950972229,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 40398
+    },
+    {
+      "epoch": 0.9616062077501666,
+      "grad_norm": 0.7128194973989432,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 40399
+    },
+    {
+      "epoch": 0.9616300104731982,
+      "grad_norm": 3.8031600028435513,
+      "learning_rate": 2e-05,
+      "loss": 0.4666,
+      "step": 40400
+    },
+    {
+      "epoch": 0.9616538131962297,
+      "grad_norm": 4.498629577171368,
+      "learning_rate": 2e-05,
+      "loss": 0.4154,
+      "step": 40401
+    },
+    {
+      "epoch": 0.9616776159192612,
+      "grad_norm": 4.315248569875782,
+      "learning_rate": 2e-05,
+      "loss": 0.4836,
+      "step": 40402
+    },
+    {
+      "epoch": 0.9617014186422926,
+      "grad_norm": 16.934716481477945,
+      "learning_rate": 2e-05,
+      "loss": 0.8504,
+      "step": 40403
+    },
+    {
+      "epoch": 0.9617252213653242,
+      "grad_norm": 3.511017121374071,
+      "learning_rate": 2e-05,
+      "loss": 0.2124,
+      "step": 40404
+    },
+    {
+      "epoch": 0.9617490240883557,
+      "grad_norm": 1.0049224286601826,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 40405
+    },
+    {
+      "epoch": 0.9617728268113872,
+      "grad_norm": 2.332986758087305,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 40406
+    },
+    {
+      "epoch": 0.9617966295344187,
+      "grad_norm": 5.470866005679271,
+      "learning_rate": 2e-05,
+      "loss": 1.036,
+      "step": 40407
+    },
+    {
+      "epoch": 0.9618204322574503,
+      "grad_norm": 6.11933577590114,
+      "learning_rate": 2e-05,
+      "loss": 1.9368,
+      "step": 40408
+    },
+    {
+      "epoch": 0.9618442349804818,
+      "grad_norm": 11.904903263737918,
+      "learning_rate": 2e-05,
+      "loss": 1.113,
+      "step": 40409
+    },
+    {
+      "epoch": 0.9618680377035133,
+      "grad_norm": 3.2982123771314966,
+      "learning_rate": 2e-05,
+      "loss": 0.5768,
+      "step": 40410
+    },
+    {
+      "epoch": 0.9618918404265447,
+      "grad_norm": 3.525092746841193,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 40411
+    },
+    {
+      "epoch": 0.9619156431495763,
+      "grad_norm": 5.05907191469281,
+      "learning_rate": 2e-05,
+      "loss": 0.7293,
+      "step": 40412
+    },
+    {
+      "epoch": 0.9619394458726078,
+      "grad_norm": 2.4842449637511836,
+      "learning_rate": 2e-05,
+      "loss": 0.2215,
+      "step": 40413
+    },
+    {
+      "epoch": 0.9619632485956393,
+      "grad_norm": 4.815144351404576,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 40414
+    },
+    {
+      "epoch": 0.9619870513186709,
+      "grad_norm": 1.068816526052139,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 40415
+    },
+    {
+      "epoch": 0.9620108540417024,
+      "grad_norm": 4.498772246605026,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 40416
+    },
+    {
+      "epoch": 0.9620346567647339,
+      "grad_norm": 1.9519358130810323,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 40417
+    },
+    {
+      "epoch": 0.9620584594877654,
+      "grad_norm": 3.4054982507861804,
+      "learning_rate": 2e-05,
+      "loss": 0.6047,
+      "step": 40418
+    },
+    {
+      "epoch": 0.962082262210797,
+      "grad_norm": 2.486538216422247,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 40419
+    },
+    {
+      "epoch": 0.9621060649338284,
+      "grad_norm": 1.2100909714611008,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 40420
+    },
+    {
+      "epoch": 0.9621298676568599,
+      "grad_norm": 3.771094419305128,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 40421
+    },
+    {
+      "epoch": 0.9621536703798914,
+      "grad_norm": 0.42562596667326646,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 40422
+    },
+    {
+      "epoch": 0.962177473102923,
+      "grad_norm": 0.18977534708380686,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 40423
+    },
+    {
+      "epoch": 0.9622012758259545,
+      "grad_norm": 4.429787966955745,
+      "learning_rate": 2e-05,
+      "loss": 0.2797,
+      "step": 40424
+    },
+    {
+      "epoch": 0.962225078548986,
+      "grad_norm": 3.854528745508154,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 40425
+    },
+    {
+      "epoch": 0.9622488812720175,
+      "grad_norm": 2.38109302341924,
+      "learning_rate": 2e-05,
+      "loss": 0.029,
+      "step": 40426
+    },
+    {
+      "epoch": 0.9622726839950491,
+      "grad_norm": 0.20988141409345118,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 40427
+    },
+    {
+      "epoch": 0.9622964867180805,
+      "grad_norm": 1.5963529369663103,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 40428
+    },
+    {
+      "epoch": 0.962320289441112,
+      "grad_norm": 2.779714356567942,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 40429
+    },
+    {
+      "epoch": 0.9623440921641436,
+      "grad_norm": 0.17405717356745376,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 40430
+    },
+    {
+      "epoch": 0.9623678948871751,
+      "grad_norm": 2.9542347950716676,
+      "learning_rate": 2e-05,
+      "loss": 0.5139,
+      "step": 40431
+    },
+    {
+      "epoch": 0.9623916976102066,
+      "grad_norm": 2.788628722724616,
+      "learning_rate": 2e-05,
+      "loss": 0.3818,
+      "step": 40432
+    },
+    {
+      "epoch": 0.9624155003332381,
+      "grad_norm": 3.54170511342483,
+      "learning_rate": 2e-05,
+      "loss": 0.0694,
+      "step": 40433
+    },
+    {
+      "epoch": 0.9624393030562697,
+      "grad_norm": 0.06639764645236895,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 40434
+    },
+    {
+      "epoch": 0.9624631057793012,
+      "grad_norm": 0.9122857115343876,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 40435
+    },
+    {
+      "epoch": 0.9624869085023326,
+      "grad_norm": 0.8346132345337703,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 40436
+    },
+    {
+      "epoch": 0.9625107112253641,
+      "grad_norm": 5.366800214647987,
+      "learning_rate": 2e-05,
+      "loss": 1.0533,
+      "step": 40437
+    },
+    {
+      "epoch": 0.9625345139483957,
+      "grad_norm": 10.760535630048272,
+      "learning_rate": 2e-05,
+      "loss": 0.9865,
+      "step": 40438
+    },
+    {
+      "epoch": 0.9625583166714272,
+      "grad_norm": 2.5209592257766347,
+      "learning_rate": 2e-05,
+      "loss": 0.4121,
+      "step": 40439
+    },
+    {
+      "epoch": 0.9625821193944587,
+      "grad_norm": 0.36664271890031025,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 40440
+    },
+    {
+      "epoch": 0.9626059221174902,
+      "grad_norm": 8.02927651878283,
+      "learning_rate": 2e-05,
+      "loss": 1.2868,
+      "step": 40441
+    },
+    {
+      "epoch": 0.9626297248405218,
+      "grad_norm": 0.042331230503554716,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40442
+    },
+    {
+      "epoch": 0.9626535275635533,
+      "grad_norm": 3.023580602713549,
+      "learning_rate": 2e-05,
+      "loss": 0.4993,
+      "step": 40443
+    },
+    {
+      "epoch": 0.9626773302865848,
+      "grad_norm": 10.518970303385972,
+      "learning_rate": 2e-05,
+      "loss": 1.1504,
+      "step": 40444
+    },
+    {
+      "epoch": 0.9627011330096163,
+      "grad_norm": 0.2452521554389615,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 40445
+    },
+    {
+      "epoch": 0.9627249357326478,
+      "grad_norm": 3.469595616003468,
+      "learning_rate": 2e-05,
+      "loss": 0.8979,
+      "step": 40446
+    },
+    {
+      "epoch": 0.9627487384556793,
+      "grad_norm": 9.296260838129248,
+      "learning_rate": 2e-05,
+      "loss": 0.2463,
+      "step": 40447
+    },
+    {
+      "epoch": 0.9627725411787108,
+      "grad_norm": 4.669770644635417,
+      "learning_rate": 2e-05,
+      "loss": 0.5497,
+      "step": 40448
+    },
+    {
+      "epoch": 0.9627963439017424,
+      "grad_norm": 0.6694363943722961,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 40449
+    },
+    {
+      "epoch": 0.9628201466247739,
+      "grad_norm": 3.9550069370736423,
+      "learning_rate": 2e-05,
+      "loss": 0.0372,
+      "step": 40450
+    },
+    {
+      "epoch": 0.9628439493478054,
+      "grad_norm": 1.3682893828353093,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 40451
+    },
+    {
+      "epoch": 0.9628677520708369,
+      "grad_norm": 3.4252150533411028,
+      "learning_rate": 2e-05,
+      "loss": 0.2146,
+      "step": 40452
+    },
+    {
+      "epoch": 0.9628915547938685,
+      "grad_norm": 2.5248975708759365,
+      "learning_rate": 2e-05,
+      "loss": 0.1552,
+      "step": 40453
+    },
+    {
+      "epoch": 0.9629153575168999,
+      "grad_norm": 16.819417286569404,
+      "learning_rate": 2e-05,
+      "loss": 0.9226,
+      "step": 40454
+    },
+    {
+      "epoch": 0.9629391602399314,
+      "grad_norm": 0.9826417944493566,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 40455
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 0.04439870969055972,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40456
+    },
+    {
+      "epoch": 0.9629867656859945,
+      "grad_norm": 0.04681218781030826,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40457
+    },
+    {
+      "epoch": 0.963010568409026,
+      "grad_norm": 0.6423358330222075,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 40458
+    },
+    {
+      "epoch": 0.9630343711320575,
+      "grad_norm": 0.41939730594227953,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 40459
+    },
+    {
+      "epoch": 0.9630581738550891,
+      "grad_norm": 2.011942773096288,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 40460
+    },
+    {
+      "epoch": 0.9630819765781206,
+      "grad_norm": 10.749916006670434,
+      "learning_rate": 2e-05,
+      "loss": 0.9723,
+      "step": 40461
+    },
+    {
+      "epoch": 0.963105779301152,
+      "grad_norm": 0.16603407429376688,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 40462
+    },
+    {
+      "epoch": 0.9631295820241835,
+      "grad_norm": 0.022253774952711733,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40463
+    },
+    {
+      "epoch": 0.9631533847472151,
+      "grad_norm": 5.032743152071339,
+      "learning_rate": 2e-05,
+      "loss": 0.9889,
+      "step": 40464
+    },
+    {
+      "epoch": 0.9631771874702466,
+      "grad_norm": 2.487228942697464,
+      "learning_rate": 2e-05,
+      "loss": 0.4319,
+      "step": 40465
+    },
+    {
+      "epoch": 0.9632009901932781,
+      "grad_norm": 3.9433647051273333,
+      "learning_rate": 2e-05,
+      "loss": 0.8388,
+      "step": 40466
+    },
+    {
+      "epoch": 0.9632247929163096,
+      "grad_norm": 2.5309996104757824,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 40467
+    },
+    {
+      "epoch": 0.9632485956393412,
+      "grad_norm": 1.0973274496332344,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 40468
+    },
+    {
+      "epoch": 0.9632723983623727,
+      "grad_norm": 4.188383852843145,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 40469
+    },
+    {
+      "epoch": 0.9632962010854041,
+      "grad_norm": 1.2951794527405711,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 40470
+    },
+    {
+      "epoch": 0.9633200038084356,
+      "grad_norm": 9.652770645416142,
+      "learning_rate": 2e-05,
+      "loss": 0.1359,
+      "step": 40471
+    },
+    {
+      "epoch": 0.9633438065314672,
+      "grad_norm": 7.007474991587474,
+      "learning_rate": 2e-05,
+      "loss": 0.7845,
+      "step": 40472
+    },
+    {
+      "epoch": 0.9633676092544987,
+      "grad_norm": 0.21763944433520485,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 40473
+    },
+    {
+      "epoch": 0.9633914119775302,
+      "grad_norm": 5.420503101537084,
+      "learning_rate": 2e-05,
+      "loss": 0.7692,
+      "step": 40474
+    },
+    {
+      "epoch": 0.9634152147005618,
+      "grad_norm": 20.002609227255345,
+      "learning_rate": 2e-05,
+      "loss": 0.3775,
+      "step": 40475
+    },
+    {
+      "epoch": 0.9634390174235933,
+      "grad_norm": 6.96736039213991,
+      "learning_rate": 2e-05,
+      "loss": 0.0875,
+      "step": 40476
+    },
+    {
+      "epoch": 0.9634628201466248,
+      "grad_norm": 1.6477440780435475,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 40477
+    },
+    {
+      "epoch": 0.9634866228696563,
+      "grad_norm": 3.8835258007804776,
+      "learning_rate": 2e-05,
+      "loss": 0.1701,
+      "step": 40478
+    },
+    {
+      "epoch": 0.9635104255926878,
+      "grad_norm": 4.777821904233687,
+      "learning_rate": 2e-05,
+      "loss": 1.1853,
+      "step": 40479
+    },
+    {
+      "epoch": 0.9635342283157193,
+      "grad_norm": 1.9293717695042274,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 40480
+    },
+    {
+      "epoch": 0.9635580310387508,
+      "grad_norm": 2.8477428851319853,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 40481
+    },
+    {
+      "epoch": 0.9635818337617823,
+      "grad_norm": 1.8320279776364943,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 40482
+    },
+    {
+      "epoch": 0.9636056364848139,
+      "grad_norm": 3.315279243504734,
+      "learning_rate": 2e-05,
+      "loss": 0.5897,
+      "step": 40483
+    },
+    {
+      "epoch": 0.9636294392078454,
+      "grad_norm": 1.5043610838448604,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 40484
+    },
+    {
+      "epoch": 0.9636532419308769,
+      "grad_norm": 2.9557474334545675,
+      "learning_rate": 2e-05,
+      "loss": 0.415,
+      "step": 40485
+    },
+    {
+      "epoch": 0.9636770446539084,
+      "grad_norm": 0.06533394812864211,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40486
+    },
+    {
+      "epoch": 0.96370084737694,
+      "grad_norm": 0.11991052115072681,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 40487
+    },
+    {
+      "epoch": 0.9637246500999714,
+      "grad_norm": 3.7701670747187386,
+      "learning_rate": 2e-05,
+      "loss": 0.3928,
+      "step": 40488
+    },
+    {
+      "epoch": 0.9637484528230029,
+      "grad_norm": 0.04478316410983291,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 40489
+    },
+    {
+      "epoch": 0.9637722555460345,
+      "grad_norm": 3.542941301620529,
+      "learning_rate": 2e-05,
+      "loss": 0.6115,
+      "step": 40490
+    },
+    {
+      "epoch": 0.963796058269066,
+      "grad_norm": 6.199908681567813,
+      "learning_rate": 2e-05,
+      "loss": 1.3425,
+      "step": 40491
+    },
+    {
+      "epoch": 0.9638198609920975,
+      "grad_norm": 5.8398511985131885,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 40492
+    },
+    {
+      "epoch": 0.963843663715129,
+      "grad_norm": 0.5981028624373865,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 40493
+    },
+    {
+      "epoch": 0.9638674664381606,
+      "grad_norm": 4.070815815560372,
+      "learning_rate": 2e-05,
+      "loss": 0.5754,
+      "step": 40494
+    },
+    {
+      "epoch": 0.963891269161192,
+      "grad_norm": 1.5502991279234224,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 40495
+    },
+    {
+      "epoch": 0.9639150718842235,
+      "grad_norm": 3.5873105330917303,
+      "learning_rate": 2e-05,
+      "loss": 0.56,
+      "step": 40496
+    },
+    {
+      "epoch": 0.963938874607255,
+      "grad_norm": 0.022627539063387655,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40497
+    },
+    {
+      "epoch": 0.9639626773302866,
+      "grad_norm": 5.996663700560486,
+      "learning_rate": 2e-05,
+      "loss": 1.5106,
+      "step": 40498
+    },
+    {
+      "epoch": 0.9639864800533181,
+      "grad_norm": 4.07788320148418,
+      "learning_rate": 2e-05,
+      "loss": 0.4733,
+      "step": 40499
+    },
+    {
+      "epoch": 0.9640102827763496,
+      "grad_norm": 4.426001331246136,
+      "learning_rate": 2e-05,
+      "loss": 0.5549,
+      "step": 40500
+    },
+    {
+      "epoch": 0.9640340854993811,
+      "grad_norm": 0.08826112170074878,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40501
+    },
+    {
+      "epoch": 0.9640578882224127,
+      "grad_norm": 6.669493718334394,
+      "learning_rate": 2e-05,
+      "loss": 0.79,
+      "step": 40502
+    },
+    {
+      "epoch": 0.9640816909454442,
+      "grad_norm": 11.798134466541766,
+      "learning_rate": 2e-05,
+      "loss": 1.5809,
+      "step": 40503
+    },
+    {
+      "epoch": 0.9641054936684756,
+      "grad_norm": 0.029873212228366256,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40504
+    },
+    {
+      "epoch": 0.9641292963915072,
+      "grad_norm": 0.029417898084978675,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40505
+    },
+    {
+      "epoch": 0.9641530991145387,
+      "grad_norm": 0.043237850253295185,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40506
+    },
+    {
+      "epoch": 0.9641769018375702,
+      "grad_norm": 3.7621256504149416,
+      "learning_rate": 2e-05,
+      "loss": 0.6441,
+      "step": 40507
+    },
+    {
+      "epoch": 0.9642007045606017,
+      "grad_norm": 1.6278611444779918,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 40508
+    },
+    {
+      "epoch": 0.9642245072836333,
+      "grad_norm": 0.7753043195827466,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 40509
+    },
+    {
+      "epoch": 0.9642483100066648,
+      "grad_norm": 0.7750682379616068,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 40510
+    },
+    {
+      "epoch": 0.9642721127296963,
+      "grad_norm": 6.89242720688689,
+      "learning_rate": 2e-05,
+      "loss": 0.1637,
+      "step": 40511
+    },
+    {
+      "epoch": 0.9642959154527277,
+      "grad_norm": 6.669498223578104,
+      "learning_rate": 2e-05,
+      "loss": 1.2432,
+      "step": 40512
+    },
+    {
+      "epoch": 0.9643197181757593,
+      "grad_norm": 3.936525670539855,
+      "learning_rate": 2e-05,
+      "loss": 1.0259,
+      "step": 40513
+    },
+    {
+      "epoch": 0.9643435208987908,
+      "grad_norm": 0.0649911078467399,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40514
+    },
+    {
+      "epoch": 0.9643673236218223,
+      "grad_norm": 0.11870189067683909,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40515
+    },
+    {
+      "epoch": 0.9643911263448538,
+      "grad_norm": 3.2003280295320446,
+      "learning_rate": 2e-05,
+      "loss": 0.0798,
+      "step": 40516
+    },
+    {
+      "epoch": 0.9644149290678854,
+      "grad_norm": 11.230326468921538,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 40517
+    },
+    {
+      "epoch": 0.9644387317909169,
+      "grad_norm": 1.2773131046262491,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 40518
+    },
+    {
+      "epoch": 0.9644625345139484,
+      "grad_norm": 1.9714361860511163,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 40519
+    },
+    {
+      "epoch": 0.96448633723698,
+      "grad_norm": 4.07483139786181,
+      "learning_rate": 2e-05,
+      "loss": 0.6786,
+      "step": 40520
+    },
+    {
+      "epoch": 0.9645101399600114,
+      "grad_norm": 0.28026505931529866,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 40521
+    },
+    {
+      "epoch": 0.9645339426830429,
+      "grad_norm": 3.648593610560668,
+      "learning_rate": 2e-05,
+      "loss": 0.7818,
+      "step": 40522
+    },
+    {
+      "epoch": 0.9645577454060744,
+      "grad_norm": 6.587844136352926,
+      "learning_rate": 2e-05,
+      "loss": 1.446,
+      "step": 40523
+    },
+    {
+      "epoch": 0.964581548129106,
+      "grad_norm": 7.565689484210238,
+      "learning_rate": 2e-05,
+      "loss": 0.6465,
+      "step": 40524
+    },
+    {
+      "epoch": 0.9646053508521375,
+      "grad_norm": 2.387912288547325,
+      "learning_rate": 2e-05,
+      "loss": 0.43,
+      "step": 40525
+    },
+    {
+      "epoch": 0.964629153575169,
+      "grad_norm": 0.9289627917613978,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 40526
+    },
+    {
+      "epoch": 0.9646529562982005,
+      "grad_norm": 3.679042148105428,
+      "learning_rate": 2e-05,
+      "loss": 0.748,
+      "step": 40527
+    },
+    {
+      "epoch": 0.9646767590212321,
+      "grad_norm": 9.021484529767154,
+      "learning_rate": 2e-05,
+      "loss": 0.0707,
+      "step": 40528
+    },
+    {
+      "epoch": 0.9647005617442636,
+      "grad_norm": 8.165943550986261,
+      "learning_rate": 2e-05,
+      "loss": 1.0916,
+      "step": 40529
+    },
+    {
+      "epoch": 0.964724364467295,
+      "grad_norm": 0.18647134694843323,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 40530
+    },
+    {
+      "epoch": 0.9647481671903265,
+      "grad_norm": 24.649107764606278,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 40531
+    },
+    {
+      "epoch": 0.9647719699133581,
+      "grad_norm": 3.4539341377379595,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 40532
+    },
+    {
+      "epoch": 0.9647957726363896,
+      "grad_norm": 4.679001653325276,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 40533
+    },
+    {
+      "epoch": 0.9648195753594211,
+      "grad_norm": 5.89522730488311,
+      "learning_rate": 2e-05,
+      "loss": 1.0042,
+      "step": 40534
+    },
+    {
+      "epoch": 0.9648433780824527,
+      "grad_norm": 0.037763536700030985,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40535
+    },
+    {
+      "epoch": 0.9648671808054842,
+      "grad_norm": 0.002204022325676786,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 40536
+    },
+    {
+      "epoch": 0.9648909835285157,
+      "grad_norm": 0.1365185806544339,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 40537
+    },
+    {
+      "epoch": 0.9649147862515471,
+      "grad_norm": 2.839994047782756,
+      "learning_rate": 2e-05,
+      "loss": 0.5982,
+      "step": 40538
+    },
+    {
+      "epoch": 0.9649385889745787,
+      "grad_norm": 6.424537217483306,
+      "learning_rate": 2e-05,
+      "loss": 0.1616,
+      "step": 40539
+    },
+    {
+      "epoch": 0.9649623916976102,
+      "grad_norm": 3.2621162733344957,
+      "learning_rate": 2e-05,
+      "loss": 0.5621,
+      "step": 40540
+    },
+    {
+      "epoch": 0.9649861944206417,
+      "grad_norm": 0.3504710096515393,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 40541
+    },
+    {
+      "epoch": 0.9650099971436732,
+      "grad_norm": 5.309443270849376,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 40542
+    },
+    {
+      "epoch": 0.9650337998667048,
+      "grad_norm": 6.588115856150449,
+      "learning_rate": 2e-05,
+      "loss": 0.2772,
+      "step": 40543
+    },
+    {
+      "epoch": 0.9650576025897363,
+      "grad_norm": 4.366055159938278,
+      "learning_rate": 2e-05,
+      "loss": 0.8654,
+      "step": 40544
+    },
+    {
+      "epoch": 0.9650814053127678,
+      "grad_norm": 6.366495210888355,
+      "learning_rate": 2e-05,
+      "loss": 0.0565,
+      "step": 40545
+    },
+    {
+      "epoch": 0.9651052080357992,
+      "grad_norm": 11.888019670569488,
+      "learning_rate": 2e-05,
+      "loss": 0.81,
+      "step": 40546
+    },
+    {
+      "epoch": 0.9651290107588308,
+      "grad_norm": 5.042874997516361,
+      "learning_rate": 2e-05,
+      "loss": 0.951,
+      "step": 40547
+    },
+    {
+      "epoch": 0.9651528134818623,
+      "grad_norm": 1.3835564287616076,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 40548
+    },
+    {
+      "epoch": 0.9651766162048938,
+      "grad_norm": 4.604751338129195,
+      "learning_rate": 2e-05,
+      "loss": 0.8624,
+      "step": 40549
+    },
+    {
+      "epoch": 0.9652004189279254,
+      "grad_norm": 17.685466390102174,
+      "learning_rate": 2e-05,
+      "loss": 0.1771,
+      "step": 40550
+    },
+    {
+      "epoch": 0.9652242216509569,
+      "grad_norm": 4.807982278662922,
+      "learning_rate": 2e-05,
+      "loss": 0.5497,
+      "step": 40551
+    },
+    {
+      "epoch": 0.9652480243739884,
+      "grad_norm": 0.44797176302905173,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 40552
+    },
+    {
+      "epoch": 0.9652718270970199,
+      "grad_norm": 2.7240362287041244,
+      "learning_rate": 2e-05,
+      "loss": 0.3739,
+      "step": 40553
+    },
+    {
+      "epoch": 0.9652956298200515,
+      "grad_norm": 5.1796180569490025,
+      "learning_rate": 2e-05,
+      "loss": 0.8405,
+      "step": 40554
+    },
+    {
+      "epoch": 0.965319432543083,
+      "grad_norm": 3.2319150311744513,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 40555
+    },
+    {
+      "epoch": 0.9653432352661144,
+      "grad_norm": 7.461042861950744,
+      "learning_rate": 2e-05,
+      "loss": 1.0759,
+      "step": 40556
+    },
+    {
+      "epoch": 0.9653670379891459,
+      "grad_norm": 5.278408459438122,
+      "learning_rate": 2e-05,
+      "loss": 1.6549,
+      "step": 40557
+    },
+    {
+      "epoch": 0.9653908407121775,
+      "grad_norm": 4.212812094174636,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 40558
+    },
+    {
+      "epoch": 0.965414643435209,
+      "grad_norm": 3.399701473986167,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 40559
+    },
+    {
+      "epoch": 0.9654384461582405,
+      "grad_norm": 1.5606611308690976,
+      "learning_rate": 2e-05,
+      "loss": 0.0553,
+      "step": 40560
+    },
+    {
+      "epoch": 0.965462248881272,
+      "grad_norm": 6.337732490960237,
+      "learning_rate": 2e-05,
+      "loss": 1.2693,
+      "step": 40561
+    },
+    {
+      "epoch": 0.9654860516043036,
+      "grad_norm": 0.2062170861128202,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 40562
+    },
+    {
+      "epoch": 0.965509854327335,
+      "grad_norm": 2.30368200743129,
+      "learning_rate": 2e-05,
+      "loss": 0.1857,
+      "step": 40563
+    },
+    {
+      "epoch": 0.9655336570503665,
+      "grad_norm": 3.0618227563092013,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 40564
+    },
+    {
+      "epoch": 0.9655574597733981,
+      "grad_norm": 0.742580380647283,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 40565
+    },
+    {
+      "epoch": 0.9655812624964296,
+      "grad_norm": 8.796050060252636,
+      "learning_rate": 2e-05,
+      "loss": 0.7209,
+      "step": 40566
+    },
+    {
+      "epoch": 0.9656050652194611,
+      "grad_norm": 2.3562312328750843,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 40567
+    },
+    {
+      "epoch": 0.9656288679424926,
+      "grad_norm": 0.21448538436928752,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 40568
+    },
+    {
+      "epoch": 0.9656526706655242,
+      "grad_norm": 3.982267181882445,
+      "learning_rate": 2e-05,
+      "loss": 0.5162,
+      "step": 40569
+    },
+    {
+      "epoch": 0.9656764733885557,
+      "grad_norm": 5.801954552790475,
+      "learning_rate": 2e-05,
+      "loss": 0.6454,
+      "step": 40570
+    },
+    {
+      "epoch": 0.9657002761115872,
+      "grad_norm": 2.611212672727823,
+      "learning_rate": 2e-05,
+      "loss": 0.5156,
+      "step": 40571
+    },
+    {
+      "epoch": 0.9657240788346186,
+      "grad_norm": 9.000035180727206,
+      "learning_rate": 2e-05,
+      "loss": 1.1217,
+      "step": 40572
+    },
+    {
+      "epoch": 0.9657478815576502,
+      "grad_norm": 4.8931770270351285,
+      "learning_rate": 2e-05,
+      "loss": 0.695,
+      "step": 40573
+    },
+    {
+      "epoch": 0.9657716842806817,
+      "grad_norm": 3.1906707309535633,
+      "learning_rate": 2e-05,
+      "loss": 0.6743,
+      "step": 40574
+    },
+    {
+      "epoch": 0.9657954870037132,
+      "grad_norm": 0.2720022774803053,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40575
+    },
+    {
+      "epoch": 0.9658192897267447,
+      "grad_norm": 5.344315983782517,
+      "learning_rate": 2e-05,
+      "loss": 0.1013,
+      "step": 40576
+    },
+    {
+      "epoch": 0.9658430924497763,
+      "grad_norm": 3.7835094440523616,
+      "learning_rate": 2e-05,
+      "loss": 0.8786,
+      "step": 40577
+    },
+    {
+      "epoch": 0.9658668951728078,
+      "grad_norm": 5.851916160187753,
+      "learning_rate": 2e-05,
+      "loss": 0.2464,
+      "step": 40578
+    },
+    {
+      "epoch": 0.9658906978958393,
+      "grad_norm": 7.938928571440595,
+      "learning_rate": 2e-05,
+      "loss": 0.066,
+      "step": 40579
+    },
+    {
+      "epoch": 0.9659145006188709,
+      "grad_norm": 3.4324520061246426,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 40580
+    },
+    {
+      "epoch": 0.9659383033419023,
+      "grad_norm": 0.0668923460695471,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40581
+    },
+    {
+      "epoch": 0.9659621060649338,
+      "grad_norm": 0.457983875307195,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 40582
+    },
+    {
+      "epoch": 0.9659859087879653,
+      "grad_norm": 0.024127002272238914,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40583
+    },
+    {
+      "epoch": 0.9660097115109969,
+      "grad_norm": 0.27105421140100344,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40584
+    },
+    {
+      "epoch": 0.9660335142340284,
+      "grad_norm": 13.082195796189106,
+      "learning_rate": 2e-05,
+      "loss": 1.3598,
+      "step": 40585
+    },
+    {
+      "epoch": 0.9660573169570599,
+      "grad_norm": 1.0051897014756865,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 40586
+    },
+    {
+      "epoch": 0.9660811196800914,
+      "grad_norm": 14.041380744693,
+      "learning_rate": 2e-05,
+      "loss": 0.1415,
+      "step": 40587
+    },
+    {
+      "epoch": 0.966104922403123,
+      "grad_norm": 3.0497162578055885,
+      "learning_rate": 2e-05,
+      "loss": 0.3442,
+      "step": 40588
+    },
+    {
+      "epoch": 0.9661287251261544,
+      "grad_norm": 4.36181593847481,
+      "learning_rate": 2e-05,
+      "loss": 0.608,
+      "step": 40589
+    },
+    {
+      "epoch": 0.9661525278491859,
+      "grad_norm": 3.3589186681271808,
+      "learning_rate": 2e-05,
+      "loss": 0.5307,
+      "step": 40590
+    },
+    {
+      "epoch": 0.9661763305722174,
+      "grad_norm": 0.113182963160838,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40591
+    },
+    {
+      "epoch": 0.966200133295249,
+      "grad_norm": 10.556960991219562,
+      "learning_rate": 2e-05,
+      "loss": 0.4748,
+      "step": 40592
+    },
+    {
+      "epoch": 0.9662239360182805,
+      "grad_norm": 4.010280156778007,
+      "learning_rate": 2e-05,
+      "loss": 0.7196,
+      "step": 40593
+    },
+    {
+      "epoch": 0.966247738741312,
+      "grad_norm": 2.5326890093038377,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 40594
+    },
+    {
+      "epoch": 0.9662715414643436,
+      "grad_norm": 0.6694603768460999,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 40595
+    },
+    {
+      "epoch": 0.9662953441873751,
+      "grad_norm": 0.24606766276870662,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 40596
+    },
+    {
+      "epoch": 0.9663191469104065,
+      "grad_norm": 0.4004355747733959,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 40597
+    },
+    {
+      "epoch": 0.966342949633438,
+      "grad_norm": 7.675295581601244,
+      "learning_rate": 2e-05,
+      "loss": 0.2898,
+      "step": 40598
+    },
+    {
+      "epoch": 0.9663667523564696,
+      "grad_norm": 0.027041541008286443,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40599
+    },
+    {
+      "epoch": 0.9663905550795011,
+      "grad_norm": 4.689969087469115,
+      "learning_rate": 2e-05,
+      "loss": 0.7216,
+      "step": 40600
+    },
+    {
+      "epoch": 0.9664143578025326,
+      "grad_norm": 8.10918906850356,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 40601
+    },
+    {
+      "epoch": 0.9664381605255641,
+      "grad_norm": 1.9740526414085549,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 40602
+    },
+    {
+      "epoch": 0.9664619632485957,
+      "grad_norm": 0.6194551065200343,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 40603
+    },
+    {
+      "epoch": 0.9664857659716272,
+      "grad_norm": 2.2317403974024628,
+      "learning_rate": 2e-05,
+      "loss": 0.0437,
+      "step": 40604
+    },
+    {
+      "epoch": 0.9665095686946586,
+      "grad_norm": 2.385291519235179,
+      "learning_rate": 2e-05,
+      "loss": 0.1168,
+      "step": 40605
+    },
+    {
+      "epoch": 0.9665333714176901,
+      "grad_norm": 2.975534592370195,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 40606
+    },
+    {
+      "epoch": 0.9665571741407217,
+      "grad_norm": 4.922239239193293,
+      "learning_rate": 2e-05,
+      "loss": 0.9099,
+      "step": 40607
+    },
+    {
+      "epoch": 0.9665809768637532,
+      "grad_norm": 3.902502410430441,
+      "learning_rate": 2e-05,
+      "loss": 0.3826,
+      "step": 40608
+    },
+    {
+      "epoch": 0.9666047795867847,
+      "grad_norm": 1.1282195052989699,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 40609
+    },
+    {
+      "epoch": 0.9666285823098163,
+      "grad_norm": 0.017738628207088762,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40610
+    },
+    {
+      "epoch": 0.9666523850328478,
+      "grad_norm": 5.233901015059805,
+      "learning_rate": 2e-05,
+      "loss": 0.6472,
+      "step": 40611
+    },
+    {
+      "epoch": 0.9666761877558793,
+      "grad_norm": 0.02119880349562389,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40612
+    },
+    {
+      "epoch": 0.9666999904789108,
+      "grad_norm": 3.870588112554105,
+      "learning_rate": 2e-05,
+      "loss": 0.4529,
+      "step": 40613
+    },
+    {
+      "epoch": 0.9667237932019423,
+      "grad_norm": 5.682084606708353,
+      "learning_rate": 2e-05,
+      "loss": 0.6054,
+      "step": 40614
+    },
+    {
+      "epoch": 0.9667475959249738,
+      "grad_norm": 0.16776353115031406,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40615
+    },
+    {
+      "epoch": 0.9667713986480053,
+      "grad_norm": 2.2704591340357743,
+      "learning_rate": 2e-05,
+      "loss": 0.133,
+      "step": 40616
+    },
+    {
+      "epoch": 0.9667952013710368,
+      "grad_norm": 0.29676811816340837,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 40617
+    },
+    {
+      "epoch": 0.9668190040940684,
+      "grad_norm": 4.4616871587677345,
+      "learning_rate": 2e-05,
+      "loss": 0.8352,
+      "step": 40618
+    },
+    {
+      "epoch": 0.9668428068170999,
+      "grad_norm": 0.3517220107618878,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 40619
+    },
+    {
+      "epoch": 0.9668666095401314,
+      "grad_norm": 3.051997603150838,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 40620
+    },
+    {
+      "epoch": 0.9668904122631629,
+      "grad_norm": 2.9882632937031675,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 40621
+    },
+    {
+      "epoch": 0.9669142149861945,
+      "grad_norm": 6.032320212020753,
+      "learning_rate": 2e-05,
+      "loss": 0.2732,
+      "step": 40622
+    },
+    {
+      "epoch": 0.9669380177092259,
+      "grad_norm": 1.609673530622746,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 40623
+    },
+    {
+      "epoch": 0.9669618204322574,
+      "grad_norm": 0.05307895714183127,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40624
+    },
+    {
+      "epoch": 0.966985623155289,
+      "grad_norm": 10.33837221242297,
+      "learning_rate": 2e-05,
+      "loss": 0.3761,
+      "step": 40625
+    },
+    {
+      "epoch": 0.9670094258783205,
+      "grad_norm": 6.685828413268415,
+      "learning_rate": 2e-05,
+      "loss": 1.0462,
+      "step": 40626
+    },
+    {
+      "epoch": 0.967033228601352,
+      "grad_norm": 3.5759496303883744,
+      "learning_rate": 2e-05,
+      "loss": 0.4893,
+      "step": 40627
+    },
+    {
+      "epoch": 0.9670570313243835,
+      "grad_norm": 8.41549352844582,
+      "learning_rate": 2e-05,
+      "loss": 0.3312,
+      "step": 40628
+    },
+    {
+      "epoch": 0.9670808340474151,
+      "grad_norm": 0.05613855212980426,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40629
+    },
+    {
+      "epoch": 0.9671046367704466,
+      "grad_norm": 0.03898500289790856,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40630
+    },
+    {
+      "epoch": 0.967128439493478,
+      "grad_norm": 3.800602926354189,
+      "learning_rate": 2e-05,
+      "loss": 0.4253,
+      "step": 40631
+    },
+    {
+      "epoch": 0.9671522422165095,
+      "grad_norm": 3.5421233118653603,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 40632
+    },
+    {
+      "epoch": 0.9671760449395411,
+      "grad_norm": 10.117933386564156,
+      "learning_rate": 2e-05,
+      "loss": 0.1306,
+      "step": 40633
+    },
+    {
+      "epoch": 0.9671998476625726,
+      "grad_norm": 0.4036937479465553,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 40634
+    },
+    {
+      "epoch": 0.9672236503856041,
+      "grad_norm": 4.090825703401517,
+      "learning_rate": 2e-05,
+      "loss": 0.1016,
+      "step": 40635
+    },
+    {
+      "epoch": 0.9672474531086356,
+      "grad_norm": 4.24327814024679,
+      "learning_rate": 2e-05,
+      "loss": 0.0948,
+      "step": 40636
+    },
+    {
+      "epoch": 0.9672712558316672,
+      "grad_norm": 3.8784901837635073,
+      "learning_rate": 2e-05,
+      "loss": 0.8,
+      "step": 40637
+    },
+    {
+      "epoch": 0.9672950585546987,
+      "grad_norm": 4.2288795241644195,
+      "learning_rate": 2e-05,
+      "loss": 1.0666,
+      "step": 40638
+    },
+    {
+      "epoch": 0.9673188612777301,
+      "grad_norm": 8.18517592300439,
+      "learning_rate": 2e-05,
+      "loss": 0.2156,
+      "step": 40639
+    },
+    {
+      "epoch": 0.9673426640007616,
+      "grad_norm": 1.0031959182609138,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 40640
+    },
+    {
+      "epoch": 0.9673664667237932,
+      "grad_norm": 2.9531307367603334,
+      "learning_rate": 2e-05,
+      "loss": 0.2828,
+      "step": 40641
+    },
+    {
+      "epoch": 0.9673902694468247,
+      "grad_norm": 0.5051170796101782,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 40642
+    },
+    {
+      "epoch": 0.9674140721698562,
+      "grad_norm": 3.354440433525227,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 40643
+    },
+    {
+      "epoch": 0.9674378748928878,
+      "grad_norm": 1.3591673775559565,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 40644
+    },
+    {
+      "epoch": 0.9674616776159193,
+      "grad_norm": 2.011163456913428,
+      "learning_rate": 2e-05,
+      "loss": 0.0521,
+      "step": 40645
+    },
+    {
+      "epoch": 0.9674854803389508,
+      "grad_norm": 0.09727154552830122,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 40646
+    },
+    {
+      "epoch": 0.9675092830619823,
+      "grad_norm": 3.7965011249883975,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 40647
+    },
+    {
+      "epoch": 0.9675330857850138,
+      "grad_norm": 14.268989790497148,
+      "learning_rate": 2e-05,
+      "loss": 0.5644,
+      "step": 40648
+    },
+    {
+      "epoch": 0.9675568885080453,
+      "grad_norm": 2.900307878491425,
+      "learning_rate": 2e-05,
+      "loss": 0.5034,
+      "step": 40649
+    },
+    {
+      "epoch": 0.9675806912310768,
+      "grad_norm": 5.974840075434006,
+      "learning_rate": 2e-05,
+      "loss": 0.9806,
+      "step": 40650
+    },
+    {
+      "epoch": 0.9676044939541083,
+      "grad_norm": 0.8354538302000879,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 40651
+    },
+    {
+      "epoch": 0.9676282966771399,
+      "grad_norm": 0.8214109956463052,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 40652
+    },
+    {
+      "epoch": 0.9676520994001714,
+      "grad_norm": 4.8568108936909695,
+      "learning_rate": 2e-05,
+      "loss": 0.7289,
+      "step": 40653
+    },
+    {
+      "epoch": 0.9676759021232029,
+      "grad_norm": 4.902814829496007,
+      "learning_rate": 2e-05,
+      "loss": 0.5015,
+      "step": 40654
+    },
+    {
+      "epoch": 0.9676997048462344,
+      "grad_norm": 0.18815140121769408,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 40655
+    },
+    {
+      "epoch": 0.967723507569266,
+      "grad_norm": 0.33534298973377386,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 40656
+    },
+    {
+      "epoch": 0.9677473102922974,
+      "grad_norm": 0.8639491394046069,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 40657
+    },
+    {
+      "epoch": 0.9677711130153289,
+      "grad_norm": 3.4302209797303735,
+      "learning_rate": 2e-05,
+      "loss": 0.038,
+      "step": 40658
+    },
+    {
+      "epoch": 0.9677949157383605,
+      "grad_norm": 1.4054055002394468,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 40659
+    },
+    {
+      "epoch": 0.967818718461392,
+      "grad_norm": 1.8040845509952765,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 40660
+    },
+    {
+      "epoch": 0.9678425211844235,
+      "grad_norm": 2.5583380585190323,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 40661
+    },
+    {
+      "epoch": 0.967866323907455,
+      "grad_norm": 0.009957033185755831,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 40662
+    },
+    {
+      "epoch": 0.9678901266304866,
+      "grad_norm": 10.834791417292765,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 40663
+    },
+    {
+      "epoch": 0.967913929353518,
+      "grad_norm": 9.591574785602138,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 40664
+    },
+    {
+      "epoch": 0.9679377320765495,
+      "grad_norm": 0.17054131634082176,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 40665
+    },
+    {
+      "epoch": 0.967961534799581,
+      "grad_norm": 4.048618558775894,
+      "learning_rate": 2e-05,
+      "loss": 0.0634,
+      "step": 40666
+    },
+    {
+      "epoch": 0.9679853375226126,
+      "grad_norm": 0.7048236935993466,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 40667
+    },
+    {
+      "epoch": 0.9680091402456441,
+      "grad_norm": 5.400968802693223,
+      "learning_rate": 2e-05,
+      "loss": 0.6756,
+      "step": 40668
+    },
+    {
+      "epoch": 0.9680329429686756,
+      "grad_norm": 3.923055930498399,
+      "learning_rate": 2e-05,
+      "loss": 0.9017,
+      "step": 40669
+    },
+    {
+      "epoch": 0.9680567456917071,
+      "grad_norm": 7.795323316194058,
+      "learning_rate": 2e-05,
+      "loss": 0.6591,
+      "step": 40670
+    },
+    {
+      "epoch": 0.9680805484147387,
+      "grad_norm": 2.2224820897040187,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 40671
+    },
+    {
+      "epoch": 0.9681043511377702,
+      "grad_norm": 1.8785585227681751,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 40672
+    },
+    {
+      "epoch": 0.9681281538608016,
+      "grad_norm": 3.5780897930383273,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 40673
+    },
+    {
+      "epoch": 0.9681519565838332,
+      "grad_norm": 0.029058263893370198,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40674
+    },
+    {
+      "epoch": 0.9681757593068647,
+      "grad_norm": 4.7724211388483075,
+      "learning_rate": 2e-05,
+      "loss": 0.0972,
+      "step": 40675
+    },
+    {
+      "epoch": 0.9681995620298962,
+      "grad_norm": 3.6520884870999017,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 40676
+    },
+    {
+      "epoch": 0.9682233647529277,
+      "grad_norm": 3.136141444261161,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 40677
+    },
+    {
+      "epoch": 0.9682471674759593,
+      "grad_norm": 3.072892999396107,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 40678
+    },
+    {
+      "epoch": 0.9682709701989908,
+      "grad_norm": 2.9818505850223533,
+      "learning_rate": 2e-05,
+      "loss": 0.4148,
+      "step": 40679
+    },
+    {
+      "epoch": 0.9682947729220223,
+      "grad_norm": 0.03877408918682028,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40680
+    },
+    {
+      "epoch": 0.9683185756450537,
+      "grad_norm": 0.19920552020778315,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 40681
+    },
+    {
+      "epoch": 0.9683423783680853,
+      "grad_norm": 23.970165957941365,
+      "learning_rate": 2e-05,
+      "loss": 0.8499,
+      "step": 40682
+    },
+    {
+      "epoch": 0.9683661810911168,
+      "grad_norm": 9.578416707556956,
+      "learning_rate": 2e-05,
+      "loss": 0.875,
+      "step": 40683
+    },
+    {
+      "epoch": 0.9683899838141483,
+      "grad_norm": 7.255158837865431,
+      "learning_rate": 2e-05,
+      "loss": 0.2436,
+      "step": 40684
+    },
+    {
+      "epoch": 0.9684137865371798,
+      "grad_norm": 3.879368686960576,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 40685
+    },
+    {
+      "epoch": 0.9684375892602114,
+      "grad_norm": 5.398097995733772,
+      "learning_rate": 2e-05,
+      "loss": 1.0755,
+      "step": 40686
+    },
+    {
+      "epoch": 0.9684613919832429,
+      "grad_norm": 6.327380026253459,
+      "learning_rate": 2e-05,
+      "loss": 1.2945,
+      "step": 40687
+    },
+    {
+      "epoch": 0.9684851947062744,
+      "grad_norm": 7.349478055296043,
+      "learning_rate": 2e-05,
+      "loss": 0.8007,
+      "step": 40688
+    },
+    {
+      "epoch": 0.968508997429306,
+      "grad_norm": 2.744725733288159,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 40689
+    },
+    {
+      "epoch": 0.9685328001523374,
+      "grad_norm": 2.760255279181883,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 40690
+    },
+    {
+      "epoch": 0.9685566028753689,
+      "grad_norm": 5.249329293612056,
+      "learning_rate": 2e-05,
+      "loss": 0.4297,
+      "step": 40691
+    },
+    {
+      "epoch": 0.9685804055984004,
+      "grad_norm": 0.1339546778213608,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40692
+    },
+    {
+      "epoch": 0.968604208321432,
+      "grad_norm": 2.1006600091843928,
+      "learning_rate": 2e-05,
+      "loss": 0.1265,
+      "step": 40693
+    },
+    {
+      "epoch": 0.9686280110444635,
+      "grad_norm": 5.311288613192592,
+      "learning_rate": 2e-05,
+      "loss": 1.4282,
+      "step": 40694
+    },
+    {
+      "epoch": 0.968651813767495,
+      "grad_norm": 5.498219068348364,
+      "learning_rate": 2e-05,
+      "loss": 0.5334,
+      "step": 40695
+    },
+    {
+      "epoch": 0.9686756164905265,
+      "grad_norm": 0.4525860566443597,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 40696
+    },
+    {
+      "epoch": 0.9686994192135581,
+      "grad_norm": 0.6189312797438696,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 40697
+    },
+    {
+      "epoch": 0.9687232219365896,
+      "grad_norm": 16.196754535240913,
+      "learning_rate": 2e-05,
+      "loss": 0.7584,
+      "step": 40698
+    },
+    {
+      "epoch": 0.968747024659621,
+      "grad_norm": 1.0262817112376932,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 40699
+    },
+    {
+      "epoch": 0.9687708273826525,
+      "grad_norm": 9.090087354091473,
+      "learning_rate": 2e-05,
+      "loss": 0.363,
+      "step": 40700
+    },
+    {
+      "epoch": 0.9687946301056841,
+      "grad_norm": 15.016484339040561,
+      "learning_rate": 2e-05,
+      "loss": 0.5793,
+      "step": 40701
+    },
+    {
+      "epoch": 0.9688184328287156,
+      "grad_norm": 2.3329429648915205,
+      "learning_rate": 2e-05,
+      "loss": 0.0391,
+      "step": 40702
+    },
+    {
+      "epoch": 0.9688422355517471,
+      "grad_norm": 0.015019116225554107,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 40703
+    },
+    {
+      "epoch": 0.9688660382747787,
+      "grad_norm": 7.072026919862828,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 40704
+    },
+    {
+      "epoch": 0.9688898409978102,
+      "grad_norm": 5.857469715580093,
+      "learning_rate": 2e-05,
+      "loss": 0.7369,
+      "step": 40705
+    },
+    {
+      "epoch": 0.9689136437208417,
+      "grad_norm": 7.304768677453677,
+      "learning_rate": 2e-05,
+      "loss": 0.9697,
+      "step": 40706
+    },
+    {
+      "epoch": 0.9689374464438731,
+      "grad_norm": 4.9577665500348305,
+      "learning_rate": 2e-05,
+      "loss": 0.8997,
+      "step": 40707
+    },
+    {
+      "epoch": 0.9689612491669047,
+      "grad_norm": 0.5385108390840724,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 40708
+    },
+    {
+      "epoch": 0.9689850518899362,
+      "grad_norm": 2.1116014154537917,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 40709
+    },
+    {
+      "epoch": 0.9690088546129677,
+      "grad_norm": 4.301699250505975,
+      "learning_rate": 2e-05,
+      "loss": 0.7582,
+      "step": 40710
+    },
+    {
+      "epoch": 0.9690326573359992,
+      "grad_norm": 3.8301548643269365,
+      "learning_rate": 2e-05,
+      "loss": 0.5444,
+      "step": 40711
+    },
+    {
+      "epoch": 0.9690564600590308,
+      "grad_norm": 5.562777381535771,
+      "learning_rate": 2e-05,
+      "loss": 0.266,
+      "step": 40712
+    },
+    {
+      "epoch": 0.9690802627820623,
+      "grad_norm": 1.943188260805589,
+      "learning_rate": 2e-05,
+      "loss": 0.2164,
+      "step": 40713
+    },
+    {
+      "epoch": 0.9691040655050938,
+      "grad_norm": 0.0769764670246804,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 40714
+    },
+    {
+      "epoch": 0.9691278682281252,
+      "grad_norm": 7.326616571188302,
+      "learning_rate": 2e-05,
+      "loss": 0.6628,
+      "step": 40715
+    },
+    {
+      "epoch": 0.9691516709511568,
+      "grad_norm": 11.762327160830063,
+      "learning_rate": 2e-05,
+      "loss": 0.4517,
+      "step": 40716
+    },
+    {
+      "epoch": 0.9691754736741883,
+      "grad_norm": 11.480224687563222,
+      "learning_rate": 2e-05,
+      "loss": 0.9021,
+      "step": 40717
+    },
+    {
+      "epoch": 0.9691992763972198,
+      "grad_norm": 4.637138397838696,
+      "learning_rate": 2e-05,
+      "loss": 0.6834,
+      "step": 40718
+    },
+    {
+      "epoch": 0.9692230791202514,
+      "grad_norm": 2.882364869604943,
+      "learning_rate": 2e-05,
+      "loss": 0.5394,
+      "step": 40719
+    },
+    {
+      "epoch": 0.9692468818432829,
+      "grad_norm": 2.5612136826175065,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 40720
+    },
+    {
+      "epoch": 0.9692706845663144,
+      "grad_norm": 3.2021769663731465,
+      "learning_rate": 2e-05,
+      "loss": 0.163,
+      "step": 40721
+    },
+    {
+      "epoch": 0.9692944872893459,
+      "grad_norm": 3.6054255850316026,
+      "learning_rate": 2e-05,
+      "loss": 0.7152,
+      "step": 40722
+    },
+    {
+      "epoch": 0.9693182900123775,
+      "grad_norm": 0.24185405830969522,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 40723
+    },
+    {
+      "epoch": 0.9693420927354089,
+      "grad_norm": 0.5841308318514689,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 40724
+    },
+    {
+      "epoch": 0.9693658954584404,
+      "grad_norm": 3.666278469286052,
+      "learning_rate": 2e-05,
+      "loss": 0.8205,
+      "step": 40725
+    },
+    {
+      "epoch": 0.9693896981814719,
+      "grad_norm": 2.547899707241982,
+      "learning_rate": 2e-05,
+      "loss": 0.0573,
+      "step": 40726
+    },
+    {
+      "epoch": 0.9694135009045035,
+      "grad_norm": 2.3081836275746133,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 40727
+    },
+    {
+      "epoch": 0.969437303627535,
+      "grad_norm": 4.639950025900819,
+      "learning_rate": 2e-05,
+      "loss": 1.1259,
+      "step": 40728
+    },
+    {
+      "epoch": 0.9694611063505665,
+      "grad_norm": 1.9346125544005242,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 40729
+    },
+    {
+      "epoch": 0.969484909073598,
+      "grad_norm": 0.46609954982251933,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 40730
+    },
+    {
+      "epoch": 0.9695087117966296,
+      "grad_norm": 2.9128606762394993,
+      "learning_rate": 2e-05,
+      "loss": 0.4754,
+      "step": 40731
+    },
+    {
+      "epoch": 0.969532514519661,
+      "grad_norm": 2.5580202989638634,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 40732
+    },
+    {
+      "epoch": 0.9695563172426925,
+      "grad_norm": 4.139299731604868,
+      "learning_rate": 2e-05,
+      "loss": 1.0096,
+      "step": 40733
+    },
+    {
+      "epoch": 0.9695801199657241,
+      "grad_norm": 0.24088454621992858,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 40734
+    },
+    {
+      "epoch": 0.9696039226887556,
+      "grad_norm": 0.06904933610997378,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40735
+    },
+    {
+      "epoch": 0.9696277254117871,
+      "grad_norm": 0.6550790278884755,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 40736
+    },
+    {
+      "epoch": 0.9696515281348186,
+      "grad_norm": 1.2018230862348256,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 40737
+    },
+    {
+      "epoch": 0.9696753308578502,
+      "grad_norm": 0.04983874240594521,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40738
+    },
+    {
+      "epoch": 0.9696991335808817,
+      "grad_norm": 0.10469072441647197,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 40739
+    },
+    {
+      "epoch": 0.9697229363039132,
+      "grad_norm": 0.42802078202050037,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 40740
+    },
+    {
+      "epoch": 0.9697467390269446,
+      "grad_norm": 0.2339392666909812,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 40741
+    },
+    {
+      "epoch": 0.9697705417499762,
+      "grad_norm": 2.9718335879535154,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 40742
+    },
+    {
+      "epoch": 0.9697943444730077,
+      "grad_norm": 5.8406903791806055,
+      "learning_rate": 2e-05,
+      "loss": 1.6261,
+      "step": 40743
+    },
+    {
+      "epoch": 0.9698181471960392,
+      "grad_norm": 3.294143881757487,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 40744
+    },
+    {
+      "epoch": 0.9698419499190707,
+      "grad_norm": 0.6530849876606032,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 40745
+    },
+    {
+      "epoch": 0.9698657526421023,
+      "grad_norm": 3.8154302678879293,
+      "learning_rate": 2e-05,
+      "loss": 0.501,
+      "step": 40746
+    },
+    {
+      "epoch": 0.9698895553651338,
+      "grad_norm": 4.445062734842024,
+      "learning_rate": 2e-05,
+      "loss": 0.8341,
+      "step": 40747
+    },
+    {
+      "epoch": 0.9699133580881653,
+      "grad_norm": 0.20988738001965448,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40748
+    },
+    {
+      "epoch": 0.9699371608111969,
+      "grad_norm": 2.726495457301301,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 40749
+    },
+    {
+      "epoch": 0.9699609635342283,
+      "grad_norm": 0.03791343356262358,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40750
+    },
+    {
+      "epoch": 0.9699847662572598,
+      "grad_norm": 0.10122545091852611,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 40751
+    },
+    {
+      "epoch": 0.9700085689802913,
+      "grad_norm": 2.620206232594728,
+      "learning_rate": 2e-05,
+      "loss": 0.4979,
+      "step": 40752
+    },
+    {
+      "epoch": 0.9700323717033229,
+      "grad_norm": 0.5414286229933708,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 40753
+    },
+    {
+      "epoch": 0.9700561744263544,
+      "grad_norm": 6.28373864416582,
+      "learning_rate": 2e-05,
+      "loss": 1.073,
+      "step": 40754
+    },
+    {
+      "epoch": 0.9700799771493859,
+      "grad_norm": 12.613713310520204,
+      "learning_rate": 2e-05,
+      "loss": 0.8533,
+      "step": 40755
+    },
+    {
+      "epoch": 0.9701037798724174,
+      "grad_norm": 2.2074923375912188,
+      "learning_rate": 2e-05,
+      "loss": 0.4674,
+      "step": 40756
+    },
+    {
+      "epoch": 0.970127582595449,
+      "grad_norm": 2.154844211644248,
+      "learning_rate": 2e-05,
+      "loss": 0.0225,
+      "step": 40757
+    },
+    {
+      "epoch": 0.9701513853184804,
+      "grad_norm": 0.2421175973330709,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 40758
+    },
+    {
+      "epoch": 0.9701751880415119,
+      "grad_norm": 3.1486605244564956,
+      "learning_rate": 2e-05,
+      "loss": 0.6423,
+      "step": 40759
+    },
+    {
+      "epoch": 0.9701989907645434,
+      "grad_norm": 4.537594994643794,
+      "learning_rate": 2e-05,
+      "loss": 0.8755,
+      "step": 40760
+    },
+    {
+      "epoch": 0.970222793487575,
+      "grad_norm": 1.6447905348137843,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 40761
+    },
+    {
+      "epoch": 0.9702465962106065,
+      "grad_norm": 3.2535204072749027,
+      "learning_rate": 2e-05,
+      "loss": 0.3882,
+      "step": 40762
+    },
+    {
+      "epoch": 0.970270398933638,
+      "grad_norm": 4.281334497970108,
+      "learning_rate": 2e-05,
+      "loss": 0.9481,
+      "step": 40763
+    },
+    {
+      "epoch": 0.9702942016566696,
+      "grad_norm": 0.18325043956798467,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 40764
+    },
+    {
+      "epoch": 0.9703180043797011,
+      "grad_norm": 4.171895811250781,
+      "learning_rate": 2e-05,
+      "loss": 0.61,
+      "step": 40765
+    },
+    {
+      "epoch": 0.9703418071027325,
+      "grad_norm": 0.05065021265896844,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40766
+    },
+    {
+      "epoch": 0.970365609825764,
+      "grad_norm": 1.0898332536370394,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 40767
+    },
+    {
+      "epoch": 0.9703894125487956,
+      "grad_norm": 0.06008274687961332,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40768
+    },
+    {
+      "epoch": 0.9704132152718271,
+      "grad_norm": 4.647848888395978,
+      "learning_rate": 2e-05,
+      "loss": 0.2756,
+      "step": 40769
+    },
+    {
+      "epoch": 0.9704370179948586,
+      "grad_norm": 13.930670140546487,
+      "learning_rate": 2e-05,
+      "loss": 0.9763,
+      "step": 40770
+    },
+    {
+      "epoch": 0.9704608207178901,
+      "grad_norm": 3.036471124664221,
+      "learning_rate": 2e-05,
+      "loss": 0.0618,
+      "step": 40771
+    },
+    {
+      "epoch": 0.9704846234409217,
+      "grad_norm": 0.020446243683472233,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 40772
+    },
+    {
+      "epoch": 0.9705084261639532,
+      "grad_norm": 22.954366945782404,
+      "learning_rate": 2e-05,
+      "loss": 0.6794,
+      "step": 40773
+    },
+    {
+      "epoch": 0.9705322288869846,
+      "grad_norm": 4.296716743663957,
+      "learning_rate": 2e-05,
+      "loss": 0.9246,
+      "step": 40774
+    },
+    {
+      "epoch": 0.9705560316100161,
+      "grad_norm": 3.1078457290503128,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 40775
+    },
+    {
+      "epoch": 0.9705798343330477,
+      "grad_norm": 13.28022763794569,
+      "learning_rate": 2e-05,
+      "loss": 1.3256,
+      "step": 40776
+    },
+    {
+      "epoch": 0.9706036370560792,
+      "grad_norm": 29.27804045694028,
+      "learning_rate": 2e-05,
+      "loss": 0.4489,
+      "step": 40777
+    },
+    {
+      "epoch": 0.9706274397791107,
+      "grad_norm": 4.5407580978989905,
+      "learning_rate": 2e-05,
+      "loss": 0.8304,
+      "step": 40778
+    },
+    {
+      "epoch": 0.9706512425021423,
+      "grad_norm": 3.476017573365969,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 40779
+    },
+    {
+      "epoch": 0.9706750452251738,
+      "grad_norm": 2.3959257969616723,
+      "learning_rate": 2e-05,
+      "loss": 0.2862,
+      "step": 40780
+    },
+    {
+      "epoch": 0.9706988479482053,
+      "grad_norm": 0.4777002196519848,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 40781
+    },
+    {
+      "epoch": 0.9707226506712368,
+      "grad_norm": 4.8806018858207345,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 40782
+    },
+    {
+      "epoch": 0.9707464533942683,
+      "grad_norm": 1.0220264775260182,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 40783
+    },
+    {
+      "epoch": 0.9707702561172998,
+      "grad_norm": 0.08105882718024354,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 40784
+    },
+    {
+      "epoch": 0.9707940588403313,
+      "grad_norm": 8.039564862862857,
+      "learning_rate": 2e-05,
+      "loss": 0.6835,
+      "step": 40785
+    },
+    {
+      "epoch": 0.9708178615633628,
+      "grad_norm": 10.4877785348813,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 40786
+    },
+    {
+      "epoch": 0.9708416642863944,
+      "grad_norm": 2.5191185707086285,
+      "learning_rate": 2e-05,
+      "loss": 0.4522,
+      "step": 40787
+    },
+    {
+      "epoch": 0.9708654670094259,
+      "grad_norm": 0.7027877217911259,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 40788
+    },
+    {
+      "epoch": 0.9708892697324574,
+      "grad_norm": 2.1955702763865172,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 40789
+    },
+    {
+      "epoch": 0.9709130724554889,
+      "grad_norm": 5.122890140486679,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 40790
+    },
+    {
+      "epoch": 0.9709368751785205,
+      "grad_norm": 2.0446522391044955,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 40791
+    },
+    {
+      "epoch": 0.9709606779015519,
+      "grad_norm": 3.5663156360257844,
+      "learning_rate": 2e-05,
+      "loss": 0.5101,
+      "step": 40792
+    },
+    {
+      "epoch": 0.9709844806245834,
+      "grad_norm": 5.283760742524156,
+      "learning_rate": 2e-05,
+      "loss": 0.129,
+      "step": 40793
+    },
+    {
+      "epoch": 0.971008283347615,
+      "grad_norm": 0.19980339656074847,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40794
+    },
+    {
+      "epoch": 0.9710320860706465,
+      "grad_norm": 5.874313297940063,
+      "learning_rate": 2e-05,
+      "loss": 1.2204,
+      "step": 40795
+    },
+    {
+      "epoch": 0.971055888793678,
+      "grad_norm": 0.039756645688418975,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40796
+    },
+    {
+      "epoch": 0.9710796915167095,
+      "grad_norm": 5.2169687161607925,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 40797
+    },
+    {
+      "epoch": 0.9711034942397411,
+      "grad_norm": 3.806485444041112,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 40798
+    },
+    {
+      "epoch": 0.9711272969627726,
+      "grad_norm": 3.49535506208943,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 40799
+    },
+    {
+      "epoch": 0.971151099685804,
+      "grad_norm": 1.0587531456002655,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 40800
+    },
+    {
+      "epoch": 0.9711749024088355,
+      "grad_norm": 0.05441265665028388,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40801
+    },
+    {
+      "epoch": 0.9711987051318671,
+      "grad_norm": 3.9786783589053245,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 40802
+    },
+    {
+      "epoch": 0.9712225078548986,
+      "grad_norm": 1.1172342188591025,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 40803
+    },
+    {
+      "epoch": 0.9712463105779301,
+      "grad_norm": 4.6130923800676165,
+      "learning_rate": 2e-05,
+      "loss": 0.5331,
+      "step": 40804
+    },
+    {
+      "epoch": 0.9712701133009616,
+      "grad_norm": 1.790110493223047,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 40805
+    },
+    {
+      "epoch": 0.9712939160239932,
+      "grad_norm": 3.1880244830540936,
+      "learning_rate": 2e-05,
+      "loss": 0.4667,
+      "step": 40806
+    },
+    {
+      "epoch": 0.9713177187470247,
+      "grad_norm": 4.504118696705063,
+      "learning_rate": 2e-05,
+      "loss": 1.0007,
+      "step": 40807
+    },
+    {
+      "epoch": 0.9713415214700561,
+      "grad_norm": 4.724911903002557,
+      "learning_rate": 2e-05,
+      "loss": 0.9838,
+      "step": 40808
+    },
+    {
+      "epoch": 0.9713653241930877,
+      "grad_norm": 5.6931463578338715,
+      "learning_rate": 2e-05,
+      "loss": 0.9619,
+      "step": 40809
+    },
+    {
+      "epoch": 0.9713891269161192,
+      "grad_norm": 2.614095867674792,
+      "learning_rate": 2e-05,
+      "loss": 0.2665,
+      "step": 40810
+    },
+    {
+      "epoch": 0.9714129296391507,
+      "grad_norm": 5.66853664527022,
+      "learning_rate": 2e-05,
+      "loss": 0.0906,
+      "step": 40811
+    },
+    {
+      "epoch": 0.9714367323621822,
+      "grad_norm": 3.8279397630094936,
+      "learning_rate": 2e-05,
+      "loss": 0.0505,
+      "step": 40812
+    },
+    {
+      "epoch": 0.9714605350852138,
+      "grad_norm": 4.931135887671304,
+      "learning_rate": 2e-05,
+      "loss": 0.759,
+      "step": 40813
+    },
+    {
+      "epoch": 0.9714843378082453,
+      "grad_norm": 0.293470601397482,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 40814
+    },
+    {
+      "epoch": 0.9715081405312768,
+      "grad_norm": 2.876358530366671,
+      "learning_rate": 2e-05,
+      "loss": 0.0899,
+      "step": 40815
+    },
+    {
+      "epoch": 0.9715319432543083,
+      "grad_norm": 0.12372366429085194,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 40816
+    },
+    {
+      "epoch": 0.9715557459773398,
+      "grad_norm": 0.0769480412874705,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 40817
+    },
+    {
+      "epoch": 0.9715795487003713,
+      "grad_norm": 3.49239744349098,
+      "learning_rate": 2e-05,
+      "loss": 0.2915,
+      "step": 40818
+    },
+    {
+      "epoch": 0.9716033514234028,
+      "grad_norm": 10.060940931161571,
+      "learning_rate": 2e-05,
+      "loss": 0.6821,
+      "step": 40819
+    },
+    {
+      "epoch": 0.9716271541464343,
+      "grad_norm": 0.2176109813380957,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 40820
+    },
+    {
+      "epoch": 0.9716509568694659,
+      "grad_norm": 4.224192445591008,
+      "learning_rate": 2e-05,
+      "loss": 0.1093,
+      "step": 40821
+    },
+    {
+      "epoch": 0.9716747595924974,
+      "grad_norm": 0.31569267563164455,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 40822
+    },
+    {
+      "epoch": 0.9716985623155289,
+      "grad_norm": 4.416427834690131,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 40823
+    },
+    {
+      "epoch": 0.9717223650385605,
+      "grad_norm": 0.08782770866833031,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 40824
+    },
+    {
+      "epoch": 0.971746167761592,
+      "grad_norm": 0.13393908752568293,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40825
+    },
+    {
+      "epoch": 0.9717699704846234,
+      "grad_norm": 4.2013042855941025,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 40826
+    },
+    {
+      "epoch": 0.9717937732076549,
+      "grad_norm": 4.88100232594088,
+      "learning_rate": 2e-05,
+      "loss": 1.4178,
+      "step": 40827
+    },
+    {
+      "epoch": 0.9718175759306865,
+      "grad_norm": 0.26231408270012224,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 40828
+    },
+    {
+      "epoch": 0.971841378653718,
+      "grad_norm": 2.894185196416864,
+      "learning_rate": 2e-05,
+      "loss": 0.4913,
+      "step": 40829
+    },
+    {
+      "epoch": 0.9718651813767495,
+      "grad_norm": 7.40291498214539,
+      "learning_rate": 2e-05,
+      "loss": 1.0047,
+      "step": 40830
+    },
+    {
+      "epoch": 0.971888984099781,
+      "grad_norm": 10.998998374975061,
+      "learning_rate": 2e-05,
+      "loss": 0.3779,
+      "step": 40831
+    },
+    {
+      "epoch": 0.9719127868228126,
+      "grad_norm": 5.851136448293704,
+      "learning_rate": 2e-05,
+      "loss": 0.8285,
+      "step": 40832
+    },
+    {
+      "epoch": 0.971936589545844,
+      "grad_norm": 5.467667833313598,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 40833
+    },
+    {
+      "epoch": 0.9719603922688755,
+      "grad_norm": 5.118670218335703,
+      "learning_rate": 2e-05,
+      "loss": 0.0884,
+      "step": 40834
+    },
+    {
+      "epoch": 0.971984194991907,
+      "grad_norm": 0.11813650839861538,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40835
+    },
+    {
+      "epoch": 0.9720079977149386,
+      "grad_norm": 6.106187417367525,
+      "learning_rate": 2e-05,
+      "loss": 0.9946,
+      "step": 40836
+    },
+    {
+      "epoch": 0.9720318004379701,
+      "grad_norm": 0.17966528523831127,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 40837
+    },
+    {
+      "epoch": 0.9720556031610016,
+      "grad_norm": 12.097121054217247,
+      "learning_rate": 2e-05,
+      "loss": 0.2254,
+      "step": 40838
+    },
+    {
+      "epoch": 0.9720794058840332,
+      "grad_norm": 0.017349446963997553,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40839
+    },
+    {
+      "epoch": 0.9721032086070647,
+      "grad_norm": 2.500451029029103,
+      "learning_rate": 2e-05,
+      "loss": 0.0349,
+      "step": 40840
+    },
+    {
+      "epoch": 0.9721270113300962,
+      "grad_norm": 2.993127399144752,
+      "learning_rate": 2e-05,
+      "loss": 0.3165,
+      "step": 40841
+    },
+    {
+      "epoch": 0.9721508140531276,
+      "grad_norm": 2.464172224484718,
+      "learning_rate": 2e-05,
+      "loss": 0.27,
+      "step": 40842
+    },
+    {
+      "epoch": 0.9721746167761592,
+      "grad_norm": 0.1325395319105681,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 40843
+    },
+    {
+      "epoch": 0.9721984194991907,
+      "grad_norm": 3.642938576626947,
+      "learning_rate": 2e-05,
+      "loss": 0.5142,
+      "step": 40844
+    },
+    {
+      "epoch": 0.9722222222222222,
+      "grad_norm": 0.12859906931698256,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 40845
+    },
+    {
+      "epoch": 0.9722460249452537,
+      "grad_norm": 10.479669039557239,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 40846
+    },
+    {
+      "epoch": 0.9722698276682853,
+      "grad_norm": 0.1329553262205755,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 40847
+    },
+    {
+      "epoch": 0.9722936303913168,
+      "grad_norm": 2.069338842868037,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 40848
+    },
+    {
+      "epoch": 0.9723174331143483,
+      "grad_norm": 4.3926488227898055,
+      "learning_rate": 2e-05,
+      "loss": 0.6033,
+      "step": 40849
+    },
+    {
+      "epoch": 0.9723412358373797,
+      "grad_norm": 0.2947390499428734,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40850
+    },
+    {
+      "epoch": 0.9723650385604113,
+      "grad_norm": 5.145077118502177,
+      "learning_rate": 2e-05,
+      "loss": 1.0699,
+      "step": 40851
+    },
+    {
+      "epoch": 0.9723888412834428,
+      "grad_norm": 5.792860992093295,
+      "learning_rate": 2e-05,
+      "loss": 0.1642,
+      "step": 40852
+    },
+    {
+      "epoch": 0.9724126440064743,
+      "grad_norm": 0.15330795734424355,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 40853
+    },
+    {
+      "epoch": 0.9724364467295059,
+      "grad_norm": 0.04726155058753521,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 40854
+    },
+    {
+      "epoch": 0.9724602494525374,
+      "grad_norm": 0.3779015948553706,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 40855
+    },
+    {
+      "epoch": 0.9724840521755689,
+      "grad_norm": 0.17998956008900946,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 40856
+    },
+    {
+      "epoch": 0.9725078548986004,
+      "grad_norm": 4.416427096072201,
+      "learning_rate": 2e-05,
+      "loss": 0.4174,
+      "step": 40857
+    },
+    {
+      "epoch": 0.972531657621632,
+      "grad_norm": 0.06264010195621592,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 40858
+    },
+    {
+      "epoch": 0.9725554603446634,
+      "grad_norm": 1.7920291050779658,
+      "learning_rate": 2e-05,
+      "loss": 0.2938,
+      "step": 40859
+    },
+    {
+      "epoch": 0.9725792630676949,
+      "grad_norm": 0.032263348186478184,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40860
+    },
+    {
+      "epoch": 0.9726030657907264,
+      "grad_norm": 2.678208301606178,
+      "learning_rate": 2e-05,
+      "loss": 0.0675,
+      "step": 40861
+    },
+    {
+      "epoch": 0.972626868513758,
+      "grad_norm": 0.0890763959712288,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40862
+    },
+    {
+      "epoch": 0.9726506712367895,
+      "grad_norm": 5.653260895279791,
+      "learning_rate": 2e-05,
+      "loss": 0.6273,
+      "step": 40863
+    },
+    {
+      "epoch": 0.972674473959821,
+      "grad_norm": 18.49480078774092,
+      "learning_rate": 2e-05,
+      "loss": 0.1418,
+      "step": 40864
+    },
+    {
+      "epoch": 0.9726982766828525,
+      "grad_norm": 5.968267826932666,
+      "learning_rate": 2e-05,
+      "loss": 0.9694,
+      "step": 40865
+    },
+    {
+      "epoch": 0.9727220794058841,
+      "grad_norm": 4.573920966465213,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 40866
+    },
+    {
+      "epoch": 0.9727458821289156,
+      "grad_norm": 7.553799163738661,
+      "learning_rate": 2e-05,
+      "loss": 1.0476,
+      "step": 40867
+    },
+    {
+      "epoch": 0.972769684851947,
+      "grad_norm": 4.9389141627663475,
+      "learning_rate": 2e-05,
+      "loss": 1.1532,
+      "step": 40868
+    },
+    {
+      "epoch": 0.9727934875749786,
+      "grad_norm": 3.8815761395344435,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 40869
+    },
+    {
+      "epoch": 0.9728172902980101,
+      "grad_norm": 0.056192602759142285,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40870
+    },
+    {
+      "epoch": 0.9728410930210416,
+      "grad_norm": 6.337876966936605,
+      "learning_rate": 2e-05,
+      "loss": 0.1563,
+      "step": 40871
+    },
+    {
+      "epoch": 0.9728648957440731,
+      "grad_norm": 3.407616923615134,
+      "learning_rate": 2e-05,
+      "loss": 0.481,
+      "step": 40872
+    },
+    {
+      "epoch": 0.9728886984671047,
+      "grad_norm": 2.9826260084187353,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 40873
+    },
+    {
+      "epoch": 0.9729125011901362,
+      "grad_norm": 1.3554720735941168,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 40874
+    },
+    {
+      "epoch": 0.9729363039131677,
+      "grad_norm": 7.294439104884268,
+      "learning_rate": 2e-05,
+      "loss": 0.2942,
+      "step": 40875
+    },
+    {
+      "epoch": 0.9729601066361991,
+      "grad_norm": 6.225327743665442,
+      "learning_rate": 2e-05,
+      "loss": 1.7819,
+      "step": 40876
+    },
+    {
+      "epoch": 0.9729839093592307,
+      "grad_norm": 1.909486578367363,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 40877
+    },
+    {
+      "epoch": 0.9730077120822622,
+      "grad_norm": 4.259803472421253,
+      "learning_rate": 2e-05,
+      "loss": 0.958,
+      "step": 40878
+    },
+    {
+      "epoch": 0.9730315148052937,
+      "grad_norm": 0.5986081581790055,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 40879
+    },
+    {
+      "epoch": 0.9730553175283252,
+      "grad_norm": 6.117850756349916,
+      "learning_rate": 2e-05,
+      "loss": 1.1572,
+      "step": 40880
+    },
+    {
+      "epoch": 0.9730791202513568,
+      "grad_norm": 13.954794931170346,
+      "learning_rate": 2e-05,
+      "loss": 0.9609,
+      "step": 40881
+    },
+    {
+      "epoch": 0.9731029229743883,
+      "grad_norm": 0.16243987601915952,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 40882
+    },
+    {
+      "epoch": 0.9731267256974198,
+      "grad_norm": 9.747382764298914,
+      "learning_rate": 2e-05,
+      "loss": 0.4196,
+      "step": 40883
+    },
+    {
+      "epoch": 0.9731505284204514,
+      "grad_norm": 3.949731957008204,
+      "learning_rate": 2e-05,
+      "loss": 0.452,
+      "step": 40884
+    },
+    {
+      "epoch": 0.9731743311434828,
+      "grad_norm": 16.47521998971039,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 40885
+    },
+    {
+      "epoch": 0.9731981338665143,
+      "grad_norm": 3.4523779383546827,
+      "learning_rate": 2e-05,
+      "loss": 0.6306,
+      "step": 40886
+    },
+    {
+      "epoch": 0.9732219365895458,
+      "grad_norm": 5.333353319014516,
+      "learning_rate": 2e-05,
+      "loss": 1.0612,
+      "step": 40887
+    },
+    {
+      "epoch": 0.9732457393125774,
+      "grad_norm": 5.143563405112168,
+      "learning_rate": 2e-05,
+      "loss": 1.3794,
+      "step": 40888
+    },
+    {
+      "epoch": 0.9732695420356089,
+      "grad_norm": 2.819374018170651,
+      "learning_rate": 2e-05,
+      "loss": 0.2666,
+      "step": 40889
+    },
+    {
+      "epoch": 0.9732933447586404,
+      "grad_norm": 0.8145245652296216,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 40890
+    },
+    {
+      "epoch": 0.9733171474816719,
+      "grad_norm": 9.307753490546073,
+      "learning_rate": 2e-05,
+      "loss": 0.4589,
+      "step": 40891
+    },
+    {
+      "epoch": 0.9733409502047035,
+      "grad_norm": 0.8832148625752877,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 40892
+    },
+    {
+      "epoch": 0.9733647529277349,
+      "grad_norm": 4.4953220097583895,
+      "learning_rate": 2e-05,
+      "loss": 0.6733,
+      "step": 40893
+    },
+    {
+      "epoch": 0.9733885556507664,
+      "grad_norm": 3.4986796094497894,
+      "learning_rate": 2e-05,
+      "loss": 0.0729,
+      "step": 40894
+    },
+    {
+      "epoch": 0.9734123583737979,
+      "grad_norm": 9.799688172623425,
+      "learning_rate": 2e-05,
+      "loss": 0.5616,
+      "step": 40895
+    },
+    {
+      "epoch": 0.9734361610968295,
+      "grad_norm": 4.353961717020526,
+      "learning_rate": 2e-05,
+      "loss": 0.5376,
+      "step": 40896
+    },
+    {
+      "epoch": 0.973459963819861,
+      "grad_norm": 0.3599033016167925,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 40897
+    },
+    {
+      "epoch": 0.9734837665428925,
+      "grad_norm": 0.035930603973421404,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 40898
+    },
+    {
+      "epoch": 0.973507569265924,
+      "grad_norm": 10.159542639591825,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 40899
+    },
+    {
+      "epoch": 0.9735313719889556,
+      "grad_norm": 11.048871527064225,
+      "learning_rate": 2e-05,
+      "loss": 0.2219,
+      "step": 40900
+    },
+    {
+      "epoch": 0.973555174711987,
+      "grad_norm": 1.7934492010801746,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 40901
+    },
+    {
+      "epoch": 0.9735789774350185,
+      "grad_norm": 5.616941069748532,
+      "learning_rate": 2e-05,
+      "loss": 1.7232,
+      "step": 40902
+    },
+    {
+      "epoch": 0.9736027801580501,
+      "grad_norm": 4.634568588127869,
+      "learning_rate": 2e-05,
+      "loss": 0.7106,
+      "step": 40903
+    },
+    {
+      "epoch": 0.9736265828810816,
+      "grad_norm": 8.296901530677781,
+      "learning_rate": 2e-05,
+      "loss": 0.375,
+      "step": 40904
+    },
+    {
+      "epoch": 0.9736503856041131,
+      "grad_norm": 3.962906452062876,
+      "learning_rate": 2e-05,
+      "loss": 0.1244,
+      "step": 40905
+    },
+    {
+      "epoch": 0.9736741883271446,
+      "grad_norm": 0.37129390116794364,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 40906
+    },
+    {
+      "epoch": 0.9736979910501762,
+      "grad_norm": 0.2772074876513252,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 40907
+    },
+    {
+      "epoch": 0.9737217937732077,
+      "grad_norm": 3.310599999958194,
+      "learning_rate": 2e-05,
+      "loss": 0.57,
+      "step": 40908
+    },
+    {
+      "epoch": 0.9737455964962392,
+      "grad_norm": 0.0377060068356068,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40909
+    },
+    {
+      "epoch": 0.9737693992192706,
+      "grad_norm": 5.822582303411092,
+      "learning_rate": 2e-05,
+      "loss": 1.9408,
+      "step": 40910
+    },
+    {
+      "epoch": 0.9737932019423022,
+      "grad_norm": 2.435892904427311,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 40911
+    },
+    {
+      "epoch": 0.9738170046653337,
+      "grad_norm": 5.548956730036701,
+      "learning_rate": 2e-05,
+      "loss": 0.4383,
+      "step": 40912
+    },
+    {
+      "epoch": 0.9738408073883652,
+      "grad_norm": 0.6632089560694555,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 40913
+    },
+    {
+      "epoch": 0.9738646101113967,
+      "grad_norm": 0.3467746880962987,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 40914
+    },
+    {
+      "epoch": 0.9738884128344283,
+      "grad_norm": 4.318645203357817,
+      "learning_rate": 2e-05,
+      "loss": 0.261,
+      "step": 40915
+    },
+    {
+      "epoch": 0.9739122155574598,
+      "grad_norm": 4.62737991879807,
+      "learning_rate": 2e-05,
+      "loss": 1.3087,
+      "step": 40916
+    },
+    {
+      "epoch": 0.9739360182804913,
+      "grad_norm": 2.3353642165680135,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 40917
+    },
+    {
+      "epoch": 0.9739598210035229,
+      "grad_norm": 0.11876449614742453,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 40918
+    },
+    {
+      "epoch": 0.9739836237265543,
+      "grad_norm": 1.7120672582027214,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 40919
+    },
+    {
+      "epoch": 0.9740074264495858,
+      "grad_norm": 8.814598896484542,
+      "learning_rate": 2e-05,
+      "loss": 1.4833,
+      "step": 40920
+    },
+    {
+      "epoch": 0.9740312291726173,
+      "grad_norm": 5.432985186112787,
+      "learning_rate": 2e-05,
+      "loss": 0.2103,
+      "step": 40921
+    },
+    {
+      "epoch": 0.9740550318956489,
+      "grad_norm": 15.430812949214012,
+      "learning_rate": 2e-05,
+      "loss": 0.7549,
+      "step": 40922
+    },
+    {
+      "epoch": 0.9740788346186804,
+      "grad_norm": 6.265408402931084,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 40923
+    },
+    {
+      "epoch": 0.9741026373417119,
+      "grad_norm": 3.089577018753649,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 40924
+    },
+    {
+      "epoch": 0.9741264400647434,
+      "grad_norm": 0.4086168424623162,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 40925
+    },
+    {
+      "epoch": 0.974150242787775,
+      "grad_norm": 3.3159896220636247,
+      "learning_rate": 2e-05,
+      "loss": 0.6254,
+      "step": 40926
+    },
+    {
+      "epoch": 0.9741740455108064,
+      "grad_norm": 0.733749529468283,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 40927
+    },
+    {
+      "epoch": 0.9741978482338379,
+      "grad_norm": 0.04669998320096604,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40928
+    },
+    {
+      "epoch": 0.9742216509568694,
+      "grad_norm": 3.732859139977757,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 40929
+    },
+    {
+      "epoch": 0.974245453679901,
+      "grad_norm": 7.352923884825686,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 40930
+    },
+    {
+      "epoch": 0.9742692564029325,
+      "grad_norm": 0.3452981068441353,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 40931
+    },
+    {
+      "epoch": 0.974293059125964,
+      "grad_norm": 4.833610572797045,
+      "learning_rate": 2e-05,
+      "loss": 0.5496,
+      "step": 40932
+    },
+    {
+      "epoch": 0.9743168618489956,
+      "grad_norm": 12.933624779790177,
+      "learning_rate": 2e-05,
+      "loss": 0.6517,
+      "step": 40933
+    },
+    {
+      "epoch": 0.9743406645720271,
+      "grad_norm": 5.133588628363518,
+      "learning_rate": 2e-05,
+      "loss": 1.1022,
+      "step": 40934
+    },
+    {
+      "epoch": 0.9743644672950585,
+      "grad_norm": 0.8350247060259504,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 40935
+    },
+    {
+      "epoch": 0.97438827001809,
+      "grad_norm": 4.509684631884501,
+      "learning_rate": 2e-05,
+      "loss": 1.1323,
+      "step": 40936
+    },
+    {
+      "epoch": 0.9744120727411216,
+      "grad_norm": 9.136537427928605,
+      "learning_rate": 2e-05,
+      "loss": 0.4353,
+      "step": 40937
+    },
+    {
+      "epoch": 0.9744358754641531,
+      "grad_norm": 6.137146834974573,
+      "learning_rate": 2e-05,
+      "loss": 1.3509,
+      "step": 40938
+    },
+    {
+      "epoch": 0.9744596781871846,
+      "grad_norm": 0.27616546769735667,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 40939
+    },
+    {
+      "epoch": 0.9744834809102161,
+      "grad_norm": 0.080955460729586,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 40940
+    },
+    {
+      "epoch": 0.9745072836332477,
+      "grad_norm": 0.17288907765984154,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 40941
+    },
+    {
+      "epoch": 0.9745310863562792,
+      "grad_norm": 0.42303444657744677,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 40942
+    },
+    {
+      "epoch": 0.9745548890793106,
+      "grad_norm": 0.04787986749722594,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40943
+    },
+    {
+      "epoch": 0.9745786918023421,
+      "grad_norm": 6.314502908590885,
+      "learning_rate": 2e-05,
+      "loss": 0.5641,
+      "step": 40944
+    },
+    {
+      "epoch": 0.9746024945253737,
+      "grad_norm": 1.279206381397836,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 40945
+    },
+    {
+      "epoch": 0.9746262972484052,
+      "grad_norm": 3.376476385324613,
+      "learning_rate": 2e-05,
+      "loss": 0.6995,
+      "step": 40946
+    },
+    {
+      "epoch": 0.9746500999714367,
+      "grad_norm": 0.010446749190232852,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 40947
+    },
+    {
+      "epoch": 0.9746739026944683,
+      "grad_norm": 3.4420436922450772,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 40948
+    },
+    {
+      "epoch": 0.9746977054174998,
+      "grad_norm": 0.6341063697923316,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 40949
+    },
+    {
+      "epoch": 0.9747215081405313,
+      "grad_norm": 2.2124976806040055,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 40950
+    },
+    {
+      "epoch": 0.9747453108635628,
+      "grad_norm": 4.721672627245825,
+      "learning_rate": 2e-05,
+      "loss": 0.923,
+      "step": 40951
+    },
+    {
+      "epoch": 0.9747691135865943,
+      "grad_norm": 0.815452930516317,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 40952
+    },
+    {
+      "epoch": 0.9747929163096258,
+      "grad_norm": 2.8761984010199306,
+      "learning_rate": 2e-05,
+      "loss": 0.2724,
+      "step": 40953
+    },
+    {
+      "epoch": 0.9748167190326573,
+      "grad_norm": 0.057318248249049596,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 40954
+    },
+    {
+      "epoch": 0.9748405217556888,
+      "grad_norm": 8.844077708642653,
+      "learning_rate": 2e-05,
+      "loss": 0.8048,
+      "step": 40955
+    },
+    {
+      "epoch": 0.9748643244787204,
+      "grad_norm": 2.490426773471463,
+      "learning_rate": 2e-05,
+      "loss": 0.4398,
+      "step": 40956
+    },
+    {
+      "epoch": 0.9748881272017519,
+      "grad_norm": 0.3301323220559546,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 40957
+    },
+    {
+      "epoch": 0.9749119299247834,
+      "grad_norm": 8.756622798055508,
+      "learning_rate": 2e-05,
+      "loss": 1.3111,
+      "step": 40958
+    },
+    {
+      "epoch": 0.9749357326478149,
+      "grad_norm": 0.7573507979554117,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 40959
+    },
+    {
+      "epoch": 0.9749595353708465,
+      "grad_norm": 5.907524894252202,
+      "learning_rate": 2e-05,
+      "loss": 1.1738,
+      "step": 40960
+    },
+    {
+      "epoch": 0.9749833380938779,
+      "grad_norm": 5.976792628585573,
+      "learning_rate": 2e-05,
+      "loss": 0.0971,
+      "step": 40961
+    },
+    {
+      "epoch": 0.9750071408169094,
+      "grad_norm": 0.034673870344085804,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 40962
+    },
+    {
+      "epoch": 0.975030943539941,
+      "grad_norm": 2.4089829363531163,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 40963
+    },
+    {
+      "epoch": 0.9750547462629725,
+      "grad_norm": 0.030974088823252224,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 40964
+    },
+    {
+      "epoch": 0.975078548986004,
+      "grad_norm": 4.063024333524057,
+      "learning_rate": 2e-05,
+      "loss": 0.8602,
+      "step": 40965
+    },
+    {
+      "epoch": 0.9751023517090355,
+      "grad_norm": 2.600598132493526,
+      "learning_rate": 2e-05,
+      "loss": 0.1166,
+      "step": 40966
+    },
+    {
+      "epoch": 0.9751261544320671,
+      "grad_norm": 0.08762345561348323,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 40967
+    },
+    {
+      "epoch": 0.9751499571550986,
+      "grad_norm": 2.9419616027309994,
+      "learning_rate": 2e-05,
+      "loss": 0.3577,
+      "step": 40968
+    },
+    {
+      "epoch": 0.97517375987813,
+      "grad_norm": 2.59929694380968,
+      "learning_rate": 2e-05,
+      "loss": 0.3939,
+      "step": 40969
+    },
+    {
+      "epoch": 0.9751975626011615,
+      "grad_norm": 1.6375237137480185,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 40970
+    },
+    {
+      "epoch": 0.9752213653241931,
+      "grad_norm": 5.116637634425031,
+      "learning_rate": 2e-05,
+      "loss": 0.4791,
+      "step": 40971
+    },
+    {
+      "epoch": 0.9752451680472246,
+      "grad_norm": 3.360187431705413,
+      "learning_rate": 2e-05,
+      "loss": 0.6656,
+      "step": 40972
+    },
+    {
+      "epoch": 0.9752689707702561,
+      "grad_norm": 2.681009698827676,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 40973
+    },
+    {
+      "epoch": 0.9752927734932876,
+      "grad_norm": 12.058673596697778,
+      "learning_rate": 2e-05,
+      "loss": 0.1836,
+      "step": 40974
+    },
+    {
+      "epoch": 0.9753165762163192,
+      "grad_norm": 0.16495976370402862,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 40975
+    },
+    {
+      "epoch": 0.9753403789393507,
+      "grad_norm": 1.675907158851628,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 40976
+    },
+    {
+      "epoch": 0.9753641816623821,
+      "grad_norm": 1.1179310748671167,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 40977
+    },
+    {
+      "epoch": 0.9753879843854137,
+      "grad_norm": 3.886786765011971,
+      "learning_rate": 2e-05,
+      "loss": 0.219,
+      "step": 40978
+    },
+    {
+      "epoch": 0.9754117871084452,
+      "grad_norm": 2.7657364110815013,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 40979
+    },
+    {
+      "epoch": 0.9754355898314767,
+      "grad_norm": 2.33316261145004,
+      "learning_rate": 2e-05,
+      "loss": 0.0448,
+      "step": 40980
+    },
+    {
+      "epoch": 0.9754593925545082,
+      "grad_norm": 3.13180651338373,
+      "learning_rate": 2e-05,
+      "loss": 0.3504,
+      "step": 40981
+    },
+    {
+      "epoch": 0.9754831952775398,
+      "grad_norm": 2.3558203251303556,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 40982
+    },
+    {
+      "epoch": 0.9755069980005713,
+      "grad_norm": 0.11306604749936082,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 40983
+    },
+    {
+      "epoch": 0.9755308007236028,
+      "grad_norm": 3.159039921482881,
+      "learning_rate": 2e-05,
+      "loss": 0.5232,
+      "step": 40984
+    },
+    {
+      "epoch": 0.9755546034466342,
+      "grad_norm": 5.213691056650627,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 40985
+    },
+    {
+      "epoch": 0.9755784061696658,
+      "grad_norm": 0.6654725743154253,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 40986
+    },
+    {
+      "epoch": 0.9756022088926973,
+      "grad_norm": 0.6917264474901541,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 40987
+    },
+    {
+      "epoch": 0.9756260116157288,
+      "grad_norm": 2.2064020168303764,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 40988
+    },
+    {
+      "epoch": 0.9756498143387603,
+      "grad_norm": 3.3320164524432174,
+      "learning_rate": 2e-05,
+      "loss": 0.2105,
+      "step": 40989
+    },
+    {
+      "epoch": 0.9756736170617919,
+      "grad_norm": 3.8789867599876486,
+      "learning_rate": 2e-05,
+      "loss": 0.6285,
+      "step": 40990
+    },
+    {
+      "epoch": 0.9756974197848234,
+      "grad_norm": 3.9936891859206445,
+      "learning_rate": 2e-05,
+      "loss": 0.5205,
+      "step": 40991
+    },
+    {
+      "epoch": 0.9757212225078549,
+      "grad_norm": 0.854995217589471,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 40992
+    },
+    {
+      "epoch": 0.9757450252308865,
+      "grad_norm": 0.11206243155670399,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 40993
+    },
+    {
+      "epoch": 0.975768827953918,
+      "grad_norm": 0.6643668547014447,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 40994
+    },
+    {
+      "epoch": 0.9757926306769494,
+      "grad_norm": 8.998308320351594,
+      "learning_rate": 2e-05,
+      "loss": 1.1823,
+      "step": 40995
+    },
+    {
+      "epoch": 0.9758164333999809,
+      "grad_norm": 8.28503229910656,
+      "learning_rate": 2e-05,
+      "loss": 0.3896,
+      "step": 40996
+    },
+    {
+      "epoch": 0.9758402361230125,
+      "grad_norm": 3.057888762933789,
+      "learning_rate": 2e-05,
+      "loss": 0.5532,
+      "step": 40997
+    },
+    {
+      "epoch": 0.975864038846044,
+      "grad_norm": 3.189303040620175,
+      "learning_rate": 2e-05,
+      "loss": 0.4469,
+      "step": 40998
+    },
+    {
+      "epoch": 0.9758878415690755,
+      "grad_norm": 0.23628213115669566,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 40999
+    },
+    {
+      "epoch": 0.975911644292107,
+      "grad_norm": 3.8663217649331356,
+      "learning_rate": 2e-05,
+      "loss": 0.4972,
+      "step": 41000
+    },
+    {
+      "epoch": 0.9759354470151386,
+      "grad_norm": 4.326347241638385,
+      "learning_rate": 2e-05,
+      "loss": 0.6394,
+      "step": 41001
+    },
+    {
+      "epoch": 0.97595924973817,
+      "grad_norm": 0.1575329162097126,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41002
+    },
+    {
+      "epoch": 0.9759830524612015,
+      "grad_norm": 2.760542176004626,
+      "learning_rate": 2e-05,
+      "loss": 0.0422,
+      "step": 41003
+    },
+    {
+      "epoch": 0.976006855184233,
+      "grad_norm": 0.04884608638510966,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 41004
+    },
+    {
+      "epoch": 0.9760306579072646,
+      "grad_norm": 3.6652429033377047,
+      "learning_rate": 2e-05,
+      "loss": 0.858,
+      "step": 41005
+    },
+    {
+      "epoch": 0.9760544606302961,
+      "grad_norm": 1.7838167338917852,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 41006
+    },
+    {
+      "epoch": 0.9760782633533276,
+      "grad_norm": 1.169478732221286,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 41007
+    },
+    {
+      "epoch": 0.9761020660763592,
+      "grad_norm": 6.061276203058151,
+      "learning_rate": 2e-05,
+      "loss": 0.1211,
+      "step": 41008
+    },
+    {
+      "epoch": 0.9761258687993907,
+      "grad_norm": 0.37137837106972466,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 41009
+    },
+    {
+      "epoch": 0.9761496715224222,
+      "grad_norm": 4.78012148447131,
+      "learning_rate": 2e-05,
+      "loss": 0.7767,
+      "step": 41010
+    },
+    {
+      "epoch": 0.9761734742454536,
+      "grad_norm": 3.4978009663702094,
+      "learning_rate": 2e-05,
+      "loss": 0.5665,
+      "step": 41011
+    },
+    {
+      "epoch": 0.9761972769684852,
+      "grad_norm": 2.850991658167238,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 41012
+    },
+    {
+      "epoch": 0.9762210796915167,
+      "grad_norm": 4.078351556304818,
+      "learning_rate": 2e-05,
+      "loss": 0.4614,
+      "step": 41013
+    },
+    {
+      "epoch": 0.9762448824145482,
+      "grad_norm": 0.5437803329011477,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 41014
+    },
+    {
+      "epoch": 0.9762686851375797,
+      "grad_norm": 0.03539463190774023,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41015
+    },
+    {
+      "epoch": 0.9762924878606113,
+      "grad_norm": 2.859117672789744,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 41016
+    },
+    {
+      "epoch": 0.9763162905836428,
+      "grad_norm": 0.3099607549994699,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41017
+    },
+    {
+      "epoch": 0.9763400933066743,
+      "grad_norm": 0.24019001383008373,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 41018
+    },
+    {
+      "epoch": 0.9763638960297057,
+      "grad_norm": 0.47701744918817246,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 41019
+    },
+    {
+      "epoch": 0.9763876987527373,
+      "grad_norm": 3.611760570430754,
+      "learning_rate": 2e-05,
+      "loss": 0.2433,
+      "step": 41020
+    },
+    {
+      "epoch": 0.9764115014757688,
+      "grad_norm": 3.7693445614887686,
+      "learning_rate": 2e-05,
+      "loss": 0.5061,
+      "step": 41021
+    },
+    {
+      "epoch": 0.9764353041988003,
+      "grad_norm": 1.3331308555521595,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 41022
+    },
+    {
+      "epoch": 0.9764591069218319,
+      "grad_norm": 9.446495821988997,
+      "learning_rate": 2e-05,
+      "loss": 0.7774,
+      "step": 41023
+    },
+    {
+      "epoch": 0.9764829096448634,
+      "grad_norm": 24.506680802621847,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 41024
+    },
+    {
+      "epoch": 0.9765067123678949,
+      "grad_norm": 4.711005597349108,
+      "learning_rate": 2e-05,
+      "loss": 0.8024,
+      "step": 41025
+    },
+    {
+      "epoch": 0.9765305150909264,
+      "grad_norm": 5.409432034454823,
+      "learning_rate": 2e-05,
+      "loss": 0.6304,
+      "step": 41026
+    },
+    {
+      "epoch": 0.976554317813958,
+      "grad_norm": 4.728616881875367,
+      "learning_rate": 2e-05,
+      "loss": 0.7431,
+      "step": 41027
+    },
+    {
+      "epoch": 0.9765781205369894,
+      "grad_norm": 0.6150210212712597,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 41028
+    },
+    {
+      "epoch": 0.9766019232600209,
+      "grad_norm": 2.813175330328674,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 41029
+    },
+    {
+      "epoch": 0.9766257259830524,
+      "grad_norm": 11.63327894187123,
+      "learning_rate": 2e-05,
+      "loss": 0.1962,
+      "step": 41030
+    },
+    {
+      "epoch": 0.976649528706084,
+      "grad_norm": 0.862811236484454,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 41031
+    },
+    {
+      "epoch": 0.9766733314291155,
+      "grad_norm": 0.976850646829143,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 41032
+    },
+    {
+      "epoch": 0.976697134152147,
+      "grad_norm": 3.9399596977273044,
+      "learning_rate": 2e-05,
+      "loss": 0.4385,
+      "step": 41033
+    },
+    {
+      "epoch": 0.9767209368751785,
+      "grad_norm": 1.3010885433671786,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 41034
+    },
+    {
+      "epoch": 0.9767447395982101,
+      "grad_norm": 0.0124157731855497,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41035
+    },
+    {
+      "epoch": 0.9767685423212416,
+      "grad_norm": 4.842871573261174,
+      "learning_rate": 2e-05,
+      "loss": 0.6748,
+      "step": 41036
+    },
+    {
+      "epoch": 0.976792345044273,
+      "grad_norm": 3.108948107527775,
+      "learning_rate": 2e-05,
+      "loss": 0.3059,
+      "step": 41037
+    },
+    {
+      "epoch": 0.9768161477673046,
+      "grad_norm": 0.02065406730656688,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41038
+    },
+    {
+      "epoch": 0.9768399504903361,
+      "grad_norm": 4.5328804384286645,
+      "learning_rate": 2e-05,
+      "loss": 0.6758,
+      "step": 41039
+    },
+    {
+      "epoch": 0.9768637532133676,
+      "grad_norm": 0.18103701595802107,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41040
+    },
+    {
+      "epoch": 0.9768875559363991,
+      "grad_norm": 8.098666298113095,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 41041
+    },
+    {
+      "epoch": 0.9769113586594307,
+      "grad_norm": 1.1320352746463294,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 41042
+    },
+    {
+      "epoch": 0.9769351613824622,
+      "grad_norm": 3.925179518280964,
+      "learning_rate": 2e-05,
+      "loss": 0.7637,
+      "step": 41043
+    },
+    {
+      "epoch": 0.9769589641054937,
+      "grad_norm": 5.594233874935256,
+      "learning_rate": 2e-05,
+      "loss": 1.2156,
+      "step": 41044
+    },
+    {
+      "epoch": 0.9769827668285251,
+      "grad_norm": 7.249090342524771,
+      "learning_rate": 2e-05,
+      "loss": 0.4719,
+      "step": 41045
+    },
+    {
+      "epoch": 0.9770065695515567,
+      "grad_norm": 2.3230772229433776,
+      "learning_rate": 2e-05,
+      "loss": 0.314,
+      "step": 41046
+    },
+    {
+      "epoch": 0.9770303722745882,
+      "grad_norm": 6.043558669864889,
+      "learning_rate": 2e-05,
+      "loss": 0.7816,
+      "step": 41047
+    },
+    {
+      "epoch": 0.9770541749976197,
+      "grad_norm": 5.350594234132138,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 41048
+    },
+    {
+      "epoch": 0.9770779777206512,
+      "grad_norm": 0.07456224125941141,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41049
+    },
+    {
+      "epoch": 0.9771017804436828,
+      "grad_norm": 4.390688746725302,
+      "learning_rate": 2e-05,
+      "loss": 0.7653,
+      "step": 41050
+    },
+    {
+      "epoch": 0.9771255831667143,
+      "grad_norm": 9.288036919773013,
+      "learning_rate": 2e-05,
+      "loss": 0.5566,
+      "step": 41051
+    },
+    {
+      "epoch": 0.9771493858897458,
+      "grad_norm": 9.056999764849262,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 41052
+    },
+    {
+      "epoch": 0.9771731886127774,
+      "grad_norm": 4.064403095431501,
+      "learning_rate": 2e-05,
+      "loss": 0.5225,
+      "step": 41053
+    },
+    {
+      "epoch": 0.9771969913358088,
+      "grad_norm": 5.137293967955595,
+      "learning_rate": 2e-05,
+      "loss": 0.5046,
+      "step": 41054
+    },
+    {
+      "epoch": 0.9772207940588403,
+      "grad_norm": 4.21242368995238,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 41055
+    },
+    {
+      "epoch": 0.9772445967818718,
+      "grad_norm": 0.9459079879887214,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 41056
+    },
+    {
+      "epoch": 0.9772683995049034,
+      "grad_norm": 0.07603150585298178,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 41057
+    },
+    {
+      "epoch": 0.9772922022279349,
+      "grad_norm": 0.1723432473420142,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 41058
+    },
+    {
+      "epoch": 0.9773160049509664,
+      "grad_norm": 0.009713096579090082,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 41059
+    },
+    {
+      "epoch": 0.9773398076739979,
+      "grad_norm": 2.5974715026742548,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 41060
+    },
+    {
+      "epoch": 0.9773636103970295,
+      "grad_norm": 0.013791516974503615,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41061
+    },
+    {
+      "epoch": 0.9773874131200609,
+      "grad_norm": 5.2203826942610485,
+      "learning_rate": 2e-05,
+      "loss": 1.076,
+      "step": 41062
+    },
+    {
+      "epoch": 0.9774112158430924,
+      "grad_norm": 7.215868309703834,
+      "learning_rate": 2e-05,
+      "loss": 0.7629,
+      "step": 41063
+    },
+    {
+      "epoch": 0.9774350185661239,
+      "grad_norm": 2.751125544868662,
+      "learning_rate": 2e-05,
+      "loss": 0.4323,
+      "step": 41064
+    },
+    {
+      "epoch": 0.9774588212891555,
+      "grad_norm": 7.5301058501438565,
+      "learning_rate": 2e-05,
+      "loss": 0.1106,
+      "step": 41065
+    },
+    {
+      "epoch": 0.977482624012187,
+      "grad_norm": 2.0527280352469015,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 41066
+    },
+    {
+      "epoch": 0.9775064267352185,
+      "grad_norm": 0.14772846091458755,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41067
+    },
+    {
+      "epoch": 0.9775302294582501,
+      "grad_norm": 0.048031904884814146,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41068
+    },
+    {
+      "epoch": 0.9775540321812816,
+      "grad_norm": 0.7031976300273248,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 41069
+    },
+    {
+      "epoch": 0.977577834904313,
+      "grad_norm": 0.08886156298198947,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41070
+    },
+    {
+      "epoch": 0.9776016376273445,
+      "grad_norm": 16.14561919936151,
+      "learning_rate": 2e-05,
+      "loss": 0.7655,
+      "step": 41071
+    },
+    {
+      "epoch": 0.9776254403503761,
+      "grad_norm": 2.066915980456165,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 41072
+    },
+    {
+      "epoch": 0.9776492430734076,
+      "grad_norm": 4.126398261473037,
+      "learning_rate": 2e-05,
+      "loss": 0.1769,
+      "step": 41073
+    },
+    {
+      "epoch": 0.9776730457964391,
+      "grad_norm": 11.483066384746728,
+      "learning_rate": 2e-05,
+      "loss": 0.7011,
+      "step": 41074
+    },
+    {
+      "epoch": 0.9776968485194706,
+      "grad_norm": 0.023632632534464536,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41075
+    },
+    {
+      "epoch": 0.9777206512425022,
+      "grad_norm": 5.89950895681938,
+      "learning_rate": 2e-05,
+      "loss": 1.3827,
+      "step": 41076
+    },
+    {
+      "epoch": 0.9777444539655337,
+      "grad_norm": 6.14219003443877,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 41077
+    },
+    {
+      "epoch": 0.9777682566885652,
+      "grad_norm": 12.462738831796974,
+      "learning_rate": 2e-05,
+      "loss": 0.2179,
+      "step": 41078
+    },
+    {
+      "epoch": 0.9777920594115966,
+      "grad_norm": 4.485469001715263,
+      "learning_rate": 2e-05,
+      "loss": 0.0945,
+      "step": 41079
+    },
+    {
+      "epoch": 0.9778158621346282,
+      "grad_norm": 2.7145581014163325,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 41080
+    },
+    {
+      "epoch": 0.9778396648576597,
+      "grad_norm": 0.25879711359213575,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 41081
+    },
+    {
+      "epoch": 0.9778634675806912,
+      "grad_norm": 4.912948480881075,
+      "learning_rate": 2e-05,
+      "loss": 0.8207,
+      "step": 41082
+    },
+    {
+      "epoch": 0.9778872703037228,
+      "grad_norm": 5.5584826221642585,
+      "learning_rate": 2e-05,
+      "loss": 1.1395,
+      "step": 41083
+    },
+    {
+      "epoch": 0.9779110730267543,
+      "grad_norm": 7.565115515311699,
+      "learning_rate": 2e-05,
+      "loss": 0.8371,
+      "step": 41084
+    },
+    {
+      "epoch": 0.9779348757497858,
+      "grad_norm": 1.0599783560057743,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 41085
+    },
+    {
+      "epoch": 0.9779586784728173,
+      "grad_norm": 9.596433222169729,
+      "learning_rate": 2e-05,
+      "loss": 0.4527,
+      "step": 41086
+    },
+    {
+      "epoch": 0.9779824811958489,
+      "grad_norm": 0.34118350730426633,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 41087
+    },
+    {
+      "epoch": 0.9780062839188803,
+      "grad_norm": 0.4433797093830303,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 41088
+    },
+    {
+      "epoch": 0.9780300866419118,
+      "grad_norm": 0.37818850810202037,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 41089
+    },
+    {
+      "epoch": 0.9780538893649433,
+      "grad_norm": 0.5840202866779417,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 41090
+    },
+    {
+      "epoch": 0.9780776920879749,
+      "grad_norm": 13.074498650771083,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 41091
+    },
+    {
+      "epoch": 0.9781014948110064,
+      "grad_norm": 3.4039042219304636,
+      "learning_rate": 2e-05,
+      "loss": 0.1104,
+      "step": 41092
+    },
+    {
+      "epoch": 0.9781252975340379,
+      "grad_norm": 3.7094939128754936,
+      "learning_rate": 2e-05,
+      "loss": 0.4259,
+      "step": 41093
+    },
+    {
+      "epoch": 0.9781491002570694,
+      "grad_norm": 10.53355616894135,
+      "learning_rate": 2e-05,
+      "loss": 0.4457,
+      "step": 41094
+    },
+    {
+      "epoch": 0.978172902980101,
+      "grad_norm": 0.02219398434222015,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41095
+    },
+    {
+      "epoch": 0.9781967057031324,
+      "grad_norm": 6.622348935478576,
+      "learning_rate": 2e-05,
+      "loss": 1.6472,
+      "step": 41096
+    },
+    {
+      "epoch": 0.9782205084261639,
+      "grad_norm": 8.548328515108437,
+      "learning_rate": 2e-05,
+      "loss": 0.1722,
+      "step": 41097
+    },
+    {
+      "epoch": 0.9782443111491955,
+      "grad_norm": 9.779424625754821,
+      "learning_rate": 2e-05,
+      "loss": 0.4068,
+      "step": 41098
+    },
+    {
+      "epoch": 0.978268113872227,
+      "grad_norm": 3.279602350833204,
+      "learning_rate": 2e-05,
+      "loss": 0.1501,
+      "step": 41099
+    },
+    {
+      "epoch": 0.9782919165952585,
+      "grad_norm": 0.7724696796250922,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 41100
+    },
+    {
+      "epoch": 0.97831571931829,
+      "grad_norm": 5.000810329067238,
+      "learning_rate": 2e-05,
+      "loss": 0.4006,
+      "step": 41101
+    },
+    {
+      "epoch": 0.9783395220413216,
+      "grad_norm": 2.233198009254206,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 41102
+    },
+    {
+      "epoch": 0.9783633247643531,
+      "grad_norm": 3.5323872856597904,
+      "learning_rate": 2e-05,
+      "loss": 0.4311,
+      "step": 41103
+    },
+    {
+      "epoch": 0.9783871274873845,
+      "grad_norm": 4.5316540304607065,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 41104
+    },
+    {
+      "epoch": 0.978410930210416,
+      "grad_norm": 19.934663315852884,
+      "learning_rate": 2e-05,
+      "loss": 0.7077,
+      "step": 41105
+    },
+    {
+      "epoch": 0.9784347329334476,
+      "grad_norm": 3.8484015995418672,
+      "learning_rate": 2e-05,
+      "loss": 0.6387,
+      "step": 41106
+    },
+    {
+      "epoch": 0.9784585356564791,
+      "grad_norm": 2.3091944704018306,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 41107
+    },
+    {
+      "epoch": 0.9784823383795106,
+      "grad_norm": 2.510572296250896,
+      "learning_rate": 2e-05,
+      "loss": 0.4259,
+      "step": 41108
+    },
+    {
+      "epoch": 0.9785061411025421,
+      "grad_norm": 0.4510056387040068,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 41109
+    },
+    {
+      "epoch": 0.9785299438255737,
+      "grad_norm": 3.55413300551345,
+      "learning_rate": 2e-05,
+      "loss": 0.4815,
+      "step": 41110
+    },
+    {
+      "epoch": 0.9785537465486052,
+      "grad_norm": 11.380491897783108,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 41111
+    },
+    {
+      "epoch": 0.9785775492716366,
+      "grad_norm": 0.006989012720204055,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41112
+    },
+    {
+      "epoch": 0.9786013519946682,
+      "grad_norm": 3.7516280261390014,
+      "learning_rate": 2e-05,
+      "loss": 0.41,
+      "step": 41113
+    },
+    {
+      "epoch": 0.9786251547176997,
+      "grad_norm": 6.362182172171887,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 41114
+    },
+    {
+      "epoch": 0.9786489574407312,
+      "grad_norm": 0.09876138511084528,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41115
+    },
+    {
+      "epoch": 0.9786727601637627,
+      "grad_norm": 10.361392824994729,
+      "learning_rate": 2e-05,
+      "loss": 0.4243,
+      "step": 41116
+    },
+    {
+      "epoch": 0.9786965628867943,
+      "grad_norm": 11.080779230929378,
+      "learning_rate": 2e-05,
+      "loss": 0.3316,
+      "step": 41117
+    },
+    {
+      "epoch": 0.9787203656098258,
+      "grad_norm": 3.636425518159371,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 41118
+    },
+    {
+      "epoch": 0.9787441683328573,
+      "grad_norm": 2.5680678421404286,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 41119
+    },
+    {
+      "epoch": 0.9787679710558888,
+      "grad_norm": 14.098269833864476,
+      "learning_rate": 2e-05,
+      "loss": 0.9315,
+      "step": 41120
+    },
+    {
+      "epoch": 0.9787917737789203,
+      "grad_norm": 0.10007116989357366,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 41121
+    },
+    {
+      "epoch": 0.9788155765019518,
+      "grad_norm": 0.4552984707968152,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 41122
+    },
+    {
+      "epoch": 0.9788393792249833,
+      "grad_norm": 4.972634509327252,
+      "learning_rate": 2e-05,
+      "loss": 1.1076,
+      "step": 41123
+    },
+    {
+      "epoch": 0.9788631819480148,
+      "grad_norm": 7.579839386200773,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 41124
+    },
+    {
+      "epoch": 0.9788869846710464,
+      "grad_norm": 1.95137600260135,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 41125
+    },
+    {
+      "epoch": 0.9789107873940779,
+      "grad_norm": 3.425230827893079,
+      "learning_rate": 2e-05,
+      "loss": 0.1054,
+      "step": 41126
+    },
+    {
+      "epoch": 0.9789345901171094,
+      "grad_norm": 5.17417979710842,
+      "learning_rate": 2e-05,
+      "loss": 0.6402,
+      "step": 41127
+    },
+    {
+      "epoch": 0.978958392840141,
+      "grad_norm": 2.6121725067177355,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 41128
+    },
+    {
+      "epoch": 0.9789821955631725,
+      "grad_norm": 2.8047837029182476,
+      "learning_rate": 2e-05,
+      "loss": 0.5137,
+      "step": 41129
+    },
+    {
+      "epoch": 0.9790059982862039,
+      "grad_norm": 0.8581300858624165,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 41130
+    },
+    {
+      "epoch": 0.9790298010092354,
+      "grad_norm": 9.55689212596649,
+      "learning_rate": 2e-05,
+      "loss": 1.7348,
+      "step": 41131
+    },
+    {
+      "epoch": 0.979053603732267,
+      "grad_norm": 4.358470878118524,
+      "learning_rate": 2e-05,
+      "loss": 0.0916,
+      "step": 41132
+    },
+    {
+      "epoch": 0.9790774064552985,
+      "grad_norm": 1.876445959767177,
+      "learning_rate": 2e-05,
+      "loss": 0.0413,
+      "step": 41133
+    },
+    {
+      "epoch": 0.97910120917833,
+      "grad_norm": 2.968809387196754,
+      "learning_rate": 2e-05,
+      "loss": 0.4288,
+      "step": 41134
+    },
+    {
+      "epoch": 0.9791250119013615,
+      "grad_norm": 4.227415557716084,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 41135
+    },
+    {
+      "epoch": 0.9791488146243931,
+      "grad_norm": 2.8109091505733,
+      "learning_rate": 2e-05,
+      "loss": 0.5739,
+      "step": 41136
+    },
+    {
+      "epoch": 0.9791726173474246,
+      "grad_norm": 0.10524262904294963,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 41137
+    },
+    {
+      "epoch": 0.979196420070456,
+      "grad_norm": 0.2398450896822245,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 41138
+    },
+    {
+      "epoch": 0.9792202227934875,
+      "grad_norm": 0.008498734106793273,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41139
+    },
+    {
+      "epoch": 0.9792440255165191,
+      "grad_norm": 2.1071467939903856,
+      "learning_rate": 2e-05,
+      "loss": 0.1522,
+      "step": 41140
+    },
+    {
+      "epoch": 0.9792678282395506,
+      "grad_norm": 0.45404657268918014,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 41141
+    },
+    {
+      "epoch": 0.9792916309625821,
+      "grad_norm": 4.674425530267739,
+      "learning_rate": 2e-05,
+      "loss": 0.6429,
+      "step": 41142
+    },
+    {
+      "epoch": 0.9793154336856136,
+      "grad_norm": 4.2691956778377715,
+      "learning_rate": 2e-05,
+      "loss": 0.8914,
+      "step": 41143
+    },
+    {
+      "epoch": 0.9793392364086452,
+      "grad_norm": 4.537905055307135,
+      "learning_rate": 2e-05,
+      "loss": 0.1082,
+      "step": 41144
+    },
+    {
+      "epoch": 0.9793630391316767,
+      "grad_norm": 5.745429929733953,
+      "learning_rate": 2e-05,
+      "loss": 0.9964,
+      "step": 41145
+    },
+    {
+      "epoch": 0.9793868418547081,
+      "grad_norm": 5.20691469337401,
+      "learning_rate": 2e-05,
+      "loss": 0.1023,
+      "step": 41146
+    },
+    {
+      "epoch": 0.9794106445777397,
+      "grad_norm": 0.46538769583929274,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 41147
+    },
+    {
+      "epoch": 0.9794344473007712,
+      "grad_norm": 9.464020176472372,
+      "learning_rate": 2e-05,
+      "loss": 0.1587,
+      "step": 41148
+    },
+    {
+      "epoch": 0.9794582500238027,
+      "grad_norm": 4.212217996427218,
+      "learning_rate": 2e-05,
+      "loss": 0.4699,
+      "step": 41149
+    },
+    {
+      "epoch": 0.9794820527468342,
+      "grad_norm": 6.2543650735367535,
+      "learning_rate": 2e-05,
+      "loss": 0.5802,
+      "step": 41150
+    },
+    {
+      "epoch": 0.9795058554698658,
+      "grad_norm": 4.4001627504014875,
+      "learning_rate": 2e-05,
+      "loss": 0.6193,
+      "step": 41151
+    },
+    {
+      "epoch": 0.9795296581928973,
+      "grad_norm": 16.492628873516484,
+      "learning_rate": 2e-05,
+      "loss": 0.6936,
+      "step": 41152
+    },
+    {
+      "epoch": 0.9795534609159288,
+      "grad_norm": 1.7750402878867961,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 41153
+    },
+    {
+      "epoch": 0.9795772636389602,
+      "grad_norm": 4.773818183524567,
+      "learning_rate": 2e-05,
+      "loss": 0.907,
+      "step": 41154
+    },
+    {
+      "epoch": 0.9796010663619918,
+      "grad_norm": 2.281098058108198,
+      "learning_rate": 2e-05,
+      "loss": 0.465,
+      "step": 41155
+    },
+    {
+      "epoch": 0.9796248690850233,
+      "grad_norm": 5.126221605342948,
+      "learning_rate": 2e-05,
+      "loss": 1.3015,
+      "step": 41156
+    },
+    {
+      "epoch": 0.9796486718080548,
+      "grad_norm": 4.076874975376378,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 41157
+    },
+    {
+      "epoch": 0.9796724745310863,
+      "grad_norm": 0.17245310931101449,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 41158
+    },
+    {
+      "epoch": 0.9796962772541179,
+      "grad_norm": 4.382739963130859,
+      "learning_rate": 2e-05,
+      "loss": 1.3438,
+      "step": 41159
+    },
+    {
+      "epoch": 0.9797200799771494,
+      "grad_norm": 3.834289245514436,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 41160
+    },
+    {
+      "epoch": 0.9797438827001809,
+      "grad_norm": 13.233120166939976,
+      "learning_rate": 2e-05,
+      "loss": 0.7655,
+      "step": 41161
+    },
+    {
+      "epoch": 0.9797676854232125,
+      "grad_norm": 2.6008062469192774,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 41162
+    },
+    {
+      "epoch": 0.979791488146244,
+      "grad_norm": 3.8178278989888463,
+      "learning_rate": 2e-05,
+      "loss": 0.7796,
+      "step": 41163
+    },
+    {
+      "epoch": 0.9798152908692754,
+      "grad_norm": 0.2915661974546229,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 41164
+    },
+    {
+      "epoch": 0.9798390935923069,
+      "grad_norm": 5.3523752787191965,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 41165
+    },
+    {
+      "epoch": 0.9798628963153385,
+      "grad_norm": 0.02484660225207502,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41166
+    },
+    {
+      "epoch": 0.97988669903837,
+      "grad_norm": 0.48028736630716734,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 41167
+    },
+    {
+      "epoch": 0.9799105017614015,
+      "grad_norm": 1.240940543713073,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 41168
+    },
+    {
+      "epoch": 0.979934304484433,
+      "grad_norm": 4.92611153842489,
+      "learning_rate": 2e-05,
+      "loss": 0.1385,
+      "step": 41169
+    },
+    {
+      "epoch": 0.9799581072074646,
+      "grad_norm": 3.830350679301771,
+      "learning_rate": 2e-05,
+      "loss": 0.5199,
+      "step": 41170
+    },
+    {
+      "epoch": 0.979981909930496,
+      "grad_norm": 0.7985524390727916,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 41171
+    },
+    {
+      "epoch": 0.9800057126535275,
+      "grad_norm": 0.10844164408193409,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 41172
+    },
+    {
+      "epoch": 0.980029515376559,
+      "grad_norm": 4.666637617203303,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 41173
+    },
+    {
+      "epoch": 0.9800533180995906,
+      "grad_norm": 0.22639207200562542,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 41174
+    },
+    {
+      "epoch": 0.9800771208226221,
+      "grad_norm": 2.3977646152844563,
+      "learning_rate": 2e-05,
+      "loss": 0.4303,
+      "step": 41175
+    },
+    {
+      "epoch": 0.9801009235456536,
+      "grad_norm": 6.053175387313199,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 41176
+    },
+    {
+      "epoch": 0.9801247262686852,
+      "grad_norm": 6.405771380688554,
+      "learning_rate": 2e-05,
+      "loss": 0.8286,
+      "step": 41177
+    },
+    {
+      "epoch": 0.9801485289917167,
+      "grad_norm": 5.373984680684259,
+      "learning_rate": 2e-05,
+      "loss": 0.208,
+      "step": 41178
+    },
+    {
+      "epoch": 0.9801723317147482,
+      "grad_norm": 7.801742742695353,
+      "learning_rate": 2e-05,
+      "loss": 0.2095,
+      "step": 41179
+    },
+    {
+      "epoch": 0.9801961344377796,
+      "grad_norm": 0.5669359950551535,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 41180
+    },
+    {
+      "epoch": 0.9802199371608112,
+      "grad_norm": 3.6629486635330704,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 41181
+    },
+    {
+      "epoch": 0.9802437398838427,
+      "grad_norm": 4.52218976699271,
+      "learning_rate": 2e-05,
+      "loss": 1.1178,
+      "step": 41182
+    },
+    {
+      "epoch": 0.9802675426068742,
+      "grad_norm": 0.4083805250025527,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 41183
+    },
+    {
+      "epoch": 0.9802913453299057,
+      "grad_norm": 6.357691830014282,
+      "learning_rate": 2e-05,
+      "loss": 0.642,
+      "step": 41184
+    },
+    {
+      "epoch": 0.9803151480529373,
+      "grad_norm": 2.9396139247420323,
+      "learning_rate": 2e-05,
+      "loss": 0.5618,
+      "step": 41185
+    },
+    {
+      "epoch": 0.9803389507759688,
+      "grad_norm": 0.4633316845575743,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 41186
+    },
+    {
+      "epoch": 0.9803627534990003,
+      "grad_norm": 2.8616495114965437,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 41187
+    },
+    {
+      "epoch": 0.9803865562220317,
+      "grad_norm": 4.578587276781393,
+      "learning_rate": 2e-05,
+      "loss": 1.0514,
+      "step": 41188
+    },
+    {
+      "epoch": 0.9804103589450633,
+      "grad_norm": 3.009740297799812,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 41189
+    },
+    {
+      "epoch": 0.9804341616680948,
+      "grad_norm": 10.68797122578404,
+      "learning_rate": 2e-05,
+      "loss": 1.0258,
+      "step": 41190
+    },
+    {
+      "epoch": 0.9804579643911263,
+      "grad_norm": 2.1214758590898737,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 41191
+    },
+    {
+      "epoch": 0.9804817671141579,
+      "grad_norm": 5.3358790740141275,
+      "learning_rate": 2e-05,
+      "loss": 0.8348,
+      "step": 41192
+    },
+    {
+      "epoch": 0.9805055698371894,
+      "grad_norm": 7.734068583472031,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 41193
+    },
+    {
+      "epoch": 0.9805293725602209,
+      "grad_norm": 3.7106467382052912,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 41194
+    },
+    {
+      "epoch": 0.9805531752832524,
+      "grad_norm": 4.301401887456353,
+      "learning_rate": 2e-05,
+      "loss": 0.6013,
+      "step": 41195
+    },
+    {
+      "epoch": 0.980576978006284,
+      "grad_norm": 7.641560180787623,
+      "learning_rate": 2e-05,
+      "loss": 1.0006,
+      "step": 41196
+    },
+    {
+      "epoch": 0.9806007807293154,
+      "grad_norm": 0.010579657812067165,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41197
+    },
+    {
+      "epoch": 0.9806245834523469,
+      "grad_norm": 0.12832580780504163,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41198
+    },
+    {
+      "epoch": 0.9806483861753784,
+      "grad_norm": 6.520030517374747,
+      "learning_rate": 2e-05,
+      "loss": 0.9263,
+      "step": 41199
+    },
+    {
+      "epoch": 0.98067218889841,
+      "grad_norm": 0.5994133251364787,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 41200
+    },
+    {
+      "epoch": 0.9806959916214415,
+      "grad_norm": 2.865630277329166,
+      "learning_rate": 2e-05,
+      "loss": 0.4158,
+      "step": 41201
+    },
+    {
+      "epoch": 0.980719794344473,
+      "grad_norm": 1.1321298960776889,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 41202
+    },
+    {
+      "epoch": 0.9807435970675045,
+      "grad_norm": 8.053836287979436,
+      "learning_rate": 2e-05,
+      "loss": 0.8833,
+      "step": 41203
+    },
+    {
+      "epoch": 0.9807673997905361,
+      "grad_norm": 9.8181338297843,
+      "learning_rate": 2e-05,
+      "loss": 0.7834,
+      "step": 41204
+    },
+    {
+      "epoch": 0.9807912025135676,
+      "grad_norm": 0.3541699907572282,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 41205
+    },
+    {
+      "epoch": 0.980815005236599,
+      "grad_norm": 0.25046454103180066,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 41206
+    },
+    {
+      "epoch": 0.9808388079596306,
+      "grad_norm": 2.9739009099345233,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 41207
+    },
+    {
+      "epoch": 0.9808626106826621,
+      "grad_norm": 4.9269173875190955,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 41208
+    },
+    {
+      "epoch": 0.9808864134056936,
+      "grad_norm": 2.8500754645578787,
+      "learning_rate": 2e-05,
+      "loss": 0.2898,
+      "step": 41209
+    },
+    {
+      "epoch": 0.9809102161287251,
+      "grad_norm": 9.80290354546484,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 41210
+    },
+    {
+      "epoch": 0.9809340188517567,
+      "grad_norm": 5.732813266985021,
+      "learning_rate": 2e-05,
+      "loss": 0.5927,
+      "step": 41211
+    },
+    {
+      "epoch": 0.9809578215747882,
+      "grad_norm": 10.927003826984661,
+      "learning_rate": 2e-05,
+      "loss": 0.8709,
+      "step": 41212
+    },
+    {
+      "epoch": 0.9809816242978197,
+      "grad_norm": 1.297614384777295,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 41213
+    },
+    {
+      "epoch": 0.9810054270208511,
+      "grad_norm": 3.2464345224977373,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 41214
+    },
+    {
+      "epoch": 0.9810292297438827,
+      "grad_norm": 3.1308797071664083,
+      "learning_rate": 2e-05,
+      "loss": 0.2641,
+      "step": 41215
+    },
+    {
+      "epoch": 0.9810530324669142,
+      "grad_norm": 2.0146278039585805,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 41216
+    },
+    {
+      "epoch": 0.9810768351899457,
+      "grad_norm": 0.14302814830346566,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 41217
+    },
+    {
+      "epoch": 0.9811006379129772,
+      "grad_norm": 3.724052871627458,
+      "learning_rate": 2e-05,
+      "loss": 0.5389,
+      "step": 41218
+    },
+    {
+      "epoch": 0.9811244406360088,
+      "grad_norm": 0.2690098956814066,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 41219
+    },
+    {
+      "epoch": 0.9811482433590403,
+      "grad_norm": 1.9041079818483477,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 41220
+    },
+    {
+      "epoch": 0.9811720460820718,
+      "grad_norm": 0.4663548423776555,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 41221
+    },
+    {
+      "epoch": 0.9811958488051034,
+      "grad_norm": 3.0726343097081545,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 41222
+    },
+    {
+      "epoch": 0.9812196515281348,
+      "grad_norm": 3.271297353202286,
+      "learning_rate": 2e-05,
+      "loss": 0.3479,
+      "step": 41223
+    },
+    {
+      "epoch": 0.9812434542511663,
+      "grad_norm": 16.161047183483902,
+      "learning_rate": 2e-05,
+      "loss": 1.2938,
+      "step": 41224
+    },
+    {
+      "epoch": 0.9812672569741978,
+      "grad_norm": 1.5703892872727998,
+      "learning_rate": 2e-05,
+      "loss": 0.0255,
+      "step": 41225
+    },
+    {
+      "epoch": 0.9812910596972294,
+      "grad_norm": 5.190503007275341,
+      "learning_rate": 2e-05,
+      "loss": 0.3567,
+      "step": 41226
+    },
+    {
+      "epoch": 0.9813148624202609,
+      "grad_norm": 0.06189411858589,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 41227
+    },
+    {
+      "epoch": 0.9813386651432924,
+      "grad_norm": 3.3247704251591235,
+      "learning_rate": 2e-05,
+      "loss": 0.6305,
+      "step": 41228
+    },
+    {
+      "epoch": 0.9813624678663239,
+      "grad_norm": 6.297670704262266,
+      "learning_rate": 2e-05,
+      "loss": 0.2607,
+      "step": 41229
+    },
+    {
+      "epoch": 0.9813862705893555,
+      "grad_norm": 14.236281698981488,
+      "learning_rate": 2e-05,
+      "loss": 0.8764,
+      "step": 41230
+    },
+    {
+      "epoch": 0.9814100733123869,
+      "grad_norm": 2.1214180801590388,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 41231
+    },
+    {
+      "epoch": 0.9814338760354184,
+      "grad_norm": 2.857850452596521,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 41232
+    },
+    {
+      "epoch": 0.9814576787584499,
+      "grad_norm": 0.9495539045122627,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 41233
+    },
+    {
+      "epoch": 0.9814814814814815,
+      "grad_norm": 9.658239135513972,
+      "learning_rate": 2e-05,
+      "loss": 0.4787,
+      "step": 41234
+    },
+    {
+      "epoch": 0.981505284204513,
+      "grad_norm": 0.9577514668899989,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 41235
+    },
+    {
+      "epoch": 0.9815290869275445,
+      "grad_norm": 0.5303769350328409,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 41236
+    },
+    {
+      "epoch": 0.9815528896505761,
+      "grad_norm": 1.6509749765406043,
+      "learning_rate": 2e-05,
+      "loss": 0.0382,
+      "step": 41237
+    },
+    {
+      "epoch": 0.9815766923736076,
+      "grad_norm": 0.6398042055637306,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 41238
+    },
+    {
+      "epoch": 0.981600495096639,
+      "grad_norm": 4.79398172750628,
+      "learning_rate": 2e-05,
+      "loss": 0.1942,
+      "step": 41239
+    },
+    {
+      "epoch": 0.9816242978196705,
+      "grad_norm": 8.889462157821768,
+      "learning_rate": 2e-05,
+      "loss": 1.2046,
+      "step": 41240
+    },
+    {
+      "epoch": 0.9816481005427021,
+      "grad_norm": 13.93677160930207,
+      "learning_rate": 2e-05,
+      "loss": 1.1539,
+      "step": 41241
+    },
+    {
+      "epoch": 0.9816719032657336,
+      "grad_norm": 4.681855882081594,
+      "learning_rate": 2e-05,
+      "loss": 0.9821,
+      "step": 41242
+    },
+    {
+      "epoch": 0.9816957059887651,
+      "grad_norm": 0.08842445555944019,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 41243
+    },
+    {
+      "epoch": 0.9817195087117966,
+      "grad_norm": 5.422915836140502,
+      "learning_rate": 2e-05,
+      "loss": 1.4009,
+      "step": 41244
+    },
+    {
+      "epoch": 0.9817433114348282,
+      "grad_norm": 0.3587141740061314,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 41245
+    },
+    {
+      "epoch": 0.9817671141578597,
+      "grad_norm": 3.0112762780699933,
+      "learning_rate": 2e-05,
+      "loss": 0.5047,
+      "step": 41246
+    },
+    {
+      "epoch": 0.9817909168808912,
+      "grad_norm": 7.143748674355789,
+      "learning_rate": 2e-05,
+      "loss": 0.8245,
+      "step": 41247
+    },
+    {
+      "epoch": 0.9818147196039226,
+      "grad_norm": 0.05234578888290992,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 41248
+    },
+    {
+      "epoch": 0.9818385223269542,
+      "grad_norm": 8.438873690606341,
+      "learning_rate": 2e-05,
+      "loss": 0.7104,
+      "step": 41249
+    },
+    {
+      "epoch": 0.9818623250499857,
+      "grad_norm": 8.324974993280854,
+      "learning_rate": 2e-05,
+      "loss": 0.591,
+      "step": 41250
+    },
+    {
+      "epoch": 0.9818861277730172,
+      "grad_norm": 0.20556588002273363,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 41251
+    },
+    {
+      "epoch": 0.9819099304960488,
+      "grad_norm": 2.3609591455046255,
+      "learning_rate": 2e-05,
+      "loss": 0.2729,
+      "step": 41252
+    },
+    {
+      "epoch": 0.9819337332190803,
+      "grad_norm": 4.04970922173363,
+      "learning_rate": 2e-05,
+      "loss": 0.5248,
+      "step": 41253
+    },
+    {
+      "epoch": 0.9819575359421118,
+      "grad_norm": 0.14116540480913167,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 41254
+    },
+    {
+      "epoch": 0.9819813386651433,
+      "grad_norm": 2.24477125725844,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 41255
+    },
+    {
+      "epoch": 0.9820051413881749,
+      "grad_norm": 0.05009118591137081,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41256
+    },
+    {
+      "epoch": 0.9820289441112063,
+      "grad_norm": 0.0485636726210208,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41257
+    },
+    {
+      "epoch": 0.9820527468342378,
+      "grad_norm": 2.3670981513197114,
+      "learning_rate": 2e-05,
+      "loss": 0.0235,
+      "step": 41258
+    },
+    {
+      "epoch": 0.9820765495572693,
+      "grad_norm": 4.356216499801887,
+      "learning_rate": 2e-05,
+      "loss": 0.0937,
+      "step": 41259
+    },
+    {
+      "epoch": 0.9821003522803009,
+      "grad_norm": 0.2993904647363732,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 41260
+    },
+    {
+      "epoch": 0.9821241550033324,
+      "grad_norm": 0.033820984045684795,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41261
+    },
+    {
+      "epoch": 0.9821479577263639,
+      "grad_norm": 6.385290860253634,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 41262
+    },
+    {
+      "epoch": 0.9821717604493954,
+      "grad_norm": 3.3732622944295967,
+      "learning_rate": 2e-05,
+      "loss": 0.5652,
+      "step": 41263
+    },
+    {
+      "epoch": 0.982195563172427,
+      "grad_norm": 7.537619989872585,
+      "learning_rate": 2e-05,
+      "loss": 0.8937,
+      "step": 41264
+    },
+    {
+      "epoch": 0.9822193658954584,
+      "grad_norm": 0.9358947101538793,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 41265
+    },
+    {
+      "epoch": 0.9822431686184899,
+      "grad_norm": 0.039488420984379005,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41266
+    },
+    {
+      "epoch": 0.9822669713415215,
+      "grad_norm": 1.1503486213308864,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 41267
+    },
+    {
+      "epoch": 0.982290774064553,
+      "grad_norm": 9.502370473681967,
+      "learning_rate": 2e-05,
+      "loss": 0.8503,
+      "step": 41268
+    },
+    {
+      "epoch": 0.9823145767875845,
+      "grad_norm": 3.3784951285278244,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 41269
+    },
+    {
+      "epoch": 0.982338379510616,
+      "grad_norm": 0.4966886478390716,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 41270
+    },
+    {
+      "epoch": 0.9823621822336476,
+      "grad_norm": 1.5548392368686226,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 41271
+    },
+    {
+      "epoch": 0.9823859849566791,
+      "grad_norm": 3.2945736566701447,
+      "learning_rate": 2e-05,
+      "loss": 0.6467,
+      "step": 41272
+    },
+    {
+      "epoch": 0.9824097876797105,
+      "grad_norm": 2.882486096676787,
+      "learning_rate": 2e-05,
+      "loss": 0.4207,
+      "step": 41273
+    },
+    {
+      "epoch": 0.982433590402742,
+      "grad_norm": 3.113876612636899,
+      "learning_rate": 2e-05,
+      "loss": 0.0973,
+      "step": 41274
+    },
+    {
+      "epoch": 0.9824573931257736,
+      "grad_norm": 2.097181855143009,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 41275
+    },
+    {
+      "epoch": 0.9824811958488051,
+      "grad_norm": 0.13165523161760379,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 41276
+    },
+    {
+      "epoch": 0.9825049985718366,
+      "grad_norm": 1.497228791031796,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 41277
+    },
+    {
+      "epoch": 0.9825288012948681,
+      "grad_norm": 5.458407067465739,
+      "learning_rate": 2e-05,
+      "loss": 0.8019,
+      "step": 41278
+    },
+    {
+      "epoch": 0.9825526040178997,
+      "grad_norm": 4.389888154233075,
+      "learning_rate": 2e-05,
+      "loss": 0.4631,
+      "step": 41279
+    },
+    {
+      "epoch": 0.9825764067409312,
+      "grad_norm": 1.581378342992967,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 41280
+    },
+    {
+      "epoch": 0.9826002094639626,
+      "grad_norm": 4.086100264032696,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 41281
+    },
+    {
+      "epoch": 0.9826240121869942,
+      "grad_norm": 6.762357536070735,
+      "learning_rate": 2e-05,
+      "loss": 0.2983,
+      "step": 41282
+    },
+    {
+      "epoch": 0.9826478149100257,
+      "grad_norm": 8.730211896398641,
+      "learning_rate": 2e-05,
+      "loss": 0.2007,
+      "step": 41283
+    },
+    {
+      "epoch": 0.9826716176330572,
+      "grad_norm": 4.946069215090498,
+      "learning_rate": 2e-05,
+      "loss": 1.1754,
+      "step": 41284
+    },
+    {
+      "epoch": 0.9826954203560887,
+      "grad_norm": 0.49188915939207606,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 41285
+    },
+    {
+      "epoch": 0.9827192230791203,
+      "grad_norm": 2.949440556269933,
+      "learning_rate": 2e-05,
+      "loss": 0.1198,
+      "step": 41286
+    },
+    {
+      "epoch": 0.9827430258021518,
+      "grad_norm": 0.2959261602491427,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 41287
+    },
+    {
+      "epoch": 0.9827668285251833,
+      "grad_norm": 4.301368297705719,
+      "learning_rate": 2e-05,
+      "loss": 0.1207,
+      "step": 41288
+    },
+    {
+      "epoch": 0.9827906312482148,
+      "grad_norm": 10.887747378251728,
+      "learning_rate": 2e-05,
+      "loss": 0.9312,
+      "step": 41289
+    },
+    {
+      "epoch": 0.9828144339712463,
+      "grad_norm": 4.308540818156427,
+      "learning_rate": 2e-05,
+      "loss": 0.1049,
+      "step": 41290
+    },
+    {
+      "epoch": 0.9828382366942778,
+      "grad_norm": 0.28060153663073073,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 41291
+    },
+    {
+      "epoch": 0.9828620394173093,
+      "grad_norm": 2.7934491000076656,
+      "learning_rate": 2e-05,
+      "loss": 0.0741,
+      "step": 41292
+    },
+    {
+      "epoch": 0.9828858421403408,
+      "grad_norm": 0.9834941743000782,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 41293
+    },
+    {
+      "epoch": 0.9829096448633724,
+      "grad_norm": 1.4408007768937727,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 41294
+    },
+    {
+      "epoch": 0.9829334475864039,
+      "grad_norm": 8.110683838619776,
+      "learning_rate": 2e-05,
+      "loss": 0.0606,
+      "step": 41295
+    },
+    {
+      "epoch": 0.9829572503094354,
+      "grad_norm": 3.0774892482699157,
+      "learning_rate": 2e-05,
+      "loss": 0.5484,
+      "step": 41296
+    },
+    {
+      "epoch": 0.982981053032467,
+      "grad_norm": 3.39139675602791,
+      "learning_rate": 2e-05,
+      "loss": 0.6063,
+      "step": 41297
+    },
+    {
+      "epoch": 0.9830048557554985,
+      "grad_norm": 0.3023665716253122,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 41298
+    },
+    {
+      "epoch": 0.9830286584785299,
+      "grad_norm": 0.04163264673935466,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41299
+    },
+    {
+      "epoch": 0.9830524612015614,
+      "grad_norm": 13.493190931294677,
+      "learning_rate": 2e-05,
+      "loss": 0.2452,
+      "step": 41300
+    },
+    {
+      "epoch": 0.983076263924593,
+      "grad_norm": 1.1529330952964711,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 41301
+    },
+    {
+      "epoch": 0.9831000666476245,
+      "grad_norm": 1.1278472456829458,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 41302
+    },
+    {
+      "epoch": 0.983123869370656,
+      "grad_norm": 0.048338027848040054,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41303
+    },
+    {
+      "epoch": 0.9831476720936875,
+      "grad_norm": 4.883561802095306,
+      "learning_rate": 2e-05,
+      "loss": 0.1107,
+      "step": 41304
+    },
+    {
+      "epoch": 0.9831714748167191,
+      "grad_norm": 0.9967771038218902,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 41305
+    },
+    {
+      "epoch": 0.9831952775397506,
+      "grad_norm": 12.879368858625345,
+      "learning_rate": 2e-05,
+      "loss": 0.5335,
+      "step": 41306
+    },
+    {
+      "epoch": 0.983219080262782,
+      "grad_norm": 4.775375164753277,
+      "learning_rate": 2e-05,
+      "loss": 1.2343,
+      "step": 41307
+    },
+    {
+      "epoch": 0.9832428829858135,
+      "grad_norm": 3.94645329236292,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 41308
+    },
+    {
+      "epoch": 0.9832666857088451,
+      "grad_norm": 5.111869265078851,
+      "learning_rate": 2e-05,
+      "loss": 0.5001,
+      "step": 41309
+    },
+    {
+      "epoch": 0.9832904884318766,
+      "grad_norm": 10.534622393627295,
+      "learning_rate": 2e-05,
+      "loss": 0.3826,
+      "step": 41310
+    },
+    {
+      "epoch": 0.9833142911549081,
+      "grad_norm": 4.627322483564966,
+      "learning_rate": 2e-05,
+      "loss": 0.0561,
+      "step": 41311
+    },
+    {
+      "epoch": 0.9833380938779397,
+      "grad_norm": 2.284342333102071,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 41312
+    },
+    {
+      "epoch": 0.9833618966009712,
+      "grad_norm": 0.15260864624666248,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 41313
+    },
+    {
+      "epoch": 0.9833856993240027,
+      "grad_norm": 1.2977973268791814,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 41314
+    },
+    {
+      "epoch": 0.9834095020470341,
+      "grad_norm": 2.9622060078698182,
+      "learning_rate": 2e-05,
+      "loss": 0.4831,
+      "step": 41315
+    },
+    {
+      "epoch": 0.9834333047700657,
+      "grad_norm": 5.1421101460162415,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 41316
+    },
+    {
+      "epoch": 0.9834571074930972,
+      "grad_norm": 10.774982886190466,
+      "learning_rate": 2e-05,
+      "loss": 0.1758,
+      "step": 41317
+    },
+    {
+      "epoch": 0.9834809102161287,
+      "grad_norm": 7.5074218133759,
+      "learning_rate": 2e-05,
+      "loss": 0.227,
+      "step": 41318
+    },
+    {
+      "epoch": 0.9835047129391602,
+      "grad_norm": 0.018246297470450187,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41319
+    },
+    {
+      "epoch": 0.9835285156621918,
+      "grad_norm": 0.07727906950012212,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41320
+    },
+    {
+      "epoch": 0.9835523183852233,
+      "grad_norm": 6.762989434137754,
+      "learning_rate": 2e-05,
+      "loss": 0.3298,
+      "step": 41321
+    },
+    {
+      "epoch": 0.9835761211082548,
+      "grad_norm": 2.69297128284704,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 41322
+    },
+    {
+      "epoch": 0.9835999238312862,
+      "grad_norm": 6.239368016106249,
+      "learning_rate": 2e-05,
+      "loss": 0.2244,
+      "step": 41323
+    },
+    {
+      "epoch": 0.9836237265543178,
+      "grad_norm": 7.997050698108368,
+      "learning_rate": 2e-05,
+      "loss": 0.2046,
+      "step": 41324
+    },
+    {
+      "epoch": 0.9836475292773493,
+      "grad_norm": 8.56695652877044,
+      "learning_rate": 2e-05,
+      "loss": 0.3792,
+      "step": 41325
+    },
+    {
+      "epoch": 0.9836713320003808,
+      "grad_norm": 7.5380548740585125,
+      "learning_rate": 2e-05,
+      "loss": 0.343,
+      "step": 41326
+    },
+    {
+      "epoch": 0.9836951347234124,
+      "grad_norm": 0.028285390791052856,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41327
+    },
+    {
+      "epoch": 0.9837189374464439,
+      "grad_norm": 0.3768658720171368,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 41328
+    },
+    {
+      "epoch": 0.9837427401694754,
+      "grad_norm": 0.015861422123963933,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41329
+    },
+    {
+      "epoch": 0.9837665428925069,
+      "grad_norm": 3.8941277142729924,
+      "learning_rate": 2e-05,
+      "loss": 0.8868,
+      "step": 41330
+    },
+    {
+      "epoch": 0.9837903456155385,
+      "grad_norm": 3.1755548272870446,
+      "learning_rate": 2e-05,
+      "loss": 0.1763,
+      "step": 41331
+    },
+    {
+      "epoch": 0.98381414833857,
+      "grad_norm": 0.01565052536970901,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41332
+    },
+    {
+      "epoch": 0.9838379510616014,
+      "grad_norm": 3.95302726915579,
+      "learning_rate": 2e-05,
+      "loss": 0.4684,
+      "step": 41333
+    },
+    {
+      "epoch": 0.9838617537846329,
+      "grad_norm": 0.07993583437983579,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41334
+    },
+    {
+      "epoch": 0.9838855565076645,
+      "grad_norm": 0.059524370171417026,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41335
+    },
+    {
+      "epoch": 0.983909359230696,
+      "grad_norm": 4.802904242091441,
+      "learning_rate": 2e-05,
+      "loss": 1.0387,
+      "step": 41336
+    },
+    {
+      "epoch": 0.9839331619537275,
+      "grad_norm": 0.2815314715392743,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 41337
+    },
+    {
+      "epoch": 0.983956964676759,
+      "grad_norm": 19.614691143191788,
+      "learning_rate": 2e-05,
+      "loss": 0.8864,
+      "step": 41338
+    },
+    {
+      "epoch": 0.9839807673997906,
+      "grad_norm": 3.993306001524258,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 41339
+    },
+    {
+      "epoch": 0.984004570122822,
+      "grad_norm": 4.087032124317565,
+      "learning_rate": 2e-05,
+      "loss": 0.5861,
+      "step": 41340
+    },
+    {
+      "epoch": 0.9840283728458535,
+      "grad_norm": 7.709767880756911,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 41341
+    },
+    {
+      "epoch": 0.9840521755688851,
+      "grad_norm": 3.8939258300831256,
+      "learning_rate": 2e-05,
+      "loss": 0.4661,
+      "step": 41342
+    },
+    {
+      "epoch": 0.9840759782919166,
+      "grad_norm": 1.5652071650076627,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 41343
+    },
+    {
+      "epoch": 0.9840997810149481,
+      "grad_norm": 3.0463378187582206,
+      "learning_rate": 2e-05,
+      "loss": 0.55,
+      "step": 41344
+    },
+    {
+      "epoch": 0.9841235837379796,
+      "grad_norm": 2.70261707169206,
+      "learning_rate": 2e-05,
+      "loss": 0.4555,
+      "step": 41345
+    },
+    {
+      "epoch": 0.9841473864610112,
+      "grad_norm": 4.6852067617775095,
+      "learning_rate": 2e-05,
+      "loss": 0.2162,
+      "step": 41346
+    },
+    {
+      "epoch": 0.9841711891840427,
+      "grad_norm": 5.9392376911387545,
+      "learning_rate": 2e-05,
+      "loss": 0.067,
+      "step": 41347
+    },
+    {
+      "epoch": 0.9841949919070742,
+      "grad_norm": 1.436370623681612,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 41348
+    },
+    {
+      "epoch": 0.9842187946301056,
+      "grad_norm": 1.6401993490117583,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 41349
+    },
+    {
+      "epoch": 0.9842425973531372,
+      "grad_norm": 3.3101162001227666,
+      "learning_rate": 2e-05,
+      "loss": 0.704,
+      "step": 41350
+    },
+    {
+      "epoch": 0.9842664000761687,
+      "grad_norm": 3.2570466293421636,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 41351
+    },
+    {
+      "epoch": 0.9842902027992002,
+      "grad_norm": 3.122673732831315,
+      "learning_rate": 2e-05,
+      "loss": 0.677,
+      "step": 41352
+    },
+    {
+      "epoch": 0.9843140055222317,
+      "grad_norm": 3.6448810348979674,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 41353
+    },
+    {
+      "epoch": 0.9843378082452633,
+      "grad_norm": 0.30700930129323656,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 41354
+    },
+    {
+      "epoch": 0.9843616109682948,
+      "grad_norm": 4.959881138492779,
+      "learning_rate": 2e-05,
+      "loss": 0.7911,
+      "step": 41355
+    },
+    {
+      "epoch": 0.9843854136913263,
+      "grad_norm": 13.37893896909871,
+      "learning_rate": 2e-05,
+      "loss": 0.8168,
+      "step": 41356
+    },
+    {
+      "epoch": 0.9844092164143579,
+      "grad_norm": 4.520849979144643,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 41357
+    },
+    {
+      "epoch": 0.9844330191373893,
+      "grad_norm": 0.15250470824191636,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41358
+    },
+    {
+      "epoch": 0.9844568218604208,
+      "grad_norm": 2.394881463257428,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 41359
+    },
+    {
+      "epoch": 0.9844806245834523,
+      "grad_norm": 0.0957960870053194,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41360
+    },
+    {
+      "epoch": 0.9845044273064839,
+      "grad_norm": 0.3365617259076645,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 41361
+    },
+    {
+      "epoch": 0.9845282300295154,
+      "grad_norm": 3.2554383432607423,
+      "learning_rate": 2e-05,
+      "loss": 0.5402,
+      "step": 41362
+    },
+    {
+      "epoch": 0.9845520327525469,
+      "grad_norm": 7.241243914812559,
+      "learning_rate": 2e-05,
+      "loss": 1.3617,
+      "step": 41363
+    },
+    {
+      "epoch": 0.9845758354755784,
+      "grad_norm": 0.11156265288301279,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41364
+    },
+    {
+      "epoch": 0.98459963819861,
+      "grad_norm": 4.784528578418391,
+      "learning_rate": 2e-05,
+      "loss": 0.5576,
+      "step": 41365
+    },
+    {
+      "epoch": 0.9846234409216414,
+      "grad_norm": 3.3579844532076315,
+      "learning_rate": 2e-05,
+      "loss": 0.4822,
+      "step": 41366
+    },
+    {
+      "epoch": 0.9846472436446729,
+      "grad_norm": 7.799923902674982,
+      "learning_rate": 2e-05,
+      "loss": 0.6273,
+      "step": 41367
+    },
+    {
+      "epoch": 0.9846710463677044,
+      "grad_norm": 11.449274628398097,
+      "learning_rate": 2e-05,
+      "loss": 0.3442,
+      "step": 41368
+    },
+    {
+      "epoch": 0.984694849090736,
+      "grad_norm": 0.7138185035751102,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 41369
+    },
+    {
+      "epoch": 0.9847186518137675,
+      "grad_norm": 5.096512694617467,
+      "learning_rate": 2e-05,
+      "loss": 0.6475,
+      "step": 41370
+    },
+    {
+      "epoch": 0.984742454536799,
+      "grad_norm": 0.0926952651183318,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41371
+    },
+    {
+      "epoch": 0.9847662572598306,
+      "grad_norm": 0.06144311763076133,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41372
+    },
+    {
+      "epoch": 0.9847900599828621,
+      "grad_norm": 8.210601617415703,
+      "learning_rate": 2e-05,
+      "loss": 0.8156,
+      "step": 41373
+    },
+    {
+      "epoch": 0.9848138627058936,
+      "grad_norm": 2.342668036523799,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 41374
+    },
+    {
+      "epoch": 0.984837665428925,
+      "grad_norm": 17.951604377553206,
+      "learning_rate": 2e-05,
+      "loss": 0.9588,
+      "step": 41375
+    },
+    {
+      "epoch": 0.9848614681519566,
+      "grad_norm": 0.07540827604667677,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 41376
+    },
+    {
+      "epoch": 0.9848852708749881,
+      "grad_norm": 5.072742427098856,
+      "learning_rate": 2e-05,
+      "loss": 0.0866,
+      "step": 41377
+    },
+    {
+      "epoch": 0.9849090735980196,
+      "grad_norm": 3.2647672923222735,
+      "learning_rate": 2e-05,
+      "loss": 0.1197,
+      "step": 41378
+    },
+    {
+      "epoch": 0.9849328763210511,
+      "grad_norm": 3.034264244071978,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 41379
+    },
+    {
+      "epoch": 0.9849566790440827,
+      "grad_norm": 3.1982917538207905,
+      "learning_rate": 2e-05,
+      "loss": 0.5178,
+      "step": 41380
+    },
+    {
+      "epoch": 0.9849804817671142,
+      "grad_norm": 14.33275417230325,
+      "learning_rate": 2e-05,
+      "loss": 1.2787,
+      "step": 41381
+    },
+    {
+      "epoch": 0.9850042844901457,
+      "grad_norm": 5.2014659562032035,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 41382
+    },
+    {
+      "epoch": 0.9850280872131771,
+      "grad_norm": 7.272792035380216,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 41383
+    },
+    {
+      "epoch": 0.9850518899362087,
+      "grad_norm": 0.05963415755825076,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41384
+    },
+    {
+      "epoch": 0.9850756926592402,
+      "grad_norm": 1.0313477424751014,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 41385
+    },
+    {
+      "epoch": 0.9850994953822717,
+      "grad_norm": 0.1817153644093458,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 41386
+    },
+    {
+      "epoch": 0.9851232981053032,
+      "grad_norm": 4.199490869291541,
+      "learning_rate": 2e-05,
+      "loss": 0.5748,
+      "step": 41387
+    },
+    {
+      "epoch": 0.9851471008283348,
+      "grad_norm": 0.17696739601995823,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 41388
+    },
+    {
+      "epoch": 0.9851709035513663,
+      "grad_norm": 3.61155276058491,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 41389
+    },
+    {
+      "epoch": 0.9851947062743978,
+      "grad_norm": 2.012387431796076,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 41390
+    },
+    {
+      "epoch": 0.9852185089974294,
+      "grad_norm": 1.1884658342561552,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 41391
+    },
+    {
+      "epoch": 0.9852423117204608,
+      "grad_norm": 2.9589121258721325,
+      "learning_rate": 2e-05,
+      "loss": 0.0537,
+      "step": 41392
+    },
+    {
+      "epoch": 0.9852661144434923,
+      "grad_norm": 2.655260643212758,
+      "learning_rate": 2e-05,
+      "loss": 0.1655,
+      "step": 41393
+    },
+    {
+      "epoch": 0.9852899171665238,
+      "grad_norm": 3.5438396894506865,
+      "learning_rate": 2e-05,
+      "loss": 0.645,
+      "step": 41394
+    },
+    {
+      "epoch": 0.9853137198895554,
+      "grad_norm": 3.449872753066346,
+      "learning_rate": 2e-05,
+      "loss": 0.6436,
+      "step": 41395
+    },
+    {
+      "epoch": 0.9853375226125869,
+      "grad_norm": 5.22867003456953,
+      "learning_rate": 2e-05,
+      "loss": 1.1587,
+      "step": 41396
+    },
+    {
+      "epoch": 0.9853613253356184,
+      "grad_norm": 7.446735406509802,
+      "learning_rate": 2e-05,
+      "loss": 0.171,
+      "step": 41397
+    },
+    {
+      "epoch": 0.9853851280586499,
+      "grad_norm": 2.1778129308846,
+      "learning_rate": 2e-05,
+      "loss": 0.346,
+      "step": 41398
+    },
+    {
+      "epoch": 0.9854089307816815,
+      "grad_norm": 1.1809776307980293,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 41399
+    },
+    {
+      "epoch": 0.9854327335047129,
+      "grad_norm": 4.380995719470485,
+      "learning_rate": 2e-05,
+      "loss": 0.8233,
+      "step": 41400
+    },
+    {
+      "epoch": 0.9854565362277444,
+      "grad_norm": 3.275638980018746,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 41401
+    },
+    {
+      "epoch": 0.9854803389507759,
+      "grad_norm": 11.63209485560377,
+      "learning_rate": 2e-05,
+      "loss": 0.2153,
+      "step": 41402
+    },
+    {
+      "epoch": 0.9855041416738075,
+      "grad_norm": 3.1696049865438805,
+      "learning_rate": 2e-05,
+      "loss": 0.3205,
+      "step": 41403
+    },
+    {
+      "epoch": 0.985527944396839,
+      "grad_norm": 0.32900237943999683,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 41404
+    },
+    {
+      "epoch": 0.9855517471198705,
+      "grad_norm": 5.731551095188636,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 41405
+    },
+    {
+      "epoch": 0.9855755498429021,
+      "grad_norm": 4.039515461909371,
+      "learning_rate": 2e-05,
+      "loss": 0.6017,
+      "step": 41406
+    },
+    {
+      "epoch": 0.9855993525659336,
+      "grad_norm": 0.29263343438846545,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 41407
+    },
+    {
+      "epoch": 0.985623155288965,
+      "grad_norm": 6.009997945424186,
+      "learning_rate": 2e-05,
+      "loss": 1.0139,
+      "step": 41408
+    },
+    {
+      "epoch": 0.9856469580119965,
+      "grad_norm": 4.217768897066718,
+      "learning_rate": 2e-05,
+      "loss": 0.6088,
+      "step": 41409
+    },
+    {
+      "epoch": 0.9856707607350281,
+      "grad_norm": 0.11013367897504522,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41410
+    },
+    {
+      "epoch": 0.9856945634580596,
+      "grad_norm": 2.7507444893145934,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 41411
+    },
+    {
+      "epoch": 0.9857183661810911,
+      "grad_norm": 4.4748056323138,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 41412
+    },
+    {
+      "epoch": 0.9857421689041226,
+      "grad_norm": 0.7602746401680246,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 41413
+    },
+    {
+      "epoch": 0.9857659716271542,
+      "grad_norm": 3.5342456555247606,
+      "learning_rate": 2e-05,
+      "loss": 0.7109,
+      "step": 41414
+    },
+    {
+      "epoch": 0.9857897743501857,
+      "grad_norm": 0.40970746767847316,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 41415
+    },
+    {
+      "epoch": 0.9858135770732172,
+      "grad_norm": 3.8721300707219366,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 41416
+    },
+    {
+      "epoch": 0.9858373797962486,
+      "grad_norm": 2.9607188984522628,
+      "learning_rate": 2e-05,
+      "loss": 0.4321,
+      "step": 41417
+    },
+    {
+      "epoch": 0.9858611825192802,
+      "grad_norm": 0.19312969549101483,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 41418
+    },
+    {
+      "epoch": 0.9858849852423117,
+      "grad_norm": 6.231805021767332,
+      "learning_rate": 2e-05,
+      "loss": 0.3118,
+      "step": 41419
+    },
+    {
+      "epoch": 0.9859087879653432,
+      "grad_norm": 1.8232268669380516,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 41420
+    },
+    {
+      "epoch": 0.9859325906883748,
+      "grad_norm": 1.9666511775724025,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 41421
+    },
+    {
+      "epoch": 0.9859563934114063,
+      "grad_norm": 5.677919855085266,
+      "learning_rate": 2e-05,
+      "loss": 0.7954,
+      "step": 41422
+    },
+    {
+      "epoch": 0.9859801961344378,
+      "grad_norm": 0.17277102049905088,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 41423
+    },
+    {
+      "epoch": 0.9860039988574693,
+      "grad_norm": 14.518767872831921,
+      "learning_rate": 2e-05,
+      "loss": 1.3253,
+      "step": 41424
+    },
+    {
+      "epoch": 0.9860278015805009,
+      "grad_norm": 6.5830552074947475,
+      "learning_rate": 2e-05,
+      "loss": 0.7198,
+      "step": 41425
+    },
+    {
+      "epoch": 0.9860516043035323,
+      "grad_norm": 3.098286477331043,
+      "learning_rate": 2e-05,
+      "loss": 0.3952,
+      "step": 41426
+    },
+    {
+      "epoch": 0.9860754070265638,
+      "grad_norm": 9.632710413486071,
+      "learning_rate": 2e-05,
+      "loss": 0.5826,
+      "step": 41427
+    },
+    {
+      "epoch": 0.9860992097495953,
+      "grad_norm": 6.178982317024816,
+      "learning_rate": 2e-05,
+      "loss": 0.9108,
+      "step": 41428
+    },
+    {
+      "epoch": 0.9861230124726269,
+      "grad_norm": 3.9699491199188235,
+      "learning_rate": 2e-05,
+      "loss": 0.1238,
+      "step": 41429
+    },
+    {
+      "epoch": 0.9861468151956584,
+      "grad_norm": 0.03144299354956074,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41430
+    },
+    {
+      "epoch": 0.9861706179186899,
+      "grad_norm": 8.722879688831107,
+      "learning_rate": 2e-05,
+      "loss": 0.093,
+      "step": 41431
+    },
+    {
+      "epoch": 0.9861944206417214,
+      "grad_norm": 5.6248738879193025,
+      "learning_rate": 2e-05,
+      "loss": 0.6664,
+      "step": 41432
+    },
+    {
+      "epoch": 0.986218223364753,
+      "grad_norm": 0.049458288889302905,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41433
+    },
+    {
+      "epoch": 0.9862420260877844,
+      "grad_norm": 6.755797249143344,
+      "learning_rate": 2e-05,
+      "loss": 1.4036,
+      "step": 41434
+    },
+    {
+      "epoch": 0.9862658288108159,
+      "grad_norm": 4.141678462059238,
+      "learning_rate": 2e-05,
+      "loss": 0.7021,
+      "step": 41435
+    },
+    {
+      "epoch": 0.9862896315338475,
+      "grad_norm": 13.396290234287873,
+      "learning_rate": 2e-05,
+      "loss": 1.048,
+      "step": 41436
+    },
+    {
+      "epoch": 0.986313434256879,
+      "grad_norm": 0.5094779576712205,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 41437
+    },
+    {
+      "epoch": 0.9863372369799105,
+      "grad_norm": 2.940563668658212,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 41438
+    },
+    {
+      "epoch": 0.986361039702942,
+      "grad_norm": 0.17770841490345315,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 41439
+    },
+    {
+      "epoch": 0.9863848424259736,
+      "grad_norm": 4.333380798496925,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 41440
+    },
+    {
+      "epoch": 0.9864086451490051,
+      "grad_norm": 4.747910969270155,
+      "learning_rate": 2e-05,
+      "loss": 0.5307,
+      "step": 41441
+    },
+    {
+      "epoch": 0.9864324478720365,
+      "grad_norm": 1.7334387594616323,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 41442
+    },
+    {
+      "epoch": 0.986456250595068,
+      "grad_norm": 4.545649450689663,
+      "learning_rate": 2e-05,
+      "loss": 0.1812,
+      "step": 41443
+    },
+    {
+      "epoch": 0.9864800533180996,
+      "grad_norm": 2.1763902635790515,
+      "learning_rate": 2e-05,
+      "loss": 0.0558,
+      "step": 41444
+    },
+    {
+      "epoch": 0.9865038560411311,
+      "grad_norm": 10.583494092207056,
+      "learning_rate": 2e-05,
+      "loss": 0.565,
+      "step": 41445
+    },
+    {
+      "epoch": 0.9865276587641626,
+      "grad_norm": 1.4371822394920513,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 41446
+    },
+    {
+      "epoch": 0.9865514614871941,
+      "grad_norm": 1.9434107604523148,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 41447
+    },
+    {
+      "epoch": 0.9865752642102257,
+      "grad_norm": 1.9006301471135552,
+      "learning_rate": 2e-05,
+      "loss": 0.1073,
+      "step": 41448
+    },
+    {
+      "epoch": 0.9865990669332572,
+      "grad_norm": 0.12714594827696404,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41449
+    },
+    {
+      "epoch": 0.9866228696562886,
+      "grad_norm": 3.762926474519484,
+      "learning_rate": 2e-05,
+      "loss": 0.7742,
+      "step": 41450
+    },
+    {
+      "epoch": 0.9866466723793202,
+      "grad_norm": 1.5238771065767136,
+      "learning_rate": 2e-05,
+      "loss": 0.2112,
+      "step": 41451
+    },
+    {
+      "epoch": 0.9866704751023517,
+      "grad_norm": 7.094037964836542,
+      "learning_rate": 2e-05,
+      "loss": 0.1652,
+      "step": 41452
+    },
+    {
+      "epoch": 0.9866942778253832,
+      "grad_norm": 1.2112148337198205,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 41453
+    },
+    {
+      "epoch": 0.9867180805484147,
+      "grad_norm": 5.777662152985479,
+      "learning_rate": 2e-05,
+      "loss": 0.7528,
+      "step": 41454
+    },
+    {
+      "epoch": 0.9867418832714463,
+      "grad_norm": 2.397905216775675,
+      "learning_rate": 2e-05,
+      "loss": 0.2942,
+      "step": 41455
+    },
+    {
+      "epoch": 0.9867656859944778,
+      "grad_norm": 3.141434220988667,
+      "learning_rate": 2e-05,
+      "loss": 0.4541,
+      "step": 41456
+    },
+    {
+      "epoch": 0.9867894887175093,
+      "grad_norm": 3.4259934874239075,
+      "learning_rate": 2e-05,
+      "loss": 0.5611,
+      "step": 41457
+    },
+    {
+      "epoch": 0.9868132914405408,
+      "grad_norm": 3.191992221443181,
+      "learning_rate": 2e-05,
+      "loss": 0.5599,
+      "step": 41458
+    },
+    {
+      "epoch": 0.9868370941635723,
+      "grad_norm": 0.46405635542175727,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 41459
+    },
+    {
+      "epoch": 0.9868608968866038,
+      "grad_norm": 23.365710469818403,
+      "learning_rate": 2e-05,
+      "loss": 0.8541,
+      "step": 41460
+    },
+    {
+      "epoch": 0.9868846996096353,
+      "grad_norm": 5.18141305227402,
+      "learning_rate": 2e-05,
+      "loss": 0.0758,
+      "step": 41461
+    },
+    {
+      "epoch": 0.9869085023326668,
+      "grad_norm": 2.9938684542699066,
+      "learning_rate": 2e-05,
+      "loss": 0.4861,
+      "step": 41462
+    },
+    {
+      "epoch": 0.9869323050556984,
+      "grad_norm": 11.360955256797801,
+      "learning_rate": 2e-05,
+      "loss": 0.8264,
+      "step": 41463
+    },
+    {
+      "epoch": 0.9869561077787299,
+      "grad_norm": 0.1916188463704311,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 41464
+    },
+    {
+      "epoch": 0.9869799105017614,
+      "grad_norm": 2.5981576843969285,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 41465
+    },
+    {
+      "epoch": 0.987003713224793,
+      "grad_norm": 0.2663467024197367,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41466
+    },
+    {
+      "epoch": 0.9870275159478245,
+      "grad_norm": 0.07622968025813479,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41467
+    },
+    {
+      "epoch": 0.9870513186708559,
+      "grad_norm": 1.3348160139739123,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 41468
+    },
+    {
+      "epoch": 0.9870751213938874,
+      "grad_norm": 0.5620745243215441,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 41469
+    },
+    {
+      "epoch": 0.987098924116919,
+      "grad_norm": 4.302913576084382,
+      "learning_rate": 2e-05,
+      "loss": 0.4605,
+      "step": 41470
+    },
+    {
+      "epoch": 0.9871227268399505,
+      "grad_norm": 2.5236644635951584,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 41471
+    },
+    {
+      "epoch": 0.987146529562982,
+      "grad_norm": 9.848152538495633,
+      "learning_rate": 2e-05,
+      "loss": 1.069,
+      "step": 41472
+    },
+    {
+      "epoch": 0.9871703322860135,
+      "grad_norm": 8.508543085626158,
+      "learning_rate": 2e-05,
+      "loss": 0.1163,
+      "step": 41473
+    },
+    {
+      "epoch": 0.9871941350090451,
+      "grad_norm": 0.19350410046712554,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 41474
+    },
+    {
+      "epoch": 0.9872179377320766,
+      "grad_norm": 2.4997498374627884,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 41475
+    },
+    {
+      "epoch": 0.987241740455108,
+      "grad_norm": 0.03311933044428204,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41476
+    },
+    {
+      "epoch": 0.9872655431781395,
+      "grad_norm": 0.5190579295175457,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 41477
+    },
+    {
+      "epoch": 0.9872893459011711,
+      "grad_norm": 7.090282290723934,
+      "learning_rate": 2e-05,
+      "loss": 0.4707,
+      "step": 41478
+    },
+    {
+      "epoch": 0.9873131486242026,
+      "grad_norm": 3.33522818900762,
+      "learning_rate": 2e-05,
+      "loss": 0.7366,
+      "step": 41479
+    },
+    {
+      "epoch": 0.9873369513472341,
+      "grad_norm": 0.07756090168124918,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41480
+    },
+    {
+      "epoch": 0.9873607540702657,
+      "grad_norm": 7.058379576907685,
+      "learning_rate": 2e-05,
+      "loss": 1.0558,
+      "step": 41481
+    },
+    {
+      "epoch": 0.9873845567932972,
+      "grad_norm": 2.5240179790050306,
+      "learning_rate": 2e-05,
+      "loss": 0.1711,
+      "step": 41482
+    },
+    {
+      "epoch": 0.9874083595163287,
+      "grad_norm": 6.679011317985202,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 41483
+    },
+    {
+      "epoch": 0.9874321622393601,
+      "grad_norm": 4.6866845493511375,
+      "learning_rate": 2e-05,
+      "loss": 0.4901,
+      "step": 41484
+    },
+    {
+      "epoch": 0.9874559649623917,
+      "grad_norm": 0.13086264958406604,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 41485
+    },
+    {
+      "epoch": 0.9874797676854232,
+      "grad_norm": 0.05094549314761601,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 41486
+    },
+    {
+      "epoch": 0.9875035704084547,
+      "grad_norm": 7.335098581975023,
+      "learning_rate": 2e-05,
+      "loss": 2.1039,
+      "step": 41487
+    },
+    {
+      "epoch": 0.9875273731314862,
+      "grad_norm": 2.860066311786477,
+      "learning_rate": 2e-05,
+      "loss": 0.3135,
+      "step": 41488
+    },
+    {
+      "epoch": 0.9875511758545178,
+      "grad_norm": 4.4081539386003845,
+      "learning_rate": 2e-05,
+      "loss": 0.4897,
+      "step": 41489
+    },
+    {
+      "epoch": 0.9875749785775493,
+      "grad_norm": 16.399767294068635,
+      "learning_rate": 2e-05,
+      "loss": 0.3794,
+      "step": 41490
+    },
+    {
+      "epoch": 0.9875987813005808,
+      "grad_norm": 5.058606738167223,
+      "learning_rate": 2e-05,
+      "loss": 0.7226,
+      "step": 41491
+    },
+    {
+      "epoch": 0.9876225840236122,
+      "grad_norm": 1.5334828950844013,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 41492
+    },
+    {
+      "epoch": 0.9876463867466438,
+      "grad_norm": 0.39784276191050627,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 41493
+    },
+    {
+      "epoch": 0.9876701894696753,
+      "grad_norm": 3.081766855246526,
+      "learning_rate": 2e-05,
+      "loss": 0.0663,
+      "step": 41494
+    },
+    {
+      "epoch": 0.9876939921927068,
+      "grad_norm": 1.8778751342262991,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 41495
+    },
+    {
+      "epoch": 0.9877177949157384,
+      "grad_norm": 0.2550771085176841,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 41496
+    },
+    {
+      "epoch": 0.9877415976387699,
+      "grad_norm": 3.6840858953207203,
+      "learning_rate": 2e-05,
+      "loss": 0.5463,
+      "step": 41497
+    },
+    {
+      "epoch": 0.9877654003618014,
+      "grad_norm": 4.173762014024084,
+      "learning_rate": 2e-05,
+      "loss": 0.7361,
+      "step": 41498
+    },
+    {
+      "epoch": 0.9877892030848329,
+      "grad_norm": 2.0005993072161554,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 41499
+    },
+    {
+      "epoch": 0.9878130058078645,
+      "grad_norm": 7.353666852307207,
+      "learning_rate": 2e-05,
+      "loss": 0.6278,
+      "step": 41500
+    },
+    {
+      "epoch": 0.987836808530896,
+      "grad_norm": 21.78329091791308,
+      "learning_rate": 2e-05,
+      "loss": 0.3045,
+      "step": 41501
+    },
+    {
+      "epoch": 0.9878606112539274,
+      "grad_norm": 0.1065117282518785,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41502
+    },
+    {
+      "epoch": 0.9878844139769589,
+      "grad_norm": 3.6693486490557166,
+      "learning_rate": 2e-05,
+      "loss": 0.4713,
+      "step": 41503
+    },
+    {
+      "epoch": 0.9879082166999905,
+      "grad_norm": 2.4706069425429193,
+      "learning_rate": 2e-05,
+      "loss": 0.0571,
+      "step": 41504
+    },
+    {
+      "epoch": 0.987932019423022,
+      "grad_norm": 10.49206883807733,
+      "learning_rate": 2e-05,
+      "loss": 0.8247,
+      "step": 41505
+    },
+    {
+      "epoch": 0.9879558221460535,
+      "grad_norm": 1.4152102648967422,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 41506
+    },
+    {
+      "epoch": 0.987979624869085,
+      "grad_norm": 0.15904668684381665,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 41507
+    },
+    {
+      "epoch": 0.9880034275921166,
+      "grad_norm": 0.051001817995619037,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 41508
+    },
+    {
+      "epoch": 0.988027230315148,
+      "grad_norm": 0.11778188491584525,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 41509
+    },
+    {
+      "epoch": 0.9880510330381795,
+      "grad_norm": 5.751369468472335,
+      "learning_rate": 2e-05,
+      "loss": 0.1411,
+      "step": 41510
+    },
+    {
+      "epoch": 0.9880748357612111,
+      "grad_norm": 6.560086025049482,
+      "learning_rate": 2e-05,
+      "loss": 0.1605,
+      "step": 41511
+    },
+    {
+      "epoch": 0.9880986384842426,
+      "grad_norm": 3.073186677664059,
+      "learning_rate": 2e-05,
+      "loss": 0.6434,
+      "step": 41512
+    },
+    {
+      "epoch": 0.9881224412072741,
+      "grad_norm": 0.05227694985932722,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41513
+    },
+    {
+      "epoch": 0.9881462439303056,
+      "grad_norm": 7.720693694053072,
+      "learning_rate": 2e-05,
+      "loss": 1.4612,
+      "step": 41514
+    },
+    {
+      "epoch": 0.9881700466533372,
+      "grad_norm": 8.664682548772866,
+      "learning_rate": 2e-05,
+      "loss": 0.3732,
+      "step": 41515
+    },
+    {
+      "epoch": 0.9881938493763687,
+      "grad_norm": 0.5158282854507503,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 41516
+    },
+    {
+      "epoch": 0.9882176520994002,
+      "grad_norm": 3.7601123450396807,
+      "learning_rate": 2e-05,
+      "loss": 0.9051,
+      "step": 41517
+    },
+    {
+      "epoch": 0.9882414548224316,
+      "grad_norm": 3.3935358757082907,
+      "learning_rate": 2e-05,
+      "loss": 0.6051,
+      "step": 41518
+    },
+    {
+      "epoch": 0.9882652575454632,
+      "grad_norm": 2.3238457694975008,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 41519
+    },
+    {
+      "epoch": 0.9882890602684947,
+      "grad_norm": 4.240076062696413,
+      "learning_rate": 2e-05,
+      "loss": 0.878,
+      "step": 41520
+    },
+    {
+      "epoch": 0.9883128629915262,
+      "grad_norm": 4.476509651350763,
+      "learning_rate": 2e-05,
+      "loss": 0.2488,
+      "step": 41521
+    },
+    {
+      "epoch": 0.9883366657145577,
+      "grad_norm": 0.11547881613314485,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 41522
+    },
+    {
+      "epoch": 0.9883604684375893,
+      "grad_norm": 2.296801869426163,
+      "learning_rate": 2e-05,
+      "loss": 0.1836,
+      "step": 41523
+    },
+    {
+      "epoch": 0.9883842711606208,
+      "grad_norm": 5.379463006418792,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 41524
+    },
+    {
+      "epoch": 0.9884080738836523,
+      "grad_norm": 4.203580804351535,
+      "learning_rate": 2e-05,
+      "loss": 0.9625,
+      "step": 41525
+    },
+    {
+      "epoch": 0.9884318766066839,
+      "grad_norm": 3.6333694385923576,
+      "learning_rate": 2e-05,
+      "loss": 0.1089,
+      "step": 41526
+    },
+    {
+      "epoch": 0.9884556793297153,
+      "grad_norm": 0.05556818630078563,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41527
+    },
+    {
+      "epoch": 0.9884794820527468,
+      "grad_norm": 4.684125783120018,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 41528
+    },
+    {
+      "epoch": 0.9885032847757783,
+      "grad_norm": 0.7178960507077827,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 41529
+    },
+    {
+      "epoch": 0.9885270874988099,
+      "grad_norm": 0.01721011554026227,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41530
+    },
+    {
+      "epoch": 0.9885508902218414,
+      "grad_norm": 6.715079096281821,
+      "learning_rate": 2e-05,
+      "loss": 0.636,
+      "step": 41531
+    },
+    {
+      "epoch": 0.9885746929448729,
+      "grad_norm": 22.62331582832744,
+      "learning_rate": 2e-05,
+      "loss": 0.2702,
+      "step": 41532
+    },
+    {
+      "epoch": 0.9885984956679044,
+      "grad_norm": 3.580771434754249,
+      "learning_rate": 2e-05,
+      "loss": 0.9406,
+      "step": 41533
+    },
+    {
+      "epoch": 0.988622298390936,
+      "grad_norm": 0.3305362950957035,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 41534
+    },
+    {
+      "epoch": 0.9886461011139674,
+      "grad_norm": 1.4508757639381673,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 41535
+    },
+    {
+      "epoch": 0.9886699038369989,
+      "grad_norm": 8.60329701465531,
+      "learning_rate": 2e-05,
+      "loss": 0.8147,
+      "step": 41536
+    },
+    {
+      "epoch": 0.9886937065600304,
+      "grad_norm": 1.6374294040294421,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 41537
+    },
+    {
+      "epoch": 0.988717509283062,
+      "grad_norm": 4.9721027592356695,
+      "learning_rate": 2e-05,
+      "loss": 0.097,
+      "step": 41538
+    },
+    {
+      "epoch": 0.9887413120060935,
+      "grad_norm": 0.022965465287770947,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41539
+    },
+    {
+      "epoch": 0.988765114729125,
+      "grad_norm": 2.754508621086344,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 41540
+    },
+    {
+      "epoch": 0.9887889174521566,
+      "grad_norm": 8.859632477444448,
+      "learning_rate": 2e-05,
+      "loss": 0.1591,
+      "step": 41541
+    },
+    {
+      "epoch": 0.9888127201751881,
+      "grad_norm": 0.6378896788880697,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 41542
+    },
+    {
+      "epoch": 0.9888365228982195,
+      "grad_norm": 0.014884689847585501,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41543
+    },
+    {
+      "epoch": 0.988860325621251,
+      "grad_norm": 3.7373865181969794,
+      "learning_rate": 2e-05,
+      "loss": 0.602,
+      "step": 41544
+    },
+    {
+      "epoch": 0.9888841283442826,
+      "grad_norm": 3.7274208486915446,
+      "learning_rate": 2e-05,
+      "loss": 0.3215,
+      "step": 41545
+    },
+    {
+      "epoch": 0.9889079310673141,
+      "grad_norm": 3.336399739863272,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 41546
+    },
+    {
+      "epoch": 0.9889317337903456,
+      "grad_norm": 0.9838664068710926,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 41547
+    },
+    {
+      "epoch": 0.9889555365133771,
+      "grad_norm": 0.21429268835856186,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 41548
+    },
+    {
+      "epoch": 0.9889793392364087,
+      "grad_norm": 3.77023064007621,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 41549
+    },
+    {
+      "epoch": 0.9890031419594402,
+      "grad_norm": 4.491100816241101,
+      "learning_rate": 2e-05,
+      "loss": 0.1536,
+      "step": 41550
+    },
+    {
+      "epoch": 0.9890269446824717,
+      "grad_norm": 3.174545810965502,
+      "learning_rate": 2e-05,
+      "loss": 0.354,
+      "step": 41551
+    },
+    {
+      "epoch": 0.9890507474055031,
+      "grad_norm": 2.570364268509558,
+      "learning_rate": 2e-05,
+      "loss": 0.3319,
+      "step": 41552
+    },
+    {
+      "epoch": 0.9890745501285347,
+      "grad_norm": 2.9621083782543085,
+      "learning_rate": 2e-05,
+      "loss": 0.1045,
+      "step": 41553
+    },
+    {
+      "epoch": 0.9890983528515662,
+      "grad_norm": 4.218824442059303,
+      "learning_rate": 2e-05,
+      "loss": 0.5051,
+      "step": 41554
+    },
+    {
+      "epoch": 0.9891221555745977,
+      "grad_norm": 4.888050830058431,
+      "learning_rate": 2e-05,
+      "loss": 0.3249,
+      "step": 41555
+    },
+    {
+      "epoch": 0.9891459582976293,
+      "grad_norm": 10.329631615445168,
+      "learning_rate": 2e-05,
+      "loss": 0.8393,
+      "step": 41556
+    },
+    {
+      "epoch": 0.9891697610206608,
+      "grad_norm": 8.053178153055164,
+      "learning_rate": 2e-05,
+      "loss": 1.3938,
+      "step": 41557
+    },
+    {
+      "epoch": 0.9891935637436923,
+      "grad_norm": 1.7677553534164538,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 41558
+    },
+    {
+      "epoch": 0.9892173664667238,
+      "grad_norm": 2.5729564956746453,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 41559
+    },
+    {
+      "epoch": 0.9892411691897554,
+      "grad_norm": 3.2477208145902363,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 41560
+    },
+    {
+      "epoch": 0.9892649719127868,
+      "grad_norm": 2.078165038419398,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 41561
+    },
+    {
+      "epoch": 0.9892887746358183,
+      "grad_norm": 0.06454856526474925,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41562
+    },
+    {
+      "epoch": 0.9893125773588498,
+      "grad_norm": 3.297943079599787,
+      "learning_rate": 2e-05,
+      "loss": 0.1279,
+      "step": 41563
+    },
+    {
+      "epoch": 0.9893363800818814,
+      "grad_norm": 5.964957601462458,
+      "learning_rate": 2e-05,
+      "loss": 0.3239,
+      "step": 41564
+    },
+    {
+      "epoch": 0.9893601828049129,
+      "grad_norm": 2.768775505977863,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 41565
+    },
+    {
+      "epoch": 0.9893839855279444,
+      "grad_norm": 2.7108203807985918,
+      "learning_rate": 2e-05,
+      "loss": 0.3776,
+      "step": 41566
+    },
+    {
+      "epoch": 0.9894077882509759,
+      "grad_norm": 3.4959560700800227,
+      "learning_rate": 2e-05,
+      "loss": 0.7856,
+      "step": 41567
+    },
+    {
+      "epoch": 0.9894315909740075,
+      "grad_norm": 3.2260138820419315,
+      "learning_rate": 2e-05,
+      "loss": 0.7524,
+      "step": 41568
+    },
+    {
+      "epoch": 0.9894553936970389,
+      "grad_norm": 0.19075470468401273,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 41569
+    },
+    {
+      "epoch": 0.9894791964200704,
+      "grad_norm": 3.247955224244002,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 41570
+    },
+    {
+      "epoch": 0.989502999143102,
+      "grad_norm": 0.1288808627240136,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41571
+    },
+    {
+      "epoch": 0.9895268018661335,
+      "grad_norm": 5.615668954487389,
+      "learning_rate": 2e-05,
+      "loss": 0.7069,
+      "step": 41572
+    },
+    {
+      "epoch": 0.989550604589165,
+      "grad_norm": 6.822595065102426,
+      "learning_rate": 2e-05,
+      "loss": 0.3862,
+      "step": 41573
+    },
+    {
+      "epoch": 0.9895744073121965,
+      "grad_norm": 4.115941947945355,
+      "learning_rate": 2e-05,
+      "loss": 0.8397,
+      "step": 41574
+    },
+    {
+      "epoch": 0.9895982100352281,
+      "grad_norm": 4.112616934330476,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 41575
+    },
+    {
+      "epoch": 0.9896220127582596,
+      "grad_norm": 1.412400371680369,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 41576
+    },
+    {
+      "epoch": 0.989645815481291,
+      "grad_norm": 2.67942955044836,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 41577
+    },
+    {
+      "epoch": 0.9896696182043225,
+      "grad_norm": 2.9308373890916144,
+      "learning_rate": 2e-05,
+      "loss": 0.3954,
+      "step": 41578
+    },
+    {
+      "epoch": 0.9896934209273541,
+      "grad_norm": 0.043526666010162675,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41579
+    },
+    {
+      "epoch": 0.9897172236503856,
+      "grad_norm": 0.72124967131361,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 41580
+    },
+    {
+      "epoch": 0.9897410263734171,
+      "grad_norm": 9.470880978476238,
+      "learning_rate": 2e-05,
+      "loss": 1.0846,
+      "step": 41581
+    },
+    {
+      "epoch": 0.9897648290964486,
+      "grad_norm": 6.098635899909041,
+      "learning_rate": 2e-05,
+      "loss": 0.2543,
+      "step": 41582
+    },
+    {
+      "epoch": 0.9897886318194802,
+      "grad_norm": 0.04481961379517561,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41583
+    },
+    {
+      "epoch": 0.9898124345425117,
+      "grad_norm": 0.2651877763351479,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 41584
+    },
+    {
+      "epoch": 0.9898362372655432,
+      "grad_norm": 3.524766328505278,
+      "learning_rate": 2e-05,
+      "loss": 0.0737,
+      "step": 41585
+    },
+    {
+      "epoch": 0.9898600399885747,
+      "grad_norm": 7.656429010811385,
+      "learning_rate": 2e-05,
+      "loss": 0.119,
+      "step": 41586
+    },
+    {
+      "epoch": 0.9898838427116062,
+      "grad_norm": 0.018571451338419278,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41587
+    },
+    {
+      "epoch": 0.9899076454346377,
+      "grad_norm": 3.909931379432414,
+      "learning_rate": 2e-05,
+      "loss": 0.3505,
+      "step": 41588
+    },
+    {
+      "epoch": 0.9899314481576692,
+      "grad_norm": 2.3768473096369345,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 41589
+    },
+    {
+      "epoch": 0.9899552508807008,
+      "grad_norm": 2.4526704904063696,
+      "learning_rate": 2e-05,
+      "loss": 0.2995,
+      "step": 41590
+    },
+    {
+      "epoch": 0.9899790536037323,
+      "grad_norm": 3.598974886438454,
+      "learning_rate": 2e-05,
+      "loss": 0.3964,
+      "step": 41591
+    },
+    {
+      "epoch": 0.9900028563267638,
+      "grad_norm": 4.652175191770845,
+      "learning_rate": 2e-05,
+      "loss": 0.7695,
+      "step": 41592
+    },
+    {
+      "epoch": 0.9900266590497953,
+      "grad_norm": 5.972033894308248,
+      "learning_rate": 2e-05,
+      "loss": 1.7296,
+      "step": 41593
+    },
+    {
+      "epoch": 0.9900504617728269,
+      "grad_norm": 3.1644678275317952,
+      "learning_rate": 2e-05,
+      "loss": 0.7061,
+      "step": 41594
+    },
+    {
+      "epoch": 0.9900742644958583,
+      "grad_norm": 15.962575374154408,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 41595
+    },
+    {
+      "epoch": 0.9900980672188898,
+      "grad_norm": 0.403268283311049,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 41596
+    },
+    {
+      "epoch": 0.9901218699419213,
+      "grad_norm": 4.664735340729467,
+      "learning_rate": 2e-05,
+      "loss": 0.6469,
+      "step": 41597
+    },
+    {
+      "epoch": 0.9901456726649529,
+      "grad_norm": 6.106229111224216,
+      "learning_rate": 2e-05,
+      "loss": 1.2766,
+      "step": 41598
+    },
+    {
+      "epoch": 0.9901694753879844,
+      "grad_norm": 2.49075120804702,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 41599
+    },
+    {
+      "epoch": 0.9901932781110159,
+      "grad_norm": 0.444226693300435,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 41600
+    },
+    {
+      "epoch": 0.9902170808340475,
+      "grad_norm": 2.141507272815373,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 41601
+    },
+    {
+      "epoch": 0.990240883557079,
+      "grad_norm": 17.32953262775578,
+      "learning_rate": 2e-05,
+      "loss": 0.228,
+      "step": 41602
+    },
+    {
+      "epoch": 0.9902646862801104,
+      "grad_norm": 0.044112098714998105,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 41603
+    },
+    {
+      "epoch": 0.9902884890031419,
+      "grad_norm": 5.258106579624157,
+      "learning_rate": 2e-05,
+      "loss": 1.2462,
+      "step": 41604
+    },
+    {
+      "epoch": 0.9903122917261735,
+      "grad_norm": 8.00285565230659,
+      "learning_rate": 2e-05,
+      "loss": 0.8638,
+      "step": 41605
+    },
+    {
+      "epoch": 0.990336094449205,
+      "grad_norm": 0.013224858145843232,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41606
+    },
+    {
+      "epoch": 0.9903598971722365,
+      "grad_norm": 6.803536135061403,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 41607
+    },
+    {
+      "epoch": 0.990383699895268,
+      "grad_norm": 8.365319019600907,
+      "learning_rate": 2e-05,
+      "loss": 2.3108,
+      "step": 41608
+    },
+    {
+      "epoch": 0.9904075026182996,
+      "grad_norm": 2.977915389550201,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 41609
+    },
+    {
+      "epoch": 0.9904313053413311,
+      "grad_norm": 1.3966674346419765,
+      "learning_rate": 2e-05,
+      "loss": 0.0317,
+      "step": 41610
+    },
+    {
+      "epoch": 0.9904551080643625,
+      "grad_norm": 4.940721645006161,
+      "learning_rate": 2e-05,
+      "loss": 0.7141,
+      "step": 41611
+    },
+    {
+      "epoch": 0.990478910787394,
+      "grad_norm": 6.174598715431118,
+      "learning_rate": 2e-05,
+      "loss": 0.5036,
+      "step": 41612
+    },
+    {
+      "epoch": 0.9905027135104256,
+      "grad_norm": 2.0131912909780016,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 41613
+    },
+    {
+      "epoch": 0.9905265162334571,
+      "grad_norm": 1.1967736560418964,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 41614
+    },
+    {
+      "epoch": 0.9905503189564886,
+      "grad_norm": 4.095296187458067,
+      "learning_rate": 2e-05,
+      "loss": 0.7539,
+      "step": 41615
+    },
+    {
+      "epoch": 0.9905741216795202,
+      "grad_norm": 0.7559228830495702,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 41616
+    },
+    {
+      "epoch": 0.9905979244025517,
+      "grad_norm": 3.935647161419653,
+      "learning_rate": 2e-05,
+      "loss": 0.5657,
+      "step": 41617
+    },
+    {
+      "epoch": 0.9906217271255832,
+      "grad_norm": 0.09790396427914547,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41618
+    },
+    {
+      "epoch": 0.9906455298486146,
+      "grad_norm": 0.1896874604129919,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 41619
+    },
+    {
+      "epoch": 0.9906693325716462,
+      "grad_norm": 0.6315044745749152,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 41620
+    },
+    {
+      "epoch": 0.9906931352946777,
+      "grad_norm": 10.348889404530127,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 41621
+    },
+    {
+      "epoch": 0.9907169380177092,
+      "grad_norm": 0.3552977200540998,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 41622
+    },
+    {
+      "epoch": 0.9907407407407407,
+      "grad_norm": 0.6851652527685742,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 41623
+    },
+    {
+      "epoch": 0.9907645434637723,
+      "grad_norm": 0.19292447554810038,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 41624
+    },
+    {
+      "epoch": 0.9907883461868038,
+      "grad_norm": 0.7892266918906471,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 41625
+    },
+    {
+      "epoch": 0.9908121489098353,
+      "grad_norm": 0.051389627916319576,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41626
+    },
+    {
+      "epoch": 0.9908359516328668,
+      "grad_norm": 0.06793471183206208,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 41627
+    },
+    {
+      "epoch": 0.9908597543558983,
+      "grad_norm": 3.2122604912669916,
+      "learning_rate": 2e-05,
+      "loss": 0.6792,
+      "step": 41628
+    },
+    {
+      "epoch": 0.9908835570789298,
+      "grad_norm": 4.19848130653215,
+      "learning_rate": 2e-05,
+      "loss": 0.8933,
+      "step": 41629
+    },
+    {
+      "epoch": 0.9909073598019613,
+      "grad_norm": 0.43329767177149453,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 41630
+    },
+    {
+      "epoch": 0.9909311625249929,
+      "grad_norm": 4.951860986622889,
+      "learning_rate": 2e-05,
+      "loss": 1.1608,
+      "step": 41631
+    },
+    {
+      "epoch": 0.9909549652480244,
+      "grad_norm": 5.83572451098027,
+      "learning_rate": 2e-05,
+      "loss": 1.348,
+      "step": 41632
+    },
+    {
+      "epoch": 0.9909787679710559,
+      "grad_norm": 3.4563352410302604,
+      "learning_rate": 2e-05,
+      "loss": 0.4452,
+      "step": 41633
+    },
+    {
+      "epoch": 0.9910025706940874,
+      "grad_norm": 3.811062279090206,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 41634
+    },
+    {
+      "epoch": 0.991026373417119,
+      "grad_norm": 0.6510367826544108,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 41635
+    },
+    {
+      "epoch": 0.9910501761401505,
+      "grad_norm": 0.055886088690254326,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 41636
+    },
+    {
+      "epoch": 0.9910739788631819,
+      "grad_norm": 3.67217384562891,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 41637
+    },
+    {
+      "epoch": 0.9910977815862134,
+      "grad_norm": 2.573859482772772,
+      "learning_rate": 2e-05,
+      "loss": 0.5338,
+      "step": 41638
+    },
+    {
+      "epoch": 0.991121584309245,
+      "grad_norm": 1.7146063914119898,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 41639
+    },
+    {
+      "epoch": 0.9911453870322765,
+      "grad_norm": 0.044272748994027784,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41640
+    },
+    {
+      "epoch": 0.991169189755308,
+      "grad_norm": 4.447246328517268,
+      "learning_rate": 2e-05,
+      "loss": 0.8786,
+      "step": 41641
+    },
+    {
+      "epoch": 0.9911929924783395,
+      "grad_norm": 5.093848556259436,
+      "learning_rate": 2e-05,
+      "loss": 0.6381,
+      "step": 41642
+    },
+    {
+      "epoch": 0.9912167952013711,
+      "grad_norm": 0.007952932924142281,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41643
+    },
+    {
+      "epoch": 0.9912405979244026,
+      "grad_norm": 0.05053606062452358,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 41644
+    },
+    {
+      "epoch": 0.991264400647434,
+      "grad_norm": 0.2948276556745734,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 41645
+    },
+    {
+      "epoch": 0.9912882033704655,
+      "grad_norm": 0.9663058877258005,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 41646
+    },
+    {
+      "epoch": 0.9913120060934971,
+      "grad_norm": 9.489574614320196,
+      "learning_rate": 2e-05,
+      "loss": 0.6474,
+      "step": 41647
+    },
+    {
+      "epoch": 0.9913358088165286,
+      "grad_norm": 7.352416754103496,
+      "learning_rate": 2e-05,
+      "loss": 0.9233,
+      "step": 41648
+    },
+    {
+      "epoch": 0.9913596115395601,
+      "grad_norm": 0.06068542634040921,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41649
+    },
+    {
+      "epoch": 0.9913834142625917,
+      "grad_norm": 0.3581704077000674,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 41650
+    },
+    {
+      "epoch": 0.9914072169856232,
+      "grad_norm": 11.597278125595162,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 41651
+    },
+    {
+      "epoch": 0.9914310197086547,
+      "grad_norm": 0.8070723227133678,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 41652
+    },
+    {
+      "epoch": 0.9914548224316861,
+      "grad_norm": 6.913674418995278,
+      "learning_rate": 2e-05,
+      "loss": 1.2201,
+      "step": 41653
+    },
+    {
+      "epoch": 0.9914786251547177,
+      "grad_norm": 12.27176533386563,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 41654
+    },
+    {
+      "epoch": 0.9915024278777492,
+      "grad_norm": 4.568415586671445,
+      "learning_rate": 2e-05,
+      "loss": 0.8876,
+      "step": 41655
+    },
+    {
+      "epoch": 0.9915262306007807,
+      "grad_norm": 0.23856738270970276,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 41656
+    },
+    {
+      "epoch": 0.9915500333238122,
+      "grad_norm": 5.95838315259675,
+      "learning_rate": 2e-05,
+      "loss": 0.9903,
+      "step": 41657
+    },
+    {
+      "epoch": 0.9915738360468438,
+      "grad_norm": 0.15791830772760293,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 41658
+    },
+    {
+      "epoch": 0.9915976387698753,
+      "grad_norm": 0.3017809177564358,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 41659
+    },
+    {
+      "epoch": 0.9916214414929068,
+      "grad_norm": 2.2692561898132158,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 41660
+    },
+    {
+      "epoch": 0.9916452442159382,
+      "grad_norm": 4.624308149541649,
+      "learning_rate": 2e-05,
+      "loss": 0.1371,
+      "step": 41661
+    },
+    {
+      "epoch": 0.9916690469389698,
+      "grad_norm": 9.091857581402225,
+      "learning_rate": 2e-05,
+      "loss": 0.8162,
+      "step": 41662
+    },
+    {
+      "epoch": 0.9916928496620013,
+      "grad_norm": 4.211741324031861,
+      "learning_rate": 2e-05,
+      "loss": 0.5661,
+      "step": 41663
+    },
+    {
+      "epoch": 0.9917166523850328,
+      "grad_norm": 0.1533220636521171,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 41664
+    },
+    {
+      "epoch": 0.9917404551080644,
+      "grad_norm": 0.9658280537633009,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 41665
+    },
+    {
+      "epoch": 0.9917642578310959,
+      "grad_norm": 10.16037460120853,
+      "learning_rate": 2e-05,
+      "loss": 0.7379,
+      "step": 41666
+    },
+    {
+      "epoch": 0.9917880605541274,
+      "grad_norm": 3.333343056218552,
+      "learning_rate": 2e-05,
+      "loss": 0.6451,
+      "step": 41667
+    },
+    {
+      "epoch": 0.9918118632771589,
+      "grad_norm": 4.854138370181494,
+      "learning_rate": 2e-05,
+      "loss": 0.4958,
+      "step": 41668
+    },
+    {
+      "epoch": 0.9918356660001905,
+      "grad_norm": 0.049826770975087295,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41669
+    },
+    {
+      "epoch": 0.991859468723222,
+      "grad_norm": 4.165849323186857,
+      "learning_rate": 2e-05,
+      "loss": 0.7453,
+      "step": 41670
+    },
+    {
+      "epoch": 0.9918832714462534,
+      "grad_norm": 2.752929692204423,
+      "learning_rate": 2e-05,
+      "loss": 0.4841,
+      "step": 41671
+    },
+    {
+      "epoch": 0.9919070741692849,
+      "grad_norm": 3.820324948744018,
+      "learning_rate": 2e-05,
+      "loss": 0.6283,
+      "step": 41672
+    },
+    {
+      "epoch": 0.9919308768923165,
+      "grad_norm": 0.3268372209972,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 41673
+    },
+    {
+      "epoch": 0.991954679615348,
+      "grad_norm": 0.7173534964803712,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 41674
+    },
+    {
+      "epoch": 0.9919784823383795,
+      "grad_norm": 2.6255430107789643,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 41675
+    },
+    {
+      "epoch": 0.992002285061411,
+      "grad_norm": 1.6548273026205291,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 41676
+    },
+    {
+      "epoch": 0.9920260877844426,
+      "grad_norm": 0.029256830747864965,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41677
+    },
+    {
+      "epoch": 0.992049890507474,
+      "grad_norm": 3.233233260874144,
+      "learning_rate": 2e-05,
+      "loss": 0.2091,
+      "step": 41678
+    },
+    {
+      "epoch": 0.9920736932305055,
+      "grad_norm": 3.971791174201029,
+      "learning_rate": 2e-05,
+      "loss": 0.8892,
+      "step": 41679
+    },
+    {
+      "epoch": 0.9920974959535371,
+      "grad_norm": 0.028994807027453023,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41680
+    },
+    {
+      "epoch": 0.9921212986765686,
+      "grad_norm": 5.916020742596956,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 41681
+    },
+    {
+      "epoch": 0.9921451013996001,
+      "grad_norm": 12.551972416139012,
+      "learning_rate": 2e-05,
+      "loss": 0.2994,
+      "step": 41682
+    },
+    {
+      "epoch": 0.9921689041226316,
+      "grad_norm": 4.919337197695693,
+      "learning_rate": 2e-05,
+      "loss": 0.0825,
+      "step": 41683
+    },
+    {
+      "epoch": 0.9921927068456632,
+      "grad_norm": 3.320186124573383,
+      "learning_rate": 2e-05,
+      "loss": 0.3024,
+      "step": 41684
+    },
+    {
+      "epoch": 0.9922165095686947,
+      "grad_norm": 3.530587957791555,
+      "learning_rate": 2e-05,
+      "loss": 0.701,
+      "step": 41685
+    },
+    {
+      "epoch": 0.9922403122917262,
+      "grad_norm": 1.5865078537076012,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 41686
+    },
+    {
+      "epoch": 0.9922641150147576,
+      "grad_norm": 0.026584138731470613,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41687
+    },
+    {
+      "epoch": 0.9922879177377892,
+      "grad_norm": 0.24136151430716865,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41688
+    },
+    {
+      "epoch": 0.9923117204608207,
+      "grad_norm": 3.4520255241491222,
+      "learning_rate": 2e-05,
+      "loss": 0.3433,
+      "step": 41689
+    },
+    {
+      "epoch": 0.9923355231838522,
+      "grad_norm": 3.9129919171397542,
+      "learning_rate": 2e-05,
+      "loss": 0.0435,
+      "step": 41690
+    },
+    {
+      "epoch": 0.9923593259068837,
+      "grad_norm": 0.5087006907054803,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 41691
+    },
+    {
+      "epoch": 0.9923831286299153,
+      "grad_norm": 2.429779637075731,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 41692
+    },
+    {
+      "epoch": 0.9924069313529468,
+      "grad_norm": 13.30165815548588,
+      "learning_rate": 2e-05,
+      "loss": 1.7083,
+      "step": 41693
+    },
+    {
+      "epoch": 0.9924307340759783,
+      "grad_norm": 4.025929812990769,
+      "learning_rate": 2e-05,
+      "loss": 1.0791,
+      "step": 41694
+    },
+    {
+      "epoch": 0.9924545367990099,
+      "grad_norm": 6.906315445394127,
+      "learning_rate": 2e-05,
+      "loss": 0.5088,
+      "step": 41695
+    },
+    {
+      "epoch": 0.9924783395220413,
+      "grad_norm": 6.202187710935164,
+      "learning_rate": 2e-05,
+      "loss": 0.9328,
+      "step": 41696
+    },
+    {
+      "epoch": 0.9925021422450728,
+      "grad_norm": 0.7609468922753985,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 41697
+    },
+    {
+      "epoch": 0.9925259449681043,
+      "grad_norm": 9.554185789013838,
+      "learning_rate": 2e-05,
+      "loss": 0.5359,
+      "step": 41698
+    },
+    {
+      "epoch": 0.9925497476911359,
+      "grad_norm": 0.639464913349937,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 41699
+    },
+    {
+      "epoch": 0.9925735504141674,
+      "grad_norm": 22.813241302205494,
+      "learning_rate": 2e-05,
+      "loss": 0.075,
+      "step": 41700
+    },
+    {
+      "epoch": 0.9925973531371989,
+      "grad_norm": 0.5004980744542606,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 41701
+    },
+    {
+      "epoch": 0.9926211558602304,
+      "grad_norm": 3.2013938115936043,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 41702
+    },
+    {
+      "epoch": 0.992644958583262,
+      "grad_norm": 0.17035122615117645,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41703
+    },
+    {
+      "epoch": 0.9926687613062934,
+      "grad_norm": 0.02504770301801399,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41704
+    },
+    {
+      "epoch": 0.9926925640293249,
+      "grad_norm": 5.043522177686043,
+      "learning_rate": 2e-05,
+      "loss": 0.157,
+      "step": 41705
+    },
+    {
+      "epoch": 0.9927163667523564,
+      "grad_norm": 3.4949031156578863,
+      "learning_rate": 2e-05,
+      "loss": 0.8891,
+      "step": 41706
+    },
+    {
+      "epoch": 0.992740169475388,
+      "grad_norm": 0.08464267699205057,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41707
+    },
+    {
+      "epoch": 0.9927639721984195,
+      "grad_norm": 10.542388579572636,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 41708
+    },
+    {
+      "epoch": 0.992787774921451,
+      "grad_norm": 3.925800683893249,
+      "learning_rate": 2e-05,
+      "loss": 0.7428,
+      "step": 41709
+    },
+    {
+      "epoch": 0.9928115776444826,
+      "grad_norm": 5.07117728863256,
+      "learning_rate": 2e-05,
+      "loss": 0.4185,
+      "step": 41710
+    },
+    {
+      "epoch": 0.9928353803675141,
+      "grad_norm": 0.14393396766034028,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 41711
+    },
+    {
+      "epoch": 0.9928591830905455,
+      "grad_norm": 4.322277904559076,
+      "learning_rate": 2e-05,
+      "loss": 0.1005,
+      "step": 41712
+    },
+    {
+      "epoch": 0.992882985813577,
+      "grad_norm": 0.43553421263335107,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 41713
+    },
+    {
+      "epoch": 0.9929067885366086,
+      "grad_norm": 0.28260796159556933,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 41714
+    },
+    {
+      "epoch": 0.9929305912596401,
+      "grad_norm": 3.454876804131496,
+      "learning_rate": 2e-05,
+      "loss": 0.6331,
+      "step": 41715
+    },
+    {
+      "epoch": 0.9929543939826716,
+      "grad_norm": 0.6208074349662638,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 41716
+    },
+    {
+      "epoch": 0.9929781967057031,
+      "grad_norm": 3.9595054642743293,
+      "learning_rate": 2e-05,
+      "loss": 0.5981,
+      "step": 41717
+    },
+    {
+      "epoch": 0.9930019994287347,
+      "grad_norm": 3.596011534763967,
+      "learning_rate": 2e-05,
+      "loss": 0.5259,
+      "step": 41718
+    },
+    {
+      "epoch": 0.9930258021517662,
+      "grad_norm": 1.9632827036100764,
+      "learning_rate": 2e-05,
+      "loss": 0.1811,
+      "step": 41719
+    },
+    {
+      "epoch": 0.9930496048747977,
+      "grad_norm": 2.981455002737849,
+      "learning_rate": 2e-05,
+      "loss": 0.6501,
+      "step": 41720
+    },
+    {
+      "epoch": 0.9930734075978291,
+      "grad_norm": 3.4725191580599697,
+      "learning_rate": 2e-05,
+      "loss": 0.4015,
+      "step": 41721
+    },
+    {
+      "epoch": 0.9930972103208607,
+      "grad_norm": 5.831039042434025,
+      "learning_rate": 2e-05,
+      "loss": 0.1925,
+      "step": 41722
+    },
+    {
+      "epoch": 0.9931210130438922,
+      "grad_norm": 4.250206158276237,
+      "learning_rate": 2e-05,
+      "loss": 0.5244,
+      "step": 41723
+    },
+    {
+      "epoch": 0.9931448157669237,
+      "grad_norm": 3.3299447227714176,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 41724
+    },
+    {
+      "epoch": 0.9931686184899553,
+      "grad_norm": 3.999540270014593,
+      "learning_rate": 2e-05,
+      "loss": 0.295,
+      "step": 41725
+    },
+    {
+      "epoch": 0.9931924212129868,
+      "grad_norm": 3.8416036063962884,
+      "learning_rate": 2e-05,
+      "loss": 0.6075,
+      "step": 41726
+    },
+    {
+      "epoch": 0.9932162239360183,
+      "grad_norm": 0.14844240619247046,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41727
+    },
+    {
+      "epoch": 0.9932400266590498,
+      "grad_norm": 3.763395213194266,
+      "learning_rate": 2e-05,
+      "loss": 0.2399,
+      "step": 41728
+    },
+    {
+      "epoch": 0.9932638293820814,
+      "grad_norm": 6.707546988475458,
+      "learning_rate": 2e-05,
+      "loss": 0.8298,
+      "step": 41729
+    },
+    {
+      "epoch": 0.9932876321051128,
+      "grad_norm": 0.2401337081073397,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 41730
+    },
+    {
+      "epoch": 0.9933114348281443,
+      "grad_norm": 2.4510031107327963,
+      "learning_rate": 2e-05,
+      "loss": 0.6237,
+      "step": 41731
+    },
+    {
+      "epoch": 0.9933352375511758,
+      "grad_norm": 2.532172313338644,
+      "learning_rate": 2e-05,
+      "loss": 0.4631,
+      "step": 41732
+    },
+    {
+      "epoch": 0.9933590402742074,
+      "grad_norm": 2.496894001573286,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 41733
+    },
+    {
+      "epoch": 0.9933828429972389,
+      "grad_norm": 14.871418494518487,
+      "learning_rate": 2e-05,
+      "loss": 1.3681,
+      "step": 41734
+    },
+    {
+      "epoch": 0.9934066457202704,
+      "grad_norm": 6.128312244460261,
+      "learning_rate": 2e-05,
+      "loss": 1.9385,
+      "step": 41735
+    },
+    {
+      "epoch": 0.9934304484433019,
+      "grad_norm": 6.670887512862235,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 41736
+    },
+    {
+      "epoch": 0.9934542511663335,
+      "grad_norm": 12.326708167857495,
+      "learning_rate": 2e-05,
+      "loss": 0.7739,
+      "step": 41737
+    },
+    {
+      "epoch": 0.9934780538893649,
+      "grad_norm": 0.003711212531448116,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 41738
+    },
+    {
+      "epoch": 0.9935018566123964,
+      "grad_norm": 4.664417477027233,
+      "learning_rate": 2e-05,
+      "loss": 0.8557,
+      "step": 41739
+    },
+    {
+      "epoch": 0.993525659335428,
+      "grad_norm": 9.733730358860823,
+      "learning_rate": 2e-05,
+      "loss": 1.4264,
+      "step": 41740
+    },
+    {
+      "epoch": 0.9935494620584595,
+      "grad_norm": 4.801138963462475,
+      "learning_rate": 2e-05,
+      "loss": 0.0598,
+      "step": 41741
+    },
+    {
+      "epoch": 0.993573264781491,
+      "grad_norm": 4.576877561444842,
+      "learning_rate": 2e-05,
+      "loss": 0.5275,
+      "step": 41742
+    },
+    {
+      "epoch": 0.9935970675045225,
+      "grad_norm": 4.893401695639422,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 41743
+    },
+    {
+      "epoch": 0.9936208702275541,
+      "grad_norm": 2.8923671422553583,
+      "learning_rate": 2e-05,
+      "loss": 0.3844,
+      "step": 41744
+    },
+    {
+      "epoch": 0.9936446729505856,
+      "grad_norm": 6.303349457621776,
+      "learning_rate": 2e-05,
+      "loss": 0.6872,
+      "step": 41745
+    },
+    {
+      "epoch": 0.993668475673617,
+      "grad_norm": 1.4042762639174582,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 41746
+    },
+    {
+      "epoch": 0.9936922783966485,
+      "grad_norm": 2.9539895872970585,
+      "learning_rate": 2e-05,
+      "loss": 0.4243,
+      "step": 41747
+    },
+    {
+      "epoch": 0.9937160811196801,
+      "grad_norm": 16.305779327631914,
+      "learning_rate": 2e-05,
+      "loss": 2.0742,
+      "step": 41748
+    },
+    {
+      "epoch": 0.9937398838427116,
+      "grad_norm": 0.6403416586535623,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 41749
+    },
+    {
+      "epoch": 0.9937636865657431,
+      "grad_norm": 10.136056092688236,
+      "learning_rate": 2e-05,
+      "loss": 1.3601,
+      "step": 41750
+    },
+    {
+      "epoch": 0.9937874892887746,
+      "grad_norm": 4.259529072178582,
+      "learning_rate": 2e-05,
+      "loss": 0.6427,
+      "step": 41751
+    },
+    {
+      "epoch": 0.9938112920118062,
+      "grad_norm": 2.05994379262743,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 41752
+    },
+    {
+      "epoch": 0.9938350947348377,
+      "grad_norm": 2.8992428707883175,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 41753
+    },
+    {
+      "epoch": 0.9938588974578692,
+      "grad_norm": 14.948476069621345,
+      "learning_rate": 2e-05,
+      "loss": 0.5426,
+      "step": 41754
+    },
+    {
+      "epoch": 0.9938827001809007,
+      "grad_norm": 3.2775284077307787,
+      "learning_rate": 2e-05,
+      "loss": 0.4773,
+      "step": 41755
+    },
+    {
+      "epoch": 0.9939065029039322,
+      "grad_norm": 3.8326459722873336,
+      "learning_rate": 2e-05,
+      "loss": 0.7111,
+      "step": 41756
+    },
+    {
+      "epoch": 0.9939303056269637,
+      "grad_norm": 3.5914880924729267,
+      "learning_rate": 2e-05,
+      "loss": 0.5069,
+      "step": 41757
+    },
+    {
+      "epoch": 0.9939541083499952,
+      "grad_norm": 19.150025798145464,
+      "learning_rate": 2e-05,
+      "loss": 0.5704,
+      "step": 41758
+    },
+    {
+      "epoch": 0.9939779110730268,
+      "grad_norm": 0.039487905369224606,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41759
+    },
+    {
+      "epoch": 0.9940017137960583,
+      "grad_norm": 0.30071091405517975,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 41760
+    },
+    {
+      "epoch": 0.9940255165190898,
+      "grad_norm": 1.659868551947969,
+      "learning_rate": 2e-05,
+      "loss": 0.1196,
+      "step": 41761
+    },
+    {
+      "epoch": 0.9940493192421213,
+      "grad_norm": 1.09888285981911,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 41762
+    },
+    {
+      "epoch": 0.9940731219651529,
+      "grad_norm": 6.230717582418606,
+      "learning_rate": 2e-05,
+      "loss": 1.2418,
+      "step": 41763
+    },
+    {
+      "epoch": 0.9940969246881843,
+      "grad_norm": 5.870711524852933,
+      "learning_rate": 2e-05,
+      "loss": 1.0669,
+      "step": 41764
+    },
+    {
+      "epoch": 0.9941207274112158,
+      "grad_norm": 0.18488225181282741,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 41765
+    },
+    {
+      "epoch": 0.9941445301342473,
+      "grad_norm": 0.07618119050370993,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41766
+    },
+    {
+      "epoch": 0.9941683328572789,
+      "grad_norm": 2.288530089953973,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 41767
+    },
+    {
+      "epoch": 0.9941921355803104,
+      "grad_norm": 6.077977314360727,
+      "learning_rate": 2e-05,
+      "loss": 0.2088,
+      "step": 41768
+    },
+    {
+      "epoch": 0.9942159383033419,
+      "grad_norm": 0.5750788880029641,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 41769
+    },
+    {
+      "epoch": 0.9942397410263735,
+      "grad_norm": 1.8527284184690898,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 41770
+    },
+    {
+      "epoch": 0.994263543749405,
+      "grad_norm": 5.357964767734592,
+      "learning_rate": 2e-05,
+      "loss": 0.4271,
+      "step": 41771
+    },
+    {
+      "epoch": 0.9942873464724364,
+      "grad_norm": 16.757978408476955,
+      "learning_rate": 2e-05,
+      "loss": 0.7346,
+      "step": 41772
+    },
+    {
+      "epoch": 0.9943111491954679,
+      "grad_norm": 3.8294184953768093,
+      "learning_rate": 2e-05,
+      "loss": 0.3663,
+      "step": 41773
+    },
+    {
+      "epoch": 0.9943349519184995,
+      "grad_norm": 15.160135951094526,
+      "learning_rate": 2e-05,
+      "loss": 0.4247,
+      "step": 41774
+    },
+    {
+      "epoch": 0.994358754641531,
+      "grad_norm": 11.776308609504628,
+      "learning_rate": 2e-05,
+      "loss": 0.4002,
+      "step": 41775
+    },
+    {
+      "epoch": 0.9943825573645625,
+      "grad_norm": 1.711516591728484,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 41776
+    },
+    {
+      "epoch": 0.994406360087594,
+      "grad_norm": 2.7010237032471345,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 41777
+    },
+    {
+      "epoch": 0.9944301628106256,
+      "grad_norm": 4.435981662748229,
+      "learning_rate": 2e-05,
+      "loss": 1.0729,
+      "step": 41778
+    },
+    {
+      "epoch": 0.9944539655336571,
+      "grad_norm": 6.302286036121788,
+      "learning_rate": 2e-05,
+      "loss": 1.2287,
+      "step": 41779
+    },
+    {
+      "epoch": 0.9944777682566885,
+      "grad_norm": 2.9975189898439947,
+      "learning_rate": 2e-05,
+      "loss": 0.1612,
+      "step": 41780
+    },
+    {
+      "epoch": 0.99450157097972,
+      "grad_norm": 0.2902095361911329,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 41781
+    },
+    {
+      "epoch": 0.9945253737027516,
+      "grad_norm": 5.0661321559022285,
+      "learning_rate": 2e-05,
+      "loss": 0.9082,
+      "step": 41782
+    },
+    {
+      "epoch": 0.9945491764257831,
+      "grad_norm": 11.473852390695162,
+      "learning_rate": 2e-05,
+      "loss": 1.7944,
+      "step": 41783
+    },
+    {
+      "epoch": 0.9945729791488146,
+      "grad_norm": 0.9710831792324752,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 41784
+    },
+    {
+      "epoch": 0.9945967818718462,
+      "grad_norm": 5.87978793954802,
+      "learning_rate": 2e-05,
+      "loss": 0.6512,
+      "step": 41785
+    },
+    {
+      "epoch": 0.9946205845948777,
+      "grad_norm": 2.7359817372787276,
+      "learning_rate": 2e-05,
+      "loss": 0.4518,
+      "step": 41786
+    },
+    {
+      "epoch": 0.9946443873179092,
+      "grad_norm": 3.9132973056533733,
+      "learning_rate": 2e-05,
+      "loss": 0.6147,
+      "step": 41787
+    },
+    {
+      "epoch": 0.9946681900409406,
+      "grad_norm": 7.223177811758636,
+      "learning_rate": 2e-05,
+      "loss": 1.6302,
+      "step": 41788
+    },
+    {
+      "epoch": 0.9946919927639722,
+      "grad_norm": 0.05234734306788168,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41789
+    },
+    {
+      "epoch": 0.9947157954870037,
+      "grad_norm": 7.17744601917486,
+      "learning_rate": 2e-05,
+      "loss": 0.4733,
+      "step": 41790
+    },
+    {
+      "epoch": 0.9947395982100352,
+      "grad_norm": 4.238609643393393,
+      "learning_rate": 2e-05,
+      "loss": 0.7354,
+      "step": 41791
+    },
+    {
+      "epoch": 0.9947634009330667,
+      "grad_norm": 4.248238726459343,
+      "learning_rate": 2e-05,
+      "loss": 0.1647,
+      "step": 41792
+    },
+    {
+      "epoch": 0.9947872036560983,
+      "grad_norm": 3.948489619556626,
+      "learning_rate": 2e-05,
+      "loss": 0.4509,
+      "step": 41793
+    },
+    {
+      "epoch": 0.9948110063791298,
+      "grad_norm": 0.010737746658544807,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41794
+    },
+    {
+      "epoch": 0.9948348091021613,
+      "grad_norm": 1.486676797162977,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 41795
+    },
+    {
+      "epoch": 0.9948586118251928,
+      "grad_norm": 3.3928476131264946,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 41796
+    },
+    {
+      "epoch": 0.9948824145482243,
+      "grad_norm": 6.3617188159214955,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 41797
+    },
+    {
+      "epoch": 0.9949062172712558,
+      "grad_norm": 9.062305606555741,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 41798
+    },
+    {
+      "epoch": 0.9949300199942873,
+      "grad_norm": 35.08826020460128,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 41799
+    },
+    {
+      "epoch": 0.9949538227173189,
+      "grad_norm": 0.6539731591552465,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 41800
+    },
+    {
+      "epoch": 0.9949776254403504,
+      "grad_norm": 13.180974237684065,
+      "learning_rate": 2e-05,
+      "loss": 0.3824,
+      "step": 41801
+    },
+    {
+      "epoch": 0.9950014281633819,
+      "grad_norm": 0.7847554527244718,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 41802
+    },
+    {
+      "epoch": 0.9950252308864134,
+      "grad_norm": 4.338926914446211,
+      "learning_rate": 2e-05,
+      "loss": 0.1024,
+      "step": 41803
+    },
+    {
+      "epoch": 0.995049033609445,
+      "grad_norm": 0.9465136568324654,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 41804
+    },
+    {
+      "epoch": 0.9950728363324765,
+      "grad_norm": 6.0630855834616675,
+      "learning_rate": 2e-05,
+      "loss": 0.2685,
+      "step": 41805
+    },
+    {
+      "epoch": 0.9950966390555079,
+      "grad_norm": 5.091590602938097,
+      "learning_rate": 2e-05,
+      "loss": 1.1599,
+      "step": 41806
+    },
+    {
+      "epoch": 0.9951204417785394,
+      "grad_norm": 22.14730907813962,
+      "learning_rate": 2e-05,
+      "loss": 0.5657,
+      "step": 41807
+    },
+    {
+      "epoch": 0.995144244501571,
+      "grad_norm": 0.25289879016162664,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 41808
+    },
+    {
+      "epoch": 0.9951680472246025,
+      "grad_norm": 2.2578293430397145,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 41809
+    },
+    {
+      "epoch": 0.995191849947634,
+      "grad_norm": 0.3619833702990627,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 41810
+    },
+    {
+      "epoch": 0.9952156526706655,
+      "grad_norm": 0.6131966946908861,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 41811
+    },
+    {
+      "epoch": 0.9952394553936971,
+      "grad_norm": 12.499773290314073,
+      "learning_rate": 2e-05,
+      "loss": 1.2374,
+      "step": 41812
+    },
+    {
+      "epoch": 0.9952632581167286,
+      "grad_norm": 4.322388141078533,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 41813
+    },
+    {
+      "epoch": 0.99528706083976,
+      "grad_norm": 0.16372114538101903,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 41814
+    },
+    {
+      "epoch": 0.9953108635627916,
+      "grad_norm": 2.530897769185347,
+      "learning_rate": 2e-05,
+      "loss": 0.1966,
+      "step": 41815
+    },
+    {
+      "epoch": 0.9953346662858231,
+      "grad_norm": 5.346787664192613,
+      "learning_rate": 2e-05,
+      "loss": 0.7644,
+      "step": 41816
+    },
+    {
+      "epoch": 0.9953584690088546,
+      "grad_norm": 3.5490703939065082,
+      "learning_rate": 2e-05,
+      "loss": 0.2597,
+      "step": 41817
+    },
+    {
+      "epoch": 0.9953822717318861,
+      "grad_norm": 0.42024651088089326,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 41818
+    },
+    {
+      "epoch": 0.9954060744549177,
+      "grad_norm": 4.249630292879949,
+      "learning_rate": 2e-05,
+      "loss": 0.6547,
+      "step": 41819
+    },
+    {
+      "epoch": 0.9954298771779492,
+      "grad_norm": 7.321410476453338,
+      "learning_rate": 2e-05,
+      "loss": 0.4615,
+      "step": 41820
+    },
+    {
+      "epoch": 0.9954536799009807,
+      "grad_norm": 2.08441715336682,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 41821
+    },
+    {
+      "epoch": 0.9954774826240121,
+      "grad_norm": 0.6697286671129122,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 41822
+    },
+    {
+      "epoch": 0.9955012853470437,
+      "grad_norm": 2.283212533544095,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 41823
+    },
+    {
+      "epoch": 0.9955250880700752,
+      "grad_norm": 7.222677930470971,
+      "learning_rate": 2e-05,
+      "loss": 0.3147,
+      "step": 41824
+    },
+    {
+      "epoch": 0.9955488907931067,
+      "grad_norm": 11.180961615485362,
+      "learning_rate": 2e-05,
+      "loss": 0.5692,
+      "step": 41825
+    },
+    {
+      "epoch": 0.9955726935161382,
+      "grad_norm": 4.219784732619632,
+      "learning_rate": 2e-05,
+      "loss": 0.7602,
+      "step": 41826
+    },
+    {
+      "epoch": 0.9955964962391698,
+      "grad_norm": 4.020728570493974,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 41827
+    },
+    {
+      "epoch": 0.9956202989622013,
+      "grad_norm": 1.273754018318327,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 41828
+    },
+    {
+      "epoch": 0.9956441016852328,
+      "grad_norm": 4.109018500856641,
+      "learning_rate": 2e-05,
+      "loss": 0.5321,
+      "step": 41829
+    },
+    {
+      "epoch": 0.9956679044082644,
+      "grad_norm": 10.321892030026534,
+      "learning_rate": 2e-05,
+      "loss": 0.1925,
+      "step": 41830
+    },
+    {
+      "epoch": 0.9956917071312958,
+      "grad_norm": 1.055259549538444,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 41831
+    },
+    {
+      "epoch": 0.9957155098543273,
+      "grad_norm": 0.05837596890182064,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41832
+    },
+    {
+      "epoch": 0.9957393125773588,
+      "grad_norm": 3.638203262817846,
+      "learning_rate": 2e-05,
+      "loss": 0.4584,
+      "step": 41833
+    },
+    {
+      "epoch": 0.9957631153003904,
+      "grad_norm": 5.055938120605914,
+      "learning_rate": 2e-05,
+      "loss": 0.6815,
+      "step": 41834
+    },
+    {
+      "epoch": 0.9957869180234219,
+      "grad_norm": 19.56472349779406,
+      "learning_rate": 2e-05,
+      "loss": 0.3411,
+      "step": 41835
+    },
+    {
+      "epoch": 0.9958107207464534,
+      "grad_norm": 5.001480595340353,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 41836
+    },
+    {
+      "epoch": 0.9958345234694849,
+      "grad_norm": 0.02073780957970132,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 41837
+    },
+    {
+      "epoch": 0.9958583261925165,
+      "grad_norm": 14.032565592479996,
+      "learning_rate": 2e-05,
+      "loss": 1.2104,
+      "step": 41838
+    },
+    {
+      "epoch": 0.995882128915548,
+      "grad_norm": 1.2880234387956422,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 41839
+    },
+    {
+      "epoch": 0.9959059316385794,
+      "grad_norm": 3.4478796306152657,
+      "learning_rate": 2e-05,
+      "loss": 0.5032,
+      "step": 41840
+    },
+    {
+      "epoch": 0.9959297343616109,
+      "grad_norm": 3.202976249048668,
+      "learning_rate": 2e-05,
+      "loss": 0.0813,
+      "step": 41841
+    },
+    {
+      "epoch": 0.9959535370846425,
+      "grad_norm": 0.19871618931461071,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41842
+    },
+    {
+      "epoch": 0.995977339807674,
+      "grad_norm": 0.16224537149776075,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41843
+    },
+    {
+      "epoch": 0.9960011425307055,
+      "grad_norm": 3.509286578006191,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 41844
+    },
+    {
+      "epoch": 0.9960249452537371,
+      "grad_norm": 14.945468963737488,
+      "learning_rate": 2e-05,
+      "loss": 0.9264,
+      "step": 41845
+    },
+    {
+      "epoch": 0.9960487479767686,
+      "grad_norm": 0.7030780901471023,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 41846
+    },
+    {
+      "epoch": 0.9960725506998,
+      "grad_norm": 4.4819160252146615,
+      "learning_rate": 2e-05,
+      "loss": 1.018,
+      "step": 41847
+    },
+    {
+      "epoch": 0.9960963534228315,
+      "grad_norm": 0.30176854468907366,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 41848
+    },
+    {
+      "epoch": 0.9961201561458631,
+      "grad_norm": 0.12603199400153847,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 41849
+    },
+    {
+      "epoch": 0.9961439588688946,
+      "grad_norm": 5.243143031938031,
+      "learning_rate": 2e-05,
+      "loss": 1.1259,
+      "step": 41850
+    },
+    {
+      "epoch": 0.9961677615919261,
+      "grad_norm": 6.569315578492233,
+      "learning_rate": 2e-05,
+      "loss": 0.7545,
+      "step": 41851
+    },
+    {
+      "epoch": 0.9961915643149576,
+      "grad_norm": 0.22802373331704087,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 41852
+    },
+    {
+      "epoch": 0.9962153670379892,
+      "grad_norm": 5.983137028992095,
+      "learning_rate": 2e-05,
+      "loss": 1.8231,
+      "step": 41853
+    },
+    {
+      "epoch": 0.9962391697610207,
+      "grad_norm": 2.3270208576021396,
+      "learning_rate": 2e-05,
+      "loss": 0.0316,
+      "step": 41854
+    },
+    {
+      "epoch": 0.9962629724840522,
+      "grad_norm": 2.6015588279081054,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 41855
+    },
+    {
+      "epoch": 0.9962867752070836,
+      "grad_norm": 3.1685867732384776,
+      "learning_rate": 2e-05,
+      "loss": 0.1225,
+      "step": 41856
+    },
+    {
+      "epoch": 0.9963105779301152,
+      "grad_norm": 3.805902341126496,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 41857
+    },
+    {
+      "epoch": 0.9963343806531467,
+      "grad_norm": 2.0093397069183063,
+      "learning_rate": 2e-05,
+      "loss": 0.1784,
+      "step": 41858
+    },
+    {
+      "epoch": 0.9963581833761782,
+      "grad_norm": 4.612712141908599,
+      "learning_rate": 2e-05,
+      "loss": 0.6445,
+      "step": 41859
+    },
+    {
+      "epoch": 0.9963819860992098,
+      "grad_norm": 0.8191358092454332,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 41860
+    },
+    {
+      "epoch": 0.9964057888222413,
+      "grad_norm": 0.10107070298841477,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 41861
+    },
+    {
+      "epoch": 0.9964295915452728,
+      "grad_norm": 2.496870095338591,
+      "learning_rate": 2e-05,
+      "loss": 0.2589,
+      "step": 41862
+    },
+    {
+      "epoch": 0.9964533942683043,
+      "grad_norm": 3.361192980349698,
+      "learning_rate": 2e-05,
+      "loss": 0.4306,
+      "step": 41863
+    },
+    {
+      "epoch": 0.9964771969913359,
+      "grad_norm": 0.6811698578524883,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 41864
+    },
+    {
+      "epoch": 0.9965009997143673,
+      "grad_norm": 2.860029896611719,
+      "learning_rate": 2e-05,
+      "loss": 0.282,
+      "step": 41865
+    },
+    {
+      "epoch": 0.9965248024373988,
+      "grad_norm": 5.641903398606728,
+      "learning_rate": 2e-05,
+      "loss": 0.9553,
+      "step": 41866
+    },
+    {
+      "epoch": 0.9965486051604303,
+      "grad_norm": 0.07316070354686925,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 41867
+    },
+    {
+      "epoch": 0.9965724078834619,
+      "grad_norm": 2.5526328080700877,
+      "learning_rate": 2e-05,
+      "loss": 0.5117,
+      "step": 41868
+    },
+    {
+      "epoch": 0.9965962106064934,
+      "grad_norm": 0.07020985791144017,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 41869
+    },
+    {
+      "epoch": 0.9966200133295249,
+      "grad_norm": 11.252256049593658,
+      "learning_rate": 2e-05,
+      "loss": 0.601,
+      "step": 41870
+    },
+    {
+      "epoch": 0.9966438160525564,
+      "grad_norm": 0.5773215575553566,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 41871
+    },
+    {
+      "epoch": 0.996667618775588,
+      "grad_norm": 20.233491753224985,
+      "learning_rate": 2e-05,
+      "loss": 0.5019,
+      "step": 41872
+    },
+    {
+      "epoch": 0.9966914214986194,
+      "grad_norm": 0.05993103337991694,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41873
+    },
+    {
+      "epoch": 0.9967152242216509,
+      "grad_norm": 0.16504790466536806,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 41874
+    },
+    {
+      "epoch": 0.9967390269446825,
+      "grad_norm": 4.111756655977646,
+      "learning_rate": 2e-05,
+      "loss": 0.1356,
+      "step": 41875
+    },
+    {
+      "epoch": 0.996762829667714,
+      "grad_norm": 4.210910915543133,
+      "learning_rate": 2e-05,
+      "loss": 0.6497,
+      "step": 41876
+    },
+    {
+      "epoch": 0.9967866323907455,
+      "grad_norm": 3.6102550593035883,
+      "learning_rate": 2e-05,
+      "loss": 0.0576,
+      "step": 41877
+    },
+    {
+      "epoch": 0.996810435113777,
+      "grad_norm": 6.139990994896258,
+      "learning_rate": 2e-05,
+      "loss": 0.5789,
+      "step": 41878
+    },
+    {
+      "epoch": 0.9968342378368086,
+      "grad_norm": 5.201565076231447,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 41879
+    },
+    {
+      "epoch": 0.9968580405598401,
+      "grad_norm": 4.03283005363065,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 41880
+    },
+    {
+      "epoch": 0.9968818432828715,
+      "grad_norm": 13.612586456584008,
+      "learning_rate": 2e-05,
+      "loss": 0.2594,
+      "step": 41881
+    },
+    {
+      "epoch": 0.996905646005903,
+      "grad_norm": 9.085153042712895,
+      "learning_rate": 2e-05,
+      "loss": 0.3821,
+      "step": 41882
+    },
+    {
+      "epoch": 0.9969294487289346,
+      "grad_norm": 3.6478582448872463,
+      "learning_rate": 2e-05,
+      "loss": 0.4316,
+      "step": 41883
+    },
+    {
+      "epoch": 0.9969532514519661,
+      "grad_norm": 13.05110057478638,
+      "learning_rate": 2e-05,
+      "loss": 0.1721,
+      "step": 41884
+    },
+    {
+      "epoch": 0.9969770541749976,
+      "grad_norm": 0.8935981576974874,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 41885
+    },
+    {
+      "epoch": 0.9970008568980291,
+      "grad_norm": 3.632970013614414,
+      "learning_rate": 2e-05,
+      "loss": 0.6652,
+      "step": 41886
+    },
+    {
+      "epoch": 0.9970246596210607,
+      "grad_norm": 0.5769022291421547,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 41887
+    },
+    {
+      "epoch": 0.9970484623440922,
+      "grad_norm": 0.3050933566646009,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41888
+    },
+    {
+      "epoch": 0.9970722650671237,
+      "grad_norm": 0.6658825049899267,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 41889
+    },
+    {
+      "epoch": 0.9970960677901551,
+      "grad_norm": 3.49698912276204,
+      "learning_rate": 2e-05,
+      "loss": 0.3937,
+      "step": 41890
+    },
+    {
+      "epoch": 0.9971198705131867,
+      "grad_norm": 0.3678972514213727,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 41891
+    },
+    {
+      "epoch": 0.9971436732362182,
+      "grad_norm": 0.1034319628240287,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 41892
+    },
+    {
+      "epoch": 0.9971674759592497,
+      "grad_norm": 5.463353392861651,
+      "learning_rate": 2e-05,
+      "loss": 0.4152,
+      "step": 41893
+    },
+    {
+      "epoch": 0.9971912786822813,
+      "grad_norm": 3.5399678421310123,
+      "learning_rate": 2e-05,
+      "loss": 0.3516,
+      "step": 41894
+    },
+    {
+      "epoch": 0.9972150814053128,
+      "grad_norm": 7.8208450056923695,
+      "learning_rate": 2e-05,
+      "loss": 0.7993,
+      "step": 41895
+    },
+    {
+      "epoch": 0.9972388841283443,
+      "grad_norm": 0.010474820355385477,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41896
+    },
+    {
+      "epoch": 0.9972626868513758,
+      "grad_norm": 3.477764387536683,
+      "learning_rate": 2e-05,
+      "loss": 0.6115,
+      "step": 41897
+    },
+    {
+      "epoch": 0.9972864895744074,
+      "grad_norm": 2.905937336875268,
+      "learning_rate": 2e-05,
+      "loss": 0.57,
+      "step": 41898
+    },
+    {
+      "epoch": 0.9973102922974388,
+      "grad_norm": 6.305042209844787,
+      "learning_rate": 2e-05,
+      "loss": 0.7317,
+      "step": 41899
+    },
+    {
+      "epoch": 0.9973340950204703,
+      "grad_norm": 6.428939757296015,
+      "learning_rate": 2e-05,
+      "loss": 1.0456,
+      "step": 41900
+    },
+    {
+      "epoch": 0.9973578977435018,
+      "grad_norm": 5.690395612804621,
+      "learning_rate": 2e-05,
+      "loss": 1.1356,
+      "step": 41901
+    },
+    {
+      "epoch": 0.9973817004665334,
+      "grad_norm": 2.603082288441015,
+      "learning_rate": 2e-05,
+      "loss": 0.183,
+      "step": 41902
+    },
+    {
+      "epoch": 0.9974055031895649,
+      "grad_norm": 1.582834053373479,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 41903
+    },
+    {
+      "epoch": 0.9974293059125964,
+      "grad_norm": 1.1243349665373092,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 41904
+    },
+    {
+      "epoch": 0.9974531086356279,
+      "grad_norm": 0.03634303206356234,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41905
+    },
+    {
+      "epoch": 0.9974769113586595,
+      "grad_norm": 1.4944152312292904,
+      "learning_rate": 2e-05,
+      "loss": 0.0167,
+      "step": 41906
+    },
+    {
+      "epoch": 0.9975007140816909,
+      "grad_norm": 0.1435942276536644,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 41907
+    },
+    {
+      "epoch": 0.9975245168047224,
+      "grad_norm": 5.25251935871094,
+      "learning_rate": 2e-05,
+      "loss": 0.8544,
+      "step": 41908
+    },
+    {
+      "epoch": 0.997548319527754,
+      "grad_norm": 2.9593878079152187,
+      "learning_rate": 2e-05,
+      "loss": 0.0364,
+      "step": 41909
+    },
+    {
+      "epoch": 0.9975721222507855,
+      "grad_norm": 4.193050313045584,
+      "learning_rate": 2e-05,
+      "loss": 0.4589,
+      "step": 41910
+    },
+    {
+      "epoch": 0.997595924973817,
+      "grad_norm": 0.11726782179758626,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 41911
+    },
+    {
+      "epoch": 0.9976197276968485,
+      "grad_norm": 5.476437984344874,
+      "learning_rate": 2e-05,
+      "loss": 0.6899,
+      "step": 41912
+    },
+    {
+      "epoch": 0.9976435304198801,
+      "grad_norm": 3.1957255323677543,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 41913
+    },
+    {
+      "epoch": 0.9976673331429116,
+      "grad_norm": 0.32974771585497054,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 41914
+    },
+    {
+      "epoch": 0.997691135865943,
+      "grad_norm": 4.852450105652488,
+      "learning_rate": 2e-05,
+      "loss": 1.0584,
+      "step": 41915
+    },
+    {
+      "epoch": 0.9977149385889745,
+      "grad_norm": 4.314453248259222,
+      "learning_rate": 2e-05,
+      "loss": 0.4037,
+      "step": 41916
+    },
+    {
+      "epoch": 0.9977387413120061,
+      "grad_norm": 10.370820407907047,
+      "learning_rate": 2e-05,
+      "loss": 0.5556,
+      "step": 41917
+    },
+    {
+      "epoch": 0.9977625440350376,
+      "grad_norm": 5.779838845799119,
+      "learning_rate": 2e-05,
+      "loss": 0.8413,
+      "step": 41918
+    },
+    {
+      "epoch": 0.9977863467580691,
+      "grad_norm": 2.9787080431235715,
+      "learning_rate": 2e-05,
+      "loss": 0.1435,
+      "step": 41919
+    },
+    {
+      "epoch": 0.9978101494811006,
+      "grad_norm": 2.6054963491280896,
+      "learning_rate": 2e-05,
+      "loss": 0.337,
+      "step": 41920
+    },
+    {
+      "epoch": 0.9978339522041322,
+      "grad_norm": 7.34028770529258,
+      "learning_rate": 2e-05,
+      "loss": 0.7861,
+      "step": 41921
+    },
+    {
+      "epoch": 0.9978577549271637,
+      "grad_norm": 8.324055449466638,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 41922
+    },
+    {
+      "epoch": 0.9978815576501952,
+      "grad_norm": 8.110889996226156,
+      "learning_rate": 2e-05,
+      "loss": 0.1937,
+      "step": 41923
+    },
+    {
+      "epoch": 0.9979053603732267,
+      "grad_norm": 3.7218921202734063,
+      "learning_rate": 2e-05,
+      "loss": 0.8428,
+      "step": 41924
+    },
+    {
+      "epoch": 0.9979291630962582,
+      "grad_norm": 3.501867341143941,
+      "learning_rate": 2e-05,
+      "loss": 0.3379,
+      "step": 41925
+    },
+    {
+      "epoch": 0.9979529658192897,
+      "grad_norm": 2.818282488716618,
+      "learning_rate": 2e-05,
+      "loss": 0.3712,
+      "step": 41926
+    },
+    {
+      "epoch": 0.9979767685423212,
+      "grad_norm": 4.02327862638404,
+      "learning_rate": 2e-05,
+      "loss": 0.5514,
+      "step": 41927
+    },
+    {
+      "epoch": 0.9980005712653528,
+      "grad_norm": 0.8342237758400913,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 41928
+    },
+    {
+      "epoch": 0.9980243739883843,
+      "grad_norm": 0.05104608597179725,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41929
+    },
+    {
+      "epoch": 0.9980481767114158,
+      "grad_norm": 4.368512794440543,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 41930
+    },
+    {
+      "epoch": 0.9980719794344473,
+      "grad_norm": 3.4946816871322928,
+      "learning_rate": 2e-05,
+      "loss": 0.0812,
+      "step": 41931
+    },
+    {
+      "epoch": 0.9980957821574789,
+      "grad_norm": 0.38792808799332756,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41932
+    },
+    {
+      "epoch": 0.9981195848805103,
+      "grad_norm": 0.4383220376469375,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 41933
+    },
+    {
+      "epoch": 0.9981433876035418,
+      "grad_norm": 0.9990498154852606,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 41934
+    },
+    {
+      "epoch": 0.9981671903265733,
+      "grad_norm": 0.9906110451500665,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 41935
+    },
+    {
+      "epoch": 0.9981909930496049,
+      "grad_norm": 35.013863849664574,
+      "learning_rate": 2e-05,
+      "loss": 0.319,
+      "step": 41936
+    },
+    {
+      "epoch": 0.9982147957726364,
+      "grad_norm": 0.045061232726790905,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 41937
+    },
+    {
+      "epoch": 0.9982385984956679,
+      "grad_norm": 53.8353793405796,
+      "learning_rate": 2e-05,
+      "loss": 0.1994,
+      "step": 41938
+    },
+    {
+      "epoch": 0.9982624012186995,
+      "grad_norm": 6.1367078600637175,
+      "learning_rate": 2e-05,
+      "loss": 0.8059,
+      "step": 41939
+    },
+    {
+      "epoch": 0.998286203941731,
+      "grad_norm": 0.014105257808521895,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41940
+    },
+    {
+      "epoch": 0.9983100066647624,
+      "grad_norm": 2.614382823327609,
+      "learning_rate": 2e-05,
+      "loss": 0.3496,
+      "step": 41941
+    },
+    {
+      "epoch": 0.9983338093877939,
+      "grad_norm": 2.519388944876944,
+      "learning_rate": 2e-05,
+      "loss": 0.5241,
+      "step": 41942
+    },
+    {
+      "epoch": 0.9983576121108255,
+      "grad_norm": 5.368166935542842,
+      "learning_rate": 2e-05,
+      "loss": 0.5966,
+      "step": 41943
+    },
+    {
+      "epoch": 0.998381414833857,
+      "grad_norm": 3.475518395134874,
+      "learning_rate": 2e-05,
+      "loss": 0.4988,
+      "step": 41944
+    },
+    {
+      "epoch": 0.9984052175568885,
+      "grad_norm": 2.9415971425911667,
+      "learning_rate": 2e-05,
+      "loss": 0.3496,
+      "step": 41945
+    },
+    {
+      "epoch": 0.99842902027992,
+      "grad_norm": 10.486834432996169,
+      "learning_rate": 2e-05,
+      "loss": 0.2719,
+      "step": 41946
+    },
+    {
+      "epoch": 0.9984528230029516,
+      "grad_norm": 2.3551521116548204,
+      "learning_rate": 2e-05,
+      "loss": 0.2965,
+      "step": 41947
+    },
+    {
+      "epoch": 0.9984766257259831,
+      "grad_norm": 12.221803763587355,
+      "learning_rate": 2e-05,
+      "loss": 0.5736,
+      "step": 41948
+    },
+    {
+      "epoch": 0.9985004284490145,
+      "grad_norm": 6.640517245314454,
+      "learning_rate": 2e-05,
+      "loss": 0.6605,
+      "step": 41949
+    },
+    {
+      "epoch": 0.998524231172046,
+      "grad_norm": 9.70320791195739,
+      "learning_rate": 2e-05,
+      "loss": 0.5051,
+      "step": 41950
+    },
+    {
+      "epoch": 0.9985480338950776,
+      "grad_norm": 3.3282537316391747,
+      "learning_rate": 2e-05,
+      "loss": 0.3895,
+      "step": 41951
+    },
+    {
+      "epoch": 0.9985718366181091,
+      "grad_norm": 3.4040291931244075,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 41952
+    },
+    {
+      "epoch": 0.9985956393411406,
+      "grad_norm": 2.9664733501219174,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 41953
+    },
+    {
+      "epoch": 0.9986194420641722,
+      "grad_norm": 0.7818436728877636,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 41954
+    },
+    {
+      "epoch": 0.9986432447872037,
+      "grad_norm": 0.01647455244008356,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 41955
+    },
+    {
+      "epoch": 0.9986670475102352,
+      "grad_norm": 7.618532368935579,
+      "learning_rate": 2e-05,
+      "loss": 0.9144,
+      "step": 41956
+    },
+    {
+      "epoch": 0.9986908502332666,
+      "grad_norm": 3.56310049867075,
+      "learning_rate": 2e-05,
+      "loss": 0.2178,
+      "step": 41957
+    },
+    {
+      "epoch": 0.9987146529562982,
+      "grad_norm": 5.8758344494092585,
+      "learning_rate": 2e-05,
+      "loss": 0.3131,
+      "step": 41958
+    },
+    {
+      "epoch": 0.9987384556793297,
+      "grad_norm": 4.600634951411233,
+      "learning_rate": 2e-05,
+      "loss": 1.1662,
+      "step": 41959
+    },
+    {
+      "epoch": 0.9987622584023612,
+      "grad_norm": 4.123251208135007,
+      "learning_rate": 2e-05,
+      "loss": 0.768,
+      "step": 41960
+    },
+    {
+      "epoch": 0.9987860611253927,
+      "grad_norm": 0.2682392003244921,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 41961
+    },
+    {
+      "epoch": 0.9988098638484243,
+      "grad_norm": 5.860179885660092,
+      "learning_rate": 2e-05,
+      "loss": 0.2378,
+      "step": 41962
+    },
+    {
+      "epoch": 0.9988336665714558,
+      "grad_norm": 1.3442157760696485,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 41963
+    },
+    {
+      "epoch": 0.9988574692944873,
+      "grad_norm": 8.053706339658197,
+      "learning_rate": 2e-05,
+      "loss": 0.452,
+      "step": 41964
+    },
+    {
+      "epoch": 0.9988812720175188,
+      "grad_norm": 5.910189782999295,
+      "learning_rate": 2e-05,
+      "loss": 1.2346,
+      "step": 41965
+    },
+    {
+      "epoch": 0.9989050747405503,
+      "grad_norm": 0.037515306061677585,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41966
+    },
+    {
+      "epoch": 0.9989288774635818,
+      "grad_norm": 0.5240398870698195,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 41967
+    },
+    {
+      "epoch": 0.9989526801866133,
+      "grad_norm": 4.516296032517625,
+      "learning_rate": 2e-05,
+      "loss": 0.3653,
+      "step": 41968
+    },
+    {
+      "epoch": 0.9989764829096449,
+      "grad_norm": 2.797866684347089,
+      "learning_rate": 2e-05,
+      "loss": 0.5656,
+      "step": 41969
+    },
+    {
+      "epoch": 0.9990002856326764,
+      "grad_norm": 0.659856964835762,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 41970
+    },
+    {
+      "epoch": 0.9990240883557079,
+      "grad_norm": 3.318417465722451,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 41971
+    },
+    {
+      "epoch": 0.9990478910787394,
+      "grad_norm": 0.04745083857439468,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 41972
+    },
+    {
+      "epoch": 0.999071693801771,
+      "grad_norm": 4.148714174337312,
+      "learning_rate": 2e-05,
+      "loss": 0.4265,
+      "step": 41973
+    },
+    {
+      "epoch": 0.9990954965248025,
+      "grad_norm": 18.154555711821047,
+      "learning_rate": 2e-05,
+      "loss": 0.6151,
+      "step": 41974
+    },
+    {
+      "epoch": 0.9991192992478339,
+      "grad_norm": 4.324078445306906,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 41975
+    },
+    {
+      "epoch": 0.9991431019708654,
+      "grad_norm": 1.6746043385429181,
+      "learning_rate": 2e-05,
+      "loss": 0.0408,
+      "step": 41976
+    },
+    {
+      "epoch": 0.999166904693897,
+      "grad_norm": 8.13808772916221,
+      "learning_rate": 2e-05,
+      "loss": 1.2939,
+      "step": 41977
+    },
+    {
+      "epoch": 0.9991907074169285,
+      "grad_norm": 6.396571922885469,
+      "learning_rate": 2e-05,
+      "loss": 1.1262,
+      "step": 41978
+    },
+    {
+      "epoch": 0.99921451013996,
+      "grad_norm": 6.496557304908596,
+      "learning_rate": 2e-05,
+      "loss": 0.2551,
+      "step": 41979
+    },
+    {
+      "epoch": 0.9992383128629915,
+      "grad_norm": 0.011910527919640194,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 41980
+    },
+    {
+      "epoch": 0.9992621155860231,
+      "grad_norm": 3.4350790414510364,
+      "learning_rate": 2e-05,
+      "loss": 0.6713,
+      "step": 41981
+    },
+    {
+      "epoch": 0.9992859183090546,
+      "grad_norm": 5.3179877584007915,
+      "learning_rate": 2e-05,
+      "loss": 0.8721,
+      "step": 41982
+    },
+    {
+      "epoch": 0.999309721032086,
+      "grad_norm": 0.04525028593726619,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41983
+    },
+    {
+      "epoch": 0.9993335237551176,
+      "grad_norm": 3.526330376768117,
+      "learning_rate": 2e-05,
+      "loss": 0.3104,
+      "step": 41984
+    },
+    {
+      "epoch": 0.9993573264781491,
+      "grad_norm": 0.692165971053514,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 41985
+    },
+    {
+      "epoch": 0.9993811292011806,
+      "grad_norm": 3.174815794997834,
+      "learning_rate": 2e-05,
+      "loss": 0.6982,
+      "step": 41986
+    },
+    {
+      "epoch": 0.9994049319242121,
+      "grad_norm": 2.567166418256363,
+      "learning_rate": 2e-05,
+      "loss": 0.0648,
+      "step": 41987
+    },
+    {
+      "epoch": 0.9994287346472437,
+      "grad_norm": 0.7992921334141416,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 41988
+    },
+    {
+      "epoch": 0.9994525373702752,
+      "grad_norm": 0.8894660929878585,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 41989
+    },
+    {
+      "epoch": 0.9994763400933067,
+      "grad_norm": 0.5310493050108411,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 41990
+    },
+    {
+      "epoch": 0.9995001428163381,
+      "grad_norm": 3.9301833178578938,
+      "learning_rate": 2e-05,
+      "loss": 0.4805,
+      "step": 41991
+    },
+    {
+      "epoch": 0.9995239455393697,
+      "grad_norm": 0.029091529919548872,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 41992
+    },
+    {
+      "epoch": 0.9995477482624012,
+      "grad_norm": 0.06651722479711793,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 41993
+    },
+    {
+      "epoch": 0.9995715509854327,
+      "grad_norm": 2.8451671684360083,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 41994
+    },
+    {
+      "epoch": 0.9995953537084642,
+      "grad_norm": 6.526741160327181,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 41995
+    },
+    {
+      "epoch": 0.9996191564314958,
+      "grad_norm": 3.168291935944319,
+      "learning_rate": 2e-05,
+      "loss": 0.6162,
+      "step": 41996
+    },
+    {
+      "epoch": 0.9996429591545273,
+      "grad_norm": 0.1093451030965356,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 41997
+    },
+    {
+      "epoch": 0.9996667618775588,
+      "grad_norm": 3.757291372478014,
+      "learning_rate": 2e-05,
+      "loss": 0.457,
+      "step": 41998
+    },
+    {
+      "epoch": 0.9996905646005904,
+      "grad_norm": 27.04550025882549,
+      "learning_rate": 2e-05,
+      "loss": 1.2213,
+      "step": 41999
+    },
+    {
+      "epoch": 0.9997143673236218,
+      "grad_norm": 2.0938081047916715,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 42000
+    },
+    {
+      "epoch": 0.9997381700466533,
+      "grad_norm": 4.210071718279809,
+      "learning_rate": 2e-05,
+      "loss": 0.7547,
+      "step": 42001
+    },
+    {
+      "epoch": 0.9997619727696848,
+      "grad_norm": 3.1230344156296637,
+      "learning_rate": 2e-05,
+      "loss": 0.6743,
+      "step": 42002
+    },
+    {
+      "epoch": 0.9997857754927164,
+      "grad_norm": 2.247208128075217,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 42003
+    },
+    {
+      "epoch": 0.9998095782157479,
+      "grad_norm": 2.8968116416496255,
+      "learning_rate": 2e-05,
+      "loss": 0.2776,
+      "step": 42004
+    },
+    {
+      "epoch": 0.9998333809387794,
+      "grad_norm": 6.441546494568778,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 42005
+    },
+    {
+      "epoch": 0.9998571836618109,
+      "grad_norm": 2.292850452721807,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 42006
+    },
+    {
+      "epoch": 0.9998809863848425,
+      "grad_norm": 5.386334939508943,
+      "learning_rate": 2e-05,
+      "loss": 0.6911,
+      "step": 42007
+    },
+    {
+      "epoch": 0.999904789107874,
+      "grad_norm": 0.05095991159156273,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42008
+    },
+    {
+      "epoch": 0.9999285918309054,
+      "grad_norm": 0.057734881826406896,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42009
+    },
+    {
+      "epoch": 0.9999523945539369,
+      "grad_norm": 9.181283392248385,
+      "learning_rate": 2e-05,
+      "loss": 1.3,
+      "step": 42010
+    },
+    {
+      "epoch": 0.9999761972769685,
+      "grad_norm": 8.282803060726835,
+      "learning_rate": 2e-05,
+      "loss": 0.4432,
+      "step": 42011
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8787536250995851,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 42012
+    },
+    {
+      "epoch": 1.0000238027230315,
+      "grad_norm": 2.8243846071405936,
+      "learning_rate": 2e-05,
+      "loss": 0.4803,
+      "step": 42013
+    },
+    {
+      "epoch": 1.000047605446063,
+      "grad_norm": 0.365546491018229,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 42014
+    },
+    {
+      "epoch": 1.0000714081690945,
+      "grad_norm": 4.063267215946576,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 42015
+    },
+    {
+      "epoch": 1.0000952108921262,
+      "grad_norm": 0.862273759165547,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 42016
+    },
+    {
+      "epoch": 1.0001190136151576,
+      "grad_norm": 2.8812865586370315,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 42017
+    },
+    {
+      "epoch": 1.0001428163381891,
+      "grad_norm": 6.346947752601828,
+      "learning_rate": 2e-05,
+      "loss": 0.4742,
+      "step": 42018
+    },
+    {
+      "epoch": 1.0001666190612206,
+      "grad_norm": 15.53721340380417,
+      "learning_rate": 2e-05,
+      "loss": 0.1851,
+      "step": 42019
+    },
+    {
+      "epoch": 1.000190421784252,
+      "grad_norm": 9.810936634497208,
+      "learning_rate": 2e-05,
+      "loss": 0.887,
+      "step": 42020
+    },
+    {
+      "epoch": 1.0002142245072836,
+      "grad_norm": 1.3744173804255049,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 42021
+    },
+    {
+      "epoch": 1.000238027230315,
+      "grad_norm": 1.472426890506062,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 42022
+    },
+    {
+      "epoch": 1.0002618299533466,
+      "grad_norm": 6.331181681652273,
+      "learning_rate": 2e-05,
+      "loss": 0.0656,
+      "step": 42023
+    },
+    {
+      "epoch": 1.0002856326763783,
+      "grad_norm": 2.9119485616645533,
+      "learning_rate": 2e-05,
+      "loss": 0.4177,
+      "step": 42024
+    },
+    {
+      "epoch": 1.0003094353994098,
+      "grad_norm": 3.859912199423558,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 42025
+    },
+    {
+      "epoch": 1.0003332381224412,
+      "grad_norm": 3.029299178378598,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 42026
+    },
+    {
+      "epoch": 1.0003570408454727,
+      "grad_norm": 3.6475127581126556,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 42027
+    },
+    {
+      "epoch": 1.0003808435685042,
+      "grad_norm": 0.050376284877626194,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 42028
+    },
+    {
+      "epoch": 1.0004046462915357,
+      "grad_norm": 13.632906128929875,
+      "learning_rate": 2e-05,
+      "loss": 0.914,
+      "step": 42029
+    },
+    {
+      "epoch": 1.0004284490145672,
+      "grad_norm": 0.43397322484660167,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 42030
+    },
+    {
+      "epoch": 1.0004522517375989,
+      "grad_norm": 1.4778597777431872,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 42031
+    },
+    {
+      "epoch": 1.0004760544606304,
+      "grad_norm": 0.709434221375567,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 42032
+    },
+    {
+      "epoch": 1.0004998571836619,
+      "grad_norm": 1.4189454007509008,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 42033
+    },
+    {
+      "epoch": 1.0005236599066933,
+      "grad_norm": 6.630424376483203,
+      "learning_rate": 2e-05,
+      "loss": 0.7295,
+      "step": 42034
+    },
+    {
+      "epoch": 1.0005474626297248,
+      "grad_norm": 4.517397178875143,
+      "learning_rate": 2e-05,
+      "loss": 0.1596,
+      "step": 42035
+    },
+    {
+      "epoch": 1.0005712653527563,
+      "grad_norm": 1.4712067994187168,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 42036
+    },
+    {
+      "epoch": 1.0005950680757878,
+      "grad_norm": 2.929715106498087,
+      "learning_rate": 2e-05,
+      "loss": 0.3597,
+      "step": 42037
+    },
+    {
+      "epoch": 1.0006188707988193,
+      "grad_norm": 2.9011881967561215,
+      "learning_rate": 2e-05,
+      "loss": 0.2988,
+      "step": 42038
+    },
+    {
+      "epoch": 1.000642673521851,
+      "grad_norm": 1.894725724918773,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 42039
+    },
+    {
+      "epoch": 1.0006664762448825,
+      "grad_norm": 3.355936784888671,
+      "learning_rate": 2e-05,
+      "loss": 0.0534,
+      "step": 42040
+    },
+    {
+      "epoch": 1.000690278967914,
+      "grad_norm": 0.22824124732171958,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 42041
+    },
+    {
+      "epoch": 1.0007140816909454,
+      "grad_norm": 4.810128042784289,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 42042
+    },
+    {
+      "epoch": 1.000737884413977,
+      "grad_norm": 4.303599555198814,
+      "learning_rate": 2e-05,
+      "loss": 1.0779,
+      "step": 42043
+    },
+    {
+      "epoch": 1.0007616871370084,
+      "grad_norm": 1.4536950371982422,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 42044
+    },
+    {
+      "epoch": 1.00078548986004,
+      "grad_norm": 1.6788225984679939,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 42045
+    },
+    {
+      "epoch": 1.0008092925830716,
+      "grad_norm": 0.17597118650623414,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 42046
+    },
+    {
+      "epoch": 1.000833095306103,
+      "grad_norm": 3.1869910947954785,
+      "learning_rate": 2e-05,
+      "loss": 0.3789,
+      "step": 42047
+    },
+    {
+      "epoch": 1.0008568980291346,
+      "grad_norm": 11.898400946397581,
+      "learning_rate": 2e-05,
+      "loss": 1.3235,
+      "step": 42048
+    },
+    {
+      "epoch": 1.000880700752166,
+      "grad_norm": 2.287377641844372,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 42049
+    },
+    {
+      "epoch": 1.0009045034751975,
+      "grad_norm": 3.6980673544418896,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 42050
+    },
+    {
+      "epoch": 1.000928306198229,
+      "grad_norm": 12.151635564696969,
+      "learning_rate": 2e-05,
+      "loss": 0.7902,
+      "step": 42051
+    },
+    {
+      "epoch": 1.0009521089212605,
+      "grad_norm": 2.520672137123659,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 42052
+    },
+    {
+      "epoch": 1.000975911644292,
+      "grad_norm": 0.2248353151712574,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 42053
+    },
+    {
+      "epoch": 1.0009997143673237,
+      "grad_norm": 2.737219591266855,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 42054
+    },
+    {
+      "epoch": 1.0010235170903552,
+      "grad_norm": 7.09174127869257,
+      "learning_rate": 2e-05,
+      "loss": 0.6789,
+      "step": 42055
+    },
+    {
+      "epoch": 1.0010473198133867,
+      "grad_norm": 3.505228471025583,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 42056
+    },
+    {
+      "epoch": 1.0010711225364182,
+      "grad_norm": 0.1260932452607932,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42057
+    },
+    {
+      "epoch": 1.0010949252594497,
+      "grad_norm": 2.172874394016276,
+      "learning_rate": 2e-05,
+      "loss": 0.3268,
+      "step": 42058
+    },
+    {
+      "epoch": 1.0011187279824811,
+      "grad_norm": 2.2467757054527113,
+      "learning_rate": 2e-05,
+      "loss": 0.0351,
+      "step": 42059
+    },
+    {
+      "epoch": 1.0011425307055126,
+      "grad_norm": 1.8962322393846396,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 42060
+    },
+    {
+      "epoch": 1.0011663334285443,
+      "grad_norm": 0.4174745469068792,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 42061
+    },
+    {
+      "epoch": 1.0011901361515758,
+      "grad_norm": 0.6438793749426384,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 42062
+    },
+    {
+      "epoch": 1.0012139388746073,
+      "grad_norm": 3.628304781973115,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 42063
+    },
+    {
+      "epoch": 1.0012377415976388,
+      "grad_norm": 3.1886325891594667,
+      "learning_rate": 2e-05,
+      "loss": 0.4121,
+      "step": 42064
+    },
+    {
+      "epoch": 1.0012615443206703,
+      "grad_norm": 12.279931913370964,
+      "learning_rate": 2e-05,
+      "loss": 0.9441,
+      "step": 42065
+    },
+    {
+      "epoch": 1.0012853470437018,
+      "grad_norm": 2.563949227713267,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 42066
+    },
+    {
+      "epoch": 1.0013091497667332,
+      "grad_norm": 3.4312598429541517,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 42067
+    },
+    {
+      "epoch": 1.0013329524897647,
+      "grad_norm": 4.7216488563737595,
+      "learning_rate": 2e-05,
+      "loss": 0.1622,
+      "step": 42068
+    },
+    {
+      "epoch": 1.0013567552127964,
+      "grad_norm": 5.2723563071979305,
+      "learning_rate": 2e-05,
+      "loss": 0.6982,
+      "step": 42069
+    },
+    {
+      "epoch": 1.001380557935828,
+      "grad_norm": 1.220481491154166,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 42070
+    },
+    {
+      "epoch": 1.0014043606588594,
+      "grad_norm": 0.03142423058540872,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42071
+    },
+    {
+      "epoch": 1.001428163381891,
+      "grad_norm": 0.5416045092165113,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 42072
+    },
+    {
+      "epoch": 1.0014519661049224,
+      "grad_norm": 8.260328263908502,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 42073
+    },
+    {
+      "epoch": 1.0014757688279539,
+      "grad_norm": 2.197355447166973,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 42074
+    },
+    {
+      "epoch": 1.0014995715509853,
+      "grad_norm": 1.9508697335749727,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 42075
+    },
+    {
+      "epoch": 1.001523374274017,
+      "grad_norm": 3.1616594659899038,
+      "learning_rate": 2e-05,
+      "loss": 0.3613,
+      "step": 42076
+    },
+    {
+      "epoch": 1.0015471769970485,
+      "grad_norm": 5.011263214360804,
+      "learning_rate": 2e-05,
+      "loss": 0.9002,
+      "step": 42077
+    },
+    {
+      "epoch": 1.00157097972008,
+      "grad_norm": 1.295178794090023,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 42078
+    },
+    {
+      "epoch": 1.0015947824431115,
+      "grad_norm": 3.944871538531607,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 42079
+    },
+    {
+      "epoch": 1.001618585166143,
+      "grad_norm": 4.324413751590336,
+      "learning_rate": 2e-05,
+      "loss": 0.6191,
+      "step": 42080
+    },
+    {
+      "epoch": 1.0016423878891745,
+      "grad_norm": 4.525622173436378,
+      "learning_rate": 2e-05,
+      "loss": 0.5674,
+      "step": 42081
+    },
+    {
+      "epoch": 1.001666190612206,
+      "grad_norm": 7.1993399468380135,
+      "learning_rate": 2e-05,
+      "loss": 0.4033,
+      "step": 42082
+    },
+    {
+      "epoch": 1.0016899933352374,
+      "grad_norm": 0.4126773177313258,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 42083
+    },
+    {
+      "epoch": 1.0017137960582692,
+      "grad_norm": 0.8919256449100635,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 42084
+    },
+    {
+      "epoch": 1.0017375987813006,
+      "grad_norm": 2.4351892068119025,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 42085
+    },
+    {
+      "epoch": 1.0017614015043321,
+      "grad_norm": 10.15938585569576,
+      "learning_rate": 2e-05,
+      "loss": 1.3523,
+      "step": 42086
+    },
+    {
+      "epoch": 1.0017852042273636,
+      "grad_norm": 4.180822820685869,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 42087
+    },
+    {
+      "epoch": 1.001809006950395,
+      "grad_norm": 3.5587698945642736,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 42088
+    },
+    {
+      "epoch": 1.0018328096734266,
+      "grad_norm": 3.1763746529428363,
+      "learning_rate": 2e-05,
+      "loss": 0.0496,
+      "step": 42089
+    },
+    {
+      "epoch": 1.001856612396458,
+      "grad_norm": 2.4103924780792565,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 42090
+    },
+    {
+      "epoch": 1.0018804151194898,
+      "grad_norm": 2.427750854050789,
+      "learning_rate": 2e-05,
+      "loss": 0.0337,
+      "step": 42091
+    },
+    {
+      "epoch": 1.0019042178425213,
+      "grad_norm": 4.67145476443579,
+      "learning_rate": 2e-05,
+      "loss": 0.7212,
+      "step": 42092
+    },
+    {
+      "epoch": 1.0019280205655527,
+      "grad_norm": 3.4794507499099274,
+      "learning_rate": 2e-05,
+      "loss": 0.1705,
+      "step": 42093
+    },
+    {
+      "epoch": 1.0019518232885842,
+      "grad_norm": 0.29033265395919466,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 42094
+    },
+    {
+      "epoch": 1.0019756260116157,
+      "grad_norm": 3.8391748748673638,
+      "learning_rate": 2e-05,
+      "loss": 0.0911,
+      "step": 42095
+    },
+    {
+      "epoch": 1.0019994287346472,
+      "grad_norm": 4.045464856036634,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 42096
+    },
+    {
+      "epoch": 1.0020232314576787,
+      "grad_norm": 3.000218328973641,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 42097
+    },
+    {
+      "epoch": 1.0020470341807102,
+      "grad_norm": 5.3059018535212585,
+      "learning_rate": 2e-05,
+      "loss": 0.4209,
+      "step": 42098
+    },
+    {
+      "epoch": 1.0020708369037419,
+      "grad_norm": 1.3578571264139836,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 42099
+    },
+    {
+      "epoch": 1.0020946396267734,
+      "grad_norm": 2.2204539612259726,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 42100
+    },
+    {
+      "epoch": 1.0021184423498048,
+      "grad_norm": 0.12535111670047638,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 42101
+    },
+    {
+      "epoch": 1.0021422450728363,
+      "grad_norm": 0.8594184433271717,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 42102
+    },
+    {
+      "epoch": 1.0021660477958678,
+      "grad_norm": 2.8848109132738524,
+      "learning_rate": 2e-05,
+      "loss": 0.3282,
+      "step": 42103
+    },
+    {
+      "epoch": 1.0021898505188993,
+      "grad_norm": 1.8245463655200829,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 42104
+    },
+    {
+      "epoch": 1.0022136532419308,
+      "grad_norm": 7.666419830638175,
+      "learning_rate": 2e-05,
+      "loss": 0.5708,
+      "step": 42105
+    },
+    {
+      "epoch": 1.0022374559649625,
+      "grad_norm": 1.0661849673085284,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 42106
+    },
+    {
+      "epoch": 1.002261258687994,
+      "grad_norm": 3.937344707195808,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 42107
+    },
+    {
+      "epoch": 1.0022850614110255,
+      "grad_norm": 5.314160832401938,
+      "learning_rate": 2e-05,
+      "loss": 0.6886,
+      "step": 42108
+    },
+    {
+      "epoch": 1.002308864134057,
+      "grad_norm": 2.174275208243931,
+      "learning_rate": 2e-05,
+      "loss": 0.338,
+      "step": 42109
+    },
+    {
+      "epoch": 1.0023326668570884,
+      "grad_norm": 2.9208276114334444,
+      "learning_rate": 2e-05,
+      "loss": 0.2848,
+      "step": 42110
+    },
+    {
+      "epoch": 1.00235646958012,
+      "grad_norm": 2.479069445159075,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 42111
+    },
+    {
+      "epoch": 1.0023802723031514,
+      "grad_norm": 3.9594441289764424,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 42112
+    },
+    {
+      "epoch": 1.002404075026183,
+      "grad_norm": 1.2765220504300636,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 42113
+    },
+    {
+      "epoch": 1.0024278777492146,
+      "grad_norm": 1.635320839936101,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 42114
+    },
+    {
+      "epoch": 1.002451680472246,
+      "grad_norm": 4.406909648305181,
+      "learning_rate": 2e-05,
+      "loss": 0.6791,
+      "step": 42115
+    },
+    {
+      "epoch": 1.0024754831952776,
+      "grad_norm": 4.144761937033041,
+      "learning_rate": 2e-05,
+      "loss": 0.5105,
+      "step": 42116
+    },
+    {
+      "epoch": 1.002499285918309,
+      "grad_norm": 4.385130946660548,
+      "learning_rate": 2e-05,
+      "loss": 0.6689,
+      "step": 42117
+    },
+    {
+      "epoch": 1.0025230886413405,
+      "grad_norm": 0.07263668091508615,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42118
+    },
+    {
+      "epoch": 1.002546891364372,
+      "grad_norm": 3.635414652424698,
+      "learning_rate": 2e-05,
+      "loss": 0.4617,
+      "step": 42119
+    },
+    {
+      "epoch": 1.0025706940874035,
+      "grad_norm": 7.0275737668148,
+      "learning_rate": 2e-05,
+      "loss": 0.6407,
+      "step": 42120
+    },
+    {
+      "epoch": 1.0025944968104352,
+      "grad_norm": 4.114274516999334,
+      "learning_rate": 2e-05,
+      "loss": 0.5807,
+      "step": 42121
+    },
+    {
+      "epoch": 1.0026182995334667,
+      "grad_norm": 4.580472608533627,
+      "learning_rate": 2e-05,
+      "loss": 0.8948,
+      "step": 42122
+    },
+    {
+      "epoch": 1.0026421022564982,
+      "grad_norm": 3.3567413113293396,
+      "learning_rate": 2e-05,
+      "loss": 0.2459,
+      "step": 42123
+    },
+    {
+      "epoch": 1.0026659049795297,
+      "grad_norm": 1.854813369134885,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 42124
+    },
+    {
+      "epoch": 1.0026897077025612,
+      "grad_norm": 6.353845351217454,
+      "learning_rate": 2e-05,
+      "loss": 0.2884,
+      "step": 42125
+    },
+    {
+      "epoch": 1.0027135104255926,
+      "grad_norm": 0.18225904845042018,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 42126
+    },
+    {
+      "epoch": 1.0027373131486241,
+      "grad_norm": 10.41628568507139,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 42127
+    },
+    {
+      "epoch": 1.0027611158716556,
+      "grad_norm": 6.875912362831163,
+      "learning_rate": 2e-05,
+      "loss": 1.3094,
+      "step": 42128
+    },
+    {
+      "epoch": 1.0027849185946873,
+      "grad_norm": 0.09101565683008496,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 42129
+    },
+    {
+      "epoch": 1.0028087213177188,
+      "grad_norm": 0.027910922663128668,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42130
+    },
+    {
+      "epoch": 1.0028325240407503,
+      "grad_norm": 0.044230907973916746,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 42131
+    },
+    {
+      "epoch": 1.0028563267637818,
+      "grad_norm": 2.9201446672707325,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 42132
+    },
+    {
+      "epoch": 1.0028801294868133,
+      "grad_norm": 4.204802500207527,
+      "learning_rate": 2e-05,
+      "loss": 0.4059,
+      "step": 42133
+    },
+    {
+      "epoch": 1.0029039322098448,
+      "grad_norm": 0.533879575798514,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 42134
+    },
+    {
+      "epoch": 1.0029277349328762,
+      "grad_norm": 0.08565848746632453,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42135
+    },
+    {
+      "epoch": 1.002951537655908,
+      "grad_norm": 4.760826555352765,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 42136
+    },
+    {
+      "epoch": 1.0029753403789394,
+      "grad_norm": 1.6561997929081274,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 42137
+    },
+    {
+      "epoch": 1.002999143101971,
+      "grad_norm": 0.1637938075054743,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42138
+    },
+    {
+      "epoch": 1.0030229458250024,
+      "grad_norm": 6.530042956585714,
+      "learning_rate": 2e-05,
+      "loss": 0.8078,
+      "step": 42139
+    },
+    {
+      "epoch": 1.0030467485480339,
+      "grad_norm": 1.6855835006631121,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 42140
+    },
+    {
+      "epoch": 1.0030705512710654,
+      "grad_norm": 4.597539762908189,
+      "learning_rate": 2e-05,
+      "loss": 0.4683,
+      "step": 42141
+    },
+    {
+      "epoch": 1.0030943539940969,
+      "grad_norm": 2.3545705036696765,
+      "learning_rate": 2e-05,
+      "loss": 0.0173,
+      "step": 42142
+    },
+    {
+      "epoch": 1.0031181567171283,
+      "grad_norm": 2.954601967481068,
+      "learning_rate": 2e-05,
+      "loss": 0.4889,
+      "step": 42143
+    },
+    {
+      "epoch": 1.00314195944016,
+      "grad_norm": 0.1787632185410061,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 42144
+    },
+    {
+      "epoch": 1.0031657621631915,
+      "grad_norm": 1.3744995158613944,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 42145
+    },
+    {
+      "epoch": 1.003189564886223,
+      "grad_norm": 1.5984631015911117,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 42146
+    },
+    {
+      "epoch": 1.0032133676092545,
+      "grad_norm": 4.283810350688683,
+      "learning_rate": 2e-05,
+      "loss": 0.3142,
+      "step": 42147
+    },
+    {
+      "epoch": 1.003237170332286,
+      "grad_norm": 2.440866629205093,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 42148
+    },
+    {
+      "epoch": 1.0032609730553175,
+      "grad_norm": 0.18625673435636883,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42149
+    },
+    {
+      "epoch": 1.003284775778349,
+      "grad_norm": 2.238379729816324,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 42150
+    },
+    {
+      "epoch": 1.0033085785013807,
+      "grad_norm": 3.7206883070455308,
+      "learning_rate": 2e-05,
+      "loss": 0.4643,
+      "step": 42151
+    },
+    {
+      "epoch": 1.0033323812244122,
+      "grad_norm": 2.2885242869485722,
+      "learning_rate": 2e-05,
+      "loss": 0.0433,
+      "step": 42152
+    },
+    {
+      "epoch": 1.0033561839474436,
+      "grad_norm": 0.3995231786121172,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 42153
+    },
+    {
+      "epoch": 1.0033799866704751,
+      "grad_norm": 5.205023314545382,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 42154
+    },
+    {
+      "epoch": 1.0034037893935066,
+      "grad_norm": 8.155845501432003,
+      "learning_rate": 2e-05,
+      "loss": 0.3569,
+      "step": 42155
+    },
+    {
+      "epoch": 1.003427592116538,
+      "grad_norm": 0.3601394060431743,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 42156
+    },
+    {
+      "epoch": 1.0034513948395696,
+      "grad_norm": 2.3073969103066982,
+      "learning_rate": 2e-05,
+      "loss": 0.0664,
+      "step": 42157
+    },
+    {
+      "epoch": 1.003475197562601,
+      "grad_norm": 8.712631047951856,
+      "learning_rate": 2e-05,
+      "loss": 0.822,
+      "step": 42158
+    },
+    {
+      "epoch": 1.0034990002856328,
+      "grad_norm": 6.495107963082726,
+      "learning_rate": 2e-05,
+      "loss": 0.3151,
+      "step": 42159
+    },
+    {
+      "epoch": 1.0035228030086643,
+      "grad_norm": 5.174664770935269,
+      "learning_rate": 2e-05,
+      "loss": 0.2802,
+      "step": 42160
+    },
+    {
+      "epoch": 1.0035466057316957,
+      "grad_norm": 0.03467143658003943,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42161
+    },
+    {
+      "epoch": 1.0035704084547272,
+      "grad_norm": 0.1118616128196778,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 42162
+    },
+    {
+      "epoch": 1.0035942111777587,
+      "grad_norm": 1.8559710679857846,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 42163
+    },
+    {
+      "epoch": 1.0036180139007902,
+      "grad_norm": 5.388432455350064,
+      "learning_rate": 2e-05,
+      "loss": 0.8217,
+      "step": 42164
+    },
+    {
+      "epoch": 1.0036418166238217,
+      "grad_norm": 6.333963632877528,
+      "learning_rate": 2e-05,
+      "loss": 0.5414,
+      "step": 42165
+    },
+    {
+      "epoch": 1.0036656193468534,
+      "grad_norm": 4.924693470877548,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 42166
+    },
+    {
+      "epoch": 1.0036894220698849,
+      "grad_norm": 6.89780927159437,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 42167
+    },
+    {
+      "epoch": 1.0037132247929164,
+      "grad_norm": 5.412672858003995,
+      "learning_rate": 2e-05,
+      "loss": 0.6199,
+      "step": 42168
+    },
+    {
+      "epoch": 1.0037370275159478,
+      "grad_norm": 18.50853801530468,
+      "learning_rate": 2e-05,
+      "loss": 0.7024,
+      "step": 42169
+    },
+    {
+      "epoch": 1.0037608302389793,
+      "grad_norm": 3.422905459762887,
+      "learning_rate": 2e-05,
+      "loss": 0.0575,
+      "step": 42170
+    },
+    {
+      "epoch": 1.0037846329620108,
+      "grad_norm": 3.3575303766177753,
+      "learning_rate": 2e-05,
+      "loss": 0.1273,
+      "step": 42171
+    },
+    {
+      "epoch": 1.0038084356850423,
+      "grad_norm": 3.433587255145536,
+      "learning_rate": 2e-05,
+      "loss": 0.4608,
+      "step": 42172
+    },
+    {
+      "epoch": 1.0038322384080738,
+      "grad_norm": 4.3752909574568575,
+      "learning_rate": 2e-05,
+      "loss": 0.3777,
+      "step": 42173
+    },
+    {
+      "epoch": 1.0038560411311055,
+      "grad_norm": 2.7159702071381697,
+      "learning_rate": 2e-05,
+      "loss": 0.3873,
+      "step": 42174
+    },
+    {
+      "epoch": 1.003879843854137,
+      "grad_norm": 3.328510030431339,
+      "learning_rate": 2e-05,
+      "loss": 0.2358,
+      "step": 42175
+    },
+    {
+      "epoch": 1.0039036465771685,
+      "grad_norm": 0.40450222847954387,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 42176
+    },
+    {
+      "epoch": 1.0039274493002,
+      "grad_norm": 1.8399651845495952,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 42177
+    },
+    {
+      "epoch": 1.0039512520232314,
+      "grad_norm": 0.3163286988106822,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 42178
+    },
+    {
+      "epoch": 1.003975054746263,
+      "grad_norm": 2.062005793405833,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 42179
+    },
+    {
+      "epoch": 1.0039988574692944,
+      "grad_norm": 5.6955158249079485,
+      "learning_rate": 2e-05,
+      "loss": 0.367,
+      "step": 42180
+    },
+    {
+      "epoch": 1.004022660192326,
+      "grad_norm": 0.2996937104409129,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 42181
+    },
+    {
+      "epoch": 1.0040464629153576,
+      "grad_norm": 0.016599516505854842,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42182
+    },
+    {
+      "epoch": 1.004070265638389,
+      "grad_norm": 0.08465003854345925,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42183
+    },
+    {
+      "epoch": 1.0040940683614206,
+      "grad_norm": 3.1148758480783236,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 42184
+    },
+    {
+      "epoch": 1.004117871084452,
+      "grad_norm": 3.353280373387079,
+      "learning_rate": 2e-05,
+      "loss": 0.3033,
+      "step": 42185
+    },
+    {
+      "epoch": 1.0041416738074835,
+      "grad_norm": 3.02928084023946,
+      "learning_rate": 2e-05,
+      "loss": 0.2751,
+      "step": 42186
+    },
+    {
+      "epoch": 1.004165476530515,
+      "grad_norm": 9.631738248481435,
+      "learning_rate": 2e-05,
+      "loss": 0.572,
+      "step": 42187
+    },
+    {
+      "epoch": 1.0041892792535465,
+      "grad_norm": 0.6824416844814558,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 42188
+    },
+    {
+      "epoch": 1.0042130819765782,
+      "grad_norm": 4.731634113077736,
+      "learning_rate": 2e-05,
+      "loss": 0.6989,
+      "step": 42189
+    },
+    {
+      "epoch": 1.0042368846996097,
+      "grad_norm": 16.32976997418691,
+      "learning_rate": 2e-05,
+      "loss": 0.9685,
+      "step": 42190
+    },
+    {
+      "epoch": 1.0042606874226412,
+      "grad_norm": 4.207825679090327,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 42191
+    },
+    {
+      "epoch": 1.0042844901456727,
+      "grad_norm": 12.874260946695683,
+      "learning_rate": 2e-05,
+      "loss": 0.7877,
+      "step": 42192
+    },
+    {
+      "epoch": 1.0043082928687042,
+      "grad_norm": 2.958599279910122,
+      "learning_rate": 2e-05,
+      "loss": 0.3674,
+      "step": 42193
+    },
+    {
+      "epoch": 1.0043320955917356,
+      "grad_norm": 2.554381941190253,
+      "learning_rate": 2e-05,
+      "loss": 0.228,
+      "step": 42194
+    },
+    {
+      "epoch": 1.0043558983147671,
+      "grad_norm": 13.146148311097196,
+      "learning_rate": 2e-05,
+      "loss": 0.4545,
+      "step": 42195
+    },
+    {
+      "epoch": 1.0043797010377988,
+      "grad_norm": 5.577613907996947,
+      "learning_rate": 2e-05,
+      "loss": 0.668,
+      "step": 42196
+    },
+    {
+      "epoch": 1.0044035037608303,
+      "grad_norm": 8.228665240562195,
+      "learning_rate": 2e-05,
+      "loss": 0.8096,
+      "step": 42197
+    },
+    {
+      "epoch": 1.0044273064838618,
+      "grad_norm": 4.013899037353749,
+      "learning_rate": 2e-05,
+      "loss": 0.5658,
+      "step": 42198
+    },
+    {
+      "epoch": 1.0044511092068933,
+      "grad_norm": 5.694297831267048,
+      "learning_rate": 2e-05,
+      "loss": 1.1176,
+      "step": 42199
+    },
+    {
+      "epoch": 1.0044749119299248,
+      "grad_norm": 7.201621022798028,
+      "learning_rate": 2e-05,
+      "loss": 0.1363,
+      "step": 42200
+    },
+    {
+      "epoch": 1.0044987146529563,
+      "grad_norm": 0.39765550534825767,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 42201
+    },
+    {
+      "epoch": 1.0045225173759877,
+      "grad_norm": 0.555506362883386,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 42202
+    },
+    {
+      "epoch": 1.0045463200990192,
+      "grad_norm": 0.11656531608397301,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42203
+    },
+    {
+      "epoch": 1.004570122822051,
+      "grad_norm": 0.29479316126230004,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 42204
+    },
+    {
+      "epoch": 1.0045939255450824,
+      "grad_norm": 0.797057128809175,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 42205
+    },
+    {
+      "epoch": 1.004617728268114,
+      "grad_norm": 1.5746543443192522,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 42206
+    },
+    {
+      "epoch": 1.0046415309911454,
+      "grad_norm": 2.2548447829658373,
+      "learning_rate": 2e-05,
+      "loss": 0.0471,
+      "step": 42207
+    },
+    {
+      "epoch": 1.0046653337141769,
+      "grad_norm": 3.162339225879179,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 42208
+    },
+    {
+      "epoch": 1.0046891364372084,
+      "grad_norm": 0.7153075688941295,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 42209
+    },
+    {
+      "epoch": 1.0047129391602398,
+      "grad_norm": 3.5138798300631393,
+      "learning_rate": 2e-05,
+      "loss": 0.1685,
+      "step": 42210
+    },
+    {
+      "epoch": 1.0047367418832716,
+      "grad_norm": 2.4768627087215145,
+      "learning_rate": 2e-05,
+      "loss": 0.0509,
+      "step": 42211
+    },
+    {
+      "epoch": 1.004760544606303,
+      "grad_norm": 0.2142522624140924,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 42212
+    },
+    {
+      "epoch": 1.0047843473293345,
+      "grad_norm": 5.951118134577499,
+      "learning_rate": 2e-05,
+      "loss": 0.2214,
+      "step": 42213
+    },
+    {
+      "epoch": 1.004808150052366,
+      "grad_norm": 3.8836144562961064,
+      "learning_rate": 2e-05,
+      "loss": 0.3957,
+      "step": 42214
+    },
+    {
+      "epoch": 1.0048319527753975,
+      "grad_norm": 4.312842018669225,
+      "learning_rate": 2e-05,
+      "loss": 0.7306,
+      "step": 42215
+    },
+    {
+      "epoch": 1.004855755498429,
+      "grad_norm": 2.5018064169420238,
+      "learning_rate": 2e-05,
+      "loss": 0.1844,
+      "step": 42216
+    },
+    {
+      "epoch": 1.0048795582214605,
+      "grad_norm": 2.235786694047728,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 42217
+    },
+    {
+      "epoch": 1.004903360944492,
+      "grad_norm": 11.911836780327006,
+      "learning_rate": 2e-05,
+      "loss": 1.3345,
+      "step": 42218
+    },
+    {
+      "epoch": 1.0049271636675237,
+      "grad_norm": 2.534387912107639,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 42219
+    },
+    {
+      "epoch": 1.0049509663905551,
+      "grad_norm": 3.7833831971942087,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 42220
+    },
+    {
+      "epoch": 1.0049747691135866,
+      "grad_norm": 11.008947407702454,
+      "learning_rate": 2e-05,
+      "loss": 0.539,
+      "step": 42221
+    },
+    {
+      "epoch": 1.0049985718366181,
+      "grad_norm": 2.54922538493345,
+      "learning_rate": 2e-05,
+      "loss": 0.3393,
+      "step": 42222
+    },
+    {
+      "epoch": 1.0050223745596496,
+      "grad_norm": 3.162967858944718,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 42223
+    },
+    {
+      "epoch": 1.005046177282681,
+      "grad_norm": 3.4953095729769887,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 42224
+    },
+    {
+      "epoch": 1.0050699800057126,
+      "grad_norm": 3.7582897456276165,
+      "learning_rate": 2e-05,
+      "loss": 0.4718,
+      "step": 42225
+    },
+    {
+      "epoch": 1.0050937827287443,
+      "grad_norm": 6.965189395613274,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 42226
+    },
+    {
+      "epoch": 1.0051175854517758,
+      "grad_norm": 0.11238282104049026,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 42227
+    },
+    {
+      "epoch": 1.0051413881748072,
+      "grad_norm": 19.09032491535812,
+      "learning_rate": 2e-05,
+      "loss": 0.3176,
+      "step": 42228
+    },
+    {
+      "epoch": 1.0051651908978387,
+      "grad_norm": 2.377197410373405,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 42229
+    },
+    {
+      "epoch": 1.0051889936208702,
+      "grad_norm": 5.68302596518993,
+      "learning_rate": 2e-05,
+      "loss": 0.8637,
+      "step": 42230
+    },
+    {
+      "epoch": 1.0052127963439017,
+      "grad_norm": 2.5627062929495943,
+      "learning_rate": 2e-05,
+      "loss": 0.1734,
+      "step": 42231
+    },
+    {
+      "epoch": 1.0052365990669332,
+      "grad_norm": 0.2768005245374271,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 42232
+    },
+    {
+      "epoch": 1.0052604017899647,
+      "grad_norm": 0.4040263152808112,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 42233
+    },
+    {
+      "epoch": 1.0052842045129964,
+      "grad_norm": 3.0164243743061405,
+      "learning_rate": 2e-05,
+      "loss": 0.3323,
+      "step": 42234
+    },
+    {
+      "epoch": 1.0053080072360279,
+      "grad_norm": 0.03437337643130979,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42235
+    },
+    {
+      "epoch": 1.0053318099590594,
+      "grad_norm": 3.9271772174633215,
+      "learning_rate": 2e-05,
+      "loss": 0.5289,
+      "step": 42236
+    },
+    {
+      "epoch": 1.0053556126820908,
+      "grad_norm": 0.26539743375730934,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 42237
+    },
+    {
+      "epoch": 1.0053794154051223,
+      "grad_norm": 10.325796242307625,
+      "learning_rate": 2e-05,
+      "loss": 0.5404,
+      "step": 42238
+    },
+    {
+      "epoch": 1.0054032181281538,
+      "grad_norm": 4.241448743849926,
+      "learning_rate": 2e-05,
+      "loss": 0.5853,
+      "step": 42239
+    },
+    {
+      "epoch": 1.0054270208511853,
+      "grad_norm": 2.7079130501921376,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 42240
+    },
+    {
+      "epoch": 1.005450823574217,
+      "grad_norm": 5.596678962291144,
+      "learning_rate": 2e-05,
+      "loss": 0.1942,
+      "step": 42241
+    },
+    {
+      "epoch": 1.0054746262972485,
+      "grad_norm": 2.330686085005806,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 42242
+    },
+    {
+      "epoch": 1.00549842902028,
+      "grad_norm": 0.5470172012588335,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 42243
+    },
+    {
+      "epoch": 1.0055222317433115,
+      "grad_norm": 0.7271135453293721,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 42244
+    },
+    {
+      "epoch": 1.005546034466343,
+      "grad_norm": 0.03476575075695612,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42245
+    },
+    {
+      "epoch": 1.0055698371893744,
+      "grad_norm": 5.523476670426594,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 42246
+    },
+    {
+      "epoch": 1.005593639912406,
+      "grad_norm": 3.4170502396803446,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 42247
+    },
+    {
+      "epoch": 1.0056174426354374,
+      "grad_norm": 0.02552624428161439,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42248
+    },
+    {
+      "epoch": 1.005641245358469,
+      "grad_norm": 0.08875295398651607,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42249
+    },
+    {
+      "epoch": 1.0056650480815006,
+      "grad_norm": 0.13775654437453846,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42250
+    },
+    {
+      "epoch": 1.005688850804532,
+      "grad_norm": 1.5452164738401213,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 42251
+    },
+    {
+      "epoch": 1.0057126535275636,
+      "grad_norm": 0.17779134816344874,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 42252
+    },
+    {
+      "epoch": 1.005736456250595,
+      "grad_norm": 3.0437989317064402,
+      "learning_rate": 2e-05,
+      "loss": 0.2606,
+      "step": 42253
+    },
+    {
+      "epoch": 1.0057602589736265,
+      "grad_norm": 2.835829393319778,
+      "learning_rate": 2e-05,
+      "loss": 0.1081,
+      "step": 42254
+    },
+    {
+      "epoch": 1.005784061696658,
+      "grad_norm": 7.764637518422773,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 42255
+    },
+    {
+      "epoch": 1.0058078644196895,
+      "grad_norm": 0.2741886352468886,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42256
+    },
+    {
+      "epoch": 1.0058316671427212,
+      "grad_norm": 0.16609909393512975,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 42257
+    },
+    {
+      "epoch": 1.0058554698657527,
+      "grad_norm": 2.7513961930736297,
+      "learning_rate": 2e-05,
+      "loss": 0.0569,
+      "step": 42258
+    },
+    {
+      "epoch": 1.0058792725887842,
+      "grad_norm": 6.242052952953388,
+      "learning_rate": 2e-05,
+      "loss": 1.2773,
+      "step": 42259
+    },
+    {
+      "epoch": 1.0059030753118157,
+      "grad_norm": 2.5306575144736554,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 42260
+    },
+    {
+      "epoch": 1.0059268780348471,
+      "grad_norm": 2.2411403353408454,
+      "learning_rate": 2e-05,
+      "loss": 0.3286,
+      "step": 42261
+    },
+    {
+      "epoch": 1.0059506807578786,
+      "grad_norm": 5.514345985919839,
+      "learning_rate": 2e-05,
+      "loss": 0.0591,
+      "step": 42262
+    },
+    {
+      "epoch": 1.0059744834809101,
+      "grad_norm": 5.783981565426625,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 42263
+    },
+    {
+      "epoch": 1.0059982862039418,
+      "grad_norm": 10.786633716959145,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 42264
+    },
+    {
+      "epoch": 1.0060220889269733,
+      "grad_norm": 4.717580185967754,
+      "learning_rate": 2e-05,
+      "loss": 0.26,
+      "step": 42265
+    },
+    {
+      "epoch": 1.0060458916500048,
+      "grad_norm": 6.582743253460342,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 42266
+    },
+    {
+      "epoch": 1.0060696943730363,
+      "grad_norm": 2.7422516505169225,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 42267
+    },
+    {
+      "epoch": 1.0060934970960678,
+      "grad_norm": 7.78370988375366,
+      "learning_rate": 2e-05,
+      "loss": 0.3036,
+      "step": 42268
+    },
+    {
+      "epoch": 1.0061172998190993,
+      "grad_norm": 1.809653126480091,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 42269
+    },
+    {
+      "epoch": 1.0061411025421307,
+      "grad_norm": 10.724375176517238,
+      "learning_rate": 2e-05,
+      "loss": 0.7984,
+      "step": 42270
+    },
+    {
+      "epoch": 1.0061649052651624,
+      "grad_norm": 0.25362694471615627,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42271
+    },
+    {
+      "epoch": 1.006188707988194,
+      "grad_norm": 1.3748238356181894,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 42272
+    },
+    {
+      "epoch": 1.0062125107112254,
+      "grad_norm": 0.37939706024532155,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 42273
+    },
+    {
+      "epoch": 1.006236313434257,
+      "grad_norm": 3.846034205162699,
+      "learning_rate": 2e-05,
+      "loss": 0.2633,
+      "step": 42274
+    },
+    {
+      "epoch": 1.0062601161572884,
+      "grad_norm": 11.571719585853767,
+      "learning_rate": 2e-05,
+      "loss": 0.2263,
+      "step": 42275
+    },
+    {
+      "epoch": 1.0062839188803199,
+      "grad_norm": 3.7379165638316945,
+      "learning_rate": 2e-05,
+      "loss": 0.0283,
+      "step": 42276
+    },
+    {
+      "epoch": 1.0063077216033514,
+      "grad_norm": 3.2869641981249806,
+      "learning_rate": 2e-05,
+      "loss": 0.3142,
+      "step": 42277
+    },
+    {
+      "epoch": 1.0063315243263828,
+      "grad_norm": 0.1383018998199551,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 42278
+    },
+    {
+      "epoch": 1.0063553270494145,
+      "grad_norm": 0.10193598651183568,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 42279
+    },
+    {
+      "epoch": 1.006379129772446,
+      "grad_norm": 3.9632571040063937,
+      "learning_rate": 2e-05,
+      "loss": 0.4577,
+      "step": 42280
+    },
+    {
+      "epoch": 1.0064029324954775,
+      "grad_norm": 7.769471569712765,
+      "learning_rate": 2e-05,
+      "loss": 0.5665,
+      "step": 42281
+    },
+    {
+      "epoch": 1.006426735218509,
+      "grad_norm": 5.013829397026867,
+      "learning_rate": 2e-05,
+      "loss": 0.9713,
+      "step": 42282
+    },
+    {
+      "epoch": 1.0064505379415405,
+      "grad_norm": 1.798232554682655,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 42283
+    },
+    {
+      "epoch": 1.006474340664572,
+      "grad_norm": 5.543485205962221,
+      "learning_rate": 2e-05,
+      "loss": 0.648,
+      "step": 42284
+    },
+    {
+      "epoch": 1.0064981433876035,
+      "grad_norm": 5.253040664552662,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 42285
+    },
+    {
+      "epoch": 1.006521946110635,
+      "grad_norm": 0.7234048788442183,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 42286
+    },
+    {
+      "epoch": 1.0065457488336667,
+      "grad_norm": 0.23524166468610483,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 42287
+    },
+    {
+      "epoch": 1.0065695515566981,
+      "grad_norm": 3.790375307181171,
+      "learning_rate": 2e-05,
+      "loss": 0.3872,
+      "step": 42288
+    },
+    {
+      "epoch": 1.0065933542797296,
+      "grad_norm": 3.4295991743927847,
+      "learning_rate": 2e-05,
+      "loss": 0.0517,
+      "step": 42289
+    },
+    {
+      "epoch": 1.006617157002761,
+      "grad_norm": 0.3056713890634817,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 42290
+    },
+    {
+      "epoch": 1.0066409597257926,
+      "grad_norm": 8.431289247171165,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 42291
+    },
+    {
+      "epoch": 1.006664762448824,
+      "grad_norm": 0.11202912318761059,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42292
+    },
+    {
+      "epoch": 1.0066885651718556,
+      "grad_norm": 5.840350914018151,
+      "learning_rate": 2e-05,
+      "loss": 0.5865,
+      "step": 42293
+    },
+    {
+      "epoch": 1.0067123678948873,
+      "grad_norm": 3.7231501507956466,
+      "learning_rate": 2e-05,
+      "loss": 0.4378,
+      "step": 42294
+    },
+    {
+      "epoch": 1.0067361706179188,
+      "grad_norm": 3.0849033519231313,
+      "learning_rate": 2e-05,
+      "loss": 0.3244,
+      "step": 42295
+    },
+    {
+      "epoch": 1.0067599733409502,
+      "grad_norm": 9.128748969676199,
+      "learning_rate": 2e-05,
+      "loss": 0.1525,
+      "step": 42296
+    },
+    {
+      "epoch": 1.0067837760639817,
+      "grad_norm": 11.425968980238265,
+      "learning_rate": 2e-05,
+      "loss": 0.6728,
+      "step": 42297
+    },
+    {
+      "epoch": 1.0068075787870132,
+      "grad_norm": 27.563409950146735,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 42298
+    },
+    {
+      "epoch": 1.0068313815100447,
+      "grad_norm": 0.5135303234018002,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 42299
+    },
+    {
+      "epoch": 1.0068551842330762,
+      "grad_norm": 0.11782260730842964,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 42300
+    },
+    {
+      "epoch": 1.0068789869561077,
+      "grad_norm": 0.09694335522764587,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42301
+    },
+    {
+      "epoch": 1.0069027896791394,
+      "grad_norm": 2.272071307678006,
+      "learning_rate": 2e-05,
+      "loss": 0.1158,
+      "step": 42302
+    },
+    {
+      "epoch": 1.0069265924021709,
+      "grad_norm": 5.096419654571333,
+      "learning_rate": 2e-05,
+      "loss": 0.2071,
+      "step": 42303
+    },
+    {
+      "epoch": 1.0069503951252023,
+      "grad_norm": 12.995069954187098,
+      "learning_rate": 2e-05,
+      "loss": 1.0672,
+      "step": 42304
+    },
+    {
+      "epoch": 1.0069741978482338,
+      "grad_norm": 1.0655521427248327,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 42305
+    },
+    {
+      "epoch": 1.0069980005712653,
+      "grad_norm": 3.515576730130969,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 42306
+    },
+    {
+      "epoch": 1.0070218032942968,
+      "grad_norm": 4.624322734301888,
+      "learning_rate": 2e-05,
+      "loss": 0.6665,
+      "step": 42307
+    },
+    {
+      "epoch": 1.0070456060173283,
+      "grad_norm": 0.03362316113903618,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42308
+    },
+    {
+      "epoch": 1.00706940874036,
+      "grad_norm": 9.230527633407087,
+      "learning_rate": 2e-05,
+      "loss": 0.492,
+      "step": 42309
+    },
+    {
+      "epoch": 1.0070932114633915,
+      "grad_norm": 2.682848986338659,
+      "learning_rate": 2e-05,
+      "loss": 0.0285,
+      "step": 42310
+    },
+    {
+      "epoch": 1.007117014186423,
+      "grad_norm": 5.440577619290911,
+      "learning_rate": 2e-05,
+      "loss": 0.5618,
+      "step": 42311
+    },
+    {
+      "epoch": 1.0071408169094545,
+      "grad_norm": 3.1648973059700163,
+      "learning_rate": 2e-05,
+      "loss": 0.0508,
+      "step": 42312
+    },
+    {
+      "epoch": 1.007164619632486,
+      "grad_norm": 2.801313803270453,
+      "learning_rate": 2e-05,
+      "loss": 0.1674,
+      "step": 42313
+    },
+    {
+      "epoch": 1.0071884223555174,
+      "grad_norm": 4.007962310874104,
+      "learning_rate": 2e-05,
+      "loss": 0.4194,
+      "step": 42314
+    },
+    {
+      "epoch": 1.007212225078549,
+      "grad_norm": 4.305263551088163,
+      "learning_rate": 2e-05,
+      "loss": 0.5223,
+      "step": 42315
+    },
+    {
+      "epoch": 1.0072360278015804,
+      "grad_norm": 8.953463541247478,
+      "learning_rate": 2e-05,
+      "loss": 0.4663,
+      "step": 42316
+    },
+    {
+      "epoch": 1.007259830524612,
+      "grad_norm": 4.5665825513748395,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 42317
+    },
+    {
+      "epoch": 1.0072836332476436,
+      "grad_norm": 0.07901725697798809,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42318
+    },
+    {
+      "epoch": 1.007307435970675,
+      "grad_norm": 3.5207368506869,
+      "learning_rate": 2e-05,
+      "loss": 0.0616,
+      "step": 42319
+    },
+    {
+      "epoch": 1.0073312386937066,
+      "grad_norm": 6.525097820661662,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 42320
+    },
+    {
+      "epoch": 1.007355041416738,
+      "grad_norm": 4.817863206263378,
+      "learning_rate": 2e-05,
+      "loss": 0.7393,
+      "step": 42321
+    },
+    {
+      "epoch": 1.0073788441397695,
+      "grad_norm": 1.3905488562990966,
+      "learning_rate": 2e-05,
+      "loss": 0.0311,
+      "step": 42322
+    },
+    {
+      "epoch": 1.007402646862801,
+      "grad_norm": 1.071599571174291,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 42323
+    },
+    {
+      "epoch": 1.0074264495858327,
+      "grad_norm": 0.2613882640898946,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 42324
+    },
+    {
+      "epoch": 1.0074502523088642,
+      "grad_norm": 2.654085228907959,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 42325
+    },
+    {
+      "epoch": 1.0074740550318957,
+      "grad_norm": 4.163512520024912,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 42326
+    },
+    {
+      "epoch": 1.0074978577549272,
+      "grad_norm": 5.893029985743723,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 42327
+    },
+    {
+      "epoch": 1.0075216604779587,
+      "grad_norm": 0.02246767734689774,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42328
+    },
+    {
+      "epoch": 1.0075454632009901,
+      "grad_norm": 6.739663928694314,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 42329
+    },
+    {
+      "epoch": 1.0075692659240216,
+      "grad_norm": 2.7113476434258788,
+      "learning_rate": 2e-05,
+      "loss": 0.4254,
+      "step": 42330
+    },
+    {
+      "epoch": 1.0075930686470531,
+      "grad_norm": 12.818722362414762,
+      "learning_rate": 2e-05,
+      "loss": 0.8907,
+      "step": 42331
+    },
+    {
+      "epoch": 1.0076168713700848,
+      "grad_norm": 5.317125085021502,
+      "learning_rate": 2e-05,
+      "loss": 0.7783,
+      "step": 42332
+    },
+    {
+      "epoch": 1.0076406740931163,
+      "grad_norm": 8.619141352161783,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 42333
+    },
+    {
+      "epoch": 1.0076644768161478,
+      "grad_norm": 6.501249479193873,
+      "learning_rate": 2e-05,
+      "loss": 0.9019,
+      "step": 42334
+    },
+    {
+      "epoch": 1.0076882795391793,
+      "grad_norm": 9.267798269183732,
+      "learning_rate": 2e-05,
+      "loss": 0.3253,
+      "step": 42335
+    },
+    {
+      "epoch": 1.0077120822622108,
+      "grad_norm": 5.586738164985266,
+      "learning_rate": 2e-05,
+      "loss": 1.0639,
+      "step": 42336
+    },
+    {
+      "epoch": 1.0077358849852422,
+      "grad_norm": 4.80921530483416,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 42337
+    },
+    {
+      "epoch": 1.0077596877082737,
+      "grad_norm": 61.84987101288701,
+      "learning_rate": 2e-05,
+      "loss": 0.6301,
+      "step": 42338
+    },
+    {
+      "epoch": 1.0077834904313054,
+      "grad_norm": 0.7329686534250709,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 42339
+    },
+    {
+      "epoch": 1.007807293154337,
+      "grad_norm": 0.7528666098403881,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 42340
+    },
+    {
+      "epoch": 1.0078310958773684,
+      "grad_norm": 3.1790260626558298,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 42341
+    },
+    {
+      "epoch": 1.0078548986004,
+      "grad_norm": 4.500330956850518,
+      "learning_rate": 2e-05,
+      "loss": 0.5979,
+      "step": 42342
+    },
+    {
+      "epoch": 1.0078787013234314,
+      "grad_norm": 5.243662534195253,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 42343
+    },
+    {
+      "epoch": 1.0079025040464629,
+      "grad_norm": 6.597134420376995,
+      "learning_rate": 2e-05,
+      "loss": 0.1327,
+      "step": 42344
+    },
+    {
+      "epoch": 1.0079263067694944,
+      "grad_norm": 7.022756048402097,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 42345
+    },
+    {
+      "epoch": 1.0079501094925258,
+      "grad_norm": 3.9876844185654323,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 42346
+    },
+    {
+      "epoch": 1.0079739122155575,
+      "grad_norm": 0.10690690140400315,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 42347
+    },
+    {
+      "epoch": 1.007997714938589,
+      "grad_norm": 4.224879098932283,
+      "learning_rate": 2e-05,
+      "loss": 0.5465,
+      "step": 42348
+    },
+    {
+      "epoch": 1.0080215176616205,
+      "grad_norm": 3.7583938873977574,
+      "learning_rate": 2e-05,
+      "loss": 0.1154,
+      "step": 42349
+    },
+    {
+      "epoch": 1.008045320384652,
+      "grad_norm": 4.565652956722255,
+      "learning_rate": 2e-05,
+      "loss": 0.0398,
+      "step": 42350
+    },
+    {
+      "epoch": 1.0080691231076835,
+      "grad_norm": 5.57474346719161,
+      "learning_rate": 2e-05,
+      "loss": 0.8319,
+      "step": 42351
+    },
+    {
+      "epoch": 1.008092925830715,
+      "grad_norm": 0.01577681584432005,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42352
+    },
+    {
+      "epoch": 1.0081167285537465,
+      "grad_norm": 5.110006697688142,
+      "learning_rate": 2e-05,
+      "loss": 0.0597,
+      "step": 42353
+    },
+    {
+      "epoch": 1.0081405312767782,
+      "grad_norm": 3.1262018997049257,
+      "learning_rate": 2e-05,
+      "loss": 0.3336,
+      "step": 42354
+    },
+    {
+      "epoch": 1.0081643339998096,
+      "grad_norm": 0.1625623707037543,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 42355
+    },
+    {
+      "epoch": 1.0081881367228411,
+      "grad_norm": 0.09948701517569353,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 42356
+    },
+    {
+      "epoch": 1.0082119394458726,
+      "grad_norm": 8.310463480644382,
+      "learning_rate": 2e-05,
+      "loss": 0.6224,
+      "step": 42357
+    },
+    {
+      "epoch": 1.008235742168904,
+      "grad_norm": 3.4943705234511713,
+      "learning_rate": 2e-05,
+      "loss": 0.3074,
+      "step": 42358
+    },
+    {
+      "epoch": 1.0082595448919356,
+      "grad_norm": 3.6097960730741763,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 42359
+    },
+    {
+      "epoch": 1.008283347614967,
+      "grad_norm": 4.109018668183361,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 42360
+    },
+    {
+      "epoch": 1.0083071503379986,
+      "grad_norm": 4.401836393857154,
+      "learning_rate": 2e-05,
+      "loss": 0.0292,
+      "step": 42361
+    },
+    {
+      "epoch": 1.0083309530610303,
+      "grad_norm": 0.6813349459206367,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 42362
+    },
+    {
+      "epoch": 1.0083547557840618,
+      "grad_norm": 1.6148577699447635,
+      "learning_rate": 2e-05,
+      "loss": 0.021,
+      "step": 42363
+    },
+    {
+      "epoch": 1.0083785585070932,
+      "grad_norm": 4.204427901051843,
+      "learning_rate": 2e-05,
+      "loss": 0.1032,
+      "step": 42364
+    },
+    {
+      "epoch": 1.0084023612301247,
+      "grad_norm": 4.115102834291107,
+      "learning_rate": 2e-05,
+      "loss": 0.0483,
+      "step": 42365
+    },
+    {
+      "epoch": 1.0084261639531562,
+      "grad_norm": 0.06844259491224776,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 42366
+    },
+    {
+      "epoch": 1.0084499666761877,
+      "grad_norm": 1.129827685518197,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 42367
+    },
+    {
+      "epoch": 1.0084737693992192,
+      "grad_norm": 3.253228033019678,
+      "learning_rate": 2e-05,
+      "loss": 0.3309,
+      "step": 42368
+    },
+    {
+      "epoch": 1.0084975721222509,
+      "grad_norm": 1.3023205714242985,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 42369
+    },
+    {
+      "epoch": 1.0085213748452824,
+      "grad_norm": 1.0364368648092257,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 42370
+    },
+    {
+      "epoch": 1.0085451775683139,
+      "grad_norm": 0.11292725616855476,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42371
+    },
+    {
+      "epoch": 1.0085689802913453,
+      "grad_norm": 0.027246309162834242,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42372
+    },
+    {
+      "epoch": 1.0085927830143768,
+      "grad_norm": 0.49414362056351113,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 42373
+    },
+    {
+      "epoch": 1.0086165857374083,
+      "grad_norm": 0.3501622407698159,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 42374
+    },
+    {
+      "epoch": 1.0086403884604398,
+      "grad_norm": 0.007654396337376773,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42375
+    },
+    {
+      "epoch": 1.0086641911834713,
+      "grad_norm": 2.822603718200275,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 42376
+    },
+    {
+      "epoch": 1.008687993906503,
+      "grad_norm": 3.7426536091704565,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 42377
+    },
+    {
+      "epoch": 1.0087117966295345,
+      "grad_norm": 0.0321172915906319,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 42378
+    },
+    {
+      "epoch": 1.008735599352566,
+      "grad_norm": 0.09946815745636017,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42379
+    },
+    {
+      "epoch": 1.0087594020755974,
+      "grad_norm": 3.0545730064942096,
+      "learning_rate": 2e-05,
+      "loss": 0.3074,
+      "step": 42380
+    },
+    {
+      "epoch": 1.008783204798629,
+      "grad_norm": 0.018767152640807912,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42381
+    },
+    {
+      "epoch": 1.0088070075216604,
+      "grad_norm": 4.611721783236999,
+      "learning_rate": 2e-05,
+      "loss": 0.4882,
+      "step": 42382
+    },
+    {
+      "epoch": 1.008830810244692,
+      "grad_norm": 0.010572590148166338,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42383
+    },
+    {
+      "epoch": 1.0088546129677236,
+      "grad_norm": 7.073313741433177,
+      "learning_rate": 2e-05,
+      "loss": 0.4626,
+      "step": 42384
+    },
+    {
+      "epoch": 1.008878415690755,
+      "grad_norm": 12.95095780710531,
+      "learning_rate": 2e-05,
+      "loss": 1.0502,
+      "step": 42385
+    },
+    {
+      "epoch": 1.0089022184137866,
+      "grad_norm": 3.6607839404107754,
+      "learning_rate": 2e-05,
+      "loss": 0.4378,
+      "step": 42386
+    },
+    {
+      "epoch": 1.008926021136818,
+      "grad_norm": 0.013623961903558945,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42387
+    },
+    {
+      "epoch": 1.0089498238598495,
+      "grad_norm": 2.7454353404997742,
+      "learning_rate": 2e-05,
+      "loss": 0.3892,
+      "step": 42388
+    },
+    {
+      "epoch": 1.008973626582881,
+      "grad_norm": 5.614008926199676,
+      "learning_rate": 2e-05,
+      "loss": 0.6775,
+      "step": 42389
+    },
+    {
+      "epoch": 1.0089974293059125,
+      "grad_norm": 3.1226207330209594,
+      "learning_rate": 2e-05,
+      "loss": 0.4299,
+      "step": 42390
+    },
+    {
+      "epoch": 1.009021232028944,
+      "grad_norm": 2.8615022630361167,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 42391
+    },
+    {
+      "epoch": 1.0090450347519757,
+      "grad_norm": 2.7056416970685855,
+      "learning_rate": 2e-05,
+      "loss": 0.2272,
+      "step": 42392
+    },
+    {
+      "epoch": 1.0090688374750072,
+      "grad_norm": 6.499388433311442,
+      "learning_rate": 2e-05,
+      "loss": 0.1345,
+      "step": 42393
+    },
+    {
+      "epoch": 1.0090926401980387,
+      "grad_norm": 2.5586542818609526,
+      "learning_rate": 2e-05,
+      "loss": 0.285,
+      "step": 42394
+    },
+    {
+      "epoch": 1.0091164429210702,
+      "grad_norm": 0.10656019883513898,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42395
+    },
+    {
+      "epoch": 1.0091402456441017,
+      "grad_norm": 3.0542785555339997,
+      "learning_rate": 2e-05,
+      "loss": 0.192,
+      "step": 42396
+    },
+    {
+      "epoch": 1.0091640483671331,
+      "grad_norm": 0.010436478895721609,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42397
+    },
+    {
+      "epoch": 1.0091878510901646,
+      "grad_norm": 4.275549330906933,
+      "learning_rate": 2e-05,
+      "loss": 0.3198,
+      "step": 42398
+    },
+    {
+      "epoch": 1.0092116538131963,
+      "grad_norm": 2.974322219469216,
+      "learning_rate": 2e-05,
+      "loss": 0.3433,
+      "step": 42399
+    },
+    {
+      "epoch": 1.0092354565362278,
+      "grad_norm": 3.394562178478522,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 42400
+    },
+    {
+      "epoch": 1.0092592592592593,
+      "grad_norm": 11.242545420515777,
+      "learning_rate": 2e-05,
+      "loss": 0.3234,
+      "step": 42401
+    },
+    {
+      "epoch": 1.0092830619822908,
+      "grad_norm": 3.5698186500526727,
+      "learning_rate": 2e-05,
+      "loss": 0.3611,
+      "step": 42402
+    },
+    {
+      "epoch": 1.0093068647053223,
+      "grad_norm": 0.039407487414933544,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42403
+    },
+    {
+      "epoch": 1.0093306674283538,
+      "grad_norm": 5.976493597736415,
+      "learning_rate": 2e-05,
+      "loss": 0.6763,
+      "step": 42404
+    },
+    {
+      "epoch": 1.0093544701513852,
+      "grad_norm": 9.288707667965172,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 42405
+    },
+    {
+      "epoch": 1.0093782728744167,
+      "grad_norm": 1.2781228859160556,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 42406
+    },
+    {
+      "epoch": 1.0094020755974484,
+      "grad_norm": 0.013987479631737826,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42407
+    },
+    {
+      "epoch": 1.00942587832048,
+      "grad_norm": 1.9734641169990055,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 42408
+    },
+    {
+      "epoch": 1.0094496810435114,
+      "grad_norm": 3.6752008265050877,
+      "learning_rate": 2e-05,
+      "loss": 0.4384,
+      "step": 42409
+    },
+    {
+      "epoch": 1.009473483766543,
+      "grad_norm": 3.3937416478381297,
+      "learning_rate": 2e-05,
+      "loss": 0.109,
+      "step": 42410
+    },
+    {
+      "epoch": 1.0094972864895744,
+      "grad_norm": 3.3666465212415906,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 42411
+    },
+    {
+      "epoch": 1.0095210892126059,
+      "grad_norm": 3.125834588519598,
+      "learning_rate": 2e-05,
+      "loss": 0.2804,
+      "step": 42412
+    },
+    {
+      "epoch": 1.0095448919356373,
+      "grad_norm": 0.018709426947779353,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42413
+    },
+    {
+      "epoch": 1.009568694658669,
+      "grad_norm": 2.5097345906727053,
+      "learning_rate": 2e-05,
+      "loss": 0.0631,
+      "step": 42414
+    },
+    {
+      "epoch": 1.0095924973817005,
+      "grad_norm": 3.6380114625863147,
+      "learning_rate": 2e-05,
+      "loss": 0.4579,
+      "step": 42415
+    },
+    {
+      "epoch": 1.009616300104732,
+      "grad_norm": 2.5124096918803134,
+      "learning_rate": 2e-05,
+      "loss": 0.198,
+      "step": 42416
+    },
+    {
+      "epoch": 1.0096401028277635,
+      "grad_norm": 0.3685234897780209,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 42417
+    },
+    {
+      "epoch": 1.009663905550795,
+      "grad_norm": 11.625612340162181,
+      "learning_rate": 2e-05,
+      "loss": 0.7489,
+      "step": 42418
+    },
+    {
+      "epoch": 1.0096877082738265,
+      "grad_norm": 0.06687259691688023,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42419
+    },
+    {
+      "epoch": 1.009711510996858,
+      "grad_norm": 3.00601701161061,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 42420
+    },
+    {
+      "epoch": 1.0097353137198894,
+      "grad_norm": 1.9204783786123398,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 42421
+    },
+    {
+      "epoch": 1.0097591164429212,
+      "grad_norm": 9.53440714637073,
+      "learning_rate": 2e-05,
+      "loss": 0.711,
+      "step": 42422
+    },
+    {
+      "epoch": 1.0097829191659526,
+      "grad_norm": 7.761049161842319,
+      "learning_rate": 2e-05,
+      "loss": 0.7353,
+      "step": 42423
+    },
+    {
+      "epoch": 1.0098067218889841,
+      "grad_norm": 3.6861556680635785,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 42424
+    },
+    {
+      "epoch": 1.0098305246120156,
+      "grad_norm": 4.838832298549031,
+      "learning_rate": 2e-05,
+      "loss": 0.4869,
+      "step": 42425
+    },
+    {
+      "epoch": 1.009854327335047,
+      "grad_norm": 3.1440852919054834,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 42426
+    },
+    {
+      "epoch": 1.0098781300580786,
+      "grad_norm": 2.2020018929769543,
+      "learning_rate": 2e-05,
+      "loss": 0.0479,
+      "step": 42427
+    },
+    {
+      "epoch": 1.00990193278111,
+      "grad_norm": 4.504952017184662,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 42428
+    },
+    {
+      "epoch": 1.0099257355041418,
+      "grad_norm": 3.0767824029291844,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 42429
+    },
+    {
+      "epoch": 1.0099495382271733,
+      "grad_norm": 13.255713669579501,
+      "learning_rate": 2e-05,
+      "loss": 0.4883,
+      "step": 42430
+    },
+    {
+      "epoch": 1.0099733409502047,
+      "grad_norm": 3.7727829091295546,
+      "learning_rate": 2e-05,
+      "loss": 0.3919,
+      "step": 42431
+    },
+    {
+      "epoch": 1.0099971436732362,
+      "grad_norm": 6.709035534648655,
+      "learning_rate": 2e-05,
+      "loss": 0.569,
+      "step": 42432
+    },
+    {
+      "epoch": 1.0100209463962677,
+      "grad_norm": 0.021448853782386684,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42433
+    },
+    {
+      "epoch": 1.0100447491192992,
+      "grad_norm": 0.03788017692427141,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42434
+    },
+    {
+      "epoch": 1.0100685518423307,
+      "grad_norm": 2.6826724409555402,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 42435
+    },
+    {
+      "epoch": 1.0100923545653622,
+      "grad_norm": 0.07924577185920523,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42436
+    },
+    {
+      "epoch": 1.0101161572883939,
+      "grad_norm": 0.7515474812092797,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 42437
+    },
+    {
+      "epoch": 1.0101399600114254,
+      "grad_norm": 8.580453902914437,
+      "learning_rate": 2e-05,
+      "loss": 1.5757,
+      "step": 42438
+    },
+    {
+      "epoch": 1.0101637627344568,
+      "grad_norm": 0.6069126016665767,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 42439
+    },
+    {
+      "epoch": 1.0101875654574883,
+      "grad_norm": 0.14613157570385832,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42440
+    },
+    {
+      "epoch": 1.0102113681805198,
+      "grad_norm": 3.1185722119697616,
+      "learning_rate": 2e-05,
+      "loss": 0.4855,
+      "step": 42441
+    },
+    {
+      "epoch": 1.0102351709035513,
+      "grad_norm": 0.02382023546837013,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42442
+    },
+    {
+      "epoch": 1.0102589736265828,
+      "grad_norm": 2.4501740190461545,
+      "learning_rate": 2e-05,
+      "loss": 0.1539,
+      "step": 42443
+    },
+    {
+      "epoch": 1.0102827763496145,
+      "grad_norm": 7.583647172785432,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 42444
+    },
+    {
+      "epoch": 1.010306579072646,
+      "grad_norm": 3.988453828958679,
+      "learning_rate": 2e-05,
+      "loss": 0.3644,
+      "step": 42445
+    },
+    {
+      "epoch": 1.0103303817956775,
+      "grad_norm": 2.669337079236396,
+      "learning_rate": 2e-05,
+      "loss": 0.4117,
+      "step": 42446
+    },
+    {
+      "epoch": 1.010354184518709,
+      "grad_norm": 3.6371452672905367,
+      "learning_rate": 2e-05,
+      "loss": 0.2323,
+      "step": 42447
+    },
+    {
+      "epoch": 1.0103779872417404,
+      "grad_norm": 0.07600451339921242,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 42448
+    },
+    {
+      "epoch": 1.010401789964772,
+      "grad_norm": 1.8969205280647194,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 42449
+    },
+    {
+      "epoch": 1.0104255926878034,
+      "grad_norm": 0.0349779537787417,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42450
+    },
+    {
+      "epoch": 1.010449395410835,
+      "grad_norm": 1.0803899864739535,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 42451
+    },
+    {
+      "epoch": 1.0104731981338666,
+      "grad_norm": 2.2021850706517636,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 42452
+    },
+    {
+      "epoch": 1.010497000856898,
+      "grad_norm": 0.6036728331460612,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 42453
+    },
+    {
+      "epoch": 1.0105208035799296,
+      "grad_norm": 3.6446022791586445,
+      "learning_rate": 2e-05,
+      "loss": 0.3812,
+      "step": 42454
+    },
+    {
+      "epoch": 1.010544606302961,
+      "grad_norm": 3.834686493484153,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 42455
+    },
+    {
+      "epoch": 1.0105684090259925,
+      "grad_norm": 7.978744145455109,
+      "learning_rate": 2e-05,
+      "loss": 0.3198,
+      "step": 42456
+    },
+    {
+      "epoch": 1.010592211749024,
+      "grad_norm": 2.777121250192738,
+      "learning_rate": 2e-05,
+      "loss": 0.1874,
+      "step": 42457
+    },
+    {
+      "epoch": 1.0106160144720555,
+      "grad_norm": 4.135024940646227,
+      "learning_rate": 2e-05,
+      "loss": 0.542,
+      "step": 42458
+    },
+    {
+      "epoch": 1.0106398171950872,
+      "grad_norm": 0.028113245367646537,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42459
+    },
+    {
+      "epoch": 1.0106636199181187,
+      "grad_norm": 4.147762787294909,
+      "learning_rate": 2e-05,
+      "loss": 0.3905,
+      "step": 42460
+    },
+    {
+      "epoch": 1.0106874226411502,
+      "grad_norm": 5.351193869469386,
+      "learning_rate": 2e-05,
+      "loss": 0.1524,
+      "step": 42461
+    },
+    {
+      "epoch": 1.0107112253641817,
+      "grad_norm": 2.422447225464298,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 42462
+    },
+    {
+      "epoch": 1.0107350280872132,
+      "grad_norm": 3.050556214475047,
+      "learning_rate": 2e-05,
+      "loss": 0.4273,
+      "step": 42463
+    },
+    {
+      "epoch": 1.0107588308102446,
+      "grad_norm": 6.038610562292157,
+      "learning_rate": 2e-05,
+      "loss": 0.2984,
+      "step": 42464
+    },
+    {
+      "epoch": 1.0107826335332761,
+      "grad_norm": 3.5384038416936137,
+      "learning_rate": 2e-05,
+      "loss": 0.3507,
+      "step": 42465
+    },
+    {
+      "epoch": 1.0108064362563076,
+      "grad_norm": 3.014575184597992,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 42466
+    },
+    {
+      "epoch": 1.0108302389793393,
+      "grad_norm": 0.7815303889717945,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 42467
+    },
+    {
+      "epoch": 1.0108540417023708,
+      "grad_norm": 2.5711063055766865,
+      "learning_rate": 2e-05,
+      "loss": 0.0928,
+      "step": 42468
+    },
+    {
+      "epoch": 1.0108778444254023,
+      "grad_norm": 3.76889480430755,
+      "learning_rate": 2e-05,
+      "loss": 0.5931,
+      "step": 42469
+    },
+    {
+      "epoch": 1.0109016471484338,
+      "grad_norm": 1.74239696799821,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 42470
+    },
+    {
+      "epoch": 1.0109254498714653,
+      "grad_norm": 0.038354997360598775,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42471
+    },
+    {
+      "epoch": 1.0109492525944968,
+      "grad_norm": 0.048333055210439305,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42472
+    },
+    {
+      "epoch": 1.0109730553175282,
+      "grad_norm": 8.330603547515171,
+      "learning_rate": 2e-05,
+      "loss": 0.9587,
+      "step": 42473
+    },
+    {
+      "epoch": 1.01099685804056,
+      "grad_norm": 0.6079390973225993,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 42474
+    },
+    {
+      "epoch": 1.0110206607635914,
+      "grad_norm": 4.248385146009237,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 42475
+    },
+    {
+      "epoch": 1.011044463486623,
+      "grad_norm": 5.225368861863647,
+      "learning_rate": 2e-05,
+      "loss": 0.5925,
+      "step": 42476
+    },
+    {
+      "epoch": 1.0110682662096544,
+      "grad_norm": 3.065228723506821,
+      "learning_rate": 2e-05,
+      "loss": 0.4233,
+      "step": 42477
+    },
+    {
+      "epoch": 1.0110920689326859,
+      "grad_norm": 0.5377529938703927,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 42478
+    },
+    {
+      "epoch": 1.0111158716557174,
+      "grad_norm": 0.008625037926680627,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42479
+    },
+    {
+      "epoch": 1.0111396743787489,
+      "grad_norm": 2.429797139040663,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 42480
+    },
+    {
+      "epoch": 1.0111634771017803,
+      "grad_norm": 3.119060682272455,
+      "learning_rate": 2e-05,
+      "loss": 0.2079,
+      "step": 42481
+    },
+    {
+      "epoch": 1.011187279824812,
+      "grad_norm": 0.4865684454253946,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 42482
+    },
+    {
+      "epoch": 1.0112110825478435,
+      "grad_norm": 0.09481810432790068,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 42483
+    },
+    {
+      "epoch": 1.011234885270875,
+      "grad_norm": 14.954806260440428,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 42484
+    },
+    {
+      "epoch": 1.0112586879939065,
+      "grad_norm": 3.0659366366622365,
+      "learning_rate": 2e-05,
+      "loss": 0.263,
+      "step": 42485
+    },
+    {
+      "epoch": 1.011282490716938,
+      "grad_norm": 0.0951829985886385,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42486
+    },
+    {
+      "epoch": 1.0113062934399695,
+      "grad_norm": 0.09754505639323204,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42487
+    },
+    {
+      "epoch": 1.011330096163001,
+      "grad_norm": 1.4595142813821194,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 42488
+    },
+    {
+      "epoch": 1.0113538988860327,
+      "grad_norm": 3.3111286215004365,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 42489
+    },
+    {
+      "epoch": 1.0113777016090642,
+      "grad_norm": 2.2109123883374027,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 42490
+    },
+    {
+      "epoch": 1.0114015043320956,
+      "grad_norm": 11.461555299689584,
+      "learning_rate": 2e-05,
+      "loss": 0.3687,
+      "step": 42491
+    },
+    {
+      "epoch": 1.0114253070551271,
+      "grad_norm": 2.750432831496829,
+      "learning_rate": 2e-05,
+      "loss": 0.2866,
+      "step": 42492
+    },
+    {
+      "epoch": 1.0114491097781586,
+      "grad_norm": 11.037366780649709,
+      "learning_rate": 2e-05,
+      "loss": 0.2232,
+      "step": 42493
+    },
+    {
+      "epoch": 1.01147291250119,
+      "grad_norm": 3.4413175026876712,
+      "learning_rate": 2e-05,
+      "loss": 0.3223,
+      "step": 42494
+    },
+    {
+      "epoch": 1.0114967152242216,
+      "grad_norm": 0.6894841297846946,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 42495
+    },
+    {
+      "epoch": 1.011520517947253,
+      "grad_norm": 52.07263112352999,
+      "learning_rate": 2e-05,
+      "loss": 0.6059,
+      "step": 42496
+    },
+    {
+      "epoch": 1.0115443206702848,
+      "grad_norm": 2.8249859043676686,
+      "learning_rate": 2e-05,
+      "loss": 0.449,
+      "step": 42497
+    },
+    {
+      "epoch": 1.0115681233933163,
+      "grad_norm": 3.6715197903786843,
+      "learning_rate": 2e-05,
+      "loss": 0.2542,
+      "step": 42498
+    },
+    {
+      "epoch": 1.0115919261163477,
+      "grad_norm": 0.013908304155927819,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42499
+    },
+    {
+      "epoch": 1.0116157288393792,
+      "grad_norm": 1.4031440770960248,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 42500
+    },
+    {
+      "epoch": 1.0116395315624107,
+      "grad_norm": 0.012977662799839628,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42501
+    },
+    {
+      "epoch": 1.0116633342854422,
+      "grad_norm": 1.0596501073575544,
+      "learning_rate": 2e-05,
+      "loss": 0.0176,
+      "step": 42502
+    },
+    {
+      "epoch": 1.0116871370084737,
+      "grad_norm": 0.13816350156778123,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 42503
+    },
+    {
+      "epoch": 1.0117109397315054,
+      "grad_norm": 24.280950955987215,
+      "learning_rate": 2e-05,
+      "loss": 0.9042,
+      "step": 42504
+    },
+    {
+      "epoch": 1.0117347424545369,
+      "grad_norm": 0.27621541681262013,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 42505
+    },
+    {
+      "epoch": 1.0117585451775684,
+      "grad_norm": 0.4540947858988106,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 42506
+    },
+    {
+      "epoch": 1.0117823479005998,
+      "grad_norm": 7.06974266133653,
+      "learning_rate": 2e-05,
+      "loss": 0.9011,
+      "step": 42507
+    },
+    {
+      "epoch": 1.0118061506236313,
+      "grad_norm": 3.2907248044423554,
+      "learning_rate": 2e-05,
+      "loss": 0.2833,
+      "step": 42508
+    },
+    {
+      "epoch": 1.0118299533466628,
+      "grad_norm": 4.1405898679689725,
+      "learning_rate": 2e-05,
+      "loss": 0.6239,
+      "step": 42509
+    },
+    {
+      "epoch": 1.0118537560696943,
+      "grad_norm": 1.7587379866482102,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 42510
+    },
+    {
+      "epoch": 1.0118775587927258,
+      "grad_norm": 2.6076744734532835,
+      "learning_rate": 2e-05,
+      "loss": 0.3317,
+      "step": 42511
+    },
+    {
+      "epoch": 1.0119013615157575,
+      "grad_norm": 9.5387263688685,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 42512
+    },
+    {
+      "epoch": 1.011925164238789,
+      "grad_norm": 17.239300507625693,
+      "learning_rate": 2e-05,
+      "loss": 0.1076,
+      "step": 42513
+    },
+    {
+      "epoch": 1.0119489669618205,
+      "grad_norm": 2.7017716055272527,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 42514
+    },
+    {
+      "epoch": 1.011972769684852,
+      "grad_norm": 0.054414111411904886,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 42515
+    },
+    {
+      "epoch": 1.0119965724078834,
+      "grad_norm": 17.08551671404807,
+      "learning_rate": 2e-05,
+      "loss": 0.3491,
+      "step": 42516
+    },
+    {
+      "epoch": 1.012020375130915,
+      "grad_norm": 1.8348139577414913,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 42517
+    },
+    {
+      "epoch": 1.0120441778539464,
+      "grad_norm": 0.01196992396477024,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42518
+    },
+    {
+      "epoch": 1.012067980576978,
+      "grad_norm": 4.856727096134656,
+      "learning_rate": 2e-05,
+      "loss": 0.6163,
+      "step": 42519
+    },
+    {
+      "epoch": 1.0120917833000096,
+      "grad_norm": 0.05538360470202886,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42520
+    },
+    {
+      "epoch": 1.012115586023041,
+      "grad_norm": 8.644237695035901,
+      "learning_rate": 2e-05,
+      "loss": 0.5561,
+      "step": 42521
+    },
+    {
+      "epoch": 1.0121393887460726,
+      "grad_norm": 5.85989485512182,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 42522
+    },
+    {
+      "epoch": 1.012163191469104,
+      "grad_norm": 0.0089472849584035,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42523
+    },
+    {
+      "epoch": 1.0121869941921355,
+      "grad_norm": 6.194288041024353,
+      "learning_rate": 2e-05,
+      "loss": 0.4933,
+      "step": 42524
+    },
+    {
+      "epoch": 1.012210796915167,
+      "grad_norm": 0.21352409354307683,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42525
+    },
+    {
+      "epoch": 1.0122345996381985,
+      "grad_norm": 4.9237365096191095,
+      "learning_rate": 2e-05,
+      "loss": 0.5286,
+      "step": 42526
+    },
+    {
+      "epoch": 1.0122584023612302,
+      "grad_norm": 0.07866376435198451,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42527
+    },
+    {
+      "epoch": 1.0122822050842617,
+      "grad_norm": 0.02023691270396235,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42528
+    },
+    {
+      "epoch": 1.0123060078072932,
+      "grad_norm": 0.06328287187059198,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42529
+    },
+    {
+      "epoch": 1.0123298105303247,
+      "grad_norm": 0.23275165529109954,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42530
+    },
+    {
+      "epoch": 1.0123536132533562,
+      "grad_norm": 0.11731435454674041,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42531
+    },
+    {
+      "epoch": 1.0123774159763876,
+      "grad_norm": 4.008584047614229,
+      "learning_rate": 2e-05,
+      "loss": 0.5448,
+      "step": 42532
+    },
+    {
+      "epoch": 1.0124012186994191,
+      "grad_norm": 0.26126639567443183,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 42533
+    },
+    {
+      "epoch": 1.0124250214224508,
+      "grad_norm": 0.004648313386949189,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42534
+    },
+    {
+      "epoch": 1.0124488241454823,
+      "grad_norm": 2.3337185472550908,
+      "learning_rate": 2e-05,
+      "loss": 0.1153,
+      "step": 42535
+    },
+    {
+      "epoch": 1.0124726268685138,
+      "grad_norm": 1.1400614065037762,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 42536
+    },
+    {
+      "epoch": 1.0124964295915453,
+      "grad_norm": 18.756956533309967,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 42537
+    },
+    {
+      "epoch": 1.0125202323145768,
+      "grad_norm": 4.8437280759159,
+      "learning_rate": 2e-05,
+      "loss": 0.3269,
+      "step": 42538
+    },
+    {
+      "epoch": 1.0125440350376083,
+      "grad_norm": 0.0944331854092294,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42539
+    },
+    {
+      "epoch": 1.0125678377606397,
+      "grad_norm": 0.1412640924354247,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 42540
+    },
+    {
+      "epoch": 1.0125916404836712,
+      "grad_norm": 9.682945885147229,
+      "learning_rate": 2e-05,
+      "loss": 0.1417,
+      "step": 42541
+    },
+    {
+      "epoch": 1.012615443206703,
+      "grad_norm": 6.363532909959433,
+      "learning_rate": 2e-05,
+      "loss": 0.9174,
+      "step": 42542
+    },
+    {
+      "epoch": 1.0126392459297344,
+      "grad_norm": 6.4299130505362685,
+      "learning_rate": 2e-05,
+      "loss": 0.09,
+      "step": 42543
+    },
+    {
+      "epoch": 1.012663048652766,
+      "grad_norm": 9.48560992182906,
+      "learning_rate": 2e-05,
+      "loss": 0.6942,
+      "step": 42544
+    },
+    {
+      "epoch": 1.0126868513757974,
+      "grad_norm": 0.9964344662618247,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 42545
+    },
+    {
+      "epoch": 1.0127106540988289,
+      "grad_norm": 5.858632057503473,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 42546
+    },
+    {
+      "epoch": 1.0127344568218604,
+      "grad_norm": 3.765244426700817,
+      "learning_rate": 2e-05,
+      "loss": 0.2285,
+      "step": 42547
+    },
+    {
+      "epoch": 1.0127582595448918,
+      "grad_norm": 0.13743571593132323,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42548
+    },
+    {
+      "epoch": 1.0127820622679236,
+      "grad_norm": 6.6915725508957005,
+      "learning_rate": 2e-05,
+      "loss": 0.5662,
+      "step": 42549
+    },
+    {
+      "epoch": 1.012805864990955,
+      "grad_norm": 0.2212878797611875,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 42550
+    },
+    {
+      "epoch": 1.0128296677139865,
+      "grad_norm": 0.17412450700980944,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 42551
+    },
+    {
+      "epoch": 1.012853470437018,
+      "grad_norm": 0.6431501868247677,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 42552
+    },
+    {
+      "epoch": 1.0128772731600495,
+      "grad_norm": 0.904525422558635,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 42553
+    },
+    {
+      "epoch": 1.012901075883081,
+      "grad_norm": 0.13728620527638968,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 42554
+    },
+    {
+      "epoch": 1.0129248786061125,
+      "grad_norm": 4.123467586634454,
+      "learning_rate": 2e-05,
+      "loss": 0.2819,
+      "step": 42555
+    },
+    {
+      "epoch": 1.012948681329144,
+      "grad_norm": 2.271058931802454,
+      "learning_rate": 2e-05,
+      "loss": 0.2231,
+      "step": 42556
+    },
+    {
+      "epoch": 1.0129724840521757,
+      "grad_norm": 0.6587556236310166,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 42557
+    },
+    {
+      "epoch": 1.0129962867752071,
+      "grad_norm": 0.04912191242084146,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42558
+    },
+    {
+      "epoch": 1.0130200894982386,
+      "grad_norm": 5.055367140923363,
+      "learning_rate": 2e-05,
+      "loss": 0.4801,
+      "step": 42559
+    },
+    {
+      "epoch": 1.0130438922212701,
+      "grad_norm": 18.177807796660364,
+      "learning_rate": 2e-05,
+      "loss": 0.7351,
+      "step": 42560
+    },
+    {
+      "epoch": 1.0130676949443016,
+      "grad_norm": 4.637341999559619,
+      "learning_rate": 2e-05,
+      "loss": 0.9236,
+      "step": 42561
+    },
+    {
+      "epoch": 1.013091497667333,
+      "grad_norm": 0.7103833544092542,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 42562
+    },
+    {
+      "epoch": 1.0131153003903646,
+      "grad_norm": 4.663596455696108,
+      "learning_rate": 2e-05,
+      "loss": 0.116,
+      "step": 42563
+    },
+    {
+      "epoch": 1.0131391031133963,
+      "grad_norm": 2.089246528979283,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 42564
+    },
+    {
+      "epoch": 1.0131629058364278,
+      "grad_norm": 0.7049441693647859,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 42565
+    },
+    {
+      "epoch": 1.0131867085594592,
+      "grad_norm": 3.956405534341836,
+      "learning_rate": 2e-05,
+      "loss": 0.2293,
+      "step": 42566
+    },
+    {
+      "epoch": 1.0132105112824907,
+      "grad_norm": 0.007489866207728188,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42567
+    },
+    {
+      "epoch": 1.0132343140055222,
+      "grad_norm": 0.0017941589694412724,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 42568
+    },
+    {
+      "epoch": 1.0132581167285537,
+      "grad_norm": 0.07908447383092124,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42569
+    },
+    {
+      "epoch": 1.0132819194515852,
+      "grad_norm": 2.1181525000644954,
+      "learning_rate": 2e-05,
+      "loss": 0.1785,
+      "step": 42570
+    },
+    {
+      "epoch": 1.0133057221746167,
+      "grad_norm": 1.2966179470075339,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 42571
+    },
+    {
+      "epoch": 1.0133295248976484,
+      "grad_norm": 5.891997625041443,
+      "learning_rate": 2e-05,
+      "loss": 0.8285,
+      "step": 42572
+    },
+    {
+      "epoch": 1.0133533276206799,
+      "grad_norm": 3.3620146138647855,
+      "learning_rate": 2e-05,
+      "loss": 0.5288,
+      "step": 42573
+    },
+    {
+      "epoch": 1.0133771303437114,
+      "grad_norm": 2.172100722709937,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 42574
+    },
+    {
+      "epoch": 1.0134009330667428,
+      "grad_norm": 0.02833397255609316,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42575
+    },
+    {
+      "epoch": 1.0134247357897743,
+      "grad_norm": 0.1902690790828541,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42576
+    },
+    {
+      "epoch": 1.0134485385128058,
+      "grad_norm": 0.6121591828716667,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 42577
+    },
+    {
+      "epoch": 1.0134723412358373,
+      "grad_norm": 1.9966622648697303,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 42578
+    },
+    {
+      "epoch": 1.013496143958869,
+      "grad_norm": 3.1245557707596885,
+      "learning_rate": 2e-05,
+      "loss": 0.6137,
+      "step": 42579
+    },
+    {
+      "epoch": 1.0135199466819005,
+      "grad_norm": 2.4674312772896934,
+      "learning_rate": 2e-05,
+      "loss": 0.2088,
+      "step": 42580
+    },
+    {
+      "epoch": 1.013543749404932,
+      "grad_norm": 1.3415054380104727,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 42581
+    },
+    {
+      "epoch": 1.0135675521279635,
+      "grad_norm": 3.415573275113524,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 42582
+    },
+    {
+      "epoch": 1.013591354850995,
+      "grad_norm": 1.3866564567830673,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 42583
+    },
+    {
+      "epoch": 1.0136151575740264,
+      "grad_norm": 0.0426397972485682,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42584
+    },
+    {
+      "epoch": 1.013638960297058,
+      "grad_norm": 7.558960589815078,
+      "learning_rate": 2e-05,
+      "loss": 1.4712,
+      "step": 42585
+    },
+    {
+      "epoch": 1.0136627630200894,
+      "grad_norm": 4.461748795294994,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 42586
+    },
+    {
+      "epoch": 1.013686565743121,
+      "grad_norm": 1.3707417200658085,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 42587
+    },
+    {
+      "epoch": 1.0137103684661526,
+      "grad_norm": 0.6018446964164943,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 42588
+    },
+    {
+      "epoch": 1.013734171189184,
+      "grad_norm": 5.042719962204662,
+      "learning_rate": 2e-05,
+      "loss": 0.5784,
+      "step": 42589
+    },
+    {
+      "epoch": 1.0137579739122156,
+      "grad_norm": 6.522968377253518,
+      "learning_rate": 2e-05,
+      "loss": 1.0957,
+      "step": 42590
+    },
+    {
+      "epoch": 1.013781776635247,
+      "grad_norm": 12.193074326623696,
+      "learning_rate": 2e-05,
+      "loss": 1.9145,
+      "step": 42591
+    },
+    {
+      "epoch": 1.0138055793582785,
+      "grad_norm": 1.739468330538792,
+      "learning_rate": 2e-05,
+      "loss": 0.1316,
+      "step": 42592
+    },
+    {
+      "epoch": 1.01382938208131,
+      "grad_norm": 6.299551881231803,
+      "learning_rate": 2e-05,
+      "loss": 0.4314,
+      "step": 42593
+    },
+    {
+      "epoch": 1.0138531848043417,
+      "grad_norm": 3.227301952962973,
+      "learning_rate": 2e-05,
+      "loss": 0.4689,
+      "step": 42594
+    },
+    {
+      "epoch": 1.0138769875273732,
+      "grad_norm": 5.388612245323326,
+      "learning_rate": 2e-05,
+      "loss": 0.4355,
+      "step": 42595
+    },
+    {
+      "epoch": 1.0139007902504047,
+      "grad_norm": 0.6229176142778913,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 42596
+    },
+    {
+      "epoch": 1.0139245929734362,
+      "grad_norm": 9.01841222583197,
+      "learning_rate": 2e-05,
+      "loss": 0.6093,
+      "step": 42597
+    },
+    {
+      "epoch": 1.0139483956964677,
+      "grad_norm": 0.042254652823382574,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 42598
+    },
+    {
+      "epoch": 1.0139721984194991,
+      "grad_norm": 0.01075733392704586,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42599
+    },
+    {
+      "epoch": 1.0139960011425306,
+      "grad_norm": 3.846082713344507,
+      "learning_rate": 2e-05,
+      "loss": 0.2918,
+      "step": 42600
+    },
+    {
+      "epoch": 1.0140198038655621,
+      "grad_norm": 1.144076610700777,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 42601
+    },
+    {
+      "epoch": 1.0140436065885938,
+      "grad_norm": 19.632511550721258,
+      "learning_rate": 2e-05,
+      "loss": 0.625,
+      "step": 42602
+    },
+    {
+      "epoch": 1.0140674093116253,
+      "grad_norm": 0.5562329141295871,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 42603
+    },
+    {
+      "epoch": 1.0140912120346568,
+      "grad_norm": 5.7378439734135025,
+      "learning_rate": 2e-05,
+      "loss": 0.4711,
+      "step": 42604
+    },
+    {
+      "epoch": 1.0141150147576883,
+      "grad_norm": 2.7848829450816766,
+      "learning_rate": 2e-05,
+      "loss": 0.0637,
+      "step": 42605
+    },
+    {
+      "epoch": 1.0141388174807198,
+      "grad_norm": 3.7740269167631157,
+      "learning_rate": 2e-05,
+      "loss": 0.5285,
+      "step": 42606
+    },
+    {
+      "epoch": 1.0141626202037513,
+      "grad_norm": 5.357029817911479,
+      "learning_rate": 2e-05,
+      "loss": 0.3005,
+      "step": 42607
+    },
+    {
+      "epoch": 1.0141864229267827,
+      "grad_norm": 0.5874455045159754,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 42608
+    },
+    {
+      "epoch": 1.0142102256498144,
+      "grad_norm": 6.357196415693321,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 42609
+    },
+    {
+      "epoch": 1.014234028372846,
+      "grad_norm": 5.582582822358449,
+      "learning_rate": 2e-05,
+      "loss": 0.5226,
+      "step": 42610
+    },
+    {
+      "epoch": 1.0142578310958774,
+      "grad_norm": 1.91110249957441,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 42611
+    },
+    {
+      "epoch": 1.014281633818909,
+      "grad_norm": 3.025084637871346,
+      "learning_rate": 2e-05,
+      "loss": 0.2705,
+      "step": 42612
+    },
+    {
+      "epoch": 1.0143054365419404,
+      "grad_norm": 1.095776118293645,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 42613
+    },
+    {
+      "epoch": 1.0143292392649719,
+      "grad_norm": 4.046608648611889,
+      "learning_rate": 2e-05,
+      "loss": 0.5435,
+      "step": 42614
+    },
+    {
+      "epoch": 1.0143530419880034,
+      "grad_norm": 0.38364094429417483,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 42615
+    },
+    {
+      "epoch": 1.0143768447110348,
+      "grad_norm": 3.6017918125550894,
+      "learning_rate": 2e-05,
+      "loss": 0.3473,
+      "step": 42616
+    },
+    {
+      "epoch": 1.0144006474340665,
+      "grad_norm": 3.2666846214565473,
+      "learning_rate": 2e-05,
+      "loss": 0.5707,
+      "step": 42617
+    },
+    {
+      "epoch": 1.014424450157098,
+      "grad_norm": 0.002223452960140153,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 42618
+    },
+    {
+      "epoch": 1.0144482528801295,
+      "grad_norm": 7.992545753782574,
+      "learning_rate": 2e-05,
+      "loss": 0.8012,
+      "step": 42619
+    },
+    {
+      "epoch": 1.014472055603161,
+      "grad_norm": 3.4969127668429163,
+      "learning_rate": 2e-05,
+      "loss": 0.3801,
+      "step": 42620
+    },
+    {
+      "epoch": 1.0144958583261925,
+      "grad_norm": 2.108458236683165,
+      "learning_rate": 2e-05,
+      "loss": 0.2143,
+      "step": 42621
+    },
+    {
+      "epoch": 1.014519661049224,
+      "grad_norm": 2.686396990955305,
+      "learning_rate": 2e-05,
+      "loss": 0.058,
+      "step": 42622
+    },
+    {
+      "epoch": 1.0145434637722555,
+      "grad_norm": 10.672478028188163,
+      "learning_rate": 2e-05,
+      "loss": 0.5811,
+      "step": 42623
+    },
+    {
+      "epoch": 1.0145672664952872,
+      "grad_norm": 0.06367332090530933,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42624
+    },
+    {
+      "epoch": 1.0145910692183187,
+      "grad_norm": 13.739995849962982,
+      "learning_rate": 2e-05,
+      "loss": 0.1194,
+      "step": 42625
+    },
+    {
+      "epoch": 1.0146148719413501,
+      "grad_norm": 7.387947258637042,
+      "learning_rate": 2e-05,
+      "loss": 0.2805,
+      "step": 42626
+    },
+    {
+      "epoch": 1.0146386746643816,
+      "grad_norm": 3.4492883331446342,
+      "learning_rate": 2e-05,
+      "loss": 0.4241,
+      "step": 42627
+    },
+    {
+      "epoch": 1.014662477387413,
+      "grad_norm": 0.008463369501542955,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42628
+    },
+    {
+      "epoch": 1.0146862801104446,
+      "grad_norm": 0.9764754813085085,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 42629
+    },
+    {
+      "epoch": 1.014710082833476,
+      "grad_norm": 8.083349306069112,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 42630
+    },
+    {
+      "epoch": 1.0147338855565076,
+      "grad_norm": 0.021535901318111263,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42631
+    },
+    {
+      "epoch": 1.0147576882795393,
+      "grad_norm": 3.4604752486417243,
+      "learning_rate": 2e-05,
+      "loss": 0.465,
+      "step": 42632
+    },
+    {
+      "epoch": 1.0147814910025708,
+      "grad_norm": 7.463547701347165,
+      "learning_rate": 2e-05,
+      "loss": 1.8491,
+      "step": 42633
+    },
+    {
+      "epoch": 1.0148052937256022,
+      "grad_norm": 0.18366261516767254,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 42634
+    },
+    {
+      "epoch": 1.0148290964486337,
+      "grad_norm": 0.6664943331580176,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 42635
+    },
+    {
+      "epoch": 1.0148528991716652,
+      "grad_norm": 4.503085516472433,
+      "learning_rate": 2e-05,
+      "loss": 0.2253,
+      "step": 42636
+    },
+    {
+      "epoch": 1.0148767018946967,
+      "grad_norm": 4.707566513048666,
+      "learning_rate": 2e-05,
+      "loss": 0.037,
+      "step": 42637
+    },
+    {
+      "epoch": 1.0149005046177282,
+      "grad_norm": 5.400102703111516,
+      "learning_rate": 2e-05,
+      "loss": 0.3416,
+      "step": 42638
+    },
+    {
+      "epoch": 1.01492430734076,
+      "grad_norm": 4.498050507396707,
+      "learning_rate": 2e-05,
+      "loss": 0.1911,
+      "step": 42639
+    },
+    {
+      "epoch": 1.0149481100637914,
+      "grad_norm": 10.985958142835361,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 42640
+    },
+    {
+      "epoch": 1.0149719127868229,
+      "grad_norm": 3.3945282101962473,
+      "learning_rate": 2e-05,
+      "loss": 0.3777,
+      "step": 42641
+    },
+    {
+      "epoch": 1.0149957155098543,
+      "grad_norm": 0.38115879199245367,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 42642
+    },
+    {
+      "epoch": 1.0150195182328858,
+      "grad_norm": 3.0295286661272263,
+      "learning_rate": 2e-05,
+      "loss": 0.167,
+      "step": 42643
+    },
+    {
+      "epoch": 1.0150433209559173,
+      "grad_norm": 0.008284941250135727,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42644
+    },
+    {
+      "epoch": 1.0150671236789488,
+      "grad_norm": 6.556510899590301,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 42645
+    },
+    {
+      "epoch": 1.0150909264019803,
+      "grad_norm": 6.834652127449415,
+      "learning_rate": 2e-05,
+      "loss": 0.3308,
+      "step": 42646
+    },
+    {
+      "epoch": 1.015114729125012,
+      "grad_norm": 0.041071574627407674,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 42647
+    },
+    {
+      "epoch": 1.0151385318480435,
+      "grad_norm": 1.7239250207958345,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 42648
+    },
+    {
+      "epoch": 1.015162334571075,
+      "grad_norm": 5.217558320103229,
+      "learning_rate": 2e-05,
+      "loss": 0.4479,
+      "step": 42649
+    },
+    {
+      "epoch": 1.0151861372941064,
+      "grad_norm": 5.355816406454944,
+      "learning_rate": 2e-05,
+      "loss": 0.9858,
+      "step": 42650
+    },
+    {
+      "epoch": 1.015209940017138,
+      "grad_norm": 1.0808018198739784,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 42651
+    },
+    {
+      "epoch": 1.0152337427401694,
+      "grad_norm": 4.746293334035978,
+      "learning_rate": 2e-05,
+      "loss": 0.3948,
+      "step": 42652
+    },
+    {
+      "epoch": 1.015257545463201,
+      "grad_norm": 4.817007750877202,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 42653
+    },
+    {
+      "epoch": 1.0152813481862326,
+      "grad_norm": 2.8703443009864795,
+      "learning_rate": 2e-05,
+      "loss": 0.2365,
+      "step": 42654
+    },
+    {
+      "epoch": 1.015305150909264,
+      "grad_norm": 7.230433242951658,
+      "learning_rate": 2e-05,
+      "loss": 1.4122,
+      "step": 42655
+    },
+    {
+      "epoch": 1.0153289536322956,
+      "grad_norm": 4.614636656676154,
+      "learning_rate": 2e-05,
+      "loss": 0.8457,
+      "step": 42656
+    },
+    {
+      "epoch": 1.015352756355327,
+      "grad_norm": 6.773900660947109,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 42657
+    },
+    {
+      "epoch": 1.0153765590783586,
+      "grad_norm": 3.326491575339743,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 42658
+    },
+    {
+      "epoch": 1.01540036180139,
+      "grad_norm": 12.79738810308379,
+      "learning_rate": 2e-05,
+      "loss": 0.2134,
+      "step": 42659
+    },
+    {
+      "epoch": 1.0154241645244215,
+      "grad_norm": 3.1282728132739765,
+      "learning_rate": 2e-05,
+      "loss": 0.3517,
+      "step": 42660
+    },
+    {
+      "epoch": 1.015447967247453,
+      "grad_norm": 0.09456515444234566,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42661
+    },
+    {
+      "epoch": 1.0154717699704847,
+      "grad_norm": 6.004754051248574,
+      "learning_rate": 2e-05,
+      "loss": 1.0986,
+      "step": 42662
+    },
+    {
+      "epoch": 1.0154955726935162,
+      "grad_norm": 3.8943117895634454,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 42663
+    },
+    {
+      "epoch": 1.0155193754165477,
+      "grad_norm": 0.0216534326289992,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42664
+    },
+    {
+      "epoch": 1.0155431781395792,
+      "grad_norm": 0.01996169787595187,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42665
+    },
+    {
+      "epoch": 1.0155669808626107,
+      "grad_norm": 6.998248694670087,
+      "learning_rate": 2e-05,
+      "loss": 0.2983,
+      "step": 42666
+    },
+    {
+      "epoch": 1.0155907835856421,
+      "grad_norm": 0.5366382949215968,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 42667
+    },
+    {
+      "epoch": 1.0156145863086736,
+      "grad_norm": 3.8400522745861947,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 42668
+    },
+    {
+      "epoch": 1.0156383890317053,
+      "grad_norm": 3.78586961462851,
+      "learning_rate": 2e-05,
+      "loss": 0.277,
+      "step": 42669
+    },
+    {
+      "epoch": 1.0156621917547368,
+      "grad_norm": 3.7799596548427785,
+      "learning_rate": 2e-05,
+      "loss": 0.3962,
+      "step": 42670
+    },
+    {
+      "epoch": 1.0156859944777683,
+      "grad_norm": 2.6161175723849706,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 42671
+    },
+    {
+      "epoch": 1.0157097972007998,
+      "grad_norm": 6.247852612120286,
+      "learning_rate": 2e-05,
+      "loss": 1.1095,
+      "step": 42672
+    },
+    {
+      "epoch": 1.0157335999238313,
+      "grad_norm": 0.18455139113658606,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42673
+    },
+    {
+      "epoch": 1.0157574026468628,
+      "grad_norm": 3.140599725075943,
+      "learning_rate": 2e-05,
+      "loss": 0.2275,
+      "step": 42674
+    },
+    {
+      "epoch": 1.0157812053698942,
+      "grad_norm": 1.589703278331017,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 42675
+    },
+    {
+      "epoch": 1.0158050080929257,
+      "grad_norm": 3.773142753315258,
+      "learning_rate": 2e-05,
+      "loss": 0.0768,
+      "step": 42676
+    },
+    {
+      "epoch": 1.0158288108159574,
+      "grad_norm": 2.996629292254941,
+      "learning_rate": 2e-05,
+      "loss": 0.2436,
+      "step": 42677
+    },
+    {
+      "epoch": 1.015852613538989,
+      "grad_norm": 0.006175989961179859,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42678
+    },
+    {
+      "epoch": 1.0158764162620204,
+      "grad_norm": 6.437227208756853,
+      "learning_rate": 2e-05,
+      "loss": 0.921,
+      "step": 42679
+    },
+    {
+      "epoch": 1.015900218985052,
+      "grad_norm": 0.21100588065567225,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42680
+    },
+    {
+      "epoch": 1.0159240217080834,
+      "grad_norm": 1.727814562714922,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 42681
+    },
+    {
+      "epoch": 1.0159478244311149,
+      "grad_norm": 0.20347781717221494,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 42682
+    },
+    {
+      "epoch": 1.0159716271541464,
+      "grad_norm": 3.9973772723030567,
+      "learning_rate": 2e-05,
+      "loss": 0.3558,
+      "step": 42683
+    },
+    {
+      "epoch": 1.015995429877178,
+      "grad_norm": 6.605584250640178,
+      "learning_rate": 2e-05,
+      "loss": 0.6205,
+      "step": 42684
+    },
+    {
+      "epoch": 1.0160192326002095,
+      "grad_norm": 8.21007268527643,
+      "learning_rate": 2e-05,
+      "loss": 0.5859,
+      "step": 42685
+    },
+    {
+      "epoch": 1.016043035323241,
+      "grad_norm": 2.088489618793869,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 42686
+    },
+    {
+      "epoch": 1.0160668380462725,
+      "grad_norm": 5.441629965277861,
+      "learning_rate": 2e-05,
+      "loss": 0.7288,
+      "step": 42687
+    },
+    {
+      "epoch": 1.016090640769304,
+      "grad_norm": 2.05096831362003,
+      "learning_rate": 2e-05,
+      "loss": 0.1074,
+      "step": 42688
+    },
+    {
+      "epoch": 1.0161144434923355,
+      "grad_norm": 4.622830226967375,
+      "learning_rate": 2e-05,
+      "loss": 0.5114,
+      "step": 42689
+    },
+    {
+      "epoch": 1.016138246215367,
+      "grad_norm": 4.210010363477054,
+      "learning_rate": 2e-05,
+      "loss": 0.0864,
+      "step": 42690
+    },
+    {
+      "epoch": 1.0161620489383985,
+      "grad_norm": 0.400567450484814,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 42691
+    },
+    {
+      "epoch": 1.0161858516614302,
+      "grad_norm": 5.2368018404042305,
+      "learning_rate": 2e-05,
+      "loss": 0.4015,
+      "step": 42692
+    },
+    {
+      "epoch": 1.0162096543844616,
+      "grad_norm": 15.494834547652895,
+      "learning_rate": 2e-05,
+      "loss": 0.8137,
+      "step": 42693
+    },
+    {
+      "epoch": 1.0162334571074931,
+      "grad_norm": 0.009241514456929457,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42694
+    },
+    {
+      "epoch": 1.0162572598305246,
+      "grad_norm": 3.85357393724474,
+      "learning_rate": 2e-05,
+      "loss": 0.4438,
+      "step": 42695
+    },
+    {
+      "epoch": 1.016281062553556,
+      "grad_norm": 4.587451791824462,
+      "learning_rate": 2e-05,
+      "loss": 0.2594,
+      "step": 42696
+    },
+    {
+      "epoch": 1.0163048652765876,
+      "grad_norm": 3.878843342346602,
+      "learning_rate": 2e-05,
+      "loss": 0.1739,
+      "step": 42697
+    },
+    {
+      "epoch": 1.016328667999619,
+      "grad_norm": 22.577959203997036,
+      "learning_rate": 2e-05,
+      "loss": 1.2531,
+      "step": 42698
+    },
+    {
+      "epoch": 1.0163524707226508,
+      "grad_norm": 0.05598684398908869,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42699
+    },
+    {
+      "epoch": 1.0163762734456823,
+      "grad_norm": 3.1076640099905433,
+      "learning_rate": 2e-05,
+      "loss": 0.0274,
+      "step": 42700
+    },
+    {
+      "epoch": 1.0164000761687138,
+      "grad_norm": 4.399372966447732,
+      "learning_rate": 2e-05,
+      "loss": 0.3631,
+      "step": 42701
+    },
+    {
+      "epoch": 1.0164238788917452,
+      "grad_norm": 3.578651253022692,
+      "learning_rate": 2e-05,
+      "loss": 0.3981,
+      "step": 42702
+    },
+    {
+      "epoch": 1.0164476816147767,
+      "grad_norm": 1.4608992062838784,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 42703
+    },
+    {
+      "epoch": 1.0164714843378082,
+      "grad_norm": 0.24251724151913842,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42704
+    },
+    {
+      "epoch": 1.0164952870608397,
+      "grad_norm": 5.345510261752405,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 42705
+    },
+    {
+      "epoch": 1.0165190897838712,
+      "grad_norm": 0.009613091573541534,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42706
+    },
+    {
+      "epoch": 1.0165428925069029,
+      "grad_norm": 0.15525615445333193,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 42707
+    },
+    {
+      "epoch": 1.0165666952299344,
+      "grad_norm": 0.011856354892566555,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42708
+    },
+    {
+      "epoch": 1.0165904979529659,
+      "grad_norm": 0.03291912710645888,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42709
+    },
+    {
+      "epoch": 1.0166143006759973,
+      "grad_norm": 0.06767452004487973,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42710
+    },
+    {
+      "epoch": 1.0166381033990288,
+      "grad_norm": 9.580936542094308,
+      "learning_rate": 2e-05,
+      "loss": 0.4791,
+      "step": 42711
+    },
+    {
+      "epoch": 1.0166619061220603,
+      "grad_norm": 3.933067898388818,
+      "learning_rate": 2e-05,
+      "loss": 0.3325,
+      "step": 42712
+    },
+    {
+      "epoch": 1.0166857088450918,
+      "grad_norm": 0.4476840864139417,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 42713
+    },
+    {
+      "epoch": 1.0167095115681235,
+      "grad_norm": 1.5717004243824364,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 42714
+    },
+    {
+      "epoch": 1.016733314291155,
+      "grad_norm": 3.7491595472345467,
+      "learning_rate": 2e-05,
+      "loss": 0.5681,
+      "step": 42715
+    },
+    {
+      "epoch": 1.0167571170141865,
+      "grad_norm": 5.306339486092621,
+      "learning_rate": 2e-05,
+      "loss": 0.5343,
+      "step": 42716
+    },
+    {
+      "epoch": 1.016780919737218,
+      "grad_norm": 2.627163311535282,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 42717
+    },
+    {
+      "epoch": 1.0168047224602494,
+      "grad_norm": 2.96801948636634,
+      "learning_rate": 2e-05,
+      "loss": 0.4237,
+      "step": 42718
+    },
+    {
+      "epoch": 1.016828525183281,
+      "grad_norm": 6.21972772196758,
+      "learning_rate": 2e-05,
+      "loss": 0.2776,
+      "step": 42719
+    },
+    {
+      "epoch": 1.0168523279063124,
+      "grad_norm": 1.665533018338392,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 42720
+    },
+    {
+      "epoch": 1.016876130629344,
+      "grad_norm": 8.206871383014846,
+      "learning_rate": 2e-05,
+      "loss": 0.7748,
+      "step": 42721
+    },
+    {
+      "epoch": 1.0168999333523756,
+      "grad_norm": 2.900017210312983,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 42722
+    },
+    {
+      "epoch": 1.016923736075407,
+      "grad_norm": 4.765061026334117,
+      "learning_rate": 2e-05,
+      "loss": 0.1445,
+      "step": 42723
+    },
+    {
+      "epoch": 1.0169475387984386,
+      "grad_norm": 4.436993532205155,
+      "learning_rate": 2e-05,
+      "loss": 0.55,
+      "step": 42724
+    },
+    {
+      "epoch": 1.01697134152147,
+      "grad_norm": 0.5784324609711207,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 42725
+    },
+    {
+      "epoch": 1.0169951442445015,
+      "grad_norm": 0.3916628630241645,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 42726
+    },
+    {
+      "epoch": 1.017018946967533,
+      "grad_norm": 0.10365194289183685,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42727
+    },
+    {
+      "epoch": 1.0170427496905645,
+      "grad_norm": 0.06528524329796326,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 42728
+    },
+    {
+      "epoch": 1.017066552413596,
+      "grad_norm": 0.13102082161346448,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 42729
+    },
+    {
+      "epoch": 1.0170903551366277,
+      "grad_norm": 5.943355088367133,
+      "learning_rate": 2e-05,
+      "loss": 0.7074,
+      "step": 42730
+    },
+    {
+      "epoch": 1.0171141578596592,
+      "grad_norm": 5.696619334593697,
+      "learning_rate": 2e-05,
+      "loss": 0.7815,
+      "step": 42731
+    },
+    {
+      "epoch": 1.0171379605826907,
+      "grad_norm": 6.163890110771197,
+      "learning_rate": 2e-05,
+      "loss": 0.3364,
+      "step": 42732
+    },
+    {
+      "epoch": 1.0171617633057222,
+      "grad_norm": 5.137846946162085,
+      "learning_rate": 2e-05,
+      "loss": 0.5384,
+      "step": 42733
+    },
+    {
+      "epoch": 1.0171855660287537,
+      "grad_norm": 11.909215272231362,
+      "learning_rate": 2e-05,
+      "loss": 0.5129,
+      "step": 42734
+    },
+    {
+      "epoch": 1.0172093687517851,
+      "grad_norm": 2.177187217646647,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 42735
+    },
+    {
+      "epoch": 1.0172331714748166,
+      "grad_norm": 2.2263529717746082,
+      "learning_rate": 2e-05,
+      "loss": 0.1542,
+      "step": 42736
+    },
+    {
+      "epoch": 1.0172569741978483,
+      "grad_norm": 4.573578157874571,
+      "learning_rate": 2e-05,
+      "loss": 0.8261,
+      "step": 42737
+    },
+    {
+      "epoch": 1.0172807769208798,
+      "grad_norm": 0.18675578157219042,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 42738
+    },
+    {
+      "epoch": 1.0173045796439113,
+      "grad_norm": 5.091577444340633,
+      "learning_rate": 2e-05,
+      "loss": 0.3757,
+      "step": 42739
+    },
+    {
+      "epoch": 1.0173283823669428,
+      "grad_norm": 4.127606829426731,
+      "learning_rate": 2e-05,
+      "loss": 0.7401,
+      "step": 42740
+    },
+    {
+      "epoch": 1.0173521850899743,
+      "grad_norm": 0.3416889217596333,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42741
+    },
+    {
+      "epoch": 1.0173759878130058,
+      "grad_norm": 7.997996877639681,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 42742
+    },
+    {
+      "epoch": 1.0173997905360372,
+      "grad_norm": 6.900457108035159,
+      "learning_rate": 2e-05,
+      "loss": 0.2686,
+      "step": 42743
+    },
+    {
+      "epoch": 1.017423593259069,
+      "grad_norm": 5.960900811054959,
+      "learning_rate": 2e-05,
+      "loss": 0.7439,
+      "step": 42744
+    },
+    {
+      "epoch": 1.0174473959821004,
+      "grad_norm": 0.8316691372316922,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 42745
+    },
+    {
+      "epoch": 1.017471198705132,
+      "grad_norm": 2.8689396622497485,
+      "learning_rate": 2e-05,
+      "loss": 0.0585,
+      "step": 42746
+    },
+    {
+      "epoch": 1.0174950014281634,
+      "grad_norm": 0.02320463981582736,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42747
+    },
+    {
+      "epoch": 1.0175188041511949,
+      "grad_norm": 0.9032258087062446,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 42748
+    },
+    {
+      "epoch": 1.0175426068742264,
+      "grad_norm": 4.716217293819517,
+      "learning_rate": 2e-05,
+      "loss": 0.4297,
+      "step": 42749
+    },
+    {
+      "epoch": 1.0175664095972579,
+      "grad_norm": 0.14627269224729939,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 42750
+    },
+    {
+      "epoch": 1.0175902123202893,
+      "grad_norm": 29.81713594868255,
+      "learning_rate": 2e-05,
+      "loss": 0.5258,
+      "step": 42751
+    },
+    {
+      "epoch": 1.017614015043321,
+      "grad_norm": 2.320544637038276,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 42752
+    },
+    {
+      "epoch": 1.0176378177663525,
+      "grad_norm": 3.8734981233441643,
+      "learning_rate": 2e-05,
+      "loss": 0.5297,
+      "step": 42753
+    },
+    {
+      "epoch": 1.017661620489384,
+      "grad_norm": 3.792265700706797,
+      "learning_rate": 2e-05,
+      "loss": 0.5807,
+      "step": 42754
+    },
+    {
+      "epoch": 1.0176854232124155,
+      "grad_norm": 15.445441184364904,
+      "learning_rate": 2e-05,
+      "loss": 0.0708,
+      "step": 42755
+    },
+    {
+      "epoch": 1.017709225935447,
+      "grad_norm": 3.288954610124084,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 42756
+    },
+    {
+      "epoch": 1.0177330286584785,
+      "grad_norm": 30.31450047319022,
+      "learning_rate": 2e-05,
+      "loss": 0.509,
+      "step": 42757
+    },
+    {
+      "epoch": 1.01775683138151,
+      "grad_norm": 99.48686691850595,
+      "learning_rate": 2e-05,
+      "loss": 0.2044,
+      "step": 42758
+    },
+    {
+      "epoch": 1.0177806341045414,
+      "grad_norm": 0.09778039421673158,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 42759
+    },
+    {
+      "epoch": 1.0178044368275732,
+      "grad_norm": 4.553758884281699,
+      "learning_rate": 2e-05,
+      "loss": 0.6897,
+      "step": 42760
+    },
+    {
+      "epoch": 1.0178282395506046,
+      "grad_norm": 0.5529595225316954,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 42761
+    },
+    {
+      "epoch": 1.0178520422736361,
+      "grad_norm": 0.004632006887919,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42762
+    },
+    {
+      "epoch": 1.0178758449966676,
+      "grad_norm": 3.1056888515864043,
+      "learning_rate": 2e-05,
+      "loss": 0.0935,
+      "step": 42763
+    },
+    {
+      "epoch": 1.017899647719699,
+      "grad_norm": 1.376077171386095,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 42764
+    },
+    {
+      "epoch": 1.0179234504427306,
+      "grad_norm": 0.5359581550073568,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 42765
+    },
+    {
+      "epoch": 1.017947253165762,
+      "grad_norm": 2.3433807883364155,
+      "learning_rate": 2e-05,
+      "loss": 0.0415,
+      "step": 42766
+    },
+    {
+      "epoch": 1.0179710558887938,
+      "grad_norm": 1.3542805466232113,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 42767
+    },
+    {
+      "epoch": 1.0179948586118253,
+      "grad_norm": 3.5857636964199355,
+      "learning_rate": 2e-05,
+      "loss": 0.2922,
+      "step": 42768
+    },
+    {
+      "epoch": 1.0180186613348567,
+      "grad_norm": 4.691244610354092,
+      "learning_rate": 2e-05,
+      "loss": 0.2315,
+      "step": 42769
+    },
+    {
+      "epoch": 1.0180424640578882,
+      "grad_norm": 4.600934981416188,
+      "learning_rate": 2e-05,
+      "loss": 0.5669,
+      "step": 42770
+    },
+    {
+      "epoch": 1.0180662667809197,
+      "grad_norm": 5.236106251669863,
+      "learning_rate": 2e-05,
+      "loss": 0.8019,
+      "step": 42771
+    },
+    {
+      "epoch": 1.0180900695039512,
+      "grad_norm": 0.6823854032759243,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 42772
+    },
+    {
+      "epoch": 1.0181138722269827,
+      "grad_norm": 4.315562376859674,
+      "learning_rate": 2e-05,
+      "loss": 0.2642,
+      "step": 42773
+    },
+    {
+      "epoch": 1.0181376749500144,
+      "grad_norm": 6.557952417968439,
+      "learning_rate": 2e-05,
+      "loss": 0.8937,
+      "step": 42774
+    },
+    {
+      "epoch": 1.0181614776730459,
+      "grad_norm": 0.952066584516839,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 42775
+    },
+    {
+      "epoch": 1.0181852803960774,
+      "grad_norm": 3.3919441409814697,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 42776
+    },
+    {
+      "epoch": 1.0182090831191088,
+      "grad_norm": 3.136776681676452,
+      "learning_rate": 2e-05,
+      "loss": 0.3034,
+      "step": 42777
+    },
+    {
+      "epoch": 1.0182328858421403,
+      "grad_norm": 1.4678243471685823,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 42778
+    },
+    {
+      "epoch": 1.0182566885651718,
+      "grad_norm": 5.63777654668131,
+      "learning_rate": 2e-05,
+      "loss": 1.1135,
+      "step": 42779
+    },
+    {
+      "epoch": 1.0182804912882033,
+      "grad_norm": 3.7571082096884076,
+      "learning_rate": 2e-05,
+      "loss": 0.5659,
+      "step": 42780
+    },
+    {
+      "epoch": 1.0183042940112348,
+      "grad_norm": 0.039712710516980845,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42781
+    },
+    {
+      "epoch": 1.0183280967342665,
+      "grad_norm": 1.0882132802471847,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 42782
+    },
+    {
+      "epoch": 1.018351899457298,
+      "grad_norm": 2.149864251251454,
+      "learning_rate": 2e-05,
+      "loss": 0.108,
+      "step": 42783
+    },
+    {
+      "epoch": 1.0183757021803295,
+      "grad_norm": 0.3428357520082147,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 42784
+    },
+    {
+      "epoch": 1.018399504903361,
+      "grad_norm": 3.7174386408098745,
+      "learning_rate": 2e-05,
+      "loss": 0.5154,
+      "step": 42785
+    },
+    {
+      "epoch": 1.0184233076263924,
+      "grad_norm": 0.3809247290680166,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 42786
+    },
+    {
+      "epoch": 1.018447110349424,
+      "grad_norm": 1.3726575598508857,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 42787
+    },
+    {
+      "epoch": 1.0184709130724554,
+      "grad_norm": 1.6246264698913164,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 42788
+    },
+    {
+      "epoch": 1.018494715795487,
+      "grad_norm": 4.296213735544043,
+      "learning_rate": 2e-05,
+      "loss": 0.5111,
+      "step": 42789
+    },
+    {
+      "epoch": 1.0185185185185186,
+      "grad_norm": 19.985416157883357,
+      "learning_rate": 2e-05,
+      "loss": 0.1746,
+      "step": 42790
+    },
+    {
+      "epoch": 1.01854232124155,
+      "grad_norm": 4.002330491972274,
+      "learning_rate": 2e-05,
+      "loss": 0.4976,
+      "step": 42791
+    },
+    {
+      "epoch": 1.0185661239645816,
+      "grad_norm": 0.6882372472170637,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 42792
+    },
+    {
+      "epoch": 1.018589926687613,
+      "grad_norm": 0.46066789476617054,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 42793
+    },
+    {
+      "epoch": 1.0186137294106445,
+      "grad_norm": 8.04920395653601,
+      "learning_rate": 2e-05,
+      "loss": 0.2193,
+      "step": 42794
+    },
+    {
+      "epoch": 1.018637532133676,
+      "grad_norm": 0.006031899332163068,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42795
+    },
+    {
+      "epoch": 1.0186613348567075,
+      "grad_norm": 0.14985240926566573,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 42796
+    },
+    {
+      "epoch": 1.0186851375797392,
+      "grad_norm": 1.9955256583151721,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 42797
+    },
+    {
+      "epoch": 1.0187089403027707,
+      "grad_norm": 0.01886500664650574,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42798
+    },
+    {
+      "epoch": 1.0187327430258022,
+      "grad_norm": 0.18895344923728974,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 42799
+    },
+    {
+      "epoch": 1.0187565457488337,
+      "grad_norm": 3.5120133669405633,
+      "learning_rate": 2e-05,
+      "loss": 0.323,
+      "step": 42800
+    },
+    {
+      "epoch": 1.0187803484718652,
+      "grad_norm": 0.10487166700173116,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42801
+    },
+    {
+      "epoch": 1.0188041511948966,
+      "grad_norm": 0.04572085903012357,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42802
+    },
+    {
+      "epoch": 1.0188279539179281,
+      "grad_norm": 3.193786704167099,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 42803
+    },
+    {
+      "epoch": 1.0188517566409598,
+      "grad_norm": 5.9097661058165585,
+      "learning_rate": 2e-05,
+      "loss": 0.5849,
+      "step": 42804
+    },
+    {
+      "epoch": 1.0188755593639913,
+      "grad_norm": 6.207476881271635,
+      "learning_rate": 2e-05,
+      "loss": 0.5344,
+      "step": 42805
+    },
+    {
+      "epoch": 1.0188993620870228,
+      "grad_norm": 3.0657283980578445,
+      "learning_rate": 2e-05,
+      "loss": 0.0531,
+      "step": 42806
+    },
+    {
+      "epoch": 1.0189231648100543,
+      "grad_norm": 0.869082152682366,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 42807
+    },
+    {
+      "epoch": 1.0189469675330858,
+      "grad_norm": 2.7509417289970437,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 42808
+    },
+    {
+      "epoch": 1.0189707702561173,
+      "grad_norm": 0.09671758009752585,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42809
+    },
+    {
+      "epoch": 1.0189945729791487,
+      "grad_norm": 1.6490053082484926,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 42810
+    },
+    {
+      "epoch": 1.0190183757021802,
+      "grad_norm": 0.012487640410237548,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42811
+    },
+    {
+      "epoch": 1.019042178425212,
+      "grad_norm": 1.0763190520286945,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 42812
+    },
+    {
+      "epoch": 1.0190659811482434,
+      "grad_norm": 0.0867908823124459,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 42813
+    },
+    {
+      "epoch": 1.019089783871275,
+      "grad_norm": 1.1923029749248333,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 42814
+    },
+    {
+      "epoch": 1.0191135865943064,
+      "grad_norm": 11.316944711755806,
+      "learning_rate": 2e-05,
+      "loss": 0.4652,
+      "step": 42815
+    },
+    {
+      "epoch": 1.0191373893173379,
+      "grad_norm": 0.05252783996484636,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42816
+    },
+    {
+      "epoch": 1.0191611920403694,
+      "grad_norm": 5.8398397096540675,
+      "learning_rate": 2e-05,
+      "loss": 0.9833,
+      "step": 42817
+    },
+    {
+      "epoch": 1.0191849947634009,
+      "grad_norm": 8.631112090225885,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 42818
+    },
+    {
+      "epoch": 1.0192087974864323,
+      "grad_norm": 5.872048849708782,
+      "learning_rate": 2e-05,
+      "loss": 0.395,
+      "step": 42819
+    },
+    {
+      "epoch": 1.019232600209464,
+      "grad_norm": 0.5950140218392429,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 42820
+    },
+    {
+      "epoch": 1.0192564029324955,
+      "grad_norm": 32.69824666640751,
+      "learning_rate": 2e-05,
+      "loss": 1.1265,
+      "step": 42821
+    },
+    {
+      "epoch": 1.019280205655527,
+      "grad_norm": 2.9643512794389273,
+      "learning_rate": 2e-05,
+      "loss": 0.4103,
+      "step": 42822
+    },
+    {
+      "epoch": 1.0193040083785585,
+      "grad_norm": 7.141684044550355,
+      "learning_rate": 2e-05,
+      "loss": 0.18,
+      "step": 42823
+    },
+    {
+      "epoch": 1.01932781110159,
+      "grad_norm": 5.370202223886832,
+      "learning_rate": 2e-05,
+      "loss": 0.404,
+      "step": 42824
+    },
+    {
+      "epoch": 1.0193516138246215,
+      "grad_norm": 0.3739125482990409,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 42825
+    },
+    {
+      "epoch": 1.019375416547653,
+      "grad_norm": 18.646464387989077,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 42826
+    },
+    {
+      "epoch": 1.0193992192706847,
+      "grad_norm": 7.329564418581355,
+      "learning_rate": 2e-05,
+      "loss": 0.5985,
+      "step": 42827
+    },
+    {
+      "epoch": 1.0194230219937161,
+      "grad_norm": 1.8031512994284982,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 42828
+    },
+    {
+      "epoch": 1.0194468247167476,
+      "grad_norm": 4.459498118492361,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 42829
+    },
+    {
+      "epoch": 1.0194706274397791,
+      "grad_norm": 1.9451830267769297,
+      "learning_rate": 2e-05,
+      "loss": 0.0839,
+      "step": 42830
+    },
+    {
+      "epoch": 1.0194944301628106,
+      "grad_norm": 0.5317423892637464,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 42831
+    },
+    {
+      "epoch": 1.019518232885842,
+      "grad_norm": 4.461552487868646,
+      "learning_rate": 2e-05,
+      "loss": 0.1817,
+      "step": 42832
+    },
+    {
+      "epoch": 1.0195420356088736,
+      "grad_norm": 0.027754757705141027,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42833
+    },
+    {
+      "epoch": 1.019565838331905,
+      "grad_norm": 12.364746310820093,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 42834
+    },
+    {
+      "epoch": 1.0195896410549368,
+      "grad_norm": 4.522964641784243,
+      "learning_rate": 2e-05,
+      "loss": 0.4383,
+      "step": 42835
+    },
+    {
+      "epoch": 1.0196134437779683,
+      "grad_norm": 1.0494108347593174,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 42836
+    },
+    {
+      "epoch": 1.0196372465009997,
+      "grad_norm": 0.3488364103828445,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 42837
+    },
+    {
+      "epoch": 1.0196610492240312,
+      "grad_norm": 8.540085081213768,
+      "learning_rate": 2e-05,
+      "loss": 0.8287,
+      "step": 42838
+    },
+    {
+      "epoch": 1.0196848519470627,
+      "grad_norm": 0.43285746490691424,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 42839
+    },
+    {
+      "epoch": 1.0197086546700942,
+      "grad_norm": 22.30684577550706,
+      "learning_rate": 2e-05,
+      "loss": 1.0341,
+      "step": 42840
+    },
+    {
+      "epoch": 1.0197324573931257,
+      "grad_norm": 1.1999460154497552,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 42841
+    },
+    {
+      "epoch": 1.0197562601161574,
+      "grad_norm": 0.021610284529261,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42842
+    },
+    {
+      "epoch": 1.0197800628391889,
+      "grad_norm": 3.3165107270121092,
+      "learning_rate": 2e-05,
+      "loss": 0.3441,
+      "step": 42843
+    },
+    {
+      "epoch": 1.0198038655622204,
+      "grad_norm": 3.477213448671191,
+      "learning_rate": 2e-05,
+      "loss": 0.0579,
+      "step": 42844
+    },
+    {
+      "epoch": 1.0198276682852518,
+      "grad_norm": 5.337947236248203,
+      "learning_rate": 2e-05,
+      "loss": 0.6845,
+      "step": 42845
+    },
+    {
+      "epoch": 1.0198514710082833,
+      "grad_norm": 0.3086338755074522,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 42846
+    },
+    {
+      "epoch": 1.0198752737313148,
+      "grad_norm": 0.04727325831077605,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 42847
+    },
+    {
+      "epoch": 1.0198990764543463,
+      "grad_norm": 7.963052361287885,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 42848
+    },
+    {
+      "epoch": 1.0199228791773778,
+      "grad_norm": 4.1046012087323955,
+      "learning_rate": 2e-05,
+      "loss": 0.5663,
+      "step": 42849
+    },
+    {
+      "epoch": 1.0199466819004095,
+      "grad_norm": 4.673971175307504,
+      "learning_rate": 2e-05,
+      "loss": 0.5106,
+      "step": 42850
+    },
+    {
+      "epoch": 1.019970484623441,
+      "grad_norm": 0.7414326949505515,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 42851
+    },
+    {
+      "epoch": 1.0199942873464725,
+      "grad_norm": 0.7792361253996233,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 42852
+    },
+    {
+      "epoch": 1.020018090069504,
+      "grad_norm": 5.970553810021946,
+      "learning_rate": 2e-05,
+      "loss": 0.2641,
+      "step": 42853
+    },
+    {
+      "epoch": 1.0200418927925354,
+      "grad_norm": 0.05377232193874375,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42854
+    },
+    {
+      "epoch": 1.020065695515567,
+      "grad_norm": 5.640880034250456,
+      "learning_rate": 2e-05,
+      "loss": 0.8013,
+      "step": 42855
+    },
+    {
+      "epoch": 1.0200894982385984,
+      "grad_norm": 0.16563068023392274,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 42856
+    },
+    {
+      "epoch": 1.02011330096163,
+      "grad_norm": 6.974815583356467,
+      "learning_rate": 2e-05,
+      "loss": 0.0747,
+      "step": 42857
+    },
+    {
+      "epoch": 1.0201371036846616,
+      "grad_norm": 0.2535564806923099,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 42858
+    },
+    {
+      "epoch": 1.020160906407693,
+      "grad_norm": 3.4363060780206163,
+      "learning_rate": 2e-05,
+      "loss": 0.3261,
+      "step": 42859
+    },
+    {
+      "epoch": 1.0201847091307246,
+      "grad_norm": 4.30879927940807,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 42860
+    },
+    {
+      "epoch": 1.020208511853756,
+      "grad_norm": 0.4074561454333189,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 42861
+    },
+    {
+      "epoch": 1.0202323145767875,
+      "grad_norm": 0.012032900463138263,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42862
+    },
+    {
+      "epoch": 1.020256117299819,
+      "grad_norm": 19.822003156283255,
+      "learning_rate": 2e-05,
+      "loss": 0.2725,
+      "step": 42863
+    },
+    {
+      "epoch": 1.0202799200228505,
+      "grad_norm": 6.744116664056468,
+      "learning_rate": 2e-05,
+      "loss": 0.053,
+      "step": 42864
+    },
+    {
+      "epoch": 1.0203037227458822,
+      "grad_norm": 8.966831075083459,
+      "learning_rate": 2e-05,
+      "loss": 1.1247,
+      "step": 42865
+    },
+    {
+      "epoch": 1.0203275254689137,
+      "grad_norm": 12.849374953773536,
+      "learning_rate": 2e-05,
+      "loss": 0.7224,
+      "step": 42866
+    },
+    {
+      "epoch": 1.0203513281919452,
+      "grad_norm": 4.001664015690829,
+      "learning_rate": 2e-05,
+      "loss": 0.4203,
+      "step": 42867
+    },
+    {
+      "epoch": 1.0203751309149767,
+      "grad_norm": 3.1160391352593204,
+      "learning_rate": 2e-05,
+      "loss": 0.1352,
+      "step": 42868
+    },
+    {
+      "epoch": 1.0203989336380082,
+      "grad_norm": 1.917419826502667,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 42869
+    },
+    {
+      "epoch": 1.0204227363610396,
+      "grad_norm": 0.4235572859643444,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 42870
+    },
+    {
+      "epoch": 1.0204465390840711,
+      "grad_norm": 13.763900661680976,
+      "learning_rate": 2e-05,
+      "loss": 1.892,
+      "step": 42871
+    },
+    {
+      "epoch": 1.0204703418071028,
+      "grad_norm": 7.373669334359081,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 42872
+    },
+    {
+      "epoch": 1.0204941445301343,
+      "grad_norm": 0.04984900123363056,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42873
+    },
+    {
+      "epoch": 1.0205179472531658,
+      "grad_norm": 1.3892081744279385,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 42874
+    },
+    {
+      "epoch": 1.0205417499761973,
+      "grad_norm": 5.619963424807037,
+      "learning_rate": 2e-05,
+      "loss": 0.5801,
+      "step": 42875
+    },
+    {
+      "epoch": 1.0205655526992288,
+      "grad_norm": 19.654055121947636,
+      "learning_rate": 2e-05,
+      "loss": 0.7699,
+      "step": 42876
+    },
+    {
+      "epoch": 1.0205893554222603,
+      "grad_norm": 0.01751290601494387,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42877
+    },
+    {
+      "epoch": 1.0206131581452917,
+      "grad_norm": 5.95841311243115,
+      "learning_rate": 2e-05,
+      "loss": 0.1638,
+      "step": 42878
+    },
+    {
+      "epoch": 1.0206369608683232,
+      "grad_norm": 9.543452593663307,
+      "learning_rate": 2e-05,
+      "loss": 1.1179,
+      "step": 42879
+    },
+    {
+      "epoch": 1.020660763591355,
+      "grad_norm": 2.7521007488389095,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 42880
+    },
+    {
+      "epoch": 1.0206845663143864,
+      "grad_norm": 0.23180357351346656,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 42881
+    },
+    {
+      "epoch": 1.020708369037418,
+      "grad_norm": 5.415074608932646,
+      "learning_rate": 2e-05,
+      "loss": 1.1108,
+      "step": 42882
+    },
+    {
+      "epoch": 1.0207321717604494,
+      "grad_norm": 2.2785104895196637,
+      "learning_rate": 2e-05,
+      "loss": 0.1997,
+      "step": 42883
+    },
+    {
+      "epoch": 1.0207559744834809,
+      "grad_norm": 4.9571115275119375,
+      "learning_rate": 2e-05,
+      "loss": 0.3556,
+      "step": 42884
+    },
+    {
+      "epoch": 1.0207797772065124,
+      "grad_norm": 0.06594576944797759,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 42885
+    },
+    {
+      "epoch": 1.0208035799295438,
+      "grad_norm": 2.625509789025863,
+      "learning_rate": 2e-05,
+      "loss": 0.4766,
+      "step": 42886
+    },
+    {
+      "epoch": 1.0208273826525756,
+      "grad_norm": 7.023111687683935,
+      "learning_rate": 2e-05,
+      "loss": 0.4858,
+      "step": 42887
+    },
+    {
+      "epoch": 1.020851185375607,
+      "grad_norm": 2.2863183734885038,
+      "learning_rate": 2e-05,
+      "loss": 0.3466,
+      "step": 42888
+    },
+    {
+      "epoch": 1.0208749880986385,
+      "grad_norm": 2.374522096607728,
+      "learning_rate": 2e-05,
+      "loss": 0.0365,
+      "step": 42889
+    },
+    {
+      "epoch": 1.02089879082167,
+      "grad_norm": 6.562388162381208,
+      "learning_rate": 2e-05,
+      "loss": 0.2815,
+      "step": 42890
+    },
+    {
+      "epoch": 1.0209225935447015,
+      "grad_norm": 4.38705103582836,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 42891
+    },
+    {
+      "epoch": 1.020946396267733,
+      "grad_norm": 2.0355876773443358,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 42892
+    },
+    {
+      "epoch": 1.0209701989907645,
+      "grad_norm": 3.8043193255354204,
+      "learning_rate": 2e-05,
+      "loss": 0.4594,
+      "step": 42893
+    },
+    {
+      "epoch": 1.020994001713796,
+      "grad_norm": 3.413197127630223,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 42894
+    },
+    {
+      "epoch": 1.0210178044368277,
+      "grad_norm": 0.14308586040304294,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 42895
+    },
+    {
+      "epoch": 1.0210416071598591,
+      "grad_norm": 6.95937697777709,
+      "learning_rate": 2e-05,
+      "loss": 0.6562,
+      "step": 42896
+    },
+    {
+      "epoch": 1.0210654098828906,
+      "grad_norm": 2.8500469137086566,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 42897
+    },
+    {
+      "epoch": 1.0210892126059221,
+      "grad_norm": 7.732496929259591,
+      "learning_rate": 2e-05,
+      "loss": 1.1059,
+      "step": 42898
+    },
+    {
+      "epoch": 1.0211130153289536,
+      "grad_norm": 4.5285595622837365,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 42899
+    },
+    {
+      "epoch": 1.021136818051985,
+      "grad_norm": 0.942336468961627,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 42900
+    },
+    {
+      "epoch": 1.0211606207750166,
+      "grad_norm": 8.13901821033426,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 42901
+    },
+    {
+      "epoch": 1.0211844234980483,
+      "grad_norm": 2.786260493203588,
+      "learning_rate": 2e-05,
+      "loss": 0.1286,
+      "step": 42902
+    },
+    {
+      "epoch": 1.0212082262210798,
+      "grad_norm": 21.9306770973768,
+      "learning_rate": 2e-05,
+      "loss": 0.5887,
+      "step": 42903
+    },
+    {
+      "epoch": 1.0212320289441112,
+      "grad_norm": 3.190372971162756,
+      "learning_rate": 2e-05,
+      "loss": 0.3357,
+      "step": 42904
+    },
+    {
+      "epoch": 1.0212558316671427,
+      "grad_norm": 2.90025189883184,
+      "learning_rate": 2e-05,
+      "loss": 0.1275,
+      "step": 42905
+    },
+    {
+      "epoch": 1.0212796343901742,
+      "grad_norm": 0.11510336028429431,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 42906
+    },
+    {
+      "epoch": 1.0213034371132057,
+      "grad_norm": 3.997940012765767,
+      "learning_rate": 2e-05,
+      "loss": 0.6209,
+      "step": 42907
+    },
+    {
+      "epoch": 1.0213272398362372,
+      "grad_norm": 0.3968967994237455,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 42908
+    },
+    {
+      "epoch": 1.0213510425592687,
+      "grad_norm": 3.756536654003028,
+      "learning_rate": 2e-05,
+      "loss": 0.2208,
+      "step": 42909
+    },
+    {
+      "epoch": 1.0213748452823004,
+      "grad_norm": 4.326493152983201,
+      "learning_rate": 2e-05,
+      "loss": 0.178,
+      "step": 42910
+    },
+    {
+      "epoch": 1.0213986480053319,
+      "grad_norm": 7.223944857944806,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 42911
+    },
+    {
+      "epoch": 1.0214224507283634,
+      "grad_norm": 8.817204809574516,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 42912
+    },
+    {
+      "epoch": 1.0214462534513948,
+      "grad_norm": 12.837675885764588,
+      "learning_rate": 2e-05,
+      "loss": 0.1353,
+      "step": 42913
+    },
+    {
+      "epoch": 1.0214700561744263,
+      "grad_norm": 2.418760344109082,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 42914
+    },
+    {
+      "epoch": 1.0214938588974578,
+      "grad_norm": 3.108306654537264,
+      "learning_rate": 2e-05,
+      "loss": 0.3454,
+      "step": 42915
+    },
+    {
+      "epoch": 1.0215176616204893,
+      "grad_norm": 5.403558448030794,
+      "learning_rate": 2e-05,
+      "loss": 0.561,
+      "step": 42916
+    },
+    {
+      "epoch": 1.021541464343521,
+      "grad_norm": 3.0847279126170433,
+      "learning_rate": 2e-05,
+      "loss": 0.0221,
+      "step": 42917
+    },
+    {
+      "epoch": 1.0215652670665525,
+      "grad_norm": 0.5727360506537695,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 42918
+    },
+    {
+      "epoch": 1.021589069789584,
+      "grad_norm": 2.5755586454004438,
+      "learning_rate": 2e-05,
+      "loss": 0.2282,
+      "step": 42919
+    },
+    {
+      "epoch": 1.0216128725126155,
+      "grad_norm": 0.14385060519252857,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42920
+    },
+    {
+      "epoch": 1.021636675235647,
+      "grad_norm": 4.7317423819158835,
+      "learning_rate": 2e-05,
+      "loss": 0.7265,
+      "step": 42921
+    },
+    {
+      "epoch": 1.0216604779586784,
+      "grad_norm": 6.951769877545907,
+      "learning_rate": 2e-05,
+      "loss": 0.6121,
+      "step": 42922
+    },
+    {
+      "epoch": 1.02168428068171,
+      "grad_norm": 4.741583829667924,
+      "learning_rate": 2e-05,
+      "loss": 0.4197,
+      "step": 42923
+    },
+    {
+      "epoch": 1.0217080834047414,
+      "grad_norm": 4.150494989270398,
+      "learning_rate": 2e-05,
+      "loss": 0.1452,
+      "step": 42924
+    },
+    {
+      "epoch": 1.021731886127773,
+      "grad_norm": 0.10125737073983983,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 42925
+    },
+    {
+      "epoch": 1.0217556888508046,
+      "grad_norm": 0.9496492675270283,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 42926
+    },
+    {
+      "epoch": 1.021779491573836,
+      "grad_norm": 1.881046301593703,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 42927
+    },
+    {
+      "epoch": 1.0218032942968676,
+      "grad_norm": 0.3739015784407189,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 42928
+    },
+    {
+      "epoch": 1.021827097019899,
+      "grad_norm": 2.8636576568854473,
+      "learning_rate": 2e-05,
+      "loss": 0.1661,
+      "step": 42929
+    },
+    {
+      "epoch": 1.0218508997429305,
+      "grad_norm": 2.2396107589371113,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 42930
+    },
+    {
+      "epoch": 1.021874702465962,
+      "grad_norm": 0.35520410096797467,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 42931
+    },
+    {
+      "epoch": 1.0218985051889937,
+      "grad_norm": 12.429097128214982,
+      "learning_rate": 2e-05,
+      "loss": 0.4597,
+      "step": 42932
+    },
+    {
+      "epoch": 1.0219223079120252,
+      "grad_norm": 0.03155249238483275,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 42933
+    },
+    {
+      "epoch": 1.0219461106350567,
+      "grad_norm": 0.2339117187188423,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 42934
+    },
+    {
+      "epoch": 1.0219699133580882,
+      "grad_norm": 0.057383825643060354,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 42935
+    },
+    {
+      "epoch": 1.0219937160811197,
+      "grad_norm": 5.744274622162273,
+      "learning_rate": 2e-05,
+      "loss": 0.0799,
+      "step": 42936
+    },
+    {
+      "epoch": 1.0220175188041511,
+      "grad_norm": 2.6029357776084967,
+      "learning_rate": 2e-05,
+      "loss": 0.3767,
+      "step": 42937
+    },
+    {
+      "epoch": 1.0220413215271826,
+      "grad_norm": 0.18865839118725738,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 42938
+    },
+    {
+      "epoch": 1.0220651242502141,
+      "grad_norm": 10.871879744376029,
+      "learning_rate": 2e-05,
+      "loss": 0.6877,
+      "step": 42939
+    },
+    {
+      "epoch": 1.0220889269732458,
+      "grad_norm": 11.71421523462698,
+      "learning_rate": 2e-05,
+      "loss": 0.3166,
+      "step": 42940
+    },
+    {
+      "epoch": 1.0221127296962773,
+      "grad_norm": 9.503527843667092,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 42941
+    },
+    {
+      "epoch": 1.0221365324193088,
+      "grad_norm": 3.748935168864752,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 42942
+    },
+    {
+      "epoch": 1.0221603351423403,
+      "grad_norm": 0.6989299471137185,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 42943
+    },
+    {
+      "epoch": 1.0221841378653718,
+      "grad_norm": 9.809381757236224,
+      "learning_rate": 2e-05,
+      "loss": 0.6051,
+      "step": 42944
+    },
+    {
+      "epoch": 1.0222079405884033,
+      "grad_norm": 0.01546444193094077,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 42945
+    },
+    {
+      "epoch": 1.0222317433114347,
+      "grad_norm": 3.0480184654109683,
+      "learning_rate": 2e-05,
+      "loss": 0.1572,
+      "step": 42946
+    },
+    {
+      "epoch": 1.0222555460344664,
+      "grad_norm": 6.050186154879345,
+      "learning_rate": 2e-05,
+      "loss": 0.3125,
+      "step": 42947
+    },
+    {
+      "epoch": 1.022279348757498,
+      "grad_norm": 6.447029686748869,
+      "learning_rate": 2e-05,
+      "loss": 0.5383,
+      "step": 42948
+    },
+    {
+      "epoch": 1.0223031514805294,
+      "grad_norm": 5.376499210946384,
+      "learning_rate": 2e-05,
+      "loss": 0.219,
+      "step": 42949
+    },
+    {
+      "epoch": 1.022326954203561,
+      "grad_norm": 2.9316800960436105,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 42950
+    },
+    {
+      "epoch": 1.0223507569265924,
+      "grad_norm": 0.2873063004441544,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 42951
+    },
+    {
+      "epoch": 1.0223745596496239,
+      "grad_norm": 0.0704282696512497,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 42952
+    },
+    {
+      "epoch": 1.0223983623726554,
+      "grad_norm": 3.1247128418702874,
+      "learning_rate": 2e-05,
+      "loss": 0.08,
+      "step": 42953
+    },
+    {
+      "epoch": 1.0224221650956868,
+      "grad_norm": 0.04758909146093255,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 42954
+    },
+    {
+      "epoch": 1.0224459678187185,
+      "grad_norm": 4.443884759729466,
+      "learning_rate": 2e-05,
+      "loss": 0.4747,
+      "step": 42955
+    },
+    {
+      "epoch": 1.02246977054175,
+      "grad_norm": 3.631941426774211,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 42956
+    },
+    {
+      "epoch": 1.0224935732647815,
+      "grad_norm": 9.510713259572322,
+      "learning_rate": 2e-05,
+      "loss": 0.3763,
+      "step": 42957
+    },
+    {
+      "epoch": 1.022517375987813,
+      "grad_norm": 3.0859890108128556,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 42958
+    },
+    {
+      "epoch": 1.0225411787108445,
+      "grad_norm": 8.925414495784914,
+      "learning_rate": 2e-05,
+      "loss": 0.9722,
+      "step": 42959
+    },
+    {
+      "epoch": 1.022564981433876,
+      "grad_norm": 0.9404011605074865,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 42960
+    },
+    {
+      "epoch": 1.0225887841569075,
+      "grad_norm": 0.695772227296205,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 42961
+    },
+    {
+      "epoch": 1.0226125868799392,
+      "grad_norm": 1.6548498392656417,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 42962
+    },
+    {
+      "epoch": 1.0226363896029707,
+      "grad_norm": 1.677105852961157,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 42963
+    },
+    {
+      "epoch": 1.0226601923260021,
+      "grad_norm": 3.3301030945761507,
+      "learning_rate": 2e-05,
+      "loss": 0.4618,
+      "step": 42964
+    },
+    {
+      "epoch": 1.0226839950490336,
+      "grad_norm": 4.837291776386138,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 42965
+    },
+    {
+      "epoch": 1.022707797772065,
+      "grad_norm": 0.030339762428244994,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42966
+    },
+    {
+      "epoch": 1.0227316004950966,
+      "grad_norm": 17.1569512424598,
+      "learning_rate": 2e-05,
+      "loss": 0.5113,
+      "step": 42967
+    },
+    {
+      "epoch": 1.022755403218128,
+      "grad_norm": 2.210743278828661,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 42968
+    },
+    {
+      "epoch": 1.0227792059411596,
+      "grad_norm": 5.40267439797968,
+      "learning_rate": 2e-05,
+      "loss": 0.3152,
+      "step": 42969
+    },
+    {
+      "epoch": 1.0228030086641913,
+      "grad_norm": 3.4445611095262163,
+      "learning_rate": 2e-05,
+      "loss": 0.2634,
+      "step": 42970
+    },
+    {
+      "epoch": 1.0228268113872228,
+      "grad_norm": 7.155181182378581,
+      "learning_rate": 2e-05,
+      "loss": 0.6983,
+      "step": 42971
+    },
+    {
+      "epoch": 1.0228506141102542,
+      "grad_norm": 3.0298977890869367,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 42972
+    },
+    {
+      "epoch": 1.0228744168332857,
+      "grad_norm": 3.429691229109432,
+      "learning_rate": 2e-05,
+      "loss": 0.146,
+      "step": 42973
+    },
+    {
+      "epoch": 1.0228982195563172,
+      "grad_norm": 0.1505083950640253,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 42974
+    },
+    {
+      "epoch": 1.0229220222793487,
+      "grad_norm": 3.6384676212263254,
+      "learning_rate": 2e-05,
+      "loss": 0.4277,
+      "step": 42975
+    },
+    {
+      "epoch": 1.0229458250023802,
+      "grad_norm": 0.010459110759826012,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 42976
+    },
+    {
+      "epoch": 1.022969627725412,
+      "grad_norm": 0.6107856414634649,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 42977
+    },
+    {
+      "epoch": 1.0229934304484434,
+      "grad_norm": 0.02015018390211144,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42978
+    },
+    {
+      "epoch": 1.0230172331714749,
+      "grad_norm": 4.009335384512755,
+      "learning_rate": 2e-05,
+      "loss": 0.5153,
+      "step": 42979
+    },
+    {
+      "epoch": 1.0230410358945063,
+      "grad_norm": 0.031615285727248615,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42980
+    },
+    {
+      "epoch": 1.0230648386175378,
+      "grad_norm": 23.607004087818623,
+      "learning_rate": 2e-05,
+      "loss": 0.9701,
+      "step": 42981
+    },
+    {
+      "epoch": 1.0230886413405693,
+      "grad_norm": 4.303926254083325,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 42982
+    },
+    {
+      "epoch": 1.0231124440636008,
+      "grad_norm": 11.851644647530248,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 42983
+    },
+    {
+      "epoch": 1.0231362467866323,
+      "grad_norm": 2.4355566017566885,
+      "learning_rate": 2e-05,
+      "loss": 0.2755,
+      "step": 42984
+    },
+    {
+      "epoch": 1.023160049509664,
+      "grad_norm": 0.08401261365697678,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 42985
+    },
+    {
+      "epoch": 1.0231838522326955,
+      "grad_norm": 2.7999920507997507,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 42986
+    },
+    {
+      "epoch": 1.023207654955727,
+      "grad_norm": 0.19631977631625502,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 42987
+    },
+    {
+      "epoch": 1.0232314576787584,
+      "grad_norm": 0.03507505064883148,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 42988
+    },
+    {
+      "epoch": 1.02325526040179,
+      "grad_norm": 2.708417412670678,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 42989
+    },
+    {
+      "epoch": 1.0232790631248214,
+      "grad_norm": 3.5575877251868215,
+      "learning_rate": 2e-05,
+      "loss": 0.4153,
+      "step": 42990
+    },
+    {
+      "epoch": 1.023302865847853,
+      "grad_norm": 7.190310734721902,
+      "learning_rate": 2e-05,
+      "loss": 1.0021,
+      "step": 42991
+    },
+    {
+      "epoch": 1.0233266685708846,
+      "grad_norm": 3.5994290642743496,
+      "learning_rate": 2e-05,
+      "loss": 0.2739,
+      "step": 42992
+    },
+    {
+      "epoch": 1.023350471293916,
+      "grad_norm": 8.64127526744918,
+      "learning_rate": 2e-05,
+      "loss": 1.0866,
+      "step": 42993
+    },
+    {
+      "epoch": 1.0233742740169476,
+      "grad_norm": 3.8506655042816127,
+      "learning_rate": 2e-05,
+      "loss": 0.3983,
+      "step": 42994
+    },
+    {
+      "epoch": 1.023398076739979,
+      "grad_norm": 0.7083392479302147,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 42995
+    },
+    {
+      "epoch": 1.0234218794630106,
+      "grad_norm": 3.6370713860143233,
+      "learning_rate": 2e-05,
+      "loss": 0.3676,
+      "step": 42996
+    },
+    {
+      "epoch": 1.023445682186042,
+      "grad_norm": 0.0018901379553446956,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 42997
+    },
+    {
+      "epoch": 1.0234694849090735,
+      "grad_norm": 17.531693486081416,
+      "learning_rate": 2e-05,
+      "loss": 1.4486,
+      "step": 42998
+    },
+    {
+      "epoch": 1.023493287632105,
+      "grad_norm": 4.612941365733405,
+      "learning_rate": 2e-05,
+      "loss": 0.1853,
+      "step": 42999
+    },
+    {
+      "epoch": 1.0235170903551367,
+      "grad_norm": 0.1867190676649928,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43000
+    },
+    {
+      "epoch": 1.0235408930781682,
+      "grad_norm": 0.003820152705041331,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43001
+    },
+    {
+      "epoch": 1.0235646958011997,
+      "grad_norm": 3.1699469726479137,
+      "learning_rate": 2e-05,
+      "loss": 0.2355,
+      "step": 43002
+    },
+    {
+      "epoch": 1.0235884985242312,
+      "grad_norm": 3.505467030234058,
+      "learning_rate": 2e-05,
+      "loss": 0.1533,
+      "step": 43003
+    },
+    {
+      "epoch": 1.0236123012472627,
+      "grad_norm": 4.004355879209747,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 43004
+    },
+    {
+      "epoch": 1.0236361039702941,
+      "grad_norm": 5.547891447109857,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 43005
+    },
+    {
+      "epoch": 1.0236599066933256,
+      "grad_norm": 4.571781628747637,
+      "learning_rate": 2e-05,
+      "loss": 0.3523,
+      "step": 43006
+    },
+    {
+      "epoch": 1.0236837094163573,
+      "grad_norm": 1.4266177455767635,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 43007
+    },
+    {
+      "epoch": 1.0237075121393888,
+      "grad_norm": 8.938107471848722,
+      "learning_rate": 2e-05,
+      "loss": 0.9745,
+      "step": 43008
+    },
+    {
+      "epoch": 1.0237313148624203,
+      "grad_norm": 0.022812449353012985,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43009
+    },
+    {
+      "epoch": 1.0237551175854518,
+      "grad_norm": 5.742537883717789,
+      "learning_rate": 2e-05,
+      "loss": 0.2135,
+      "step": 43010
+    },
+    {
+      "epoch": 1.0237789203084833,
+      "grad_norm": 6.136930464494072,
+      "learning_rate": 2e-05,
+      "loss": 0.4227,
+      "step": 43011
+    },
+    {
+      "epoch": 1.0238027230315148,
+      "grad_norm": 0.020491567145263675,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43012
+    },
+    {
+      "epoch": 1.0238265257545462,
+      "grad_norm": 13.0903370055581,
+      "learning_rate": 2e-05,
+      "loss": 0.8075,
+      "step": 43013
+    },
+    {
+      "epoch": 1.0238503284775777,
+      "grad_norm": 0.7952748471235432,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 43014
+    },
+    {
+      "epoch": 1.0238741312006094,
+      "grad_norm": 0.08148952397623578,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43015
+    },
+    {
+      "epoch": 1.023897933923641,
+      "grad_norm": 3.7598500605394944,
+      "learning_rate": 2e-05,
+      "loss": 0.2989,
+      "step": 43016
+    },
+    {
+      "epoch": 1.0239217366466724,
+      "grad_norm": 0.06885068635589044,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43017
+    },
+    {
+      "epoch": 1.023945539369704,
+      "grad_norm": 1.5799044508404474,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 43018
+    },
+    {
+      "epoch": 1.0239693420927354,
+      "grad_norm": 1.137218879887109,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 43019
+    },
+    {
+      "epoch": 1.0239931448157669,
+      "grad_norm": 5.79493800557809,
+      "learning_rate": 2e-05,
+      "loss": 0.6989,
+      "step": 43020
+    },
+    {
+      "epoch": 1.0240169475387983,
+      "grad_norm": 0.14995817137079637,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 43021
+    },
+    {
+      "epoch": 1.02404075026183,
+      "grad_norm": 5.49032090824105,
+      "learning_rate": 2e-05,
+      "loss": 0.5382,
+      "step": 43022
+    },
+    {
+      "epoch": 1.0240645529848615,
+      "grad_norm": 0.21082560277051285,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 43023
+    },
+    {
+      "epoch": 1.024088355707893,
+      "grad_norm": 0.02744919552824591,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43024
+    },
+    {
+      "epoch": 1.0241121584309245,
+      "grad_norm": 11.48150771468292,
+      "learning_rate": 2e-05,
+      "loss": 0.2643,
+      "step": 43025
+    },
+    {
+      "epoch": 1.024135961153956,
+      "grad_norm": 3.5780485011441048,
+      "learning_rate": 2e-05,
+      "loss": 0.3113,
+      "step": 43026
+    },
+    {
+      "epoch": 1.0241597638769875,
+      "grad_norm": 4.682424792298836,
+      "learning_rate": 2e-05,
+      "loss": 0.2177,
+      "step": 43027
+    },
+    {
+      "epoch": 1.024183566600019,
+      "grad_norm": 4.1202373521917695,
+      "learning_rate": 2e-05,
+      "loss": 0.6083,
+      "step": 43028
+    },
+    {
+      "epoch": 1.0242073693230505,
+      "grad_norm": 5.03289325099839,
+      "learning_rate": 2e-05,
+      "loss": 0.334,
+      "step": 43029
+    },
+    {
+      "epoch": 1.0242311720460822,
+      "grad_norm": 2.4867436504786045,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 43030
+    },
+    {
+      "epoch": 1.0242549747691136,
+      "grad_norm": 2.268937365468175,
+      "learning_rate": 2e-05,
+      "loss": 0.1611,
+      "step": 43031
+    },
+    {
+      "epoch": 1.0242787774921451,
+      "grad_norm": 5.847217052772523,
+      "learning_rate": 2e-05,
+      "loss": 0.145,
+      "step": 43032
+    },
+    {
+      "epoch": 1.0243025802151766,
+      "grad_norm": 0.3617366274606756,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 43033
+    },
+    {
+      "epoch": 1.024326382938208,
+      "grad_norm": 0.9437036119611396,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 43034
+    },
+    {
+      "epoch": 1.0243501856612396,
+      "grad_norm": 0.01985790677716553,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43035
+    },
+    {
+      "epoch": 1.024373988384271,
+      "grad_norm": 10.838838851528473,
+      "learning_rate": 2e-05,
+      "loss": 0.1728,
+      "step": 43036
+    },
+    {
+      "epoch": 1.0243977911073028,
+      "grad_norm": 4.935348093158699,
+      "learning_rate": 2e-05,
+      "loss": 0.5611,
+      "step": 43037
+    },
+    {
+      "epoch": 1.0244215938303343,
+      "grad_norm": 2.372760253571879,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 43038
+    },
+    {
+      "epoch": 1.0244453965533658,
+      "grad_norm": 3.8206495569151424,
+      "learning_rate": 2e-05,
+      "loss": 0.3032,
+      "step": 43039
+    },
+    {
+      "epoch": 1.0244691992763972,
+      "grad_norm": 1.2453763739719554,
+      "learning_rate": 2e-05,
+      "loss": 0.0295,
+      "step": 43040
+    },
+    {
+      "epoch": 1.0244930019994287,
+      "grad_norm": 2.1719429023481265,
+      "learning_rate": 2e-05,
+      "loss": 0.023,
+      "step": 43041
+    },
+    {
+      "epoch": 1.0245168047224602,
+      "grad_norm": 0.2064973874487792,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43042
+    },
+    {
+      "epoch": 1.0245406074454917,
+      "grad_norm": 2.8516143384856893,
+      "learning_rate": 2e-05,
+      "loss": 0.2485,
+      "step": 43043
+    },
+    {
+      "epoch": 1.0245644101685232,
+      "grad_norm": 15.163955310169214,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 43044
+    },
+    {
+      "epoch": 1.0245882128915549,
+      "grad_norm": 2.454141632456098,
+      "learning_rate": 2e-05,
+      "loss": 0.307,
+      "step": 43045
+    },
+    {
+      "epoch": 1.0246120156145864,
+      "grad_norm": 5.3773587853546125,
+      "learning_rate": 2e-05,
+      "loss": 0.6835,
+      "step": 43046
+    },
+    {
+      "epoch": 1.0246358183376179,
+      "grad_norm": 11.806971903257152,
+      "learning_rate": 2e-05,
+      "loss": 0.6419,
+      "step": 43047
+    },
+    {
+      "epoch": 1.0246596210606493,
+      "grad_norm": 10.434349815179104,
+      "learning_rate": 2e-05,
+      "loss": 0.7372,
+      "step": 43048
+    },
+    {
+      "epoch": 1.0246834237836808,
+      "grad_norm": 2.5935042310023153,
+      "learning_rate": 2e-05,
+      "loss": 0.0246,
+      "step": 43049
+    },
+    {
+      "epoch": 1.0247072265067123,
+      "grad_norm": 3.41598845649607,
+      "learning_rate": 2e-05,
+      "loss": 0.0486,
+      "step": 43050
+    },
+    {
+      "epoch": 1.0247310292297438,
+      "grad_norm": 5.584250734051653,
+      "learning_rate": 2e-05,
+      "loss": 0.8867,
+      "step": 43051
+    },
+    {
+      "epoch": 1.0247548319527755,
+      "grad_norm": 5.045651230873095,
+      "learning_rate": 2e-05,
+      "loss": 0.7487,
+      "step": 43052
+    },
+    {
+      "epoch": 1.024778634675807,
+      "grad_norm": 2.9202592866706185,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 43053
+    },
+    {
+      "epoch": 1.0248024373988385,
+      "grad_norm": 0.4264890328412883,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 43054
+    },
+    {
+      "epoch": 1.02482624012187,
+      "grad_norm": 5.258088215406457,
+      "learning_rate": 2e-05,
+      "loss": 0.5464,
+      "step": 43055
+    },
+    {
+      "epoch": 1.0248500428449014,
+      "grad_norm": 7.011268707540622,
+      "learning_rate": 2e-05,
+      "loss": 0.1241,
+      "step": 43056
+    },
+    {
+      "epoch": 1.024873845567933,
+      "grad_norm": 5.646224031594321,
+      "learning_rate": 2e-05,
+      "loss": 0.7921,
+      "step": 43057
+    },
+    {
+      "epoch": 1.0248976482909644,
+      "grad_norm": 0.06678897055492057,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43058
+    },
+    {
+      "epoch": 1.024921451013996,
+      "grad_norm": 3.24987977288388,
+      "learning_rate": 2e-05,
+      "loss": 0.0343,
+      "step": 43059
+    },
+    {
+      "epoch": 1.0249452537370276,
+      "grad_norm": 4.614398604014577,
+      "learning_rate": 2e-05,
+      "loss": 0.4847,
+      "step": 43060
+    },
+    {
+      "epoch": 1.024969056460059,
+      "grad_norm": 0.6316625243346956,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 43061
+    },
+    {
+      "epoch": 1.0249928591830906,
+      "grad_norm": 4.735965052464638,
+      "learning_rate": 2e-05,
+      "loss": 0.6383,
+      "step": 43062
+    },
+    {
+      "epoch": 1.025016661906122,
+      "grad_norm": 13.106713838422744,
+      "learning_rate": 2e-05,
+      "loss": 0.1421,
+      "step": 43063
+    },
+    {
+      "epoch": 1.0250404646291535,
+      "grad_norm": 4.444658476371819,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 43064
+    },
+    {
+      "epoch": 1.025064267352185,
+      "grad_norm": 4.716741810827677,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 43065
+    },
+    {
+      "epoch": 1.0250880700752165,
+      "grad_norm": 2.231110604456979,
+      "learning_rate": 2e-05,
+      "loss": 0.2341,
+      "step": 43066
+    },
+    {
+      "epoch": 1.0251118727982482,
+      "grad_norm": 5.6854087496264265,
+      "learning_rate": 2e-05,
+      "loss": 0.9376,
+      "step": 43067
+    },
+    {
+      "epoch": 1.0251356755212797,
+      "grad_norm": 3.6183168373097243,
+      "learning_rate": 2e-05,
+      "loss": 0.3268,
+      "step": 43068
+    },
+    {
+      "epoch": 1.0251594782443112,
+      "grad_norm": 5.342163786515696,
+      "learning_rate": 2e-05,
+      "loss": 0.6129,
+      "step": 43069
+    },
+    {
+      "epoch": 1.0251832809673427,
+      "grad_norm": 1.043305848464887,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 43070
+    },
+    {
+      "epoch": 1.0252070836903742,
+      "grad_norm": 0.9284629117821391,
+      "learning_rate": 2e-05,
+      "loss": 0.0188,
+      "step": 43071
+    },
+    {
+      "epoch": 1.0252308864134057,
+      "grad_norm": 6.892963917959387,
+      "learning_rate": 2e-05,
+      "loss": 0.1031,
+      "step": 43072
+    },
+    {
+      "epoch": 1.0252546891364371,
+      "grad_norm": 1.9293510067752728,
+      "learning_rate": 2e-05,
+      "loss": 0.2912,
+      "step": 43073
+    },
+    {
+      "epoch": 1.0252784918594686,
+      "grad_norm": 3.6660547110019435,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 43074
+    },
+    {
+      "epoch": 1.0253022945825003,
+      "grad_norm": 7.509092218035608,
+      "learning_rate": 2e-05,
+      "loss": 0.086,
+      "step": 43075
+    },
+    {
+      "epoch": 1.0253260973055318,
+      "grad_norm": 0.15660848538201874,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43076
+    },
+    {
+      "epoch": 1.0253499000285633,
+      "grad_norm": 0.08851627674753243,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43077
+    },
+    {
+      "epoch": 1.0253737027515948,
+      "grad_norm": 3.5753161667737183,
+      "learning_rate": 2e-05,
+      "loss": 0.4121,
+      "step": 43078
+    },
+    {
+      "epoch": 1.0253975054746263,
+      "grad_norm": 5.466173258649228,
+      "learning_rate": 2e-05,
+      "loss": 0.5352,
+      "step": 43079
+    },
+    {
+      "epoch": 1.0254213081976578,
+      "grad_norm": 0.8670402808081341,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 43080
+    },
+    {
+      "epoch": 1.0254451109206892,
+      "grad_norm": 2.1743461764690104,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 43081
+    },
+    {
+      "epoch": 1.025468913643721,
+      "grad_norm": 0.43562173117850633,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 43082
+    },
+    {
+      "epoch": 1.0254927163667524,
+      "grad_norm": 2.9701498447457815,
+      "learning_rate": 2e-05,
+      "loss": 0.3315,
+      "step": 43083
+    },
+    {
+      "epoch": 1.025516519089784,
+      "grad_norm": 11.292823565170577,
+      "learning_rate": 2e-05,
+      "loss": 0.1923,
+      "step": 43084
+    },
+    {
+      "epoch": 1.0255403218128154,
+      "grad_norm": 2.036277072158003,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 43085
+    },
+    {
+      "epoch": 1.0255641245358469,
+      "grad_norm": 3.2183666730802574,
+      "learning_rate": 2e-05,
+      "loss": 0.3475,
+      "step": 43086
+    },
+    {
+      "epoch": 1.0255879272588784,
+      "grad_norm": 0.097795885555403,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43087
+    },
+    {
+      "epoch": 1.0256117299819099,
+      "grad_norm": 3.049732139702765,
+      "learning_rate": 2e-05,
+      "loss": 0.2981,
+      "step": 43088
+    },
+    {
+      "epoch": 1.0256355327049413,
+      "grad_norm": 2.382699102676473,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 43089
+    },
+    {
+      "epoch": 1.025659335427973,
+      "grad_norm": 2.1667220701691767,
+      "learning_rate": 2e-05,
+      "loss": 0.148,
+      "step": 43090
+    },
+    {
+      "epoch": 1.0256831381510045,
+      "grad_norm": 34.85229165305888,
+      "learning_rate": 2e-05,
+      "loss": 0.6213,
+      "step": 43091
+    },
+    {
+      "epoch": 1.025706940874036,
+      "grad_norm": 3.774428593135724,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 43092
+    },
+    {
+      "epoch": 1.0257307435970675,
+      "grad_norm": 0.16896152921182073,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43093
+    },
+    {
+      "epoch": 1.025754546320099,
+      "grad_norm": 0.0162690536591626,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43094
+    },
+    {
+      "epoch": 1.0257783490431305,
+      "grad_norm": 0.36276362155489855,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 43095
+    },
+    {
+      "epoch": 1.025802151766162,
+      "grad_norm": 2.8250644177648816,
+      "learning_rate": 2e-05,
+      "loss": 0.099,
+      "step": 43096
+    },
+    {
+      "epoch": 1.0258259544891937,
+      "grad_norm": 0.4641885533165349,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 43097
+    },
+    {
+      "epoch": 1.0258497572122252,
+      "grad_norm": 0.1336633579501335,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43098
+    },
+    {
+      "epoch": 1.0258735599352566,
+      "grad_norm": 4.193588345837011,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 43099
+    },
+    {
+      "epoch": 1.0258973626582881,
+      "grad_norm": 6.244374288320527,
+      "learning_rate": 2e-05,
+      "loss": 0.8543,
+      "step": 43100
+    },
+    {
+      "epoch": 1.0259211653813196,
+      "grad_norm": 0.5564309111736236,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 43101
+    },
+    {
+      "epoch": 1.025944968104351,
+      "grad_norm": 2.8836779640480494,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 43102
+    },
+    {
+      "epoch": 1.0259687708273826,
+      "grad_norm": 0.712621827184808,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 43103
+    },
+    {
+      "epoch": 1.025992573550414,
+      "grad_norm": 0.6275213009007109,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 43104
+    },
+    {
+      "epoch": 1.0260163762734458,
+      "grad_norm": 6.533838941414008,
+      "learning_rate": 2e-05,
+      "loss": 0.6461,
+      "step": 43105
+    },
+    {
+      "epoch": 1.0260401789964773,
+      "grad_norm": 5.079257331758414,
+      "learning_rate": 2e-05,
+      "loss": 0.5291,
+      "step": 43106
+    },
+    {
+      "epoch": 1.0260639817195087,
+      "grad_norm": 3.774390615345785,
+      "learning_rate": 2e-05,
+      "loss": 0.6624,
+      "step": 43107
+    },
+    {
+      "epoch": 1.0260877844425402,
+      "grad_norm": 0.12753269508112725,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43108
+    },
+    {
+      "epoch": 1.0261115871655717,
+      "grad_norm": 3.144944260876912,
+      "learning_rate": 2e-05,
+      "loss": 0.4503,
+      "step": 43109
+    },
+    {
+      "epoch": 1.0261353898886032,
+      "grad_norm": 3.4801916005415268,
+      "learning_rate": 2e-05,
+      "loss": 0.4107,
+      "step": 43110
+    },
+    {
+      "epoch": 1.0261591926116347,
+      "grad_norm": 0.036319262913798195,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43111
+    },
+    {
+      "epoch": 1.0261829953346664,
+      "grad_norm": 7.766787227975632,
+      "learning_rate": 2e-05,
+      "loss": 1.0923,
+      "step": 43112
+    },
+    {
+      "epoch": 1.0262067980576979,
+      "grad_norm": 5.7345870909786765,
+      "learning_rate": 2e-05,
+      "loss": 0.2905,
+      "step": 43113
+    },
+    {
+      "epoch": 1.0262306007807294,
+      "grad_norm": 0.6891913760106746,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 43114
+    },
+    {
+      "epoch": 1.0262544035037608,
+      "grad_norm": 6.583280687034022,
+      "learning_rate": 2e-05,
+      "loss": 0.1567,
+      "step": 43115
+    },
+    {
+      "epoch": 1.0262782062267923,
+      "grad_norm": 4.2518400666922265,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 43116
+    },
+    {
+      "epoch": 1.0263020089498238,
+      "grad_norm": 3.4928771946906667,
+      "learning_rate": 2e-05,
+      "loss": 0.5306,
+      "step": 43117
+    },
+    {
+      "epoch": 1.0263258116728553,
+      "grad_norm": 0.054696801191424164,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43118
+    },
+    {
+      "epoch": 1.0263496143958868,
+      "grad_norm": 0.02129864611726702,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43119
+    },
+    {
+      "epoch": 1.0263734171189185,
+      "grad_norm": 0.1535259018270345,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43120
+    },
+    {
+      "epoch": 1.02639721984195,
+      "grad_norm": 6.618430375663579,
+      "learning_rate": 2e-05,
+      "loss": 0.3313,
+      "step": 43121
+    },
+    {
+      "epoch": 1.0264210225649815,
+      "grad_norm": 3.5937569177944555,
+      "learning_rate": 2e-05,
+      "loss": 0.1591,
+      "step": 43122
+    },
+    {
+      "epoch": 1.026444825288013,
+      "grad_norm": 0.11162037017258171,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43123
+    },
+    {
+      "epoch": 1.0264686280110444,
+      "grad_norm": 0.008927783959676677,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43124
+    },
+    {
+      "epoch": 1.026492430734076,
+      "grad_norm": 0.191285396153532,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 43125
+    },
+    {
+      "epoch": 1.0265162334571074,
+      "grad_norm": 0.671513958652283,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 43126
+    },
+    {
+      "epoch": 1.0265400361801391,
+      "grad_norm": 2.2309537418815397,
+      "learning_rate": 2e-05,
+      "loss": 0.2805,
+      "step": 43127
+    },
+    {
+      "epoch": 1.0265638389031706,
+      "grad_norm": 1.458196034579756,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 43128
+    },
+    {
+      "epoch": 1.026587641626202,
+      "grad_norm": 16.267764457519174,
+      "learning_rate": 2e-05,
+      "loss": 0.8329,
+      "step": 43129
+    },
+    {
+      "epoch": 1.0266114443492336,
+      "grad_norm": 4.235995033934685,
+      "learning_rate": 2e-05,
+      "loss": 0.404,
+      "step": 43130
+    },
+    {
+      "epoch": 1.026635247072265,
+      "grad_norm": 0.0016939125403366683,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 43131
+    },
+    {
+      "epoch": 1.0266590497952965,
+      "grad_norm": 0.08504005127759173,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43132
+    },
+    {
+      "epoch": 1.026682852518328,
+      "grad_norm": 6.291322860736794,
+      "learning_rate": 2e-05,
+      "loss": 0.5573,
+      "step": 43133
+    },
+    {
+      "epoch": 1.0267066552413595,
+      "grad_norm": 4.478190583746405,
+      "learning_rate": 2e-05,
+      "loss": 0.6574,
+      "step": 43134
+    },
+    {
+      "epoch": 1.0267304579643912,
+      "grad_norm": 1.2375617333882194,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 43135
+    },
+    {
+      "epoch": 1.0267542606874227,
+      "grad_norm": 4.35157220155214,
+      "learning_rate": 2e-05,
+      "loss": 0.044,
+      "step": 43136
+    },
+    {
+      "epoch": 1.0267780634104542,
+      "grad_norm": 0.22931204190367493,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 43137
+    },
+    {
+      "epoch": 1.0268018661334857,
+      "grad_norm": 5.558382432568929,
+      "learning_rate": 2e-05,
+      "loss": 0.4318,
+      "step": 43138
+    },
+    {
+      "epoch": 1.0268256688565172,
+      "grad_norm": 4.653577848834373,
+      "learning_rate": 2e-05,
+      "loss": 0.6293,
+      "step": 43139
+    },
+    {
+      "epoch": 1.0268494715795486,
+      "grad_norm": 10.700863718397791,
+      "learning_rate": 2e-05,
+      "loss": 0.7661,
+      "step": 43140
+    },
+    {
+      "epoch": 1.0268732743025801,
+      "grad_norm": 6.405788189205653,
+      "learning_rate": 2e-05,
+      "loss": 0.9489,
+      "step": 43141
+    },
+    {
+      "epoch": 1.0268970770256118,
+      "grad_norm": 44.492112582633325,
+      "learning_rate": 2e-05,
+      "loss": 0.4005,
+      "step": 43142
+    },
+    {
+      "epoch": 1.0269208797486433,
+      "grad_norm": 0.9872511335447618,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 43143
+    },
+    {
+      "epoch": 1.0269446824716748,
+      "grad_norm": 7.729655976663755,
+      "learning_rate": 2e-05,
+      "loss": 0.0743,
+      "step": 43144
+    },
+    {
+      "epoch": 1.0269684851947063,
+      "grad_norm": 4.334380026893057,
+      "learning_rate": 2e-05,
+      "loss": 0.5042,
+      "step": 43145
+    },
+    {
+      "epoch": 1.0269922879177378,
+      "grad_norm": 0.08505179882183517,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 43146
+    },
+    {
+      "epoch": 1.0270160906407693,
+      "grad_norm": 0.9488633122113351,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 43147
+    },
+    {
+      "epoch": 1.0270398933638007,
+      "grad_norm": 9.568540549027842,
+      "learning_rate": 2e-05,
+      "loss": 0.2953,
+      "step": 43148
+    },
+    {
+      "epoch": 1.0270636960868322,
+      "grad_norm": 24.915753729525015,
+      "learning_rate": 2e-05,
+      "loss": 0.4,
+      "step": 43149
+    },
+    {
+      "epoch": 1.027087498809864,
+      "grad_norm": 0.7966057871745346,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 43150
+    },
+    {
+      "epoch": 1.0271113015328954,
+      "grad_norm": 0.002373521291969538,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43151
+    },
+    {
+      "epoch": 1.027135104255927,
+      "grad_norm": 2.4547063295659965,
+      "learning_rate": 2e-05,
+      "loss": 0.0989,
+      "step": 43152
+    },
+    {
+      "epoch": 1.0271589069789584,
+      "grad_norm": 4.8986261164638725,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 43153
+    },
+    {
+      "epoch": 1.0271827097019899,
+      "grad_norm": 0.10761752607453487,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43154
+    },
+    {
+      "epoch": 1.0272065124250214,
+      "grad_norm": 3.0548880801236535,
+      "learning_rate": 2e-05,
+      "loss": 0.3185,
+      "step": 43155
+    },
+    {
+      "epoch": 1.0272303151480529,
+      "grad_norm": 1.006559709960544,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 43156
+    },
+    {
+      "epoch": 1.0272541178710846,
+      "grad_norm": 5.183857124021552,
+      "learning_rate": 2e-05,
+      "loss": 0.6469,
+      "step": 43157
+    },
+    {
+      "epoch": 1.027277920594116,
+      "grad_norm": 0.07054468948767895,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43158
+    },
+    {
+      "epoch": 1.0273017233171475,
+      "grad_norm": 0.04583774426570908,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43159
+    },
+    {
+      "epoch": 1.027325526040179,
+      "grad_norm": 0.24190031591818567,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 43160
+    },
+    {
+      "epoch": 1.0273493287632105,
+      "grad_norm": 1.232937583859818,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 43161
+    },
+    {
+      "epoch": 1.027373131486242,
+      "grad_norm": 4.443844985323433,
+      "learning_rate": 2e-05,
+      "loss": 0.1765,
+      "step": 43162
+    },
+    {
+      "epoch": 1.0273969342092735,
+      "grad_norm": 4.071345716416923,
+      "learning_rate": 2e-05,
+      "loss": 0.4006,
+      "step": 43163
+    },
+    {
+      "epoch": 1.027420736932305,
+      "grad_norm": 17.55011408745756,
+      "learning_rate": 2e-05,
+      "loss": 0.5202,
+      "step": 43164
+    },
+    {
+      "epoch": 1.0274445396553367,
+      "grad_norm": 3.0127159277248063,
+      "learning_rate": 2e-05,
+      "loss": 0.2087,
+      "step": 43165
+    },
+    {
+      "epoch": 1.0274683423783681,
+      "grad_norm": 4.320571610497883,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 43166
+    },
+    {
+      "epoch": 1.0274921451013996,
+      "grad_norm": 4.770792228464618,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 43167
+    },
+    {
+      "epoch": 1.0275159478244311,
+      "grad_norm": 2.391252191955057,
+      "learning_rate": 2e-05,
+      "loss": 0.0254,
+      "step": 43168
+    },
+    {
+      "epoch": 1.0275397505474626,
+      "grad_norm": 12.299859326554163,
+      "learning_rate": 2e-05,
+      "loss": 1.0291,
+      "step": 43169
+    },
+    {
+      "epoch": 1.027563553270494,
+      "grad_norm": 0.0879358806245918,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43170
+    },
+    {
+      "epoch": 1.0275873559935256,
+      "grad_norm": 4.5893757191939955,
+      "learning_rate": 2e-05,
+      "loss": 0.6325,
+      "step": 43171
+    },
+    {
+      "epoch": 1.0276111587165573,
+      "grad_norm": 4.67449581508503,
+      "learning_rate": 2e-05,
+      "loss": 0.1595,
+      "step": 43172
+    },
+    {
+      "epoch": 1.0276349614395888,
+      "grad_norm": 2.116774948945317,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 43173
+    },
+    {
+      "epoch": 1.0276587641626203,
+      "grad_norm": 2.6291839020903267,
+      "learning_rate": 2e-05,
+      "loss": 0.1839,
+      "step": 43174
+    },
+    {
+      "epoch": 1.0276825668856517,
+      "grad_norm": 0.007221895610189336,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43175
+    },
+    {
+      "epoch": 1.0277063696086832,
+      "grad_norm": 2.8948617551838502,
+      "learning_rate": 2e-05,
+      "loss": 0.0563,
+      "step": 43176
+    },
+    {
+      "epoch": 1.0277301723317147,
+      "grad_norm": 0.0055892057922440515,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43177
+    },
+    {
+      "epoch": 1.0277539750547462,
+      "grad_norm": 6.335573196068212,
+      "learning_rate": 2e-05,
+      "loss": 0.1251,
+      "step": 43178
+    },
+    {
+      "epoch": 1.0277777777777777,
+      "grad_norm": 4.009331890705462,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 43179
+    },
+    {
+      "epoch": 1.0278015805008094,
+      "grad_norm": 0.0032163253962059334,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43180
+    },
+    {
+      "epoch": 1.0278253832238409,
+      "grad_norm": 29.693349936524523,
+      "learning_rate": 2e-05,
+      "loss": 0.2986,
+      "step": 43181
+    },
+    {
+      "epoch": 1.0278491859468724,
+      "grad_norm": 2.4401089480878544,
+      "learning_rate": 2e-05,
+      "loss": 0.057,
+      "step": 43182
+    },
+    {
+      "epoch": 1.0278729886699038,
+      "grad_norm": 3.8622845779697523,
+      "learning_rate": 2e-05,
+      "loss": 0.6081,
+      "step": 43183
+    },
+    {
+      "epoch": 1.0278967913929353,
+      "grad_norm": 0.08343047454731245,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43184
+    },
+    {
+      "epoch": 1.0279205941159668,
+      "grad_norm": 14.244546123713654,
+      "learning_rate": 2e-05,
+      "loss": 0.5421,
+      "step": 43185
+    },
+    {
+      "epoch": 1.0279443968389983,
+      "grad_norm": 14.116285228028227,
+      "learning_rate": 2e-05,
+      "loss": 0.7765,
+      "step": 43186
+    },
+    {
+      "epoch": 1.02796819956203,
+      "grad_norm": 5.126917483516366,
+      "learning_rate": 2e-05,
+      "loss": 0.5945,
+      "step": 43187
+    },
+    {
+      "epoch": 1.0279920022850615,
+      "grad_norm": 2.719535578301123,
+      "learning_rate": 2e-05,
+      "loss": 0.309,
+      "step": 43188
+    },
+    {
+      "epoch": 1.028015805008093,
+      "grad_norm": 5.617422545029019,
+      "learning_rate": 2e-05,
+      "loss": 0.8426,
+      "step": 43189
+    },
+    {
+      "epoch": 1.0280396077311245,
+      "grad_norm": 2.0892676729993664,
+      "learning_rate": 2e-05,
+      "loss": 0.0334,
+      "step": 43190
+    },
+    {
+      "epoch": 1.028063410454156,
+      "grad_norm": 2.5895771783255253,
+      "learning_rate": 2e-05,
+      "loss": 0.3545,
+      "step": 43191
+    },
+    {
+      "epoch": 1.0280872131771874,
+      "grad_norm": 2.756905755862824,
+      "learning_rate": 2e-05,
+      "loss": 0.2257,
+      "step": 43192
+    },
+    {
+      "epoch": 1.028111015900219,
+      "grad_norm": 7.52612564363187,
+      "learning_rate": 2e-05,
+      "loss": 1.1126,
+      "step": 43193
+    },
+    {
+      "epoch": 1.0281348186232504,
+      "grad_norm": 5.616970057319963,
+      "learning_rate": 2e-05,
+      "loss": 0.7286,
+      "step": 43194
+    },
+    {
+      "epoch": 1.028158621346282,
+      "grad_norm": 1.7480294277638793,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 43195
+    },
+    {
+      "epoch": 1.0281824240693136,
+      "grad_norm": 0.011441472582020646,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43196
+    },
+    {
+      "epoch": 1.028206226792345,
+      "grad_norm": 0.017204982283180542,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43197
+    },
+    {
+      "epoch": 1.0282300295153766,
+      "grad_norm": 0.41418048947134306,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 43198
+    },
+    {
+      "epoch": 1.028253832238408,
+      "grad_norm": 0.30307866490332136,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 43199
+    },
+    {
+      "epoch": 1.0282776349614395,
+      "grad_norm": 1.1045806762914074,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 43200
+    },
+    {
+      "epoch": 1.028301437684471,
+      "grad_norm": 3.635899898651005,
+      "learning_rate": 2e-05,
+      "loss": 0.339,
+      "step": 43201
+    },
+    {
+      "epoch": 1.0283252404075027,
+      "grad_norm": 4.38144064394012,
+      "learning_rate": 2e-05,
+      "loss": 0.2528,
+      "step": 43202
+    },
+    {
+      "epoch": 1.0283490431305342,
+      "grad_norm": 2.192170829688753,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 43203
+    },
+    {
+      "epoch": 1.0283728458535657,
+      "grad_norm": 0.07254025347240393,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43204
+    },
+    {
+      "epoch": 1.0283966485765972,
+      "grad_norm": 5.085252970336673,
+      "learning_rate": 2e-05,
+      "loss": 0.8052,
+      "step": 43205
+    },
+    {
+      "epoch": 1.0284204512996287,
+      "grad_norm": 1.214470083018713,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 43206
+    },
+    {
+      "epoch": 1.0284442540226602,
+      "grad_norm": 0.03129048612024469,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43207
+    },
+    {
+      "epoch": 1.0284680567456916,
+      "grad_norm": 6.680009277353015,
+      "learning_rate": 2e-05,
+      "loss": 1.2141,
+      "step": 43208
+    },
+    {
+      "epoch": 1.0284918594687231,
+      "grad_norm": 3.9548038664297347,
+      "learning_rate": 2e-05,
+      "loss": 0.6604,
+      "step": 43209
+    },
+    {
+      "epoch": 1.0285156621917548,
+      "grad_norm": 3.7202745996644446,
+      "learning_rate": 2e-05,
+      "loss": 0.215,
+      "step": 43210
+    },
+    {
+      "epoch": 1.0285394649147863,
+      "grad_norm": 0.5169398301800158,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 43211
+    },
+    {
+      "epoch": 1.0285632676378178,
+      "grad_norm": 7.218151480188208,
+      "learning_rate": 2e-05,
+      "loss": 1.2514,
+      "step": 43212
+    },
+    {
+      "epoch": 1.0285870703608493,
+      "grad_norm": 3.808694550725293,
+      "learning_rate": 2e-05,
+      "loss": 0.2898,
+      "step": 43213
+    },
+    {
+      "epoch": 1.0286108730838808,
+      "grad_norm": 2.0357552457647565,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 43214
+    },
+    {
+      "epoch": 1.0286346758069123,
+      "grad_norm": 3.781072905678405,
+      "learning_rate": 2e-05,
+      "loss": 0.5698,
+      "step": 43215
+    },
+    {
+      "epoch": 1.0286584785299437,
+      "grad_norm": 5.072267103430771,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 43216
+    },
+    {
+      "epoch": 1.0286822812529755,
+      "grad_norm": 0.07037171180382157,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43217
+    },
+    {
+      "epoch": 1.028706083976007,
+      "grad_norm": 0.010766573438562438,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43218
+    },
+    {
+      "epoch": 1.0287298866990384,
+      "grad_norm": 7.283515325027115,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 43219
+    },
+    {
+      "epoch": 1.02875368942207,
+      "grad_norm": 2.0128136952510705,
+      "learning_rate": 2e-05,
+      "loss": 0.0388,
+      "step": 43220
+    },
+    {
+      "epoch": 1.0287774921451014,
+      "grad_norm": 0.5545173425700869,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 43221
+    },
+    {
+      "epoch": 1.0288012948681329,
+      "grad_norm": 6.076527503835938,
+      "learning_rate": 2e-05,
+      "loss": 0.1048,
+      "step": 43222
+    },
+    {
+      "epoch": 1.0288250975911644,
+      "grad_norm": 6.6888702864235805,
+      "learning_rate": 2e-05,
+      "loss": 0.1896,
+      "step": 43223
+    },
+    {
+      "epoch": 1.0288489003141958,
+      "grad_norm": 0.02844030726271224,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43224
+    },
+    {
+      "epoch": 1.0288727030372276,
+      "grad_norm": 3.8103771832313225,
+      "learning_rate": 2e-05,
+      "loss": 0.5198,
+      "step": 43225
+    },
+    {
+      "epoch": 1.028896505760259,
+      "grad_norm": 3.081022313967263,
+      "learning_rate": 2e-05,
+      "loss": 0.197,
+      "step": 43226
+    },
+    {
+      "epoch": 1.0289203084832905,
+      "grad_norm": 4.6567041985879,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 43227
+    },
+    {
+      "epoch": 1.028944111206322,
+      "grad_norm": 1.0237149465339108,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 43228
+    },
+    {
+      "epoch": 1.0289679139293535,
+      "grad_norm": 4.480637509350244,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 43229
+    },
+    {
+      "epoch": 1.028991716652385,
+      "grad_norm": 0.07160179873342826,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43230
+    },
+    {
+      "epoch": 1.0290155193754165,
+      "grad_norm": 10.374083112651716,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 43231
+    },
+    {
+      "epoch": 1.029039322098448,
+      "grad_norm": 0.009305192474806651,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43232
+    },
+    {
+      "epoch": 1.0290631248214797,
+      "grad_norm": 3.9095130453311566,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 43233
+    },
+    {
+      "epoch": 1.0290869275445111,
+      "grad_norm": 5.532293300760162,
+      "learning_rate": 2e-05,
+      "loss": 0.8134,
+      "step": 43234
+    },
+    {
+      "epoch": 1.0291107302675426,
+      "grad_norm": 0.017500626387452492,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43235
+    },
+    {
+      "epoch": 1.0291345329905741,
+      "grad_norm": 3.732651801872157,
+      "learning_rate": 2e-05,
+      "loss": 0.3453,
+      "step": 43236
+    },
+    {
+      "epoch": 1.0291583357136056,
+      "grad_norm": 7.341007380931179,
+      "learning_rate": 2e-05,
+      "loss": 0.2342,
+      "step": 43237
+    },
+    {
+      "epoch": 1.029182138436637,
+      "grad_norm": 3.756744957696931,
+      "learning_rate": 2e-05,
+      "loss": 0.3003,
+      "step": 43238
+    },
+    {
+      "epoch": 1.0292059411596686,
+      "grad_norm": 3.611279554366291,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 43239
+    },
+    {
+      "epoch": 1.0292297438827003,
+      "grad_norm": 0.3017210014466295,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 43240
+    },
+    {
+      "epoch": 1.0292535466057318,
+      "grad_norm": 0.029783599484811435,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43241
+    },
+    {
+      "epoch": 1.0292773493287632,
+      "grad_norm": 0.039537381892041784,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43242
+    },
+    {
+      "epoch": 1.0293011520517947,
+      "grad_norm": 0.18355608729903453,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43243
+    },
+    {
+      "epoch": 1.0293249547748262,
+      "grad_norm": 2.888496983414523,
+      "learning_rate": 2e-05,
+      "loss": 0.0908,
+      "step": 43244
+    },
+    {
+      "epoch": 1.0293487574978577,
+      "grad_norm": 4.621297519937242,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 43245
+    },
+    {
+      "epoch": 1.0293725602208892,
+      "grad_norm": 2.640786644534251,
+      "learning_rate": 2e-05,
+      "loss": 0.288,
+      "step": 43246
+    },
+    {
+      "epoch": 1.029396362943921,
+      "grad_norm": 3.4995064290005296,
+      "learning_rate": 2e-05,
+      "loss": 0.4516,
+      "step": 43247
+    },
+    {
+      "epoch": 1.0294201656669524,
+      "grad_norm": 0.3751782462223346,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 43248
+    },
+    {
+      "epoch": 1.0294439683899839,
+      "grad_norm": 5.529886401921196,
+      "learning_rate": 2e-05,
+      "loss": 0.4986,
+      "step": 43249
+    },
+    {
+      "epoch": 1.0294677711130154,
+      "grad_norm": 0.08674341654201162,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43250
+    },
+    {
+      "epoch": 1.0294915738360468,
+      "grad_norm": 6.370117691327759,
+      "learning_rate": 2e-05,
+      "loss": 0.6404,
+      "step": 43251
+    },
+    {
+      "epoch": 1.0295153765590783,
+      "grad_norm": 0.2787782714599428,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 43252
+    },
+    {
+      "epoch": 1.0295391792821098,
+      "grad_norm": 7.548695347224952,
+      "learning_rate": 2e-05,
+      "loss": 0.3005,
+      "step": 43253
+    },
+    {
+      "epoch": 1.0295629820051413,
+      "grad_norm": 8.150655559211422,
+      "learning_rate": 2e-05,
+      "loss": 0.7228,
+      "step": 43254
+    },
+    {
+      "epoch": 1.029586784728173,
+      "grad_norm": 3.130902803889142,
+      "learning_rate": 2e-05,
+      "loss": 0.2276,
+      "step": 43255
+    },
+    {
+      "epoch": 1.0296105874512045,
+      "grad_norm": 3.500637156648554,
+      "learning_rate": 2e-05,
+      "loss": 0.3195,
+      "step": 43256
+    },
+    {
+      "epoch": 1.029634390174236,
+      "grad_norm": 9.269633881189804,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 43257
+    },
+    {
+      "epoch": 1.0296581928972675,
+      "grad_norm": 2.791792222464034,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 43258
+    },
+    {
+      "epoch": 1.029681995620299,
+      "grad_norm": 2.6932165712801783,
+      "learning_rate": 2e-05,
+      "loss": 0.304,
+      "step": 43259
+    },
+    {
+      "epoch": 1.0297057983433304,
+      "grad_norm": 0.4051974181937567,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 43260
+    },
+    {
+      "epoch": 1.029729601066362,
+      "grad_norm": 1.2560559278162267,
+      "learning_rate": 2e-05,
+      "loss": 0.0321,
+      "step": 43261
+    },
+    {
+      "epoch": 1.0297534037893934,
+      "grad_norm": 2.512266667114301,
+      "learning_rate": 2e-05,
+      "loss": 0.0863,
+      "step": 43262
+    },
+    {
+      "epoch": 1.029777206512425,
+      "grad_norm": 11.223190043855121,
+      "learning_rate": 2e-05,
+      "loss": 0.7719,
+      "step": 43263
+    },
+    {
+      "epoch": 1.0298010092354566,
+      "grad_norm": 6.537743720929478,
+      "learning_rate": 2e-05,
+      "loss": 0.5264,
+      "step": 43264
+    },
+    {
+      "epoch": 1.029824811958488,
+      "grad_norm": 5.85637142512227,
+      "learning_rate": 2e-05,
+      "loss": 0.776,
+      "step": 43265
+    },
+    {
+      "epoch": 1.0298486146815196,
+      "grad_norm": 0.016007406766063412,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43266
+    },
+    {
+      "epoch": 1.029872417404551,
+      "grad_norm": 3.472505419815917,
+      "learning_rate": 2e-05,
+      "loss": 0.6813,
+      "step": 43267
+    },
+    {
+      "epoch": 1.0298962201275825,
+      "grad_norm": 0.05051610259990304,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43268
+    },
+    {
+      "epoch": 1.029920022850614,
+      "grad_norm": 5.49710886030627,
+      "learning_rate": 2e-05,
+      "loss": 0.3706,
+      "step": 43269
+    },
+    {
+      "epoch": 1.0299438255736457,
+      "grad_norm": 12.996941920425702,
+      "learning_rate": 2e-05,
+      "loss": 0.7241,
+      "step": 43270
+    },
+    {
+      "epoch": 1.0299676282966772,
+      "grad_norm": 0.021656719355062568,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43271
+    },
+    {
+      "epoch": 1.0299914310197087,
+      "grad_norm": 0.16619199066371723,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 43272
+    },
+    {
+      "epoch": 1.0300152337427402,
+      "grad_norm": 1.6504195845696712,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 43273
+    },
+    {
+      "epoch": 1.0300390364657717,
+      "grad_norm": 6.479974123947015,
+      "learning_rate": 2e-05,
+      "loss": 0.1391,
+      "step": 43274
+    },
+    {
+      "epoch": 1.0300628391888031,
+      "grad_norm": 0.3962176204125003,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 43275
+    },
+    {
+      "epoch": 1.0300866419118346,
+      "grad_norm": 2.276088042114369,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 43276
+    },
+    {
+      "epoch": 1.0301104446348663,
+      "grad_norm": 5.34278560064223,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 43277
+    },
+    {
+      "epoch": 1.0301342473578978,
+      "grad_norm": 8.57276322491858,
+      "learning_rate": 2e-05,
+      "loss": 0.3972,
+      "step": 43278
+    },
+    {
+      "epoch": 1.0301580500809293,
+      "grad_norm": 3.4061786640645475,
+      "learning_rate": 2e-05,
+      "loss": 0.1309,
+      "step": 43279
+    },
+    {
+      "epoch": 1.0301818528039608,
+      "grad_norm": 6.61026949339364,
+      "learning_rate": 2e-05,
+      "loss": 0.8694,
+      "step": 43280
+    },
+    {
+      "epoch": 1.0302056555269923,
+      "grad_norm": 6.610537506008429,
+      "learning_rate": 2e-05,
+      "loss": 0.5727,
+      "step": 43281
+    },
+    {
+      "epoch": 1.0302294582500238,
+      "grad_norm": 3.6523913792466804,
+      "learning_rate": 2e-05,
+      "loss": 0.4455,
+      "step": 43282
+    },
+    {
+      "epoch": 1.0302532609730553,
+      "grad_norm": 2.56804388702847,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 43283
+    },
+    {
+      "epoch": 1.0302770636960867,
+      "grad_norm": 0.3241043461263926,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 43284
+    },
+    {
+      "epoch": 1.0303008664191184,
+      "grad_norm": 3.4266355708919973,
+      "learning_rate": 2e-05,
+      "loss": 0.1969,
+      "step": 43285
+    },
+    {
+      "epoch": 1.03032466914215,
+      "grad_norm": 5.290909120397192,
+      "learning_rate": 2e-05,
+      "loss": 0.8293,
+      "step": 43286
+    },
+    {
+      "epoch": 1.0303484718651814,
+      "grad_norm": 4.711439743570349,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 43287
+    },
+    {
+      "epoch": 1.030372274588213,
+      "grad_norm": 3.6916637875321285,
+      "learning_rate": 2e-05,
+      "loss": 0.2657,
+      "step": 43288
+    },
+    {
+      "epoch": 1.0303960773112444,
+      "grad_norm": 0.04464702810525536,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43289
+    },
+    {
+      "epoch": 1.0304198800342759,
+      "grad_norm": 18.54100438683457,
+      "learning_rate": 2e-05,
+      "loss": 0.3898,
+      "step": 43290
+    },
+    {
+      "epoch": 1.0304436827573074,
+      "grad_norm": 2.1511482929853987,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 43291
+    },
+    {
+      "epoch": 1.0304674854803388,
+      "grad_norm": 1.0517260707229905,
+      "learning_rate": 2e-05,
+      "loss": 0.0106,
+      "step": 43292
+    },
+    {
+      "epoch": 1.0304912882033705,
+      "grad_norm": 10.574254623253177,
+      "learning_rate": 2e-05,
+      "loss": 0.8701,
+      "step": 43293
+    },
+    {
+      "epoch": 1.030515090926402,
+      "grad_norm": 0.9483782893269175,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 43294
+    },
+    {
+      "epoch": 1.0305388936494335,
+      "grad_norm": 0.013369815338343466,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43295
+    },
+    {
+      "epoch": 1.030562696372465,
+      "grad_norm": 5.554961199235479,
+      "learning_rate": 2e-05,
+      "loss": 0.7076,
+      "step": 43296
+    },
+    {
+      "epoch": 1.0305864990954965,
+      "grad_norm": 0.31391790018043436,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 43297
+    },
+    {
+      "epoch": 1.030610301818528,
+      "grad_norm": 0.0231027012049399,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43298
+    },
+    {
+      "epoch": 1.0306341045415595,
+      "grad_norm": 3.6039041071527196,
+      "learning_rate": 2e-05,
+      "loss": 0.4136,
+      "step": 43299
+    },
+    {
+      "epoch": 1.0306579072645912,
+      "grad_norm": 4.906709589973241,
+      "learning_rate": 2e-05,
+      "loss": 0.1001,
+      "step": 43300
+    },
+    {
+      "epoch": 1.0306817099876227,
+      "grad_norm": 0.140929713424411,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43301
+    },
+    {
+      "epoch": 1.0307055127106541,
+      "grad_norm": 6.507863325559236,
+      "learning_rate": 2e-05,
+      "loss": 0.4545,
+      "step": 43302
+    },
+    {
+      "epoch": 1.0307293154336856,
+      "grad_norm": 4.898313084945225,
+      "learning_rate": 2e-05,
+      "loss": 0.7555,
+      "step": 43303
+    },
+    {
+      "epoch": 1.030753118156717,
+      "grad_norm": 2.6353470948792053,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 43304
+    },
+    {
+      "epoch": 1.0307769208797486,
+      "grad_norm": 3.3406021997090103,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 43305
+    },
+    {
+      "epoch": 1.03080072360278,
+      "grad_norm": 5.330857566165081,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 43306
+    },
+    {
+      "epoch": 1.0308245263258118,
+      "grad_norm": 2.0468473738731667,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 43307
+    },
+    {
+      "epoch": 1.0308483290488433,
+      "grad_norm": 4.943094922128169,
+      "learning_rate": 2e-05,
+      "loss": 0.8975,
+      "step": 43308
+    },
+    {
+      "epoch": 1.0308721317718748,
+      "grad_norm": 4.089993689497961,
+      "learning_rate": 2e-05,
+      "loss": 0.5161,
+      "step": 43309
+    },
+    {
+      "epoch": 1.0308959344949062,
+      "grad_norm": 0.14608647572273062,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43310
+    },
+    {
+      "epoch": 1.0309197372179377,
+      "grad_norm": 3.4890822072156453,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 43311
+    },
+    {
+      "epoch": 1.0309435399409692,
+      "grad_norm": 2.696519679192939,
+      "learning_rate": 2e-05,
+      "loss": 0.303,
+      "step": 43312
+    },
+    {
+      "epoch": 1.0309673426640007,
+      "grad_norm": 6.563052231040611,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 43313
+    },
+    {
+      "epoch": 1.0309911453870322,
+      "grad_norm": 5.192088366668486,
+      "learning_rate": 2e-05,
+      "loss": 0.5348,
+      "step": 43314
+    },
+    {
+      "epoch": 1.031014948110064,
+      "grad_norm": 11.014440363182983,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 43315
+    },
+    {
+      "epoch": 1.0310387508330954,
+      "grad_norm": 5.448001893174172,
+      "learning_rate": 2e-05,
+      "loss": 0.4797,
+      "step": 43316
+    },
+    {
+      "epoch": 1.0310625535561269,
+      "grad_norm": 4.2919782873803545,
+      "learning_rate": 2e-05,
+      "loss": 0.6769,
+      "step": 43317
+    },
+    {
+      "epoch": 1.0310863562791583,
+      "grad_norm": 3.939633727630899,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 43318
+    },
+    {
+      "epoch": 1.0311101590021898,
+      "grad_norm": 3.093523570581147,
+      "learning_rate": 2e-05,
+      "loss": 0.1512,
+      "step": 43319
+    },
+    {
+      "epoch": 1.0311339617252213,
+      "grad_norm": 6.151594386032852,
+      "learning_rate": 2e-05,
+      "loss": 0.4386,
+      "step": 43320
+    },
+    {
+      "epoch": 1.0311577644482528,
+      "grad_norm": 9.564404638911654,
+      "learning_rate": 2e-05,
+      "loss": 0.8677,
+      "step": 43321
+    },
+    {
+      "epoch": 1.0311815671712843,
+      "grad_norm": 2.403122145778574,
+      "learning_rate": 2e-05,
+      "loss": 0.048,
+      "step": 43322
+    },
+    {
+      "epoch": 1.031205369894316,
+      "grad_norm": 5.7950376787063576,
+      "learning_rate": 2e-05,
+      "loss": 0.382,
+      "step": 43323
+    },
+    {
+      "epoch": 1.0312291726173475,
+      "grad_norm": 4.826009782810488,
+      "learning_rate": 2e-05,
+      "loss": 0.3384,
+      "step": 43324
+    },
+    {
+      "epoch": 1.031252975340379,
+      "grad_norm": 1.356197332803698,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 43325
+    },
+    {
+      "epoch": 1.0312767780634104,
+      "grad_norm": 4.443051545427368,
+      "learning_rate": 2e-05,
+      "loss": 0.4217,
+      "step": 43326
+    },
+    {
+      "epoch": 1.031300580786442,
+      "grad_norm": 4.498634496499665,
+      "learning_rate": 2e-05,
+      "loss": 0.6532,
+      "step": 43327
+    },
+    {
+      "epoch": 1.0313243835094734,
+      "grad_norm": 5.165863111441997,
+      "learning_rate": 2e-05,
+      "loss": 0.7603,
+      "step": 43328
+    },
+    {
+      "epoch": 1.031348186232505,
+      "grad_norm": 4.771432420389172,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 43329
+    },
+    {
+      "epoch": 1.0313719889555366,
+      "grad_norm": 4.543649247359767,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 43330
+    },
+    {
+      "epoch": 1.031395791678568,
+      "grad_norm": 9.939476537838084,
+      "learning_rate": 2e-05,
+      "loss": 0.4398,
+      "step": 43331
+    },
+    {
+      "epoch": 1.0314195944015996,
+      "grad_norm": 3.5615178192129173,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 43332
+    },
+    {
+      "epoch": 1.031443397124631,
+      "grad_norm": 3.4407644910121538,
+      "learning_rate": 2e-05,
+      "loss": 0.3243,
+      "step": 43333
+    },
+    {
+      "epoch": 1.0314671998476626,
+      "grad_norm": 2.370849759761432,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 43334
+    },
+    {
+      "epoch": 1.031491002570694,
+      "grad_norm": 5.94061576451102,
+      "learning_rate": 2e-05,
+      "loss": 0.4138,
+      "step": 43335
+    },
+    {
+      "epoch": 1.0315148052937255,
+      "grad_norm": 3.591054981778054,
+      "learning_rate": 2e-05,
+      "loss": 0.5472,
+      "step": 43336
+    },
+    {
+      "epoch": 1.0315386080167572,
+      "grad_norm": 0.07481184489296842,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43337
+    },
+    {
+      "epoch": 1.0315624107397887,
+      "grad_norm": 5.404518176050921,
+      "learning_rate": 2e-05,
+      "loss": 1.01,
+      "step": 43338
+    },
+    {
+      "epoch": 1.0315862134628202,
+      "grad_norm": 14.72190475367258,
+      "learning_rate": 2e-05,
+      "loss": 0.2459,
+      "step": 43339
+    },
+    {
+      "epoch": 1.0316100161858517,
+      "grad_norm": 0.38027696889590434,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 43340
+    },
+    {
+      "epoch": 1.0316338189088832,
+      "grad_norm": 2.1487637516347355,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 43341
+    },
+    {
+      "epoch": 1.0316576216319147,
+      "grad_norm": 1.070437526490214,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 43342
+    },
+    {
+      "epoch": 1.0316814243549461,
+      "grad_norm": 0.1191636067248589,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43343
+    },
+    {
+      "epoch": 1.0317052270779776,
+      "grad_norm": 2.585611034642385,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 43344
+    },
+    {
+      "epoch": 1.0317290298010093,
+      "grad_norm": 2.4628710490645256,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 43345
+    },
+    {
+      "epoch": 1.0317528325240408,
+      "grad_norm": 0.6016855812063018,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 43346
+    },
+    {
+      "epoch": 1.0317766352470723,
+      "grad_norm": 0.1565340002837409,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 43347
+    },
+    {
+      "epoch": 1.0318004379701038,
+      "grad_norm": 0.13505428093337052,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 43348
+    },
+    {
+      "epoch": 1.0318242406931353,
+      "grad_norm": 3.140221241456137,
+      "learning_rate": 2e-05,
+      "loss": 0.2304,
+      "step": 43349
+    },
+    {
+      "epoch": 1.0318480434161668,
+      "grad_norm": 2.457729727420002,
+      "learning_rate": 2e-05,
+      "loss": 0.0307,
+      "step": 43350
+    },
+    {
+      "epoch": 1.0318718461391982,
+      "grad_norm": 0.109233512929428,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 43351
+    },
+    {
+      "epoch": 1.0318956488622297,
+      "grad_norm": 3.4187797009456355,
+      "learning_rate": 2e-05,
+      "loss": 0.3911,
+      "step": 43352
+    },
+    {
+      "epoch": 1.0319194515852614,
+      "grad_norm": 0.3023438288165618,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 43353
+    },
+    {
+      "epoch": 1.031943254308293,
+      "grad_norm": 1.73205089058181,
+      "learning_rate": 2e-05,
+      "loss": 0.0968,
+      "step": 43354
+    },
+    {
+      "epoch": 1.0319670570313244,
+      "grad_norm": 0.24468787812741866,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 43355
+    },
+    {
+      "epoch": 1.031990859754356,
+      "grad_norm": 3.0976627083394135,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 43356
+    },
+    {
+      "epoch": 1.0320146624773874,
+      "grad_norm": 4.262055838601047,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 43357
+    },
+    {
+      "epoch": 1.0320384652004189,
+      "grad_norm": 15.595256258194727,
+      "learning_rate": 2e-05,
+      "loss": 0.4561,
+      "step": 43358
+    },
+    {
+      "epoch": 1.0320622679234503,
+      "grad_norm": 2.904675162316892,
+      "learning_rate": 2e-05,
+      "loss": 0.291,
+      "step": 43359
+    },
+    {
+      "epoch": 1.032086070646482,
+      "grad_norm": 4.582231754852637,
+      "learning_rate": 2e-05,
+      "loss": 0.455,
+      "step": 43360
+    },
+    {
+      "epoch": 1.0321098733695135,
+      "grad_norm": 9.381883093608224,
+      "learning_rate": 2e-05,
+      "loss": 0.6919,
+      "step": 43361
+    },
+    {
+      "epoch": 1.032133676092545,
+      "grad_norm": 2.270335766268709,
+      "learning_rate": 2e-05,
+      "loss": 0.3527,
+      "step": 43362
+    },
+    {
+      "epoch": 1.0321574788155765,
+      "grad_norm": 0.5638374069747614,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 43363
+    },
+    {
+      "epoch": 1.032181281538608,
+      "grad_norm": 3.3767495388113367,
+      "learning_rate": 2e-05,
+      "loss": 0.2026,
+      "step": 43364
+    },
+    {
+      "epoch": 1.0322050842616395,
+      "grad_norm": 0.017081309986118655,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43365
+    },
+    {
+      "epoch": 1.032228886984671,
+      "grad_norm": 6.478862629571245,
+      "learning_rate": 2e-05,
+      "loss": 0.7457,
+      "step": 43366
+    },
+    {
+      "epoch": 1.0322526897077027,
+      "grad_norm": 7.009634205129026,
+      "learning_rate": 2e-05,
+      "loss": 1.1607,
+      "step": 43367
+    },
+    {
+      "epoch": 1.0322764924307342,
+      "grad_norm": 0.05760492574458356,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43368
+    },
+    {
+      "epoch": 1.0323002951537656,
+      "grad_norm": 1.4918787200966017,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 43369
+    },
+    {
+      "epoch": 1.0323240978767971,
+      "grad_norm": 0.24903330067895185,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 43370
+    },
+    {
+      "epoch": 1.0323479005998286,
+      "grad_norm": 4.150394801609643,
+      "learning_rate": 2e-05,
+      "loss": 0.0635,
+      "step": 43371
+    },
+    {
+      "epoch": 1.03237170332286,
+      "grad_norm": 2.6368540398369342,
+      "learning_rate": 2e-05,
+      "loss": 0.3225,
+      "step": 43372
+    },
+    {
+      "epoch": 1.0323955060458916,
+      "grad_norm": 13.734673713323888,
+      "learning_rate": 2e-05,
+      "loss": 0.5993,
+      "step": 43373
+    },
+    {
+      "epoch": 1.032419308768923,
+      "grad_norm": 3.3540283286478574,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 43374
+    },
+    {
+      "epoch": 1.0324431114919548,
+      "grad_norm": 0.13619708320283153,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43375
+    },
+    {
+      "epoch": 1.0324669142149863,
+      "grad_norm": 2.6288800742864202,
+      "learning_rate": 2e-05,
+      "loss": 0.256,
+      "step": 43376
+    },
+    {
+      "epoch": 1.0324907169380177,
+      "grad_norm": 9.192051962117661,
+      "learning_rate": 2e-05,
+      "loss": 0.5282,
+      "step": 43377
+    },
+    {
+      "epoch": 1.0325145196610492,
+      "grad_norm": 0.4828604299838061,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 43378
+    },
+    {
+      "epoch": 1.0325383223840807,
+      "grad_norm": 0.4602507827604215,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 43379
+    },
+    {
+      "epoch": 1.0325621251071122,
+      "grad_norm": 0.2180351406190717,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 43380
+    },
+    {
+      "epoch": 1.0325859278301437,
+      "grad_norm": 3.293371468305127,
+      "learning_rate": 2e-05,
+      "loss": 0.3295,
+      "step": 43381
+    },
+    {
+      "epoch": 1.0326097305531752,
+      "grad_norm": 3.035215069171534,
+      "learning_rate": 2e-05,
+      "loss": 0.3264,
+      "step": 43382
+    },
+    {
+      "epoch": 1.0326335332762069,
+      "grad_norm": 4.044494830064227,
+      "learning_rate": 2e-05,
+      "loss": 0.088,
+      "step": 43383
+    },
+    {
+      "epoch": 1.0326573359992384,
+      "grad_norm": 8.476986738193693,
+      "learning_rate": 2e-05,
+      "loss": 0.183,
+      "step": 43384
+    },
+    {
+      "epoch": 1.0326811387222699,
+      "grad_norm": 0.28898869248875186,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 43385
+    },
+    {
+      "epoch": 1.0327049414453013,
+      "grad_norm": 0.5172915161704384,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 43386
+    },
+    {
+      "epoch": 1.0327287441683328,
+      "grad_norm": 4.271683763568879,
+      "learning_rate": 2e-05,
+      "loss": 0.1292,
+      "step": 43387
+    },
+    {
+      "epoch": 1.0327525468913643,
+      "grad_norm": 11.937987709854417,
+      "learning_rate": 2e-05,
+      "loss": 0.1018,
+      "step": 43388
+    },
+    {
+      "epoch": 1.0327763496143958,
+      "grad_norm": 2.488343013258772,
+      "learning_rate": 2e-05,
+      "loss": 0.224,
+      "step": 43389
+    },
+    {
+      "epoch": 1.0328001523374275,
+      "grad_norm": 9.133589362428825,
+      "learning_rate": 2e-05,
+      "loss": 1.0909,
+      "step": 43390
+    },
+    {
+      "epoch": 1.032823955060459,
+      "grad_norm": 5.19014999802488,
+      "learning_rate": 2e-05,
+      "loss": 0.5267,
+      "step": 43391
+    },
+    {
+      "epoch": 1.0328477577834905,
+      "grad_norm": 7.469056848362036,
+      "learning_rate": 2e-05,
+      "loss": 0.2565,
+      "step": 43392
+    },
+    {
+      "epoch": 1.032871560506522,
+      "grad_norm": 0.039674698816020385,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43393
+    },
+    {
+      "epoch": 1.0328953632295534,
+      "grad_norm": 38.681447969247245,
+      "learning_rate": 2e-05,
+      "loss": 0.3107,
+      "step": 43394
+    },
+    {
+      "epoch": 1.032919165952585,
+      "grad_norm": 5.075509727321792,
+      "learning_rate": 2e-05,
+      "loss": 0.612,
+      "step": 43395
+    },
+    {
+      "epoch": 1.0329429686756164,
+      "grad_norm": 0.6514667982800506,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 43396
+    },
+    {
+      "epoch": 1.032966771398648,
+      "grad_norm": 1.922253411023235,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 43397
+    },
+    {
+      "epoch": 1.0329905741216796,
+      "grad_norm": 5.917708520724098,
+      "learning_rate": 2e-05,
+      "loss": 0.814,
+      "step": 43398
+    },
+    {
+      "epoch": 1.033014376844711,
+      "grad_norm": 2.0072675326709737,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 43399
+    },
+    {
+      "epoch": 1.0330381795677426,
+      "grad_norm": 0.035655076963325996,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43400
+    },
+    {
+      "epoch": 1.033061982290774,
+      "grad_norm": 1.4202882846527294,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 43401
+    },
+    {
+      "epoch": 1.0330857850138055,
+      "grad_norm": 2.505565039153893,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 43402
+    },
+    {
+      "epoch": 1.033109587736837,
+      "grad_norm": 0.13697262567256085,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 43403
+    },
+    {
+      "epoch": 1.0331333904598685,
+      "grad_norm": 3.16225966264701,
+      "learning_rate": 2e-05,
+      "loss": 0.43,
+      "step": 43404
+    },
+    {
+      "epoch": 1.0331571931829002,
+      "grad_norm": 0.12114197089103126,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43405
+    },
+    {
+      "epoch": 1.0331809959059317,
+      "grad_norm": 9.957248117141317,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 43406
+    },
+    {
+      "epoch": 1.0332047986289632,
+      "grad_norm": 5.056152225190775,
+      "learning_rate": 2e-05,
+      "loss": 0.6031,
+      "step": 43407
+    },
+    {
+      "epoch": 1.0332286013519947,
+      "grad_norm": 0.09057746072797467,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43408
+    },
+    {
+      "epoch": 1.0332524040750262,
+      "grad_norm": 1.3161259709057176,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 43409
+    },
+    {
+      "epoch": 1.0332762067980577,
+      "grad_norm": 0.05124522026380585,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43410
+    },
+    {
+      "epoch": 1.0333000095210891,
+      "grad_norm": 0.02055731959909863,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43411
+    },
+    {
+      "epoch": 1.0333238122441206,
+      "grad_norm": 6.588012405033924,
+      "learning_rate": 2e-05,
+      "loss": 0.8985,
+      "step": 43412
+    },
+    {
+      "epoch": 1.0333476149671523,
+      "grad_norm": 0.2659045218112874,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 43413
+    },
+    {
+      "epoch": 1.0333714176901838,
+      "grad_norm": 0.4140150734219279,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 43414
+    },
+    {
+      "epoch": 1.0333952204132153,
+      "grad_norm": 0.05468881217466397,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43415
+    },
+    {
+      "epoch": 1.0334190231362468,
+      "grad_norm": 0.023704033055587687,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43416
+    },
+    {
+      "epoch": 1.0334428258592783,
+      "grad_norm": 0.2384490793904644,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 43417
+    },
+    {
+      "epoch": 1.0334666285823098,
+      "grad_norm": 0.4555696042787654,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 43418
+    },
+    {
+      "epoch": 1.0334904313053412,
+      "grad_norm": 32.04887540645297,
+      "learning_rate": 2e-05,
+      "loss": 0.6567,
+      "step": 43419
+    },
+    {
+      "epoch": 1.033514234028373,
+      "grad_norm": 4.795741263291784,
+      "learning_rate": 2e-05,
+      "loss": 0.2854,
+      "step": 43420
+    },
+    {
+      "epoch": 1.0335380367514044,
+      "grad_norm": 0.07592856768515265,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43421
+    },
+    {
+      "epoch": 1.033561839474436,
+      "grad_norm": 7.66786874603464,
+      "learning_rate": 2e-05,
+      "loss": 0.2419,
+      "step": 43422
+    },
+    {
+      "epoch": 1.0335856421974674,
+      "grad_norm": 14.115185269478676,
+      "learning_rate": 2e-05,
+      "loss": 0.7218,
+      "step": 43423
+    },
+    {
+      "epoch": 1.0336094449204989,
+      "grad_norm": 0.14297017488805722,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43424
+    },
+    {
+      "epoch": 1.0336332476435304,
+      "grad_norm": 0.26127232064194456,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 43425
+    },
+    {
+      "epoch": 1.0336570503665619,
+      "grad_norm": 7.073628218150816,
+      "learning_rate": 2e-05,
+      "loss": 0.4242,
+      "step": 43426
+    },
+    {
+      "epoch": 1.0336808530895933,
+      "grad_norm": 5.87940822612689,
+      "learning_rate": 2e-05,
+      "loss": 0.5079,
+      "step": 43427
+    },
+    {
+      "epoch": 1.033704655812625,
+      "grad_norm": 0.5973540421297813,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 43428
+    },
+    {
+      "epoch": 1.0337284585356565,
+      "grad_norm": 6.140421874927942,
+      "learning_rate": 2e-05,
+      "loss": 0.5512,
+      "step": 43429
+    },
+    {
+      "epoch": 1.033752261258688,
+      "grad_norm": 0.7164418283500009,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 43430
+    },
+    {
+      "epoch": 1.0337760639817195,
+      "grad_norm": 1.2084753115059947,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 43431
+    },
+    {
+      "epoch": 1.033799866704751,
+      "grad_norm": 0.12579181516056892,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43432
+    },
+    {
+      "epoch": 1.0338236694277825,
+      "grad_norm": 0.2814991029967223,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 43433
+    },
+    {
+      "epoch": 1.033847472150814,
+      "grad_norm": 0.03829949784487461,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43434
+    },
+    {
+      "epoch": 1.0338712748738457,
+      "grad_norm": 0.4133985822323409,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 43435
+    },
+    {
+      "epoch": 1.0338950775968772,
+      "grad_norm": 14.97979862858627,
+      "learning_rate": 2e-05,
+      "loss": 0.5277,
+      "step": 43436
+    },
+    {
+      "epoch": 1.0339188803199086,
+      "grad_norm": 3.2053731835968646,
+      "learning_rate": 2e-05,
+      "loss": 0.2306,
+      "step": 43437
+    },
+    {
+      "epoch": 1.0339426830429401,
+      "grad_norm": 4.3320956183396575,
+      "learning_rate": 2e-05,
+      "loss": 0.3013,
+      "step": 43438
+    },
+    {
+      "epoch": 1.0339664857659716,
+      "grad_norm": 4.154404458305237,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 43439
+    },
+    {
+      "epoch": 1.033990288489003,
+      "grad_norm": 0.1419795072762879,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 43440
+    },
+    {
+      "epoch": 1.0340140912120346,
+      "grad_norm": 3.3833542444649978,
+      "learning_rate": 2e-05,
+      "loss": 0.0298,
+      "step": 43441
+    },
+    {
+      "epoch": 1.034037893935066,
+      "grad_norm": 5.088287310927975,
+      "learning_rate": 2e-05,
+      "loss": 0.7548,
+      "step": 43442
+    },
+    {
+      "epoch": 1.0340616966580978,
+      "grad_norm": 1.4612907699704187,
+      "learning_rate": 2e-05,
+      "loss": 0.0139,
+      "step": 43443
+    },
+    {
+      "epoch": 1.0340854993811293,
+      "grad_norm": 6.927374575837763,
+      "learning_rate": 2e-05,
+      "loss": 0.6716,
+      "step": 43444
+    },
+    {
+      "epoch": 1.0341093021041607,
+      "grad_norm": 0.07883255332066141,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43445
+    },
+    {
+      "epoch": 1.0341331048271922,
+      "grad_norm": 3.3898808406533263,
+      "learning_rate": 2e-05,
+      "loss": 0.4166,
+      "step": 43446
+    },
+    {
+      "epoch": 1.0341569075502237,
+      "grad_norm": 2.7551597026946113,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 43447
+    },
+    {
+      "epoch": 1.0341807102732552,
+      "grad_norm": 0.06449434417198033,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43448
+    },
+    {
+      "epoch": 1.0342045129962867,
+      "grad_norm": 10.063571940367979,
+      "learning_rate": 2e-05,
+      "loss": 0.7562,
+      "step": 43449
+    },
+    {
+      "epoch": 1.0342283157193184,
+      "grad_norm": 3.1737896503281613,
+      "learning_rate": 2e-05,
+      "loss": 0.4635,
+      "step": 43450
+    },
+    {
+      "epoch": 1.0342521184423499,
+      "grad_norm": 3.257078026975573,
+      "learning_rate": 2e-05,
+      "loss": 0.4772,
+      "step": 43451
+    },
+    {
+      "epoch": 1.0342759211653814,
+      "grad_norm": 25.614206704913222,
+      "learning_rate": 2e-05,
+      "loss": 0.9504,
+      "step": 43452
+    },
+    {
+      "epoch": 1.0342997238884128,
+      "grad_norm": 3.5998091621671957,
+      "learning_rate": 2e-05,
+      "loss": 0.5029,
+      "step": 43453
+    },
+    {
+      "epoch": 1.0343235266114443,
+      "grad_norm": 3.082942290036363,
+      "learning_rate": 2e-05,
+      "loss": 0.2443,
+      "step": 43454
+    },
+    {
+      "epoch": 1.0343473293344758,
+      "grad_norm": 4.894973856252363,
+      "learning_rate": 2e-05,
+      "loss": 0.3064,
+      "step": 43455
+    },
+    {
+      "epoch": 1.0343711320575073,
+      "grad_norm": 0.050239414215682766,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43456
+    },
+    {
+      "epoch": 1.0343949347805388,
+      "grad_norm": 2.268220518372666,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 43457
+    },
+    {
+      "epoch": 1.0344187375035705,
+      "grad_norm": 3.7187924807691406,
+      "learning_rate": 2e-05,
+      "loss": 0.3405,
+      "step": 43458
+    },
+    {
+      "epoch": 1.034442540226602,
+      "grad_norm": 0.1868996453322918,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 43459
+    },
+    {
+      "epoch": 1.0344663429496335,
+      "grad_norm": 6.77230208703201,
+      "learning_rate": 2e-05,
+      "loss": 0.3555,
+      "step": 43460
+    },
+    {
+      "epoch": 1.034490145672665,
+      "grad_norm": 3.2225030087951287,
+      "learning_rate": 2e-05,
+      "loss": 0.1936,
+      "step": 43461
+    },
+    {
+      "epoch": 1.0345139483956964,
+      "grad_norm": 3.9720083858004984,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 43462
+    },
+    {
+      "epoch": 1.034537751118728,
+      "grad_norm": 3.705188110050889,
+      "learning_rate": 2e-05,
+      "loss": 0.458,
+      "step": 43463
+    },
+    {
+      "epoch": 1.0345615538417594,
+      "grad_norm": 2.7300811962244076,
+      "learning_rate": 2e-05,
+      "loss": 0.275,
+      "step": 43464
+    },
+    {
+      "epoch": 1.0345853565647911,
+      "grad_norm": 0.02247552502407085,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43465
+    },
+    {
+      "epoch": 1.0346091592878226,
+      "grad_norm": 0.07629289179621017,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43466
+    },
+    {
+      "epoch": 1.034632962010854,
+      "grad_norm": 5.063223598657518,
+      "learning_rate": 2e-05,
+      "loss": 0.73,
+      "step": 43467
+    },
+    {
+      "epoch": 1.0346567647338856,
+      "grad_norm": 0.23186134424771085,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43468
+    },
+    {
+      "epoch": 1.034680567456917,
+      "grad_norm": 4.539902151731347,
+      "learning_rate": 2e-05,
+      "loss": 0.5362,
+      "step": 43469
+    },
+    {
+      "epoch": 1.0347043701799485,
+      "grad_norm": 4.660275059164263,
+      "learning_rate": 2e-05,
+      "loss": 0.064,
+      "step": 43470
+    },
+    {
+      "epoch": 1.03472817290298,
+      "grad_norm": 5.0012782859493115,
+      "learning_rate": 2e-05,
+      "loss": 0.6894,
+      "step": 43471
+    },
+    {
+      "epoch": 1.0347519756260115,
+      "grad_norm": 13.370493011329206,
+      "learning_rate": 2e-05,
+      "loss": 0.5628,
+      "step": 43472
+    },
+    {
+      "epoch": 1.0347757783490432,
+      "grad_norm": 5.473509400362069,
+      "learning_rate": 2e-05,
+      "loss": 0.0919,
+      "step": 43473
+    },
+    {
+      "epoch": 1.0347995810720747,
+      "grad_norm": 3.395006339954602,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 43474
+    },
+    {
+      "epoch": 1.0348233837951062,
+      "grad_norm": 3.686792800992287,
+      "learning_rate": 2e-05,
+      "loss": 0.1389,
+      "step": 43475
+    },
+    {
+      "epoch": 1.0348471865181377,
+      "grad_norm": 4.375828409494979,
+      "learning_rate": 2e-05,
+      "loss": 0.0751,
+      "step": 43476
+    },
+    {
+      "epoch": 1.0348709892411692,
+      "grad_norm": 1.4348967853053576,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 43477
+    },
+    {
+      "epoch": 1.0348947919642006,
+      "grad_norm": 0.2518140171234961,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 43478
+    },
+    {
+      "epoch": 1.0349185946872321,
+      "grad_norm": 2.4577611191875834,
+      "learning_rate": 2e-05,
+      "loss": 0.0485,
+      "step": 43479
+    },
+    {
+      "epoch": 1.0349423974102638,
+      "grad_norm": 0.7453741292165719,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 43480
+    },
+    {
+      "epoch": 1.0349662001332953,
+      "grad_norm": 14.508830100774091,
+      "learning_rate": 2e-05,
+      "loss": 0.4388,
+      "step": 43481
+    },
+    {
+      "epoch": 1.0349900028563268,
+      "grad_norm": 3.847021824408859,
+      "learning_rate": 2e-05,
+      "loss": 0.2846,
+      "step": 43482
+    },
+    {
+      "epoch": 1.0350138055793583,
+      "grad_norm": 2.193182771970576,
+      "learning_rate": 2e-05,
+      "loss": 0.0779,
+      "step": 43483
+    },
+    {
+      "epoch": 1.0350376083023898,
+      "grad_norm": 4.562549509264713,
+      "learning_rate": 2e-05,
+      "loss": 0.4688,
+      "step": 43484
+    },
+    {
+      "epoch": 1.0350614110254213,
+      "grad_norm": 9.832197050327323,
+      "learning_rate": 2e-05,
+      "loss": 0.1736,
+      "step": 43485
+    },
+    {
+      "epoch": 1.0350852137484527,
+      "grad_norm": 0.02428062208356818,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43486
+    },
+    {
+      "epoch": 1.0351090164714842,
+      "grad_norm": 2.9256067470008777,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 43487
+    },
+    {
+      "epoch": 1.035132819194516,
+      "grad_norm": 0.5491615342747047,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 43488
+    },
+    {
+      "epoch": 1.0351566219175474,
+      "grad_norm": 3.599078432947167,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 43489
+    },
+    {
+      "epoch": 1.035180424640579,
+      "grad_norm": 7.728700352711406,
+      "learning_rate": 2e-05,
+      "loss": 0.7961,
+      "step": 43490
+    },
+    {
+      "epoch": 1.0352042273636104,
+      "grad_norm": 5.693159307079992,
+      "learning_rate": 2e-05,
+      "loss": 0.6413,
+      "step": 43491
+    },
+    {
+      "epoch": 1.0352280300866419,
+      "grad_norm": 1.0827104302429769,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 43492
+    },
+    {
+      "epoch": 1.0352518328096734,
+      "grad_norm": 2.3377992690034652,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 43493
+    },
+    {
+      "epoch": 1.0352756355327049,
+      "grad_norm": 0.47452306579444575,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 43494
+    },
+    {
+      "epoch": 1.0352994382557366,
+      "grad_norm": 0.8079452729913659,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 43495
+    },
+    {
+      "epoch": 1.035323240978768,
+      "grad_norm": 0.12372275560676894,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 43496
+    },
+    {
+      "epoch": 1.0353470437017995,
+      "grad_norm": 1.290226970537052,
+      "learning_rate": 2e-05,
+      "loss": 0.025,
+      "step": 43497
+    },
+    {
+      "epoch": 1.035370846424831,
+      "grad_norm": 0.05299185135287244,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43498
+    },
+    {
+      "epoch": 1.0353946491478625,
+      "grad_norm": 2.8607225150009383,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 43499
+    },
+    {
+      "epoch": 1.035418451870894,
+      "grad_norm": 2.2133142506543164,
+      "learning_rate": 2e-05,
+      "loss": 0.1918,
+      "step": 43500
+    },
+    {
+      "epoch": 1.0354422545939255,
+      "grad_norm": 6.736921287708766,
+      "learning_rate": 2e-05,
+      "loss": 0.0626,
+      "step": 43501
+    },
+    {
+      "epoch": 1.035466057316957,
+      "grad_norm": 9.870619416025201,
+      "learning_rate": 2e-05,
+      "loss": 0.3624,
+      "step": 43502
+    },
+    {
+      "epoch": 1.0354898600399887,
+      "grad_norm": 0.5498562129828923,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 43503
+    },
+    {
+      "epoch": 1.0355136627630201,
+      "grad_norm": 0.4228794607111406,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 43504
+    },
+    {
+      "epoch": 1.0355374654860516,
+      "grad_norm": 10.77886680218521,
+      "learning_rate": 2e-05,
+      "loss": 0.1471,
+      "step": 43505
+    },
+    {
+      "epoch": 1.0355612682090831,
+      "grad_norm": 0.09282243953091054,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 43506
+    },
+    {
+      "epoch": 1.0355850709321146,
+      "grad_norm": 2.5806160227779875,
+      "learning_rate": 2e-05,
+      "loss": 0.296,
+      "step": 43507
+    },
+    {
+      "epoch": 1.035608873655146,
+      "grad_norm": 1.5404415778766365,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 43508
+    },
+    {
+      "epoch": 1.0356326763781776,
+      "grad_norm": 4.244485442105647,
+      "learning_rate": 2e-05,
+      "loss": 0.3774,
+      "step": 43509
+    },
+    {
+      "epoch": 1.0356564791012093,
+      "grad_norm": 6.11495629555765,
+      "learning_rate": 2e-05,
+      "loss": 0.5936,
+      "step": 43510
+    },
+    {
+      "epoch": 1.0356802818242408,
+      "grad_norm": 0.12142982424338589,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43511
+    },
+    {
+      "epoch": 1.0357040845472723,
+      "grad_norm": 9.579376221190039,
+      "learning_rate": 2e-05,
+      "loss": 1.0656,
+      "step": 43512
+    },
+    {
+      "epoch": 1.0357278872703037,
+      "grad_norm": 0.032795946639196175,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43513
+    },
+    {
+      "epoch": 1.0357516899933352,
+      "grad_norm": 0.02465789505273346,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43514
+    },
+    {
+      "epoch": 1.0357754927163667,
+      "grad_norm": 0.02438252474994515,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43515
+    },
+    {
+      "epoch": 1.0357992954393982,
+      "grad_norm": 0.8863198176128044,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 43516
+    },
+    {
+      "epoch": 1.0358230981624297,
+      "grad_norm": 0.12595414751539374,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43517
+    },
+    {
+      "epoch": 1.0358469008854614,
+      "grad_norm": 4.219825596429335,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 43518
+    },
+    {
+      "epoch": 1.0358707036084929,
+      "grad_norm": 3.017744978105864,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 43519
+    },
+    {
+      "epoch": 1.0358945063315244,
+      "grad_norm": 1.6264714136777774,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 43520
+    },
+    {
+      "epoch": 1.0359183090545558,
+      "grad_norm": 0.06787226712591873,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43521
+    },
+    {
+      "epoch": 1.0359421117775873,
+      "grad_norm": 5.163790041067031,
+      "learning_rate": 2e-05,
+      "loss": 0.3474,
+      "step": 43522
+    },
+    {
+      "epoch": 1.0359659145006188,
+      "grad_norm": 3.867910302750886,
+      "learning_rate": 2e-05,
+      "loss": 0.5946,
+      "step": 43523
+    },
+    {
+      "epoch": 1.0359897172236503,
+      "grad_norm": 0.06932924492246052,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43524
+    },
+    {
+      "epoch": 1.036013519946682,
+      "grad_norm": 0.7931912514015216,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 43525
+    },
+    {
+      "epoch": 1.0360373226697135,
+      "grad_norm": 0.03209391116991874,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43526
+    },
+    {
+      "epoch": 1.036061125392745,
+      "grad_norm": 14.730164480652146,
+      "learning_rate": 2e-05,
+      "loss": 1.0949,
+      "step": 43527
+    },
+    {
+      "epoch": 1.0360849281157765,
+      "grad_norm": 20.736955518347955,
+      "learning_rate": 2e-05,
+      "loss": 0.2954,
+      "step": 43528
+    },
+    {
+      "epoch": 1.036108730838808,
+      "grad_norm": 5.348222444555686,
+      "learning_rate": 2e-05,
+      "loss": 0.5296,
+      "step": 43529
+    },
+    {
+      "epoch": 1.0361325335618394,
+      "grad_norm": 0.9676497345954167,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 43530
+    },
+    {
+      "epoch": 1.036156336284871,
+      "grad_norm": 0.570483927334478,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 43531
+    },
+    {
+      "epoch": 1.0361801390079024,
+      "grad_norm": 6.062005755083919,
+      "learning_rate": 2e-05,
+      "loss": 0.7437,
+      "step": 43532
+    },
+    {
+      "epoch": 1.036203941730934,
+      "grad_norm": 5.492687191272652,
+      "learning_rate": 2e-05,
+      "loss": 0.6722,
+      "step": 43533
+    },
+    {
+      "epoch": 1.0362277444539656,
+      "grad_norm": 2.2234695113305527,
+      "learning_rate": 2e-05,
+      "loss": 0.2245,
+      "step": 43534
+    },
+    {
+      "epoch": 1.036251547176997,
+      "grad_norm": 7.240262591945455,
+      "learning_rate": 2e-05,
+      "loss": 0.3256,
+      "step": 43535
+    },
+    {
+      "epoch": 1.0362753499000286,
+      "grad_norm": 2.9866680346000547,
+      "learning_rate": 2e-05,
+      "loss": 0.2885,
+      "step": 43536
+    },
+    {
+      "epoch": 1.03629915262306,
+      "grad_norm": 5.020658491633772,
+      "learning_rate": 2e-05,
+      "loss": 0.7252,
+      "step": 43537
+    },
+    {
+      "epoch": 1.0363229553460915,
+      "grad_norm": 0.014237078812010943,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43538
+    },
+    {
+      "epoch": 1.036346758069123,
+      "grad_norm": 2.613338526203411,
+      "learning_rate": 2e-05,
+      "loss": 0.1231,
+      "step": 43539
+    },
+    {
+      "epoch": 1.0363705607921547,
+      "grad_norm": 3.712407933708081,
+      "learning_rate": 2e-05,
+      "loss": 0.2813,
+      "step": 43540
+    },
+    {
+      "epoch": 1.0363943635151862,
+      "grad_norm": 5.311088119783934,
+      "learning_rate": 2e-05,
+      "loss": 0.2713,
+      "step": 43541
+    },
+    {
+      "epoch": 1.0364181662382177,
+      "grad_norm": 4.621123873559884,
+      "learning_rate": 2e-05,
+      "loss": 0.6381,
+      "step": 43542
+    },
+    {
+      "epoch": 1.0364419689612492,
+      "grad_norm": 3.8876081768128525,
+      "learning_rate": 2e-05,
+      "loss": 0.5694,
+      "step": 43543
+    },
+    {
+      "epoch": 1.0364657716842807,
+      "grad_norm": 0.32248938515913,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 43544
+    },
+    {
+      "epoch": 1.0364895744073122,
+      "grad_norm": 9.901206980338658,
+      "learning_rate": 2e-05,
+      "loss": 0.7786,
+      "step": 43545
+    },
+    {
+      "epoch": 1.0365133771303436,
+      "grad_norm": 2.4897633318529473,
+      "learning_rate": 2e-05,
+      "loss": 0.3919,
+      "step": 43546
+    },
+    {
+      "epoch": 1.0365371798533751,
+      "grad_norm": 3.5498589190486376,
+      "learning_rate": 2e-05,
+      "loss": 0.376,
+      "step": 43547
+    },
+    {
+      "epoch": 1.0365609825764068,
+      "grad_norm": 5.982180919771063,
+      "learning_rate": 2e-05,
+      "loss": 0.9137,
+      "step": 43548
+    },
+    {
+      "epoch": 1.0365847852994383,
+      "grad_norm": 0.009735513822541106,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43549
+    },
+    {
+      "epoch": 1.0366085880224698,
+      "grad_norm": 0.2867759389837598,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 43550
+    },
+    {
+      "epoch": 1.0366323907455013,
+      "grad_norm": 0.035756873770965786,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43551
+    },
+    {
+      "epoch": 1.0366561934685328,
+      "grad_norm": 0.1082662508263923,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43552
+    },
+    {
+      "epoch": 1.0366799961915643,
+      "grad_norm": 0.8171377166412612,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 43553
+    },
+    {
+      "epoch": 1.0367037989145957,
+      "grad_norm": 0.12755001281389097,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 43554
+    },
+    {
+      "epoch": 1.0367276016376274,
+      "grad_norm": 5.282121325166886,
+      "learning_rate": 2e-05,
+      "loss": 0.6597,
+      "step": 43555
+    },
+    {
+      "epoch": 1.036751404360659,
+      "grad_norm": 0.09768581763817603,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 43556
+    },
+    {
+      "epoch": 1.0367752070836904,
+      "grad_norm": 4.54959815621802,
+      "learning_rate": 2e-05,
+      "loss": 0.5459,
+      "step": 43557
+    },
+    {
+      "epoch": 1.036799009806722,
+      "grad_norm": 3.6961860272238014,
+      "learning_rate": 2e-05,
+      "loss": 0.5764,
+      "step": 43558
+    },
+    {
+      "epoch": 1.0368228125297534,
+      "grad_norm": 3.3388024095251096,
+      "learning_rate": 2e-05,
+      "loss": 0.4546,
+      "step": 43559
+    },
+    {
+      "epoch": 1.0368466152527849,
+      "grad_norm": 0.070734730825581,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43560
+    },
+    {
+      "epoch": 1.0368704179758164,
+      "grad_norm": 2.792960080591673,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 43561
+    },
+    {
+      "epoch": 1.0368942206988478,
+      "grad_norm": 0.1560228213742849,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43562
+    },
+    {
+      "epoch": 1.0369180234218796,
+      "grad_norm": 4.859700251668991,
+      "learning_rate": 2e-05,
+      "loss": 0.7438,
+      "step": 43563
+    },
+    {
+      "epoch": 1.036941826144911,
+      "grad_norm": 0.022857945739303664,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43564
+    },
+    {
+      "epoch": 1.0369656288679425,
+      "grad_norm": 0.03398501777895518,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43565
+    },
+    {
+      "epoch": 1.036989431590974,
+      "grad_norm": 6.241158987523648,
+      "learning_rate": 2e-05,
+      "loss": 0.5388,
+      "step": 43566
+    },
+    {
+      "epoch": 1.0370132343140055,
+      "grad_norm": 0.14237729381855702,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 43567
+    },
+    {
+      "epoch": 1.037037037037037,
+      "grad_norm": 6.879248393448326,
+      "learning_rate": 2e-05,
+      "loss": 0.4892,
+      "step": 43568
+    },
+    {
+      "epoch": 1.0370608397600685,
+      "grad_norm": 6.662297742028418,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 43569
+    },
+    {
+      "epoch": 1.0370846424831002,
+      "grad_norm": 0.09881109058276455,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43570
+    },
+    {
+      "epoch": 1.0371084452061317,
+      "grad_norm": 2.48643448143641,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 43571
+    },
+    {
+      "epoch": 1.0371322479291631,
+      "grad_norm": 7.417684307625957,
+      "learning_rate": 2e-05,
+      "loss": 0.1519,
+      "step": 43572
+    },
+    {
+      "epoch": 1.0371560506521946,
+      "grad_norm": 2.0488757227493055,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 43573
+    },
+    {
+      "epoch": 1.0371798533752261,
+      "grad_norm": 0.11409558580059423,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43574
+    },
+    {
+      "epoch": 1.0372036560982576,
+      "grad_norm": 2.8604593025998137,
+      "learning_rate": 2e-05,
+      "loss": 0.2023,
+      "step": 43575
+    },
+    {
+      "epoch": 1.037227458821289,
+      "grad_norm": 1.5392684135103987,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 43576
+    },
+    {
+      "epoch": 1.0372512615443206,
+      "grad_norm": 1.0446329501516765,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 43577
+    },
+    {
+      "epoch": 1.0372750642673523,
+      "grad_norm": 0.07709451225928475,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43578
+    },
+    {
+      "epoch": 1.0372988669903838,
+      "grad_norm": 6.331244608497099,
+      "learning_rate": 2e-05,
+      "loss": 0.4035,
+      "step": 43579
+    },
+    {
+      "epoch": 1.0373226697134152,
+      "grad_norm": 4.571218252512112,
+      "learning_rate": 2e-05,
+      "loss": 0.0477,
+      "step": 43580
+    },
+    {
+      "epoch": 1.0373464724364467,
+      "grad_norm": 9.699323008972954,
+      "learning_rate": 2e-05,
+      "loss": 0.8759,
+      "step": 43581
+    },
+    {
+      "epoch": 1.0373702751594782,
+      "grad_norm": 3.939381350955841,
+      "learning_rate": 2e-05,
+      "loss": 0.2692,
+      "step": 43582
+    },
+    {
+      "epoch": 1.0373940778825097,
+      "grad_norm": 3.3436485009922174,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 43583
+    },
+    {
+      "epoch": 1.0374178806055412,
+      "grad_norm": 4.0156736924113545,
+      "learning_rate": 2e-05,
+      "loss": 0.5581,
+      "step": 43584
+    },
+    {
+      "epoch": 1.037441683328573,
+      "grad_norm": 2.9319927930335323,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 43585
+    },
+    {
+      "epoch": 1.0374654860516044,
+      "grad_norm": 0.34723625419926757,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 43586
+    },
+    {
+      "epoch": 1.0374892887746359,
+      "grad_norm": 5.628489797323099,
+      "learning_rate": 2e-05,
+      "loss": 0.5678,
+      "step": 43587
+    },
+    {
+      "epoch": 1.0375130914976674,
+      "grad_norm": 3.2030235075632105,
+      "learning_rate": 2e-05,
+      "loss": 0.4545,
+      "step": 43588
+    },
+    {
+      "epoch": 1.0375368942206988,
+      "grad_norm": 3.782543755627258,
+      "learning_rate": 2e-05,
+      "loss": 0.3299,
+      "step": 43589
+    },
+    {
+      "epoch": 1.0375606969437303,
+      "grad_norm": 1.5111245005449478,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 43590
+    },
+    {
+      "epoch": 1.0375844996667618,
+      "grad_norm": 1.5988661063348737,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 43591
+    },
+    {
+      "epoch": 1.0376083023897933,
+      "grad_norm": 4.2114093411256475,
+      "learning_rate": 2e-05,
+      "loss": 0.5224,
+      "step": 43592
+    },
+    {
+      "epoch": 1.037632105112825,
+      "grad_norm": 0.03228174020208668,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43593
+    },
+    {
+      "epoch": 1.0376559078358565,
+      "grad_norm": 0.5774344856877919,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 43594
+    },
+    {
+      "epoch": 1.037679710558888,
+      "grad_norm": 4.316848349931704,
+      "learning_rate": 2e-05,
+      "loss": 0.2913,
+      "step": 43595
+    },
+    {
+      "epoch": 1.0377035132819195,
+      "grad_norm": 7.549684324288885,
+      "learning_rate": 2e-05,
+      "loss": 0.2783,
+      "step": 43596
+    },
+    {
+      "epoch": 1.037727316004951,
+      "grad_norm": 1.9162055033283019,
+      "learning_rate": 2e-05,
+      "loss": 0.0981,
+      "step": 43597
+    },
+    {
+      "epoch": 1.0377511187279824,
+      "grad_norm": 0.1743343907132373,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 43598
+    },
+    {
+      "epoch": 1.037774921451014,
+      "grad_norm": 3.4471732253587146,
+      "learning_rate": 2e-05,
+      "loss": 0.6628,
+      "step": 43599
+    },
+    {
+      "epoch": 1.0377987241740456,
+      "grad_norm": 3.819865930883308,
+      "learning_rate": 2e-05,
+      "loss": 0.3178,
+      "step": 43600
+    },
+    {
+      "epoch": 1.037822526897077,
+      "grad_norm": 11.63169893547258,
+      "learning_rate": 2e-05,
+      "loss": 0.6136,
+      "step": 43601
+    },
+    {
+      "epoch": 1.0378463296201086,
+      "grad_norm": 1.2481979318720517,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 43602
+    },
+    {
+      "epoch": 1.03787013234314,
+      "grad_norm": 3.9186707602613793,
+      "learning_rate": 2e-05,
+      "loss": 0.3511,
+      "step": 43603
+    },
+    {
+      "epoch": 1.0378939350661716,
+      "grad_norm": 1.2977119367253316,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 43604
+    },
+    {
+      "epoch": 1.037917737789203,
+      "grad_norm": 22.284483374194647,
+      "learning_rate": 2e-05,
+      "loss": 0.3622,
+      "step": 43605
+    },
+    {
+      "epoch": 1.0379415405122345,
+      "grad_norm": 0.37088478258251534,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 43606
+    },
+    {
+      "epoch": 1.037965343235266,
+      "grad_norm": 2.6027860680170356,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 43607
+    },
+    {
+      "epoch": 1.0379891459582977,
+      "grad_norm": 1.231432604307212,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 43608
+    },
+    {
+      "epoch": 1.0380129486813292,
+      "grad_norm": 14.038020078188218,
+      "learning_rate": 2e-05,
+      "loss": 0.3559,
+      "step": 43609
+    },
+    {
+      "epoch": 1.0380367514043607,
+      "grad_norm": 17.052577280247224,
+      "learning_rate": 2e-05,
+      "loss": 1.1198,
+      "step": 43610
+    },
+    {
+      "epoch": 1.0380605541273922,
+      "grad_norm": 0.06905093580933744,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43611
+    },
+    {
+      "epoch": 1.0380843568504237,
+      "grad_norm": 4.371069082060042,
+      "learning_rate": 2e-05,
+      "loss": 0.7748,
+      "step": 43612
+    },
+    {
+      "epoch": 1.0381081595734551,
+      "grad_norm": 3.164987448487146,
+      "learning_rate": 2e-05,
+      "loss": 0.3722,
+      "step": 43613
+    },
+    {
+      "epoch": 1.0381319622964866,
+      "grad_norm": 3.467117100492025,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 43614
+    },
+    {
+      "epoch": 1.0381557650195183,
+      "grad_norm": 3.0061227963933335,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 43615
+    },
+    {
+      "epoch": 1.0381795677425498,
+      "grad_norm": 1.3732039628279145,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 43616
+    },
+    {
+      "epoch": 1.0382033704655813,
+      "grad_norm": 4.189165243645289,
+      "learning_rate": 2e-05,
+      "loss": 0.3039,
+      "step": 43617
+    },
+    {
+      "epoch": 1.0382271731886128,
+      "grad_norm": 8.697468535031602,
+      "learning_rate": 2e-05,
+      "loss": 0.3637,
+      "step": 43618
+    },
+    {
+      "epoch": 1.0382509759116443,
+      "grad_norm": 4.9764722299125586,
+      "learning_rate": 2e-05,
+      "loss": 0.2194,
+      "step": 43619
+    },
+    {
+      "epoch": 1.0382747786346758,
+      "grad_norm": 0.03182919455007236,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43620
+    },
+    {
+      "epoch": 1.0382985813577073,
+      "grad_norm": 10.155316364275311,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 43621
+    },
+    {
+      "epoch": 1.0383223840807387,
+      "grad_norm": 8.823416536600373,
+      "learning_rate": 2e-05,
+      "loss": 0.6578,
+      "step": 43622
+    },
+    {
+      "epoch": 1.0383461868037704,
+      "grad_norm": 2.7722656452094148,
+      "learning_rate": 2e-05,
+      "loss": 0.2813,
+      "step": 43623
+    },
+    {
+      "epoch": 1.038369989526802,
+      "grad_norm": 5.399182625043855,
+      "learning_rate": 2e-05,
+      "loss": 0.6073,
+      "step": 43624
+    },
+    {
+      "epoch": 1.0383937922498334,
+      "grad_norm": 3.2224323872568315,
+      "learning_rate": 2e-05,
+      "loss": 0.2947,
+      "step": 43625
+    },
+    {
+      "epoch": 1.038417594972865,
+      "grad_norm": 0.9072937004204864,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 43626
+    },
+    {
+      "epoch": 1.0384413976958964,
+      "grad_norm": 4.4439518408006355,
+      "learning_rate": 2e-05,
+      "loss": 0.433,
+      "step": 43627
+    },
+    {
+      "epoch": 1.0384652004189279,
+      "grad_norm": 5.457237360650175,
+      "learning_rate": 2e-05,
+      "loss": 0.7664,
+      "step": 43628
+    },
+    {
+      "epoch": 1.0384890031419594,
+      "grad_norm": 4.008026591425659,
+      "learning_rate": 2e-05,
+      "loss": 0.2017,
+      "step": 43629
+    },
+    {
+      "epoch": 1.038512805864991,
+      "grad_norm": 0.08848859990195211,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43630
+    },
+    {
+      "epoch": 1.0385366085880225,
+      "grad_norm": 7.4669904049613605,
+      "learning_rate": 2e-05,
+      "loss": 1.3993,
+      "step": 43631
+    },
+    {
+      "epoch": 1.038560411311054,
+      "grad_norm": 7.88386002797484,
+      "learning_rate": 2e-05,
+      "loss": 1.111,
+      "step": 43632
+    },
+    {
+      "epoch": 1.0385842140340855,
+      "grad_norm": 4.139129569973821,
+      "learning_rate": 2e-05,
+      "loss": 0.5457,
+      "step": 43633
+    },
+    {
+      "epoch": 1.038608016757117,
+      "grad_norm": 2.343949422902286,
+      "learning_rate": 2e-05,
+      "loss": 0.1788,
+      "step": 43634
+    },
+    {
+      "epoch": 1.0386318194801485,
+      "grad_norm": 4.332105171660778,
+      "learning_rate": 2e-05,
+      "loss": 0.54,
+      "step": 43635
+    },
+    {
+      "epoch": 1.03865562220318,
+      "grad_norm": 2.441930684475454,
+      "learning_rate": 2e-05,
+      "loss": 0.1752,
+      "step": 43636
+    },
+    {
+      "epoch": 1.0386794249262115,
+      "grad_norm": 5.246320243456175,
+      "learning_rate": 2e-05,
+      "loss": 0.6552,
+      "step": 43637
+    },
+    {
+      "epoch": 1.0387032276492432,
+      "grad_norm": 5.290131578039002,
+      "learning_rate": 2e-05,
+      "loss": 0.5391,
+      "step": 43638
+    },
+    {
+      "epoch": 1.0387270303722747,
+      "grad_norm": 7.583296190416946,
+      "learning_rate": 2e-05,
+      "loss": 0.9391,
+      "step": 43639
+    },
+    {
+      "epoch": 1.0387508330953061,
+      "grad_norm": 4.225432390033369,
+      "learning_rate": 2e-05,
+      "loss": 0.106,
+      "step": 43640
+    },
+    {
+      "epoch": 1.0387746358183376,
+      "grad_norm": 0.04430563933260135,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43641
+    },
+    {
+      "epoch": 1.038798438541369,
+      "grad_norm": 0.23444834962552255,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 43642
+    },
+    {
+      "epoch": 1.0388222412644006,
+      "grad_norm": 0.4178430742944527,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 43643
+    },
+    {
+      "epoch": 1.038846043987432,
+      "grad_norm": 0.2566742068971955,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 43644
+    },
+    {
+      "epoch": 1.0388698467104638,
+      "grad_norm": 22.525376711569333,
+      "learning_rate": 2e-05,
+      "loss": 0.6936,
+      "step": 43645
+    },
+    {
+      "epoch": 1.0388936494334953,
+      "grad_norm": 28.669378657981166,
+      "learning_rate": 2e-05,
+      "loss": 0.0792,
+      "step": 43646
+    },
+    {
+      "epoch": 1.0389174521565268,
+      "grad_norm": 6.901977586005583,
+      "learning_rate": 2e-05,
+      "loss": 0.6596,
+      "step": 43647
+    },
+    {
+      "epoch": 1.0389412548795582,
+      "grad_norm": 4.207076807342747,
+      "learning_rate": 2e-05,
+      "loss": 0.3568,
+      "step": 43648
+    },
+    {
+      "epoch": 1.0389650576025897,
+      "grad_norm": 2.6279994244934977,
+      "learning_rate": 2e-05,
+      "loss": 0.3969,
+      "step": 43649
+    },
+    {
+      "epoch": 1.0389888603256212,
+      "grad_norm": 0.07609969572292914,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43650
+    },
+    {
+      "epoch": 1.0390126630486527,
+      "grad_norm": 14.724609638258478,
+      "learning_rate": 2e-05,
+      "loss": 0.1585,
+      "step": 43651
+    },
+    {
+      "epoch": 1.0390364657716842,
+      "grad_norm": 7.347715379924204,
+      "learning_rate": 2e-05,
+      "loss": 0.1941,
+      "step": 43652
+    },
+    {
+      "epoch": 1.0390602684947159,
+      "grad_norm": 3.246935633134691,
+      "learning_rate": 2e-05,
+      "loss": 0.03,
+      "step": 43653
+    },
+    {
+      "epoch": 1.0390840712177474,
+      "grad_norm": 0.13297901708826,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 43654
+    },
+    {
+      "epoch": 1.0391078739407789,
+      "grad_norm": 1.3629996115054899,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 43655
+    },
+    {
+      "epoch": 1.0391316766638103,
+      "grad_norm": 4.114729373292003,
+      "learning_rate": 2e-05,
+      "loss": 0.5794,
+      "step": 43656
+    },
+    {
+      "epoch": 1.0391554793868418,
+      "grad_norm": 6.053299182893789,
+      "learning_rate": 2e-05,
+      "loss": 1.3849,
+      "step": 43657
+    },
+    {
+      "epoch": 1.0391792821098733,
+      "grad_norm": 8.895353812911958,
+      "learning_rate": 2e-05,
+      "loss": 0.5145,
+      "step": 43658
+    },
+    {
+      "epoch": 1.0392030848329048,
+      "grad_norm": 0.5250794742505507,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 43659
+    },
+    {
+      "epoch": 1.0392268875559365,
+      "grad_norm": 0.033960473569585446,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43660
+    },
+    {
+      "epoch": 1.039250690278968,
+      "grad_norm": 0.0439199470210535,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43661
+    },
+    {
+      "epoch": 1.0392744930019995,
+      "grad_norm": 8.677011591067597,
+      "learning_rate": 2e-05,
+      "loss": 0.1101,
+      "step": 43662
+    },
+    {
+      "epoch": 1.039298295725031,
+      "grad_norm": 0.00842455959323524,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43663
+    },
+    {
+      "epoch": 1.0393220984480624,
+      "grad_norm": 0.8684992649664547,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 43664
+    },
+    {
+      "epoch": 1.039345901171094,
+      "grad_norm": 2.760971961890759,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 43665
+    },
+    {
+      "epoch": 1.0393697038941254,
+      "grad_norm": 12.884475548965401,
+      "learning_rate": 2e-05,
+      "loss": 0.4844,
+      "step": 43666
+    },
+    {
+      "epoch": 1.039393506617157,
+      "grad_norm": 3.1248087313233324,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 43667
+    },
+    {
+      "epoch": 1.0394173093401886,
+      "grad_norm": 5.553336008037336,
+      "learning_rate": 2e-05,
+      "loss": 0.4997,
+      "step": 43668
+    },
+    {
+      "epoch": 1.03944111206322,
+      "grad_norm": 15.095620047241718,
+      "learning_rate": 2e-05,
+      "loss": 1.254,
+      "step": 43669
+    },
+    {
+      "epoch": 1.0394649147862516,
+      "grad_norm": 2.847532538937855,
+      "learning_rate": 2e-05,
+      "loss": 0.2753,
+      "step": 43670
+    },
+    {
+      "epoch": 1.039488717509283,
+      "grad_norm": 0.6519390041483846,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 43671
+    },
+    {
+      "epoch": 1.0395125202323146,
+      "grad_norm": 0.12984600283937064,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 43672
+    },
+    {
+      "epoch": 1.039536322955346,
+      "grad_norm": 5.62168158756323,
+      "learning_rate": 2e-05,
+      "loss": 0.3626,
+      "step": 43673
+    },
+    {
+      "epoch": 1.0395601256783775,
+      "grad_norm": 4.449410520089508,
+      "learning_rate": 2e-05,
+      "loss": 0.3112,
+      "step": 43674
+    },
+    {
+      "epoch": 1.039583928401409,
+      "grad_norm": 0.2508144465959236,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 43675
+    },
+    {
+      "epoch": 1.0396077311244407,
+      "grad_norm": 4.075611613522556,
+      "learning_rate": 2e-05,
+      "loss": 0.2168,
+      "step": 43676
+    },
+    {
+      "epoch": 1.0396315338474722,
+      "grad_norm": 0.07327599577298295,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43677
+    },
+    {
+      "epoch": 1.0396553365705037,
+      "grad_norm": 0.5494980148070379,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 43678
+    },
+    {
+      "epoch": 1.0396791392935352,
+      "grad_norm": 5.767182859858499,
+      "learning_rate": 2e-05,
+      "loss": 0.9827,
+      "step": 43679
+    },
+    {
+      "epoch": 1.0397029420165667,
+      "grad_norm": 2.3741243734503543,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 43680
+    },
+    {
+      "epoch": 1.0397267447395981,
+      "grad_norm": 0.6281144184153364,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 43681
+    },
+    {
+      "epoch": 1.0397505474626296,
+      "grad_norm": 5.542122749627525,
+      "learning_rate": 2e-05,
+      "loss": 1.1488,
+      "step": 43682
+    },
+    {
+      "epoch": 1.0397743501856613,
+      "grad_norm": 5.873213685952879,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 43683
+    },
+    {
+      "epoch": 1.0397981529086928,
+      "grad_norm": 3.7368069247437186,
+      "learning_rate": 2e-05,
+      "loss": 0.4565,
+      "step": 43684
+    },
+    {
+      "epoch": 1.0398219556317243,
+      "grad_norm": 0.026875842551869485,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43685
+    },
+    {
+      "epoch": 1.0398457583547558,
+      "grad_norm": 1.183946501473149,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 43686
+    },
+    {
+      "epoch": 1.0398695610777873,
+      "grad_norm": 4.21798082522093,
+      "learning_rate": 2e-05,
+      "loss": 0.4272,
+      "step": 43687
+    },
+    {
+      "epoch": 1.0398933638008188,
+      "grad_norm": 9.058621326734677,
+      "learning_rate": 2e-05,
+      "loss": 0.4486,
+      "step": 43688
+    },
+    {
+      "epoch": 1.0399171665238502,
+      "grad_norm": 0.13977615435016735,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 43689
+    },
+    {
+      "epoch": 1.039940969246882,
+      "grad_norm": 3.8542941869932097,
+      "learning_rate": 2e-05,
+      "loss": 0.4726,
+      "step": 43690
+    },
+    {
+      "epoch": 1.0399647719699134,
+      "grad_norm": 4.844801206793682,
+      "learning_rate": 2e-05,
+      "loss": 0.5722,
+      "step": 43691
+    },
+    {
+      "epoch": 1.039988574692945,
+      "grad_norm": 0.10607138863921432,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 43692
+    },
+    {
+      "epoch": 1.0400123774159764,
+      "grad_norm": 0.02518975772962033,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43693
+    },
+    {
+      "epoch": 1.040036180139008,
+      "grad_norm": 2.1943837317674215,
+      "learning_rate": 2e-05,
+      "loss": 0.0363,
+      "step": 43694
+    },
+    {
+      "epoch": 1.0400599828620394,
+      "grad_norm": 4.905845422045444,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 43695
+    },
+    {
+      "epoch": 1.0400837855850709,
+      "grad_norm": 0.43154169210215554,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 43696
+    },
+    {
+      "epoch": 1.0401075883081023,
+      "grad_norm": 5.511094858642601,
+      "learning_rate": 2e-05,
+      "loss": 0.0679,
+      "step": 43697
+    },
+    {
+      "epoch": 1.040131391031134,
+      "grad_norm": 0.06328803850102806,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43698
+    },
+    {
+      "epoch": 1.0401551937541655,
+      "grad_norm": 0.3840096722766891,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 43699
+    },
+    {
+      "epoch": 1.040178996477197,
+      "grad_norm": 1.5432273729764379,
+      "learning_rate": 2e-05,
+      "loss": 0.0157,
+      "step": 43700
+    },
+    {
+      "epoch": 1.0402027992002285,
+      "grad_norm": 3.860921616105987,
+      "learning_rate": 2e-05,
+      "loss": 0.0253,
+      "step": 43701
+    },
+    {
+      "epoch": 1.04022660192326,
+      "grad_norm": 0.1717374116726929,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 43702
+    },
+    {
+      "epoch": 1.0402504046462915,
+      "grad_norm": 8.82726754338355,
+      "learning_rate": 2e-05,
+      "loss": 0.631,
+      "step": 43703
+    },
+    {
+      "epoch": 1.040274207369323,
+      "grad_norm": 4.040785032939957,
+      "learning_rate": 2e-05,
+      "loss": 0.4609,
+      "step": 43704
+    },
+    {
+      "epoch": 1.0402980100923545,
+      "grad_norm": 5.409708181654304,
+      "learning_rate": 2e-05,
+      "loss": 0.713,
+      "step": 43705
+    },
+    {
+      "epoch": 1.0403218128153862,
+      "grad_norm": 2.8039674157902867,
+      "learning_rate": 2e-05,
+      "loss": 0.2158,
+      "step": 43706
+    },
+    {
+      "epoch": 1.0403456155384176,
+      "grad_norm": 4.642814025362658,
+      "learning_rate": 2e-05,
+      "loss": 0.9359,
+      "step": 43707
+    },
+    {
+      "epoch": 1.0403694182614491,
+      "grad_norm": 0.5004369375422181,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 43708
+    },
+    {
+      "epoch": 1.0403932209844806,
+      "grad_norm": 4.0568545113168035,
+      "learning_rate": 2e-05,
+      "loss": 0.1742,
+      "step": 43709
+    },
+    {
+      "epoch": 1.040417023707512,
+      "grad_norm": 2.2428847038659403,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 43710
+    },
+    {
+      "epoch": 1.0404408264305436,
+      "grad_norm": 0.13376133144280816,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43711
+    },
+    {
+      "epoch": 1.040464629153575,
+      "grad_norm": 0.8975478093814191,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 43712
+    },
+    {
+      "epoch": 1.0404884318766068,
+      "grad_norm": 5.7393668160299836,
+      "learning_rate": 2e-05,
+      "loss": 0.7867,
+      "step": 43713
+    },
+    {
+      "epoch": 1.0405122345996383,
+      "grad_norm": 3.861020812540205,
+      "learning_rate": 2e-05,
+      "loss": 0.0697,
+      "step": 43714
+    },
+    {
+      "epoch": 1.0405360373226697,
+      "grad_norm": 0.8501577910525365,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 43715
+    },
+    {
+      "epoch": 1.0405598400457012,
+      "grad_norm": 3.048214389169072,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 43716
+    },
+    {
+      "epoch": 1.0405836427687327,
+      "grad_norm": 2.5022885540200437,
+      "learning_rate": 2e-05,
+      "loss": 0.1414,
+      "step": 43717
+    },
+    {
+      "epoch": 1.0406074454917642,
+      "grad_norm": 1.6680578906923968,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 43718
+    },
+    {
+      "epoch": 1.0406312482147957,
+      "grad_norm": 1.7241471203554457,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 43719
+    },
+    {
+      "epoch": 1.0406550509378274,
+      "grad_norm": 3.0513323009299342,
+      "learning_rate": 2e-05,
+      "loss": 0.4198,
+      "step": 43720
+    },
+    {
+      "epoch": 1.0406788536608589,
+      "grad_norm": 0.027488508905470485,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43721
+    },
+    {
+      "epoch": 1.0407026563838904,
+      "grad_norm": 6.078480521764146,
+      "learning_rate": 2e-05,
+      "loss": 1.1692,
+      "step": 43722
+    },
+    {
+      "epoch": 1.0407264591069219,
+      "grad_norm": 4.098806153210778,
+      "learning_rate": 2e-05,
+      "loss": 0.4751,
+      "step": 43723
+    },
+    {
+      "epoch": 1.0407502618299533,
+      "grad_norm": 1.0764342226931216,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 43724
+    },
+    {
+      "epoch": 1.0407740645529848,
+      "grad_norm": 2.6373048449925784,
+      "learning_rate": 2e-05,
+      "loss": 0.071,
+      "step": 43725
+    },
+    {
+      "epoch": 1.0407978672760163,
+      "grad_norm": 1.7941816342849262,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 43726
+    },
+    {
+      "epoch": 1.0408216699990478,
+      "grad_norm": 8.52245433783057,
+      "learning_rate": 2e-05,
+      "loss": 0.8193,
+      "step": 43727
+    },
+    {
+      "epoch": 1.0408454727220795,
+      "grad_norm": 3.871351783056532,
+      "learning_rate": 2e-05,
+      "loss": 0.5178,
+      "step": 43728
+    },
+    {
+      "epoch": 1.040869275445111,
+      "grad_norm": 0.19188297518962305,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 43729
+    },
+    {
+      "epoch": 1.0408930781681425,
+      "grad_norm": 2.254503665091132,
+      "learning_rate": 2e-05,
+      "loss": 0.1863,
+      "step": 43730
+    },
+    {
+      "epoch": 1.040916880891174,
+      "grad_norm": 6.064781322288363,
+      "learning_rate": 2e-05,
+      "loss": 0.1597,
+      "step": 43731
+    },
+    {
+      "epoch": 1.0409406836142054,
+      "grad_norm": 0.8080465127350906,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 43732
+    },
+    {
+      "epoch": 1.040964486337237,
+      "grad_norm": 1.502112216829801,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 43733
+    },
+    {
+      "epoch": 1.0409882890602684,
+      "grad_norm": 0.011118749908632672,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43734
+    },
+    {
+      "epoch": 1.0410120917833,
+      "grad_norm": 0.38885823161198074,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 43735
+    },
+    {
+      "epoch": 1.0410358945063316,
+      "grad_norm": 18.33558904666785,
+      "learning_rate": 2e-05,
+      "loss": 0.5499,
+      "step": 43736
+    },
+    {
+      "epoch": 1.041059697229363,
+      "grad_norm": 2.244406955914461,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 43737
+    },
+    {
+      "epoch": 1.0410834999523946,
+      "grad_norm": 0.1448880076621713,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 43738
+    },
+    {
+      "epoch": 1.041107302675426,
+      "grad_norm": 5.161877631055312,
+      "learning_rate": 2e-05,
+      "loss": 0.068,
+      "step": 43739
+    },
+    {
+      "epoch": 1.0411311053984575,
+      "grad_norm": 0.009008120515763511,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43740
+    },
+    {
+      "epoch": 1.041154908121489,
+      "grad_norm": 9.741278370530491,
+      "learning_rate": 2e-05,
+      "loss": 0.9148,
+      "step": 43741
+    },
+    {
+      "epoch": 1.0411787108445205,
+      "grad_norm": 0.0018787063644040887,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 43742
+    },
+    {
+      "epoch": 1.0412025135675522,
+      "grad_norm": 1.2868756414471514,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 43743
+    },
+    {
+      "epoch": 1.0412263162905837,
+      "grad_norm": 5.779239746800999,
+      "learning_rate": 2e-05,
+      "loss": 0.4492,
+      "step": 43744
+    },
+    {
+      "epoch": 1.0412501190136152,
+      "grad_norm": 2.969594314571506,
+      "learning_rate": 2e-05,
+      "loss": 0.3108,
+      "step": 43745
+    },
+    {
+      "epoch": 1.0412739217366467,
+      "grad_norm": 6.942245160143766,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 43746
+    },
+    {
+      "epoch": 1.0412977244596782,
+      "grad_norm": 0.028461012620082878,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43747
+    },
+    {
+      "epoch": 1.0413215271827096,
+      "grad_norm": 0.9898137437528942,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 43748
+    },
+    {
+      "epoch": 1.0413453299057411,
+      "grad_norm": 3.06175727271558,
+      "learning_rate": 2e-05,
+      "loss": 0.1716,
+      "step": 43749
+    },
+    {
+      "epoch": 1.0413691326287728,
+      "grad_norm": 5.628268071336509,
+      "learning_rate": 2e-05,
+      "loss": 0.4794,
+      "step": 43750
+    },
+    {
+      "epoch": 1.0413929353518043,
+      "grad_norm": 13.05488180067638,
+      "learning_rate": 2e-05,
+      "loss": 0.5884,
+      "step": 43751
+    },
+    {
+      "epoch": 1.0414167380748358,
+      "grad_norm": 0.03058819605364092,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 43752
+    },
+    {
+      "epoch": 1.0414405407978673,
+      "grad_norm": 0.8385487717898847,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 43753
+    },
+    {
+      "epoch": 1.0414643435208988,
+      "grad_norm": 1.9867926503200333,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 43754
+    },
+    {
+      "epoch": 1.0414881462439303,
+      "grad_norm": 2.446580317154049,
+      "learning_rate": 2e-05,
+      "loss": 0.1097,
+      "step": 43755
+    },
+    {
+      "epoch": 1.0415119489669618,
+      "grad_norm": 6.914672185364458,
+      "learning_rate": 2e-05,
+      "loss": 1.1504,
+      "step": 43756
+    },
+    {
+      "epoch": 1.0415357516899932,
+      "grad_norm": 3.9238148749493216,
+      "learning_rate": 2e-05,
+      "loss": 0.0749,
+      "step": 43757
+    },
+    {
+      "epoch": 1.041559554413025,
+      "grad_norm": 5.857491966375781,
+      "learning_rate": 2e-05,
+      "loss": 0.2345,
+      "step": 43758
+    },
+    {
+      "epoch": 1.0415833571360564,
+      "grad_norm": 6.6593869696745,
+      "learning_rate": 2e-05,
+      "loss": 0.4679,
+      "step": 43759
+    },
+    {
+      "epoch": 1.041607159859088,
+      "grad_norm": 4.1768975630041325,
+      "learning_rate": 2e-05,
+      "loss": 0.6223,
+      "step": 43760
+    },
+    {
+      "epoch": 1.0416309625821194,
+      "grad_norm": 0.15976594324063115,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 43761
+    },
+    {
+      "epoch": 1.0416547653051509,
+      "grad_norm": 3.5687457855600524,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 43762
+    },
+    {
+      "epoch": 1.0416785680281824,
+      "grad_norm": 5.2984043895453485,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 43763
+    },
+    {
+      "epoch": 1.0417023707512139,
+      "grad_norm": 4.74988244304641,
+      "learning_rate": 2e-05,
+      "loss": 0.1771,
+      "step": 43764
+    },
+    {
+      "epoch": 1.0417261734742453,
+      "grad_norm": 0.07953503174919814,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43765
+    },
+    {
+      "epoch": 1.041749976197277,
+      "grad_norm": 0.010708869739744315,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43766
+    },
+    {
+      "epoch": 1.0417737789203085,
+      "grad_norm": 11.651691654039404,
+      "learning_rate": 2e-05,
+      "loss": 0.6416,
+      "step": 43767
+    },
+    {
+      "epoch": 1.04179758164334,
+      "grad_norm": 3.266405551733374,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 43768
+    },
+    {
+      "epoch": 1.0418213843663715,
+      "grad_norm": 1.6892900203618184,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 43769
+    },
+    {
+      "epoch": 1.041845187089403,
+      "grad_norm": 3.4519492543778054,
+      "learning_rate": 2e-05,
+      "loss": 0.2301,
+      "step": 43770
+    },
+    {
+      "epoch": 1.0418689898124345,
+      "grad_norm": 3.2346894215825914,
+      "learning_rate": 2e-05,
+      "loss": 0.5051,
+      "step": 43771
+    },
+    {
+      "epoch": 1.041892792535466,
+      "grad_norm": 6.665850984915281,
+      "learning_rate": 2e-05,
+      "loss": 0.0695,
+      "step": 43772
+    },
+    {
+      "epoch": 1.0419165952584977,
+      "grad_norm": 10.067906633595626,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 43773
+    },
+    {
+      "epoch": 1.0419403979815292,
+      "grad_norm": 3.5469192720807077,
+      "learning_rate": 2e-05,
+      "loss": 0.568,
+      "step": 43774
+    },
+    {
+      "epoch": 1.0419642007045606,
+      "grad_norm": 2.5138230127078174,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 43775
+    },
+    {
+      "epoch": 1.0419880034275921,
+      "grad_norm": 3.663954550532745,
+      "learning_rate": 2e-05,
+      "loss": 0.3896,
+      "step": 43776
+    },
+    {
+      "epoch": 1.0420118061506236,
+      "grad_norm": 3.208483398066012,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 43777
+    },
+    {
+      "epoch": 1.042035608873655,
+      "grad_norm": 0.08434920137001334,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43778
+    },
+    {
+      "epoch": 1.0420594115966866,
+      "grad_norm": 3.195591494775816,
+      "learning_rate": 2e-05,
+      "loss": 0.5189,
+      "step": 43779
+    },
+    {
+      "epoch": 1.0420832143197183,
+      "grad_norm": 4.962173520320593,
+      "learning_rate": 2e-05,
+      "loss": 0.7743,
+      "step": 43780
+    },
+    {
+      "epoch": 1.0421070170427498,
+      "grad_norm": 4.169825575969713,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 43781
+    },
+    {
+      "epoch": 1.0421308197657813,
+      "grad_norm": 12.580024328536071,
+      "learning_rate": 2e-05,
+      "loss": 1.028,
+      "step": 43782
+    },
+    {
+      "epoch": 1.0421546224888127,
+      "grad_norm": 5.471853558421644,
+      "learning_rate": 2e-05,
+      "loss": 0.0782,
+      "step": 43783
+    },
+    {
+      "epoch": 1.0421784252118442,
+      "grad_norm": 3.672066555791755,
+      "learning_rate": 2e-05,
+      "loss": 0.2052,
+      "step": 43784
+    },
+    {
+      "epoch": 1.0422022279348757,
+      "grad_norm": 0.05870327270386524,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43785
+    },
+    {
+      "epoch": 1.0422260306579072,
+      "grad_norm": 0.8925627548606798,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 43786
+    },
+    {
+      "epoch": 1.0422498333809387,
+      "grad_norm": 36.410556687997556,
+      "learning_rate": 2e-05,
+      "loss": 0.4599,
+      "step": 43787
+    },
+    {
+      "epoch": 1.0422736361039704,
+      "grad_norm": 2.350832526139626,
+      "learning_rate": 2e-05,
+      "loss": 0.2741,
+      "step": 43788
+    },
+    {
+      "epoch": 1.0422974388270019,
+      "grad_norm": 4.188759333793528,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 43789
+    },
+    {
+      "epoch": 1.0423212415500334,
+      "grad_norm": 3.4397681026536677,
+      "learning_rate": 2e-05,
+      "loss": 0.4018,
+      "step": 43790
+    },
+    {
+      "epoch": 1.0423450442730648,
+      "grad_norm": 6.744736345211562,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 43791
+    },
+    {
+      "epoch": 1.0423688469960963,
+      "grad_norm": 1.245260164058064,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 43792
+    },
+    {
+      "epoch": 1.0423926497191278,
+      "grad_norm": 0.8032596405232469,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 43793
+    },
+    {
+      "epoch": 1.0424164524421593,
+      "grad_norm": 2.961132913688595,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 43794
+    },
+    {
+      "epoch": 1.0424402551651908,
+      "grad_norm": 0.06845589325341457,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43795
+    },
+    {
+      "epoch": 1.0424640578882225,
+      "grad_norm": 64.84195428002205,
+      "learning_rate": 2e-05,
+      "loss": 1.1758,
+      "step": 43796
+    },
+    {
+      "epoch": 1.042487860611254,
+      "grad_norm": 1.350068749601673,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 43797
+    },
+    {
+      "epoch": 1.0425116633342855,
+      "grad_norm": 3.947336297796828,
+      "learning_rate": 2e-05,
+      "loss": 0.495,
+      "step": 43798
+    },
+    {
+      "epoch": 1.042535466057317,
+      "grad_norm": 9.576344270165055,
+      "learning_rate": 2e-05,
+      "loss": 0.3443,
+      "step": 43799
+    },
+    {
+      "epoch": 1.0425592687803484,
+      "grad_norm": 0.68410961279575,
+      "learning_rate": 2e-05,
+      "loss": 0.0061,
+      "step": 43800
+    },
+    {
+      "epoch": 1.04258307150338,
+      "grad_norm": 2.9624976644335694,
+      "learning_rate": 2e-05,
+      "loss": 0.3708,
+      "step": 43801
+    },
+    {
+      "epoch": 1.0426068742264114,
+      "grad_norm": 0.2332951169074393,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 43802
+    },
+    {
+      "epoch": 1.0426306769494431,
+      "grad_norm": 10.067408413395727,
+      "learning_rate": 2e-05,
+      "loss": 1.4946,
+      "step": 43803
+    },
+    {
+      "epoch": 1.0426544796724746,
+      "grad_norm": 13.386872967449527,
+      "learning_rate": 2e-05,
+      "loss": 0.8487,
+      "step": 43804
+    },
+    {
+      "epoch": 1.042678282395506,
+      "grad_norm": 0.32152577585215836,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 43805
+    },
+    {
+      "epoch": 1.0427020851185376,
+      "grad_norm": 4.495602382867101,
+      "learning_rate": 2e-05,
+      "loss": 0.1376,
+      "step": 43806
+    },
+    {
+      "epoch": 1.042725887841569,
+      "grad_norm": 8.818826609857553,
+      "learning_rate": 2e-05,
+      "loss": 1.1313,
+      "step": 43807
+    },
+    {
+      "epoch": 1.0427496905646005,
+      "grad_norm": 5.007959947980155,
+      "learning_rate": 2e-05,
+      "loss": 0.2305,
+      "step": 43808
+    },
+    {
+      "epoch": 1.042773493287632,
+      "grad_norm": 4.771374555368266,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 43809
+    },
+    {
+      "epoch": 1.0427972960106637,
+      "grad_norm": 4.979876068324984,
+      "learning_rate": 2e-05,
+      "loss": 0.1448,
+      "step": 43810
+    },
+    {
+      "epoch": 1.0428210987336952,
+      "grad_norm": 3.1881017288870805,
+      "learning_rate": 2e-05,
+      "loss": 0.54,
+      "step": 43811
+    },
+    {
+      "epoch": 1.0428449014567267,
+      "grad_norm": 5.579518972685026,
+      "learning_rate": 2e-05,
+      "loss": 0.5094,
+      "step": 43812
+    },
+    {
+      "epoch": 1.0428687041797582,
+      "grad_norm": 0.008062324770468617,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43813
+    },
+    {
+      "epoch": 1.0428925069027897,
+      "grad_norm": 8.533317195856922,
+      "learning_rate": 2e-05,
+      "loss": 0.2474,
+      "step": 43814
+    },
+    {
+      "epoch": 1.0429163096258212,
+      "grad_norm": 6.900430092994204,
+      "learning_rate": 2e-05,
+      "loss": 1.5575,
+      "step": 43815
+    },
+    {
+      "epoch": 1.0429401123488526,
+      "grad_norm": 0.2766987394370999,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 43816
+    },
+    {
+      "epoch": 1.0429639150718841,
+      "grad_norm": 3.7490683109697605,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 43817
+    },
+    {
+      "epoch": 1.0429877177949158,
+      "grad_norm": 0.12437426092453799,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43818
+    },
+    {
+      "epoch": 1.0430115205179473,
+      "grad_norm": 10.999441343913846,
+      "learning_rate": 2e-05,
+      "loss": 0.961,
+      "step": 43819
+    },
+    {
+      "epoch": 1.0430353232409788,
+      "grad_norm": 14.704314818121643,
+      "learning_rate": 2e-05,
+      "loss": 0.6239,
+      "step": 43820
+    },
+    {
+      "epoch": 1.0430591259640103,
+      "grad_norm": 2.2543572395846736,
+      "learning_rate": 2e-05,
+      "loss": 0.0186,
+      "step": 43821
+    },
+    {
+      "epoch": 1.0430829286870418,
+      "grad_norm": 4.0903142462846125,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 43822
+    },
+    {
+      "epoch": 1.0431067314100733,
+      "grad_norm": 3.2768751807482404,
+      "learning_rate": 2e-05,
+      "loss": 0.0472,
+      "step": 43823
+    },
+    {
+      "epoch": 1.0431305341331047,
+      "grad_norm": 0.0951141479380722,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43824
+    },
+    {
+      "epoch": 1.0431543368561362,
+      "grad_norm": 1.1188812412962914,
+      "learning_rate": 2e-05,
+      "loss": 0.0086,
+      "step": 43825
+    },
+    {
+      "epoch": 1.043178139579168,
+      "grad_norm": 2.965368835228957,
+      "learning_rate": 2e-05,
+      "loss": 0.0367,
+      "step": 43826
+    },
+    {
+      "epoch": 1.0432019423021994,
+      "grad_norm": 3.0749366721769635,
+      "learning_rate": 2e-05,
+      "loss": 0.3743,
+      "step": 43827
+    },
+    {
+      "epoch": 1.043225745025231,
+      "grad_norm": 3.624234356727034,
+      "learning_rate": 2e-05,
+      "loss": 0.472,
+      "step": 43828
+    },
+    {
+      "epoch": 1.0432495477482624,
+      "grad_norm": 3.977021324589919,
+      "learning_rate": 2e-05,
+      "loss": 0.2905,
+      "step": 43829
+    },
+    {
+      "epoch": 1.0432733504712939,
+      "grad_norm": 5.464741214928137,
+      "learning_rate": 2e-05,
+      "loss": 0.359,
+      "step": 43830
+    },
+    {
+      "epoch": 1.0432971531943254,
+      "grad_norm": 2.341639264655641,
+      "learning_rate": 2e-05,
+      "loss": 0.0725,
+      "step": 43831
+    },
+    {
+      "epoch": 1.0433209559173569,
+      "grad_norm": 5.336390656790262,
+      "learning_rate": 2e-05,
+      "loss": 0.5219,
+      "step": 43832
+    },
+    {
+      "epoch": 1.0433447586403886,
+      "grad_norm": 1.9218311053547756,
+      "learning_rate": 2e-05,
+      "loss": 0.0387,
+      "step": 43833
+    },
+    {
+      "epoch": 1.04336856136342,
+      "grad_norm": 5.894260624620863,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 43834
+    },
+    {
+      "epoch": 1.0433923640864515,
+      "grad_norm": 7.882472738111161,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 43835
+    },
+    {
+      "epoch": 1.043416166809483,
+      "grad_norm": 7.167898463744532,
+      "learning_rate": 2e-05,
+      "loss": 1.2631,
+      "step": 43836
+    },
+    {
+      "epoch": 1.0434399695325145,
+      "grad_norm": 2.1219020969142433,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 43837
+    },
+    {
+      "epoch": 1.043463772255546,
+      "grad_norm": 6.837312057055441,
+      "learning_rate": 2e-05,
+      "loss": 0.1532,
+      "step": 43838
+    },
+    {
+      "epoch": 1.0434875749785775,
+      "grad_norm": 0.12786615772366675,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43839
+    },
+    {
+      "epoch": 1.0435113777016092,
+      "grad_norm": 11.893669921451593,
+      "learning_rate": 2e-05,
+      "loss": 1.3117,
+      "step": 43840
+    },
+    {
+      "epoch": 1.0435351804246407,
+      "grad_norm": 8.255380878798434,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 43841
+    },
+    {
+      "epoch": 1.0435589831476721,
+      "grad_norm": 4.295731074327589,
+      "learning_rate": 2e-05,
+      "loss": 0.6282,
+      "step": 43842
+    },
+    {
+      "epoch": 1.0435827858707036,
+      "grad_norm": 0.8401124090999682,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 43843
+    },
+    {
+      "epoch": 1.0436065885937351,
+      "grad_norm": 1.6812287999910398,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 43844
+    },
+    {
+      "epoch": 1.0436303913167666,
+      "grad_norm": 3.558824944232069,
+      "learning_rate": 2e-05,
+      "loss": 0.0455,
+      "step": 43845
+    },
+    {
+      "epoch": 1.043654194039798,
+      "grad_norm": 4.439494566028615,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 43846
+    },
+    {
+      "epoch": 1.0436779967628296,
+      "grad_norm": 6.608807596482354,
+      "learning_rate": 2e-05,
+      "loss": 0.3757,
+      "step": 43847
+    },
+    {
+      "epoch": 1.0437017994858613,
+      "grad_norm": 0.3512501989781327,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 43848
+    },
+    {
+      "epoch": 1.0437256022088928,
+      "grad_norm": 4.3392524193009,
+      "learning_rate": 2e-05,
+      "loss": 0.488,
+      "step": 43849
+    },
+    {
+      "epoch": 1.0437494049319243,
+      "grad_norm": 13.546004522096181,
+      "learning_rate": 2e-05,
+      "loss": 0.9935,
+      "step": 43850
+    },
+    {
+      "epoch": 1.0437732076549557,
+      "grad_norm": 0.019830005781628513,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43851
+    },
+    {
+      "epoch": 1.0437970103779872,
+      "grad_norm": 0.006717363301303234,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43852
+    },
+    {
+      "epoch": 1.0438208131010187,
+      "grad_norm": 1.9336178519936145,
+      "learning_rate": 2e-05,
+      "loss": 0.0248,
+      "step": 43853
+    },
+    {
+      "epoch": 1.0438446158240502,
+      "grad_norm": 2.6106196898023875,
+      "learning_rate": 2e-05,
+      "loss": 0.2494,
+      "step": 43854
+    },
+    {
+      "epoch": 1.0438684185470817,
+      "grad_norm": 2.13486851683728,
+      "learning_rate": 2e-05,
+      "loss": 0.2351,
+      "step": 43855
+    },
+    {
+      "epoch": 1.0438922212701134,
+      "grad_norm": 2.8095496272402944,
+      "learning_rate": 2e-05,
+      "loss": 0.2225,
+      "step": 43856
+    },
+    {
+      "epoch": 1.0439160239931449,
+      "grad_norm": 0.33562729067798525,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 43857
+    },
+    {
+      "epoch": 1.0439398267161764,
+      "grad_norm": 2.4459317142993156,
+      "learning_rate": 2e-05,
+      "loss": 0.0319,
+      "step": 43858
+    },
+    {
+      "epoch": 1.0439636294392078,
+      "grad_norm": 0.29188218535907945,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 43859
+    },
+    {
+      "epoch": 1.0439874321622393,
+      "grad_norm": 0.960438602139947,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 43860
+    },
+    {
+      "epoch": 1.0440112348852708,
+      "grad_norm": 1.5694355737745824,
+      "learning_rate": 2e-05,
+      "loss": 0.0609,
+      "step": 43861
+    },
+    {
+      "epoch": 1.0440350376083023,
+      "grad_norm": 4.566242032112864,
+      "learning_rate": 2e-05,
+      "loss": 0.1337,
+      "step": 43862
+    },
+    {
+      "epoch": 1.044058840331334,
+      "grad_norm": 0.04593426055311325,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 43863
+    },
+    {
+      "epoch": 1.0440826430543655,
+      "grad_norm": 2.437029183187474,
+      "learning_rate": 2e-05,
+      "loss": 0.0596,
+      "step": 43864
+    },
+    {
+      "epoch": 1.044106445777397,
+      "grad_norm": 0.028165611751252314,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43865
+    },
+    {
+      "epoch": 1.0441302485004285,
+      "grad_norm": 3.285666417566372,
+      "learning_rate": 2e-05,
+      "loss": 0.2911,
+      "step": 43866
+    },
+    {
+      "epoch": 1.04415405122346,
+      "grad_norm": 15.94048819672088,
+      "learning_rate": 2e-05,
+      "loss": 0.5712,
+      "step": 43867
+    },
+    {
+      "epoch": 1.0441778539464914,
+      "grad_norm": 9.671172370676144,
+      "learning_rate": 2e-05,
+      "loss": 0.7627,
+      "step": 43868
+    },
+    {
+      "epoch": 1.044201656669523,
+      "grad_norm": 4.471309956530969,
+      "learning_rate": 2e-05,
+      "loss": 0.5774,
+      "step": 43869
+    },
+    {
+      "epoch": 1.0442254593925546,
+      "grad_norm": 0.18454562571705233,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43870
+    },
+    {
+      "epoch": 1.044249262115586,
+      "grad_norm": 0.02434843882485834,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43871
+    },
+    {
+      "epoch": 1.0442730648386176,
+      "grad_norm": 0.3766767155415063,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 43872
+    },
+    {
+      "epoch": 1.044296867561649,
+      "grad_norm": 0.06660525146622494,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43873
+    },
+    {
+      "epoch": 1.0443206702846806,
+      "grad_norm": 0.020508500283304625,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 43874
+    },
+    {
+      "epoch": 1.044344473007712,
+      "grad_norm": 1.984928324749969,
+      "learning_rate": 2e-05,
+      "loss": 0.1075,
+      "step": 43875
+    },
+    {
+      "epoch": 1.0443682757307435,
+      "grad_norm": 3.5197659797661465,
+      "learning_rate": 2e-05,
+      "loss": 0.2603,
+      "step": 43876
+    },
+    {
+      "epoch": 1.044392078453775,
+      "grad_norm": 2.443866004566764,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 43877
+    },
+    {
+      "epoch": 1.0444158811768067,
+      "grad_norm": 1.1981931437007551,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 43878
+    },
+    {
+      "epoch": 1.0444396838998382,
+      "grad_norm": 5.336065978665478,
+      "learning_rate": 2e-05,
+      "loss": 0.6074,
+      "step": 43879
+    },
+    {
+      "epoch": 1.0444634866228697,
+      "grad_norm": 2.4882590203456214,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 43880
+    },
+    {
+      "epoch": 1.0444872893459012,
+      "grad_norm": 8.937463922067415,
+      "learning_rate": 2e-05,
+      "loss": 0.4582,
+      "step": 43881
+    },
+    {
+      "epoch": 1.0445110920689327,
+      "grad_norm": 6.593162725224257,
+      "learning_rate": 2e-05,
+      "loss": 0.8021,
+      "step": 43882
+    },
+    {
+      "epoch": 1.0445348947919642,
+      "grad_norm": 4.102138399919269,
+      "learning_rate": 2e-05,
+      "loss": 0.3759,
+      "step": 43883
+    },
+    {
+      "epoch": 1.0445586975149956,
+      "grad_norm": 3.888336823143534,
+      "learning_rate": 2e-05,
+      "loss": 0.3176,
+      "step": 43884
+    },
+    {
+      "epoch": 1.0445825002380271,
+      "grad_norm": 10.380372958666392,
+      "learning_rate": 2e-05,
+      "loss": 0.6912,
+      "step": 43885
+    },
+    {
+      "epoch": 1.0446063029610588,
+      "grad_norm": 1.1715066547815727,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 43886
+    },
+    {
+      "epoch": 1.0446301056840903,
+      "grad_norm": 1.037713366567206,
+      "learning_rate": 2e-05,
+      "loss": 0.0156,
+      "step": 43887
+    },
+    {
+      "epoch": 1.0446539084071218,
+      "grad_norm": 0.573505163647598,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 43888
+    },
+    {
+      "epoch": 1.0446777111301533,
+      "grad_norm": 0.04817992925554748,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43889
+    },
+    {
+      "epoch": 1.0447015138531848,
+      "grad_norm": 3.28272060048777,
+      "learning_rate": 2e-05,
+      "loss": 0.3017,
+      "step": 43890
+    },
+    {
+      "epoch": 1.0447253165762163,
+      "grad_norm": 2.1533931818441223,
+      "learning_rate": 2e-05,
+      "loss": 0.043,
+      "step": 43891
+    },
+    {
+      "epoch": 1.0447491192992477,
+      "grad_norm": 5.206869864721076,
+      "learning_rate": 2e-05,
+      "loss": 1.0619,
+      "step": 43892
+    },
+    {
+      "epoch": 1.0447729220222794,
+      "grad_norm": 5.209395418457949,
+      "learning_rate": 2e-05,
+      "loss": 0.5952,
+      "step": 43893
+    },
+    {
+      "epoch": 1.044796724745311,
+      "grad_norm": 1.6486779630166042,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 43894
+    },
+    {
+      "epoch": 1.0448205274683424,
+      "grad_norm": 4.781003032417008,
+      "learning_rate": 2e-05,
+      "loss": 0.8232,
+      "step": 43895
+    },
+    {
+      "epoch": 1.044844330191374,
+      "grad_norm": 1.889090323472931,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 43896
+    },
+    {
+      "epoch": 1.0448681329144054,
+      "grad_norm": 3.215586162658906,
+      "learning_rate": 2e-05,
+      "loss": 0.2592,
+      "step": 43897
+    },
+    {
+      "epoch": 1.0448919356374369,
+      "grad_norm": 2.287886454072014,
+      "learning_rate": 2e-05,
+      "loss": 0.2022,
+      "step": 43898
+    },
+    {
+      "epoch": 1.0449157383604684,
+      "grad_norm": 2.9031985308598984,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 43899
+    },
+    {
+      "epoch": 1.0449395410834998,
+      "grad_norm": 0.13109553944892138,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 43900
+    },
+    {
+      "epoch": 1.0449633438065316,
+      "grad_norm": 0.019043836761288725,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43901
+    },
+    {
+      "epoch": 1.044987146529563,
+      "grad_norm": 1.6703931368676428,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 43902
+    },
+    {
+      "epoch": 1.0450109492525945,
+      "grad_norm": 1.3575376771893954,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 43903
+    },
+    {
+      "epoch": 1.045034751975626,
+      "grad_norm": 0.15419988534442067,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 43904
+    },
+    {
+      "epoch": 1.0450585546986575,
+      "grad_norm": 0.07291302899832401,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43905
+    },
+    {
+      "epoch": 1.045082357421689,
+      "grad_norm": 9.551363695410704,
+      "learning_rate": 2e-05,
+      "loss": 0.3666,
+      "step": 43906
+    },
+    {
+      "epoch": 1.0451061601447205,
+      "grad_norm": 4.081813174944981,
+      "learning_rate": 2e-05,
+      "loss": 0.0599,
+      "step": 43907
+    },
+    {
+      "epoch": 1.0451299628677522,
+      "grad_norm": 6.1515133260841175,
+      "learning_rate": 2e-05,
+      "loss": 0.4093,
+      "step": 43908
+    },
+    {
+      "epoch": 1.0451537655907837,
+      "grad_norm": 0.014930066218807336,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43909
+    },
+    {
+      "epoch": 1.0451775683138151,
+      "grad_norm": 1.2215899754600439,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 43910
+    },
+    {
+      "epoch": 1.0452013710368466,
+      "grad_norm": 3.824411651965249,
+      "learning_rate": 2e-05,
+      "loss": 0.4917,
+      "step": 43911
+    },
+    {
+      "epoch": 1.0452251737598781,
+      "grad_norm": 3.775867312726173,
+      "learning_rate": 2e-05,
+      "loss": 0.6357,
+      "step": 43912
+    },
+    {
+      "epoch": 1.0452489764829096,
+      "grad_norm": 2.0365968912813286,
+      "learning_rate": 2e-05,
+      "loss": 0.1468,
+      "step": 43913
+    },
+    {
+      "epoch": 1.045272779205941,
+      "grad_norm": 0.042218789949297626,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 43914
+    },
+    {
+      "epoch": 1.0452965819289726,
+      "grad_norm": 0.0032335122456016753,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 43915
+    },
+    {
+      "epoch": 1.0453203846520043,
+      "grad_norm": 0.430615850166202,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 43916
+    },
+    {
+      "epoch": 1.0453441873750358,
+      "grad_norm": 0.13864118985812615,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 43917
+    },
+    {
+      "epoch": 1.0453679900980672,
+      "grad_norm": 0.06083649967422286,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43918
+    },
+    {
+      "epoch": 1.0453917928210987,
+      "grad_norm": 2.594846981744478,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 43919
+    },
+    {
+      "epoch": 1.0454155955441302,
+      "grad_norm": 3.4589704620227626,
+      "learning_rate": 2e-05,
+      "loss": 0.3818,
+      "step": 43920
+    },
+    {
+      "epoch": 1.0454393982671617,
+      "grad_norm": 3.6160148300788704,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 43921
+    },
+    {
+      "epoch": 1.0454632009901932,
+      "grad_norm": 0.035126343198178527,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 43922
+    },
+    {
+      "epoch": 1.045487003713225,
+      "grad_norm": 4.0719605484753965,
+      "learning_rate": 2e-05,
+      "loss": 0.4235,
+      "step": 43923
+    },
+    {
+      "epoch": 1.0455108064362564,
+      "grad_norm": 5.514703106560612,
+      "learning_rate": 2e-05,
+      "loss": 0.5663,
+      "step": 43924
+    },
+    {
+      "epoch": 1.0455346091592879,
+      "grad_norm": 15.80309570864874,
+      "learning_rate": 2e-05,
+      "loss": 0.1633,
+      "step": 43925
+    },
+    {
+      "epoch": 1.0455584118823193,
+      "grad_norm": 0.05724047851497922,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43926
+    },
+    {
+      "epoch": 1.0455822146053508,
+      "grad_norm": 8.663605153545118,
+      "learning_rate": 2e-05,
+      "loss": 0.4047,
+      "step": 43927
+    },
+    {
+      "epoch": 1.0456060173283823,
+      "grad_norm": 0.04035088011719376,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43928
+    },
+    {
+      "epoch": 1.0456298200514138,
+      "grad_norm": 5.734605606423056,
+      "learning_rate": 2e-05,
+      "loss": 0.6751,
+      "step": 43929
+    },
+    {
+      "epoch": 1.0456536227744453,
+      "grad_norm": 0.9367360193002802,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 43930
+    },
+    {
+      "epoch": 1.045677425497477,
+      "grad_norm": 3.3918441300618576,
+      "learning_rate": 2e-05,
+      "loss": 0.3419,
+      "step": 43931
+    },
+    {
+      "epoch": 1.0457012282205085,
+      "grad_norm": 0.3196443583560876,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 43932
+    },
+    {
+      "epoch": 1.04572503094354,
+      "grad_norm": 0.13882436845597876,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 43933
+    },
+    {
+      "epoch": 1.0457488336665715,
+      "grad_norm": 4.56718587647245,
+      "learning_rate": 2e-05,
+      "loss": 0.5543,
+      "step": 43934
+    },
+    {
+      "epoch": 1.045772636389603,
+      "grad_norm": 0.2649284896172801,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 43935
+    },
+    {
+      "epoch": 1.0457964391126344,
+      "grad_norm": 6.139081878808668,
+      "learning_rate": 2e-05,
+      "loss": 0.6738,
+      "step": 43936
+    },
+    {
+      "epoch": 1.045820241835666,
+      "grad_norm": 13.008749933746842,
+      "learning_rate": 2e-05,
+      "loss": 0.1425,
+      "step": 43937
+    },
+    {
+      "epoch": 1.0458440445586976,
+      "grad_norm": 5.933114840554793,
+      "learning_rate": 2e-05,
+      "loss": 0.0965,
+      "step": 43938
+    },
+    {
+      "epoch": 1.045867847281729,
+      "grad_norm": 3.516705929795478,
+      "learning_rate": 2e-05,
+      "loss": 0.3528,
+      "step": 43939
+    },
+    {
+      "epoch": 1.0458916500047606,
+      "grad_norm": 10580.108691930562,
+      "learning_rate": 2e-05,
+      "loss": 0.4668,
+      "step": 43940
+    },
+    {
+      "epoch": 1.045915452727792,
+      "grad_norm": 1.0050674223324065,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 43941
+    },
+    {
+      "epoch": 1.0459392554508236,
+      "grad_norm": 1.5116408441292464,
+      "learning_rate": 2e-05,
+      "loss": 0.0243,
+      "step": 43942
+    },
+    {
+      "epoch": 1.045963058173855,
+      "grad_norm": 9.061598814198506,
+      "learning_rate": 2e-05,
+      "loss": 0.1301,
+      "step": 43943
+    },
+    {
+      "epoch": 1.0459868608968865,
+      "grad_norm": 0.285384735452631,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 43944
+    },
+    {
+      "epoch": 1.046010663619918,
+      "grad_norm": 0.7868268473882827,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 43945
+    },
+    {
+      "epoch": 1.0460344663429497,
+      "grad_norm": 4.799334136132934,
+      "learning_rate": 2e-05,
+      "loss": 0.5448,
+      "step": 43946
+    },
+    {
+      "epoch": 1.0460582690659812,
+      "grad_norm": 4.179361730878672,
+      "learning_rate": 2e-05,
+      "loss": 0.9105,
+      "step": 43947
+    },
+    {
+      "epoch": 1.0460820717890127,
+      "grad_norm": 4.534868142284109,
+      "learning_rate": 2e-05,
+      "loss": 0.3461,
+      "step": 43948
+    },
+    {
+      "epoch": 1.0461058745120442,
+      "grad_norm": 9.4585629173267,
+      "learning_rate": 2e-05,
+      "loss": 0.6539,
+      "step": 43949
+    },
+    {
+      "epoch": 1.0461296772350757,
+      "grad_norm": 3.2010688461741776,
+      "learning_rate": 2e-05,
+      "loss": 0.6486,
+      "step": 43950
+    },
+    {
+      "epoch": 1.0461534799581071,
+      "grad_norm": 0.45791405722480083,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 43951
+    },
+    {
+      "epoch": 1.0461772826811386,
+      "grad_norm": 0.7199916870236394,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 43952
+    },
+    {
+      "epoch": 1.0462010854041703,
+      "grad_norm": 0.0858860543790603,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43953
+    },
+    {
+      "epoch": 1.0462248881272018,
+      "grad_norm": 9.510413939578221,
+      "learning_rate": 2e-05,
+      "loss": 0.1329,
+      "step": 43954
+    },
+    {
+      "epoch": 1.0462486908502333,
+      "grad_norm": 3.4951140300608636,
+      "learning_rate": 2e-05,
+      "loss": 0.2818,
+      "step": 43955
+    },
+    {
+      "epoch": 1.0462724935732648,
+      "grad_norm": 6.065626377494067,
+      "learning_rate": 2e-05,
+      "loss": 0.6034,
+      "step": 43956
+    },
+    {
+      "epoch": 1.0462962962962963,
+      "grad_norm": 12.314152694168023,
+      "learning_rate": 2e-05,
+      "loss": 0.9628,
+      "step": 43957
+    },
+    {
+      "epoch": 1.0463200990193278,
+      "grad_norm": 1.195373468603996,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 43958
+    },
+    {
+      "epoch": 1.0463439017423593,
+      "grad_norm": 2.890893629070151,
+      "learning_rate": 2e-05,
+      "loss": 0.0428,
+      "step": 43959
+    },
+    {
+      "epoch": 1.0463677044653907,
+      "grad_norm": 2.2093971725730737,
+      "learning_rate": 2e-05,
+      "loss": 0.1859,
+      "step": 43960
+    },
+    {
+      "epoch": 1.0463915071884224,
+      "grad_norm": 3.492478938039481,
+      "learning_rate": 2e-05,
+      "loss": 0.0366,
+      "step": 43961
+    },
+    {
+      "epoch": 1.046415309911454,
+      "grad_norm": 2.998640512109719,
+      "learning_rate": 2e-05,
+      "loss": 0.3429,
+      "step": 43962
+    },
+    {
+      "epoch": 1.0464391126344854,
+      "grad_norm": 11.788803206054189,
+      "learning_rate": 2e-05,
+      "loss": 0.1961,
+      "step": 43963
+    },
+    {
+      "epoch": 1.046462915357517,
+      "grad_norm": 1.9953284763169181,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 43964
+    },
+    {
+      "epoch": 1.0464867180805484,
+      "grad_norm": 0.008411840714695883,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43965
+    },
+    {
+      "epoch": 1.0465105208035799,
+      "grad_norm": 0.010093170880983883,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43966
+    },
+    {
+      "epoch": 1.0465343235266114,
+      "grad_norm": 0.04827550773502833,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 43967
+    },
+    {
+      "epoch": 1.046558126249643,
+      "grad_norm": 0.3450243913702315,
+      "learning_rate": 2e-05,
+      "loss": 0.0081,
+      "step": 43968
+    },
+    {
+      "epoch": 1.0465819289726745,
+      "grad_norm": 4.686396562257199,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 43969
+    },
+    {
+      "epoch": 1.046605731695706,
+      "grad_norm": 3.2827505742436713,
+      "learning_rate": 2e-05,
+      "loss": 0.3632,
+      "step": 43970
+    },
+    {
+      "epoch": 1.0466295344187375,
+      "grad_norm": 2.9897446980835136,
+      "learning_rate": 2e-05,
+      "loss": 0.0416,
+      "step": 43971
+    },
+    {
+      "epoch": 1.046653337141769,
+      "grad_norm": 0.053302897281128886,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 43972
+    },
+    {
+      "epoch": 1.0466771398648005,
+      "grad_norm": 5.630235647204708,
+      "learning_rate": 2e-05,
+      "loss": 0.2931,
+      "step": 43973
+    },
+    {
+      "epoch": 1.046700942587832,
+      "grad_norm": 0.016513381243351027,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 43974
+    },
+    {
+      "epoch": 1.0467247453108635,
+      "grad_norm": 4.074019667040502,
+      "learning_rate": 2e-05,
+      "loss": 0.3938,
+      "step": 43975
+    },
+    {
+      "epoch": 1.0467485480338952,
+      "grad_norm": 0.04520224083592767,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43976
+    },
+    {
+      "epoch": 1.0467723507569267,
+      "grad_norm": 9.569633463237297,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 43977
+    },
+    {
+      "epoch": 1.0467961534799581,
+      "grad_norm": 5.577971378906394,
+      "learning_rate": 2e-05,
+      "loss": 1.3434,
+      "step": 43978
+    },
+    {
+      "epoch": 1.0468199562029896,
+      "grad_norm": 0.23762817647492307,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 43979
+    },
+    {
+      "epoch": 1.046843758926021,
+      "grad_norm": 5.040428628220145,
+      "learning_rate": 2e-05,
+      "loss": 0.4328,
+      "step": 43980
+    },
+    {
+      "epoch": 1.0468675616490526,
+      "grad_norm": 7.80031705525068,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 43981
+    },
+    {
+      "epoch": 1.046891364372084,
+      "grad_norm": 2.5937224758064983,
+      "learning_rate": 2e-05,
+      "loss": 0.3444,
+      "step": 43982
+    },
+    {
+      "epoch": 1.0469151670951158,
+      "grad_norm": 7.4293810255148705,
+      "learning_rate": 2e-05,
+      "loss": 0.1679,
+      "step": 43983
+    },
+    {
+      "epoch": 1.0469389698181473,
+      "grad_norm": 0.25389208952158465,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 43984
+    },
+    {
+      "epoch": 1.0469627725411788,
+      "grad_norm": 2.9566898332923777,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 43985
+    },
+    {
+      "epoch": 1.0469865752642102,
+      "grad_norm": 7.5054746131452665,
+      "learning_rate": 2e-05,
+      "loss": 0.1309,
+      "step": 43986
+    },
+    {
+      "epoch": 1.0470103779872417,
+      "grad_norm": 0.0705447285400039,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 43987
+    },
+    {
+      "epoch": 1.0470341807102732,
+      "grad_norm": 14.16907408636055,
+      "learning_rate": 2e-05,
+      "loss": 0.4313,
+      "step": 43988
+    },
+    {
+      "epoch": 1.0470579834333047,
+      "grad_norm": 0.9889552250121485,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 43989
+    },
+    {
+      "epoch": 1.0470817861563362,
+      "grad_norm": 4.536737663382653,
+      "learning_rate": 2e-05,
+      "loss": 0.0832,
+      "step": 43990
+    },
+    {
+      "epoch": 1.0471055888793679,
+      "grad_norm": 4.186776012705878,
+      "learning_rate": 2e-05,
+      "loss": 0.2262,
+      "step": 43991
+    },
+    {
+      "epoch": 1.0471293916023994,
+      "grad_norm": 2.4213101266652184,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 43992
+    },
+    {
+      "epoch": 1.0471531943254309,
+      "grad_norm": 4.631693363668581,
+      "learning_rate": 2e-05,
+      "loss": 0.3186,
+      "step": 43993
+    },
+    {
+      "epoch": 1.0471769970484623,
+      "grad_norm": 0.012386575316571263,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 43994
+    },
+    {
+      "epoch": 1.0472007997714938,
+      "grad_norm": 2.737459965289557,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 43995
+    },
+    {
+      "epoch": 1.0472246024945253,
+      "grad_norm": 9.110731439124852,
+      "learning_rate": 2e-05,
+      "loss": 0.062,
+      "step": 43996
+    },
+    {
+      "epoch": 1.0472484052175568,
+      "grad_norm": 5.675060441446148,
+      "learning_rate": 2e-05,
+      "loss": 0.5086,
+      "step": 43997
+    },
+    {
+      "epoch": 1.0472722079405885,
+      "grad_norm": 3.822010625423844,
+      "learning_rate": 2e-05,
+      "loss": 0.0396,
+      "step": 43998
+    },
+    {
+      "epoch": 1.04729601066362,
+      "grad_norm": 2.0299357349854605,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 43999
+    },
+    {
+      "epoch": 1.0473198133866515,
+      "grad_norm": 1.764693331440729,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 44000
+    },
+    {
+      "epoch": 1.047343616109683,
+      "grad_norm": 10.571734882154258,
+      "learning_rate": 2e-05,
+      "loss": 0.6608,
+      "step": 44001
+    },
+    {
+      "epoch": 1.0473674188327144,
+      "grad_norm": 0.08896897500194488,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44002
+    },
+    {
+      "epoch": 1.047391221555746,
+      "grad_norm": 0.3005228185075326,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 44003
+    },
+    {
+      "epoch": 1.0474150242787774,
+      "grad_norm": 0.3608292743391849,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 44004
+    },
+    {
+      "epoch": 1.047438827001809,
+      "grad_norm": 9.068564085579988,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 44005
+    },
+    {
+      "epoch": 1.0474626297248406,
+      "grad_norm": 3.0026936962476034,
+      "learning_rate": 2e-05,
+      "loss": 0.0991,
+      "step": 44006
+    },
+    {
+      "epoch": 1.047486432447872,
+      "grad_norm": 0.20616901374630472,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 44007
+    },
+    {
+      "epoch": 1.0475102351709036,
+      "grad_norm": 0.014972024005530805,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44008
+    },
+    {
+      "epoch": 1.047534037893935,
+      "grad_norm": 0.5233142585132906,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 44009
+    },
+    {
+      "epoch": 1.0475578406169666,
+      "grad_norm": 10.417504079812655,
+      "learning_rate": 2e-05,
+      "loss": 1.3745,
+      "step": 44010
+    },
+    {
+      "epoch": 1.047581643339998,
+      "grad_norm": 0.14875612697694116,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44011
+    },
+    {
+      "epoch": 1.0476054460630295,
+      "grad_norm": 0.025998637169713498,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44012
+    },
+    {
+      "epoch": 1.0476292487860612,
+      "grad_norm": 3.821156443050749,
+      "learning_rate": 2e-05,
+      "loss": 0.1281,
+      "step": 44013
+    },
+    {
+      "epoch": 1.0476530515090927,
+      "grad_norm": 3.024292249373153,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 44014
+    },
+    {
+      "epoch": 1.0476768542321242,
+      "grad_norm": 3.217382594946894,
+      "learning_rate": 2e-05,
+      "loss": 0.083,
+      "step": 44015
+    },
+    {
+      "epoch": 1.0477006569551557,
+      "grad_norm": 0.009531627411456902,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44016
+    },
+    {
+      "epoch": 1.0477244596781872,
+      "grad_norm": 0.3839331658919753,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 44017
+    },
+    {
+      "epoch": 1.0477482624012187,
+      "grad_norm": 9.69798796028164,
+      "learning_rate": 2e-05,
+      "loss": 0.1215,
+      "step": 44018
+    },
+    {
+      "epoch": 1.0477720651242501,
+      "grad_norm": 0.05390637060602709,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44019
+    },
+    {
+      "epoch": 1.0477958678472816,
+      "grad_norm": 5.634781263383748,
+      "learning_rate": 2e-05,
+      "loss": 0.0897,
+      "step": 44020
+    },
+    {
+      "epoch": 1.0478196705703133,
+      "grad_norm": 15.803588958087166,
+      "learning_rate": 2e-05,
+      "loss": 0.6375,
+      "step": 44021
+    },
+    {
+      "epoch": 1.0478434732933448,
+      "grad_norm": 2.615801604189538,
+      "learning_rate": 2e-05,
+      "loss": 0.2216,
+      "step": 44022
+    },
+    {
+      "epoch": 1.0478672760163763,
+      "grad_norm": 3.929602485134495,
+      "learning_rate": 2e-05,
+      "loss": 0.4674,
+      "step": 44023
+    },
+    {
+      "epoch": 1.0478910787394078,
+      "grad_norm": 14.461375367482443,
+      "learning_rate": 2e-05,
+      "loss": 0.6649,
+      "step": 44024
+    },
+    {
+      "epoch": 1.0479148814624393,
+      "grad_norm": 3.8504116535574964,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 44025
+    },
+    {
+      "epoch": 1.0479386841854708,
+      "grad_norm": 2.2659476851551,
+      "learning_rate": 2e-05,
+      "loss": 0.1847,
+      "step": 44026
+    },
+    {
+      "epoch": 1.0479624869085022,
+      "grad_norm": 1.709966235029343,
+      "learning_rate": 2e-05,
+      "loss": 0.1625,
+      "step": 44027
+    },
+    {
+      "epoch": 1.047986289631534,
+      "grad_norm": 5.468246321395572,
+      "learning_rate": 2e-05,
+      "loss": 0.6142,
+      "step": 44028
+    },
+    {
+      "epoch": 1.0480100923545654,
+      "grad_norm": 6.922784791346569,
+      "learning_rate": 2e-05,
+      "loss": 1.4687,
+      "step": 44029
+    },
+    {
+      "epoch": 1.048033895077597,
+      "grad_norm": 12.073654663022516,
+      "learning_rate": 2e-05,
+      "loss": 1.0512,
+      "step": 44030
+    },
+    {
+      "epoch": 1.0480576978006284,
+      "grad_norm": 2.710636368973027,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 44031
+    },
+    {
+      "epoch": 1.04808150052366,
+      "grad_norm": 0.8458325661522342,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 44032
+    },
+    {
+      "epoch": 1.0481053032466914,
+      "grad_norm": 0.8316871283488362,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 44033
+    },
+    {
+      "epoch": 1.0481291059697229,
+      "grad_norm": 5.807764277152989,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 44034
+    },
+    {
+      "epoch": 1.0481529086927543,
+      "grad_norm": 9.665395036923012,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 44035
+    },
+    {
+      "epoch": 1.048176711415786,
+      "grad_norm": 7.563734478064725,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 44036
+    },
+    {
+      "epoch": 1.0482005141388175,
+      "grad_norm": 0.470007557010476,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 44037
+    },
+    {
+      "epoch": 1.048224316861849,
+      "grad_norm": 0.06182102956583959,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44038
+    },
+    {
+      "epoch": 1.0482481195848805,
+      "grad_norm": 4.088867634596568,
+      "learning_rate": 2e-05,
+      "loss": 0.514,
+      "step": 44039
+    },
+    {
+      "epoch": 1.048271922307912,
+      "grad_norm": 0.037479343147211144,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44040
+    },
+    {
+      "epoch": 1.0482957250309435,
+      "grad_norm": 0.1718341349455799,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 44041
+    },
+    {
+      "epoch": 1.048319527753975,
+      "grad_norm": 3.4817215277989315,
+      "learning_rate": 2e-05,
+      "loss": 0.5187,
+      "step": 44042
+    },
+    {
+      "epoch": 1.0483433304770067,
+      "grad_norm": 0.02079665708846417,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44043
+    },
+    {
+      "epoch": 1.0483671332000382,
+      "grad_norm": 0.6691690566791532,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 44044
+    },
+    {
+      "epoch": 1.0483909359230696,
+      "grad_norm": 3.8699061611067678,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 44045
+    },
+    {
+      "epoch": 1.0484147386461011,
+      "grad_norm": 2.4999389991230307,
+      "learning_rate": 2e-05,
+      "loss": 0.0264,
+      "step": 44046
+    },
+    {
+      "epoch": 1.0484385413691326,
+      "grad_norm": 2.940550298702186,
+      "learning_rate": 2e-05,
+      "loss": 0.2171,
+      "step": 44047
+    },
+    {
+      "epoch": 1.048462344092164,
+      "grad_norm": 3.375414520141851,
+      "learning_rate": 2e-05,
+      "loss": 0.3234,
+      "step": 44048
+    },
+    {
+      "epoch": 1.0484861468151956,
+      "grad_norm": 4.960805308272169,
+      "learning_rate": 2e-05,
+      "loss": 0.6262,
+      "step": 44049
+    },
+    {
+      "epoch": 1.048509949538227,
+      "grad_norm": 0.34360347012307624,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 44050
+    },
+    {
+      "epoch": 1.0485337522612588,
+      "grad_norm": 3.679777899859286,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 44051
+    },
+    {
+      "epoch": 1.0485575549842903,
+      "grad_norm": 0.7800380085874058,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 44052
+    },
+    {
+      "epoch": 1.0485813577073217,
+      "grad_norm": 0.08410448449460421,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44053
+    },
+    {
+      "epoch": 1.0486051604303532,
+      "grad_norm": 2.2799380030203884,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 44054
+    },
+    {
+      "epoch": 1.0486289631533847,
+      "grad_norm": 2.2679032222030036,
+      "learning_rate": 2e-05,
+      "loss": 0.175,
+      "step": 44055
+    },
+    {
+      "epoch": 1.0486527658764162,
+      "grad_norm": 3.2739341705344307,
+      "learning_rate": 2e-05,
+      "loss": 0.2751,
+      "step": 44056
+    },
+    {
+      "epoch": 1.0486765685994477,
+      "grad_norm": 4.216443674358591,
+      "learning_rate": 2e-05,
+      "loss": 0.2072,
+      "step": 44057
+    },
+    {
+      "epoch": 1.0487003713224794,
+      "grad_norm": 3.0370219257501088,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 44058
+    },
+    {
+      "epoch": 1.0487241740455109,
+      "grad_norm": 0.17615082189784112,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44059
+    },
+    {
+      "epoch": 1.0487479767685424,
+      "grad_norm": 3.4907352596529595,
+      "learning_rate": 2e-05,
+      "loss": 0.0302,
+      "step": 44060
+    },
+    {
+      "epoch": 1.0487717794915739,
+      "grad_norm": 2.5254579087180025,
+      "learning_rate": 2e-05,
+      "loss": 0.2901,
+      "step": 44061
+    },
+    {
+      "epoch": 1.0487955822146053,
+      "grad_norm": 3.802709047066893,
+      "learning_rate": 2e-05,
+      "loss": 0.4902,
+      "step": 44062
+    },
+    {
+      "epoch": 1.0488193849376368,
+      "grad_norm": 0.16815240287093058,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 44063
+    },
+    {
+      "epoch": 1.0488431876606683,
+      "grad_norm": 18.00609565175671,
+      "learning_rate": 2e-05,
+      "loss": 0.0816,
+      "step": 44064
+    },
+    {
+      "epoch": 1.0488669903836998,
+      "grad_norm": 0.14861166980043597,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 44065
+    },
+    {
+      "epoch": 1.0488907931067315,
+      "grad_norm": 0.169709335664908,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 44066
+    },
+    {
+      "epoch": 1.048914595829763,
+      "grad_norm": 8.579764376768148,
+      "learning_rate": 2e-05,
+      "loss": 0.3783,
+      "step": 44067
+    },
+    {
+      "epoch": 1.0489383985527945,
+      "grad_norm": 4.9350385119760345,
+      "learning_rate": 2e-05,
+      "loss": 0.6843,
+      "step": 44068
+    },
+    {
+      "epoch": 1.048962201275826,
+      "grad_norm": 0.038240195970373245,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44069
+    },
+    {
+      "epoch": 1.0489860039988574,
+      "grad_norm": 1.1541992737813742,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 44070
+    },
+    {
+      "epoch": 1.049009806721889,
+      "grad_norm": 0.33937945081863896,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44071
+    },
+    {
+      "epoch": 1.0490336094449204,
+      "grad_norm": 4.276046767722922,
+      "learning_rate": 2e-05,
+      "loss": 0.3991,
+      "step": 44072
+    },
+    {
+      "epoch": 1.0490574121679521,
+      "grad_norm": 1.0614134694882877,
+      "learning_rate": 2e-05,
+      "loss": 0.0151,
+      "step": 44073
+    },
+    {
+      "epoch": 1.0490812148909836,
+      "grad_norm": 5.386768336151782,
+      "learning_rate": 2e-05,
+      "loss": 0.8615,
+      "step": 44074
+    },
+    {
+      "epoch": 1.049105017614015,
+      "grad_norm": 3.568449064870162,
+      "learning_rate": 2e-05,
+      "loss": 0.2771,
+      "step": 44075
+    },
+    {
+      "epoch": 1.0491288203370466,
+      "grad_norm": 2.8054491009040627,
+      "learning_rate": 2e-05,
+      "loss": 0.2301,
+      "step": 44076
+    },
+    {
+      "epoch": 1.049152623060078,
+      "grad_norm": 3.0996926517622616,
+      "learning_rate": 2e-05,
+      "loss": 0.3683,
+      "step": 44077
+    },
+    {
+      "epoch": 1.0491764257831095,
+      "grad_norm": 1.5262631936300455,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 44078
+    },
+    {
+      "epoch": 1.049200228506141,
+      "grad_norm": 6.698371007132304,
+      "learning_rate": 2e-05,
+      "loss": 0.701,
+      "step": 44079
+    },
+    {
+      "epoch": 1.0492240312291725,
+      "grad_norm": 6.484403127050285,
+      "learning_rate": 2e-05,
+      "loss": 0.2288,
+      "step": 44080
+    },
+    {
+      "epoch": 1.0492478339522042,
+      "grad_norm": 0.005761913887669267,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44081
+    },
+    {
+      "epoch": 1.0492716366752357,
+      "grad_norm": 1.1521965511190855,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 44082
+    },
+    {
+      "epoch": 1.0492954393982672,
+      "grad_norm": 5.0723256300648485,
+      "learning_rate": 2e-05,
+      "loss": 0.7826,
+      "step": 44083
+    },
+    {
+      "epoch": 1.0493192421212987,
+      "grad_norm": 18.402034698678733,
+      "learning_rate": 2e-05,
+      "loss": 0.7482,
+      "step": 44084
+    },
+    {
+      "epoch": 1.0493430448443302,
+      "grad_norm": 4.583848236346955,
+      "learning_rate": 2e-05,
+      "loss": 0.6206,
+      "step": 44085
+    },
+    {
+      "epoch": 1.0493668475673616,
+      "grad_norm": 5.33360187086066,
+      "learning_rate": 2e-05,
+      "loss": 0.6323,
+      "step": 44086
+    },
+    {
+      "epoch": 1.0493906502903931,
+      "grad_norm": 0.016564163380996193,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44087
+    },
+    {
+      "epoch": 1.0494144530134248,
+      "grad_norm": 0.08229072403863284,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44088
+    },
+    {
+      "epoch": 1.0494382557364563,
+      "grad_norm": 0.9876888366817569,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 44089
+    },
+    {
+      "epoch": 1.0494620584594878,
+      "grad_norm": 2.2166964192494145,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 44090
+    },
+    {
+      "epoch": 1.0494858611825193,
+      "grad_norm": 12.602780519232425,
+      "learning_rate": 2e-05,
+      "loss": 0.9249,
+      "step": 44091
+    },
+    {
+      "epoch": 1.0495096639055508,
+      "grad_norm": 4.437973312550484,
+      "learning_rate": 2e-05,
+      "loss": 0.4101,
+      "step": 44092
+    },
+    {
+      "epoch": 1.0495334666285823,
+      "grad_norm": 3.219568576059812,
+      "learning_rate": 2e-05,
+      "loss": 0.2996,
+      "step": 44093
+    },
+    {
+      "epoch": 1.0495572693516138,
+      "grad_norm": 0.21064695449356807,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 44094
+    },
+    {
+      "epoch": 1.0495810720746452,
+      "grad_norm": 0.006427647272678806,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44095
+    },
+    {
+      "epoch": 1.049604874797677,
+      "grad_norm": 2.3929438840104207,
+      "learning_rate": 2e-05,
+      "loss": 0.0361,
+      "step": 44096
+    },
+    {
+      "epoch": 1.0496286775207084,
+      "grad_norm": 10.328094544405475,
+      "learning_rate": 2e-05,
+      "loss": 0.5122,
+      "step": 44097
+    },
+    {
+      "epoch": 1.04965248024374,
+      "grad_norm": 4.852697029810181,
+      "learning_rate": 2e-05,
+      "loss": 0.3237,
+      "step": 44098
+    },
+    {
+      "epoch": 1.0496762829667714,
+      "grad_norm": 2.167751928792867,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 44099
+    },
+    {
+      "epoch": 1.0497000856898029,
+      "grad_norm": 0.1513408454899502,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44100
+    },
+    {
+      "epoch": 1.0497238884128344,
+      "grad_norm": 0.00538132690636581,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44101
+    },
+    {
+      "epoch": 1.0497476911358659,
+      "grad_norm": 5.637041410047497,
+      "learning_rate": 2e-05,
+      "loss": 0.4855,
+      "step": 44102
+    },
+    {
+      "epoch": 1.0497714938588976,
+      "grad_norm": 0.5567461748111988,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 44103
+    },
+    {
+      "epoch": 1.049795296581929,
+      "grad_norm": 0.776302063827368,
+      "learning_rate": 2e-05,
+      "loss": 0.0154,
+      "step": 44104
+    },
+    {
+      "epoch": 1.0498190993049605,
+      "grad_norm": 3.683560852248956,
+      "learning_rate": 2e-05,
+      "loss": 0.1219,
+      "step": 44105
+    },
+    {
+      "epoch": 1.049842902027992,
+      "grad_norm": 3.4730504519029943,
+      "learning_rate": 2e-05,
+      "loss": 0.3167,
+      "step": 44106
+    },
+    {
+      "epoch": 1.0498667047510235,
+      "grad_norm": 4.053122648783415,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 44107
+    },
+    {
+      "epoch": 1.049890507474055,
+      "grad_norm": 1.082659337255419,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 44108
+    },
+    {
+      "epoch": 1.0499143101970865,
+      "grad_norm": 15.231181937500452,
+      "learning_rate": 2e-05,
+      "loss": 0.3007,
+      "step": 44109
+    },
+    {
+      "epoch": 1.049938112920118,
+      "grad_norm": 0.017858010201834185,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44110
+    },
+    {
+      "epoch": 1.0499619156431497,
+      "grad_norm": 3.9526463938838505,
+      "learning_rate": 2e-05,
+      "loss": 0.1714,
+      "step": 44111
+    },
+    {
+      "epoch": 1.0499857183661812,
+      "grad_norm": 3.983236647133372,
+      "learning_rate": 2e-05,
+      "loss": 0.2645,
+      "step": 44112
+    },
+    {
+      "epoch": 1.0500095210892126,
+      "grad_norm": 2.9140246019127494,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 44113
+    },
+    {
+      "epoch": 1.0500333238122441,
+      "grad_norm": 0.0615044089187702,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 44114
+    },
+    {
+      "epoch": 1.0500571265352756,
+      "grad_norm": 1.363899779704017,
+      "learning_rate": 2e-05,
+      "loss": 0.0125,
+      "step": 44115
+    },
+    {
+      "epoch": 1.050080929258307,
+      "grad_norm": 0.0532891507271516,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44116
+    },
+    {
+      "epoch": 1.0501047319813386,
+      "grad_norm": 4.647207124396165,
+      "learning_rate": 2e-05,
+      "loss": 0.2369,
+      "step": 44117
+    },
+    {
+      "epoch": 1.0501285347043703,
+      "grad_norm": 4.5872577660476415,
+      "learning_rate": 2e-05,
+      "loss": 0.6542,
+      "step": 44118
+    },
+    {
+      "epoch": 1.0501523374274018,
+      "grad_norm": 0.23494801834680515,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 44119
+    },
+    {
+      "epoch": 1.0501761401504333,
+      "grad_norm": 3.704673930299725,
+      "learning_rate": 2e-05,
+      "loss": 0.4438,
+      "step": 44120
+    },
+    {
+      "epoch": 1.0501999428734647,
+      "grad_norm": 4.405733875916863,
+      "learning_rate": 2e-05,
+      "loss": 0.3503,
+      "step": 44121
+    },
+    {
+      "epoch": 1.0502237455964962,
+      "grad_norm": 0.05023088458199612,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44122
+    },
+    {
+      "epoch": 1.0502475483195277,
+      "grad_norm": 0.7999282059263785,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 44123
+    },
+    {
+      "epoch": 1.0502713510425592,
+      "grad_norm": 3.204226968601663,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 44124
+    },
+    {
+      "epoch": 1.0502951537655907,
+      "grad_norm": 1.2887252555090614,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 44125
+    },
+    {
+      "epoch": 1.0503189564886224,
+      "grad_norm": 0.4870913176985327,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 44126
+    },
+    {
+      "epoch": 1.0503427592116539,
+      "grad_norm": 5.331290281103923,
+      "learning_rate": 2e-05,
+      "loss": 0.0624,
+      "step": 44127
+    },
+    {
+      "epoch": 1.0503665619346854,
+      "grad_norm": 13.502074127860881,
+      "learning_rate": 2e-05,
+      "loss": 0.3212,
+      "step": 44128
+    },
+    {
+      "epoch": 1.0503903646577168,
+      "grad_norm": 0.18673320929499823,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44129
+    },
+    {
+      "epoch": 1.0504141673807483,
+      "grad_norm": 3.4258691309853098,
+      "learning_rate": 2e-05,
+      "loss": 0.3373,
+      "step": 44130
+    },
+    {
+      "epoch": 1.0504379701037798,
+      "grad_norm": 8.213292242810324,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 44131
+    },
+    {
+      "epoch": 1.0504617728268113,
+      "grad_norm": 4.323269099575004,
+      "learning_rate": 2e-05,
+      "loss": 0.4788,
+      "step": 44132
+    },
+    {
+      "epoch": 1.050485575549843,
+      "grad_norm": 4.277673132389688,
+      "learning_rate": 2e-05,
+      "loss": 0.4611,
+      "step": 44133
+    },
+    {
+      "epoch": 1.0505093782728745,
+      "grad_norm": 0.03270280238348921,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44134
+    },
+    {
+      "epoch": 1.050533180995906,
+      "grad_norm": 3.224602106502383,
+      "learning_rate": 2e-05,
+      "loss": 0.3217,
+      "step": 44135
+    },
+    {
+      "epoch": 1.0505569837189375,
+      "grad_norm": 0.32638156964470855,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 44136
+    },
+    {
+      "epoch": 1.050580786441969,
+      "grad_norm": 6.647711536414831,
+      "learning_rate": 2e-05,
+      "loss": 0.6829,
+      "step": 44137
+    },
+    {
+      "epoch": 1.0506045891650004,
+      "grad_norm": 0.12315154330970117,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44138
+    },
+    {
+      "epoch": 1.050628391888032,
+      "grad_norm": 6.070518891039312,
+      "learning_rate": 2e-05,
+      "loss": 0.7627,
+      "step": 44139
+    },
+    {
+      "epoch": 1.0506521946110634,
+      "grad_norm": 0.12068060221843552,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44140
+    },
+    {
+      "epoch": 1.0506759973340951,
+      "grad_norm": 56.202740555692216,
+      "learning_rate": 2e-05,
+      "loss": 0.5705,
+      "step": 44141
+    },
+    {
+      "epoch": 1.0506998000571266,
+      "grad_norm": 5.032084433005284,
+      "learning_rate": 2e-05,
+      "loss": 0.7479,
+      "step": 44142
+    },
+    {
+      "epoch": 1.050723602780158,
+      "grad_norm": 0.20935819612301237,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44143
+    },
+    {
+      "epoch": 1.0507474055031896,
+      "grad_norm": 0.08597752938319918,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44144
+    },
+    {
+      "epoch": 1.050771208226221,
+      "grad_norm": 43.04061522030002,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 44145
+    },
+    {
+      "epoch": 1.0507950109492525,
+      "grad_norm": 7.11143612567595,
+      "learning_rate": 2e-05,
+      "loss": 0.5169,
+      "step": 44146
+    },
+    {
+      "epoch": 1.050818813672284,
+      "grad_norm": 0.16558601127310787,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 44147
+    },
+    {
+      "epoch": 1.0508426163953155,
+      "grad_norm": 0.026937431295723544,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44148
+    },
+    {
+      "epoch": 1.0508664191183472,
+      "grad_norm": 4.164333696338786,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 44149
+    },
+    {
+      "epoch": 1.0508902218413787,
+      "grad_norm": 0.019604975987015105,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44150
+    },
+    {
+      "epoch": 1.0509140245644102,
+      "grad_norm": 6.840933190325524,
+      "learning_rate": 2e-05,
+      "loss": 0.6278,
+      "step": 44151
+    },
+    {
+      "epoch": 1.0509378272874417,
+      "grad_norm": 4.738589651836026,
+      "learning_rate": 2e-05,
+      "loss": 0.4491,
+      "step": 44152
+    },
+    {
+      "epoch": 1.0509616300104732,
+      "grad_norm": 7.215797330880582,
+      "learning_rate": 2e-05,
+      "loss": 0.5685,
+      "step": 44153
+    },
+    {
+      "epoch": 1.0509854327335046,
+      "grad_norm": 5.009142930330889,
+      "learning_rate": 2e-05,
+      "loss": 0.4908,
+      "step": 44154
+    },
+    {
+      "epoch": 1.0510092354565361,
+      "grad_norm": 2.114518117558548,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 44155
+    },
+    {
+      "epoch": 1.0510330381795678,
+      "grad_norm": 5.679425222189179,
+      "learning_rate": 2e-05,
+      "loss": 0.599,
+      "step": 44156
+    },
+    {
+      "epoch": 1.0510568409025993,
+      "grad_norm": 2.604092676779969,
+      "learning_rate": 2e-05,
+      "loss": 0.103,
+      "step": 44157
+    },
+    {
+      "epoch": 1.0510806436256308,
+      "grad_norm": 6.593050221293145,
+      "learning_rate": 2e-05,
+      "loss": 0.5069,
+      "step": 44158
+    },
+    {
+      "epoch": 1.0511044463486623,
+      "grad_norm": 0.1321438542754887,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 44159
+    },
+    {
+      "epoch": 1.0511282490716938,
+      "grad_norm": 3.2696030173535053,
+      "learning_rate": 2e-05,
+      "loss": 0.4973,
+      "step": 44160
+    },
+    {
+      "epoch": 1.0511520517947253,
+      "grad_norm": 0.25532775652159895,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 44161
+    },
+    {
+      "epoch": 1.0511758545177567,
+      "grad_norm": 3.2942184926008062,
+      "learning_rate": 2e-05,
+      "loss": 0.2055,
+      "step": 44162
+    },
+    {
+      "epoch": 1.0511996572407885,
+      "grad_norm": 0.38660476632876517,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 44163
+    },
+    {
+      "epoch": 1.05122345996382,
+      "grad_norm": 8.457446027444345,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 44164
+    },
+    {
+      "epoch": 1.0512472626868514,
+      "grad_norm": 0.10741477711314239,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44165
+    },
+    {
+      "epoch": 1.051271065409883,
+      "grad_norm": 2.1175784718560595,
+      "learning_rate": 2e-05,
+      "loss": 0.1747,
+      "step": 44166
+    },
+    {
+      "epoch": 1.0512948681329144,
+      "grad_norm": 0.13673940857969913,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44167
+    },
+    {
+      "epoch": 1.0513186708559459,
+      "grad_norm": 3.217800419077061,
+      "learning_rate": 2e-05,
+      "loss": 0.0384,
+      "step": 44168
+    },
+    {
+      "epoch": 1.0513424735789774,
+      "grad_norm": 4.155746322097752,
+      "learning_rate": 2e-05,
+      "loss": 0.4445,
+      "step": 44169
+    },
+    {
+      "epoch": 1.0513662763020089,
+      "grad_norm": 3.0651681496702348,
+      "learning_rate": 2e-05,
+      "loss": 0.3019,
+      "step": 44170
+    },
+    {
+      "epoch": 1.0513900790250406,
+      "grad_norm": 16.831638650272374,
+      "learning_rate": 2e-05,
+      "loss": 0.7217,
+      "step": 44171
+    },
+    {
+      "epoch": 1.051413881748072,
+      "grad_norm": 4.1946375186934794,
+      "learning_rate": 2e-05,
+      "loss": 0.6595,
+      "step": 44172
+    },
+    {
+      "epoch": 1.0514376844711035,
+      "grad_norm": 10.641186187618317,
+      "learning_rate": 2e-05,
+      "loss": 0.806,
+      "step": 44173
+    },
+    {
+      "epoch": 1.051461487194135,
+      "grad_norm": 1.6995823236337448,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 44174
+    },
+    {
+      "epoch": 1.0514852899171665,
+      "grad_norm": 0.04998527091067612,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44175
+    },
+    {
+      "epoch": 1.051509092640198,
+      "grad_norm": 0.6982753674225984,
+      "learning_rate": 2e-05,
+      "loss": 0.0082,
+      "step": 44176
+    },
+    {
+      "epoch": 1.0515328953632295,
+      "grad_norm": 0.011120531242402299,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44177
+    },
+    {
+      "epoch": 1.051556698086261,
+      "grad_norm": 4.618049971940441,
+      "learning_rate": 2e-05,
+      "loss": 0.7938,
+      "step": 44178
+    },
+    {
+      "epoch": 1.0515805008092927,
+      "grad_norm": 0.2900322731278906,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 44179
+    },
+    {
+      "epoch": 1.0516043035323241,
+      "grad_norm": 0.09229126908308548,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44180
+    },
+    {
+      "epoch": 1.0516281062553556,
+      "grad_norm": 4.2900390744988135,
+      "learning_rate": 2e-05,
+      "loss": 0.2374,
+      "step": 44181
+    },
+    {
+      "epoch": 1.0516519089783871,
+      "grad_norm": 7.405826483821777,
+      "learning_rate": 2e-05,
+      "loss": 0.2393,
+      "step": 44182
+    },
+    {
+      "epoch": 1.0516757117014186,
+      "grad_norm": 13.438666215456061,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 44183
+    },
+    {
+      "epoch": 1.05169951442445,
+      "grad_norm": 13.9172064220242,
+      "learning_rate": 2e-05,
+      "loss": 0.7158,
+      "step": 44184
+    },
+    {
+      "epoch": 1.0517233171474816,
+      "grad_norm": 6.502466356340714,
+      "learning_rate": 2e-05,
+      "loss": 1.0049,
+      "step": 44185
+    },
+    {
+      "epoch": 1.0517471198705133,
+      "grad_norm": 0.10027439499786209,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44186
+    },
+    {
+      "epoch": 1.0517709225935448,
+      "grad_norm": 10.296519091744505,
+      "learning_rate": 2e-05,
+      "loss": 0.9104,
+      "step": 44187
+    },
+    {
+      "epoch": 1.0517947253165763,
+      "grad_norm": 9.906822522741704,
+      "learning_rate": 2e-05,
+      "loss": 1.0731,
+      "step": 44188
+    },
+    {
+      "epoch": 1.0518185280396077,
+      "grad_norm": 4.26898596912615,
+      "learning_rate": 2e-05,
+      "loss": 0.3062,
+      "step": 44189
+    },
+    {
+      "epoch": 1.0518423307626392,
+      "grad_norm": 5.9641423387938515,
+      "learning_rate": 2e-05,
+      "loss": 0.0837,
+      "step": 44190
+    },
+    {
+      "epoch": 1.0518661334856707,
+      "grad_norm": 0.8635230247658287,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 44191
+    },
+    {
+      "epoch": 1.0518899362087022,
+      "grad_norm": 0.36878239662152523,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 44192
+    },
+    {
+      "epoch": 1.051913738931734,
+      "grad_norm": 0.32453147753185835,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 44193
+    },
+    {
+      "epoch": 1.0519375416547654,
+      "grad_norm": 0.509874762690742,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 44194
+    },
+    {
+      "epoch": 1.0519613443777969,
+      "grad_norm": 8.668885448570919,
+      "learning_rate": 2e-05,
+      "loss": 0.6055,
+      "step": 44195
+    },
+    {
+      "epoch": 1.0519851471008284,
+      "grad_norm": 3.3869432120604666,
+      "learning_rate": 2e-05,
+      "loss": 0.3951,
+      "step": 44196
+    },
+    {
+      "epoch": 1.0520089498238598,
+      "grad_norm": 5.624695778677072,
+      "learning_rate": 2e-05,
+      "loss": 0.2737,
+      "step": 44197
+    },
+    {
+      "epoch": 1.0520327525468913,
+      "grad_norm": 6.16253773800436,
+      "learning_rate": 2e-05,
+      "loss": 0.821,
+      "step": 44198
+    },
+    {
+      "epoch": 1.0520565552699228,
+      "grad_norm": 0.5939938674189976,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 44199
+    },
+    {
+      "epoch": 1.0520803579929543,
+      "grad_norm": 0.6020726501832195,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 44200
+    },
+    {
+      "epoch": 1.052104160715986,
+      "grad_norm": 7.156127495453876,
+      "learning_rate": 2e-05,
+      "loss": 0.0653,
+      "step": 44201
+    },
+    {
+      "epoch": 1.0521279634390175,
+      "grad_norm": 6.093421706994958,
+      "learning_rate": 2e-05,
+      "loss": 0.8212,
+      "step": 44202
+    },
+    {
+      "epoch": 1.052151766162049,
+      "grad_norm": 0.004676616626718496,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44203
+    },
+    {
+      "epoch": 1.0521755688850805,
+      "grad_norm": 3.2095576648863964,
+      "learning_rate": 2e-05,
+      "loss": 0.1135,
+      "step": 44204
+    },
+    {
+      "epoch": 1.052199371608112,
+      "grad_norm": 9.146553280778521,
+      "learning_rate": 2e-05,
+      "loss": 0.4051,
+      "step": 44205
+    },
+    {
+      "epoch": 1.0522231743311434,
+      "grad_norm": 2.4499536844987926,
+      "learning_rate": 2e-05,
+      "loss": 0.191,
+      "step": 44206
+    },
+    {
+      "epoch": 1.052246977054175,
+      "grad_norm": 1.0760730057922925,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 44207
+    },
+    {
+      "epoch": 1.0522707797772064,
+      "grad_norm": 0.009427188779834277,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44208
+    },
+    {
+      "epoch": 1.052294582500238,
+      "grad_norm": 3.2541968594661435,
+      "learning_rate": 2e-05,
+      "loss": 0.2457,
+      "step": 44209
+    },
+    {
+      "epoch": 1.0523183852232696,
+      "grad_norm": 0.05178308054277903,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44210
+    },
+    {
+      "epoch": 1.052342187946301,
+      "grad_norm": 2.706633047103706,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 44211
+    },
+    {
+      "epoch": 1.0523659906693326,
+      "grad_norm": 2.972991545258773,
+      "learning_rate": 2e-05,
+      "loss": 0.2701,
+      "step": 44212
+    },
+    {
+      "epoch": 1.052389793392364,
+      "grad_norm": 4.606708800630045,
+      "learning_rate": 2e-05,
+      "loss": 0.724,
+      "step": 44213
+    },
+    {
+      "epoch": 1.0524135961153955,
+      "grad_norm": 1.1589744889751494,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 44214
+    },
+    {
+      "epoch": 1.052437398838427,
+      "grad_norm": 0.057927644625904204,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44215
+    },
+    {
+      "epoch": 1.0524612015614587,
+      "grad_norm": 7.457124978982094,
+      "learning_rate": 2e-05,
+      "loss": 0.2286,
+      "step": 44216
+    },
+    {
+      "epoch": 1.0524850042844902,
+      "grad_norm": 0.48680531810706984,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 44217
+    },
+    {
+      "epoch": 1.0525088070075217,
+      "grad_norm": 0.3522868601762493,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 44218
+    },
+    {
+      "epoch": 1.0525326097305532,
+      "grad_norm": 1.3675498725364896,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 44219
+    },
+    {
+      "epoch": 1.0525564124535847,
+      "grad_norm": 4.102603605397563,
+      "learning_rate": 2e-05,
+      "loss": 0.4824,
+      "step": 44220
+    },
+    {
+      "epoch": 1.0525802151766162,
+      "grad_norm": 7.087074498004337,
+      "learning_rate": 2e-05,
+      "loss": 0.9275,
+      "step": 44221
+    },
+    {
+      "epoch": 1.0526040178996476,
+      "grad_norm": 3.4324623692898935,
+      "learning_rate": 2e-05,
+      "loss": 0.07,
+      "step": 44222
+    },
+    {
+      "epoch": 1.0526278206226793,
+      "grad_norm": 1.1621746391090446,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 44223
+    },
+    {
+      "epoch": 1.0526516233457108,
+      "grad_norm": 3.28182675798117,
+      "learning_rate": 2e-05,
+      "loss": 0.332,
+      "step": 44224
+    },
+    {
+      "epoch": 1.0526754260687423,
+      "grad_norm": 3.8811124059099,
+      "learning_rate": 2e-05,
+      "loss": 0.4399,
+      "step": 44225
+    },
+    {
+      "epoch": 1.0526992287917738,
+      "grad_norm": 0.5940893499018794,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 44226
+    },
+    {
+      "epoch": 1.0527230315148053,
+      "grad_norm": 6.4841794190295,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 44227
+    },
+    {
+      "epoch": 1.0527468342378368,
+      "grad_norm": 0.18139702567619193,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44228
+    },
+    {
+      "epoch": 1.0527706369608683,
+      "grad_norm": 6.302535542924335,
+      "learning_rate": 2e-05,
+      "loss": 0.4494,
+      "step": 44229
+    },
+    {
+      "epoch": 1.0527944396838997,
+      "grad_norm": 0.1915928049432795,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 44230
+    },
+    {
+      "epoch": 1.0528182424069314,
+      "grad_norm": 0.21420190310120174,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44231
+    },
+    {
+      "epoch": 1.052842045129963,
+      "grad_norm": 8.900738499575734,
+      "learning_rate": 2e-05,
+      "loss": 0.3204,
+      "step": 44232
+    },
+    {
+      "epoch": 1.0528658478529944,
+      "grad_norm": 8.31708958339111,
+      "learning_rate": 2e-05,
+      "loss": 0.2445,
+      "step": 44233
+    },
+    {
+      "epoch": 1.052889650576026,
+      "grad_norm": 0.11150026802860588,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44234
+    },
+    {
+      "epoch": 1.0529134532990574,
+      "grad_norm": 0.01615795108258645,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44235
+    },
+    {
+      "epoch": 1.0529372560220889,
+      "grad_norm": 4.221661832058543,
+      "learning_rate": 2e-05,
+      "loss": 0.1441,
+      "step": 44236
+    },
+    {
+      "epoch": 1.0529610587451204,
+      "grad_norm": 5.74058896509404,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 44237
+    },
+    {
+      "epoch": 1.0529848614681518,
+      "grad_norm": 0.2324650270639326,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 44238
+    },
+    {
+      "epoch": 1.0530086641911836,
+      "grad_norm": 4.518582098016148,
+      "learning_rate": 2e-05,
+      "loss": 0.4428,
+      "step": 44239
+    },
+    {
+      "epoch": 1.053032466914215,
+      "grad_norm": 0.06908592049561121,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44240
+    },
+    {
+      "epoch": 1.0530562696372465,
+      "grad_norm": 5.169758180426145,
+      "learning_rate": 2e-05,
+      "loss": 0.3458,
+      "step": 44241
+    },
+    {
+      "epoch": 1.053080072360278,
+      "grad_norm": 0.15437494282000067,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 44242
+    },
+    {
+      "epoch": 1.0531038750833095,
+      "grad_norm": 0.8197043695887707,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 44243
+    },
+    {
+      "epoch": 1.053127677806341,
+      "grad_norm": 1.9109016111686734,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 44244
+    },
+    {
+      "epoch": 1.0531514805293725,
+      "grad_norm": 6.810780784321841,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 44245
+    },
+    {
+      "epoch": 1.0531752832524042,
+      "grad_norm": 18.77103163891519,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 44246
+    },
+    {
+      "epoch": 1.0531990859754357,
+      "grad_norm": 2.7758145895259143,
+      "learning_rate": 2e-05,
+      "loss": 0.2832,
+      "step": 44247
+    },
+    {
+      "epoch": 1.0532228886984671,
+      "grad_norm": 0.017508935590377552,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44248
+    },
+    {
+      "epoch": 1.0532466914214986,
+      "grad_norm": 3.325919508905273,
+      "learning_rate": 2e-05,
+      "loss": 0.0228,
+      "step": 44249
+    },
+    {
+      "epoch": 1.0532704941445301,
+      "grad_norm": 3.7782234036486964,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 44250
+    },
+    {
+      "epoch": 1.0532942968675616,
+      "grad_norm": 5.7956893857800384,
+      "learning_rate": 2e-05,
+      "loss": 1.3164,
+      "step": 44251
+    },
+    {
+      "epoch": 1.053318099590593,
+      "grad_norm": 0.03634058765112157,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44252
+    },
+    {
+      "epoch": 1.0533419023136248,
+      "grad_norm": 3.3431576800618052,
+      "learning_rate": 2e-05,
+      "loss": 0.5904,
+      "step": 44253
+    },
+    {
+      "epoch": 1.0533657050366563,
+      "grad_norm": 5.047431558878642,
+      "learning_rate": 2e-05,
+      "loss": 0.6438,
+      "step": 44254
+    },
+    {
+      "epoch": 1.0533895077596878,
+      "grad_norm": 3.34900511494492,
+      "learning_rate": 2e-05,
+      "loss": 0.3884,
+      "step": 44255
+    },
+    {
+      "epoch": 1.0534133104827192,
+      "grad_norm": 4.26224374659298,
+      "learning_rate": 2e-05,
+      "loss": 0.3242,
+      "step": 44256
+    },
+    {
+      "epoch": 1.0534371132057507,
+      "grad_norm": 4.233781240920194,
+      "learning_rate": 2e-05,
+      "loss": 0.1323,
+      "step": 44257
+    },
+    {
+      "epoch": 1.0534609159287822,
+      "grad_norm": 4.533939171111295,
+      "learning_rate": 2e-05,
+      "loss": 0.8034,
+      "step": 44258
+    },
+    {
+      "epoch": 1.0534847186518137,
+      "grad_norm": 2.409263254017026,
+      "learning_rate": 2e-05,
+      "loss": 0.1368,
+      "step": 44259
+    },
+    {
+      "epoch": 1.0535085213748452,
+      "grad_norm": 0.037073423692725774,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44260
+    },
+    {
+      "epoch": 1.053532324097877,
+      "grad_norm": 14.235233444696272,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 44261
+    },
+    {
+      "epoch": 1.0535561268209084,
+      "grad_norm": 0.5366378725925106,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 44262
+    },
+    {
+      "epoch": 1.0535799295439399,
+      "grad_norm": 3.861820026288808,
+      "learning_rate": 2e-05,
+      "loss": 0.4921,
+      "step": 44263
+    },
+    {
+      "epoch": 1.0536037322669713,
+      "grad_norm": 5.592958760946398,
+      "learning_rate": 2e-05,
+      "loss": 0.9253,
+      "step": 44264
+    },
+    {
+      "epoch": 1.0536275349900028,
+      "grad_norm": 6.985933362863116,
+      "learning_rate": 2e-05,
+      "loss": 0.8596,
+      "step": 44265
+    },
+    {
+      "epoch": 1.0536513377130343,
+      "grad_norm": 0.1202318735704724,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 44266
+    },
+    {
+      "epoch": 1.0536751404360658,
+      "grad_norm": 1.379082846287287,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 44267
+    },
+    {
+      "epoch": 1.0536989431590973,
+      "grad_norm": 0.8490442611392707,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 44268
+    },
+    {
+      "epoch": 1.053722745882129,
+      "grad_norm": 9.216935589947164,
+      "learning_rate": 2e-05,
+      "loss": 1.6785,
+      "step": 44269
+    },
+    {
+      "epoch": 1.0537465486051605,
+      "grad_norm": 4.778196022953446,
+      "learning_rate": 2e-05,
+      "loss": 0.6733,
+      "step": 44270
+    },
+    {
+      "epoch": 1.053770351328192,
+      "grad_norm": 2.601842799944885,
+      "learning_rate": 2e-05,
+      "loss": 0.159,
+      "step": 44271
+    },
+    {
+      "epoch": 1.0537941540512235,
+      "grad_norm": 0.18140174377906107,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44272
+    },
+    {
+      "epoch": 1.053817956774255,
+      "grad_norm": 3.4560518447687643,
+      "learning_rate": 2e-05,
+      "loss": 0.5175,
+      "step": 44273
+    },
+    {
+      "epoch": 1.0538417594972864,
+      "grad_norm": 4.689944265865147,
+      "learning_rate": 2e-05,
+      "loss": 0.6392,
+      "step": 44274
+    },
+    {
+      "epoch": 1.053865562220318,
+      "grad_norm": 5.195514144005946,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 44275
+    },
+    {
+      "epoch": 1.0538893649433496,
+      "grad_norm": 3.7661740138886106,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 44276
+    },
+    {
+      "epoch": 1.053913167666381,
+      "grad_norm": 0.6155889314287443,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 44277
+    },
+    {
+      "epoch": 1.0539369703894126,
+      "grad_norm": 4.274570259354302,
+      "learning_rate": 2e-05,
+      "loss": 0.5227,
+      "step": 44278
+    },
+    {
+      "epoch": 1.053960773112444,
+      "grad_norm": 3.4105274104590513,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 44279
+    },
+    {
+      "epoch": 1.0539845758354756,
+      "grad_norm": 4.812237568800555,
+      "learning_rate": 2e-05,
+      "loss": 0.1373,
+      "step": 44280
+    },
+    {
+      "epoch": 1.054008378558507,
+      "grad_norm": 3.6075047451079754,
+      "learning_rate": 2e-05,
+      "loss": 0.721,
+      "step": 44281
+    },
+    {
+      "epoch": 1.0540321812815385,
+      "grad_norm": 5.204115041512039,
+      "learning_rate": 2e-05,
+      "loss": 0.4625,
+      "step": 44282
+    },
+    {
+      "epoch": 1.0540559840045702,
+      "grad_norm": 4.941352806882416,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 44283
+    },
+    {
+      "epoch": 1.0540797867276017,
+      "grad_norm": 5.379093089623145,
+      "learning_rate": 2e-05,
+      "loss": 0.4666,
+      "step": 44284
+    },
+    {
+      "epoch": 1.0541035894506332,
+      "grad_norm": 0.08977116173103007,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44285
+    },
+    {
+      "epoch": 1.0541273921736647,
+      "grad_norm": 0.005497311810711939,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44286
+    },
+    {
+      "epoch": 1.0541511948966962,
+      "grad_norm": 3.216094208937364,
+      "learning_rate": 2e-05,
+      "loss": 0.0787,
+      "step": 44287
+    },
+    {
+      "epoch": 1.0541749976197277,
+      "grad_norm": 2.556971225424701,
+      "learning_rate": 2e-05,
+      "loss": 0.2143,
+      "step": 44288
+    },
+    {
+      "epoch": 1.0541988003427591,
+      "grad_norm": 1.9314501258046668,
+      "learning_rate": 2e-05,
+      "loss": 0.1225,
+      "step": 44289
+    },
+    {
+      "epoch": 1.0542226030657906,
+      "grad_norm": 13.649595945857135,
+      "learning_rate": 2e-05,
+      "loss": 0.9159,
+      "step": 44290
+    },
+    {
+      "epoch": 1.0542464057888223,
+      "grad_norm": 3.5407201114553426,
+      "learning_rate": 2e-05,
+      "loss": 0.4021,
+      "step": 44291
+    },
+    {
+      "epoch": 1.0542702085118538,
+      "grad_norm": 10.721981795649443,
+      "learning_rate": 2e-05,
+      "loss": 1.1716,
+      "step": 44292
+    },
+    {
+      "epoch": 1.0542940112348853,
+      "grad_norm": 2.3234313342125916,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 44293
+    },
+    {
+      "epoch": 1.0543178139579168,
+      "grad_norm": 2.1550158544030924,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 44294
+    },
+    {
+      "epoch": 1.0543416166809483,
+      "grad_norm": 2.8679650055524,
+      "learning_rate": 2e-05,
+      "loss": 0.2602,
+      "step": 44295
+    },
+    {
+      "epoch": 1.0543654194039798,
+      "grad_norm": 4.354997645739889,
+      "learning_rate": 2e-05,
+      "loss": 0.1767,
+      "step": 44296
+    },
+    {
+      "epoch": 1.0543892221270112,
+      "grad_norm": 0.1449832790916996,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44297
+    },
+    {
+      "epoch": 1.0544130248500427,
+      "grad_norm": 3.6825590899779153,
+      "learning_rate": 2e-05,
+      "loss": 0.2296,
+      "step": 44298
+    },
+    {
+      "epoch": 1.0544368275730744,
+      "grad_norm": 0.07623002346398557,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44299
+    },
+    {
+      "epoch": 1.054460630296106,
+      "grad_norm": 0.1640853431941195,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44300
+    },
+    {
+      "epoch": 1.0544844330191374,
+      "grad_norm": 5.303933243631947,
+      "learning_rate": 2e-05,
+      "loss": 0.3756,
+      "step": 44301
+    },
+    {
+      "epoch": 1.054508235742169,
+      "grad_norm": 3.1238120051623315,
+      "learning_rate": 2e-05,
+      "loss": 0.4025,
+      "step": 44302
+    },
+    {
+      "epoch": 1.0545320384652004,
+      "grad_norm": 0.10663983295579227,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44303
+    },
+    {
+      "epoch": 1.0545558411882319,
+      "grad_norm": 3.117381226498429,
+      "learning_rate": 2e-05,
+      "loss": 0.1962,
+      "step": 44304
+    },
+    {
+      "epoch": 1.0545796439112634,
+      "grad_norm": 3.106956084878993,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 44305
+    },
+    {
+      "epoch": 1.054603446634295,
+      "grad_norm": 3.162946395936402,
+      "learning_rate": 2e-05,
+      "loss": 0.2624,
+      "step": 44306
+    },
+    {
+      "epoch": 1.0546272493573265,
+      "grad_norm": 0.07150536676948047,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 44307
+    },
+    {
+      "epoch": 1.054651052080358,
+      "grad_norm": 3.9088810435838304,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 44308
+    },
+    {
+      "epoch": 1.0546748548033895,
+      "grad_norm": 0.018828427567013695,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44309
+    },
+    {
+      "epoch": 1.054698657526421,
+      "grad_norm": 4.975622881988556,
+      "learning_rate": 2e-05,
+      "loss": 0.8669,
+      "step": 44310
+    },
+    {
+      "epoch": 1.0547224602494525,
+      "grad_norm": 7.272541165632384,
+      "learning_rate": 2e-05,
+      "loss": 0.0711,
+      "step": 44311
+    },
+    {
+      "epoch": 1.054746262972484,
+      "grad_norm": 2.537656042457004,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 44312
+    },
+    {
+      "epoch": 1.0547700656955157,
+      "grad_norm": 0.3872369100753712,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 44313
+    },
+    {
+      "epoch": 1.0547938684185472,
+      "grad_norm": 5.329644758754061,
+      "learning_rate": 2e-05,
+      "loss": 0.2385,
+      "step": 44314
+    },
+    {
+      "epoch": 1.0548176711415786,
+      "grad_norm": 1.6050314800478476,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 44315
+    },
+    {
+      "epoch": 1.0548414738646101,
+      "grad_norm": 1.7840354098513653,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 44316
+    },
+    {
+      "epoch": 1.0548652765876416,
+      "grad_norm": 0.6778816847516607,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 44317
+    },
+    {
+      "epoch": 1.054889079310673,
+      "grad_norm": 6.693790901887591,
+      "learning_rate": 2e-05,
+      "loss": 0.7253,
+      "step": 44318
+    },
+    {
+      "epoch": 1.0549128820337046,
+      "grad_norm": 4.529368461912317,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 44319
+    },
+    {
+      "epoch": 1.054936684756736,
+      "grad_norm": 3.812447341095085,
+      "learning_rate": 2e-05,
+      "loss": 0.3769,
+      "step": 44320
+    },
+    {
+      "epoch": 1.0549604874797678,
+      "grad_norm": 6.137538621182214,
+      "learning_rate": 2e-05,
+      "loss": 0.3225,
+      "step": 44321
+    },
+    {
+      "epoch": 1.0549842902027993,
+      "grad_norm": 5.032353684762162,
+      "learning_rate": 2e-05,
+      "loss": 0.421,
+      "step": 44322
+    },
+    {
+      "epoch": 1.0550080929258308,
+      "grad_norm": 0.13349065742581379,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44323
+    },
+    {
+      "epoch": 1.0550318956488622,
+      "grad_norm": 20.076513361062656,
+      "learning_rate": 2e-05,
+      "loss": 0.226,
+      "step": 44324
+    },
+    {
+      "epoch": 1.0550556983718937,
+      "grad_norm": 0.8526635136586129,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 44325
+    },
+    {
+      "epoch": 1.0550795010949252,
+      "grad_norm": 0.0188074677797585,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44326
+    },
+    {
+      "epoch": 1.0551033038179567,
+      "grad_norm": 0.7457411080216814,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 44327
+    },
+    {
+      "epoch": 1.0551271065409882,
+      "grad_norm": 0.16104773244836948,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 44328
+    },
+    {
+      "epoch": 1.0551509092640199,
+      "grad_norm": 0.5772473131350863,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 44329
+    },
+    {
+      "epoch": 1.0551747119870514,
+      "grad_norm": 4.159537738204976,
+      "learning_rate": 2e-05,
+      "loss": 0.1303,
+      "step": 44330
+    },
+    {
+      "epoch": 1.0551985147100829,
+      "grad_norm": 4.385695765366459,
+      "learning_rate": 2e-05,
+      "loss": 0.0917,
+      "step": 44331
+    },
+    {
+      "epoch": 1.0552223174331143,
+      "grad_norm": 6.943805427918738,
+      "learning_rate": 2e-05,
+      "loss": 0.691,
+      "step": 44332
+    },
+    {
+      "epoch": 1.0552461201561458,
+      "grad_norm": 0.11819010844682049,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44333
+    },
+    {
+      "epoch": 1.0552699228791773,
+      "grad_norm": 0.03831808957710155,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44334
+    },
+    {
+      "epoch": 1.0552937256022088,
+      "grad_norm": 0.022369508750095884,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44335
+    },
+    {
+      "epoch": 1.0553175283252405,
+      "grad_norm": 0.11966612848240735,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44336
+    },
+    {
+      "epoch": 1.055341331048272,
+      "grad_norm": 6.264937321856633,
+      "learning_rate": 2e-05,
+      "loss": 0.0662,
+      "step": 44337
+    },
+    {
+      "epoch": 1.0553651337713035,
+      "grad_norm": 0.03068012267610254,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44338
+    },
+    {
+      "epoch": 1.055388936494335,
+      "grad_norm": 0.005541913403271536,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44339
+    },
+    {
+      "epoch": 1.0554127392173664,
+      "grad_norm": 5.426570179818726,
+      "learning_rate": 2e-05,
+      "loss": 0.6085,
+      "step": 44340
+    },
+    {
+      "epoch": 1.055436541940398,
+      "grad_norm": 6.024150383784426,
+      "learning_rate": 2e-05,
+      "loss": 0.4332,
+      "step": 44341
+    },
+    {
+      "epoch": 1.0554603446634294,
+      "grad_norm": 0.0007856371885875795,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 44342
+    },
+    {
+      "epoch": 1.0554841473864611,
+      "grad_norm": 4.61727420573822,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 44343
+    },
+    {
+      "epoch": 1.0555079501094926,
+      "grad_norm": 5.55906521887773,
+      "learning_rate": 2e-05,
+      "loss": 0.5709,
+      "step": 44344
+    },
+    {
+      "epoch": 1.055531752832524,
+      "grad_norm": 0.04795342584508964,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44345
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 4.002304930670575,
+      "learning_rate": 2e-05,
+      "loss": 0.3066,
+      "step": 44346
+    },
+    {
+      "epoch": 1.055579358278587,
+      "grad_norm": 9.87169299671697,
+      "learning_rate": 2e-05,
+      "loss": 0.1482,
+      "step": 44347
+    },
+    {
+      "epoch": 1.0556031610016186,
+      "grad_norm": 3.391833278692541,
+      "learning_rate": 2e-05,
+      "loss": 0.1879,
+      "step": 44348
+    },
+    {
+      "epoch": 1.05562696372465,
+      "grad_norm": 0.034094801494531694,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44349
+    },
+    {
+      "epoch": 1.0556507664476815,
+      "grad_norm": 0.1843311661250857,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44350
+    },
+    {
+      "epoch": 1.0556745691707132,
+      "grad_norm": 3.971089069073622,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 44351
+    },
+    {
+      "epoch": 1.0556983718937447,
+      "grad_norm": 6.912779062185704,
+      "learning_rate": 2e-05,
+      "loss": 0.2469,
+      "step": 44352
+    },
+    {
+      "epoch": 1.0557221746167762,
+      "grad_norm": 1.7292998352975721,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 44353
+    },
+    {
+      "epoch": 1.0557459773398077,
+      "grad_norm": 0.25102484099980477,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 44354
+    },
+    {
+      "epoch": 1.0557697800628392,
+      "grad_norm": 2.510977860085748,
+      "learning_rate": 2e-05,
+      "loss": 0.0286,
+      "step": 44355
+    },
+    {
+      "epoch": 1.0557935827858707,
+      "grad_norm": 6.286244622649743,
+      "learning_rate": 2e-05,
+      "loss": 0.0484,
+      "step": 44356
+    },
+    {
+      "epoch": 1.0558173855089021,
+      "grad_norm": 0.02546636325215863,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44357
+    },
+    {
+      "epoch": 1.0558411882319336,
+      "grad_norm": 1.018046623189003,
+      "learning_rate": 2e-05,
+      "loss": 0.0187,
+      "step": 44358
+    },
+    {
+      "epoch": 1.0558649909549653,
+      "grad_norm": 5.666295305584185,
+      "learning_rate": 2e-05,
+      "loss": 0.1902,
+      "step": 44359
+    },
+    {
+      "epoch": 1.0558887936779968,
+      "grad_norm": 1.964635950586878,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 44360
+    },
+    {
+      "epoch": 1.0559125964010283,
+      "grad_norm": 21.272563793569557,
+      "learning_rate": 2e-05,
+      "loss": 0.1014,
+      "step": 44361
+    },
+    {
+      "epoch": 1.0559363991240598,
+      "grad_norm": 0.01283665115790309,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44362
+    },
+    {
+      "epoch": 1.0559602018470913,
+      "grad_norm": 3.9239062661268487,
+      "learning_rate": 2e-05,
+      "loss": 0.3559,
+      "step": 44363
+    },
+    {
+      "epoch": 1.0559840045701228,
+      "grad_norm": 3.591285603552304,
+      "learning_rate": 2e-05,
+      "loss": 0.3803,
+      "step": 44364
+    },
+    {
+      "epoch": 1.0560078072931542,
+      "grad_norm": 0.07472551834616735,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44365
+    },
+    {
+      "epoch": 1.056031610016186,
+      "grad_norm": 2.4140115580393426,
+      "learning_rate": 2e-05,
+      "loss": 0.0819,
+      "step": 44366
+    },
+    {
+      "epoch": 1.0560554127392174,
+      "grad_norm": 6.850641709674856,
+      "learning_rate": 2e-05,
+      "loss": 1.139,
+      "step": 44367
+    },
+    {
+      "epoch": 1.056079215462249,
+      "grad_norm": 12.879197910149863,
+      "learning_rate": 2e-05,
+      "loss": 0.3202,
+      "step": 44368
+    },
+    {
+      "epoch": 1.0561030181852804,
+      "grad_norm": 2.705320778368535,
+      "learning_rate": 2e-05,
+      "loss": 0.0755,
+      "step": 44369
+    },
+    {
+      "epoch": 1.056126820908312,
+      "grad_norm": 3.8189108436481267,
+      "learning_rate": 2e-05,
+      "loss": 0.3558,
+      "step": 44370
+    },
+    {
+      "epoch": 1.0561506236313434,
+      "grad_norm": 1.757773904996597,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 44371
+    },
+    {
+      "epoch": 1.0561744263543749,
+      "grad_norm": 0.10378867295380365,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44372
+    },
+    {
+      "epoch": 1.0561982290774066,
+      "grad_norm": 6.173880418122236,
+      "learning_rate": 2e-05,
+      "loss": 0.543,
+      "step": 44373
+    },
+    {
+      "epoch": 1.056222031800438,
+      "grad_norm": 10.199662383585434,
+      "learning_rate": 2e-05,
+      "loss": 0.9316,
+      "step": 44374
+    },
+    {
+      "epoch": 1.0562458345234695,
+      "grad_norm": 7.40635327481967,
+      "learning_rate": 2e-05,
+      "loss": 0.6048,
+      "step": 44375
+    },
+    {
+      "epoch": 1.056269637246501,
+      "grad_norm": 2.84742677032532,
+      "learning_rate": 2e-05,
+      "loss": 0.0297,
+      "step": 44376
+    },
+    {
+      "epoch": 1.0562934399695325,
+      "grad_norm": 4.175441784687836,
+      "learning_rate": 2e-05,
+      "loss": 0.383,
+      "step": 44377
+    },
+    {
+      "epoch": 1.056317242692564,
+      "grad_norm": 3.367117645465265,
+      "learning_rate": 2e-05,
+      "loss": 0.5255,
+      "step": 44378
+    },
+    {
+      "epoch": 1.0563410454155955,
+      "grad_norm": 2.0291071860166547,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 44379
+    },
+    {
+      "epoch": 1.056364848138627,
+      "grad_norm": 2.6478218473477524,
+      "learning_rate": 2e-05,
+      "loss": 0.0761,
+      "step": 44380
+    },
+    {
+      "epoch": 1.0563886508616587,
+      "grad_norm": 3.4151665653229695,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 44381
+    },
+    {
+      "epoch": 1.0564124535846902,
+      "grad_norm": 7.697026906699264,
+      "learning_rate": 2e-05,
+      "loss": 0.1823,
+      "step": 44382
+    },
+    {
+      "epoch": 1.0564362563077216,
+      "grad_norm": 3.1237747497282884,
+      "learning_rate": 2e-05,
+      "loss": 0.1133,
+      "step": 44383
+    },
+    {
+      "epoch": 1.0564600590307531,
+      "grad_norm": 3.3946861010472955,
+      "learning_rate": 2e-05,
+      "loss": 0.0402,
+      "step": 44384
+    },
+    {
+      "epoch": 1.0564838617537846,
+      "grad_norm": 9.03845572417822,
+      "learning_rate": 2e-05,
+      "loss": 0.2138,
+      "step": 44385
+    },
+    {
+      "epoch": 1.056507664476816,
+      "grad_norm": 3.2041400624466405,
+      "learning_rate": 2e-05,
+      "loss": 0.0466,
+      "step": 44386
+    },
+    {
+      "epoch": 1.0565314671998476,
+      "grad_norm": 4.502698273162663,
+      "learning_rate": 2e-05,
+      "loss": 0.3359,
+      "step": 44387
+    },
+    {
+      "epoch": 1.056555269922879,
+      "grad_norm": 4.102200686843728,
+      "learning_rate": 2e-05,
+      "loss": 0.4647,
+      "step": 44388
+    },
+    {
+      "epoch": 1.0565790726459108,
+      "grad_norm": 3.745617205671525,
+      "learning_rate": 2e-05,
+      "loss": 0.4519,
+      "step": 44389
+    },
+    {
+      "epoch": 1.0566028753689423,
+      "grad_norm": 2.3644360546731784,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 44390
+    },
+    {
+      "epoch": 1.0566266780919737,
+      "grad_norm": 0.10370734405743903,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44391
+    },
+    {
+      "epoch": 1.0566504808150052,
+      "grad_norm": 3.5189556018687806,
+      "learning_rate": 2e-05,
+      "loss": 0.6438,
+      "step": 44392
+    },
+    {
+      "epoch": 1.0566742835380367,
+      "grad_norm": 3.777445730721017,
+      "learning_rate": 2e-05,
+      "loss": 0.3996,
+      "step": 44393
+    },
+    {
+      "epoch": 1.0566980862610682,
+      "grad_norm": 2.7551717127875768,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 44394
+    },
+    {
+      "epoch": 1.0567218889840997,
+      "grad_norm": 12.746094300012299,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 44395
+    },
+    {
+      "epoch": 1.0567456917071314,
+      "grad_norm": 3.8423532092836425,
+      "learning_rate": 2e-05,
+      "loss": 0.5664,
+      "step": 44396
+    },
+    {
+      "epoch": 1.0567694944301629,
+      "grad_norm": 4.100020652651845,
+      "learning_rate": 2e-05,
+      "loss": 0.0661,
+      "step": 44397
+    },
+    {
+      "epoch": 1.0567932971531944,
+      "grad_norm": 3.7887186244139293,
+      "learning_rate": 2e-05,
+      "loss": 0.4877,
+      "step": 44398
+    },
+    {
+      "epoch": 1.0568170998762259,
+      "grad_norm": 0.6866888787117971,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 44399
+    },
+    {
+      "epoch": 1.0568409025992573,
+      "grad_norm": 0.06334160871030944,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44400
+    },
+    {
+      "epoch": 1.0568647053222888,
+      "grad_norm": 2.695791221968377,
+      "learning_rate": 2e-05,
+      "loss": 0.295,
+      "step": 44401
+    },
+    {
+      "epoch": 1.0568885080453203,
+      "grad_norm": 3.8964503064998883,
+      "learning_rate": 2e-05,
+      "loss": 0.4525,
+      "step": 44402
+    },
+    {
+      "epoch": 1.0569123107683518,
+      "grad_norm": 0.6675774186583687,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 44403
+    },
+    {
+      "epoch": 1.0569361134913835,
+      "grad_norm": 3.8159500578764147,
+      "learning_rate": 2e-05,
+      "loss": 0.4127,
+      "step": 44404
+    },
+    {
+      "epoch": 1.056959916214415,
+      "grad_norm": 3.1792120798796915,
+      "learning_rate": 2e-05,
+      "loss": 0.0468,
+      "step": 44405
+    },
+    {
+      "epoch": 1.0569837189374465,
+      "grad_norm": 10.684891747053168,
+      "learning_rate": 2e-05,
+      "loss": 0.5367,
+      "step": 44406
+    },
+    {
+      "epoch": 1.057007521660478,
+      "grad_norm": 0.2011427707749914,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 44407
+    },
+    {
+      "epoch": 1.0570313243835094,
+      "grad_norm": 4.3993229639038205,
+      "learning_rate": 2e-05,
+      "loss": 0.0946,
+      "step": 44408
+    },
+    {
+      "epoch": 1.057055127106541,
+      "grad_norm": 0.4133413091619892,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 44409
+    },
+    {
+      "epoch": 1.0570789298295724,
+      "grad_norm": 12.675477389537871,
+      "learning_rate": 2e-05,
+      "loss": 0.1119,
+      "step": 44410
+    },
+    {
+      "epoch": 1.0571027325526041,
+      "grad_norm": 0.1958598233216804,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 44411
+    },
+    {
+      "epoch": 1.0571265352756356,
+      "grad_norm": 4.663833442867429,
+      "learning_rate": 2e-05,
+      "loss": 0.5916,
+      "step": 44412
+    },
+    {
+      "epoch": 1.057150337998667,
+      "grad_norm": 8.574146296428882,
+      "learning_rate": 2e-05,
+      "loss": 0.8486,
+      "step": 44413
+    },
+    {
+      "epoch": 1.0571741407216986,
+      "grad_norm": 5.588658414653093,
+      "learning_rate": 2e-05,
+      "loss": 0.4821,
+      "step": 44414
+    },
+    {
+      "epoch": 1.05719794344473,
+      "grad_norm": 0.12274464911696079,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44415
+    },
+    {
+      "epoch": 1.0572217461677615,
+      "grad_norm": 0.6766811766463193,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 44416
+    },
+    {
+      "epoch": 1.057245548890793,
+      "grad_norm": 2.419961569192363,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 44417
+    },
+    {
+      "epoch": 1.0572693516138245,
+      "grad_norm": 3.4502999292959684,
+      "learning_rate": 2e-05,
+      "loss": 0.098,
+      "step": 44418
+    },
+    {
+      "epoch": 1.0572931543368562,
+      "grad_norm": 3.8327415887832963,
+      "learning_rate": 2e-05,
+      "loss": 0.4421,
+      "step": 44419
+    },
+    {
+      "epoch": 1.0573169570598877,
+      "grad_norm": 4.835744264455631,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 44420
+    },
+    {
+      "epoch": 1.0573407597829192,
+      "grad_norm": 7.693496937527854,
+      "learning_rate": 2e-05,
+      "loss": 0.0907,
+      "step": 44421
+    },
+    {
+      "epoch": 1.0573645625059507,
+      "grad_norm": 3.818510125821528,
+      "learning_rate": 2e-05,
+      "loss": 0.4119,
+      "step": 44422
+    },
+    {
+      "epoch": 1.0573883652289822,
+      "grad_norm": 13.816473011003215,
+      "learning_rate": 2e-05,
+      "loss": 0.3782,
+      "step": 44423
+    },
+    {
+      "epoch": 1.0574121679520136,
+      "grad_norm": 18.999553722745386,
+      "learning_rate": 2e-05,
+      "loss": 0.1617,
+      "step": 44424
+    },
+    {
+      "epoch": 1.0574359706750451,
+      "grad_norm": 0.2228410243401826,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 44425
+    },
+    {
+      "epoch": 1.0574597733980768,
+      "grad_norm": 0.8764608569355432,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 44426
+    },
+    {
+      "epoch": 1.0574835761211083,
+      "grad_norm": 0.839862366285438,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 44427
+    },
+    {
+      "epoch": 1.0575073788441398,
+      "grad_norm": 1.0826345899524914,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 44428
+    },
+    {
+      "epoch": 1.0575311815671713,
+      "grad_norm": 2.191283529812708,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 44429
+    },
+    {
+      "epoch": 1.0575549842902028,
+      "grad_norm": 9.31503553220119,
+      "learning_rate": 2e-05,
+      "loss": 1.0756,
+      "step": 44430
+    },
+    {
+      "epoch": 1.0575787870132343,
+      "grad_norm": 6.832612039964796,
+      "learning_rate": 2e-05,
+      "loss": 0.7565,
+      "step": 44431
+    },
+    {
+      "epoch": 1.0576025897362658,
+      "grad_norm": 0.3507723837608351,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 44432
+    },
+    {
+      "epoch": 1.0576263924592972,
+      "grad_norm": 3.0699535479023563,
+      "learning_rate": 2e-05,
+      "loss": 0.0512,
+      "step": 44433
+    },
+    {
+      "epoch": 1.057650195182329,
+      "grad_norm": 4.929576378011144,
+      "learning_rate": 2e-05,
+      "loss": 0.4417,
+      "step": 44434
+    },
+    {
+      "epoch": 1.0576739979053604,
+      "grad_norm": 4.029490310299241,
+      "learning_rate": 2e-05,
+      "loss": 0.4113,
+      "step": 44435
+    },
+    {
+      "epoch": 1.057697800628392,
+      "grad_norm": 3.4329068517378887,
+      "learning_rate": 2e-05,
+      "loss": 0.3778,
+      "step": 44436
+    },
+    {
+      "epoch": 1.0577216033514234,
+      "grad_norm": 0.6058143412354987,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 44437
+    },
+    {
+      "epoch": 1.0577454060744549,
+      "grad_norm": 0.020182417301422093,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44438
+    },
+    {
+      "epoch": 1.0577692087974864,
+      "grad_norm": 3.0505840463745453,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 44439
+    },
+    {
+      "epoch": 1.0577930115205179,
+      "grad_norm": 3.165099006787688,
+      "learning_rate": 2e-05,
+      "loss": 0.1988,
+      "step": 44440
+    },
+    {
+      "epoch": 1.0578168142435496,
+      "grad_norm": 2.664648137237923,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 44441
+    },
+    {
+      "epoch": 1.057840616966581,
+      "grad_norm": 0.9606936795118028,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 44442
+    },
+    {
+      "epoch": 1.0578644196896125,
+      "grad_norm": 0.5234144317252214,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 44443
+    },
+    {
+      "epoch": 1.057888222412644,
+      "grad_norm": 10.956454872951403,
+      "learning_rate": 2e-05,
+      "loss": 1.021,
+      "step": 44444
+    },
+    {
+      "epoch": 1.0579120251356755,
+      "grad_norm": 1.8372615875880631,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 44445
+    },
+    {
+      "epoch": 1.057935827858707,
+      "grad_norm": 3.8432411498369956,
+      "learning_rate": 2e-05,
+      "loss": 0.3492,
+      "step": 44446
+    },
+    {
+      "epoch": 1.0579596305817385,
+      "grad_norm": 19.83054362224949,
+      "learning_rate": 2e-05,
+      "loss": 1.203,
+      "step": 44447
+    },
+    {
+      "epoch": 1.05798343330477,
+      "grad_norm": 0.5773435023680171,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 44448
+    },
+    {
+      "epoch": 1.0580072360278017,
+      "grad_norm": 3.2911113874722724,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 44449
+    },
+    {
+      "epoch": 1.0580310387508332,
+      "grad_norm": 1.0708169702586752,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 44450
+    },
+    {
+      "epoch": 1.0580548414738646,
+      "grad_norm": 4.936093024985507,
+      "learning_rate": 2e-05,
+      "loss": 0.6855,
+      "step": 44451
+    },
+    {
+      "epoch": 1.0580786441968961,
+      "grad_norm": 6.569852234778257,
+      "learning_rate": 2e-05,
+      "loss": 0.2498,
+      "step": 44452
+    },
+    {
+      "epoch": 1.0581024469199276,
+      "grad_norm": 1.1628176933007055,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 44453
+    },
+    {
+      "epoch": 1.058126249642959,
+      "grad_norm": 4.200937739389533,
+      "learning_rate": 2e-05,
+      "loss": 0.3513,
+      "step": 44454
+    },
+    {
+      "epoch": 1.0581500523659906,
+      "grad_norm": 4.633257382276801,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 44455
+    },
+    {
+      "epoch": 1.0581738550890223,
+      "grad_norm": 3.6977812904112497,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 44456
+    },
+    {
+      "epoch": 1.0581976578120538,
+      "grad_norm": 4.653565962727117,
+      "learning_rate": 2e-05,
+      "loss": 0.4244,
+      "step": 44457
+    },
+    {
+      "epoch": 1.0582214605350853,
+      "grad_norm": 0.022917055410686955,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44458
+    },
+    {
+      "epoch": 1.0582452632581167,
+      "grad_norm": 18.453372028592625,
+      "learning_rate": 2e-05,
+      "loss": 0.5053,
+      "step": 44459
+    },
+    {
+      "epoch": 1.0582690659811482,
+      "grad_norm": 4.736126451128271,
+      "learning_rate": 2e-05,
+      "loss": 0.5347,
+      "step": 44460
+    },
+    {
+      "epoch": 1.0582928687041797,
+      "grad_norm": 5.822561888326299,
+      "learning_rate": 2e-05,
+      "loss": 0.7941,
+      "step": 44461
+    },
+    {
+      "epoch": 1.0583166714272112,
+      "grad_norm": 5.31080781923222,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 44462
+    },
+    {
+      "epoch": 1.0583404741502427,
+      "grad_norm": 3.7496313443090874,
+      "learning_rate": 2e-05,
+      "loss": 0.1261,
+      "step": 44463
+    },
+    {
+      "epoch": 1.0583642768732744,
+      "grad_norm": 0.798341844151226,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 44464
+    },
+    {
+      "epoch": 1.0583880795963059,
+      "grad_norm": 4.738359358986287,
+      "learning_rate": 2e-05,
+      "loss": 0.7053,
+      "step": 44465
+    },
+    {
+      "epoch": 1.0584118823193374,
+      "grad_norm": 10.245551897652103,
+      "learning_rate": 2e-05,
+      "loss": 0.6092,
+      "step": 44466
+    },
+    {
+      "epoch": 1.0584356850423688,
+      "grad_norm": 11.247800113942452,
+      "learning_rate": 2e-05,
+      "loss": 0.5461,
+      "step": 44467
+    },
+    {
+      "epoch": 1.0584594877654003,
+      "grad_norm": 0.09731415631091547,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44468
+    },
+    {
+      "epoch": 1.0584832904884318,
+      "grad_norm": 2.330143189435597,
+      "learning_rate": 2e-05,
+      "loss": 0.1829,
+      "step": 44469
+    },
+    {
+      "epoch": 1.0585070932114633,
+      "grad_norm": 0.6584990375011365,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 44470
+    },
+    {
+      "epoch": 1.058530895934495,
+      "grad_norm": 2.4533388986159688,
+      "learning_rate": 2e-05,
+      "loss": 0.0301,
+      "step": 44471
+    },
+    {
+      "epoch": 1.0585546986575265,
+      "grad_norm": 0.5799030858005042,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 44472
+    },
+    {
+      "epoch": 1.058578501380558,
+      "grad_norm": 6.614700209089009,
+      "learning_rate": 2e-05,
+      "loss": 0.5624,
+      "step": 44473
+    },
+    {
+      "epoch": 1.0586023041035895,
+      "grad_norm": 9.634854867696381,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 44474
+    },
+    {
+      "epoch": 1.058626106826621,
+      "grad_norm": 0.2413172475546092,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 44475
+    },
+    {
+      "epoch": 1.0586499095496524,
+      "grad_norm": 2.8062640433458457,
+      "learning_rate": 2e-05,
+      "loss": 0.2456,
+      "step": 44476
+    },
+    {
+      "epoch": 1.058673712272684,
+      "grad_norm": 6.0598966787424775,
+      "learning_rate": 2e-05,
+      "loss": 0.0543,
+      "step": 44477
+    },
+    {
+      "epoch": 1.0586975149957154,
+      "grad_norm": 0.012326471538036277,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44478
+    },
+    {
+      "epoch": 1.0587213177187471,
+      "grad_norm": 0.115396492026779,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44479
+    },
+    {
+      "epoch": 1.0587451204417786,
+      "grad_norm": 9.584361665803591,
+      "learning_rate": 2e-05,
+      "loss": 0.3248,
+      "step": 44480
+    },
+    {
+      "epoch": 1.05876892316481,
+      "grad_norm": 3.4592934934326163,
+      "learning_rate": 2e-05,
+      "loss": 0.1778,
+      "step": 44481
+    },
+    {
+      "epoch": 1.0587927258878416,
+      "grad_norm": 0.6944347274771042,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 44482
+    },
+    {
+      "epoch": 1.058816528610873,
+      "grad_norm": 0.035826636725717356,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44483
+    },
+    {
+      "epoch": 1.0588403313339045,
+      "grad_norm": 4.694506336969043,
+      "learning_rate": 2e-05,
+      "loss": 0.1558,
+      "step": 44484
+    },
+    {
+      "epoch": 1.058864134056936,
+      "grad_norm": 7.117445133049257,
+      "learning_rate": 2e-05,
+      "loss": 0.8622,
+      "step": 44485
+    },
+    {
+      "epoch": 1.0588879367799677,
+      "grad_norm": 5.182236589241034,
+      "learning_rate": 2e-05,
+      "loss": 0.6494,
+      "step": 44486
+    },
+    {
+      "epoch": 1.0589117395029992,
+      "grad_norm": 6.710035658137216,
+      "learning_rate": 2e-05,
+      "loss": 0.7855,
+      "step": 44487
+    },
+    {
+      "epoch": 1.0589355422260307,
+      "grad_norm": 3.6198092553441112,
+      "learning_rate": 2e-05,
+      "loss": 0.4619,
+      "step": 44488
+    },
+    {
+      "epoch": 1.0589593449490622,
+      "grad_norm": 5.495077278648874,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 44489
+    },
+    {
+      "epoch": 1.0589831476720937,
+      "grad_norm": 0.008397958246292008,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44490
+    },
+    {
+      "epoch": 1.0590069503951252,
+      "grad_norm": 1.0656781893914302,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 44491
+    },
+    {
+      "epoch": 1.0590307531181566,
+      "grad_norm": 2.3849629931772776,
+      "learning_rate": 2e-05,
+      "loss": 0.1067,
+      "step": 44492
+    },
+    {
+      "epoch": 1.0590545558411881,
+      "grad_norm": 3.983326900583724,
+      "learning_rate": 2e-05,
+      "loss": 0.4167,
+      "step": 44493
+    },
+    {
+      "epoch": 1.0590783585642198,
+      "grad_norm": 7.4184764347179195,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 44494
+    },
+    {
+      "epoch": 1.0591021612872513,
+      "grad_norm": 6.033240064978403,
+      "learning_rate": 2e-05,
+      "loss": 0.6263,
+      "step": 44495
+    },
+    {
+      "epoch": 1.0591259640102828,
+      "grad_norm": 3.15833431294038,
+      "learning_rate": 2e-05,
+      "loss": 0.4135,
+      "step": 44496
+    },
+    {
+      "epoch": 1.0591497667333143,
+      "grad_norm": 3.3838666982994825,
+      "learning_rate": 2e-05,
+      "loss": 0.3075,
+      "step": 44497
+    },
+    {
+      "epoch": 1.0591735694563458,
+      "grad_norm": 5.119681591403885,
+      "learning_rate": 2e-05,
+      "loss": 0.4428,
+      "step": 44498
+    },
+    {
+      "epoch": 1.0591973721793773,
+      "grad_norm": 11.11924913389089,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 44499
+    },
+    {
+      "epoch": 1.0592211749024087,
+      "grad_norm": 5.500969774005045,
+      "learning_rate": 2e-05,
+      "loss": 0.4579,
+      "step": 44500
+    },
+    {
+      "epoch": 1.0592449776254405,
+      "grad_norm": 2.333617170104105,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 44501
+    },
+    {
+      "epoch": 1.059268780348472,
+      "grad_norm": 0.9648080880706722,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 44502
+    },
+    {
+      "epoch": 1.0592925830715034,
+      "grad_norm": 4.538877433863977,
+      "learning_rate": 2e-05,
+      "loss": 0.4325,
+      "step": 44503
+    },
+    {
+      "epoch": 1.059316385794535,
+      "grad_norm": 0.33287706288136193,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 44504
+    },
+    {
+      "epoch": 1.0593401885175664,
+      "grad_norm": 4.038274199448346,
+      "learning_rate": 2e-05,
+      "loss": 0.703,
+      "step": 44505
+    },
+    {
+      "epoch": 1.0593639912405979,
+      "grad_norm": 3.6977540027049103,
+      "learning_rate": 2e-05,
+      "loss": 0.0689,
+      "step": 44506
+    },
+    {
+      "epoch": 1.0593877939636294,
+      "grad_norm": 0.14968816539083876,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44507
+    },
+    {
+      "epoch": 1.0594115966866609,
+      "grad_norm": 4.3078367940882805,
+      "learning_rate": 2e-05,
+      "loss": 0.4319,
+      "step": 44508
+    },
+    {
+      "epoch": 1.0594353994096926,
+      "grad_norm": 1.6930459028791134,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 44509
+    },
+    {
+      "epoch": 1.059459202132724,
+      "grad_norm": 4.561654593289447,
+      "learning_rate": 2e-05,
+      "loss": 0.5915,
+      "step": 44510
+    },
+    {
+      "epoch": 1.0594830048557555,
+      "grad_norm": 5.625565501228672,
+      "learning_rate": 2e-05,
+      "loss": 0.4081,
+      "step": 44511
+    },
+    {
+      "epoch": 1.059506807578787,
+      "grad_norm": 0.03895503108331899,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44512
+    },
+    {
+      "epoch": 1.0595306103018185,
+      "grad_norm": 0.016054627256412723,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44513
+    },
+    {
+      "epoch": 1.05955441302485,
+      "grad_norm": 5.05091575487534,
+      "learning_rate": 2e-05,
+      "loss": 0.7369,
+      "step": 44514
+    },
+    {
+      "epoch": 1.0595782157478815,
+      "grad_norm": 3.7637631762019637,
+      "learning_rate": 2e-05,
+      "loss": 0.148,
+      "step": 44515
+    },
+    {
+      "epoch": 1.0596020184709132,
+      "grad_norm": 2.8465471839220666,
+      "learning_rate": 2e-05,
+      "loss": 0.3406,
+      "step": 44516
+    },
+    {
+      "epoch": 1.0596258211939447,
+      "grad_norm": 3.640429141377037,
+      "learning_rate": 2e-05,
+      "loss": 0.2875,
+      "step": 44517
+    },
+    {
+      "epoch": 1.0596496239169761,
+      "grad_norm": 3.524392608645111,
+      "learning_rate": 2e-05,
+      "loss": 0.3761,
+      "step": 44518
+    },
+    {
+      "epoch": 1.0596734266400076,
+      "grad_norm": 4.943052785401978,
+      "learning_rate": 2e-05,
+      "loss": 0.487,
+      "step": 44519
+    },
+    {
+      "epoch": 1.0596972293630391,
+      "grad_norm": 0.8498827136623803,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 44520
+    },
+    {
+      "epoch": 1.0597210320860706,
+      "grad_norm": 4.5214127258104035,
+      "learning_rate": 2e-05,
+      "loss": 0.3492,
+      "step": 44521
+    },
+    {
+      "epoch": 1.059744834809102,
+      "grad_norm": 0.045619264951209604,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44522
+    },
+    {
+      "epoch": 1.0597686375321336,
+      "grad_norm": 6.924285011244253,
+      "learning_rate": 2e-05,
+      "loss": 0.7018,
+      "step": 44523
+    },
+    {
+      "epoch": 1.0597924402551653,
+      "grad_norm": 3.2091444808984866,
+      "learning_rate": 2e-05,
+      "loss": 0.4054,
+      "step": 44524
+    },
+    {
+      "epoch": 1.0598162429781968,
+      "grad_norm": 4.011928248069565,
+      "learning_rate": 2e-05,
+      "loss": 0.0665,
+      "step": 44525
+    },
+    {
+      "epoch": 1.0598400457012283,
+      "grad_norm": 3.5065743471910915,
+      "learning_rate": 2e-05,
+      "loss": 0.1333,
+      "step": 44526
+    },
+    {
+      "epoch": 1.0598638484242597,
+      "grad_norm": 3.388909367258565,
+      "learning_rate": 2e-05,
+      "loss": 0.3227,
+      "step": 44527
+    },
+    {
+      "epoch": 1.0598876511472912,
+      "grad_norm": 0.22024371582946922,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 44528
+    },
+    {
+      "epoch": 1.0599114538703227,
+      "grad_norm": 7.00632356679641,
+      "learning_rate": 2e-05,
+      "loss": 0.4245,
+      "step": 44529
+    },
+    {
+      "epoch": 1.0599352565933542,
+      "grad_norm": 10.135504289835094,
+      "learning_rate": 2e-05,
+      "loss": 0.9641,
+      "step": 44530
+    },
+    {
+      "epoch": 1.059959059316386,
+      "grad_norm": 3.131384650937634,
+      "learning_rate": 2e-05,
+      "loss": 0.3415,
+      "step": 44531
+    },
+    {
+      "epoch": 1.0599828620394174,
+      "grad_norm": 5.822379625280965,
+      "learning_rate": 2e-05,
+      "loss": 0.4131,
+      "step": 44532
+    },
+    {
+      "epoch": 1.0600066647624489,
+      "grad_norm": 0.034080005698792486,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44533
+    },
+    {
+      "epoch": 1.0600304674854804,
+      "grad_norm": 0.16735130253380776,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 44534
+    },
+    {
+      "epoch": 1.0600542702085118,
+      "grad_norm": 24.554481965955222,
+      "learning_rate": 2e-05,
+      "loss": 0.9102,
+      "step": 44535
+    },
+    {
+      "epoch": 1.0600780729315433,
+      "grad_norm": 1.7836303934764157,
+      "learning_rate": 2e-05,
+      "loss": 0.0723,
+      "step": 44536
+    },
+    {
+      "epoch": 1.0601018756545748,
+      "grad_norm": 9.78533960902054,
+      "learning_rate": 2e-05,
+      "loss": 1.361,
+      "step": 44537
+    },
+    {
+      "epoch": 1.0601256783776063,
+      "grad_norm": 0.3870041487039057,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 44538
+    },
+    {
+      "epoch": 1.060149481100638,
+      "grad_norm": 0.7994468783952685,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 44539
+    },
+    {
+      "epoch": 1.0601732838236695,
+      "grad_norm": 0.7405559399884969,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 44540
+    },
+    {
+      "epoch": 1.060197086546701,
+      "grad_norm": 2.1275264931824758,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 44541
+    },
+    {
+      "epoch": 1.0602208892697325,
+      "grad_norm": 0.002889085268079058,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44542
+    },
+    {
+      "epoch": 1.060244691992764,
+      "grad_norm": 7.542405523109585,
+      "learning_rate": 2e-05,
+      "loss": 0.0584,
+      "step": 44543
+    },
+    {
+      "epoch": 1.0602684947157954,
+      "grad_norm": 0.04702413366720724,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44544
+    },
+    {
+      "epoch": 1.060292297438827,
+      "grad_norm": 0.159195070793178,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 44545
+    },
+    {
+      "epoch": 1.0603161001618586,
+      "grad_norm": 5.364159429634957,
+      "learning_rate": 2e-05,
+      "loss": 0.9118,
+      "step": 44546
+    },
+    {
+      "epoch": 1.06033990288489,
+      "grad_norm": 0.03814819553823924,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44547
+    },
+    {
+      "epoch": 1.0603637056079216,
+      "grad_norm": 5.087753704744743,
+      "learning_rate": 2e-05,
+      "loss": 0.4608,
+      "step": 44548
+    },
+    {
+      "epoch": 1.060387508330953,
+      "grad_norm": 0.8055541788249719,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 44549
+    },
+    {
+      "epoch": 1.0604113110539846,
+      "grad_norm": 6.312254620770337,
+      "learning_rate": 2e-05,
+      "loss": 0.8497,
+      "step": 44550
+    },
+    {
+      "epoch": 1.060435113777016,
+      "grad_norm": 0.6846495649261747,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 44551
+    },
+    {
+      "epoch": 1.0604589165000475,
+      "grad_norm": 0.042647585459445164,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44552
+    },
+    {
+      "epoch": 1.060482719223079,
+      "grad_norm": 3.3122816116498632,
+      "learning_rate": 2e-05,
+      "loss": 0.5138,
+      "step": 44553
+    },
+    {
+      "epoch": 1.0605065219461107,
+      "grad_norm": 6.568402399625456,
+      "learning_rate": 2e-05,
+      "loss": 1.0089,
+      "step": 44554
+    },
+    {
+      "epoch": 1.0605303246691422,
+      "grad_norm": 3.5224335099433755,
+      "learning_rate": 2e-05,
+      "loss": 0.302,
+      "step": 44555
+    },
+    {
+      "epoch": 1.0605541273921737,
+      "grad_norm": 0.037221227673497906,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44556
+    },
+    {
+      "epoch": 1.0605779301152052,
+      "grad_norm": 4.132416198967282,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 44557
+    },
+    {
+      "epoch": 1.0606017328382367,
+      "grad_norm": 4.057337705102271,
+      "learning_rate": 2e-05,
+      "loss": 0.0784,
+      "step": 44558
+    },
+    {
+      "epoch": 1.0606255355612682,
+      "grad_norm": 0.013971917401688166,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44559
+    },
+    {
+      "epoch": 1.0606493382842996,
+      "grad_norm": 10.595981006306562,
+      "learning_rate": 2e-05,
+      "loss": 0.1438,
+      "step": 44560
+    },
+    {
+      "epoch": 1.0606731410073313,
+      "grad_norm": 5.763061234667785,
+      "learning_rate": 2e-05,
+      "loss": 0.4528,
+      "step": 44561
+    },
+    {
+      "epoch": 1.0606969437303628,
+      "grad_norm": 4.000605062690708,
+      "learning_rate": 2e-05,
+      "loss": 0.4552,
+      "step": 44562
+    },
+    {
+      "epoch": 1.0607207464533943,
+      "grad_norm": 0.17041055860207846,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44563
+    },
+    {
+      "epoch": 1.0607445491764258,
+      "grad_norm": 5.215797950423242,
+      "learning_rate": 2e-05,
+      "loss": 0.6603,
+      "step": 44564
+    },
+    {
+      "epoch": 1.0607683518994573,
+      "grad_norm": 2.652413331917684,
+      "learning_rate": 2e-05,
+      "loss": 0.361,
+      "step": 44565
+    },
+    {
+      "epoch": 1.0607921546224888,
+      "grad_norm": 0.015861913031164278,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44566
+    },
+    {
+      "epoch": 1.0608159573455203,
+      "grad_norm": 5.555719891335965,
+      "learning_rate": 2e-05,
+      "loss": 0.4845,
+      "step": 44567
+    },
+    {
+      "epoch": 1.0608397600685517,
+      "grad_norm": 8.600034742337664,
+      "learning_rate": 2e-05,
+      "loss": 0.665,
+      "step": 44568
+    },
+    {
+      "epoch": 1.0608635627915834,
+      "grad_norm": 0.003516330223113457,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44569
+    },
+    {
+      "epoch": 1.060887365514615,
+      "grad_norm": 2.48519820826884,
+      "learning_rate": 2e-05,
+      "loss": 0.2383,
+      "step": 44570
+    },
+    {
+      "epoch": 1.0609111682376464,
+      "grad_norm": 14.631821759496061,
+      "learning_rate": 2e-05,
+      "loss": 0.3697,
+      "step": 44571
+    },
+    {
+      "epoch": 1.060934970960678,
+      "grad_norm": 0.013998030494624109,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44572
+    },
+    {
+      "epoch": 1.0609587736837094,
+      "grad_norm": 0.29668375522074203,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 44573
+    },
+    {
+      "epoch": 1.0609825764067409,
+      "grad_norm": 0.17940635006469047,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 44574
+    },
+    {
+      "epoch": 1.0610063791297724,
+      "grad_norm": 0.753462405719241,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 44575
+    },
+    {
+      "epoch": 1.061030181852804,
+      "grad_norm": 0.05981810017697024,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44576
+    },
+    {
+      "epoch": 1.0610539845758356,
+      "grad_norm": 4.3316252944633735,
+      "learning_rate": 2e-05,
+      "loss": 0.4656,
+      "step": 44577
+    },
+    {
+      "epoch": 1.061077787298867,
+      "grad_norm": 5.133561232430727,
+      "learning_rate": 2e-05,
+      "loss": 0.5972,
+      "step": 44578
+    },
+    {
+      "epoch": 1.0611015900218985,
+      "grad_norm": 0.028815822759112382,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44579
+    },
+    {
+      "epoch": 1.06112539274493,
+      "grad_norm": 0.024567791661316916,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44580
+    },
+    {
+      "epoch": 1.0611491954679615,
+      "grad_norm": 0.027651782941443645,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44581
+    },
+    {
+      "epoch": 1.061172998190993,
+      "grad_norm": 3.3557434744946795,
+      "learning_rate": 2e-05,
+      "loss": 0.3369,
+      "step": 44582
+    },
+    {
+      "epoch": 1.0611968009140245,
+      "grad_norm": 3.168140304280606,
+      "learning_rate": 2e-05,
+      "loss": 0.1484,
+      "step": 44583
+    },
+    {
+      "epoch": 1.0612206036370562,
+      "grad_norm": 1.7533767476798388,
+      "learning_rate": 2e-05,
+      "loss": 0.0281,
+      "step": 44584
+    },
+    {
+      "epoch": 1.0612444063600877,
+      "grad_norm": 0.004307413478305241,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44585
+    },
+    {
+      "epoch": 1.0612682090831191,
+      "grad_norm": 11.907710957347756,
+      "learning_rate": 2e-05,
+      "loss": 0.385,
+      "step": 44586
+    },
+    {
+      "epoch": 1.0612920118061506,
+      "grad_norm": 2.8836327609773442,
+      "learning_rate": 2e-05,
+      "loss": 0.1873,
+      "step": 44587
+    },
+    {
+      "epoch": 1.0613158145291821,
+      "grad_norm": 4.9246273177979605,
+      "learning_rate": 2e-05,
+      "loss": 0.9722,
+      "step": 44588
+    },
+    {
+      "epoch": 1.0613396172522136,
+      "grad_norm": 0.5213414385137061,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 44589
+    },
+    {
+      "epoch": 1.061363419975245,
+      "grad_norm": 0.10242845351174697,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 44590
+    },
+    {
+      "epoch": 1.0613872226982768,
+      "grad_norm": 6.724105095228671,
+      "learning_rate": 2e-05,
+      "loss": 0.8173,
+      "step": 44591
+    },
+    {
+      "epoch": 1.0614110254213083,
+      "grad_norm": 0.005606251201208246,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44592
+    },
+    {
+      "epoch": 1.0614348281443398,
+      "grad_norm": 0.07698239511453188,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44593
+    },
+    {
+      "epoch": 1.0614586308673712,
+      "grad_norm": 0.9426705482501235,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44594
+    },
+    {
+      "epoch": 1.0614824335904027,
+      "grad_norm": 5.688903984964302,
+      "learning_rate": 2e-05,
+      "loss": 0.2812,
+      "step": 44595
+    },
+    {
+      "epoch": 1.0615062363134342,
+      "grad_norm": 8.980158336531934,
+      "learning_rate": 2e-05,
+      "loss": 0.7262,
+      "step": 44596
+    },
+    {
+      "epoch": 1.0615300390364657,
+      "grad_norm": 5.518233386800181,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 44597
+    },
+    {
+      "epoch": 1.0615538417594972,
+      "grad_norm": 8.79974395943535,
+      "learning_rate": 2e-05,
+      "loss": 0.1551,
+      "step": 44598
+    },
+    {
+      "epoch": 1.061577644482529,
+      "grad_norm": 2.0276918899367815,
+      "learning_rate": 2e-05,
+      "loss": 0.0533,
+      "step": 44599
+    },
+    {
+      "epoch": 1.0616014472055604,
+      "grad_norm": 6.223309627241183,
+      "learning_rate": 2e-05,
+      "loss": 0.983,
+      "step": 44600
+    },
+    {
+      "epoch": 1.0616252499285919,
+      "grad_norm": 6.445361334820218,
+      "learning_rate": 2e-05,
+      "loss": 0.195,
+      "step": 44601
+    },
+    {
+      "epoch": 1.0616490526516233,
+      "grad_norm": 0.04368131112622355,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44602
+    },
+    {
+      "epoch": 1.0616728553746548,
+      "grad_norm": 3.662341591545541,
+      "learning_rate": 2e-05,
+      "loss": 0.3998,
+      "step": 44603
+    },
+    {
+      "epoch": 1.0616966580976863,
+      "grad_norm": 0.04118716783952469,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44604
+    },
+    {
+      "epoch": 1.0617204608207178,
+      "grad_norm": 3.2855669354811385,
+      "learning_rate": 2e-05,
+      "loss": 0.4306,
+      "step": 44605
+    },
+    {
+      "epoch": 1.0617442635437495,
+      "grad_norm": 3.6363796347233004,
+      "learning_rate": 2e-05,
+      "loss": 0.1591,
+      "step": 44606
+    },
+    {
+      "epoch": 1.061768066266781,
+      "grad_norm": 2.2372401613581303,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 44607
+    },
+    {
+      "epoch": 1.0617918689898125,
+      "grad_norm": 5.410454548802274,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 44608
+    },
+    {
+      "epoch": 1.061815671712844,
+      "grad_norm": 2.607516079617351,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 44609
+    },
+    {
+      "epoch": 1.0618394744358755,
+      "grad_norm": 4.854693686392207,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 44610
+    },
+    {
+      "epoch": 1.061863277158907,
+      "grad_norm": 0.16994916184242592,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 44611
+    },
+    {
+      "epoch": 1.0618870798819384,
+      "grad_norm": 5.133775358350732,
+      "learning_rate": 2e-05,
+      "loss": 0.6276,
+      "step": 44612
+    },
+    {
+      "epoch": 1.06191088260497,
+      "grad_norm": 0.060089577132790575,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44613
+    },
+    {
+      "epoch": 1.0619346853280016,
+      "grad_norm": 3.069122817331105,
+      "learning_rate": 2e-05,
+      "loss": 0.1659,
+      "step": 44614
+    },
+    {
+      "epoch": 1.061958488051033,
+      "grad_norm": 3.437820787972997,
+      "learning_rate": 2e-05,
+      "loss": 0.3873,
+      "step": 44615
+    },
+    {
+      "epoch": 1.0619822907740646,
+      "grad_norm": 5.072444263777249,
+      "learning_rate": 2e-05,
+      "loss": 0.4916,
+      "step": 44616
+    },
+    {
+      "epoch": 1.062006093497096,
+      "grad_norm": 4.47804539414551,
+      "learning_rate": 2e-05,
+      "loss": 0.1997,
+      "step": 44617
+    },
+    {
+      "epoch": 1.0620298962201276,
+      "grad_norm": 12.969584792473855,
+      "learning_rate": 2e-05,
+      "loss": 0.3898,
+      "step": 44618
+    },
+    {
+      "epoch": 1.062053698943159,
+      "grad_norm": 0.3411401677990578,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 44619
+    },
+    {
+      "epoch": 1.0620775016661905,
+      "grad_norm": 4.05507355179504,
+      "learning_rate": 2e-05,
+      "loss": 0.0992,
+      "step": 44620
+    },
+    {
+      "epoch": 1.0621013043892222,
+      "grad_norm": 3.4597003994692037,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 44621
+    },
+    {
+      "epoch": 1.0621251071122537,
+      "grad_norm": 0.18825899837413926,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44622
+    },
+    {
+      "epoch": 1.0621489098352852,
+      "grad_norm": 1.289551786221308,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 44623
+    },
+    {
+      "epoch": 1.0621727125583167,
+      "grad_norm": 3.0561264277098856,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 44624
+    },
+    {
+      "epoch": 1.0621965152813482,
+      "grad_norm": 9.55593601631464,
+      "learning_rate": 2e-05,
+      "loss": 0.1589,
+      "step": 44625
+    },
+    {
+      "epoch": 1.0622203180043797,
+      "grad_norm": 0.02954756929718015,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44626
+    },
+    {
+      "epoch": 1.0622441207274111,
+      "grad_norm": 5.050620410322476,
+      "learning_rate": 2e-05,
+      "loss": 0.6033,
+      "step": 44627
+    },
+    {
+      "epoch": 1.0622679234504426,
+      "grad_norm": 9.826238581468429,
+      "learning_rate": 2e-05,
+      "loss": 0.7524,
+      "step": 44628
+    },
+    {
+      "epoch": 1.0622917261734743,
+      "grad_norm": 8.303379150058854,
+      "learning_rate": 2e-05,
+      "loss": 0.7515,
+      "step": 44629
+    },
+    {
+      "epoch": 1.0623155288965058,
+      "grad_norm": 5.324501763845186,
+      "learning_rate": 2e-05,
+      "loss": 0.0678,
+      "step": 44630
+    },
+    {
+      "epoch": 1.0623393316195373,
+      "grad_norm": 3.1559321137849667,
+      "learning_rate": 2e-05,
+      "loss": 0.5057,
+      "step": 44631
+    },
+    {
+      "epoch": 1.0623631343425688,
+      "grad_norm": 3.3098441341514806,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 44632
+    },
+    {
+      "epoch": 1.0623869370656003,
+      "grad_norm": 2.6381458769711803,
+      "learning_rate": 2e-05,
+      "loss": 0.0555,
+      "step": 44633
+    },
+    {
+      "epoch": 1.0624107397886318,
+      "grad_norm": 1.2234671795540617,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 44634
+    },
+    {
+      "epoch": 1.0624345425116632,
+      "grad_norm": 1.1231756531397636,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 44635
+    },
+    {
+      "epoch": 1.062458345234695,
+      "grad_norm": 3.6581659240025552,
+      "learning_rate": 2e-05,
+      "loss": 0.4092,
+      "step": 44636
+    },
+    {
+      "epoch": 1.0624821479577264,
+      "grad_norm": 1.8023317491181907,
+      "learning_rate": 2e-05,
+      "loss": 0.0172,
+      "step": 44637
+    },
+    {
+      "epoch": 1.062505950680758,
+      "grad_norm": 5.88531222195505,
+      "learning_rate": 2e-05,
+      "loss": 0.7193,
+      "step": 44638
+    },
+    {
+      "epoch": 1.0625297534037894,
+      "grad_norm": 0.17083583923050136,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 44639
+    },
+    {
+      "epoch": 1.062553556126821,
+      "grad_norm": 3.298615091891224,
+      "learning_rate": 2e-05,
+      "loss": 0.072,
+      "step": 44640
+    },
+    {
+      "epoch": 1.0625773588498524,
+      "grad_norm": 3.0110718400193033,
+      "learning_rate": 2e-05,
+      "loss": 0.0548,
+      "step": 44641
+    },
+    {
+      "epoch": 1.0626011615728839,
+      "grad_norm": 4.53875793426311,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 44642
+    },
+    {
+      "epoch": 1.0626249642959154,
+      "grad_norm": 20.81589870656154,
+      "learning_rate": 2e-05,
+      "loss": 0.5922,
+      "step": 44643
+    },
+    {
+      "epoch": 1.062648767018947,
+      "grad_norm": 2.8231675394277698,
+      "learning_rate": 2e-05,
+      "loss": 0.2209,
+      "step": 44644
+    },
+    {
+      "epoch": 1.0626725697419785,
+      "grad_norm": 5.165779460238943,
+      "learning_rate": 2e-05,
+      "loss": 1.1967,
+      "step": 44645
+    },
+    {
+      "epoch": 1.06269637246501,
+      "grad_norm": 1.516076298068283,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 44646
+    },
+    {
+      "epoch": 1.0627201751880415,
+      "grad_norm": 0.20168770263604385,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 44647
+    },
+    {
+      "epoch": 1.062743977911073,
+      "grad_norm": 3.892714410967781,
+      "learning_rate": 2e-05,
+      "loss": 0.6055,
+      "step": 44648
+    },
+    {
+      "epoch": 1.0627677806341045,
+      "grad_norm": 5.488871082048536,
+      "learning_rate": 2e-05,
+      "loss": 0.4508,
+      "step": 44649
+    },
+    {
+      "epoch": 1.062791583357136,
+      "grad_norm": 5.37020670061924,
+      "learning_rate": 2e-05,
+      "loss": 0.1834,
+      "step": 44650
+    },
+    {
+      "epoch": 1.0628153860801675,
+      "grad_norm": 0.021563656131503576,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44651
+    },
+    {
+      "epoch": 1.0628391888031992,
+      "grad_norm": 1.3145289149550798,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 44652
+    },
+    {
+      "epoch": 1.0628629915262306,
+      "grad_norm": 0.08753314952651857,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44653
+    },
+    {
+      "epoch": 1.0628867942492621,
+      "grad_norm": 4.718881743482303,
+      "learning_rate": 2e-05,
+      "loss": 0.6302,
+      "step": 44654
+    },
+    {
+      "epoch": 1.0629105969722936,
+      "grad_norm": 2.72584068382011,
+      "learning_rate": 2e-05,
+      "loss": 0.369,
+      "step": 44655
+    },
+    {
+      "epoch": 1.062934399695325,
+      "grad_norm": 0.05273407886149064,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 44656
+    },
+    {
+      "epoch": 1.0629582024183566,
+      "grad_norm": 4.479916951543396,
+      "learning_rate": 2e-05,
+      "loss": 0.4731,
+      "step": 44657
+    },
+    {
+      "epoch": 1.062982005141388,
+      "grad_norm": 1.6652268018014826,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 44658
+    },
+    {
+      "epoch": 1.0630058078644198,
+      "grad_norm": 1.5327877326442465,
+      "learning_rate": 2e-05,
+      "loss": 0.0294,
+      "step": 44659
+    },
+    {
+      "epoch": 1.0630296105874513,
+      "grad_norm": 5.1735371984218315,
+      "learning_rate": 2e-05,
+      "loss": 0.1463,
+      "step": 44660
+    },
+    {
+      "epoch": 1.0630534133104828,
+      "grad_norm": 3.2198304413475287,
+      "learning_rate": 2e-05,
+      "loss": 0.1852,
+      "step": 44661
+    },
+    {
+      "epoch": 1.0630772160335142,
+      "grad_norm": 1.6141831572565473,
+      "learning_rate": 2e-05,
+      "loss": 0.0179,
+      "step": 44662
+    },
+    {
+      "epoch": 1.0631010187565457,
+      "grad_norm": 1.2744175828842175,
+      "learning_rate": 2e-05,
+      "loss": 0.0149,
+      "step": 44663
+    },
+    {
+      "epoch": 1.0631248214795772,
+      "grad_norm": 2.017292573757733,
+      "learning_rate": 2e-05,
+      "loss": 0.1724,
+      "step": 44664
+    },
+    {
+      "epoch": 1.0631486242026087,
+      "grad_norm": 0.008781202030709614,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44665
+    },
+    {
+      "epoch": 1.0631724269256404,
+      "grad_norm": 3.356218229585302,
+      "learning_rate": 2e-05,
+      "loss": 0.2818,
+      "step": 44666
+    },
+    {
+      "epoch": 1.0631962296486719,
+      "grad_norm": 0.44045445359507235,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 44667
+    },
+    {
+      "epoch": 1.0632200323717034,
+      "grad_norm": 10.489179994642274,
+      "learning_rate": 2e-05,
+      "loss": 1.2268,
+      "step": 44668
+    },
+    {
+      "epoch": 1.0632438350947349,
+      "grad_norm": 0.012488728960370286,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44669
+    },
+    {
+      "epoch": 1.0632676378177663,
+      "grad_norm": 3.729185300042186,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 44670
+    },
+    {
+      "epoch": 1.0632914405407978,
+      "grad_norm": 0.6906873221202315,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 44671
+    },
+    {
+      "epoch": 1.0633152432638293,
+      "grad_norm": 2.366165636720089,
+      "learning_rate": 2e-05,
+      "loss": 0.0395,
+      "step": 44672
+    },
+    {
+      "epoch": 1.0633390459868608,
+      "grad_norm": 3.679096194027022,
+      "learning_rate": 2e-05,
+      "loss": 0.4383,
+      "step": 44673
+    },
+    {
+      "epoch": 1.0633628487098925,
+      "grad_norm": 6.07250610571886,
+      "learning_rate": 2e-05,
+      "loss": 0.444,
+      "step": 44674
+    },
+    {
+      "epoch": 1.063386651432924,
+      "grad_norm": 6.368231893060062,
+      "learning_rate": 2e-05,
+      "loss": 0.0401,
+      "step": 44675
+    },
+    {
+      "epoch": 1.0634104541559555,
+      "grad_norm": 5.515959928357162,
+      "learning_rate": 2e-05,
+      "loss": 0.6384,
+      "step": 44676
+    },
+    {
+      "epoch": 1.063434256878987,
+      "grad_norm": 0.3528800074412855,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 44677
+    },
+    {
+      "epoch": 1.0634580596020184,
+      "grad_norm": 14.300177732358703,
+      "learning_rate": 2e-05,
+      "loss": 1.2809,
+      "step": 44678
+    },
+    {
+      "epoch": 1.06348186232505,
+      "grad_norm": 5.322374176367566,
+      "learning_rate": 2e-05,
+      "loss": 0.7417,
+      "step": 44679
+    },
+    {
+      "epoch": 1.0635056650480814,
+      "grad_norm": 1.3013034259779492,
+      "learning_rate": 2e-05,
+      "loss": 0.0211,
+      "step": 44680
+    },
+    {
+      "epoch": 1.063529467771113,
+      "grad_norm": 1.570917144272076,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 44681
+    },
+    {
+      "epoch": 1.0635532704941446,
+      "grad_norm": 0.5328518767021908,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 44682
+    },
+    {
+      "epoch": 1.063577073217176,
+      "grad_norm": 1.1723655774702906,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 44683
+    },
+    {
+      "epoch": 1.0636008759402076,
+      "grad_norm": 0.16759761450835306,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44684
+    },
+    {
+      "epoch": 1.063624678663239,
+      "grad_norm": 7.677887986747295,
+      "learning_rate": 2e-05,
+      "loss": 0.2071,
+      "step": 44685
+    },
+    {
+      "epoch": 1.0636484813862706,
+      "grad_norm": 3.9436173323770527,
+      "learning_rate": 2e-05,
+      "loss": 0.1072,
+      "step": 44686
+    },
+    {
+      "epoch": 1.063672284109302,
+      "grad_norm": 0.3589899171000086,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 44687
+    },
+    {
+      "epoch": 1.0636960868323335,
+      "grad_norm": 0.09685453904025451,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44688
+    },
+    {
+      "epoch": 1.0637198895553652,
+      "grad_norm": 0.14017601883579592,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 44689
+    },
+    {
+      "epoch": 1.0637436922783967,
+      "grad_norm": 11.022840169520062,
+      "learning_rate": 2e-05,
+      "loss": 0.3214,
+      "step": 44690
+    },
+    {
+      "epoch": 1.0637674950014282,
+      "grad_norm": 8.087358739575837,
+      "learning_rate": 2e-05,
+      "loss": 0.1588,
+      "step": 44691
+    },
+    {
+      "epoch": 1.0637912977244597,
+      "grad_norm": 1.199082073324185,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 44692
+    },
+    {
+      "epoch": 1.0638151004474912,
+      "grad_norm": 2.8691948247370584,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 44693
+    },
+    {
+      "epoch": 1.0638389031705227,
+      "grad_norm": 2.5312047809504525,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 44694
+    },
+    {
+      "epoch": 1.0638627058935541,
+      "grad_norm": 3.0948895875079976,
+      "learning_rate": 2e-05,
+      "loss": 0.3861,
+      "step": 44695
+    },
+    {
+      "epoch": 1.0638865086165858,
+      "grad_norm": 1.6131385754706782,
+      "learning_rate": 2e-05,
+      "loss": 0.0284,
+      "step": 44696
+    },
+    {
+      "epoch": 1.0639103113396173,
+      "grad_norm": 5.033563996222254,
+      "learning_rate": 2e-05,
+      "loss": 0.199,
+      "step": 44697
+    },
+    {
+      "epoch": 1.0639341140626488,
+      "grad_norm": 0.3840604502501438,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 44698
+    },
+    {
+      "epoch": 1.0639579167856803,
+      "grad_norm": 0.38060903369640325,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 44699
+    },
+    {
+      "epoch": 1.0639817195087118,
+      "grad_norm": 4.90771137598507,
+      "learning_rate": 2e-05,
+      "loss": 0.0586,
+      "step": 44700
+    },
+    {
+      "epoch": 1.0640055222317433,
+      "grad_norm": 5.9700581942239355,
+      "learning_rate": 2e-05,
+      "loss": 0.9399,
+      "step": 44701
+    },
+    {
+      "epoch": 1.0640293249547748,
+      "grad_norm": 1.5398187537104568,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 44702
+    },
+    {
+      "epoch": 1.0640531276778062,
+      "grad_norm": 4.716475436026637,
+      "learning_rate": 2e-05,
+      "loss": 0.3387,
+      "step": 44703
+    },
+    {
+      "epoch": 1.064076930400838,
+      "grad_norm": 4.484938884542067,
+      "learning_rate": 2e-05,
+      "loss": 0.7035,
+      "step": 44704
+    },
+    {
+      "epoch": 1.0641007331238694,
+      "grad_norm": 4.19846437962451,
+      "learning_rate": 2e-05,
+      "loss": 0.1234,
+      "step": 44705
+    },
+    {
+      "epoch": 1.064124535846901,
+      "grad_norm": 1.7206681946678442,
+      "learning_rate": 2e-05,
+      "loss": 0.0353,
+      "step": 44706
+    },
+    {
+      "epoch": 1.0641483385699324,
+      "grad_norm": 1.5560338579738857,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 44707
+    },
+    {
+      "epoch": 1.064172141292964,
+      "grad_norm": 3.542348798988931,
+      "learning_rate": 2e-05,
+      "loss": 0.3071,
+      "step": 44708
+    },
+    {
+      "epoch": 1.0641959440159954,
+      "grad_norm": 5.515481461016738,
+      "learning_rate": 2e-05,
+      "loss": 0.1802,
+      "step": 44709
+    },
+    {
+      "epoch": 1.0642197467390269,
+      "grad_norm": 0.622241696450868,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 44710
+    },
+    {
+      "epoch": 1.0642435494620583,
+      "grad_norm": 0.09214297416924283,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44711
+    },
+    {
+      "epoch": 1.06426735218509,
+      "grad_norm": 2.182243347235063,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 44712
+    },
+    {
+      "epoch": 1.0642911549081215,
+      "grad_norm": 2.899777689714965,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 44713
+    },
+    {
+      "epoch": 1.064314957631153,
+      "grad_norm": 2.3603124346777533,
+      "learning_rate": 2e-05,
+      "loss": 0.1325,
+      "step": 44714
+    },
+    {
+      "epoch": 1.0643387603541845,
+      "grad_norm": 0.6434122970898595,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 44715
+    },
+    {
+      "epoch": 1.064362563077216,
+      "grad_norm": 2.431903395712902,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 44716
+    },
+    {
+      "epoch": 1.0643863658002475,
+      "grad_norm": 4.9743217159942805,
+      "learning_rate": 2e-05,
+      "loss": 0.4553,
+      "step": 44717
+    },
+    {
+      "epoch": 1.064410168523279,
+      "grad_norm": 7.220486193533254,
+      "learning_rate": 2e-05,
+      "loss": 0.3487,
+      "step": 44718
+    },
+    {
+      "epoch": 1.0644339712463107,
+      "grad_norm": 1.8566692586649398,
+      "learning_rate": 2e-05,
+      "loss": 0.0303,
+      "step": 44719
+    },
+    {
+      "epoch": 1.0644577739693422,
+      "grad_norm": 1.4007188210634272,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 44720
+    },
+    {
+      "epoch": 1.0644815766923736,
+      "grad_norm": 4.152119975227064,
+      "learning_rate": 2e-05,
+      "loss": 0.4888,
+      "step": 44721
+    },
+    {
+      "epoch": 1.0645053794154051,
+      "grad_norm": 7.107943111276124,
+      "learning_rate": 2e-05,
+      "loss": 0.2524,
+      "step": 44722
+    },
+    {
+      "epoch": 1.0645291821384366,
+      "grad_norm": 8.865710594229052,
+      "learning_rate": 2e-05,
+      "loss": 0.281,
+      "step": 44723
+    },
+    {
+      "epoch": 1.064552984861468,
+      "grad_norm": 0.024171140604320344,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44724
+    },
+    {
+      "epoch": 1.0645767875844996,
+      "grad_norm": 0.13547269657477262,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 44725
+    },
+    {
+      "epoch": 1.0646005903075313,
+      "grad_norm": 2.616035738011503,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 44726
+    },
+    {
+      "epoch": 1.0646243930305628,
+      "grad_norm": 0.6357867139193982,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 44727
+    },
+    {
+      "epoch": 1.0646481957535943,
+      "grad_norm": 0.14805750688845495,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 44728
+    },
+    {
+      "epoch": 1.0646719984766257,
+      "grad_norm": 3.9331878966457943,
+      "learning_rate": 2e-05,
+      "loss": 0.2295,
+      "step": 44729
+    },
+    {
+      "epoch": 1.0646958011996572,
+      "grad_norm": 0.02079486008393317,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44730
+    },
+    {
+      "epoch": 1.0647196039226887,
+      "grad_norm": 5.9752298033441775,
+      "learning_rate": 2e-05,
+      "loss": 0.567,
+      "step": 44731
+    },
+    {
+      "epoch": 1.0647434066457202,
+      "grad_norm": 8.8223640235575,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 44732
+    },
+    {
+      "epoch": 1.0647672093687517,
+      "grad_norm": 3.5563828840629825,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 44733
+    },
+    {
+      "epoch": 1.0647910120917834,
+      "grad_norm": 2.301058543546855,
+      "learning_rate": 2e-05,
+      "loss": 0.0962,
+      "step": 44734
+    },
+    {
+      "epoch": 1.0648148148148149,
+      "grad_norm": 8.094827701847045,
+      "learning_rate": 2e-05,
+      "loss": 0.5457,
+      "step": 44735
+    },
+    {
+      "epoch": 1.0648386175378464,
+      "grad_norm": 0.9504199477586777,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 44736
+    },
+    {
+      "epoch": 1.0648624202608779,
+      "grad_norm": 6.453776517097653,
+      "learning_rate": 2e-05,
+      "loss": 0.6606,
+      "step": 44737
+    },
+    {
+      "epoch": 1.0648862229839093,
+      "grad_norm": 0.019676177117778453,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44738
+    },
+    {
+      "epoch": 1.0649100257069408,
+      "grad_norm": 0.2848111768203717,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 44739
+    },
+    {
+      "epoch": 1.0649338284299723,
+      "grad_norm": 1.6269616367686843,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 44740
+    },
+    {
+      "epoch": 1.0649576311530038,
+      "grad_norm": 3.9569618954316197,
+      "learning_rate": 2e-05,
+      "loss": 0.453,
+      "step": 44741
+    },
+    {
+      "epoch": 1.0649814338760355,
+      "grad_norm": 2.755745530877933,
+      "learning_rate": 2e-05,
+      "loss": 0.4484,
+      "step": 44742
+    },
+    {
+      "epoch": 1.065005236599067,
+      "grad_norm": 0.6965003941342148,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 44743
+    },
+    {
+      "epoch": 1.0650290393220985,
+      "grad_norm": 2.8469624583085658,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 44744
+    },
+    {
+      "epoch": 1.06505284204513,
+      "grad_norm": 3.0910537007194767,
+      "learning_rate": 2e-05,
+      "loss": 0.2895,
+      "step": 44745
+    },
+    {
+      "epoch": 1.0650766447681614,
+      "grad_norm": 0.03097251052385433,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44746
+    },
+    {
+      "epoch": 1.065100447491193,
+      "grad_norm": 4.91416147746289,
+      "learning_rate": 2e-05,
+      "loss": 0.3239,
+      "step": 44747
+    },
+    {
+      "epoch": 1.0651242502142244,
+      "grad_norm": 0.029711940911416313,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44748
+    },
+    {
+      "epoch": 1.0651480529372561,
+      "grad_norm": 2.450210492261827,
+      "learning_rate": 2e-05,
+      "loss": 0.0842,
+      "step": 44749
+    },
+    {
+      "epoch": 1.0651718556602876,
+      "grad_norm": 1.965081342354475,
+      "learning_rate": 2e-05,
+      "loss": 0.1422,
+      "step": 44750
+    },
+    {
+      "epoch": 1.065195658383319,
+      "grad_norm": 6.4439027442071835,
+      "learning_rate": 2e-05,
+      "loss": 0.2883,
+      "step": 44751
+    },
+    {
+      "epoch": 1.0652194611063506,
+      "grad_norm": 3.8846890915315044,
+      "learning_rate": 2e-05,
+      "loss": 0.4953,
+      "step": 44752
+    },
+    {
+      "epoch": 1.065243263829382,
+      "grad_norm": 3.6513111905281392,
+      "learning_rate": 2e-05,
+      "loss": 0.3888,
+      "step": 44753
+    },
+    {
+      "epoch": 1.0652670665524135,
+      "grad_norm": 6.924103859133428,
+      "learning_rate": 2e-05,
+      "loss": 0.8116,
+      "step": 44754
+    },
+    {
+      "epoch": 1.065290869275445,
+      "grad_norm": 3.3378829118388054,
+      "learning_rate": 2e-05,
+      "loss": 0.051,
+      "step": 44755
+    },
+    {
+      "epoch": 1.0653146719984767,
+      "grad_norm": 0.6222874840718124,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 44756
+    },
+    {
+      "epoch": 1.0653384747215082,
+      "grad_norm": 3.627082829642306,
+      "learning_rate": 2e-05,
+      "loss": 0.2258,
+      "step": 44757
+    },
+    {
+      "epoch": 1.0653622774445397,
+      "grad_norm": 0.5523594828006114,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 44758
+    },
+    {
+      "epoch": 1.0653860801675712,
+      "grad_norm": 8.932574264169078,
+      "learning_rate": 2e-05,
+      "loss": 0.438,
+      "step": 44759
+    },
+    {
+      "epoch": 1.0654098828906027,
+      "grad_norm": 16.71115680425966,
+      "learning_rate": 2e-05,
+      "loss": 0.6597,
+      "step": 44760
+    },
+    {
+      "epoch": 1.0654336856136342,
+      "grad_norm": 1.108248228306962,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 44761
+    },
+    {
+      "epoch": 1.0654574883366656,
+      "grad_norm": 6.791717686342274,
+      "learning_rate": 2e-05,
+      "loss": 0.7537,
+      "step": 44762
+    },
+    {
+      "epoch": 1.0654812910596971,
+      "grad_norm": 0.9316939369384589,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 44763
+    },
+    {
+      "epoch": 1.0655050937827288,
+      "grad_norm": 4.4930513601447215,
+      "learning_rate": 2e-05,
+      "loss": 0.2277,
+      "step": 44764
+    },
+    {
+      "epoch": 1.0655288965057603,
+      "grad_norm": 6.701339209608715,
+      "learning_rate": 2e-05,
+      "loss": 1.1918,
+      "step": 44765
+    },
+    {
+      "epoch": 1.0655526992287918,
+      "grad_norm": 9.664639902988823,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 44766
+    },
+    {
+      "epoch": 1.0655765019518233,
+      "grad_norm": 3.026853023016471,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 44767
+    },
+    {
+      "epoch": 1.0656003046748548,
+      "grad_norm": 0.5307699997855798,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 44768
+    },
+    {
+      "epoch": 1.0656241073978863,
+      "grad_norm": 2.5081468751426663,
+      "learning_rate": 2e-05,
+      "loss": 0.0627,
+      "step": 44769
+    },
+    {
+      "epoch": 1.0656479101209178,
+      "grad_norm": 5.9215433690908,
+      "learning_rate": 2e-05,
+      "loss": 0.6044,
+      "step": 44770
+    },
+    {
+      "epoch": 1.0656717128439492,
+      "grad_norm": 4.166533625023452,
+      "learning_rate": 2e-05,
+      "loss": 0.3285,
+      "step": 44771
+    },
+    {
+      "epoch": 1.065695515566981,
+      "grad_norm": 0.10068825008611311,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 44772
+    },
+    {
+      "epoch": 1.0657193182900124,
+      "grad_norm": 0.014588772414772865,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44773
+    },
+    {
+      "epoch": 1.065743121013044,
+      "grad_norm": 0.12864730530721746,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44774
+    },
+    {
+      "epoch": 1.0657669237360754,
+      "grad_norm": 0.4994501560227274,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 44775
+    },
+    {
+      "epoch": 1.0657907264591069,
+      "grad_norm": 2.803811601641993,
+      "learning_rate": 2e-05,
+      "loss": 0.137,
+      "step": 44776
+    },
+    {
+      "epoch": 1.0658145291821384,
+      "grad_norm": 6.6896929065665365,
+      "learning_rate": 2e-05,
+      "loss": 0.363,
+      "step": 44777
+    },
+    {
+      "epoch": 1.0658383319051699,
+      "grad_norm": 4.634854858719998,
+      "learning_rate": 2e-05,
+      "loss": 0.5284,
+      "step": 44778
+    },
+    {
+      "epoch": 1.0658621346282016,
+      "grad_norm": 1.379618822551267,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 44779
+    },
+    {
+      "epoch": 1.065885937351233,
+      "grad_norm": 3.066893637880125,
+      "learning_rate": 2e-05,
+      "loss": 0.3947,
+      "step": 44780
+    },
+    {
+      "epoch": 1.0659097400742645,
+      "grad_norm": 1.022968179834888,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 44781
+    },
+    {
+      "epoch": 1.065933542797296,
+      "grad_norm": 0.7964572369714381,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 44782
+    },
+    {
+      "epoch": 1.0659573455203275,
+      "grad_norm": 0.06406289699314457,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44783
+    },
+    {
+      "epoch": 1.065981148243359,
+      "grad_norm": 8.088252603029812,
+      "learning_rate": 2e-05,
+      "loss": 0.7122,
+      "step": 44784
+    },
+    {
+      "epoch": 1.0660049509663905,
+      "grad_norm": 0.013451890865069,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44785
+    },
+    {
+      "epoch": 1.0660287536894222,
+      "grad_norm": 4.68102253460078,
+      "learning_rate": 2e-05,
+      "loss": 0.2565,
+      "step": 44786
+    },
+    {
+      "epoch": 1.0660525564124537,
+      "grad_norm": 5.674074132302522,
+      "learning_rate": 2e-05,
+      "loss": 0.0822,
+      "step": 44787
+    },
+    {
+      "epoch": 1.0660763591354852,
+      "grad_norm": 18.48749761381928,
+      "learning_rate": 2e-05,
+      "loss": 0.3522,
+      "step": 44788
+    },
+    {
+      "epoch": 1.0661001618585166,
+      "grad_norm": 6.048434586564535,
+      "learning_rate": 2e-05,
+      "loss": 0.3494,
+      "step": 44789
+    },
+    {
+      "epoch": 1.0661239645815481,
+      "grad_norm": 0.2798603425678237,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 44790
+    },
+    {
+      "epoch": 1.0661477673045796,
+      "grad_norm": 5.11185525666024,
+      "learning_rate": 2e-05,
+      "loss": 0.3513,
+      "step": 44791
+    },
+    {
+      "epoch": 1.066171570027611,
+      "grad_norm": 3.6381443741245754,
+      "learning_rate": 2e-05,
+      "loss": 0.2462,
+      "step": 44792
+    },
+    {
+      "epoch": 1.0661953727506426,
+      "grad_norm": 8.090908003856013,
+      "learning_rate": 2e-05,
+      "loss": 0.3233,
+      "step": 44793
+    },
+    {
+      "epoch": 1.0662191754736743,
+      "grad_norm": 0.34161076270554097,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 44794
+    },
+    {
+      "epoch": 1.0662429781967058,
+      "grad_norm": 1.0594168810126083,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 44795
+    },
+    {
+      "epoch": 1.0662667809197373,
+      "grad_norm": 0.13731163989079892,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 44796
+    },
+    {
+      "epoch": 1.0662905836427687,
+      "grad_norm": 6.504294282235219,
+      "learning_rate": 2e-05,
+      "loss": 0.9479,
+      "step": 44797
+    },
+    {
+      "epoch": 1.0663143863658002,
+      "grad_norm": 2.785469873954981,
+      "learning_rate": 2e-05,
+      "loss": 0.0953,
+      "step": 44798
+    },
+    {
+      "epoch": 1.0663381890888317,
+      "grad_norm": 2.645059377849055,
+      "learning_rate": 2e-05,
+      "loss": 0.0223,
+      "step": 44799
+    },
+    {
+      "epoch": 1.0663619918118632,
+      "grad_norm": 6.372658142190367,
+      "learning_rate": 2e-05,
+      "loss": 1.0831,
+      "step": 44800
+    },
+    {
+      "epoch": 1.0663857945348947,
+      "grad_norm": 2.813944899038073,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 44801
+    },
+    {
+      "epoch": 1.0664095972579264,
+      "grad_norm": 2.2655847007860594,
+      "learning_rate": 2e-05,
+      "loss": 0.14,
+      "step": 44802
+    },
+    {
+      "epoch": 1.0664333999809579,
+      "grad_norm": 2.522379168569292,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 44803
+    },
+    {
+      "epoch": 1.0664572027039894,
+      "grad_norm": 3.459226814973724,
+      "learning_rate": 2e-05,
+      "loss": 0.5802,
+      "step": 44804
+    },
+    {
+      "epoch": 1.0664810054270208,
+      "grad_norm": 0.26613507658153285,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 44805
+    },
+    {
+      "epoch": 1.0665048081500523,
+      "grad_norm": 4.534905140092684,
+      "learning_rate": 2e-05,
+      "loss": 0.5741,
+      "step": 44806
+    },
+    {
+      "epoch": 1.0665286108730838,
+      "grad_norm": 10.079346559427437,
+      "learning_rate": 2e-05,
+      "loss": 0.4889,
+      "step": 44807
+    },
+    {
+      "epoch": 1.0665524135961153,
+      "grad_norm": 4.140775751260834,
+      "learning_rate": 2e-05,
+      "loss": 0.4914,
+      "step": 44808
+    },
+    {
+      "epoch": 1.066576216319147,
+      "grad_norm": 2.5020592635189787,
+      "learning_rate": 2e-05,
+      "loss": 0.203,
+      "step": 44809
+    },
+    {
+      "epoch": 1.0666000190421785,
+      "grad_norm": 5.080365060587175,
+      "learning_rate": 2e-05,
+      "loss": 0.5136,
+      "step": 44810
+    },
+    {
+      "epoch": 1.06662382176521,
+      "grad_norm": 2.881288135135692,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 44811
+    },
+    {
+      "epoch": 1.0666476244882415,
+      "grad_norm": 0.03073775515645492,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 44812
+    },
+    {
+      "epoch": 1.066671427211273,
+      "grad_norm": 1.042481049821986,
+      "learning_rate": 2e-05,
+      "loss": 0.0208,
+      "step": 44813
+    },
+    {
+      "epoch": 1.0666952299343044,
+      "grad_norm": 9.13567023234815,
+      "learning_rate": 2e-05,
+      "loss": 1.1671,
+      "step": 44814
+    },
+    {
+      "epoch": 1.066719032657336,
+      "grad_norm": 0.4947614246037226,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 44815
+    },
+    {
+      "epoch": 1.0667428353803676,
+      "grad_norm": 2.859260300091645,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 44816
+    },
+    {
+      "epoch": 1.0667666381033991,
+      "grad_norm": 7.120505945547396,
+      "learning_rate": 2e-05,
+      "loss": 1.1791,
+      "step": 44817
+    },
+    {
+      "epoch": 1.0667904408264306,
+      "grad_norm": 3.6388893387992693,
+      "learning_rate": 2e-05,
+      "loss": 0.2569,
+      "step": 44818
+    },
+    {
+      "epoch": 1.066814243549462,
+      "grad_norm": 0.4454483467592826,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 44819
+    },
+    {
+      "epoch": 1.0668380462724936,
+      "grad_norm": 2.6499609850976436,
+      "learning_rate": 2e-05,
+      "loss": 0.2037,
+      "step": 44820
+    },
+    {
+      "epoch": 1.066861848995525,
+      "grad_norm": 5.157643949629726,
+      "learning_rate": 2e-05,
+      "loss": 0.6282,
+      "step": 44821
+    },
+    {
+      "epoch": 1.0668856517185565,
+      "grad_norm": 2.4881545082018017,
+      "learning_rate": 2e-05,
+      "loss": 0.1801,
+      "step": 44822
+    },
+    {
+      "epoch": 1.066909454441588,
+      "grad_norm": 3.8643928438173094,
+      "learning_rate": 2e-05,
+      "loss": 0.1553,
+      "step": 44823
+    },
+    {
+      "epoch": 1.0669332571646197,
+      "grad_norm": 0.46082294390916584,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 44824
+    },
+    {
+      "epoch": 1.0669570598876512,
+      "grad_norm": 2.8763742050169623,
+      "learning_rate": 2e-05,
+      "loss": 0.5477,
+      "step": 44825
+    },
+    {
+      "epoch": 1.0669808626106827,
+      "grad_norm": 2.827236146913892,
+      "learning_rate": 2e-05,
+      "loss": 0.7146,
+      "step": 44826
+    },
+    {
+      "epoch": 1.0670046653337142,
+      "grad_norm": 0.019991597757758703,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44827
+    },
+    {
+      "epoch": 1.0670284680567457,
+      "grad_norm": 8.366111035757074,
+      "learning_rate": 2e-05,
+      "loss": 0.3691,
+      "step": 44828
+    },
+    {
+      "epoch": 1.0670522707797772,
+      "grad_norm": 9.179381635036622,
+      "learning_rate": 2e-05,
+      "loss": 0.6828,
+      "step": 44829
+    },
+    {
+      "epoch": 1.0670760735028086,
+      "grad_norm": 0.07724224963177148,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44830
+    },
+    {
+      "epoch": 1.0670998762258401,
+      "grad_norm": 1.5033652526728796,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 44831
+    },
+    {
+      "epoch": 1.0671236789488718,
+      "grad_norm": 2.741779200348206,
+      "learning_rate": 2e-05,
+      "loss": 0.1693,
+      "step": 44832
+    },
+    {
+      "epoch": 1.0671474816719033,
+      "grad_norm": 3.0719752754731173,
+      "learning_rate": 2e-05,
+      "loss": 0.2889,
+      "step": 44833
+    },
+    {
+      "epoch": 1.0671712843949348,
+      "grad_norm": 3.4086771869764627,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 44834
+    },
+    {
+      "epoch": 1.0671950871179663,
+      "grad_norm": 5.5035407374922745,
+      "learning_rate": 2e-05,
+      "loss": 0.527,
+      "step": 44835
+    },
+    {
+      "epoch": 1.0672188898409978,
+      "grad_norm": 3.530064459899805,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 44836
+    },
+    {
+      "epoch": 1.0672426925640293,
+      "grad_norm": 4.877892288293273,
+      "learning_rate": 2e-05,
+      "loss": 0.0603,
+      "step": 44837
+    },
+    {
+      "epoch": 1.0672664952870607,
+      "grad_norm": 5.495283284167792,
+      "learning_rate": 2e-05,
+      "loss": 0.4134,
+      "step": 44838
+    },
+    {
+      "epoch": 1.0672902980100925,
+      "grad_norm": 3.3742266209084883,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 44839
+    },
+    {
+      "epoch": 1.067314100733124,
+      "grad_norm": 3.8170833442481125,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 44840
+    },
+    {
+      "epoch": 1.0673379034561554,
+      "grad_norm": 0.08954518168344539,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44841
+    },
+    {
+      "epoch": 1.067361706179187,
+      "grad_norm": 0.013830980687621756,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44842
+    },
+    {
+      "epoch": 1.0673855089022184,
+      "grad_norm": 3.560364659336459,
+      "learning_rate": 2e-05,
+      "loss": 0.2513,
+      "step": 44843
+    },
+    {
+      "epoch": 1.0674093116252499,
+      "grad_norm": 0.4442596862216703,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 44844
+    },
+    {
+      "epoch": 1.0674331143482814,
+      "grad_norm": 2.8466462616049526,
+      "learning_rate": 2e-05,
+      "loss": 0.0445,
+      "step": 44845
+    },
+    {
+      "epoch": 1.067456917071313,
+      "grad_norm": 1.9430825120553796,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 44846
+    },
+    {
+      "epoch": 1.0674807197943446,
+      "grad_norm": 1.6173727200749892,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 44847
+    },
+    {
+      "epoch": 1.067504522517376,
+      "grad_norm": 2.132910612908479,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 44848
+    },
+    {
+      "epoch": 1.0675283252404075,
+      "grad_norm": 3.599795725645096,
+      "learning_rate": 2e-05,
+      "loss": 0.3067,
+      "step": 44849
+    },
+    {
+      "epoch": 1.067552127963439,
+      "grad_norm": 1.1277076316869485,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 44850
+    },
+    {
+      "epoch": 1.0675759306864705,
+      "grad_norm": 2.6245091886194456,
+      "learning_rate": 2e-05,
+      "loss": 0.3185,
+      "step": 44851
+    },
+    {
+      "epoch": 1.067599733409502,
+      "grad_norm": 2.211539699820915,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 44852
+    },
+    {
+      "epoch": 1.0676235361325335,
+      "grad_norm": 7.343484078158462,
+      "learning_rate": 2e-05,
+      "loss": 0.273,
+      "step": 44853
+    },
+    {
+      "epoch": 1.0676473388555652,
+      "grad_norm": 3.3892401194603163,
+      "learning_rate": 2e-05,
+      "loss": 0.5067,
+      "step": 44854
+    },
+    {
+      "epoch": 1.0676711415785967,
+      "grad_norm": 5.813195734538993,
+      "learning_rate": 2e-05,
+      "loss": 0.935,
+      "step": 44855
+    },
+    {
+      "epoch": 1.0676949443016281,
+      "grad_norm": 5.395012839436707,
+      "learning_rate": 2e-05,
+      "loss": 0.3698,
+      "step": 44856
+    },
+    {
+      "epoch": 1.0677187470246596,
+      "grad_norm": 5.527260906334515,
+      "learning_rate": 2e-05,
+      "loss": 0.9693,
+      "step": 44857
+    },
+    {
+      "epoch": 1.0677425497476911,
+      "grad_norm": 0.08309114339065819,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44858
+    },
+    {
+      "epoch": 1.0677663524707226,
+      "grad_norm": 0.5087463945235153,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 44859
+    },
+    {
+      "epoch": 1.067790155193754,
+      "grad_norm": 0.9650666028682383,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 44860
+    },
+    {
+      "epoch": 1.0678139579167856,
+      "grad_norm": 3.7170553930143972,
+      "learning_rate": 2e-05,
+      "loss": 0.4252,
+      "step": 44861
+    },
+    {
+      "epoch": 1.0678377606398173,
+      "grad_norm": 1.1942060008879352,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 44862
+    },
+    {
+      "epoch": 1.0678615633628488,
+      "grad_norm": 0.13089842868304485,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44863
+    },
+    {
+      "epoch": 1.0678853660858802,
+      "grad_norm": 3.6583613282696286,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 44864
+    },
+    {
+      "epoch": 1.0679091688089117,
+      "grad_norm": 7.005333735957334,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 44865
+    },
+    {
+      "epoch": 1.0679329715319432,
+      "grad_norm": 0.012333753842335475,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 44866
+    },
+    {
+      "epoch": 1.0679567742549747,
+      "grad_norm": 0.00616377545500158,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 44867
+    },
+    {
+      "epoch": 1.0679805769780062,
+      "grad_norm": 2.5260543825983834,
+      "learning_rate": 2e-05,
+      "loss": 0.2614,
+      "step": 44868
+    },
+    {
+      "epoch": 1.068004379701038,
+      "grad_norm": 4.26379542589775,
+      "learning_rate": 2e-05,
+      "loss": 0.4019,
+      "step": 44869
+    },
+    {
+      "epoch": 1.0680281824240694,
+      "grad_norm": 0.15641780369652308,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 44870
+    },
+    {
+      "epoch": 1.0680519851471009,
+      "grad_norm": 0.17449214945452907,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 44871
+    },
+    {
+      "epoch": 1.0680757878701324,
+      "grad_norm": 28.49519144574918,
+      "learning_rate": 2e-05,
+      "loss": 0.5376,
+      "step": 44872
+    },
+    {
+      "epoch": 1.0680995905931638,
+      "grad_norm": 3.350726380214968,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 44873
+    },
+    {
+      "epoch": 1.0681233933161953,
+      "grad_norm": 2.6875618178727363,
+      "learning_rate": 2e-05,
+      "loss": 0.024,
+      "step": 44874
+    },
+    {
+      "epoch": 1.0681471960392268,
+      "grad_norm": 2.7829973228916085,
+      "learning_rate": 2e-05,
+      "loss": 0.3328,
+      "step": 44875
+    },
+    {
+      "epoch": 1.0681709987622585,
+      "grad_norm": 0.12445292630451575,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44876
+    },
+    {
+      "epoch": 1.06819480148529,
+      "grad_norm": 2.3703307978587467,
+      "learning_rate": 2e-05,
+      "loss": 0.3194,
+      "step": 44877
+    },
+    {
+      "epoch": 1.0682186042083215,
+      "grad_norm": 7.29813218214568,
+      "learning_rate": 2e-05,
+      "loss": 0.6354,
+      "step": 44878
+    },
+    {
+      "epoch": 1.068242406931353,
+      "grad_norm": 3.2492132759730135,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 44879
+    },
+    {
+      "epoch": 1.0682662096543845,
+      "grad_norm": 0.6253224060562214,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 44880
+    },
+    {
+      "epoch": 1.068290012377416,
+      "grad_norm": 0.07002796911202036,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44881
+    },
+    {
+      "epoch": 1.0683138151004474,
+      "grad_norm": 2.944162126836555,
+      "learning_rate": 2e-05,
+      "loss": 0.2028,
+      "step": 44882
+    },
+    {
+      "epoch": 1.068337617823479,
+      "grad_norm": 2.119033225950385,
+      "learning_rate": 2e-05,
+      "loss": 0.2204,
+      "step": 44883
+    },
+    {
+      "epoch": 1.0683614205465106,
+      "grad_norm": 0.431562678429672,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 44884
+    },
+    {
+      "epoch": 1.068385223269542,
+      "grad_norm": 0.3926818560860389,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 44885
+    },
+    {
+      "epoch": 1.0684090259925736,
+      "grad_norm": 0.013645254438684053,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44886
+    },
+    {
+      "epoch": 1.068432828715605,
+      "grad_norm": 3.6413638777578696,
+      "learning_rate": 2e-05,
+      "loss": 0.2092,
+      "step": 44887
+    },
+    {
+      "epoch": 1.0684566314386366,
+      "grad_norm": 2.5662437798615465,
+      "learning_rate": 2e-05,
+      "loss": 0.2844,
+      "step": 44888
+    },
+    {
+      "epoch": 1.068480434161668,
+      "grad_norm": 0.5735805433575977,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 44889
+    },
+    {
+      "epoch": 1.0685042368846995,
+      "grad_norm": 5.197597374067563,
+      "learning_rate": 2e-05,
+      "loss": 0.4111,
+      "step": 44890
+    },
+    {
+      "epoch": 1.068528039607731,
+      "grad_norm": 3.1395742383354834,
+      "learning_rate": 2e-05,
+      "loss": 0.2876,
+      "step": 44891
+    },
+    {
+      "epoch": 1.0685518423307627,
+      "grad_norm": 5.173727536459524,
+      "learning_rate": 2e-05,
+      "loss": 0.5377,
+      "step": 44892
+    },
+    {
+      "epoch": 1.0685756450537942,
+      "grad_norm": 3.3488437788274656,
+      "learning_rate": 2e-05,
+      "loss": 0.2948,
+      "step": 44893
+    },
+    {
+      "epoch": 1.0685994477768257,
+      "grad_norm": 1.2459617957843085,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 44894
+    },
+    {
+      "epoch": 1.0686232504998572,
+      "grad_norm": 0.06200664659299088,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 44895
+    },
+    {
+      "epoch": 1.0686470532228887,
+      "grad_norm": 6.576870826563543,
+      "learning_rate": 2e-05,
+      "loss": 0.7514,
+      "step": 44896
+    },
+    {
+      "epoch": 1.0686708559459202,
+      "grad_norm": 0.07661478153656728,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 44897
+    },
+    {
+      "epoch": 1.0686946586689516,
+      "grad_norm": 1.3277167969535502,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 44898
+    },
+    {
+      "epoch": 1.0687184613919833,
+      "grad_norm": 11.40022807590567,
+      "learning_rate": 2e-05,
+      "loss": 0.5613,
+      "step": 44899
+    },
+    {
+      "epoch": 1.0687422641150148,
+      "grad_norm": 0.09292338592971436,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 44900
+    },
+    {
+      "epoch": 1.0687660668380463,
+      "grad_norm": 24.07033466018825,
+      "learning_rate": 2e-05,
+      "loss": 0.3345,
+      "step": 44901
+    },
+    {
+      "epoch": 1.0687898695610778,
+      "grad_norm": 11.279191804942542,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 44902
+    },
+    {
+      "epoch": 1.0688136722841093,
+      "grad_norm": 5.322939445872234,
+      "learning_rate": 2e-05,
+      "loss": 0.9534,
+      "step": 44903
+    },
+    {
+      "epoch": 1.0688374750071408,
+      "grad_norm": 0.04553347967902517,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 44904
+    },
+    {
+      "epoch": 1.0688612777301723,
+      "grad_norm": 5.1313146651635035,
+      "learning_rate": 2e-05,
+      "loss": 0.347,
+      "step": 44905
+    },
+    {
+      "epoch": 1.068885080453204,
+      "grad_norm": 5.62797324958204,
+      "learning_rate": 2e-05,
+      "loss": 0.92,
+      "step": 44906
+    },
+    {
+      "epoch": 1.0689088831762354,
+      "grad_norm": 0.12426761637891914,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 44907
+    },
+    {
+      "epoch": 1.068932685899267,
+      "grad_norm": 6.007478781397699,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 44908
+    },
+    {
+      "epoch": 1.0689564886222984,
+      "grad_norm": 12.242646407974368,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 44909
+    },
+    {
+      "epoch": 1.06898029134533,
+      "grad_norm": 8.909887454190333,
+      "learning_rate": 2e-05,
+      "loss": 0.7096,
+      "step": 44910
+    },
+    {
+      "epoch": 1.0690040940683614,
+      "grad_norm": 14.678233073001152,
+      "learning_rate": 2e-05,
+      "loss": 0.7124,
+      "step": 44911
+    },
+    {
+      "epoch": 1.0690278967913929,
+      "grad_norm": 0.12613046752458848,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44912
+    },
+    {
+      "epoch": 1.0690516995144244,
+      "grad_norm": 2.885508611881603,
+      "learning_rate": 2e-05,
+      "loss": 0.2375,
+      "step": 44913
+    },
+    {
+      "epoch": 1.069075502237456,
+      "grad_norm": 2.580359160250365,
+      "learning_rate": 2e-05,
+      "loss": 0.3148,
+      "step": 44914
+    },
+    {
+      "epoch": 1.0690993049604876,
+      "grad_norm": 1.9959307146124774,
+      "learning_rate": 2e-05,
+      "loss": 0.2828,
+      "step": 44915
+    },
+    {
+      "epoch": 1.069123107683519,
+      "grad_norm": 2.683643941753027,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 44916
+    },
+    {
+      "epoch": 1.0691469104065505,
+      "grad_norm": 4.203962108402757,
+      "learning_rate": 2e-05,
+      "loss": 0.0734,
+      "step": 44917
+    },
+    {
+      "epoch": 1.069170713129582,
+      "grad_norm": 1.1682525650715496,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 44918
+    },
+    {
+      "epoch": 1.0691945158526135,
+      "grad_norm": 8.0065079185293,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 44919
+    },
+    {
+      "epoch": 1.069218318575645,
+      "grad_norm": 5.577401225277206,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 44920
+    },
+    {
+      "epoch": 1.0692421212986765,
+      "grad_norm": 1.3226961718346646,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 44921
+    },
+    {
+      "epoch": 1.0692659240217082,
+      "grad_norm": 2.054071661692408,
+      "learning_rate": 2e-05,
+      "loss": 0.1146,
+      "step": 44922
+    },
+    {
+      "epoch": 1.0692897267447397,
+      "grad_norm": 15.521865769334244,
+      "learning_rate": 2e-05,
+      "loss": 0.1882,
+      "step": 44923
+    },
+    {
+      "epoch": 1.0693135294677711,
+      "grad_norm": 6.170047793803545,
+      "learning_rate": 2e-05,
+      "loss": 1.1281,
+      "step": 44924
+    },
+    {
+      "epoch": 1.0693373321908026,
+      "grad_norm": 3.0215732487136413,
+      "learning_rate": 2e-05,
+      "loss": 0.4072,
+      "step": 44925
+    },
+    {
+      "epoch": 1.069361134913834,
+      "grad_norm": 10.712435843658142,
+      "learning_rate": 2e-05,
+      "loss": 0.5899,
+      "step": 44926
+    },
+    {
+      "epoch": 1.0693849376368656,
+      "grad_norm": 3.813284261574643,
+      "learning_rate": 2e-05,
+      "loss": 0.6813,
+      "step": 44927
+    },
+    {
+      "epoch": 1.069408740359897,
+      "grad_norm": 1.44624913557285,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 44928
+    },
+    {
+      "epoch": 1.0694325430829288,
+      "grad_norm": 5.955039022992793,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 44929
+    },
+    {
+      "epoch": 1.0694563458059603,
+      "grad_norm": 0.5440938320434943,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 44930
+    },
+    {
+      "epoch": 1.0694801485289918,
+      "grad_norm": 12.432645694796413,
+      "learning_rate": 2e-05,
+      "loss": 0.9362,
+      "step": 44931
+    },
+    {
+      "epoch": 1.0695039512520232,
+      "grad_norm": 5.540683458149935,
+      "learning_rate": 2e-05,
+      "loss": 0.5836,
+      "step": 44932
+    },
+    {
+      "epoch": 1.0695277539750547,
+      "grad_norm": 3.438207506715931,
+      "learning_rate": 2e-05,
+      "loss": 0.3752,
+      "step": 44933
+    },
+    {
+      "epoch": 1.0695515566980862,
+      "grad_norm": 8.791765094811698,
+      "learning_rate": 2e-05,
+      "loss": 0.5082,
+      "step": 44934
+    },
+    {
+      "epoch": 1.0695753594211177,
+      "grad_norm": 1.6295322749975318,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 44935
+    },
+    {
+      "epoch": 1.0695991621441494,
+      "grad_norm": 0.4073189838557331,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 44936
+    },
+    {
+      "epoch": 1.069622964867181,
+      "grad_norm": 7.4696126391610935,
+      "learning_rate": 2e-05,
+      "loss": 0.1381,
+      "step": 44937
+    },
+    {
+      "epoch": 1.0696467675902124,
+      "grad_norm": 0.026644211250705232,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44938
+    },
+    {
+      "epoch": 1.0696705703132439,
+      "grad_norm": 5.703220154253129,
+      "learning_rate": 2e-05,
+      "loss": 0.6023,
+      "step": 44939
+    },
+    {
+      "epoch": 1.0696943730362753,
+      "grad_norm": 7.235641340871147,
+      "learning_rate": 2e-05,
+      "loss": 0.1794,
+      "step": 44940
+    },
+    {
+      "epoch": 1.0697181757593068,
+      "grad_norm": 3.2818838373734653,
+      "learning_rate": 2e-05,
+      "loss": 0.2895,
+      "step": 44941
+    },
+    {
+      "epoch": 1.0697419784823383,
+      "grad_norm": 4.240077878013485,
+      "learning_rate": 2e-05,
+      "loss": 0.4264,
+      "step": 44942
+    },
+    {
+      "epoch": 1.0697657812053698,
+      "grad_norm": 0.9420011107051762,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 44943
+    },
+    {
+      "epoch": 1.0697895839284015,
+      "grad_norm": 2.8665253058787274,
+      "learning_rate": 2e-05,
+      "loss": 0.3974,
+      "step": 44944
+    },
+    {
+      "epoch": 1.069813386651433,
+      "grad_norm": 0.7863643377673786,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 44945
+    },
+    {
+      "epoch": 1.0698371893744645,
+      "grad_norm": 3.602584057349023,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 44946
+    },
+    {
+      "epoch": 1.069860992097496,
+      "grad_norm": 6.205932913922628,
+      "learning_rate": 2e-05,
+      "loss": 0.7913,
+      "step": 44947
+    },
+    {
+      "epoch": 1.0698847948205275,
+      "grad_norm": 2.6598044867647004,
+      "learning_rate": 2e-05,
+      "loss": 0.1203,
+      "step": 44948
+    },
+    {
+      "epoch": 1.069908597543559,
+      "grad_norm": 13.993914737104411,
+      "learning_rate": 2e-05,
+      "loss": 1.1124,
+      "step": 44949
+    },
+    {
+      "epoch": 1.0699324002665904,
+      "grad_norm": 6.82869691055491,
+      "learning_rate": 2e-05,
+      "loss": 0.9199,
+      "step": 44950
+    },
+    {
+      "epoch": 1.069956202989622,
+      "grad_norm": 2.283686252484524,
+      "learning_rate": 2e-05,
+      "loss": 0.0152,
+      "step": 44951
+    },
+    {
+      "epoch": 1.0699800057126536,
+      "grad_norm": 0.36773129708567165,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 44952
+    },
+    {
+      "epoch": 1.070003808435685,
+      "grad_norm": 2.6776341767381426,
+      "learning_rate": 2e-05,
+      "loss": 0.0623,
+      "step": 44953
+    },
+    {
+      "epoch": 1.0700276111587166,
+      "grad_norm": 4.608845320997499,
+      "learning_rate": 2e-05,
+      "loss": 0.5453,
+      "step": 44954
+    },
+    {
+      "epoch": 1.070051413881748,
+      "grad_norm": 0.3333387417715003,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 44955
+    },
+    {
+      "epoch": 1.0700752166047796,
+      "grad_norm": 2.3751826628472426,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 44956
+    },
+    {
+      "epoch": 1.070099019327811,
+      "grad_norm": 4.04335610886858,
+      "learning_rate": 2e-05,
+      "loss": 0.4884,
+      "step": 44957
+    },
+    {
+      "epoch": 1.0701228220508425,
+      "grad_norm": 3.5015417058942364,
+      "learning_rate": 2e-05,
+      "loss": 0.5214,
+      "step": 44958
+    },
+    {
+      "epoch": 1.0701466247738742,
+      "grad_norm": 4.96131035471111,
+      "learning_rate": 2e-05,
+      "loss": 0.7636,
+      "step": 44959
+    },
+    {
+      "epoch": 1.0701704274969057,
+      "grad_norm": 4.022879716453445,
+      "learning_rate": 2e-05,
+      "loss": 0.1201,
+      "step": 44960
+    },
+    {
+      "epoch": 1.0701942302199372,
+      "grad_norm": 0.3678232056991612,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 44961
+    },
+    {
+      "epoch": 1.0702180329429687,
+      "grad_norm": 6.015375703965612,
+      "learning_rate": 2e-05,
+      "loss": 0.7876,
+      "step": 44962
+    },
+    {
+      "epoch": 1.0702418356660002,
+      "grad_norm": 0.510049177787683,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 44963
+    },
+    {
+      "epoch": 1.0702656383890317,
+      "grad_norm": 2.972835439016149,
+      "learning_rate": 2e-05,
+      "loss": 0.3909,
+      "step": 44964
+    },
+    {
+      "epoch": 1.0702894411120631,
+      "grad_norm": 1.4237646546508642,
+      "learning_rate": 2e-05,
+      "loss": 0.0326,
+      "step": 44965
+    },
+    {
+      "epoch": 1.0703132438350949,
+      "grad_norm": 2.407368016758403,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 44966
+    },
+    {
+      "epoch": 1.0703370465581263,
+      "grad_norm": 2.9953274165488106,
+      "learning_rate": 2e-05,
+      "loss": 0.4173,
+      "step": 44967
+    },
+    {
+      "epoch": 1.0703608492811578,
+      "grad_norm": 0.006136151391624907,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 44968
+    },
+    {
+      "epoch": 1.0703846520041893,
+      "grad_norm": 5.448867244901498,
+      "learning_rate": 2e-05,
+      "loss": 0.3942,
+      "step": 44969
+    },
+    {
+      "epoch": 1.0704084547272208,
+      "grad_norm": 5.616431721304949,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 44970
+    },
+    {
+      "epoch": 1.0704322574502523,
+      "grad_norm": 6.5176611538771105,
+      "learning_rate": 2e-05,
+      "loss": 0.6533,
+      "step": 44971
+    },
+    {
+      "epoch": 1.0704560601732838,
+      "grad_norm": 2.1243350694747463,
+      "learning_rate": 2e-05,
+      "loss": 0.0181,
+      "step": 44972
+    },
+    {
+      "epoch": 1.0704798628963152,
+      "grad_norm": 0.024866909164186157,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44973
+    },
+    {
+      "epoch": 1.070503665619347,
+      "grad_norm": 7.472090156292255,
+      "learning_rate": 2e-05,
+      "loss": 0.2241,
+      "step": 44974
+    },
+    {
+      "epoch": 1.0705274683423784,
+      "grad_norm": 3.4138389644484475,
+      "learning_rate": 2e-05,
+      "loss": 0.1571,
+      "step": 44975
+    },
+    {
+      "epoch": 1.07055127106541,
+      "grad_norm": 0.0675565655935551,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 44976
+    },
+    {
+      "epoch": 1.0705750737884414,
+      "grad_norm": 0.019338217760513535,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 44977
+    },
+    {
+      "epoch": 1.070598876511473,
+      "grad_norm": 0.0426635391206228,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44978
+    },
+    {
+      "epoch": 1.0706226792345044,
+      "grad_norm": 0.22970854045005012,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 44979
+    },
+    {
+      "epoch": 1.0706464819575359,
+      "grad_norm": 6.299842272292069,
+      "learning_rate": 2e-05,
+      "loss": 0.9879,
+      "step": 44980
+    },
+    {
+      "epoch": 1.0706702846805674,
+      "grad_norm": 0.14168545771003585,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 44981
+    },
+    {
+      "epoch": 1.070694087403599,
+      "grad_norm": 8.248998274368857,
+      "learning_rate": 2e-05,
+      "loss": 0.4506,
+      "step": 44982
+    },
+    {
+      "epoch": 1.0707178901266305,
+      "grad_norm": 3.760609453268343,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 44983
+    },
+    {
+      "epoch": 1.070741692849662,
+      "grad_norm": 3.1913385260873866,
+      "learning_rate": 2e-05,
+      "loss": 0.3446,
+      "step": 44984
+    },
+    {
+      "epoch": 1.0707654955726935,
+      "grad_norm": 13.761416856747195,
+      "learning_rate": 2e-05,
+      "loss": 1.1179,
+      "step": 44985
+    },
+    {
+      "epoch": 1.070789298295725,
+      "grad_norm": 3.434610034412446,
+      "learning_rate": 2e-05,
+      "loss": 0.7106,
+      "step": 44986
+    },
+    {
+      "epoch": 1.0708131010187565,
+      "grad_norm": 12.961737888962633,
+      "learning_rate": 2e-05,
+      "loss": 0.3192,
+      "step": 44987
+    },
+    {
+      "epoch": 1.070836903741788,
+      "grad_norm": 1.2394962688728988,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 44988
+    },
+    {
+      "epoch": 1.0708607064648197,
+      "grad_norm": 4.504963306839001,
+      "learning_rate": 2e-05,
+      "loss": 0.6388,
+      "step": 44989
+    },
+    {
+      "epoch": 1.0708845091878512,
+      "grad_norm": 2.676199232922878,
+      "learning_rate": 2e-05,
+      "loss": 0.2666,
+      "step": 44990
+    },
+    {
+      "epoch": 1.0709083119108826,
+      "grad_norm": 3.225873693204601,
+      "learning_rate": 2e-05,
+      "loss": 0.2226,
+      "step": 44991
+    },
+    {
+      "epoch": 1.0709321146339141,
+      "grad_norm": 3.097738607633718,
+      "learning_rate": 2e-05,
+      "loss": 0.3846,
+      "step": 44992
+    },
+    {
+      "epoch": 1.0709559173569456,
+      "grad_norm": 3.3016143323071696,
+      "learning_rate": 2e-05,
+      "loss": 0.0874,
+      "step": 44993
+    },
+    {
+      "epoch": 1.070979720079977,
+      "grad_norm": 2.5121024389652806,
+      "learning_rate": 2e-05,
+      "loss": 0.1493,
+      "step": 44994
+    },
+    {
+      "epoch": 1.0710035228030086,
+      "grad_norm": 0.029813710997732404,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 44995
+    },
+    {
+      "epoch": 1.0710273255260403,
+      "grad_norm": 3.2736909103186935,
+      "learning_rate": 2e-05,
+      "loss": 0.2106,
+      "step": 44996
+    },
+    {
+      "epoch": 1.0710511282490718,
+      "grad_norm": 2.706151765475859,
+      "learning_rate": 2e-05,
+      "loss": 0.0692,
+      "step": 44997
+    },
+    {
+      "epoch": 1.0710749309721033,
+      "grad_norm": 0.5056170609992336,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 44998
+    },
+    {
+      "epoch": 1.0710987336951348,
+      "grad_norm": 29.546526963470573,
+      "learning_rate": 2e-05,
+      "loss": 0.3015,
+      "step": 44999
+    },
+    {
+      "epoch": 1.0711225364181662,
+      "grad_norm": 0.43476684506517055,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 45000
+    },
+    {
+      "epoch": 1.0711463391411977,
+      "grad_norm": 0.06458695182926176,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45001
+    },
+    {
+      "epoch": 1.0711701418642292,
+      "grad_norm": 9.177953128611126,
+      "learning_rate": 2e-05,
+      "loss": 0.9287,
+      "step": 45002
+    },
+    {
+      "epoch": 1.0711939445872607,
+      "grad_norm": 4.244541676833767,
+      "learning_rate": 2e-05,
+      "loss": 0.5606,
+      "step": 45003
+    },
+    {
+      "epoch": 1.0712177473102924,
+      "grad_norm": 8.176448473831746,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 45004
+    },
+    {
+      "epoch": 1.0712415500333239,
+      "grad_norm": 5.14830169635116,
+      "learning_rate": 2e-05,
+      "loss": 0.6137,
+      "step": 45005
+    },
+    {
+      "epoch": 1.0712653527563554,
+      "grad_norm": 2.3514843843300066,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 45006
+    },
+    {
+      "epoch": 1.0712891554793869,
+      "grad_norm": 17.635219046908407,
+      "learning_rate": 2e-05,
+      "loss": 0.3456,
+      "step": 45007
+    },
+    {
+      "epoch": 1.0713129582024183,
+      "grad_norm": 3.0293317188034794,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 45008
+    },
+    {
+      "epoch": 1.0713367609254498,
+      "grad_norm": 3.3540994535715414,
+      "learning_rate": 2e-05,
+      "loss": 0.4327,
+      "step": 45009
+    },
+    {
+      "epoch": 1.0713605636484813,
+      "grad_norm": 4.848544470948757,
+      "learning_rate": 2e-05,
+      "loss": 0.5154,
+      "step": 45010
+    },
+    {
+      "epoch": 1.0713843663715128,
+      "grad_norm": 19.088300570731654,
+      "learning_rate": 2e-05,
+      "loss": 0.4611,
+      "step": 45011
+    },
+    {
+      "epoch": 1.0714081690945445,
+      "grad_norm": 0.46398201777501624,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 45012
+    },
+    {
+      "epoch": 1.071431971817576,
+      "grad_norm": 2.803014309681737,
+      "learning_rate": 2e-05,
+      "loss": 0.2595,
+      "step": 45013
+    },
+    {
+      "epoch": 1.0714557745406075,
+      "grad_norm": 3.1003981555222584,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 45014
+    },
+    {
+      "epoch": 1.071479577263639,
+      "grad_norm": 0.005199380211429061,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45015
+    },
+    {
+      "epoch": 1.0715033799866704,
+      "grad_norm": 5.0735322733775945,
+      "learning_rate": 2e-05,
+      "loss": 0.077,
+      "step": 45016
+    },
+    {
+      "epoch": 1.071527182709702,
+      "grad_norm": 13.89103329104522,
+      "learning_rate": 2e-05,
+      "loss": 0.2047,
+      "step": 45017
+    },
+    {
+      "epoch": 1.0715509854327334,
+      "grad_norm": 6.881211728392296,
+      "learning_rate": 2e-05,
+      "loss": 0.3047,
+      "step": 45018
+    },
+    {
+      "epoch": 1.0715747881557651,
+      "grad_norm": 3.4452811332226694,
+      "learning_rate": 2e-05,
+      "loss": 0.4982,
+      "step": 45019
+    },
+    {
+      "epoch": 1.0715985908787966,
+      "grad_norm": 3.1196918177671225,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 45020
+    },
+    {
+      "epoch": 1.071622393601828,
+      "grad_norm": 5.081173705522905,
+      "learning_rate": 2e-05,
+      "loss": 0.4377,
+      "step": 45021
+    },
+    {
+      "epoch": 1.0716461963248596,
+      "grad_norm": 7.88142584631009,
+      "learning_rate": 2e-05,
+      "loss": 0.2763,
+      "step": 45022
+    },
+    {
+      "epoch": 1.071669999047891,
+      "grad_norm": 0.18327560277298807,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 45023
+    },
+    {
+      "epoch": 1.0716938017709225,
+      "grad_norm": 3.448479911842294,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 45024
+    },
+    {
+      "epoch": 1.071717604493954,
+      "grad_norm": 4.725550793333565,
+      "learning_rate": 2e-05,
+      "loss": 0.1066,
+      "step": 45025
+    },
+    {
+      "epoch": 1.0717414072169855,
+      "grad_norm": 4.3332063265228875,
+      "learning_rate": 2e-05,
+      "loss": 0.3366,
+      "step": 45026
+    },
+    {
+      "epoch": 1.0717652099400172,
+      "grad_norm": 8.110999340656484,
+      "learning_rate": 2e-05,
+      "loss": 0.1398,
+      "step": 45027
+    },
+    {
+      "epoch": 1.0717890126630487,
+      "grad_norm": 3.9403879735038916,
+      "learning_rate": 2e-05,
+      "loss": 0.0523,
+      "step": 45028
+    },
+    {
+      "epoch": 1.0718128153860802,
+      "grad_norm": 49.371547179755424,
+      "learning_rate": 2e-05,
+      "loss": 0.6078,
+      "step": 45029
+    },
+    {
+      "epoch": 1.0718366181091117,
+      "grad_norm": 2.836208811838756,
+      "learning_rate": 2e-05,
+      "loss": 0.1582,
+      "step": 45030
+    },
+    {
+      "epoch": 1.0718604208321432,
+      "grad_norm": 3.270659908521554,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 45031
+    },
+    {
+      "epoch": 1.0718842235551747,
+      "grad_norm": 0.7208918075846635,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 45032
+    },
+    {
+      "epoch": 1.0719080262782061,
+      "grad_norm": 0.018949039087782723,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45033
+    },
+    {
+      "epoch": 1.0719318290012376,
+      "grad_norm": 4.903001542117337,
+      "learning_rate": 2e-05,
+      "loss": 0.7787,
+      "step": 45034
+    },
+    {
+      "epoch": 1.0719556317242693,
+      "grad_norm": 3.4087022695945044,
+      "learning_rate": 2e-05,
+      "loss": 0.0346,
+      "step": 45035
+    },
+    {
+      "epoch": 1.0719794344473008,
+      "grad_norm": 5.778972880699177,
+      "learning_rate": 2e-05,
+      "loss": 0.927,
+      "step": 45036
+    },
+    {
+      "epoch": 1.0720032371703323,
+      "grad_norm": 0.21007966226523667,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45037
+    },
+    {
+      "epoch": 1.0720270398933638,
+      "grad_norm": 3.356893930600384,
+      "learning_rate": 2e-05,
+      "loss": 0.6242,
+      "step": 45038
+    },
+    {
+      "epoch": 1.0720508426163953,
+      "grad_norm": 5.0594019588202785,
+      "learning_rate": 2e-05,
+      "loss": 0.4582,
+      "step": 45039
+    },
+    {
+      "epoch": 1.0720746453394268,
+      "grad_norm": 0.06634847258431675,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45040
+    },
+    {
+      "epoch": 1.0720984480624582,
+      "grad_norm": 0.10916566956433416,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45041
+    },
+    {
+      "epoch": 1.07212225078549,
+      "grad_norm": 0.26270686840064955,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 45042
+    },
+    {
+      "epoch": 1.0721460535085214,
+      "grad_norm": 2.178171936518852,
+      "learning_rate": 2e-05,
+      "loss": 0.1419,
+      "step": 45043
+    },
+    {
+      "epoch": 1.072169856231553,
+      "grad_norm": 1.5281872811549888,
+      "learning_rate": 2e-05,
+      "loss": 0.0123,
+      "step": 45044
+    },
+    {
+      "epoch": 1.0721936589545844,
+      "grad_norm": 12.947515881310247,
+      "learning_rate": 2e-05,
+      "loss": 0.0848,
+      "step": 45045
+    },
+    {
+      "epoch": 1.072217461677616,
+      "grad_norm": 2.952473119622471,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 45046
+    },
+    {
+      "epoch": 1.0722412644006474,
+      "grad_norm": 4.040645653520893,
+      "learning_rate": 2e-05,
+      "loss": 0.4926,
+      "step": 45047
+    },
+    {
+      "epoch": 1.0722650671236789,
+      "grad_norm": 2.8311733460859294,
+      "learning_rate": 2e-05,
+      "loss": 0.3452,
+      "step": 45048
+    },
+    {
+      "epoch": 1.0722888698467106,
+      "grad_norm": 1.6135628186753397,
+      "learning_rate": 2e-05,
+      "loss": 0.0315,
+      "step": 45049
+    },
+    {
+      "epoch": 1.072312672569742,
+      "grad_norm": 1.0394838139761806,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 45050
+    },
+    {
+      "epoch": 1.0723364752927735,
+      "grad_norm": 3.7341651011844954,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 45051
+    },
+    {
+      "epoch": 1.072360278015805,
+      "grad_norm": 0.050884863504352944,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45052
+    },
+    {
+      "epoch": 1.0723840807388365,
+      "grad_norm": 3.055071669821172,
+      "learning_rate": 2e-05,
+      "loss": 0.1907,
+      "step": 45053
+    },
+    {
+      "epoch": 1.072407883461868,
+      "grad_norm": 6.727373205473217,
+      "learning_rate": 2e-05,
+      "loss": 0.0843,
+      "step": 45054
+    },
+    {
+      "epoch": 1.0724316861848995,
+      "grad_norm": 0.029387794448735967,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45055
+    },
+    {
+      "epoch": 1.072455488907931,
+      "grad_norm": 9.870688980523912,
+      "learning_rate": 2e-05,
+      "loss": 1.0957,
+      "step": 45056
+    },
+    {
+      "epoch": 1.0724792916309627,
+      "grad_norm": 1.915007515057026,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 45057
+    },
+    {
+      "epoch": 1.0725030943539942,
+      "grad_norm": 2.244474230818159,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 45058
+    },
+    {
+      "epoch": 1.0725268970770256,
+      "grad_norm": 0.023890546686613,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45059
+    },
+    {
+      "epoch": 1.0725506998000571,
+      "grad_norm": 0.03638059120321279,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45060
+    },
+    {
+      "epoch": 1.0725745025230886,
+      "grad_norm": 2.2060329788184845,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 45061
+    },
+    {
+      "epoch": 1.07259830524612,
+      "grad_norm": 1.6748515693453276,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 45062
+    },
+    {
+      "epoch": 1.0726221079691516,
+      "grad_norm": 5.556600787108011,
+      "learning_rate": 2e-05,
+      "loss": 0.3822,
+      "step": 45063
+    },
+    {
+      "epoch": 1.072645910692183,
+      "grad_norm": 0.08843642461957726,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45064
+    },
+    {
+      "epoch": 1.0726697134152148,
+      "grad_norm": 0.7401552756844354,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 45065
+    },
+    {
+      "epoch": 1.0726935161382463,
+      "grad_norm": 15.459834761936579,
+      "learning_rate": 2e-05,
+      "loss": 0.7332,
+      "step": 45066
+    },
+    {
+      "epoch": 1.0727173188612777,
+      "grad_norm": 4.010244567083515,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 45067
+    },
+    {
+      "epoch": 1.0727411215843092,
+      "grad_norm": 3.5971212146287534,
+      "learning_rate": 2e-05,
+      "loss": 0.1478,
+      "step": 45068
+    },
+    {
+      "epoch": 1.0727649243073407,
+      "grad_norm": 0.013236861838858641,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45069
+    },
+    {
+      "epoch": 1.0727887270303722,
+      "grad_norm": 0.13444193741202373,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 45070
+    },
+    {
+      "epoch": 1.0728125297534037,
+      "grad_norm": 7.295295217204875,
+      "learning_rate": 2e-05,
+      "loss": 0.4091,
+      "step": 45071
+    },
+    {
+      "epoch": 1.0728363324764354,
+      "grad_norm": 8.863075954065812,
+      "learning_rate": 2e-05,
+      "loss": 0.4013,
+      "step": 45072
+    },
+    {
+      "epoch": 1.0728601351994669,
+      "grad_norm": 1.3628409797474237,
+      "learning_rate": 2e-05,
+      "loss": 0.0158,
+      "step": 45073
+    },
+    {
+      "epoch": 1.0728839379224984,
+      "grad_norm": 5.528774502824384,
+      "learning_rate": 2e-05,
+      "loss": 0.4413,
+      "step": 45074
+    },
+    {
+      "epoch": 1.0729077406455299,
+      "grad_norm": 3.2006272054294755,
+      "learning_rate": 2e-05,
+      "loss": 0.0715,
+      "step": 45075
+    },
+    {
+      "epoch": 1.0729315433685613,
+      "grad_norm": 5.583504258057568,
+      "learning_rate": 2e-05,
+      "loss": 0.6238,
+      "step": 45076
+    },
+    {
+      "epoch": 1.0729553460915928,
+      "grad_norm": 0.02857740854201257,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45077
+    },
+    {
+      "epoch": 1.0729791488146243,
+      "grad_norm": 0.7851392476128051,
+      "learning_rate": 2e-05,
+      "loss": 0.0117,
+      "step": 45078
+    },
+    {
+      "epoch": 1.073002951537656,
+      "grad_norm": 0.12275141926045327,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45079
+    },
+    {
+      "epoch": 1.0730267542606875,
+      "grad_norm": 0.44179042155056325,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 45080
+    },
+    {
+      "epoch": 1.073050556983719,
+      "grad_norm": 0.23227162442164448,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 45081
+    },
+    {
+      "epoch": 1.0730743597067505,
+      "grad_norm": 2.857242145176698,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 45082
+    },
+    {
+      "epoch": 1.073098162429782,
+      "grad_norm": 6.802374158377842,
+      "learning_rate": 2e-05,
+      "loss": 0.2234,
+      "step": 45083
+    },
+    {
+      "epoch": 1.0731219651528134,
+      "grad_norm": 15.455967507574938,
+      "learning_rate": 2e-05,
+      "loss": 0.7053,
+      "step": 45084
+    },
+    {
+      "epoch": 1.073145767875845,
+      "grad_norm": 0.05153566587753122,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45085
+    },
+    {
+      "epoch": 1.0731695705988764,
+      "grad_norm": 2.234333059194782,
+      "learning_rate": 2e-05,
+      "loss": 0.115,
+      "step": 45086
+    },
+    {
+      "epoch": 1.0731933733219081,
+      "grad_norm": 3.2314522889220694,
+      "learning_rate": 2e-05,
+      "loss": 0.1051,
+      "step": 45087
+    },
+    {
+      "epoch": 1.0732171760449396,
+      "grad_norm": 1.8333208623174138,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 45088
+    },
+    {
+      "epoch": 1.073240978767971,
+      "grad_norm": 3.513251150071855,
+      "learning_rate": 2e-05,
+      "loss": 0.2363,
+      "step": 45089
+    },
+    {
+      "epoch": 1.0732647814910026,
+      "grad_norm": 2.7410671552037282,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 45090
+    },
+    {
+      "epoch": 1.073288584214034,
+      "grad_norm": 4.0480931930935515,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 45091
+    },
+    {
+      "epoch": 1.0733123869370655,
+      "grad_norm": 0.014118466951972764,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45092
+    },
+    {
+      "epoch": 1.073336189660097,
+      "grad_norm": 0.27434308096131027,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 45093
+    },
+    {
+      "epoch": 1.0733599923831285,
+      "grad_norm": 2.7344146513578678,
+      "learning_rate": 2e-05,
+      "loss": 0.1375,
+      "step": 45094
+    },
+    {
+      "epoch": 1.0733837951061602,
+      "grad_norm": 3.4410399943632672,
+      "learning_rate": 2e-05,
+      "loss": 0.5912,
+      "step": 45095
+    },
+    {
+      "epoch": 1.0734075978291917,
+      "grad_norm": 1.3289214314438371,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 45096
+    },
+    {
+      "epoch": 1.0734314005522232,
+      "grad_norm": 2.589934667067985,
+      "learning_rate": 2e-05,
+      "loss": 0.2274,
+      "step": 45097
+    },
+    {
+      "epoch": 1.0734552032752547,
+      "grad_norm": 2.52400554480027,
+      "learning_rate": 2e-05,
+      "loss": 0.1432,
+      "step": 45098
+    },
+    {
+      "epoch": 1.0734790059982862,
+      "grad_norm": 2.671672429686085,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 45099
+    },
+    {
+      "epoch": 1.0735028087213176,
+      "grad_norm": 2.3868825734406163,
+      "learning_rate": 2e-05,
+      "loss": 0.2338,
+      "step": 45100
+    },
+    {
+      "epoch": 1.0735266114443491,
+      "grad_norm": 0.8860549075938325,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 45101
+    },
+    {
+      "epoch": 1.0735504141673808,
+      "grad_norm": 0.04162934122290395,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45102
+    },
+    {
+      "epoch": 1.0735742168904123,
+      "grad_norm": 2.8266649274539204,
+      "learning_rate": 2e-05,
+      "loss": 0.3646,
+      "step": 45103
+    },
+    {
+      "epoch": 1.0735980196134438,
+      "grad_norm": 2.411356861937131,
+      "learning_rate": 2e-05,
+      "loss": 0.1561,
+      "step": 45104
+    },
+    {
+      "epoch": 1.0736218223364753,
+      "grad_norm": 0.1992318068845013,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 45105
+    },
+    {
+      "epoch": 1.0736456250595068,
+      "grad_norm": 0.0866039146254041,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45106
+    },
+    {
+      "epoch": 1.0736694277825383,
+      "grad_norm": 0.25322390787016524,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 45107
+    },
+    {
+      "epoch": 1.0736932305055698,
+      "grad_norm": 0.9679597777641739,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 45108
+    },
+    {
+      "epoch": 1.0737170332286015,
+      "grad_norm": 5.881280980137443,
+      "learning_rate": 2e-05,
+      "loss": 0.8545,
+      "step": 45109
+    },
+    {
+      "epoch": 1.073740835951633,
+      "grad_norm": 0.5362115106359875,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 45110
+    },
+    {
+      "epoch": 1.0737646386746644,
+      "grad_norm": 0.11466913599587554,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 45111
+    },
+    {
+      "epoch": 1.073788441397696,
+      "grad_norm": 8.288856111151095,
+      "learning_rate": 2e-05,
+      "loss": 0.1703,
+      "step": 45112
+    },
+    {
+      "epoch": 1.0738122441207274,
+      "grad_norm": 5.2440402629891105,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 45113
+    },
+    {
+      "epoch": 1.0738360468437589,
+      "grad_norm": 2.7652588370300433,
+      "learning_rate": 2e-05,
+      "loss": 0.2098,
+      "step": 45114
+    },
+    {
+      "epoch": 1.0738598495667904,
+      "grad_norm": 3.4153654221939918,
+      "learning_rate": 2e-05,
+      "loss": 0.1514,
+      "step": 45115
+    },
+    {
+      "epoch": 1.0738836522898219,
+      "grad_norm": 0.8408550363468078,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 45116
+    },
+    {
+      "epoch": 1.0739074550128536,
+      "grad_norm": 14.1523564285062,
+      "learning_rate": 2e-05,
+      "loss": 0.3089,
+      "step": 45117
+    },
+    {
+      "epoch": 1.073931257735885,
+      "grad_norm": 5.313123106331898,
+      "learning_rate": 2e-05,
+      "loss": 0.1039,
+      "step": 45118
+    },
+    {
+      "epoch": 1.0739550604589165,
+      "grad_norm": 0.014856901904854372,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45119
+    },
+    {
+      "epoch": 1.073978863181948,
+      "grad_norm": 0.27084610617403115,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45120
+    },
+    {
+      "epoch": 1.0740026659049795,
+      "grad_norm": 0.18830341859241928,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 45121
+    },
+    {
+      "epoch": 1.074026468628011,
+      "grad_norm": 0.014603486608072392,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45122
+    },
+    {
+      "epoch": 1.0740502713510425,
+      "grad_norm": 3.813055703938403,
+      "learning_rate": 2e-05,
+      "loss": 0.2039,
+      "step": 45123
+    },
+    {
+      "epoch": 1.074074074074074,
+      "grad_norm": 0.5955253399329147,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 45124
+    },
+    {
+      "epoch": 1.0740978767971057,
+      "grad_norm": 2.3514824625493227,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 45125
+    },
+    {
+      "epoch": 1.0741216795201372,
+      "grad_norm": 3.6353069967992044,
+      "learning_rate": 2e-05,
+      "loss": 0.0516,
+      "step": 45126
+    },
+    {
+      "epoch": 1.0741454822431686,
+      "grad_norm": 5.215180751151517,
+      "learning_rate": 2e-05,
+      "loss": 0.6664,
+      "step": 45127
+    },
+    {
+      "epoch": 1.0741692849662001,
+      "grad_norm": 4.262361252254771,
+      "learning_rate": 2e-05,
+      "loss": 0.1147,
+      "step": 45128
+    },
+    {
+      "epoch": 1.0741930876892316,
+      "grad_norm": 0.05291233113709445,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45129
+    },
+    {
+      "epoch": 1.074216890412263,
+      "grad_norm": 3.325696894640884,
+      "learning_rate": 2e-05,
+      "loss": 0.0482,
+      "step": 45130
+    },
+    {
+      "epoch": 1.0742406931352946,
+      "grad_norm": 7.741027558021726,
+      "learning_rate": 2e-05,
+      "loss": 0.3423,
+      "step": 45131
+    },
+    {
+      "epoch": 1.0742644958583263,
+      "grad_norm": 0.032159704443258254,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45132
+    },
+    {
+      "epoch": 1.0742882985813578,
+      "grad_norm": 3.2823639162793974,
+      "learning_rate": 2e-05,
+      "loss": 0.0578,
+      "step": 45133
+    },
+    {
+      "epoch": 1.0743121013043893,
+      "grad_norm": 4.042857643616388,
+      "learning_rate": 2e-05,
+      "loss": 0.3277,
+      "step": 45134
+    },
+    {
+      "epoch": 1.0743359040274207,
+      "grad_norm": 1.846678100768461,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 45135
+    },
+    {
+      "epoch": 1.0743597067504522,
+      "grad_norm": 3.897862779161478,
+      "learning_rate": 2e-05,
+      "loss": 0.2897,
+      "step": 45136
+    },
+    {
+      "epoch": 1.0743835094734837,
+      "grad_norm": 0.02062185105049056,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45137
+    },
+    {
+      "epoch": 1.0744073121965152,
+      "grad_norm": 0.7511260832423527,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 45138
+    },
+    {
+      "epoch": 1.074431114919547,
+      "grad_norm": 2.8768243778684557,
+      "learning_rate": 2e-05,
+      "loss": 0.377,
+      "step": 45139
+    },
+    {
+      "epoch": 1.0744549176425784,
+      "grad_norm": 5.220489394425504,
+      "learning_rate": 2e-05,
+      "loss": 0.5837,
+      "step": 45140
+    },
+    {
+      "epoch": 1.0744787203656099,
+      "grad_norm": 0.09719666225112342,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 45141
+    },
+    {
+      "epoch": 1.0745025230886414,
+      "grad_norm": 4.374898172940238,
+      "learning_rate": 2e-05,
+      "loss": 0.805,
+      "step": 45142
+    },
+    {
+      "epoch": 1.0745263258116728,
+      "grad_norm": 4.2897627595467736,
+      "learning_rate": 2e-05,
+      "loss": 0.1155,
+      "step": 45143
+    },
+    {
+      "epoch": 1.0745501285347043,
+      "grad_norm": 0.20758829650431387,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 45144
+    },
+    {
+      "epoch": 1.0745739312577358,
+      "grad_norm": 0.021037163934142088,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45145
+    },
+    {
+      "epoch": 1.0745977339807673,
+      "grad_norm": 0.8556853699812262,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 45146
+    },
+    {
+      "epoch": 1.074621536703799,
+      "grad_norm": 6.820884377724129,
+      "learning_rate": 2e-05,
+      "loss": 0.8269,
+      "step": 45147
+    },
+    {
+      "epoch": 1.0746453394268305,
+      "grad_norm": 4.465239372820224,
+      "learning_rate": 2e-05,
+      "loss": 0.729,
+      "step": 45148
+    },
+    {
+      "epoch": 1.074669142149862,
+      "grad_norm": 4.134021468798416,
+      "learning_rate": 2e-05,
+      "loss": 0.4232,
+      "step": 45149
+    },
+    {
+      "epoch": 1.0746929448728935,
+      "grad_norm": 11.52450573151411,
+      "learning_rate": 2e-05,
+      "loss": 0.9072,
+      "step": 45150
+    },
+    {
+      "epoch": 1.074716747595925,
+      "grad_norm": 9.546035788316505,
+      "learning_rate": 2e-05,
+      "loss": 0.0817,
+      "step": 45151
+    },
+    {
+      "epoch": 1.0747405503189564,
+      "grad_norm": 4.059496139828307,
+      "learning_rate": 2e-05,
+      "loss": 0.0444,
+      "step": 45152
+    },
+    {
+      "epoch": 1.074764353041988,
+      "grad_norm": 0.04608468031406841,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45153
+    },
+    {
+      "epoch": 1.0747881557650194,
+      "grad_norm": 3.3099817181017435,
+      "learning_rate": 2e-05,
+      "loss": 0.1744,
+      "step": 45154
+    },
+    {
+      "epoch": 1.0748119584880511,
+      "grad_norm": 4.583014381773661,
+      "learning_rate": 2e-05,
+      "loss": 0.6958,
+      "step": 45155
+    },
+    {
+      "epoch": 1.0748357612110826,
+      "grad_norm": 0.07478933810269882,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45156
+    },
+    {
+      "epoch": 1.074859563934114,
+      "grad_norm": 9.388794197985444,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 45157
+    },
+    {
+      "epoch": 1.0748833666571456,
+      "grad_norm": 0.03473406774609831,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45158
+    },
+    {
+      "epoch": 1.074907169380177,
+      "grad_norm": 3.765882142339768,
+      "learning_rate": 2e-05,
+      "loss": 0.2491,
+      "step": 45159
+    },
+    {
+      "epoch": 1.0749309721032085,
+      "grad_norm": 0.1608286361552099,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45160
+    },
+    {
+      "epoch": 1.07495477482624,
+      "grad_norm": 0.34002057090570365,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45161
+    },
+    {
+      "epoch": 1.0749785775492717,
+      "grad_norm": 0.37736314984909114,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 45162
+    },
+    {
+      "epoch": 1.0750023802723032,
+      "grad_norm": 2.4917157129864536,
+      "learning_rate": 2e-05,
+      "loss": 0.0645,
+      "step": 45163
+    },
+    {
+      "epoch": 1.0750261829953347,
+      "grad_norm": 5.383851793531096,
+      "learning_rate": 2e-05,
+      "loss": 0.3048,
+      "step": 45164
+    },
+    {
+      "epoch": 1.0750499857183662,
+      "grad_norm": 0.7879756638556592,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 45165
+    },
+    {
+      "epoch": 1.0750737884413977,
+      "grad_norm": 6.284658062248196,
+      "learning_rate": 2e-05,
+      "loss": 0.4946,
+      "step": 45166
+    },
+    {
+      "epoch": 1.0750975911644292,
+      "grad_norm": 3.760480029372955,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 45167
+    },
+    {
+      "epoch": 1.0751213938874606,
+      "grad_norm": 1.1794504280786557,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 45168
+    },
+    {
+      "epoch": 1.0751451966104923,
+      "grad_norm": 1.4074545034978274,
+      "learning_rate": 2e-05,
+      "loss": 0.0127,
+      "step": 45169
+    },
+    {
+      "epoch": 1.0751689993335238,
+      "grad_norm": 3.2078221357318846,
+      "learning_rate": 2e-05,
+      "loss": 0.4119,
+      "step": 45170
+    },
+    {
+      "epoch": 1.0751928020565553,
+      "grad_norm": 6.163796405767615,
+      "learning_rate": 2e-05,
+      "loss": 0.4975,
+      "step": 45171
+    },
+    {
+      "epoch": 1.0752166047795868,
+      "grad_norm": 0.03693406763131601,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45172
+    },
+    {
+      "epoch": 1.0752404075026183,
+      "grad_norm": 3.4838634285051673,
+      "learning_rate": 2e-05,
+      "loss": 0.5293,
+      "step": 45173
+    },
+    {
+      "epoch": 1.0752642102256498,
+      "grad_norm": 3.7862925880966034,
+      "learning_rate": 2e-05,
+      "loss": 0.4195,
+      "step": 45174
+    },
+    {
+      "epoch": 1.0752880129486813,
+      "grad_norm": 3.3497393730561744,
+      "learning_rate": 2e-05,
+      "loss": 0.026,
+      "step": 45175
+    },
+    {
+      "epoch": 1.0753118156717127,
+      "grad_norm": 6.622825188741395,
+      "learning_rate": 2e-05,
+      "loss": 0.6351,
+      "step": 45176
+    },
+    {
+      "epoch": 1.0753356183947445,
+      "grad_norm": 3.92166155428071,
+      "learning_rate": 2e-05,
+      "loss": 0.1095,
+      "step": 45177
+    },
+    {
+      "epoch": 1.075359421117776,
+      "grad_norm": 0.03541426726910651,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45178
+    },
+    {
+      "epoch": 1.0753832238408074,
+      "grad_norm": 2.3981205092131916,
+      "learning_rate": 2e-05,
+      "loss": 0.1843,
+      "step": 45179
+    },
+    {
+      "epoch": 1.075407026563839,
+      "grad_norm": 6.168619095626887,
+      "learning_rate": 2e-05,
+      "loss": 0.537,
+      "step": 45180
+    },
+    {
+      "epoch": 1.0754308292868704,
+      "grad_norm": 6.105041584885287,
+      "learning_rate": 2e-05,
+      "loss": 0.5934,
+      "step": 45181
+    },
+    {
+      "epoch": 1.0754546320099019,
+      "grad_norm": 0.06120896141032399,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45182
+    },
+    {
+      "epoch": 1.0754784347329334,
+      "grad_norm": 0.005871098459672924,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45183
+    },
+    {
+      "epoch": 1.0755022374559648,
+      "grad_norm": 2.6858682866756047,
+      "learning_rate": 2e-05,
+      "loss": 0.0566,
+      "step": 45184
+    },
+    {
+      "epoch": 1.0755260401789966,
+      "grad_norm": 4.415032297192314,
+      "learning_rate": 2e-05,
+      "loss": 0.0905,
+      "step": 45185
+    },
+    {
+      "epoch": 1.075549842902028,
+      "grad_norm": 2.3469402807726087,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 45186
+    },
+    {
+      "epoch": 1.0755736456250595,
+      "grad_norm": 5.180187233420187,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 45187
+    },
+    {
+      "epoch": 1.075597448348091,
+      "grad_norm": 0.024475875312028463,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45188
+    },
+    {
+      "epoch": 1.0756212510711225,
+      "grad_norm": 0.27436848444579537,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45189
+    },
+    {
+      "epoch": 1.075645053794154,
+      "grad_norm": 0.07318403500602552,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45190
+    },
+    {
+      "epoch": 1.0756688565171855,
+      "grad_norm": 2.780836969287068,
+      "learning_rate": 2e-05,
+      "loss": 0.1581,
+      "step": 45191
+    },
+    {
+      "epoch": 1.0756926592402172,
+      "grad_norm": 2.576890574358364,
+      "learning_rate": 2e-05,
+      "loss": 0.0299,
+      "step": 45192
+    },
+    {
+      "epoch": 1.0757164619632487,
+      "grad_norm": 3.1699477441846606,
+      "learning_rate": 2e-05,
+      "loss": 0.2677,
+      "step": 45193
+    },
+    {
+      "epoch": 1.0757402646862801,
+      "grad_norm": 0.05771136323866317,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45194
+    },
+    {
+      "epoch": 1.0757640674093116,
+      "grad_norm": 0.10308494224951978,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45195
+    },
+    {
+      "epoch": 1.0757878701323431,
+      "grad_norm": 0.06080990894820059,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45196
+    },
+    {
+      "epoch": 1.0758116728553746,
+      "grad_norm": 5.285403045282893,
+      "learning_rate": 2e-05,
+      "loss": 0.6748,
+      "step": 45197
+    },
+    {
+      "epoch": 1.075835475578406,
+      "grad_norm": 3.5423065703749432,
+      "learning_rate": 2e-05,
+      "loss": 0.4221,
+      "step": 45198
+    },
+    {
+      "epoch": 1.0758592783014378,
+      "grad_norm": 0.929894081315518,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 45199
+    },
+    {
+      "epoch": 1.0758830810244693,
+      "grad_norm": 0.02866680390371627,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45200
+    },
+    {
+      "epoch": 1.0759068837475008,
+      "grad_norm": 5.83156123322221,
+      "learning_rate": 2e-05,
+      "loss": 1.3176,
+      "step": 45201
+    },
+    {
+      "epoch": 1.0759306864705322,
+      "grad_norm": 3.5782664932183796,
+      "learning_rate": 2e-05,
+      "loss": 0.5031,
+      "step": 45202
+    },
+    {
+      "epoch": 1.0759544891935637,
+      "grad_norm": 0.7856345720993572,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 45203
+    },
+    {
+      "epoch": 1.0759782919165952,
+      "grad_norm": 11.797028774381355,
+      "learning_rate": 2e-05,
+      "loss": 0.8911,
+      "step": 45204
+    },
+    {
+      "epoch": 1.0760020946396267,
+      "grad_norm": 2.7148819692915884,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 45205
+    },
+    {
+      "epoch": 1.0760258973626582,
+      "grad_norm": 4.2407989271640645,
+      "learning_rate": 2e-05,
+      "loss": 0.4085,
+      "step": 45206
+    },
+    {
+      "epoch": 1.07604970008569,
+      "grad_norm": 7.372366684040638,
+      "learning_rate": 2e-05,
+      "loss": 0.2564,
+      "step": 45207
+    },
+    {
+      "epoch": 1.0760735028087214,
+      "grad_norm": 0.02122771113933063,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45208
+    },
+    {
+      "epoch": 1.0760973055317529,
+      "grad_norm": 4.451062500802264,
+      "learning_rate": 2e-05,
+      "loss": 0.3266,
+      "step": 45209
+    },
+    {
+      "epoch": 1.0761211082547844,
+      "grad_norm": 15.635089272515108,
+      "learning_rate": 2e-05,
+      "loss": 0.1094,
+      "step": 45210
+    },
+    {
+      "epoch": 1.0761449109778158,
+      "grad_norm": 1.7622681022960855,
+      "learning_rate": 2e-05,
+      "loss": 0.1065,
+      "step": 45211
+    },
+    {
+      "epoch": 1.0761687137008473,
+      "grad_norm": 5.402858092561353,
+      "learning_rate": 2e-05,
+      "loss": 0.5477,
+      "step": 45212
+    },
+    {
+      "epoch": 1.0761925164238788,
+      "grad_norm": 4.395370380269612,
+      "learning_rate": 2e-05,
+      "loss": 0.3762,
+      "step": 45213
+    },
+    {
+      "epoch": 1.0762163191469103,
+      "grad_norm": 4.3201555630297745,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 45214
+    },
+    {
+      "epoch": 1.076240121869942,
+      "grad_norm": 0.43885897154332987,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 45215
+    },
+    {
+      "epoch": 1.0762639245929735,
+      "grad_norm": 0.07020801501530567,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45216
+    },
+    {
+      "epoch": 1.076287727316005,
+      "grad_norm": 0.003341224744771667,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45217
+    },
+    {
+      "epoch": 1.0763115300390365,
+      "grad_norm": 1.6084608790496309,
+      "learning_rate": 2e-05,
+      "loss": 0.02,
+      "step": 45218
+    },
+    {
+      "epoch": 1.076335332762068,
+      "grad_norm": 3.831772004948512,
+      "learning_rate": 2e-05,
+      "loss": 0.0791,
+      "step": 45219
+    },
+    {
+      "epoch": 1.0763591354850994,
+      "grad_norm": 5.5837600068097,
+      "learning_rate": 2e-05,
+      "loss": 0.8614,
+      "step": 45220
+    },
+    {
+      "epoch": 1.076382938208131,
+      "grad_norm": 3.6454878634460055,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 45221
+    },
+    {
+      "epoch": 1.0764067409311626,
+      "grad_norm": 2.7433496563971964,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 45222
+    },
+    {
+      "epoch": 1.076430543654194,
+      "grad_norm": 6.617714122221036,
+      "learning_rate": 2e-05,
+      "loss": 0.631,
+      "step": 45223
+    },
+    {
+      "epoch": 1.0764543463772256,
+      "grad_norm": 3.104981760419617,
+      "learning_rate": 2e-05,
+      "loss": 0.1848,
+      "step": 45224
+    },
+    {
+      "epoch": 1.076478149100257,
+      "grad_norm": 2.427103684769767,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 45225
+    },
+    {
+      "epoch": 1.0765019518232886,
+      "grad_norm": 4.6070114983478465,
+      "learning_rate": 2e-05,
+      "loss": 0.6138,
+      "step": 45226
+    },
+    {
+      "epoch": 1.07652575454632,
+      "grad_norm": 3.2401473417814515,
+      "learning_rate": 2e-05,
+      "loss": 0.3118,
+      "step": 45227
+    },
+    {
+      "epoch": 1.0765495572693515,
+      "grad_norm": 8.198446841589528,
+      "learning_rate": 2e-05,
+      "loss": 0.4223,
+      "step": 45228
+    },
+    {
+      "epoch": 1.0765733599923832,
+      "grad_norm": 4.298107262301464,
+      "learning_rate": 2e-05,
+      "loss": 0.6392,
+      "step": 45229
+    },
+    {
+      "epoch": 1.0765971627154147,
+      "grad_norm": 4.522153925605831,
+      "learning_rate": 2e-05,
+      "loss": 0.5842,
+      "step": 45230
+    },
+    {
+      "epoch": 1.0766209654384462,
+      "grad_norm": 3.696508856360113,
+      "learning_rate": 2e-05,
+      "loss": 0.5282,
+      "step": 45231
+    },
+    {
+      "epoch": 1.0766447681614777,
+      "grad_norm": 2.340367810584045,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 45232
+    },
+    {
+      "epoch": 1.0766685708845092,
+      "grad_norm": 6.2393906731311,
+      "learning_rate": 2e-05,
+      "loss": 0.1436,
+      "step": 45233
+    },
+    {
+      "epoch": 1.0766923736075407,
+      "grad_norm": 1.9844772429567037,
+      "learning_rate": 2e-05,
+      "loss": 0.0502,
+      "step": 45234
+    },
+    {
+      "epoch": 1.0767161763305722,
+      "grad_norm": 5.834224162693148,
+      "learning_rate": 2e-05,
+      "loss": 0.0852,
+      "step": 45235
+    },
+    {
+      "epoch": 1.0767399790536036,
+      "grad_norm": 4.755507043752084,
+      "learning_rate": 2e-05,
+      "loss": 0.6591,
+      "step": 45236
+    },
+    {
+      "epoch": 1.0767637817766353,
+      "grad_norm": 3.852606047378601,
+      "learning_rate": 2e-05,
+      "loss": 0.1973,
+      "step": 45237
+    },
+    {
+      "epoch": 1.0767875844996668,
+      "grad_norm": 0.8900811806240813,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 45238
+    },
+    {
+      "epoch": 1.0768113872226983,
+      "grad_norm": 3.734266609764805,
+      "learning_rate": 2e-05,
+      "loss": 0.2713,
+      "step": 45239
+    },
+    {
+      "epoch": 1.0768351899457298,
+      "grad_norm": 0.07349803344825756,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 45240
+    },
+    {
+      "epoch": 1.0768589926687613,
+      "grad_norm": 6.063237231490434,
+      "learning_rate": 2e-05,
+      "loss": 0.7085,
+      "step": 45241
+    },
+    {
+      "epoch": 1.0768827953917928,
+      "grad_norm": 4.15043010200688,
+      "learning_rate": 2e-05,
+      "loss": 0.6008,
+      "step": 45242
+    },
+    {
+      "epoch": 1.0769065981148243,
+      "grad_norm": 1.4688827102567925,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 45243
+    },
+    {
+      "epoch": 1.0769304008378557,
+      "grad_norm": 4.062466023573515,
+      "learning_rate": 2e-05,
+      "loss": 0.6381,
+      "step": 45244
+    },
+    {
+      "epoch": 1.0769542035608874,
+      "grad_norm": 5.131169474520399,
+      "learning_rate": 2e-05,
+      "loss": 0.3573,
+      "step": 45245
+    },
+    {
+      "epoch": 1.076978006283919,
+      "grad_norm": 4.867200173874895,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 45246
+    },
+    {
+      "epoch": 1.0770018090069504,
+      "grad_norm": 0.14700348460039642,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 45247
+    },
+    {
+      "epoch": 1.077025611729982,
+      "grad_norm": 0.03380001632254735,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45248
+    },
+    {
+      "epoch": 1.0770494144530134,
+      "grad_norm": 0.6315570208345137,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 45249
+    },
+    {
+      "epoch": 1.0770732171760449,
+      "grad_norm": 0.29852663736333,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 45250
+    },
+    {
+      "epoch": 1.0770970198990764,
+      "grad_norm": 0.08307553541880572,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45251
+    },
+    {
+      "epoch": 1.077120822622108,
+      "grad_norm": 3.344235910096425,
+      "learning_rate": 2e-05,
+      "loss": 0.4021,
+      "step": 45252
+    },
+    {
+      "epoch": 1.0771446253451396,
+      "grad_norm": 4.032261617587501,
+      "learning_rate": 2e-05,
+      "loss": 0.4591,
+      "step": 45253
+    },
+    {
+      "epoch": 1.077168428068171,
+      "grad_norm": 2.177207132154821,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 45254
+    },
+    {
+      "epoch": 1.0771922307912025,
+      "grad_norm": 11.43544124133767,
+      "learning_rate": 2e-05,
+      "loss": 0.4563,
+      "step": 45255
+    },
+    {
+      "epoch": 1.077216033514234,
+      "grad_norm": 3.6378598152987753,
+      "learning_rate": 2e-05,
+      "loss": 0.218,
+      "step": 45256
+    },
+    {
+      "epoch": 1.0772398362372655,
+      "grad_norm": 5.389208528069145,
+      "learning_rate": 2e-05,
+      "loss": 0.065,
+      "step": 45257
+    },
+    {
+      "epoch": 1.077263638960297,
+      "grad_norm": 11.4348289699933,
+      "learning_rate": 2e-05,
+      "loss": 0.6954,
+      "step": 45258
+    },
+    {
+      "epoch": 1.0772874416833287,
+      "grad_norm": 4.364221647002444,
+      "learning_rate": 2e-05,
+      "loss": 0.5131,
+      "step": 45259
+    },
+    {
+      "epoch": 1.0773112444063602,
+      "grad_norm": 0.021614337727803593,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45260
+    },
+    {
+      "epoch": 1.0773350471293917,
+      "grad_norm": 1.3482900816879217,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 45261
+    },
+    {
+      "epoch": 1.0773588498524231,
+      "grad_norm": 0.18511090743319622,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 45262
+    },
+    {
+      "epoch": 1.0773826525754546,
+      "grad_norm": 0.05951499791645491,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45263
+    },
+    {
+      "epoch": 1.077406455298486,
+      "grad_norm": 8.455453803352206,
+      "learning_rate": 2e-05,
+      "loss": 1.1701,
+      "step": 45264
+    },
+    {
+      "epoch": 1.0774302580215176,
+      "grad_norm": 1.8196780183849546,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 45265
+    },
+    {
+      "epoch": 1.077454060744549,
+      "grad_norm": 3.446514331938888,
+      "learning_rate": 2e-05,
+      "loss": 0.4339,
+      "step": 45266
+    },
+    {
+      "epoch": 1.0774778634675808,
+      "grad_norm": 0.5301610851254215,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 45267
+    },
+    {
+      "epoch": 1.0775016661906123,
+      "grad_norm": 0.5819787297784995,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 45268
+    },
+    {
+      "epoch": 1.0775254689136438,
+      "grad_norm": 4.104894184106721,
+      "learning_rate": 2e-05,
+      "loss": 0.8018,
+      "step": 45269
+    },
+    {
+      "epoch": 1.0775492716366752,
+      "grad_norm": 0.6264153696666137,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 45270
+    },
+    {
+      "epoch": 1.0775730743597067,
+      "grad_norm": 3.3498953317659694,
+      "learning_rate": 2e-05,
+      "loss": 0.2879,
+      "step": 45271
+    },
+    {
+      "epoch": 1.0775968770827382,
+      "grad_norm": 0.5597697674010008,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 45272
+    },
+    {
+      "epoch": 1.0776206798057697,
+      "grad_norm": 10.173599652966784,
+      "learning_rate": 2e-05,
+      "loss": 0.3853,
+      "step": 45273
+    },
+    {
+      "epoch": 1.0776444825288012,
+      "grad_norm": 3.7985396479032274,
+      "learning_rate": 2e-05,
+      "loss": 0.3344,
+      "step": 45274
+    },
+    {
+      "epoch": 1.077668285251833,
+      "grad_norm": 3.7031219143112963,
+      "learning_rate": 2e-05,
+      "loss": 0.4665,
+      "step": 45275
+    },
+    {
+      "epoch": 1.0776920879748644,
+      "grad_norm": 30.3106991535817,
+      "learning_rate": 2e-05,
+      "loss": 0.2662,
+      "step": 45276
+    },
+    {
+      "epoch": 1.0777158906978959,
+      "grad_norm": 3.118296176374066,
+      "learning_rate": 2e-05,
+      "loss": 0.1865,
+      "step": 45277
+    },
+    {
+      "epoch": 1.0777396934209273,
+      "grad_norm": 0.07334634404612252,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45278
+    },
+    {
+      "epoch": 1.0777634961439588,
+      "grad_norm": 3.6832366706646384,
+      "learning_rate": 2e-05,
+      "loss": 0.0851,
+      "step": 45279
+    },
+    {
+      "epoch": 1.0777872988669903,
+      "grad_norm": 4.162065929711235,
+      "learning_rate": 2e-05,
+      "loss": 0.0769,
+      "step": 45280
+    },
+    {
+      "epoch": 1.0778111015900218,
+      "grad_norm": 0.03525532161571293,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45281
+    },
+    {
+      "epoch": 1.0778349043130535,
+      "grad_norm": 8.65009073032027,
+      "learning_rate": 2e-05,
+      "loss": 0.4261,
+      "step": 45282
+    },
+    {
+      "epoch": 1.077858707036085,
+      "grad_norm": 4.872107403081799,
+      "learning_rate": 2e-05,
+      "loss": 0.7932,
+      "step": 45283
+    },
+    {
+      "epoch": 1.0778825097591165,
+      "grad_norm": 11.373778225142019,
+      "learning_rate": 2e-05,
+      "loss": 0.8824,
+      "step": 45284
+    },
+    {
+      "epoch": 1.077906312482148,
+      "grad_norm": 8.99737700650994,
+      "learning_rate": 2e-05,
+      "loss": 0.2917,
+      "step": 45285
+    },
+    {
+      "epoch": 1.0779301152051795,
+      "grad_norm": 4.671173505146678,
+      "learning_rate": 2e-05,
+      "loss": 0.1382,
+      "step": 45286
+    },
+    {
+      "epoch": 1.077953917928211,
+      "grad_norm": 0.08844623758874179,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 45287
+    },
+    {
+      "epoch": 1.0779777206512424,
+      "grad_norm": 0.014764053471601243,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45288
+    },
+    {
+      "epoch": 1.0780015233742741,
+      "grad_norm": 0.021903985055432056,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45289
+    },
+    {
+      "epoch": 1.0780253260973056,
+      "grad_norm": 0.2711080378863114,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 45290
+    },
+    {
+      "epoch": 1.078049128820337,
+      "grad_norm": 3.1635465684437865,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 45291
+    },
+    {
+      "epoch": 1.0780729315433686,
+      "grad_norm": 0.09812407767389568,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 45292
+    },
+    {
+      "epoch": 1.0780967342664,
+      "grad_norm": 1.109926909813784,
+      "learning_rate": 2e-05,
+      "loss": 0.0132,
+      "step": 45293
+    },
+    {
+      "epoch": 1.0781205369894316,
+      "grad_norm": 3.208522977106589,
+      "learning_rate": 2e-05,
+      "loss": 0.4462,
+      "step": 45294
+    },
+    {
+      "epoch": 1.078144339712463,
+      "grad_norm": 0.12446071431163282,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 45295
+    },
+    {
+      "epoch": 1.0781681424354945,
+      "grad_norm": 0.3403802860901896,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 45296
+    },
+    {
+      "epoch": 1.0781919451585262,
+      "grad_norm": 4.71455768140344,
+      "learning_rate": 2e-05,
+      "loss": 0.5644,
+      "step": 45297
+    },
+    {
+      "epoch": 1.0782157478815577,
+      "grad_norm": 8.998396697197013,
+      "learning_rate": 2e-05,
+      "loss": 0.3423,
+      "step": 45298
+    },
+    {
+      "epoch": 1.0782395506045892,
+      "grad_norm": 4.487759549760224,
+      "learning_rate": 2e-05,
+      "loss": 0.1933,
+      "step": 45299
+    },
+    {
+      "epoch": 1.0782633533276207,
+      "grad_norm": 6.851130481558015,
+      "learning_rate": 2e-05,
+      "loss": 0.1777,
+      "step": 45300
+    },
+    {
+      "epoch": 1.0782871560506522,
+      "grad_norm": 0.31688554058851587,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 45301
+    },
+    {
+      "epoch": 1.0783109587736837,
+      "grad_norm": 5.688120675266002,
+      "learning_rate": 2e-05,
+      "loss": 0.7908,
+      "step": 45302
+    },
+    {
+      "epoch": 1.0783347614967151,
+      "grad_norm": 0.1704043713832887,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 45303
+    },
+    {
+      "epoch": 1.0783585642197466,
+      "grad_norm": 4.65598742841998,
+      "learning_rate": 2e-05,
+      "loss": 0.2481,
+      "step": 45304
+    },
+    {
+      "epoch": 1.0783823669427783,
+      "grad_norm": 8.343495658665917,
+      "learning_rate": 2e-05,
+      "loss": 0.6938,
+      "step": 45305
+    },
+    {
+      "epoch": 1.0784061696658098,
+      "grad_norm": 2.0861766471062118,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 45306
+    },
+    {
+      "epoch": 1.0784299723888413,
+      "grad_norm": 3.788909680470455,
+      "learning_rate": 2e-05,
+      "loss": 0.3396,
+      "step": 45307
+    },
+    {
+      "epoch": 1.0784537751118728,
+      "grad_norm": 11.82970545255003,
+      "learning_rate": 2e-05,
+      "loss": 0.5688,
+      "step": 45308
+    },
+    {
+      "epoch": 1.0784775778349043,
+      "grad_norm": 3.2071427648589,
+      "learning_rate": 2e-05,
+      "loss": 0.4389,
+      "step": 45309
+    },
+    {
+      "epoch": 1.0785013805579358,
+      "grad_norm": 0.9847310776021426,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 45310
+    },
+    {
+      "epoch": 1.0785251832809672,
+      "grad_norm": 3.6557879418541512,
+      "learning_rate": 2e-05,
+      "loss": 0.5267,
+      "step": 45311
+    },
+    {
+      "epoch": 1.078548986003999,
+      "grad_norm": 10.140817222589972,
+      "learning_rate": 2e-05,
+      "loss": 0.2231,
+      "step": 45312
+    },
+    {
+      "epoch": 1.0785727887270304,
+      "grad_norm": 1.209799232456728,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 45313
+    },
+    {
+      "epoch": 1.078596591450062,
+      "grad_norm": 0.3584424109446921,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 45314
+    },
+    {
+      "epoch": 1.0786203941730934,
+      "grad_norm": 4.322157505349772,
+      "learning_rate": 2e-05,
+      "loss": 0.0766,
+      "step": 45315
+    },
+    {
+      "epoch": 1.078644196896125,
+      "grad_norm": 4.863010861151317,
+      "learning_rate": 2e-05,
+      "loss": 0.3533,
+      "step": 45316
+    },
+    {
+      "epoch": 1.0786679996191564,
+      "grad_norm": 3.9730710681826564,
+      "learning_rate": 2e-05,
+      "loss": 0.23,
+      "step": 45317
+    },
+    {
+      "epoch": 1.0786918023421879,
+      "grad_norm": 4.539446555068996,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 45318
+    },
+    {
+      "epoch": 1.0787156050652196,
+      "grad_norm": 2.51370216726316,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 45319
+    },
+    {
+      "epoch": 1.078739407788251,
+      "grad_norm": 4.080956900221881,
+      "learning_rate": 2e-05,
+      "loss": 0.5,
+      "step": 45320
+    },
+    {
+      "epoch": 1.0787632105112825,
+      "grad_norm": 0.008065672924373593,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45321
+    },
+    {
+      "epoch": 1.078787013234314,
+      "grad_norm": 16.966981186661318,
+      "learning_rate": 2e-05,
+      "loss": 0.8561,
+      "step": 45322
+    },
+    {
+      "epoch": 1.0788108159573455,
+      "grad_norm": 4.589840974614353,
+      "learning_rate": 2e-05,
+      "loss": 0.4605,
+      "step": 45323
+    },
+    {
+      "epoch": 1.078834618680377,
+      "grad_norm": 0.041675452087137674,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45324
+    },
+    {
+      "epoch": 1.0788584214034085,
+      "grad_norm": 4.347795872432017,
+      "learning_rate": 2e-05,
+      "loss": 0.1326,
+      "step": 45325
+    },
+    {
+      "epoch": 1.07888222412644,
+      "grad_norm": 0.11778055834160205,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45326
+    },
+    {
+      "epoch": 1.0789060268494717,
+      "grad_norm": 14.2224734851094,
+      "learning_rate": 2e-05,
+      "loss": 1.2353,
+      "step": 45327
+    },
+    {
+      "epoch": 1.0789298295725032,
+      "grad_norm": 0.017068475362325,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45328
+    },
+    {
+      "epoch": 1.0789536322955346,
+      "grad_norm": 8.9384011706416,
+      "learning_rate": 2e-05,
+      "loss": 0.9884,
+      "step": 45329
+    },
+    {
+      "epoch": 1.0789774350185661,
+      "grad_norm": 2.191011374023882,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 45330
+    },
+    {
+      "epoch": 1.0790012377415976,
+      "grad_norm": 8.221155807858558,
+      "learning_rate": 2e-05,
+      "loss": 0.2207,
+      "step": 45331
+    },
+    {
+      "epoch": 1.079025040464629,
+      "grad_norm": 17.29232179924515,
+      "learning_rate": 2e-05,
+      "loss": 0.2462,
+      "step": 45332
+    },
+    {
+      "epoch": 1.0790488431876606,
+      "grad_norm": 0.06498475931288111,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45333
+    },
+    {
+      "epoch": 1.079072645910692,
+      "grad_norm": 0.21429273122154546,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 45334
+    },
+    {
+      "epoch": 1.0790964486337238,
+      "grad_norm": 0.011576776151443325,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45335
+    },
+    {
+      "epoch": 1.0791202513567553,
+      "grad_norm": 1.68368008762459,
+      "learning_rate": 2e-05,
+      "loss": 0.0638,
+      "step": 45336
+    },
+    {
+      "epoch": 1.0791440540797868,
+      "grad_norm": 4.6742790704417665,
+      "learning_rate": 2e-05,
+      "loss": 0.1862,
+      "step": 45337
+    },
+    {
+      "epoch": 1.0791678568028182,
+      "grad_norm": 8.897292816363871,
+      "learning_rate": 2e-05,
+      "loss": 0.4847,
+      "step": 45338
+    },
+    {
+      "epoch": 1.0791916595258497,
+      "grad_norm": 3.9931428945107776,
+      "learning_rate": 2e-05,
+      "loss": 0.3916,
+      "step": 45339
+    },
+    {
+      "epoch": 1.0792154622488812,
+      "grad_norm": 8.415680948244585,
+      "learning_rate": 2e-05,
+      "loss": 0.2449,
+      "step": 45340
+    },
+    {
+      "epoch": 1.0792392649719127,
+      "grad_norm": 0.4327273911582689,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 45341
+    },
+    {
+      "epoch": 1.0792630676949444,
+      "grad_norm": 4.066755541069646,
+      "learning_rate": 2e-05,
+      "loss": 0.5212,
+      "step": 45342
+    },
+    {
+      "epoch": 1.0792868704179759,
+      "grad_norm": 0.11227672217269992,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 45343
+    },
+    {
+      "epoch": 1.0793106731410074,
+      "grad_norm": 0.2830484913849911,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 45344
+    },
+    {
+      "epoch": 1.0793344758640389,
+      "grad_norm": 0.025346847388869886,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45345
+    },
+    {
+      "epoch": 1.0793582785870703,
+      "grad_norm": 0.017183408026284153,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45346
+    },
+    {
+      "epoch": 1.0793820813101018,
+      "grad_norm": 3.8395684107980887,
+      "learning_rate": 2e-05,
+      "loss": 0.3072,
+      "step": 45347
+    },
+    {
+      "epoch": 1.0794058840331333,
+      "grad_norm": 4.734247310829403,
+      "learning_rate": 2e-05,
+      "loss": 0.0405,
+      "step": 45348
+    },
+    {
+      "epoch": 1.079429686756165,
+      "grad_norm": 5.6427881505100395,
+      "learning_rate": 2e-05,
+      "loss": 0.6272,
+      "step": 45349
+    },
+    {
+      "epoch": 1.0794534894791965,
+      "grad_norm": 2.86484690598646,
+      "learning_rate": 2e-05,
+      "loss": 0.4879,
+      "step": 45350
+    },
+    {
+      "epoch": 1.079477292202228,
+      "grad_norm": 3.7978444182844515,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 45351
+    },
+    {
+      "epoch": 1.0795010949252595,
+      "grad_norm": 3.9485515724089284,
+      "learning_rate": 2e-05,
+      "loss": 0.2637,
+      "step": 45352
+    },
+    {
+      "epoch": 1.079524897648291,
+      "grad_norm": 3.745521614091919,
+      "learning_rate": 2e-05,
+      "loss": 0.3442,
+      "step": 45353
+    },
+    {
+      "epoch": 1.0795487003713224,
+      "grad_norm": 2.2237924178680024,
+      "learning_rate": 2e-05,
+      "loss": 0.0794,
+      "step": 45354
+    },
+    {
+      "epoch": 1.079572503094354,
+      "grad_norm": 2.799692522336581,
+      "learning_rate": 2e-05,
+      "loss": 0.0421,
+      "step": 45355
+    },
+    {
+      "epoch": 1.0795963058173854,
+      "grad_norm": 9.42702477053052,
+      "learning_rate": 2e-05,
+      "loss": 0.9604,
+      "step": 45356
+    },
+    {
+      "epoch": 1.0796201085404171,
+      "grad_norm": 3.9155499905059896,
+      "learning_rate": 2e-05,
+      "loss": 0.3494,
+      "step": 45357
+    },
+    {
+      "epoch": 1.0796439112634486,
+      "grad_norm": 2.5419316246050445,
+      "learning_rate": 2e-05,
+      "loss": 0.4426,
+      "step": 45358
+    },
+    {
+      "epoch": 1.07966771398648,
+      "grad_norm": 10.056196793408681,
+      "learning_rate": 2e-05,
+      "loss": 0.4042,
+      "step": 45359
+    },
+    {
+      "epoch": 1.0796915167095116,
+      "grad_norm": 2.111398211544383,
+      "learning_rate": 2e-05,
+      "loss": 0.0164,
+      "step": 45360
+    },
+    {
+      "epoch": 1.079715319432543,
+      "grad_norm": 3.7296036400828787,
+      "learning_rate": 2e-05,
+      "loss": 0.4045,
+      "step": 45361
+    },
+    {
+      "epoch": 1.0797391221555745,
+      "grad_norm": 0.07521487114747048,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45362
+    },
+    {
+      "epoch": 1.079762924878606,
+      "grad_norm": 0.1629567289112205,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 45363
+    },
+    {
+      "epoch": 1.0797867276016375,
+      "grad_norm": 1.7980747062499096,
+      "learning_rate": 2e-05,
+      "loss": 0.0481,
+      "step": 45364
+    },
+    {
+      "epoch": 1.0798105303246692,
+      "grad_norm": 0.2643393395259559,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 45365
+    },
+    {
+      "epoch": 1.0798343330477007,
+      "grad_norm": 4.56389079620376,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 45366
+    },
+    {
+      "epoch": 1.0798581357707322,
+      "grad_norm": 0.47236296760178914,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 45367
+    },
+    {
+      "epoch": 1.0798819384937637,
+      "grad_norm": 10.477330458534864,
+      "learning_rate": 2e-05,
+      "loss": 0.6258,
+      "step": 45368
+    },
+    {
+      "epoch": 1.0799057412167952,
+      "grad_norm": 0.019352837644289436,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45369
+    },
+    {
+      "epoch": 1.0799295439398267,
+      "grad_norm": 4.82974365081095,
+      "learning_rate": 2e-05,
+      "loss": 0.4566,
+      "step": 45370
+    },
+    {
+      "epoch": 1.0799533466628581,
+      "grad_norm": 0.24372558900298927,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 45371
+    },
+    {
+      "epoch": 1.0799771493858898,
+      "grad_norm": 2.9093434733976773,
+      "learning_rate": 2e-05,
+      "loss": 0.3552,
+      "step": 45372
+    },
+    {
+      "epoch": 1.0800009521089213,
+      "grad_norm": 15.328416651511407,
+      "learning_rate": 2e-05,
+      "loss": 0.9589,
+      "step": 45373
+    },
+    {
+      "epoch": 1.0800247548319528,
+      "grad_norm": 0.529459469395006,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 45374
+    },
+    {
+      "epoch": 1.0800485575549843,
+      "grad_norm": 0.42085847270269877,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 45375
+    },
+    {
+      "epoch": 1.0800723602780158,
+      "grad_norm": 4.5305698944984645,
+      "learning_rate": 2e-05,
+      "loss": 0.2902,
+      "step": 45376
+    },
+    {
+      "epoch": 1.0800961630010473,
+      "grad_norm": 0.38504069941172037,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 45377
+    },
+    {
+      "epoch": 1.0801199657240788,
+      "grad_norm": 3.3092911827799307,
+      "learning_rate": 2e-05,
+      "loss": 0.1682,
+      "step": 45378
+    },
+    {
+      "epoch": 1.0801437684471105,
+      "grad_norm": 1.9677178394715522,
+      "learning_rate": 2e-05,
+      "loss": 0.2066,
+      "step": 45379
+    },
+    {
+      "epoch": 1.080167571170142,
+      "grad_norm": 0.029396760782483735,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45380
+    },
+    {
+      "epoch": 1.0801913738931734,
+      "grad_norm": 0.6193221575538085,
+      "learning_rate": 2e-05,
+      "loss": 0.0092,
+      "step": 45381
+    },
+    {
+      "epoch": 1.080215176616205,
+      "grad_norm": 3.1241063314962467,
+      "learning_rate": 2e-05,
+      "loss": 0.4871,
+      "step": 45382
+    },
+    {
+      "epoch": 1.0802389793392364,
+      "grad_norm": 21.710969944810355,
+      "learning_rate": 2e-05,
+      "loss": 0.465,
+      "step": 45383
+    },
+    {
+      "epoch": 1.080262782062268,
+      "grad_norm": 0.5986605705775819,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 45384
+    },
+    {
+      "epoch": 1.0802865847852994,
+      "grad_norm": 0.7439202177327477,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 45385
+    },
+    {
+      "epoch": 1.0803103875083309,
+      "grad_norm": 2.235691690326008,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 45386
+    },
+    {
+      "epoch": 1.0803341902313626,
+      "grad_norm": 3.342801506909379,
+      "learning_rate": 2e-05,
+      "loss": 0.3738,
+      "step": 45387
+    },
+    {
+      "epoch": 1.080357992954394,
+      "grad_norm": 3.413014347687078,
+      "learning_rate": 2e-05,
+      "loss": 0.4827,
+      "step": 45388
+    },
+    {
+      "epoch": 1.0803817956774255,
+      "grad_norm": 2.7726549235358875,
+      "learning_rate": 2e-05,
+      "loss": 0.2683,
+      "step": 45389
+    },
+    {
+      "epoch": 1.080405598400457,
+      "grad_norm": 4.714324598327141,
+      "learning_rate": 2e-05,
+      "loss": 0.5758,
+      "step": 45390
+    },
+    {
+      "epoch": 1.0804294011234885,
+      "grad_norm": 3.8600522861823605,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 45391
+    },
+    {
+      "epoch": 1.08045320384652,
+      "grad_norm": 3.6922361639185626,
+      "learning_rate": 2e-05,
+      "loss": 0.533,
+      "step": 45392
+    },
+    {
+      "epoch": 1.0804770065695515,
+      "grad_norm": 0.36599735492513236,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45393
+    },
+    {
+      "epoch": 1.080500809292583,
+      "grad_norm": 0.37418740338597245,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45394
+    },
+    {
+      "epoch": 1.0805246120156147,
+      "grad_norm": 0.07680657331296446,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45395
+    },
+    {
+      "epoch": 1.0805484147386462,
+      "grad_norm": 29.62248810074519,
+      "learning_rate": 2e-05,
+      "loss": 0.7548,
+      "step": 45396
+    },
+    {
+      "epoch": 1.0805722174616776,
+      "grad_norm": 0.037267983282326195,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45397
+    },
+    {
+      "epoch": 1.0805960201847091,
+      "grad_norm": 3.0983742088704442,
+      "learning_rate": 2e-05,
+      "loss": 0.4069,
+      "step": 45398
+    },
+    {
+      "epoch": 1.0806198229077406,
+      "grad_norm": 0.7702640284479337,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 45399
+    },
+    {
+      "epoch": 1.080643625630772,
+      "grad_norm": 3.79090453950649,
+      "learning_rate": 2e-05,
+      "loss": 0.0377,
+      "step": 45400
+    },
+    {
+      "epoch": 1.0806674283538036,
+      "grad_norm": 7.007219406077506,
+      "learning_rate": 2e-05,
+      "loss": 1.1395,
+      "step": 45401
+    },
+    {
+      "epoch": 1.0806912310768353,
+      "grad_norm": 5.426903048976697,
+      "learning_rate": 2e-05,
+      "loss": 1.1667,
+      "step": 45402
+    },
+    {
+      "epoch": 1.0807150337998668,
+      "grad_norm": 5.96276082649047,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 45403
+    },
+    {
+      "epoch": 1.0807388365228983,
+      "grad_norm": 6.555443363762821,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 45404
+    },
+    {
+      "epoch": 1.0807626392459297,
+      "grad_norm": 11.95692134200577,
+      "learning_rate": 2e-05,
+      "loss": 0.7426,
+      "step": 45405
+    },
+    {
+      "epoch": 1.0807864419689612,
+      "grad_norm": 5.495587029695466,
+      "learning_rate": 2e-05,
+      "loss": 0.6572,
+      "step": 45406
+    },
+    {
+      "epoch": 1.0808102446919927,
+      "grad_norm": 15.48080944327734,
+      "learning_rate": 2e-05,
+      "loss": 0.338,
+      "step": 45407
+    },
+    {
+      "epoch": 1.0808340474150242,
+      "grad_norm": 5.046089755075301,
+      "learning_rate": 2e-05,
+      "loss": 0.4456,
+      "step": 45408
+    },
+    {
+      "epoch": 1.080857850138056,
+      "grad_norm": 4.525197072931684,
+      "learning_rate": 2e-05,
+      "loss": 0.5961,
+      "step": 45409
+    },
+    {
+      "epoch": 1.0808816528610874,
+      "grad_norm": 0.03712413871789715,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45410
+    },
+    {
+      "epoch": 1.0809054555841189,
+      "grad_norm": 12.898726370891184,
+      "learning_rate": 2e-05,
+      "loss": 0.1149,
+      "step": 45411
+    },
+    {
+      "epoch": 1.0809292583071504,
+      "grad_norm": 7.115453330655951,
+      "learning_rate": 2e-05,
+      "loss": 0.0607,
+      "step": 45412
+    },
+    {
+      "epoch": 1.0809530610301818,
+      "grad_norm": 4.0314213899150095,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 45413
+    },
+    {
+      "epoch": 1.0809768637532133,
+      "grad_norm": 11.725606966390194,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 45414
+    },
+    {
+      "epoch": 1.0810006664762448,
+      "grad_norm": 2.946881980707033,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 45415
+    },
+    {
+      "epoch": 1.0810244691992763,
+      "grad_norm": 11.97499446220667,
+      "learning_rate": 2e-05,
+      "loss": 0.9518,
+      "step": 45416
+    },
+    {
+      "epoch": 1.081048271922308,
+      "grad_norm": 4.687998313724622,
+      "learning_rate": 2e-05,
+      "loss": 0.4811,
+      "step": 45417
+    },
+    {
+      "epoch": 1.0810720746453395,
+      "grad_norm": 0.04841293397184284,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45418
+    },
+    {
+      "epoch": 1.081095877368371,
+      "grad_norm": 6.7950127802207465,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 45419
+    },
+    {
+      "epoch": 1.0811196800914025,
+      "grad_norm": 7.184181573603486,
+      "learning_rate": 2e-05,
+      "loss": 0.1849,
+      "step": 45420
+    },
+    {
+      "epoch": 1.081143482814434,
+      "grad_norm": 16.701924098392173,
+      "learning_rate": 2e-05,
+      "loss": 0.158,
+      "step": 45421
+    },
+    {
+      "epoch": 1.0811672855374654,
+      "grad_norm": 4.276309329996251,
+      "learning_rate": 2e-05,
+      "loss": 0.0647,
+      "step": 45422
+    },
+    {
+      "epoch": 1.081191088260497,
+      "grad_norm": 4.373224615815516,
+      "learning_rate": 2e-05,
+      "loss": 0.1554,
+      "step": 45423
+    },
+    {
+      "epoch": 1.0812148909835284,
+      "grad_norm": 0.9998005405351602,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 45424
+    },
+    {
+      "epoch": 1.0812386937065601,
+      "grad_norm": 3.410080633565804,
+      "learning_rate": 2e-05,
+      "loss": 0.0891,
+      "step": 45425
+    },
+    {
+      "epoch": 1.0812624964295916,
+      "grad_norm": 2.567437970858121,
+      "learning_rate": 2e-05,
+      "loss": 0.0279,
+      "step": 45426
+    },
+    {
+      "epoch": 1.081286299152623,
+      "grad_norm": 0.15758771181102169,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 45427
+    },
+    {
+      "epoch": 1.0813101018756546,
+      "grad_norm": 1.1349338075148943,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 45428
+    },
+    {
+      "epoch": 1.081333904598686,
+      "grad_norm": 2.901289719814969,
+      "learning_rate": 2e-05,
+      "loss": 0.233,
+      "step": 45429
+    },
+    {
+      "epoch": 1.0813577073217175,
+      "grad_norm": 0.1517600903042771,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45430
+    },
+    {
+      "epoch": 1.081381510044749,
+      "grad_norm": 7.36559735542158,
+      "learning_rate": 2e-05,
+      "loss": 0.8414,
+      "step": 45431
+    },
+    {
+      "epoch": 1.0814053127677807,
+      "grad_norm": 7.150855887181076,
+      "learning_rate": 2e-05,
+      "loss": 0.9028,
+      "step": 45432
+    },
+    {
+      "epoch": 1.0814291154908122,
+      "grad_norm": 0.031939463464969814,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45433
+    },
+    {
+      "epoch": 1.0814529182138437,
+      "grad_norm": 1.197586786994236,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 45434
+    },
+    {
+      "epoch": 1.0814767209368752,
+      "grad_norm": 4.764286928712361,
+      "learning_rate": 2e-05,
+      "loss": 0.1274,
+      "step": 45435
+    },
+    {
+      "epoch": 1.0815005236599067,
+      "grad_norm": 0.24086151920510757,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 45436
+    },
+    {
+      "epoch": 1.0815243263829382,
+      "grad_norm": 10.661704349512867,
+      "learning_rate": 2e-05,
+      "loss": 0.5949,
+      "step": 45437
+    },
+    {
+      "epoch": 1.0815481291059696,
+      "grad_norm": 4.223050422331231,
+      "learning_rate": 2e-05,
+      "loss": 0.4689,
+      "step": 45438
+    },
+    {
+      "epoch": 1.0815719318290014,
+      "grad_norm": 6.059979863383225,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 45439
+    },
+    {
+      "epoch": 1.0815957345520328,
+      "grad_norm": 4.08978765411432,
+      "learning_rate": 2e-05,
+      "loss": 0.0855,
+      "step": 45440
+    },
+    {
+      "epoch": 1.0816195372750643,
+      "grad_norm": 4.355727254155083,
+      "learning_rate": 2e-05,
+      "loss": 0.644,
+      "step": 45441
+    },
+    {
+      "epoch": 1.0816433399980958,
+      "grad_norm": 0.05151141490279917,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45442
+    },
+    {
+      "epoch": 1.0816671427211273,
+      "grad_norm": 0.29759624201447643,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 45443
+    },
+    {
+      "epoch": 1.0816909454441588,
+      "grad_norm": 1.6552070766934999,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 45444
+    },
+    {
+      "epoch": 1.0817147481671903,
+      "grad_norm": 5.0754710779136065,
+      "learning_rate": 2e-05,
+      "loss": 0.474,
+      "step": 45445
+    },
+    {
+      "epoch": 1.0817385508902218,
+      "grad_norm": 3.483455934486966,
+      "learning_rate": 2e-05,
+      "loss": 0.0659,
+      "step": 45446
+    },
+    {
+      "epoch": 1.0817623536132535,
+      "grad_norm": 9.414376021287033,
+      "learning_rate": 2e-05,
+      "loss": 0.416,
+      "step": 45447
+    },
+    {
+      "epoch": 1.081786156336285,
+      "grad_norm": 0.026927421453161637,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45448
+    },
+    {
+      "epoch": 1.0818099590593164,
+      "grad_norm": 0.2844534671047113,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 45449
+    },
+    {
+      "epoch": 1.081833761782348,
+      "grad_norm": 3.6723171211564583,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 45450
+    },
+    {
+      "epoch": 1.0818575645053794,
+      "grad_norm": 14.539922539151304,
+      "learning_rate": 2e-05,
+      "loss": 0.9367,
+      "step": 45451
+    },
+    {
+      "epoch": 1.0818813672284109,
+      "grad_norm": 5.549178660332209,
+      "learning_rate": 2e-05,
+      "loss": 0.6515,
+      "step": 45452
+    },
+    {
+      "epoch": 1.0819051699514424,
+      "grad_norm": 0.0921539419574663,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45453
+    },
+    {
+      "epoch": 1.0819289726744739,
+      "grad_norm": 0.04017444004294511,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45454
+    },
+    {
+      "epoch": 1.0819527753975056,
+      "grad_norm": 0.602879816269237,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45455
+    },
+    {
+      "epoch": 1.081976578120537,
+      "grad_norm": 0.23490660649667458,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 45456
+    },
+    {
+      "epoch": 1.0820003808435685,
+      "grad_norm": 0.12186627335677773,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45457
+    },
+    {
+      "epoch": 1.0820241835666,
+      "grad_norm": 6.866583463638033,
+      "learning_rate": 2e-05,
+      "loss": 0.1152,
+      "step": 45458
+    },
+    {
+      "epoch": 1.0820479862896315,
+      "grad_norm": 4.093827075240282,
+      "learning_rate": 2e-05,
+      "loss": 0.1935,
+      "step": 45459
+    },
+    {
+      "epoch": 1.082071789012663,
+      "grad_norm": 2.1692067812116953,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 45460
+    },
+    {
+      "epoch": 1.0820955917356945,
+      "grad_norm": 10.704929853343895,
+      "learning_rate": 2e-05,
+      "loss": 0.5133,
+      "step": 45461
+    },
+    {
+      "epoch": 1.0821193944587262,
+      "grad_norm": 4.218890551673295,
+      "learning_rate": 2e-05,
+      "loss": 0.4153,
+      "step": 45462
+    },
+    {
+      "epoch": 1.0821431971817577,
+      "grad_norm": 0.017436179994680646,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45463
+    },
+    {
+      "epoch": 1.0821669999047892,
+      "grad_norm": 0.11946585986813038,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45464
+    },
+    {
+      "epoch": 1.0821908026278206,
+      "grad_norm": 0.09707568921607891,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45465
+    },
+    {
+      "epoch": 1.0822146053508521,
+      "grad_norm": 5.0736339040557255,
+      "learning_rate": 2e-05,
+      "loss": 0.0654,
+      "step": 45466
+    },
+    {
+      "epoch": 1.0822384080738836,
+      "grad_norm": 7.094650429887152,
+      "learning_rate": 2e-05,
+      "loss": 0.8209,
+      "step": 45467
+    },
+    {
+      "epoch": 1.082262210796915,
+      "grad_norm": 3.0583738907671236,
+      "learning_rate": 2e-05,
+      "loss": 0.4313,
+      "step": 45468
+    },
+    {
+      "epoch": 1.0822860135199468,
+      "grad_norm": 4.329900713086018,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 45469
+    },
+    {
+      "epoch": 1.0823098162429783,
+      "grad_norm": 0.0589692792953149,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45470
+    },
+    {
+      "epoch": 1.0823336189660098,
+      "grad_norm": 0.37953569920508373,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 45471
+    },
+    {
+      "epoch": 1.0823574216890413,
+      "grad_norm": 2.05667353080917,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 45472
+    },
+    {
+      "epoch": 1.0823812244120727,
+      "grad_norm": 3.814306569153644,
+      "learning_rate": 2e-05,
+      "loss": 0.0287,
+      "step": 45473
+    },
+    {
+      "epoch": 1.0824050271351042,
+      "grad_norm": 0.41510868383562743,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 45474
+    },
+    {
+      "epoch": 1.0824288298581357,
+      "grad_norm": 0.06420349087704849,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45475
+    },
+    {
+      "epoch": 1.0824526325811672,
+      "grad_norm": 3.726721937435302,
+      "learning_rate": 2e-05,
+      "loss": 0.1632,
+      "step": 45476
+    },
+    {
+      "epoch": 1.082476435304199,
+      "grad_norm": 1.656878367461669,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 45477
+    },
+    {
+      "epoch": 1.0825002380272304,
+      "grad_norm": 3.527719705813169,
+      "learning_rate": 2e-05,
+      "loss": 0.3563,
+      "step": 45478
+    },
+    {
+      "epoch": 1.0825240407502619,
+      "grad_norm": 0.4995389006601603,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 45479
+    },
+    {
+      "epoch": 1.0825478434732934,
+      "grad_norm": 2.760315135341828,
+      "learning_rate": 2e-05,
+      "loss": 0.0862,
+      "step": 45480
+    },
+    {
+      "epoch": 1.0825716461963248,
+      "grad_norm": 5.804165321099304,
+      "learning_rate": 2e-05,
+      "loss": 0.7188,
+      "step": 45481
+    },
+    {
+      "epoch": 1.0825954489193563,
+      "grad_norm": 5.121824696614476,
+      "learning_rate": 2e-05,
+      "loss": 0.6363,
+      "step": 45482
+    },
+    {
+      "epoch": 1.0826192516423878,
+      "grad_norm": 4.579552955455817,
+      "learning_rate": 2e-05,
+      "loss": 0.2728,
+      "step": 45483
+    },
+    {
+      "epoch": 1.0826430543654193,
+      "grad_norm": 3.108618790078428,
+      "learning_rate": 2e-05,
+      "loss": 0.3879,
+      "step": 45484
+    },
+    {
+      "epoch": 1.082666857088451,
+      "grad_norm": 6.993815693878853,
+      "learning_rate": 2e-05,
+      "loss": 0.546,
+      "step": 45485
+    },
+    {
+      "epoch": 1.0826906598114825,
+      "grad_norm": 0.005468474676696797,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45486
+    },
+    {
+      "epoch": 1.082714462534514,
+      "grad_norm": 7.699223613248108,
+      "learning_rate": 2e-05,
+      "loss": 0.3849,
+      "step": 45487
+    },
+    {
+      "epoch": 1.0827382652575455,
+      "grad_norm": 3.576470249520566,
+      "learning_rate": 2e-05,
+      "loss": 0.449,
+      "step": 45488
+    },
+    {
+      "epoch": 1.082762067980577,
+      "grad_norm": 6.990864668803391,
+      "learning_rate": 2e-05,
+      "loss": 0.3921,
+      "step": 45489
+    },
+    {
+      "epoch": 1.0827858707036084,
+      "grad_norm": 19.162229080962366,
+      "learning_rate": 2e-05,
+      "loss": 0.4105,
+      "step": 45490
+    },
+    {
+      "epoch": 1.08280967342664,
+      "grad_norm": 7.170718173068114,
+      "learning_rate": 2e-05,
+      "loss": 1.1292,
+      "step": 45491
+    },
+    {
+      "epoch": 1.0828334761496716,
+      "grad_norm": 3.4375846941420347,
+      "learning_rate": 2e-05,
+      "loss": 0.3871,
+      "step": 45492
+    },
+    {
+      "epoch": 1.082857278872703,
+      "grad_norm": 0.6488459679081543,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 45493
+    },
+    {
+      "epoch": 1.0828810815957346,
+      "grad_norm": 2.946839495767851,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 45494
+    },
+    {
+      "epoch": 1.082904884318766,
+      "grad_norm": 0.5719933577436717,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45495
+    },
+    {
+      "epoch": 1.0829286870417976,
+      "grad_norm": 13.267350091513354,
+      "learning_rate": 2e-05,
+      "loss": 1.1668,
+      "step": 45496
+    },
+    {
+      "epoch": 1.082952489764829,
+      "grad_norm": 0.5675178557182333,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 45497
+    },
+    {
+      "epoch": 1.0829762924878605,
+      "grad_norm": 3.3135316799069536,
+      "learning_rate": 2e-05,
+      "loss": 0.392,
+      "step": 45498
+    },
+    {
+      "epoch": 1.083000095210892,
+      "grad_norm": 4.4864741138343485,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 45499
+    },
+    {
+      "epoch": 1.0830238979339237,
+      "grad_norm": 2.8705764179761384,
+      "learning_rate": 2e-05,
+      "loss": 0.2741,
+      "step": 45500
+    },
+    {
+      "epoch": 1.0830477006569552,
+      "grad_norm": 2.7232968680265177,
+      "learning_rate": 2e-05,
+      "loss": 0.0383,
+      "step": 45501
+    },
+    {
+      "epoch": 1.0830715033799867,
+      "grad_norm": 4.865979468575473,
+      "learning_rate": 2e-05,
+      "loss": 0.4267,
+      "step": 45502
+    },
+    {
+      "epoch": 1.0830953061030182,
+      "grad_norm": 5.824028295591088,
+      "learning_rate": 2e-05,
+      "loss": 0.7483,
+      "step": 45503
+    },
+    {
+      "epoch": 1.0831191088260497,
+      "grad_norm": 0.005862128056924724,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45504
+    },
+    {
+      "epoch": 1.0831429115490812,
+      "grad_norm": 4.811469655006753,
+      "learning_rate": 2e-05,
+      "loss": 0.0849,
+      "step": 45505
+    },
+    {
+      "epoch": 1.0831667142721126,
+      "grad_norm": 4.813302743359032,
+      "learning_rate": 2e-05,
+      "loss": 0.6593,
+      "step": 45506
+    },
+    {
+      "epoch": 1.0831905169951443,
+      "grad_norm": 0.14920261203693602,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45507
+    },
+    {
+      "epoch": 1.0832143197181758,
+      "grad_norm": 8.569963249385967,
+      "learning_rate": 2e-05,
+      "loss": 0.6169,
+      "step": 45508
+    },
+    {
+      "epoch": 1.0832381224412073,
+      "grad_norm": 0.28568951850399205,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 45509
+    },
+    {
+      "epoch": 1.0832619251642388,
+      "grad_norm": 0.0860267732193617,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45510
+    },
+    {
+      "epoch": 1.0832857278872703,
+      "grad_norm": 0.051232613249704996,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45511
+    },
+    {
+      "epoch": 1.0833095306103018,
+      "grad_norm": 2.2474271477345167,
+      "learning_rate": 2e-05,
+      "loss": 0.0465,
+      "step": 45512
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 0.6228187338228526,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 45513
+    },
+    {
+      "epoch": 1.0833571360563647,
+      "grad_norm": 0.27297899643368806,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 45514
+    },
+    {
+      "epoch": 1.0833809387793965,
+      "grad_norm": 1.3054674474812624,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 45515
+    },
+    {
+      "epoch": 1.083404741502428,
+      "grad_norm": 0.0790635260420678,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 45516
+    },
+    {
+      "epoch": 1.0834285442254594,
+      "grad_norm": 5.412071780071576,
+      "learning_rate": 2e-05,
+      "loss": 0.0491,
+      "step": 45517
+    },
+    {
+      "epoch": 1.083452346948491,
+      "grad_norm": 7.796523147988369,
+      "learning_rate": 2e-05,
+      "loss": 0.3863,
+      "step": 45518
+    },
+    {
+      "epoch": 1.0834761496715224,
+      "grad_norm": 0.043622246402511757,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45519
+    },
+    {
+      "epoch": 1.0834999523945539,
+      "grad_norm": 2.2703083250933234,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 45520
+    },
+    {
+      "epoch": 1.0835237551175854,
+      "grad_norm": 3.208240309675795,
+      "learning_rate": 2e-05,
+      "loss": 0.4503,
+      "step": 45521
+    },
+    {
+      "epoch": 1.083547557840617,
+      "grad_norm": 0.6774995360468329,
+      "learning_rate": 2e-05,
+      "loss": 0.011,
+      "step": 45522
+    },
+    {
+      "epoch": 1.0835713605636486,
+      "grad_norm": 12.14819320936766,
+      "learning_rate": 2e-05,
+      "loss": 1.2864,
+      "step": 45523
+    },
+    {
+      "epoch": 1.08359516328668,
+      "grad_norm": 0.03808227426770547,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45524
+    },
+    {
+      "epoch": 1.0836189660097115,
+      "grad_norm": 0.4272622052157699,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 45525
+    },
+    {
+      "epoch": 1.083642768732743,
+      "grad_norm": 5.914333652209977,
+      "learning_rate": 2e-05,
+      "loss": 0.5794,
+      "step": 45526
+    },
+    {
+      "epoch": 1.0836665714557745,
+      "grad_norm": 0.0410336608426196,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45527
+    },
+    {
+      "epoch": 1.083690374178806,
+      "grad_norm": 0.5056764082213427,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45528
+    },
+    {
+      "epoch": 1.0837141769018375,
+      "grad_norm": 0.06292989711800107,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45529
+    },
+    {
+      "epoch": 1.0837379796248692,
+      "grad_norm": 5.654017801749038,
+      "learning_rate": 2e-05,
+      "loss": 0.8642,
+      "step": 45530
+    },
+    {
+      "epoch": 1.0837617823479007,
+      "grad_norm": 4.2139855296447095,
+      "learning_rate": 2e-05,
+      "loss": 0.2552,
+      "step": 45531
+    },
+    {
+      "epoch": 1.0837855850709321,
+      "grad_norm": 1.1915227565847657,
+      "learning_rate": 2e-05,
+      "loss": 0.0238,
+      "step": 45532
+    },
+    {
+      "epoch": 1.0838093877939636,
+      "grad_norm": 8.624087440387083,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 45533
+    },
+    {
+      "epoch": 1.0838331905169951,
+      "grad_norm": 0.07285376483356232,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45534
+    },
+    {
+      "epoch": 1.0838569932400266,
+      "grad_norm": 6.353314923492867,
+      "learning_rate": 2e-05,
+      "loss": 0.0446,
+      "step": 45535
+    },
+    {
+      "epoch": 1.083880795963058,
+      "grad_norm": 5.09830302327993,
+      "learning_rate": 2e-05,
+      "loss": 0.2447,
+      "step": 45536
+    },
+    {
+      "epoch": 1.0839045986860896,
+      "grad_norm": 3.129331889760412,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 45537
+    },
+    {
+      "epoch": 1.0839284014091213,
+      "grad_norm": 0.03833723012872643,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45538
+    },
+    {
+      "epoch": 1.0839522041321528,
+      "grad_norm": 1.9705371507655398,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 45539
+    },
+    {
+      "epoch": 1.0839760068551842,
+      "grad_norm": 11.869119861528501,
+      "learning_rate": 2e-05,
+      "loss": 0.5077,
+      "step": 45540
+    },
+    {
+      "epoch": 1.0839998095782157,
+      "grad_norm": 3.213371108326105,
+      "learning_rate": 2e-05,
+      "loss": 0.3331,
+      "step": 45541
+    },
+    {
+      "epoch": 1.0840236123012472,
+      "grad_norm": 0.031765406350561284,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45542
+    },
+    {
+      "epoch": 1.0840474150242787,
+      "grad_norm": 3.819694018776258,
+      "learning_rate": 2e-05,
+      "loss": 0.0942,
+      "step": 45543
+    },
+    {
+      "epoch": 1.0840712177473102,
+      "grad_norm": 5.460156604367818,
+      "learning_rate": 2e-05,
+      "loss": 0.499,
+      "step": 45544
+    },
+    {
+      "epoch": 1.084095020470342,
+      "grad_norm": 0.07529040274974008,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45545
+    },
+    {
+      "epoch": 1.0841188231933734,
+      "grad_norm": 11.816035644807295,
+      "learning_rate": 2e-05,
+      "loss": 0.733,
+      "step": 45546
+    },
+    {
+      "epoch": 1.0841426259164049,
+      "grad_norm": 1.0605146884985763,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 45547
+    },
+    {
+      "epoch": 1.0841664286394364,
+      "grad_norm": 1.2825639282426187,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 45548
+    },
+    {
+      "epoch": 1.0841902313624678,
+      "grad_norm": 5.4888454807314,
+      "learning_rate": 2e-05,
+      "loss": 0.2567,
+      "step": 45549
+    },
+    {
+      "epoch": 1.0842140340854993,
+      "grad_norm": 6.3193196699447505,
+      "learning_rate": 2e-05,
+      "loss": 1.0668,
+      "step": 45550
+    },
+    {
+      "epoch": 1.0842378368085308,
+      "grad_norm": 0.17601189529010086,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45551
+    },
+    {
+      "epoch": 1.0842616395315625,
+      "grad_norm": 9.577123332236352,
+      "learning_rate": 2e-05,
+      "loss": 0.8356,
+      "step": 45552
+    },
+    {
+      "epoch": 1.084285442254594,
+      "grad_norm": 0.05464832639062954,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45553
+    },
+    {
+      "epoch": 1.0843092449776255,
+      "grad_norm": 6.4876093375969255,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 45554
+    },
+    {
+      "epoch": 1.084333047700657,
+      "grad_norm": 3.7710353372896668,
+      "learning_rate": 2e-05,
+      "loss": 0.5335,
+      "step": 45555
+    },
+    {
+      "epoch": 1.0843568504236885,
+      "grad_norm": 8.147911889285238,
+      "learning_rate": 2e-05,
+      "loss": 0.4377,
+      "step": 45556
+    },
+    {
+      "epoch": 1.08438065314672,
+      "grad_norm": 5.2330262357870145,
+      "learning_rate": 2e-05,
+      "loss": 0.9293,
+      "step": 45557
+    },
+    {
+      "epoch": 1.0844044558697514,
+      "grad_norm": 1.4830997159635102,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 45558
+    },
+    {
+      "epoch": 1.084428258592783,
+      "grad_norm": 2.2651205670253463,
+      "learning_rate": 2e-05,
+      "loss": 0.1556,
+      "step": 45559
+    },
+    {
+      "epoch": 1.0844520613158146,
+      "grad_norm": 7.50636710008571,
+      "learning_rate": 2e-05,
+      "loss": 0.6509,
+      "step": 45560
+    },
+    {
+      "epoch": 1.084475864038846,
+      "grad_norm": 0.10661389667103278,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45561
+    },
+    {
+      "epoch": 1.0844996667618776,
+      "grad_norm": 1.0675077942769853,
+      "learning_rate": 2e-05,
+      "loss": 0.0342,
+      "step": 45562
+    },
+    {
+      "epoch": 1.084523469484909,
+      "grad_norm": 12.192922564335321,
+      "learning_rate": 2e-05,
+      "loss": 1.0252,
+      "step": 45563
+    },
+    {
+      "epoch": 1.0845472722079406,
+      "grad_norm": 0.17137348520047394,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 45564
+    },
+    {
+      "epoch": 1.084571074930972,
+      "grad_norm": 0.31475733869555994,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 45565
+    },
+    {
+      "epoch": 1.0845948776540035,
+      "grad_norm": 1.8524607884408608,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 45566
+    },
+    {
+      "epoch": 1.084618680377035,
+      "grad_norm": 8.807082093514154,
+      "learning_rate": 2e-05,
+      "loss": 0.0957,
+      "step": 45567
+    },
+    {
+      "epoch": 1.0846424831000667,
+      "grad_norm": 5.373571364732211,
+      "learning_rate": 2e-05,
+      "loss": 0.61,
+      "step": 45568
+    },
+    {
+      "epoch": 1.0846662858230982,
+      "grad_norm": 5.873689739944672,
+      "learning_rate": 2e-05,
+      "loss": 0.5719,
+      "step": 45569
+    },
+    {
+      "epoch": 1.0846900885461297,
+      "grad_norm": 8.01208735624008,
+      "learning_rate": 2e-05,
+      "loss": 1.3068,
+      "step": 45570
+    },
+    {
+      "epoch": 1.0847138912691612,
+      "grad_norm": 6.196656728764918,
+      "learning_rate": 2e-05,
+      "loss": 0.4931,
+      "step": 45571
+    },
+    {
+      "epoch": 1.0847376939921927,
+      "grad_norm": 0.026340620589267012,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45572
+    },
+    {
+      "epoch": 1.0847614967152241,
+      "grad_norm": 3.8746538896870546,
+      "learning_rate": 2e-05,
+      "loss": 0.3418,
+      "step": 45573
+    },
+    {
+      "epoch": 1.0847852994382556,
+      "grad_norm": 8.763799069900118,
+      "learning_rate": 2e-05,
+      "loss": 1.1661,
+      "step": 45574
+    },
+    {
+      "epoch": 1.0848091021612873,
+      "grad_norm": 0.06438220700208476,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45575
+    },
+    {
+      "epoch": 1.0848329048843188,
+      "grad_norm": 0.07090327694487217,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 45576
+    },
+    {
+      "epoch": 1.0848567076073503,
+      "grad_norm": 6.391069959238068,
+      "learning_rate": 2e-05,
+      "loss": 0.4213,
+      "step": 45577
+    },
+    {
+      "epoch": 1.0848805103303818,
+      "grad_norm": 0.4386306532337705,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 45578
+    },
+    {
+      "epoch": 1.0849043130534133,
+      "grad_norm": 0.15685987060803303,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45579
+    },
+    {
+      "epoch": 1.0849281157764448,
+      "grad_norm": 0.019032182263618448,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45580
+    },
+    {
+      "epoch": 1.0849519184994763,
+      "grad_norm": 5.342308534116846,
+      "learning_rate": 2e-05,
+      "loss": 0.609,
+      "step": 45581
+    },
+    {
+      "epoch": 1.084975721222508,
+      "grad_norm": 1.58841984154484,
+      "learning_rate": 2e-05,
+      "loss": 0.0124,
+      "step": 45582
+    },
+    {
+      "epoch": 1.0849995239455394,
+      "grad_norm": 6.0244063129841114,
+      "learning_rate": 2e-05,
+      "loss": 0.7851,
+      "step": 45583
+    },
+    {
+      "epoch": 1.085023326668571,
+      "grad_norm": 1.6462941145124816,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 45584
+    },
+    {
+      "epoch": 1.0850471293916024,
+      "grad_norm": 0.014866816749304474,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45585
+    },
+    {
+      "epoch": 1.085070932114634,
+      "grad_norm": 1.714679864566717,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 45586
+    },
+    {
+      "epoch": 1.0850947348376654,
+      "grad_norm": 3.1641754664681523,
+      "learning_rate": 2e-05,
+      "loss": 0.2139,
+      "step": 45587
+    },
+    {
+      "epoch": 1.0851185375606969,
+      "grad_norm": 14.243204429749843,
+      "learning_rate": 2e-05,
+      "loss": 0.5765,
+      "step": 45588
+    },
+    {
+      "epoch": 1.0851423402837284,
+      "grad_norm": 3.8045154529765495,
+      "learning_rate": 2e-05,
+      "loss": 0.3574,
+      "step": 45589
+    },
+    {
+      "epoch": 1.08516614300676,
+      "grad_norm": 2.010180902705354,
+      "learning_rate": 2e-05,
+      "loss": 0.0191,
+      "step": 45590
+    },
+    {
+      "epoch": 1.0851899457297915,
+      "grad_norm": 8.03516371892763,
+      "learning_rate": 2e-05,
+      "loss": 0.2672,
+      "step": 45591
+    },
+    {
+      "epoch": 1.085213748452823,
+      "grad_norm": 0.09830583843057412,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45592
+    },
+    {
+      "epoch": 1.0852375511758545,
+      "grad_norm": 14.915782332138066,
+      "learning_rate": 2e-05,
+      "loss": 0.4632,
+      "step": 45593
+    },
+    {
+      "epoch": 1.085261353898886,
+      "grad_norm": 3.7521263055353056,
+      "learning_rate": 2e-05,
+      "loss": 0.3866,
+      "step": 45594
+    },
+    {
+      "epoch": 1.0852851566219175,
+      "grad_norm": 15.55424077630381,
+      "learning_rate": 2e-05,
+      "loss": 0.4078,
+      "step": 45595
+    },
+    {
+      "epoch": 1.085308959344949,
+      "grad_norm": 0.028900120432306674,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45596
+    },
+    {
+      "epoch": 1.0853327620679805,
+      "grad_norm": 0.17861285591518763,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 45597
+    },
+    {
+      "epoch": 1.0853565647910122,
+      "grad_norm": 0.21861594272393783,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45598
+    },
+    {
+      "epoch": 1.0853803675140437,
+      "grad_norm": 1.2386153721618889,
+      "learning_rate": 2e-05,
+      "loss": 0.0213,
+      "step": 45599
+    },
+    {
+      "epoch": 1.0854041702370751,
+      "grad_norm": 2.038427906075719,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 45600
+    },
+    {
+      "epoch": 1.0854279729601066,
+      "grad_norm": 1.5373231589375005,
+      "learning_rate": 2e-05,
+      "loss": 0.0165,
+      "step": 45601
+    },
+    {
+      "epoch": 1.085451775683138,
+      "grad_norm": 5.878417718317063,
+      "learning_rate": 2e-05,
+      "loss": 0.7934,
+      "step": 45602
+    },
+    {
+      "epoch": 1.0854755784061696,
+      "grad_norm": 0.04976287495466641,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45603
+    },
+    {
+      "epoch": 1.085499381129201,
+      "grad_norm": 3.717476931024409,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 45604
+    },
+    {
+      "epoch": 1.0855231838522328,
+      "grad_norm": 0.7856607801652513,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 45605
+    },
+    {
+      "epoch": 1.0855469865752643,
+      "grad_norm": 0.010890183768987694,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45606
+    },
+    {
+      "epoch": 1.0855707892982958,
+      "grad_norm": 3.4728348071100354,
+      "learning_rate": 2e-05,
+      "loss": 0.2245,
+      "step": 45607
+    },
+    {
+      "epoch": 1.0855945920213272,
+      "grad_norm": 0.08856020642853281,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45608
+    },
+    {
+      "epoch": 1.0856183947443587,
+      "grad_norm": 0.007423157744791626,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45609
+    },
+    {
+      "epoch": 1.0856421974673902,
+      "grad_norm": 3.6470298160477577,
+      "learning_rate": 2e-05,
+      "loss": 0.4284,
+      "step": 45610
+    },
+    {
+      "epoch": 1.0856660001904217,
+      "grad_norm": 4.657217304083864,
+      "learning_rate": 2e-05,
+      "loss": 0.6505,
+      "step": 45611
+    },
+    {
+      "epoch": 1.0856898029134534,
+      "grad_norm": 4.316594498592697,
+      "learning_rate": 2e-05,
+      "loss": 0.2311,
+      "step": 45612
+    },
+    {
+      "epoch": 1.085713605636485,
+      "grad_norm": 0.3609750298057234,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45613
+    },
+    {
+      "epoch": 1.0857374083595164,
+      "grad_norm": 15.04551077720812,
+      "learning_rate": 2e-05,
+      "loss": 0.6622,
+      "step": 45614
+    },
+    {
+      "epoch": 1.0857612110825479,
+      "grad_norm": 5.085518275392253,
+      "learning_rate": 2e-05,
+      "loss": 0.4782,
+      "step": 45615
+    },
+    {
+      "epoch": 1.0857850138055793,
+      "grad_norm": 1.079557589101616,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 45616
+    },
+    {
+      "epoch": 1.0858088165286108,
+      "grad_norm": 0.0056782353575611485,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45617
+    },
+    {
+      "epoch": 1.0858326192516423,
+      "grad_norm": 4.49943130674547,
+      "learning_rate": 2e-05,
+      "loss": 0.2342,
+      "step": 45618
+    },
+    {
+      "epoch": 1.0858564219746738,
+      "grad_norm": 4.665801809736563,
+      "learning_rate": 2e-05,
+      "loss": 0.5381,
+      "step": 45619
+    },
+    {
+      "epoch": 1.0858802246977055,
+      "grad_norm": 0.02552654555751062,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45620
+    },
+    {
+      "epoch": 1.085904027420737,
+      "grad_norm": 3.648282727900156,
+      "learning_rate": 2e-05,
+      "loss": 0.3766,
+      "step": 45621
+    },
+    {
+      "epoch": 1.0859278301437685,
+      "grad_norm": 1.9391251288250742,
+      "learning_rate": 2e-05,
+      "loss": 0.0219,
+      "step": 45622
+    },
+    {
+      "epoch": 1.0859516328668,
+      "grad_norm": 9.015307043132532,
+      "learning_rate": 2e-05,
+      "loss": 0.1369,
+      "step": 45623
+    },
+    {
+      "epoch": 1.0859754355898315,
+      "grad_norm": 2.0314530800891366,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 45624
+    },
+    {
+      "epoch": 1.085999238312863,
+      "grad_norm": 4.519794662938995,
+      "learning_rate": 2e-05,
+      "loss": 0.6557,
+      "step": 45625
+    },
+    {
+      "epoch": 1.0860230410358944,
+      "grad_norm": 4.086628889721455,
+      "learning_rate": 2e-05,
+      "loss": 0.5003,
+      "step": 45626
+    },
+    {
+      "epoch": 1.086046843758926,
+      "grad_norm": 0.006068308226629143,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45627
+    },
+    {
+      "epoch": 1.0860706464819576,
+      "grad_norm": 1.7896590574249742,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 45628
+    },
+    {
+      "epoch": 1.086094449204989,
+      "grad_norm": 2.0729222962016594,
+      "learning_rate": 2e-05,
+      "loss": 0.0163,
+      "step": 45629
+    },
+    {
+      "epoch": 1.0861182519280206,
+      "grad_norm": 0.6913704745961389,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 45630
+    },
+    {
+      "epoch": 1.086142054651052,
+      "grad_norm": 6.908143142853197,
+      "learning_rate": 2e-05,
+      "loss": 1.0921,
+      "step": 45631
+    },
+    {
+      "epoch": 1.0861658573740836,
+      "grad_norm": 5.6742337057974925,
+      "learning_rate": 2e-05,
+      "loss": 0.4496,
+      "step": 45632
+    },
+    {
+      "epoch": 1.086189660097115,
+      "grad_norm": 7.072759294625225,
+      "learning_rate": 2e-05,
+      "loss": 0.6499,
+      "step": 45633
+    },
+    {
+      "epoch": 1.0862134628201465,
+      "grad_norm": 11.168323916433863,
+      "learning_rate": 2e-05,
+      "loss": 0.5508,
+      "step": 45634
+    },
+    {
+      "epoch": 1.0862372655431782,
+      "grad_norm": 0.5873130962646709,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 45635
+    },
+    {
+      "epoch": 1.0862610682662097,
+      "grad_norm": 3.845553077996961,
+      "learning_rate": 2e-05,
+      "loss": 0.1931,
+      "step": 45636
+    },
+    {
+      "epoch": 1.0862848709892412,
+      "grad_norm": 6.053539712561049,
+      "learning_rate": 2e-05,
+      "loss": 0.1404,
+      "step": 45637
+    },
+    {
+      "epoch": 1.0863086737122727,
+      "grad_norm": 136.93937287124496,
+      "learning_rate": 2e-05,
+      "loss": 0.3341,
+      "step": 45638
+    },
+    {
+      "epoch": 1.0863324764353042,
+      "grad_norm": 5.701498281263092,
+      "learning_rate": 2e-05,
+      "loss": 0.9037,
+      "step": 45639
+    },
+    {
+      "epoch": 1.0863562791583357,
+      "grad_norm": 5.047188623684351,
+      "learning_rate": 2e-05,
+      "loss": 0.0986,
+      "step": 45640
+    },
+    {
+      "epoch": 1.0863800818813671,
+      "grad_norm": 7.107382348091959,
+      "learning_rate": 2e-05,
+      "loss": 0.8572,
+      "step": 45641
+    },
+    {
+      "epoch": 1.0864038846043989,
+      "grad_norm": 0.0077848170154288365,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45642
+    },
+    {
+      "epoch": 1.0864276873274303,
+      "grad_norm": 8.298392937828167,
+      "learning_rate": 2e-05,
+      "loss": 0.6335,
+      "step": 45643
+    },
+    {
+      "epoch": 1.0864514900504618,
+      "grad_norm": 0.08441307872442792,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45644
+    },
+    {
+      "epoch": 1.0864752927734933,
+      "grad_norm": 0.25579083088904603,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 45645
+    },
+    {
+      "epoch": 1.0864990954965248,
+      "grad_norm": 0.006075622636794378,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45646
+    },
+    {
+      "epoch": 1.0865228982195563,
+      "grad_norm": 5.289377503993368,
+      "learning_rate": 2e-05,
+      "loss": 0.7222,
+      "step": 45647
+    },
+    {
+      "epoch": 1.0865467009425878,
+      "grad_norm": 5.27998476656483,
+      "learning_rate": 2e-05,
+      "loss": 0.5544,
+      "step": 45648
+    },
+    {
+      "epoch": 1.0865705036656192,
+      "grad_norm": 0.06531677935295388,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45649
+    },
+    {
+      "epoch": 1.086594306388651,
+      "grad_norm": 0.020551547671376187,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45650
+    },
+    {
+      "epoch": 1.0866181091116824,
+      "grad_norm": 7.661145591591164,
+      "learning_rate": 2e-05,
+      "loss": 0.9944,
+      "step": 45651
+    },
+    {
+      "epoch": 1.086641911834714,
+      "grad_norm": 0.04604001202999421,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45652
+    },
+    {
+      "epoch": 1.0866657145577454,
+      "grad_norm": 5.179814217143766,
+      "learning_rate": 2e-05,
+      "loss": 0.4255,
+      "step": 45653
+    },
+    {
+      "epoch": 1.086689517280777,
+      "grad_norm": 0.17204272782602648,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 45654
+    },
+    {
+      "epoch": 1.0867133200038084,
+      "grad_norm": 3.462753076090668,
+      "learning_rate": 2e-05,
+      "loss": 0.1476,
+      "step": 45655
+    },
+    {
+      "epoch": 1.0867371227268399,
+      "grad_norm": 4.881374325441754,
+      "learning_rate": 2e-05,
+      "loss": 0.5411,
+      "step": 45656
+    },
+    {
+      "epoch": 1.0867609254498714,
+      "grad_norm": 10.253223118097448,
+      "learning_rate": 2e-05,
+      "loss": 1.3962,
+      "step": 45657
+    },
+    {
+      "epoch": 1.086784728172903,
+      "grad_norm": 4.863405074003474,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 45658
+    },
+    {
+      "epoch": 1.0868085308959345,
+      "grad_norm": 3.122618772468214,
+      "learning_rate": 2e-05,
+      "loss": 0.3542,
+      "step": 45659
+    },
+    {
+      "epoch": 1.086832333618966,
+      "grad_norm": 2.0246747211698395,
+      "learning_rate": 2e-05,
+      "loss": 0.0175,
+      "step": 45660
+    },
+    {
+      "epoch": 1.0868561363419975,
+      "grad_norm": 3.4917135044075973,
+      "learning_rate": 2e-05,
+      "loss": 0.3051,
+      "step": 45661
+    },
+    {
+      "epoch": 1.086879939065029,
+      "grad_norm": 5.96912271198668,
+      "learning_rate": 2e-05,
+      "loss": 0.287,
+      "step": 45662
+    },
+    {
+      "epoch": 1.0869037417880605,
+      "grad_norm": 5.927788394112543,
+      "learning_rate": 2e-05,
+      "loss": 0.6816,
+      "step": 45663
+    },
+    {
+      "epoch": 1.086927544511092,
+      "grad_norm": 2.9998892686532987,
+      "learning_rate": 2e-05,
+      "loss": 0.352,
+      "step": 45664
+    },
+    {
+      "epoch": 1.0869513472341237,
+      "grad_norm": 0.06651620654277637,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45665
+    },
+    {
+      "epoch": 1.0869751499571552,
+      "grad_norm": 0.008118929815021023,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45666
+    },
+    {
+      "epoch": 1.0869989526801866,
+      "grad_norm": 3.5588012617168947,
+      "learning_rate": 2e-05,
+      "loss": 0.0469,
+      "step": 45667
+    },
+    {
+      "epoch": 1.0870227554032181,
+      "grad_norm": 0.2675369145750076,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 45668
+    },
+    {
+      "epoch": 1.0870465581262496,
+      "grad_norm": 1.158794177541317,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 45669
+    },
+    {
+      "epoch": 1.087070360849281,
+      "grad_norm": 13.58518908590657,
+      "learning_rate": 2e-05,
+      "loss": 0.7823,
+      "step": 45670
+    },
+    {
+      "epoch": 1.0870941635723126,
+      "grad_norm": 5.330824890475084,
+      "learning_rate": 2e-05,
+      "loss": 0.1035,
+      "step": 45671
+    },
+    {
+      "epoch": 1.0871179662953443,
+      "grad_norm": 21.95120731265745,
+      "learning_rate": 2e-05,
+      "loss": 0.299,
+      "step": 45672
+    },
+    {
+      "epoch": 1.0871417690183758,
+      "grad_norm": 10.216024697885498,
+      "learning_rate": 2e-05,
+      "loss": 0.3612,
+      "step": 45673
+    },
+    {
+      "epoch": 1.0871655717414073,
+      "grad_norm": 2.38040771041876,
+      "learning_rate": 2e-05,
+      "loss": 0.2162,
+      "step": 45674
+    },
+    {
+      "epoch": 1.0871893744644388,
+      "grad_norm": 4.3084060783827365,
+      "learning_rate": 2e-05,
+      "loss": 0.2218,
+      "step": 45675
+    },
+    {
+      "epoch": 1.0872131771874702,
+      "grad_norm": 2.3456389738255257,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 45676
+    },
+    {
+      "epoch": 1.0872369799105017,
+      "grad_norm": 0.6109200497844582,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 45677
+    },
+    {
+      "epoch": 1.0872607826335332,
+      "grad_norm": 6.388556715609621,
+      "learning_rate": 2e-05,
+      "loss": 0.3431,
+      "step": 45678
+    },
+    {
+      "epoch": 1.0872845853565647,
+      "grad_norm": 2.8868792445693288,
+      "learning_rate": 2e-05,
+      "loss": 0.2045,
+      "step": 45679
+    },
+    {
+      "epoch": 1.0873083880795964,
+      "grad_norm": 3.9855520292483617,
+      "learning_rate": 2e-05,
+      "loss": 0.3392,
+      "step": 45680
+    },
+    {
+      "epoch": 1.0873321908026279,
+      "grad_norm": 0.6142696846969612,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 45681
+    },
+    {
+      "epoch": 1.0873559935256594,
+      "grad_norm": 4.069012030037257,
+      "learning_rate": 2e-05,
+      "loss": 0.6036,
+      "step": 45682
+    },
+    {
+      "epoch": 1.0873797962486909,
+      "grad_norm": 4.8574265610894365,
+      "learning_rate": 2e-05,
+      "loss": 0.4632,
+      "step": 45683
+    },
+    {
+      "epoch": 1.0874035989717223,
+      "grad_norm": 0.894123475234112,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 45684
+    },
+    {
+      "epoch": 1.0874274016947538,
+      "grad_norm": 3.2345177212810294,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 45685
+    },
+    {
+      "epoch": 1.0874512044177853,
+      "grad_norm": 4.221809415759925,
+      "learning_rate": 2e-05,
+      "loss": 0.3211,
+      "step": 45686
+    },
+    {
+      "epoch": 1.0874750071408168,
+      "grad_norm": 3.8628953814162945,
+      "learning_rate": 2e-05,
+      "loss": 0.5548,
+      "step": 45687
+    },
+    {
+      "epoch": 1.0874988098638485,
+      "grad_norm": 0.03092621010774867,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45688
+    },
+    {
+      "epoch": 1.08752261258688,
+      "grad_norm": 0.08329675023096521,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45689
+    },
+    {
+      "epoch": 1.0875464153099115,
+      "grad_norm": 2.417594020533126,
+      "learning_rate": 2e-05,
+      "loss": 0.0583,
+      "step": 45690
+    },
+    {
+      "epoch": 1.087570218032943,
+      "grad_norm": 2.0560815380933404,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 45691
+    },
+    {
+      "epoch": 1.0875940207559744,
+      "grad_norm": 3.631800917263243,
+      "learning_rate": 2e-05,
+      "loss": 0.0542,
+      "step": 45692
+    },
+    {
+      "epoch": 1.087617823479006,
+      "grad_norm": 7.797256824686476,
+      "learning_rate": 2e-05,
+      "loss": 0.1895,
+      "step": 45693
+    },
+    {
+      "epoch": 1.0876416262020374,
+      "grad_norm": 4.428567859247222,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 45694
+    },
+    {
+      "epoch": 1.0876654289250691,
+      "grad_norm": 1.0863726456256406,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 45695
+    },
+    {
+      "epoch": 1.0876892316481006,
+      "grad_norm": 0.04841500099400562,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45696
+    },
+    {
+      "epoch": 1.087713034371132,
+      "grad_norm": 3.7122422954946805,
+      "learning_rate": 2e-05,
+      "loss": 0.4639,
+      "step": 45697
+    },
+    {
+      "epoch": 1.0877368370941636,
+      "grad_norm": 1.5583324936370027,
+      "learning_rate": 2e-05,
+      "loss": 0.0923,
+      "step": 45698
+    },
+    {
+      "epoch": 1.087760639817195,
+      "grad_norm": 6.643687474064296,
+      "learning_rate": 2e-05,
+      "loss": 0.1086,
+      "step": 45699
+    },
+    {
+      "epoch": 1.0877844425402265,
+      "grad_norm": 3.046939493037335,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 45700
+    },
+    {
+      "epoch": 1.087808245263258,
+      "grad_norm": 12.18351566820301,
+      "learning_rate": 2e-05,
+      "loss": 0.3964,
+      "step": 45701
+    },
+    {
+      "epoch": 1.0878320479862897,
+      "grad_norm": 0.38160000173941155,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 45702
+    },
+    {
+      "epoch": 1.0878558507093212,
+      "grad_norm": 1.037295713710932,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 45703
+    },
+    {
+      "epoch": 1.0878796534323527,
+      "grad_norm": 8.584618107793132,
+      "learning_rate": 2e-05,
+      "loss": 0.4754,
+      "step": 45704
+    },
+    {
+      "epoch": 1.0879034561553842,
+      "grad_norm": 0.21217230683692687,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 45705
+    },
+    {
+      "epoch": 1.0879272588784157,
+      "grad_norm": 0.7734884954135073,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 45706
+    },
+    {
+      "epoch": 1.0879510616014472,
+      "grad_norm": 0.05967065370437153,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 45707
+    },
+    {
+      "epoch": 1.0879748643244787,
+      "grad_norm": 0.7575065112589873,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 45708
+    },
+    {
+      "epoch": 1.0879986670475101,
+      "grad_norm": 1.3690847733576992,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 45709
+    },
+    {
+      "epoch": 1.0880224697705418,
+      "grad_norm": 5.0442575646486025,
+      "learning_rate": 2e-05,
+      "loss": 0.6619,
+      "step": 45710
+    },
+    {
+      "epoch": 1.0880462724935733,
+      "grad_norm": 0.18251498814390965,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 45711
+    },
+    {
+      "epoch": 1.0880700752166048,
+      "grad_norm": 0.8646535494475663,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 45712
+    },
+    {
+      "epoch": 1.0880938779396363,
+      "grad_norm": 5.837252138934308,
+      "learning_rate": 2e-05,
+      "loss": 0.4116,
+      "step": 45713
+    },
+    {
+      "epoch": 1.0881176806626678,
+      "grad_norm": 0.4830199602701925,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 45714
+    },
+    {
+      "epoch": 1.0881414833856993,
+      "grad_norm": 2.7518384671875435,
+      "learning_rate": 2e-05,
+      "loss": 0.027,
+      "step": 45715
+    },
+    {
+      "epoch": 1.0881652861087308,
+      "grad_norm": 0.2036171102549689,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 45716
+    },
+    {
+      "epoch": 1.0881890888317622,
+      "grad_norm": 0.007021087459023614,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45717
+    },
+    {
+      "epoch": 1.088212891554794,
+      "grad_norm": 4.877366767531838,
+      "learning_rate": 2e-05,
+      "loss": 0.6283,
+      "step": 45718
+    },
+    {
+      "epoch": 1.0882366942778254,
+      "grad_norm": 2.7158649426036363,
+      "learning_rate": 2e-05,
+      "loss": 0.0293,
+      "step": 45719
+    },
+    {
+      "epoch": 1.088260497000857,
+      "grad_norm": 4.991415047030838,
+      "learning_rate": 2e-05,
+      "loss": 0.6463,
+      "step": 45720
+    },
+    {
+      "epoch": 1.0882842997238884,
+      "grad_norm": 5.932293397625218,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 45721
+    },
+    {
+      "epoch": 1.08830810244692,
+      "grad_norm": 0.3833300600068694,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 45722
+    },
+    {
+      "epoch": 1.0883319051699514,
+      "grad_norm": 8.512995582862372,
+      "learning_rate": 2e-05,
+      "loss": 0.9501,
+      "step": 45723
+    },
+    {
+      "epoch": 1.0883557078929829,
+      "grad_norm": 5.076098174493568,
+      "learning_rate": 2e-05,
+      "loss": 1.1312,
+      "step": 45724
+    },
+    {
+      "epoch": 1.0883795106160146,
+      "grad_norm": 17.474412568905237,
+      "learning_rate": 2e-05,
+      "loss": 0.4266,
+      "step": 45725
+    },
+    {
+      "epoch": 1.088403313339046,
+      "grad_norm": 5.45767353802713,
+      "learning_rate": 2e-05,
+      "loss": 0.4351,
+      "step": 45726
+    },
+    {
+      "epoch": 1.0884271160620775,
+      "grad_norm": 6.098935494384228,
+      "learning_rate": 2e-05,
+      "loss": 0.3139,
+      "step": 45727
+    },
+    {
+      "epoch": 1.088450918785109,
+      "grad_norm": 7.25651012666082,
+      "learning_rate": 2e-05,
+      "loss": 0.049,
+      "step": 45728
+    },
+    {
+      "epoch": 1.0884747215081405,
+      "grad_norm": 2.965493352145813,
+      "learning_rate": 2e-05,
+      "loss": 0.3954,
+      "step": 45729
+    },
+    {
+      "epoch": 1.088498524231172,
+      "grad_norm": 0.09470544843452315,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45730
+    },
+    {
+      "epoch": 1.0885223269542035,
+      "grad_norm": 5.621447587888111,
+      "learning_rate": 2e-05,
+      "loss": 0.0418,
+      "step": 45731
+    },
+    {
+      "epoch": 1.0885461296772352,
+      "grad_norm": 6.42767149242046,
+      "learning_rate": 2e-05,
+      "loss": 0.9812,
+      "step": 45732
+    },
+    {
+      "epoch": 1.0885699324002667,
+      "grad_norm": 3.0329987581115327,
+      "learning_rate": 2e-05,
+      "loss": 0.3043,
+      "step": 45733
+    },
+    {
+      "epoch": 1.0885937351232982,
+      "grad_norm": 9.979734136952072,
+      "learning_rate": 2e-05,
+      "loss": 0.643,
+      "step": 45734
+    },
+    {
+      "epoch": 1.0886175378463296,
+      "grad_norm": 10.107887696644562,
+      "learning_rate": 2e-05,
+      "loss": 0.6376,
+      "step": 45735
+    },
+    {
+      "epoch": 1.0886413405693611,
+      "grad_norm": 10.05363844729806,
+      "learning_rate": 2e-05,
+      "loss": 0.7237,
+      "step": 45736
+    },
+    {
+      "epoch": 1.0886651432923926,
+      "grad_norm": 2.7162641922922623,
+      "learning_rate": 2e-05,
+      "loss": 0.2683,
+      "step": 45737
+    },
+    {
+      "epoch": 1.088688946015424,
+      "grad_norm": 1.1250737501832808,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 45738
+    },
+    {
+      "epoch": 1.0887127487384556,
+      "grad_norm": 4.590564230646295,
+      "learning_rate": 2e-05,
+      "loss": 0.3424,
+      "step": 45739
+    },
+    {
+      "epoch": 1.0887365514614873,
+      "grad_norm": 1.0035047922308498,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 45740
+    },
+    {
+      "epoch": 1.0887603541845188,
+      "grad_norm": 0.04634814781420058,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45741
+    },
+    {
+      "epoch": 1.0887841569075503,
+      "grad_norm": 1.5737416622152653,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 45742
+    },
+    {
+      "epoch": 1.0888079596305817,
+      "grad_norm": 2.620447100680095,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 45743
+    },
+    {
+      "epoch": 1.0888317623536132,
+      "grad_norm": 1.3145994095628828,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 45744
+    },
+    {
+      "epoch": 1.0888555650766447,
+      "grad_norm": 3.5084996581615795,
+      "learning_rate": 2e-05,
+      "loss": 0.3108,
+      "step": 45745
+    },
+    {
+      "epoch": 1.0888793677996762,
+      "grad_norm": 3.15589678498284,
+      "learning_rate": 2e-05,
+      "loss": 0.0407,
+      "step": 45746
+    },
+    {
+      "epoch": 1.0889031705227077,
+      "grad_norm": 0.1525605762123196,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45747
+    },
+    {
+      "epoch": 1.0889269732457394,
+      "grad_norm": 0.0037057603228961883,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45748
+    },
+    {
+      "epoch": 1.0889507759687709,
+      "grad_norm": 8.510186440643896,
+      "learning_rate": 2e-05,
+      "loss": 0.5987,
+      "step": 45749
+    },
+    {
+      "epoch": 1.0889745786918024,
+      "grad_norm": 2.8272578851765795,
+      "learning_rate": 2e-05,
+      "loss": 0.3379,
+      "step": 45750
+    },
+    {
+      "epoch": 1.0889983814148338,
+      "grad_norm": 3.6446445332468524,
+      "learning_rate": 2e-05,
+      "loss": 0.6127,
+      "step": 45751
+    },
+    {
+      "epoch": 1.0890221841378653,
+      "grad_norm": 0.9303779675169233,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 45752
+    },
+    {
+      "epoch": 1.0890459868608968,
+      "grad_norm": 2.5387178844426996,
+      "learning_rate": 2e-05,
+      "loss": 0.2619,
+      "step": 45753
+    },
+    {
+      "epoch": 1.0890697895839283,
+      "grad_norm": 0.04438363055655211,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45754
+    },
+    {
+      "epoch": 1.08909359230696,
+      "grad_norm": 6.138310913835229,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 45755
+    },
+    {
+      "epoch": 1.0891173950299915,
+      "grad_norm": 3.715094370585549,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 45756
+    },
+    {
+      "epoch": 1.089141197753023,
+      "grad_norm": 0.04594503699408883,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45757
+    },
+    {
+      "epoch": 1.0891650004760545,
+      "grad_norm": 3.0244462186140635,
+      "learning_rate": 2e-05,
+      "loss": 0.3275,
+      "step": 45758
+    },
+    {
+      "epoch": 1.089188803199086,
+      "grad_norm": 4.189353679262816,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 45759
+    },
+    {
+      "epoch": 1.0892126059221174,
+      "grad_norm": 9.195548823892377,
+      "learning_rate": 2e-05,
+      "loss": 0.7479,
+      "step": 45760
+    },
+    {
+      "epoch": 1.089236408645149,
+      "grad_norm": 3.797662449723247,
+      "learning_rate": 2e-05,
+      "loss": 0.3982,
+      "step": 45761
+    },
+    {
+      "epoch": 1.0892602113681806,
+      "grad_norm": 10.640178345487648,
+      "learning_rate": 2e-05,
+      "loss": 0.6261,
+      "step": 45762
+    },
+    {
+      "epoch": 1.0892840140912121,
+      "grad_norm": 5.041240646816903,
+      "learning_rate": 2e-05,
+      "loss": 0.9471,
+      "step": 45763
+    },
+    {
+      "epoch": 1.0893078168142436,
+      "grad_norm": 16.617472471272112,
+      "learning_rate": 2e-05,
+      "loss": 0.1348,
+      "step": 45764
+    },
+    {
+      "epoch": 1.089331619537275,
+      "grad_norm": 7.901910398756744,
+      "learning_rate": 2e-05,
+      "loss": 0.7333,
+      "step": 45765
+    },
+    {
+      "epoch": 1.0893554222603066,
+      "grad_norm": 3.6426014106880733,
+      "learning_rate": 2e-05,
+      "loss": 0.3749,
+      "step": 45766
+    },
+    {
+      "epoch": 1.089379224983338,
+      "grad_norm": 4.718884453666236,
+      "learning_rate": 2e-05,
+      "loss": 0.1839,
+      "step": 45767
+    },
+    {
+      "epoch": 1.0894030277063695,
+      "grad_norm": 4.433514929747164,
+      "learning_rate": 2e-05,
+      "loss": 0.1059,
+      "step": 45768
+    },
+    {
+      "epoch": 1.089426830429401,
+      "grad_norm": 1.3177613819396996,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 45769
+    },
+    {
+      "epoch": 1.0894506331524327,
+      "grad_norm": 3.7326268412946373,
+      "learning_rate": 2e-05,
+      "loss": 0.4413,
+      "step": 45770
+    },
+    {
+      "epoch": 1.0894744358754642,
+      "grad_norm": 5.637838307049691,
+      "learning_rate": 2e-05,
+      "loss": 0.4646,
+      "step": 45771
+    },
+    {
+      "epoch": 1.0894982385984957,
+      "grad_norm": 0.8229039200434934,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 45772
+    },
+    {
+      "epoch": 1.0895220413215272,
+      "grad_norm": 2.508353059299344,
+      "learning_rate": 2e-05,
+      "loss": 0.0323,
+      "step": 45773
+    },
+    {
+      "epoch": 1.0895458440445587,
+      "grad_norm": 0.11156980511068385,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 45774
+    },
+    {
+      "epoch": 1.0895696467675902,
+      "grad_norm": 1.135570327725924,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 45775
+    },
+    {
+      "epoch": 1.0895934494906216,
+      "grad_norm": 4.027040777589261,
+      "learning_rate": 2e-05,
+      "loss": 0.0414,
+      "step": 45776
+    },
+    {
+      "epoch": 1.0896172522136531,
+      "grad_norm": 0.011802868317798774,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45777
+    },
+    {
+      "epoch": 1.0896410549366848,
+      "grad_norm": 2.8692946047166266,
+      "learning_rate": 2e-05,
+      "loss": 0.1957,
+      "step": 45778
+    },
+    {
+      "epoch": 1.0896648576597163,
+      "grad_norm": 3.376186322314594,
+      "learning_rate": 2e-05,
+      "loss": 0.1071,
+      "step": 45779
+    },
+    {
+      "epoch": 1.0896886603827478,
+      "grad_norm": 3.771936208485078,
+      "learning_rate": 2e-05,
+      "loss": 0.5025,
+      "step": 45780
+    },
+    {
+      "epoch": 1.0897124631057793,
+      "grad_norm": 4.031035764608825,
+      "learning_rate": 2e-05,
+      "loss": 0.7366,
+      "step": 45781
+    },
+    {
+      "epoch": 1.0897362658288108,
+      "grad_norm": 0.4158434520482896,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 45782
+    },
+    {
+      "epoch": 1.0897600685518423,
+      "grad_norm": 0.6160431646641877,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 45783
+    },
+    {
+      "epoch": 1.0897838712748737,
+      "grad_norm": 3.2767788346414584,
+      "learning_rate": 2e-05,
+      "loss": 0.2918,
+      "step": 45784
+    },
+    {
+      "epoch": 1.0898076739979055,
+      "grad_norm": 4.870420644266569,
+      "learning_rate": 2e-05,
+      "loss": 0.1169,
+      "step": 45785
+    },
+    {
+      "epoch": 1.089831476720937,
+      "grad_norm": 2.6397299449811533,
+      "learning_rate": 2e-05,
+      "loss": 0.2362,
+      "step": 45786
+    },
+    {
+      "epoch": 1.0898552794439684,
+      "grad_norm": 2.3372815842233083,
+      "learning_rate": 2e-05,
+      "loss": 0.0674,
+      "step": 45787
+    },
+    {
+      "epoch": 1.089879082167,
+      "grad_norm": 6.213724822972659,
+      "learning_rate": 2e-05,
+      "loss": 0.1666,
+      "step": 45788
+    },
+    {
+      "epoch": 1.0899028848900314,
+      "grad_norm": 2.1843649253835156,
+      "learning_rate": 2e-05,
+      "loss": 0.0617,
+      "step": 45789
+    },
+    {
+      "epoch": 1.0899266876130629,
+      "grad_norm": 6.582829274257963,
+      "learning_rate": 2e-05,
+      "loss": 0.5717,
+      "step": 45790
+    },
+    {
+      "epoch": 1.0899504903360944,
+      "grad_norm": 3.013497492392032,
+      "learning_rate": 2e-05,
+      "loss": 0.3675,
+      "step": 45791
+    },
+    {
+      "epoch": 1.089974293059126,
+      "grad_norm": 3.258819002125361,
+      "learning_rate": 2e-05,
+      "loss": 0.4463,
+      "step": 45792
+    },
+    {
+      "epoch": 1.0899980957821576,
+      "grad_norm": 0.3420467655925828,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 45793
+    },
+    {
+      "epoch": 1.090021898505189,
+      "grad_norm": 6.177083073020423,
+      "learning_rate": 2e-05,
+      "loss": 0.2527,
+      "step": 45794
+    },
+    {
+      "epoch": 1.0900457012282205,
+      "grad_norm": 2.127283366211475,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 45795
+    },
+    {
+      "epoch": 1.090069503951252,
+      "grad_norm": 4.56733061641686,
+      "learning_rate": 2e-05,
+      "loss": 0.7032,
+      "step": 45796
+    },
+    {
+      "epoch": 1.0900933066742835,
+      "grad_norm": 0.6734877217097134,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 45797
+    },
+    {
+      "epoch": 1.090117109397315,
+      "grad_norm": 4.6503688483313415,
+      "learning_rate": 2e-05,
+      "loss": 0.5514,
+      "step": 45798
+    },
+    {
+      "epoch": 1.0901409121203465,
+      "grad_norm": 3.9853128161905076,
+      "learning_rate": 2e-05,
+      "loss": 0.2944,
+      "step": 45799
+    },
+    {
+      "epoch": 1.0901647148433782,
+      "grad_norm": 0.2174194089661663,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45800
+    },
+    {
+      "epoch": 1.0901885175664097,
+      "grad_norm": 63.14070287729562,
+      "learning_rate": 2e-05,
+      "loss": 0.2338,
+      "step": 45801
+    },
+    {
+      "epoch": 1.0902123202894412,
+      "grad_norm": 0.42166046045278277,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 45802
+    },
+    {
+      "epoch": 1.0902361230124726,
+      "grad_norm": 0.25380820976106966,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 45803
+    },
+    {
+      "epoch": 1.0902599257355041,
+      "grad_norm": 0.2897664671213656,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 45804
+    },
+    {
+      "epoch": 1.0902837284585356,
+      "grad_norm": 3.503968248382806,
+      "learning_rate": 2e-05,
+      "loss": 0.321,
+      "step": 45805
+    },
+    {
+      "epoch": 1.090307531181567,
+      "grad_norm": 0.5040806400743251,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 45806
+    },
+    {
+      "epoch": 1.0903313339045986,
+      "grad_norm": 3.7805862247630024,
+      "learning_rate": 2e-05,
+      "loss": 0.5169,
+      "step": 45807
+    },
+    {
+      "epoch": 1.0903551366276303,
+      "grad_norm": 0.06270367646451239,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45808
+    },
+    {
+      "epoch": 1.0903789393506618,
+      "grad_norm": 5.1656486657167715,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 45809
+    },
+    {
+      "epoch": 1.0904027420736933,
+      "grad_norm": 2.22480723321306,
+      "learning_rate": 2e-05,
+      "loss": 0.0339,
+      "step": 45810
+    },
+    {
+      "epoch": 1.0904265447967247,
+      "grad_norm": 0.008077673915962735,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45811
+    },
+    {
+      "epoch": 1.0904503475197562,
+      "grad_norm": 1.7699516181455386,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 45812
+    },
+    {
+      "epoch": 1.0904741502427877,
+      "grad_norm": 7.093303271376752,
+      "learning_rate": 2e-05,
+      "loss": 0.6538,
+      "step": 45813
+    },
+    {
+      "epoch": 1.0904979529658192,
+      "grad_norm": 0.3178690343170994,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 45814
+    },
+    {
+      "epoch": 1.090521755688851,
+      "grad_norm": 0.0329563170419941,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45815
+    },
+    {
+      "epoch": 1.0905455584118824,
+      "grad_norm": 7.486837061032163,
+      "learning_rate": 2e-05,
+      "loss": 0.2079,
+      "step": 45816
+    },
+    {
+      "epoch": 1.0905693611349139,
+      "grad_norm": 11.877899303413162,
+      "learning_rate": 2e-05,
+      "loss": 0.4606,
+      "step": 45817
+    },
+    {
+      "epoch": 1.0905931638579454,
+      "grad_norm": 6.85872496227014,
+      "learning_rate": 2e-05,
+      "loss": 0.5813,
+      "step": 45818
+    },
+    {
+      "epoch": 1.0906169665809768,
+      "grad_norm": 2.2837228218129466,
+      "learning_rate": 2e-05,
+      "loss": 0.2778,
+      "step": 45819
+    },
+    {
+      "epoch": 1.0906407693040083,
+      "grad_norm": 2.222986660530147,
+      "learning_rate": 2e-05,
+      "loss": 0.1515,
+      "step": 45820
+    },
+    {
+      "epoch": 1.0906645720270398,
+      "grad_norm": 3.8721366504698342,
+      "learning_rate": 2e-05,
+      "loss": 0.2887,
+      "step": 45821
+    },
+    {
+      "epoch": 1.0906883747500715,
+      "grad_norm": 7.431777838983326,
+      "learning_rate": 2e-05,
+      "loss": 0.6516,
+      "step": 45822
+    },
+    {
+      "epoch": 1.090712177473103,
+      "grad_norm": 6.467487016762213,
+      "learning_rate": 2e-05,
+      "loss": 0.8334,
+      "step": 45823
+    },
+    {
+      "epoch": 1.0907359801961345,
+      "grad_norm": 1.6444234232790425,
+      "learning_rate": 2e-05,
+      "loss": 0.0308,
+      "step": 45824
+    },
+    {
+      "epoch": 1.090759782919166,
+      "grad_norm": 0.3714350546375775,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 45825
+    },
+    {
+      "epoch": 1.0907835856421975,
+      "grad_norm": 4.766649958033223,
+      "learning_rate": 2e-05,
+      "loss": 0.523,
+      "step": 45826
+    },
+    {
+      "epoch": 1.090807388365229,
+      "grad_norm": 8.190744294041703,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 45827
+    },
+    {
+      "epoch": 1.0908311910882604,
+      "grad_norm": 0.7718473207076713,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 45828
+    },
+    {
+      "epoch": 1.090854993811292,
+      "grad_norm": 0.22107826272240372,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 45829
+    },
+    {
+      "epoch": 1.0908787965343236,
+      "grad_norm": 4.314171992463969,
+      "learning_rate": 2e-05,
+      "loss": 0.3607,
+      "step": 45830
+    },
+    {
+      "epoch": 1.090902599257355,
+      "grad_norm": 0.22070609926148513,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45831
+    },
+    {
+      "epoch": 1.0909264019803866,
+      "grad_norm": 0.049395412876961996,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45832
+    },
+    {
+      "epoch": 1.090950204703418,
+      "grad_norm": 2.157193294398395,
+      "learning_rate": 2e-05,
+      "loss": 0.1899,
+      "step": 45833
+    },
+    {
+      "epoch": 1.0909740074264496,
+      "grad_norm": 3.8128914537262486,
+      "learning_rate": 2e-05,
+      "loss": 0.1517,
+      "step": 45834
+    },
+    {
+      "epoch": 1.090997810149481,
+      "grad_norm": 5.829396240018917,
+      "learning_rate": 2e-05,
+      "loss": 0.4309,
+      "step": 45835
+    },
+    {
+      "epoch": 1.0910216128725125,
+      "grad_norm": 8.722761411650598,
+      "learning_rate": 2e-05,
+      "loss": 0.668,
+      "step": 45836
+    },
+    {
+      "epoch": 1.091045415595544,
+      "grad_norm": 2.9764923096749776,
+      "learning_rate": 2e-05,
+      "loss": 0.3203,
+      "step": 45837
+    },
+    {
+      "epoch": 1.0910692183185757,
+      "grad_norm": 2.746659669472947,
+      "learning_rate": 2e-05,
+      "loss": 0.4202,
+      "step": 45838
+    },
+    {
+      "epoch": 1.0910930210416072,
+      "grad_norm": 0.0130823529989934,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45839
+    },
+    {
+      "epoch": 1.0911168237646387,
+      "grad_norm": 7.555188321075523,
+      "learning_rate": 2e-05,
+      "loss": 0.4311,
+      "step": 45840
+    },
+    {
+      "epoch": 1.0911406264876702,
+      "grad_norm": 5.0719250664724695,
+      "learning_rate": 2e-05,
+      "loss": 0.5594,
+      "step": 45841
+    },
+    {
+      "epoch": 1.0911644292107017,
+      "grad_norm": 4.521837050213722,
+      "learning_rate": 2e-05,
+      "loss": 0.1428,
+      "step": 45842
+    },
+    {
+      "epoch": 1.0911882319337332,
+      "grad_norm": 8.079069432091531,
+      "learning_rate": 2e-05,
+      "loss": 0.4427,
+      "step": 45843
+    },
+    {
+      "epoch": 1.0912120346567646,
+      "grad_norm": 3.722529958550539,
+      "learning_rate": 2e-05,
+      "loss": 0.3469,
+      "step": 45844
+    },
+    {
+      "epoch": 1.0912358373797963,
+      "grad_norm": 6.872681708495698,
+      "learning_rate": 2e-05,
+      "loss": 0.577,
+      "step": 45845
+    },
+    {
+      "epoch": 1.0912596401028278,
+      "grad_norm": 0.036621013120505204,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45846
+    },
+    {
+      "epoch": 1.0912834428258593,
+      "grad_norm": 0.08158282191165124,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45847
+    },
+    {
+      "epoch": 1.0913072455488908,
+      "grad_norm": 5.1593750340315765,
+      "learning_rate": 2e-05,
+      "loss": 0.4416,
+      "step": 45848
+    },
+    {
+      "epoch": 1.0913310482719223,
+      "grad_norm": 5.402416889513455,
+      "learning_rate": 2e-05,
+      "loss": 0.0535,
+      "step": 45849
+    },
+    {
+      "epoch": 1.0913548509949538,
+      "grad_norm": 0.022194324172212444,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45850
+    },
+    {
+      "epoch": 1.0913786537179853,
+      "grad_norm": 0.0380192282936877,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45851
+    },
+    {
+      "epoch": 1.091402456441017,
+      "grad_norm": 0.019983993753496812,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45852
+    },
+    {
+      "epoch": 1.0914262591640485,
+      "grad_norm": 7.154199078751954,
+      "learning_rate": 2e-05,
+      "loss": 0.2136,
+      "step": 45853
+    },
+    {
+      "epoch": 1.09145006188708,
+      "grad_norm": 0.09914770230149059,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45854
+    },
+    {
+      "epoch": 1.0914738646101114,
+      "grad_norm": 13.562899973937862,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 45855
+    },
+    {
+      "epoch": 1.091497667333143,
+      "grad_norm": 7.660855446768535,
+      "learning_rate": 2e-05,
+      "loss": 0.4043,
+      "step": 45856
+    },
+    {
+      "epoch": 1.0915214700561744,
+      "grad_norm": 0.08857005524778618,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 45857
+    },
+    {
+      "epoch": 1.0915452727792059,
+      "grad_norm": 6.207299986542186,
+      "learning_rate": 2e-05,
+      "loss": 0.3512,
+      "step": 45858
+    },
+    {
+      "epoch": 1.0915690755022374,
+      "grad_norm": 3.7029520620704797,
+      "learning_rate": 2e-05,
+      "loss": 0.5226,
+      "step": 45859
+    },
+    {
+      "epoch": 1.091592878225269,
+      "grad_norm": 10.326389787933149,
+      "learning_rate": 2e-05,
+      "loss": 1.0287,
+      "step": 45860
+    },
+    {
+      "epoch": 1.0916166809483006,
+      "grad_norm": 0.1039051899625078,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45861
+    },
+    {
+      "epoch": 1.091640483671332,
+      "grad_norm": 3.849292661966942,
+      "learning_rate": 2e-05,
+      "loss": 0.4671,
+      "step": 45862
+    },
+    {
+      "epoch": 1.0916642863943635,
+      "grad_norm": 0.4484108797008655,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 45863
+    },
+    {
+      "epoch": 1.091688089117395,
+      "grad_norm": 4.287190348920875,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 45864
+    },
+    {
+      "epoch": 1.0917118918404265,
+      "grad_norm": 4.127654018400155,
+      "learning_rate": 2e-05,
+      "loss": 0.5052,
+      "step": 45865
+    },
+    {
+      "epoch": 1.091735694563458,
+      "grad_norm": 4.195977127422452,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 45866
+    },
+    {
+      "epoch": 1.0917594972864895,
+      "grad_norm": 15.07311192815503,
+      "learning_rate": 2e-05,
+      "loss": 0.688,
+      "step": 45867
+    },
+    {
+      "epoch": 1.0917833000095212,
+      "grad_norm": 4.452447423484981,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 45868
+    },
+    {
+      "epoch": 1.0918071027325527,
+      "grad_norm": 4.0901378825637,
+      "learning_rate": 2e-05,
+      "loss": 0.4697,
+      "step": 45869
+    },
+    {
+      "epoch": 1.0918309054555841,
+      "grad_norm": 5.296108460065425,
+      "learning_rate": 2e-05,
+      "loss": 0.2319,
+      "step": 45870
+    },
+    {
+      "epoch": 1.0918547081786156,
+      "grad_norm": 2.2072343391223366,
+      "learning_rate": 2e-05,
+      "loss": 0.0467,
+      "step": 45871
+    },
+    {
+      "epoch": 1.0918785109016471,
+      "grad_norm": 3.6127303176809753,
+      "learning_rate": 2e-05,
+      "loss": 0.1984,
+      "step": 45872
+    },
+    {
+      "epoch": 1.0919023136246786,
+      "grad_norm": 5.601989578705549,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 45873
+    },
+    {
+      "epoch": 1.09192611634771,
+      "grad_norm": 17.539413170405286,
+      "learning_rate": 2e-05,
+      "loss": 0.8618,
+      "step": 45874
+    },
+    {
+      "epoch": 1.0919499190707418,
+      "grad_norm": 1.877071071478199,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 45875
+    },
+    {
+      "epoch": 1.0919737217937733,
+      "grad_norm": 0.013018336537989252,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45876
+    },
+    {
+      "epoch": 1.0919975245168048,
+      "grad_norm": 2.14321224245556,
+      "learning_rate": 2e-05,
+      "loss": 0.0411,
+      "step": 45877
+    },
+    {
+      "epoch": 1.0920213272398362,
+      "grad_norm": 0.33793655849512205,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 45878
+    },
+    {
+      "epoch": 1.0920451299628677,
+      "grad_norm": 11.919229109492141,
+      "learning_rate": 2e-05,
+      "loss": 0.3371,
+      "step": 45879
+    },
+    {
+      "epoch": 1.0920689326858992,
+      "grad_norm": 4.227953654199508,
+      "learning_rate": 2e-05,
+      "loss": 0.0687,
+      "step": 45880
+    },
+    {
+      "epoch": 1.0920927354089307,
+      "grad_norm": 13.201605403365502,
+      "learning_rate": 2e-05,
+      "loss": 0.2157,
+      "step": 45881
+    },
+    {
+      "epoch": 1.0921165381319624,
+      "grad_norm": 11.72911053418689,
+      "learning_rate": 2e-05,
+      "loss": 0.5091,
+      "step": 45882
+    },
+    {
+      "epoch": 1.092140340854994,
+      "grad_norm": 6.744443207001852,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 45883
+    },
+    {
+      "epoch": 1.0921641435780254,
+      "grad_norm": 0.04758753775816107,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 45884
+    },
+    {
+      "epoch": 1.0921879463010569,
+      "grad_norm": 0.09408293002600943,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45885
+    },
+    {
+      "epoch": 1.0922117490240884,
+      "grad_norm": 6.122828241494517,
+      "learning_rate": 2e-05,
+      "loss": 0.1394,
+      "step": 45886
+    },
+    {
+      "epoch": 1.0922355517471198,
+      "grad_norm": 0.006938292684600101,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 45887
+    },
+    {
+      "epoch": 1.0922593544701513,
+      "grad_norm": 6.220715842542087,
+      "learning_rate": 2e-05,
+      "loss": 0.6953,
+      "step": 45888
+    },
+    {
+      "epoch": 1.0922831571931828,
+      "grad_norm": 0.015648272910098317,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45889
+    },
+    {
+      "epoch": 1.0923069599162145,
+      "grad_norm": 0.227809446851059,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45890
+    },
+    {
+      "epoch": 1.092330762639246,
+      "grad_norm": 0.2224740557583411,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 45891
+    },
+    {
+      "epoch": 1.0923545653622775,
+      "grad_norm": 5.297942557291056,
+      "learning_rate": 2e-05,
+      "loss": 0.8223,
+      "step": 45892
+    },
+    {
+      "epoch": 1.092378368085309,
+      "grad_norm": 7.686492077151185,
+      "learning_rate": 2e-05,
+      "loss": 0.0611,
+      "step": 45893
+    },
+    {
+      "epoch": 1.0924021708083405,
+      "grad_norm": 0.09948870682625907,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45894
+    },
+    {
+      "epoch": 1.092425973531372,
+      "grad_norm": 0.04493945161013276,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45895
+    },
+    {
+      "epoch": 1.0924497762544034,
+      "grad_norm": 5.8623534662821095,
+      "learning_rate": 2e-05,
+      "loss": 0.6152,
+      "step": 45896
+    },
+    {
+      "epoch": 1.092473578977435,
+      "grad_norm": 0.8944760633109452,
+      "learning_rate": 2e-05,
+      "loss": 0.009,
+      "step": 45897
+    },
+    {
+      "epoch": 1.0924973817004666,
+      "grad_norm": 0.20278221243140274,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 45898
+    },
+    {
+      "epoch": 1.092521184423498,
+      "grad_norm": 2.870216500504445,
+      "learning_rate": 2e-05,
+      "loss": 0.0658,
+      "step": 45899
+    },
+    {
+      "epoch": 1.0925449871465296,
+      "grad_norm": 5.469632211205731,
+      "learning_rate": 2e-05,
+      "loss": 0.1715,
+      "step": 45900
+    },
+    {
+      "epoch": 1.092568789869561,
+      "grad_norm": 0.826487196222409,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 45901
+    },
+    {
+      "epoch": 1.0925925925925926,
+      "grad_norm": 12.310096463640974,
+      "learning_rate": 2e-05,
+      "loss": 0.9804,
+      "step": 45902
+    },
+    {
+      "epoch": 1.092616395315624,
+      "grad_norm": 2.311226702037506,
+      "learning_rate": 2e-05,
+      "loss": 0.2674,
+      "step": 45903
+    },
+    {
+      "epoch": 1.0926401980386555,
+      "grad_norm": 10.67359350952305,
+      "learning_rate": 2e-05,
+      "loss": 1.2464,
+      "step": 45904
+    },
+    {
+      "epoch": 1.0926640007616872,
+      "grad_norm": 1.0410816488271641,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 45905
+    },
+    {
+      "epoch": 1.0926878034847187,
+      "grad_norm": 5.639764078881702,
+      "learning_rate": 2e-05,
+      "loss": 0.7206,
+      "step": 45906
+    },
+    {
+      "epoch": 1.0927116062077502,
+      "grad_norm": 0.2536051352032084,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 45907
+    },
+    {
+      "epoch": 1.0927354089307817,
+      "grad_norm": 4.585525347880753,
+      "learning_rate": 2e-05,
+      "loss": 0.1322,
+      "step": 45908
+    },
+    {
+      "epoch": 1.0927592116538132,
+      "grad_norm": 3.492538836975745,
+      "learning_rate": 2e-05,
+      "loss": 0.1565,
+      "step": 45909
+    },
+    {
+      "epoch": 1.0927830143768447,
+      "grad_norm": 3.569780906705907,
+      "learning_rate": 2e-05,
+      "loss": 0.34,
+      "step": 45910
+    },
+    {
+      "epoch": 1.0928068170998761,
+      "grad_norm": 1.5955851085213226,
+      "learning_rate": 2e-05,
+      "loss": 0.0209,
+      "step": 45911
+    },
+    {
+      "epoch": 1.0928306198229079,
+      "grad_norm": 0.05456891095212601,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45912
+    },
+    {
+      "epoch": 1.0928544225459393,
+      "grad_norm": 0.15844616259676803,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45913
+    },
+    {
+      "epoch": 1.0928782252689708,
+      "grad_norm": 9.967955912531831,
+      "learning_rate": 2e-05,
+      "loss": 0.3468,
+      "step": 45914
+    },
+    {
+      "epoch": 1.0929020279920023,
+      "grad_norm": 3.962125298374596,
+      "learning_rate": 2e-05,
+      "loss": 0.4268,
+      "step": 45915
+    },
+    {
+      "epoch": 1.0929258307150338,
+      "grad_norm": 6.9589578933794165,
+      "learning_rate": 2e-05,
+      "loss": 0.776,
+      "step": 45916
+    },
+    {
+      "epoch": 1.0929496334380653,
+      "grad_norm": 4.860245197256223,
+      "learning_rate": 2e-05,
+      "loss": 0.6957,
+      "step": 45917
+    },
+    {
+      "epoch": 1.0929734361610968,
+      "grad_norm": 4.129718694978383,
+      "learning_rate": 2e-05,
+      "loss": 0.3391,
+      "step": 45918
+    },
+    {
+      "epoch": 1.0929972388841283,
+      "grad_norm": 0.03248386293047207,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 45919
+    },
+    {
+      "epoch": 1.09302104160716,
+      "grad_norm": 5.59728034398614,
+      "learning_rate": 2e-05,
+      "loss": 1.0013,
+      "step": 45920
+    },
+    {
+      "epoch": 1.0930448443301914,
+      "grad_norm": 9.974352587705383,
+      "learning_rate": 2e-05,
+      "loss": 0.1388,
+      "step": 45921
+    },
+    {
+      "epoch": 1.093068647053223,
+      "grad_norm": 6.596548764424409,
+      "learning_rate": 2e-05,
+      "loss": 0.7027,
+      "step": 45922
+    },
+    {
+      "epoch": 1.0930924497762544,
+      "grad_norm": 0.4583975005570088,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 45923
+    },
+    {
+      "epoch": 1.093116252499286,
+      "grad_norm": 1.76763766896143,
+      "learning_rate": 2e-05,
+      "loss": 0.0328,
+      "step": 45924
+    },
+    {
+      "epoch": 1.0931400552223174,
+      "grad_norm": 2.2764729232846985,
+      "learning_rate": 2e-05,
+      "loss": 0.2115,
+      "step": 45925
+    },
+    {
+      "epoch": 1.0931638579453489,
+      "grad_norm": 14.725356984069197,
+      "learning_rate": 2e-05,
+      "loss": 0.3381,
+      "step": 45926
+    },
+    {
+      "epoch": 1.0931876606683804,
+      "grad_norm": 0.02490634831826723,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45927
+    },
+    {
+      "epoch": 1.093211463391412,
+      "grad_norm": 12.807199454520855,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 45928
+    },
+    {
+      "epoch": 1.0932352661144435,
+      "grad_norm": 3.656748774313683,
+      "learning_rate": 2e-05,
+      "loss": 0.4124,
+      "step": 45929
+    },
+    {
+      "epoch": 1.093259068837475,
+      "grad_norm": 0.5167895790107117,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 45930
+    },
+    {
+      "epoch": 1.0932828715605065,
+      "grad_norm": 0.13332573801269265,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 45931
+    },
+    {
+      "epoch": 1.093306674283538,
+      "grad_norm": 0.015509004646362465,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 45932
+    },
+    {
+      "epoch": 1.0933304770065695,
+      "grad_norm": 0.011539429177246844,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 45933
+    },
+    {
+      "epoch": 1.093354279729601,
+      "grad_norm": 3.885868170508814,
+      "learning_rate": 2e-05,
+      "loss": 0.2984,
+      "step": 45934
+    },
+    {
+      "epoch": 1.0933780824526327,
+      "grad_norm": 4.426167528311326,
+      "learning_rate": 2e-05,
+      "loss": 0.0312,
+      "step": 45935
+    },
+    {
+      "epoch": 1.0934018851756642,
+      "grad_norm": 10.200836325012201,
+      "learning_rate": 2e-05,
+      "loss": 0.0652,
+      "step": 45936
+    },
+    {
+      "epoch": 1.0934256878986957,
+      "grad_norm": 4.959029635085296,
+      "learning_rate": 2e-05,
+      "loss": 0.0592,
+      "step": 45937
+    },
+    {
+      "epoch": 1.0934494906217271,
+      "grad_norm": 11.37625709794269,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 45938
+    },
+    {
+      "epoch": 1.0934732933447586,
+      "grad_norm": 4.382660463891192,
+      "learning_rate": 2e-05,
+      "loss": 0.4736,
+      "step": 45939
+    },
+    {
+      "epoch": 1.09349709606779,
+      "grad_norm": 0.13778996430570728,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 45940
+    },
+    {
+      "epoch": 1.0935208987908216,
+      "grad_norm": 3.8753729136373867,
+      "learning_rate": 2e-05,
+      "loss": 0.3572,
+      "step": 45941
+    },
+    {
+      "epoch": 1.0935447015138533,
+      "grad_norm": 3.3508635553489277,
+      "learning_rate": 2e-05,
+      "loss": 0.0344,
+      "step": 45942
+    },
+    {
+      "epoch": 1.0935685042368848,
+      "grad_norm": 0.07021826161345485,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 45943
+    },
+    {
+      "epoch": 1.0935923069599163,
+      "grad_norm": 0.9963689037744807,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 45944
+    },
+    {
+      "epoch": 1.0936161096829478,
+      "grad_norm": 0.11460145741315018,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45945
+    },
+    {
+      "epoch": 1.0936399124059792,
+      "grad_norm": 2.874691029785093,
+      "learning_rate": 2e-05,
+      "loss": 0.2758,
+      "step": 45946
+    },
+    {
+      "epoch": 1.0936637151290107,
+      "grad_norm": 1.350633583519813,
+      "learning_rate": 2e-05,
+      "loss": 0.0196,
+      "step": 45947
+    },
+    {
+      "epoch": 1.0936875178520422,
+      "grad_norm": 74.50817689903657,
+      "learning_rate": 2e-05,
+      "loss": 0.2368,
+      "step": 45948
+    },
+    {
+      "epoch": 1.0937113205750737,
+      "grad_norm": 4.32207931002195,
+      "learning_rate": 2e-05,
+      "loss": 0.4514,
+      "step": 45949
+    },
+    {
+      "epoch": 1.0937351232981054,
+      "grad_norm": 0.8723234464322557,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 45950
+    },
+    {
+      "epoch": 1.093758926021137,
+      "grad_norm": 5.043890505697215,
+      "learning_rate": 2e-05,
+      "loss": 0.1386,
+      "step": 45951
+    },
+    {
+      "epoch": 1.0937827287441684,
+      "grad_norm": 0.08450414643911845,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 45952
+    },
+    {
+      "epoch": 1.0938065314671999,
+      "grad_norm": 1.398557306293118,
+      "learning_rate": 2e-05,
+      "loss": 0.0177,
+      "step": 45953
+    },
+    {
+      "epoch": 1.0938303341902313,
+      "grad_norm": 13.652227234531088,
+      "learning_rate": 2e-05,
+      "loss": 1.0517,
+      "step": 45954
+    },
+    {
+      "epoch": 1.0938541369132628,
+      "grad_norm": 1.5640603874745578,
+      "learning_rate": 2e-05,
+      "loss": 0.0122,
+      "step": 45955
+    },
+    {
+      "epoch": 1.0938779396362943,
+      "grad_norm": 1.8452526210447302,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 45956
+    },
+    {
+      "epoch": 1.0939017423593258,
+      "grad_norm": 0.14778968569758075,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 45957
+    },
+    {
+      "epoch": 1.0939255450823575,
+      "grad_norm": 2.045820096909296,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 45958
+    },
+    {
+      "epoch": 1.093949347805389,
+      "grad_norm": 0.019397492387612153,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 45959
+    },
+    {
+      "epoch": 1.0939731505284205,
+      "grad_norm": 3.4713746217564005,
+      "learning_rate": 2e-05,
+      "loss": 0.0507,
+      "step": 45960
+    },
+    {
+      "epoch": 1.093996953251452,
+      "grad_norm": 5.432304940497878,
+      "learning_rate": 2e-05,
+      "loss": 0.8338,
+      "step": 45961
+    },
+    {
+      "epoch": 1.0940207559744834,
+      "grad_norm": 1.9287357327141494,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 45962
+    },
+    {
+      "epoch": 1.094044558697515,
+      "grad_norm": 11.541128727818922,
+      "learning_rate": 2e-05,
+      "loss": 0.2947,
+      "step": 45963
+    },
+    {
+      "epoch": 1.0940683614205464,
+      "grad_norm": 5.528505646396321,
+      "learning_rate": 2e-05,
+      "loss": 0.1459,
+      "step": 45964
+    },
+    {
+      "epoch": 1.0940921641435781,
+      "grad_norm": 21.11734680559834,
+      "learning_rate": 2e-05,
+      "loss": 0.4806,
+      "step": 45965
+    },
+    {
+      "epoch": 1.0941159668666096,
+      "grad_norm": 2.907517937476836,
+      "learning_rate": 2e-05,
+      "loss": 0.2533,
+      "step": 45966
+    },
+    {
+      "epoch": 1.094139769589641,
+      "grad_norm": 0.4201517623400896,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 45967
+    },
+    {
+      "epoch": 1.0941635723126726,
+      "grad_norm": 7.960063734105227,
+      "learning_rate": 2e-05,
+      "loss": 1.8556,
+      "step": 45968
+    },
+    {
+      "epoch": 1.094187375035704,
+      "grad_norm": 0.8990205184088186,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 45969
+    },
+    {
+      "epoch": 1.0942111777587356,
+      "grad_norm": 2.216883158214928,
+      "learning_rate": 2e-05,
+      "loss": 0.0194,
+      "step": 45970
+    },
+    {
+      "epoch": 1.094234980481767,
+      "grad_norm": 0.16521705185256544,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 45971
+    },
+    {
+      "epoch": 1.0942587832047987,
+      "grad_norm": 7.332263238770355,
+      "learning_rate": 2e-05,
+      "loss": 0.7862,
+      "step": 45972
+    },
+    {
+      "epoch": 1.0942825859278302,
+      "grad_norm": 7.1873852174252875,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 45973
+    },
+    {
+      "epoch": 1.0943063886508617,
+      "grad_norm": 0.020914228855119812,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45974
+    },
+    {
+      "epoch": 1.0943301913738932,
+      "grad_norm": 2.858309619982985,
+      "learning_rate": 2e-05,
+      "loss": 0.1639,
+      "step": 45975
+    },
+    {
+      "epoch": 1.0943539940969247,
+      "grad_norm": 3.9920819196909294,
+      "learning_rate": 2e-05,
+      "loss": 0.1308,
+      "step": 45976
+    },
+    {
+      "epoch": 1.0943777968199562,
+      "grad_norm": 3.739717757726337,
+      "learning_rate": 2e-05,
+      "loss": 0.033,
+      "step": 45977
+    },
+    {
+      "epoch": 1.0944015995429877,
+      "grad_norm": 2.878217066279352,
+      "learning_rate": 2e-05,
+      "loss": 0.3583,
+      "step": 45978
+    },
+    {
+      "epoch": 1.0944254022660191,
+      "grad_norm": 0.06663363954885498,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45979
+    },
+    {
+      "epoch": 1.0944492049890509,
+      "grad_norm": 4.528567659944179,
+      "learning_rate": 2e-05,
+      "loss": 0.5653,
+      "step": 45980
+    },
+    {
+      "epoch": 1.0944730077120823,
+      "grad_norm": 3.2274049704774006,
+      "learning_rate": 2e-05,
+      "loss": 0.0625,
+      "step": 45981
+    },
+    {
+      "epoch": 1.0944968104351138,
+      "grad_norm": 0.04829342764070862,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 45982
+    },
+    {
+      "epoch": 1.0945206131581453,
+      "grad_norm": 2.192031473304659,
+      "learning_rate": 2e-05,
+      "loss": 0.036,
+      "step": 45983
+    },
+    {
+      "epoch": 1.0945444158811768,
+      "grad_norm": 0.5300789313669569,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 45984
+    },
+    {
+      "epoch": 1.0945682186042083,
+      "grad_norm": 0.8622527596623619,
+      "learning_rate": 2e-05,
+      "loss": 0.0073,
+      "step": 45985
+    },
+    {
+      "epoch": 1.0945920213272398,
+      "grad_norm": 0.230776143974891,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 45986
+    },
+    {
+      "epoch": 1.0946158240502712,
+      "grad_norm": 3.182155174213709,
+      "learning_rate": 2e-05,
+      "loss": 0.0588,
+      "step": 45987
+    },
+    {
+      "epoch": 1.094639626773303,
+      "grad_norm": 12.336873959776273,
+      "learning_rate": 2e-05,
+      "loss": 0.233,
+      "step": 45988
+    },
+    {
+      "epoch": 1.0946634294963344,
+      "grad_norm": 3.147297064325157,
+      "learning_rate": 2e-05,
+      "loss": 0.185,
+      "step": 45989
+    },
+    {
+      "epoch": 1.094687232219366,
+      "grad_norm": 13.045837227463139,
+      "learning_rate": 2e-05,
+      "loss": 1.0648,
+      "step": 45990
+    },
+    {
+      "epoch": 1.0947110349423974,
+      "grad_norm": 3.6427754091254387,
+      "learning_rate": 2e-05,
+      "loss": 0.5411,
+      "step": 45991
+    },
+    {
+      "epoch": 1.094734837665429,
+      "grad_norm": 0.012052595173130214,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 45992
+    },
+    {
+      "epoch": 1.0947586403884604,
+      "grad_norm": 0.18092004957055058,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 45993
+    },
+    {
+      "epoch": 1.0947824431114919,
+      "grad_norm": 0.18452158729359705,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45994
+    },
+    {
+      "epoch": 1.0948062458345236,
+      "grad_norm": 3.117184772921222,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 45995
+    },
+    {
+      "epoch": 1.094830048557555,
+      "grad_norm": 0.04873010824591161,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 45996
+    },
+    {
+      "epoch": 1.0948538512805865,
+      "grad_norm": 2.2829563051961466,
+      "learning_rate": 2e-05,
+      "loss": 0.1268,
+      "step": 45997
+    },
+    {
+      "epoch": 1.094877654003618,
+      "grad_norm": 0.8791981525627292,
+      "learning_rate": 2e-05,
+      "loss": 0.0079,
+      "step": 45998
+    },
+    {
+      "epoch": 1.0949014567266495,
+      "grad_norm": 0.4731876758787346,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 45999
+    },
+    {
+      "epoch": 1.094925259449681,
+      "grad_norm": 8.780478766501458,
+      "learning_rate": 2e-05,
+      "loss": 0.8785,
+      "step": 46000
+    },
+    {
+      "epoch": 1.0949490621727125,
+      "grad_norm": 0.8905933657331271,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 46001
+    },
+    {
+      "epoch": 1.094972864895744,
+      "grad_norm": 15.65342882474592,
+      "learning_rate": 2e-05,
+      "loss": 0.7189,
+      "step": 46002
+    },
+    {
+      "epoch": 1.0949966676187757,
+      "grad_norm": 5.88749136087587,
+      "learning_rate": 2e-05,
+      "loss": 0.9025,
+      "step": 46003
+    },
+    {
+      "epoch": 1.0950204703418072,
+      "grad_norm": 4.295301624847134,
+      "learning_rate": 2e-05,
+      "loss": 0.7162,
+      "step": 46004
+    },
+    {
+      "epoch": 1.0950442730648386,
+      "grad_norm": 15.435977836040577,
+      "learning_rate": 2e-05,
+      "loss": 0.1733,
+      "step": 46005
+    },
+    {
+      "epoch": 1.0950680757878701,
+      "grad_norm": 2.14258147896717,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 46006
+    },
+    {
+      "epoch": 1.0950918785109016,
+      "grad_norm": 4.564595912512786,
+      "learning_rate": 2e-05,
+      "loss": 0.5215,
+      "step": 46007
+    },
+    {
+      "epoch": 1.095115681233933,
+      "grad_norm": 12.80520633053086,
+      "learning_rate": 2e-05,
+      "loss": 0.8446,
+      "step": 46008
+    },
+    {
+      "epoch": 1.0951394839569646,
+      "grad_norm": 14.587211763352832,
+      "learning_rate": 2e-05,
+      "loss": 2.01,
+      "step": 46009
+    },
+    {
+      "epoch": 1.0951632866799963,
+      "grad_norm": 7.087768434292587,
+      "learning_rate": 2e-05,
+      "loss": 0.5268,
+      "step": 46010
+    },
+    {
+      "epoch": 1.0951870894030278,
+      "grad_norm": 6.212276326510082,
+      "learning_rate": 2e-05,
+      "loss": 1.0181,
+      "step": 46011
+    },
+    {
+      "epoch": 1.0952108921260593,
+      "grad_norm": 5.686621792855485,
+      "learning_rate": 2e-05,
+      "loss": 0.7159,
+      "step": 46012
+    },
+    {
+      "epoch": 1.0952346948490908,
+      "grad_norm": 2.10610361323893,
+      "learning_rate": 2e-05,
+      "loss": 0.1786,
+      "step": 46013
+    },
+    {
+      "epoch": 1.0952584975721222,
+      "grad_norm": 3.3343873448474586,
+      "learning_rate": 2e-05,
+      "loss": 0.0614,
+      "step": 46014
+    },
+    {
+      "epoch": 1.0952823002951537,
+      "grad_norm": 0.043674650928396376,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46015
+    },
+    {
+      "epoch": 1.0953061030181852,
+      "grad_norm": 3.7434033328916896,
+      "learning_rate": 2e-05,
+      "loss": 0.5334,
+      "step": 46016
+    },
+    {
+      "epoch": 1.0953299057412167,
+      "grad_norm": 0.0058968759173086025,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46017
+    },
+    {
+      "epoch": 1.0953537084642484,
+      "grad_norm": 10.086314516912378,
+      "learning_rate": 2e-05,
+      "loss": 0.5636,
+      "step": 46018
+    },
+    {
+      "epoch": 1.0953775111872799,
+      "grad_norm": 13.972288050996285,
+      "learning_rate": 2e-05,
+      "loss": 0.0833,
+      "step": 46019
+    },
+    {
+      "epoch": 1.0954013139103114,
+      "grad_norm": 0.1153754327350602,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46020
+    },
+    {
+      "epoch": 1.0954251166333429,
+      "grad_norm": 3.4850614798054056,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 46021
+    },
+    {
+      "epoch": 1.0954489193563743,
+      "grad_norm": 3.7163186181894003,
+      "learning_rate": 2e-05,
+      "loss": 0.039,
+      "step": 46022
+    },
+    {
+      "epoch": 1.0954727220794058,
+      "grad_norm": 2.233299708371586,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 46023
+    },
+    {
+      "epoch": 1.0954965248024373,
+      "grad_norm": 0.8294314489774739,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 46024
+    },
+    {
+      "epoch": 1.095520327525469,
+      "grad_norm": 0.35633976303082493,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46025
+    },
+    {
+      "epoch": 1.0955441302485005,
+      "grad_norm": 3.0090641566442433,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 46026
+    },
+    {
+      "epoch": 1.095567932971532,
+      "grad_norm": 0.10358965831078852,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 46027
+    },
+    {
+      "epoch": 1.0955917356945635,
+      "grad_norm": 8.121421326914874,
+      "learning_rate": 2e-05,
+      "loss": 0.272,
+      "step": 46028
+    },
+    {
+      "epoch": 1.095615538417595,
+      "grad_norm": 2.81328667478079,
+      "learning_rate": 2e-05,
+      "loss": 0.4617,
+      "step": 46029
+    },
+    {
+      "epoch": 1.0956393411406264,
+      "grad_norm": 4.595926879365331,
+      "learning_rate": 2e-05,
+      "loss": 0.1137,
+      "step": 46030
+    },
+    {
+      "epoch": 1.095663143863658,
+      "grad_norm": 0.07685570090607696,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 46031
+    },
+    {
+      "epoch": 1.0956869465866894,
+      "grad_norm": 0.44427405804374953,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 46032
+    },
+    {
+      "epoch": 1.0957107493097211,
+      "grad_norm": 0.03819590206655372,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46033
+    },
+    {
+      "epoch": 1.0957345520327526,
+      "grad_norm": 0.2340017322412383,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46034
+    },
+    {
+      "epoch": 1.095758354755784,
+      "grad_norm": 0.5272747525654083,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 46035
+    },
+    {
+      "epoch": 1.0957821574788156,
+      "grad_norm": 4.0380635533781115,
+      "learning_rate": 2e-05,
+      "loss": 0.0796,
+      "step": 46036
+    },
+    {
+      "epoch": 1.095805960201847,
+      "grad_norm": 0.46430619444600796,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 46037
+    },
+    {
+      "epoch": 1.0958297629248785,
+      "grad_norm": 5.413603950969056,
+      "learning_rate": 2e-05,
+      "loss": 0.0564,
+      "step": 46038
+    },
+    {
+      "epoch": 1.09585356564791,
+      "grad_norm": 6.563271029209963,
+      "learning_rate": 2e-05,
+      "loss": 0.6161,
+      "step": 46039
+    },
+    {
+      "epoch": 1.0958773683709415,
+      "grad_norm": 6.981004389459016,
+      "learning_rate": 2e-05,
+      "loss": 1.2897,
+      "step": 46040
+    },
+    {
+      "epoch": 1.0959011710939732,
+      "grad_norm": 2.779681514419479,
+      "learning_rate": 2e-05,
+      "loss": 0.1068,
+      "step": 46041
+    },
+    {
+      "epoch": 1.0959249738170047,
+      "grad_norm": 4.7423553463000525,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 46042
+    },
+    {
+      "epoch": 1.0959487765400362,
+      "grad_norm": 2.8812485780803425,
+      "learning_rate": 2e-05,
+      "loss": 0.1271,
+      "step": 46043
+    },
+    {
+      "epoch": 1.0959725792630677,
+      "grad_norm": 7.514614497089442,
+      "learning_rate": 2e-05,
+      "loss": 0.7818,
+      "step": 46044
+    },
+    {
+      "epoch": 1.0959963819860992,
+      "grad_norm": 11.005771586673744,
+      "learning_rate": 2e-05,
+      "loss": 0.8697,
+      "step": 46045
+    },
+    {
+      "epoch": 1.0960201847091307,
+      "grad_norm": 4.322629391534993,
+      "learning_rate": 2e-05,
+      "loss": 0.5078,
+      "step": 46046
+    },
+    {
+      "epoch": 1.0960439874321621,
+      "grad_norm": 1.9314403314393138,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 46047
+    },
+    {
+      "epoch": 1.0960677901551938,
+      "grad_norm": 0.48533176290396846,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46048
+    },
+    {
+      "epoch": 1.0960915928782253,
+      "grad_norm": 10.394267478404743,
+      "learning_rate": 2e-05,
+      "loss": 1.104,
+      "step": 46049
+    },
+    {
+      "epoch": 1.0961153956012568,
+      "grad_norm": 0.04216651620122308,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46050
+    },
+    {
+      "epoch": 1.0961391983242883,
+      "grad_norm": 5.934326228686577,
+      "learning_rate": 2e-05,
+      "loss": 0.3447,
+      "step": 46051
+    },
+    {
+      "epoch": 1.0961630010473198,
+      "grad_norm": 2.3501091649809656,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 46052
+    },
+    {
+      "epoch": 1.0961868037703513,
+      "grad_norm": 0.04184600133311589,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46053
+    },
+    {
+      "epoch": 1.0962106064933828,
+      "grad_norm": 0.1642549045890776,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46054
+    },
+    {
+      "epoch": 1.0962344092164145,
+      "grad_norm": 0.0381186851874058,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46055
+    },
+    {
+      "epoch": 1.096258211939446,
+      "grad_norm": 3.8781917158374246,
+      "learning_rate": 2e-05,
+      "loss": 0.3375,
+      "step": 46056
+    },
+    {
+      "epoch": 1.0962820146624774,
+      "grad_norm": 4.04185951500917,
+      "learning_rate": 2e-05,
+      "loss": 0.6192,
+      "step": 46057
+    },
+    {
+      "epoch": 1.096305817385509,
+      "grad_norm": 0.00869219649713209,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46058
+    },
+    {
+      "epoch": 1.0963296201085404,
+      "grad_norm": 0.454051148359203,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 46059
+    },
+    {
+      "epoch": 1.0963534228315719,
+      "grad_norm": 0.019646752821540416,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46060
+    },
+    {
+      "epoch": 1.0963772255546034,
+      "grad_norm": 2.8312471584720464,
+      "learning_rate": 2e-05,
+      "loss": 0.2363,
+      "step": 46061
+    },
+    {
+      "epoch": 1.0964010282776349,
+      "grad_norm": 3.220969921157229,
+      "learning_rate": 2e-05,
+      "loss": 0.3128,
+      "step": 46062
+    },
+    {
+      "epoch": 1.0964248310006666,
+      "grad_norm": 3.5346122769108503,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 46063
+    },
+    {
+      "epoch": 1.096448633723698,
+      "grad_norm": 14.131801553757763,
+      "learning_rate": 2e-05,
+      "loss": 0.5248,
+      "step": 46064
+    },
+    {
+      "epoch": 1.0964724364467295,
+      "grad_norm": 6.836196457119248,
+      "learning_rate": 2e-05,
+      "loss": 1.0434,
+      "step": 46065
+    },
+    {
+      "epoch": 1.096496239169761,
+      "grad_norm": 0.6134038482576071,
+      "learning_rate": 2e-05,
+      "loss": 0.0108,
+      "step": 46066
+    },
+    {
+      "epoch": 1.0965200418927925,
+      "grad_norm": 2.412989889150199,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 46067
+    },
+    {
+      "epoch": 1.096543844615824,
+      "grad_norm": 4.352138570111692,
+      "learning_rate": 2e-05,
+      "loss": 0.3592,
+      "step": 46068
+    },
+    {
+      "epoch": 1.0965676473388555,
+      "grad_norm": 5.210127574485944,
+      "learning_rate": 2e-05,
+      "loss": 0.0731,
+      "step": 46069
+    },
+    {
+      "epoch": 1.096591450061887,
+      "grad_norm": 2.9141768880879777,
+      "learning_rate": 2e-05,
+      "loss": 0.2082,
+      "step": 46070
+    },
+    {
+      "epoch": 1.0966152527849187,
+      "grad_norm": 1.99645404215369,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 46071
+    },
+    {
+      "epoch": 1.0966390555079502,
+      "grad_norm": 5.2126542936246905,
+      "learning_rate": 2e-05,
+      "loss": 0.2128,
+      "step": 46072
+    },
+    {
+      "epoch": 1.0966628582309816,
+      "grad_norm": 2.5131117050235643,
+      "learning_rate": 2e-05,
+      "loss": 0.2634,
+      "step": 46073
+    },
+    {
+      "epoch": 1.0966866609540131,
+      "grad_norm": 3.805997891727559,
+      "learning_rate": 2e-05,
+      "loss": 0.3893,
+      "step": 46074
+    },
+    {
+      "epoch": 1.0967104636770446,
+      "grad_norm": 3.1719323041902165,
+      "learning_rate": 2e-05,
+      "loss": 0.1124,
+      "step": 46075
+    },
+    {
+      "epoch": 1.096734266400076,
+      "grad_norm": 6.36229814046504,
+      "learning_rate": 2e-05,
+      "loss": 0.5414,
+      "step": 46076
+    },
+    {
+      "epoch": 1.0967580691231076,
+      "grad_norm": 1.3071633546955563,
+      "learning_rate": 2e-05,
+      "loss": 0.0091,
+      "step": 46077
+    },
+    {
+      "epoch": 1.0967818718461393,
+      "grad_norm": 3.9525681594915096,
+      "learning_rate": 2e-05,
+      "loss": 0.5969,
+      "step": 46078
+    },
+    {
+      "epoch": 1.0968056745691708,
+      "grad_norm": 5.368916699870205,
+      "learning_rate": 2e-05,
+      "loss": 0.9787,
+      "step": 46079
+    },
+    {
+      "epoch": 1.0968294772922023,
+      "grad_norm": 4.195518981855723,
+      "learning_rate": 2e-05,
+      "loss": 0.0554,
+      "step": 46080
+    },
+    {
+      "epoch": 1.0968532800152337,
+      "grad_norm": 0.8913769555827163,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 46081
+    },
+    {
+      "epoch": 1.0968770827382652,
+      "grad_norm": 0.27136819369213616,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 46082
+    },
+    {
+      "epoch": 1.0969008854612967,
+      "grad_norm": 0.04395888520814902,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46083
+    },
+    {
+      "epoch": 1.0969246881843282,
+      "grad_norm": 6.512512602025418,
+      "learning_rate": 2e-05,
+      "loss": 0.5096,
+      "step": 46084
+    },
+    {
+      "epoch": 1.09694849090736,
+      "grad_norm": 0.3694750691804691,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 46085
+    },
+    {
+      "epoch": 1.0969722936303914,
+      "grad_norm": 4.04948842615438,
+      "learning_rate": 2e-05,
+      "loss": 0.2518,
+      "step": 46086
+    },
+    {
+      "epoch": 1.0969960963534229,
+      "grad_norm": 7.731367067753092,
+      "learning_rate": 2e-05,
+      "loss": 1.0357,
+      "step": 46087
+    },
+    {
+      "epoch": 1.0970198990764544,
+      "grad_norm": 3.3797259738029655,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 46088
+    },
+    {
+      "epoch": 1.0970437017994858,
+      "grad_norm": 3.4896287299911735,
+      "learning_rate": 2e-05,
+      "loss": 0.3679,
+      "step": 46089
+    },
+    {
+      "epoch": 1.0970675045225173,
+      "grad_norm": 13.080901287239454,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 46090
+    },
+    {
+      "epoch": 1.0970913072455488,
+      "grad_norm": 0.2307462470758481,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 46091
+    },
+    {
+      "epoch": 1.0971151099685803,
+      "grad_norm": 6.157011807473785,
+      "learning_rate": 2e-05,
+      "loss": 0.1216,
+      "step": 46092
+    },
+    {
+      "epoch": 1.097138912691612,
+      "grad_norm": 2.4238786377166806,
+      "learning_rate": 2e-05,
+      "loss": 0.2403,
+      "step": 46093
+    },
+    {
+      "epoch": 1.0971627154146435,
+      "grad_norm": 7.136631859814205,
+      "learning_rate": 2e-05,
+      "loss": 0.1134,
+      "step": 46094
+    },
+    {
+      "epoch": 1.097186518137675,
+      "grad_norm": 6.821431971752466,
+      "learning_rate": 2e-05,
+      "loss": 0.1176,
+      "step": 46095
+    },
+    {
+      "epoch": 1.0972103208607065,
+      "grad_norm": 4.925227163935156,
+      "learning_rate": 2e-05,
+      "loss": 0.0667,
+      "step": 46096
+    },
+    {
+      "epoch": 1.097234123583738,
+      "grad_norm": 5.209675122218455,
+      "learning_rate": 2e-05,
+      "loss": 0.4279,
+      "step": 46097
+    },
+    {
+      "epoch": 1.0972579263067694,
+      "grad_norm": 5.966076243154365,
+      "learning_rate": 2e-05,
+      "loss": 0.0793,
+      "step": 46098
+    },
+    {
+      "epoch": 1.097281729029801,
+      "grad_norm": 7.470919553963394,
+      "learning_rate": 2e-05,
+      "loss": 0.7928,
+      "step": 46099
+    },
+    {
+      "epoch": 1.0973055317528324,
+      "grad_norm": 4.961447523539067,
+      "learning_rate": 2e-05,
+      "loss": 0.6446,
+      "step": 46100
+    },
+    {
+      "epoch": 1.0973293344758641,
+      "grad_norm": 0.479370348763277,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 46101
+    },
+    {
+      "epoch": 1.0973531371988956,
+      "grad_norm": 0.7939774329739847,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 46102
+    },
+    {
+      "epoch": 1.097376939921927,
+      "grad_norm": 5.042630849884517,
+      "learning_rate": 2e-05,
+      "loss": 0.5898,
+      "step": 46103
+    },
+    {
+      "epoch": 1.0974007426449586,
+      "grad_norm": 3.929883605800059,
+      "learning_rate": 2e-05,
+      "loss": 0.2826,
+      "step": 46104
+    },
+    {
+      "epoch": 1.09742454536799,
+      "grad_norm": 0.008336178157508192,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46105
+    },
+    {
+      "epoch": 1.0974483480910215,
+      "grad_norm": 0.12318860872936314,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46106
+    },
+    {
+      "epoch": 1.097472150814053,
+      "grad_norm": 2.964892143871962,
+      "learning_rate": 2e-05,
+      "loss": 0.0529,
+      "step": 46107
+    },
+    {
+      "epoch": 1.0974959535370847,
+      "grad_norm": 0.06404476872057188,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46108
+    },
+    {
+      "epoch": 1.0975197562601162,
+      "grad_norm": 1.1396693906085194,
+      "learning_rate": 2e-05,
+      "loss": 0.0113,
+      "step": 46109
+    },
+    {
+      "epoch": 1.0975435589831477,
+      "grad_norm": 4.671920859300829,
+      "learning_rate": 2e-05,
+      "loss": 0.1212,
+      "step": 46110
+    },
+    {
+      "epoch": 1.0975673617061792,
+      "grad_norm": 0.05626329328683818,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46111
+    },
+    {
+      "epoch": 1.0975911644292107,
+      "grad_norm": 1.1549527658232182,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 46112
+    },
+    {
+      "epoch": 1.0976149671522422,
+      "grad_norm": 1.767886094791714,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 46113
+    },
+    {
+      "epoch": 1.0976387698752736,
+      "grad_norm": 0.4739053548229928,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 46114
+    },
+    {
+      "epoch": 1.0976625725983054,
+      "grad_norm": 5.723739914788197,
+      "learning_rate": 2e-05,
+      "loss": 0.6786,
+      "step": 46115
+    },
+    {
+      "epoch": 1.0976863753213368,
+      "grad_norm": 5.858287830151839,
+      "learning_rate": 2e-05,
+      "loss": 0.5194,
+      "step": 46116
+    },
+    {
+      "epoch": 1.0977101780443683,
+      "grad_norm": 0.1643761830801659,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46117
+    },
+    {
+      "epoch": 1.0977339807673998,
+      "grad_norm": 5.752117049051982,
+      "learning_rate": 2e-05,
+      "loss": 0.0721,
+      "step": 46118
+    },
+    {
+      "epoch": 1.0977577834904313,
+      "grad_norm": 0.806872748402225,
+      "learning_rate": 2e-05,
+      "loss": 0.0098,
+      "step": 46119
+    },
+    {
+      "epoch": 1.0977815862134628,
+      "grad_norm": 0.28706762783748935,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46120
+    },
+    {
+      "epoch": 1.0978053889364943,
+      "grad_norm": 3.3170019543363805,
+      "learning_rate": 2e-05,
+      "loss": 0.3914,
+      "step": 46121
+    },
+    {
+      "epoch": 1.0978291916595257,
+      "grad_norm": 0.9052533011825508,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 46122
+    },
+    {
+      "epoch": 1.0978529943825575,
+      "grad_norm": 0.017686745039605257,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46123
+    },
+    {
+      "epoch": 1.097876797105589,
+      "grad_norm": 2.6325065158980405,
+      "learning_rate": 2e-05,
+      "loss": 0.0918,
+      "step": 46124
+    },
+    {
+      "epoch": 1.0979005998286204,
+      "grad_norm": 7.023852287269683,
+      "learning_rate": 2e-05,
+      "loss": 0.7694,
+      "step": 46125
+    },
+    {
+      "epoch": 1.097924402551652,
+      "grad_norm": 0.8681612747388828,
+      "learning_rate": 2e-05,
+      "loss": 0.0134,
+      "step": 46126
+    },
+    {
+      "epoch": 1.0979482052746834,
+      "grad_norm": 10.63530389892865,
+      "learning_rate": 2e-05,
+      "loss": 0.7042,
+      "step": 46127
+    },
+    {
+      "epoch": 1.0979720079977149,
+      "grad_norm": 3.380774941517005,
+      "learning_rate": 2e-05,
+      "loss": 0.4884,
+      "step": 46128
+    },
+    {
+      "epoch": 1.0979958107207464,
+      "grad_norm": 0.1173649404824398,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46129
+    },
+    {
+      "epoch": 1.0980196134437779,
+      "grad_norm": 11.682726820926966,
+      "learning_rate": 2e-05,
+      "loss": 0.1529,
+      "step": 46130
+    },
+    {
+      "epoch": 1.0980434161668096,
+      "grad_norm": 0.025041538508353254,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46131
+    },
+    {
+      "epoch": 1.098067218889841,
+      "grad_norm": 0.08278939477090429,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 46132
+    },
+    {
+      "epoch": 1.0980910216128725,
+      "grad_norm": 0.04552266818104565,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46133
+    },
+    {
+      "epoch": 1.098114824335904,
+      "grad_norm": 8.091551013371769,
+      "learning_rate": 2e-05,
+      "loss": 0.4356,
+      "step": 46134
+    },
+    {
+      "epoch": 1.0981386270589355,
+      "grad_norm": 3.738366085784148,
+      "learning_rate": 2e-05,
+      "loss": 0.4249,
+      "step": 46135
+    },
+    {
+      "epoch": 1.098162429781967,
+      "grad_norm": 11.626354720205072,
+      "learning_rate": 2e-05,
+      "loss": 0.3551,
+      "step": 46136
+    },
+    {
+      "epoch": 1.0981862325049985,
+      "grad_norm": 2.899025655927057,
+      "learning_rate": 2e-05,
+      "loss": 0.3362,
+      "step": 46137
+    },
+    {
+      "epoch": 1.0982100352280302,
+      "grad_norm": 0.28779095319832987,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 46138
+    },
+    {
+      "epoch": 1.0982338379510617,
+      "grad_norm": 2.690086039612825,
+      "learning_rate": 2e-05,
+      "loss": 0.2649,
+      "step": 46139
+    },
+    {
+      "epoch": 1.0982576406740931,
+      "grad_norm": 0.058568524022405644,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46140
+    },
+    {
+      "epoch": 1.0982814433971246,
+      "grad_norm": 0.9966794490217195,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 46141
+    },
+    {
+      "epoch": 1.0983052461201561,
+      "grad_norm": 5.260485125922456,
+      "learning_rate": 2e-05,
+      "loss": 0.5447,
+      "step": 46142
+    },
+    {
+      "epoch": 1.0983290488431876,
+      "grad_norm": 0.08367911232420429,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46143
+    },
+    {
+      "epoch": 1.098352851566219,
+      "grad_norm": 3.5988651642404084,
+      "learning_rate": 2e-05,
+      "loss": 0.2979,
+      "step": 46144
+    },
+    {
+      "epoch": 1.0983766542892508,
+      "grad_norm": 0.004389134576495687,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46145
+    },
+    {
+      "epoch": 1.0984004570122823,
+      "grad_norm": 0.4969047515687644,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 46146
+    },
+    {
+      "epoch": 1.0984242597353138,
+      "grad_norm": 0.049375105837361884,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46147
+    },
+    {
+      "epoch": 1.0984480624583453,
+      "grad_norm": 3.3363849115992528,
+      "learning_rate": 2e-05,
+      "loss": 0.1884,
+      "step": 46148
+    },
+    {
+      "epoch": 1.0984718651813767,
+      "grad_norm": 9.640894287518714,
+      "learning_rate": 2e-05,
+      "loss": 0.6744,
+      "step": 46149
+    },
+    {
+      "epoch": 1.0984956679044082,
+      "grad_norm": 13.599439078788405,
+      "learning_rate": 2e-05,
+      "loss": 0.528,
+      "step": 46150
+    },
+    {
+      "epoch": 1.0985194706274397,
+      "grad_norm": 6.367562262713738,
+      "learning_rate": 2e-05,
+      "loss": 1.2226,
+      "step": 46151
+    },
+    {
+      "epoch": 1.0985432733504712,
+      "grad_norm": 1.3800191168830096,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 46152
+    },
+    {
+      "epoch": 1.098567076073503,
+      "grad_norm": 3.4942641078941254,
+      "learning_rate": 2e-05,
+      "loss": 0.5613,
+      "step": 46153
+    },
+    {
+      "epoch": 1.0985908787965344,
+      "grad_norm": 0.017144840074425766,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46154
+    },
+    {
+      "epoch": 1.0986146815195659,
+      "grad_norm": 0.062058231769474184,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46155
+    },
+    {
+      "epoch": 1.0986384842425974,
+      "grad_norm": 6.261867862212733,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 46156
+    },
+    {
+      "epoch": 1.0986622869656288,
+      "grad_norm": 7.238902830626511,
+      "learning_rate": 2e-05,
+      "loss": 0.9764,
+      "step": 46157
+    },
+    {
+      "epoch": 1.0986860896886603,
+      "grad_norm": 3.6677518520654178,
+      "learning_rate": 2e-05,
+      "loss": 0.0847,
+      "step": 46158
+    },
+    {
+      "epoch": 1.0987098924116918,
+      "grad_norm": 0.15559000594875225,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46159
+    },
+    {
+      "epoch": 1.0987336951347233,
+      "grad_norm": 0.026951262108040778,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46160
+    },
+    {
+      "epoch": 1.098757497857755,
+      "grad_norm": 4.539334875208458,
+      "learning_rate": 2e-05,
+      "loss": 0.453,
+      "step": 46161
+    },
+    {
+      "epoch": 1.0987813005807865,
+      "grad_norm": 5.0648044879828795,
+      "learning_rate": 2e-05,
+      "loss": 0.4159,
+      "step": 46162
+    },
+    {
+      "epoch": 1.098805103303818,
+      "grad_norm": 2.0543322241345843,
+      "learning_rate": 2e-05,
+      "loss": 0.0403,
+      "step": 46163
+    },
+    {
+      "epoch": 1.0988289060268495,
+      "grad_norm": 7.5618759560703195,
+      "learning_rate": 2e-05,
+      "loss": 0.7346,
+      "step": 46164
+    },
+    {
+      "epoch": 1.098852708749881,
+      "grad_norm": 4.2029780011486375,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 46165
+    },
+    {
+      "epoch": 1.0988765114729124,
+      "grad_norm": 7.670905765351225,
+      "learning_rate": 2e-05,
+      "loss": 0.9789,
+      "step": 46166
+    },
+    {
+      "epoch": 1.098900314195944,
+      "grad_norm": 3.4842013920552115,
+      "learning_rate": 2e-05,
+      "loss": 0.535,
+      "step": 46167
+    },
+    {
+      "epoch": 1.0989241169189756,
+      "grad_norm": 3.7943661735051237,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 46168
+    },
+    {
+      "epoch": 1.098947919642007,
+      "grad_norm": 5.719595391591197,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 46169
+    },
+    {
+      "epoch": 1.0989717223650386,
+      "grad_norm": 3.6645750228838208,
+      "learning_rate": 2e-05,
+      "loss": 0.2283,
+      "step": 46170
+    },
+    {
+      "epoch": 1.09899552508807,
+      "grad_norm": 0.08056471763321432,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46171
+    },
+    {
+      "epoch": 1.0990193278111016,
+      "grad_norm": 5.992461512748164,
+      "learning_rate": 2e-05,
+      "loss": 0.405,
+      "step": 46172
+    },
+    {
+      "epoch": 1.099043130534133,
+      "grad_norm": 6.3302190848996265,
+      "learning_rate": 2e-05,
+      "loss": 0.0978,
+      "step": 46173
+    },
+    {
+      "epoch": 1.0990669332571645,
+      "grad_norm": 2.5255992174139474,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 46174
+    },
+    {
+      "epoch": 1.0990907359801962,
+      "grad_norm": 0.0726229315997947,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46175
+    },
+    {
+      "epoch": 1.0991145387032277,
+      "grad_norm": 4.660771061416184,
+      "learning_rate": 2e-05,
+      "loss": 0.0304,
+      "step": 46176
+    },
+    {
+      "epoch": 1.0991383414262592,
+      "grad_norm": 4.57581513921685,
+      "learning_rate": 2e-05,
+      "loss": 0.1619,
+      "step": 46177
+    },
+    {
+      "epoch": 1.0991621441492907,
+      "grad_norm": 4.7508166513761285,
+      "learning_rate": 2e-05,
+      "loss": 0.5194,
+      "step": 46178
+    },
+    {
+      "epoch": 1.0991859468723222,
+      "grad_norm": 1.1825998568172016,
+      "learning_rate": 2e-05,
+      "loss": 0.0206,
+      "step": 46179
+    },
+    {
+      "epoch": 1.0992097495953537,
+      "grad_norm": 6.971952892223256,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 46180
+    },
+    {
+      "epoch": 1.0992335523183852,
+      "grad_norm": 1.926522739067255,
+      "learning_rate": 2e-05,
+      "loss": 0.1078,
+      "step": 46181
+    },
+    {
+      "epoch": 1.0992573550414166,
+      "grad_norm": 8.626316032588598,
+      "learning_rate": 2e-05,
+      "loss": 0.2629,
+      "step": 46182
+    },
+    {
+      "epoch": 1.0992811577644483,
+      "grad_norm": 7.350031649080753,
+      "learning_rate": 2e-05,
+      "loss": 0.7527,
+      "step": 46183
+    },
+    {
+      "epoch": 1.0993049604874798,
+      "grad_norm": 2.458564104363506,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 46184
+    },
+    {
+      "epoch": 1.0993287632105113,
+      "grad_norm": 0.043472297570452643,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46185
+    },
+    {
+      "epoch": 1.0993525659335428,
+      "grad_norm": 3.4814859518382577,
+      "learning_rate": 2e-05,
+      "loss": 0.1027,
+      "step": 46186
+    },
+    {
+      "epoch": 1.0993763686565743,
+      "grad_norm": 0.08846110063960498,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46187
+    },
+    {
+      "epoch": 1.0994001713796058,
+      "grad_norm": 1.9516613669965306,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 46188
+    },
+    {
+      "epoch": 1.0994239741026373,
+      "grad_norm": 6.248174871890073,
+      "learning_rate": 2e-05,
+      "loss": 0.0781,
+      "step": 46189
+    },
+    {
+      "epoch": 1.0994477768256687,
+      "grad_norm": 2.7530972587870597,
+      "learning_rate": 2e-05,
+      "loss": 0.0456,
+      "step": 46190
+    },
+    {
+      "epoch": 1.0994715795487005,
+      "grad_norm": 0.2920827470484122,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 46191
+    },
+    {
+      "epoch": 1.099495382271732,
+      "grad_norm": 0.007857647848220635,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46192
+    },
+    {
+      "epoch": 1.0995191849947634,
+      "grad_norm": 3.649897591163476,
+      "learning_rate": 2e-05,
+      "loss": 0.0831,
+      "step": 46193
+    },
+    {
+      "epoch": 1.099542987717795,
+      "grad_norm": 5.175774650212297,
+      "learning_rate": 2e-05,
+      "loss": 0.9667,
+      "step": 46194
+    },
+    {
+      "epoch": 1.0995667904408264,
+      "grad_norm": 0.003823849051250019,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46195
+    },
+    {
+      "epoch": 1.0995905931638579,
+      "grad_norm": 0.027602236707781565,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46196
+    },
+    {
+      "epoch": 1.0996143958868894,
+      "grad_norm": 6.3259980967204195,
+      "learning_rate": 2e-05,
+      "loss": 0.1061,
+      "step": 46197
+    },
+    {
+      "epoch": 1.099638198609921,
+      "grad_norm": 3.3513256872390644,
+      "learning_rate": 2e-05,
+      "loss": 0.3206,
+      "step": 46198
+    },
+    {
+      "epoch": 1.0996620013329526,
+      "grad_norm": 0.02550837473142559,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46199
+    },
+    {
+      "epoch": 1.099685804055984,
+      "grad_norm": 0.052171520574554674,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46200
+    },
+    {
+      "epoch": 1.0997096067790155,
+      "grad_norm": 8.04708330671251,
+      "learning_rate": 2e-05,
+      "loss": 0.3612,
+      "step": 46201
+    },
+    {
+      "epoch": 1.099733409502047,
+      "grad_norm": 0.8364283632799123,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 46202
+    },
+    {
+      "epoch": 1.0997572122250785,
+      "grad_norm": 4.525892408208284,
+      "learning_rate": 2e-05,
+      "loss": 0.3818,
+      "step": 46203
+    },
+    {
+      "epoch": 1.09978101494811,
+      "grad_norm": 4.067712534557621,
+      "learning_rate": 2e-05,
+      "loss": 0.3353,
+      "step": 46204
+    },
+    {
+      "epoch": 1.0998048176711417,
+      "grad_norm": 0.08524541941373585,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46205
+    },
+    {
+      "epoch": 1.0998286203941732,
+      "grad_norm": 4.405910939657206,
+      "learning_rate": 2e-05,
+      "loss": 0.7333,
+      "step": 46206
+    },
+    {
+      "epoch": 1.0998524231172047,
+      "grad_norm": 0.6419890682974513,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 46207
+    },
+    {
+      "epoch": 1.0998762258402361,
+      "grad_norm": 6.105819236288721,
+      "learning_rate": 2e-05,
+      "loss": 0.5572,
+      "step": 46208
+    },
+    {
+      "epoch": 1.0999000285632676,
+      "grad_norm": 3.762658204870821,
+      "learning_rate": 2e-05,
+      "loss": 0.0956,
+      "step": 46209
+    },
+    {
+      "epoch": 1.0999238312862991,
+      "grad_norm": 0.4987477994417857,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 46210
+    },
+    {
+      "epoch": 1.0999476340093306,
+      "grad_norm": 3.3736402050132153,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 46211
+    },
+    {
+      "epoch": 1.099971436732362,
+      "grad_norm": 0.032232269818059656,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46212
+    },
+    {
+      "epoch": 1.0999952394553938,
+      "grad_norm": 4.775190741078,
+      "learning_rate": 2e-05,
+      "loss": 0.6889,
+      "step": 46213
+    },
+    {
+      "epoch": 1.1000190421784253,
+      "grad_norm": 0.45638799224663296,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 46214
+    },
+    {
+      "epoch": 1.1000428449014568,
+      "grad_norm": 4.2862075291812705,
+      "learning_rate": 2e-05,
+      "loss": 0.155,
+      "step": 46215
+    },
+    {
+      "epoch": 1.1000666476244882,
+      "grad_norm": 0.014808898180162925,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46216
+    },
+    {
+      "epoch": 1.1000904503475197,
+      "grad_norm": 1.5060109905370904,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 46217
+    },
+    {
+      "epoch": 1.1001142530705512,
+      "grad_norm": 4.14741625502346,
+      "learning_rate": 2e-05,
+      "loss": 0.7413,
+      "step": 46218
+    },
+    {
+      "epoch": 1.1001380557935827,
+      "grad_norm": 3.061891430300467,
+      "learning_rate": 2e-05,
+      "loss": 0.1583,
+      "step": 46219
+    },
+    {
+      "epoch": 1.1001618585166142,
+      "grad_norm": 0.04093892768104011,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46220
+    },
+    {
+      "epoch": 1.100185661239646,
+      "grad_norm": 0.1108867216075773,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46221
+    },
+    {
+      "epoch": 1.1002094639626774,
+      "grad_norm": 1.844017802613393,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 46222
+    },
+    {
+      "epoch": 1.1002332666857089,
+      "grad_norm": 10.695452008404398,
+      "learning_rate": 2e-05,
+      "loss": 1.3808,
+      "step": 46223
+    },
+    {
+      "epoch": 1.1002570694087404,
+      "grad_norm": 0.3403823618412369,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 46224
+    },
+    {
+      "epoch": 1.1002808721317718,
+      "grad_norm": 3.8928858326078664,
+      "learning_rate": 2e-05,
+      "loss": 0.211,
+      "step": 46225
+    },
+    {
+      "epoch": 1.1003046748548033,
+      "grad_norm": 3.2387834360530356,
+      "learning_rate": 2e-05,
+      "loss": 0.2201,
+      "step": 46226
+    },
+    {
+      "epoch": 1.1003284775778348,
+      "grad_norm": 0.3166818722397139,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 46227
+    },
+    {
+      "epoch": 1.1003522803008665,
+      "grad_norm": 4.212446934315594,
+      "learning_rate": 2e-05,
+      "loss": 0.069,
+      "step": 46228
+    },
+    {
+      "epoch": 1.100376083023898,
+      "grad_norm": 0.1660401914240218,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 46229
+    },
+    {
+      "epoch": 1.1003998857469295,
+      "grad_norm": 35.63996423857563,
+      "learning_rate": 2e-05,
+      "loss": 0.8929,
+      "step": 46230
+    },
+    {
+      "epoch": 1.100423688469961,
+      "grad_norm": 0.026904333587085966,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46231
+    },
+    {
+      "epoch": 1.1004474911929925,
+      "grad_norm": 1.57886310497887,
+      "learning_rate": 2e-05,
+      "loss": 0.0673,
+      "step": 46232
+    },
+    {
+      "epoch": 1.100471293916024,
+      "grad_norm": 0.08018182980082837,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46233
+    },
+    {
+      "epoch": 1.1004950966390554,
+      "grad_norm": 1.016135231485496,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 46234
+    },
+    {
+      "epoch": 1.1005188993620871,
+      "grad_norm": 0.0076130442378035575,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46235
+    },
+    {
+      "epoch": 1.1005427020851186,
+      "grad_norm": 3.8184856268965164,
+      "learning_rate": 2e-05,
+      "loss": 0.3484,
+      "step": 46236
+    },
+    {
+      "epoch": 1.10056650480815,
+      "grad_norm": 4.720490536560132,
+      "learning_rate": 2e-05,
+      "loss": 0.3179,
+      "step": 46237
+    },
+    {
+      "epoch": 1.1005903075311816,
+      "grad_norm": 29.606949681785043,
+      "learning_rate": 2e-05,
+      "loss": 0.3122,
+      "step": 46238
+    },
+    {
+      "epoch": 1.100614110254213,
+      "grad_norm": 0.9144725994323977,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 46239
+    },
+    {
+      "epoch": 1.1006379129772446,
+      "grad_norm": 2.8217221438635107,
+      "learning_rate": 2e-05,
+      "loss": 0.2729,
+      "step": 46240
+    },
+    {
+      "epoch": 1.100661715700276,
+      "grad_norm": 0.021235903337090568,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46241
+    },
+    {
+      "epoch": 1.1006855184233075,
+      "grad_norm": 4.307049415530484,
+      "learning_rate": 2e-05,
+      "loss": 0.6374,
+      "step": 46242
+    },
+    {
+      "epoch": 1.1007093211463392,
+      "grad_norm": 1.7460318553996823,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 46243
+    },
+    {
+      "epoch": 1.1007331238693707,
+      "grad_norm": 15.053703744912886,
+      "learning_rate": 2e-05,
+      "loss": 0.9917,
+      "step": 46244
+    },
+    {
+      "epoch": 1.1007569265924022,
+      "grad_norm": 3.087118038905793,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 46245
+    },
+    {
+      "epoch": 1.1007807293154337,
+      "grad_norm": 4.743415279117196,
+      "learning_rate": 2e-05,
+      "loss": 0.4742,
+      "step": 46246
+    },
+    {
+      "epoch": 1.1008045320384652,
+      "grad_norm": 0.024126396406637673,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46247
+    },
+    {
+      "epoch": 1.1008283347614967,
+      "grad_norm": 6.781748686797932,
+      "learning_rate": 2e-05,
+      "loss": 0.0427,
+      "step": 46248
+    },
+    {
+      "epoch": 1.1008521374845281,
+      "grad_norm": 4.079732291258487,
+      "learning_rate": 2e-05,
+      "loss": 0.487,
+      "step": 46249
+    },
+    {
+      "epoch": 1.1008759402075596,
+      "grad_norm": 2.421984159522426,
+      "learning_rate": 2e-05,
+      "loss": 0.0275,
+      "step": 46250
+    },
+    {
+      "epoch": 1.1008997429305913,
+      "grad_norm": 3.728399983315053,
+      "learning_rate": 2e-05,
+      "loss": 0.4789,
+      "step": 46251
+    },
+    {
+      "epoch": 1.1009235456536228,
+      "grad_norm": 1.7486716583031838,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 46252
+    },
+    {
+      "epoch": 1.1009473483766543,
+      "grad_norm": 4.760600444223887,
+      "learning_rate": 2e-05,
+      "loss": 0.3586,
+      "step": 46253
+    },
+    {
+      "epoch": 1.1009711510996858,
+      "grad_norm": 0.5055286668386626,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 46254
+    },
+    {
+      "epoch": 1.1009949538227173,
+      "grad_norm": 2.8594711640658566,
+      "learning_rate": 2e-05,
+      "loss": 0.3675,
+      "step": 46255
+    },
+    {
+      "epoch": 1.1010187565457488,
+      "grad_norm": 1.9636735231555555,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 46256
+    },
+    {
+      "epoch": 1.1010425592687803,
+      "grad_norm": 0.018126578502739823,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46257
+    },
+    {
+      "epoch": 1.101066361991812,
+      "grad_norm": 4.936874254140075,
+      "learning_rate": 2e-05,
+      "loss": 0.7515,
+      "step": 46258
+    },
+    {
+      "epoch": 1.1010901647148434,
+      "grad_norm": 0.9238264769669702,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 46259
+    },
+    {
+      "epoch": 1.101113967437875,
+      "grad_norm": 2.571810143309626,
+      "learning_rate": 2e-05,
+      "loss": 0.2366,
+      "step": 46260
+    },
+    {
+      "epoch": 1.1011377701609064,
+      "grad_norm": 2.3691468388427257,
+      "learning_rate": 2e-05,
+      "loss": 0.1601,
+      "step": 46261
+    },
+    {
+      "epoch": 1.101161572883938,
+      "grad_norm": 6.004837605593897,
+      "learning_rate": 2e-05,
+      "loss": 0.4969,
+      "step": 46262
+    },
+    {
+      "epoch": 1.1011853756069694,
+      "grad_norm": 4.378074364111288,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 46263
+    },
+    {
+      "epoch": 1.1012091783300009,
+      "grad_norm": 0.012294482514275971,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46264
+    },
+    {
+      "epoch": 1.1012329810530326,
+      "grad_norm": 6.059850580088658,
+      "learning_rate": 2e-05,
+      "loss": 0.8739,
+      "step": 46265
+    },
+    {
+      "epoch": 1.101256783776064,
+      "grad_norm": 0.0194769003312446,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46266
+    },
+    {
+      "epoch": 1.1012805864990955,
+      "grad_norm": 0.288053465580589,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 46267
+    },
+    {
+      "epoch": 1.101304389222127,
+      "grad_norm": 1.4023181090572627,
+      "learning_rate": 2e-05,
+      "loss": 0.0975,
+      "step": 46268
+    },
+    {
+      "epoch": 1.1013281919451585,
+      "grad_norm": 6.477531715890351,
+      "learning_rate": 2e-05,
+      "loss": 0.7071,
+      "step": 46269
+    },
+    {
+      "epoch": 1.10135199466819,
+      "grad_norm": 5.475439108666281,
+      "learning_rate": 2e-05,
+      "loss": 0.3604,
+      "step": 46270
+    },
+    {
+      "epoch": 1.1013757973912215,
+      "grad_norm": 1.2877219728097893,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 46271
+    },
+    {
+      "epoch": 1.101399600114253,
+      "grad_norm": 0.6081169926252312,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 46272
+    },
+    {
+      "epoch": 1.1014234028372847,
+      "grad_norm": 10.451155915194226,
+      "learning_rate": 2e-05,
+      "loss": 1.2609,
+      "step": 46273
+    },
+    {
+      "epoch": 1.1014472055603162,
+      "grad_norm": 0.049108622874485794,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46274
+    },
+    {
+      "epoch": 1.1014710082833477,
+      "grad_norm": 3.196896499123776,
+      "learning_rate": 2e-05,
+      "loss": 0.0999,
+      "step": 46275
+    },
+    {
+      "epoch": 1.1014948110063791,
+      "grad_norm": 0.041432136574270646,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46276
+    },
+    {
+      "epoch": 1.1015186137294106,
+      "grad_norm": 5.827051577581867,
+      "learning_rate": 2e-05,
+      "loss": 0.4978,
+      "step": 46277
+    },
+    {
+      "epoch": 1.101542416452442,
+      "grad_norm": 5.5772889370076335,
+      "learning_rate": 2e-05,
+      "loss": 0.5791,
+      "step": 46278
+    },
+    {
+      "epoch": 1.1015662191754736,
+      "grad_norm": 11.021595140998249,
+      "learning_rate": 2e-05,
+      "loss": 1.1262,
+      "step": 46279
+    },
+    {
+      "epoch": 1.101590021898505,
+      "grad_norm": 3.9992700158585213,
+      "learning_rate": 2e-05,
+      "loss": 0.4941,
+      "step": 46280
+    },
+    {
+      "epoch": 1.1016138246215368,
+      "grad_norm": 5.342892306231382,
+      "learning_rate": 2e-05,
+      "loss": 0.1671,
+      "step": 46281
+    },
+    {
+      "epoch": 1.1016376273445683,
+      "grad_norm": 0.04032437703488414,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46282
+    },
+    {
+      "epoch": 1.1016614300675998,
+      "grad_norm": 0.280349426918252,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46283
+    },
+    {
+      "epoch": 1.1016852327906312,
+      "grad_norm": 0.024259509350329653,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46284
+    },
+    {
+      "epoch": 1.1017090355136627,
+      "grad_norm": 0.15929575624520736,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46285
+    },
+    {
+      "epoch": 1.1017328382366942,
+      "grad_norm": 0.8155739480004419,
+      "learning_rate": 2e-05,
+      "loss": 0.0067,
+      "step": 46286
+    },
+    {
+      "epoch": 1.1017566409597257,
+      "grad_norm": 5.457717802013499,
+      "learning_rate": 2e-05,
+      "loss": 0.4046,
+      "step": 46287
+    },
+    {
+      "epoch": 1.1017804436827574,
+      "grad_norm": 0.055744956634069816,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46288
+    },
+    {
+      "epoch": 1.101804246405789,
+      "grad_norm": 6.941107542879695,
+      "learning_rate": 2e-05,
+      "loss": 0.6698,
+      "step": 46289
+    },
+    {
+      "epoch": 1.1018280491288204,
+      "grad_norm": 5.485757284543494,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 46290
+    },
+    {
+      "epoch": 1.1018518518518519,
+      "grad_norm": 0.5581371764950005,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 46291
+    },
+    {
+      "epoch": 1.1018756545748833,
+      "grad_norm": 7.52358748027699,
+      "learning_rate": 2e-05,
+      "loss": 0.5649,
+      "step": 46292
+    },
+    {
+      "epoch": 1.1018994572979148,
+      "grad_norm": 3.653362551177384,
+      "learning_rate": 2e-05,
+      "loss": 0.5154,
+      "step": 46293
+    },
+    {
+      "epoch": 1.1019232600209463,
+      "grad_norm": 5.4731722063027926,
+      "learning_rate": 2e-05,
+      "loss": 0.0628,
+      "step": 46294
+    },
+    {
+      "epoch": 1.101947062743978,
+      "grad_norm": 8.6101544520377,
+      "learning_rate": 2e-05,
+      "loss": 0.4542,
+      "step": 46295
+    },
+    {
+      "epoch": 1.1019708654670095,
+      "grad_norm": 2.4040876343947333,
+      "learning_rate": 2e-05,
+      "loss": 0.0236,
+      "step": 46296
+    },
+    {
+      "epoch": 1.101994668190041,
+      "grad_norm": 3.3776527782125254,
+      "learning_rate": 2e-05,
+      "loss": 0.2982,
+      "step": 46297
+    },
+    {
+      "epoch": 1.1020184709130725,
+      "grad_norm": 0.03483570811735794,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46298
+    },
+    {
+      "epoch": 1.102042273636104,
+      "grad_norm": 0.07526425909669276,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46299
+    },
+    {
+      "epoch": 1.1020660763591354,
+      "grad_norm": 0.16316532560795521,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46300
+    },
+    {
+      "epoch": 1.102089879082167,
+      "grad_norm": 4.493513329857355,
+      "learning_rate": 2e-05,
+      "loss": 0.5975,
+      "step": 46301
+    },
+    {
+      "epoch": 1.1021136818051984,
+      "grad_norm": 7.701568469690287,
+      "learning_rate": 2e-05,
+      "loss": 1.1504,
+      "step": 46302
+    },
+    {
+      "epoch": 1.1021374845282301,
+      "grad_norm": 5.618860089446336,
+      "learning_rate": 2e-05,
+      "loss": 0.9781,
+      "step": 46303
+    },
+    {
+      "epoch": 1.1021612872512616,
+      "grad_norm": 1.4018375965355538,
+      "learning_rate": 2e-05,
+      "loss": 0.0426,
+      "step": 46304
+    },
+    {
+      "epoch": 1.102185089974293,
+      "grad_norm": 1.4317732875070237,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 46305
+    },
+    {
+      "epoch": 1.1022088926973246,
+      "grad_norm": 2.9643543733011506,
+      "learning_rate": 2e-05,
+      "loss": 0.1283,
+      "step": 46306
+    },
+    {
+      "epoch": 1.102232695420356,
+      "grad_norm": 12.461637582546965,
+      "learning_rate": 2e-05,
+      "loss": 0.2122,
+      "step": 46307
+    },
+    {
+      "epoch": 1.1022564981433876,
+      "grad_norm": 5.714313695302759,
+      "learning_rate": 2e-05,
+      "loss": 0.3967,
+      "step": 46308
+    },
+    {
+      "epoch": 1.102280300866419,
+      "grad_norm": 6.03970362300242,
+      "learning_rate": 2e-05,
+      "loss": 0.0841,
+      "step": 46309
+    },
+    {
+      "epoch": 1.1023041035894505,
+      "grad_norm": 2.4587485816531327,
+      "learning_rate": 2e-05,
+      "loss": 0.2235,
+      "step": 46310
+    },
+    {
+      "epoch": 1.1023279063124822,
+      "grad_norm": 4.817035592894973,
+      "learning_rate": 2e-05,
+      "loss": 0.4044,
+      "step": 46311
+    },
+    {
+      "epoch": 1.1023517090355137,
+      "grad_norm": 0.025607617501364435,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46312
+    },
+    {
+      "epoch": 1.1023755117585452,
+      "grad_norm": 6.349866238992866,
+      "learning_rate": 2e-05,
+      "loss": 0.0462,
+      "step": 46313
+    },
+    {
+      "epoch": 1.1023993144815767,
+      "grad_norm": 0.03928648008867749,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46314
+    },
+    {
+      "epoch": 1.1024231172046082,
+      "grad_norm": 8.483366917195013,
+      "learning_rate": 2e-05,
+      "loss": 0.9613,
+      "step": 46315
+    },
+    {
+      "epoch": 1.1024469199276397,
+      "grad_norm": 4.147813345722702,
+      "learning_rate": 2e-05,
+      "loss": 0.3765,
+      "step": 46316
+    },
+    {
+      "epoch": 1.1024707226506711,
+      "grad_norm": 1.2342905417009336,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 46317
+    },
+    {
+      "epoch": 1.1024945253737028,
+      "grad_norm": 2.5198066632649243,
+      "learning_rate": 2e-05,
+      "loss": 0.2154,
+      "step": 46318
+    },
+    {
+      "epoch": 1.1025183280967343,
+      "grad_norm": 6.638931694785707,
+      "learning_rate": 2e-05,
+      "loss": 0.0567,
+      "step": 46319
+    },
+    {
+      "epoch": 1.1025421308197658,
+      "grad_norm": 0.2300188981387483,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46320
+    },
+    {
+      "epoch": 1.1025659335427973,
+      "grad_norm": 0.18201645335404587,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 46321
+    },
+    {
+      "epoch": 1.1025897362658288,
+      "grad_norm": 3.126443433229837,
+      "learning_rate": 2e-05,
+      "loss": 0.2198,
+      "step": 46322
+    },
+    {
+      "epoch": 1.1026135389888603,
+      "grad_norm": 5.087573459879395,
+      "learning_rate": 2e-05,
+      "loss": 0.4909,
+      "step": 46323
+    },
+    {
+      "epoch": 1.1026373417118918,
+      "grad_norm": 19.795760348687466,
+      "learning_rate": 2e-05,
+      "loss": 0.2292,
+      "step": 46324
+    },
+    {
+      "epoch": 1.1026611444349235,
+      "grad_norm": 6.454417269064032,
+      "learning_rate": 2e-05,
+      "loss": 0.0629,
+      "step": 46325
+    },
+    {
+      "epoch": 1.102684947157955,
+      "grad_norm": 3.0946671539079866,
+      "learning_rate": 2e-05,
+      "loss": 0.1012,
+      "step": 46326
+    },
+    {
+      "epoch": 1.1027087498809864,
+      "grad_norm": 4.737269310272644,
+      "learning_rate": 2e-05,
+      "loss": 0.6241,
+      "step": 46327
+    },
+    {
+      "epoch": 1.102732552604018,
+      "grad_norm": 0.1537196981489292,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46328
+    },
+    {
+      "epoch": 1.1027563553270494,
+      "grad_norm": 2.944887759202235,
+      "learning_rate": 2e-05,
+      "loss": 0.204,
+      "step": 46329
+    },
+    {
+      "epoch": 1.102780158050081,
+      "grad_norm": 0.14202827445051042,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46330
+    },
+    {
+      "epoch": 1.1028039607731124,
+      "grad_norm": 0.7331923117393037,
+      "learning_rate": 2e-05,
+      "loss": 0.0126,
+      "step": 46331
+    },
+    {
+      "epoch": 1.1028277634961439,
+      "grad_norm": 0.013407200451645588,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46332
+    },
+    {
+      "epoch": 1.1028515662191756,
+      "grad_norm": 0.43797495092517036,
+      "learning_rate": 2e-05,
+      "loss": 0.0075,
+      "step": 46333
+    },
+    {
+      "epoch": 1.102875368942207,
+      "grad_norm": 3.573313004690054,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 46334
+    },
+    {
+      "epoch": 1.1028991716652385,
+      "grad_norm": 0.08081223792832051,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46335
+    },
+    {
+      "epoch": 1.10292297438827,
+      "grad_norm": 0.05729890171971518,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46336
+    },
+    {
+      "epoch": 1.1029467771113015,
+      "grad_norm": 0.0265845324151827,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46337
+    },
+    {
+      "epoch": 1.102970579834333,
+      "grad_norm": 0.0831396279207636,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46338
+    },
+    {
+      "epoch": 1.1029943825573645,
+      "grad_norm": 3.364559404349463,
+      "learning_rate": 2e-05,
+      "loss": 0.2438,
+      "step": 46339
+    },
+    {
+      "epoch": 1.103018185280396,
+      "grad_norm": 0.03896845809150736,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46340
+    },
+    {
+      "epoch": 1.1030419880034277,
+      "grad_norm": 0.10016820222122347,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46341
+    },
+    {
+      "epoch": 1.1030657907264592,
+      "grad_norm": 5.60123189795173,
+      "learning_rate": 2e-05,
+      "loss": 0.462,
+      "step": 46342
+    },
+    {
+      "epoch": 1.1030895934494906,
+      "grad_norm": 8.58848522246209,
+      "learning_rate": 2e-05,
+      "loss": 0.6167,
+      "step": 46343
+    },
+    {
+      "epoch": 1.1031133961725221,
+      "grad_norm": 0.0792113511168443,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46344
+    },
+    {
+      "epoch": 1.1031371988955536,
+      "grad_norm": 0.014606111203163996,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46345
+    },
+    {
+      "epoch": 1.103161001618585,
+      "grad_norm": 3.833088162313981,
+      "learning_rate": 2e-05,
+      "loss": 0.2187,
+      "step": 46346
+    },
+    {
+      "epoch": 1.1031848043416166,
+      "grad_norm": 0.2263022268593967,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 46347
+    },
+    {
+      "epoch": 1.1032086070646483,
+      "grad_norm": 0.07182376834552506,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46348
+    },
+    {
+      "epoch": 1.1032324097876798,
+      "grad_norm": 0.0840077423334832,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46349
+    },
+    {
+      "epoch": 1.1032562125107113,
+      "grad_norm": 0.08309445060854784,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46350
+    },
+    {
+      "epoch": 1.1032800152337428,
+      "grad_norm": 5.554065720624401,
+      "learning_rate": 2e-05,
+      "loss": 0.1042,
+      "step": 46351
+    },
+    {
+      "epoch": 1.1033038179567742,
+      "grad_norm": 0.009524808079101523,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46352
+    },
+    {
+      "epoch": 1.1033276206798057,
+      "grad_norm": 8.182604498640611,
+      "learning_rate": 2e-05,
+      "loss": 0.2866,
+      "step": 46353
+    },
+    {
+      "epoch": 1.1033514234028372,
+      "grad_norm": 0.0372356600976082,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46354
+    },
+    {
+      "epoch": 1.103375226125869,
+      "grad_norm": 10.056567548586006,
+      "learning_rate": 2e-05,
+      "loss": 0.4442,
+      "step": 46355
+    },
+    {
+      "epoch": 1.1033990288489004,
+      "grad_norm": 6.042423335411992,
+      "learning_rate": 2e-05,
+      "loss": 0.4201,
+      "step": 46356
+    },
+    {
+      "epoch": 1.1034228315719319,
+      "grad_norm": 2.9232861561200147,
+      "learning_rate": 2e-05,
+      "loss": 0.2796,
+      "step": 46357
+    },
+    {
+      "epoch": 1.1034466342949634,
+      "grad_norm": 0.5498894132616836,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 46358
+    },
+    {
+      "epoch": 1.1034704370179949,
+      "grad_norm": 0.26732412656103294,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 46359
+    },
+    {
+      "epoch": 1.1034942397410263,
+      "grad_norm": 0.4684535802936882,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 46360
+    },
+    {
+      "epoch": 1.1035180424640578,
+      "grad_norm": 0.8454170980718989,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 46361
+    },
+    {
+      "epoch": 1.1035418451870893,
+      "grad_norm": 0.05181135639195774,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46362
+    },
+    {
+      "epoch": 1.103565647910121,
+      "grad_norm": 0.0031722619223548544,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 46363
+    },
+    {
+      "epoch": 1.1035894506331525,
+      "grad_norm": 1.4285439010075356,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 46364
+    },
+    {
+      "epoch": 1.103613253356184,
+      "grad_norm": 0.17906404182320743,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46365
+    },
+    {
+      "epoch": 1.1036370560792155,
+      "grad_norm": 0.9686447874657205,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 46366
+    },
+    {
+      "epoch": 1.103660858802247,
+      "grad_norm": 8.271089466956267,
+      "learning_rate": 2e-05,
+      "loss": 1.2159,
+      "step": 46367
+    },
+    {
+      "epoch": 1.1036846615252784,
+      "grad_norm": 0.008501416459133975,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46368
+    },
+    {
+      "epoch": 1.10370846424831,
+      "grad_norm": 5.363003441546514,
+      "learning_rate": 2e-05,
+      "loss": 0.6613,
+      "step": 46369
+    },
+    {
+      "epoch": 1.1037322669713414,
+      "grad_norm": 4.020684384649597,
+      "learning_rate": 2e-05,
+      "loss": 0.4619,
+      "step": 46370
+    },
+    {
+      "epoch": 1.1037560696943731,
+      "grad_norm": 4.305256575357964,
+      "learning_rate": 2e-05,
+      "loss": 0.4338,
+      "step": 46371
+    },
+    {
+      "epoch": 1.1037798724174046,
+      "grad_norm": 7.44829755661339,
+      "learning_rate": 2e-05,
+      "loss": 0.0753,
+      "step": 46372
+    },
+    {
+      "epoch": 1.103803675140436,
+      "grad_norm": 0.0106020261726522,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46373
+    },
+    {
+      "epoch": 1.1038274778634676,
+      "grad_norm": 1.0480567515513957,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 46374
+    },
+    {
+      "epoch": 1.103851280586499,
+      "grad_norm": 0.00741659220312971,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46375
+    },
+    {
+      "epoch": 1.1038750833095305,
+      "grad_norm": 5.146268606173193,
+      "learning_rate": 2e-05,
+      "loss": 0.22,
+      "step": 46376
+    },
+    {
+      "epoch": 1.103898886032562,
+      "grad_norm": 5.72878754326739,
+      "learning_rate": 2e-05,
+      "loss": 0.8834,
+      "step": 46377
+    },
+    {
+      "epoch": 1.1039226887555937,
+      "grad_norm": 0.07844387787255318,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 46378
+    },
+    {
+      "epoch": 1.1039464914786252,
+      "grad_norm": 0.15243654508630158,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46379
+    },
+    {
+      "epoch": 1.1039702942016567,
+      "grad_norm": 8.183792944645424,
+      "learning_rate": 2e-05,
+      "loss": 0.7421,
+      "step": 46380
+    },
+    {
+      "epoch": 1.1039940969246882,
+      "grad_norm": 1.0752698006693786,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 46381
+    },
+    {
+      "epoch": 1.1040178996477197,
+      "grad_norm": 6.352025527840363,
+      "learning_rate": 2e-05,
+      "loss": 0.189,
+      "step": 46382
+    },
+    {
+      "epoch": 1.1040417023707512,
+      "grad_norm": 6.231762460952157,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 46383
+    },
+    {
+      "epoch": 1.1040655050937827,
+      "grad_norm": 5.859616638634295,
+      "learning_rate": 2e-05,
+      "loss": 0.5434,
+      "step": 46384
+    },
+    {
+      "epoch": 1.1040893078168144,
+      "grad_norm": 5.526867608916892,
+      "learning_rate": 2e-05,
+      "loss": 0.4749,
+      "step": 46385
+    },
+    {
+      "epoch": 1.1041131105398458,
+      "grad_norm": 11.057448504962922,
+      "learning_rate": 2e-05,
+      "loss": 0.1585,
+      "step": 46386
+    },
+    {
+      "epoch": 1.1041369132628773,
+      "grad_norm": 0.005915398660085459,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46387
+    },
+    {
+      "epoch": 1.1041607159859088,
+      "grad_norm": 0.216168908366522,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 46388
+    },
+    {
+      "epoch": 1.1041845187089403,
+      "grad_norm": 12.09026934398842,
+      "learning_rate": 2e-05,
+      "loss": 0.5132,
+      "step": 46389
+    },
+    {
+      "epoch": 1.1042083214319718,
+      "grad_norm": 1.3847420967774802,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 46390
+    },
+    {
+      "epoch": 1.1042321241550033,
+      "grad_norm": 0.07364692769076035,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46391
+    },
+    {
+      "epoch": 1.1042559268780348,
+      "grad_norm": 5.889044298751456,
+      "learning_rate": 2e-05,
+      "loss": 0.5828,
+      "step": 46392
+    },
+    {
+      "epoch": 1.1042797296010665,
+      "grad_norm": 0.011838777976680362,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46393
+    },
+    {
+      "epoch": 1.104303532324098,
+      "grad_norm": 5.178999049265101,
+      "learning_rate": 2e-05,
+      "loss": 0.2866,
+      "step": 46394
+    },
+    {
+      "epoch": 1.1043273350471294,
+      "grad_norm": 0.14511319986065419,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46395
+    },
+    {
+      "epoch": 1.104351137770161,
+      "grad_norm": 0.015007042997521688,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46396
+    },
+    {
+      "epoch": 1.1043749404931924,
+      "grad_norm": 3.0583451272375513,
+      "learning_rate": 2e-05,
+      "loss": 0.315,
+      "step": 46397
+    },
+    {
+      "epoch": 1.1043987432162239,
+      "grad_norm": 6.4226003383443135,
+      "learning_rate": 2e-05,
+      "loss": 0.3669,
+      "step": 46398
+    },
+    {
+      "epoch": 1.1044225459392554,
+      "grad_norm": 2.0432946277256527,
+      "learning_rate": 2e-05,
+      "loss": 0.0476,
+      "step": 46399
+    },
+    {
+      "epoch": 1.1044463486622869,
+      "grad_norm": 20.865021412775846,
+      "learning_rate": 2e-05,
+      "loss": 1.134,
+      "step": 46400
+    },
+    {
+      "epoch": 1.1044701513853186,
+      "grad_norm": 4.63890495054683,
+      "learning_rate": 2e-05,
+      "loss": 0.4835,
+      "step": 46401
+    },
+    {
+      "epoch": 1.10449395410835,
+      "grad_norm": 2.0835055723187197,
+      "learning_rate": 2e-05,
+      "loss": 0.0168,
+      "step": 46402
+    },
+    {
+      "epoch": 1.1045177568313815,
+      "grad_norm": 3.389444595794532,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 46403
+    },
+    {
+      "epoch": 1.104541559554413,
+      "grad_norm": 0.03456291864183123,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46404
+    },
+    {
+      "epoch": 1.1045653622774445,
+      "grad_norm": 7.79611000942407,
+      "learning_rate": 2e-05,
+      "loss": 0.9477,
+      "step": 46405
+    },
+    {
+      "epoch": 1.104589165000476,
+      "grad_norm": 4.594087379089141,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 46406
+    },
+    {
+      "epoch": 1.1046129677235075,
+      "grad_norm": 2.664479742710523,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 46407
+    },
+    {
+      "epoch": 1.1046367704465392,
+      "grad_norm": 3.310491583645931,
+      "learning_rate": 2e-05,
+      "loss": 0.0288,
+      "step": 46408
+    },
+    {
+      "epoch": 1.1046605731695707,
+      "grad_norm": 4.1244626042108825,
+      "learning_rate": 2e-05,
+      "loss": 0.1648,
+      "step": 46409
+    },
+    {
+      "epoch": 1.1046843758926022,
+      "grad_norm": 8.976897227946669,
+      "learning_rate": 2e-05,
+      "loss": 0.6889,
+      "step": 46410
+    },
+    {
+      "epoch": 1.1047081786156336,
+      "grad_norm": 4.395302943488702,
+      "learning_rate": 2e-05,
+      "loss": 0.2203,
+      "step": 46411
+    },
+    {
+      "epoch": 1.1047319813386651,
+      "grad_norm": 0.004812536912008707,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46412
+    },
+    {
+      "epoch": 1.1047557840616966,
+      "grad_norm": 0.600462332562383,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 46413
+    },
+    {
+      "epoch": 1.104779586784728,
+      "grad_norm": 3.8895678362450177,
+      "learning_rate": 2e-05,
+      "loss": 0.1668,
+      "step": 46414
+    },
+    {
+      "epoch": 1.1048033895077598,
+      "grad_norm": 7.310537843664085,
+      "learning_rate": 2e-05,
+      "loss": 0.4083,
+      "step": 46415
+    },
+    {
+      "epoch": 1.1048271922307913,
+      "grad_norm": 2.9657144011547323,
+      "learning_rate": 2e-05,
+      "loss": 0.0568,
+      "step": 46416
+    },
+    {
+      "epoch": 1.1048509949538228,
+      "grad_norm": 0.13540530393029518,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 46417
+    },
+    {
+      "epoch": 1.1048747976768543,
+      "grad_norm": 2.78456566285233,
+      "learning_rate": 2e-05,
+      "loss": 0.3708,
+      "step": 46418
+    },
+    {
+      "epoch": 1.1048986003998857,
+      "grad_norm": 5.658638405265789,
+      "learning_rate": 2e-05,
+      "loss": 0.104,
+      "step": 46419
+    },
+    {
+      "epoch": 1.1049224031229172,
+      "grad_norm": 2.5023061192425335,
+      "learning_rate": 2e-05,
+      "loss": 0.0824,
+      "step": 46420
+    },
+    {
+      "epoch": 1.1049462058459487,
+      "grad_norm": 0.12273178725378661,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 46421
+    },
+    {
+      "epoch": 1.1049700085689802,
+      "grad_norm": 0.309943347679687,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 46422
+    },
+    {
+      "epoch": 1.104993811292012,
+      "grad_norm": 3.128026089400186,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 46423
+    },
+    {
+      "epoch": 1.1050176140150434,
+      "grad_norm": 2.3977706983813114,
+      "learning_rate": 2e-05,
+      "loss": 0.0289,
+      "step": 46424
+    },
+    {
+      "epoch": 1.1050414167380749,
+      "grad_norm": 0.04747025253423831,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46425
+    },
+    {
+      "epoch": 1.1050652194611064,
+      "grad_norm": 3.177262517533186,
+      "learning_rate": 2e-05,
+      "loss": 0.2432,
+      "step": 46426
+    },
+    {
+      "epoch": 1.1050890221841378,
+      "grad_norm": 2.3158087044128273,
+      "learning_rate": 2e-05,
+      "loss": 0.0371,
+      "step": 46427
+    },
+    {
+      "epoch": 1.1051128249071693,
+      "grad_norm": 0.022671640611457364,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46428
+    },
+    {
+      "epoch": 1.1051366276302008,
+      "grad_norm": 2.4797396963730316,
+      "learning_rate": 2e-05,
+      "loss": 0.3201,
+      "step": 46429
+    },
+    {
+      "epoch": 1.1051604303532323,
+      "grad_norm": 7.1971738727035435,
+      "learning_rate": 2e-05,
+      "loss": 0.3685,
+      "step": 46430
+    },
+    {
+      "epoch": 1.105184233076264,
+      "grad_norm": 11.868037610498588,
+      "learning_rate": 2e-05,
+      "loss": 0.1277,
+      "step": 46431
+    },
+    {
+      "epoch": 1.1052080357992955,
+      "grad_norm": 9.755201436612237,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 46432
+    },
+    {
+      "epoch": 1.105231838522327,
+      "grad_norm": 1.4677296235160742,
+      "learning_rate": 2e-05,
+      "loss": 0.0171,
+      "step": 46433
+    },
+    {
+      "epoch": 1.1052556412453585,
+      "grad_norm": 7.416261754969928,
+      "learning_rate": 2e-05,
+      "loss": 0.3109,
+      "step": 46434
+    },
+    {
+      "epoch": 1.10527944396839,
+      "grad_norm": 3.6749284400849263,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 46435
+    },
+    {
+      "epoch": 1.1053032466914214,
+      "grad_norm": 6.577873430308312,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 46436
+    },
+    {
+      "epoch": 1.105327049414453,
+      "grad_norm": 4.852296903459684,
+      "learning_rate": 2e-05,
+      "loss": 0.5487,
+      "step": 46437
+    },
+    {
+      "epoch": 1.1053508521374846,
+      "grad_norm": 0.017375270395460732,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46438
+    },
+    {
+      "epoch": 1.1053746548605161,
+      "grad_norm": 2.656254498392279,
+      "learning_rate": 2e-05,
+      "loss": 0.249,
+      "step": 46439
+    },
+    {
+      "epoch": 1.1053984575835476,
+      "grad_norm": 7.160756927965441,
+      "learning_rate": 2e-05,
+      "loss": 1.1015,
+      "step": 46440
+    },
+    {
+      "epoch": 1.105422260306579,
+      "grad_norm": 0.2886972789305163,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 46441
+    },
+    {
+      "epoch": 1.1054460630296106,
+      "grad_norm": 0.05245411770237049,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46442
+    },
+    {
+      "epoch": 1.105469865752642,
+      "grad_norm": 3.2293108150308885,
+      "learning_rate": 2e-05,
+      "loss": 0.2775,
+      "step": 46443
+    },
+    {
+      "epoch": 1.1054936684756735,
+      "grad_norm": 0.4102434917538791,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 46444
+    },
+    {
+      "epoch": 1.1055174711987052,
+      "grad_norm": 4.895739998713172,
+      "learning_rate": 2e-05,
+      "loss": 0.4656,
+      "step": 46445
+    },
+    {
+      "epoch": 1.1055412739217367,
+      "grad_norm": 4.177695125711693,
+      "learning_rate": 2e-05,
+      "loss": 0.4622,
+      "step": 46446
+    },
+    {
+      "epoch": 1.1055650766447682,
+      "grad_norm": 0.05451893551555748,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46447
+    },
+    {
+      "epoch": 1.1055888793677997,
+      "grad_norm": 2.3618708317156902,
+      "learning_rate": 2e-05,
+      "loss": 0.3581,
+      "step": 46448
+    },
+    {
+      "epoch": 1.1056126820908312,
+      "grad_norm": 0.2822214100603205,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 46449
+    },
+    {
+      "epoch": 1.1056364848138627,
+      "grad_norm": 0.028427356683440697,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46450
+    },
+    {
+      "epoch": 1.1056602875368942,
+      "grad_norm": 0.2817266025392406,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 46451
+    },
+    {
+      "epoch": 1.1056840902599256,
+      "grad_norm": 3.952872066938306,
+      "learning_rate": 2e-05,
+      "loss": 0.3892,
+      "step": 46452
+    },
+    {
+      "epoch": 1.1057078929829574,
+      "grad_norm": 8.863905314699604,
+      "learning_rate": 2e-05,
+      "loss": 1.7335,
+      "step": 46453
+    },
+    {
+      "epoch": 1.1057316957059888,
+      "grad_norm": 3.538466664519448,
+      "learning_rate": 2e-05,
+      "loss": 0.4557,
+      "step": 46454
+    },
+    {
+      "epoch": 1.1057554984290203,
+      "grad_norm": 0.7231577123631667,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 46455
+    },
+    {
+      "epoch": 1.1057793011520518,
+      "grad_norm": 4.799966588269547,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 46456
+    },
+    {
+      "epoch": 1.1058031038750833,
+      "grad_norm": 6.567246638034832,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 46457
+    },
+    {
+      "epoch": 1.1058269065981148,
+      "grad_norm": 4.778355745760493,
+      "learning_rate": 2e-05,
+      "loss": 0.714,
+      "step": 46458
+    },
+    {
+      "epoch": 1.1058507093211463,
+      "grad_norm": 25.640461828188158,
+      "learning_rate": 2e-05,
+      "loss": 0.8903,
+      "step": 46459
+    },
+    {
+      "epoch": 1.1058745120441777,
+      "grad_norm": 1.018094803086887,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 46460
+    },
+    {
+      "epoch": 1.1058983147672095,
+      "grad_norm": 0.0069050389830073835,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46461
+    },
+    {
+      "epoch": 1.105922117490241,
+      "grad_norm": 3.4565765287305164,
+      "learning_rate": 2e-05,
+      "loss": 0.1297,
+      "step": 46462
+    },
+    {
+      "epoch": 1.1059459202132724,
+      "grad_norm": 2.6538521547083223,
+      "learning_rate": 2e-05,
+      "loss": 0.0224,
+      "step": 46463
+    },
+    {
+      "epoch": 1.105969722936304,
+      "grad_norm": 3.5385232817489656,
+      "learning_rate": 2e-05,
+      "loss": 0.4856,
+      "step": 46464
+    },
+    {
+      "epoch": 1.1059935256593354,
+      "grad_norm": 3.5425287967758705,
+      "learning_rate": 2e-05,
+      "loss": 0.3941,
+      "step": 46465
+    },
+    {
+      "epoch": 1.1060173283823669,
+      "grad_norm": 0.32118226444180414,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 46466
+    },
+    {
+      "epoch": 1.1060411311053984,
+      "grad_norm": 0.5022706455238991,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 46467
+    },
+    {
+      "epoch": 1.10606493382843,
+      "grad_norm": 10.997067667815264,
+      "learning_rate": 2e-05,
+      "loss": 0.7093,
+      "step": 46468
+    },
+    {
+      "epoch": 1.1060887365514616,
+      "grad_norm": 8.131172780397037,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 46469
+    },
+    {
+      "epoch": 1.106112539274493,
+      "grad_norm": 2.3747158244701256,
+      "learning_rate": 2e-05,
+      "loss": 0.0335,
+      "step": 46470
+    },
+    {
+      "epoch": 1.1061363419975245,
+      "grad_norm": 0.018954220699142862,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46471
+    },
+    {
+      "epoch": 1.106160144720556,
+      "grad_norm": 0.742942705864624,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 46472
+    },
+    {
+      "epoch": 1.1061839474435875,
+      "grad_norm": 0.49717616222439115,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46473
+    },
+    {
+      "epoch": 1.106207750166619,
+      "grad_norm": 6.513709819587393,
+      "learning_rate": 2e-05,
+      "loss": 0.8473,
+      "step": 46474
+    },
+    {
+      "epoch": 1.1062315528896507,
+      "grad_norm": 0.015025676680417344,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46475
+    },
+    {
+      "epoch": 1.1062553556126822,
+      "grad_norm": 6.313931504479485,
+      "learning_rate": 2e-05,
+      "loss": 0.1496,
+      "step": 46476
+    },
+    {
+      "epoch": 1.1062791583357137,
+      "grad_norm": 0.8716829947438168,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 46477
+    },
+    {
+      "epoch": 1.1063029610587451,
+      "grad_norm": 0.1252019748935703,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 46478
+    },
+    {
+      "epoch": 1.1063267637817766,
+      "grad_norm": 3.8812163669491433,
+      "learning_rate": 2e-05,
+      "loss": 0.1615,
+      "step": 46479
+    },
+    {
+      "epoch": 1.1063505665048081,
+      "grad_norm": 2.4872725693992797,
+      "learning_rate": 2e-05,
+      "loss": 0.182,
+      "step": 46480
+    },
+    {
+      "epoch": 1.1063743692278396,
+      "grad_norm": 0.19451183208229472,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46481
+    },
+    {
+      "epoch": 1.106398171950871,
+      "grad_norm": 4.7538362401237215,
+      "learning_rate": 2e-05,
+      "loss": 0.7528,
+      "step": 46482
+    },
+    {
+      "epoch": 1.1064219746739028,
+      "grad_norm": 1.7838099322181058,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 46483
+    },
+    {
+      "epoch": 1.1064457773969343,
+      "grad_norm": 4.49729888425978,
+      "learning_rate": 2e-05,
+      "loss": 0.0805,
+      "step": 46484
+    },
+    {
+      "epoch": 1.1064695801199658,
+      "grad_norm": 2.331337848645111,
+      "learning_rate": 2e-05,
+      "loss": 0.0392,
+      "step": 46485
+    },
+    {
+      "epoch": 1.1064933828429973,
+      "grad_norm": 5.4402990493690675,
+      "learning_rate": 2e-05,
+      "loss": 0.1096,
+      "step": 46486
+    },
+    {
+      "epoch": 1.1065171855660287,
+      "grad_norm": 10.283042945644128,
+      "learning_rate": 2e-05,
+      "loss": 0.8204,
+      "step": 46487
+    },
+    {
+      "epoch": 1.1065409882890602,
+      "grad_norm": 3.553698950506131,
+      "learning_rate": 2e-05,
+      "loss": 0.0739,
+      "step": 46488
+    },
+    {
+      "epoch": 1.1065647910120917,
+      "grad_norm": 6.072032774318332,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 46489
+    },
+    {
+      "epoch": 1.1065885937351232,
+      "grad_norm": 12.16358744308644,
+      "learning_rate": 2e-05,
+      "loss": 0.7798,
+      "step": 46490
+    },
+    {
+      "epoch": 1.106612396458155,
+      "grad_norm": 9.363235828806829,
+      "learning_rate": 2e-05,
+      "loss": 1.0516,
+      "step": 46491
+    },
+    {
+      "epoch": 1.1066361991811864,
+      "grad_norm": 0.458896621516669,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 46492
+    },
+    {
+      "epoch": 1.1066600019042179,
+      "grad_norm": 2.3404365801407,
+      "learning_rate": 2e-05,
+      "loss": 0.1092,
+      "step": 46493
+    },
+    {
+      "epoch": 1.1066838046272494,
+      "grad_norm": 0.3666681092408414,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 46494
+    },
+    {
+      "epoch": 1.1067076073502808,
+      "grad_norm": 1.299642369869688,
+      "learning_rate": 2e-05,
+      "loss": 0.0143,
+      "step": 46495
+    },
+    {
+      "epoch": 1.1067314100733123,
+      "grad_norm": 2.8372668253012705,
+      "learning_rate": 2e-05,
+      "loss": 0.0394,
+      "step": 46496
+    },
+    {
+      "epoch": 1.1067552127963438,
+      "grad_norm": 3.7084934171302475,
+      "learning_rate": 2e-05,
+      "loss": 0.4749,
+      "step": 46497
+    },
+    {
+      "epoch": 1.1067790155193755,
+      "grad_norm": 0.23324773694952727,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 46498
+    },
+    {
+      "epoch": 1.106802818242407,
+      "grad_norm": 0.025922320814727902,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46499
+    },
+    {
+      "epoch": 1.1068266209654385,
+      "grad_norm": 0.1389276731462342,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46500
+    },
+    {
+      "epoch": 1.10685042368847,
+      "grad_norm": 10.319736877314158,
+      "learning_rate": 2e-05,
+      "loss": 1.3053,
+      "step": 46501
+    },
+    {
+      "epoch": 1.1068742264115015,
+      "grad_norm": 4.059276208610704,
+      "learning_rate": 2e-05,
+      "loss": 0.3918,
+      "step": 46502
+    },
+    {
+      "epoch": 1.106898029134533,
+      "grad_norm": 5.569988604832132,
+      "learning_rate": 2e-05,
+      "loss": 0.9295,
+      "step": 46503
+    },
+    {
+      "epoch": 1.1069218318575644,
+      "grad_norm": 5.253230433327308,
+      "learning_rate": 2e-05,
+      "loss": 1.1509,
+      "step": 46504
+    },
+    {
+      "epoch": 1.106945634580596,
+      "grad_norm": 0.04216906253013237,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46505
+    },
+    {
+      "epoch": 1.1069694373036276,
+      "grad_norm": 3.8403888587454102,
+      "learning_rate": 2e-05,
+      "loss": 0.2625,
+      "step": 46506
+    },
+    {
+      "epoch": 1.106993240026659,
+      "grad_norm": 3.3113308044742675,
+      "learning_rate": 2e-05,
+      "loss": 0.2505,
+      "step": 46507
+    },
+    {
+      "epoch": 1.1070170427496906,
+      "grad_norm": 0.6703706235461963,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 46508
+    },
+    {
+      "epoch": 1.107040845472722,
+      "grad_norm": 6.426251129586686,
+      "learning_rate": 2e-05,
+      "loss": 0.7387,
+      "step": 46509
+    },
+    {
+      "epoch": 1.1070646481957536,
+      "grad_norm": 0.11411394377570386,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 46510
+    },
+    {
+      "epoch": 1.107088450918785,
+      "grad_norm": 15.491620499674378,
+      "learning_rate": 2e-05,
+      "loss": 0.1102,
+      "step": 46511
+    },
+    {
+      "epoch": 1.1071122536418165,
+      "grad_norm": 0.05263029669385063,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46512
+    },
+    {
+      "epoch": 1.1071360563648482,
+      "grad_norm": 3.2820964720338424,
+      "learning_rate": 2e-05,
+      "loss": 0.1829,
+      "step": 46513
+    },
+    {
+      "epoch": 1.1071598590878797,
+      "grad_norm": 4.280165401820944,
+      "learning_rate": 2e-05,
+      "loss": 0.4524,
+      "step": 46514
+    },
+    {
+      "epoch": 1.1071836618109112,
+      "grad_norm": 0.5039248156388502,
+      "learning_rate": 2e-05,
+      "loss": 0.0078,
+      "step": 46515
+    },
+    {
+      "epoch": 1.1072074645339427,
+      "grad_norm": 2.0125265870428835,
+      "learning_rate": 2e-05,
+      "loss": 0.0148,
+      "step": 46516
+    },
+    {
+      "epoch": 1.1072312672569742,
+      "grad_norm": 3.395238792641586,
+      "learning_rate": 2e-05,
+      "loss": 0.3563,
+      "step": 46517
+    },
+    {
+      "epoch": 1.1072550699800057,
+      "grad_norm": 3.5461268462237476,
+      "learning_rate": 2e-05,
+      "loss": 0.1126,
+      "step": 46518
+    },
+    {
+      "epoch": 1.1072788727030372,
+      "grad_norm": 0.13069607304963016,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46519
+    },
+    {
+      "epoch": 1.1073026754260686,
+      "grad_norm": 15.079650448025719,
+      "learning_rate": 2e-05,
+      "loss": 0.8685,
+      "step": 46520
+    },
+    {
+      "epoch": 1.1073264781491003,
+      "grad_norm": 26.855625659379584,
+      "learning_rate": 2e-05,
+      "loss": 0.2841,
+      "step": 46521
+    },
+    {
+      "epoch": 1.1073502808721318,
+      "grad_norm": 0.027668646231072305,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46522
+    },
+    {
+      "epoch": 1.1073740835951633,
+      "grad_norm": 0.9769254422740531,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 46523
+    },
+    {
+      "epoch": 1.1073978863181948,
+      "grad_norm": 1.0579221634851455,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 46524
+    },
+    {
+      "epoch": 1.1074216890412263,
+      "grad_norm": 3.8424016053129635,
+      "learning_rate": 2e-05,
+      "loss": 0.4914,
+      "step": 46525
+    },
+    {
+      "epoch": 1.1074454917642578,
+      "grad_norm": 0.8907145616253445,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 46526
+    },
+    {
+      "epoch": 1.1074692944872893,
+      "grad_norm": 8.428473555449317,
+      "learning_rate": 2e-05,
+      "loss": 0.2181,
+      "step": 46527
+    },
+    {
+      "epoch": 1.107493097210321,
+      "grad_norm": 3.4808815817592302,
+      "learning_rate": 2e-05,
+      "loss": 0.0429,
+      "step": 46528
+    },
+    {
+      "epoch": 1.1075168999333524,
+      "grad_norm": 0.09138111812280834,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 46529
+    },
+    {
+      "epoch": 1.107540702656384,
+      "grad_norm": 3.0738139890940275,
+      "learning_rate": 2e-05,
+      "loss": 0.0276,
+      "step": 46530
+    },
+    {
+      "epoch": 1.1075645053794154,
+      "grad_norm": 3.9841624711139763,
+      "learning_rate": 2e-05,
+      "loss": 0.6733,
+      "step": 46531
+    },
+    {
+      "epoch": 1.107588308102447,
+      "grad_norm": 8.294417136667327,
+      "learning_rate": 2e-05,
+      "loss": 0.3993,
+      "step": 46532
+    },
+    {
+      "epoch": 1.1076121108254784,
+      "grad_norm": 12.406462068046169,
+      "learning_rate": 2e-05,
+      "loss": 0.5881,
+      "step": 46533
+    },
+    {
+      "epoch": 1.1076359135485099,
+      "grad_norm": 0.10108067426772648,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 46534
+    },
+    {
+      "epoch": 1.1076597162715414,
+      "grad_norm": 10.586814741347135,
+      "learning_rate": 2e-05,
+      "loss": 0.3903,
+      "step": 46535
+    },
+    {
+      "epoch": 1.107683518994573,
+      "grad_norm": 7.596265668463237,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 46536
+    },
+    {
+      "epoch": 1.1077073217176046,
+      "grad_norm": 5.156702449900755,
+      "learning_rate": 2e-05,
+      "loss": 0.2401,
+      "step": 46537
+    },
+    {
+      "epoch": 1.107731124440636,
+      "grad_norm": 3.4600382840007025,
+      "learning_rate": 2e-05,
+      "loss": 0.3055,
+      "step": 46538
+    },
+    {
+      "epoch": 1.1077549271636675,
+      "grad_norm": 2.1781914741314092,
+      "learning_rate": 2e-05,
+      "loss": 0.2329,
+      "step": 46539
+    },
+    {
+      "epoch": 1.107778729886699,
+      "grad_norm": 12.529338643002392,
+      "learning_rate": 2e-05,
+      "loss": 0.9847,
+      "step": 46540
+    },
+    {
+      "epoch": 1.1078025326097305,
+      "grad_norm": 0.045370227032907356,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46541
+    },
+    {
+      "epoch": 1.107826335332762,
+      "grad_norm": 8.289444187908309,
+      "learning_rate": 2e-05,
+      "loss": 0.8737,
+      "step": 46542
+    },
+    {
+      "epoch": 1.1078501380557935,
+      "grad_norm": 9.022605447066953,
+      "learning_rate": 2e-05,
+      "loss": 0.1269,
+      "step": 46543
+    },
+    {
+      "epoch": 1.1078739407788252,
+      "grad_norm": 0.11226907515711151,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46544
+    },
+    {
+      "epoch": 1.1078977435018567,
+      "grad_norm": 0.15250693642605664,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 46545
+    },
+    {
+      "epoch": 1.1079215462248881,
+      "grad_norm": 3.4352618531720727,
+      "learning_rate": 2e-05,
+      "loss": 0.4168,
+      "step": 46546
+    },
+    {
+      "epoch": 1.1079453489479196,
+      "grad_norm": 4.425032004234761,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 46547
+    },
+    {
+      "epoch": 1.1079691516709511,
+      "grad_norm": 3.8401524045766755,
+      "learning_rate": 2e-05,
+      "loss": 0.2985,
+      "step": 46548
+    },
+    {
+      "epoch": 1.1079929543939826,
+      "grad_norm": 5.172359947642147,
+      "learning_rate": 2e-05,
+      "loss": 0.3901,
+      "step": 46549
+    },
+    {
+      "epoch": 1.108016757117014,
+      "grad_norm": 3.904035608823269,
+      "learning_rate": 2e-05,
+      "loss": 0.2116,
+      "step": 46550
+    },
+    {
+      "epoch": 1.1080405598400458,
+      "grad_norm": 0.2922215601065073,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46551
+    },
+    {
+      "epoch": 1.1080643625630773,
+      "grad_norm": 1.0493612901423641,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 46552
+    },
+    {
+      "epoch": 1.1080881652861088,
+      "grad_norm": 8.712472594090562,
+      "learning_rate": 2e-05,
+      "loss": 0.2964,
+      "step": 46553
+    },
+    {
+      "epoch": 1.1081119680091402,
+      "grad_norm": 4.203955748506031,
+      "learning_rate": 2e-05,
+      "loss": 0.4724,
+      "step": 46554
+    },
+    {
+      "epoch": 1.1081357707321717,
+      "grad_norm": 0.07667075483911831,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46555
+    },
+    {
+      "epoch": 1.1081595734552032,
+      "grad_norm": 0.03363480503789434,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46556
+    },
+    {
+      "epoch": 1.1081833761782347,
+      "grad_norm": 4.422510640304411,
+      "learning_rate": 2e-05,
+      "loss": 0.3014,
+      "step": 46557
+    },
+    {
+      "epoch": 1.1082071789012664,
+      "grad_norm": 0.37589318139194056,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 46558
+    },
+    {
+      "epoch": 1.108230981624298,
+      "grad_norm": 0.009143457525725433,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46559
+    },
+    {
+      "epoch": 1.1082547843473294,
+      "grad_norm": 3.751496711313371,
+      "learning_rate": 2e-05,
+      "loss": 0.052,
+      "step": 46560
+    },
+    {
+      "epoch": 1.1082785870703609,
+      "grad_norm": 3.7445460034937996,
+      "learning_rate": 2e-05,
+      "loss": 0.3846,
+      "step": 46561
+    },
+    {
+      "epoch": 1.1083023897933924,
+      "grad_norm": 3.8671116584755825,
+      "learning_rate": 2e-05,
+      "loss": 0.105,
+      "step": 46562
+    },
+    {
+      "epoch": 1.1083261925164238,
+      "grad_norm": 0.08495260383049638,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 46563
+    },
+    {
+      "epoch": 1.1083499952394553,
+      "grad_norm": 0.2727375694024449,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 46564
+    },
+    {
+      "epoch": 1.1083737979624868,
+      "grad_norm": 3.723906064016535,
+      "learning_rate": 2e-05,
+      "loss": 0.1145,
+      "step": 46565
+    },
+    {
+      "epoch": 1.1083976006855185,
+      "grad_norm": 12.895342451503382,
+      "learning_rate": 2e-05,
+      "loss": 0.4241,
+      "step": 46566
+    },
+    {
+      "epoch": 1.10842140340855,
+      "grad_norm": 2.7900106141941308,
+      "learning_rate": 2e-05,
+      "loss": 0.2757,
+      "step": 46567
+    },
+    {
+      "epoch": 1.1084452061315815,
+      "grad_norm": 4.433902265661682,
+      "learning_rate": 2e-05,
+      "loss": 0.3955,
+      "step": 46568
+    },
+    {
+      "epoch": 1.108469008854613,
+      "grad_norm": 0.036558945613994115,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46569
+    },
+    {
+      "epoch": 1.1084928115776445,
+      "grad_norm": 3.5532754918650333,
+      "learning_rate": 2e-05,
+      "loss": 0.0385,
+      "step": 46570
+    },
+    {
+      "epoch": 1.108516614300676,
+      "grad_norm": 5.332897054927856,
+      "learning_rate": 2e-05,
+      "loss": 0.5431,
+      "step": 46571
+    },
+    {
+      "epoch": 1.1085404170237074,
+      "grad_norm": 4.490164315488176,
+      "learning_rate": 2e-05,
+      "loss": 0.3194,
+      "step": 46572
+    },
+    {
+      "epoch": 1.108564219746739,
+      "grad_norm": 2.6893120790717595,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 46573
+    },
+    {
+      "epoch": 1.1085880224697706,
+      "grad_norm": 0.9752579495334863,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 46574
+    },
+    {
+      "epoch": 1.108611825192802,
+      "grad_norm": 0.29467947399274075,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 46575
+    },
+    {
+      "epoch": 1.1086356279158336,
+      "grad_norm": 1.01701783788168,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 46576
+    },
+    {
+      "epoch": 1.108659430638865,
+      "grad_norm": 3.0932961488711284,
+      "learning_rate": 2e-05,
+      "loss": 0.4701,
+      "step": 46577
+    },
+    {
+      "epoch": 1.1086832333618966,
+      "grad_norm": 7.718129746625833,
+      "learning_rate": 2e-05,
+      "loss": 0.5134,
+      "step": 46578
+    },
+    {
+      "epoch": 1.108707036084928,
+      "grad_norm": 3.8472611004672137,
+      "learning_rate": 2e-05,
+      "loss": 0.4248,
+      "step": 46579
+    },
+    {
+      "epoch": 1.1087308388079595,
+      "grad_norm": 1.286827838141207,
+      "learning_rate": 2e-05,
+      "loss": 0.0198,
+      "step": 46580
+    },
+    {
+      "epoch": 1.1087546415309912,
+      "grad_norm": 3.893358114817352,
+      "learning_rate": 2e-05,
+      "loss": 0.2408,
+      "step": 46581
+    },
+    {
+      "epoch": 1.1087784442540227,
+      "grad_norm": 3.346101189062095,
+      "learning_rate": 2e-05,
+      "loss": 0.3251,
+      "step": 46582
+    },
+    {
+      "epoch": 1.1088022469770542,
+      "grad_norm": 11.861981678865138,
+      "learning_rate": 2e-05,
+      "loss": 0.4436,
+      "step": 46583
+    },
+    {
+      "epoch": 1.1088260497000857,
+      "grad_norm": 3.278158456791114,
+      "learning_rate": 2e-05,
+      "loss": 0.3078,
+      "step": 46584
+    },
+    {
+      "epoch": 1.1088498524231172,
+      "grad_norm": 13.413489013560335,
+      "learning_rate": 2e-05,
+      "loss": 0.0795,
+      "step": 46585
+    },
+    {
+      "epoch": 1.1088736551461487,
+      "grad_norm": 1.9602111764470667,
+      "learning_rate": 2e-05,
+      "loss": 0.1523,
+      "step": 46586
+    },
+    {
+      "epoch": 1.1088974578691801,
+      "grad_norm": 3.3970510259255566,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 46587
+    },
+    {
+      "epoch": 1.1089212605922119,
+      "grad_norm": 8.248608088271428,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 46588
+    },
+    {
+      "epoch": 1.1089450633152433,
+      "grad_norm": 9.922342769443423,
+      "learning_rate": 2e-05,
+      "loss": 0.1602,
+      "step": 46589
+    },
+    {
+      "epoch": 1.1089688660382748,
+      "grad_norm": 0.9621184143741512,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 46590
+    },
+    {
+      "epoch": 1.1089926687613063,
+      "grad_norm": 3.8252872117804597,
+      "learning_rate": 2e-05,
+      "loss": 0.0633,
+      "step": 46591
+    },
+    {
+      "epoch": 1.1090164714843378,
+      "grad_norm": 2.244794092366305,
+      "learning_rate": 2e-05,
+      "loss": 0.0277,
+      "step": 46592
+    },
+    {
+      "epoch": 1.1090402742073693,
+      "grad_norm": 13.765806496822599,
+      "learning_rate": 2e-05,
+      "loss": 0.8224,
+      "step": 46593
+    },
+    {
+      "epoch": 1.1090640769304008,
+      "grad_norm": 4.59286475896293,
+      "learning_rate": 2e-05,
+      "loss": 0.6704,
+      "step": 46594
+    },
+    {
+      "epoch": 1.1090878796534323,
+      "grad_norm": 1.202151867936154,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 46595
+    },
+    {
+      "epoch": 1.109111682376464,
+      "grad_norm": 4.7349219384531525,
+      "learning_rate": 2e-05,
+      "loss": 0.5163,
+      "step": 46596
+    },
+    {
+      "epoch": 1.1091354850994954,
+      "grad_norm": 5.154999640996011,
+      "learning_rate": 2e-05,
+      "loss": 0.2742,
+      "step": 46597
+    },
+    {
+      "epoch": 1.109159287822527,
+      "grad_norm": 6.863553119535747,
+      "learning_rate": 2e-05,
+      "loss": 0.4531,
+      "step": 46598
+    },
+    {
+      "epoch": 1.1091830905455584,
+      "grad_norm": 3.5325955498236006,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 46599
+    },
+    {
+      "epoch": 1.10920689326859,
+      "grad_norm": 2.485750178527968,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 46600
+    },
+    {
+      "epoch": 1.1092306959916214,
+      "grad_norm": 6.454701901296223,
+      "learning_rate": 2e-05,
+      "loss": 0.5303,
+      "step": 46601
+    },
+    {
+      "epoch": 1.1092544987146529,
+      "grad_norm": 1.8878171722122254,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 46602
+    },
+    {
+      "epoch": 1.1092783014376844,
+      "grad_norm": 4.847897755802959,
+      "learning_rate": 2e-05,
+      "loss": 0.3746,
+      "step": 46603
+    },
+    {
+      "epoch": 1.109302104160716,
+      "grad_norm": 3.5388685182077495,
+      "learning_rate": 2e-05,
+      "loss": 0.089,
+      "step": 46604
+    },
+    {
+      "epoch": 1.1093259068837475,
+      "grad_norm": 5.765606464250518,
+      "learning_rate": 2e-05,
+      "loss": 0.4669,
+      "step": 46605
+    },
+    {
+      "epoch": 1.109349709606779,
+      "grad_norm": 0.86517269064584,
+      "learning_rate": 2e-05,
+      "loss": 0.0048,
+      "step": 46606
+    },
+    {
+      "epoch": 1.1093735123298105,
+      "grad_norm": 5.431526863185205,
+      "learning_rate": 2e-05,
+      "loss": 0.5552,
+      "step": 46607
+    },
+    {
+      "epoch": 1.109397315052842,
+      "grad_norm": 5.857499389375628,
+      "learning_rate": 2e-05,
+      "loss": 0.1079,
+      "step": 46608
+    },
+    {
+      "epoch": 1.1094211177758735,
+      "grad_norm": 15.115440259243769,
+      "learning_rate": 2e-05,
+      "loss": 0.3449,
+      "step": 46609
+    },
+    {
+      "epoch": 1.109444920498905,
+      "grad_norm": 1.5824172353594737,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 46610
+    },
+    {
+      "epoch": 1.1094687232219367,
+      "grad_norm": 0.7766759371290605,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 46611
+    },
+    {
+      "epoch": 1.1094925259449682,
+      "grad_norm": 7.963803658886265,
+      "learning_rate": 2e-05,
+      "loss": 0.62,
+      "step": 46612
+    },
+    {
+      "epoch": 1.1095163286679997,
+      "grad_norm": 0.7716890582141206,
+      "learning_rate": 2e-05,
+      "loss": 0.0119,
+      "step": 46613
+    },
+    {
+      "epoch": 1.1095401313910311,
+      "grad_norm": 6.1203879755924,
+      "learning_rate": 2e-05,
+      "loss": 0.583,
+      "step": 46614
+    },
+    {
+      "epoch": 1.1095639341140626,
+      "grad_norm": 3.694903725948332,
+      "learning_rate": 2e-05,
+      "loss": 0.073,
+      "step": 46615
+    },
+    {
+      "epoch": 1.109587736837094,
+      "grad_norm": 3.521274717635011,
+      "learning_rate": 2e-05,
+      "loss": 0.3272,
+      "step": 46616
+    },
+    {
+      "epoch": 1.1096115395601256,
+      "grad_norm": 2.104639853572304,
+      "learning_rate": 2e-05,
+      "loss": 0.1143,
+      "step": 46617
+    },
+    {
+      "epoch": 1.1096353422831573,
+      "grad_norm": 3.5038556532396186,
+      "learning_rate": 2e-05,
+      "loss": 0.3228,
+      "step": 46618
+    },
+    {
+      "epoch": 1.1096591450061888,
+      "grad_norm": 3.079843022300031,
+      "learning_rate": 2e-05,
+      "loss": 0.0997,
+      "step": 46619
+    },
+    {
+      "epoch": 1.1096829477292203,
+      "grad_norm": 0.018966230815031394,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46620
+    },
+    {
+      "epoch": 1.1097067504522518,
+      "grad_norm": 10.281791004681779,
+      "learning_rate": 2e-05,
+      "loss": 0.5292,
+      "step": 46621
+    },
+    {
+      "epoch": 1.1097305531752832,
+      "grad_norm": 5.416011946108758,
+      "learning_rate": 2e-05,
+      "loss": 0.2807,
+      "step": 46622
+    },
+    {
+      "epoch": 1.1097543558983147,
+      "grad_norm": 7.202244742688868,
+      "learning_rate": 2e-05,
+      "loss": 0.3859,
+      "step": 46623
+    },
+    {
+      "epoch": 1.1097781586213462,
+      "grad_norm": 7.535124990386117,
+      "learning_rate": 2e-05,
+      "loss": 0.8031,
+      "step": 46624
+    },
+    {
+      "epoch": 1.1098019613443777,
+      "grad_norm": 4.695702224028928,
+      "learning_rate": 2e-05,
+      "loss": 0.388,
+      "step": 46625
+    },
+    {
+      "epoch": 1.1098257640674094,
+      "grad_norm": 6.203302811201464,
+      "learning_rate": 2e-05,
+      "loss": 1.2723,
+      "step": 46626
+    },
+    {
+      "epoch": 1.109849566790441,
+      "grad_norm": 6.317355188398933,
+      "learning_rate": 2e-05,
+      "loss": 0.6719,
+      "step": 46627
+    },
+    {
+      "epoch": 1.1098733695134724,
+      "grad_norm": 0.0302478303761554,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46628
+    },
+    {
+      "epoch": 1.1098971722365039,
+      "grad_norm": 4.865746964956769,
+      "learning_rate": 2e-05,
+      "loss": 0.5056,
+      "step": 46629
+    },
+    {
+      "epoch": 1.1099209749595353,
+      "grad_norm": 0.0814101472340127,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46630
+    },
+    {
+      "epoch": 1.1099447776825668,
+      "grad_norm": 1.4368975137662754,
+      "learning_rate": 2e-05,
+      "loss": 0.0077,
+      "step": 46631
+    },
+    {
+      "epoch": 1.1099685804055983,
+      "grad_norm": 3.309983941911705,
+      "learning_rate": 2e-05,
+      "loss": 0.0888,
+      "step": 46632
+    },
+    {
+      "epoch": 1.1099923831286298,
+      "grad_norm": 3.593601109421788,
+      "learning_rate": 2e-05,
+      "loss": 0.0424,
+      "step": 46633
+    },
+    {
+      "epoch": 1.1100161858516615,
+      "grad_norm": 0.5596088552157472,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 46634
+    },
+    {
+      "epoch": 1.110039988574693,
+      "grad_norm": 4.821713228465861,
+      "learning_rate": 2e-05,
+      "loss": 0.6872,
+      "step": 46635
+    },
+    {
+      "epoch": 1.1100637912977245,
+      "grad_norm": 6.988090751108088,
+      "learning_rate": 2e-05,
+      "loss": 0.6629,
+      "step": 46636
+    },
+    {
+      "epoch": 1.110087594020756,
+      "grad_norm": 3.0076264990218706,
+      "learning_rate": 2e-05,
+      "loss": 0.0574,
+      "step": 46637
+    },
+    {
+      "epoch": 1.1101113967437874,
+      "grad_norm": 1.1815826994468746,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 46638
+    },
+    {
+      "epoch": 1.110135199466819,
+      "grad_norm": 0.013114168459467968,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46639
+    },
+    {
+      "epoch": 1.1101590021898504,
+      "grad_norm": 0.9470811431519108,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 46640
+    },
+    {
+      "epoch": 1.1101828049128821,
+      "grad_norm": 3.156797650965759,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 46641
+    },
+    {
+      "epoch": 1.1102066076359136,
+      "grad_norm": 5.316442097261342,
+      "learning_rate": 2e-05,
+      "loss": 0.4815,
+      "step": 46642
+    },
+    {
+      "epoch": 1.110230410358945,
+      "grad_norm": 4.300962648137254,
+      "learning_rate": 2e-05,
+      "loss": 0.316,
+      "step": 46643
+    },
+    {
+      "epoch": 1.1102542130819766,
+      "grad_norm": 5.884614693353937,
+      "learning_rate": 2e-05,
+      "loss": 0.1729,
+      "step": 46644
+    },
+    {
+      "epoch": 1.110278015805008,
+      "grad_norm": 2.959839864509424,
+      "learning_rate": 2e-05,
+      "loss": 0.2612,
+      "step": 46645
+    },
+    {
+      "epoch": 1.1103018185280396,
+      "grad_norm": 0.04643088794297296,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46646
+    },
+    {
+      "epoch": 1.110325621251071,
+      "grad_norm": 4.3692555393416255,
+      "learning_rate": 2e-05,
+      "loss": 0.0644,
+      "step": 46647
+    },
+    {
+      "epoch": 1.1103494239741027,
+      "grad_norm": 2.6631778507943187,
+      "learning_rate": 2e-05,
+      "loss": 0.2301,
+      "step": 46648
+    },
+    {
+      "epoch": 1.1103732266971342,
+      "grad_norm": 8.298069658258905,
+      "learning_rate": 2e-05,
+      "loss": 0.2754,
+      "step": 46649
+    },
+    {
+      "epoch": 1.1103970294201657,
+      "grad_norm": 4.960364085451323,
+      "learning_rate": 2e-05,
+      "loss": 0.4502,
+      "step": 46650
+    },
+    {
+      "epoch": 1.1104208321431972,
+      "grad_norm": 0.10214377815217895,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46651
+    },
+    {
+      "epoch": 1.1104446348662287,
+      "grad_norm": 6.41735606713891,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 46652
+    },
+    {
+      "epoch": 1.1104684375892602,
+      "grad_norm": 1.132119261034675,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 46653
+    },
+    {
+      "epoch": 1.1104922403122917,
+      "grad_norm": 3.567002283304996,
+      "learning_rate": 2e-05,
+      "loss": 0.0771,
+      "step": 46654
+    },
+    {
+      "epoch": 1.1105160430353231,
+      "grad_norm": 0.06785460742048789,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46655
+    },
+    {
+      "epoch": 1.1105398457583548,
+      "grad_norm": 0.7021610140961738,
+      "learning_rate": 2e-05,
+      "loss": 0.0053,
+      "step": 46656
+    },
+    {
+      "epoch": 1.1105636484813863,
+      "grad_norm": 0.3176956858123974,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 46657
+    },
+    {
+      "epoch": 1.1105874512044178,
+      "grad_norm": 3.846735312853305,
+      "learning_rate": 2e-05,
+      "loss": 0.1282,
+      "step": 46658
+    },
+    {
+      "epoch": 1.1106112539274493,
+      "grad_norm": 9.543914074810518,
+      "learning_rate": 2e-05,
+      "loss": 0.5344,
+      "step": 46659
+    },
+    {
+      "epoch": 1.1106350566504808,
+      "grad_norm": 9.213141499471392,
+      "learning_rate": 2e-05,
+      "loss": 0.186,
+      "step": 46660
+    },
+    {
+      "epoch": 1.1106588593735123,
+      "grad_norm": 6.717203875071944,
+      "learning_rate": 2e-05,
+      "loss": 0.9394,
+      "step": 46661
+    },
+    {
+      "epoch": 1.1106826620965438,
+      "grad_norm": 0.053221735709707695,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46662
+    },
+    {
+      "epoch": 1.1107064648195752,
+      "grad_norm": 0.013491093150771334,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46663
+    },
+    {
+      "epoch": 1.110730267542607,
+      "grad_norm": 0.014203707825040582,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46664
+    },
+    {
+      "epoch": 1.1107540702656384,
+      "grad_norm": 9.558988930224425,
+      "learning_rate": 2e-05,
+      "loss": 0.946,
+      "step": 46665
+    },
+    {
+      "epoch": 1.11077787298867,
+      "grad_norm": 2.3419539756680576,
+      "learning_rate": 2e-05,
+      "loss": 0.0306,
+      "step": 46666
+    },
+    {
+      "epoch": 1.1108016757117014,
+      "grad_norm": 4.183492955004737,
+      "learning_rate": 2e-05,
+      "loss": 0.3681,
+      "step": 46667
+    },
+    {
+      "epoch": 1.110825478434733,
+      "grad_norm": 11.150070712535316,
+      "learning_rate": 2e-05,
+      "loss": 0.7396,
+      "step": 46668
+    },
+    {
+      "epoch": 1.1108492811577644,
+      "grad_norm": 1.720663961039073,
+      "learning_rate": 2e-05,
+      "loss": 0.0748,
+      "step": 46669
+    },
+    {
+      "epoch": 1.1108730838807959,
+      "grad_norm": 4.0319776050456495,
+      "learning_rate": 2e-05,
+      "loss": 0.2362,
+      "step": 46670
+    },
+    {
+      "epoch": 1.1108968866038276,
+      "grad_norm": 0.06014028051194075,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46671
+    },
+    {
+      "epoch": 1.110920689326859,
+      "grad_norm": 0.27131259325966545,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46672
+    },
+    {
+      "epoch": 1.1109444920498905,
+      "grad_norm": 18.818913260987756,
+      "learning_rate": 2e-05,
+      "loss": 0.576,
+      "step": 46673
+    },
+    {
+      "epoch": 1.110968294772922,
+      "grad_norm": 0.005452037088078503,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46674
+    },
+    {
+      "epoch": 1.1109920974959535,
+      "grad_norm": 0.06987429605298545,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46675
+    },
+    {
+      "epoch": 1.111015900218985,
+      "grad_norm": 2.202054531871808,
+      "learning_rate": 2e-05,
+      "loss": 0.1509,
+      "step": 46676
+    },
+    {
+      "epoch": 1.1110397029420165,
+      "grad_norm": 1.6199963038071221,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 46677
+    },
+    {
+      "epoch": 1.1110635056650482,
+      "grad_norm": 0.840143686711352,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 46678
+    },
+    {
+      "epoch": 1.1110873083880797,
+      "grad_norm": 4.668539155643729,
+      "learning_rate": 2e-05,
+      "loss": 0.2209,
+      "step": 46679
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.16923191427381554,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 46680
+    },
+    {
+      "epoch": 1.1111349138341426,
+      "grad_norm": 5.789027268874916,
+      "learning_rate": 2e-05,
+      "loss": 1.2378,
+      "step": 46681
+    },
+    {
+      "epoch": 1.1111587165571741,
+      "grad_norm": 3.2722758899594537,
+      "learning_rate": 2e-05,
+      "loss": 0.463,
+      "step": 46682
+    },
+    {
+      "epoch": 1.1111825192802056,
+      "grad_norm": 0.17872195095200488,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 46683
+    },
+    {
+      "epoch": 1.111206322003237,
+      "grad_norm": 2.3044026359387324,
+      "learning_rate": 2e-05,
+      "loss": 0.2097,
+      "step": 46684
+    },
+    {
+      "epoch": 1.1112301247262686,
+      "grad_norm": 0.8817865495427024,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 46685
+    },
+    {
+      "epoch": 1.1112539274493003,
+      "grad_norm": 11.707055654751093,
+      "learning_rate": 2e-05,
+      "loss": 0.1952,
+      "step": 46686
+    },
+    {
+      "epoch": 1.1112777301723318,
+      "grad_norm": 3.5284867632625057,
+      "learning_rate": 2e-05,
+      "loss": 0.0524,
+      "step": 46687
+    },
+    {
+      "epoch": 1.1113015328953633,
+      "grad_norm": 0.23957798344853382,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 46688
+    },
+    {
+      "epoch": 1.1113253356183947,
+      "grad_norm": 4.8611963980702635,
+      "learning_rate": 2e-05,
+      "loss": 0.3498,
+      "step": 46689
+    },
+    {
+      "epoch": 1.1113491383414262,
+      "grad_norm": 0.9377412196783259,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 46690
+    },
+    {
+      "epoch": 1.1113729410644577,
+      "grad_norm": 3.6686312732183803,
+      "learning_rate": 2e-05,
+      "loss": 0.2402,
+      "step": 46691
+    },
+    {
+      "epoch": 1.1113967437874892,
+      "grad_norm": 4.374496319166625,
+      "learning_rate": 2e-05,
+      "loss": 0.7386,
+      "step": 46692
+    },
+    {
+      "epoch": 1.1114205465105207,
+      "grad_norm": 4.168479823482845,
+      "learning_rate": 2e-05,
+      "loss": 0.5297,
+      "step": 46693
+    },
+    {
+      "epoch": 1.1114443492335524,
+      "grad_norm": 5.461157697190591,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 46694
+    },
+    {
+      "epoch": 1.1114681519565839,
+      "grad_norm": 3.7097450360830067,
+      "learning_rate": 2e-05,
+      "loss": 0.2114,
+      "step": 46695
+    },
+    {
+      "epoch": 1.1114919546796154,
+      "grad_norm": 2.165873371602723,
+      "learning_rate": 2e-05,
+      "loss": 0.1332,
+      "step": 46696
+    },
+    {
+      "epoch": 1.1115157574026469,
+      "grad_norm": 1.6033897736935447,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 46697
+    },
+    {
+      "epoch": 1.1115395601256783,
+      "grad_norm": 0.28762799446507686,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 46698
+    },
+    {
+      "epoch": 1.1115633628487098,
+      "grad_norm": 4.449724304310086,
+      "learning_rate": 2e-05,
+      "loss": 0.4161,
+      "step": 46699
+    },
+    {
+      "epoch": 1.1115871655717413,
+      "grad_norm": 1.8092185342446148,
+      "learning_rate": 2e-05,
+      "loss": 0.1548,
+      "step": 46700
+    },
+    {
+      "epoch": 1.111610968294773,
+      "grad_norm": 4.693355421554288,
+      "learning_rate": 2e-05,
+      "loss": 0.3847,
+      "step": 46701
+    },
+    {
+      "epoch": 1.1116347710178045,
+      "grad_norm": 2.381804013496709,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 46702
+    },
+    {
+      "epoch": 1.111658573740836,
+      "grad_norm": 9.08411099342244,
+      "learning_rate": 2e-05,
+      "loss": 0.1684,
+      "step": 46703
+    },
+    {
+      "epoch": 1.1116823764638675,
+      "grad_norm": 3.2023651538603732,
+      "learning_rate": 2e-05,
+      "loss": 0.1892,
+      "step": 46704
+    },
+    {
+      "epoch": 1.111706179186899,
+      "grad_norm": 8.953066012160852,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 46705
+    },
+    {
+      "epoch": 1.1117299819099304,
+      "grad_norm": 1.3365893781856046,
+      "learning_rate": 2e-05,
+      "loss": 0.017,
+      "step": 46706
+    },
+    {
+      "epoch": 1.111753784632962,
+      "grad_norm": 0.049590688566083674,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46707
+    },
+    {
+      "epoch": 1.1117775873559936,
+      "grad_norm": 1.6523638706366788,
+      "learning_rate": 2e-05,
+      "loss": 0.0239,
+      "step": 46708
+    },
+    {
+      "epoch": 1.1118013900790251,
+      "grad_norm": 0.3096669178638585,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 46709
+    },
+    {
+      "epoch": 1.1118251928020566,
+      "grad_norm": 29.405204068283798,
+      "learning_rate": 2e-05,
+      "loss": 1.2478,
+      "step": 46710
+    },
+    {
+      "epoch": 1.111848995525088,
+      "grad_norm": 0.33945368845109675,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46711
+    },
+    {
+      "epoch": 1.1118727982481196,
+      "grad_norm": 8.207004198036644,
+      "learning_rate": 2e-05,
+      "loss": 1.0177,
+      "step": 46712
+    },
+    {
+      "epoch": 1.111896600971151,
+      "grad_norm": 2.749714227012013,
+      "learning_rate": 2e-05,
+      "loss": 0.2531,
+      "step": 46713
+    },
+    {
+      "epoch": 1.1119204036941825,
+      "grad_norm": 0.051328019067750565,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46714
+    },
+    {
+      "epoch": 1.111944206417214,
+      "grad_norm": 0.3653289246726691,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 46715
+    },
+    {
+      "epoch": 1.1119680091402457,
+      "grad_norm": 4.108584485904541,
+      "learning_rate": 2e-05,
+      "loss": 0.4765,
+      "step": 46716
+    },
+    {
+      "epoch": 1.1119918118632772,
+      "grad_norm": 0.011842953162940918,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46717
+    },
+    {
+      "epoch": 1.1120156145863087,
+      "grad_norm": 0.15580465208979283,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 46718
+    },
+    {
+      "epoch": 1.1120394173093402,
+      "grad_norm": 2.937444581744729,
+      "learning_rate": 2e-05,
+      "loss": 0.3704,
+      "step": 46719
+    },
+    {
+      "epoch": 1.1120632200323717,
+      "grad_norm": 12.08232081304096,
+      "learning_rate": 2e-05,
+      "loss": 1.7583,
+      "step": 46720
+    },
+    {
+      "epoch": 1.1120870227554032,
+      "grad_norm": 4.374342768738262,
+      "learning_rate": 2e-05,
+      "loss": 0.3408,
+      "step": 46721
+    },
+    {
+      "epoch": 1.1121108254784347,
+      "grad_norm": 3.1033755653980593,
+      "learning_rate": 2e-05,
+      "loss": 0.4003,
+      "step": 46722
+    },
+    {
+      "epoch": 1.1121346282014661,
+      "grad_norm": 5.78959135492518,
+      "learning_rate": 2e-05,
+      "loss": 0.4811,
+      "step": 46723
+    },
+    {
+      "epoch": 1.1121584309244978,
+      "grad_norm": 1.8798715486817505,
+      "learning_rate": 2e-05,
+      "loss": 0.0434,
+      "step": 46724
+    },
+    {
+      "epoch": 1.1121822336475293,
+      "grad_norm": 0.04324117564184036,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46725
+    },
+    {
+      "epoch": 1.1122060363705608,
+      "grad_norm": 4.449172582847271,
+      "learning_rate": 2e-05,
+      "loss": 0.3935,
+      "step": 46726
+    },
+    {
+      "epoch": 1.1122298390935923,
+      "grad_norm": 8.001279373010664,
+      "learning_rate": 2e-05,
+      "loss": 0.8467,
+      "step": 46727
+    },
+    {
+      "epoch": 1.1122536418166238,
+      "grad_norm": 3.8437120060898993,
+      "learning_rate": 2e-05,
+      "loss": 0.311,
+      "step": 46728
+    },
+    {
+      "epoch": 1.1122774445396553,
+      "grad_norm": 4.4517936455715486,
+      "learning_rate": 2e-05,
+      "loss": 0.1832,
+      "step": 46729
+    },
+    {
+      "epoch": 1.1123012472626868,
+      "grad_norm": 4.456364366595726,
+      "learning_rate": 2e-05,
+      "loss": 0.4529,
+      "step": 46730
+    },
+    {
+      "epoch": 1.1123250499857185,
+      "grad_norm": 3.976940068434659,
+      "learning_rate": 2e-05,
+      "loss": 0.3175,
+      "step": 46731
+    },
+    {
+      "epoch": 1.11234885270875,
+      "grad_norm": 2.602580741055478,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 46732
+    },
+    {
+      "epoch": 1.1123726554317814,
+      "grad_norm": 3.584513496469759,
+      "learning_rate": 2e-05,
+      "loss": 0.5703,
+      "step": 46733
+    },
+    {
+      "epoch": 1.112396458154813,
+      "grad_norm": 3.3094619984114324,
+      "learning_rate": 2e-05,
+      "loss": 0.0546,
+      "step": 46734
+    },
+    {
+      "epoch": 1.1124202608778444,
+      "grad_norm": 1.249230966014836,
+      "learning_rate": 2e-05,
+      "loss": 0.0135,
+      "step": 46735
+    },
+    {
+      "epoch": 1.1124440636008759,
+      "grad_norm": 5.316358489101085,
+      "learning_rate": 2e-05,
+      "loss": 0.6084,
+      "step": 46736
+    },
+    {
+      "epoch": 1.1124678663239074,
+      "grad_norm": 1.0861515498133478,
+      "learning_rate": 2e-05,
+      "loss": 0.0201,
+      "step": 46737
+    },
+    {
+      "epoch": 1.112491669046939,
+      "grad_norm": 3.456763838626235,
+      "learning_rate": 2e-05,
+      "loss": 0.2271,
+      "step": 46738
+    },
+    {
+      "epoch": 1.1125154717699706,
+      "grad_norm": 10.467163479965999,
+      "learning_rate": 2e-05,
+      "loss": 0.2127,
+      "step": 46739
+    },
+    {
+      "epoch": 1.112539274493002,
+      "grad_norm": 9.349377554704944,
+      "learning_rate": 2e-05,
+      "loss": 0.1531,
+      "step": 46740
+    },
+    {
+      "epoch": 1.1125630772160335,
+      "grad_norm": 3.572186811002452,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 46741
+    },
+    {
+      "epoch": 1.112586879939065,
+      "grad_norm": 0.4221068857384551,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 46742
+    },
+    {
+      "epoch": 1.1126106826620965,
+      "grad_norm": 5.962729602818454,
+      "learning_rate": 2e-05,
+      "loss": 1.1803,
+      "step": 46743
+    },
+    {
+      "epoch": 1.112634485385128,
+      "grad_norm": 0.04265018673644462,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46744
+    },
+    {
+      "epoch": 1.1126582881081595,
+      "grad_norm": 2.937780149917962,
+      "learning_rate": 2e-05,
+      "loss": 0.3689,
+      "step": 46745
+    },
+    {
+      "epoch": 1.1126820908311912,
+      "grad_norm": 0.7782723308484946,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 46746
+    },
+    {
+      "epoch": 1.1127058935542227,
+      "grad_norm": 0.31928700618561584,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46747
+    },
+    {
+      "epoch": 1.1127296962772542,
+      "grad_norm": 8.677973011668852,
+      "learning_rate": 2e-05,
+      "loss": 0.677,
+      "step": 46748
+    },
+    {
+      "epoch": 1.1127534990002856,
+      "grad_norm": 0.6243806382438318,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 46749
+    },
+    {
+      "epoch": 1.1127773017233171,
+      "grad_norm": 2.6174420555852462,
+      "learning_rate": 2e-05,
+      "loss": 0.1881,
+      "step": 46750
+    },
+    {
+      "epoch": 1.1128011044463486,
+      "grad_norm": 4.396450411393808,
+      "learning_rate": 2e-05,
+      "loss": 0.5242,
+      "step": 46751
+    },
+    {
+      "epoch": 1.11282490716938,
+      "grad_norm": 0.07605285059268799,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46752
+    },
+    {
+      "epoch": 1.1128487098924116,
+      "grad_norm": 0.057639400930956944,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46753
+    },
+    {
+      "epoch": 1.1128725126154433,
+      "grad_norm": 6.3489180610651434,
+      "learning_rate": 2e-05,
+      "loss": 0.4901,
+      "step": 46754
+    },
+    {
+      "epoch": 1.1128963153384748,
+      "grad_norm": 0.04601928831727351,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46755
+    },
+    {
+      "epoch": 1.1129201180615063,
+      "grad_norm": 5.569953863249848,
+      "learning_rate": 2e-05,
+      "loss": 0.0736,
+      "step": 46756
+    },
+    {
+      "epoch": 1.1129439207845377,
+      "grad_norm": 0.08134602535596087,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46757
+    },
+    {
+      "epoch": 1.1129677235075692,
+      "grad_norm": 2.879074212037117,
+      "learning_rate": 2e-05,
+      "loss": 0.2733,
+      "step": 46758
+    },
+    {
+      "epoch": 1.1129915262306007,
+      "grad_norm": 5.729332245610046,
+      "learning_rate": 2e-05,
+      "loss": 0.7783,
+      "step": 46759
+    },
+    {
+      "epoch": 1.1130153289536322,
+      "grad_norm": 4.745617497549848,
+      "learning_rate": 2e-05,
+      "loss": 0.6558,
+      "step": 46760
+    },
+    {
+      "epoch": 1.113039131676664,
+      "grad_norm": 8.264368941170058,
+      "learning_rate": 2e-05,
+      "loss": 0.0622,
+      "step": 46761
+    },
+    {
+      "epoch": 1.1130629343996954,
+      "grad_norm": 0.10100057224083948,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 46762
+    },
+    {
+      "epoch": 1.1130867371227269,
+      "grad_norm": 3.2821219465147204,
+      "learning_rate": 2e-05,
+      "loss": 0.0322,
+      "step": 46763
+    },
+    {
+      "epoch": 1.1131105398457584,
+      "grad_norm": 2.197053729557546,
+      "learning_rate": 2e-05,
+      "loss": 0.0296,
+      "step": 46764
+    },
+    {
+      "epoch": 1.1131343425687898,
+      "grad_norm": 3.958185108728947,
+      "learning_rate": 2e-05,
+      "loss": 0.3099,
+      "step": 46765
+    },
+    {
+      "epoch": 1.1131581452918213,
+      "grad_norm": 0.7674381251604382,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 46766
+    },
+    {
+      "epoch": 1.1131819480148528,
+      "grad_norm": 2.7288053114757247,
+      "learning_rate": 2e-05,
+      "loss": 0.0425,
+      "step": 46767
+    },
+    {
+      "epoch": 1.1132057507378845,
+      "grad_norm": 0.007828010391831454,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46768
+    },
+    {
+      "epoch": 1.113229553460916,
+      "grad_norm": 0.9658111786754819,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 46769
+    },
+    {
+      "epoch": 1.1132533561839475,
+      "grad_norm": 0.07261012921332094,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46770
+    },
+    {
+      "epoch": 1.113277158906979,
+      "grad_norm": 0.011331387266190361,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46771
+    },
+    {
+      "epoch": 1.1133009616300105,
+      "grad_norm": 0.39096986970364456,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 46772
+    },
+    {
+      "epoch": 1.113324764353042,
+      "grad_norm": 2.6494900230146765,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 46773
+    },
+    {
+      "epoch": 1.1133485670760734,
+      "grad_norm": 4.97843756968145,
+      "learning_rate": 2e-05,
+      "loss": 0.2806,
+      "step": 46774
+    },
+    {
+      "epoch": 1.113372369799105,
+      "grad_norm": 0.012803650108463368,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46775
+    },
+    {
+      "epoch": 1.1133961725221366,
+      "grad_norm": 0.21968453162385929,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 46776
+    },
+    {
+      "epoch": 1.1134199752451681,
+      "grad_norm": 0.49294303692669644,
+      "learning_rate": 2e-05,
+      "loss": 0.0044,
+      "step": 46777
+    },
+    {
+      "epoch": 1.1134437779681996,
+      "grad_norm": 0.005570896342459212,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46778
+    },
+    {
+      "epoch": 1.113467580691231,
+      "grad_norm": 0.15684572834778954,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 46779
+    },
+    {
+      "epoch": 1.1134913834142626,
+      "grad_norm": 14.020166541685178,
+      "learning_rate": 2e-05,
+      "loss": 0.5906,
+      "step": 46780
+    },
+    {
+      "epoch": 1.113515186137294,
+      "grad_norm": 1.4027023087514725,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 46781
+    },
+    {
+      "epoch": 1.1135389888603255,
+      "grad_norm": 3.890410487323528,
+      "learning_rate": 2e-05,
+      "loss": 0.4539,
+      "step": 46782
+    },
+    {
+      "epoch": 1.113562791583357,
+      "grad_norm": 0.24879061663042595,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 46783
+    },
+    {
+      "epoch": 1.1135865943063887,
+      "grad_norm": 0.08795941742684815,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 46784
+    },
+    {
+      "epoch": 1.1136103970294202,
+      "grad_norm": 3.3505974179903264,
+      "learning_rate": 2e-05,
+      "loss": 0.268,
+      "step": 46785
+    },
+    {
+      "epoch": 1.1136341997524517,
+      "grad_norm": 4.010438862375497,
+      "learning_rate": 2e-05,
+      "loss": 0.0262,
+      "step": 46786
+    },
+    {
+      "epoch": 1.1136580024754832,
+      "grad_norm": 0.2350334970387326,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 46787
+    },
+    {
+      "epoch": 1.1136818051985147,
+      "grad_norm": 4.213907831073414,
+      "learning_rate": 2e-05,
+      "loss": 0.0218,
+      "step": 46788
+    },
+    {
+      "epoch": 1.1137056079215462,
+      "grad_norm": 19.935621117002935,
+      "learning_rate": 2e-05,
+      "loss": 0.542,
+      "step": 46789
+    },
+    {
+      "epoch": 1.1137294106445776,
+      "grad_norm": 4.889949147360489,
+      "learning_rate": 2e-05,
+      "loss": 0.3997,
+      "step": 46790
+    },
+    {
+      "epoch": 1.1137532133676094,
+      "grad_norm": 0.5364052590284349,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46791
+    },
+    {
+      "epoch": 1.1137770160906408,
+      "grad_norm": 0.06026197077924949,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46792
+    },
+    {
+      "epoch": 1.1138008188136723,
+      "grad_norm": 4.778638670495711,
+      "learning_rate": 2e-05,
+      "loss": 0.0726,
+      "step": 46793
+    },
+    {
+      "epoch": 1.1138246215367038,
+      "grad_norm": 6.620630962441975,
+      "learning_rate": 2e-05,
+      "loss": 0.9028,
+      "step": 46794
+    },
+    {
+      "epoch": 1.1138484242597353,
+      "grad_norm": 3.774243350070021,
+      "learning_rate": 2e-05,
+      "loss": 0.5412,
+      "step": 46795
+    },
+    {
+      "epoch": 1.1138722269827668,
+      "grad_norm": 3.1143412115300957,
+      "learning_rate": 2e-05,
+      "loss": 0.4408,
+      "step": 46796
+    },
+    {
+      "epoch": 1.1138960297057983,
+      "grad_norm": 0.12353042201952306,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 46797
+    },
+    {
+      "epoch": 1.11391983242883,
+      "grad_norm": 1.863776309957817,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 46798
+    },
+    {
+      "epoch": 1.1139436351518615,
+      "grad_norm": 5.328406672479277,
+      "learning_rate": 2e-05,
+      "loss": 0.6181,
+      "step": 46799
+    },
+    {
+      "epoch": 1.113967437874893,
+      "grad_norm": 0.023061672821938904,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46800
+    },
+    {
+      "epoch": 1.1139912405979244,
+      "grad_norm": 10.273250770825426,
+      "learning_rate": 2e-05,
+      "loss": 1.4943,
+      "step": 46801
+    },
+    {
+      "epoch": 1.114015043320956,
+      "grad_norm": 3.303685531002666,
+      "learning_rate": 2e-05,
+      "loss": 0.3587,
+      "step": 46802
+    },
+    {
+      "epoch": 1.1140388460439874,
+      "grad_norm": 3.4722488767949717,
+      "learning_rate": 2e-05,
+      "loss": 0.5022,
+      "step": 46803
+    },
+    {
+      "epoch": 1.1140626487670189,
+      "grad_norm": 0.4399725557462977,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46804
+    },
+    {
+      "epoch": 1.1140864514900504,
+      "grad_norm": 3.30447100182804,
+      "learning_rate": 2e-05,
+      "loss": 0.3951,
+      "step": 46805
+    },
+    {
+      "epoch": 1.114110254213082,
+      "grad_norm": 1.3862401184953947,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 46806
+    },
+    {
+      "epoch": 1.1141340569361136,
+      "grad_norm": 0.2857025027642795,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 46807
+    },
+    {
+      "epoch": 1.114157859659145,
+      "grad_norm": 3.4286981297182177,
+      "learning_rate": 2e-05,
+      "loss": 0.085,
+      "step": 46808
+    },
+    {
+      "epoch": 1.1141816623821765,
+      "grad_norm": 3.469080618583888,
+      "learning_rate": 2e-05,
+      "loss": 0.6742,
+      "step": 46809
+    },
+    {
+      "epoch": 1.114205465105208,
+      "grad_norm": 0.1436373599314587,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 46810
+    },
+    {
+      "epoch": 1.1142292678282395,
+      "grad_norm": 0.03676923513030646,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46811
+    },
+    {
+      "epoch": 1.114253070551271,
+      "grad_norm": 0.3437508700763049,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 46812
+    },
+    {
+      "epoch": 1.1142768732743025,
+      "grad_norm": 0.8049163696402204,
+      "learning_rate": 2e-05,
+      "loss": 0.0096,
+      "step": 46813
+    },
+    {
+      "epoch": 1.1143006759973342,
+      "grad_norm": 2.078688678343984,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 46814
+    },
+    {
+      "epoch": 1.1143244787203657,
+      "grad_norm": 3.080755993394977,
+      "learning_rate": 2e-05,
+      "loss": 0.0489,
+      "step": 46815
+    },
+    {
+      "epoch": 1.1143482814433971,
+      "grad_norm": 5.7322859435425775,
+      "learning_rate": 2e-05,
+      "loss": 0.5856,
+      "step": 46816
+    },
+    {
+      "epoch": 1.1143720841664286,
+      "grad_norm": 4.9115698980845535,
+      "learning_rate": 2e-05,
+      "loss": 0.0756,
+      "step": 46817
+    },
+    {
+      "epoch": 1.1143958868894601,
+      "grad_norm": 0.02479557301876693,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46818
+    },
+    {
+      "epoch": 1.1144196896124916,
+      "grad_norm": 0.4121059584299458,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 46819
+    },
+    {
+      "epoch": 1.114443492335523,
+      "grad_norm": 0.021470238566884705,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46820
+    },
+    {
+      "epoch": 1.1144672950585548,
+      "grad_norm": 3.550386973449636,
+      "learning_rate": 2e-05,
+      "loss": 0.3728,
+      "step": 46821
+    },
+    {
+      "epoch": 1.1144910977815863,
+      "grad_norm": 6.726987454955135,
+      "learning_rate": 2e-05,
+      "loss": 0.4743,
+      "step": 46822
+    },
+    {
+      "epoch": 1.1145149005046178,
+      "grad_norm": 1.256880086400361,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 46823
+    },
+    {
+      "epoch": 1.1145387032276493,
+      "grad_norm": 0.03695856507333405,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46824
+    },
+    {
+      "epoch": 1.1145625059506807,
+      "grad_norm": 3.0436679871222068,
+      "learning_rate": 2e-05,
+      "loss": 0.298,
+      "step": 46825
+    },
+    {
+      "epoch": 1.1145863086737122,
+      "grad_norm": 0.24468370183328655,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46826
+    },
+    {
+      "epoch": 1.1146101113967437,
+      "grad_norm": 4.254734543860517,
+      "learning_rate": 2e-05,
+      "loss": 0.6958,
+      "step": 46827
+    },
+    {
+      "epoch": 1.1146339141197754,
+      "grad_norm": 0.03560781563790087,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46828
+    },
+    {
+      "epoch": 1.114657716842807,
+      "grad_norm": 14.145602386702507,
+      "learning_rate": 2e-05,
+      "loss": 1.5983,
+      "step": 46829
+    },
+    {
+      "epoch": 1.1146815195658384,
+      "grad_norm": 1.167500565855021,
+      "learning_rate": 2e-05,
+      "loss": 0.0114,
+      "step": 46830
+    },
+    {
+      "epoch": 1.1147053222888699,
+      "grad_norm": 5.761511617261688,
+      "learning_rate": 2e-05,
+      "loss": 0.4335,
+      "step": 46831
+    },
+    {
+      "epoch": 1.1147291250119014,
+      "grad_norm": 0.027614952467621866,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46832
+    },
+    {
+      "epoch": 1.1147529277349328,
+      "grad_norm": 0.7600119349995209,
+      "learning_rate": 2e-05,
+      "loss": 0.0042,
+      "step": 46833
+    },
+    {
+      "epoch": 1.1147767304579643,
+      "grad_norm": 3.104323231005101,
+      "learning_rate": 2e-05,
+      "loss": 0.306,
+      "step": 46834
+    },
+    {
+      "epoch": 1.1148005331809958,
+      "grad_norm": 3.174466668146546,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 46835
+    },
+    {
+      "epoch": 1.1148243359040275,
+      "grad_norm": 3.044422110796367,
+      "learning_rate": 2e-05,
+      "loss": 0.2113,
+      "step": 46836
+    },
+    {
+      "epoch": 1.114848138627059,
+      "grad_norm": 4.790398745145532,
+      "learning_rate": 2e-05,
+      "loss": 0.2224,
+      "step": 46837
+    },
+    {
+      "epoch": 1.1148719413500905,
+      "grad_norm": 0.5599633971634198,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46838
+    },
+    {
+      "epoch": 1.114895744073122,
+      "grad_norm": 3.8334175064020304,
+      "learning_rate": 2e-05,
+      "loss": 0.5753,
+      "step": 46839
+    },
+    {
+      "epoch": 1.1149195467961535,
+      "grad_norm": 0.5358241894895294,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46840
+    },
+    {
+      "epoch": 1.114943349519185,
+      "grad_norm": 5.303342267747095,
+      "learning_rate": 2e-05,
+      "loss": 0.7661,
+      "step": 46841
+    },
+    {
+      "epoch": 1.1149671522422164,
+      "grad_norm": 4.981908726838301,
+      "learning_rate": 2e-05,
+      "loss": 0.402,
+      "step": 46842
+    },
+    {
+      "epoch": 1.114990954965248,
+      "grad_norm": 1.452885410308488,
+      "learning_rate": 2e-05,
+      "loss": 0.004,
+      "step": 46843
+    },
+    {
+      "epoch": 1.1150147576882796,
+      "grad_norm": 0.03479858826915725,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46844
+    },
+    {
+      "epoch": 1.115038560411311,
+      "grad_norm": 6.075248904980307,
+      "learning_rate": 2e-05,
+      "loss": 0.7105,
+      "step": 46845
+    },
+    {
+      "epoch": 1.1150623631343426,
+      "grad_norm": 0.42507685649912275,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 46846
+    },
+    {
+      "epoch": 1.115086165857374,
+      "grad_norm": 0.18331635075801686,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46847
+    },
+    {
+      "epoch": 1.1151099685804056,
+      "grad_norm": 7.585520286066604,
+      "learning_rate": 2e-05,
+      "loss": 0.691,
+      "step": 46848
+    },
+    {
+      "epoch": 1.115133771303437,
+      "grad_norm": 0.00866678878268883,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46849
+    },
+    {
+      "epoch": 1.1151575740264685,
+      "grad_norm": 4.465826965911264,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 46850
+    },
+    {
+      "epoch": 1.1151813767495002,
+      "grad_norm": 0.03580831436580382,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46851
+    },
+    {
+      "epoch": 1.1152051794725317,
+      "grad_norm": 5.829807590328273,
+      "learning_rate": 2e-05,
+      "loss": 0.4216,
+      "step": 46852
+    },
+    {
+      "epoch": 1.1152289821955632,
+      "grad_norm": 3.116765001099335,
+      "learning_rate": 2e-05,
+      "loss": 0.0702,
+      "step": 46853
+    },
+    {
+      "epoch": 1.1152527849185947,
+      "grad_norm": 8.482198539446063,
+      "learning_rate": 2e-05,
+      "loss": 0.8517,
+      "step": 46854
+    },
+    {
+      "epoch": 1.1152765876416262,
+      "grad_norm": 4.996439598764924,
+      "learning_rate": 2e-05,
+      "loss": 0.4217,
+      "step": 46855
+    },
+    {
+      "epoch": 1.1153003903646577,
+      "grad_norm": 0.5228781235251584,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 46856
+    },
+    {
+      "epoch": 1.1153241930876892,
+      "grad_norm": 0.06791546962942654,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 46857
+    },
+    {
+      "epoch": 1.1153479958107209,
+      "grad_norm": 5.948212579999503,
+      "learning_rate": 2e-05,
+      "loss": 0.6858,
+      "step": 46858
+    },
+    {
+      "epoch": 1.1153717985337523,
+      "grad_norm": 3.9225324140522004,
+      "learning_rate": 2e-05,
+      "loss": 0.471,
+      "step": 46859
+    },
+    {
+      "epoch": 1.1153956012567838,
+      "grad_norm": 4.00413128806506,
+      "learning_rate": 2e-05,
+      "loss": 0.0374,
+      "step": 46860
+    },
+    {
+      "epoch": 1.1154194039798153,
+      "grad_norm": 12.82557612878648,
+      "learning_rate": 2e-05,
+      "loss": 0.4979,
+      "step": 46861
+    },
+    {
+      "epoch": 1.1154432067028468,
+      "grad_norm": 3.0381659716563725,
+      "learning_rate": 2e-05,
+      "loss": 0.4032,
+      "step": 46862
+    },
+    {
+      "epoch": 1.1154670094258783,
+      "grad_norm": 0.05648141924663838,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46863
+    },
+    {
+      "epoch": 1.1154908121489098,
+      "grad_norm": 1.436264383065742,
+      "learning_rate": 2e-05,
+      "loss": 0.0182,
+      "step": 46864
+    },
+    {
+      "epoch": 1.1155146148719413,
+      "grad_norm": 0.029411468231320435,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46865
+    },
+    {
+      "epoch": 1.115538417594973,
+      "grad_norm": 5.642057965650954,
+      "learning_rate": 2e-05,
+      "loss": 0.0904,
+      "step": 46866
+    },
+    {
+      "epoch": 1.1155622203180044,
+      "grad_norm": 3.544962865834042,
+      "learning_rate": 2e-05,
+      "loss": 0.1309,
+      "step": 46867
+    },
+    {
+      "epoch": 1.115586023041036,
+      "grad_norm": 3.230590056554159,
+      "learning_rate": 2e-05,
+      "loss": 0.3098,
+      "step": 46868
+    },
+    {
+      "epoch": 1.1156098257640674,
+      "grad_norm": 3.661139257475226,
+      "learning_rate": 2e-05,
+      "loss": 0.3734,
+      "step": 46869
+    },
+    {
+      "epoch": 1.115633628487099,
+      "grad_norm": 1.7261519703568904,
+      "learning_rate": 2e-05,
+      "loss": 0.0118,
+      "step": 46870
+    },
+    {
+      "epoch": 1.1156574312101304,
+      "grad_norm": 1.259822624150094,
+      "learning_rate": 2e-05,
+      "loss": 0.012,
+      "step": 46871
+    },
+    {
+      "epoch": 1.1156812339331619,
+      "grad_norm": 5.655852791512025,
+      "learning_rate": 2e-05,
+      "loss": 0.4853,
+      "step": 46872
+    },
+    {
+      "epoch": 1.1157050366561934,
+      "grad_norm": 11.073662593521997,
+      "learning_rate": 2e-05,
+      "loss": 0.6746,
+      "step": 46873
+    },
+    {
+      "epoch": 1.115728839379225,
+      "grad_norm": 25.364952165922507,
+      "learning_rate": 2e-05,
+      "loss": 0.6507,
+      "step": 46874
+    },
+    {
+      "epoch": 1.1157526421022566,
+      "grad_norm": 3.9975334850657003,
+      "learning_rate": 2e-05,
+      "loss": 0.5372,
+      "step": 46875
+    },
+    {
+      "epoch": 1.115776444825288,
+      "grad_norm": 0.03328976983191621,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46876
+    },
+    {
+      "epoch": 1.1158002475483195,
+      "grad_norm": 3.6672654135218217,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 46877
+    },
+    {
+      "epoch": 1.115824050271351,
+      "grad_norm": 0.17942834987696446,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 46878
+    },
+    {
+      "epoch": 1.1158478529943825,
+      "grad_norm": 5.315912632146345,
+      "learning_rate": 2e-05,
+      "loss": 0.4368,
+      "step": 46879
+    },
+    {
+      "epoch": 1.115871655717414,
+      "grad_norm": 14.43913829142869,
+      "learning_rate": 2e-05,
+      "loss": 0.3241,
+      "step": 46880
+    },
+    {
+      "epoch": 1.1158954584404457,
+      "grad_norm": 0.4439403700433464,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46881
+    },
+    {
+      "epoch": 1.1159192611634772,
+      "grad_norm": 0.0742658001010934,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46882
+    },
+    {
+      "epoch": 1.1159430638865087,
+      "grad_norm": 5.119700707769544,
+      "learning_rate": 2e-05,
+      "loss": 0.1534,
+      "step": 46883
+    },
+    {
+      "epoch": 1.1159668666095401,
+      "grad_norm": 0.4804064141637527,
+      "learning_rate": 2e-05,
+      "loss": 0.0028,
+      "step": 46884
+    },
+    {
+      "epoch": 1.1159906693325716,
+      "grad_norm": 1.7077090391828185,
+      "learning_rate": 2e-05,
+      "loss": 0.0141,
+      "step": 46885
+    },
+    {
+      "epoch": 1.1160144720556031,
+      "grad_norm": 3.1981016291457287,
+      "learning_rate": 2e-05,
+      "loss": 0.0587,
+      "step": 46886
+    },
+    {
+      "epoch": 1.1160382747786346,
+      "grad_norm": 7.556349332715626,
+      "learning_rate": 2e-05,
+      "loss": 0.4415,
+      "step": 46887
+    },
+    {
+      "epoch": 1.1160620775016663,
+      "grad_norm": 0.9270501650015305,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 46888
+    },
+    {
+      "epoch": 1.1160858802246978,
+      "grad_norm": 1.3873411617650184,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 46889
+    },
+    {
+      "epoch": 1.1161096829477293,
+      "grad_norm": 5.97036909451349,
+      "learning_rate": 2e-05,
+      "loss": 0.3484,
+      "step": 46890
+    },
+    {
+      "epoch": 1.1161334856707608,
+      "grad_norm": 6.089274193430157,
+      "learning_rate": 2e-05,
+      "loss": 0.7299,
+      "step": 46891
+    },
+    {
+      "epoch": 1.1161572883937922,
+      "grad_norm": 0.0340174437068009,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46892
+    },
+    {
+      "epoch": 1.1161810911168237,
+      "grad_norm": 2.574466357463761,
+      "learning_rate": 2e-05,
+      "loss": 0.3977,
+      "step": 46893
+    },
+    {
+      "epoch": 1.1162048938398552,
+      "grad_norm": 7.727606810959999,
+      "learning_rate": 2e-05,
+      "loss": 0.2853,
+      "step": 46894
+    },
+    {
+      "epoch": 1.1162286965628867,
+      "grad_norm": 0.09298540090912862,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 46895
+    },
+    {
+      "epoch": 1.1162524992859184,
+      "grad_norm": 3.156712357239193,
+      "learning_rate": 2e-05,
+      "loss": 0.2167,
+      "step": 46896
+    },
+    {
+      "epoch": 1.11627630200895,
+      "grad_norm": 3.568814392301458,
+      "learning_rate": 2e-05,
+      "loss": 0.6804,
+      "step": 46897
+    },
+    {
+      "epoch": 1.1163001047319814,
+      "grad_norm": 1.6953326075257464,
+      "learning_rate": 2e-05,
+      "loss": 0.0227,
+      "step": 46898
+    },
+    {
+      "epoch": 1.1163239074550129,
+      "grad_norm": 0.03834015514855789,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46899
+    },
+    {
+      "epoch": 1.1163477101780444,
+      "grad_norm": 3.7450137675575927,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 46900
+    },
+    {
+      "epoch": 1.1163715129010758,
+      "grad_norm": 0.4906671012844041,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 46901
+    },
+    {
+      "epoch": 1.1163953156241073,
+      "grad_norm": 1.615771794320958,
+      "learning_rate": 2e-05,
+      "loss": 0.0184,
+      "step": 46902
+    },
+    {
+      "epoch": 1.1164191183471388,
+      "grad_norm": 4.4100673517731295,
+      "learning_rate": 2e-05,
+      "loss": 0.4255,
+      "step": 46903
+    },
+    {
+      "epoch": 1.1164429210701705,
+      "grad_norm": 16.2670409054259,
+      "learning_rate": 2e-05,
+      "loss": 0.7001,
+      "step": 46904
+    },
+    {
+      "epoch": 1.116466723793202,
+      "grad_norm": 0.017756920095309065,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 46905
+    },
+    {
+      "epoch": 1.1164905265162335,
+      "grad_norm": 4.868497848272205,
+      "learning_rate": 2e-05,
+      "loss": 0.5181,
+      "step": 46906
+    },
+    {
+      "epoch": 1.116514329239265,
+      "grad_norm": 0.060300831001009754,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46907
+    },
+    {
+      "epoch": 1.1165381319622965,
+      "grad_norm": 4.3269662183746105,
+      "learning_rate": 2e-05,
+      "loss": 0.0269,
+      "step": 46908
+    },
+    {
+      "epoch": 1.116561934685328,
+      "grad_norm": 40.37519017238886,
+      "learning_rate": 2e-05,
+      "loss": 0.4612,
+      "step": 46909
+    },
+    {
+      "epoch": 1.1165857374083594,
+      "grad_norm": 2.3662972217403393,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 46910
+    },
+    {
+      "epoch": 1.1166095401313911,
+      "grad_norm": 8.559828315635858,
+      "learning_rate": 2e-05,
+      "loss": 1.3344,
+      "step": 46911
+    },
+    {
+      "epoch": 1.1166333428544226,
+      "grad_norm": 4.039142278410087,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 46912
+    },
+    {
+      "epoch": 1.116657145577454,
+      "grad_norm": 0.010636349616833221,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46913
+    },
+    {
+      "epoch": 1.1166809483004856,
+      "grad_norm": 0.33374700424029285,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 46914
+    },
+    {
+      "epoch": 1.116704751023517,
+      "grad_norm": 0.3639416290744143,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 46915
+    },
+    {
+      "epoch": 1.1167285537465486,
+      "grad_norm": 2.903155528273494,
+      "learning_rate": 2e-05,
+      "loss": 0.1046,
+      "step": 46916
+    },
+    {
+      "epoch": 1.11675235646958,
+      "grad_norm": 0.23260655568091565,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46917
+    },
+    {
+      "epoch": 1.1167761591926118,
+      "grad_norm": 12.523162852907818,
+      "learning_rate": 2e-05,
+      "loss": 0.298,
+      "step": 46918
+    },
+    {
+      "epoch": 1.1167999619156432,
+      "grad_norm": 0.7899488289669735,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 46919
+    },
+    {
+      "epoch": 1.1168237646386747,
+      "grad_norm": 15.411069872869323,
+      "learning_rate": 2e-05,
+      "loss": 0.1167,
+      "step": 46920
+    },
+    {
+      "epoch": 1.1168475673617062,
+      "grad_norm": 6.21593756491444,
+      "learning_rate": 2e-05,
+      "loss": 0.5195,
+      "step": 46921
+    },
+    {
+      "epoch": 1.1168713700847377,
+      "grad_norm": 0.13485874715991142,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46922
+    },
+    {
+      "epoch": 1.1168951728077692,
+      "grad_norm": 0.007209137052828352,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 46923
+    },
+    {
+      "epoch": 1.1169189755308007,
+      "grad_norm": 0.05046205610125108,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46924
+    },
+    {
+      "epoch": 1.1169427782538321,
+      "grad_norm": 0.02325079325841244,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 46925
+    },
+    {
+      "epoch": 1.1169665809768639,
+      "grad_norm": 1.3800662350875164,
+      "learning_rate": 2e-05,
+      "loss": 0.013,
+      "step": 46926
+    },
+    {
+      "epoch": 1.1169903836998953,
+      "grad_norm": 0.1015787579910621,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 46927
+    },
+    {
+      "epoch": 1.1170141864229268,
+      "grad_norm": 9.032780706454554,
+      "learning_rate": 2e-05,
+      "loss": 0.3596,
+      "step": 46928
+    },
+    {
+      "epoch": 1.1170379891459583,
+      "grad_norm": 1.6637773146604737,
+      "learning_rate": 2e-05,
+      "loss": 0.0192,
+      "step": 46929
+    },
+    {
+      "epoch": 1.1170617918689898,
+      "grad_norm": 0.07642372084180968,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 46930
+    },
+    {
+      "epoch": 1.1170855945920213,
+      "grad_norm": 3.522922417802815,
+      "learning_rate": 2e-05,
+      "loss": 0.3428,
+      "step": 46931
+    },
+    {
+      "epoch": 1.1171093973150528,
+      "grad_norm": 1.9464592866976773,
+      "learning_rate": 2e-05,
+      "loss": 0.0099,
+      "step": 46932
+    },
+    {
+      "epoch": 1.1171332000380843,
+      "grad_norm": 2.321087678268462,
+      "learning_rate": 2e-05,
+      "loss": 0.1437,
+      "step": 46933
+    },
+    {
+      "epoch": 1.117157002761116,
+      "grad_norm": 0.0058429719215287905,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 46934
+    },
+    {
+      "epoch": 1.1171808054841474,
+      "grad_norm": 2.5329420104217664,
+      "learning_rate": 2e-05,
+      "loss": 0.2159,
+      "step": 46935
+    },
+    {
+      "epoch": 1.117204608207179,
+      "grad_norm": 3.8880304180794476,
+      "learning_rate": 2e-05,
+      "loss": 0.1294,
+      "step": 46936
+    },
+    {
+      "epoch": 1.1172284109302104,
+      "grad_norm": 4.398606957290489,
+      "learning_rate": 2e-05,
+      "loss": 0.4217,
+      "step": 46937
+    },
+    {
+      "epoch": 1.117252213653242,
+      "grad_norm": 4.354650833534538,
+      "learning_rate": 2e-05,
+      "loss": 0.054,
+      "step": 46938
+    },
+    {
+      "epoch": 1.1172760163762734,
+      "grad_norm": 8.784945049031087,
+      "learning_rate": 2e-05,
+      "loss": 0.4208,
+      "step": 46939
+    },
+    {
+      "epoch": 1.1172998190993049,
+      "grad_norm": 2.990682068370793,
+      "learning_rate": 2e-05,
+      "loss": 0.3226,
+      "step": 46940
+    },
+    {
+      "epoch": 1.1173236218223366,
+      "grad_norm": 4.666802341475099,
+      "learning_rate": 2e-05,
+      "loss": 0.3379,
+      "step": 46941
+    },
+    {
+      "epoch": 1.117347424545368,
+      "grad_norm": 0.24747595354447277,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 46942
+    },
+    {
+      "epoch": 1.1173712272683995,
+      "grad_norm": 3.101138507061853,
+      "learning_rate": 2e-05,
+      "loss": 0.0432,
+      "step": 46943
+    },
+    {
+      "epoch": 1.117395029991431,
+      "grad_norm": 4.166374465929632,
+      "learning_rate": 2e-05,
+      "loss": 0.6031,
+      "step": 46944
+    },
+    {
+      "epoch": 1.1174188327144625,
+      "grad_norm": 5.5365395682545735,
+      "learning_rate": 2e-05,
+      "loss": 0.9812,
+      "step": 46945
+    },
+    {
+      "epoch": 1.117442635437494,
+      "grad_norm": 4.458330239439385,
+      "learning_rate": 2e-05,
+      "loss": 0.6074,
+      "step": 46946
+    },
+    {
+      "epoch": 1.1174664381605255,
+      "grad_norm": 0.10452794452438718,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46947
+    },
+    {
+      "epoch": 1.1174902408835572,
+      "grad_norm": 0.60477061407893,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 46948
+    },
+    {
+      "epoch": 1.1175140436065887,
+      "grad_norm": 0.40251215001673507,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 46949
+    },
+    {
+      "epoch": 1.1175378463296202,
+      "grad_norm": 0.04359219653388818,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46950
+    },
+    {
+      "epoch": 1.1175616490526517,
+      "grad_norm": 2.6507060195472154,
+      "learning_rate": 2e-05,
+      "loss": 0.2836,
+      "step": 46951
+    },
+    {
+      "epoch": 1.1175854517756831,
+      "grad_norm": 6.640833302937184,
+      "learning_rate": 2e-05,
+      "loss": 0.5783,
+      "step": 46952
+    },
+    {
+      "epoch": 1.1176092544987146,
+      "grad_norm": 0.03355128313417304,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 46953
+    },
+    {
+      "epoch": 1.117633057221746,
+      "grad_norm": 2.5752641690175913,
+      "learning_rate": 2e-05,
+      "loss": 0.3742,
+      "step": 46954
+    },
+    {
+      "epoch": 1.1176568599447776,
+      "grad_norm": 4.477524583676956,
+      "learning_rate": 2e-05,
+      "loss": 0.2246,
+      "step": 46955
+    },
+    {
+      "epoch": 1.1176806626678093,
+      "grad_norm": 6.165270726652403,
+      "learning_rate": 2e-05,
+      "loss": 0.8573,
+      "step": 46956
+    },
+    {
+      "epoch": 1.1177044653908408,
+      "grad_norm": 0.07244020650106653,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 46957
+    },
+    {
+      "epoch": 1.1177282681138723,
+      "grad_norm": 9.069710275445475,
+      "learning_rate": 2e-05,
+      "loss": 2.1496,
+      "step": 46958
+    },
+    {
+      "epoch": 1.1177520708369038,
+      "grad_norm": 1.3554961824546325,
+      "learning_rate": 2e-05,
+      "loss": 0.0153,
+      "step": 46959
+    },
+    {
+      "epoch": 1.1177758735599352,
+      "grad_norm": 6.054654226577543,
+      "learning_rate": 2e-05,
+      "loss": 0.8969,
+      "step": 46960
+    },
+    {
+      "epoch": 1.1177996762829667,
+      "grad_norm": 6.042261399283995,
+      "learning_rate": 2e-05,
+      "loss": 0.8259,
+      "step": 46961
+    },
+    {
+      "epoch": 1.1178234790059982,
+      "grad_norm": 0.1237280238915267,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46962
+    },
+    {
+      "epoch": 1.1178472817290297,
+      "grad_norm": 0.04425028935863279,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 46963
+    },
+    {
+      "epoch": 1.1178710844520614,
+      "grad_norm": 4.487344885647564,
+      "learning_rate": 2e-05,
+      "loss": 0.3015,
+      "step": 46964
+    },
+    {
+      "epoch": 1.1178948871750929,
+      "grad_norm": 3.679341574960611,
+      "learning_rate": 2e-05,
+      "loss": 0.4675,
+      "step": 46965
+    },
+    {
+      "epoch": 1.1179186898981244,
+      "grad_norm": 1.762476310006981,
+      "learning_rate": 2e-05,
+      "loss": 0.1396,
+      "step": 46966
+    },
+    {
+      "epoch": 1.1179424926211559,
+      "grad_norm": 2.89628848012398,
+      "learning_rate": 2e-05,
+      "loss": 0.1689,
+      "step": 46967
+    },
+    {
+      "epoch": 1.1179662953441873,
+      "grad_norm": 2.8229131919579245,
+      "learning_rate": 2e-05,
+      "loss": 0.291,
+      "step": 46968
+    },
+    {
+      "epoch": 1.1179900980672188,
+      "grad_norm": 1.540804565713018,
+      "learning_rate": 2e-05,
+      "loss": 0.0222,
+      "step": 46969
+    },
+    {
+      "epoch": 1.1180139007902503,
+      "grad_norm": 4.189174874206157,
+      "learning_rate": 2e-05,
+      "loss": 0.4937,
+      "step": 46970
+    },
+    {
+      "epoch": 1.118037703513282,
+      "grad_norm": 3.999301011796878,
+      "learning_rate": 2e-05,
+      "loss": 0.5768,
+      "step": 46971
+    },
+    {
+      "epoch": 1.1180615062363135,
+      "grad_norm": 2.8842890522468623,
+      "learning_rate": 2e-05,
+      "loss": 0.2332,
+      "step": 46972
+    },
+    {
+      "epoch": 1.118085308959345,
+      "grad_norm": 3.3319570618846224,
+      "learning_rate": 2e-05,
+      "loss": 0.2298,
+      "step": 46973
+    },
+    {
+      "epoch": 1.1181091116823765,
+      "grad_norm": 10.207475841111696,
+      "learning_rate": 2e-05,
+      "loss": 0.6525,
+      "step": 46974
+    },
+    {
+      "epoch": 1.118132914405408,
+      "grad_norm": 2.5444165371359273,
+      "learning_rate": 2e-05,
+      "loss": 0.0345,
+      "step": 46975
+    },
+    {
+      "epoch": 1.1181567171284394,
+      "grad_norm": 0.12537925391103963,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 46976
+    },
+    {
+      "epoch": 1.118180519851471,
+      "grad_norm": 5.244096068095516,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 46977
+    },
+    {
+      "epoch": 1.1182043225745026,
+      "grad_norm": 26.4980380215844,
+      "learning_rate": 2e-05,
+      "loss": 1.7662,
+      "step": 46978
+    },
+    {
+      "epoch": 1.1182281252975341,
+      "grad_norm": 0.02072923857582262,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 46979
+    },
+    {
+      "epoch": 1.1182519280205656,
+      "grad_norm": 7.995838339531158,
+      "learning_rate": 2e-05,
+      "loss": 0.0961,
+      "step": 46980
+    },
+    {
+      "epoch": 1.118275730743597,
+      "grad_norm": 4.745774744658234,
+      "learning_rate": 2e-05,
+      "loss": 0.3053,
+      "step": 46981
+    },
+    {
+      "epoch": 1.1182995334666286,
+      "grad_norm": 2.5710809962542327,
+      "learning_rate": 2e-05,
+      "loss": 0.0267,
+      "step": 46982
+    },
+    {
+      "epoch": 1.11832333618966,
+      "grad_norm": 0.20069896073345644,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 46983
+    },
+    {
+      "epoch": 1.1183471389126916,
+      "grad_norm": 6.0958710602819455,
+      "learning_rate": 2e-05,
+      "loss": 0.0538,
+      "step": 46984
+    },
+    {
+      "epoch": 1.118370941635723,
+      "grad_norm": 2.3972403729184006,
+      "learning_rate": 2e-05,
+      "loss": 0.334,
+      "step": 46985
+    },
+    {
+      "epoch": 1.1183947443587547,
+      "grad_norm": 11.089478754130445,
+      "learning_rate": 2e-05,
+      "loss": 0.4319,
+      "step": 46986
+    },
+    {
+      "epoch": 1.1184185470817862,
+      "grad_norm": 2.4513231524530616,
+      "learning_rate": 2e-05,
+      "loss": 0.0605,
+      "step": 46987
+    },
+    {
+      "epoch": 1.1184423498048177,
+      "grad_norm": 0.14613968870203298,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 46988
+    },
+    {
+      "epoch": 1.1184661525278492,
+      "grad_norm": 3.506202192675637,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 46989
+    },
+    {
+      "epoch": 1.1184899552508807,
+      "grad_norm": 0.8911138898081968,
+      "learning_rate": 2e-05,
+      "loss": 0.0088,
+      "step": 46990
+    },
+    {
+      "epoch": 1.1185137579739122,
+      "grad_norm": 0.11421221932814452,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 46991
+    },
+    {
+      "epoch": 1.1185375606969437,
+      "grad_norm": 0.10291897089448822,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 46992
+    },
+    {
+      "epoch": 1.1185613634199751,
+      "grad_norm": 7.704242278412133,
+      "learning_rate": 2e-05,
+      "loss": 0.58,
+      "step": 46993
+    },
+    {
+      "epoch": 1.1185851661430068,
+      "grad_norm": 1.987767877591031,
+      "learning_rate": 2e-05,
+      "loss": 0.152,
+      "step": 46994
+    },
+    {
+      "epoch": 1.1186089688660383,
+      "grad_norm": 3.2698196616292363,
+      "learning_rate": 2e-05,
+      "loss": 0.3475,
+      "step": 46995
+    },
+    {
+      "epoch": 1.1186327715890698,
+      "grad_norm": 0.11271815412545491,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 46996
+    },
+    {
+      "epoch": 1.1186565743121013,
+      "grad_norm": 3.100481002298795,
+      "learning_rate": 2e-05,
+      "loss": 0.2597,
+      "step": 46997
+    },
+    {
+      "epoch": 1.1186803770351328,
+      "grad_norm": 3.284475505579065,
+      "learning_rate": 2e-05,
+      "loss": 0.0803,
+      "step": 46998
+    },
+    {
+      "epoch": 1.1187041797581643,
+      "grad_norm": 0.9553065180698308,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 46999
+    },
+    {
+      "epoch": 1.1187279824811958,
+      "grad_norm": 3.146591985073169,
+      "learning_rate": 2e-05,
+      "loss": 0.2716,
+      "step": 47000
+    },
+    {
+      "epoch": 1.1187517852042275,
+      "grad_norm": 7.0036389367101055,
+      "learning_rate": 2e-05,
+      "loss": 0.1259,
+      "step": 47001
+    },
+    {
+      "epoch": 1.118775587927259,
+      "grad_norm": 71.57979584534738,
+      "learning_rate": 2e-05,
+      "loss": 0.3541,
+      "step": 47002
+    },
+    {
+      "epoch": 1.1187993906502904,
+      "grad_norm": 3.6463274258346443,
+      "learning_rate": 2e-05,
+      "loss": 0.0487,
+      "step": 47003
+    },
+    {
+      "epoch": 1.118823193373322,
+      "grad_norm": 0.08061858862642121,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47004
+    },
+    {
+      "epoch": 1.1188469960963534,
+      "grad_norm": 16.790184268691586,
+      "learning_rate": 2e-05,
+      "loss": 0.1148,
+      "step": 47005
+    },
+    {
+      "epoch": 1.118870798819385,
+      "grad_norm": 5.444667234027648,
+      "learning_rate": 2e-05,
+      "loss": 0.5367,
+      "step": 47006
+    },
+    {
+      "epoch": 1.1188946015424164,
+      "grad_norm": 5.921427596356901,
+      "learning_rate": 2e-05,
+      "loss": 0.5295,
+      "step": 47007
+    },
+    {
+      "epoch": 1.1189184042654479,
+      "grad_norm": 16.965858693158328,
+      "learning_rate": 2e-05,
+      "loss": 0.6811,
+      "step": 47008
+    },
+    {
+      "epoch": 1.1189422069884796,
+      "grad_norm": 12.911739274486838,
+      "learning_rate": 2e-05,
+      "loss": 0.0515,
+      "step": 47009
+    },
+    {
+      "epoch": 1.118966009711511,
+      "grad_norm": 12.24605475565806,
+      "learning_rate": 2e-05,
+      "loss": 0.2889,
+      "step": 47010
+    },
+    {
+      "epoch": 1.1189898124345425,
+      "grad_norm": 0.8158981748629505,
+      "learning_rate": 2e-05,
+      "loss": 0.018,
+      "step": 47011
+    },
+    {
+      "epoch": 1.119013615157574,
+      "grad_norm": 5.334692184752187,
+      "learning_rate": 2e-05,
+      "loss": 0.565,
+      "step": 47012
+    },
+    {
+      "epoch": 1.1190374178806055,
+      "grad_norm": 3.111343433129088,
+      "learning_rate": 2e-05,
+      "loss": 0.0215,
+      "step": 47013
+    },
+    {
+      "epoch": 1.119061220603637,
+      "grad_norm": 1.1924988764334532,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 47014
+    },
+    {
+      "epoch": 1.1190850233266685,
+      "grad_norm": 3.655812511336498,
+      "learning_rate": 2e-05,
+      "loss": 0.3277,
+      "step": 47015
+    },
+    {
+      "epoch": 1.1191088260497002,
+      "grad_norm": 1.7576384477130689,
+      "learning_rate": 2e-05,
+      "loss": 0.1233,
+      "step": 47016
+    },
+    {
+      "epoch": 1.1191326287727317,
+      "grad_norm": 3.7369937214755304,
+      "learning_rate": 2e-05,
+      "loss": 0.2575,
+      "step": 47017
+    },
+    {
+      "epoch": 1.1191564314957632,
+      "grad_norm": 1.6316477734733263,
+      "learning_rate": 2e-05,
+      "loss": 0.1842,
+      "step": 47018
+    },
+    {
+      "epoch": 1.1191802342187946,
+      "grad_norm": 3.4042075193155226,
+      "learning_rate": 2e-05,
+      "loss": 0.0536,
+      "step": 47019
+    },
+    {
+      "epoch": 1.1192040369418261,
+      "grad_norm": 3.9165625926857373,
+      "learning_rate": 2e-05,
+      "loss": 0.2086,
+      "step": 47020
+    },
+    {
+      "epoch": 1.1192278396648576,
+      "grad_norm": 0.7988457274867771,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 47021
+    },
+    {
+      "epoch": 1.119251642387889,
+      "grad_norm": 3.6181626462878933,
+      "learning_rate": 2e-05,
+      "loss": 0.1466,
+      "step": 47022
+    },
+    {
+      "epoch": 1.1192754451109206,
+      "grad_norm": 3.4113375655177354,
+      "learning_rate": 2e-05,
+      "loss": 0.2859,
+      "step": 47023
+    },
+    {
+      "epoch": 1.1192992478339523,
+      "grad_norm": 0.063557177899517,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47024
+    },
+    {
+      "epoch": 1.1193230505569838,
+      "grad_norm": 14.756752100806986,
+      "learning_rate": 2e-05,
+      "loss": 0.148,
+      "step": 47025
+    },
+    {
+      "epoch": 1.1193468532800153,
+      "grad_norm": 0.1309689411430484,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47026
+    },
+    {
+      "epoch": 1.1193706560030467,
+      "grad_norm": 9.960295712009106,
+      "learning_rate": 2e-05,
+      "loss": 0.187,
+      "step": 47027
+    },
+    {
+      "epoch": 1.1193944587260782,
+      "grad_norm": 9.591902645137825,
+      "learning_rate": 2e-05,
+      "loss": 0.414,
+      "step": 47028
+    },
+    {
+      "epoch": 1.1194182614491097,
+      "grad_norm": 3.4717735859099346,
+      "learning_rate": 2e-05,
+      "loss": 0.549,
+      "step": 47029
+    },
+    {
+      "epoch": 1.1194420641721412,
+      "grad_norm": 0.23255182876487032,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 47030
+    },
+    {
+      "epoch": 1.119465866895173,
+      "grad_norm": 4.98671741660743,
+      "learning_rate": 2e-05,
+      "loss": 0.6514,
+      "step": 47031
+    },
+    {
+      "epoch": 1.1194896696182044,
+      "grad_norm": 2.882050491367749,
+      "learning_rate": 2e-05,
+      "loss": 0.3272,
+      "step": 47032
+    },
+    {
+      "epoch": 1.1195134723412359,
+      "grad_norm": 5.481046173511082,
+      "learning_rate": 2e-05,
+      "loss": 0.5176,
+      "step": 47033
+    },
+    {
+      "epoch": 1.1195372750642674,
+      "grad_norm": 9.577255913895815,
+      "learning_rate": 2e-05,
+      "loss": 0.8576,
+      "step": 47034
+    },
+    {
+      "epoch": 1.1195610777872989,
+      "grad_norm": 272.72717907426596,
+      "learning_rate": 2e-05,
+      "loss": 0.1688,
+      "step": 47035
+    },
+    {
+      "epoch": 1.1195848805103303,
+      "grad_norm": 0.0369162112081382,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47036
+    },
+    {
+      "epoch": 1.1196086832333618,
+      "grad_norm": 2.4724301796113375,
+      "learning_rate": 2e-05,
+      "loss": 0.2114,
+      "step": 47037
+    },
+    {
+      "epoch": 1.1196324859563933,
+      "grad_norm": 4.4986140146001325,
+      "learning_rate": 2e-05,
+      "loss": 0.5445,
+      "step": 47038
+    },
+    {
+      "epoch": 1.119656288679425,
+      "grad_norm": 3.663128249729411,
+      "learning_rate": 2e-05,
+      "loss": 0.5217,
+      "step": 47039
+    },
+    {
+      "epoch": 1.1196800914024565,
+      "grad_norm": 7.07161312913519,
+      "learning_rate": 2e-05,
+      "loss": 0.6482,
+      "step": 47040
+    },
+    {
+      "epoch": 1.119703894125488,
+      "grad_norm": 0.04061284002064106,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47041
+    },
+    {
+      "epoch": 1.1197276968485195,
+      "grad_norm": 16.338028448124014,
+      "learning_rate": 2e-05,
+      "loss": 0.3862,
+      "step": 47042
+    },
+    {
+      "epoch": 1.119751499571551,
+      "grad_norm": 14.481200834508165,
+      "learning_rate": 2e-05,
+      "loss": 0.2818,
+      "step": 47043
+    },
+    {
+      "epoch": 1.1197753022945824,
+      "grad_norm": 3.6899257917790846,
+      "learning_rate": 2e-05,
+      "loss": 0.6012,
+      "step": 47044
+    },
+    {
+      "epoch": 1.119799105017614,
+      "grad_norm": 2.915507187765266,
+      "learning_rate": 2e-05,
+      "loss": 0.0397,
+      "step": 47045
+    },
+    {
+      "epoch": 1.1198229077406454,
+      "grad_norm": 7.740059130602268,
+      "learning_rate": 2e-05,
+      "loss": 0.2371,
+      "step": 47046
+    },
+    {
+      "epoch": 1.1198467104636771,
+      "grad_norm": 5.168902988708532,
+      "learning_rate": 2e-05,
+      "loss": 0.5134,
+      "step": 47047
+    },
+    {
+      "epoch": 1.1198705131867086,
+      "grad_norm": 3.786817055825433,
+      "learning_rate": 2e-05,
+      "loss": 0.4567,
+      "step": 47048
+    },
+    {
+      "epoch": 1.11989431590974,
+      "grad_norm": 0.3458949208768839,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47049
+    },
+    {
+      "epoch": 1.1199181186327716,
+      "grad_norm": 2.1988708760594946,
+      "learning_rate": 2e-05,
+      "loss": 0.176,
+      "step": 47050
+    },
+    {
+      "epoch": 1.119941921355803,
+      "grad_norm": 0.017958893911328933,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47051
+    },
+    {
+      "epoch": 1.1199657240788345,
+      "grad_norm": 1.3369515693403209,
+      "learning_rate": 2e-05,
+      "loss": 0.0085,
+      "step": 47052
+    },
+    {
+      "epoch": 1.119989526801866,
+      "grad_norm": 0.06513171695848358,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47053
+    },
+    {
+      "epoch": 1.1200133295248977,
+      "grad_norm": 0.49202594817410267,
+      "learning_rate": 2e-05,
+      "loss": 0.0065,
+      "step": 47054
+    },
+    {
+      "epoch": 1.1200371322479292,
+      "grad_norm": 8.809089508550924,
+      "learning_rate": 2e-05,
+      "loss": 0.2668,
+      "step": 47055
+    },
+    {
+      "epoch": 1.1200609349709607,
+      "grad_norm": 5.770916844716327,
+      "learning_rate": 2e-05,
+      "loss": 0.6643,
+      "step": 47056
+    },
+    {
+      "epoch": 1.1200847376939922,
+      "grad_norm": 3.104341168808112,
+      "learning_rate": 2e-05,
+      "loss": 0.0877,
+      "step": 47057
+    },
+    {
+      "epoch": 1.1201085404170237,
+      "grad_norm": 0.4757599832802922,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 47058
+    },
+    {
+      "epoch": 1.1201323431400552,
+      "grad_norm": 4.892198192045842,
+      "learning_rate": 2e-05,
+      "loss": 0.4656,
+      "step": 47059
+    },
+    {
+      "epoch": 1.1201561458630866,
+      "grad_norm": 10.19028229548342,
+      "learning_rate": 2e-05,
+      "loss": 0.5387,
+      "step": 47060
+    },
+    {
+      "epoch": 1.1201799485861184,
+      "grad_norm": 1.9548077494259233,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 47061
+    },
+    {
+      "epoch": 1.1202037513091498,
+      "grad_norm": 5.211971781539667,
+      "learning_rate": 2e-05,
+      "loss": 0.5252,
+      "step": 47062
+    },
+    {
+      "epoch": 1.1202275540321813,
+      "grad_norm": 7.46517167164001,
+      "learning_rate": 2e-05,
+      "loss": 0.8179,
+      "step": 47063
+    },
+    {
+      "epoch": 1.1202513567552128,
+      "grad_norm": 0.8161013602590207,
+      "learning_rate": 2e-05,
+      "loss": 0.0072,
+      "step": 47064
+    },
+    {
+      "epoch": 1.1202751594782443,
+      "grad_norm": 0.0839471338831363,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47065
+    },
+    {
+      "epoch": 1.1202989622012758,
+      "grad_norm": 0.32933295337643986,
+      "learning_rate": 2e-05,
+      "loss": 0.0055,
+      "step": 47066
+    },
+    {
+      "epoch": 1.1203227649243073,
+      "grad_norm": 3.6475372706213274,
+      "learning_rate": 2e-05,
+      "loss": 0.4858,
+      "step": 47067
+    },
+    {
+      "epoch": 1.1203465676473388,
+      "grad_norm": 390.36656987166964,
+      "learning_rate": 2e-05,
+      "loss": 0.3886,
+      "step": 47068
+    },
+    {
+      "epoch": 1.1203703703703705,
+      "grad_norm": 3.73090035192089,
+      "learning_rate": 2e-05,
+      "loss": 0.4693,
+      "step": 47069
+    },
+    {
+      "epoch": 1.120394173093402,
+      "grad_norm": 5.252847821117913,
+      "learning_rate": 2e-05,
+      "loss": 0.0368,
+      "step": 47070
+    },
+    {
+      "epoch": 1.1204179758164334,
+      "grad_norm": 4.6095310717522295,
+      "learning_rate": 2e-05,
+      "loss": 0.7967,
+      "step": 47071
+    },
+    {
+      "epoch": 1.120441778539465,
+      "grad_norm": 6.877725333296785,
+      "learning_rate": 2e-05,
+      "loss": 0.13,
+      "step": 47072
+    },
+    {
+      "epoch": 1.1204655812624964,
+      "grad_norm": 18.378867975498707,
+      "learning_rate": 2e-05,
+      "loss": 0.2769,
+      "step": 47073
+    },
+    {
+      "epoch": 1.1204893839855279,
+      "grad_norm": 0.151301137069757,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47074
+    },
+    {
+      "epoch": 1.1205131867085594,
+      "grad_norm": 9.265362383953642,
+      "learning_rate": 2e-05,
+      "loss": 0.4567,
+      "step": 47075
+    },
+    {
+      "epoch": 1.1205369894315909,
+      "grad_norm": 0.09478323559058531,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 47076
+    },
+    {
+      "epoch": 1.1205607921546226,
+      "grad_norm": 3.678082963016357,
+      "learning_rate": 2e-05,
+      "loss": 0.3043,
+      "step": 47077
+    },
+    {
+      "epoch": 1.120584594877654,
+      "grad_norm": 3.7083332396902304,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 47078
+    },
+    {
+      "epoch": 1.1206083976006855,
+      "grad_norm": 0.0418296491767142,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47079
+    },
+    {
+      "epoch": 1.120632200323717,
+      "grad_norm": 0.08853823040615312,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 47080
+    },
+    {
+      "epoch": 1.1206560030467485,
+      "grad_norm": 6.6378173711581505,
+      "learning_rate": 2e-05,
+      "loss": 0.7488,
+      "step": 47081
+    },
+    {
+      "epoch": 1.12067980576978,
+      "grad_norm": 3.214793241499251,
+      "learning_rate": 2e-05,
+      "loss": 0.042,
+      "step": 47082
+    },
+    {
+      "epoch": 1.1207036084928115,
+      "grad_norm": 1.2483631983870154,
+      "learning_rate": 2e-05,
+      "loss": 0.0203,
+      "step": 47083
+    },
+    {
+      "epoch": 1.1207274112158432,
+      "grad_norm": 2.8364606059645783,
+      "learning_rate": 2e-05,
+      "loss": 0.0265,
+      "step": 47084
+    },
+    {
+      "epoch": 1.1207512139388747,
+      "grad_norm": 1.6016194944069382,
+      "learning_rate": 2e-05,
+      "loss": 0.0325,
+      "step": 47085
+    },
+    {
+      "epoch": 1.1207750166619062,
+      "grad_norm": 11.31933414191766,
+      "learning_rate": 2e-05,
+      "loss": 0.4927,
+      "step": 47086
+    },
+    {
+      "epoch": 1.1207988193849376,
+      "grad_norm": 0.06087515336598494,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47087
+    },
+    {
+      "epoch": 1.1208226221079691,
+      "grad_norm": 1.0189862386998487,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 47088
+    },
+    {
+      "epoch": 1.1208464248310006,
+      "grad_norm": 3.3158840179777846,
+      "learning_rate": 2e-05,
+      "loss": 0.1935,
+      "step": 47089
+    },
+    {
+      "epoch": 1.120870227554032,
+      "grad_norm": 6.37867360659211,
+      "learning_rate": 2e-05,
+      "loss": 0.0818,
+      "step": 47090
+    },
+    {
+      "epoch": 1.1208940302770638,
+      "grad_norm": 3.7729044217479526,
+      "learning_rate": 2e-05,
+      "loss": 0.0954,
+      "step": 47091
+    },
+    {
+      "epoch": 1.1209178330000953,
+      "grad_norm": 0.03708503408299208,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47092
+    },
+    {
+      "epoch": 1.1209416357231268,
+      "grad_norm": 2.90362501102357,
+      "learning_rate": 2e-05,
+      "loss": 0.0406,
+      "step": 47093
+    },
+    {
+      "epoch": 1.1209654384461583,
+      "grad_norm": 5.1584436800153854,
+      "learning_rate": 2e-05,
+      "loss": 0.0244,
+      "step": 47094
+    },
+    {
+      "epoch": 1.1209892411691897,
+      "grad_norm": 3.8328358085396563,
+      "learning_rate": 2e-05,
+      "loss": 0.6143,
+      "step": 47095
+    },
+    {
+      "epoch": 1.1210130438922212,
+      "grad_norm": 0.3392537639423346,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47096
+    },
+    {
+      "epoch": 1.1210368466152527,
+      "grad_norm": 0.05453272503502931,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47097
+    },
+    {
+      "epoch": 1.1210606493382842,
+      "grad_norm": 5.407245669693155,
+      "learning_rate": 2e-05,
+      "loss": 0.5404,
+      "step": 47098
+    },
+    {
+      "epoch": 1.121084452061316,
+      "grad_norm": 4.258866089428166,
+      "learning_rate": 2e-05,
+      "loss": 0.2502,
+      "step": 47099
+    },
+    {
+      "epoch": 1.1211082547843474,
+      "grad_norm": 3.765986079585758,
+      "learning_rate": 2e-05,
+      "loss": 0.1047,
+      "step": 47100
+    },
+    {
+      "epoch": 1.1211320575073789,
+      "grad_norm": 3.2747994125828632,
+      "learning_rate": 2e-05,
+      "loss": 0.3928,
+      "step": 47101
+    },
+    {
+      "epoch": 1.1211558602304104,
+      "grad_norm": 6.265936692924152,
+      "learning_rate": 2e-05,
+      "loss": 0.5354,
+      "step": 47102
+    },
+    {
+      "epoch": 1.1211796629534418,
+      "grad_norm": 7.864351222691368,
+      "learning_rate": 2e-05,
+      "loss": 0.5658,
+      "step": 47103
+    },
+    {
+      "epoch": 1.1212034656764733,
+      "grad_norm": 0.9427344302685178,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 47104
+    },
+    {
+      "epoch": 1.1212272683995048,
+      "grad_norm": 4.513904548048067,
+      "learning_rate": 2e-05,
+      "loss": 0.3614,
+      "step": 47105
+    },
+    {
+      "epoch": 1.1212510711225363,
+      "grad_norm": 2.4886400698392146,
+      "learning_rate": 2e-05,
+      "loss": 0.3438,
+      "step": 47106
+    },
+    {
+      "epoch": 1.121274873845568,
+      "grad_norm": 5.100169856793969,
+      "learning_rate": 2e-05,
+      "loss": 0.5698,
+      "step": 47107
+    },
+    {
+      "epoch": 1.1212986765685995,
+      "grad_norm": 3.387688601171746,
+      "learning_rate": 2e-05,
+      "loss": 0.4216,
+      "step": 47108
+    },
+    {
+      "epoch": 1.121322479291631,
+      "grad_norm": 4.314496060254935,
+      "learning_rate": 2e-05,
+      "loss": 0.0438,
+      "step": 47109
+    },
+    {
+      "epoch": 1.1213462820146625,
+      "grad_norm": 7.215563917226154,
+      "learning_rate": 2e-05,
+      "loss": 1.4305,
+      "step": 47110
+    },
+    {
+      "epoch": 1.121370084737694,
+      "grad_norm": 2.3736847504300385,
+      "learning_rate": 2e-05,
+      "loss": 0.3022,
+      "step": 47111
+    },
+    {
+      "epoch": 1.1213938874607254,
+      "grad_norm": 0.04656187615834635,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47112
+    },
+    {
+      "epoch": 1.121417690183757,
+      "grad_norm": 9.311279944448652,
+      "learning_rate": 2e-05,
+      "loss": 0.0282,
+      "step": 47113
+    },
+    {
+      "epoch": 1.1214414929067886,
+      "grad_norm": 0.1693214086909933,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47114
+    },
+    {
+      "epoch": 1.1214652956298201,
+      "grad_norm": 0.19252783820331942,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47115
+    },
+    {
+      "epoch": 1.1214890983528516,
+      "grad_norm": 15.426346709459667,
+      "learning_rate": 2e-05,
+      "loss": 0.5417,
+      "step": 47116
+    },
+    {
+      "epoch": 1.121512901075883,
+      "grad_norm": 0.22273392555384564,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 47117
+    },
+    {
+      "epoch": 1.1215367037989146,
+      "grad_norm": 3.628864703809571,
+      "learning_rate": 2e-05,
+      "loss": 0.4484,
+      "step": 47118
+    },
+    {
+      "epoch": 1.121560506521946,
+      "grad_norm": 2.4184207663592736,
+      "learning_rate": 2e-05,
+      "loss": 0.3012,
+      "step": 47119
+    },
+    {
+      "epoch": 1.1215843092449775,
+      "grad_norm": 5.546297746019757,
+      "learning_rate": 2e-05,
+      "loss": 0.5072,
+      "step": 47120
+    },
+    {
+      "epoch": 1.1216081119680092,
+      "grad_norm": 2.3363194798258493,
+      "learning_rate": 2e-05,
+      "loss": 0.0464,
+      "step": 47121
+    },
+    {
+      "epoch": 1.1216319146910407,
+      "grad_norm": 4.027600580974043,
+      "learning_rate": 2e-05,
+      "loss": 0.427,
+      "step": 47122
+    },
+    {
+      "epoch": 1.1216557174140722,
+      "grad_norm": 0.12245086761643352,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47123
+    },
+    {
+      "epoch": 1.1216795201371037,
+      "grad_norm": 4.193291419891839,
+      "learning_rate": 2e-05,
+      "loss": 0.3103,
+      "step": 47124
+    },
+    {
+      "epoch": 1.1217033228601352,
+      "grad_norm": 0.516303884429376,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 47125
+    },
+    {
+      "epoch": 1.1217271255831667,
+      "grad_norm": 2.8368565799514838,
+      "learning_rate": 2e-05,
+      "loss": 0.0195,
+      "step": 47126
+    },
+    {
+      "epoch": 1.1217509283061982,
+      "grad_norm": 2.710387196729083,
+      "learning_rate": 2e-05,
+      "loss": 0.079,
+      "step": 47127
+    },
+    {
+      "epoch": 1.1217747310292296,
+      "grad_norm": 6.314806856336853,
+      "learning_rate": 2e-05,
+      "loss": 0.6145,
+      "step": 47128
+    },
+    {
+      "epoch": 1.1217985337522614,
+      "grad_norm": 6.254740759895251,
+      "learning_rate": 2e-05,
+      "loss": 0.2334,
+      "step": 47129
+    },
+    {
+      "epoch": 1.1218223364752928,
+      "grad_norm": 0.05019144210487431,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47130
+    },
+    {
+      "epoch": 1.1218461391983243,
+      "grad_norm": 0.8245874874736275,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 47131
+    },
+    {
+      "epoch": 1.1218699419213558,
+      "grad_norm": 2.8246681035606125,
+      "learning_rate": 2e-05,
+      "loss": 0.0324,
+      "step": 47132
+    },
+    {
+      "epoch": 1.1218937446443873,
+      "grad_norm": 24.99589567862362,
+      "learning_rate": 2e-05,
+      "loss": 0.552,
+      "step": 47133
+    },
+    {
+      "epoch": 1.1219175473674188,
+      "grad_norm": 0.058789508265075265,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47134
+    },
+    {
+      "epoch": 1.1219413500904503,
+      "grad_norm": 5.224581228383757,
+      "learning_rate": 2e-05,
+      "loss": 0.6356,
+      "step": 47135
+    },
+    {
+      "epoch": 1.1219651528134817,
+      "grad_norm": 3.4674524022297937,
+      "learning_rate": 2e-05,
+      "loss": 0.32,
+      "step": 47136
+    },
+    {
+      "epoch": 1.1219889555365135,
+      "grad_norm": 2.1047576474394374,
+      "learning_rate": 2e-05,
+      "loss": 0.1592,
+      "step": 47137
+    },
+    {
+      "epoch": 1.122012758259545,
+      "grad_norm": 3.1598814354063007,
+      "learning_rate": 2e-05,
+      "loss": 0.305,
+      "step": 47138
+    },
+    {
+      "epoch": 1.1220365609825764,
+      "grad_norm": 3.562570623822941,
+      "learning_rate": 2e-05,
+      "loss": 0.4101,
+      "step": 47139
+    },
+    {
+      "epoch": 1.122060363705608,
+      "grad_norm": 3.6773532056672926,
+      "learning_rate": 2e-05,
+      "loss": 0.056,
+      "step": 47140
+    },
+    {
+      "epoch": 1.1220841664286394,
+      "grad_norm": 1.145387948844812,
+      "learning_rate": 2e-05,
+      "loss": 0.0137,
+      "step": 47141
+    },
+    {
+      "epoch": 1.1221079691516709,
+      "grad_norm": 3.55849485279594,
+      "learning_rate": 2e-05,
+      "loss": 0.4122,
+      "step": 47142
+    },
+    {
+      "epoch": 1.1221317718747024,
+      "grad_norm": 2.0878504573534387,
+      "learning_rate": 2e-05,
+      "loss": 0.2424,
+      "step": 47143
+    },
+    {
+      "epoch": 1.122155574597734,
+      "grad_norm": 2.8746605016121034,
+      "learning_rate": 2e-05,
+      "loss": 0.4106,
+      "step": 47144
+    },
+    {
+      "epoch": 1.1221793773207656,
+      "grad_norm": 3.8762565275655874,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 47145
+    },
+    {
+      "epoch": 1.122203180043797,
+      "grad_norm": 5.7921981636047875,
+      "learning_rate": 2e-05,
+      "loss": 0.7684,
+      "step": 47146
+    },
+    {
+      "epoch": 1.1222269827668285,
+      "grad_norm": 0.01954818351803079,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47147
+    },
+    {
+      "epoch": 1.12225078548986,
+      "grad_norm": 2.3054436784378507,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 47148
+    },
+    {
+      "epoch": 1.1222745882128915,
+      "grad_norm": 4.263210793622711,
+      "learning_rate": 2e-05,
+      "loss": 0.2509,
+      "step": 47149
+    },
+    {
+      "epoch": 1.122298390935923,
+      "grad_norm": 3.7532804689491037,
+      "learning_rate": 2e-05,
+      "loss": 0.6064,
+      "step": 47150
+    },
+    {
+      "epoch": 1.1223221936589547,
+      "grad_norm": 1.9634506959284972,
+      "learning_rate": 2e-05,
+      "loss": 0.0379,
+      "step": 47151
+    },
+    {
+      "epoch": 1.1223459963819862,
+      "grad_norm": 1.4471815629847782,
+      "learning_rate": 2e-05,
+      "loss": 0.0278,
+      "step": 47152
+    },
+    {
+      "epoch": 1.1223697991050177,
+      "grad_norm": 3.0901363181948627,
+      "learning_rate": 2e-05,
+      "loss": 0.2598,
+      "step": 47153
+    },
+    {
+      "epoch": 1.1223936018280491,
+      "grad_norm": 0.05509366093408295,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47154
+    },
+    {
+      "epoch": 1.1224174045510806,
+      "grad_norm": 2.6998008509380536,
+      "learning_rate": 2e-05,
+      "loss": 0.1776,
+      "step": 47155
+    },
+    {
+      "epoch": 1.1224412072741121,
+      "grad_norm": 3.9278002375412684,
+      "learning_rate": 2e-05,
+      "loss": 0.0838,
+      "step": 47156
+    },
+    {
+      "epoch": 1.1224650099971436,
+      "grad_norm": 8.503239745687727,
+      "learning_rate": 2e-05,
+      "loss": 0.1253,
+      "step": 47157
+    },
+    {
+      "epoch": 1.122488812720175,
+      "grad_norm": 0.24962301348032476,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 47158
+    },
+    {
+      "epoch": 1.1225126154432068,
+      "grad_norm": 1.7057385677319647,
+      "learning_rate": 2e-05,
+      "loss": 0.014,
+      "step": 47159
+    },
+    {
+      "epoch": 1.1225364181662383,
+      "grad_norm": 5.108757392468787,
+      "learning_rate": 2e-05,
+      "loss": 0.5763,
+      "step": 47160
+    },
+    {
+      "epoch": 1.1225602208892698,
+      "grad_norm": 3.1799792210462705,
+      "learning_rate": 2e-05,
+      "loss": 0.0811,
+      "step": 47161
+    },
+    {
+      "epoch": 1.1225840236123013,
+      "grad_norm": 2.4966608099386054,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 47162
+    },
+    {
+      "epoch": 1.1226078263353327,
+      "grad_norm": 0.057248734438840314,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47163
+    },
+    {
+      "epoch": 1.1226316290583642,
+      "grad_norm": 0.5584491864764325,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 47164
+    },
+    {
+      "epoch": 1.1226554317813957,
+      "grad_norm": 22.03126119156119,
+      "learning_rate": 2e-05,
+      "loss": 1.4496,
+      "step": 47165
+    },
+    {
+      "epoch": 1.1226792345044272,
+      "grad_norm": 6.53568816234983,
+      "learning_rate": 2e-05,
+      "loss": 0.6935,
+      "step": 47166
+    },
+    {
+      "epoch": 1.122703037227459,
+      "grad_norm": 3.6836196536848576,
+      "learning_rate": 2e-05,
+      "loss": 0.3837,
+      "step": 47167
+    },
+    {
+      "epoch": 1.1227268399504904,
+      "grad_norm": 0.6010178297509347,
+      "learning_rate": 2e-05,
+      "loss": 0.0155,
+      "step": 47168
+    },
+    {
+      "epoch": 1.1227506426735219,
+      "grad_norm": 8.280781474195754,
+      "learning_rate": 2e-05,
+      "loss": 0.6754,
+      "step": 47169
+    },
+    {
+      "epoch": 1.1227744453965534,
+      "grad_norm": 0.5143914649790896,
+      "learning_rate": 2e-05,
+      "loss": 0.0046,
+      "step": 47170
+    },
+    {
+      "epoch": 1.1227982481195848,
+      "grad_norm": 7.568604878156757,
+      "learning_rate": 2e-05,
+      "loss": 0.1177,
+      "step": 47171
+    },
+    {
+      "epoch": 1.1228220508426163,
+      "grad_norm": 2.372775661733708,
+      "learning_rate": 2e-05,
+      "loss": 0.028,
+      "step": 47172
+    },
+    {
+      "epoch": 1.1228458535656478,
+      "grad_norm": 3.5878517776346257,
+      "learning_rate": 2e-05,
+      "loss": 0.186,
+      "step": 47173
+    },
+    {
+      "epoch": 1.1228696562886795,
+      "grad_norm": 6.112525001125128,
+      "learning_rate": 2e-05,
+      "loss": 0.4164,
+      "step": 47174
+    },
+    {
+      "epoch": 1.122893459011711,
+      "grad_norm": 0.138641350412808,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47175
+    },
+    {
+      "epoch": 1.1229172617347425,
+      "grad_norm": 5.54107553142175,
+      "learning_rate": 2e-05,
+      "loss": 0.1762,
+      "step": 47176
+    },
+    {
+      "epoch": 1.122941064457774,
+      "grad_norm": 4.4760237626884045,
+      "learning_rate": 2e-05,
+      "loss": 0.2939,
+      "step": 47177
+    },
+    {
+      "epoch": 1.1229648671808055,
+      "grad_norm": 4.060770372662508,
+      "learning_rate": 2e-05,
+      "loss": 0.2711,
+      "step": 47178
+    },
+    {
+      "epoch": 1.122988669903837,
+      "grad_norm": 4.671773876368905,
+      "learning_rate": 2e-05,
+      "loss": 0.0649,
+      "step": 47179
+    },
+    {
+      "epoch": 1.1230124726268684,
+      "grad_norm": 1.6109348988405114,
+      "learning_rate": 2e-05,
+      "loss": 0.0359,
+      "step": 47180
+    },
+    {
+      "epoch": 1.1230362753499001,
+      "grad_norm": 0.49400378056110095,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 47181
+    },
+    {
+      "epoch": 1.1230600780729316,
+      "grad_norm": 15.090860752586423,
+      "learning_rate": 2e-05,
+      "loss": 0.3294,
+      "step": 47182
+    },
+    {
+      "epoch": 1.123083880795963,
+      "grad_norm": 0.15713379621275056,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47183
+    },
+    {
+      "epoch": 1.1231076835189946,
+      "grad_norm": 4.054924792135517,
+      "learning_rate": 2e-05,
+      "loss": 0.5574,
+      "step": 47184
+    },
+    {
+      "epoch": 1.123131486242026,
+      "grad_norm": 3.1223267292060712,
+      "learning_rate": 2e-05,
+      "loss": 0.4183,
+      "step": 47185
+    },
+    {
+      "epoch": 1.1231552889650576,
+      "grad_norm": 3.411939504565325,
+      "learning_rate": 2e-05,
+      "loss": 0.0774,
+      "step": 47186
+    },
+    {
+      "epoch": 1.123179091688089,
+      "grad_norm": 14.540388794057783,
+      "learning_rate": 2e-05,
+      "loss": 0.3306,
+      "step": 47187
+    },
+    {
+      "epoch": 1.1232028944111205,
+      "grad_norm": 4.212526628853759,
+      "learning_rate": 2e-05,
+      "loss": 0.1946,
+      "step": 47188
+    },
+    {
+      "epoch": 1.1232266971341522,
+      "grad_norm": 4.421523080994907,
+      "learning_rate": 2e-05,
+      "loss": 0.3101,
+      "step": 47189
+    },
+    {
+      "epoch": 1.1232504998571837,
+      "grad_norm": 0.09058450054860283,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47190
+    },
+    {
+      "epoch": 1.1232743025802152,
+      "grad_norm": 9.504079520541982,
+      "learning_rate": 2e-05,
+      "loss": 0.3515,
+      "step": 47191
+    },
+    {
+      "epoch": 1.1232981053032467,
+      "grad_norm": 2.0317639388965367,
+      "learning_rate": 2e-05,
+      "loss": 0.0258,
+      "step": 47192
+    },
+    {
+      "epoch": 1.1233219080262782,
+      "grad_norm": 0.8157791582239542,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 47193
+    },
+    {
+      "epoch": 1.1233457107493097,
+      "grad_norm": 8.025036630447088,
+      "learning_rate": 2e-05,
+      "loss": 0.1574,
+      "step": 47194
+    },
+    {
+      "epoch": 1.1233695134723412,
+      "grad_norm": 13.9735046301599,
+      "learning_rate": 2e-05,
+      "loss": 0.9374,
+      "step": 47195
+    },
+    {
+      "epoch": 1.1233933161953726,
+      "grad_norm": 5.184640519506464,
+      "learning_rate": 2e-05,
+      "loss": 0.745,
+      "step": 47196
+    },
+    {
+      "epoch": 1.1234171189184043,
+      "grad_norm": 0.03986118161330948,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47197
+    },
+    {
+      "epoch": 1.1234409216414358,
+      "grad_norm": 0.53412778735674,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 47198
+    },
+    {
+      "epoch": 1.1234647243644673,
+      "grad_norm": 3.791469057588212,
+      "learning_rate": 2e-05,
+      "loss": 0.4526,
+      "step": 47199
+    },
+    {
+      "epoch": 1.1234885270874988,
+      "grad_norm": 2.753463970560368,
+      "learning_rate": 2e-05,
+      "loss": 0.2073,
+      "step": 47200
+    },
+    {
+      "epoch": 1.1235123298105303,
+      "grad_norm": 1.0376453758642576,
+      "learning_rate": 2e-05,
+      "loss": 0.0129,
+      "step": 47201
+    },
+    {
+      "epoch": 1.1235361325335618,
+      "grad_norm": 0.18406985193343928,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47202
+    },
+    {
+      "epoch": 1.1235599352565933,
+      "grad_norm": 0.031974351615091,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47203
+    },
+    {
+      "epoch": 1.123583737979625,
+      "grad_norm": 0.08887465385029412,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47204
+    },
+    {
+      "epoch": 1.1236075407026564,
+      "grad_norm": 5.394218725917712,
+      "learning_rate": 2e-05,
+      "loss": 0.2377,
+      "step": 47205
+    },
+    {
+      "epoch": 1.123631343425688,
+      "grad_norm": 0.09651932167740611,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47206
+    },
+    {
+      "epoch": 1.1236551461487194,
+      "grad_norm": 3.486974361807668,
+      "learning_rate": 2e-05,
+      "loss": 0.0868,
+      "step": 47207
+    },
+    {
+      "epoch": 1.123678948871751,
+      "grad_norm": 0.09277357688978462,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 47208
+    },
+    {
+      "epoch": 1.1237027515947824,
+      "grad_norm": 0.3042160345948216,
+      "learning_rate": 2e-05,
+      "loss": 0.0064,
+      "step": 47209
+    },
+    {
+      "epoch": 1.1237265543178139,
+      "grad_norm": 0.1568554442056972,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 47210
+    },
+    {
+      "epoch": 1.1237503570408456,
+      "grad_norm": 5.868931565321518,
+      "learning_rate": 2e-05,
+      "loss": 0.6407,
+      "step": 47211
+    },
+    {
+      "epoch": 1.123774159763877,
+      "grad_norm": 5.052449236322525,
+      "learning_rate": 2e-05,
+      "loss": 0.391,
+      "step": 47212
+    },
+    {
+      "epoch": 1.1237979624869086,
+      "grad_norm": 2.955418188384756,
+      "learning_rate": 2e-05,
+      "loss": 0.3672,
+      "step": 47213
+    },
+    {
+      "epoch": 1.12382176520994,
+      "grad_norm": 0.12151773790013344,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 47214
+    },
+    {
+      "epoch": 1.1238455679329715,
+      "grad_norm": 6.692847461558111,
+      "learning_rate": 2e-05,
+      "loss": 0.1174,
+      "step": 47215
+    },
+    {
+      "epoch": 1.123869370656003,
+      "grad_norm": 20.5655572083611,
+      "learning_rate": 2e-05,
+      "loss": 0.5967,
+      "step": 47216
+    },
+    {
+      "epoch": 1.1238931733790345,
+      "grad_norm": 0.058211861989606366,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47217
+    },
+    {
+      "epoch": 1.123916976102066,
+      "grad_norm": 6.166098736190923,
+      "learning_rate": 2e-05,
+      "loss": 0.9018,
+      "step": 47218
+    },
+    {
+      "epoch": 1.1239407788250977,
+      "grad_norm": 1.335224819530413,
+      "learning_rate": 2e-05,
+      "loss": 0.031,
+      "step": 47219
+    },
+    {
+      "epoch": 1.1239645815481292,
+      "grad_norm": 0.4491897530983239,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 47220
+    },
+    {
+      "epoch": 1.1239883842711607,
+      "grad_norm": 8.72177048915439,
+      "learning_rate": 2e-05,
+      "loss": 0.378,
+      "step": 47221
+    },
+    {
+      "epoch": 1.1240121869941921,
+      "grad_norm": 13.433050895997198,
+      "learning_rate": 2e-05,
+      "loss": 0.229,
+      "step": 47222
+    },
+    {
+      "epoch": 1.1240359897172236,
+      "grad_norm": 2.717631642292671,
+      "learning_rate": 2e-05,
+      "loss": 0.1505,
+      "step": 47223
+    },
+    {
+      "epoch": 1.1240597924402551,
+      "grad_norm": 3.1206270942484315,
+      "learning_rate": 2e-05,
+      "loss": 0.3335,
+      "step": 47224
+    },
+    {
+      "epoch": 1.1240835951632866,
+      "grad_norm": 4.691434051907848,
+      "learning_rate": 2e-05,
+      "loss": 0.7483,
+      "step": 47225
+    },
+    {
+      "epoch": 1.124107397886318,
+      "grad_norm": 0.014571059936477506,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47226
+    },
+    {
+      "epoch": 1.1241312006093498,
+      "grad_norm": 3.4765863288384433,
+      "learning_rate": 2e-05,
+      "loss": 0.3462,
+      "step": 47227
+    },
+    {
+      "epoch": 1.1241550033323813,
+      "grad_norm": 0.015230886309308791,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47228
+    },
+    {
+      "epoch": 1.1241788060554128,
+      "grad_norm": 1.648088361933342,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 47229
+    },
+    {
+      "epoch": 1.1242026087784442,
+      "grad_norm": 5.245887005870407,
+      "learning_rate": 2e-05,
+      "loss": 0.8741,
+      "step": 47230
+    },
+    {
+      "epoch": 1.1242264115014757,
+      "grad_norm": 2.86663752008417,
+      "learning_rate": 2e-05,
+      "loss": 0.1816,
+      "step": 47231
+    },
+    {
+      "epoch": 1.1242502142245072,
+      "grad_norm": 3.087316352062294,
+      "learning_rate": 2e-05,
+      "loss": 0.4129,
+      "step": 47232
+    },
+    {
+      "epoch": 1.1242740169475387,
+      "grad_norm": 1.0047563578939442,
+      "learning_rate": 2e-05,
+      "loss": 0.0136,
+      "step": 47233
+    },
+    {
+      "epoch": 1.1242978196705704,
+      "grad_norm": 0.19536939875726717,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 47234
+    },
+    {
+      "epoch": 1.124321622393602,
+      "grad_norm": 3.7553683886763793,
+      "learning_rate": 2e-05,
+      "loss": 0.3998,
+      "step": 47235
+    },
+    {
+      "epoch": 1.1243454251166334,
+      "grad_norm": 5.89213007339072,
+      "learning_rate": 2e-05,
+      "loss": 0.4245,
+      "step": 47236
+    },
+    {
+      "epoch": 1.1243692278396649,
+      "grad_norm": 0.03879998168910634,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47237
+    },
+    {
+      "epoch": 1.1243930305626963,
+      "grad_norm": 0.5313461662797324,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 47238
+    },
+    {
+      "epoch": 1.1244168332857278,
+      "grad_norm": 4.84053924181412,
+      "learning_rate": 2e-05,
+      "loss": 0.6225,
+      "step": 47239
+    },
+    {
+      "epoch": 1.1244406360087593,
+      "grad_norm": 5.218992571389397,
+      "learning_rate": 2e-05,
+      "loss": 0.5207,
+      "step": 47240
+    },
+    {
+      "epoch": 1.124464438731791,
+      "grad_norm": 0.04348801661562801,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47241
+    },
+    {
+      "epoch": 1.1244882414548225,
+      "grad_norm": 6.175178981082319,
+      "learning_rate": 2e-05,
+      "loss": 0.5759,
+      "step": 47242
+    },
+    {
+      "epoch": 1.124512044177854,
+      "grad_norm": 2.168105241142166,
+      "learning_rate": 2e-05,
+      "loss": 0.0357,
+      "step": 47243
+    },
+    {
+      "epoch": 1.1245358469008855,
+      "grad_norm": 0.5019340352869516,
+      "learning_rate": 2e-05,
+      "loss": 0.005,
+      "step": 47244
+    },
+    {
+      "epoch": 1.124559649623917,
+      "grad_norm": 5.0310617527142725,
+      "learning_rate": 2e-05,
+      "loss": 1.0993,
+      "step": 47245
+    },
+    {
+      "epoch": 1.1245834523469485,
+      "grad_norm": 0.03556025776938995,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47246
+    },
+    {
+      "epoch": 1.12460725506998,
+      "grad_norm": 5.284356501250324,
+      "learning_rate": 2e-05,
+      "loss": 0.9342,
+      "step": 47247
+    },
+    {
+      "epoch": 1.1246310577930114,
+      "grad_norm": 0.13268220086026958,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47248
+    },
+    {
+      "epoch": 1.1246548605160431,
+      "grad_norm": 0.6234044927068477,
+      "learning_rate": 2e-05,
+      "loss": 0.0057,
+      "step": 47249
+    },
+    {
+      "epoch": 1.1246786632390746,
+      "grad_norm": 3.3217691015558675,
+      "learning_rate": 2e-05,
+      "loss": 0.0355,
+      "step": 47250
+    },
+    {
+      "epoch": 1.124702465962106,
+      "grad_norm": 4.0512513224950615,
+      "learning_rate": 2e-05,
+      "loss": 0.1408,
+      "step": 47251
+    },
+    {
+      "epoch": 1.1247262686851376,
+      "grad_norm": 2.2224111005319998,
+      "learning_rate": 2e-05,
+      "loss": 0.1799,
+      "step": 47252
+    },
+    {
+      "epoch": 1.124750071408169,
+      "grad_norm": 8.66628939031702,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 47253
+    },
+    {
+      "epoch": 1.1247738741312006,
+      "grad_norm": 0.9662204266706507,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 47254
+    },
+    {
+      "epoch": 1.124797676854232,
+      "grad_norm": 0.2181701415861426,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 47255
+    },
+    {
+      "epoch": 1.1248214795772635,
+      "grad_norm": 16.07742313802719,
+      "learning_rate": 2e-05,
+      "loss": 0.2737,
+      "step": 47256
+    },
+    {
+      "epoch": 1.1248452823002952,
+      "grad_norm": 0.6397854776884055,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47257
+    },
+    {
+      "epoch": 1.1248690850233267,
+      "grad_norm": 0.02668167888352422,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47258
+    },
+    {
+      "epoch": 1.1248928877463582,
+      "grad_norm": 0.013760984157519309,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47259
+    },
+    {
+      "epoch": 1.1249166904693897,
+      "grad_norm": 4.046246544626521,
+      "learning_rate": 2e-05,
+      "loss": 0.1723,
+      "step": 47260
+    },
+    {
+      "epoch": 1.1249404931924212,
+      "grad_norm": 0.047086379154747934,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47261
+    },
+    {
+      "epoch": 1.1249642959154527,
+      "grad_norm": 0.04518672195217406,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47262
+    },
+    {
+      "epoch": 1.1249880986384841,
+      "grad_norm": 0.006454760205204387,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47263
+    },
+    {
+      "epoch": 1.1250119013615159,
+      "grad_norm": 0.11565133900167272,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47264
+    },
+    {
+      "epoch": 1.1250357040845473,
+      "grad_norm": 3.7086809482825474,
+      "learning_rate": 2e-05,
+      "loss": 0.0457,
+      "step": 47265
+    },
+    {
+      "epoch": 1.1250595068075788,
+      "grad_norm": 0.43744307073583605,
+      "learning_rate": 2e-05,
+      "loss": 0.0076,
+      "step": 47266
+    },
+    {
+      "epoch": 1.1250833095306103,
+      "grad_norm": 0.09686248249561873,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47267
+    },
+    {
+      "epoch": 1.1251071122536418,
+      "grad_norm": 0.17043495569269143,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 47268
+    },
+    {
+      "epoch": 1.1251309149766733,
+      "grad_norm": 1.9842329711962847,
+      "learning_rate": 2e-05,
+      "loss": 0.0202,
+      "step": 47269
+    },
+    {
+      "epoch": 1.1251547176997048,
+      "grad_norm": 4.282493324314506,
+      "learning_rate": 2e-05,
+      "loss": 0.39,
+      "step": 47270
+    },
+    {
+      "epoch": 1.1251785204227365,
+      "grad_norm": 0.09578397067735728,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 47271
+    },
+    {
+      "epoch": 1.125202323145768,
+      "grad_norm": 10.715955769608371,
+      "learning_rate": 2e-05,
+      "loss": 1.1554,
+      "step": 47272
+    },
+    {
+      "epoch": 1.1252261258687994,
+      "grad_norm": 0.13196306394776675,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47273
+    },
+    {
+      "epoch": 1.125249928591831,
+      "grad_norm": 0.5967420800380789,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 47274
+    },
+    {
+      "epoch": 1.1252737313148624,
+      "grad_norm": 3.0726265464769025,
+      "learning_rate": 2e-05,
+      "loss": 0.4286,
+      "step": 47275
+    },
+    {
+      "epoch": 1.125297534037894,
+      "grad_norm": 3.5656774421232678,
+      "learning_rate": 2e-05,
+      "loss": 0.0693,
+      "step": 47276
+    },
+    {
+      "epoch": 1.1253213367609254,
+      "grad_norm": 0.17576681029758523,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 47277
+    },
+    {
+      "epoch": 1.1253451394839569,
+      "grad_norm": 4.479913563505314,
+      "learning_rate": 2e-05,
+      "loss": 0.0745,
+      "step": 47278
+    },
+    {
+      "epoch": 1.1253689422069886,
+      "grad_norm": 5.861586681776344,
+      "learning_rate": 2e-05,
+      "loss": 0.5793,
+      "step": 47279
+    },
+    {
+      "epoch": 1.12539274493002,
+      "grad_norm": 0.6372660401495197,
+      "learning_rate": 2e-05,
+      "loss": 0.0045,
+      "step": 47280
+    },
+    {
+      "epoch": 1.1254165476530515,
+      "grad_norm": 4.065089450205105,
+      "learning_rate": 2e-05,
+      "loss": 0.6132,
+      "step": 47281
+    },
+    {
+      "epoch": 1.125440350376083,
+      "grad_norm": 2.253546526652035,
+      "learning_rate": 2e-05,
+      "loss": 0.2162,
+      "step": 47282
+    },
+    {
+      "epoch": 1.1254641530991145,
+      "grad_norm": 0.8709021295807112,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 47283
+    },
+    {
+      "epoch": 1.125487955822146,
+      "grad_norm": 9.790285432479529,
+      "learning_rate": 2e-05,
+      "loss": 0.7287,
+      "step": 47284
+    },
+    {
+      "epoch": 1.1255117585451775,
+      "grad_norm": 2.0439189599649557,
+      "learning_rate": 2e-05,
+      "loss": 0.0109,
+      "step": 47285
+    },
+    {
+      "epoch": 1.125535561268209,
+      "grad_norm": 0.19073927432095103,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 47286
+    },
+    {
+      "epoch": 1.1255593639912407,
+      "grad_norm": 0.7067877558766323,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 47287
+    },
+    {
+      "epoch": 1.1255831667142722,
+      "grad_norm": 3.4580709307392667,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 47288
+    },
+    {
+      "epoch": 1.1256069694373037,
+      "grad_norm": 0.6153431247288197,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 47289
+    },
+    {
+      "epoch": 1.1256307721603351,
+      "grad_norm": 4.285125967409504,
+      "learning_rate": 2e-05,
+      "loss": 0.3536,
+      "step": 47290
+    },
+    {
+      "epoch": 1.1256545748833666,
+      "grad_norm": 3.5163668435964035,
+      "learning_rate": 2e-05,
+      "loss": 0.0589,
+      "step": 47291
+    },
+    {
+      "epoch": 1.125678377606398,
+      "grad_norm": 0.34502541315614726,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 47292
+    },
+    {
+      "epoch": 1.1257021803294296,
+      "grad_norm": 5.833259752896454,
+      "learning_rate": 2e-05,
+      "loss": 0.4823,
+      "step": 47293
+    },
+    {
+      "epoch": 1.1257259830524613,
+      "grad_norm": 0.05334202034475169,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47294
+    },
+    {
+      "epoch": 1.1257497857754928,
+      "grad_norm": 12.21185735730302,
+      "learning_rate": 2e-05,
+      "loss": 1.019,
+      "step": 47295
+    },
+    {
+      "epoch": 1.1257735884985243,
+      "grad_norm": 1.1935188237701313,
+      "learning_rate": 2e-05,
+      "loss": 0.0229,
+      "step": 47296
+    },
+    {
+      "epoch": 1.1257973912215558,
+      "grad_norm": 0.006330408356919021,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47297
+    },
+    {
+      "epoch": 1.1258211939445872,
+      "grad_norm": 2.3715935216291903,
+      "learning_rate": 2e-05,
+      "loss": 0.0727,
+      "step": 47298
+    },
+    {
+      "epoch": 1.1258449966676187,
+      "grad_norm": 0.09061563958272523,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47299
+    },
+    {
+      "epoch": 1.1258687993906502,
+      "grad_norm": 5.56783522979461,
+      "learning_rate": 2e-05,
+      "loss": 0.0369,
+      "step": 47300
+    },
+    {
+      "epoch": 1.125892602113682,
+      "grad_norm": 2.587784504780944,
+      "learning_rate": 2e-05,
+      "loss": 0.1257,
+      "step": 47301
+    },
+    {
+      "epoch": 1.1259164048367134,
+      "grad_norm": 6.626440900146258,
+      "learning_rate": 2e-05,
+      "loss": 0.7166,
+      "step": 47302
+    },
+    {
+      "epoch": 1.1259402075597449,
+      "grad_norm": 0.03206941535517975,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47303
+    },
+    {
+      "epoch": 1.1259640102827764,
+      "grad_norm": 0.8872244772660002,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 47304
+    },
+    {
+      "epoch": 1.1259878130058079,
+      "grad_norm": 0.3681264463026622,
+      "learning_rate": 2e-05,
+      "loss": 0.0043,
+      "step": 47305
+    },
+    {
+      "epoch": 1.1260116157288393,
+      "grad_norm": 7.955509260755811,
+      "learning_rate": 2e-05,
+      "loss": 0.724,
+      "step": 47306
+    },
+    {
+      "epoch": 1.1260354184518708,
+      "grad_norm": 0.045547523732662756,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47307
+    },
+    {
+      "epoch": 1.1260592211749023,
+      "grad_norm": 0.027893155404220396,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47308
+    },
+    {
+      "epoch": 1.1260830238979338,
+      "grad_norm": 6.450482326172791,
+      "learning_rate": 2e-05,
+      "loss": 0.557,
+      "step": 47309
+    },
+    {
+      "epoch": 1.1261068266209655,
+      "grad_norm": 3.2210041239086404,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 47310
+    },
+    {
+      "epoch": 1.126130629343997,
+      "grad_norm": 4.136014723499874,
+      "learning_rate": 2e-05,
+      "loss": 0.4603,
+      "step": 47311
+    },
+    {
+      "epoch": 1.1261544320670285,
+      "grad_norm": 0.022557433874093338,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47312
+    },
+    {
+      "epoch": 1.12617823479006,
+      "grad_norm": 0.48768102629700333,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 47313
+    },
+    {
+      "epoch": 1.1262020375130914,
+      "grad_norm": 2.0752691512481896,
+      "learning_rate": 2e-05,
+      "loss": 0.0207,
+      "step": 47314
+    },
+    {
+      "epoch": 1.126225840236123,
+      "grad_norm": 1.6485360024708364,
+      "learning_rate": 2e-05,
+      "loss": 0.0233,
+      "step": 47315
+    },
+    {
+      "epoch": 1.1262496429591544,
+      "grad_norm": 0.41269040832278,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 47316
+    },
+    {
+      "epoch": 1.1262734456821861,
+      "grad_norm": 7.067750817439049,
+      "learning_rate": 2e-05,
+      "loss": 0.4355,
+      "step": 47317
+    },
+    {
+      "epoch": 1.1262972484052176,
+      "grad_norm": 0.0513846229769265,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47318
+    },
+    {
+      "epoch": 1.126321051128249,
+      "grad_norm": 2.490947959753373,
+      "learning_rate": 2e-05,
+      "loss": 0.0174,
+      "step": 47319
+    },
+    {
+      "epoch": 1.1263448538512806,
+      "grad_norm": 2.2799953955280117,
+      "learning_rate": 2e-05,
+      "loss": 0.1424,
+      "step": 47320
+    },
+    {
+      "epoch": 1.126368656574312,
+      "grad_norm": 0.09517395763545977,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47321
+    },
+    {
+      "epoch": 1.1263924592973436,
+      "grad_norm": 4.651168611209864,
+      "learning_rate": 2e-05,
+      "loss": 0.3757,
+      "step": 47322
+    },
+    {
+      "epoch": 1.126416262020375,
+      "grad_norm": 3.3637220318870367,
+      "learning_rate": 2e-05,
+      "loss": 0.3409,
+      "step": 47323
+    },
+    {
+      "epoch": 1.1264400647434067,
+      "grad_norm": 4.458279318040209,
+      "learning_rate": 2e-05,
+      "loss": 0.5962,
+      "step": 47324
+    },
+    {
+      "epoch": 1.1264638674664382,
+      "grad_norm": 2.4869559474057406,
+      "learning_rate": 2e-05,
+      "loss": 0.0655,
+      "step": 47325
+    },
+    {
+      "epoch": 1.1264876701894697,
+      "grad_norm": 0.016292783588849252,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47326
+    },
+    {
+      "epoch": 1.1265114729125012,
+      "grad_norm": 6.43096936540715,
+      "learning_rate": 2e-05,
+      "loss": 0.7183,
+      "step": 47327
+    },
+    {
+      "epoch": 1.1265352756355327,
+      "grad_norm": 1.5799849122515195,
+      "learning_rate": 2e-05,
+      "loss": 0.0199,
+      "step": 47328
+    },
+    {
+      "epoch": 1.1265590783585642,
+      "grad_norm": 8.156048521357704,
+      "learning_rate": 2e-05,
+      "loss": 0.0773,
+      "step": 47329
+    },
+    {
+      "epoch": 1.1265828810815957,
+      "grad_norm": 3.0700355269330077,
+      "learning_rate": 2e-05,
+      "loss": 0.1948,
+      "step": 47330
+    },
+    {
+      "epoch": 1.1266066838046274,
+      "grad_norm": 0.028544989570833512,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47331
+    },
+    {
+      "epoch": 1.1266304865276588,
+      "grad_norm": 5.915927299162236,
+      "learning_rate": 2e-05,
+      "loss": 0.6902,
+      "step": 47332
+    },
+    {
+      "epoch": 1.1266542892506903,
+      "grad_norm": 0.16512588792676353,
+      "learning_rate": 2e-05,
+      "loss": 0.0021,
+      "step": 47333
+    },
+    {
+      "epoch": 1.1266780919737218,
+      "grad_norm": 3.905206578791707,
+      "learning_rate": 2e-05,
+      "loss": 0.4238,
+      "step": 47334
+    },
+    {
+      "epoch": 1.1267018946967533,
+      "grad_norm": 3.4083742204003955,
+      "learning_rate": 2e-05,
+      "loss": 0.4987,
+      "step": 47335
+    },
+    {
+      "epoch": 1.1267256974197848,
+      "grad_norm": 7.020615177498571,
+      "learning_rate": 2e-05,
+      "loss": 0.4145,
+      "step": 47336
+    },
+    {
+      "epoch": 1.1267495001428163,
+      "grad_norm": 0.2789451776287134,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 47337
+    },
+    {
+      "epoch": 1.1267733028658478,
+      "grad_norm": 0.2289999785482042,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47338
+    },
+    {
+      "epoch": 1.1267971055888792,
+      "grad_norm": 1.475352017044391,
+      "learning_rate": 2e-05,
+      "loss": 0.0256,
+      "step": 47339
+    },
+    {
+      "epoch": 1.126820908311911,
+      "grad_norm": 3.9959042876130977,
+      "learning_rate": 2e-05,
+      "loss": 0.2917,
+      "step": 47340
+    },
+    {
+      "epoch": 1.1268447110349424,
+      "grad_norm": 2.2187613867345095,
+      "learning_rate": 2e-05,
+      "loss": 0.2065,
+      "step": 47341
+    },
+    {
+      "epoch": 1.126868513757974,
+      "grad_norm": 0.25905855023619473,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47342
+    },
+    {
+      "epoch": 1.1268923164810054,
+      "grad_norm": 9.14466510679947,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 47343
+    },
+    {
+      "epoch": 1.126916119204037,
+      "grad_norm": 0.10495212652555688,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47344
+    },
+    {
+      "epoch": 1.1269399219270684,
+      "grad_norm": 9.637393159139535,
+      "learning_rate": 2e-05,
+      "loss": 0.7888,
+      "step": 47345
+    },
+    {
+      "epoch": 1.1269637246500999,
+      "grad_norm": 0.005903550365120532,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47346
+    },
+    {
+      "epoch": 1.1269875273731316,
+      "grad_norm": 2.9893217660788536,
+      "learning_rate": 2e-05,
+      "loss": 0.1355,
+      "step": 47347
+    },
+    {
+      "epoch": 1.127011330096163,
+      "grad_norm": 2.2949030695262937,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 47348
+    },
+    {
+      "epoch": 1.1270351328191945,
+      "grad_norm": 0.03388850089481817,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47349
+    },
+    {
+      "epoch": 1.127058935542226,
+      "grad_norm": 3.5127516529228844,
+      "learning_rate": 2e-05,
+      "loss": 0.3518,
+      "step": 47350
+    },
+    {
+      "epoch": 1.1270827382652575,
+      "grad_norm": 3.388153699525384,
+      "learning_rate": 2e-05,
+      "loss": 0.2325,
+      "step": 47351
+    },
+    {
+      "epoch": 1.127106540988289,
+      "grad_norm": 0.15990120113119455,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 47352
+    },
+    {
+      "epoch": 1.1271303437113205,
+      "grad_norm": 0.2692452333441713,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47353
+    },
+    {
+      "epoch": 1.1271541464343522,
+      "grad_norm": 13.151245546619874,
+      "learning_rate": 2e-05,
+      "loss": 0.3137,
+      "step": 47354
+    },
+    {
+      "epoch": 1.1271779491573837,
+      "grad_norm": 0.04336661574808725,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47355
+    },
+    {
+      "epoch": 1.1272017518804152,
+      "grad_norm": 4.486491226978167,
+      "learning_rate": 2e-05,
+      "loss": 0.716,
+      "step": 47356
+    },
+    {
+      "epoch": 1.1272255546034466,
+      "grad_norm": 4.083388058104225,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 47357
+    },
+    {
+      "epoch": 1.1272493573264781,
+      "grad_norm": 1.1434572798625262,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 47358
+    },
+    {
+      "epoch": 1.1272731600495096,
+      "grad_norm": 0.33246749820376437,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 47359
+    },
+    {
+      "epoch": 1.127296962772541,
+      "grad_norm": 3.81021616968907,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 47360
+    },
+    {
+      "epoch": 1.1273207654955728,
+      "grad_norm": 0.047501570811588005,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47361
+    },
+    {
+      "epoch": 1.1273445682186043,
+      "grad_norm": 5.680085402045212,
+      "learning_rate": 2e-05,
+      "loss": 0.1664,
+      "step": 47362
+    },
+    {
+      "epoch": 1.1273683709416358,
+      "grad_norm": 0.7310819742669263,
+      "learning_rate": 2e-05,
+      "loss": 0.0068,
+      "step": 47363
+    },
+    {
+      "epoch": 1.1273921736646673,
+      "grad_norm": 1.1432398571697526,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 47364
+    },
+    {
+      "epoch": 1.1274159763876987,
+      "grad_norm": 0.0668918766411665,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47365
+    },
+    {
+      "epoch": 1.1274397791107302,
+      "grad_norm": 0.010616177639436479,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47366
+    },
+    {
+      "epoch": 1.1274635818337617,
+      "grad_norm": 0.9540338642888667,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47367
+    },
+    {
+      "epoch": 1.1274873845567932,
+      "grad_norm": 4.9668253040884975,
+      "learning_rate": 2e-05,
+      "loss": 0.3015,
+      "step": 47368
+    },
+    {
+      "epoch": 1.1275111872798247,
+      "grad_norm": 5.4650077139226,
+      "learning_rate": 2e-05,
+      "loss": 0.406,
+      "step": 47369
+    },
+    {
+      "epoch": 1.1275349900028564,
+      "grad_norm": 1.1039930826587572,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 47370
+    },
+    {
+      "epoch": 1.1275587927258879,
+      "grad_norm": 4.297200220524038,
+      "learning_rate": 2e-05,
+      "loss": 0.1062,
+      "step": 47371
+    },
+    {
+      "epoch": 1.1275825954489194,
+      "grad_norm": 5.2188476805679835,
+      "learning_rate": 2e-05,
+      "loss": 0.587,
+      "step": 47372
+    },
+    {
+      "epoch": 1.1276063981719509,
+      "grad_norm": 5.115653805914536,
+      "learning_rate": 2e-05,
+      "loss": 0.4602,
+      "step": 47373
+    },
+    {
+      "epoch": 1.1276302008949823,
+      "grad_norm": 0.9940477718909387,
+      "learning_rate": 2e-05,
+      "loss": 0.0083,
+      "step": 47374
+    },
+    {
+      "epoch": 1.1276540036180138,
+      "grad_norm": 3.291721023653368,
+      "learning_rate": 2e-05,
+      "loss": 0.2069,
+      "step": 47375
+    },
+    {
+      "epoch": 1.1276778063410453,
+      "grad_norm": 2.135330880151772,
+      "learning_rate": 2e-05,
+      "loss": 0.1449,
+      "step": 47376
+    },
+    {
+      "epoch": 1.127701609064077,
+      "grad_norm": 1.000446951183163,
+      "learning_rate": 2e-05,
+      "loss": 0.0128,
+      "step": 47377
+    },
+    {
+      "epoch": 1.1277254117871085,
+      "grad_norm": 0.835444233610134,
+      "learning_rate": 2e-05,
+      "loss": 0.007,
+      "step": 47378
+    },
+    {
+      "epoch": 1.12774921451014,
+      "grad_norm": 2.4025804280236436,
+      "learning_rate": 2e-05,
+      "loss": 0.18,
+      "step": 47379
+    },
+    {
+      "epoch": 1.1277730172331715,
+      "grad_norm": 5.9735171045987805,
+      "learning_rate": 2e-05,
+      "loss": 0.8175,
+      "step": 47380
+    },
+    {
+      "epoch": 1.127796819956203,
+      "grad_norm": 0.014420481069824607,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47381
+    },
+    {
+      "epoch": 1.1278206226792344,
+      "grad_norm": 1.8314128440859925,
+      "learning_rate": 2e-05,
+      "loss": 0.1751,
+      "step": 47382
+    },
+    {
+      "epoch": 1.127844425402266,
+      "grad_norm": 9.547657393322776,
+      "learning_rate": 2e-05,
+      "loss": 0.9976,
+      "step": 47383
+    },
+    {
+      "epoch": 1.1278682281252976,
+      "grad_norm": 4.156511416310206,
+      "learning_rate": 2e-05,
+      "loss": 0.0451,
+      "step": 47384
+    },
+    {
+      "epoch": 1.1278920308483291,
+      "grad_norm": 4.203620038756683,
+      "learning_rate": 2e-05,
+      "loss": 0.6015,
+      "step": 47385
+    },
+    {
+      "epoch": 1.1279158335713606,
+      "grad_norm": 4.217603059940237,
+      "learning_rate": 2e-05,
+      "loss": 0.2032,
+      "step": 47386
+    },
+    {
+      "epoch": 1.127939636294392,
+      "grad_norm": 1.0432889222734025,
+      "learning_rate": 2e-05,
+      "loss": 0.0138,
+      "step": 47387
+    },
+    {
+      "epoch": 1.1279634390174236,
+      "grad_norm": 22.42268430202323,
+      "learning_rate": 2e-05,
+      "loss": 0.1828,
+      "step": 47388
+    },
+    {
+      "epoch": 1.127987241740455,
+      "grad_norm": 4.15538314849845,
+      "learning_rate": 2e-05,
+      "loss": 0.4511,
+      "step": 47389
+    },
+    {
+      "epoch": 1.1280110444634865,
+      "grad_norm": 1.3825152222521673,
+      "learning_rate": 2e-05,
+      "loss": 0.0272,
+      "step": 47390
+    },
+    {
+      "epoch": 1.1280348471865183,
+      "grad_norm": 0.190423835844216,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47391
+    },
+    {
+      "epoch": 1.1280586499095497,
+      "grad_norm": 5.658098032877712,
+      "learning_rate": 2e-05,
+      "loss": 0.6263,
+      "step": 47392
+    },
+    {
+      "epoch": 1.1280824526325812,
+      "grad_norm": 4.338930054080598,
+      "learning_rate": 2e-05,
+      "loss": 0.4282,
+      "step": 47393
+    },
+    {
+      "epoch": 1.1281062553556127,
+      "grad_norm": 1.6659454679478292,
+      "learning_rate": 2e-05,
+      "loss": 0.0409,
+      "step": 47394
+    },
+    {
+      "epoch": 1.1281300580786442,
+      "grad_norm": 0.396298761030266,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 47395
+    },
+    {
+      "epoch": 1.1281538608016757,
+      "grad_norm": 0.20998256440391228,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 47396
+    },
+    {
+      "epoch": 1.1281776635247072,
+      "grad_norm": 3.8042438701843144,
+      "learning_rate": 2e-05,
+      "loss": 0.4486,
+      "step": 47397
+    },
+    {
+      "epoch": 1.1282014662477386,
+      "grad_norm": 3.7281678448241404,
+      "learning_rate": 2e-05,
+      "loss": 0.0931,
+      "step": 47398
+    },
+    {
+      "epoch": 1.1282252689707701,
+      "grad_norm": 1.72476393682477,
+      "learning_rate": 2e-05,
+      "loss": 0.0147,
+      "step": 47399
+    },
+    {
+      "epoch": 1.1282490716938018,
+      "grad_norm": 27.077214840035282,
+      "learning_rate": 2e-05,
+      "loss": 1.0846,
+      "step": 47400
+    },
+    {
+      "epoch": 1.1282728744168333,
+      "grad_norm": 5.565633104196515,
+      "learning_rate": 2e-05,
+      "loss": 0.3727,
+      "step": 47401
+    },
+    {
+      "epoch": 1.1282966771398648,
+      "grad_norm": 2.8415055693320235,
+      "learning_rate": 2e-05,
+      "loss": 0.0183,
+      "step": 47402
+    },
+    {
+      "epoch": 1.1283204798628963,
+      "grad_norm": 4.36306736359697,
+      "learning_rate": 2e-05,
+      "loss": 0.0834,
+      "step": 47403
+    },
+    {
+      "epoch": 1.1283442825859278,
+      "grad_norm": 3.124828507450684,
+      "learning_rate": 2e-05,
+      "loss": 0.1004,
+      "step": 47404
+    },
+    {
+      "epoch": 1.1283680853089593,
+      "grad_norm": 7.397787382076116,
+      "learning_rate": 2e-05,
+      "loss": 0.147,
+      "step": 47405
+    },
+    {
+      "epoch": 1.1283918880319908,
+      "grad_norm": 1.1924609527373553,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 47406
+    },
+    {
+      "epoch": 1.1284156907550225,
+      "grad_norm": 7.642187190435954,
+      "learning_rate": 2e-05,
+      "loss": 0.6364,
+      "step": 47407
+    },
+    {
+      "epoch": 1.128439493478054,
+      "grad_norm": 4.655794797718942,
+      "learning_rate": 2e-05,
+      "loss": 0.7366,
+      "step": 47408
+    },
+    {
+      "epoch": 1.1284632962010854,
+      "grad_norm": 1.867784219892978,
+      "learning_rate": 2e-05,
+      "loss": 0.0318,
+      "step": 47409
+    },
+    {
+      "epoch": 1.128487098924117,
+      "grad_norm": 10.49377303199798,
+      "learning_rate": 2e-05,
+      "loss": 0.1755,
+      "step": 47410
+    },
+    {
+      "epoch": 1.1285109016471484,
+      "grad_norm": 1.6734083103451556,
+      "learning_rate": 2e-05,
+      "loss": 0.0501,
+      "step": 47411
+    },
+    {
+      "epoch": 1.1285347043701799,
+      "grad_norm": 0.009281906498619936,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47412
+    },
+    {
+      "epoch": 1.1285585070932114,
+      "grad_norm": 8.72399027936557,
+      "learning_rate": 2e-05,
+      "loss": 0.9294,
+      "step": 47413
+    },
+    {
+      "epoch": 1.128582309816243,
+      "grad_norm": 0.03537951967120368,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47414
+    },
+    {
+      "epoch": 1.1286061125392746,
+      "grad_norm": 1.93829547910225,
+      "learning_rate": 2e-05,
+      "loss": 0.0245,
+      "step": 47415
+    },
+    {
+      "epoch": 1.128629915262306,
+      "grad_norm": 16.152226777387014,
+      "learning_rate": 2e-05,
+      "loss": 0.483,
+      "step": 47416
+    },
+    {
+      "epoch": 1.1286537179853375,
+      "grad_norm": 5.410126780834547,
+      "learning_rate": 2e-05,
+      "loss": 0.4743,
+      "step": 47417
+    },
+    {
+      "epoch": 1.128677520708369,
+      "grad_norm": 6.116789532257386,
+      "learning_rate": 2e-05,
+      "loss": 1.027,
+      "step": 47418
+    },
+    {
+      "epoch": 1.1287013234314005,
+      "grad_norm": 4.692136716915018,
+      "learning_rate": 2e-05,
+      "loss": 0.6464,
+      "step": 47419
+    },
+    {
+      "epoch": 1.128725126154432,
+      "grad_norm": 3.5869554306968503,
+      "learning_rate": 2e-05,
+      "loss": 0.5289,
+      "step": 47420
+    },
+    {
+      "epoch": 1.1287489288774637,
+      "grad_norm": 0.032279986187020666,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47421
+    },
+    {
+      "epoch": 1.1287727316004952,
+      "grad_norm": 3.1307093044583576,
+      "learning_rate": 2e-05,
+      "loss": 0.2311,
+      "step": 47422
+    },
+    {
+      "epoch": 1.1287965343235267,
+      "grad_norm": 0.17442414407038997,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47423
+    },
+    {
+      "epoch": 1.1288203370465582,
+      "grad_norm": 0.7337454755086801,
+      "learning_rate": 2e-05,
+      "loss": 0.0107,
+      "step": 47424
+    },
+    {
+      "epoch": 1.1288441397695896,
+      "grad_norm": 4.571623150310622,
+      "learning_rate": 2e-05,
+      "loss": 0.3634,
+      "step": 47425
+    },
+    {
+      "epoch": 1.1288679424926211,
+      "grad_norm": 1.1504279735411775,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 47426
+    },
+    {
+      "epoch": 1.1288917452156526,
+      "grad_norm": 1.435267903057464,
+      "learning_rate": 2e-05,
+      "loss": 0.0231,
+      "step": 47427
+    },
+    {
+      "epoch": 1.128915547938684,
+      "grad_norm": 4.0937447511908625,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 47428
+    },
+    {
+      "epoch": 1.1289393506617156,
+      "grad_norm": 3.6887680777021115,
+      "learning_rate": 2e-05,
+      "loss": 0.6266,
+      "step": 47429
+    },
+    {
+      "epoch": 1.1289631533847473,
+      "grad_norm": 0.03031515102982231,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47430
+    },
+    {
+      "epoch": 1.1289869561077788,
+      "grad_norm": 2.7233734648948635,
+      "learning_rate": 2e-05,
+      "loss": 0.2261,
+      "step": 47431
+    },
+    {
+      "epoch": 1.1290107588308103,
+      "grad_norm": 0.09606368701241301,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 47432
+    },
+    {
+      "epoch": 1.1290345615538417,
+      "grad_norm": 0.04659504615908179,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47433
+    },
+    {
+      "epoch": 1.1290583642768732,
+      "grad_norm": 0.9063591086571356,
+      "learning_rate": 2e-05,
+      "loss": 0.0074,
+      "step": 47434
+    },
+    {
+      "epoch": 1.1290821669999047,
+      "grad_norm": 6.706174850793004,
+      "learning_rate": 2e-05,
+      "loss": 0.7045,
+      "step": 47435
+    },
+    {
+      "epoch": 1.1291059697229362,
+      "grad_norm": 5.442878470851988,
+      "learning_rate": 2e-05,
+      "loss": 0.0772,
+      "step": 47436
+    },
+    {
+      "epoch": 1.129129772445968,
+      "grad_norm": 3.925655909944585,
+      "learning_rate": 2e-05,
+      "loss": 0.2852,
+      "step": 47437
+    },
+    {
+      "epoch": 1.1291535751689994,
+      "grad_norm": 0.09644899065521267,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 47438
+    },
+    {
+      "epoch": 1.1291773778920309,
+      "grad_norm": 2.293997049463706,
+      "learning_rate": 2e-05,
+      "loss": 0.0979,
+      "step": 47439
+    },
+    {
+      "epoch": 1.1292011806150624,
+      "grad_norm": 15.507042286999841,
+      "learning_rate": 2e-05,
+      "loss": 0.9692,
+      "step": 47440
+    },
+    {
+      "epoch": 1.1292249833380938,
+      "grad_norm": 0.24603609413815813,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 47441
+    },
+    {
+      "epoch": 1.1292487860611253,
+      "grad_norm": 4.463791580085996,
+      "learning_rate": 2e-05,
+      "loss": 0.7734,
+      "step": 47442
+    },
+    {
+      "epoch": 1.1292725887841568,
+      "grad_norm": 0.013812512373479281,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47443
+    },
+    {
+      "epoch": 1.1292963915071885,
+      "grad_norm": 4.906264030534738,
+      "learning_rate": 2e-05,
+      "loss": 0.4,
+      "step": 47444
+    },
+    {
+      "epoch": 1.12932019423022,
+      "grad_norm": 0.11375173529500905,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 47445
+    },
+    {
+      "epoch": 1.1293439969532515,
+      "grad_norm": 5.329970989781331,
+      "learning_rate": 2e-05,
+      "loss": 0.6381,
+      "step": 47446
+    },
+    {
+      "epoch": 1.129367799676283,
+      "grad_norm": 14.745926390842577,
+      "learning_rate": 2e-05,
+      "loss": 0.756,
+      "step": 47447
+    },
+    {
+      "epoch": 1.1293916023993145,
+      "grad_norm": 7.371262853090681,
+      "learning_rate": 2e-05,
+      "loss": 0.7094,
+      "step": 47448
+    },
+    {
+      "epoch": 1.129415405122346,
+      "grad_norm": 3.0690632718663178,
+      "learning_rate": 2e-05,
+      "loss": 0.0449,
+      "step": 47449
+    },
+    {
+      "epoch": 1.1294392078453774,
+      "grad_norm": 3.5535699518120314,
+      "learning_rate": 2e-05,
+      "loss": 0.4143,
+      "step": 47450
+    },
+    {
+      "epoch": 1.1294630105684091,
+      "grad_norm": 4.24641357371745,
+      "learning_rate": 2e-05,
+      "loss": 0.5958,
+      "step": 47451
+    },
+    {
+      "epoch": 1.1294868132914406,
+      "grad_norm": 0.8810190705053603,
+      "learning_rate": 2e-05,
+      "loss": 0.0146,
+      "step": 47452
+    },
+    {
+      "epoch": 1.1295106160144721,
+      "grad_norm": 1.0843289192980965,
+      "learning_rate": 2e-05,
+      "loss": 0.0121,
+      "step": 47453
+    },
+    {
+      "epoch": 1.1295344187375036,
+      "grad_norm": 0.19640820426581754,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 47454
+    },
+    {
+      "epoch": 1.129558221460535,
+      "grad_norm": 3.167986754354058,
+      "learning_rate": 2e-05,
+      "loss": 0.4522,
+      "step": 47455
+    },
+    {
+      "epoch": 1.1295820241835666,
+      "grad_norm": 9.363278590540439,
+      "learning_rate": 2e-05,
+      "loss": 0.8814,
+      "step": 47456
+    },
+    {
+      "epoch": 1.129605826906598,
+      "grad_norm": 9.13492636489521,
+      "learning_rate": 2e-05,
+      "loss": 0.2428,
+      "step": 47457
+    },
+    {
+      "epoch": 1.1296296296296295,
+      "grad_norm": 6.530844531033467,
+      "learning_rate": 2e-05,
+      "loss": 0.1485,
+      "step": 47458
+    },
+    {
+      "epoch": 1.129653432352661,
+      "grad_norm": 5.231635364241597,
+      "learning_rate": 2e-05,
+      "loss": 1.0659,
+      "step": 47459
+    },
+    {
+      "epoch": 1.1296772350756927,
+      "grad_norm": 11.119281529435144,
+      "learning_rate": 2e-05,
+      "loss": 0.8324,
+      "step": 47460
+    },
+    {
+      "epoch": 1.1297010377987242,
+      "grad_norm": 0.0359010516046348,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47461
+    },
+    {
+      "epoch": 1.1297248405217557,
+      "grad_norm": 0.013421967024988468,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47462
+    },
+    {
+      "epoch": 1.1297486432447872,
+      "grad_norm": 5.035244480460254,
+      "learning_rate": 2e-05,
+      "loss": 0.4441,
+      "step": 47463
+    },
+    {
+      "epoch": 1.1297724459678187,
+      "grad_norm": 2.850987220958367,
+      "learning_rate": 2e-05,
+      "loss": 0.0657,
+      "step": 47464
+    },
+    {
+      "epoch": 1.1297962486908502,
+      "grad_norm": 0.810949509839203,
+      "learning_rate": 2e-05,
+      "loss": 0.0093,
+      "step": 47465
+    },
+    {
+      "epoch": 1.1298200514138816,
+      "grad_norm": 3.358722205370283,
+      "learning_rate": 2e-05,
+      "loss": 0.0461,
+      "step": 47466
+    },
+    {
+      "epoch": 1.1298438541369134,
+      "grad_norm": 0.050339167688399744,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47467
+    },
+    {
+      "epoch": 1.1298676568599448,
+      "grad_norm": 3.9720451517890423,
+      "learning_rate": 2e-05,
+      "loss": 0.5271,
+      "step": 47468
+    },
+    {
+      "epoch": 1.1298914595829763,
+      "grad_norm": 3.573078632178549,
+      "learning_rate": 2e-05,
+      "loss": 0.3547,
+      "step": 47469
+    },
+    {
+      "epoch": 1.1299152623060078,
+      "grad_norm": 2.2996427184969255,
+      "learning_rate": 2e-05,
+      "loss": 0.0204,
+      "step": 47470
+    },
+    {
+      "epoch": 1.1299390650290393,
+      "grad_norm": 1.1415071092355453,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 47471
+    },
+    {
+      "epoch": 1.1299628677520708,
+      "grad_norm": 2.32542289003526,
+      "learning_rate": 2e-05,
+      "loss": 0.2041,
+      "step": 47472
+    },
+    {
+      "epoch": 1.1299866704751023,
+      "grad_norm": 0.10739670961757747,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 47473
+    },
+    {
+      "epoch": 1.130010473198134,
+      "grad_norm": 3.5943423796581024,
+      "learning_rate": 2e-05,
+      "loss": 0.5737,
+      "step": 47474
+    },
+    {
+      "epoch": 1.1300342759211655,
+      "grad_norm": 0.9730276065195638,
+      "learning_rate": 2e-05,
+      "loss": 0.0103,
+      "step": 47475
+    },
+    {
+      "epoch": 1.130058078644197,
+      "grad_norm": 6.820269650529443,
+      "learning_rate": 2e-05,
+      "loss": 0.4669,
+      "step": 47476
+    },
+    {
+      "epoch": 1.1300818813672284,
+      "grad_norm": 13.913670890250799,
+      "learning_rate": 2e-05,
+      "loss": 1.2762,
+      "step": 47477
+    },
+    {
+      "epoch": 1.13010568409026,
+      "grad_norm": 0.036392034217152876,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47478
+    },
+    {
+      "epoch": 1.1301294868132914,
+      "grad_norm": 3.4685291093637125,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 47479
+    },
+    {
+      "epoch": 1.1301532895363229,
+      "grad_norm": 9.200876121724136,
+      "learning_rate": 2e-05,
+      "loss": 0.8682,
+      "step": 47480
+    },
+    {
+      "epoch": 1.1301770922593546,
+      "grad_norm": 2.5054537265155665,
+      "learning_rate": 2e-05,
+      "loss": 0.3952,
+      "step": 47481
+    },
+    {
+      "epoch": 1.130200894982386,
+      "grad_norm": 0.2421715409751545,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47482
+    },
+    {
+      "epoch": 1.1302246977054176,
+      "grad_norm": 1.2030239412730341,
+      "learning_rate": 2e-05,
+      "loss": 0.0111,
+      "step": 47483
+    },
+    {
+      "epoch": 1.130248500428449,
+      "grad_norm": 10.61804511144649,
+      "learning_rate": 2e-05,
+      "loss": 0.4595,
+      "step": 47484
+    },
+    {
+      "epoch": 1.1302723031514805,
+      "grad_norm": 0.44542616366674215,
+      "learning_rate": 2e-05,
+      "loss": 0.0051,
+      "step": 47485
+    },
+    {
+      "epoch": 1.130296105874512,
+      "grad_norm": 0.11360976561096815,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 47486
+    },
+    {
+      "epoch": 1.1303199085975435,
+      "grad_norm": 2.6253115841661177,
+      "learning_rate": 2e-05,
+      "loss": 0.2477,
+      "step": 47487
+    },
+    {
+      "epoch": 1.130343711320575,
+      "grad_norm": 4.249883100889533,
+      "learning_rate": 2e-05,
+      "loss": 0.3717,
+      "step": 47488
+    },
+    {
+      "epoch": 1.1303675140436065,
+      "grad_norm": 0.2825791689904762,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47489
+    },
+    {
+      "epoch": 1.1303913167666382,
+      "grad_norm": 5.014980753419084,
+      "learning_rate": 2e-05,
+      "loss": 0.7893,
+      "step": 47490
+    },
+    {
+      "epoch": 1.1304151194896697,
+      "grad_norm": 0.06542064574289326,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47491
+    },
+    {
+      "epoch": 1.1304389222127011,
+      "grad_norm": 5.05424094732248,
+      "learning_rate": 2e-05,
+      "loss": 0.8453,
+      "step": 47492
+    },
+    {
+      "epoch": 1.1304627249357326,
+      "grad_norm": 14.295386900420043,
+      "learning_rate": 2e-05,
+      "loss": 0.3617,
+      "step": 47493
+    },
+    {
+      "epoch": 1.1304865276587641,
+      "grad_norm": 0.9093963016639216,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 47494
+    },
+    {
+      "epoch": 1.1305103303817956,
+      "grad_norm": 3.9170175751570238,
+      "learning_rate": 2e-05,
+      "loss": 0.2475,
+      "step": 47495
+    },
+    {
+      "epoch": 1.130534133104827,
+      "grad_norm": 0.1564121397910029,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47496
+    },
+    {
+      "epoch": 1.1305579358278588,
+      "grad_norm": 0.01492799115481713,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47497
+    },
+    {
+      "epoch": 1.1305817385508903,
+      "grad_norm": 0.021391543208386495,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47498
+    },
+    {
+      "epoch": 1.1306055412739218,
+      "grad_norm": 5.139102403314801,
+      "learning_rate": 2e-05,
+      "loss": 0.1766,
+      "step": 47499
+    },
+    {
+      "epoch": 1.1306293439969533,
+      "grad_norm": 0.2019345478075436,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 47500
+    },
+    {
+      "epoch": 1.1306531467199847,
+      "grad_norm": 6.1200667640724795,
+      "learning_rate": 2e-05,
+      "loss": 0.556,
+      "step": 47501
+    },
+    {
+      "epoch": 1.1306769494430162,
+      "grad_norm": 5.845839093202793,
+      "learning_rate": 2e-05,
+      "loss": 1.0966,
+      "step": 47502
+    },
+    {
+      "epoch": 1.1307007521660477,
+      "grad_norm": 0.011101560586110473,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47503
+    },
+    {
+      "epoch": 1.1307245548890794,
+      "grad_norm": 4.464320974654531,
+      "learning_rate": 2e-05,
+      "loss": 0.7641,
+      "step": 47504
+    },
+    {
+      "epoch": 1.130748357612111,
+      "grad_norm": 5.279659329803204,
+      "learning_rate": 2e-05,
+      "loss": 0.168,
+      "step": 47505
+    },
+    {
+      "epoch": 1.1307721603351424,
+      "grad_norm": 2.413711106828016,
+      "learning_rate": 2e-05,
+      "loss": 0.0259,
+      "step": 47506
+    },
+    {
+      "epoch": 1.1307959630581739,
+      "grad_norm": 0.07617877677988802,
+      "learning_rate": 2e-05,
+      "loss": 0.0016,
+      "step": 47507
+    },
+    {
+      "epoch": 1.1308197657812054,
+      "grad_norm": 0.13970730928958672,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47508
+    },
+    {
+      "epoch": 1.1308435685042368,
+      "grad_norm": 6.8415041165958295,
+      "learning_rate": 2e-05,
+      "loss": 0.0347,
+      "step": 47509
+    },
+    {
+      "epoch": 1.1308673712272683,
+      "grad_norm": 39.1978232170198,
+      "learning_rate": 2e-05,
+      "loss": 0.2614,
+      "step": 47510
+    },
+    {
+      "epoch": 1.1308911739503,
+      "grad_norm": 1.314786645831268,
+      "learning_rate": 2e-05,
+      "loss": 0.0095,
+      "step": 47511
+    },
+    {
+      "epoch": 1.1309149766733315,
+      "grad_norm": 3.1739969215703567,
+      "learning_rate": 2e-05,
+      "loss": 0.0641,
+      "step": 47512
+    },
+    {
+      "epoch": 1.130938779396363,
+      "grad_norm": 0.01123445635198413,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47513
+    },
+    {
+      "epoch": 1.1309625821193945,
+      "grad_norm": 7.181978527958951,
+      "learning_rate": 2e-05,
+      "loss": 0.4094,
+      "step": 47514
+    },
+    {
+      "epoch": 1.130986384842426,
+      "grad_norm": 3.254484611365742,
+      "learning_rate": 2e-05,
+      "loss": 0.2103,
+      "step": 47515
+    },
+    {
+      "epoch": 1.1310101875654575,
+      "grad_norm": 0.09229545822422724,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47516
+    },
+    {
+      "epoch": 1.131033990288489,
+      "grad_norm": 7.470183669557602,
+      "learning_rate": 2e-05,
+      "loss": 0.7489,
+      "step": 47517
+    },
+    {
+      "epoch": 1.1310577930115204,
+      "grad_norm": 1.8492004752310933,
+      "learning_rate": 2e-05,
+      "loss": 0.0226,
+      "step": 47518
+    },
+    {
+      "epoch": 1.131081595734552,
+      "grad_norm": 7.901447635662442,
+      "learning_rate": 2e-05,
+      "loss": 0.4852,
+      "step": 47519
+    },
+    {
+      "epoch": 1.1311053984575836,
+      "grad_norm": 0.2690020385766001,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47520
+    },
+    {
+      "epoch": 1.131129201180615,
+      "grad_norm": 6.165969203698888,
+      "learning_rate": 2e-05,
+      "loss": 0.553,
+      "step": 47521
+    },
+    {
+      "epoch": 1.1311530039036466,
+      "grad_norm": 5.126955640043889,
+      "learning_rate": 2e-05,
+      "loss": 0.594,
+      "step": 47522
+    },
+    {
+      "epoch": 1.131176806626678,
+      "grad_norm": 1.2799196009913933,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 47523
+    },
+    {
+      "epoch": 1.1312006093497096,
+      "grad_norm": 0.3062192701358715,
+      "learning_rate": 2e-05,
+      "loss": 0.0033,
+      "step": 47524
+    },
+    {
+      "epoch": 1.131224412072741,
+      "grad_norm": 0.6172133986399059,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 47525
+    },
+    {
+      "epoch": 1.1312482147957725,
+      "grad_norm": 0.7960362615585141,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 47526
+    },
+    {
+      "epoch": 1.1312720175188042,
+      "grad_norm": 0.029308902619900035,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47527
+    },
+    {
+      "epoch": 1.1312958202418357,
+      "grad_norm": 4.066487080708652,
+      "learning_rate": 2e-05,
+      "loss": 0.4401,
+      "step": 47528
+    },
+    {
+      "epoch": 1.1313196229648672,
+      "grad_norm": 0.0753333624230353,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47529
+    },
+    {
+      "epoch": 1.1313434256878987,
+      "grad_norm": 1.462679809926033,
+      "learning_rate": 2e-05,
+      "loss": 0.0161,
+      "step": 47530
+    },
+    {
+      "epoch": 1.1313672284109302,
+      "grad_norm": 6.221133195875709,
+      "learning_rate": 2e-05,
+      "loss": 0.1702,
+      "step": 47531
+    },
+    {
+      "epoch": 1.1313910311339617,
+      "grad_norm": 6.249607178886242,
+      "learning_rate": 2e-05,
+      "loss": 0.1547,
+      "step": 47532
+    },
+    {
+      "epoch": 1.1314148338569932,
+      "grad_norm": 6.688186224174634,
+      "learning_rate": 2e-05,
+      "loss": 0.0854,
+      "step": 47533
+    },
+    {
+      "epoch": 1.1314386365800249,
+      "grad_norm": 8.56411465335469,
+      "learning_rate": 2e-05,
+      "loss": 0.4956,
+      "step": 47534
+    },
+    {
+      "epoch": 1.1314624393030563,
+      "grad_norm": 0.1635857633411422,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 47535
+    },
+    {
+      "epoch": 1.1314862420260878,
+      "grad_norm": 7.755913515220412,
+      "learning_rate": 2e-05,
+      "loss": 1.4769,
+      "step": 47536
+    },
+    {
+      "epoch": 1.1315100447491193,
+      "grad_norm": 3.0623959719022307,
+      "learning_rate": 2e-05,
+      "loss": 0.2669,
+      "step": 47537
+    },
+    {
+      "epoch": 1.1315338474721508,
+      "grad_norm": 1.8777276223530042,
+      "learning_rate": 2e-05,
+      "loss": 0.0232,
+      "step": 47538
+    },
+    {
+      "epoch": 1.1315576501951823,
+      "grad_norm": 7.461262038238712,
+      "learning_rate": 2e-05,
+      "loss": 0.7109,
+      "step": 47539
+    },
+    {
+      "epoch": 1.1315814529182138,
+      "grad_norm": 3.535566792854564,
+      "learning_rate": 2e-05,
+      "loss": 0.3347,
+      "step": 47540
+    },
+    {
+      "epoch": 1.1316052556412455,
+      "grad_norm": 0.1566209128516003,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47541
+    },
+    {
+      "epoch": 1.131629058364277,
+      "grad_norm": 0.8836091483482044,
+      "learning_rate": 2e-05,
+      "loss": 0.008,
+      "step": 47542
+    },
+    {
+      "epoch": 1.1316528610873084,
+      "grad_norm": 6.402672347437925,
+      "learning_rate": 2e-05,
+      "loss": 0.3459,
+      "step": 47543
+    },
+    {
+      "epoch": 1.13167666381034,
+      "grad_norm": 1.3412138351979923,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 47544
+    },
+    {
+      "epoch": 1.1317004665333714,
+      "grad_norm": 0.025169401093995195,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47545
+    },
+    {
+      "epoch": 1.131724269256403,
+      "grad_norm": 5.7084923510029215,
+      "learning_rate": 2e-05,
+      "loss": 0.1653,
+      "step": 47546
+    },
+    {
+      "epoch": 1.1317480719794344,
+      "grad_norm": 7.377891382547934,
+      "learning_rate": 2e-05,
+      "loss": 0.9203,
+      "step": 47547
+    },
+    {
+      "epoch": 1.1317718747024659,
+      "grad_norm": 0.22209550086600077,
+      "learning_rate": 2e-05,
+      "loss": 0.003,
+      "step": 47548
+    },
+    {
+      "epoch": 1.1317956774254974,
+      "grad_norm": 7.633324286858472,
+      "learning_rate": 2e-05,
+      "loss": 0.1343,
+      "step": 47549
+    },
+    {
+      "epoch": 1.131819480148529,
+      "grad_norm": 7.2486943095399035,
+      "learning_rate": 2e-05,
+      "loss": 0.6539,
+      "step": 47550
+    },
+    {
+      "epoch": 1.1318432828715606,
+      "grad_norm": 5.37754581960705,
+      "learning_rate": 2e-05,
+      "loss": 1.0261,
+      "step": 47551
+    },
+    {
+      "epoch": 1.131867085594592,
+      "grad_norm": 11.079317980440283,
+      "learning_rate": 2e-05,
+      "loss": 0.2321,
+      "step": 47552
+    },
+    {
+      "epoch": 1.1318908883176235,
+      "grad_norm": 1.1219156232667544,
+      "learning_rate": 2e-05,
+      "loss": 0.0212,
+      "step": 47553
+    },
+    {
+      "epoch": 1.131914691040655,
+      "grad_norm": 4.3104917909417395,
+      "learning_rate": 2e-05,
+      "loss": 0.5049,
+      "step": 47554
+    },
+    {
+      "epoch": 1.1319384937636865,
+      "grad_norm": 3.1398147246625987,
+      "learning_rate": 2e-05,
+      "loss": 0.0356,
+      "step": 47555
+    },
+    {
+      "epoch": 1.131962296486718,
+      "grad_norm": 1.8902873390495925,
+      "learning_rate": 2e-05,
+      "loss": 0.0241,
+      "step": 47556
+    },
+    {
+      "epoch": 1.1319860992097497,
+      "grad_norm": 0.011931048606470354,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47557
+    },
+    {
+      "epoch": 1.1320099019327812,
+      "grad_norm": 2.5541427983328675,
+      "learning_rate": 2e-05,
+      "loss": 0.2613,
+      "step": 47558
+    },
+    {
+      "epoch": 1.1320337046558127,
+      "grad_norm": 4.596014481538593,
+      "learning_rate": 2e-05,
+      "loss": 0.7175,
+      "step": 47559
+    },
+    {
+      "epoch": 1.1320575073788441,
+      "grad_norm": 6.235440331892009,
+      "learning_rate": 2e-05,
+      "loss": 1.0707,
+      "step": 47560
+    },
+    {
+      "epoch": 1.1320813101018756,
+      "grad_norm": 2.3713284894604443,
+      "learning_rate": 2e-05,
+      "loss": 0.193,
+      "step": 47561
+    },
+    {
+      "epoch": 1.1321051128249071,
+      "grad_norm": 0.0927558518801994,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 47562
+    },
+    {
+      "epoch": 1.1321289155479386,
+      "grad_norm": 0.24202964284661138,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47563
+    },
+    {
+      "epoch": 1.1321527182709703,
+      "grad_norm": 2.811259143549752,
+      "learning_rate": 2e-05,
+      "loss": 0.2142,
+      "step": 47564
+    },
+    {
+      "epoch": 1.1321765209940018,
+      "grad_norm": 0.026999212032693697,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47565
+    },
+    {
+      "epoch": 1.1322003237170333,
+      "grad_norm": 15.750978088813785,
+      "learning_rate": 2e-05,
+      "loss": 1.2333,
+      "step": 47566
+    },
+    {
+      "epoch": 1.1322241264400648,
+      "grad_norm": 1.2796776443947133,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 47567
+    },
+    {
+      "epoch": 1.1322479291630962,
+      "grad_norm": 2.535234504670075,
+      "learning_rate": 2e-05,
+      "loss": 0.0197,
+      "step": 47568
+    },
+    {
+      "epoch": 1.1322717318861277,
+      "grad_norm": 0.009032305596898335,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47569
+    },
+    {
+      "epoch": 1.1322955346091592,
+      "grad_norm": 0.22986705169274924,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47570
+    },
+    {
+      "epoch": 1.132319337332191,
+      "grad_norm": 1.1417473895319337,
+      "learning_rate": 2e-05,
+      "loss": 0.0105,
+      "step": 47571
+    },
+    {
+      "epoch": 1.1323431400552224,
+      "grad_norm": 0.3752023234542014,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47572
+    },
+    {
+      "epoch": 1.132366942778254,
+      "grad_norm": 5.453722227663446,
+      "learning_rate": 2e-05,
+      "loss": 0.8074,
+      "step": 47573
+    },
+    {
+      "epoch": 1.1323907455012854,
+      "grad_norm": 4.964907041975633,
+      "learning_rate": 2e-05,
+      "loss": 0.9547,
+      "step": 47574
+    },
+    {
+      "epoch": 1.1324145482243169,
+      "grad_norm": 9.010454304931702,
+      "learning_rate": 2e-05,
+      "loss": 0.5322,
+      "step": 47575
+    },
+    {
+      "epoch": 1.1324383509473483,
+      "grad_norm": 2.4614439154635246,
+      "learning_rate": 2e-05,
+      "loss": 0.0431,
+      "step": 47576
+    },
+    {
+      "epoch": 1.1324621536703798,
+      "grad_norm": 4.322469635655646,
+      "learning_rate": 2e-05,
+      "loss": 0.3718,
+      "step": 47577
+    },
+    {
+      "epoch": 1.1324859563934113,
+      "grad_norm": 0.04350060400284363,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47578
+    },
+    {
+      "epoch": 1.1325097591164428,
+      "grad_norm": 2.04484010513583,
+      "learning_rate": 2e-05,
+      "loss": 0.0439,
+      "step": 47579
+    },
+    {
+      "epoch": 1.1325335618394745,
+      "grad_norm": 0.408796124027461,
+      "learning_rate": 2e-05,
+      "loss": 0.0059,
+      "step": 47580
+    },
+    {
+      "epoch": 1.132557364562506,
+      "grad_norm": 2.355844171690225,
+      "learning_rate": 2e-05,
+      "loss": 0.0217,
+      "step": 47581
+    },
+    {
+      "epoch": 1.1325811672855375,
+      "grad_norm": 4.0202614992690915,
+      "learning_rate": 2e-05,
+      "loss": 0.0442,
+      "step": 47582
+    },
+    {
+      "epoch": 1.132604970008569,
+      "grad_norm": 41.76372680847211,
+      "learning_rate": 2e-05,
+      "loss": 0.7998,
+      "step": 47583
+    },
+    {
+      "epoch": 1.1326287727316005,
+      "grad_norm": 2.814628804690565,
+      "learning_rate": 2e-05,
+      "loss": 0.0846,
+      "step": 47584
+    },
+    {
+      "epoch": 1.132652575454632,
+      "grad_norm": 0.039334029271327904,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47585
+    },
+    {
+      "epoch": 1.1326763781776634,
+      "grad_norm": 4.240080521998681,
+      "learning_rate": 2e-05,
+      "loss": 0.4038,
+      "step": 47586
+    },
+    {
+      "epoch": 1.1327001809006951,
+      "grad_norm": 5.772768616232191,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 47587
+    },
+    {
+      "epoch": 1.1327239836237266,
+      "grad_norm": 1.9222456083456672,
+      "learning_rate": 2e-05,
+      "loss": 0.0257,
+      "step": 47588
+    },
+    {
+      "epoch": 1.132747786346758,
+      "grad_norm": 7.969011522175886,
+      "learning_rate": 2e-05,
+      "loss": 0.1945,
+      "step": 47589
+    },
+    {
+      "epoch": 1.1327715890697896,
+      "grad_norm": 0.028908143176345786,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47590
+    },
+    {
+      "epoch": 1.132795391792821,
+      "grad_norm": 0.009321392803507923,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47591
+    },
+    {
+      "epoch": 1.1328191945158526,
+      "grad_norm": 0.006999232480853732,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47592
+    },
+    {
+      "epoch": 1.132842997238884,
+      "grad_norm": 3.78512187259721,
+      "learning_rate": 2e-05,
+      "loss": 0.0686,
+      "step": 47593
+    },
+    {
+      "epoch": 1.1328667999619157,
+      "grad_norm": 2.403174216058067,
+      "learning_rate": 2e-05,
+      "loss": 0.0608,
+      "step": 47594
+    },
+    {
+      "epoch": 1.1328906026849472,
+      "grad_norm": 4.188517480201001,
+      "learning_rate": 2e-05,
+      "loss": 0.4108,
+      "step": 47595
+    },
+    {
+      "epoch": 1.1329144054079787,
+      "grad_norm": 5.620621817827784,
+      "learning_rate": 2e-05,
+      "loss": 0.5099,
+      "step": 47596
+    },
+    {
+      "epoch": 1.1329382081310102,
+      "grad_norm": 5.056688092861137,
+      "learning_rate": 2e-05,
+      "loss": 0.316,
+      "step": 47597
+    },
+    {
+      "epoch": 1.1329620108540417,
+      "grad_norm": 5.030793331531486,
+      "learning_rate": 2e-05,
+      "loss": 0.7797,
+      "step": 47598
+    },
+    {
+      "epoch": 1.1329858135770732,
+      "grad_norm": 0.23860637401858473,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47599
+    },
+    {
+      "epoch": 1.1330096163001047,
+      "grad_norm": 2.832157487284541,
+      "learning_rate": 2e-05,
+      "loss": 0.222,
+      "step": 47600
+    },
+    {
+      "epoch": 1.1330334190231364,
+      "grad_norm": 3.53770895549855,
+      "learning_rate": 2e-05,
+      "loss": 0.3022,
+      "step": 47601
+    },
+    {
+      "epoch": 1.1330572217461679,
+      "grad_norm": 6.6023200687677015,
+      "learning_rate": 2e-05,
+      "loss": 0.6793,
+      "step": 47602
+    },
+    {
+      "epoch": 1.1330810244691993,
+      "grad_norm": 2.882510424063827,
+      "learning_rate": 2e-05,
+      "loss": 0.1069,
+      "step": 47603
+    },
+    {
+      "epoch": 1.1331048271922308,
+      "grad_norm": 8.432385379866423,
+      "learning_rate": 2e-05,
+      "loss": 0.6964,
+      "step": 47604
+    },
+    {
+      "epoch": 1.1331286299152623,
+      "grad_norm": 0.009221255580500024,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47605
+    },
+    {
+      "epoch": 1.1331524326382938,
+      "grad_norm": 4.810873566846185,
+      "learning_rate": 2e-05,
+      "loss": 0.4592,
+      "step": 47606
+    },
+    {
+      "epoch": 1.1331762353613253,
+      "grad_norm": 0.04806551475833081,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47607
+    },
+    {
+      "epoch": 1.1332000380843568,
+      "grad_norm": 5.369836229579986,
+      "learning_rate": 2e-05,
+      "loss": 0.7652,
+      "step": 47608
+    },
+    {
+      "epoch": 1.1332238408073882,
+      "grad_norm": 0.026739363746976153,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47609
+    },
+    {
+      "epoch": 1.13324764353042,
+      "grad_norm": 3.251936987876422,
+      "learning_rate": 2e-05,
+      "loss": 0.1882,
+      "step": 47610
+    },
+    {
+      "epoch": 1.1332714462534514,
+      "grad_norm": 0.46893638739645466,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 47611
+    },
+    {
+      "epoch": 1.133295248976483,
+      "grad_norm": 3.7303394083204156,
+      "learning_rate": 2e-05,
+      "loss": 0.3326,
+      "step": 47612
+    },
+    {
+      "epoch": 1.1333190516995144,
+      "grad_norm": 7.655871404017384,
+      "learning_rate": 2e-05,
+      "loss": 0.4086,
+      "step": 47613
+    },
+    {
+      "epoch": 1.133342854422546,
+      "grad_norm": 2.7131704315865264,
+      "learning_rate": 2e-05,
+      "loss": 0.1544,
+      "step": 47614
+    },
+    {
+      "epoch": 1.1333666571455774,
+      "grad_norm": 6.338837943997211,
+      "learning_rate": 2e-05,
+      "loss": 1.179,
+      "step": 47615
+    },
+    {
+      "epoch": 1.1333904598686089,
+      "grad_norm": 13.433324059121677,
+      "learning_rate": 2e-05,
+      "loss": 0.8396,
+      "step": 47616
+    },
+    {
+      "epoch": 1.1334142625916406,
+      "grad_norm": 4.048738916328935,
+      "learning_rate": 2e-05,
+      "loss": 0.551,
+      "step": 47617
+    },
+    {
+      "epoch": 1.133438065314672,
+      "grad_norm": 9.123436786468977,
+      "learning_rate": 2e-05,
+      "loss": 0.5523,
+      "step": 47618
+    },
+    {
+      "epoch": 1.1334618680377035,
+      "grad_norm": 0.20406702490775674,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 47619
+    },
+    {
+      "epoch": 1.133485670760735,
+      "grad_norm": 0.18830367586174962,
+      "learning_rate": 2e-05,
+      "loss": 0.0032,
+      "step": 47620
+    },
+    {
+      "epoch": 1.1335094734837665,
+      "grad_norm": 4.599911458369114,
+      "learning_rate": 2e-05,
+      "loss": 0.1063,
+      "step": 47621
+    },
+    {
+      "epoch": 1.133533276206798,
+      "grad_norm": 0.15708361834051665,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 47622
+    },
+    {
+      "epoch": 1.1335570789298295,
+      "grad_norm": 7.468969977648386,
+      "learning_rate": 2e-05,
+      "loss": 0.5592,
+      "step": 47623
+    },
+    {
+      "epoch": 1.1335808816528612,
+      "grad_norm": 0.8740011115341093,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 47624
+    },
+    {
+      "epoch": 1.1336046843758927,
+      "grad_norm": 4.155864563949413,
+      "learning_rate": 2e-05,
+      "loss": 0.2751,
+      "step": 47625
+    },
+    {
+      "epoch": 1.1336284870989242,
+      "grad_norm": 1.758203502108612,
+      "learning_rate": 2e-05,
+      "loss": 0.0352,
+      "step": 47626
+    },
+    {
+      "epoch": 1.1336522898219556,
+      "grad_norm": 0.5504648559741552,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47627
+    },
+    {
+      "epoch": 1.1336760925449871,
+      "grad_norm": 3.0156598946776443,
+      "learning_rate": 2e-05,
+      "loss": 0.2763,
+      "step": 47628
+    },
+    {
+      "epoch": 1.1336998952680186,
+      "grad_norm": 1.290943500280969,
+      "learning_rate": 2e-05,
+      "loss": 0.0094,
+      "step": 47629
+    },
+    {
+      "epoch": 1.13372369799105,
+      "grad_norm": 1.8718828849853062,
+      "learning_rate": 2e-05,
+      "loss": 0.0252,
+      "step": 47630
+    },
+    {
+      "epoch": 1.1337475007140818,
+      "grad_norm": 11.44313083531254,
+      "learning_rate": 2e-05,
+      "loss": 0.4935,
+      "step": 47631
+    },
+    {
+      "epoch": 1.1337713034371133,
+      "grad_norm": 0.11143201817875376,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47632
+    },
+    {
+      "epoch": 1.1337951061601448,
+      "grad_norm": 6.3532769040784105,
+      "learning_rate": 2e-05,
+      "loss": 0.5281,
+      "step": 47633
+    },
+    {
+      "epoch": 1.1338189088831763,
+      "grad_norm": 0.12447499783744914,
+      "learning_rate": 2e-05,
+      "loss": 0.0017,
+      "step": 47634
+    },
+    {
+      "epoch": 1.1338427116062078,
+      "grad_norm": 2.828083598838151,
+      "learning_rate": 2e-05,
+      "loss": 0.1886,
+      "step": 47635
+    },
+    {
+      "epoch": 1.1338665143292392,
+      "grad_norm": 5.3885767027582725,
+      "learning_rate": 2e-05,
+      "loss": 0.5693,
+      "step": 47636
+    },
+    {
+      "epoch": 1.1338903170522707,
+      "grad_norm": 3.7863244423848594,
+      "learning_rate": 2e-05,
+      "loss": 0.4997,
+      "step": 47637
+    },
+    {
+      "epoch": 1.1339141197753022,
+      "grad_norm": 0.029127823622705655,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47638
+    },
+    {
+      "epoch": 1.1339379224983337,
+      "grad_norm": 4.276160772620866,
+      "learning_rate": 2e-05,
+      "loss": 0.5079,
+      "step": 47639
+    },
+    {
+      "epoch": 1.1339617252213654,
+      "grad_norm": 7.045815785397637,
+      "learning_rate": 2e-05,
+      "loss": 0.5591,
+      "step": 47640
+    },
+    {
+      "epoch": 1.1339855279443969,
+      "grad_norm": 0.1541028164995385,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47641
+    },
+    {
+      "epoch": 1.1340093306674284,
+      "grad_norm": 0.010842161832393258,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47642
+    },
+    {
+      "epoch": 1.1340331333904599,
+      "grad_norm": 4.088541553197194,
+      "learning_rate": 2e-05,
+      "loss": 0.2858,
+      "step": 47643
+    },
+    {
+      "epoch": 1.1340569361134913,
+      "grad_norm": 3.9659729672522444,
+      "learning_rate": 2e-05,
+      "loss": 0.4191,
+      "step": 47644
+    },
+    {
+      "epoch": 1.1340807388365228,
+      "grad_norm": 9.681527010510658,
+      "learning_rate": 2e-05,
+      "loss": 0.1264,
+      "step": 47645
+    },
+    {
+      "epoch": 1.1341045415595543,
+      "grad_norm": 0.9328047728045027,
+      "learning_rate": 2e-05,
+      "loss": 0.0097,
+      "step": 47646
+    },
+    {
+      "epoch": 1.134128344282586,
+      "grad_norm": 0.6222570100968549,
+      "learning_rate": 2e-05,
+      "loss": 0.0089,
+      "step": 47647
+    },
+    {
+      "epoch": 1.1341521470056175,
+      "grad_norm": 3.007567838561596,
+      "learning_rate": 2e-05,
+      "loss": 0.209,
+      "step": 47648
+    },
+    {
+      "epoch": 1.134175949728649,
+      "grad_norm": 7.621056904694451,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 47649
+    },
+    {
+      "epoch": 1.1341997524516805,
+      "grad_norm": 0.3007152839401867,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47650
+    },
+    {
+      "epoch": 1.134223555174712,
+      "grad_norm": 4.541784886893454,
+      "learning_rate": 2e-05,
+      "loss": 0.9678,
+      "step": 47651
+    },
+    {
+      "epoch": 1.1342473578977434,
+      "grad_norm": 9.093541499449303,
+      "learning_rate": 2e-05,
+      "loss": 0.1218,
+      "step": 47652
+    },
+    {
+      "epoch": 1.134271160620775,
+      "grad_norm": 0.12289885066305616,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 47653
+    },
+    {
+      "epoch": 1.1342949633438066,
+      "grad_norm": 5.639980871516542,
+      "learning_rate": 2e-05,
+      "loss": 0.4332,
+      "step": 47654
+    },
+    {
+      "epoch": 1.1343187660668381,
+      "grad_norm": 7.589317625509686,
+      "learning_rate": 2e-05,
+      "loss": 0.3243,
+      "step": 47655
+    },
+    {
+      "epoch": 1.1343425687898696,
+      "grad_norm": 5.8911804621518025,
+      "learning_rate": 2e-05,
+      "loss": 0.0459,
+      "step": 47656
+    },
+    {
+      "epoch": 1.134366371512901,
+      "grad_norm": 0.22785008383833202,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47657
+    },
+    {
+      "epoch": 1.1343901742359326,
+      "grad_norm": 3.1645727014707985,
+      "learning_rate": 2e-05,
+      "loss": 0.1175,
+      "step": 47658
+    },
+    {
+      "epoch": 1.134413976958964,
+      "grad_norm": 4.438137574052472,
+      "learning_rate": 2e-05,
+      "loss": 0.0332,
+      "step": 47659
+    },
+    {
+      "epoch": 1.1344377796819956,
+      "grad_norm": 11.543353207587465,
+      "learning_rate": 2e-05,
+      "loss": 1.2312,
+      "step": 47660
+    },
+    {
+      "epoch": 1.1344615824050273,
+      "grad_norm": 3.824709286567855,
+      "learning_rate": 2e-05,
+      "loss": 0.3485,
+      "step": 47661
+    },
+    {
+      "epoch": 1.1344853851280587,
+      "grad_norm": 3.2533395041104356,
+      "learning_rate": 2e-05,
+      "loss": 0.0389,
+      "step": 47662
+    },
+    {
+      "epoch": 1.1345091878510902,
+      "grad_norm": 0.06315072842225039,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47663
+    },
+    {
+      "epoch": 1.1345329905741217,
+      "grad_norm": 0.6763129842409433,
+      "learning_rate": 2e-05,
+      "loss": 0.0071,
+      "step": 47664
+    },
+    {
+      "epoch": 1.1345567932971532,
+      "grad_norm": 5.781390337264227,
+      "learning_rate": 2e-05,
+      "loss": 0.035,
+      "step": 47665
+    },
+    {
+      "epoch": 1.1345805960201847,
+      "grad_norm": 1.0933112022751184,
+      "learning_rate": 2e-05,
+      "loss": 0.0159,
+      "step": 47666
+    },
+    {
+      "epoch": 1.1346043987432162,
+      "grad_norm": 1.8550326044642174,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 47667
+    },
+    {
+      "epoch": 1.1346282014662477,
+      "grad_norm": 2.500867035644616,
+      "learning_rate": 2e-05,
+      "loss": 0.0684,
+      "step": 47668
+    },
+    {
+      "epoch": 1.1346520041892791,
+      "grad_norm": 2.613230641253165,
+      "learning_rate": 2e-05,
+      "loss": 0.1951,
+      "step": 47669
+    },
+    {
+      "epoch": 1.1346758069123108,
+      "grad_norm": 0.037557796521191775,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47670
+    },
+    {
+      "epoch": 1.1346996096353423,
+      "grad_norm": 0.4701355243973769,
+      "learning_rate": 2e-05,
+      "loss": 0.0056,
+      "step": 47671
+    },
+    {
+      "epoch": 1.1347234123583738,
+      "grad_norm": 0.24015991439198328,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47672
+    },
+    {
+      "epoch": 1.1347472150814053,
+      "grad_norm": 4.535962692109239,
+      "learning_rate": 2e-05,
+      "loss": 0.3717,
+      "step": 47673
+    },
+    {
+      "epoch": 1.1347710178044368,
+      "grad_norm": 4.920301643438826,
+      "learning_rate": 2e-05,
+      "loss": 0.5146,
+      "step": 47674
+    },
+    {
+      "epoch": 1.1347948205274683,
+      "grad_norm": 4.430188665275492,
+      "learning_rate": 2e-05,
+      "loss": 0.516,
+      "step": 47675
+    },
+    {
+      "epoch": 1.1348186232504998,
+      "grad_norm": 0.07515310425094128,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47676
+    },
+    {
+      "epoch": 1.1348424259735315,
+      "grad_norm": 2.484416321087604,
+      "learning_rate": 2e-05,
+      "loss": 0.19,
+      "step": 47677
+    },
+    {
+      "epoch": 1.134866228696563,
+      "grad_norm": 3.288471024827821,
+      "learning_rate": 2e-05,
+      "loss": 0.347,
+      "step": 47678
+    },
+    {
+      "epoch": 1.1348900314195944,
+      "grad_norm": 0.054402833894004105,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47679
+    },
+    {
+      "epoch": 1.134913834142626,
+      "grad_norm": 4.723408544499205,
+      "learning_rate": 2e-05,
+      "loss": 0.7531,
+      "step": 47680
+    },
+    {
+      "epoch": 1.1349376368656574,
+      "grad_norm": 0.023284431152706595,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47681
+    },
+    {
+      "epoch": 1.134961439588689,
+      "grad_norm": 2.6915653623704623,
+      "learning_rate": 2e-05,
+      "loss": 0.2152,
+      "step": 47682
+    },
+    {
+      "epoch": 1.1349852423117204,
+      "grad_norm": 15.574669720206519,
+      "learning_rate": 2e-05,
+      "loss": 0.2376,
+      "step": 47683
+    },
+    {
+      "epoch": 1.135009045034752,
+      "grad_norm": 36.73734707181763,
+      "learning_rate": 2e-05,
+      "loss": 0.7163,
+      "step": 47684
+    },
+    {
+      "epoch": 1.1350328477577836,
+      "grad_norm": 4.390367191061431,
+      "learning_rate": 2e-05,
+      "loss": 0.5751,
+      "step": 47685
+    },
+    {
+      "epoch": 1.135056650480815,
+      "grad_norm": 3.87379790228459,
+      "learning_rate": 2e-05,
+      "loss": 0.0417,
+      "step": 47686
+    },
+    {
+      "epoch": 1.1350804532038465,
+      "grad_norm": 0.10725694001464282,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47687
+    },
+    {
+      "epoch": 1.135104255926878,
+      "grad_norm": 2.0484759520621503,
+      "learning_rate": 2e-05,
+      "loss": 0.2014,
+      "step": 47688
+    },
+    {
+      "epoch": 1.1351280586499095,
+      "grad_norm": 9.346271599469983,
+      "learning_rate": 2e-05,
+      "loss": 0.4583,
+      "step": 47689
+    },
+    {
+      "epoch": 1.135151861372941,
+      "grad_norm": 3.4766481430883505,
+      "learning_rate": 2e-05,
+      "loss": 0.3218,
+      "step": 47690
+    },
+    {
+      "epoch": 1.1351756640959727,
+      "grad_norm": 0.06324947084239095,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47691
+    },
+    {
+      "epoch": 1.1351994668190042,
+      "grad_norm": 2.758065249554013,
+      "learning_rate": 2e-05,
+      "loss": 0.0242,
+      "step": 47692
+    },
+    {
+      "epoch": 1.1352232695420357,
+      "grad_norm": 3.2759204373587876,
+      "learning_rate": 2e-05,
+      "loss": 0.4048,
+      "step": 47693
+    },
+    {
+      "epoch": 1.1352470722650672,
+      "grad_norm": 4.571337106458995,
+      "learning_rate": 2e-05,
+      "loss": 0.0169,
+      "step": 47694
+    },
+    {
+      "epoch": 1.1352708749880986,
+      "grad_norm": 0.06451981340979246,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47695
+    },
+    {
+      "epoch": 1.1352946777111301,
+      "grad_norm": 0.8300016548458526,
+      "learning_rate": 2e-05,
+      "loss": 0.0162,
+      "step": 47696
+    },
+    {
+      "epoch": 1.1353184804341616,
+      "grad_norm": 0.1756077184283662,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 47697
+    },
+    {
+      "epoch": 1.135342283157193,
+      "grad_norm": 0.06100888721378232,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47698
+    },
+    {
+      "epoch": 1.1353660858802246,
+      "grad_norm": 9.164645666628624,
+      "learning_rate": 2e-05,
+      "loss": 0.1472,
+      "step": 47699
+    },
+    {
+      "epoch": 1.1353898886032563,
+      "grad_norm": 0.01575787269919962,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47700
+    },
+    {
+      "epoch": 1.1354136913262878,
+      "grad_norm": 0.04130172141450817,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47701
+    },
+    {
+      "epoch": 1.1354374940493193,
+      "grad_norm": 8.846619950451318,
+      "learning_rate": 2e-05,
+      "loss": 0.5212,
+      "step": 47702
+    },
+    {
+      "epoch": 1.1354612967723507,
+      "grad_norm": 0.7070631918827751,
+      "learning_rate": 2e-05,
+      "loss": 0.0087,
+      "step": 47703
+    },
+    {
+      "epoch": 1.1354850994953822,
+      "grad_norm": 0.09197635123461434,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47704
+    },
+    {
+      "epoch": 1.1355089022184137,
+      "grad_norm": 3.546277070984971,
+      "learning_rate": 2e-05,
+      "loss": 0.4565,
+      "step": 47705
+    },
+    {
+      "epoch": 1.1355327049414452,
+      "grad_norm": 0.08169432917150705,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47706
+    },
+    {
+      "epoch": 1.135556507664477,
+      "grad_norm": 0.004523169025853698,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47707
+    },
+    {
+      "epoch": 1.1355803103875084,
+      "grad_norm": 2.1974346681692647,
+      "learning_rate": 2e-05,
+      "loss": 0.0178,
+      "step": 47708
+    },
+    {
+      "epoch": 1.1356041131105399,
+      "grad_norm": 2.1162768157100493,
+      "learning_rate": 2e-05,
+      "loss": 0.1098,
+      "step": 47709
+    },
+    {
+      "epoch": 1.1356279158335714,
+      "grad_norm": 0.006772598704110647,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47710
+    },
+    {
+      "epoch": 1.1356517185566029,
+      "grad_norm": 4.888550710186264,
+      "learning_rate": 2e-05,
+      "loss": 0.0783,
+      "step": 47711
+    },
+    {
+      "epoch": 1.1356755212796343,
+      "grad_norm": 0.0204559902771497,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47712
+    },
+    {
+      "epoch": 1.1356993240026658,
+      "grad_norm": 5.158864448265708,
+      "learning_rate": 2e-05,
+      "loss": 0.2173,
+      "step": 47713
+    },
+    {
+      "epoch": 1.1357231267256975,
+      "grad_norm": 0.15124171194922903,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 47714
+    },
+    {
+      "epoch": 1.135746929448729,
+      "grad_norm": 51.44614642240197,
+      "learning_rate": 2e-05,
+      "loss": 1.1414,
+      "step": 47715
+    },
+    {
+      "epoch": 1.1357707321717605,
+      "grad_norm": 0.010191920331763919,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47716
+    },
+    {
+      "epoch": 1.135794534894792,
+      "grad_norm": 3.344353978067574,
+      "learning_rate": 2e-05,
+      "loss": 0.1996,
+      "step": 47717
+    },
+    {
+      "epoch": 1.1358183376178235,
+      "grad_norm": 0.09797302414934876,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47718
+    },
+    {
+      "epoch": 1.135842140340855,
+      "grad_norm": 9.503850257513879,
+      "learning_rate": 2e-05,
+      "loss": 0.1017,
+      "step": 47719
+    },
+    {
+      "epoch": 1.1358659430638864,
+      "grad_norm": 0.3902393884664601,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47720
+    },
+    {
+      "epoch": 1.1358897457869181,
+      "grad_norm": 3.8631254556145254,
+      "learning_rate": 2e-05,
+      "loss": 0.0376,
+      "step": 47721
+    },
+    {
+      "epoch": 1.1359135485099496,
+      "grad_norm": 7.801355446268479,
+      "learning_rate": 2e-05,
+      "loss": 0.6995,
+      "step": 47722
+    },
+    {
+      "epoch": 1.1359373512329811,
+      "grad_norm": 2.7166142862224225,
+      "learning_rate": 2e-05,
+      "loss": 0.2511,
+      "step": 47723
+    },
+    {
+      "epoch": 1.1359611539560126,
+      "grad_norm": 4.10157933869502,
+      "learning_rate": 2e-05,
+      "loss": 0.3255,
+      "step": 47724
+    },
+    {
+      "epoch": 1.135984956679044,
+      "grad_norm": 0.011109102021702066,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47725
+    },
+    {
+      "epoch": 1.1360087594020756,
+      "grad_norm": 1.797386193277098,
+      "learning_rate": 2e-05,
+      "loss": 0.0691,
+      "step": 47726
+    },
+    {
+      "epoch": 1.136032562125107,
+      "grad_norm": 0.1895718799251451,
+      "learning_rate": 2e-05,
+      "loss": 0.0018,
+      "step": 47727
+    },
+    {
+      "epoch": 1.1360563648481385,
+      "grad_norm": 2.778538930518635,
+      "learning_rate": 2e-05,
+      "loss": 0.1927,
+      "step": 47728
+    },
+    {
+      "epoch": 1.13608016757117,
+      "grad_norm": 0.07141624753617018,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47729
+    },
+    {
+      "epoch": 1.1361039702942017,
+      "grad_norm": 4.265048384289374,
+      "learning_rate": 2e-05,
+      "loss": 0.6577,
+      "step": 47730
+    },
+    {
+      "epoch": 1.1361277730172332,
+      "grad_norm": 0.3674813795066139,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 47731
+    },
+    {
+      "epoch": 1.1361515757402647,
+      "grad_norm": 1.7419469306230935,
+      "learning_rate": 2e-05,
+      "loss": 0.0305,
+      "step": 47732
+    },
+    {
+      "epoch": 1.1361753784632962,
+      "grad_norm": 0.1435938835350288,
+      "learning_rate": 2e-05,
+      "loss": 0.0015,
+      "step": 47733
+    },
+    {
+      "epoch": 1.1361991811863277,
+      "grad_norm": 1.6546555635633518,
+      "learning_rate": 2e-05,
+      "loss": 0.0378,
+      "step": 47734
+    },
+    {
+      "epoch": 1.1362229839093592,
+      "grad_norm": 0.011864803792345918,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47735
+    },
+    {
+      "epoch": 1.1362467866323906,
+      "grad_norm": 0.11516600205778764,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47736
+    },
+    {
+      "epoch": 1.1362705893554224,
+      "grad_norm": 0.034668646174570245,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47737
+    },
+    {
+      "epoch": 1.1362943920784538,
+      "grad_norm": 0.021181811098269672,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47738
+    },
+    {
+      "epoch": 1.1363181948014853,
+      "grad_norm": 10.571800162366921,
+      "learning_rate": 2e-05,
+      "loss": 0.3624,
+      "step": 47739
+    },
+    {
+      "epoch": 1.1363419975245168,
+      "grad_norm": 0.032306964852222175,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47740
+    },
+    {
+      "epoch": 1.1363658002475483,
+      "grad_norm": 0.7225755478976392,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47741
+    },
+    {
+      "epoch": 1.1363896029705798,
+      "grad_norm": 0.005774740267716357,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47742
+    },
+    {
+      "epoch": 1.1364134056936113,
+      "grad_norm": 5.941725564173261,
+      "learning_rate": 2e-05,
+      "loss": 0.7904,
+      "step": 47743
+    },
+    {
+      "epoch": 1.136437208416643,
+      "grad_norm": 5.063961515510571,
+      "learning_rate": 2e-05,
+      "loss": 0.3554,
+      "step": 47744
+    },
+    {
+      "epoch": 1.1364610111396745,
+      "grad_norm": 0.015265129163017711,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47745
+    },
+    {
+      "epoch": 1.136484813862706,
+      "grad_norm": 11.408670531361965,
+      "learning_rate": 2e-05,
+      "loss": 0.2896,
+      "step": 47746
+    },
+    {
+      "epoch": 1.1365086165857374,
+      "grad_norm": 0.4870744975183463,
+      "learning_rate": 2e-05,
+      "loss": 0.0035,
+      "step": 47747
+    },
+    {
+      "epoch": 1.136532419308769,
+      "grad_norm": 1.116563823324709,
+      "learning_rate": 2e-05,
+      "loss": 0.0101,
+      "step": 47748
+    },
+    {
+      "epoch": 1.1365562220318004,
+      "grad_norm": 2.170577446490802,
+      "learning_rate": 2e-05,
+      "loss": 0.0268,
+      "step": 47749
+    },
+    {
+      "epoch": 1.1365800247548319,
+      "grad_norm": 1.657471415339841,
+      "learning_rate": 2e-05,
+      "loss": 0.0249,
+      "step": 47750
+    },
+    {
+      "epoch": 1.1366038274778634,
+      "grad_norm": 6.839803349015688,
+      "learning_rate": 2e-05,
+      "loss": 1.1591,
+      "step": 47751
+    },
+    {
+      "epoch": 1.136627630200895,
+      "grad_norm": 1.8591492204918956,
+      "learning_rate": 2e-05,
+      "loss": 0.0273,
+      "step": 47752
+    },
+    {
+      "epoch": 1.1366514329239266,
+      "grad_norm": 0.24602643400560514,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47753
+    },
+    {
+      "epoch": 1.136675235646958,
+      "grad_norm": 20.880467212337102,
+      "learning_rate": 2e-05,
+      "loss": 0.6274,
+      "step": 47754
+    },
+    {
+      "epoch": 1.1366990383699895,
+      "grad_norm": 6.193670708455023,
+      "learning_rate": 2e-05,
+      "loss": 0.2649,
+      "step": 47755
+    },
+    {
+      "epoch": 1.136722841093021,
+      "grad_norm": 12.966075746639659,
+      "learning_rate": 2e-05,
+      "loss": 0.2636,
+      "step": 47756
+    },
+    {
+      "epoch": 1.1367466438160525,
+      "grad_norm": 0.026355964768121236,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47757
+    },
+    {
+      "epoch": 1.136770446539084,
+      "grad_norm": 4.742914324099686,
+      "learning_rate": 2e-05,
+      "loss": 0.1578,
+      "step": 47758
+    },
+    {
+      "epoch": 1.1367942492621155,
+      "grad_norm": 12.003195265765566,
+      "learning_rate": 2e-05,
+      "loss": 0.3887,
+      "step": 47759
+    },
+    {
+      "epoch": 1.1368180519851472,
+      "grad_norm": 9.104123462653238,
+      "learning_rate": 2e-05,
+      "loss": 0.9693,
+      "step": 47760
+    },
+    {
+      "epoch": 1.1368418547081787,
+      "grad_norm": 4.1189643584992215,
+      "learning_rate": 2e-05,
+      "loss": 0.1488,
+      "step": 47761
+    },
+    {
+      "epoch": 1.1368656574312102,
+      "grad_norm": 2.242600016698753,
+      "learning_rate": 2e-05,
+      "loss": 0.2627,
+      "step": 47762
+    },
+    {
+      "epoch": 1.1368894601542416,
+      "grad_norm": 0.09249512749193048,
+      "learning_rate": 2e-05,
+      "loss": 0.0014,
+      "step": 47763
+    },
+    {
+      "epoch": 1.1369132628772731,
+      "grad_norm": 1.6427377431932195,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 47764
+    },
+    {
+      "epoch": 1.1369370656003046,
+      "grad_norm": 0.02917560452666539,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47765
+    },
+    {
+      "epoch": 1.136960868323336,
+      "grad_norm": 3.2824775632861574,
+      "learning_rate": 2e-05,
+      "loss": 0.2652,
+      "step": 47766
+    },
+    {
+      "epoch": 1.1369846710463678,
+      "grad_norm": 2.373621573096171,
+      "learning_rate": 2e-05,
+      "loss": 0.3092,
+      "step": 47767
+    },
+    {
+      "epoch": 1.1370084737693993,
+      "grad_norm": 4.883254642979199,
+      "learning_rate": 2e-05,
+      "loss": 0.7377,
+      "step": 47768
+    },
+    {
+      "epoch": 1.1370322764924308,
+      "grad_norm": 5.510552230713826,
+      "learning_rate": 2e-05,
+      "loss": 0.4206,
+      "step": 47769
+    },
+    {
+      "epoch": 1.1370560792154623,
+      "grad_norm": 0.7753914353183714,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 47770
+    },
+    {
+      "epoch": 1.1370798819384937,
+      "grad_norm": 4.580018762923539,
+      "learning_rate": 2e-05,
+      "loss": 0.2906,
+      "step": 47771
+    },
+    {
+      "epoch": 1.1371036846615252,
+      "grad_norm": 0.003863347796369737,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47772
+    },
+    {
+      "epoch": 1.1371274873845567,
+      "grad_norm": 0.0013380953302574398,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 47773
+    },
+    {
+      "epoch": 1.1371512901075884,
+      "grad_norm": 0.025963292639366264,
+      "learning_rate": 2e-05,
+      "loss": 0.0004,
+      "step": 47774
+    },
+    {
+      "epoch": 1.13717509283062,
+      "grad_norm": 2.1702772789285336,
+      "learning_rate": 2e-05,
+      "loss": 0.1129,
+      "step": 47775
+    },
+    {
+      "epoch": 1.1371988955536514,
+      "grad_norm": 4.672697700557139,
+      "learning_rate": 2e-05,
+      "loss": 0.356,
+      "step": 47776
+    },
+    {
+      "epoch": 1.1372226982766829,
+      "grad_norm": 5.474866041323505,
+      "learning_rate": 2e-05,
+      "loss": 0.3508,
+      "step": 47777
+    },
+    {
+      "epoch": 1.1372465009997144,
+      "grad_norm": 6.68705972908073,
+      "learning_rate": 2e-05,
+      "loss": 0.654,
+      "step": 47778
+    },
+    {
+      "epoch": 1.1372703037227458,
+      "grad_norm": 1.4617727647195957,
+      "learning_rate": 2e-05,
+      "loss": 0.0116,
+      "step": 47779
+    },
+    {
+      "epoch": 1.1372941064457773,
+      "grad_norm": 0.04348486976066085,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47780
+    },
+    {
+      "epoch": 1.1373179091688088,
+      "grad_norm": 7.814475941531866,
+      "learning_rate": 2e-05,
+      "loss": 1.1354,
+      "step": 47781
+    },
+    {
+      "epoch": 1.1373417118918405,
+      "grad_norm": 1.6535977526771701,
+      "learning_rate": 2e-05,
+      "loss": 0.0266,
+      "step": 47782
+    },
+    {
+      "epoch": 1.137365514614872,
+      "grad_norm": 0.06321015487552861,
+      "learning_rate": 2e-05,
+      "loss": 0.0008,
+      "step": 47783
+    },
+    {
+      "epoch": 1.1373893173379035,
+      "grad_norm": 0.128121905079971,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47784
+    },
+    {
+      "epoch": 1.137413120060935,
+      "grad_norm": 0.06751013631915145,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47785
+    },
+    {
+      "epoch": 1.1374369227839665,
+      "grad_norm": 2.253747158793918,
+      "learning_rate": 2e-05,
+      "loss": 0.0271,
+      "step": 47786
+    },
+    {
+      "epoch": 1.137460725506998,
+      "grad_norm": 0.14941590342759165,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47787
+    },
+    {
+      "epoch": 1.1374845282300294,
+      "grad_norm": 3.343906875482252,
+      "learning_rate": 2e-05,
+      "loss": 0.0404,
+      "step": 47788
+    },
+    {
+      "epoch": 1.137508330953061,
+      "grad_norm": 15.508962204263883,
+      "learning_rate": 2e-05,
+      "loss": 1.3766,
+      "step": 47789
+    },
+    {
+      "epoch": 1.1375321336760926,
+      "grad_norm": 2.2755275407632833,
+      "learning_rate": 2e-05,
+      "loss": 0.2314,
+      "step": 47790
+    },
+    {
+      "epoch": 1.1375559363991241,
+      "grad_norm": 0.16238329775158905,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47791
+    },
+    {
+      "epoch": 1.1375797391221556,
+      "grad_norm": 0.030158263639855434,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47792
+    },
+    {
+      "epoch": 1.137603541845187,
+      "grad_norm": 0.003557060607262222,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47793
+    },
+    {
+      "epoch": 1.1376273445682186,
+      "grad_norm": 4.261624911699238,
+      "learning_rate": 2e-05,
+      "loss": 0.3521,
+      "step": 47794
+    },
+    {
+      "epoch": 1.13765114729125,
+      "grad_norm": 4.935807725547164,
+      "learning_rate": 2e-05,
+      "loss": 0.4847,
+      "step": 47795
+    },
+    {
+      "epoch": 1.1376749500142815,
+      "grad_norm": 16.470938352357663,
+      "learning_rate": 2e-05,
+      "loss": 0.9951,
+      "step": 47796
+    },
+    {
+      "epoch": 1.1376987527373132,
+      "grad_norm": 2.4454145526192503,
+      "learning_rate": 2e-05,
+      "loss": 0.1415,
+      "step": 47797
+    },
+    {
+      "epoch": 1.1377225554603447,
+      "grad_norm": 0.294916935773109,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47798
+    },
+    {
+      "epoch": 1.1377463581833762,
+      "grad_norm": 4.707845821941817,
+      "learning_rate": 2e-05,
+      "loss": 0.7178,
+      "step": 47799
+    },
+    {
+      "epoch": 1.1377701609064077,
+      "grad_norm": 0.6339300963424384,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 47800
+    },
+    {
+      "epoch": 1.1377939636294392,
+      "grad_norm": 4.6520937304383425,
+      "learning_rate": 2e-05,
+      "loss": 0.0341,
+      "step": 47801
+    },
+    {
+      "epoch": 1.1378177663524707,
+      "grad_norm": 6.057080262592452,
+      "learning_rate": 2e-05,
+      "loss": 0.342,
+      "step": 47802
+    },
+    {
+      "epoch": 1.1378415690755022,
+      "grad_norm": 0.02002055115376315,
+      "learning_rate": 2e-05,
+      "loss": 0.0003,
+      "step": 47803
+    },
+    {
+      "epoch": 1.1378653717985339,
+      "grad_norm": 14.777728778082494,
+      "learning_rate": 2e-05,
+      "loss": 0.2444,
+      "step": 47804
+    },
+    {
+      "epoch": 1.1378891745215653,
+      "grad_norm": 0.009989926574004531,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47805
+    },
+    {
+      "epoch": 1.1379129772445968,
+      "grad_norm": 16.913413108056098,
+      "learning_rate": 2e-05,
+      "loss": 1.6653,
+      "step": 47806
+    },
+    {
+      "epoch": 1.1379367799676283,
+      "grad_norm": 1.9690445088266193,
+      "learning_rate": 2e-05,
+      "loss": 0.016,
+      "step": 47807
+    },
+    {
+      "epoch": 1.1379605826906598,
+      "grad_norm": 8.60029158672216,
+      "learning_rate": 2e-05,
+      "loss": 1.4647,
+      "step": 47808
+    },
+    {
+      "epoch": 1.1379843854136913,
+      "grad_norm": 6.4701642821864045,
+      "learning_rate": 2e-05,
+      "loss": 0.4122,
+      "step": 47809
+    },
+    {
+      "epoch": 1.1380081881367228,
+      "grad_norm": 7.704370889986556,
+      "learning_rate": 2e-05,
+      "loss": 0.1975,
+      "step": 47810
+    },
+    {
+      "epoch": 1.1380319908597543,
+      "grad_norm": 5.435345385124052,
+      "learning_rate": 2e-05,
+      "loss": 0.2774,
+      "step": 47811
+    },
+    {
+      "epoch": 1.1380557935827857,
+      "grad_norm": 1.5551626997766064,
+      "learning_rate": 2e-05,
+      "loss": 0.0541,
+      "step": 47812
+    },
+    {
+      "epoch": 1.1380795963058175,
+      "grad_norm": 5.492107739161126,
+      "learning_rate": 2e-05,
+      "loss": 0.461,
+      "step": 47813
+    },
+    {
+      "epoch": 1.138103399028849,
+      "grad_norm": 0.039350293842575365,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47814
+    },
+    {
+      "epoch": 1.1381272017518804,
+      "grad_norm": 5.951654002861288,
+      "learning_rate": 2e-05,
+      "loss": 0.7402,
+      "step": 47815
+    },
+    {
+      "epoch": 1.138151004474912,
+      "grad_norm": 8.358393795661396,
+      "learning_rate": 2e-05,
+      "loss": 0.2144,
+      "step": 47816
+    },
+    {
+      "epoch": 1.1381748071979434,
+      "grad_norm": 4.090148713497723,
+      "learning_rate": 2e-05,
+      "loss": 0.3436,
+      "step": 47817
+    },
+    {
+      "epoch": 1.1381986099209749,
+      "grad_norm": 3.519498702160454,
+      "learning_rate": 2e-05,
+      "loss": 0.0309,
+      "step": 47818
+    },
+    {
+      "epoch": 1.1382224126440064,
+      "grad_norm": 1.534556701949286,
+      "learning_rate": 2e-05,
+      "loss": 0.015,
+      "step": 47819
+    },
+    {
+      "epoch": 1.138246215367038,
+      "grad_norm": 0.0022573774397485493,
+      "learning_rate": 2e-05,
+      "loss": 0.0,
+      "step": 47820
+    },
+    {
+      "epoch": 1.1382700180900696,
+      "grad_norm": 3.7762183117708648,
+      "learning_rate": 2e-05,
+      "loss": 0.4484,
+      "step": 47821
+    },
+    {
+      "epoch": 1.138293820813101,
+      "grad_norm": 3.041016284335528,
+      "learning_rate": 2e-05,
+      "loss": 0.0572,
+      "step": 47822
+    },
+    {
+      "epoch": 1.1383176235361325,
+      "grad_norm": 0.347352228829001,
+      "learning_rate": 2e-05,
+      "loss": 0.0047,
+      "step": 47823
+    },
+    {
+      "epoch": 1.138341426259164,
+      "grad_norm": 10.781824710555629,
+      "learning_rate": 2e-05,
+      "loss": 0.4586,
+      "step": 47824
+    },
+    {
+      "epoch": 1.1383652289821955,
+      "grad_norm": 0.554906965935829,
+      "learning_rate": 2e-05,
+      "loss": 0.0058,
+      "step": 47825
+    },
+    {
+      "epoch": 1.138389031705227,
+      "grad_norm": 0.047956492673848404,
+      "learning_rate": 2e-05,
+      "loss": 0.0006,
+      "step": 47826
+    },
+    {
+      "epoch": 1.1384128344282587,
+      "grad_norm": 5.222790467985152,
+      "learning_rate": 2e-05,
+      "loss": 0.7682,
+      "step": 47827
+    },
+    {
+      "epoch": 1.1384366371512902,
+      "grad_norm": 2.6068864571268993,
+      "learning_rate": 2e-05,
+      "loss": 0.0494,
+      "step": 47828
+    },
+    {
+      "epoch": 1.1384604398743217,
+      "grad_norm": 0.18186514686478092,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 47829
+    },
+    {
+      "epoch": 1.1384842425973531,
+      "grad_norm": 5.16547837978213,
+      "learning_rate": 2e-05,
+      "loss": 0.3123,
+      "step": 47830
+    },
+    {
+      "epoch": 1.1385080453203846,
+      "grad_norm": 0.05498074541796278,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47831
+    },
+    {
+      "epoch": 1.1385318480434161,
+      "grad_norm": 5.217342328276578,
+      "learning_rate": 2e-05,
+      "loss": 0.5443,
+      "step": 47832
+    },
+    {
+      "epoch": 1.1385556507664476,
+      "grad_norm": 7.774821247771204,
+      "learning_rate": 2e-05,
+      "loss": 0.1545,
+      "step": 47833
+    },
+    {
+      "epoch": 1.1385794534894793,
+      "grad_norm": 3.3458087610144593,
+      "learning_rate": 2e-05,
+      "loss": 0.2349,
+      "step": 47834
+    },
+    {
+      "epoch": 1.1386032562125108,
+      "grad_norm": 7.677591072011891,
+      "learning_rate": 2e-05,
+      "loss": 0.9336,
+      "step": 47835
+    },
+    {
+      "epoch": 1.1386270589355423,
+      "grad_norm": 8.14205688889396,
+      "learning_rate": 2e-05,
+      "loss": 1.1305,
+      "step": 47836
+    },
+    {
+      "epoch": 1.1386508616585738,
+      "grad_norm": 3.1784958722304055,
+      "learning_rate": 2e-05,
+      "loss": 0.2287,
+      "step": 47837
+    },
+    {
+      "epoch": 1.1386746643816053,
+      "grad_norm": 10.278940217313703,
+      "learning_rate": 2e-05,
+      "loss": 0.3156,
+      "step": 47838
+    },
+    {
+      "epoch": 1.1386984671046367,
+      "grad_norm": 0.12699875564522695,
+      "learning_rate": 2e-05,
+      "loss": 0.0026,
+      "step": 47839
+    },
+    {
+      "epoch": 1.1387222698276682,
+      "grad_norm": 2.259679659675746,
+      "learning_rate": 2e-05,
+      "loss": 0.074,
+      "step": 47840
+    },
+    {
+      "epoch": 1.1387460725506997,
+      "grad_norm": 14.219807826088473,
+      "learning_rate": 2e-05,
+      "loss": 0.4098,
+      "step": 47841
+    },
+    {
+      "epoch": 1.1387698752737312,
+      "grad_norm": 10.246967686256083,
+      "learning_rate": 2e-05,
+      "loss": 0.8717,
+      "step": 47842
+    },
+    {
+      "epoch": 1.138793677996763,
+      "grad_norm": 3.848707466636422,
+      "learning_rate": 2e-05,
+      "loss": 0.2336,
+      "step": 47843
+    },
+    {
+      "epoch": 1.1388174807197944,
+      "grad_norm": 0.1534204672905601,
+      "learning_rate": 2e-05,
+      "loss": 0.002,
+      "step": 47844
+    },
+    {
+      "epoch": 1.1388412834428259,
+      "grad_norm": 0.06376046599326841,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47845
+    },
+    {
+      "epoch": 1.1388650861658574,
+      "grad_norm": 1.432334995447741,
+      "learning_rate": 2e-05,
+      "loss": 0.0263,
+      "step": 47846
+    },
+    {
+      "epoch": 1.1388888888888888,
+      "grad_norm": 3.1977819158916723,
+      "learning_rate": 2e-05,
+      "loss": 0.3988,
+      "step": 47847
+    },
+    {
+      "epoch": 1.1389126916119203,
+      "grad_norm": 4.446877518376612,
+      "learning_rate": 2e-05,
+      "loss": 0.4881,
+      "step": 47848
+    },
+    {
+      "epoch": 1.1389364943349518,
+      "grad_norm": 8.860119759208322,
+      "learning_rate": 2e-05,
+      "loss": 0.1846,
+      "step": 47849
+    },
+    {
+      "epoch": 1.1389602970579835,
+      "grad_norm": 7.47965102623078,
+      "learning_rate": 2e-05,
+      "loss": 0.4423,
+      "step": 47850
+    },
+    {
+      "epoch": 1.138984099781015,
+      "grad_norm": 6.301389145819062,
+      "learning_rate": 2e-05,
+      "loss": 0.4814,
+      "step": 47851
+    },
+    {
+      "epoch": 1.1390079025040465,
+      "grad_norm": 0.7891721511342342,
+      "learning_rate": 2e-05,
+      "loss": 0.0112,
+      "step": 47852
+    },
+    {
+      "epoch": 1.139031705227078,
+      "grad_norm": 9.82418297831316,
+      "learning_rate": 2e-05,
+      "loss": 0.6355,
+      "step": 47853
+    },
+    {
+      "epoch": 1.1390555079501095,
+      "grad_norm": 12.710223955794921,
+      "learning_rate": 2e-05,
+      "loss": 1.1337,
+      "step": 47854
+    },
+    {
+      "epoch": 1.139079310673141,
+      "grad_norm": 0.6041854951944093,
+      "learning_rate": 2e-05,
+      "loss": 0.006,
+      "step": 47855
+    },
+    {
+      "epoch": 1.1391031133961724,
+      "grad_norm": 5.906079963082944,
+      "learning_rate": 2e-05,
+      "loss": 0.0699,
+      "step": 47856
+    },
+    {
+      "epoch": 1.1391269161192041,
+      "grad_norm": 1.0012983053502955,
+      "learning_rate": 2e-05,
+      "loss": 0.0084,
+      "step": 47857
+    },
+    {
+      "epoch": 1.1391507188422356,
+      "grad_norm": 4.266804945055829,
+      "learning_rate": 2e-05,
+      "loss": 0.0582,
+      "step": 47858
+    },
+    {
+      "epoch": 1.139174521565267,
+      "grad_norm": 0.16262164180508484,
+      "learning_rate": 2e-05,
+      "loss": 0.0019,
+      "step": 47859
+    },
+    {
+      "epoch": 1.1391983242882986,
+      "grad_norm": 11.271218495948226,
+      "learning_rate": 2e-05,
+      "loss": 0.9304,
+      "step": 47860
+    },
+    {
+      "epoch": 1.13922212701133,
+      "grad_norm": 3.65362817799168,
+      "learning_rate": 2e-05,
+      "loss": 0.3376,
+      "step": 47861
+    },
+    {
+      "epoch": 1.1392459297343616,
+      "grad_norm": 0.5050253830146874,
+      "learning_rate": 2e-05,
+      "loss": 0.0038,
+      "step": 47862
+    },
+    {
+      "epoch": 1.139269732457393,
+      "grad_norm": 3.8474005419085064,
+      "learning_rate": 2e-05,
+      "loss": 0.3279,
+      "step": 47863
+    },
+    {
+      "epoch": 1.1392935351804248,
+      "grad_norm": 0.5560061361884941,
+      "learning_rate": 2e-05,
+      "loss": 0.0069,
+      "step": 47864
+    },
+    {
+      "epoch": 1.1393173379034562,
+      "grad_norm": 0.39930075850031654,
+      "learning_rate": 2e-05,
+      "loss": 0.0029,
+      "step": 47865
+    },
+    {
+      "epoch": 1.1393411406264877,
+      "grad_norm": 0.6271050859725791,
+      "learning_rate": 2e-05,
+      "loss": 0.0052,
+      "step": 47866
+    },
+    {
+      "epoch": 1.1393649433495192,
+      "grad_norm": 2.507216960519483,
+      "learning_rate": 2e-05,
+      "loss": 0.2181,
+      "step": 47867
+    },
+    {
+      "epoch": 1.1393887460725507,
+      "grad_norm": 0.0074374611694274706,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47868
+    },
+    {
+      "epoch": 1.1394125487955822,
+      "grad_norm": 3.3786530824264407,
+      "learning_rate": 2e-05,
+      "loss": 0.0314,
+      "step": 47869
+    },
+    {
+      "epoch": 1.1394363515186137,
+      "grad_norm": 5.317941572160831,
+      "learning_rate": 2e-05,
+      "loss": 0.973,
+      "step": 47870
+    },
+    {
+      "epoch": 1.1394601542416452,
+      "grad_norm": 0.011474539963160791,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47871
+    },
+    {
+      "epoch": 1.1394839569646766,
+      "grad_norm": 5.846846570424994,
+      "learning_rate": 2e-05,
+      "loss": 0.5789,
+      "step": 47872
+    },
+    {
+      "epoch": 1.1395077596877083,
+      "grad_norm": 3.607425173791676,
+      "learning_rate": 2e-05,
+      "loss": 0.3027,
+      "step": 47873
+    },
+    {
+      "epoch": 1.1395315624107398,
+      "grad_norm": 7.396220141429324,
+      "learning_rate": 2e-05,
+      "loss": 0.3368,
+      "step": 47874
+    },
+    {
+      "epoch": 1.1395553651337713,
+      "grad_norm": 0.1045777416740426,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47875
+    },
+    {
+      "epoch": 1.1395791678568028,
+      "grad_norm": 6.27041576228411,
+      "learning_rate": 2e-05,
+      "loss": 0.1429,
+      "step": 47876
+    },
+    {
+      "epoch": 1.1396029705798343,
+      "grad_norm": 9.260699953195862,
+      "learning_rate": 2e-05,
+      "loss": 0.8303,
+      "step": 47877
+    },
+    {
+      "epoch": 1.1396267733028658,
+      "grad_norm": 1.9653899037819682,
+      "learning_rate": 2e-05,
+      "loss": 0.1767,
+      "step": 47878
+    },
+    {
+      "epoch": 1.1396505760258973,
+      "grad_norm": 14.228677114437069,
+      "learning_rate": 2e-05,
+      "loss": 0.3433,
+      "step": 47879
+    },
+    {
+      "epoch": 1.139674378748929,
+      "grad_norm": 9.284000741581982,
+      "learning_rate": 2e-05,
+      "loss": 0.6991,
+      "step": 47880
+    },
+    {
+      "epoch": 1.1396981814719604,
+      "grad_norm": 16.04464227185832,
+      "learning_rate": 2e-05,
+      "loss": 0.8814,
+      "step": 47881
+    },
+    {
+      "epoch": 1.139721984194992,
+      "grad_norm": 7.615038610844672,
+      "learning_rate": 2e-05,
+      "loss": 0.6039,
+      "step": 47882
+    },
+    {
+      "epoch": 1.1397457869180234,
+      "grad_norm": 1.3742305326220645,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 47883
+    },
+    {
+      "epoch": 1.139769589641055,
+      "grad_norm": 5.452077628455291,
+      "learning_rate": 2e-05,
+      "loss": 0.0994,
+      "step": 47884
+    },
+    {
+      "epoch": 1.1397933923640864,
+      "grad_norm": 2.6582874001908334,
+      "learning_rate": 2e-05,
+      "loss": 0.2501,
+      "step": 47885
+    },
+    {
+      "epoch": 1.1398171950871179,
+      "grad_norm": 2.634666411059474,
+      "learning_rate": 2e-05,
+      "loss": 0.2652,
+      "step": 47886
+    },
+    {
+      "epoch": 1.1398409978101496,
+      "grad_norm": 4.437168885333192,
+      "learning_rate": 2e-05,
+      "loss": 0.2089,
+      "step": 47887
+    },
+    {
+      "epoch": 1.139864800533181,
+      "grad_norm": 0.4337252194747475,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 47888
+    },
+    {
+      "epoch": 1.1398886032562126,
+      "grad_norm": 0.8118586737971991,
+      "learning_rate": 2e-05,
+      "loss": 0.0131,
+      "step": 47889
+    },
+    {
+      "epoch": 1.139912405979244,
+      "grad_norm": 0.16617141496253104,
+      "learning_rate": 2e-05,
+      "loss": 0.0031,
+      "step": 47890
+    },
+    {
+      "epoch": 1.1399362087022755,
+      "grad_norm": 0.10342533845630514,
+      "learning_rate": 2e-05,
+      "loss": 0.001,
+      "step": 47891
+    },
+    {
+      "epoch": 1.139960011425307,
+      "grad_norm": 7.429506976691396,
+      "learning_rate": 2e-05,
+      "loss": 0.6402,
+      "step": 47892
+    },
+    {
+      "epoch": 1.1399838141483385,
+      "grad_norm": 8.742016345435424,
+      "learning_rate": 2e-05,
+      "loss": 0.493,
+      "step": 47893
+    },
+    {
+      "epoch": 1.1400076168713702,
+      "grad_norm": 0.031142008978353997,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47894
+    },
+    {
+      "epoch": 1.1400314195944017,
+      "grad_norm": 8.273356747312862,
+      "learning_rate": 2e-05,
+      "loss": 0.5867,
+      "step": 47895
+    },
+    {
+      "epoch": 1.1400552223174332,
+      "grad_norm": 6.041925430397998,
+      "learning_rate": 2e-05,
+      "loss": 0.511,
+      "step": 47896
+    },
+    {
+      "epoch": 1.1400790250404647,
+      "grad_norm": 7.656852657136449,
+      "learning_rate": 2e-05,
+      "loss": 0.6047,
+      "step": 47897
+    },
+    {
+      "epoch": 1.1401028277634961,
+      "grad_norm": 0.03704759237296042,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47898
+    },
+    {
+      "epoch": 1.1401266304865276,
+      "grad_norm": 2.0588742420770028,
+      "learning_rate": 2e-05,
+      "loss": 0.1112,
+      "step": 47899
+    },
+    {
+      "epoch": 1.140150433209559,
+      "grad_norm": 4.761277626961159,
+      "learning_rate": 2e-05,
+      "loss": 0.0423,
+      "step": 47900
+    },
+    {
+      "epoch": 1.1401742359325906,
+      "grad_norm": 4.328228615083804,
+      "learning_rate": 2e-05,
+      "loss": 0.3271,
+      "step": 47901
+    },
+    {
+      "epoch": 1.140198038655622,
+      "grad_norm": 0.04220255997815162,
+      "learning_rate": 2e-05,
+      "loss": 0.0009,
+      "step": 47902
+    },
+    {
+      "epoch": 1.1402218413786538,
+      "grad_norm": 0.04136059265452073,
+      "learning_rate": 2e-05,
+      "loss": 0.0007,
+      "step": 47903
+    },
+    {
+      "epoch": 1.1402456441016853,
+      "grad_norm": 6.496921405132259,
+      "learning_rate": 2e-05,
+      "loss": 0.6489,
+      "step": 47904
+    },
+    {
+      "epoch": 1.1402694468247168,
+      "grad_norm": 3.5895975568804492,
+      "learning_rate": 2e-05,
+      "loss": 0.4945,
+      "step": 47905
+    },
+    {
+      "epoch": 1.1402932495477482,
+      "grad_norm": 4.13278547304754,
+      "learning_rate": 2e-05,
+      "loss": 0.094,
+      "step": 47906
+    },
+    {
+      "epoch": 1.1403170522707797,
+      "grad_norm": 0.31641707494013854,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47907
+    },
+    {
+      "epoch": 1.1403408549938112,
+      "grad_norm": 7.080603037354272,
+      "learning_rate": 2e-05,
+      "loss": 0.3028,
+      "step": 47908
+    },
+    {
+      "epoch": 1.1403646577168427,
+      "grad_norm": 19.51915568477217,
+      "learning_rate": 2e-05,
+      "loss": 0.3616,
+      "step": 47909
+    },
+    {
+      "epoch": 1.1403884604398744,
+      "grad_norm": 0.8763854863222812,
+      "learning_rate": 2e-05,
+      "loss": 0.0142,
+      "step": 47910
+    },
+    {
+      "epoch": 1.140412263162906,
+      "grad_norm": 7.712443618470975,
+      "learning_rate": 2e-05,
+      "loss": 0.587,
+      "step": 47911
+    },
+    {
+      "epoch": 1.1404360658859374,
+      "grad_norm": 6.126352020955595,
+      "learning_rate": 2e-05,
+      "loss": 0.6998,
+      "step": 47912
+    },
+    {
+      "epoch": 1.1404598686089689,
+      "grad_norm": 0.09205446889129014,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47913
+    },
+    {
+      "epoch": 1.1404836713320003,
+      "grad_norm": 3.2450275201944594,
+      "learning_rate": 2e-05,
+      "loss": 0.1995,
+      "step": 47914
+    },
+    {
+      "epoch": 1.1405074740550318,
+      "grad_norm": 3.5285493895506965,
+      "learning_rate": 2e-05,
+      "loss": 0.034,
+      "step": 47915
+    },
+    {
+      "epoch": 1.1405312767780633,
+      "grad_norm": 4.360014791381237,
+      "learning_rate": 2e-05,
+      "loss": 0.0185,
+      "step": 47916
+    },
+    {
+      "epoch": 1.140555079501095,
+      "grad_norm": 0.007261976650890934,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47917
+    },
+    {
+      "epoch": 1.1405788822241265,
+      "grad_norm": 4.774022998574913,
+      "learning_rate": 2e-05,
+      "loss": 0.6752,
+      "step": 47918
+    },
+    {
+      "epoch": 1.140602684947158,
+      "grad_norm": 0.033357747161093104,
+      "learning_rate": 2e-05,
+      "loss": 0.0005,
+      "step": 47919
+    },
+    {
+      "epoch": 1.1406264876701895,
+      "grad_norm": 3.7206281919740802,
+      "learning_rate": 2e-05,
+      "loss": 0.4486,
+      "step": 47920
+    },
+    {
+      "epoch": 1.140650290393221,
+      "grad_norm": 0.6919068465166214,
+      "learning_rate": 2e-05,
+      "loss": 0.0054,
+      "step": 47921
+    },
+    {
+      "epoch": 1.1406740931162525,
+      "grad_norm": 9.010172048354608,
+      "learning_rate": 2e-05,
+      "loss": 0.4326,
+      "step": 47922
+    },
+    {
+      "epoch": 1.140697895839284,
+      "grad_norm": 0.18424707640715798,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 47923
+    },
+    {
+      "epoch": 1.1407216985623156,
+      "grad_norm": 0.5875695407258376,
+      "learning_rate": 2e-05,
+      "loss": 0.0104,
+      "step": 47924
+    },
+    {
+      "epoch": 1.1407455012853471,
+      "grad_norm": 9.901253341136178,
+      "learning_rate": 2e-05,
+      "loss": 0.1395,
+      "step": 47925
+    },
+    {
+      "epoch": 1.1407693040083786,
+      "grad_norm": 0.06701588533224295,
+      "learning_rate": 2e-05,
+      "loss": 0.0011,
+      "step": 47926
+    },
+    {
+      "epoch": 1.14079310673141,
+      "grad_norm": 2.96152596827969,
+      "learning_rate": 2e-05,
+      "loss": 0.087,
+      "step": 47927
+    },
+    {
+      "epoch": 1.1408169094544416,
+      "grad_norm": 7.827665765801842,
+      "learning_rate": 2e-05,
+      "loss": 0.1226,
+      "step": 47928
+    },
+    {
+      "epoch": 1.140840712177473,
+      "grad_norm": 0.4182093588998841,
+      "learning_rate": 2e-05,
+      "loss": 0.0037,
+      "step": 47929
+    },
+    {
+      "epoch": 1.1408645149005046,
+      "grad_norm": 5.643865366117965,
+      "learning_rate": 2e-05,
+      "loss": 0.5813,
+      "step": 47930
+    },
+    {
+      "epoch": 1.140888317623536,
+      "grad_norm": 3.213656880774892,
+      "learning_rate": 2e-05,
+      "loss": 0.6218,
+      "step": 47931
+    },
+    {
+      "epoch": 1.1409121203465675,
+      "grad_norm": 3.421483007497658,
+      "learning_rate": 2e-05,
+      "loss": 0.5074,
+      "step": 47932
+    },
+    {
+      "epoch": 1.1409359230695992,
+      "grad_norm": 2.4175508362569516,
+      "learning_rate": 2e-05,
+      "loss": 0.1678,
+      "step": 47933
+    },
+    {
+      "epoch": 1.1409597257926307,
+      "grad_norm": 1.0536423269486062,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 47934
+    },
+    {
+      "epoch": 1.1409835285156622,
+      "grad_norm": 1.2284698487005947,
+      "learning_rate": 2e-05,
+      "loss": 0.0145,
+      "step": 47935
+    },
+    {
+      "epoch": 1.1410073312386937,
+      "grad_norm": 5.645010886364864,
+      "learning_rate": 2e-05,
+      "loss": 0.8113,
+      "step": 47936
+    },
+    {
+      "epoch": 1.1410311339617252,
+      "grad_norm": 4.803044231442031,
+      "learning_rate": 2e-05,
+      "loss": 0.5033,
+      "step": 47937
+    },
+    {
+      "epoch": 1.1410549366847567,
+      "grad_norm": 1.236106206502604,
+      "learning_rate": 2e-05,
+      "loss": 0.0144,
+      "step": 47938
+    },
+    {
+      "epoch": 1.1410787394077881,
+      "grad_norm": 0.2601019925818292,
+      "learning_rate": 2e-05,
+      "loss": 0.0049,
+      "step": 47939
+    },
+    {
+      "epoch": 1.1411025421308199,
+      "grad_norm": 4.511453743424862,
+      "learning_rate": 2e-05,
+      "loss": 0.6059,
+      "step": 47940
+    },
+    {
+      "epoch": 1.1411263448538513,
+      "grad_norm": 3.9682193600964317,
+      "learning_rate": 2e-05,
+      "loss": 0.0913,
+      "step": 47941
+    },
+    {
+      "epoch": 1.1411501475768828,
+      "grad_norm": 4.3756904824277285,
+      "learning_rate": 2e-05,
+      "loss": 0.0514,
+      "step": 47942
+    },
+    {
+      "epoch": 1.1411739502999143,
+      "grad_norm": 4.27942678124345,
+      "learning_rate": 2e-05,
+      "loss": 0.0764,
+      "step": 47943
+    },
+    {
+      "epoch": 1.1411977530229458,
+      "grad_norm": 3.393160166284606,
+      "learning_rate": 2e-05,
+      "loss": 0.0375,
+      "step": 47944
+    },
+    {
+      "epoch": 1.1412215557459773,
+      "grad_norm": 2.3098917538533947,
+      "learning_rate": 2e-05,
+      "loss": 0.1421,
+      "step": 47945
+    },
+    {
+      "epoch": 1.1412453584690088,
+      "grad_norm": 3.2875807004410347,
+      "learning_rate": 2e-05,
+      "loss": 0.0336,
+      "step": 47946
+    },
+    {
+      "epoch": 1.1412691611920405,
+      "grad_norm": 0.16388821269552845,
+      "learning_rate": 2e-05,
+      "loss": 0.0022,
+      "step": 47947
+    },
+    {
+      "epoch": 1.141292963915072,
+      "grad_norm": 3.356421747080855,
+      "learning_rate": 2e-05,
+      "loss": 0.2183,
+      "step": 47948
+    },
+    {
+      "epoch": 1.1413167666381034,
+      "grad_norm": 3.433713924505461,
+      "learning_rate": 2e-05,
+      "loss": 0.2163,
+      "step": 47949
+    },
+    {
+      "epoch": 1.141340569361135,
+      "grad_norm": 0.3611556963240466,
+      "learning_rate": 2e-05,
+      "loss": 0.0034,
+      "step": 47950
+    },
+    {
+      "epoch": 1.1413643720841664,
+      "grad_norm": 4.182494771419251,
+      "learning_rate": 2e-05,
+      "loss": 0.4801,
+      "step": 47951
+    },
+    {
+      "epoch": 1.141388174807198,
+      "grad_norm": 0.07004038300765959,
+      "learning_rate": 2e-05,
+      "loss": 0.0012,
+      "step": 47952
+    },
+    {
+      "epoch": 1.1414119775302294,
+      "grad_norm": 5.044197583775496,
+      "learning_rate": 2e-05,
+      "loss": 0.142,
+      "step": 47953
+    },
+    {
+      "epoch": 1.141435780253261,
+      "grad_norm": 11.63471875529363,
+      "learning_rate": 2e-05,
+      "loss": 1.2894,
+      "step": 47954
+    },
+    {
+      "epoch": 1.1414595829762926,
+      "grad_norm": 0.2552051126908471,
+      "learning_rate": 2e-05,
+      "loss": 0.0039,
+      "step": 47955
+    },
+    {
+      "epoch": 1.141483385699324,
+      "grad_norm": 0.08482223131897267,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 47956
+    },
+    {
+      "epoch": 1.1415071884223555,
+      "grad_norm": 0.8055491758030385,
+      "learning_rate": 2e-05,
+      "loss": 0.0102,
+      "step": 47957
+    },
+    {
+      "epoch": 1.141530991145387,
+      "grad_norm": 0.9443277601836598,
+      "learning_rate": 2e-05,
+      "loss": 0.0066,
+      "step": 47958
+    },
+    {
+      "epoch": 1.1415547938684185,
+      "grad_norm": 2.2589332250139904,
+      "learning_rate": 2e-05,
+      "loss": 0.0559,
+      "step": 47959
+    },
+    {
+      "epoch": 1.14157859659145,
+      "grad_norm": 0.4875318147392052,
+      "learning_rate": 2e-05,
+      "loss": 0.0062,
+      "step": 47960
+    },
+    {
+      "epoch": 1.1416023993144815,
+      "grad_norm": 29.109518417022855,
+      "learning_rate": 2e-05,
+      "loss": 0.353,
+      "step": 47961
+    },
+    {
+      "epoch": 1.141626202037513,
+      "grad_norm": 4.977547051544662,
+      "learning_rate": 2e-05,
+      "loss": 0.8149,
+      "step": 47962
+    },
+    {
+      "epoch": 1.1416500047605447,
+      "grad_norm": 14.55754342564546,
+      "learning_rate": 2e-05,
+      "loss": 0.764,
+      "step": 47963
+    },
+    {
+      "epoch": 1.1416738074835762,
+      "grad_norm": 3.2927424329521346,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 47964
+    },
+    {
+      "epoch": 1.1416976102066076,
+      "grad_norm": 7.921831645248449,
+      "learning_rate": 2e-05,
+      "loss": 0.5042,
+      "step": 47965
+    },
+    {
+      "epoch": 1.1417214129296391,
+      "grad_norm": 2.168985902345191,
+      "learning_rate": 2e-05,
+      "loss": 0.1901,
+      "step": 47966
+    },
+    {
+      "epoch": 1.1417452156526706,
+      "grad_norm": 0.11201218158739702,
+      "learning_rate": 2e-05,
+      "loss": 0.0013,
+      "step": 47967
+    },
+    {
+      "epoch": 1.141769018375702,
+      "grad_norm": 0.17732630448619013,
+      "learning_rate": 2e-05,
+      "loss": 0.0027,
+      "step": 47968
+    },
+    {
+      "epoch": 1.1417928210987336,
+      "grad_norm": 0.009726172420545196,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47969
+    },
+    {
+      "epoch": 1.1418166238217653,
+      "grad_norm": 5.868687962028841,
+      "learning_rate": 2e-05,
+      "loss": 0.3647,
+      "step": 47970
+    },
+    {
+      "epoch": 1.1418404265447968,
+      "grad_norm": 6.978915179752285,
+      "learning_rate": 2e-05,
+      "loss": 0.5553,
+      "step": 47971
+    },
+    {
+      "epoch": 1.1418642292678283,
+      "grad_norm": 1.3248541570027017,
+      "learning_rate": 2e-05,
+      "loss": 0.0234,
+      "step": 47972
+    },
+    {
+      "epoch": 1.1418880319908598,
+      "grad_norm": 5.0511196721096,
+      "learning_rate": 2e-05,
+      "loss": 0.3746,
+      "step": 47973
+    },
+    {
+      "epoch": 1.1419118347138912,
+      "grad_norm": 5.425261619578091,
+      "learning_rate": 2e-05,
+      "loss": 0.0519,
+      "step": 47974
+    },
+    {
+      "epoch": 1.1419356374369227,
+      "grad_norm": 1.6897581619887996,
+      "learning_rate": 2e-05,
+      "loss": 0.0473,
+      "step": 47975
+    },
+    {
+      "epoch": 1.1419594401599542,
+      "grad_norm": 8.444327703398375,
+      "learning_rate": 2e-05,
+      "loss": 1.4852,
+      "step": 47976
+    },
+    {
+      "epoch": 1.141983242882986,
+      "grad_norm": 13.317910777641169,
+      "learning_rate": 2e-05,
+      "loss": 0.1357,
+      "step": 47977
+    },
+    {
+      "epoch": 1.1420070456060174,
+      "grad_norm": 15.813571271909767,
+      "learning_rate": 2e-05,
+      "loss": 0.6966,
+      "step": 47978
+    },
+    {
+      "epoch": 1.1420308483290489,
+      "grad_norm": 0.4624636268143477,
+      "learning_rate": 2e-05,
+      "loss": 0.0063,
+      "step": 47979
+    },
+    {
+      "epoch": 1.1420546510520804,
+      "grad_norm": 3.3621259421526433,
+      "learning_rate": 2e-05,
+      "loss": 0.3099,
+      "step": 47980
+    },
+    {
+      "epoch": 1.1420784537751119,
+      "grad_norm": 0.2807320000132581,
+      "learning_rate": 2e-05,
+      "loss": 0.0024,
+      "step": 47981
+    },
+    {
+      "epoch": 1.1421022564981433,
+      "grad_norm": 0.22932255648057626,
+      "learning_rate": 2e-05,
+      "loss": 0.0036,
+      "step": 47982
+    },
+    {
+      "epoch": 1.1421260592211748,
+      "grad_norm": 1.8863059282836667,
+      "learning_rate": 2e-05,
+      "loss": 0.1783,
+      "step": 47983
+    },
+    {
+      "epoch": 1.1421498619442065,
+      "grad_norm": 0.00917788780345385,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47984
+    },
+    {
+      "epoch": 1.142173664667238,
+      "grad_norm": 0.14045865967673812,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 47985
+    },
+    {
+      "epoch": 1.1421974673902695,
+      "grad_norm": 1.1036871567554785,
+      "learning_rate": 2e-05,
+      "loss": 0.0115,
+      "step": 47986
+    },
+    {
+      "epoch": 1.142221270113301,
+      "grad_norm": 0.004864725997862331,
+      "learning_rate": 2e-05,
+      "loss": 0.0001,
+      "step": 47987
+    },
+    {
+      "epoch": 1.1422450728363325,
+      "grad_norm": 6.945218976165436,
+      "learning_rate": 2e-05,
+      "loss": 0.7763,
+      "step": 47988
+    },
+    {
+      "epoch": 1.142268875559364,
+      "grad_norm": 2.6749068395143354,
+      "learning_rate": 2e-05,
+      "loss": 0.0358,
+      "step": 47989
+    },
+    {
+      "epoch": 1.1422926782823954,
+      "grad_norm": 0.00956194891682404,
+      "learning_rate": 2e-05,
+      "loss": 0.0002,
+      "step": 47990
+    },
+    {
+      "epoch": 1.142316481005427,
+      "grad_norm": 0.2555907299107725,
+      "learning_rate": 2e-05,
+      "loss": 0.0025,
+      "step": 47991
+    },
+    {
+      "epoch": 1.1423402837284584,
+      "grad_norm": 2.8944792632692047,
+      "learning_rate": 2e-05,
+      "loss": 0.4492,
+      "step": 47992
+    },
+    {
+      "epoch": 1.1423640864514901,
+      "grad_norm": 5.278901752704585,
+      "learning_rate": 2e-05,
+      "loss": 0.5003,
+      "step": 47993
+    },
+    {
+      "epoch": 1.1423878891745216,
+      "grad_norm": 0.149344734622936,
+      "learning_rate": 2e-05,
+      "loss": 0.0023,
+      "step": 47994
+    },
+    {
+      "epoch": 1.142411691897553,
+      "grad_norm": 4.679412421081246,
+      "learning_rate": 2e-05,
+      "loss": 0.4779,
+      "step": 47995
+    },
+    {
+      "epoch": 1.1424354946205846,
+      "grad_norm": 1.4397530862243806,
+      "learning_rate": 2e-05,
+      "loss": 0.01,
+      "step": 47996
+    },
+    {
+      "epoch": 1.142459297343616,
+      "grad_norm": 3.066739552682618,
+      "learning_rate": 2e-05,
+      "loss": 0.2516,
+      "step": 47997
+    },
+    {
+      "epoch": 1.1424831000666476,
+      "grad_norm": 2.725336762389615,
+      "learning_rate": 2e-05,
+      "loss": 0.368,
+      "step": 47998
+    },
+    {
+      "epoch": 1.142506902789679,
+      "grad_norm": 1.7241317671772394,
+      "learning_rate": 2e-05,
+      "loss": 0.0193,
+      "step": 47999
+    },
+    {
+      "epoch": 1.1425307055127107,
+      "grad_norm": 0.9781463448171965,
+      "learning_rate": 2e-05,
+      "loss": 0.0041,
+      "step": 48000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 126036,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 12000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 93823060811776.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}